Document mmmMg8dnX1O6dzXXjqbpbEj6B

P184-133771 Advance* in Exposure, Health and Environmental Effect* Studio* of PCS* ICA1K Ufa Syateae, Ine. Cleveland, Ohid 13 May 1982 U.8. DEPARTMENT OF COMMERCE Nittonil Technical Information Sorvlco MONS 026032 JCiftSystems, Jhc. ?ot>4- U5771 TR-507-137B ADVANCES IN EXPOSURE, HEALTH AND ENVIRONMENTAL EFFECTS STUDIES OF PCBs SYMPOSIUM PR0CEEDH6S May 12-13,1962 Prepared Undar Contract No. 68*01 -6564 Prefect 1247 ICAIR Work Assignment No: 30 by ICAIR JjfeSpats. Jkt Cleveland, OH 44122 for Office of Toxic Subetancee Health and Environmental Review Division U3. Environmental Protection Agency Washington, DC 20460 NATIONAL TECHNICAL INFORMATION SERVICE MONS 026033 1 REPORT DOCUMENTATION l- "a PAGE 1 LSI-TR-507-137B * |t TlIlM |dg Advances In Exposure, Health and Environmental | if facts Scudles of PCBs: Symposia# Papceedlngs I /. miASWrtt) | R. J. Davenport and B. K. Bernard (Eds.) | >. ^erfmrmlihg OrgaltMta* Nm *A* AMP ' ICAlR 1 Lift Systems, Inc. 2 WSJ lllghpolnc Road ; Clsveland, Ohio 64122 11 XuaMflUi Oitiiiittmi IM, IMiii Office of Toxic Substances . .nealch and Environmental Review Division I 0.$. Environmental Protection Agency ' '-.shington, DC 20460 * ' Nmm la AmIMATO MMN88 Ho. PB84-U5771 ' | a awe me 12733 (Date of Publlcitio la a fertanatae OrsaaiuilM Kmc at TR-507-137B |a Pwnm,rtaa/vmM um at 1247/02 6 30 Ik Canewuei ar StmO) Mt to 68-01-6354 Ml in 4 Nrm e>HPK Final 05/12 - 13/82 u mmuwmnm The proceedings is a compilation of data protested and dlscuaaad during i jvepoetva held Mar 12-13, 1982 In Betheada, Maryland, to provide as Information update :na Environmental Protection Agener' Polychlorinated Biphenyls Program. Nine papers, dlseuaalon summarise and concluding raurkj address data, published of from studies progress since 1978, In the following topic areas< , Analytical Methodologies; Exposure Studies: Environmental Residues end Bloaccumulation; Exposure Studies: Industrial Processes; Health Effects: Epidemiology; Health Effects: laboratory Studies; 1 Environmental Effects; aod i Risk Assessment Z-th contributed paper Includes a significant bibliography of cited and related ; scientific literature referencee for ehoae seeking greater detail. Pclychlordlnated Biphenyls Er/lronantal Residues Bloaccumulatlon ) a imhhui--Tm j Analytical Kathodologlaa ; Liboratory Scudlaa ' Exposure Studies Health Effects Environmental Effacte Risk Aaseeement Symposium Proceedings PCB [ t cceaTi nas/emas .i, *.*s*.a#tj HwSMa; . Release unlimited, j s"ail*ble from Metlonsl Technical Information Sar'toe, Springfield,VA . 22161 Ji-I ANU-CM.ia ta. iiniii atm enm at--<v Unclassified m emaeaianNiNa Unclassified. m tt.NRf hM 365 a. Mat imoiim, tomt m HONS 026034 TR-507-137B advances in exposure, health and environmental EFFECTS STUDIES 07 PCBe Syvpoalua Proceeding* Hay 12-13, 1982 Prepared Under Contract No. 68-01-6334 Project 12*7 ICAIR Work Aaalgnment No.: 30 8y ICAIR Lift Spates*, Inc. Cleveland, OH 44122 for Offica of Toxic Sitbatancaa Health and Environmental Review Dlviaion V.S. Envlronaental Procaction Agency Vaahlngton, DC 20460 I mqns 0^6035 DISCLAIMER Although thla Procaadlnga Doeuaant rport tha praaantationa and diacuaalon that occurred la tha PCI Syapoalum fundod by tha Dnltad Statta Envlronawntal Frotactloo Agaacy through Cootract Ho. W-01-653A vlth Llfa Syataaa. Inc., tha cootaota of tha Procaadlnga ara tha iadapandant vlava of aclaotlaea not aaaoclatad vlth EPA. Conaaquontly, tha Procaadlnga Docuaant baa not baan aubjactad to tha Agaacy'i paar raalov procaaa. It doat not nacoaaarlly raflact tha rlava of tha Agancy and no official andoraanane ahould ba Infarrad. 11 HONS 026036 FOREWORD ilmoat anyone who la likely to open thle book will know the beelc facte about polychlorinated blphenyla (PCB*)t flret described In the late 19th century, they have been produced on an industrial ecele elnce about 19JO. In the subsequent 40 year* they have been uaed extenaivaly in a variety of applicatlona becauaa of their chemical and thenaal aeablllty and good dielectric properties. Since the different commercial PCS products are not simple entitles but faalllas of compounds, the picture is coapllcated because individusl component* of these mixtures differ in their chemical, physical and biological properties. The initial concern about chlorinated aroaatle coapounds originated from reports around 1940 of chlorscna and. liver changes in workers who cam* in concact with these aeeerials during production or handling. Worldvld* atteutlon waa not given to PCBs, however, until 1966 when the journal. Haw Scientist, reported on Soven Jenaen'a finding of PCBa in pika fro* Swadlah iaitas, in the hunan brain and In an aagla which waa found dead in the Stockholm area. During tha years that followed, PCBa were found in envlronsiantal aemplaa In aaveral parte of tha world. It baeana evident that theae paralatant and bloaecusiulatlng conpounda wars ubiquitously diatributadi biological material from polar regions and othsr uninhabited and rsaota arses contained traces of PCBt. Tha O.S. nanufacturara raapondad to theae findings by Halting aalaa of PCBa to usas in enclosed aystaaa only. The praaance of PCBa In aoaa Isolated environmental aaaplae from Sweden might hevn raaainad an analytical curloelty for soaia tins had not Rachael Caraon'a book, "Sllant Spring", alortod anvironnental sciantlat* - and tha general public - aoaa years before to the poaalble danger of organoehlorlne coapounds. In 1968 accidental Ingestion of PCBa through contaalnatad rlca oil revealed a nuabar of toxic effects of PCBa In a group of Japanese people. This incident also established placental transfer of thaaa eoapounda and persistence In the allk of eontaalnetsd mother*. Comparable effects wars observed In tha 1979 PCBa poisoning incident In Taiwan. In tha aariy 1970a, it waa also shewn that hunane were directly exposed to PCBs froa such diverse aoureea aa food packaging materials and earbonleaa copying paper. Tha same time period also saw tha beginning of intensive scientific efforts to understand eh* rout* of exposure and the fat* and affacta of PCBa. Long tarn axpoauran at low lnvnla were studied and eh* phenomena of bloaccuanilatlon end food-chain magnification wore alao investigated. Prior to tudiet with PCBs, little was known aa photochemical behavior and metabolic patterns of chioroaromatic compound*. PCBs production and import is either forbidden or llmitad in moat Industrialised countries. However, moat of th* total PCBs produced to data are either still la service or found lu duaps and land fill*. To avoid further environmental contamination, safe methods for ultimata disposal have .to be found.,. ill HONS 026037 The PCBa problea will not dlaappear orarolght, Thr* u; b eaocional reaction and unfounded faar concerning PCBa, but ehara la alao the poaelblllty of tael anvlronaental probleae with potentially heaardoua coneaquencae. W oust look at the facta, evaluate thaa and conduct lnvaatlgatlona In araaa whara additional aclantiflc raaaarch la required. A eeienrlf lcally .eound bade (or prudant political daelatona aniat ba developed. Tha O.S. Environaental Pro c action Agency and Ufa Syateaa. Inc. ara to ba connandad for organlalng tbit Important Syapoalua. 0. Hutalngar ttatwareley of 'Uyxaath Bayreuth, Caraany lv HONS 026038 ACKNOWLEDGEMENTS The symposium from which this document vas prepared, "Recent Advance* In Exposure, Health, and Environmental Efface* Studies of PCI*," va* held May 12-13, 1982 at tha Holiday Inn in Bethesda, Maryland. The syaposlua va* organised by ICAIR of Ilf* Systems, Inc. The effort was performed under Contract No. 68-01-6554 froa the 0.S. Envlronaental Protection Agency, Office of Pesticides and Toxic Substances. We wish to gratefully acknowledge che administrative services of the follovlng Individuals responsible for the program organisation, execution, and subsequent preparation of these proceedings* Dr. Irvin Beuael, Or. Wllllaa Farland and Mr. 8. Clsnn Williams of the Office of Foeticides and Toxic Sub stance*, and Mr. Michael Kangas, Ms. Cynthia Patrick, Dr. Roy Reuter, Mr. Creg Schiefer, Ms. Patricia Sweeney and Ms. Cynthia Wesley of ICAIR. Special appreciation 1* extended to the syaposlua participants for the excellent quality of the presentations, the professional Banner In which the sessions ware conducted, and the stimulating nature of the discussions. It was the coablned efforts of all these Individuals which ensured the success of this syaposlua and the resultant contribution of knowledge to tha scientific coHunl'ty. I. J. Davenport B. K. Bernard MQNs 026039 CONTENTS Dlaclaloer ....... ............................................................................ Foravord............................................... lcknovUd|wiiti......................................... Chapter 1 INTRODUCTION ............................................................................................ Chaptar 2ANALYTICAL METHODOLOGIES ......................................................................... Racant Advance* on eha Analyala of PCSa In Eavlronaantal Biological Madia. E. D. Pallltzarl, M. A. Moaalay, S. 0. Coopar, J. V. Barry. B. A. Data an ......................................... Dlacuaclon Susury. Mary S. Wolff ............................................................. Chaptar 3 EXPOSURE STUDIES - ENVIRONMENTAL RESIDUES AND BIOACCUMULATION.................................................... Dlatrlbutloa and Fata of Chloroblphanyle In tha Environment. R. J. Moolanaar ... ................................. ..... Factora Af factIn* tha lloaccuaulatlon and Faralatanca of Folychlorloatad Blphanyla.M. J. Vodlenlk ........... Diacuaeion Sunary. M. Malancon ................................................................. Chapter 4 EXPOSURE STUDIES - INDUSTRIAL PROCESSES ..................................... Expoaura Studlaa Relating to laduatrial Froeeaaaa Containing Incidental FCB, K. L. lodged. T. A. Rohinaoa. C. J, Bantlor. R. G. Raley. I. Caapera, F. Evald, A. R, Plttavey ................................................................ Dlacuaaloa Suaaary. R. 0. Relay ................. Chapter S HEALTH EFFECTS - EPIDEMIOLOGY ......................................................... Racant Epldealologlc Studlaa of PCBa. W. R. Gaffay . ..................... Dlacuaaloa Suaaury. R. D. Rlabrough ......................... ........ Chaptar 6 REALTH EFFECTS LABORATOEY STUDIES ............................................. Foteatlal laaltb Xffaeta froa Expoaura to Folychlorlnated Blphanyla| laboratory Studlaa. S. L. Frlaaa, V. A. Drill, 1. V. Raya. Y. A. looala, C. B. Schaffer ............ FCBet Structure-Activity Salatlonahlpa. S. Safa, A, Farkinaon, L. Robertton, T. Sevyer, S, Beadlera, L. Safa. M. A. Caapball, M. Mull In......................................................... Dlaeuaalon Suaatary. R. D, Rlabrough ......................................................... Pag* 11 Ill v I 4 4 61 66 67 97 128 136 137 182 188 189 202 211 212 229 249 contlnuadvl WOA/5 0^6040 Cjncmti - conclcuad Ptgt Chapear 7 ENVIRONMENTAL E7TECTS........................................................................ tnvlronmancal Efface* of Polychlorinated Blphenylai Important Ecological Aapacta. R. C. Hunter........................ .... Dlacuaaioo Suanary. M. Malaneon.......................................... Chapter 8 RISK ASSESSMENT , ............................................................ Aaaasament of cha Human Rials* co PCS a Aasociatad with eh* Expected Environmental Expoauraa. R. C. Jama*. R. B. Harbiaon ................................................. ................................................ Dlacuaalon Summary. R.B. Harbiaon...................... Chaocar 9 CONCESSIONS............................................................................................ Suanary of Racant Advarvcaa In Exposure. Health and Environmental Efface* Studlaa of FCB*. O. Hutxlngar ...... Attend***........................................................... ............................................................................................................................................ 264 26S 292 300 300 337 340 340 346 351 vll MQNS 026041 CHAPTER I IHTRODDCTIOH Thi publication 1 a compilation of data praaantad and dlacuaaad during tha symposium antlclad, "Recant Advancaa In Exposure, Htalth and Environment!l Effaces Scudlas of ?CBa," ha Id May 12-13, 1982 In Bethasda, Maryland. Tha symposium was orgaalzad specifically as an Information updata for tha Envlronaanta! Protactlon Agency's (EPA'a) Polychlorinated Biphenyls (PCBs) Program. In more ganaral terms, the symposium provided an opportunity for tha exchange of Information between PCBs researchers from academia, government aganclas, industry and Independent rasaarch institutions. Tha PCBs Program was implemented by tha Office of Peseleldes and Toxic Substances (OPTS) to carry out the specific requirements of Section 6(e) of the Toxle Substances Control Act (TSCA). Section 6(e) mandates tha promulgation of rults by tha SPA for tha dlspoaal and marking of PCBa, and aatabllshaa a timetable for controlling the uae of PCBa, and for eliminating th manufacture, procettlug and dlttrlbuelon of PCBa In commarca. Aa part of tha continuing updata procasa for acquiring ralavant data for tha PCBa Program, tha OPTS requeated that a two-day aympoalua be undertaken which would provide a forum for a full and frae dlacuaalon of aclentlflc latuoa ralatad to PCBa. Sympoalum Scopt ar.d Objectives Tha date presented In the eympoelum was limited to that which haa become available tinea 1978, including both published date and data from etudlaa that ara still In progress. Ho formal consideration was glvtn tha EPA'a paat or potantlal regulatory acclvltlee concamlng PCBa. The following objectives were eectbllehed for the eympoelum: 1. To review, dltcuet end Interpret PCBs eclentlflc data published since 1978 end preliminary PCBa dace from ongoing work. 2. To provide for an exchange of views between scientists who art knowledgeable about txpoeura, health and environmental effects, and human haalth studies sssoeltted with PCBs. 3. To Identify roaaarch gsps and arses of agreement and controversy. Svmooalum Participants Tho following Individuals participated ea tha Chairperson, ea Spaekere or ee Discussion Leaders: HQHS 026042 Chairperson: Or. Otto Hutzlnger* University of Amsterdam Th Netherlands Authors: Dr. Seymour Prists Drill, Frless, Hays, Locals 6 Schaffar, Inc. Arlington, VA 12209 Dr. William Caffay Monsanto Company St. Louis, MO 63167 Mr. Kant Bodges Dow Chemical Company Midland, MI 46640 Mr. Richard Hunter D.S. Army Corps of Engineers Tulsa, OK 74121 Dr, Robert Jamest Ecology and Environment, Inc. Buffalo, NT 14235 Discussion Leaders: Dr. Raymond Harblson University of Arkansas for Medical Sclaoess Little Rock, AK 72205 Dr. Robert Kaley Monsanto Company St. Louis, MO 63167 Dr. Ranats Kimbrough^ Cantor for Disease Control Chanblas, CA 30341 Dr. Robert Moolenaar Dov Chemical Company Midland, MI 48640 Dr. Edo Pelllzzart Research Triangle Institute Research Triangle Park, NC 27706 Dr. Stephen Safe Texas A 6 M University College Staelon, TX 77843 Dr. Mary Jo Vodlcnlk Medical College of Wisconsin Milvaukee, VI 53226 Dr. Mark Melancon Medical Collage of Wisconsin Milvaukee, HI 53226 Dr. Mary Wolff Mount Sinai School of Medicine Nov York, NT 10029 Nov at University of Bayreuth, Bayreuth. Pederal Republic of Cermany. tNov at University of Arkansas, Division of Toxicology, little Rock, AR 72205. eMov at Canear -for Environmental Bealth, Atlanta. GA 30333. 2 HONS 026043 Ci-ganimation oif Symposium In order Co achieve ch above symposium objectives, th symposium was divided Into ieven sessions. The session topic*, la order of presentation 'ere a* follow*! Analytical Methodologies Exposure Studies: environmental Residues and Bloeccumulatlon Exposure Studlta: Industrial Proessses Health Effects: Epidemiology Health Sffecti: Laboratory Studies Environmental Effects Risk Assessment Each session consisted of the presentation of papers by one to three speakers, followed by a discussion period. During the discussion period, a dialogue was established between the speakers and the symposium attendee*, luastlons wera answered, and data and the Interpretations of data were ajco.nanged. Each Discussion leader moderated the discussion period for one or two sessions. Following the symposium, the Discussion Leaders prepered summerlea of the discussions from thalr notes and transcripts mads during the symposium. The chapters that follow are arranged In the order of the seven sessloos. Each chapter contains the papers that wera presented and the discussion summary, prepared by tna Dlscusalon Leader. The last chapter (Chapter 9) Is a summary of the accomplishments and results of the work praseneed and discussed during the symposium. Although vrltten from the technological viewpoint of a single participant (the symposium Chairperson, Dr. Otto Hutslnger), it provides a needed Interpreta tion of the many facts, recent technical advances and Identified data gapa thje were presented and dlacuseed during this tvo-day period. 3 0260^ HONS CHAPTER 2 ANALYTICAL METHODOLOGIES This chaptsr rsvlsvs csrtsln analytical advances in PCS* detection, Identification, and quantification. High raaoiutlon gaa chromatography la covarad In detail, with (pacific rafarancaa to capillary column development and supporting 'phasa' methodologies, datactor aysterna, and tba comparative advantages and limitations of each combination. Problems associated vlth Instrument calibration ara generally caused by the nonavailability or lack of primary PCSa standards, and the possibility of using secondary standards such as well-defined commercial PCBs mixtures. The application of these newer methods to environmental, biological, and process stream samples are discussed. The discussion summary Is a review of the advantages and limitations of the various mathodologlas. Topics Include procedures for analysis of PCBs contaminants, validation methodology, limits of detection, PCBs pattern recognition, and the cosc/beneflt ratio of packed versus capillary column gas chromatography. RECENT ADVANCES IN THE ANALYSIS OP POLYCHLORINATED BIPHENYLS IN ENVIRONMENTAL AND BIOLOGICAL MEDIA E; D. Pelllssari, M. A, Moseley, S. D. Cooper, J. V. Harry, B. Damian Analytical and Chemical Sciences Research Triangle Institute Research Triangle Park, NC 27709 M. D. Mullin 0.S. Environmental Protection Agency Large Lakes Research Station Groaae lie, MI 48118 INTRODUCTION Polychlorinated biphenyls (PCBs), terphenyls (PCTs) and quadphenyls (PCQs) had been manufactured as commercial mixtures (Aroclors* (US), Clophen* (Germany), and Ranoclor* (Japan)] for more chan four decades before they ware banned In the US In 1976. Because of their long-term heavy usage as dielectric fluids, in transformers and capacitors. In hydraulic fluids, fire retardanta, etc., and their persistence, PCBs have permeated Into practically every environmental medium throughout the world (Ruezlnger at al. 1974). Beeauee of thalr ubiquity PCBa have received considerable attention In recant yeara In thn area of analytical neaeuramenta and toxicology. Since environmental polluelon by PCBa first became apparent, a plethora of purification procedures, gas ehronatographle systems, detectors and methods for quantifying gas chromatographic responses has bean reported to determine aba conttmlnaclca of environmental and biological eamplae (Hutslnger at tl. 1974, Sharma 1973, Chau and Sampson 1973, Leo at el. 1976, Rrull 1977, Brinkman at al. 1978, Pomarants at al. 1978, Stalling at al. 1979a,b, Shares 4 MOMS 026045 1981, Albro it *1. 1979). However, until recent advances In analytical instrumentation, most of the conventional methodology yielded only serniquentitatlve data with virtually no qualitative or quantitative information on the leomer composition for PCBs in the aample. The analytical problems are complicated by the fact that there are 209 individual chlorinated biphenyl Isomers (Ballachmltter end 2ell 1980) spanning 10 homologous series (1 to 10 chlorines per biphenyl). A commercial mixcure (e.g., Aroclor) Itself may contain as many as $0 ehloroblphenyl isomers representing several homologs (Ballachmltter and Zell 1980). The conventional quentlflcaclon method Is based upon pecked column gae chromatography (GC) and reports the FOB content In environmental and biological ssmplee by referring it to a particular Aroclor mixture (e.g., Aroclor 1262, 1254 or 1260). The CC detector Is first calibrated uelng commercial Aroclor mixtures, and then ehe appropriate commercial Aroclor profile le matched to the sample profile. Using the Vebb-McCall technique or a variation of it the total FCB content is calculated (Webb end McCall 1973). This approach, hovever. Is potentially subject to considerable error. Environmental contamination may be derived from Aroclor mixtures or from incidentally generated chloroblphenyls vhose profiles do not resemble Aroclor patterns (Flttavay at al. 1981). In either case, the conventional quantifi cation method is Inadequate. As time peeaes, the "Aroclor patterns" undergo alteration In the environment since selective veatherlng and blotransformatlon and bloaccumulatlon in living organisms perturb these patterns. The problem of Aroclor pattern dissimilarity will be further aggravated as PCI monitoring continues into the future. The quantification problem is further exacerbated by production of ehloroblphenyl isomers in chemical process streams, incineration, etc., via chemical or pyrolysis reactions which are not the same ae the chemical reac tions once used for the manufacture of Aroclora (Pittavay et al. 1981). There is no reason to believe that Incidental generation produces any fixed pattern of ehloroblphenyl Isomers which the analyst can use to identify end quantify chloroblohenyla based upon pattern recognition from the packed column gee chromatogram. Thus, the qualitative and quantitative analysis of FCBs in environmental and biological samples and in eanples from process screens involves the difficult issue of having to detect, identify and quantify each individual isomer. The accuracy of PCI determination in environmental, biological and process stream saaplee la, in addition to the reasons given above, also related to the degree of variability in the analytical response of each ehloroblphenyl isomer to the detector employed (Hutslnger et el. 1974, toe end Egaaa 1979, Albro et al. 1981, 8afa et al. 1975). Recent theoretical and experimental studies have indicated that the biological properties of ehloroblphenyl isomers ere significantly influenced by the number and biphenyl ring position of the chlorine atoms (McKinney and Singh 1981, Poland and Clover 1977, Matthews et al. 1978, Kimbrough 1980). Because the toxicological properties vary considerably among Isomers (McKinney and Sir.gh 1981, Poland end Clover 1977), sort sophisticated methods capable of yielding information about he~chemleal composition at the Isomeric level are 5 MONS 026046 required In order to precisely Investigate and taaata tha toxicological consequences of PCI pollution. Obviously, tha ldtal analytical procadura la ona which ldantlflaa and measures aach Individual chloroblphanyl iaomar. Thla chapter axaalnaa tha moat racant advance* which atrive to neat tha objective of Individual chloroblphanyl laoaar identification and quantifica tion. The area* dlacuaaed arc: (a) high resolution ga chromatography; (b) detection ayatana; (c) availability of chloroblphanyl laoaar standards; and (d) application of atate-of-tha-art aethode to the analysis of environmental and biological samples. It la bayond tha scope of this paper to Include sampling techniques and isolation and purification method*. These have been adequately reviewed aliavhera. HIGH RESOLUTION CAS CHROMATOGRAPHIC TECHNOLOGY In order to analyxa for 209 Individual chloroblphanyl isoaars tha us* of high resolution chromatographic technique* 1* mandatory. In this respect. CC currently la far tuparlor to high resolution thin-layer and high resolution liquid chrooatogrspby for PCB analysis. The relative aarlt* of packad column (low raaolutlon) varaua capillary column (high resolution) GC analysis of PCB* have bean succinctly reported by Mullin end Filkine (1981). Their work provided tha Impetus for further research into the determination of an optimum capillary column(e) for the analysis of individual isomer* of PCBa, polybromlnated biphenyl*, pesticides and other helogenated hydrocarbon* in biolog ical and environmental samples. Development of Capillary Columns Moseley end Pelllctarl have recently reported on axtanalv* investigation into five variables which affect capillary performance for the analysis of PCBa (Hoseley end Pelllstetl 1982a,b, Pelllasarl at el. 1981). These variables arar (a) materiel of construction) (b) pratraatmane/daaetlvatlon procadura*; (c) stationary phase type; (d) stationary phase film thickness; and (a) capillary dimensions, Evaluation criteria employed wars: (a) separa tion number (T2) between 2,2',4*,5-tatrachlorobiphenyl and 2,2',4,4',6,6'-h*xacbloroblphanyl; (b) resolution between 2,2',5.J'-c*trachloroblph*nyl and 2,2',4,,j-c*trschloroblphenyl; (c) heighe equivalent to an affective theoret ical plate (BEET?) for 2,2',4',3-tstrachloroblphenyl; (d) adsorption character istics; (a) thermal stability; and (f) -general performance on an Aroelor 1242/1260 mixture (Itl, w/w). Capillaries were made from pyrex and soft glass, quarts, vitreous allies and fused silica materials (Moseley and Felllttari 1982s,b, Pellicseri at al. 1981). The objective of this facet of tha aeudy was to determine tha most suitable material for construction, pretreatmant/deactivation, and amenability to coating of e thin, uniform, seabl* film of leeelonary phaae. It was recognised early In this work that these variables ware not independent of each other. Except for C-87, ell stationary phases evaluated could b* coated on the flexible allies capillar!** (Palllssarl at al. 1981). Prafarsoca for the allies capillaries was attributed to their flexible nature facilitating their aeaambly Into CCs even by the novice. Thus, a major Impediment to the V** of capttlvrles -by tb* affWtyelcal eoaaimlty had been removed. 6 HONS 026047 Many pratraatment/deectlvatlon procedures vara Investigated tinea the raw construction material vaa nor aulcabla tor directly coat log with stationary phase (Moseley and Pelllzxeri 1982a,b, Pelllxxarl at ai. 1991). The procaduraa vera: (a) barium carbonaca treatment; (b) Carbovax 20M; (e) Superox-4; (d) HCi etching; (a) persllylatlon; and (f) thermally inducad polyailoxana (SE-52 or 0V-101) bonding, tha prafarrad atehoda vara polyalloxana daacclvaclon on alllca and pyrax and ptrsllylatlon on pyrax (Moatlay and Palllxxarl 1982b, Palllaxari at al. 1981). Although glaat caplllarlaa coatad with C-87 ttatlonary phaaa provlda axeallant raaolution of Individual chlorobiphany1 laoaara, uaa of C-87 eurrantly la llaltad by two faetora - lta caoparttura atablllty (>220*0 and tba currant Inability to auceaaafully coat a uniform film on allies, gacauaa of the upper temperature Unit tha C-87 phase coated caplllarlaa are Inadequate Tor the analyelt of PCTe, PCQe, PB8a and sample extracts which are wary "dirty" (e.g., flah). For thaaa reason* alternate atatlonary phaaaa ware sought. The currant literature reporta aavaral different pbaaea used In eaplliary CC for analyala of PCla (Albro at al. 1981, Mullln and PlUlae 1981, Krupclk at al. 1978, Krupclk at al. 1980, Albro-at al. 1977, Stalling at ai. 1978, Irunstra at al. 1981); however, until our research there was not a concerted effort to ayateaatlcaily evaluate pbaaaa (Moseley and Palllxxarl 1982a,b, Palllxxarl at al. 1981). Doing tha previously aantloned criteria a aatrlx designed atudy was executed In concert and while eking a comparison of phase selectivity (McRaynolda' constants) to guide the overall Investigation toward the "optimum" phaaa. Aaong the phases evaluated were C-87, SC-34, SP-2100, QP-1, SC-32. 09-101, Dexall 410, Aplexon M and Aplexon L. Early reaules predicted (fro* McRaynolda' constants) that Aplexon M would aost closely mimic tha exeallene eeparatlon pattern of C-87 (Pelllssarl at al. 1981). In fact, this correapondeoco was experimentally demonstrated (Moseley and Pelllxxarl 1982a). Several advantages of Aplexon M were recognised and will be discussed later. Finally, a complementary stationary phaaa to Aplexon M was aoughe ao that one phaaa could serve as a primary analytical column and tha other a reference column. SC-34 eoated capillaries provided a signifi cantly different resolution pattern compared to Aplexon M and SC-54 was also selected In our research (Moseley and Palllxxarl 1982a,b). It has long bean recognised that the thinner the stationary film thick ness the higher the mass tranafer coefficient. The sample capacity, however, decreases. Thus, research vaa performed to determine a film thickness which had a very high mass transfer coefficient (as measured by BEET?), adequate ample capacity for tha detection system to be employed, and atablllty to long periods of usage (Moseley and Felllxxaxl 1982a,b, Pelllxxarl at al. 1981). Stability (to solvent and thermal shock) eould, of course, be Imparted by Immobilising the film on the silica surface by a erosalinklng/surface bonding reaction (Mosalay and Palllxxarl 1982b). Two methods vera reported which Umtoblllxad phases to silica (Moseley and Pelllxxarl 1982b). One utilised a thermal technique (SE-32), the second a dleumyl peroxide reaction (SC-34), The first method easily produced the dealrad <U)2i.u film caplllarlaa for use with electron capture (CC) and negative Ion chenleal ionisation (NICI) mass spectromstric (MS) detection of 7 0260*i8 HONS PCBt. These capillaries exhibited bleed of 0.7 pA (flame Ionization detaetton) it 320*C. The second method yielded thicker film* (0.1 u) for higher capacity for use with electron Impact MS. An Aplezon M phase Immobilized to silica hat not been reported. The results of these Investigations vhleh specifically focused on resolu tion. of FCBe indicated that a elllea capillary, polyelloxane deactivated, 0.2-ai IT) x SO-* In length, coaeed irlth Apeizon M (0.025 u film) or SE-34 (0.025-u flla) la preferred (Moseley and Pelllzzsrl 1992a,b. FsHltzarl at el. 1981). An Aplezon M coated elllea eaplllary allove the elution of PCSe to occur approximately 30-40*C lover than ocher stationary phases (Moseley end Fellixzarl 1902a,b, Pelllzzarl ae al. 1981) and Is thernally stable to '28J*C. Comparison of Packed and Capillary Column. Profllea The Inherent dlffarancea beevean lov and high resolution gee chroma tographic coluans ere exaapllfled by Figures.I and 2. Figure L coaperee the profiles for e atandard mixture of Aroclor 1242/1260., Figure 2 depicts cne profiles for a atack (stationary aourca) aaapla (Mu11in 1981). Both high resolution analyaat vara parfonaad on SE-54 futad elllea capillaries. It ia avldaot In Flgura 2 that tha lev raaolntlon profile does noe readily raaaabla a coaaarclal Aroclor mixture. Thus, It vould ha highly lnaecurata to quantify tha lov rasolutlon profile using a coaaarclal Aroclor standard for lnatruaent calibration and tha Vebb-McCell method (Webb end McCall 1973). Several axaaplee demonstrating the performance of Apleson M elllea capillaries will be discussed later in this chapter. DETECTIOM ITSTEMS Modei of Detection Ose of several type* of detectors has bean reported over tha years In tha analyals of PCBe. Hovever, only a fav have made significant recent advances that are aotevorthy. These are electron capture, negative Ion chemical loultatlon mesa spectrometry (XICX-MS) and selected Ion ewnleorlng (SIM), a variant of alaetron Impact use spectrometry. Electron Capture Pataction Tha alaetron eapturo detector continues eo be ona of tha moat sensitive and hence valuable selective detectors for PCB detection. It la only recently, bovevar, that Its full potential has bean realised and incorporated into coHarclal aysterna, From a quantitative standpoint, ona of tha limitations of the alaetron capture detector la pate hat been nonlinearity of taaponsa. Until tha aid-1970s eon EC# la *ae had m linear dynamic rang* of approximately 50-100. Several papers have dealt vlth determining the proper function vhleh vould give a linear relationship vlth concentration (Felllzzarl 1974, Vanevorth and 8 MQNS 026049 Packad column -i--............. ..... j 3* INJ Tima, aln Tima, mlm s i uu0u 2 Pl|uia 1. Chromatograma of Aroclor 1242/1260. Top - packad column, bottom - capillary. 9 MONS 026050 Claaa capillary coluaa -all ml i ill J1yJUL I 100 70 Tia, aIn 30 XKJ 7i|art 2. ChroMtograaa of atack aunpla. Top - packed colon, boteoa - capillary. 10 HONS 026051 Chm 1979). It ha* bean suggastad clue Che response vaa logarithmic by analogy to light absorption (Wantvorth and Chon 1979). Hovavar, whao tha BCD was oparated In tha pulaa sampling aoda tha raactlona oceurrad primarily In tha flald fraa parted, to that tha analogy vaa not considered vary appToprlata (Wantvorth and Chan 1979). Onca a valid kinetic modal for tha electron capture process** for thermal raactlona vaa derived by Wantvorth ae al. (1979) and tha research in ttmoaphsrle ptaaauTt Ionization Base apactromatry gava direct evidence on negative and poaltlva Ion formation (undar ECD conditlona). tha solution to electron capture detector nonlinearity vaa in hand. Attention yes focused on tha BCD's concantratlon dapandance vhlch has bean of gTeet concern to anaLytircel chemistry. Analysis of che kinetic modal vaa carried out by numerical solution of differen tial aquaeions vhlch alleviated acme of tha early mathematical assumption*. Aa a result of tha experimental vork on electron capture mecnanis*., subsequent report* appeared in tha literature describing an altarnata asthod if linearization of response (Maggs at al. 1971, Patterson *t al. 1976). The fundamenta 1 prlnclpla vaa baaad upon modulating tha fraquancy of pulalng tha dacactor alactrodaa so that tha plasma currant vaa msintalnad constant. This function is given by vhara t " tha fraquancy giving tha bass currant la ths abaancs of capturing ipacias t tha fraquancy giving eha bate currant in tha praosaca of a capturing spaclss Initially thara vara rapores of a "break" in linearity, hut with improved instrument design paranatars this problem vaa eliminated (Patterson at al. 1976). Thus, tb* modern ICD.utilises nodulatad pulsed fraquancy to achieve a dynamic range of approximately four orders of magnitude, an improvement vhlch la nacaaaary for tha analysis of PCBo In environmental and biological sazq>l*a. A second limitation vao tha ECD call volume. This problem did not aurface until analysts began Investigating tha use of high resolution capil lary columns (Palllsaarl 1974). Tha original call volumes of 2-4-aL vara ade quate vhau used with packed columns vhose flovs vara high and chromatographic paak ahapa and afficiancy vara prtserved. A lav isolated raporta appeared in tha iltaratura addressing improved eall design, in particular iov volume cello, for uaa vlth capillar1st. ECD call volumes batvaan 250 and 500-ul utilising' a coaxial design still required a acavangar gas after tha capillary eelumn to reduce tbs rasidanca time in tha ECD call and preserve tha ultrahigh theoretical platas vhlch vara attainable by eaplllarlos (Pallittarl 1974). Bovevtr, it vaa only recently that commercial CC systems became routinely available vlth ECDa compatible vlth eaplllary flov rataa. With eh*a* tvo parallel davalopmauta it la now feasible to perform elegant quantitative alaetron capture detection of PCTa. 11 HONS 026052 Waaatlve loo Chemical tonlsatlou Haas Spectrometry TM* detection method It a variant of positive loo chemical lonliatloo naai spectrometry. Tha associated electronics to regulate, focus and paaa negative Iona to ao appropriately polarlaad electron multiplier vara developed In tha aid-1970a; however, It vat not until recently that loo aourea and raagtnt |aa coodltlooa vara Investigated for optimising PCB analysis (felllisari at al. 1981). NICI-M3 la uniquely aulted to measuring traea quantities of polyhalogeuated chemicals In anvlrottaectsl samples baeauaa of lta high sanaitivity for thaaa chaalcala and lta virtual transparency to otherwise potentially lotarfarlng Bolaeulaa (Kuehl at al. 1980). It alao provldaa. la addition to etusltlvity. nolaculir loo loforaacloo and thua a varlflcation of tha atructural entity being aaaaurad, a highly daalrabla faatura whan examining complex anvlronaantal and biological temples. RICJC-MS la vary iloeely analogoua to alaetron capture detection In that the loo forming raactlooa are common to both. Tha Ion forming raacelona that are Important for polychlorinated aoleculea have bean daacrlbad (Kuehl at al. 1980) . They Include; (a) tha roaldent capture of thermal electrons; (b) chloride attaehaanrt (c) daproeonatlon; and (d) oxygen exchange. Baeauaa ttlCI-MS la t novel technique, the operating paraaetera which optlalae the technique preferentially to ono of tho above aechanlema for PCB anolyalo havo recently bean etudltd In our laboratory. Invtatlgatloua In thio laboratory havo centered oa the uoa of high raoolutlon gae ehroaatography in combination vith KICI-MS while elucidating and characterising lmtruaantal paraaetera suitable for PCS analyses (Pellltaarl at al. 1981). Tbasa invaatlgatlona have included: (a) examining tha performance of two different ion source designs; (b) tha affect of source prasaura on oonaltlvlty and spectral signature; and (c) tha affect of varlouo raagant gases ou sensitivity sad fragmentation of PCS Isomers. Compared to conventional electron lapact MS. less Information about the structure of the compound is obtained. Thus, the specific elm of tha Investigation* was to study aavaral moderating and reagent gases to enhance; (1) eha formation of molecular enlona of the Individual chloroblphenyi Isomers; or (2) their dissociation to yield ehlotlda 35 and 37 Isotopic anlona (Pellltsarl at al. 1981) . Baagtot gaa stadiet vara conducted with methane, oxygen/nitrogen, nltroua oxlda/nltregam, nltroua oxlda/methane. dlfluorodlchloromethane, and tatrafluoromathane. This variety of reagent gaaea was necessary baeauaa a major problem encountered lu RIC1-MS analysis of PCSa had bean tha lack of molecular weight information obtained for tha lower molecular weight PCSa (C.-C.) under methane moderated electron capture conditions. Slaee the lower molecular valght PCSa undergo dissociative electron capture to form Cl lone under thaaa conditions, Ufa of a raagant gat which chemically raacts with Individual chloroblphauyl Isomers rathar than Just moderating electron energy lad to eha observation of useful molecular weighe information in tha apactral signature. Plaid at al., (Salt and Plaid 1980) had observed that a mixture of nitrous oxide and methane produced abundant OH- Ions under negative ion12 12 HONS 026053 condition*. The hydroxide lon war* ob**rved_to react vlth a vlda variety of compound* by proton abttractlon to fora (M-H)~ Iona. Therefore, thl* reefent |*e Mixture vee atudlad In our laboratory aa a likely candidate to provide aoleculer weight Information for individual chloroblphenyl leotsere. Experiment* were conducted with an LR.B 2091 Magnetic lector Instrument with a relatively opan aourca daalgn. Figure* 3-7 depict aaaa apecera of monochloroblphanyl, trlchloroblphenyl, haxachloroblpbanyl, octachloroblphenyl, and dacachloroblphaoyl obtained under X.O/CH, HICl condition*. Nicroue oxide va* introduced through a reagent gat.lnlet (5 x 10'3 Torr aa eaaaurad at the Fanning gauge) an^aethana (3 x 10 Torr) waa peered Into the Ion eourc* via a CC nake-up line and aaparator. Bated on calculation* aada for poaltlv* Ion Bethane Cl-the actuaL *oure*_ praaauca-exerted -by boeb--reagent gaee* va* 0.2-0.3 Torr. The MS tyttem w** optialied for /*_ 17 (OH ). A 2-uL Injec tion (MOO-pg) of a etandard aolutlon of FCB* vT*--made with a 511 apllt ratio. Aa Indicated In Figure 3 the spectrum of 2-cblorobiphanyl exhibited (M-H)' Iona. The M- peak at /*_ 188 1* no larger than expected fro* JC leotopc abundance. In addition, vary low lntenalty Iona were found for tb* (M-H+0)" ion at b/i_ 203, (M-HvHjO)" ion at n!z 231, and (M-H+MO)- Ion at a/a, 217. Tha high aa** region for trlehloroblphenyl (> /*_ 40, Figure 4) eaet spectrum waa dominated by tha (M-H)~ Iona at /*_ 257,7237, and 239. Tb* M2lntenalty waa greater than predicted on tha baeTa of l3C laotope abundance (24X v*. 132) auggaatlng that eom* atablllaatloo of aa occurring. Alao, (M-Cl)" Ion* war* evident at m/a 221 and 223. For hexaehloroblphanyl (Figure 3). tha (M-H)~ lona at / 337, 339, 361, and 363 war* praaant. Alao, substantial abundance of Ion* waa obaarvad (531 obaarvad va. 132 calculated aa 4,C). Ion cluatara for tba loa* of one chlorine (m/i_ 323) and two chlorine* (n/f_ 268} from the parent molecule were detected. --An lnteraatlng_lon eluatar Toned by the lose of chlorine and the addition of oxygen (M-19) va* obaarved (/* 338). The mate spectrum of octachloroblphenyl (Figure 6) exhibited negligible hydrogen abttractlon but vaa certainly.doalnatad by lona. Since there ar* only two hydrogen* available for abatracclon on thl* FCB hoaolog, thla raault vaa unexpected. Dacochloroblphanyl (Figure 7) vlth no hydrogen* available for abatracclon ylald* a spectrum that la due to electron capture (foraatlon of >-) and Cl". Tb* above obaaryatlont for H, 0/CH. reagent ga* Indicate that verifi cation of tha aolecuiar weight of the chloroblphenyl leoatr can be achieved In high raaolutlon QC 8ICI-MS. Furthanore, the hoaoganelty of gaa chromatographle peak* in coaplax mixture analyale can be eatabllahad aloe* non-FC> aubatancea or nonhoaologoue FCB laoaara can be dlatlngulahed. N1CI-MS will thua provide aolecuiar weight Information, which haa bean one of the atrong auite claimed for IX-MS. Another aodoratlng gae typically uaed In RICI-MS 1* aathana. Exparlaanta in our laboratorlaa on tha effaet* of aourca praaaure on aanaltlvlty indicated that aa tha aerhana praaauee-incraaaad ao alao did sensitivity. Aa a coaproalaa between naxlaua aanaltlvlty and axcaaalv* praaaure, a reagent gat prta- 13 MONS 026054 Flgura 3. HjO/CH^ NICI un ipaetrua o aonoehloroblphanyl. 14 MONS 026055 80 257 300 M/2 Fleur* 4. KjO/CH^ MCI ** pci.ru of trlchloroblphonyl. 15 HONS 026056 R e la tiv e in te n s ity 80 60 40 20 0-* .!* 4 JO 100 200 M/Z JOO PiUure S. N.,0/Cll^ NIC1 mss spectrum of heKacliloroblplieuy I. 359 4*. MONS 026057 R elaelT * JLntn*rr MQNS 026058 M/Z figure 7. M^O/CII^ NICl Mia ipectcua of decactiloroblpheny l. MGNS 026059 un of 4 s tO-5 Torr 1* normally employed (measured at diffusion pump throat, actual aourca praaaura may ba higher). Of tba raaganc and moderating gasae studied, aatbana provldaa oaa of tha mora aanalelva aoda* of oparatlon. Hovavar, eha orlnclpal mechanism of raactlon la disaoclatlv* electron captura leading to / 33 and 37 vhleh can ha uaad for quantification. ~_ Tba chemical thermodynamics In NIC1-MS ia alto affactad by tba physical daalgn of tha CC/MS aourca vhara tba Cl takaa placa. Tba asparlaaac* con ducted with a relatively opan and cloaad aourca designs gave parallel raaulta. Tba principal feature of tha opao aourca wee that It wee laaa aubjeet to ' filament carbonization and Ita sensitivity was maintained atabla fot a few weeks. Considerable abeoluta difference* In tha Holts of detection war* observed b*twain two different Instrument makes. One ayscam vac elaarly capabla of datactlng In tha high faatograa region, while tha other required two order* of aagnltuda nor* PCS material. Thu*, the advantages of both sensitivity and oolecular weight information can ba realised with HICl-MS. Tba racant studies have provided n-basis for choosing fhs approprlats operating paraaatata for analysis of PCBs. gsltctad ton Monitoring Ptllislm 11-MS Elsctron lapsee asst spactroaatry has boon a vary popular analytical tool, In addition to SCO, it a SC datactor for PCBa, Tha uas of tha conven tional scanning aod* suffer* fro* an inadsquat* limit of dateeeiow for ehloto- blphanyl iaoaara. Tha limits of datectiou aay differ batvaan It-MS and ECO by aa much as thraa ordara of aagnltuda (Pallissari 1981). Por this raaaon, EI-MS par aa has not baan widely successful unlaas there are large quantltlaa of PCBa prasant la anvlrouaaatal tad biological saaplaa. Thar* have baaa reports, however, of a apaelallscd application of EI-MS. Salactad ion aonltorlag (SIM) had baaa primarily developed for drug analyst*, but recently It has baan applisd to varifylng and quantifying PCBa (Tindall and Wlnlngsr 1980, Kartalll at al. 1981, Albro and Parker 1980, Erickson and Pallissari 1979). Xaprovsd limits of datactloa srs raallsad. Another variant of II-M3 Is Halted aass acaaalag (IMS). Tha u*a of SIM (prograaaad aoda) aad LMS paralts tha apsetroaatar to spend mor* tlma transalttlng through tba aaalysar to tba alaetroa anltlpllar Ions of Interest to yield lower Halts of dotoetloa (Tindall and Vlnlngar 1980). Both of thas* tacholqnts are under ebaputar control aad ara alto available to tho analyst whan operating la tha RIC1-MS aoda. Pulsed Positive Ion negative Ion Chemical Ionisation (PPIHIC1) Mesa isactroaatrr During tha past few years PPXXICl has baaa developedJ rapid twitching batvaan poaltlva aad aagatlva Cl (12Us) allows alaultaeoous information to ba acquired (Pallissari and Moaalay 1981). Llttla raaaarch ha* baan parforaad on ita optimisation and application to PCI aaalyalo. 19 HONS 026060 Combination of High ftaaolutlon Caa Chromatograpny od Oltraaentltlva Reaction Flgura* 1 and 2 daplet the combination of (tata-of-eha-art high resolutloo chromatography and alactron captura datactlon. It la clearly evident from tbeae profllea that modem EC detectors ara designed to preserve cbe high resolution which capillary columns are now capable of delivering. The direct coupling of high resolution GC capillaries with the Ion source of mass spectrometers has also been successfully accomplished. The flexible nature of silica capillaries has greatly faellltatad this accomplishment, sinee most M3 systems are not wall anglnaerad to accept the rigid, glass eeplllarlts. Figure 3 depicts am exempts of analysis of an Aroclor 1016/1254/1260 mixture (2.5t2.0s1.0 */w/w) utilizing ta Apiston M phase coeted (0.025-u film), PSD silica capillary and NICt-MS detection. Table I Hats the operat ing conditions which were optimized for dissociative electron capture (l.e., enhancement of m/a 35 and 37). Tha maaa chromatogram (m/z 35) axampllflte tha eloea similarity Satvaan the proportional reeponte for individual ehlorobiphenyl isomers under methane WICI condition* and thorn* obtained witb ECD. Ah aoepandad version of thle chromatogram la givan in Flguras 9-11. Chromato graphic peaks 3, 33, and 103 ara tha intarnal standard*, 1,2-dichloronaphthalana, 1.2,3,4-tatrachloronaphthalsne, and octacblorouephthalene, raspeetlvaly. The chromatographic poaka daplctad in Plguraa 9-11 lndlcata ldaal symmetry and thus tha chromatography was preserved. Tha paak residence time* ara gonorally 7-8a. Because the ECO and M1CI-MS profllea are eupetlmpoaabla, the numbering scheme shown in Figure* 9-11 was standardised for our cheracterlsatlon research end for our eroee-referencing and aerification efforts. The identity of several of the peeks will he dleeussed below. Another meene of representing RICX-MS information ia daplctad by Flguraa 12 and 13. The top treeing is the one shown in Figure 8 but it is consider ably reduced. The remaining profliaa in Flgura 12 are mate chromatogram# for m/t 188 (Cl.), 222 (Cl,), 256 (Cl,). 292 (Clt), and 326 (Cl.) raprasanting tha 7lva homologous series. It is Important to note that under methane MICI conditions, the intensity of molecular anions for Cl.-Cl, is vary waak to nouaxlatont. It ia more appropriate to use R.O/CR. for detecting which chromatographic paak balonge to a homolog. Figure*13 shows mass chromacograma for m/a 360 (Cl.), 394 (Cl.), 428 (Cl.), 462 (Cl.), and 496 (Cl., not datactad) raprasanting the remaining'homologous eerie*. In this ease all'of ehe ehloroblphanyl lsomare were detected in the Aroelor mixture. Variation of Detector leenonses A common problem with ECD, RICI-MS and El-MS (SDf, IMS) as detectors la the large variation whleh is observed between the individual chloroblphanyl lsomare, both within end between homologous series (Hutzlngar st el. 1974, loe and Egaas 1979, Albro ec al. 1981, Safa at al. 1975). Tha relative reaponse fsx'ore (RRF.) for a few individual ehloro'ilphanyl isomers obtalnsd with high -ssolurljo Cc-lCl ((GC) -ECO] have beet reported. A majority of tha laomtrs have bean analysad in our laboratory and tha ART. comparad (Tabls 2). Tha AAF^ waa calculatsd aa follows! 20 HONS 026061 n* tmmt #i tfmUt ItM/im/ttM rttt KM Im II Mill )Q|II H ** I.MM Utol! m. *.# Mm A #. I.l M*I M. I hM M U <| 21 HONS 026062 Table 1. (0C)2/NICI-HS Operating Condition* (Finnigan 4021) Parameter GC Capillary n> Length Deactivation Stationary phaae File thlckaeaa He carrier Splitleaa/apllt Teaperature Condition Fused allica 0.23 na 43 a Poly* Horace Apiezon M 0.023 (i 0.6 nL/nin (32 ca/sec) 40 sec; 10:1 100C/0.1 ain -* 260C gl.5C/nin MS Reagent gaa Foreprtaaure High vacnua preaaure Manifold teaperature Ionizer teaperature Kaiatlon current Electron energy Scan cycle Scan range cb4 0.2 torr , 4.2 * 10 3 torr 120*C 260*C 0.5 aA 70 eV 1.0 sac 30-700 daltons 22 MONS 026063 RM* itNMIfN Figut* f. Kai* ehroMtoiraa (/* 35) ot Aroclor 1016/1234/ 1260 btla4 by Cl^ MIC1-MS (Ugur* 6 xpnd4). 23 MONS 026064 M Hwuy llJtt/fl t*ilftOR lUi tVMiM fMt lx*4) n Call' tttlltlf N Mtt PM '.4 ft Mm HN itw Flgura 10. Kin ehroMt|ra (/< 35) ot Aroelor 1014/1254/1240 obtained by CR^ RICI-MS (Flgura-# axpandad--nlddl* portion). 24 HONS 026065 Mm rttwitiii>titao CIhoiini uMoBJSiu*) nN ' tii WNlM lllt/UH'tW ** Mtt !* T tapf l I# Utolr #, .# fetot . 1.9 Ia*#f f Jf, ) IU n.tn I.M Tito Figure II. Htee chronatogrea (/<_ 33) of AroeXor 1016/1234/1260 obtained by Cl^ NICI-KS (Figure 8 expended--later portion). 23 HONS 026066 9am *hiiihti ill/tM'/tltaltim,i||u<iiMtmWt/totmii naMa.**.#nm(miimaa rat. *itiiit u(tnttmfliii n* i.t imi n, i !< J lii JllLi.i.Llul.i Jilin. !.. la < Mm a* t ll ((AM Jl.flt I t.Mt II IM.tM t.Mt IM `1 |. iul. tM.tr? In. , , t.Mt itt.tt? " `1 i 11. i t.Mt n IM < mil. Ill______ iM.ttr t.Mt iim M.M itat n%m Mt mm Mat wia satt uim mm mm Mm TIM Flgura 12. Matt ehromatogran for partnt Iona of aach FOB bonolog (Clj-CIj) In Aroelor 1016/1234/1260 nlxtura. 26 MQNS 026067 fr*aa*4' iMkiai UnUi itHfIMtnCM >ltt KW UN ( Nc*UUft IUtJlSlWtUSI f<t <>> |. . UMlt 1 . . I I f. I.t l*M < tl. I lui, m UI I4M*. __________J ih JilL.i.UAi Jillii. .i. .W L_ .it, i. .i '] i_____________________________ *. , __ ii_ jJ__ t .. ill. *. W44.,JmW 1**4, J*4. III J.5M 14, .1.MM Vi nr. i- UI.1M .IN I. m.ia .IN tm li tM |>iN N WtQ 44M ** MM Ifilt MN IN INN Tim Plfur* 13. Hai chro*ato|raa for parant lens of aaeh PCS hoaolog (CIj-CIjq) in Aroelor 1016/1254/1260 nlxtura. 27 HONS 026068 Table 2.-- Relative Response Factors For Individual PCB Isoners Using (GC) /ECD Hoaolog (Xo.)` PCB Iaooer No. RRFt (XRSD) Range Mean S.D. (%RSO) Ne 1C1(3) 1 5.041 (3.9) 2 11.470 (2.3) 3 13.143 (6.8) 201(12) 4 3 7 9 10 a 12 14 15 3.555 (4.1) 2.925 (5.9) 0.523 (5.6) 0.376 (3.4) 0.619 (1.8) 2.726 (1.2) 1.231 (1.3) 0.749 (0.8) 0.142 (3.9) 301(24) 18 21 22 26 29 30 31 33 37 0.705 (3.9) 0.313 (0.3) 0.598 (4.6) 2.033 (0.9) 1.514 (2.0) 0.321 (4.4) 1.706 (2.9) 1.389 (2.2) 1.656 (2.7) 401(42) 40 42 43 44 45 46 47 48 49 SO 52 S3 54 56 56 60 61 62 63 65 66 68 69 70 0.422 (1.2) 1.252 (3.8) 0.471 (0.6) 1.086 (2.6) 1.472 (0.7) 1.585 (0.7) 0.905 (3.3) 0.460 (0.4) 1.008 (1.6) 1.150 (10) 1.176 (2.5) 1.504 (12) 0.643 (4.4) 1.004 (0.6) 1.632 (1.2) 1.160 (3.8) 0.367 (0.3) 0.918 (1.0) 1.138 (0.6) 0.479 (5.3) 1.229 (2.3) 0.812 (l.J> 0.866 (1.5) 0.978 (8.4) 3.041-13.143 9.885 t 4.27 (43) 3 0.142-3.555 1.427 t 1.28 (90) 9 0.313-2.035 1.137 t 0.(5 (58) 9 0.3(7-2.124 1.024 t 0.39 (38) 31 28 MONS 026069 Tabln 2 (continued) Homolog (Ho.)* PCB Isomer No. RRFt (%RSD) 4C1(42) (cone.) 5C1(46) 6C1(42) 71 72 73 75 77 78 81 82 83 85 86 87 88 92 93 94 96 97 98 100 101 103 104 106 108 109 110 111 112 113 114 115 116 117 118 120 121 122 123 124 126 127 128 130 1.036 (0.3) 1.039 (3.1) 0.812 (1.3) 1.096 (1.4) 2.124 (0.5) 0.831 (0.7) 1.097 (0.3) 1.123 (0.6) 0.440 (3.4) 1.358 (11.4) 1.094 (2.5) 0.999 (18.3) 0.697 (0.6) 0.904 (0.3) 1.226 (1.1) 1.456 (1.1) 1.092 (0.1) 1.047 (3.7) 1.231 (0.8) 0.944 (3.8) 1.058 (5.7) 0.857 (2.2) 0.986 (0.8) 1.324 (1.9) 1.242 (2.4) 0.450 (3.3) 1.461 (1.3) 1.059 (1.3) 0.984 (13.6) 0.875 (1.5) 1.382 (2.1) 0.879 (4.8) 0.641 (4.7) 0.901 (1.8) 1.288 (2.6) 1.031 (1.6) 0.388 (1.2) 1.243 (2.6) 0.709 (3.2) 0.509 (4.2) 8.079 HC* 1.242 (2.4) 0.952 (1.2) 0.882 (7.3) 1.384 (3.8) Range Mean t S.D. (%RSD) Na 0.440-8.079 1.206 1.23 (102) 35 29 026070 HONS Teble 2"(continued) Homolof (No.)* PCB Iioatr No. RRTt (SD) 601(42) (cont.) 701(24) 131 132 133 136 137 138 139 140 141 142 143 144 145 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 163 164 165 166 167 168 170 171 172 173 174 176 177 178 180 181 182 1.484 (0.8) 0.835 (1.0) 1.090 (2.0) 1.141 (5.7) 0.381 (5.7) 0.938 (3.6) 0.374 (2.3) 1.119 (1.5) 0.789 (4.0) 0.948 (0.7) 0.632 (0.9) 0.913 (4.4) 1.163 (8.5) 1.186 (2.9) 0.962 (1.8) 1.822 (1.6) 0.878 (0.9) 0.889 (3.7) 0.529 (3.2) 0.942 (1.9) 0.951 (0.8) 1.463 (0.6) 0.926 (1.5) 1.226 (0.7) 0.508 (13.6) 1.221 (0.7) 0.653 (1.6) 0.455 (8.5) 0.786 (1.7) 0.695 (12.2) 1.003 (1.4) 0.471 (16.3) 1.097 (3.6) 0.628 (1.0) 2.317 (0.9) 1.189 (2.7) 1.195 (5.3) 0.894 (2.8) 1.356 (3.2) 0.383 (0.8) 1.142 (7.6) 1.169 (3.9) 1.421 (1.6) 0.679 (14r7) 1.350 NC* Rente He.n 4 S.D. (WSD) N* 0.374-1.822 0.927 4 0.32 (35) 37 0.383-2.317 1.181 1 0.40 (34) 21 30 MQNS 026071 Table 2 (continued) Homoljg (No.)4 PCB Iaooer No. RWi (WSD) Range Mean t S.D. (XRSD) Ne 7C1(24) (cone.) BC1(12) 9C1(3) 10CK1) 183 184 185 188 187 188 190 191 192 193 194 195 196 197 199 200 201 202 204 205 206 207 208 209 1.079 (1.4) 1.1S9 (4.4) 1.387 (1.2) 0.771 (6.0) 1.359 (1.1) 1.S84 (2.5) 1.256 (0.6) 0.668 (9.6) 0.858 (4.4) 1.176 (1.5) 1.026 (3.9) 2.602 (7.7) 1.889 (3.0) 0.959 (8.2) 0.925 (2.5) 2.470 (1.9) 2.114 (31) 0.976 (2.7) 1.038 (4.2) 1.139 (1.2) d 1.053 (1.9) 1.816 (4.3) 1.005 (1.8) d 1.168 (2.7) d 0.925-2.602 1.514 t 0.679 (45) 1.005-1.816 1.291 0.45 (35) -- *(No.) nuaber of theoretical isoaers possible, ''internal standard 1,2-Dichloroaaphthalene. CInternal standard 1,2,3,4-Tetrachloroaaphth*lane. internal standard Octachloronaphthalene. *N nuaber of lsoaers studied; each isoaer was aeasured In triplicate. *See lallschaltter and Zell 1980. *NG not given. 10 3 - 31 HONS 026072 Aat RAF1 - i "l.s. -- x Amt l.l. where: UF relative response factor for FC8 Isomer 1 Aat11 amount PCS Isomer 1 Injected *1 peak area *1.*. peak area for Internal standard Aat^ t * Internal standard amount The percent relative stendard deviation for the MF s within a hoaologous arias ranged froa 33 to 102X., Thus, for accurate quantification of Individual chloroblphenyl lioners by (CC) SCO, the appropriate RRF auat be eaployed. The largest variation la observed with the Cl. hoaologous series. As addi tional chlorine subetltuents are introduced ofito the biphenyl nucleus, the Halts of detection do not significantly decrease and the response variation decreases since ring subeeltution pattern no longer plays a large role in determining the magnitude of response. This is not the case, of course, with the lover homologous series where the ring substitution pattern and number of chlorine substituents are important determinants of ECD response. In a parallel study, the relative aolar response (MOt,,for a/s 33) was deteralned for individual chloroblphenyl isomers using (CC)*HICI~mJ. The operating coudltloea were previously given in Table 1. The IKK was calculated as follows) ns. A1 tWi s totl.s. i.s. Wl.s. to'i where! 1. area of PCS isomer i A. area ef Internal standard KB?' molecular weight ef PCS Isomer i MV. 1 - molecular weight of internal standard Amt:*** amount of Internal standard AilJ* amount of PCI laoaar 1 The percent relative standard deviation for the SXR.s within a hoaologous arias ranged froa 33 to I25X (Table 3). This daea base Is being expanded to include all available isomers, and no conclusions should ba drawn at this tine, leverthelese, it is anticipated that appropriate IMt factors aust be employed for accurst* quantification. - A summary comparison of response factors is given In Table A for EC), JI1C1-MS. and. EI-MS .(EDO. The literature values for SI-MS unfortunately are normalised.to the lowest response withlp a homologous series. Thus, a compar ison between honologs le not possible (Martelll et al. 1981). On the other hand, it can be readily seen that EI-MS exhibits the lovest XXSD within a 32 HONS 026073 Hobolog (Ho.)* Table 3. Relative Holer Reapooac* |a/i 35) for Individual PCB ItoBcra Uaing (GC) /XICI-HS PCB IOMC HOe RRT1 (SRSD) Rang* Haas * S.D. (XRSD) X* 1C1(3) 2C1(12) 3C1(24) 1C 1(42) 5C1(46) 1 2 3 3 7 9 10 11 12 14 IS 18 21 22 26 29 30 31 40 47 44 47 53 54 35 60 61 65 69 70 72 75 77 81 85 87 93 101 106 112 114 u* 0.356 (6.8) 0.105 (6.2) 0.091 (5.0) 0.574 (10) 4.223 (ID 2.064 (0.6) 2.430 (4.4) 0.574 (4.6) 1.452 (5.0) 1.209 (1.3) 0.773 (1.0) 0.340 (12) 0.842 (3.1) 5.143 (2.6) 0.623 (3.1) 0.462 (4.3) 1.686 (4.4) 0.555 (1.9) 1.240 (6.3) 1.208 (3.0) 0.920 (4.7) 1.447 (2.5) 0.048 (16) 0.358 (0.5) 0.779 (5.3) 2.013 (1.9) 1.562 (2.3) 0.789 (0.6) 1.055 (4.6) 0.989 (8.4) 0.840 (18) 1.038 (1.7) 0.364 (3.9) 0.893 (1.7) 1.077 (2.0) 0.564 (8.0) 0.528 (4.6) 0.818 (8.0) 1.123 (4.8) 0.667 (3.1) 1.137 (4.0) 1.216 074) 0.091-0.356 0.184 t 0.15 (81) 3 0.574-4.225 1.662*1.23(74) < 0.340-5.143 1.378 * 1.72 (125) 7 0.048-2.013 0.971 * 0.48 (50) 6 0.465-1.216 0.805 t 0.27 (33) 12 33 MQNS 026074 Tbl 3 (continued) Roaolef (Mo.)1 PCB Ifoar Ho. RRIt (SD) 5C1(46) (coot.) 6C1(42) 7C1(24) 801(12) 901(3) 117 118 121 122 128 129 131 133 136 137 139 141 147 151 153 159 160 161 163 165 170 173 175 180 184 165 186 187 188 191 192 193 194 196 197 199 200 201 202 204 206 207 208 0.839 (2.5) 0.463 (8.4) 0.634 (5.5) 0.391 (2.3) 1.011 (4.8) 0.901 (8.4) 1.236 (4.8) 0.559 (2.7) 0.396 (4.9) 0.911 (4.6) 0.747 (4.1) 1.440 (8.6) 1.042 (6.8) 1.030 (7.3) 0.369 (2.9) 0.525 (0.7) 0.735 (2.5) 0.719 (2.3) 0.688 (4.4) 0.736 (3.9) 0.451 (10) 0.659 (0.4) 0.305 (2.3) 0.679 (ID 0.708 (7.0) 0.755 (16) 0.236 (4.9) 1.102 (ID 0.503 (5.1) 1.192 (3.0) 0.542 (5.0) 1.096 (8.0) 0.637 (1.3) 0.608 (3.9) 1.116 (9.0) 0.393 (6.3) 0.416 (6.1) 0.413 (3.3) 0.692 (5.9) 0.241 (9.1) 0.431 (i.O) 0.066 (26) 0.563 (20) Ran| Meu t S.D. (WSD) We 0.369-1.440 0.617 t 0.29 (36) 16 0.236-1.192 0.703 t 0.30 (43) 13 0.241-1.116 9.S73 i 0.26 (46) 8 0.066-0.565 0.354 t 0.26 (73) 34 3 HONS 026075 Table 3 (continued) Hoaolog (No.)* PCB Isomer No. RRF1 (XRSD) Range Mean t S.D. (XRSD) N lOCl(l) 209 0.418 (7.5) 4 . - - *(No.) number of theoretical isomers possible. k laterall standard * 1,2'Diehloroaaphthalene. cIoteroal standard 1,2,3,4-Tetrachloronaphthalene. internal standard * Octachloronaphthalene. N number of isomers studied; each isomer was measured in triplicate. ( Sea Ballsehaltter and 2ell 1980. - 35 MONS 026076 Table 4. Comparison of Relative Response Factors Between (CQ'-BCP, CC-B1MS (Molecular low) and (GC)*-NICIMS (n/s SS) for llonologous Series of PCBs Rtaclt* bil at*4 (ao'-i* Data t a.a. (tase) i* ICI(l) Kl(ll) SC1<4> 4CK41) KI(U) !(*#) tei(*() KI(IS) Kill) ItCI(l) IS.Mt-M.SU 411-IM4I a.ua-s.m a.iti-i.m S.Ul-S.UI .m-i.au a.tei-i.au .an-! .tai l.sM-i.ut - HU! 4 U.IS (41) 4.IH * 1.S1 (*) i.Wi a e.M (si) 1.1)4 1 1.41 (11) 1.1*4 a I.1S (111) a.sis s a.iis () i.ue a e.4 04) in* a a.tn (41) i.m a a.4i (is) i.isi i a si is s? it i* i l vtrtllt UM'M.Mi (-11*1i) 'at! a*u. Wulli, U t> . IMI. 1 mSu i( KS licU!t! | MHMfM (eo'-Micina* CC-IIHS* bi|<4 a.4S4-i.f*r Mata * M. Wit) taa|c* Mata t l.l. (lUO) 1 .i4 a . (si) 1 1.BOO-1.MO i.m* a a.a* (!) a.ui-n.iM .141 a *.)> 04) 1.M0-1.M} i.)M a a.so (ID a.m-ia.aai a.aai a s.t* on) 1 l.aea-i.ti) t.4M a a.it OD .iai-4.it) i.m* a t.ei os) It i.eee-i.iu i.sta a a.n HD .4*1-1.11* .ms a .i) (id 11 l.aaa-i.au i.m* a a.ai (*.)) .MS-1.44* .id a . (it) 1* 1.***-!.Ill i.m a a.ii (it) a.iM-i.m .)! a *. (4i) 11 - - .141-1.II* .in a (.1* (U) a i.ooa-i.is* i.iis a a.is OD .**-.1*} .is* a a.i* oi) l - - - .41* i- - vanHi .**-!).m. (-!) 1) ^411 nlui ( nlitlva U utickUnuphikilt M. *a*aa*ai wtn rtlatlva la Itwal raafaaaa (at aacli |[W|. '( ) - aiaatar ( tkaaratlcal laaaart pottlSW. 1 1* II 02607? HONS homologous series when compered Co ECS end NICI-MS. Some caution 1* needed In chle comperleon since the number of cbloroblphenyl Isomers need In Che study vis email. This observation la consistent with the expectation that EI-MS produces the smallest variation of response between chloroblphenyl Isomers. The extreme REP values determined with ECD were approximately 120:1. This ratio would be lomevnac larger If the calculations were made and expressed as RMRs (which Includes M.V.). Nevertheless, the range of RRFs Is not at large as reported by other Investigators (Butzlnger at el. 1974, Boa and Egaes 1979, Albro et el. 1981. Safe at al. 197)). Ve believe thae this Is associated with differences between Instrumental tystams, since In our laboratory we have determined RRFs on three different (GC)-ECD systems (all with the same capillary column conditions), two of which are of the same make and model. We have observed significant differences In RRF values between all systems. The extreme RMS. (calculation Includes M.W.) values determined with NICI-MS were approximately 320:1. The Importance of using appropriate response factors for quantifying Individual chLoroblphenyl Isomers cannot be overemphasised. CTlOROBIPSafTL ISOMERS AMP INSTRUMENT CALIBRATION Primary Standards The Issue of Instrument calibration for chloroblphenyl Isomer analysis In environmental, biological, and process stream samples has been recognised by Investigators. As such, a program for the synthesis of the Isomers was Initiated. Primary standards ara needed for establishing reference data (spectral) banks for RICI-MS, EI-MS, FTIR, etc. for use In qualitative analysis or for verification purposes. They are needed also to establish relative retention Indices for standardised (CC)* operating conditions. The large scale synthesis of chloroblphenyl isomers is obviously an expensive proposition, particularly If the supply must accommodate the needs of many analytical laboratories performing Isomer quantification In environ mental, biological and process stream samples. For this reason, our labora tories have been actively engaged in devising alternate techniques for Instrument calibration. Secondary Standards Our objective has been to determine whether a secondary standard for calibrating instruments for quantification could be developed to that this secondary standard could be widely distributed among laboratories Involved In chloroblphenyl Isomer analysis. The initial availability of the primary standards (209) Is the key to developing a secondary standard. Tvo approaches are currently under investigation in our laboratory. One utilizer thscharacteritatton of an Aroclor "cocktail"-using the primary standardsi the other employs a "clustering" algorithm to derive a small subset 37 MGNS 026076 of ehlorobiphenyl* froa the 209 prlaary atandard*, vith a ehlorobiphenyl aerrlng aa a aurrogate for aavaral laoaara. Bach of thaaa coactpta la dlaeuaaed Kara. Characterization of Aroclor Mixture* Sinea aapla quantltla* of varloua coanerclal Aroclor* ara available for dlatrlbutloo, the poaalblllty of prapariog an Aroclor "cocktail" which la thoroughly charactarlxad aa to laoaar apaclatlon and quantity waa undertaken. Aa Aroclor 1014/1244/1240 alxture (2.3t2.0tl,0 w/w/w) waa praparad. Tabla S llata cha laoaara tentatively Identified In thla alxture. Characterliatlon haa eaployed the uaa of relative retention tlae data for prlaary acaudarda and matching with thoaa In the.Aroclor trlpart alxture for chrea different (CC) -BCD ayatan* and one JC) -MICI-MS ayaten. In addition, the aoleeular Iona and aaaa apactra obtained froa (GC)*-MIC1-MS and (CC) -EX-MS analyala of the Aroclor trlpart alxture were uaad to eatabllah the identity of tha coaponanca In each of tha chroMtographle peaka. The peak nuabar aealgnaente in thla table eorreepoud to ehoee given In Flgurea 9-11. M indicated in Table 3 a majority of the chroaatogrephlc peaka have been aaalgned ehlorobiphenyl laoaar ldentltlee. In aoae caaea the ehroaatographie peaka are hecerogeneoue and aeverel laoaara were Identified. A fev of Che aore toxic laoaara (ayaaetrical 3,4 and 3,4,3) have been detected la thla Aroclor alxtvre. A fewpaaka reaaln ae unknown*; noat are alnor in quantity. Aa additional pure ayntbetlc ehlorobiphenyl laoaara becoaa available, the reaalnlng unknowna will be identified, if poaeible. Aroclor 1016, 1221, 1242, 1234, and 1240 are alao being characterized and will be reported elaewhere. A coaplete Aroclor "cocktail" characterItation (qualitatively and quanti tatively) would allow ita uae ae a eeeondary atandard for inetrunenc calibra tion provided that Identical high reaolutlon ehroaatographie condition* are aaployed In aampla analyala ae need for the Aroclor characterization. Cluatarad Secondary Standard Another concept being purewad la ehe development of a eeeondary atandard which containa a ana11 aubaet of Individual ehlorobiphenyl laoaara that can be uaed to develop inetrunent reeponaae for all PCS* of interact. The prellalnary reaulta of thle approach are diaeueaed here. The reaponaa factor* relative to the 1,2,3,4-tetraehloronaphthalene peak (TCM) for each PCI leaner were calculated ualng the axpreatloa given earlier. Tha calculated reaponaa factor* are Hated in Table 4. Frou the 33 llatad reaponaa factor*, a calibration table waa generated uelng a cluaterlng algorithm baaed on the Student t-ta*t. Reaponaa factora were Hated In lncreaclng retention tlae order of the correapondlng peake. For each eonaeeutlve pair of reaponaa factora the Studvnt r.-tiet wee 'perfornodl 38 MQNS 026079 Table S. M.k So.* PC8 Isomers with Retention Time Values Matching Peaks In Aroclor 1016/1254/1260 Mixture CSmlc.l Ntk . *- Ckaalcal 01 33 2.2'.3,3',4 l 34 2.2'.3,3,3',4 2 2,4 4/0 3 4/0 4 37 2.2'.3.4,S',4 1 l,2-0ichlcMfhtJ44lu (i..) 34 2,2',3,4',3,4 A/0 3,3',4,4', A/0 4 2.3 2,2*,3,3'.3,3' A/0 2,2'.3,4',3'.t s A/0 2,2',3,4',}'.4' < 3t 2,2',3,3',4,4 A/0 2,2',3,4.4',4 ; 3.3 * A/0 2,2`,3,4,3,4 to 2,2',3,3',4,4' A/0 2*.3,4,3,3* 10 2.2*.3 A/0 3.3 41 2.3,!',4,3 62 2.2',3,4,4' A/0 2,3.4.4'.3 u 43 2,3',*.4',3 u 44 2,2',3.4,A',4,4' 13 43 2,3,4,4',3 14 IS 4,4' It 2,2',3,4' 17 It 1,3* ,3 44 47 44 2,3,!',4,4' A/0 2,3,3',4,3' ,4 tt 2,2',3,4,5,5' 70 It 1,2',3,4 71 2,1',3,4,4',3 VO 2,3,3',4',3',4 20 72 11 l.f.S 73 2,2',3,1',4,S 12 2,3,4 23 I'.S.t 74 2.2',3,4,4',3 A/0 2,3,3'.4,3,4 A/0 1,3,3',4',3,4 24 7S 23 2.4,3 A/0 1,2',3,3 74 1,3,3*,4,4',4 2t 2,2'.4,S' A/0 2.2',4,S 77 27 2,2',3,S' A/0 l.S.S.t 74 2,2',3,3',4,4' VO 2,2',3,4',3,3',4 21 1,2',*,*' VO 1,3,4,4 It 30 3.4,4',4 VO 1,1',1,4' VO 1,2',3.4,4',3,4' 7 2,3',$,4,S,5 ,4 VO 2,2',3,4,4',3',4 to 2,2',3,3',4,3,4' 31 41 2,2',3,3',3,3*,4,4' 32 1,1',3,3' 33 42 2,2',3,3',4',S,4 43 2,2',3,4,4',3,4 VO 1,2',3,3',4,S.4 34 44 2,3,3',4,4,3' A/0 2,2',3,3*,4,3',4,4' 33 1,2,3,4-Tit(KUnuHUiltM (1.*.) VO 1,2* ,3,S' .4,4',4 3t 43 2.1'.3,4,4'.3,4.4' 4/0 1,3',4,4*,3,3' 37 2,1',3,3,4 VO 1,3',S,3' 44 1,1',3,3'.4,3,4.4' 31 2,3',3',4 It 47 44 40 41 2,3',4',3 VO 1,3,3',4' 42 43 It 2,2',3.4.4',3,S' VO 2,3,3',4* ,3,3',4 to 2,3,3',4,4*,3',4 tl 1,2',3,3 ,4,4',3 tt 1,1',3,3',4,S,S',4' u 2,3,3',4,4',3,4 43 3,4,4',S VO 2.2',4,4',4,4' VO 1.2',3.3,S' !4 1,2',3,3',4,4*,3,4' ts 44 1,2',3.4',4,4* tt 3.3',4,4',3,3' 47 1,2'.3,3'.4,S' 44 2,2*,4.3,3* 4/9 2,2',3,4,4,4' 4t V 1,2',3,3',4,4*,3,4 tl tt 34 1,3,3',S,4 4/0 1,2',3,4,3 100 2,2',3,3',4,4',3,3' SI 2,3,3',4,4 101 S3 a,3'S.4,S' S3 1.3,4',5,4 102 2,2',3,3*,4,4',3,3',4 103 Octllnu|kUi!ui (!) S4 *See Figures 9-11 for peak no. assignaents. A/O and/or* 39 MQNS 026080 Table 6. Relative Reepoaaa Factors for PCS Iaoaara Peak Ho.* PCB Iioaer RRFb fC 10 2,2',5 a/o 2,4,6 3.61 1.84 11 2,2*,6.6* 7.69 0.71 18 2,2,5' 2.87 0.92 23 2'.3,4 24 2,3,4' 25 2,2',5,5' a/o 2,4,5 27 2,2',3,3' a/o 2,2'3'S 30 2,4,4',6 a/o 2,2'3,4 37 2,2',3,5,6 a/o 2,3',5,5' 41 2,3',4',5 47 2,2',3,3',6,6' 48 2*2',4,5,5' 52 2,2',3,4,5' 2.20 0.76 3.61 2.98 2.30 2.61 2.13 5.30 2.85 1.81 0.09 0.03 0.67 0.26 0.51 0.17 0.40 1.57 0.21 0.13 53 2,3,4',5,6 2.22 0.16 58 2,2',3,4',5,6 a/o 3,3',4,4* 6.36 0.59 59 2,2',3,3*,4,6 2.31 0.14 61 2,3,3*,4,5 1.24 0.41 63 2,3',4,4',5 2.32 0.22 64 2,2',3,4,4',6,6' 2.13 0.50 68 2,3',4,4' 1.24 0.40 69 2,2',3,4,5,5* a/o 2,2',4,4',5,5',6 1.47 0.20 74 2,2',3',4,4',5 2.24 0.09 78 2,2',3,3',4,4', a/o 2,2',3,4',5,5',6 1.66 0.15 80 2,2',3,3',4,5,6' a/o 2,2*,3,4,5,5',6 1.11 0.14 81 2,2',3,3',5,5',6,6' 2.07 0.56 82 2,2',3,3*,4',5,6 2.08 1.13 84 2,2',3,3',4,S',6,6' 3.51 0.35 85 2,2',3,4,4',5,6,6' 1.29 0.07 86 2,2',3,3*,4,4',6,6* a/o 2,2',3,3',4,5,6,6* 1.34 0.34 89 2,2',3,4,4',5,5' a/o 2,3,3',4',5,5',6 1.58 0.55 90 2,3,3',4,4',5',6 1.07 0.33 93 2,2',3,3',4,5',6 0.34 0.13 96 2,2',3,3',4,5,5',6,6' 1.27 0.39 98 2,2*,3,3',4,4*,3,6,6' 4.47 1.00 100 2,2',3,3',4,4*,5,5' 1.27 0.18 *Peak auaber aa la rigara 9-11. *R*F relative reaponee (actor. " etandard deviation, a Maple alta. ad 6 3 3 3 3 6 3 6 3 6 3 3 3 3 3 3 3 3 3 3 6 3 3 3 3 3 3 3 3 6 3 3 9 3 3 40 HONS 026081 t - OWFt - RWin) . * . (1/Bj l/n1+irl/2 uhet# t 1* the Student C value; MF. and MP, . vara response factor* for sequential. packs; I Is poolad standard deviation, and n. and n. , ara nu&bar of runs, fooled standard deviation was calculated as; 1 1+1 . (nt - 1) CtM ~ 0 .*!! -2 ..hers sA and *i+l vara standard deviation of MF^ and respectively. If tha,calculatad t vas lover than t (taken fro* Student t tablas) for " degree* of freedom and <i 0.95, the two response factors vers .tpiacld by a coasson naan (MF^). IRTJ V Wi * Vl ni * Vi -2 1 * "i 1 'j I *----------------------I nt Vl vhara a^ Is the standard deviation for MF^. MF 1*1 After all pairs vara tasted, ths clustering procedure vas repeated to generate a second calibration table. This procedure vas continued until response factor* vara statistically different fro* the cvo nearest values. The clustered list of response factor* la presented In Table 7, Chromatographic peaks than ara Identified as PCI leoaers by (etching retention' time vltb an Aroclor calibration mixture. Semple component* vltb retention tlmoe matching peaka In the Aroclor etandard mixture vere identified as FCBs in the sample*. Each matched peak vas calculated asi PC81 - MFt x Atu X AIS If the PCI peak does not correspond to a calibrated peak, the response factor of the nearest calibrated paak vas used. The "total PCI value" fur the sample than vae-given by; PCI t FCIt 41 HONS 026082 Tabl* 7. Clustered Lite of Ralatlva Rasponaa Factora >akKo. 10 11 18-23 : 24 25-27 30-41 47-48 52 53 58 59 61 63-64 68-69 74-82 84 65-90 93 96 98 100 RRF 3.61 7.69 2.54 0.76 3.40 2.29 4.08 1.81 2.22 6.36 2.31 1.24 2.23 1.39 1.79 3.51 1.37 0.34 1.27 4.47 1.27 a 1.84 0.71 0.45 0.03 0.55 0.39 1.00 0 r 0.16 0.59 0.14 0.41 0.35 0.25 0.42 0.35 0.38 0.13 0.39 1.00 0.18 a 6 3 6 3 9 15 6 3 3 3 3 3 6 9 21 3 15 3 9 3 3 42 026083 HONS Aa example of the us* of this concept ha* been applied to quantifying the total PCB weight for an Aroclor 1016/1254/1260 mixture uilog (CC) -ECD: Observed weight: S62 Actual weight (gravlaetrlcally determined): SSO Error - 562-350 x 100 2.2X 550 By expanding the "clustering" algorithm analyala of reeponaea to include all the available chloroblphanyl laonera, a aurrogate conpoeltlon containing a eaall number of laomera (20-407) will be defined. Thue, only thla aac of chloroblphanyl laonera need* to be aynthaalxed on a large tcel*. AFPLICAII0H OF STATE-OF-THE-ART METHODS 10 PCB AWAITSIS IH ESVIKOWMEKTAI, AW) BIOLOGICAL MATRICES ~ With the#* recent parallel devalopnenta laboratorlaa have begun to cbaracterlxa and quantify chloroblphanyl laonera In environmental) biological, and proceae atream tanplea. Examplea of the combined atata-of-che-art technology described above are discussed bare. The benefit of (CC)2 over conventional packed coluam gas chromatography for analysis of sera and adipose la exemplified by figures 14 and 15, respec tively. Figure 14 depicts the profile for a human serum sample extract chromatographed on a packed column (top) and capillary (bottom). .Improved resolution and hence better signal to noise la realised with (CC)2, Figure 15 shows that the low resolution profile has many hidden Isomers In lta chromato graphic peaks which are revealed by (CC)2. Identification of specific chloroblphanyl Isomers in human tissues may be Important for tvo reasons: for assessment of long-term perslatenca and for evaluation of potential health effects aa suggested by toxicological studies on Individual isooars. Recently, the disposition of PCB laonera In occupa tionally exposed persons has bean raporeed (Volff at el. 1982). The concentra tions of FCB* la adipose and plasma vara related to duration and intensity of exposure In the workplace. It wee reported that FCB levels In adipose tissue were proportional eo those In plasma (total FCB ratio of 190:1). The distribu tion of specific chloroblphanyl Isomers between plasma and adipose tissue was reported, however, to be related to specific ring position substitution, differing among Isomers. Similar observations have been reported in our laboratory. For these reasons, and because of a 11mltad number of isomers end suboptlmal analytical conditions In the recant aeudy, research was undertaken to characterise the individual Isomers In sera and adipose of occupationally exposed Individuals using state-of-ths-art techniques. Using approximately 165 primary rafarenes standards (individual FCB Isomers) and (CC)-NICI-HS (RRT and molecular Ions) the Isomers were tentatively identified (Tables 8 and 9). Also the detected end Identified peeks vert referenced against an Aroclor 1016/1254/1260 standard nlxturs. 43 HONS 026084 0 2.0 4.0 6.0 9.0 10.0 12.0 U.O 16.0 18.0 TIm (ala) Figure 14. Chromatograms of huaan aerua sample, fop - packed column. bottom - capillary. 44 MONS 026085 TIm (ain) KCD response J 10 20 30 40 50 60 70 80 90 100 TIm (ala) Figure u. ChroMtograai of huaan adlpoae tlaaua. Top - packed coluam, bottoa capillary 45 02608b mqns Table 8. Individual PCB 1 toner> Tentatively Identified in llunan So run hm ut (yui) 1.0 fl (0) l.'VO (0.01) l.Ofl (0.10) i.ow fo.m 1.14) (0.10) 1.147 (I II) toil (0.11) till (0.10) 1.001 (0.11) 1.001 (0.11) 1.011 (0.10) 1,1,1,4'TiitackUiMipktkiltaa ' Mia HI (USD) 0.411 (O.U) 1.000 (0) 1.110 (0.01) I.OM (0.10) I.UI (0) i.m (a) i.m (o.oi) 1.110 (0.11) 1.141 (0) 1.1)0 (O.M) 1.104 (O.M) OcMCfcUrMMffckOtltM Mia oot (taw) 0.100 (0.10) 0.4)1 (0.11) I.OM (0) 0.410 (0.11) 0.111 (0.11) I.UI (0) 0.1)1 (0) 0.1)1 (0) 0.501 (0) 0.100 (0.10) 0.411 (0) l.llt (0.10) }.0 (0.10) 1.111 (OH) i.ooi (o.ti) I.101 (0.11) 1.001 (O.U) 1.011 (O.tO) 1.001 (0.11) 1.101 (0.10) 1.000 (0.10) 1.01) (0.11) i.roi (o.u) 1.101 (O.U) 1.011 (0.11) i.m (o.io) 0.011 (0.10) 4.HO (O.tO) 4.001 (0.11) 1.140 (O.M) i.m (o.ao) 1.410 (O.M) 1.441 (0) 1.14) (0.0)) I.OM (O.M) 1.4)4 (O.M) 1.401 (O.M) I.OM (O.M) l.ltl (O.M) 1.14) (O.M) 1.100 (O.M) 1.101 (O.M) 1.410 (O.M) l.il) (O.M) 1.400 (0.0)) t.))0 (0.01) 1.041 (0.01) 0.411 (0) 0.011 (O.M) o ut (0) 0.000 (0.00) 0.40) (0.00) 0.401 (O.M) 0.401 (0) 0.400 (0) 0.10) (0) 0.114 (0) 0.110 (0) 0.740 (O.M) 0.)SI (O.M) 0.)04 (O.M) 0.)tt (0) O.MI (0) 0.010 (0) 0.014 (0) Vroclor 1016/12S4/1260 nixture, see Plfures 9*11. CktarlaitiO Ckialcil CwnpM/lii 4rcUr Nik M. 1 J'OUtiUrMipktbileu 1,1,1,4-tcUacktaiaaipfctkiWai OclICfclolMIpklklltat 1.1-,S,5' i/0 1,1',1,1,1 1,1*,4,1,1', 4/0 1,1',1,4,4,4* 1.1.1'.4.4 1.1M.4.1' OM 0 1,1',1,1,1',4 !,}`,l,4',i,t 4/0 1,1',4,4* 4/0 1.1',1,4,1, 4/0 1,1',1,1',5,1' 4/0 1,)',1,4',V,4 1,1*,),)'.4.4' 4/0 l',),4.S,I> 1,1*,4.4',1 4/0 1',1.1',4.1 1.1,1',4,4', 4/0 1,1,)',4,l',4 1.I1.).4.1.1' 0. l,!\l,4,4',i 4/0 1,1,1',4',1',4 1,1',1,1',4,1 1,1',1,4,4',1' 4/0 1,1,1',4,1,4 4/0 1,1,1',4*,1,4 1,1.1',4,4',4 1,1 ,),)',4,4', 4/0 1,1',1,4',1,1',4 4/0 1,1',1,4,4',1.4' 1,1),1,4,1,1',4 4/0 l.l'.l,4,4',i',t 4/0 l.l'.i.l',4,1,4' l.l'.l 1,1',1.4,4',1,4 4/0 1,1',1,1',4,1,4 1,1*,1,4,4',1,4,4* 4/0 1,1',4,4*,1,1* 0 1,1',1,4,4',1,1' 4/0 1,1,1',4*,1,1',4 1,S',1,1',4,4',1 i,r,).i'.(,(,,i,s' 1 11 101 17 41 11 11 54 14 11 40 41 44 40 70 71 71 74 74 71 70,10 11 11 fS 17 10 01 IM HONS 026087 Table 9. Individual PCS laonera Tentatively Identified In Hunan Adlpoee Tieeue t,|*9<4 9e*<41*ne*lWlt*n hriiiHeliH ' iwii <1M#> ' MT (UM1 AllMMlM (iMIfll hM le. 4.941 (4.111 I MI <4 40 1.194 (114) 1.941 (4.11) MM (4.141 I.Uf (9.191 l.ltl (9.44) lift (9.141 1.441 (9.19) Mil (9.It) |.M1 (4.4V) 119 (4.99) MI4 (9 )1) Mil (I II) i ni 44.it) t in (t.U) Mil (4 141 1.944 (9.14) 1.411 (9.11) M?t (9.19) Mil (l.ltl |.|49 (Ml) 1.199 (9.14) l.tlf (4.Ill 1.419 (9.11) 1.491 (9.111 1.499 (9. U) 1.4H (4.U) I.M9 (9.11) l.ltl (9.19) l.ltl (9.14) 1.199 I9.lt) 9.449 CI.U) 14t) (9.191 1.449 (9.11) t.tll (9.19) 1.919 (9.141 9.499 (4.91) 9.999 (4.14) Ml) (9.11) i.m (9.14) 491 (4112 4.191 (|.l 4. IN (9-14) 4.111 (9.14) 4.I4| (l.ltl 4.1(4 (1.4)9 4.M4 (l.lll V.4I6 (4-16) 1.499 (9) l.ltl (I.H) I.Sfl It-IM 4*1 (4.44) 4.214 (9.99) 4.Ilf (9 99) Ilf (4.42) 9.441 (l.lll 1.1)4 (9.441 I MS (4.94) 1.991 (1.9)) 1.144 (4.4)1 Ml) (l.lll 1.1*4 (4) l.lft (II 1.194 (1) 1.144 (9) 1.199 (1.91) 1.119 49) 1.149 (9.94) 1.149 (9.491 I.I9S (4 94) I.H9 (4.94) i.m (i.m 1.1)9 (194) Mil 14 44) 1.499 (1) I.M* (4.It) 1.444 19.94) l.ltl (1.94) l.e> n.Ml l.ttf ., Mil (II MIS 14.99) t.tlt (4.91) I.M (9.9M MM W.MJ i.m (t in M99 (9.94) l.lll 11.94) 1.449 (1) 1.1)4 (I.M) i.m (I.m) I.m (9.94) i.m (4 i) 111* (9.9)) M9I (4.911 MM III Mil (9.99) i.m Ml i.m (i.m i.m (t.m * 9.191 If.til l.4f) (4.111 1.999 (1) 4.1)1 (91 i.m (t in i.m Mi i.m <4.id 9.19) (Ml) 4.411 (1.14) 9.419 (9) 4.141 (9) l.tlf (1) l.lll (11 l.ltl (41)1 4.144 (9.11) Ml) (1) 9.444 (4) 4.144 (1) l.ltl (1) i.m (i) 9.449 ID I.M4 (9.14) 9.414 (1) 1.419 () I MS (1) 9.911 (II 1.999 (It .<M <l ,.M, M.MI 4 441 (t) ,.M 1*1 9. Ml Ml ,.! <> .Ml () ,. n, (., .Mi Ml i.m i.m ,.e m.mi I.M ( I.M) (4) 4.H4 Ml *.! IN I.Mt Ml e.n* <> 1.191 Ml 9.419 111 4.4H (1) I MI (4) 9.449 14 42) Ml (4) MH M.MI m t,.,n 4.1)4 (1) * .... I ,M>{liireaa*tiUIM ItUeUlfiMflUaltM 1 f r,M t.i.s u% t.r.t.i wt i.rj.t.t t i,tv ,* vi i.i'.v.i f l,t`.M,)' 4/9 t.l\S.4,l,4* V# 1,1 I.U.M f vt vi ti* ,i,i,i,i* vi i,t'.uM,r 4* i.r.j. i.r.MM 1,1,1*,4,*' V 1,1,4*,49*,l i,r,i,u,r t.l*.M,*, VI |,|*,1,1,4',S* 4/4 41 i.,r,4*,M t t 1,1.1*,4,4*, I,I*,1,4-4,1 ,1 4/4 U',),M,r, VI V# M M.r.M.I' I.IM.IM,**,#.** l,l<,),4,)',U VI 1,1* .1,1* .4.1.4 <.I**>,4.4*4.M* */ t.>*,4,*\l,t* t mu l.i*!4.4,,,* 9/9 t.j.l'.i1 J.IM i.r.j.r.i.i' .* i.i*.i,r,4.4,* * i,r,i.i*..4*,,4* T |KI t (Kl) 1,1*,1,1*,4,4*,S1* *,4.1.4*.*.4*.4,1'.* ) a to it if n u &i 41 *r 44 U M M M M I m i u u i M 4 n it it H H 94 It 13 "xroclot 1016/1234/1260 mixture, eee Figure* 9-11. 47 MONS 026oaa Table 10 provide* a composite lifting of PCS Isomers tentatively Identi fied In 4 (era and 4 adlpoea of thl* population under study. The-presence of the symmetrical 3.4,3 and 3,4 lsoaers was tentatively established. These tvo Isomers are presuaably aaongst the aost toxic of the FOB lsoaers. The quantifi cation of the Individual lsoaars in this sat of about 100 Individuals will be reported elsewhere. These state-of-the-art technique* are currently being applied to the analysis of samples of rain, surface water, sedlaent, fish, human Bother's allk, asternal cord blood, etc. (MullIn 1981). Figures 16-21 depict exaaples of profiles obtained for specific cases cited. Finally, an area of considerable activity Involves the developaent of computerised data analysis systsas. With the use of automated CCs, a con siderable aaount of data Is gensratsd when using (CC) . ComputerItat Ion Is being dons not only to facilitate calculations of FCB lsoaer end total FCB levels In samples, but la also being developed to assist In answering com parative questions about saaplss (Stalling at al. 1982). Questions relating to blotransforaatlon, distribution and fat* through an eeosyatea and sources require sophisticated pattern recognition techniques for coaperlng and relat ing FCB information between environmental and biological saaplss (Stalling at si. 1982). swum Significant studies have been performed during the pest 3-4 years In developing! (e) high resolution CC capillaries tailor-made for FCB analysis! (b) altra-saneltive electron capture detectors compatible with capillary column flow rates] (e) an expended linear dynamic rang* of operation for the BCD] (d) the operating conditions for RXCI-MS necessary to asasure end characterise FCBs in environmental and biological samples; and (a) eh* devel opment of primary and secondary FCB standarde for calibration of Instrumenta tion. The research in the latter area continues as more work need* to be done. Investigators are just beginning to reap the fruits of their labor by applying this advanced analytical methodology to real world problems. 48 MONS 026089 Table 10. Tentatively Identified PC8 Isomers Detected in Four fluaan Ser3 and Four Adipose Samples by (GC1l/ECD and (GC)2/NICI-MS/DS Kl Iiau Corr?odtflS ArvcUr rtk .* /j.r.M1 a o j,j* i\s,4 1.4,3 A/0 J.J'.J.S ,42,4,4* A/0 2.2',3.4' 2,2' ,3,1,4 A/0 2,J'4,J' 1,3',4',3 A/0 2,1.3',4' . 2.2*,3.3',4,4' 2,2' ,4,3,S' A/0 2,* ,3,4,4,4' A2,3,3',5, A/0 2,2*,3,4.3 2,3,3',4,A 2,2',3,4,3* 2.3'.3.3,3'.4 2,2*,4.4*,3,A' 2,2',3,4',5,A A/0 3,3',4,4', A/0 2,2',3,3',3,3' A/0 2,2',3,4',3',A A/0 2,2',3.4,3,4' 2.2*,3,3',4,A 1/0 2,2',3,4,4',4 A/0 2,2',3,4,3,A 2,2',3,3',4,0* A/0 2*,3,4,3,3' 2,3,3',4,3 3.3',4.4',$ 2,3,3',4,4' A/0 2,3,3',4,3',A 2,2',3,4,3,}' 2,2\3,A,4',3 A/0 2,3,3',4',}',A 2,2'.3,3',4,3 2,2',3,4,4',3' A/0 2,3,3',4,3,4 A/0 2,3,3',4*,3,4 A2,3,3',4,4', _ 2,3',3.3*,4,4' A/0 2,2',3,4',3,3',A A/0 2,2',3,4,4',3,4' A A2.2',3,4,3,3', A/0 2,2',3,4,4',}', A/0 2.2',3,3',4,3,4' 2,2',3.3'.3,S'.4,4* 2.2',3,3',4*,3,4 2,2',3,4,4',3,4 A/0 2,2',3,3',A,3,A 2,2',3,4,4',3,4,4' A/0 2,3',4,4',3,3' 2.2'.3,4,4' ,3,3' A/0 2,3,3*,4*,3.3',4 2,2',3,3',4,4',3 1,2',3,3',4,3,3'.4* 2.3,3',4,4',},4 2,2*,3,3',4,4',3,4' ^ 3.3',4,4',S,3' 2,2',3,3',4,4',3,A 2.2',3,1',4,4',$,$' 2,2',3,1',4,4',3,3',A 12 23 23 30 37 40 41 42 43 u 44 SO 31 32 34 34 * 30 St 00 41 A3 40 40 71 71 73 74 73 74 74 73,10 11 12 A3 13 47 44 19 91 12 93 94 93 94 97 too 102 49 MONS 026090 Figure 16. Chromatogram of precipitation sample showing PCBs and PCTs. 50 MONS 026091 Saginaw livor: glass capillary eoluan :I u u u0 U Figure 17, Chroaaeograas of surface vatar saapla. Top - packed eoluan, bottoa - capillary. 51 MONS 026092 Saginaw Bay aadiaanc Packed colunn _L 34 Tlaa, sin 3 & 2. ft* O i w j nu & 100 70 30 WJ llaa, sin Figure IS. Chroaatograaa of aadlaane *tapla. Top - packad column, bottoa - capillary. 52 HONS 026093 Fl|ur 19. Chromatogram* of fiah aampl*. Top - packad column, bottom - capillary. 53 MONS 026094 i a u uuo 4s1 & Figure 20. ..Chromatogram of human moeher'a silk. Top * peeked column, bottom - capillary. 5* mons ^6095 Ill Figure 21. Chromatograms of blood sample. Top maternal, Bottom - cord (Matched pair). 55 MONS 0^6096 ACKNOWLEDGEMENTS The author* viih to thank Dra. J. Buney tod K. Toaer for their helpful tug|*tlona In the courae of the uu ipectroaecry raiaarch. Thi research vat aupported by 0. S. EPA Grant Ho. 807167*02 and U. S. EPA Contract Ho. 68-01-5915. 56 MONS 026097 REFERENCES Albro PW, Corbett JT, Schroeder JL. 1981. Quantitative characterisation of polychlorinated biphenyl mixtures (Aroclorsm 1248, 1234 and 1260) by gas chromatography using cspllUry columns. J. Chromatogr. 203:103. Albro PW, Haseman JR, Cleaner TA, Corbett BJ. 1977. Identification of the Individual polychlorinated blphsnyls in s aixture by gas-liquid chromatography. J. Chrooacogr. 136:147. Albro PW, Hass JR, CnnssMtt VB. 1979. Summary of ths workshop on recent dvancss In soslytlcsl techniques for halogenated aromatic compounds. Ann. N. T. Acad. Sci. 320:123. Albro, PV, Parker CE. 1980. Canaral approach to the fractionation and class Jaceralnaelou of complex mixtures of chlorinated aromatic compounds. J, Chromacogr. 197:135. eallscnvlctar K, fall M. I960. Analysis of polychlorinated hlpbanyls (PCB) by {Ujj capillary gas ehromatogrsphy. Composition of technical Aroclet- and iV..lopnea-?CB mixtures. Pressnlus Z. Ansi. Chen. 302:20. Bos B, Egaas E. 1979. Qualitative and quantitative analyses of polychlorlnaead biphenyls by gas-liquid ehromatogrsphy. J. Chromatogr. 180:127. Brinkman OAT, DeVries C, DeRok A, DeJonge AL. 1978. Dlscrimioatloo between polychlorinated naphthalenes and polychlorinated biphenyls. J. Chromatogr. 152:97. Chau AST, Seapscn CJ. 1975. Electron capture gas chromatographic methodology for the quantitation of polychlorinated biphenyls: survey and compromise. Environ. Lett. 8:89. Erlcksor. tffl, Pellissarl ED. 1979. Analysis of municipal sludge samples by Ci'.,MS/couputsr for polychlorinated biphenyls and other chlorinated organics. 3ull. Environ. Concern. Toxicol. 22:688. Huczlnger 0, Safa 8, Zitko T. 1974. The ehemlsery of PCBa. Cleveland, Ohio: CtC Press, 269 pp. Kimbrough ID. 1980, laloganated biphenyls, terphenyls, naphthalenes, dlbensodioxlns and related produces. Mow Tork: Elsevler/Horth Bo11and Biomedical Press. Krull IS. 1977. Recant advances in PCB analysis. Residua Revlavs, Gunther PA, Gunther JD, ads. Mew Tork: John Wiley 4 Sons, Inc., Pol. 66. Krupcik J, Isclercq PA, Cara] J, Slmova A. 1980. Analysis of alkylated mixtures of polychlorinated biphenyls by eaplllary gas chromatography-mass spectromecry. J. Chromatogr. 191:207. 37 MONS 026098 Krupclk J, Leclercq PA, Siaova A, Suchanak P, Collik M, Hrivnek J. 1976, Poeeibiliciaa and limleatlona of capillary gaa chromatography and maaa epeetromecry In tha analyaia of polychlorinated blphanyla. J. Chromatogr 119:217. Euehl OW, Whitaker MJ, Dougherty PC. I960. Micromathoda for toxic reeldue acraanlng by negative chemical ionlxaclon maaa apectrowatry. Anal. Cham. 52:935. Lao RC, Thomaa RS, Monkaan JL, 1976. Application of computerized gaa chromatography-maaa apactroaiatry to tha analyals of polychlorlnatad blphanyla. Dynamic Maaa Spactrom. 4:107. Magga RJ, Joynaa PL. Davlaa AJ, lovelock JK. 1971. Anal-. Cham-- Martalli CP. Caatalli MC, Panalli R. 1981. Analytical raaponaa of alo|la polychlorlnatad blphanyla analyzed by aalactad ion monitoring. Biomad. Maaa Spactrom. 8:347. Macthava I, Priaa C, Oardnar A, Carthoff L, Goldeteln J, Ku T, Moora J. 1978, Maeabollam and biochemical toxicity of PCBa and PBBa. Environ. Health Parapact. 24:147. McRlnnay JD, Singh P. 1981. Structure-activity ralatlonahipa in haloganatod blphanyla: unifying hypothaala for atructural apacificity. Cham.-Biol. Intaractlona 33:271. Moaalay MA, Pallltaari ED. 1982a. Development and evaluation of mall coatad opan-tubular column* for CC analyaia of individual polychlorlnatad blphanyl laomara. J. Ugh Baa. Chromatogr. I Chromatogr. Coaaran. 3:404. Moaalay MA. Palllaiarl ED. 1982b. Polyailoxana daactlvatad fuaad alllca caplllarlaa containing immobilized atatlonary phaaaa. J. High Raa. Chromatogr. t Chromatogr. Coamun. 3:472. Mullin MD. 1981. Prlvato Communication. Mullln MD, Pllfclna JC. 1981. Analyaia of polychlorlnatad blphanyla by glaaa capillary and packed-column chromatography. Advancaa in tha ldantlflcatlon and analyaia of organic pollutaaca in vatar. Raich LH, ad. Ann Arbor. MI: Ann Arbor Scl. Pub.. lac.* Chapter 11, p. 187. Pattaraon PL, Palton J, Fraltaa E.Bova R. 1976. 1976 Plttaburg Conf. Appl. Spactroacopy. Clavaland, 01. Palllaiarl ED. 1974. Electron capture detection in gaa chromatography. J. Chromatogr. 98:323. Pallltaari ED. 1974. Ilgh raaolutlon alactron capture gae-liquid chroma tography. J. Chromatogr. 92:299. 58 hqns 0^6099 Palliezari ED. 1981. State-of-che-re lnacruotattl organic analyals In environmental chamlatry. Environmental Haalth Chamlatry, McKinney JD, ad. Ann Arbor. MI: Ann Arbor Scl. Pub.. Inc., p. 19S. Palllzzarl ED, Tomar U, Moaaley MA. 1981. Blgh-raaolutlon |aa chroaatography and ntgatlva Ion chamlcal Ion 1: a don maaa aptctromatry of poly-' chlorinated blphanyla. Advancaa In the ldandfleadoo and analyala of organic pollucauca In waear. Kalth LH, ad. Ann Arbor, MI: Ann Arbor Scl. Pub., Inc., Chap. 12, p. 197. Pletavay Aft, O'Connor R5, Haldan EJ. 1981. A raport of a autvey ... cha lncldantal aanufaccura, procaaalng, dlatrlbutlon, and uaa of polychlorlnatad blphanyl at concantradona balov SO ppm. Cham. Kan. Aaaoc. Raport by Rag. Raa. Sap:. Poland A, Glovar E. 1977. Chlorinated blphanyl Induction of aryl hydrocarbon hydroxylata activityi a itudy of tha atructura-activity ralatlonshlp. Kolacular Pharmacol. 13:924. Posarantz I, Burka J, Plraatona 0, McKinney J, Roach J, Trottar W. 1978. Chaalaery of FCBa and PBBa. Environ. Haalth Parapact. 24:133. Safa S, Platonov H, Huttlngar 0, Jamiaaon HD. 197$, Analyala of organochlorlna natabolltaa In eruda aitraeta by high raaolutlon photoplata aaa apactroaatry. Blonad. Kaaa Spectrom. 2:201. .Sherma J. 1975. Caa-ehromatography analyala of polychlorlnatad blphanyla and othar nonpaatleldo organic pollutanta. Advancaa In Chromatography, Glddlnga JC, Gruahka 1, Collar IA, Cagoa J, ada. Nav York: Mareol Dakkar, Inc., Chap. 3. Sharma J. 1981. Manual of analytical quality control for pesticides and ralatad compounda In human and environmental ramplaa. EPA600/2-81-059. a) Salt ALC, Plaid PH. 1977, Gaaaoua anion chemistry. Pormatlon and raactlona of 0H*j raactlona of anlona vith H.Oj OH' nogatlvo chamlcal Ionization. J, Am. Cham. Soc,, 99:6471, h) i.loyd JR, Agoata WC, Plaid 79. 1980. Gasaous anion chamlatry. Nagatlva chamlcal Ionization of acyl ehlorldaa, diacyl ehlorldaa, and parfluorodlacyl ehlorldaa. J. Org. Cham. 43:1614. Stalling Dt, Huckloe JH, Patty JD. 1978. Preaeuca and potantlal algnlflcanca of o-,o'-unsubstituted PCI laomara and tract AroclorO 1248 and 1234 impurities. Hydrocarbona and haloganatad hydrocarbona in cha agnatic environment. Afghan IK, MacRay D, ada. Nav York: Planum Praaa, p. 131. Stalling DL, Hucklna JM, Patty JD, Johnaon JJ, Bandera BO. 1979a. An expanded approach to tha atudy and meaaureaent of PCIa and selected planar haloganatad aromatic environmental pollutants. Ann. N.T. Acad. Scl. 320:48. Stalling Dl, Schwartz TR, Karlov BK, Patty JD, Hogan JD, Johansson E. 1982. Applications of SIMCA to tha characterization of PCI residua profiles In tedlaent, vatar, and aquatic biota. ACS Mtg., Kanaas City, HO. 59 026100 HONS Stalling Dl, Smith Ui, Patty JD. 1979. Approaches to comprehensive analyst* of persistant haloganatad environmental contaminant*. Maaauraoant of organic pollutants in vatar and vaitavatar, ASTM STT (86. VanHall CE, ad. Am. Soc. for Tasting and Matarisls, p. 302. Tindall GW, Winlngar PE. I960. Gaa chromatography-mass sptctromttry mathod for idantifying and datarmining polychlorinatad biphenyl*. J. Chromatogr. 196:109. Truuatra L, Traag HA, VanMunstaren AJ. 1981. Datarvination of individual chlorinated biphanyla In agricultural product* by automatad capillary gas chromatography. Determination In eattla faad and Its ralatlon to milk residua*. J. Chromatogr. 204:413. Wabb *0, McCall AC. 1973. Quantitative PCS standards for alactron eaptur* gas chromatography. J. Chromatogr. Set. 11:366. Wanevorth WE, Chan ECM. 1979. Analyst* of kinetics and concantratlon dependence of electron-capture detector. Advances in chromatography. Zlatkls A, ad. Amsterdam: Elsevier Scl. Pub. Co. p. 99. Wolff NS, Thom too J, Piaehbaln A, Lilia X, Salikoff 1J. 1982. Disposition of polychlorinated biphenyl congeners in occupationally exposed persone. Toxicol, and Appl. Pharmacol. 62:294. 60 MOAjs 026101 DISCUSSION SUMMARY Mery S. Wolff Environmental Science Laboratory House Sisal School of Modlelso Saw York, NY 10029 Tha paper by Dr. Pslllzsarl described tha recent advance* In Instrumental -aaaiurananc cechnlquta for PCBe, which allow Identification of Individual FOB lsomere at low lava la of detection. The dlacuaalon period cantarad on tha faaalblllty and daalrablllty of nolnf auch technique* In routine laboratory juialyalA-of environmental sample*. Tha relevance-of Isomer specificity to current concapca of biological activity of PCBa vaa thought to bo of potential significance. Tho utlllcatlon of high resolution geo chromatography (SRCC) vaa stressed as a powerful tool In elucidation of structure activity relation ships. In addition, the natter of aethod validation was discussed, with reference to the importance of comparing data from various laboratories for PCS* In environmental aamplas. The currently developing area of chenonatrlce, and lta application to pattern recognition of PCBa In environmental eamplat vie Identified as an Important new contribution eo the field. Tha advantage! of UGC In PCI analyele include the apeclfleity ana eeneltivlty of the method. Specificity le achieved by resolution, which allows differentiation of FOBa from lneerferents such as pssticlde residues. Tbsst snelyses may be conducted routinely uelng electron capture detection (EC), but additional specificity may be Introduced by uelng various maae spectrometrle detection techniques. In particular, BGC vaa parcalvad to ba a significant lmprovemaat ovar packad column CC (UGC) for determination of lover chlorinated blphenyla (LPCB). It wee emphasised that the electron eeptnre response for LPCla, depending boeh on degree of substitution and on orientation of substitution, eould very several orders of magnitude. In general, the LPCBe have a much lower response than the higher chlorinated PCI congener* (HPCB), eo that lack of sensitivity may cause them eo be overlooked. One Investigator waa skeptical of LPCB analysis using BGC - 1C. Purther, the pattern of LPCB in environ mental samples may let be recognisable in comparison with tha uaual Aroclor standards. Whereas BPCB, using either UGC or BGC, can generally be matched approximately to an Aroclor standard, the LPCBe are often eo extenelvely degraded that visual pattern recognition is Impossible. The utility ef BGC in routine analysis was suggests! to be feasible In the lnmadiaea future by uea of ahort columns and autoaampllng davleas. Ths usa of BGC as an ancillary tool for qualitative identificetlon ef PCB con gener*, with eubeequent extrapolation to larger numbers of ssmplts analysed by IIOC, was urged by several participants. Although the reliability of this approach was objected to by some ef tha discussants, its practicability wee thought to have soma heals in the pest literature. In several paet etudlee. Identification of individual KB congener* in human or animal tissue* wee repotted. Some investlgatocs seed BGC ee an adjunct to UGC analyses, while in other cases use of afficlMT~packed columns, often in combination with 61 mgns 026102 spedtie sample preparation techniques, allowed identification of Individual coaponente. For example, non-ortho, mono-, and di- orthoehloro substituted PCBa bavn baan raporead to ba aaparabla by column chromatography on charcoal. Tfca eanaltlvity of HXCC was raportad to ba In tha range of 0.1-100 pg dapandlng on tha particular FOB isomer and on tha method of dacactlon. Using east spectrometry aa tha dacactlon aathod (MS), graacar sensitivity can ba acblavad ualng alngla Ion aonltorlng chan ualng full acan. Tha laporcaoca of coordinating preparative methodology with lnatruaantal aathodology vaa straaaad. It vaa pointed out that the achlavaaanc of low limit! of dacactlooa requires validation of claan-up aathodology aa wall aa lnatruaantal datactlon. That la, lnatruaantal datactlon of 1 pg auac ba coablnad with a aaapla claan-up aathod that paralta recovery of that amount of aaterlal In order to pacify a Halt of datactlon of 1 pg for analyala of actual biological samples. Tha application of HXCC to Identification and quantitation of FCB derived degradation producta, especially in tha presence of FCBs, was discussed briefly. In particular, the use of LXGC to Identify PCDPs la PCB alxtnras vas aantloned. Several such reports exist In the recent literature. The validity of electron capture with either LXGC or HXCC to Identify PCDFs or PCDDs in the presence of FCBs vas questioned. Dr. Stalling pointed out that In analysis of soot froa a transformer fire, tha presence of large asiounts of chlorinated blphenyleites had prevented meaningful determination of PCDFs by EC detection. Recent papers have reported the synthesis of e number of PCDP isomers as standards, as well as Identification and quantitation of these compounds in environmental aamplea ualng HXGC In combination with MS detection, following a highly specific clean-up procedure. Sensitivities ware in the same range as for PCB isomers. Dr. Stalling alao mentioned on-going research into a new detection system for FCBs, which employ* a halogen specific helium plasma dateetor, vhicb may permit differentiation of bronluatad and chlorinated biphenyls, for example. The major problems and disadvantages for widespread application of HXCC to PCB analysis vers discussed. The moat common drawbacks aaotloned were the coat end time of analysis, especially la the use of MS as the aathod of detection, for routine analysis of PCBa. Investigators involved In epidemio logic investigations were doubtful of the practicality cf HXGC analysis of large numbers of aamplea. The lack cf standards, available to all investlgatera, la a currant problem. Although several laboratories have eyutheeltad a large number of Individual PCB congeners, the availability through commercial sources la limited. Tha uniformity of purity of standards la also important. Evan with Aroelora or othor commercial PCB mixtures, tha PCB content of individual eenganars may vary from batch to batch. That several major discrepancies axlatad in recant raporta of Individual PCB conganars in Aroelora or In biological aamplea vas pointed out. PCB conganors raportad from at least four dlffaront laboratories were different even for conmercltl PCB mixtures (e.g. Atoclors) virMo.the east fav years, end it yea strongly suggested that this ,olnt be clarified, the possibility of quantitation using proportions of Individual PCB conganars la Aroelora (along the lines of the aathod of Webb 62 MOWS 026103 and McCall used lor LRGC) ha* bacon* a possibility with tht publication of such data la tha past fav yaars. Tha aacasslty of conducting lnterlsboratory studlas vis strssssd, as part of tha quality control and nathod validation of saapla analysis, vtiatever tha athod of PCI Identification and quantitation. In particular) as previously nantlonad. tha Importance of introducing such prograas into analytic nathodologlaa nov developing for BRGC analysis, was strongly aaphaslxad. Quality control should apply to clean-up proesduraa, to quantitation procedures, and to nathod validation. Tha lack of honoganalty of TCI analytic procedures reported in the literature was suggested by results of lntarlsboratory studies at CDC, vhar* quantitation was achieved using LRGC. Or. Larry Needham raportad that one lntarlaboratory study involvad analysis by various laboratories, using any nathod of their choice, of blood aarun fro* cattle fad PCBe. Tha results, as described by Dr. Maadhan, rssenblad a randon assortment of numbers. Whan tha laboratorlas vara latar asked to us* a salactad nathod, tha results vara nueh Improved, but still bad considerable variance. Dr. Susan Soncblck (Versar, Inc.) reported that tventy laboratories had recently participated in analysis of PCB* In oil matrices using LRGC. Aeeordlng to Dr. Bonchlek, several standardised clean-up procedures vare specified for different oils. Tha results should be available for evaluation vlthln a fsv nooths. Dr. Carl Manger (Balelnoro Css and Ilaetrlc Co.) reported that the ASTM vould soon publish a method for analysis of Aroclor mixtures In oil (1983 ASTM part AD), vblch vas also evaluated In an lntarlsboratory study. In terms of recent advances, little progress vas thought to hava been made In validation and standardisation of clean-up methods for biological materials. Two papers by Albro in the past three years hava emphasised tha. vlda variability Incorporated Into sample preparation procedure from methods of extraction and clean-up (add wash, column chromatography, HPtC). Implicit In such variability Is the Important factor of differential extraction and recovery of different PCB Isomers. The critical Importance of extraction efficiency of different techniques vas also pointed out In a recent publica tion, vhar* levels of 20-3000 ppb of FCDDs and PCDFs war* observed In identical fly-ash samples using seven different extraction methods. The validity of quality control methods which employ fortified tissue samples for recovery determination* ha* also been questioned by Albro, who has compared samples fortified with chemical In different solvent* to samples heving endogenous CM - labelled chemical. The importance of reliability versus recovery efficiency has also been mentioned by Albro. The use of automated gel permeation chroma tography for removal of lipid has bean used by several laboratorlas, and Is advantageous because It ean be automated, recoveries are quantitative, and lipid ts well separated from PCB residues. However. It vas pointed out that tha reliability of this method, in one report, vas poorer than another widely used technique (seetonltrlle-hsxan* partition), although the latter had slightly lover average recoveries. 63 HONS 02610* It vat aaphaslzad that In addition to tba varlanca introduced by sample preparation techniques, tba reliability of PCB quantitation baa baan reported to be no better than a factor of three to tan. ao that the combination of auch varlancai autt Inevitably lead to vide Interlaboratory variance, aa underacorad by Dr. Needham's remark*. The queatlon of reliability vee rained vlth reference to enforcement eechnoloty. vhere level* of 30 ppm in eartaln oil* are aubject to control. The analytic variance, a* veil ea the aenaltlvlty. of e given preparative and Inacmaental technique are factor* to be coneldered la netting realletlc control guideline*. The toxicological algnlflcance of Identifying Individual PCB congener* In environmental aampla* vaa discussed. In viev of the finding* preaented In other **elone of the ayapoelua, many of vhieh have also been reported In the literature, the consensus recognised the Importance of PCB Isomer types. The measurement of PCB* in human blood for example, has been performed on the premia* that incanalty of exposure, ae reflected In concentration of total PCBs. would be related to toxic effects. Thus, vlth vary few exceptions, PCB residues In human tissues have been reported as a total concentration, or the sum, of several LBCC peaks. Although In a fsv studies, Individual PCB congeners have bean identified, population base studies Invariably have reported a total number, on vhieh biological correlates have been based. In several cases, not even the type of Arnclor used to estimate PCBs has bean specified in such studies. It vaa pointed out that such discrepancies, along vlth tba methodological insufficiencies also mentioned, east doubt on eh* comparability of such studies, in tens of PCB exposure or la terms of bio* logical correlates. Clearly, even the ttCC techniques currently in us* require standardisation to provide some consistency among different studies, both for quantitation and for designation of the kind of PCB exposure. Although the state-of-the-art of PCB analysis vould dictate use of BBGC, vlth subsequent biological correlation vlth individual isomers, the advocates of this methodology veto consistently reminded of the rapidity and large volume of samples vhieh must be dealt with in many studies. Tha application of pattern recognition to SIGC analysis vaa recommended,.and It vas further discussed vhethar this computer maehod vould be applicable eo LACC, The use of H*CC eo complement large numbers of LXGC vaa also recommended. Although presently limited for the quantitative evaluation of large number* of samples, BRCC vas urged as an area of further development for routine usage in combina tion vlth patttrn raeogaltlon technique*. It vas auggastad to bo a potandal aeaat in identifying PCB congener subsets for usn in predicting rolstlve toxicity. One major hortconlng of L1GC, In estimation of PCBs, vas said to ha its failure te detect or to recognise lever chlorinated congeners. Usually, an analyst attempts to match o sample vlth an Aroclor mlxeure, vhieh la than used for quantitation. Both the lack of detection (lev detector response) and tha purported underestimate of concentration (use of standards unsulted to less chlorinated lsomets) vert cited as contributory factors. Bovevsr, tvo rsesnt papers, ons vlth humans and one vlth animals, rsportad comparable LBCC sod -race vaiuee. The uas of a method auch ae the commonly cited Webb end McCall autonetlcally Incorporate* a compansatory faetor for lower responsa of LPCBs. 64 HONS 026105 Mevarthal***, tha advantages of HBOC analysis for L?CB war* lapr*iv, ihovlng greater sensitivity and discrimination of litearfarant*. The developing ehaaoaatrle theory and It* application to PCS pattern recognition wa* luoaurlied briefly by Dr. Stalling. In collaboration with ipaclallat* In thl* field, hi* group ha* applied a tatlatlcal nathod (prin cipal component analyala) to 69 PCS congener* In HRGC analyaaa of approxiaacaly 1,000 environmental samples. Tha raault* era pra**nt*d in an eigen vector format, or principal component projection, which allow* tha analyac to determine the behavior of individual PCB component* within a eimilar type or claas of aaaplea. Some of hla finding* will be preaented in a aerie* of paper* at the forthcoming American Chemical Society meeting in Kansas City. The applicability of varlou* methoda of analyala to the industrial sector was dlscuasad, with special reference to the necesaicy of performing BRGC. Zt wee conceded that LAGC we* aufflclant to analyta undiffarantiatad Aroclors, or othar comm*rc1*1 PCB products, in varioua oil aampla*. The application of simpla methods to PCB dagradatlon products, or contaminants such as FCDFs. uaa said to ba-nonspecific. In tares of. tha discussion which suggested that MGC was rtlatlvaly lnaceurata and lnaanaltlva aa a maaua of maaaurlng LPCB, it could ba concluded that tha uaa of HBCC might be etuentlal in Identification of PCB* incidentally produced during induatrlal proeaasaa, such as l,r*s*ne*d in tha latar aasalon on incldantally producad PCB* in Induatrlal proemaaaa. 63 MQNS 026106 CHAPTER 3 EXPOSURE STUDIES - ENVIRONMENTAL RESIDUES AND BIOACCUMULATION The flr*c papar in thla chapter, ancoupaeslng production, distribution, and environmental fata of PCBa, offara data which aaclaataa the total PCBa contamination of Cba aobll* environment and demonstrate* wfaara tha hlghaat coocantratlooa and largaat quantities of tha conpound presently axlat. Tlaa-coura* studies ara prasaotad to determine tha ralaclve PCBa coneaalnetlona over a number of years. Qualitative analyse* of Individual laonara or homolog* ara parformad to evaluate the compounds which ara note likely to be environmentally or biologically degradable. Finally, environmental peopling data la praaentad which confirm tha build-up of aelactad homolog*. Tha aaeond paper artenda tbaae results to living organlana. Studies of fieh and human populaclona reveal aalactlva bloaeeuvulatlon of PCBa leoaera and homolog* which relate- to their decree of chlorination, substitution In tha para poaltlon, and.occurrence nf adjacent unoubacltutad earbona. Problem* aaaoelaead with achieving. PCBa atnady-atat# lavala la animals, followed by PCB elimination ara artentirely covered. Further dlacuaaed in detail la the affect of tha reproductive cycle and lactation in flah and humane, respec tively, on tha elimination of PCBa body-burden. we loanee of tha aoclo-madleal aapecta of lactation, thn aeructurnactlvlty relationships to motabollam, bloaeeumulatloo, and toxicity, and thn prohlama of aaaeurlng PCBa contamination in tha Great Lakaa are elaborated upon In thn dlaeuaalon aumary. 6 HONS 026107 DISTRIBUTION AND FATE OF CHLOROBIFBENYLS IN THE ENVIRONMENT R. J. Mooleuaar Dow Chemical Company Midland, MI 48640 INTRODOCTION Thla paper ravleve racent Information on tha Input of FOB producta and ehloroblphanyla to tha environment, thalr dlatrlbutlon and fata, and on raalduaa of that# aaterlala In environmental aaaplaa. Tha Information la darlvad from aatlmataa of ehloroblphenyl ralaaaa, laboratory studlaa, and raaldua analyaia. An undaratandlng of tha lnearralatlon of thaaa varlad atudlpa, and a projection to actual environmental bahavlor la extremely Important to aerve aa a baala for control action for thaaa eompounda. Kora broadly, tha ehloroblphanyla, along with aavaral othar natarlala for vhlch thara ta an axtanalva databaaa, aarva aa nodal eoarpounda for predicting tha bahavlor of tha who la apactrun of aynthatlc and natural ehamleala. It la, tharafora. Inportant that parlodle updataa ba nada of tha vhota picture regarding tha environmental bahavlor of the ehloroblphanyla. ENVIRONMENTAL EKTRT OT CHLOROBIFBEHUS Polychlorinated blphanyla, producta produced by tha chlorination of blphanyl, vara uead estanaivaly during SOa, 60a and aarly 70a. Thalr production and uaa vara aharply curtailed, flret voluntarily, and than by regulation during tha 1970a. Aa eotlmated 610 x 10 kg PCBa.vera produced and need In tha O.S. through 1975, of which an aatlaatad 82 x 10 kg haa baen ralaaaad to tha aoblla environment (HAS 1979). Tha raaalndor haa althar baen daatroyad, placed In landfllla, or la atlll In naa. PCB producta havo not boon manufactured In tha O.S. alnea 1977. Chlorobiphenyle hava bean produced in avail quantifier aa by-produeta of othar manufacturing proceeaee. Manufacturing examptlona hava bean filed with EPA for the production of prodneta containing ehloroblphanyla In concantratlona In axceaa of 50 ppm. Praaumably, total quaneltlaa produced and tha amount antarlng tha bloephere could be aatlmatad from thla Information. A recant aurvey of tha chemical lnduatry revealed 0.0062 x 10 kg of ehloroblphanyla ara generated annually aa Incidental eontamlnanta ae concantratlona bolov 50 ppm (Table 1). leae than 0.0006 x 10 kg of thla amount annually antara tha mobile environment. The major currant loan of PCBa la believed to ba leakage from capacleora, tranaformara, and othar electrical equipment in uaa. An uppar bound getlmata of euch entry Into tha environment provided by 0A la about 0.18 x 10 kg annually (47 EX 17425, April 22, 1982). ehloroblphanyla ara alao produced aa by-produeta of high temperature conbuatlon proceaaaa (Richard and Junk 1981, Junk and ford 1980, Vick at al. 1978). Thaaa Include municipal and lnduatrlal lnclneratora, coal-fired electric power planta, and tha burning of wood. Tha origin of tha ehloro- 67 HONS 026108 Table 1. Incidental Cblorobiptaenyl Production and Dlipoaal Total Annual Production Oca) Dlapoaal Inclnaratloo lalt Water Dlaehar|e landfill Ondarfround Injection Incloaad Solid Producta Other Producta TOTAL 6,200 4,100 800 700 300 200 100 6,200 Chealcal Manufacturer Aaaoclatlon, Horanbar 15, 1981. 68 MONS 026109 biphenyl* la not baliavad to bo tho coabuacioo of notarial* containing PCB, alcca addition of tofuao containing PCS to a coal flrod burner did ooe Incraaee tho aaount of chloroblphaayla la tho affluent*. Tho magnitude of chlorobiphanyl foraatloo from thla aourco la unknown. Tho saounta reported la offluonc gate* range fro* 0.0007 utin1 to 200 ug/a3 eoatparad to aablaat air lavala of 0.0001 to 0.01 ug/a . A review of ot|aole oalaalooa froa covbuttlon procaaaaa haa baaa providad by Junk and ford (1980). Still aoothar aourco of PCB to tho aoblla anvlronaont la leakage froa landfill* and equipment dump*. An aatlmatad UO x 10kg PCB bad boon dlapoaod of la landfllla and equipment duapa through 1978 (HAS 1979). Vary llttla data la available for eetltMtlng tho quantity of PCB antaring tho aoblla onvlroaadnt froa thla aourco. Tho aaount leaking into groundvatar lo ballavad to ba negligible, but avail aaounca ara hollared to rolatllise froa relatively expoeed altaa (HAS 1979). PISTB1BOTIOH- AHP PATE OP CHLOBQBrPHEHTtS Tho dlatrlbutloa and fato of chlorobiphoayla dependa on their ralaaaa patterne, partitioning aad.dlaparalon, and choir euscoptiblllty to degradation by varlona routna. Tho ehlorob1-phonyla that eonotlmtod PCB products vora raloaeod over a period of yoora aa a result of o wide variety of atea. Load aatlaMtes In tha ctvoaphara, hydroaphara and llthoiphorn for 0.8. asaufacturod PCBa that have entered tho aoblle onvironvont are shown la Table i (HAS 1979). Tho porcoatatoe are baaed on on input of 82 x 10 kg froa 0.8. production and ralaaaa. Tha aajor raaarrolr for PCB la raportad to ha tha Atlantic Oeaan. Tha ataoaphara la a vary alnor raaarrolr, but la baliorad to bare aarrad aa a aajor tranapoxt vehicle for PCB. Praahwatar rtaervolra do not contain a large fraction of tho total load, but tho hlghoat concentration* ara uaually found In froohvator acoayataaa. Much attention haa boon focuead on tha dlatrihutloo of PCBa In tho Groat Lakao. A recant auvvary haa haan provided by Elaonrolch (1982). He concludes tha praaant PCB burdoo for tho entlro Groat Lake* la 0.44 to O.S x 10 kg of which 882 t* In tha sadlaaoes. Katinatea for each lake aro ahovn la Table 3. Elaanraleh concluded that the aajor currant input to tha Upper Lokae la via tha ataoaphara. and that Input la nearly balanced by loaaaa do* to burial in the aadlaent. One* rolesaod, tho PCB product* oxpoaod to tho atvoapborlc or aquatic onvlrooaanto vor* alovly vaporised or dissolved. Farther distribution and/or dagradatlon took place depending on tha eboaleal, phyalcal, and hiologlenl proportion of tho individual eonpounda. Thaa* propartlaa vary cousldorably vithia eh* faally of chloroblphonyle, and thoy greatly influence tha ulelaaea fata of oach coapound. Thus, eho propartlaa of tho Individual ehovleals vlll ha discustod la order to oodoraeaad thair fata and tha observation* raportad on residua* of those notarial*. PHTSICAI HOPKBTIES Tha physical propartlaa of the chloroblphaayla her* recently bean raportad or auvaarltad (HAS 1979, Mackay at si. 1980, Banarjaa at al. 1980, Vaitcott at al. 1981, Italy 1983, McDonald, unpublished, Salth at al. 1984), 69 MQNS 026110 Tabla 2. Envlronaantal Load Eatloataa of FOB for tha J l, and tha Atlantic Ocaan Aaount * 109 k| Parcantaga Baaad on Total Ralaata Atooaphara 0.018 0.02 Bpdroaptiara Fraatnratar Fraatnratar aadtaant Fraatnratar biota Marina wntar Marina tadlaant Marina biota 0.012 - 0.039 1.4 - 7.1 0.03 6 - 66 0.68 - 2.7 0.03 0.01 - 0.04 1.7 - 8.6 0.04 7.3 - 73 0.7 - 3.3 0.04 Llthoaphora 0.14 - 2.8 0.2 - 3.4 has 1979. 70 HONS 026111 Table 3. PCS Burdana In tba Crate Lake** Lake PCS Aaoune X lir k| Superior Mlchlgao Huron Ontario Erla 38 57 89 122 187 *llaanrtlch 1982. 71 HONS 0Z611-2 Important propartiat ara shown in Table 4. Solubility and volatility lanarally dacraaaa vlth increased degree of chlorination, although individual isomers nay show some variation. From thaaa propartlas it la poiaibla to calculata dlatributlon eoefficlanta (Dllllng 1979, Karlckhoff at al. 1979, Chiou at al. 1977, Faith at al. 1979) for individual conganara bacvaan various media at shown in Tabla 5. Such. Information ia aaaaotlal for pradletlng vtiat part of tha anvlronaant will contain raalduaa of tha compounds, or whathar tha compounds will gat to raglona vhara degradation will occur. Aa inspection of tha propartlas and distribution eoafflclants claarly shows increasing levels of chlorination lncraaaas tha tendency for adaorptlon to soils and for bioconcantratlon. Low degrees of chlorination increase tha nobility of tha compound, both into tha atnosphara and into water, and lncraaaa tha probability for undergoing degradation. omnOHMEHTAL DECKADABILITT Several studies have bean reported describing tha blodegradablllty of ehloroblphenyl congeners (Salley ae al, 19SI, Abated and Fochc 1973, Anderson 1980, Furukava at al. 1978, Wong and Kaiser 1975, keichardt at al. 1981). Ail present a eonslstant picture. Chloroblphenyle with a low degree of chlorina tion ara readily biodegradable, while the highly chlorinated coogeners are increasingly resistant to biodegradation. .. The work of alley at al, (1981) is particularly noteworthy since *C-labeled compounds were used and good material balances were obtained in his studies. Degradation studies ware csrriad one in river water. The pattern of degradation observed for 2-chloroblpbenyl is shown in Figure 1. The parent ehloroblphenyl degraded rapidly (t .,,, approximately 2 days), one loolatable intermediate degradation product wal'obeerved (chlorobenroic acid)', which rapidly degraded to 1 cO, and presumably water and inorganic ehlorlde. These observations are conalstant with the early work of Ahmed and Focht (1973) who proposed the degradation pathway shown in Figure 2, and also the later work of Anderson (1960), who observed chlorobenroic acids as degradation. Intermediates. The half-life for degradation was similar for concentrations ranging from 1 to 100 vg/l for all three isomers of monoehloroblphenyl. By way of contrast, the tetra-subatltuted compound, 2,2',4,4*-ehloroblphenyl, studied under similar eouditloua shoved no measurable degradation during a period of SO days. Anderson (1980) found the degradability of chlorobiphenyls in lake sediments decreased with increasing degree of chlorination, with half-lives ranging from a few days for dleblorobipbenyl to 200 days for pentaehloco laomars. Biodegradability is also affected by the position of chlorine substitution on the biphenyl ring, particularly for poly substituted deriva tives. Bo degradation vat observed under anaerobic conditions for any of the chlorobiphenyls. Studies of the atmospheric degradation of chlorobiphenyls are sparse. Appleby (1976) reported that 2-chloroblphenyl degraded under simulated atmos pheric conditions vlth a half-life of a few days. Theoretical work by Hendry (1979) la eonelatent with the work of Appleby, and suggests the more highly chlorinated congenefs are more resistant to atmospheric photodegradation, kecant more definitive work by hilling (1983) confirmed the work of Appleby. 72 HONS 026113 Tabla 4. Chloroblphenyl Physical Proparclas* Chaalcal 2-CBP 4-CBP 2,2*-CBP 4,4'-CBP 2,5,2'-C8P 2,3,8'-C8P 2,5,2',3'-CBP 3,4,3,4'-CBP 2,4,2*,4'-CBP Uaear Solubility uaola/l, 25* C Vapor Praaaura an Bg, 25* C 26 8.4 > 10'3 12 4.6 * 10"3 4.9 2.1 s 10"4 0.3 1.9 X 10"5 3.9 9 a 10"5 0.3 0.09 0.3 0.23 3.7 x 10"5 8.6 x 10"5 2,3,4,J,2*-CBP 0.028 `itAS 1979, Kackay at al. 1980, Banarjaa at 1. 1980, Waatcott at al, 1981, Manly (1983), McDonald (unpubllabad data), Salth at al. 1984. 73 HONS 026114 Table 5. Environmental Oietribution Coefficient* Compound 2-CBB 4-CBP Dlatrlbution Coefficient* Alr/Watar B* Solld/jfater Kav 0.017 0.020 660 270 Flah/Watar SCFc 2,000 960 2.2'-CBP l.l'-CW 2,3,2'-CBP J.S.l'.S'-CW a.i.aM'-cBP 2,3,4,3,2'~CBF 0.002 0.004 0.003 0.023 0.020 1,100 700 3,300 16,000 33,000 29,000 3,300 2,100 12,000 31,000 60,000 32,000 *In unit* of f&(V *n *tr--, calculated froa Billing 1979. ag/l 1" water ^Aaeuaea 22 organic carbon, calculated froa Banerjea at al. 1980 and Karlckhoff at al. 1979. cCaleulated froa lanarjaa at al. 1980 and Velth at al. 1979. 74 MONS 026115 Initial eooeantratlon 10 ug/L 1 Initial l*C activity Flgura 1. llodagradatlon of 2-nonochlorobiphanyl, 75 HONS 0Z6U6 COj Cl" + ijO Flyura 2. lldi|ridtlM pathway of aonochloroblphanyla. 76 HONS 026117 He iend ltd Ch# photodegridation of #11 three aonochlorobtphenyl l#oa*r# In glee# eh#nb#r using benzene and a acabla chlorofluorocarbon aa inctmal ftandarda. Tha itudy via carried out with tvo different surface to volume ratio#, ahovlng that tha reaction vaa In tha ga* pbaa# and not surface catalyzed. Tha raaulta ara llluitrittd In Figure 3 ahovlng tha dlaappcaranca of 2-chloroblphauyl under dark and light condltlona. By conparlng tha rata of dlaappaaranca of each laoaur with that of benzene under light and dark condltlona, and knowing the rat# conatant for OB radleal attack on benzane (Perry and Pitta 1977) from previous publlabed work, Billing vaa able to calculate corresponding rate conatanta and average eropoapherlc halfllves for cb# ehr# nonochloroblphenyl laonera. Theae are ahovn In Table 6. All thr#e aonoehloroblphenyl laonera degrade under simulated atmospheric condltlona with half-Uvee of about one day. PREDICTION OP ENVIRONMENTAt RESIDWS The laboratory data ahova that Increasing lev#la of chlorine fubatltutlon on biphenyl reeult* In a greater tendency for adaorptlon to aolida, an Increase In the bloconcentratlon potential, and a greater degree of peralatence. On a qualitative basis It la clear that monocblorobiphenyl is likely to disperse and degrade rapidly, vhlle the highly chlorinated coapouooa (especially tetra and above) are likely to be peralstent and appear, as environ mental raolduae. Neely (1981) recently Integrated this data Into a nathenatlcal nodal Intended to simulate tha "average" world In order to predict the lifetime of aalected chlorobiphenyl congener* in the environment. II* eoncloelona, abovn in Plgure.A, predict a half-life of about 3 day*, for monochlorobiphenyls, 130 daya for dl, 579 daye for trl, 1044 day* for tttra, and 3443 days for pants. If ataospharlc dagradatlon la included for mouochloroblphenyla, tha balf-llfa la reduced to one day. If condltlona of temperature, microorganism concantratlon, etc. dlffar markedly from thoee aaaumad by Neely,. difference* In peralatence would be expected. However, the relative degradation rate* between different congener* should ranaln nearly constant. Nealy (1983) also modeled tha altuatlon In Lake Michigan with reepect to homolog distribution. Shown in Figure 3 are tha difference! expected between an assumed original "PCS fluid" that may have entered eh* lake, and eh* distribution of chloroblphanyla expected In boeh vacer and fish. Mono and di go from 1 and 16Z of tha chloroblphanyla la tha original fluid to about 0 end 6.41 In waeer end 0 end 2Z In fleh, respectively. On the other hand, penta goes from 10Z of the ehlerohlphenyls In tha fluid to 10Z In tha water and 48Z In the fish. The differences reflect preferential biodegradation of the mono and dl chloro compounds baaed on laboratory aeudlea of their blodagradablllty, end also tha greater tendency for bloconcentratlon for tha penta. EKVIlOWgMTAt 8ESTPW8 Reports describing the homolog distribution of chloroblphanyla found In tha environment ara conalstant with Nealy's predictions. Figure 6 shove tha distribution of homologe sold In tha 0.$. during the period 1930-1978. Trlchlorohlphenyl wee tha most ebundane homolog of the serlae, followed by tetra, pants, and dl. 77 MONS 026118 Figure 3. S Inula ted atnoepherlc degradation of 2-cli!oro|>l|>han)ri HONS 026119 Tabla 6. Ataoapharlc Dagradatlort Paraaatcra for Konochlorobiphanyli k'0H *1/2 (Tropoapharic Avaraga) 2-CB 3-CB 4-CB 5.4 x 10 c* aolacula"* c-1 1.0 x io'n 1.2 x 10'11 l.S day 0.77 0.66 Asauaat C0H)#va 106 e*-5 *DlXlln| at <1. 1963. 79 MONS 026120 Figure Environmental persistence of selected chlorinated biphenyl Isomers MONS 026121 Mono II Hgure 5. Predicted liomulog distribution of chloroblphcnyls in U> Mlcbigsn. MONS 026122 Humber of chlorines Figure 6. Chlorinated biphenyls sold In the U.S. 1930 - 1978. MGNS 026123 Chloroblphenyla found In plasma and adlpoaa tissue of capacitor manu facturer workers (Figure 7), who vert exposed to FCB products with homolog distribution at manufactured, showed a shift In homolog distribution to the more highly chlorinated compounds (Wolff et al. 1961) Terra was tha pre dominant chloroblphenyl found, followad by trl. peota and hexa. Di was reduced to very low levels, while mono was not observed. Tha ahlft In distribu tion can be axplainad by the fact that chloroblphanylt with a low degree of . chlorination, especially mono and dl, are cleared rapidly from nasnala, Figure 8 (Raley et al. 1976, Matthews and Anderson 1975). Thus, accumulation favors the highly chlorinated compounds. The additional effect of the environment on tha homolog distribution lo human tissue can be teen by examining Figures 9 and 10 (Jensen and Sundstrom 1974, Mas et al. 1980). These data were from humans exposed to ehlorobiphtnylt In the general environment rather than in tha workplace. The hcoolog distribution Is shifted to even higher degrees of chlorination, with hexa and hepta levels predominating. The disappearance of dlchlorobiphenyls and the dramatic reduction of trlchloroblphenyls indicate that transport and degradation In the environment has greatly reduced the levels of these com pounds. Thus, both the environment and maanallan organisms tend to praferantlelly degrade or eliminate biphenyls with a lov degree of chlorination coeiparad to those with high degree of chlorination. Tha preferential degradation and excretion of lower chlorinated biphenyls Is probably a key Tesson why they exhibit less potential for chronic toxicity than that shown by PCI produets or highly chlorinated biphenyls. In a 31-day gavage toxicity study la rats, approximately 100 times more monochloroblphenyl was required to produce liver effects equivalent to those noted for Aroclor 1754 (Dleti et al. 1983). Continuous flew studies (32 days) vlth fishes also showed 2-chlorobiphenyl was lass toxic ehan 2,2',4,4'-tetrachloroblphenyl by a factor of aboue 30 (Dill et al. 1962). Tha overall levels of PCS la Lake Michigan are declining (Bartlg 1961). Figure 11 shows levels of PCB in chubs taken from Lake Mlehlgen. The PCB concentration has dropped from 5.5 ppm in 1974 to 2.2 ppm in I960. Figure 12 shows PCB levels In lake trout and coho salmon are also dropping, although there are no data taken since 1976. Figure 13 indicates a dramatic drop in PCB levels in gull eggs taken from islands In Lake Michigan. The drop in PCB levels in gull eggs closely parallels the drop in PCB levels In chubs. Eissnralch (1962) concluded the residence times of PCB in Lake Superior and Lake Michigan waters are 2 to 7 years and l.S to 4 years, respectively. Besldenee times In thssa laka waters depend In part on the rate of permanent burial of theta compounds in sediments. Once burled, the PCB is not acces sible to water organisms, and does net accumulate in the food chain. No data hat beeo reported on the degradation of PCB in tha actual sediments in the lakes. Although, as discussed earlier, sediments taken into tha laboratory have shown degradation of chloroblphenyls (Anderson 1980). Atmospheric levels of PCB over Lake Superior, bated on limited sampling shown lo Table 7, arc alto decreasing (Eisenrelch. Partonal.communlcatlon). Reported levels hive droppco from s rsngeof 0.9 to 3.5 ng/n in 1978 to 0.1 to 0.6 ng/n* In 1961. It is possible to estimate yearly production and 83 026l2^ mqns LJ 50 H A0- St Legend V~A Adipose, ppa 82) Plasaa, ppb 30- 2014.A 10- 0.06 0-7 I 3 *5 67 Number of chlorine* Figure 7. PCB level* In capacitor manufacturer worker*. HONS 026125 El Initiation half-life, day* SO l.egend 40 CZJ Uver CK) Fat 30 20 10 Y///\---------- l^jonga 2S Humber of chlorlnea "Hotel Elimination of hcxachlor from fat m not obaerved. Figure 8. Bioclearance of chlorinated biphenyl# from rata. MQNS 026126 Figure 9. Distribution of chlorinated biphenyls found id selected Jwwn fet eenplee. HONS 026127 Figure 10. Dtetrlbullun of chlorinated biphenyl* found in Imman nitk. MOMS 026128 Figure II. Chlorinated biphenyl levels In Lake Michigan chub. HONS 026129 Figure 12. Chlorinated blplicnyl level* in Lake Michigan apart fl*h. MONS 026130 Figure 13. Chlorinated hi|>lienyl levels In gull egg* on take Michigan Islands. HONS Tibi* 7, PCI Coocntrtloni In Air Ovr Lika Superior* T*ar 1978 1979 1980 1981 Kite 1.9 0.9 1.0 0.3 l 0.9 - 3.3 0.4 - 1.4 0.1 - 2.3 0.1 - 0.8 S.D. n 0.8 13 0.4 8 1.0 8 0.1 9 *Elinrich, ptnul cowualcaeloa. 91 MONS 026132 release of PCB In the O.S. froe the Information rtvlevad In this paper. Tha data art picaantad graphically In figure 14a. Production of PCS droppad drastically during tha aarly 70s and lntantional production of tha alxtures andad in 1978. A email lneldantal production (<0.3 Million lbs/yr) has eontlnuad into tha 1980a. Release of PCi to tha aoblla anvlronnant is stlaatad at 3 to 12 alllion lbe/yr during tha 10 yaar parlod 1962-1972. Slnca 1978 tha ralaaaa baa baan <0.3 million lba par yaar. Data on tha levels of PCS in Lakes Michigan and Suparlor. as eutuarlzed in Flgura 14b. raflact tha raaponaa of thasa lakaa to tha raductlon in ralaasa of tha coapounda to tha aoblla anvlronaant. LavaIs bagan to drop about ona yaar aftar tha ralaaaa rata vaa raducad. and hava ahovu a ataady dacllna sines that tlaa. Slailar trands hava baan obsarvad for DDT and aarcury in tha Graat Lakaa (Hartlg 1981). Furthar raductlon vlth tlaa la expected aa raaoval of thaaa coapounda froa tha aoblla anvlronaant through dagradatlon and raaoval by burial eontlnuaa. Conclualona 1. Slnca production of PCI bagan. an aatlaatad 80 to 90 x 10* kg hava baan ralaaaad to tha aoblla anvlronaant In tha O.S. 2. PCI producta and chloroblphanyla still antar tha aoblla anvlronaant In saall aaounta (approximately 0.2 x 10 kg/yr). 3. Tha major rasarvolr for PCI ralaaaad In eha O.S. la tha Korth Atlantic Ocaan, but freshwater ayataas contain the highest eoncantratlona. 4. PCI levels In tha Graat Lakaa (air. water, and fish) hava droppad dramatic ally slnca releases vara curtailed In the ald-70s. PCI concentrations In Lake Michigan chubs hava droppad froa 5.5 ag/kg In 1974 to 2.2 ag/kg In 1980. 5. Laboratory atudles have shown that tha degradability of chloroblphanyla decreases with Increasing chlorine content. For axaspla, aooechloroblpbanyl degrades readily In lake sediments, natural waters and air with half-llvee of a few days or laaa. However, tha half-lives of tatra and higher horologe i of the order of Booths to years. 6. tealduas aaasurad la human tlseue show no aonochloroblphenyl, vary little dlchloroblphaoyl and raducad trlehloroblphenyl compared to tha hoaolog distribution in PCI products manufactured In tha D.S. This observation tends to confirm the laboratory observations oo tha relative degrad ability of the lower chlorlnatad biphenyls compared to the highly chlori nated coapounda. 92 MONS 026133 Pi|ure 14a. PCIi - Annual U.8. production and aatlaatad roleaae to tha environment. MONS 026134 Il|uti Ilk. rcia In tha Grant takaa anvlronnant. HONS 026135 REFERENCES Ahmed M, Focht DD, 1973. Can. J. Microbiol. 19:17. Andereon ML. 1980. Degradation of polychlorinated biphenyl* In aedlments of the Great Lake*. PhD Thaala. U. of Michigan, Ann Arbor, Ml. Appleby A. 1976. Atmoapharic fraone and haloganatad compound*. EPA 600/3-76-108. Bailey RE. Rhinahart VI, Conalor SJ, Batchalder TL, Mandota CG, Nealy VB. 1981. Praaantad at cht Society of Env. Toxicol, and Cham, seating, November 25, 1981, Arlington, VA. Banarja* S, Yalkovakl SH, Valvanl SC. 1980. Env. Scl. 6 Tachnol. 14:1227, Chlou CT, Freed VB, Schmedlng DV, Rohnert RL. 1977. Env. Scl. 6 Tachnol. 11:475. Dlatt FA, Quaat JF, Braun VB. 1982. Dov Chaalcal Company, unpubllahed data. Dill DC, Mayaa MA, Mandoza CG, Bogga GO, Eamitt# JA. Comparlaou of tha toxlcltlaa of blphanyl, sonochlorobiphanyl, and 2,2',4,4'-t*trachloroblphanyl to flah and daphnlda. Aquatic Toxicology and Hazard Aaaaaraent: Fifth Confaranca, ASTM 5TF 766, J. G. Faaraon, Robert Foater and V. E. Blahop, eda., American Society for Taatlng and Materlale (1982). Cllllng VL, Miracle GE, Bogga CD. 1983. Preprint* of paper* praaantad at the 186th National Meeting, ACS, Dlv. Env. Cham., Vol. 23, No. 2, pp. 343-346. Dllllng VL. 1979. Env. Scl. 6 Tachnol. 11:405. Elaeoreleh SJ. 1982. PCBa In tha Craat Lakaaf aourcaa, fink*, burden*. International Sympoalua on PCBa In the Great Lake*, Michigan Seat* Dnlvaralty. Elaanralch SJ, No date. Dnlveralty of Mlnnaaota, paraenal communication. Furukav* R, Ionomura R. Raslbayaahl A, 1978. .Appl- Environ. Microbiol. 35:773. Bartlg JH. 1981. Highlight* of vatat quality and pollution control In Michi gan. Pub. no. 4833-9804, p. 13. Hendry DG, Ranlay BA, 1979. Eap 560/12-79-001. Jenaeu 8, Suodatroa C. 1974. Aablo 3:70. Junk CA, Ford CS. 1980. Chemoaphere 9:187. Relay BG, Hick* 0, Mae* WM, Tucker ES, Miaura JP, Johannaan FR, levlnakaa CJ. 1976. Bull. Env, Dataa 6 Toxicol. 15:699. lUrlekhoff SV, Brown DS, Scott TA. 1979. water Rea. 13:241. 95 HONS 026136 Mackey D, Maacarenhaa R, Shiv VY, Valvatii SC, Yalkoveky SR. 1980. Cheaoephere 9:257. Matthave RB, Andaraon MW. 1975. Drug Mat. Dlapoe. 3:371. MeDenald RA. Ho data. Tha Dov Chemical Company, Thermal Laboratory, Midland, MI 18640. Unpubllahad data. Mat J, Davlea DJ, Lay FT. 1980. Cheaoephere 9:763. NAS. 1979. Polychlorinated blphanyla. Waehlngton, DC, Naaly VB. 1981. Chentach. 11:219. Naaly U>. 1983. Reactivity and anvlronaantal paralatanca of ?CB laoaara In Phyaleal bahavlor of PCB in tha Croat lakaa. Mackey D, Pataraon S, Elaanralch SJ, Slnona MS, ede. Ann Arbor Sclanca Publiahara, Inc., Ann Arbor, MI. Parry U, Pitta JN. 1977. J. Phya. Cbaa. 81:296. Ralchardt PB, Chadviek BL, Cola MA, Roberteon BR, Button DR. 1981. Env. Scl. 6 Taehnol. 15:7$. Richard JJ, Junk GA. 1981. Env. Sel. 6 Taehnol. 15:1095. Salth NR, Gorin G, Good VD, McCullouh JP. 1964. J. Phya. Chea. 68:940. 0SEPA. 1982, O.S. Envlronatntal Protection Agency. Polychlorinated blphanyla (PCB): uaa In alactrleal equipment. Notice of Propoaad Rule. 47 Fad. Reg. No. 78, 17425, April 22, 1982. Velth CD, DaPoa DL, Bargetedt BV. 1979. J. Fieh. Rea. Board Can. 36:1060. Vick 1, Junk C, Avary M, Svae H. 1978. Chemoaphera 7:893. Weetcott JV, Slaon CG, Bldlaaan TP. 1981. Env. Scl. 6 Taehnol. 15:1375. Wolff MS, Thornton J, Plachbaln A, Lilia R, Sallkoff IJ. 1981. Toxicol. Appl. Pharmacol. 62:294. Wong FTS, Ralaar RLE. 1975. Bull. Env. Cont. Toxicol. 13:249. 96 MQNS 026137 FACTORS AFFECTING THE BIOACCVMUUTION AND PERSISTENCE OP POLTdaORLSATED BIPHENYLS Harp Jo Vodlcnlk Department of Pharmacology and Toxicology Medical Collage of Vlfeooaln Milwaukee, HI 33226 Polychlorinated biphenyls (PCBa) are vldaapread paralatant lnduatrlal contaminant* In cha global ecoayacea. Originally used In paints, Ink*, hydraulic fluid* and electrical components, PCBa hav* been found In tha atmosphere, fresh and salt vaeer, soil and tissue* of nuaeroot animal apaclas. Because of thalr high lipid solubility, PCBa tend to accusmlat* In sdlpoaa tissue and thus bacon* blovagnlflad through the food chain. Although nunarous Investigator* hav* axaalnad th* phanaacokinctlc behavior of PCBa In aaamallan apsclaa, there hav* bean relatively few studies vhlch consider th*** phenomena In the lover vartebrato fora*. This paper attaints to coapara tha accuaulstlon, distribution, persistence and aetabollsa of polychlorinated biphenyl* in isasali and aquatic specie*. IIOACCTMPLAT ION Aa la the ease In the higher anIn*la, fleh are axposed to anvlronnental contaminants prlnsrlly through the diet and Inhalation <vla the gill). Whether a particular chealcal la absorbed and accumulated by aquatic specie* la dependent to a great extant on Its physico-chemical characteristics, aost notably, Its vatar solubility and llpophlllclty. PCBa hava been shown to blcaccueulat* In flah species and aside from Its obvious significance from an ecological standpolot, bloaccuaulatlon among lover vertebrates has been recog nised as a potential human health hatard aa veil. Bloaccuaulatlon can ba assessed by the calculation of the bloaccuaulatlon factor (concentration of chealcal In whole flsh/eoneentratlon of chealcal In vatar). This factor Is a characteristic of tha fish spaclas and duration of exposure to th* compound and la independent of ch* vatar concentration of the chealcal. Figure 1 Illustrates the time dependency and magnitude of bloaceuaulatlon of Aroclora 12SA and 1268 by channel catfish exposed to these alxturas In water (Mayer at al. 1977). It ean ba seen that even after 77 days of exposure, tha eoneantratlon of PCBa In flsb had not reached the steady scat* even though bloaccuaulatlon had proceeded to over 60,000 times the concentration of these alxturas In tha exposure vater. These data demonstrate that In order to achieve a steady staee concentration of PCBa In fish, an extended duration of axpoeura la required. Steady state concentrations of PCBs vere reached In fathead alnnovs vhlch had bean exposed to Aroclora 1262, 1268, 1256 or 1260 for a period of eight aonths (62 PI 6332, 1977), Bloaccuaulatlon factor* for Individual alnnovs ranged from 32,000 for Aroclor 1262 to 307,000 for Aroclor 1256. The hloaecuanlatloo o.OCBs In flah ha* also bacn assessed by th* us* of feeding studies. When eoho aelaon veto' fed diets containing Increasing concsntrstlons of Aroclor 1256, it vaa found ehat the higher the concentration 97 MONS 026138 60,000 SO,000 b0 ' *0,000 ! W 2 30,000 O Aroclor 125* O Aroclor 12*6 3 20,000 10,000 \A Water 26 Exposure tL 56 77 Time ( Dey* ) Figure 1. Bioseeuoulatlon of Aroclors 12*8 (5.6 u g/l) end 1254 (2 *U|/L) In channel catfleh following water exposure. 98 HONS 026139 of the PCB In th diet, th longer It took for whole body PCI concentration* to reach a ateady etate (Mayer at al. 1977). Furthermore, tha higher the concentration of Aroclor 1254 in tha diet, tha higher the ateady atata concen tration In the anlaal. Thaaa data concerning tlaa to ateady atata vlth different concentration* of PCS* Bay explain tha raaulta of Laatbarland at al. (1979) vho ahovad that PCB level* In coho aalmon fad a high PCB diet (JOO mg/kg) war* only 3.2-fold higher than in eoho fad a lov PCB diet (50 Bg/Vg). It la important to not* that data concerning bloaecumulatlon of chamleala, Including tha PCBa, can b* axpraaaad a* an abaolut* amount of chemical in tha body at any given tlm* or aa the concentration of that particular chemical In a given tlaaua. Thi* 1* an important concept line* in long tarn axpoaur* atudlaa tha flah ar* growing and on* mutt deal with the combination of cooacant Intake of chemical* and the taodancy toward dilution of chair concentration In the animal by growth. Thi* la llluatratad la Figure 2 where rainbow trout war* fad 15 part* par million (ppm) of Aroclor 1254 for 32 weak* (Liab *c al. 1974). Graph A rapraaant* the PCB concentration In flah In ppm, B rapraaant* tha total PCI content of tha flah In mlerograma. and graph C llluatratat the body weight maaaurad over the axpoaur* period. It la clear that although the concentration In the flah Vagan to level off at approximately 24 wetka of dietary expoaure, the abaolute amount of PCB* In the anlmale continued to lncreaae. Theee data Uluatrat* that although a ateady atata may he reached In tana of concentration in the animal during tha expoaure period, total PCB content In abaolute mlerograma/anlmal may continue to lncraaae. Whan rainbow trout were fed a diet containing 15 ppm of Aroelor 1254 for 16 week* and then were awltched to a control diet, the flah which continued to receive PCB* continued to accumulate than, but the trout removed fron the PCB-contalnlng diet demonatratad no decree** in PCB contant (Figure 3). It la Important to recognlae that the abaolute amount of PCBa expreaaed a* ng/flah did not change elgnlfleantly from the 16th to 32nd weak on tha control diet. Thaaa data Uluatrat* the lack of elimination of Aroclor 1254 avan vhen animal* were fed a control diet. If the data ar* expreaaed on a concentration batla however, the altuation la quite different. Thi* la llluatratad In Figure 4 where PCB concentration In lipid la plotted agalnat average flah weight and the tlm* In week* after removal from the PCB-contaln lng diet. Theee data clearly ahov that the PCB concentration decreaaed over tlm* after removal of tha flab from the PCB-contalnlng diet and thla decree** In concentration can be accounted for almoat entirely on the baala of dilution through weight gain of tha animal*. An Important factor whleh determine* tha extent to which PCBa ar* accumu lated by flab la tha degree of chlorination of the biphenyl nucleua. General ly, lowar chlorinated PCB mixture* ar* accumulated to a laaaar extant than more highly chlorinated mixture*. Thla la llluatrated In Figure 5 by ehannal catflah accumulating Aroelor* 1232 and 1248 (32 and 481 chlorine, reapectlvely) to a laaaar extent than 1254 and 1260 (54 and 60X chlorine, raapeetlvaly) (Mayer at al. 1977). The lover aceuanilatlon 1* due to more rapid matabollam tnd elimination of the lover chlorinated ehloroblphenyla, Thla ha* alao been demonatratad in fathead minnow* expoaed to Aroclor* 1248 or 1260 for 66 day* In water and tranaferred to PCB free-vater fdr 34 day* (DaFoe at el. 19.18). Figure 6 ahov* chromatogram of Aroclor 1248 realdua* In mlnnov* after the 34 day vaahout period cohered to an Aroclor 1248 atandard. It ean b* aean that the competition of PCB raalduea In minnow* 1* different 99 MOMS 026140 a Dietary expoaure else (wttki) Figure 2. lloaceuaulatioa of Aroclor 1254 In rainbow trout followlag dietary expoaura (15 ppa for 32 weeks). 100 MGNS 026141 Figure 3. PCS concentration In ralnbov trout following dietary expoaure to 15 ppn Aroclor 1254. A. 32 week expoaure; 8. 16 week expoaure followed by tranafer to FCB-free diet. 101 MONS 026142 Average fla h M ig h t (graaaa) 80 Fi|ur t. .PCI concentration in lipid and weight of rainbow trout renoved iro..dla{ containing 15 ppa Aroclor 1254 after 16 week*. 102 HONS 026143 Percent chlorine In PCB alxture Figure $. PCB concentration In channel catfleh after dietary expoaure to 24 pg/g Aroclora 1232. 124B, 1254 or 1260 for 193 daya aa related to the percent chlorine In each alxture. 103 HONS 026144 R e la tiv e In te n s ity Figure 6. Cae chroMtogreae of (A) Aroclor 1248 atandard; (I) raaiduae In fathead mianovt. 104 HONS 0261^5 from th etandard. A dlfftrant rttult vaa obtained for Aroclor 1260 where FOB residue* In alonovs wart aaaantlally ldantleal on a component bail* to Aroclor 1260. Alon| with blotraosformstloa vblch any data mint the extant to vhlch PCBa can be seeuaulatsd by fish, tbe lipid content of the animal must alto be conaldered. Female fathead mlnnovs accumulated about tvlce aa much Aroclor 1248 or 1260 from water ae aalea and thla vae due to the higher lipid content in the female anlmela (Dafoe at el. 1978). DISTRIBUTION Manama la Beceuae of the many different ehemleal and biological psepertl** of tbe individual congener* vhlch make up commercial PCS mlxturea. It hea proven extremely difficult to assess the environmental fat* and dlepeeltlon of theea chemical*. Thus, moat itudle* to dace have examined the tlaao* distribution and metabolism of Individual PCI congener*. Thla ha* not only led to a better undemanding of the pharmacokinetic behavior of polychlorinated biphenyl*, but to the elucidation of tha atrueture-actlvlty relationship* which determine their dlspoaltlon end persistence. Administration of Individual PCS congeners to leheratory aalmale has demonstretad that tha lover chlorinated species era more rapidly mataholltad than the higher chlorinated compounds and this contribute* to tb* bloaccumulatlon of th* latter in animal tlaeuee. These types of studl** have also shown that the tlssua absorption.and distribution of individual PCB component* 1* both apacla* and organ dependant. Tha examination of the distribution and excretion of individual PCB congtnsrs In rata ha* provided an Insight into th* fat* of th* nor* complex conmerclal PCB mixtures (Matthev* and Anderson 1975, luts at al. 1977). Tha Individual congeners studltd Included 4-chloroblphenyl (1-CB), *,4'-dlchloTOblphenyl (2-CB), 2,4,3,2',5'-ptntachlorobiph*nyl (5-CB), and 2,4,5,2',4',5'haxachloroblphanyl (6-CB). These particular congeners vara chosen btcaus* their degree* of ehlorlnatlon are similar to those of coMerclal PCB mixtures. It vet determined that following Intravenous Injection of a single dose of each of the congeners, tha PCBa vara removed rapidly from tha blood and Initially stored In liver and muscle. More Importantly, the rata of excretion or redistribution of tha PCBa to tlaaua* with high lipid content could be related to their degree of chlorination. Although tha storage of netaholltss did not occur to any appreciable extant, the percentage of total radioactivity accounted for by metabolite* decreased ae th* percent .chlorination of tha particular congener increased. Ninety percent ef the mono-, di- and psnttchlorobiphenyl* va* excreted during'tb* 62 day experimental period. Th* data further Indicated that lea* than 20Z of th* administered doe* of 6-CB would ever be excreted, Theee investigator* thus suggest that tha ratea of elimina tion of PCB eongenere are determined by their rates of metabolism. Tht data In Figure 7 shov the tlseu* level* of 6-CB after a single l.v, adalnlstratlon to th* ret end those in Figure 8 show the cumulative dally excretion of 1-CB, 2-CB, i B end 6-CB In feces. It Is obvious that rats* of metabolism ef the Individual PCB eongenere play a role In their capacity for bloeeeumulatlon end that this metabolism Is dependent, at least In part, on 105 MONS 026146 ?ia, hrt. Figure 7. Tliaue level* of 2,4,5,2'.4*,5'-hexachlorobiphenyl (6-CB) follqplQg, a a ingle l.v_lnjaction of 0.6 ag/kg to sale rat*. 106 MOHs 026147 Fleur# 8. Pare#nt of radioactivity of th# total doa# excreted In fee#* for 7 or 42 day# followlnf a #ln|l l.v. injection of 0.6 |/k| of 1-C (A), 2-Cl (1), 5-CB (C) or 6-C1 (0) to sale rata. 107 HONS 026148 the degree of chlorination. However, the position of the chloTlne atoms on the biphenyl nucleue eleo determines the rate at vhlch any congener la eetabollted. Chlorine substitution patterna and their role In metabolism will be discussed In a later section. The data gleaned fro* accumulation studlea ualng polychlorinated biphenyl lxturea support those vhlch have examined the behavior of individual conganers. Analyses of PCB residues fro* human tissues have denonstrated the retention of relatively few congeners when compared to the gas chroaatographlc profile of the commercial mixtures (Jensen and Sundstrom 1974, Takushljl at al, 1979). Studlsa on the pharmacokinetics of commercial PCB mixtures In experimental animals have elso denonstrated the retention of only certain congeners In adipose tissue (Ransan 1979). Fish Although a number of elegant studies have been performed investigating the tissue distribution of both coasserclal mixtures of PCBs and Individual congeners In aaamals, very few studies have been csrrled oue on the dlsposltlon of PCBs In fish tissues. The data In Tabla 1 indicate the concentration of PCB residues and Upld content in tissues of coho salmon vhlch vara fed equal amounts of three chloroblphenyls for 117 days. For most tissues exam ined, as Upld content Increases, so does PCB concsntratlen, although the data for brain, heart and spleen suggest that lipid eontene may not be the only determinant of PCB accumulation. Similar to vhat has bean described In nasals, the mors highly chlorinated congeners vers found In higher concentra tion In the tissues examined. The dets of Culney et al. (1977) suggest that disposition studies expres sing only PCB concentration Id various tissues may be misleading. The data In Figure 9 show the concentration of C-labeled 2,3,2',3'-tetrachloroblphenvl residues In tissues of rainbow.trout for up to 36 days after water exposure. The highest concentrations of C activity vers in visceral fat, bile and eyes plus associated periorbital fat. However^ when PCB residues in the sane flan ware expressed ss a percentage of total C content In the whole animal, It can be seen that carcass, sksletal muscle and skin contained the highest percentage of tctrachloroblphenyl and visceral fat, eyes and periorbital fat and bile contained the lowest (Figure 10). These considerations of tissue mass should be taken into account In any disposition studies. METABOLISM Mammals The rola of metabolism in the capacity for bloaeeumulatlon of PCBs has been alluded to previously, and the reader Is referred to a recant, compre hensive review of their uptake, disposition, bloaeeumulatlon and metabolism -(Safe 1980)-- Briefly, those-PCS congeners vlth higher chlorine content appear to be preferentially ratalned. Although routes of PCB metabolism can vary considerably depending on the species studied, a fsv generalisations can be put forth. The major metabolites of PCBs Include phenolic products, dlhydrodlols, and aulfur containing metabolites. Hydroxylatlon Is favored at the para position In the least chlorinated phenyl ring. In the lover chlorinated 106 HONS 026149 Table 1. Lipid Content and Concentration of PCB Raaiduaa in Coho Salmon Tlaauea* Tiaeue Lipid Content <vt X) 2.4,6,2',*',6'Heaechloro- blphanyl 2,4,9,2',$'Pentachloro- blphanyl (jig/g vat valghe) 3,4,3'.*'Tatrachloro- biphanyl Adlpoaa tlaeue Brain Spinal coluan Lataral line uacla Intaatlnaa Stomach and pyloric eaaca Liver Whitt muacla Heart Spleen 73.0 7.1 6.6 6.1 4.3 3.3 2.9 2.6 ,,R.D,,b N.D. 18.SO 0.38 1.40 1.90 0.82 0.9$ O.SO 0.64 1.50 2.10 19.30 0.31 1.40 1.80 0.79 0.76 0.3$ 0.63 1.20 2.00 10.60 0.13 0.92 0.77 0.30 0.43 0.25 0.29 0.98 0.85 *Cni|r at al. 197$. DNot datactabla. 109 HONS 026150 4-CB c o n c e n tra tio n (ppa) figure 9 .Hague concentration* of C In rainbow trout exposed to 14 C-i-CS in water and transferred to 4-C3-fr*e water 'day 0}. 110 HONS 026151 Percent of I n i t i a l to ta l C-4-CB residue Tlae after transfer to *-C8-free water (days) FUurc^O^ Percental* of total 14C in tissues of rainbow trout exposed to -Cl in water and transferred to 4-CB-free water (dey 0). Ill WOUs 0^615^ biphenyl congeners, th* part positions of both biphenyl ring* and carbon atom* vhlch arc para to tht chloro aubatltuant art all readily hydroxylated. la general, at tha dagrat of chlorination lncraatai on both rlcga, tha rat* of metabolism dacraaaaa. Finally, oxidative metabollam of PCBs la further facilitated by th* availability of tvo adjacent unaubatltutad carbon atone to allow arena oxide formation ar an Intermediate tep In aetebollem (Kato et al. I960). Although It ha* been definitively demonttrated that both th* degree end petition of chlorine atom* determine* both the rate* and route* of matabollem of FCB congener*, other factor* may alao be Involved. Following examination of th* pharmecokloeele* of etveral tetrachloroblphenyl laomer* in th* mouae, it vaa demonttrated that thoee laomer* which poaseeaed adjacent unaubatltuted carbon atom* were more rapidly eliminated than tboa* that did not. but rate* of elimination of th* former group of laomer* were ilgnlfieahtly different from each other (Suglura et al. 1975, Suglura et al. 1976). Thi* led to tb* auggeatlon that FCB laomer* vlth lncraaalng ortho-chlorine substitution pattern* ar* lea* planar end thu* leea eoluble in fat tlaau*. Subsequent atudle* demonttrated that th* accumulation of th* more highly chlorinated PCBs la not directly proportional to llpophlllclty (Butclnger 1976. Tulp and Butxlngar 1976). Thum, although th* number and poaltlon of chlorine atom*, and the octanol/water partition coefficient* of FC6 congener* nay euggeat their pradlapoaltlon for matabollem. there doe* not appear to be on* overriding factor which determine* th* rat* of elimination of Individual FCB congener* In mamma la. FHh Thera 1* comparatively little information available concerning FCB metabolite In flch. Varloua Investigation* hav* demonttrated that fleh ar* capable of metabolltlng FCB* to compound* more polar than the parent (Sanborn et al. 1975, Metcalf et al. 1975. Bend at al. 1976, Hint and Mattumura 1977) and th* principal nacbahlam Involved appear* to be hydroxylatlon. Trout and akate, aa wall a* the Invertebrate*, red crab and lobster, can metabolize biphenyl to both th* 4-hydroxy and 2-hydroxv,derivative* (Craaven et *1. 1965, Gruger at al. 1975). Folar metabolites of 1 C-lab*l*d 2,5,2',5'-tetrachloro blphenyl were demonstrated la th* bile of rainbow trout and one of these metabolites was Identified as a glueuronld* conjugate of 4-hydroxy-2,S,2',5'tetrachloroblphenyl (Malancon and Lech 1976). Dogfish metabolized 2,4,5,2',5'' pentaehloroblphaayl extremely slowly but polar metabolites ver* also found In tha bile of these animals (Bend at al. 1976). Although ehe effect of the position of chlorine atoms on th* biphenyl nucleus on metabolism has not been studied In fish, an effeej^of ehe number of chlorine* has been reported. Green aunflah ver* exposed to C-l*b*l*d trichloro-, tetrachloro- and pantachloroblphenyl congener* on days 0 and 9 and aacrlfleed on day 16, and th* percent total radioactivity In fleh e* parent compound and polar metabolites vat determined (Sanborn et al. 1975). Th* data In Table 2 Indicate-that, ea-tc the mammalian species, the rat* of metabollem appears to be Invar**ly related to the degree of chlorination. Furthermore, thee* dsta suggest that th* availability of more then tvo adjaeene unsubatituted carbon atoms say be necessary for an appreciable rat* of FCB metabolism. It should be noted that although some of th* generalizations regarding th* 112 HONS 026153 Table 2. Effect of dagTee of ehlerlnjtlon os PCI Mtabollra In green euofieh PCJ 2,3,2'-Trlchlero 2.5.2'.S'-Tttracbloro 2,a,5.2'.S'-Pantachloro *Sanbom at al. 1975. 2 Parent 1 99 99 X Polar Mtabolltaa ao 1 1 113 MONS 026154 metabolism of PCBa which have baas determined from mammalian system* nay ba applicable to thalr natabolla* lo flah, flah aatabollia PCBa auch sort slowly than thalr mammalian Uountarparta. An attanpt baa bate mad* to briefly dlacuaa tha factora which affaet the bloaccumulatlou, paralatanca and matsbollam of PCBa lo a mnbtr of vartabrata species. Although a number of valid fanaralltatlona have bean pot forth regarding tba pharmacokinetic behavior of PCBa, thalr fata and paralatanca lo animals-with altered physiological etatea haa not been ao eompraheoalvaly addrat ted. Of both anvlronnantal and toxicological concern la the affaet of reprodnctlve proceeeee on tha disposition, natabollam and elimination of there compounda. The remainder of thla report will, therefore, deal with tha affect of reproduction on the pharmacokinetics of apeclflc PCB congaoera lo both flah and maamala. Plah 2,5,2' iS'-Tetrachlorobiphenyl (4-CB) haa bean shown to ba one of tba major PCB eoBgenere In contaminated flah (Velth 1974). In adult rainbow trout exposed to tble PCI coogaoar in water, metabolism waa alow aa avldanead by laaa than 12 of the compound accumulated by tha flah being laolatad from blit la metabolised form (Melancon and Lech 1976). Tha affect of agg maturation oo the dletrlbution and elimination of 4-CB ratlduea in fosalo rainbow trout waa atudltd by exposing two year old flah to 1 c-labelod 4-CB In wateV for 36 houra In December with subsequent tranafar to flowing, nen-roclreulatlng prlatlna water (Culney at al. 1979). At deelgnated time* tbaraaftar. flah wart eacrlficed and aalaetod tlaauae analysed for radioactivity. In thaac anlmalt. vltellogeneela waa Initiated In June and apawnlng commenced In October. Figure 11 ahowa tha ollalnatloo of 6-CB from fomala trout on a whole body baale and weight of maturing agga at each sampling time. Prom January through August, tha allmlnatlon of 4-CB vaa slow with an apparent t 1/2 of elimination of 1.76 years. Vat weight of tha agga waa low from January through Hay, then Inereaaed gradually from tho end of Juno through Auguat. Beginning In Stptambor, there waa a rapid lncraaae lo egg weight which reached a peak In October. Tha mature agga were than spent from tha flah caualng tha vtlght of tha agga which remained In the peritoneal cavity to decrease dramat ically by December. Tho vortical dotted lino Indicates tba laat sampling time before tha rata of 4-CB elimination from tha whole flah appeared to increase. Using thla as rafaronea point for tha change In 4-CB allmlnatlon rata, it can ho toon that tho lncroaotd elimination occurred at the end of egg matura tion end thronghomt the apawnlng aeaaon. The t 1/2 of elimination vat reduced to 0.5 yeero. Tho roletloneblp betveee egg maturation In female trout and 4-CB contant of tha oggo la given In Figure 12.. The major finding waa that aa egg weight increased to peak, the percent `*C realdue In the agga also increased. When tha female ovulated end the agga vara epent from the flah In the act of spawning, the 4-CB associated with tha agga vat eliminated from tha body. Thla resulted 4n- a reduced total percent T C realdue associated with the fewer ggs that remained In tha peritoneal cavity by December when the spawning aeason vaa over. Taken together, these finding* suggest that allmlnatlon of 4-CB In tha spent agga was. In part, responsible for tha more rapid allmlnatlon of 4-CB from female rainbow trout from September to December. 114 HONS 026155 Ti*a aftar transfer Co 4-CB-frta wacar (weak*) Figura 11. Effaet of pawning on tllalnation of 14C-4-CB raalduai In fault rainbow trout and vac valghc of aggi following tranifar of aniula to 4-CB-frat watar (day 0). Tha vartleal lint rapraaanca tha laat group of fiah aaapltd bafora aliminatlon rata incraacad. 115 hqns 026156 Sg8* 10 20 30 AO 50 Tlae after transfer to A-CB-frea water (veeks)14 Figure 12. Redistribution of 14C activity fro* fat depots to eg|a of feaale rainbow front following'transfer of aniaala to A-CB^fraa water (day 0), expreased as a percentage of whole body 1 C-A-CB residues. 116 MQNS 02615? Although th 4-CB that accumulated in the aggi may hava originated from a rcdlatrlbutloo of 4-CB from aavaral compartments of tha body, a ceetporel relationship vaa found batwaan tha accumulation of 4-CB in maturing agga and tha depletion of 4-CB fro* body fat. More explicitly, the Increase In egg 4-CB content vaa aeeoclated Initially vlth a reduction In 4-CB content of aye* and periorbital fat. Tbla vae followed later In time by a dacraaae In 4-CB content of visceral fat. It la probable therefore, that Iona of the 4-CB accumulated in maturing egg* wee derived from tbaae body fat dapots. Cenerally, vet veight of eggs; eye* plua periorbital fat and vlaeeral fat lncraaead after transfer of female trout to 4-CB-free vater but the Increase In vet veight of the maturing egge far exceeded that of vlaeeral fat and ayaa plua periorbital fat. 4-CB concentration In the egge remained aaeentlally _ constant throughout the study, viille that of vlaeeral fat and eyes plus perl-' orbital fat decreased. The reduction of 4-0 concentration la these latter tvo body fat depots cannot be explained solely on the basis of an lncrsase la fat vet veight, demonstrating that other factors vsre also Involved. Johnson and Morris (1971) found a dlraet linear relationship between the FCB and lipid content of eggs stripped fro* freshwater fish in nature. This suggests that lipid content of the eggs Is a determinant of the aagnleude to vhlch FCB accumulates la the*. What is lets clearly understood in.the mechanic* of tba depletion of 4-CB fro* visceral fat and periorbital fat and eyes In the present study. The observation that the vet veight of visceral fat, eyes, sad periorbital fat resalned virtually eonsttnt during the time that 4-CB content of these tissues vat decreasing Indicates that the redistribution of 4-CB la not due to the redistribution of body fat. In so*e migratory salmonldt vhlch do not feed during the tpavnlng run, the percentage of oil In the vhole fish drops to lets than 10S of tha pretpavnlng value (Love 1970). Under these conditions, the percent FCB body burden transferred to the eggs may ba even greater. These data suggest that spawning nay provide an alternate route of elimination of persistent FCB congeners In fish speeles. Mammals Whether similar alterations occur In tha pharmacokinetics of a peralstent FCB congener during reproduction In aanmali was also examined. A number of studies hava been carried out on the transplacental passage of polychlorinated' biphenyls and thalr transfar to nursing offspring. In vlav of the fact that numerous racent monitoring studies have demonstrated the presence of FCBs In human breast *llk in various parts of the vorld (Yakushljl at al. 1978, Brevlk and BJerk 1978, Kuvabara at al. 1979, Mee and Davies 1979), It is surprising to note tha paucity of toxicological loforaatlon regarding their half-lives and disposition In pregnant and lactatlng animals and their rate of accumula tion and disposition in offspring. Previous Investigations In rodents vlth FCB alxtures hive deaonstrsted minimal transplacental passage of these chemicals when compared to levels observed in the milk of lactatlng mothers (Curley at al. 1973, Takagl at al. 1978, Ms enda-et al. 1978). Severer, use of mixtures of FCBs does not allov for an accurate appraisal of the behavior of those particular congeners known 117 HONS 026158 co b* the moat paralatant. Aa mentioned previously, the data of Matthew* and Andarton (1975) auggeat that laai than 202 of tha total adnlniatarad dole of 2(4,5,2*,4',5'-b*xachloroblph*nyl (6-CI) would aver ba axcratad from mala rata. 6--CB waa accumulated primarily bp adlpoaa tlaaua. miaela and akin in thaaa anlaala. Alao, Jaoatn and Sundatrow (1974) have ahovn 6-CB to ba tha oat prevalent PCB in human tlaaua aamplaa. It vaa thua of intaraat to amluaca whatbar thia paralatant PCB congener waa tranafarrad cranaplaeantallp or to nonlag offaprlng through milk. Although tha minimal tranaplacantal tranafar of 6-CB haa baan raportad in lea (Orbarg 1977), thia atudy did not eonaldar eha dlatrlbutlon of tha PCB in mother*, rataa of allaloatlon In milk or dlapoaltlon and rataa of accumulation In offaprlng. Aa la alao eha caaa In-tha afeaamamtioned-lnvaatlgatlon* with PCB mixture*, animala In thia axparlmant vara doaad chronically during preg nancy and/or lactation and thua tranafar of 6-CB from aeoraga dapota In tha mother could not ba determined. Thaaa atudlea do not raflact tha eltuatlon In vhleh expoaura to PCBa raaulta In tha aalectlv* eaquaetratloo of particular eonganara Into atoraga dapota prior to pregnancy through tha matabollam and/or elimination of laaa paralatant eonganara. Va have ehua atudlad tha dlapoal tlon of 6-CB in pregnant and laetatlng mice and.lta accumulation In tha fatua and nuralng offaprlng afear admlnlatratlon of 1*C-labalad 6-CB to famala mica 14 daya prior to mating the animal* (Vodlcnlk and loch I960). Tha data In Plgura, 13 ahow that mo algnlfleant dlfferaneaa war* noted in tha dlapoal tlon of activity In the tlaaua* from pregnant animal* whan compared to virgin control*. On tha day of birth however, liver, adlpoaa tlaaua and kidney 6-CB concentration* war* algnlflcantly higher In mother* than In virgin* tacrlfletd at tha cam* tint. Tha half-llvaa of alimlnatlon of 6-CB In adlpoaa tlaaua In 1-to 12-day pragnant mica and their virgin control* war* 7.5 and 11.6 daya, raapactlvaly, and in tha livar, 17.5 and 15.6 daya, raapactlvaly. During lactation, 6-CB concentration dacraaaad tharply In all tleauet taaplad from aothart (Plgura* 13, 14). Tlaaua concentration* In virgin control* aacrlflead at tb* aaa* time remained elevated at lavala tlnllar to thoa* of virgin* aacrlflead coneurrancly with day 12 pragnant animal*. Tht half-llfa of alimlnatlon of 6-CB from tha adlpoaa tlaaua of virgin* aacrlflead following eha aqulvaltnt of eh* 12tb day of pregnancy waa greater than 100 daya. Tha half-llfa la tha liver of virgin* could not ba determined due to the lack of detectable elimination of tha chemical after thia elm*. Ralf11vaa of elimination from livar and adlpoaa tlaaua fall eo 4.7 and 2.1 day*, raapactlvaly, In laetatlng notbara. little 6-CB waa tranafarrad acroaa the placenta during tha court* of pregnancy. The concentration* of 6-CB in fatuaaa from 12 and 16 day pragnant anlaala wat* 0.71 and 2.45 wg/g tlaaua, raapactlvaly. At birth, total carcaea concentration of 6-CB In pooled newborn* va* laaa than 3 ug/g tlaaua vhleh rapraaanta lata than 32 of eha body burdan In tha mothtra at birth. Throughout tha court* of lactation, auckllng offaprlng accumulated 6-CB through milk. Figure 14 ahowa the concentration of 6-CB In tha mammary gland* of lata pragnant and laetatlng mica. Thaaa data eloaaly parallel thoi* of tomach content* concentration* of 6-CI In mining offaprlng. Tha PCB accuau- 118 MONS 026159 2,0 Blood g Virgin 1.5 H Pregnant/ 1.0 Iterating 0.5 75 50 * 25 8 1 6 12 18 Birth 5 10 IS 20 Kuacle Afci'a Jkil-.n. 1 6 12 18 Birth 5 10 IS 20 Kidney 1 6 12 18 Birch 5 10 IS 20 1500 Adipoee tiaaua 1000 500 .fr-Ck-.O; 1 6 12 18 Birch S 10 15 20 30 1* Liver 20 10 fin.tiJUuJk 1 6 12 18 Birth S 10 15 20 Day of pregnancy Days poatpartus Figure 13. Concentration of 14C-6-CB in aeleeted tiaaua* froa virgin and pragnanc/laecaelng ale*. * rapreaant* aignifleant difference froa virgin eontrol (p < 0.05). 119 MQNS 026160 0 Maamary Bland of Bother* QSto*ch content* of offspring C-6-CB/g tltW M i---- 1------- 1------ 1 I---- 1 6 12 18 Birth Day of pregnancy n -n 10 15 20 Day* poatpartua Figure 14. Concentration of 14C-6-CB In the aasnary gland of pregnant/ 'ctetiog alee and ccoaach contents of nursing offspring. 120 MONS 026161 Isttd primarily In tha Uvar, muscle and skin of suckling young, Adipose tissue samples could n^j be obtslnad from offspring In great enough quantity to allow for reliable C determination, figure IS above that levels of 6-CB vers highest in the liver of offspring at day 5 postpartum and hepatic content dropped sharply by 10 days postpartum. Muscle and skin concentrations of 6-CB reached a maximum at day 10 poatpartum and decreased thereafter, corresponding to the increase in litter weight and decrease in 6-CB concentration in milk. From the design of this experiment. It was not possible to dletlngulsh whether the decreased concentration of 6-CB waa the tesult of elimination or simply a dilution affect due to rapid grovtb of the offspring at this time. In otder to determine the percentage body burden of.6-CB eliminated during lactation, virgin female mice ware Injected with `*C-6-CB 14 days prior to mating and two animals were sacrificed on day 19 of pregnancy and two father* and thalr Utters were sacrificed 5, 10, 15 and 20 days postpartum. C activity was asseased In carcasses of mothers and thalr pooled Utters. The daea In Table 3 show total carcass 6-CB content In late pregnant and lactatlng mothers and thalr offspring. Mothers eliminated 57X of the total dose of 6-CB remaining during lata pregnancy by day 5 postpartum and asaentlally tbalr entire body burden was eliminated by day 20,(> 96X1. Burs lag offspring accumulated 6-CB through milk and all of the 1 c eliminated by the mothers could be accounted for In the suckling offspring. Thasa data support those of ether Investigators who have suggested the placenta to ha an affective barrier to the passage of FCBs In experimental animals vharais the maamary gland represents an Important route of elimination (Curlay at al. 1973,-Orbarg 1977, Maauda at al. 1978). These results also support the human data of Maauda and eovorkara (1978) who demonstrated that although FCB levels In the blood of tlllbern fetuses vara much lover than those In thalr mothers, nursing Infants from the same population had higher levels of PCBa in blood than thoae In maternal blood. The higher concentrations of 6-CB In adipose tissue, liver and kidney In mothers on the day of birth In this study nay be the result of lipid mobiliza tion to the proliferating mammary gland*. That la, naiaary gland development and/or milk production may b* occurring. In part, at the axpanse of stored adlpoa* tissue lipid during late pregnancy resulting in a further concentra tion of 6-CB in fat. The Increase in liver and kidney 6-CB content at this time tuggesta soma translocation of the chemical to tha developing mammary gland eo that theaa tlaauas may have been exposed to more circulating PCB. Although the concantratlon of 6-CB wet not significantly different In tha blood of virgin and pregnant animals. It la possible that tbara was an increased turnover of 6-CB In tha blood. In support of this hypothesis is tha faet that FCBs have bean shove to baeona associated with pIssue lipoproteins la vitro (Matthava at al. 1977, Tonechka at al. 1963) and that mammary gland lipoprotein llpaaa activity Increases and adipose tlasua lipoprotein lipat* activity decreases during lactation to faellltat* trlacylglycsrol Incorpor ation into milk from chylonlerout and vary low density lipoproteins (Under st al. 1974). Whether Tatta of elimination and the percentage body burden of 6-CB sllnlnatad through breast milk In huaans are similar to those parameters observed In mica cannot ba aaeartalnad. Elimination through tha lactatlng 121 MONS 026162 Off print liver 120 110 100- W2 80 70- 2 60uI SOt4 40- 9 5o- 20- 10' n Offspring nutclc ^ Offspring skin /V/, / /, V/ Birch Day* potpar:ua Figure 13. Conetntrtlon of 14C-6-CB in ti**ua* fro offpring of 14C-6-CB pr*cr*c%d-nothar. In*t M*n litter weight* on day* o' **crifica. 122 HONS 026163 Table ). Transfer of C)6-Cg froa Bothers to nursing offspring Day ef sacrifice Mothers . _______________________________ ag 6-Ct/ Percentage total Mf 6-CB/ total dose, carcaM carcaM elialnatad* ag 6-C8/ litter Offspring* Mg 6-CB/ g litter Fercentage of aothsr'a does accuaulatad Day It pregnancy Birth Day S postpartoa Day 10 poetpartua Day IS poetpartua Day 20 poetpartua 0.162 -- 0.372 0.167 0.038 0.016 IS.12 -- I.St 3.68 0.87 0.37 0 2.7* S6.8 80.6 tS.6 98.1 -- -- 0.S6S 0.810 0.763 0.900 --- IS.31 13.97 11.18 12.SO --3.00* 63.2 96.0 86.2 106.6 Virgin fosuls ales prstreatsi! with | C)6-C8 (approiiaataly I.IS ag/anfasl) two waska prior to sating. .F.aeh value rsprssents tha assn of two carcassao. "Represents percentage elialnatad froa does reasinlng In aethers at Day It ef pregnancy. jEach value represents the aean froa the cerceeeee of two pooled lfttere. Approx lasts, froa figure IS. 123 MONS 026164 usury gland may not only depend on ths number of nursing offspring sod dl*t of the mother, but slse on ths dependence of t psrtlculsr specie* on dietary vsrsus storsd lipid for ths production of milk trlscylglycsrol. Hs hrr* obeervad ehst pareaetrlai fst stores vsrs significantly dsplstsd In 20 day poetpertum nits nursing average sited litters. Milk fat content also differs among species. It has been demonstrated that human silk contains epproalmatsly *Z lipids, whereas mouse milk contains up to 13Z milk fat (Jenoaea 197*). We have recently obtained similar results following the administration of 3,4,5,)',*',5'-h*juchloroblph*nyl to virgin female mice (Vodlenlk tad Leeh 1982) . .This latter compound was sequestered Into storage depots and minimal transplacental passage was obeervad following mating of these animals. Bowever, mothers eliminated nearly lOOl^of their body burden to nursing offspring, end tbs bulk of ellmlnsted C activity could be accounted for upon analysis of pooled litter carcasses. These data suggest that the pharmacokinetic behsvlot observed was not dependent on the planarity of the PCB Isomer. Those Investigations with tetrachloroblpbeoyl in the rainbow trout end two benechloroblphenyl isomers in the mouse indicate that once persistant PCBeare sequestered into storage depots, they can bs redistributed to other lipid-rich tissues (milk, yolk) and thus be eliminated from the animal via alternate route*. Is both fish sod tassels, the physiological processes sasoclstad with reproduction offer alternative routes of elimination which can significantly affect the pharmacokinetic behavior of PCBe. The mochanlsm(p) by which tho transfer of PCBs from storage sites in the mother to developing oocytes in fish or milk in maamala occurs remain* unknown, but it le possible that the endocrine changes associated with vitel logenesis or the onset end maintenance of lactation nay be Involved. Estrogen and prolactin In mammal* and aetrogen In fishes are responsible fot the biochemical and physiological alteration* obeervad during lactation In mammal* and during praepawnlng end spawning in fish, and havs also been shown to altar plasma lipoprotein profile* and/or turnover. Since PCB* have been shown to associate with plasma llpoproteine (Matthave at al. 1977, Vomachka *t al. 1983) , and tinea plasma lipoprotein profile* are altered during vltallogenasle In flab and late pregnancy in masals, this phenomenon may be, In pert, responsible for the obeervad redistribution of PCBe described In the present studies. In conclusion, ths physiological adaptations associated with reproduction In both flch end mammalian specie* may profoundly alter the distribution end elimination of ealaeted PCB congeners. We are presently attempting to assess the possible role of plasma lipoproteins acting as carriers of PCBs from sites of storage to tbs developing oocyte In fleb end 12* MONS 026165 REFERENCES Band JR, Hare LG, Guarlno AN, Rail DP, Pout* JR. 1976. Natlooal eonfarenct oo polychlorinated blphenyle. Waehlngton, DC: EPA, Office of Toxic Subtcaocaa, p. 292. Bravlk EM, BJark JE. 1978. Acta Pharmacol, Toxicol. 13:59. Craavan PJ, Parka DT, Vllllane RT. 196$. Blochea. J. 96:879. Curley A, Buraa VW, Grim ME. 1973. Pood. Coaaae. Toxicol. 11:171. DaPot DL, Vaith CD, Carlaoo RV. 1978. J. Flab. Raa. Bd. Canada. 35:997. Padaral Raglatar. 1977. Toxic pollutant affluent atandarda: acandarda for polychlorinated blphenyla (PCBa). Final Daclilon. Pad. Rag. 12:6532. Cruger EH, Jr, Karrlck ML, David aon AI, Cmby T. 1975. Environ. Scl. Tachnol. 9:121. Gruger EH, Jr, Earrlek Rl, Davidaon AI, Iruby T. 1975. Environ. Scl. Tachnol. 9:121. Gulnay PD, Melancon KJ, Jr, Lach JJ, Pataraon RE. 1979. Toxicol. Appl, Pharmacol. 17:261. Gulnay PD, Pataraon RE, Malancon KJ, Jr, Lack JJ. 1977. Toxicol. Appl, Pharmacol. 39:329. Hanaen LG. 1979. Ann. N.T. Read. Sel. 320:238. Hint R, Kateunura P. 1977. Bull. Environ. Contaa. Toxicol. 18:631. Huttlnger 0, Tulp KIM, Eltko T, 1978. Aquatle pollutanta: tranaforaatlona and biological affaeta. Oxford: Pargaaon Praaa, p. 13. Jannaaa R. 1971. Lactation. Rav York: Acadaale Praaa, p. 1. Janatn S, Sundatroa G. 1974, Aablo 3:70. Johnaon LG, Morris RL. 1971. gull. Environ. Contaa. Toxicol. 11:503. Rate S, McKinney JD, Matthave IB. 1980. Toxicol. Appl. Pharmacol. 53:389. Kuvabara R, Yakuahljl T, Vatanabo I, Toehida 8, Klyaaa t, Kunlta I. 1979. Bull. Environ. Contaa. Toxicol. 21:158. Laatharland J7, Sonitagard RA, loldrlont H9. 1979. Coap. Bloehaa. Phyalol. 63C: 213. Llab AJ, TllTa DD-, SinnhubBT RO. 1971. J. Agr. Pood Ghaa. 22:638. 125 HONS 026166 Lova RM. 1970. Tha chaulcal biology of fiahaa. Nav Tork: Acadamlc Brass, p. 97. .- Luca RJ, DdBrlek RL, Katthaw* HR, Ellng TE, Anderson MV. 1977. Drug Matab, Dlap. Si 36*. Maauda T, Ragava R, Kuroki a, Ruratauna M, Toshlnura T, Takl 1, Ruauda K, Taaaabltt F, Bayashl M. 1971. Food Cosuat. Toxicol. 16:543. Maauda f, Ragava R, Tokudona S, Ruratauaa M. 1978. Food Coaaat. Toxicol. 16:33. Matthava HB. Andaraon MV. 1973. Drug. Matab. Dlap. 3:371. Matcbnva HB, Surlaa JR, Carvar JC, Andaraon MV. 1977. Toxicol. Appl. Pharmacol. 41:201. Mayar FL, Mabrla PM, tandara BO. 1977. Arch. Environ. Contaa. Toxicol. 5:301. Malancon RJ, Jr, Lach JJ. 1976. lull. Environ. Contaa. Toxicol. 13:181. Maa J, Davlaa DJ. 1979. lull. Environ. Contaa. Toxicol. 21:381. Matcalf RL, Sanborn JR, Lu FT. Rya D. 1973. Arch. Environ. Contaa. Toxicol. 3:151. Orbarg J. 1977. Aablo 6:278. Safa S. 1980. Raloganatad blphanylo, tarphanyla, ndphehalanaa, dlbanzodloxlna and ralatad producta. Aaatardaa: Elsavlar/Horth-Holland Bloaadlcal Praaa, p. 81. Sanborn JR, Chlldara VF, Matcalf RL. 1975. Bull. Environ. Contaa. Toxicol. 13:209. Suglura R. lattorl M, Baba M, Goto M. 1975. Chaaoaphara 3:181. Suglnra R, Tanaka B, Takauebl R, Toyoda M, Coto M. 1976. Chaaoaphara 4:31. Takagl T, Otaka T, Rataoka M, Murata T, Aburada S, Akaaaka S, Baahlnaoto R, Vda B, Rltaura T. 1976. Toxicol. Appl. Phamaeol. 38:549. Tulp KIM, Sutalngar 0. 1978. Cbaaoapbara 7:849. Valtb CD. 197*. 1971 baaallna concantratlona of PCBa and DDT In Laka Michigan flab. Duluth, VI: EFA, Rational Vatar Quality Laboratory. Vodlcnlk MJ. Uch JJ. 1982. Tha toxleologlat. 2:136. Vodlcnlk MJ, Lach JJ. 1980. Toxicol. Appl. Fharaacol. 54:293. 126 HONS 026167 Voaachka MS, Vodicnik MJ, Laeh JJ. 1983. Toxicol. Appl. Phanaacol. In praaaT TakuahlJ1 T. Watanab* t, Ruvabara I, Toahlda S, Royaaa R, Bara I, Runita K. 1978. Arch. Eaviroo. Contaa. Toxicol. 7:493. Takuahlji T, Watanab* I, Ruvabara R, Toahida S, Royaaa R, Ruoita H. 1979. Iat. Arcb. Occup. Environ. Health 43:1. Zindar 0, Baaoah M, Flack TR Clary. Scow RO. 1974. Aaiar. J. Fhyaiol. 226:744. 127 MONS 026168 DISCUSSION SUMHART H. J. Malaneon Department of Phermaeology and Toxicology Madlcal College of Wieeonein Mllveukee, VI 53226 DR. KELANCONt Obviously tha PCB problem hat ralatad vary auch to tha aquatic environment became thla la a routa by which tha PCBe hava raachad come humane and tha way tha problem has come to attaotlon. Tha preeentetlon referrlo| In Urge extent to tha Graat Lakaa flab aarllar ahova chla atrong tie-in to tha aquatic environment. Tha explanation for tha dlaappaaranca of tha lower chlorinated PCB eongeoero waa wary lntaraatlng and ralataa to tha pattern wa are aware of that tha lower chlorinated leomera tend to dleappear. In fact, even beyond whatavar degradation occura In river water, wa are aware that many apeclaa of anlmala are capable of aetabollelnf the leaner chlorinated laomara, while flab cannot etabollta the higher chlorinated laonera. Earlier wa were talking quite a bit about analytical work and tha algnlflcance of PCI datamlnatlona. In regard to tha data froa tha Michigan Conaarvatlon Dapartment, It would be nice if wo could aaa the atatlatleal evaluation. Maybe thla can be commented on laear. Another qnaation which ease to wy nlnd la It the total PCBa arc dropping aowewhat. la thla actually a drop In PCBa In ganaral or Juac tha disappearance of tha leaaar chlorinated laonera; la tha pattern of laonera in tha flah changing froa what it waa before? Ha have dlecueeed nalnly tha Croat Lakaa thla morning. What la happening to PCBa In other aquatic ayatana aeroaa tha United Statue? I knov tha Dapartaeet of tha Interior haa a atudy on tha Mlaalaalppl River (tha upper Hlealaalppl River la a problan area for the PCBa) and haa a nationwide taatlng program which I believe covere a period froa 1973 to 1979. Maybe aoaeone could connect on thla later. Tha quantity of PCBa raachlng tha anvlronaant per year nay be leaa now bacauaa of tha varloua controla and the daeTaatad production of PCBa aa PCBa par aa rather than aa lapurltloa aa preaantly. Currently, the relatively large quantltloa Of PCBa praaent la sediments nay, In affect, be a alnk whereby they era relatively Inactive. On tha other hand, It nay be a reaervolr from which In tha future they can keep eupplying tha environment with PCBa even though tha amount of new material coning In la laaa. Are thaaa eedlnenta alnka or are they going to be probleme, aay, la rlvera bacauaa they can be stirred up at varloua tinea of tha year, eprlng flooda, or whan harbors are dredged and so on? This la something wa have to consider for tha future. A brief comment on the second presentation, tha reproduction-related PCB redistribution which can oeeur In vertebrates as different as fish and small manmals. It Is interesting that under normal conditions fish cannot do much to 128 MOMS 026169 eliminate higher chlorinated PCB laomara rapidly (T 1/2 - l.S or 2 years, probably higher for tht more highly chlorlnatad laomara), yat during saxual maturation, agg formation and 10 on, PCB laomara can ba tranaftrrad, to tha aggt and aort rapidly tllmlnatad. tn tha amall maamala, tha fatua do** not racalva tha PCB laomara. Thara vara a coupla of poaalbla raaaona glvan for that, yat during nuralng thara la a tranafar from tha mother to the milk and to tha offaprlng, Praaumably If PCBa vara balng glvan at tha time that tha agga vara being formed or nuralng going on, theae PCBa too, might ba transferred. Xf anyone vould like to atart off the questioning nov or make comment*, you are valcoma to. DR. EDOROOCBt 1 actually have eeveral quaetIona. Plrat of all I vould Ilk* to dlacuaa tha raaulta vith tha excretion of PCBa through milk and tha relevance of that to humane. Of eourae, you cannot do thesa type* of atudlaa In humane. So, vhat va have primarily dona la looked at PCBa In milk. In adlpoae eleeue and In aarum. Often that* sample* vara not collected from tha lame type of people. V* all carry vary lov body burden* of PCB* and *st vovan will excreta PCB* In milk. On tha other hand, thara haa bean the recommenda tion that vonen ought to continue nuralng beeauae of eh* benefit* of nuralng. On* raaaon for doing that la that tha concentration In milk ha* alvay* bean vary lov and alao, In looking at auccaaalv* pregnancies and the little bit of data that v* have, thara do** not aaam to ba ehla tremendous mobilisation In people. They vould navar rid their body burdens In just nuralng, and ae va* already pointed out, one reason la tha dlffsrancs In tha lipid concentration In milk from different epaelas. In tha mouse It la 13 percent, I think it la IS percent in tha rat, and in human* it rangaa somavhar* from on* to four percent. It alao depends on vhar* in the nuralng period you taka tha milk sample, the amount of milk thac la balng produced, and the grovth rat* of tha offspring. Mica and rats have a tramandoua grovth rate compared to the grovth rata of an Infant over a vhola year. Tha period of nursing In human* la also much differ ent. So, thara are a lot of differences, and If you look at tha USDA data, for inataoee, vbere there ver* successive pregnanelea, vonen produced the so-called "eolan colored babla*. Thera vara aaveral auccaaalv* pregnancies vhar* the second Infant vas, also, a cola eolorad baby. So, that la, in a aort of a roundabout vay, again a suggestion that the vonen really don't mobilise a lot of this material, and 1 vas voedarlng If anyone could comment on that! DR. SAFEt Could I make a comment on that? Thera Is a paper out by a group In Japan vhlch lookad at an occupationally exposed mother. I think It va* published In tha Bulletin of Environmental Contamination, and In that particular study tha vomaa axcratad a considerable part of har body burden of PCBa la milk (Takushljl at al. 1971). DR. BELLI (Camera1 Electric). In that study, over IS months of lacta tion, tha mother eliminated 73 percent of har body burden of PCB, and that vaa confirmed in tha milk vhleh vas collected and analysed. DR. TODICRIKi Was thara any estimate In that study of tha amount of fat In tha mother) Vaa thara a significant valght daeraaaa? Va* lipid balng mobilised? 129 HONS 026170 DS. SELL: la tha paper ltaaif that la oot consented on, but I understand fro* othar private conversations that tha lipid In tha woman did aot change substantially ovar tha parlod of lactation. So, I don't think va ara looking at a tranafar simply bacauaa aha la allalnatlng har lipid. I think va ara aaalng a vaahlng out of PCI Into har Baanary gland and than Into tha milk producad. DI. JASON: Camay Jaaon fro* EPA. Thar* la ona othar raport. ~ Wlckhalaar raeantly raportad In tha American Journal of Public Health of the Michigan group of over 2000 vo*an that vara being aaaplad for PBS level* which va* dona for fra*. Thorn* vllllag to pay had PCB lav*la determined alao. Thla va* a llttla ovar 1000 vo**n. Aa It turned out. It vaa vary lntaraatlng. In looking at tha braaat milk, PCB laval* In half of the woman approached tha tolerance Halt of tha PDA which va* 1.5 part* par Billion. Ha alao calculated tha average amount of braaat *llk taken in by an Infant In an 8*month parlod, In thla group of vonan. Given a 50th percentile growth rata In tha child, and aaaunlng that* la no affact fro* anything the child la taking In exogenously fro* the anvlronaaot, thla child ahould approach a ona part par alllloo laval by eight aontha of ago. Dt. W0LPP: On* thing about the Japan*** atudy you are talking about 1* that thla yonan nuraad for IS aontha, an nnuaually long parlod, and It vaa alao a vary large volua* of braaat milk. In any caae, certainly tha body burdan could have bean reduced. I want to cowent on tha ganaralitatlon that tha lover ehlorinatad biphenyl* arc not bioaccuaulatad. 1 think thla la aort of a atatlatleal event. Fleet of all, there are not an *any laonara among tha trlchloroblphanyla that arc appropriately eubatltuted, and I think that It 1* *uch *or* important to concentrate on that. Alao the amount of ao-callad "accumulation" depend* on hov recant the exposure vac. In tha people that va have looked at that vara occupationally exposed to the lovar chlorinated biphenyls, by far the greatest PCB congener in those people vaa-tha trlehlorohlphanyl 2,A,A'- which was alao tha on* that had tha hlghaat concentration in tha ehloroblphanyl mixture to vhleh they vara axpoaad. Ok. STALLING: I wanted to briefly saanarlaa so** information that will b* in a publication sometlM In tha not too distant future in Pesticide Monitoring Journal, fro* tha monitoring of fish residua* of PCBs. There are 100 aawpllng sitae In the DS, and this countrywide data to soae extant parallels but to ao*a extant does not parallel tha Lake Michigan situation. Tha occurrence of PCI* apparently increased between 197A and 1978/79. They vara found at 95 percent of tha atatlona sampled in 197A, 92 percent In 197S/77 and 91 percent in 1970/79. Maxlmu* concentrations did not ehanga appreciably. Concentrations approaching 100 parts par Billion vat weight, and 500 Blcrograns par gra* lipid weight were aaasured as raeantly a* 1978. Of tha four PCB Bixturas aaasured, that 1* Aroclors 12A2, 12A8, 125A and 1280, 125A was Boat widespread. It va* found at 8A parcant of tha atatlona In 1976/77, 98 percent In 197B/7-9| realdue* resembling 1260 war* present at 8A percent of tha atatlona In 1976, 80 parcant In 1978, and Arodor 1268 was prasant at 39 parcant of the atatlona in 1976 and A2 parcant in 1978/79. No results resanbl- ing Aroelor 12*2 wars found in 1976/77, down fro* about 1A parcant raportad In 1973 and about flva percent In 197A. - 130 MONS 026171 UK. WEBBER: Ian Wabber, RTE Corporation. Have you looked at the matabollc product! of PCB* either lo the environment or through animals? The reason that I aaked that queatlon vaa becauac I did lose work with tha Canadian Electrical Aeaoelatlon on trying to ellalnate PCBs from lnaulatlng olla uaing atrontlua 90 high-snsrgy beta partlclaa. There la a correlation between uaing high-energy eleetrona froa beta partlclaa, froa a Van da Craaff aachlna which HIT did, uaing cobalt BO raya which the Japaneaa triad or uaing a ehealcil proceaa like elenental aodlua, to generate eleetrona. In tha envlronaeot you vould tend to get e" aqueous attaching to tha water and apllttlng that ea OH radleala and H radical!. My system Involved PCBa in oil with leopropanol and potaaalua hydroxide, I found that the Isopropanol acted as a catalyst and tha potassium hydroxide was involved in tha overall chain by producing isopropoxide radicals. The PCB becane a radical anion which eliminated chlorine as chloride and gradually degraded, attaching hydrogen from tha leopropanol and eventually ending up as biphenyl. In the process though you got OH radicals. Thera is a competing reaction between the OB radicals and R dot. Bov, the B dot because of its mobility tends to get to other radicals much faster, and tharafora you and up with biphenyl. In soma of the cases, e substantial portion of the cases, you do find that hydroxylotion has taken place. So, the hydroxylotion then produces vstar solubility of the otherwise insoluble PCB. If you then look at the oil, you will find that you have interacted with all the PCB, and you no longer have PCB contelned in that oil. However, the water layer now contains a far worse problem than you started with because of the hydrosylated product. It can go from that by elimination of HCL if it is suitably oriented on the PCB molecule to produce dlbensofurans, and then we eome back to that problem. So I dropped that work because of the generation of dlbensofurans and a worse problem than I started with. DR. MOOLENAAA: Tour queatlon related, I think, to tba products of degradation of PCBs in the environment, and I think you aaw from the talk that we have studied in eome detail tha produete that might be produced, Intermediate producta in the biodegradation by microorganisms of cbolorobiphenyle. Not only have wa dona that, but tha University of Michigan vorkars have as well, and a conaiatant picture emerge* that really tha only laolatabl* inter mediate* ar* tha ehlorobonxole adds which chan dagrsda further. If on* thinks of the degradation of these materials in tha atmosphere which la probably initiated by OH attack, you would expect tha same tort of a degradation pattern for tbea* material* aa you ** for hydrocarbons in general, and lndaod in ch* Applaby experiment from several years ago thsy did nots soma aldehydes and ao oo aa Intermediate degradation products in that study, but most of thosa, of course, ar* also subject to OH attack. Tou vould expect the progression of degradation similar to what you sot for hydrocarbons that sntsr the atmosphere. In terms of water degradation involving fra* radicals, this has baan a subject ot tome schdy over tn* last few years, but at far aa 1 know in most 131 026172 mons cat** thi* roue* of degradation It not very laportant froa an environmental point of view for anything that haa a half Ufa of laaa than ona year or ao. Although the individual raaetlona nay b faat. tha concantratlona of radleala In tha vatar phaaa la vary low, and ao thay do not aaca to ba vary laportant raaetlona froa an anvlronaantal point of view. I would ba vary curprlaad If you would aaa much of that going on with tha ehloroblphanyla In a watar ayataa. Tou aantlonad tha hydroxylatlon of tha chloroblphanyl, If that wara to occur, forming a compound which la now hydroxylatad. wa hava incraatad tha vatar aolublllty quite a bit, and thoaa kind* of natarlala ara known to ba fairly blodagradabla. Tha phanola and tha phanyl aubatltutad phanol* ara fairly rapidly blodagradad, much nora rapidly avan than tha chlorobantolc acid* thanaalvaa. So, If hydroxylatlon vara to occur I would axpact than to dagrada rapidly via tha alcroblal procaaa. DJI, VODICHIJCj I would Ilka to naka a coanant on your data with tha chub In taka Michigan. 1 ballava It ahowad tha concantratlon of PCBa dacraaalng ovar a a lx or aaven-year period. 1 think dapandlng on tha way tha anlaal waa exposed on* ha* to conaldar In any axprasslon of bloaccuaulatlon the growth of tha anlaal. I don't know what tha grovth rata of tha chub la, but If tha animal, for axanpla, was axposad to a spill and accumulated a good deal of PCBa and than grew thara would not ba an sllalnatlon of tha PCI froa tha anlaal. Thara would just b* a dilution. So, whll* tha concantratlon In any particular tlaaua alght have dacraaaad, tha amount in tha anlaal any not hava dacraasad. However, looking at It froa the other sens* where you sea tha dacraasa In concantratlon, It could rslata to the spawning data suggesting that again tha PCBa arc being diluted out. A rainbow trout is capable of spawning up to 10,000 eggs. If thara Is a significant amount In tha eggs and that PCI Is retained by tha offaprlng there could ba that dilution affect again. So, you nay not be getting a dacraasa of PCB In the bloaass, but you are getting a dacraasa In tha concantratlon In any particular animal. DR. MOOLEWAAR: Thank you for that cotaent. I think your point Is vail taken and that 1* tha vary reason why I think tbs Stats of Michigan biologists and thoaa that work in the other laboratory that la Involved have salacted tha chub a* their test species. These arc small fish, ara relatively short-lived, and tha sampling tlaas, locations and tha slsa of tha fish taken and so on vsra all carefully salacted with these kinds of psraaatars In alnd b*causa they wanted to get a true representation of what thay thought tha fish population vat doing at a function of time In tha lake. All of theta factors wart care fully analysed, and tha biologists there faal that thay are truly getting a representative taapla of what would b*'expected to be occurring a* a function of tlaa in Lake Michigan. DR. KIMBROUGHt I have a comment and a question. At far as nursing and the nursing period In humans It concerned, froa telephone calls that 1 gat It Is not unusual in the United Statat nov that wonsn will nurse for 16 months to tvo years. Dr. Moolanaar, you briefly aantlonad In passing that there was a difference in'thr biodegradation of different Isomer*, and I was wondaring If you could eosnant on that a little bit aora. Is that the tana at v* find vlth metabolism, and which are tha lsoaart that ara tha nott persistant? 132 HONS 026173 DR. HOOLENAAR: In regard to your question on the blodegradablllty of th different ltosere, I don't recall ell of that Information In the literature precisely, but It la ay general Impression that tha pattern you aee la nanmsllan ayateaa of degradability when you have adjacent carbona without ehlorinee le what you tea In microbial ayateaa, and ao I believe tha pattern there la fairly conalatent. DR. SAFEs I would like to oake one consent about bloaccuaulatlon. We publlehed e etudy a couple of yeare ego le which we ueed five hexacbloroblphenyle, none of which had two adjacent unaubatltutad carbona. They were hexachloroblphenyls, eo they were relatively eooaetebolizable, and they all had ehlorinee on the f4 poeltlone of both rlngt. Theaa are cheaicala you would look at and eay, "They are not going to be aetabollted very quickly." The difference la the five lsoaera-vas-ln-hele-degTee-e< ortho-subetltutioev and In a eenae as we decreaae ortho-substitution, we tend to lncreaee toxicity. In that etudy, which waa a very prellalnnry 30-day atudy, wa uaed five different apeclea of anlnele. The animals that preferentially bloconcentrated the planar or lower ortho-eubetltuted laoaere were the anlnale oore auaceptlble to PCB toxicity, for exaaple, in the quail and In the guinea pig there waa a very high concentration of the highly toxic l.l.J.l'l'-S'-bexaehloroblphwnyl end In the quail the funny thing about It le that It wee the only compound of the five that vaa left. If you look at the rat, the flah and the rabbit, the five leooera were preaant In about equal eoncentratlone. There eeeae to be another factor not aaaoclated with netabollam but aaaoelated with ahape and of eourae, ehape of the nolecule la aaaoelated with toxicity. So, that la another factor that one haa to conelder, and It la apeclea dependent. DR. BROOKS> Ronald Brooke, G.E. Regarding the consent of the parallel correlation between nlcroblal degradation and aaaaullan, 1 think It la Import ant to rcaeaber that In nlcroblal ayataaa, tha literature la not very extenalve regarding apeclfle laoaera challenged with either nixed or pure culturea. Secondly, In thoae eaaea where there have been, rateo have not been a prlnary concern. What haa been a concern la whether the nolecule degraded at all and whether there ware netabolltea produced. Finally, It la laportant to reneaber that It haa not yet bean eatabllahed that all laoaera enter Into or are evellsble for nlcroblal degradation and ao with the nlcroblal ayacen you have to worry about both eolubllity and transport and the presence of a suitable activity in the organlsas. Of the 33 laoaera vbleh have.been examined there la noticeable difference among various stralna of bacteria, ao that In order to make soac specific statement additional work has to be done. DR. MOOUEMAARi That la an excellent eonaant, and 1 think It la really good that In our first presentation this morning we heard of the work going on In the analytical area to better determine these Individual Isomers. That haa really been a holdup In a lot of the studies that have been dona over the years that va really have not baeo abla to distinguish these laoaera analytically, and I think you know, the real challenge in better understanding the science of .these ehloroblphenyla In the future Is going to be In that area. DR. TOTH: My .name la Jason Toth with Tfk. I want to direct one consent to Dr. Vodlcnik. I found your data with respect to the study on the transfer 133 HONS 026174 and accumulation of PCS* In offspring of maternal mousa notavorthy, If not rasarkabla with raapact to tha Implications for human populatlona. I am vondarlng If tha conditiona under which the original dtpoaltlon of PCBa In tha aatamal aouaa occurred influenced ehe quantity of tha PCBa that vara liberated and accumulated In tha manner? gland and subsequently tranaferrad to tha offaprlng? In othar vorda, la thla phanomana Independent of tha fact of tha maternal mouea, for instance, being exposed, nay, two, three years prior to giving birth, varaua parhapa only 60 daya prior to that? DR. VOD1CHIX: It la awfully hard to expose a mouea three yaare before giving birth, but 1 undaratand your point. Juat looking at tha data we have, comparing pregnant anlmala and virgin animala you can ana that in tha flrat two vaaka there wae atlll acme redlatrlbutlon of tha PCI occurring. If 1 warn to repeat that experiment now I vould probably wait, at laaat one month for redlatrlbutlon of tha chemical. However, when we looked at 3,6,5,3',4',5'-hexachloToblphenyl we did atudy thoae anlmala for up to four montha prior to mating, and what we aaw there waa again about one month to aim weeka of redlatrlbutlon and than no further elimination from tlaauma In virgin mice and again tha fame reaulta that we aaw with 2,6,5,2',6',5'-hexachloroblphenyl. I think, alao, tha fact that thara waa no algnlflcant difference between virgin and pregnant anlmala auggeata that during thoae early phaaaa of preg nancy 1 don't think what we eew wea an artifact of when Che compound wee admlnlatered because there vere no difference* at that point. DR. TOTH: If you were to make a prediction with respect to leetatlng human mothers, Intuitively you would be inclined to say that if they had accumulated and deposited their PCS* aomeclm* prior to pregnancy you would not expect them to be liberating elmllar percentages of thoae quantities, would you, at all? DR. TODlCHIKt 1 vould not expect elmllar percentages, and thie vac brought up earlier, because of the percentage of fat In rodent versus human milk, because of the specie* differences In dietary veraue stored lipid for the synthesis of milk triglyceride, becauee of litter six* and because of growth rat* of the offspring. I vould not expect those value* to be anywhere near what v* have seen In the mouea. SUMMARYt Although level* of the lesser chlorinated PCBa appear to be decreasing, it la not elaar that thla la tha cast with the more stable, highly chlorinated PCBa, Because of the widespread distribution of PCBa, tbalr prasenca In human tlasuer and tbalr tranafar from mother to Infant via lacta tion It teems advisable te continue monitoring for PCBa and to lnvastlgatt mora fully possible effaces of PCBa on eh* developing Infant. 134 hqns REFERENCE Takuehlji T, Vataoaba I, Kuvabara R, Toahida S, Royaaa K, Hara 1. Runlta N. 1978. Long-tars atudlai of tha axcratlon of polyehlorioatad blphtnyla (PCRa) through tha Bother'I silk of an occupationally axpoaad worker. Arch. Environs. Contaa. Toxicol. 7:193. 133 MQNS 026176 CHAPTER 4 EXPOSURE STUDIES - INDUSTRIAL PROCESSES Frtvloua chapter* hava dealt with leeuee turroundlng coanercially-produced PCB (a.g., Aroclora). Hodge* *t al. preaeot the vary coaplex technical probleat aaaoclattd with the detection and quantification of trace level PCB generated at incidental product* of highly dlvara* chemical proeaaaa*. The pattern of PCB thu* produced relate* to the tpeclflc conerclal proceaa under Invattlgation. Atteapt* at Identification of the** PCB by coaparleon* to Aroclora ar* of little value. Several poaelble aelutlon* to thla dliana* era auggaatad together vlth their technical limitation*. The dlacuaalon auaanary elaborate* on tha problaaa of differential aatrlee* In which PCB mutt be Identified, tha pretence of many Interfering eubetancae, and the baela for the 50 ppa regulatory Halt. In addition to the leave* of Halt* of detection and quantification, le the difficulty aaaoclattd vlth analyelng 209 potential PCB eongantra whoa* total quantity cannot, by lav, exceed 50 ppa. The paper conclude* with an ovarvltv of the total Incidental PCB produced per year (13,800 lba) by 135 ehealcal proceeeee, and their fat* in term* of dltpoaal or eaeapaulatlon In manufactured producta. 136 HONS 026177 EXPOSUKE STUDIES RELATING TO INDUSTRIAL PROCESSES CONTAINING INCIDENTAL PCS R. L. Hodge. Dow Chealcal Coapeny Midland, MI 48640 T. A. Robln.on Vulcan Materlala/Chealeala Blrainghea, AL 35253 C. J. Renalar DuPont Daap Water, NJ 0802) X. G. Raley Monaanto Coapany St. Louie, MO (3167 B. Caapera Stauffer Chaalcal Hobbe Ferry, NC 10322 F. Evald PPG Induatrlea Barberton, OH 44203 A. R. Plttavay Heldan, Plttavay Aaaoelataa, Inc. Waahln|ton, OC 20036 137 HONS 026178 Thrs* topic* *r* discussed: (1) Technique* for the estimation of PCB Incidentally formed In chemical procaaaca, (2) th* potential for worker exposure, (3) the potential for envlronaental reltue and (4) the relative Importance of that releaae, compared to existing environmental pool* of PCB. The flnt topic 1* emphailiod. but a Chemical Manufacturer* A**oclatlon survey that cover* th* other three tople* 1* also summarised. TECKHIQPE5 POP THE ESTIMATION Of INCIDENTAL PCB Th* following 1* * quota from "Recommendation* for Improving the Reli ability and Acceptability of Analytical Chemical Data Used for Public Purpose*" by L. H. Roger*, *t. alk, who represent an Ad Hoc Subcommittee (of th* Joint Board/Councll Committee of Science of th* American Chemical Society), dealing with th* Scientific Aspects of Regulatory Measurement (Roger* at al. Ho date.) "Analytical chemical data are central to governmental and Industrial decision* relevant to regulations that can have far-reaching Impacts on society. When tha result* of ehemlcal analysis are used for a regulatory purpose, the Information must not only be reliable and creditable from a scientific point of vlev, but must also be expressed In terms that tha general public can understand sod accept. The demands placed on reliability and creditability are especially great because the general public is likely to encounter chemical data In a contentious situation where It has no basis for determining the true state of sffalrs. Analytical chemists must always emphasise to tha public that the single, most important characteristic of any analytical chemical result obtained Is an adequate statement of Its uncertainty Interval. It Is critical that the public eppreclste the existence and nature of uncertainty In aelentlfle measurements." Analytical Measurement* "As Chlorinated Biphenyls" vs "As Aroclors" The degree of uncertainty associated with PCB analysis as Aroclors has been discussed extensively In th* literature over th* past fifteen years. However, th* uncertainty In measuring incidental PCB, or non-Aroclor resembling PCB, In chemical process screens has only recently been described. Basically, this Is because the analytical science associated with Incidental PCB la only 2-1/2 years old. This paper represents the first attempts by Industrial analytical chemists to quantitate these species. The focus of this paper Is on the detection. Identification and measure ment of chlorinated biphenyls in th* complex matrices of chemical production process streams, where these substances occur ss contaminants measured In parts-per-mllllon. The analytical problems In process stream analysis are significantly more complex than the problem of analysing for Aroclor fluids and similar PCB products In environmental samples, biological fluids and transformation*. 136 MONS 026179 On* to ten chlorine atom* can b* attached to a biphenyl toleculi, thua creatlni ten ehloroblphenyl "homologa." While the ten* homo lot la uaually applied to the deacrlption of an aliphatic hydrocarbon (crlea, It* uat through out chi* paper demote* the homology of chlorine In the chloroblpheoyl aerlaa. Hence, when we refer to a homolog data, we will be talking about a given degree of chlorination. Thera are a total of 209 poaelbla congener* among the ten homologe a* ehown here (Table 1). Recognition of the fact that chloroblphenyl* are actually 209 Individual. Identifiable, chemical compound* l* Important becauae virtually every compound hat different chemical and/or phyalcal propertle*. The problem* in analyzing ehloroblphenyl* from proceee etream* can therefore Involve the detection, Identification, and quantitation of on* to 209 aeparat* chemical aubatancaa. In th* U. S., the principal manufacturer produced a epeclfled aerlaa of thee* mixture* under th* tred* name Aroclor fluid*. Each of thee* fluid* la eatimatad to contain up to 60 different ehloroblphenyl congener*. Becauae each of thee* fluid* va* manufactured under carefully controlled condltione by the direct chlorination of biphenyl, th* number. Identity, and concentration of congener* In each Aroclor fluid wa* very conelatent. When one apeak* of PCS, th* image generally created 1* that of on* or more of the** Aroclor fluid*. Th* minute amount of chlorinated biphenyl* generated during routine chemical manufacturing doe* not produce "Aroclor" fluid*. Thl* fact la of extra** Importance when considering laauca of detection. Identification, and quantitation. When chlorinated biphenyl occur* In a normal chemical proce**. It doe* not naeeaacrlly r**ult fro* the chlorination of biphenyl. Incidental generation of ehloroblphenyl* occur* through different *eta of reaction mechanlamt and eonaaquently, produce* a different congener mix of chlorohlphenyl aubetanea*. Thete proce**es may produce a tingle congener, a aerie* of specifle Isomer* within a given homolog claaa, or a random mix of various congeners, but none are the same a* th* pattern produced In Aroclor fluid manufacture. The capillary chromatogram show* in Figure 1 show* th* separation of a series of trace level chlorinated biphenyls which war* isolated fro* a chemical process stream after weak* of Intensive analytical effort. Th* major component eluting at 31.8 minutes la equivalent to 3 ppm baaed on rh* original aampl*. Th* chromatographic profile appear* to have many of the character istic* of an Aroclor fluid) however, closer observation show* that only two or three of tha component* have common retention lndlce*. After GC/MS confirma tion, It va* learned that all 26 raaolvad component* vera isomer* of th* same homolog elaas. The chemical manufacturing situation* which generate these minute quanti ties of chlorinate biphenyl* ar* extremely diverse, and th* type and quantity of congener* produced ar* unique to each process. At least 11 different chemical rtactlona ar* known which lead to th* generation of chloroblphanyle, of which direct chlorination of biphenyl 1* only oae. It 1* Important to nota that congener* which do not occur (or have not bean ohaarved) in th* direct chlorination'Of biphenyl (1;*., Aroclor fluid manufacture), can and do occur, and have bean observed In chloroblphanyl Incidental ganaratlon. 139 MONS 026180 Tabla 1. Distribution Of Chlorlna Anoug Poaslbla PCS Coogansrs 0 Chlorlna 1 Atons 2 on 3 Mug 4 15 Totali Chlorlna Atons on King A 1 2 34 5 Conganars 3631 It 18 9 3 21 36 18 6 21 18 6 63 mmm JL 9 45 81 34 20 - 209 140 HONS 026181 Figure 1. Capillary ehroaatograa of 26 PCB hoaolog liMri Isolated fro* a chaaleal aanufaeturlni process streaa. 141 MONS 026162 Status of the Identification of Chloroblphenyl Cont triers The Identification of chloroblphenyl congener* hat baen tha subject of externlve ratearch. Almost all raaaarch on laoaiar ldantlfIcatlon haa baan bated on analysis of Aroclor fluid*. For example, Webb and McCall (1972) aaparatad component Isomers fro Aroclor fluids by preparative get chroma tography and Identified the* by coaparlson vlch synthetic materials, using both GC retention tine* and Infrared spectra. Sissons and Haiti (1971) characterised tha major constituents of Aroclor 1254 by proton magnetic resonance and aas* spectrometry aftar previous seperatlons by gas chroaatography. Falllscarl at al. (1981) Identified 73 FCB congeners In a mixture of Aroclor 1016, 1254 and 1260 fluids using capillary gae chroaatography vlth electron capture detection. Confirmation was obtained using capillary gas chroaatography/nagatlve chemical Ionisation mass spectrometry. Albro at al. (1977, 1979, 1961) provided characterisations of Aroclor fluids using capil lary gas chroaatography. In Aroclor fluid* 1248, 1254 and 1260, they found 110 conganara and vara able to Identify 106, based on method of eo-chromatogrsphy (coincidence of retention lndicaa) using 43 available standards and "best fit" Batching of sussed half Indices. Further confirmation of Identi fication was obtained by separating tha Aroclor fluid Into fractions contain ing four, three, or lees than three ortho-substltutad chlorine by absorption on chercoal, as described by Jensen and Sundstroa. Albro at al. found that the retention behavior of chloroblphenyl congeners In Aroclor fluids le sufficiently different for different liquid phases that one ean generally find a phase capable of resolving any given laoaer pair. Albro and covorkers have deaonstrated a technique vhlch hold* promise for successful solution of one key problem In chloroblphenyl analysis: The successful resolution of all lsoaer mixture* Into their Individual coaponents and by providing a catalogue of 135 congener rataotlon lndicaa on a variety of coluans, and by predicting the retention Indices of the remaining congeners. Hovevst, Albro warns the users of this technique that further confirmation studies using Ik end IM1 vlth known standards is ttlll necessary. This coumant Is particularly appropriate for those congeners vhlch have never bean Identified in Aroclor fluids, but vhlch ean and do occur In process streams. little has baen published In the lleersture concerning tha Isolation and characterisation of incidentally generated ehlorebiphenyls, perhaps due to the fact that incidental chloroblphenyl eongeners are present In process streams at levele below 10 ppm and, In many cases, lass than 1 ppm. Aroclor character isations, on the other hand, have dealt vlth congeners present at the percent level with no matrix. The techniques required for the concentration, resolution snd Isolation of the 26 homolog Isomers shown earlier are typical of the sophisticated work effort required to charaetarise Incidental ehlorebiphenyls In some process streams. 142 HONS 026183 First, vacuus fractional distillation of kilogram quantities of process temple la required to concentrate the chloroblphenyls. Second, the saspla matrix Is separated froa the chloroblphcnyle by first a aacro-scale absorbent coluem using alualna, silica gel, floriail or a combination of these, and then after concentration, a micro-polishing cleanup la needed. Rlgb pressure liquid chromatography la then used to separate and isolate the Indlvlduel chlorobiphenyl congeners for subsequent identification by Infrared and nuclear magnetic resonance spectroscopy. Individual chlorobiphenyl isomer resolution for all meabers of a hoaolog class cannot generally be obtained, even on chromatographic systems with more than 30,000 theoretical plates. Therefore, spectroscopic confirmation is often difficult in the absence of standards. The problea of specific ItoMr identification has relevance to both the analysis of chlorobiphenyl In environmental asaples (where the contamination may have occurred froa Aroclor or similar type fluids) as well as the analysis of chemical process streams, Patten recognition methods are based on the Identification of a specific Aroclor fluid from the "pattern" of congeners observed, and then analysing the environmental sample as If It contained such a fluid. However, it haa been shown that the environment changes the pattern, that the toxicity of specific leomeTS varies, and that such ahort-sut methods do not show the presence or absence of the toxic components (Albro st el. 1979). Vhils the Aroclor fluids have been quite thoroughly characterised, the current state-of-the-art Is not sufficiently developed to permit positive identification of the possible 209 eongeners which might be present In e chemical procaas stream at the low ppm levels. Systematically, this can be attributed to the fact that only eighty chlorobiphenyl congeners are available commercially (Ultra Scientific, Inc.). In addition, the work effort required to separate, Isolate end Identify the quantities of chloroblphenyls present In chemical process streams is orders of magnitude more complex chan Aroclor characterisations. Reasonable scientific evidence, however, Is developing for approximately 135 congener identifications based on the co-chromatography approach of Albro. Succinctly stated, given the current state-of-the-art, the quantity of ehloroblphenyls cannot ba determined unlees the congeners present can be Identified, and a known standard for each congener exists to calibrate the instrument. Analytical Methods for Chlorobiphenyl Detection, Identification end Quantitation ere Limited All applicable methods ere based on some form of Instrumental analysis. Instrumental methods eapeble of detecting chloroblphenyls are also applicable to their quantitative measurement with varying degree* of praclalon and accuracy. Current known Instrumental method* with eom* degree of applicability to detection of chloroblphenyls ares 143 hohs Ultraviolet Spectroscopy; Infrared Spectroscopy; Nuclesr Magnetic Resonance Spectroscopy; High Pressure Liquid Chromatography; Gas Chromatography; and Haaa Spectrometry. Ultraviolet. lofrared, and nuclaar magnetic rtaonanct apactroaeopy lack tha aalacclvlty aod sensitivity to measure chloroblphenyl In tha parts-per- mllllon ranga of lntaraat In most chemical proeaaa streams. Some atudlaa hava baan aada on high praaaura liquid chromatography methods, but tha ptlaary , llaltatlon la tha lack of a datactor which la aalaetlva, and nondlaerlalnatlng for chloroblphanyla. Only gaa chroaatography and aaaa apactroaatry hava daaonatratad aufflclant aalactlvlty and datactlon lanaitivity to ba applicable to chloroblphanyl aaatutaaant In ppa concentration ranges. Analytical aathod development for chlorinated biphenyl analyela auat ba cognltanc of criteria other than tha heat aultad Instrumental technique. Analytical methoda hava two typea of eharaeterletlce (Bandal at al. 1961) - scientific and practical. Tha aclentlflc eharactarlatlca datamlna tha reliability of tha analytical data: tha practical eharaetarlatlea determine tha utility of tha method. The aclentlflc attrlbutaa of a method Include auch thlnga aa accuracy, precision, epaclflclty, and limit of rallable meaeurement; tha practical attrlbutaa Include coat of performance, time required, and level or training needed. For raaeareh purpoaet, the practical aapecta are of tecondary cone deration; for regulatory operatlona of compliance and eurvelllance. practicality la of great Importance. Llttla enforcement la poealble ualng a method which turna out one analytical value per dayl Thla la one area where lnduatrlal analytical ehealstry and regulatory lurvelllanee hae a atrong common Interest. Industry demands rapid, accurate analytical number* to control proceasea and approve shipment* of product. Regulatory lurvelllauce la considerably more rallable and effective If the data can be generated ualng a tlmple. rapid analytical method. Quantitative Analyal* of Chloroblphanyla In Froca** Stream* (LOP and LOQ) Given the datactlon and Identification of one or more chloroblphanyl Isomer*, on* eaa than proceed to measure tha amount present. Instrumental aethoda applicable to detection are generally applicable to quantitation of the aubstanc*. except that a new aerie* of technical problems Is then encouotarad. Quantitation must deal with the limit of detection (LOQ) of the detector, changes In the sensitivity of the detection mechanism to changes In concentration of a chloroblphanyl congener (linearity), variation In detector response beeveen on* chloroblphenyl and anochar (discrimination), variation In the analysis as performad from day to day, (precision) as well a* the Influence of the sample matrix on detector response Interference, recovery). Certain terms, Limit of Detection (LOD) and Limit of Quantitation (LOQ), need to be defined In order to establish a common basis for discussion of these Issues of detectability and quantitation. 144 hqns Each detector generates a background electrical output algnal which la tha aua of electronic and chemical nolaa. Thla background algnal la not conatant, but varlaa slightly ovar time. Tha alactrlcal algnal ganaratad by a aubitanc* to ba measured must ba graatar than thla variation In background algnal by a apeelflad amount In ordar to ba aura that vhat la obaarvad la truly a measurement of a aubatanca and not a random variation in tha back ground algnal. Tha LOD la tharafora daflnad aa a algnal which la at laaat thraa times tha background algnal-to-nolaa ratio. Tha aacond tarn to ba daflnad la tha LOQ. Tha LOQ la daflnad aa tha quantity of a aubatanca which auat ba praaantad to a dataetor In the praaanca of a aaapla matrix to achlava a algnal which la tan tlaaa tba dataetor algnal- to-nolaa ratio. Thla coneantratlon rapraaanta an amount which ean ba rallably aaaaurad and raproducad If tha aaaplaa wata analyaod by aany different laboratoriea. Tha LOQ definition atataa that elgnale lata than thla magnitude cannot rallably ba uaad to maaaura tha quantity of a aubatanca, nueh laaa regulata It. Tha importance of thaaa tarma cannot ba overemphaalxed In esta- bllahlng not only creditable analytical methods, but alao creditable regula tory llrnlta. .. Let'a examine for a minute what lnduatrlal analytical chemlete have bean trying to attain ,ln tarma of LOQ over tha paat 2-1/2 yaara to enanra regu latory compliance at 50 ppm total Incidental TCI. We muat maaaura the praaanca of 209 individual ehemleal apeclea and determine 11 tha aum ef thaaa aeasurementa la laaa than SO ppm. Tba LOQ tharafora for each congener muat ba SO ppm divided by tha potential 209 conganara or 0.24 ppm par congener. Tha limit of dateetloo then muat ba 0.24 ppm divided by S.33 or 72 ppb par congener to have a creditable analytical method. Thla la preaantly unattainable. In tha paat, aoclo-polltical and/or toxleologleal conaldaratlona have aometinea lad to a perceived need to have regulatory llmlte that are laaa than tha currant LCD. Thla oceura whan exceedingly email eoncantratlona are coneldered to be dangerous, constituting considerable risk to human health. In such a situation, the regulatory limit la usually sat at tha detection Halt, aa hae bean dona aa an out-growth of tha wall-known Delaney amendment. Hovevar, there are powerful scientific raasona for aattlng a regulatory limit not only above xaro, but also "reasonably" far above tha detection limit, mat, It la Impossible to prove scientifically tha absence of any given ehemleal spaclas. Tha bast that can ba accomplished by the measurement pro cess la to establish a limit below which tha analytical armr la unacceptably large. Second, apart from the measurement limitation, It Is Impossible In principle to achieve a state of complete purity for any sample. Third, It la extremely difficult to make reliable measurements, and especially so at the trace level. Fourth, one should note that analytical methods have been improving (and vlll undoubtedly continue to improve) In selectivity and In reaching lower limits of detection so that a concentration that is below tha limit today may be above It tomorrow. Hence, when tha regulatory limit la set equal to, or Just above, a current limit of detection, the timing of the analyses relative to that of regulatory enforcement introduces yat another uncertainty. Tharafora, one auat recognise the compelling scientific reasons to establish a regulatory limit above the detection limit and to define 14S ttQNS acceptable analytical tlaka (l.a. , tha documentation of falsa positive/ negatives, vail aa "statistical" errors). Electron Capture Pttactora Although gaa chromatography In conjunction with alaetron capture detection la the moat widely accepted technique In uaa today for the determination of polychlorinated blphenyla, It la a technique which la sore applicable to the determination of Aroclor fluid type naterlala than to thoaa chloroblphenyl* which are likely to form In various typaa of chemical proceaaea. The predict able leoner patterne of the different Aroclor flulda or their mixtures makea recognition of their pretence In a aampla relatively aaay. In addition, ehe varloua Aroclor flulda are atlll readily available for uaa aa atandarde and detectable coneentretlona aa low aa 2 v/g/L (or approximately 2 ppb) have bean reported (Delflno and Eatty 1979, Willlama end Benoit 1979) for auch dlvarae aaaplea aa paper mill effluenta and houaehold product*, loveat detectable coneentretlona In tamplea auch aa tranaformer oil* (Ogata at al. 1977) appear to be 1 ppm. Beeldee being sensitive to ehloroblphenyla, the electron capture detector la alto eenaltlve to chloronaphthalenea, chloroterphenyle and chlorophenylcthera aa well a* aulfur compound*, all of which may elute with the varloua Aroclor groupe (tap11a and Ave 1977, Armour and Burke 1971), While thla preaentt problem* In the analyalt of tamplea for Aroclor compound*, It la a auch more terlout problem In the analyala of chemical proceaa atream aaaplea for ehloro blphenyla alnce thoaa formed In any given chemical proceaa do not exhibit the pattern of laomer dlatrlbutlon profile* ahown by the varloua Aroclor fluid*. The capillary chromatogram at the far left of Figure 2 represent* 'tha equivalent of 1 ng of Aroclor 1212 In hexane which, relative to a sample preparation scheme, could equate to 50 ppm. The Identification, If thla were a temple extract, would be unequivocal. The middle chromatogram repreaent* three ng'e, 1 og each, of the three monochloroblphenyl leoner*, and the chromatogram at the far right repreaanca 1 ng or the equivalent of 50 ppm, of 3 chloroblphenyl. This effect of losing the chromatographic profile of FCB nakaa chloroblphenyl identification and quantitation very difficult using electron capture detection and consequently, the llmlta of detection for Incidental FCB moaaurement are tvo or three orders of magnitude higher. In addition, there la a high probability that nonchloroblphanyl components la a process stream sample would elute within tha chloroblphenyl retention time region. This makes It virtually Impossible to say, with any degree of eartalnty, which chromatographic peaks represent real chloroblphenyl* and which represent nonchloroblphanyl Interferences. The sensitivity and LOD of the electron capture detector varies over a tremendous range from homolog to homolog and from laomer to Isomer within honolog. The range of the variability 1* Illustrated In Table 2 from Jttjitalnger'a book, "The Chemlaxry of FCB*a." Hot* that the malar response of deeachloroblphenyl Is 7,000 times greater than the response to the same solar concentration of 3ychloroblphenyl. Not* 146 MQNS 02618? Figure 2. Electron capture reaponea of aonochloroblphenyls v. Aroclor 1242. 147 MONS 026188 Tbi 2. Ralatlvt Molar Reaponic* of Elaetron-Capeura Dttactort to Soat Chlorobiphtnyl* Ralatlv* Molar Raaponaa 2342,2'-01 2.4-01 2,6-01 3,3'-01 3,4-Di 4,4'-01 2,4,4'-Ttl 2,2'.4.4'-Ttta 2,2\6,6'-Tacra 3,3\4,4'-T*era J.S'.S.S'-ttera 2,3,4.5-Tttra 2,3,5,6-Tatra 2,2,,4.4,.6.*'-Hm 2l2,.3,3',M''M,>Oeta 2,2'3,3\5,5',6.6'-Oeta Daca 1.00 0.20 1.10 5.16 17.1 32.0 6.10 15.2 5.97 135 106 20.6 396 320 367 259 347 726 1,180 1,150 1,410 148 HONS 026189 l(o the difference In response between the isomers of the tetrachloro homolog, which very hr a factor of 20 for the aaa* molar quantity of chlorobiphenyl, The full range of thla variation la not known since the absence of standards haa prevented research on the Issues The importance of the variation In electron capture response must be clearly understood, for incidental PCS quantitation, each congener in the process sample must be coe^letely chereeterlted, and a standard of the characterised congener must be available in suitable purity for calibration either by purchase or by synthesis. In addition, the chlorinated biphenyl congener Bust be completely separated chronatographlcally before being preaented to the detector or coeluting Isomers say generate large false positives or nagstlves. At the present tlae, only 80 of the possible 209 chlorobiphenyl isotiers are commercially available. (Ultra Scientific, Inc.) Baking the celibratlon procedure almost hopeless because of the diversity of response among the individual laoMra with thla detector. This factor, together with the coelu tion of unkaovn interferences, acts to preclude the usa of the electron capture deteetor for the quantitative analysis of process and vasts stress samples for incidental generated chloroblphenyla. In som specific Instances, it My he possible to eellbrate an electron capture detector-based analysis procedure with a mss spectroMter for e specific process stress. However, it can be expected that both the number and amount of chlorobiphenyl Isomers in a process stress can change as process conditions vsry over else, thus calling Into question the accuracy of the calibration perforMd. Frequent recalibration with a sate spectrosetar would be necessary. Finally, one additional factor Mkes use of the electron capture detector difficult in sott process stress analyses. The deteetor response Is extresely sensitive to chlorine containing compounds, which Is.the basis for is utility in chlorobiphenyl analysis. However, when chlorobiphenyl is formed in process streams, the Mtrlx is usually organo-chlerine compounds. Since the chlorobi phenyl is present In ppm levels, a suitable sample cleanup is generally needed. Perchlorlnatlon se a Chlorobiphenyl Analysis Sample Preparation Procedure Slnee chlorobiphenyl mixtures eontain many chlorobiphenyl laoaers which are difficult to Identify and analyse, a procedure called perchlorlnatlon was developed to analyse low concentrations of Aroelor fluids. The pereblorinatlon procedure reacts the lower chlorinated homologs and isomers with antimony pentachlorlde, SbCl., to form the single compound decachloroblphenyl (KB) (Berg et al. 1972). The amount of KB so formed Is then determined by gas chroMtograpby. This proeedura was developed as a means of lowering the analytical LOD when enalytlug for trace quantities of Aroelor fluids in envlrooMutal samples. In check Instances, the typical Aroelor fluid gas chroMtographlc pattern My be recognisable, but the concentration is too low for accurate quantitation. Conversion of all the chlorobiphenyl IsoMrs Into DCB permits the Masurement of a tingle peak and the use of a known high purity and readily available standard. Perchlorlnatlon does not, however, yield any date regarding the concentration or distribution of specific chloroblphcnyl isomers. 149 HONS 026190 Id applying tha perehlorlnstlon raactlon to tha aoalyala of Aroclor fluid!, varloua raaaatchcra obaarvad that tha aathod could ylald arrooaoua rcaulta (Anaour 197J, Stratton at al. 1979, Trottar and Young 1975). While tha aptelflad raactlon condltlona yielded 90-100 parcant eonvaralona of Aroclor 125A to OCI, eonvaralona of only 30-70 parcant vara obtained for Aroclor 1242. Higher reaction temperatures reaulted In poor recovery of the aono- and dl- substituted loosera due to the apparent lnatablllty of thaaa apaclaa at tha higher temperature. Parehlorlnatlon technlquet In general are not applicable for chloroblphenyle In proceaa streams due to aerloue lnterferencea which generate falae poeltlvee. Poaltlve Interference In the perchlorlnatlon procedure* arlaet from the preaence of biphenyl, naphthalene or chlorinated naphthalene in tha aaaple. Proceaa atraaaa ean contain thaaa and other potential Interfering organic coapounda. biphenyl and naphthalene are for exaaple, heat tranafer aganta used In Industrial heat exchange equipment, and are, therefore, a potential contaminant In proceaa atraaaa (Tindall and Vlnlnger 1980, Burkhard and Araatrong 1981). In addition, polychlorobansanee praaant in aoaa proceaa atraaaa can eondenae In tha preaence of antlaony pentachlorlde to aaka chloroblphanyla which la turn would be chlorinated, reaultlng In falae poeltlvee. Elactroconductlvity Patactor Tha elactroconductlvity (Hall) detector operating in the chloride aode, haa alao bean uaed In the gaa chromatographic analyala of chloroblphanyla, but to a far leaner extent (Sawyer 1978a,b, Mlndnip 1979). In contrast to noraal operation, tha detector furnace met be at taaparaturaa above 820*C (Dolan at al. 1972) or even at high aa 1100*0 (PPG Ind., Inc.) to affect the pyrolysis of chlorinated blphenyla. The detector responda specifically to chlorine containing coapounda and baa a greater degree of selectivity for chlorobiphenyl analysis than tha electron capture detector. In addition, tha response la a direct function of tha halogen eoutant of tha eoapound and la linear. However, the sensitivity In terms of aolar response at these elavaced detector tempera tures Is at least one and possibly two orders of magnitude lower chan the electron capture detector. Suitable sample pretraataant In terae of cleanup aust be Incorporated when this detector la applied to process atraaa analysis. Caa Chromatotraphy/Haaa Spectrometry In Quantitative Chlorobiphenyl Analysis * Aa noted before, gaa chroaatography combined with either electron capture or elactroconductlvity detectors are of little value for the quantitation of trace quantities of unchsraetarlaed chlorinated biphenyls. In both cases, they require standardisation with known compounds or by a principle analytical method such me ges chromatography/mas a spectrometry. RC/MS can be used to identify and quantify individual chlorobiphenyl laomera In a process stream. However, the Identification la Halted to (1) confirmation that an eluting substance is a chlorobiphenyl, and (2) elucidation of tha chlorobiphenyl homologue class. In most cases, tha specific Isomer cannot be Identified. 150 HONS 026191 While total leant fro 100-600 atoale sail uniti are dtalrabla for confirmation during the Initial aaaplt screening, Salactad Ion Monitoring (SIM), alto known aa Single Ion Monitoring or Multiple Ion Detection la the technique Boat frequently employed for quantitation, and offara at leaat an order of aegnltude higher sensitivity than total.acan (Elebalberger at air 1974). two coBproalee technlquea between total acan and SIM have alee been deaerlbed! (Canada and Regular 1976) United Bate acannlng and aaaa chroma tography. Both approaehaa represent modification In computer data collection rates for Ion Intensity measurements, ao chloroblphenyl confirmation or selectivity will not be sacrificed In achieving high sensitivity. In spite of a higher limit of detection as compared to the electron capture detector, SIM la the moat specific method for the identification of a substanee aa being a chloroblphenyl and for the quantitation of chloroblphenyla of unknown congener and homolog distribution. The technique essentially Involves using the mass spectrometer to measure tha integrated ion lntensltlve of Ions characteristic of the species of loterest. For tha chloroblphenyls, the Ions are typically tha molecular weight lone resulting from the naturally occurring chlorine Isotope abundance ratios M and M*2. (Table 3). Thus, for mono through heptachloroblphenyl, the following mass Iona are monitored. For octa and nonachlor, the M*2 and M+4 mass Ions are tha most Intense, and for tha decaehloroblpheayl, tha M+4 and M+6 Iona are used. The intensity ratios for each homolog are used to confirm the eluting component as s chloroblphenyl. The Ions for each homolog can be monitored In groups spanning the retention time windows of the various homologa which are established by standards. The qualitative criteria for confirmation of eluting chlorinated blohenyls have been described by two authors, Tindall and Wlnlnger, (1980) Collard and Irvin (1982) Independently. They are: the chromatographic peak of the characteristic mesa Ions must maximize at the same retention time; tha peak must be In tha proper retention time vlndov; and thirdly, the relative peak intensities of the molecular ions must be within _+15X of the theoretical. (Figure 3) In eaaenca, twenty specific ion chromatograms art gsnartttd in ons 20 mlnuta temperature programmed run. This yields a high degree of specificity. In addition, all of the chlorlnatsd biphenyls form stabls molecular Ions, and the iaotoplc abundance ratios are fixed by nature, Martelll (1981) and coworkers havt evaluated the analytical response of 45 single chlorlnstsd biphenyls using SIM and have found tha following vari ability In response within tha homolog group*. (Table 4) Their study docu ments mass spectrometry as tha least discriminating detector available for use with gas chromatography, and as a result, a single chlorinated biphenyl lcomar can ha used to estimate tha quantity of all other Isomer* within Its homolog cists. Co Hard and Irvin evaluated the relative response of each homolog class at a function of concentration. The response vsa linear in all cases, hut sensitivity decreased by a factor of tan whan comparing mono to dteachloro- 151 HONS 026192 Table 3. Specific Mat* Iona with Rtlatlvt Intnltle Dead for Dataralnini Chlorinated Biphenyls Cl* Oj M (Intensity) M 2 (Intensity) M + 4 (Intensity) M 6 (Intensity) Mooo-Cl^ di-ci2 Trl-Cl3 Tetra-Cl^ Panta-Clj Bexa-Clg Bapta-Cl7 Octe-Clg Nona-Clg Deca-Cljo IBS (100X) 222 (100X) 256 (100X) 290 (77X3 324 (61X) 351 (SIX) 392 (44X) 190 (33X) 224 (6SX) 256 (98X) 292 (100X) 325 (I00X) 360 (100X) 394 (100X) 426 (68X) 462 (77X) 430 (100X) 464 (100X) 496 (100X) 500 (87X) 152 HONS 026193 elJ Ill **t V 1 1001 3 o*i* `i -i* ' - ; _: ^ : ^' /V 5 Cl? A . *-------- - >'M >iM 10:00 11 ilO l)iOO t:)0 1MI 2 ^ ... : W _ )i00 _ >:W 10:00 WiW IttulH l:00 : a. IJ:J0 : Fl|ur* 3. Coapotlta and Individual aaaa Ion ehroaotograaa of aavaral chlorlnatad blphanyl laoaari. 153 HONS 026194 Tabla 4, Molecular Ion Keaponae of Chlorinated Blphenyla to Elactron Impact Mara Spectroaetry* Boaolog No. of poaalbla etructuree No. of analytad atrueturaa Avirt|i rapontt Reaponae variability Mono D1 Tri Tatra Pasta Beta 3 12 24 42 48 42 3 1.050 4.42 10 1.738 17.2 9 1.334 12.4 11 1.577 20.7 3 1.005 0.7 7 1.153 10.8 *Martam, CP, Caatalll, MC, Fanalll, R. 1981. lo> aadlcal'Maes Spactroaatry, 8, 8, 347-350. 154 0261-95 MONS biphenyl. (Figure 4) As show to Table 5, recoveries performed In replicate at tha tvo, tao and thirty ng lavala had a rang# of 95 to 103Z. Tab la i praaanta a praclaloo study on ton rapltcata velghlogs demonstrated a relative arror of *201 at tha 95X eonfldanca level. Tha Halt of datactloo io cha complex chlorlnatad hydrocarbon matrix vat observed aa 1.0 i| of i alnfla 1sonar injected on-column which vaa equivalent to 5 ppn par laenar in tha original aanpla. Tindall and tflnlngar hava raportad conparabla eenaltlvitlea In tha abaanca of a natrlx. CS/MS/SIM allnlnataa many lntarfarsneaa which occur with alaetron capture or electroconductivity datactors Insofar as only compounds w^lch heve common masa fragmanta and aluta In tha retention time window ara potential Interfarants. The technique has a special utility lo cases of Incidental ganaratlon In chemical processes since It does not depend upon operator "pattern recogni tion" to Identify the chloroblphenyls. Ivan though SIM does net detect aa small amounts aa does electron capture or eleetroconductivlty/gaa chroma tography, tha increase In algmal-eo-nolae (S/N) ratio resulting from tha Increased specificity offaats tha decrease la absolute signal strangth when analysing process straaaa for chloroblphenyls. Sample Preparation and Clean-up Techniques Sample preparation schemes for the analysis of ehloroblpbenyla in chemical process strasms often require considerable development effort. Many materials are not amenable to direct chromatographic analysis. Crystalline aollds. polymerle materials, tart, snd corrosive liquids are exea^les. Samples such at these ara generally dissolved In a suitable solvent and analysed directly after appropriate sample cleanup or extraction of the chloroblphenyls Into a more suitable solvent for chromatographic analysis. Direct extraction of aollds with solvents has been used for pigments, biological samples and other materials, but liquid-solid extractions are often nonquantltatlvs for encapsulated chloroblphenyls, except In cases of simple adsorption of the chloroblphenyls on the solid surface. The problem of extracting the chloroblphenyls from tha matrix can consti tute tha major analytical problem in determining chloroblphenyls aloes many of the matrices ara similar to the ehloroblphaeylt In physical, chemical and chromatographic properties. Many methods are employed to remove lneerfaraoces from tha aanpla matrix. Solvent extraction or partitioning with strong add, usually sulfuric, Is successful when the lnearferents are ehemlcally dissimilar to tha chloroblphenyls, but can lead to problems with recovery of the lower chlorinated homologs (Dry Color Manufacturers Association, linear 1973). Figure 3 shows tha effect of a sulfuric add cleanup on uonoehloroblpbenyls. A hexane solution containing 10 ppm mono and 1000 ppm Aroclor 1242 was separated by capillary gas chromatography with flame Ionisation datactlon. After shaking the eolutlon for three hours with concentrated aulfurle aeld, tha monochloroblphanyls hava decreased significantly while the Aroclor fluid rename unchanged. After eighteen hours, the monoe are completely absent with only minor changes In the Aroclor composition (Figure d). This represents an MCi desttwetien rate of appcenlmacely 201 per hour at room temperature. 155 MQNS 026196 tXfHS f*fun |*t oiu'cMdiw lh(<i|N twtfrMufufclpt'AAylf CC/HI (fonM lr hi^ticMei*- lhtM|k dscg'VlorofclpfitAy I uo> Njnw|fda l|crteJ Figure 4. Linearity of response for selected aono-decachloroblpltenyl congeners using CC/HS/SIM. MONS 026197 157 Table 5. Chlorinated Biphenyl Recoveries froe Spiked Chlorinated Hydrocarbon Hetrie Begree of chlorination Amount added (ng) Run II Amount found .()* "StJ7 dev. (ng) Amount added (ng) Run It Amount found . K.(ng)* Std. dev. (ng) Amount added (ng) Run M Amount found *.("g)C Std. dev. (ng) He an percent recovery 1 2.0 1.9 0.03 10.0 9.2 0.2 20.0 19.0 1.1 96 l 5.0 3.0 0,00 15.0 14.0 0.4 30.0 29.1 1.3 99 3 4.0 4.1 0.09 20.0 19.7 0.7 40.0 39.1 l.t 99 4 3.0 2.0 0.21 15.0 15.3 0.9 30.0 27.9 0.9 96 5 2.0 1.0 0.12 10.0 10.0 0.3 20.0 11.4 0.9 93 6 3.0 2.0 0.15 15.0 15.2 0.0 30.0 29.9 1.0 97 7 1.0 1.0 0.15 5.0 3.2 0.2 20.0 9.9 0.3 10) 2.0 2.0 0.00 10.0 10.3 0.2 20.0 20.0 0.2 100 9 1.0 1.0 0.06 5.0 3.1 0.3 10.0 10.1 0.2 99 10 1.0 0.9 0.10 5.0 5.0 0.3 10.0 9.7 0.1 96 Naan value of 5 determinations. "Mean value of ) determinations. cMean value o( 2 determine!lone. MONS 026198 Bagrat of chlorination Tabic i. Relative Precision of Chlorinated Biphenyl Haeauraoanta (Baaple 2) Hanograaa Run nuaiber 15- Naan (>) RelatIvt Ptaclalon (952) 1 6.6 7 2 6.9 7.0 7.5 7 6 8 3 7.1 7 7 7.1 142 2 37.) 36.3 36 0 37.4 15.9 31.4 32 2 14 1 31.0 32 3 34.6 162 1 43.3 47.7 46 46.2 45.a 19.3 19 0 42 a 39.4 41 8 41.4 152 4 24.a 23.4 25 25.9 25.7 21.2 21 1 21 3 21.3 23 2 21.8 172 5 12.a 12.4 12 11.4 11.0 10.8 10 a II 7 10.7 II 1 11.9 182 4 a.4 3.4 9 9.6 9.8 7.a 7 9 a 1 7.8 a 1 a.a 192 2 4.0 6.0 3 5.9 6.4 4.3 4 9 4 9 4.8 4 a 3.4 252 4.6 3.3 3 5.5 3.6 4.1 4 1 4 6 4.6 4 6 4.8 192 * 3.3 3.7 5 5.9 3.a 4.3 4 4 4 8 4.7 4 9 3.2 242 10 5.7 3.9 3 6.0 6.0 4.9 3 1 3 7 3.1 5 3.6 142 MONS 026199 * r Vm Tljulrs 3; Monoehloroblphnnyl in th. princ of CONC sulfuric seid. 159 HONS 026200 TIm (hr.) Figure 6 MCB Otftructlon With HjS04 (CON) at 35C. 160 HONS 026201 In neny ceeee, column ehronatogrephy It required to remove interference*. Alumina and Florleil art commonly uaed at adaorbanta. The development of claaoup procadurta uelng column chromatography ara laborioua, tima conaumlng, and often, add considerable variability to tha analyala. Maintaining a eonatant activity for tha adaorbant and reproducing it la extremely difficult (Buttlnger at al. 1974) and la often the cauee of poor lotarlaboratory repro ducibility. In addition, the eelectlvlty and loading capacity of a given aorbent aeparatlon muat be meticulously evaluated for tha lower chlorinated homologa aa vail at tha higher onea. Thia haa not been the caae with many of the cleanup technlquea which have been daaerlbed In the literature for Aroclor quantitation. Figure 7 ahowa the eluting eharecterletlca of eome of the lower chlorinated homologa on baalc chromatographic alumina ualng baxana aa the eluant. The ratio betveen chloroblphenyl congener concentration and weight of alumina la 500/1 (alumina/congener), eo aorbent capacity la not exceeded. Thlt cleanup technique aa deecrlbed In the literature atatee that ell PC> elute In the flrat 5-6 bed volumea, where in feet, only SOX of the 2,6-dichloro, 25X of the 2,4-dlcbloro end none of the 3- end 4-ehloroblphenyl heve eluted. Thete two ereee of temple pretreetment, minerei ecld dlgeetlou end edeorbent column cleanup, reptetent ereee of publlehed PCI enelytlcel methodology which muet be reevelueted before they cem be applied to Ineldentel chlorlneted biphenyl enelyele. A multldlmenelonel cbrometogrephle technique utilizing both high performence preperetlve ecele liquid ehrometogrephy end ceplllery column gee chrometogrepby haa bean davalopad raeantly for tha datarmlnatlon of polychlorlnatad blphenyla In transformer coolent end motor olle by Cheeler et el. (1982). While thlt technique le not directly eppllcable to the datarmlnetlon of chloroblphenyle In proceee atreeme, It affords e higher degree of resolution then the conventional gravity flov edeorbent cleanup technlquea without aecrlflclng column capacity. In addition, euch syeteme ebould be much fetter cleanupe and readily lend themeelvet to automation. Tha final claanup to ramova sample matrix lntarfareneaa In almoet all chloroblphenyl anelyels schemes la the selection of the appropriate gee chromatographic column. Often the type of complexity of tha matrix will dictate whether packed column or capillary columns should be employed. The efficiency or eeperatlng power of capillary columns la sometimes not aa desirable as the selectivity which can be obtained by polar phaeee or mixed phaete ualng leomer selective modifier* euch ae tba Bentonite day*. Three end four chloroblphenyl ere extremely difficult ltomert to separata on most conventional packed columns and only slightly batter on eha bonded SE-30 and SE-S4 methyl silicone fused allies capillary columns. Their resolution is Imperative when ualng an electron capture detector since the 4-ehloroblphenyl leomer le 6 times more responsive then the 3-chloro leomer. Binary liquid phase optimisation euch ae deecrlbed by Laub and Purnell (H76>-however, ell-ow for the systematic design.of eeparatlons to resolve leomer* a* well ae matrix interference*. 161 MONS 026202 Figure 7. l.owcr chlorinated biphenyl elution character let Ice from alumina with hexane. Q26203 mons Picked or capillary column (election for the analyele of PCBe et Aroclora le really academic alnce the pattern recognition can easily be accomplished uelng cither (operation mode. For Incidental chloroblphanyl analytic, the trend obviously oeeda to be tovarda more reaolvlng povar (capillary columns) bacauee of the lover LODa required par eluting eongenet. Thle trend vlll become more apparent aa good fuaad alllca capillary columna vlth polar bonded phaaea become available. Quality Aaaurance In Qualitative and Quantitative Analyele Quality aaaurance la an eaaentlal but often neglected aapact of ehloroblphanyl analyala (MacDougall and Crummett I960). The flrat aapact of a quality aaaurance program la the development and eonalatant uaa of a fully validated method. Validation of a method raqulraa a atudy of the follovlng characterlatlca. Flrat. the linearity of reaponae for the analyte In queatlon ahould be evaluated over the dealred quantitation range. Second, aufflelent replicator of aamplea which had prevloualy beeo spiked ahould be analyaad to determine the recovery and accuracy of the method over the dealred quantita tion range-. Thla lncludee adding the analyte to the aample matrix before any aample preparation or cleanup la performed. Thirdly, a preclaion atudy on a minimum of ten replicate velghlnge analyaad on tvo different daye ahould be performed. Finally, any known'lnterferencee ahould be documented aa to their effect on the quantitation of the dealred analyte by running aufflelent aample blanke. Once a validated method haa been eetabllehed, eonalatant uaa policy. In addition to an on-golng quality aaaurance program la Important, Thla la necaaaary to document that the procedure la performing at the prevloualy determined level. Validated analytical meehodology and active quality aaaurance programa are prarequlaltea to generating creditable analytical data. For the aoalyale of ehloroblphenyla in proceee etreame, a aingle analytical method haa yet to be agreed upon, however, gaa chrosatography/maaa apectrometry with auitable aample cleanup appeara to be the only viable alternative eonalderlng the great variability of sample matrices encountered in chemical procaaa streams. The Dearea of Dncartainty Associated with Incidental FC Analysis - A Round Robin To evaluate the uncertainty In the quantitation of Incidentally generated ehloroblphenyla, a recent round robin analysis of five samples vas sponsored by the Chemical Manufacturers Association (1982). The samples were chosen to typify process stream matrices that are presently regulated by the Incidental FC> Manufacturing Rule. The following participants were asked to generate tbelr best effort using analytical methodology they have developed for the measurement of uon-Aroelor PCIs; Environmental Monitoring and Support Laboratory Office of Research and Development Environmental Protection Agency Cincinnati, Ohio E.I. DuPont DeNenoura A Company Vllmlngton, Delaware PPC Industries Barberton, Ohio HONS 026204 Surveillance and Analyala Division Region 7 Envlronaental Protection Agency Kansas City, Kansas Dow Chealcal Coapany Midland, Michigan Konaanto Coapany St. Loula, Mlaaourl Stauffer Chemical Coapany Dobbs Ferry, North Carolina Vulcan Mattrlala Coapany Wichita, Kansas Thla affort vaa designed to docuaaot tha praaant etate-of-the-art for ehloroblphanyl analyala whan a coapany or Independent lab la contracted to analyte aaaple for regulatory coapllanca. Each aaaple vaa deacrlbed aa containing ona or aora ehloroblphanyl conganara ranging froa mono to decachloroblphenyl. . Total congener eoncantratlona vara deacrlbed aa ranging froa 5 ppa to lean than 500 ppa in the follovlng aatrlcaa: A - Chlorinated lantene Watte I - Mixture of Chlorinated Bentenea C - Mixture of Chlorinated Bentenea 0 - Chlorinated Aliphatic Watte Stream E - Chlorinated Aroaatle Staple Each aaaple vee to be analyted In triplicate vlth eaparate portlona of the aaaple being taken for each replicate analyala. The Inetruaental techniques and aathoda of calibration vara utlllted for the analyala of tha five round robin aaaplee are ahovu In Table 7. Eight of the can laboratorlea elected to ute CC/MS. Seven of tbaae eight labe uaed packed coluana vhlle one utlllted an SE-30 fuaed alllca capillary column. One of the laba utlllted packed coluan GC vlth electron capture detection, vhlle another uaed packed coluan/flaae iooltatlon detection. The flame lonltatlon detector data vae deleted froa the coapotlte tunaary for statistical reasons. Sumsary of the data across sll nine laboratories Is shovn In Table 8. For aaaple A, the aeen concentration ves 283 ppa. The range of the reported aean concentrations ves froa a lov of 10A ppa to a high of 413 ppa for a length of 309. The everage coefficient of verlaelon among repllcaces vlthln laboratories ves M1.4Z relative. The coefficient of varletlon aaong labora tories vaa *321 and the expected coefficient of variation for a aaaple sent to a random laboratory vlth a single replicate concentration reported vould be expected to be 34S, For aaaple B, the range vae 70 ppa, vlth e coefficient of variation of 922 for a single replleate. For C, the range ves 105 ppa, vlth e coefficient of variation of 46X. For D, the range vas 261 ppa, and 722 and for E, the values vere 12.9 ppa vlth 532 as the coefficient of variation for a single replicate. 164 MONS 026205 Table 7. Round Robin Methodology for Incidental PCS Analytic Laboratory A B C D E r G H I J Technique GC/EC GC/MS CC/HS (Semlquantltatlve Single Replicate) GC/FID GC/MS GC/MS GC/MS GC/MS Capillary GC/MS GC/MS Calibration 18 Congenera 24 Congenera 10 Congenera (l/Hoaolog) 10 Congenera (1 Homolog) 25 Congenera Aroclor Mix Plua Mono 20 Congenera 10 Congenera (Internal Scd) 10.Congenera 24 Congenera 165 MONS 026206 Table 8. Susaary of Exparlaantal Raaults for *11 Honologa Coablnad for Round Robin Experlasnt Across Kino Laboratories Statistic A9 CD E Moan Ranga Endpoint Length CA C. cc 283 2* 10* - *13 309 11.40Z 32Z 34Z 1 - 71 70 15Z 83Z 92Z 6* 130 9.2 7 - 112 33 - 294 0.3 - 13.2 10$ 261 12.9 8Z 17Z 28Z 45Z 64Z 44Z 46Z 72Z 53Z Co--nta Hoan Maas FC> concootratlon aeroaa all laboratorlaa Range Range of RCR aaan eoncantraclona across oil laboratorlaa C. Average eoafficiant of variation aaong rtpllcatos vlthln laboratorlaa C. Coafflclant of variation aaong laboratorlaa with lntra- 0 laboratory variation removed C- Coafflclant of variation aaong PC> eoncantrationa for a C aaapla sant to a randoa laboratory with a slngla rapllcata coneontratlon raault raportad 166 HONS 02620? The ability to recover nd measure s quantity of analyte added to a ample demonstrates the accuracy of a nethod. Simple C la Identical In matrix to sample B, except the chloroblphenyl congeners were added (Table 9). The data, as reported by laboratory for these two samples, la ehovo In the table. The values found for sample C In ppa are shown In the first line from left to right. The second line represents the data for ssmple I, The difference between B and C reprasenta the amount found, aod the last Una represents the recovery baaed on the addition of 6* ppa of the 22 congeners. The data from laboratory t arrived too late to be a part of the original aanuscrlpt, but Is Included here. The reproducibility of the recovery for eight of the ten labs Is remarkable. The ability to quantitate levele of chloroblphenyls, however. Is very random. This Implies each analyst has a high degree of discipline to measure chloroblphenyl In matrices he's familiar with relative to his method. However, the sbsence of an agreed upon method or convention for the selective chloroblphenyl detection causes vldely divergent concentration to be reported. Figure 8 gives a graphical representation of the data reported for samples A and scross all labs is shown here. It appears that no single lsboratory has the ability to analyse all matrices veil; however, for each sample two or three of the labs generate data which Is In reasonable agree ment, Indicating soma coaeon familiarity with the matrix and types of Inter* farences. The data for samples C, and D are shown In Figure 9. The relative error contribution as a function of homoleg (Table 10) was evaluated as follows. For sample A, the calculated mean, the lovest value, the highest value and the resulting length of the reported data by homologue Is shown In the first five columns. The average standard deviation, expressed In ppm, to be expected for a sample sent to a random laboratory and analysed In a single replicate Is shown In column six. The calculated relative error, expressed In percent at the 95Z confidence level of each homolog'a contribution to the overall sample analysis Is shown In column seven. Tha relative error and the range as a function of homolog for each sample are shown graphically on the following figures. Fot sample A, tha relative error at the 95Z confidence level was 68Z for all homologs, 30Z for the mono, and an average of 10Z or the other homologs (Figure 10). For sample B, the relative error for all homologs was 1842 mono was 133*, and tha others averaged 28Z (Figure 11). For C, 92Z for all homo logs, 502 for mono, snd so average of 16Z or tha others (Figure 12). For D, 1*42 or all homologs, 103Z for mono, and an average of 17Z for tha others (Figure 13). For sample B, 1062 for all homologs; 65Z for mono, 43Z or dl, and 22Z or trl (Figure 1*). The matrix contribution for these types of samples affaets the ability to quantitate tha lower chlorinated homologs, especially mono the most. Other types of matrices may affect quantitation of some of the higher homologs In tha absence of suitable cleanups. The objective of the round robin was explanatory In nature and designed to determine the degree of uncertainty associated with Incidental FCB analysis. The data show that the magnitude of these uncertainties are highly algolfic*"t. Tha Impact of requesting a sample analysis for Incidental FCB froa a random laboratory could produce measurements far from tha true value, 167 MQNS 026208 Tabia 9. SuMtry of Chlorobiphtnyl Splka Saapla Laboratory A B c E t G H 1 J l Kean C 66 69 7 83 43 63 112 30 80 61 8 ii H 35 JS n n 39 il 10 Aaount Found - 30 52 - 30 39 51 41 11 59 51 45 J2 48 X Raeovary 78X SIX 78X 61X 80Z 64Z 17Z 92Z OX (Baaad on 64 ppa Addad) 7SZ 168 MONS 0Z6Z09 Display of data for sample A all homolot* combined Display of data for sample B all homologs combined Each oolnt is a PCS eoncantratlon for a replicate at a laboratory. Figure 6. Chlorliuted biphenyl concentrations as reported by laboratory for samples A and B of Che Bound Robin Analysis. 169 HONS 026210 hoaoloi* coablned hoaoloft coablned Each peine in PCI concentration for replicate at a laboratory. hoaoloi* coabined Figure 9. Chlorinated biphenyl concentration* a* reported by laboratory for tuple* C, D and of the Round Robin Analyst*. 170 MQNS 026211 Tabla 10. Sunaary Statlatlea for Saaplt "A" by Hoaolog Ranga 0* Relativeb Hoaolog Maan Low High Langth (ppai) Error (5) 1 S2 18 123 2 59 24 98 3 60 33 96 4 53 7 52 5 20 0 37 6 16 0 31 7 18 0 77 8 10 0 35 9 6 0 14 10 11 0 21 Total PCB contact * 284 ppa 105 74 61 46 37 31 77 35 14 21 142 130Z 122 115t 123 1162 118 1132 113 1 92 112 1 82 124 1172 110 1 72 1 4 1 32 1 7 1 52 *Tha avaragt atandard davlatlon (c), axprttatd in pp, to ba axpactad for a aaapla aant to a rando* laboratory and analyzed In a tingla rtplicatt, bTha ralatlv* arror, (2o 100 ), cxpraaaad In percent, iaaple Maan PCS Conttnt of aaeh hoaolog't contribution to tha overall atup la analyait. 171 MONS 026212 Sample "A" 284 PPM PCS Relative error (2) 682 302 All homologuea 152 162 II 23 132 92 1 45 172 J"__II_7LS I32 6 78 9 52 X. 10 Range (PPM) Figure 10. CMA PCS Round Robin Summary - relative error/range aa a (unction of hoaologue - sample "a."- 172 HONS 026213 1841 Saaplt "B" 24 PPM Relative error (X) 1332 33X SOX . 251 .`ill i 6X All 1 hoaologuee 1 ND not detected ND ND 89 ND 10 Figure 11. CMA PCB Round Roblo Suoaary - relative error/range aa a (unction of hoaologue - aaaple "l." 173 MONS 026214 homologucs homologuts Flgura 12. CMA PCB RounJ Robin Summary - relative error/range a function of homologua - sample "C." 17/, MOhis 026215 U4Z F. QW PCB Round Robin Sunaary - ralaciva arror/rangt at a function of hoaologua - tanpla "D." 175 HONS 026216 Range (PPM) Figure 14. CIA PCB Round Robin Sunary - relative arror/range aa a function of hoaologue - sample "E." 176 MOWS 026217 and that tha vita of *uch raaulta for eartlfleatlon of a meouf*cturlng operation, belov specified regulatory cut off (t.a., 50 ppm), could b highly misleading. "Zero Molecule*" or "Tha Ab*nc* of" a* Daflead In a Kegulatofy Contort Prom an analytical aclanca parapactlva. It la Important to focua on a apaclflad regulatory level baaad on rlak aaaaaraant. Tba regulatory level, In effect, daflnaa a goal or ala point for tha analytical chaalat. Only than can ha davalop aathod criteria aueh aa aaapla collection, preparation, cleanup, and lnetmaantal aeaaureaent. Thle raaulta la a aat of ordarly experiments which allova tha aathod criteria to be obtained. Tha reault la a validated analytical aaaaurlng device vlth a quantifiable degree of eartalnty. On tha other hand, "Zaro Holeculea" or "The Abeence of," aa daflnad la a regulatory context are eclentlflcally unquant If labia, and ara atatlatlcally uncaleulabla. McElratb and Bearman (1956) have noted tha similarity In thinking batveen edentleta and lawyer* In dealing with uncertainty. Sciential* axpreia their uncertalntlea In tarma of probabllltla* through atatlatlca. The atandardltad degree* of cartalntla* are 95Z, 99Z, or 99.9Z probabllltlaa. Tha lawyer* use tha following aueh phra*a* a* thalr coroliarla*, "preponderance of evidence," "clear and convincing," and "beyond a raaaonabla doubt" for thalr degree of certainty. Tha Implementation of regulatory level* baaad on "Zaro Molaculaa" or "Abaanca of" will undoubtedly Introduce tha following daacrlptlona of uncertainty! "lack of avldenca," "obacura and eonfualng," and "with conalderabla doubt." EXPOSURE AW> MSZ POTENTIAL kELATING TO INCIDENTAL PCB Data covering worker axpoaura and potential rlak Incidental FOB ha* on tha environment la tha subject of a recently prepared aurvey conducted by the Chemical Manufacture* Aaaoelatlon entitled "The Incidental Manufacture, Proceaalng, Dlatrlbutlon and Dae of PCB at Concentration* Below 50 PPM." (EPA 1981). Tha aurvey represent* over a year of data accumulation from 85 different chemical manufacturing firm* and thua, represent* the currant extant and fate of Incidental PCB. la thle aurvey, 135 chemical proceaaea were deacrlbad which generated a total of 13,800 lb*, of PCB per year. Ninety-five percent or 13,100 lb*, of thla material 1* dlapoeed of a* controlled va*te. Of tha 13,100 lba. dlapoeed of aa controlled waate, 9,100 lb*. (66Z) 1* Incinerated, 1,800 lba. (I3Z) la In a aalt vatar vaah dlachargad to a aalt water rnaarvolr aurfaca water ayatam, 1,600 lba. (12Z) goa* to landfill, and 600 lba. (ZZ) goon to ground Injection. The difference between the 13,800 lbe. of Incidental PCB produced and 13,100 lba. dlapoaed of, repraaont* the net pound* (700) of PCB per year contained la product* manufactured by the chemical Industry. Of thla amount, 500 pound* (75Z) ara encapsulated In solid product* aueh a* paint film* and polymer*, or deatroyed, thua Inhibiting human contact and expoaur* to the environment. No ehloroblphanyla war* reported a* going Into consumer product*. About 490 of the 660 lb*, vara reported to be In products which were listed a* 177 HONS 0262 IS "captive Intermediate*," Intermediate*, or Industrial product*. And finally, only 68 lb*, par yaar vara raportad entering tha environment a* commercial product*. ' In regard to workplace occupational exposure to chlorinated biphenyl*, tha chancae are extremely low, sine* 99X of tha poundage 1* generated and processed In "enclosed process aystaaa" or "controlled ralaaaa procasa eystams," vhere relaaie occur* under carefully controlled condition*. 178 HONS 026219 REFERENCES Albro PW, Corbett JT, Schroedar Jt. 1981. J. of Chromatography 203, 103-111. Albro PW, Baeeman JT, Cleaner TA, Corbatt IJ 1977. J. Chromatography 135, 147. Albro PW, data RJ, Cruamatt WB. 1979. Annals of Maw York Ac adm. Scl. 125-130. Albro PW, Parkar CE. 1979. J. Chromatography 169, 1(1. Armour JA. 1973, J. Aaaoc. Offle. Anal. Cham. 56, 987-993. Armour JA, Burka, JA. 1971. J.A.O.A.C. 54, 173. Bandal K5, Marco CJ, Coldberg 1, Lang ML. 1981. Tha pesticide chamlat and modern toxicology. ACS Sympoalum Sarlaa. Barg OW, Dlooady PL, Raaa CA. 1972. Bull. Environ. Contam. Toxicol. 7, 338-347. Burkhard LP, Armatroog DE. 1981. Anal. Cham. 53, 523-528. Canada DC, Regular PE. 1976. laotopa ratloa aa a charaetarlatlc aalactlon technique for maea chromatography. J. Chromatogr. Scl. 14, 149-154. Chemical Manufacturer* Aaaoelatlon. 1982. Statlatleal analyala of data from a round robin experiment on PCB aamplaa. Submitted to Environmental Protection Agency, March 18. 1982. Opta-62013 and 62014. Chaeler SW, Guenther FR, May VE, Parrla DM. 1982. SBMa for aecurato analyala of PCBa In oil. National Bureau of Standarda (to ha publlabed). Collard RS, Irvin MM. 1982. GC/MS determination of chlorinated blphenyla In complex chlorinated hydrocarbon proceaa and waste streams. The Dov Chemical Company. Dclflno JJ, Eaaty DB. 1979, Anal. Cham. 51, 2235. Dolan JV, Vail EC, Todd TM. 1972. J.A.O.A.C. 55, 537. Dry Color Manufacturers Aaaoelatlon. No date. An analytical procedure for the determination of polychlorinated biphenyls In dry phthaloeyaulna blue, phthalocyanlne green, and diaryiide yellow pigments. Elehelberger JV, larrlt IE, Budde VL. 1974. Application of gaa chronatograpby-masa spectrometry with computer controlled repetitive data acquisition from aelacted specific Iona. Anal. Cham. 46 (2) 227-232. 179 HONS 026220 EPA. 1981. Environmental Protection Agency. A report of a aurvey on th Incidental manufacture, processing, distribution and uae of polychlorinated biphenyl at concentrations below 50 ppm. November 15. Hutelnger 0, Safe S, Zltko V. 1974. The chemistry of PCBs. CRC Frees. Jensen S, Sundstron C. 1974. Aablo 3, 70. Kaplla S, Ave WA. 1977. J. Chroaatogr. Scl. 15, 569. Laub RJ. Purnell JB. 1976. J. Aaer. Cham. Soc. 96 (30), 35. Linear JL. 1973. Folychlorlnated biphenyls! their potential Interference vlth pesticide residue analysis and present analytical status. Frogr. Anal. Chea. 5. 109-131. MacDougall, Cruamett VB. 1960. Analytical Chealstry 52, (14), 2241A. Hartelll CF, Castelll MG, Fanelll R. 1961. Analytical response of single polychlorinated biphenyls analysed by selected Ion monitoring. Honed. Mess Spcctroa. 6 (8), 367-350. McElnath CV, Bearaan JB. 1956. Science 124, 589. Mlndrup X. 1979. MBS Spec. Fubl. (OS). Ogata JN, Okun JO, Hylln JV, Bevenue A. 1977. J. Chroaatogr. 189, 569. Falllttarl BO, Toaer KB, Moseley MA. 1981. Hlgh-resolutlon gas chromatography and negative Ion chemical Ionisation mass spectrometry of polychlorinated Biphenyls. In: Advances lh the identification snd Analysis of orgsnlc pollutants In water. Chap. 12. Keith LH. ed., Vol. 1, Ann Arbor, Ml: Ann Arbor Science Publishers, Inc., pp. 197-218. PPC lnd., Inc. No data. Private eowunlcation. Rogers LB, Cruamett VB, Currie LA, Falconer VS, Libby RA, Rellley CN (deceased), Shults NO, Vestgard JO, Wilson MK. No date. Recosaendatlons for Improving the reliability and acceptability of analytical chemical data used for public purposes. Sawyer LD. 1978a. J.A.O.A.C. 61, 272. Sawyer L2>. 1978b. J.A.O.A.C. 61. 282. Sissons 0, Valtl D. 1971. J. Chromatography 60, 15-37. Stratton CL, Allan JM, Stuart AV. 1979. Bull. Environ. Coutam. Toxicol. 21, 230-237. Tindall GU, Vlnlngcr FB. 1980. Gas chromatography-mass spectrometry method for Identifying and determining polychlorinated biphenyls. J. Chromatogr. 196, 109-119. 180 MQNS 026221 Tlodall CW, Wlnlnger PE. 1980. J. Chroaatogr. 196, 109-119. Trotter VJ, Toung SJV. 1975. J. Aeeoc. Offlc. Anal. Che. 58, 466-468. Ultra Scientific, Inc. Ho dot*. Hope, HI. Analaba, Horth Haven. CT. Vebb KG, McCall AC. 1972. J. Aaaoe. Offlc. Anal. Chee. 55, 746-752. Vllllane DT, Benoit FM. 1979. Bull. Environ. Contaa. Toxicol. 21. 181 HONS 026222 Robert C. Rjley, 11 Monsanto Company St. Louie, MO (3167 DISCUSSION SUMMARY The prepared presentation for thle eeeelon vee entitled Expoeure Studlee Relating to lnduetrlel Process#! Contelnlng IncldentaV PCB. The peper addressed the Incidental generation of chlorinated biphenyl* In Industrial proceeeet end the measurement of the chlorinated biphenyls In process stream* and products. The question of Incidental feneration of chlorinated biphenyls has arisen primarily In reaponae to regulation of Materials containing PCBs at treater than SO ppm. Thesa regulations necessitated the development of analyti cal techniques to identify and quantify low levels of chloroblphenyla In a vide variety of Industrial Matrices. Some of the analytical problems in the area of analysis for incidentally generated chloroblphenyla are common to the determination of PCBs in environaental or aakarel-fluld type matrices. These probleas were addressed in Seselon 1 of this syapoelua. Bowever, many of the probleas are unique to the incidental generation situation. Soae of the areas of particular concern are the vide variety of matrices which are encountered, the avallablllcy of appropriate standards, and the suitability of various analytical techniques for qualitative Identification or confirmation and quantification In the absence of Aroclor** fluid elution patterns, l.e., flcgcrprintc. SUITABILITY Of ANALYTICAL TECHNIQUES The paper presented for this session eaphaslsed the apparent superiority ef electron impact (El) gas chroaatography/Bass spectrometry (CC/MS) for the determination of Incidentally generated chloroblphenyla. Although GC/MS is sore readily available than It was sevaral years ago, electron capture gas chromatography (ECGC) la still the technique aost usually identified with ehloroblphenyi analysis. The bulk ef the literature vhlch mentions chloroblphenyl analysis has used ECGC to detect and measure cocaerciai PCB mixtures in environmental matrices such as water, sediment, or biological tissues. There is also a body of literature vhlch has discussed characterization of coaaerclal alxturas using ECGC with or without other confirmatory techniques. Most ef the methodology In the literature uses the characteristic elution patterns of the various homologs and isomers of ehloroblphenyls as the primary aeana of qualitative Identification. However, chit approach is not applicable to the determination of Incidentally generated chloroblphenyla. because ehey are formed la processes and by mechanisms unrelated to the aanufacturc of the commercial mixtures. In cases of Incidental generation, single Isomers, Isomers of a single hoaolog, or a variety of mixtures of Isomers and hoxologs can be formed at trace levels. The elution pattern of these compounds obtained by ECGC cannot be recognised as chloroblphenyla in most cases. Therefore, mass spectrometry must be used to Identify the efilorobiphenyU vhieh are present. Aroclor is a registered trademark of Monsanto Company, Sc. Louis, MO 63167. 182 MQNS 026223 An important question which la yst to b completely resolved la whether packed CC coluan* or capillary GC coluana arc the beat for the analyses In queatlon. (In general, the tarn capillary coluan la understood to Bean fused allies capillary coluan.) The presentation and discussions In Session 1 ewphaslted the advantages of capillary gas ehroaatography with respect to resolution of the chloroblphanyl lsoaera froa one another and the separation of chlorobiphenyls froa potential Inter!erencea. In fact, soae analytical chemists fael that packed CC coluana should be rejectad coapletely, Tha ability to attain upwards of 2 parts In 10,000 reproducibility of relative retention time data with capillary coluana la a strong arguaent for qualitative Identifica tion of chloroblphenyls, whether as single lsoaera or as complex alzturaa. The coablnatlon of capillary GC with asst spectroaetry la certainly the aost powerful technique applicable to tha detection of ehloroblpbcnyls. Tha other aids of tha arguaent is that, although the separating power and qualitative aspects of capillary GC are Irrefutable, there Is a paucity of validated aethodology for tha detarainatlon of chlorinated biphenyls using capillary eoluana. Since an laportant aspect of the Incidental PCB problem Is regulatory compliance, the availability and applicability of validated method ology oust be considered. The accuracy and precision of capillary GC aathods for PCBs have not been widely addressed In tha literature, especially for cases utilising tha Bass spactroaater as the detector. In the particular case of generating analytical results for coapllence purposes, the question la not so much whether to use packed or capillary CC coluans, but how to ganerata high quality analytical results.' Method valida tion involves all of the steps In a procedure, froa saaple collection through analyte ncasureaent. There Is no question that high quality data can be generated with capillary GC/MS. In aoae cases results of high quality also can be generated by methods utilizing packed eoluan GC/MS. Validation results art available for many of theta methods. In tha final assettaent, the ques tion of data quality must be demonstrated by validated analytical methodology, and aubatantlated by a thorough quality assurance and quality control program, whether the method ueee capillary or packed GC columns. MATHIS COKSIDERAIIOHS AMD COWCEHCT DISTRIBUTION One of the most serious problems in the analysis for Incidentally generated chloroblphenyla Is .the vide variety of matrices which la encountered. In general, chlorinated solvents or any ebaaleal systems vbleh contain an organic compound and a source of chlorine hive the potential to generate traea levels of PCBs. Certainly this statement Is an oversimplification and most such processes probably do net generate PCBs, but the potential Is there. The matrix or particular proceas also determines the distribution of the PCBs that could be generated. Per example, one might expect that only lover chlorinated eonganets could bo Incidentally ganatttad In a precast Involving a ncuochlorlnatad sronttlc compound. Hovnvnr, In the ease of pbthalocyanlne blue pigments, pants- and btxaehlorobiphenyl* era generated iron tha triehlorobansona vhleh la uatd at a solvent. Furthar chlorination of this ayataa to product phthalncyanlna green, pigment generates exclusively dteachloroblpbenyl. 183 HONS 026224 To large extant, the matrix lo dictate* tha elean-up methodology that mutt bt u**d to permit accurate and preelie aeaturamanc of the PCB* chat arc prtient. In the caac of relatively clean solvent ayatema, a simple "dilute end shoot" procedure often suffice*. However, In the cate of the phthelocyanlne pigment*, an exhaustive eulfurle acid digestion of th* matrix la required to fra* the PCBa from the matrix. (A more time consuming strong solvent extraction procedure 1* also available.)' Is this case, another problem has been documented. Th* sulfuric acid treatment aulfonatas th* lower chlori nated homolog*, *o that. If they had been present, they vould not have been measured by th* procedure. Another problem Is that the matrix In which particular Incidentally generated PCBe are found can change. On* manufacturer's product often is another'* raw material. Is such a casa, the chloroblphenyls may not b* measur able by th* same method used to quantitate them In th* first piece because th* matrix may have changed dramatically, in addition, th* second process may alter th* distribution of congeners. This is especially a problem If on* la trying to document th* occurrence of PCBs in a process or to track the genera tion of PCBs through th* procas*. A similar conearm arises In tha processes for chemically destroying com mercial PCB mixtures. As th* destruction process la carried out, tha mixture of chloroblphenyls changes significantly, so that monitoring tha progress or th* effectiveness of th* degradation scheme requires an ever-changing quantifi cation scheme. finally, the matrix can affect tha accuracy and precision of an analytical method in an Isomer or homolog specific way. The results of th* Chemical Manufacturers Association (CMA) PCB Analytical Task Croup round robin dlscuated In tha prepared paper may provide an illustration of this effect. In chit particular eats, tb* preelcion for tha monochlorlr.ated biphenyls was signifi cantly vora* than for the other homolog*. This could be attributed to the fact that In that particular matrix, moat intarfarenea* from the matrix occur In th* retention vlndov of the monochloroblphcnyls, affecting the quactlfleatlon of those compoundt. This affect could be axpecttd to occur for different homolog* In different metric**. In susntry, th* question! involving matrix effect! seam to b some of th* moat problematical. It saams unlikely at thl* point that any tingle analytical schema will provide accurate and precis* results for Incidentally generated PCBs In *11 the matrices In which chav might occur. OPAL ITT ASSPKA5CI AND DATA MASACgftST Th* Importance of a comprehensive quality saturate* program hss already been emphasised. It 1* essential chat only fully valldatad methods be uatd for tb* determination of Incidentally generated chloroblphenyls. Further, a continuous program of quality control suet be used to insure chat th* accuracy end precision established for th* method are maintained during routine us* of the method. 134 HONS 026225 An Important concapt vhlch map havt significant Impact on futura quality aaauranec programs, aa vail aa data msn*t***nt, la that of pattarn racognitlon. Pattarn rtcofnltIon, In tha aanaa of vlaual recognition of PCS alutloo patterns, has been Important for any years. However, tha futura applications of pattern recognition Involve computerized eystame which mathematically evaluate experimental results and cluster tha raaulta in understandable form. In tha Columbia National Fisheries Research laboratory, C.S. Fish and Vlldllfa Service, Columbia, MO. pattarn recognition 1* teen aa one of the moat critical and povarful maana of evaluating tha performance of attsbllshsd methodology and Identifying probleme when they exist. Tha "patterns" vhleb may exist for Incidentally generated PCBs are not established yet. However, It le reaeonable to aaeume that these pattern* will be Identified aa research continues In this area. As tha data base for these analyses become! established, pattern recognition could be a valuable tool to identify patterns that might exist in various materials. Tha uaa of the temple* a* training tat*, aa opposed to ualng commercial PCI mixtures, for principal component analytes, above strong potential for generating auefa data baaet. This type of systematic approach to data management for both samples and quality control sample* could help analytical chemleta make aanaa out of what othervita would bacom* "a real mesa." EXTENT OF PROBLEM - EXPOSURE AKD BIOAVAILABIUTT On* of the lituea relating to Incidentally generated chloroblphenyla la the extant of tha problem. Tha question Involves tha amounts of chloroblphanyla actually being produced, the potential exposure to the** chloroblphonyls, and tha bloavallablllty of tha chloroblphenyla. Although information regarding tha amount* of incidentally gencratad FCB* la sparse, a recant aurvty by tha CMA identified about 135 proeaaaet vhleb could product about 16,000 lbs. of Incidental PCB* at levels last than 50 ppm. The bulk of that* material* ax* dlapoaed of In an EFA regulated manner. Moat of tha remalndar la In forma not readily available for exposure, for example, the phthalocyanln* pigments mentioned tarllat. This 1* the very reason why treatments as harsh at sulfuric aeld digestion era necessary to analyse for chloroblphenyla In such matrices. Evan If the CMA estimate Is low by a factor of 100 or to, tha amount* of PCBs produced and released by Incidental genera tion art vary lov relative to the amounts still authorised for uaa In older electrical equipment. This point was discussed mots fully by Or. X. J. Mooleoaar In Station It Exposure Studies - Environmental Reeldue* and Bloaccumulatlon. Other information regarding Incidental genaratlon of chloroblphenyla ha* baan filed with the EPA In exaaptlon petitions. This Information could also help Identify specific proeaaaat or condition* which may presanc particular eoneams. lowavar, tha information has not base published and la nee readily available for review by toxicologist* to evaluate the risks or potantlal tzpoeurs. Thar* are certainly case* where tha environment, animals, and humans have baan exposed to PCB* In nonobvloua wave. Several reported example* are the Ingestion oi 'PCY-hontalnlhg anti-corroalon paint by hogs, application of PCBdiluted pesticides to cattle, leakage of PCBs Into a fat recycling ayatam and 185 HONS 026226 subsequent recycling Into animal feed, and the related polybromlnated biphenyl (PBB) incident in Michigan. Rovever, It must be emphasized that chase vert grots exposures to commercial PCIs, not exposures to Incidentally generated PCBa at levels lest than SO ppm. In summary, It appears that the amouees of ehloroblphenyle geoeratad Incidentally In chemical proctaaas la small compared to the amount authorleed for use or already In the environment. Hovevar, the toxicological and epidemio logical assessments of these chloroblphcnyls has not been carried out. THE SO ppm QUESTION Many of the questions regarding the Incidental generation of chlorinated biphenyls are concerned vlth the SO ppm regulatory cut off level for PCBa. The original rules ragardlng the manufacture of PCBa essentially stated that PCBa manufactured at leas than SO ppm In chemical processes vara not subleet to regulation. The appropriateness of this cut off level has been questioned In court and the EPA la presancly rewriting theee regulation*. According to a former EPA employee, the 50 ppm level was established for several reasons. One eoncero vea to establish a laval that vould clearly separata .environmental contamination from manufacturing activities. Since sever sludge and certain sediments shoved PCB contamination at the 20 to 30 ppm level, a rule to raqulra disposal of PCBa at a laval belov SO ppm vould have effected eovage treatment and dredging operations, for example. This vould have craatad an unmanageable situation In terna of bulk of materials to be Incinerated or landfilled. The ETA also fait that the Clean Vater Act and Clean Air Act could affectively regulate PCBa In river sediment and savagt sludge on a ca*e-by-ca*e baal*. Economic considerations vara alao Involved In the decision to set the eut off at SO ppm. Studies of costs of PCB removal shoved that costs Increased dramatically at sbouc the SO ppm level. Tor these reasons the regulatory cut off vas established at the 50 ppm level. There are still many question* which Impact on the appropriateness of the 50 ppm cut off. The extent of the problem la certainly on* concern,'. '"This vas discussed In the previous section. There are also serious analytical problems vhlch must be addressed. The** ver* the subject of Mr. Hodges' presentation for this Session, lastly, the toxicological and epidemiological aspects of the question* are yet to be fully explored. A great deal of research will be necessary before these questions can be resolved. IMPLICATIONS FOB TOXICOIOCT The Increasing aopblatlcation of analytical technology eraatas a serious dileema for toxicologist*. Mstbode utilising capillary CC column* vlth electron capture, El mass spectrematrle, or negative ion chemical Ionization (N1CX) meat spectromettic detectors can dstect subnanogran quantities of individual PCB congtoara In many cats*. Soma of these congeners are more toxic than ether*. However, very fev of these Individual congeners have bean tested for toxicity. Further, In toxlculogicel testing most exposures art to mixture* of congener*, rather then to single congeners, and at relatively high levels conparad to tht ultimata detection limits of the analytical techniques. 186 MQNS 026227 The problem for toxicologist* then becomes vhst to do with tht analytical data, vhtthtr for commercial PCB mixture* or for Incidentally generated PCB cort|tnri or statures. Th# toxicolojlat* ncad to relate dosage (or exposure level) to effect observed, From their viewpoint, the Ideal eltuatloo would be to take the reeult* of the sophisticated analytical methodology and produce eom* surro|ata value vhlch relates to doss|e or exposure. One possible approach to the above problem Is the use of a aatrlx approach to foraat tbe results of analytical studies. (This approach Is closely tied to structure - activity correlations discussed In Session 5 by Or. S. Safe.) In one example, the results are reported by the number of chlorines per biphenyl ring, as well as Identifying hov many chlorines are ortho-substituted. By analogy to chlorinated dlbento-p-dloxlne, ortho-unsubatltuted lsoaera have ore entyae induction activity and may be of particular concern texleolegleally. This type of approach may be a valuable Interface between the analytical chemists and the toxicologists. SPHHAXT .. - Tbe Issues surrounding the Incidental generation of cbloroblpbenyls are only beginning to be addressed. Although environmental analyeae for eoamerelal PCB mixtures havt advanced Impressively, tbe enelytleal approaches developed for commercial mixture# or for eevlronmantal analyeae ehould have applicability for thla art*. Aaldo from tha analytical laeuea, them are ttlll many question* to he answered with respect to th* level aod significance of exposure. In addition, th# queetlon of regulation It, as of this VTltlog, still to be ansvertd. It It to be hoped that an affective scientific approach to anaverlng thaaa questions vlll provide answers which will protect humane end th* environment, while at th* eeme time minimize economic burden*. 187 HONS 026228 CHATTER J HEALTH EFFECTS - EPIDEMIOLOGY This paper sunnarlze? ch results of seven cross-sectloaal cpldealologle studies dealing with health sffscts of PCBs, and three cohort aortallty ttudlts which contained data on the carcinogenicity of PCBs la huisans, Tha findings froa tha cross-sectional studies ara analysed according to tha following catagorlaat deraatologlc effacts, llvar function, fat setabollsa, othar objective findings, and reported ayaptoo* and lllnassas. Tha raportad cohort aortallty atudlas result In terns of nuabars of death attributed to various types of esnear. Tha paper concludes with a general discussion on tha strengths and weaknesses of each study, and the cross-study results which can be drawn forth. During tha first part of tha discussion suomary, a paper Is presented reviewing possible huaan reproductive effacts of PCBs. The data Includes results obtained froa tha Yusho Incident in Japan, and studies of several groups of Individuals previously reported on by Vllllaa Caffay. Tha discussion addresses tha aaaaured variables of days of gestation prior to birth and birth velght. A second study examining the relationship of PCBs to hypertension la presented. 138 MONS 026229 RECENT E?U>HIOLOCZC STTOIES 09 PC8* William R. Gaffty Monsanto Company SC. Louis. HO 63167 (for* the Tufho Incident in 1968, published studies of tba. health fftett of FCBt, vich eh* exception of Melga c al, (1954), vro aaaanclany clinical atudlaa of occasional accidental aavara exposures. In the following decade tan epidemiologic studies vara published (In addition to tba many raporti on tbo Tuaho lneldant ltaalf), all but ona of vbleb daalc vich oceupetlonally exposed population!, and only ona af vbleb atvdlad -tba ralatlonahlp batvaan cascar and PC* exposure (Caffay 1982). Sine* 1978 thara baa* baan tan epldamlologlc atudlaa of tba haaltb affacta of PCBs. Savan of than art erose-eectiooal Investigations of symptom* and bioehaaical parameters. Of tbaaa, fir* art occupational, ona la a atudy of both occupational and ooooccupatloaal arpoaura, and ona is uouoeeuputlonal. Tha remaining thraa taporta art cohort mortality atudlaa of occupationally axpoaad groups. In addition, thara baa baan a rapart on caneat nortallty In tha Tuaho population. Tbla papar britfly summarizes tha atatua of tba epldniolo|le evidence on health affacta aa of 1978, and reviews and evaluates tha subsequent atudlaa in detail. THE EVIDENCE AS OF 1978 The Tuaho Incident of 1968 at flrat appeared to be a claaeie example of tha effacta of a massive ingestion of PCBa. Over 1000 Japan*** becaae ill after eating a cooking oil contaminated vlth Kanecblor 400, a PCB conpound of Japanetc manufacture. Tha noat eoamon acute aynptona observed vara hyparpignentetion and acne-like lesion*, central nervous system aynptona, diacharge from tha ayea, and vomiting and diarrhea. The** aynptona vara dose related, and son* of then perslatad for aa long as six years. Laboratory teat* shoved dlscurbaoeaa of liver function and fat maeaboliam, and there vara clinical reports suggestive of tone ebnomalltiee in the children of Tueho noehere (KIOSH, 19*7). Ae of the end ef 1977, 51 deaths among Tuaho paelonta hod boon identi fied, vlth an apparant axcaaa of eanear death* (Grab* et al. 1979). flovevar, the data vara not adjusted for age, and tha eonplatanaaa of ascertainment of the deetha la not known. The elapeed tine from the original incident to tba reported deaths (ona dacada) was short enough to cast doubt on whether the cancer* could in fact have arisen iron the Tuaho expoaura. In any eats, in tha decada following tha Tuaho lneldant analytic techniques for Identifying PC*a and allied compounds improved considerably, and the cooking oil wee reanelysed. The nev analyses shoved that tba Tuaho perltne* (tad in feet ingested about the same amount ef polychlorinated dlbansofuran* (PCDFs) as of 90s, in addition to polychlorinated quatar-phanyla 189 HONS 026230 (PCQt). Current determination* of blood and tltauc level* of the** ether compound* In Yutho patient* have thovn total level* about equal to that of PCS* (Kimbrough 1980). It therefore appear* that the epidemiologic finding* in the Yuiho Incident (hould properly be attributed to the nor* toxic FCDP rather chan to PCB*. There vera nine other publlahad croca-aectlonal ttudle* of tymptoo* and biochemical parameter* related to PCB expoaure, one of which vat nonoccupatlonal, and on* occupational atudy of cancer morbidity and mortality. In lummary, the occupational atudle* found ehloraen* or ether dermatitis, and mild liver function abnormalities In the abaence of clinical llln***> to be aaeoclated with PCB expoaure. Tvo atudle* of choleatarol level* gave contradictory reaulca, and on* :udy of triglyceride levele choved an lncraaa*. Th* nonoccup*tlonal atudy ahovad no ataoclatlon between blood PCB level* and any of th* Tuaho eymptoma, but did not examine liver function or fat metabolic*. Th* cancer atudy found an apparent exceta of malignant melanoma bated on three caaea, but vea later withdrawn bacaua* of concern about whether the expoaed population had been correctly Identified. BECEKT STUDIES Th* aeven eroae-aectlonal atudle* completed alnc* 1978 addre** a broader rang* of health effect* than did th* pre-1978 ttudle*, and lo tome cate* uaed formal (tatlatlcal technique* to take account of confounding variable* tueh aa age, aex and veighc. Tvo of the ttudle*, Fltchbeln et al. (1979) and Smith at al. (1982) evaluated tbalr finding* In relation to higher chlorinated and lover chlorinated PCBa aeparately. Th* atudlea' finding* can be elatilfled under five general heading*! dermatologic effect*, liver function, fat m*taboll*m, other objective finding*, and reported aywpcom* and illneoaet. Th* cohort mortality atudle*, although they collected data on all c*uet of death, were concerned primarily with cancer. Dermatolocic Effect* Table 1 suumarltes th* reault* of th* five ctudle* that reported on dermatologic effect*. Th* negative atudy of Baker et al. (1980) examined 18 expoaed worker*, 19 member* of their faeillet, and 89 eoounlty reiidenc* with xpoaure to fertlllxer containing PCB*. Th* rest were occupational studle*. two of which found cbloracne and all of which found dermatltl*. Flaehbeln at al. (1979) reported that 50 percent of 326 capacitor manufacturing worker* reported a blatory of dermatological *yaptoas, the most coaon being a raah. Thoa* with *ymptoma had higher blood level* of high chlorinated PCBa. Karonl et al. 0981 11) reported ten ea**t of darmatltl* (five diagnotad a* active or peat ehloraen*) In SO expoaed worker*, but .provided no further data on the blood PCB level* of thoa* with dermatltlt verau* those without. Chat* at al. (1982)`found lacteased dermatitis, and tom* cbloracne, In th* most exposed group out of 120 railroad maintenance workers, but within the group found no significant association with blood or fat PCB level*. Smith et al. (1982), In a study of 92 employ*** of tvo 190 MONS 026231 Tabla 1. Studio of Chloracna and Othar Deraatiti* In Halation to Blood PCB Laval** Study Chloracna Othar Doaa Adjuatad for Dtrwatltla Htlated Othar varlabla* Flachbaln at al. (1979) N T T Bak.tr at al. (1980) N N Maronl at al. (1981 ID T T t Chat* at al. (1962) T T ? Snlth at al. (1982) N T N N Not found T Found No entry Not raportad ? Equivocal finding N N N H N 191 MONS 026232 utility companies, found no consistent assocletioc betveen dermatitis and ithsr high or lov chlorinated blood PCS levels. Although tha date are not consistent, they fuggait that there nay be a groaa doaa response relatlonahlp batweeo darnatltla and blood lavela of PCBa, poatlbly conpllcated by varlatlona In individual euaceptlblllty and in vork habit* that nay affaet absorption of PCBs through the akin. Liver function Table 2 shove tha result* of the six studies that investigated liver function. The two nonoccupatlonal studies shoved no liver function anonallas associated vith blood PCB levels. Ths first one did not adjust for confounding variables (Baker et al. 1980). The second, a study of 458 residents of a community vith high environmental levels of PCBs, found an association vhlch disappeared when age and alcohol consumption vara taken into account (Krais* at al. 1981). Of the four occupational studlstf three shoved various liver function anomalies relsted to blood PCB levels. Chase et al. (1982) adjusted the results for ags. Smith st si. (1982) adjusted the data for age and sex. The latter study shoved a significant association of liver function anomalies vith lov chlorinated blood PCBs in one of the tvo plants studied but not In the other. Maronl et al. (1981 II) found liver abnormalities associated vith elevated blood PCB levels, but the liver abnormalities vere defined as liver function anomalies or liver relsted symptoms or clinical findings of abnor mality. They state that "only In a fev cases vas a vell-deflned liver failure (presence of symptoms, hepatomegaly, and abnormal liver findings) present." Flsehbein et al. (1979) found no liver function anomalies associated vith exposure, and in fact commented on the "paucity of abnormal results" ir. their biochemical studies. With the exception of Maronl et al. 1981, the occupational studies agree in finding fev or no liver function ebnormalitles, and no associated elinieal illness. Pst Metabolism Table 3 eumaarltet the resulte of the-five studies that examined fat metabolism. Flsehbein et al. (1979) found no association betvaen cholssterol or trlglycarldea and blood PCBs. Krelsa at al. (1981) found an lnersass in cbolaatarol vith incraaalug blood PCB lavtls, but no relationship of triglycer ides to blood PCBa vhan an adjustmant vas mads for cholesterol level. Tvo ocher studies, Bekar at al. (1980) and Chase at al. (1962) agree that choltsterol la net associated vith blood PCBs but that trlglycsrldss art. Smith st si. (19B2) found eholsstsrol levels to be positively associated vith lov chlorinated PCBs In one of the tvo-plants studied. Triglyeerldes vert positively tssoclated vith high chlorinated PCBs in one plant and negatively associated in the other, vhere they vere positively associated vith lov chlorinated PCBs. 192 MONS 026233 Tabic 2. Studies of Liver Function In Relation to Blood PCB level** Study Abnormalltlee Flechbein et el. (1979) Baker et el. (1980) Rrala* et el. (1981) Moroni et el. (1981 II) Cheat et el. (1982) Smith et el. (1982) N N N T T T N Not found T Found No entry Not reported Dote Related Adjueted for Covariable* N N T TH TT TT 193 MQNS 026234 Tabl* 3. Studlaa of Ft Mctabollta In Relation to Blood PCB lavtli Study Total Triglyceride* Adjueted for Cholatterol Covarlablaa Flaehbalo t al. (1979) N N N BaVar t *1. (1980) N T N. Kralat at 1. (1981) T N T Chat* at al. (1982) N T T Smith at al. (1982) N N T * N Not found T Found 194 MONS 026235 The preponderance of avidtnea 1* that cholaatarel 1* not associated vlth blood PCBa. Thar* la no obvloui applanation for tha aaaodttlon of cholcattrol and blood PCBa found by trait* at al. (1981), Bovtvar, cbalr uaa of cholaatarol at an adjustment factor for triglycerides implies that they vlav cholaatarol aa an lndcpandant varlabla chat predict* triglycerides. The raaulta of Baker at al. '1980) and Cbaaa at al. (1982) do not support thla vlav, alnca thay ahov that the pueaclv* predictor, cholaatarol, la not aatocla tad vlth blood PCB level* vhlla trlflycerlda lavala art. The aaaoelatlon batvaao triglycerides and blood PCB lavala appaara to be ambivalent at the laaat. Other Objective findings Plaehbain at al. (1979), Baker ae al. (1980) and Haronl at al. (1981 II) examined blood chamlatry and found no aaaoelatlon vlth blood PCB lavala. trait* at al. (1981) found a ttatlatlcally significant poaltlva aaaociatlon batvaan dlaatollc blood praaaura and blood PCBa after adjutting for cholaatarol, trl|lyearldaa, smoking and race. Bovtvar, Smith at al. (1982) reported no aucb aaaoelatlon. Warahav at al. (1979) reported daeraaaad vital eapoeity In 243 eapaeltor manufacturing vorkara compared vlth published reference atandarda. Bovaver, moat of the atudy population vara current or former amokaxa, vhlla tha rafaranea ttandard la bated on a non-smoking population. Tha effecta of amoklng could bo aufflclant to axplaln the findinga. Reported Symptoma and Illnaaaaa Six of the atudloa Invaatlgatad a range of eymptome end lllneeeea. Two of them reported poaltlve findloga In oeeupaelonally axpoaad popularIona. Fiachbaln at al. (1979) raportad a hlatory of laatrointaatlnal aymptoma In 18 parcant of 326 eapaeltor manufacturing vorkara, a pravalenca of from 3,0 to IS.2 percent of verloua mutculoakelatal eymptome, end a prevalence of from 4.8 to 27.8 of verloua nanrologleal aymptoma. Thaae vara, hovavtr, nnralacad to blood PCB leval or duration of employment. Karonl at al. (1981 II) raportad eight eaeee of gaetrmlnteeclnal complaints In 80 axpoaad vorkara vlth no Indication of vhether there vaa a relationship to blood PCB level or duration of employment. Thay alao raportad tvo blaadlng haemanglomaa and one eaaa of chronic myalocytlc ldukamla. The Other four etudlaa raportad no PCB ralatad findInga. Specifically Baker at al. (1980) found no relatlonrblp vlth any of the follovlng; favar, valght loaa, anorexia, fatigue, haadacha, aya Irritation, eough, shortness of breath, nauaaa, vomiting, diarrhea, abdominal pain, arthralgia and paralatant raah. Kreiaa at al, 1(1981) raportad the aana thing for pravalenca of illntea or velghe loaa la tha pravloua year, naa of abdication, uaa of aadleal care, hlatory of heart dlaaaaa, and parentage of pregnanciea ending In mlecarriage, etlllblrth or Infant death. Chaaar-mt-al. (1902) failed to find any evidence of organ toxicity In a revlev of tha aedleaC hletorlea and phyaleal findings In a group of exposed railroad maintenance vorkara, and Smith at al. (1982) raportad no eonsleteut 195 HONS 026236 doit-depandent increaee 1b either symptoms or past Illnesses. Their inquiry covered digestive, respiratory, eantrel and peripheral nervous system, darmetologlc, and Motional symptoms. Tha evidence ahowa no FCB ralacad symptoms or illntaaai, tinea Cha tvo reports out of six chat mentioned symptoms failed to shov an association vlth level or duration of axpoaura. CarclnoaenlcltT Table 4 ahowa the major finding* of eancer mortality In three cohort studies of FCB exposed vorkers. Brown et al. (1981) studied a cohort of capacitor manufacturing vorkers In evo plants who had been exposed at least three months between 1946 end 1975 Inclusive In one plant or betveen 1940 and 1975 Inclusive In the other. Follow-up was more than 97 percent complete as of the end of 1975) and expected mortality was calculated from 0.S. population rates. Nonstatlstlcally significant axcassea vara found for rectal cancer, based on four deaths, and liver cancer, based on three deaths. The rectal cancers showed a slight Increase with an Increase In the latency period, but the liver cancers shoved no consistent trend. Thera vere no increases In mortality from these causes associated with increasing lengths of exposure. Bertmtzl et al. (1981) also seudled a cohort of capacitor manufacturing workers in a plant osar lillan. The cohort consisted of every person vho had worked for at least six months between 1946 and 1970, Inclusive, except for clerical workers. Follow-up was over 98 percent complete as of the end of 1978, and expected mortality was based on rates la the city where the plant vas located. Data were analysed separately for male (290) and female (1020) vorkers. A statistically significant axcasa mortality from all cancers, based on eight deaths, was found In male workers, and a statistically significant excess for all causes vas found In female workers, based on 15 deaths. There vere nonsignificant exeassas of lymphatic and hematopoietic eancer In both sexes, besed on four deaths, and of digestive cancer In males, based on three deaths. The sites Involved In the digestive cancer deaths were stomach, pancreas and biliary tract, do analysis by-duration or latency of exposure vet attempted because of the small numbers Involved. The authors not# that the eohort wee very young, end that their follov-up will be continued. Zeck et el. etudled all hourly mala workers who had baan amploysd In tho manufacture of FCIa in a ganaral chemical plant for at laast six months bstvten 1945 and 1965 lacluslva. Follow-up vas approximately 99 percant complete (one person was lost) as of tha and of 1977, and axpaetad mortality waa calculated from 0.8. population rates. There vara eight csnear deaths. Tha only noteworthy finding waa a non-seatlatlcally significant excess In lung cancar based on four deaths. No livar cancer deaths were found. 0ISC0SS10M Doll (1981) has suggested criteria for establishing carcinogenicity from epidemiologic evidence. They are similar to theta proposed by cha Interna-tlonal Ageneyfer Xssearcb onXencar (1980) .and are essentially ee follows: (1) there is so excess in exposed groups beyond what can be' expected by chance, (2) there Is an appropriate relationship vlth dose or duration of 196 HONS 026237 Tabl* 4. Inconalatanclaa In Mortality Studlaa of Cancar In PC8 Expoaad Population* Study Brown at al. (1981) Principal Mo. Studlad Mo. of Daatba Plodlni* 2387 183 Llvar lactun Bartatxl at al. (1981) 1310 27 Dlgaatlva Lynphatle and ha*a- topolatic lack at al. (In praparatlon) 30 Lung 197 MOWS 026238 axpoaura, (3) thara art no knovn blaalng or confounding factors, (4) tht association Is observed raptatadly In different circumstances. The develop ment of these criteria was stimulated by concern to provide generally accepted guidelines for establishing carcinogenicity, but they apply equally veil to other healeh outcomes. Criteria for establishing the abeencc of a health effect are inherently more difficult. The cliche that it is Impossible to prove a negative Is true but misleading, In the sense that it is also Impossible to prove a positive association by means of epidemiologic studies. A statistically significant positive finding allovs us to conclude that an association exlsta, vich knovn probability chat the conclusion Is vrong. Repeated positive studies reduce that probability In a manner that can be calculated, bue there remains a chance that the finding la false. On the other hand, a negative study allovs ua to conclude that there la no association, vlth a probability of error that depends on hov great an association really might exist. Repeated negative studies reduce this probability, but there Is alvays a large probability that a vary small effect might not be detected. Therefore both positive and negative studies carry a risk of error, although In both cases the strength of the evidence may be such as to make that chance negligibly small. Cross-sectional Studies There Is a reasonable consensus that dermatitis Is associated vlth occupational exposure to PCB*, although eha dose response relationship has not been established In all of the studies. Tvo nonoccupatlonal studies agree that there Is no dermatitis associated vlth environmental, l.e., nonoecupetional exposure. The difference In the results of the occupational and nonoesupaelonel studies Is consistent vlth the order-of-magultude differences In the exposures of the tvo groups. There Is s similar preponderance of evidence that mild liver function anomalies are aasoclated vlth occupational but not vlth nonoccupatlonal axpoaura. One occupational study found associated clinical sympeoms, but the raaainlng three studies found no datactabla clinical Illness. By Doll's criteria of consistency, it Is vary unlikely that clinical lllnsss Is associa ted vlth the aoomallal found In occupationally exposed population*. findings concerning cholattarol level* again ahov a preponderance of avldanea that eholaatatol levels are not associated vlth blood PCB level*, even though on* nonoccupatlonal study did report an association. Tht situation vlth respect to trlglyearld* level* la slightly mors ambiguous. Tvo out of five studies, Including oos nonoccupatlonal study, found a relationship vlth blood TCBs, so that tba possibility that such a relationship really exists must ba considered. Studies of blood chemistry vara uniformly nagatlv*. Tvo studies of diastolic blood prsssura gsvs contradictory result*. Ons study of pulmonary fbncclon suffered from a fallttri to account for smoking so that Its findings cannot bs attributed to PCB exposure. Nona of thasa sraas of study appear to hov any rslatlonshlp to PCB axpoaura. 198 MGNS 026239 Hoot of the tlx atudlaa of reported lllnaaaea and tymptoaa eonatltute vldcoca by Ooll't erltarla. Four atudlaa thoved oo relatlonahip of a vide ring* of aymptome to PCS axpotura. Tvo occupational ttudlaa reported varloua lymptoma, but did not ahov that they vara ralatad to blood PCS laval or duration of employment. Cohort Cancar Studlaa Tha note aotavorthv charaetarlitle of tha thraa mortality ttudlaa It that they do not agraa vlth aach othar. Excatt cancar of tha liver, ttetua, itomach, pancreaa. biliary tract and lung ara aach found In ona of tha ttudlaa but la nona of tha othart. Although tha povar of tha atudlaa It limited (tha moet powerful ona. Brown at al. (1981) haa a probability of about 0.33 of detecting a threefold excatt in liver cancar mortality), tha lack of a latency ralatlonehip in that ttudy and the eomplaca abaanca of liver cancar in tha other two ttudlaa atrangthaot chit power In a way thae cannot bo quantified precieeiy. Tha average duration of follow-up In each of tha thraa atudlaa vaa 13.3 yaart Brown at al. (1981), 13.1 yaara (Karonl at el. 1981) and 20.3 yaara (Zack at al.). Further follow-up, eepaeially of the flrat two atudlaa, would provide tha longer lataney and greater number of daathe that would lncraaaa tha power of the atudlaa. Nevartheleae, eha data available at tha praaant time do not provide eha eonalarant, axpotura ralatad raaulta that would juatlfy concluding that PCSa ara carcinogenic In human being*. 199 MONS 026240 REFERENCES Baker El, Landrlgsn PJ, Cluack Cl, Jack MU, llddla JA, Buraa VW, Housevorth UJ, Ntadhan 1L. 1980, Metabolic consequences of axpoaura ce polychlorinated biphenyls (PCS) In aavafc sludge. Anar. J. Eptd. 112:333. Bartaazl PA, Zocchettl C, Cuercllena S, Foglla M Delia, Ptaatori A, Rlboldl 1. 1981. Mortality study of sale and female vorkara exposed to PCBs. lot. Symposium on Prav. of Oecup. Cancer. Helsinki. Brovn DP, Jones M. 1981. Mortality and Industrial hygiene study of workers exposed to polychlorinated biphenyls. Arch. Envlr. Health 36:120. Chase KM. Wong 0, Thomas D, Barney BV, Simon RK. 19B2. Clinical and metabolic abnormalities associated with occupational exposure to polychlorlnaeed biphenyls (PCBs). Jour. Cce. Med. 24:109-114. Doll R. 1981. Relevance of epidemiology to policies for the prevention of cancer. Jour. Occ. Med. 23:601-609. Plschbeln A, Wolff MS, Lllls R. Thornton J, Sellkoff IJ. 1979. "Clinical findings among PCB exposed capacitor manufacturing workers. Ann. N.T. Acad. Scl. 320:203. Caffey WR. 1983. The epidemiology of PCBs. In PCB's: human and environmental hasards. D'ltrl FM, Kamrln MA, Buttetvorth Publisher, Woburn, MA. International Aasoclatlon for Research on Cancer (IARC). 1980. Monographs on the evaluation of the carcinogenic risk of chemicals to tumors. Pol. 19:18. Rlabrough R, ed. 1980. Halogenated biphenyls, terphanyls, naphthalenes; dlbensodloxlns and related products. Elsevler/Horth Holland Biomedical Press, Amsterdam. Chapter 9 SI. Kralss R, Zack MW, Kimbrough RD, Needham LL, Smrek AL, Jones BT. 1981. Association of blood pressure end polychlorinated biphenyl levels. JAMA 245:2503. Ksronl K, Colombi A, Arbostl C, Centonl S, Foe V. 1981. Occupational exposure to polychlorinated blphenyla II. Health effects. Brit. J. Ind. Med. 38:53. Meigs JV, Alboo JJ, Kertln BL, 1954. Chloraena from an unusual exposure to Aroclor. JAMA 154:1417. NI0SI. 1977. National Institute for Occupational Safety and Health. Criteria for a recommended standard: occupational exposure to polychlorinated biphenyls (PCBs). 0SDEW. N10SB Pub. No. 77-255 Sept. 200 MONS 026241 Smith 5CS, Schloener J, lovry IX, Saellvood AW, Llgo XU, Tanaka S, Stringer V, Jones M, Bervln A, Glueck CJ. 1982. Metabolic end health consequence* of occupational eaposure to polychlorinated biphenyls (PCBs), Accepted for publicetloo, Brit. Jour. Ind. Med. Urebe H, Rod* B, Aeehl K. 1979. Present state of Tusho petlents. Ann. H.T. Acad. Scl. 3201273. Warshaw R, Flschbeln A, Thornton J, Killer A, Sellkoff IJ, 1979. Oecress* la vital capacity In PCB-exposed workers In s capacitor auinufacturlng facility. Ann. N.T. Acad. Scl. 320:277. Zacfc JA, Musch DC. Mortality of PCB workers ee the Monsanto plant in Seuget, II. In preparation. 201 HONS 026242 DISCUSSION SUMMARY Renat* D. Kimbrough, M.D. C*ntar for Dla*( Control Cnt*r for Environmental Health Atlanta, CA 30331 Follovlng th* dlseoure* of Dr. Caffey, Dr. Philip Taylor presented pre liminary raaulta of a study h* conducted In conjunction with investigator* from th* New York State Department of Health for the National Inatltute of Occupational Safety aod Health. In thla atudy, the effect* of FCB* on repro duction in occupationally azpoaed women were Inveetlgaced. This atudy was undertaken because in th* Yusho apltode (accidental lngoation of rice oil contaminated with PCts, chlorinated dlbemofurans and quarterphenyls) 11 vomen were pregnant at tha time of exposure, two of these women had miscarriages and nine had live births. Three of th* Infants were light for age and alghe of the nine had various tymptona of (ototoxicity, such ae aya discharge, colacolored skin, and erupted teeth. In addition, fetotoxlcity has baan produced In a variety of animal species. The cohort of women that was studied were workers for the Ceneral Electric Company at two facilities In Opseate New York south of Lake George. The two facilities ware located In adjacent communities and have been Involved In the manufacture of capacitors since 1946. There have been roughly 7400 Individuals employed In those eomoltles beeveen 1943 and 1975, of whom about half were females. A cohort of women was studied whose personnel records Indicated they had taken leave because of pregnancy. Between 1950 and 1973, theta were 388 pregnancies In 334 vomen that resulted In live births within the State of Mew York, and these became eh* object of the investigation. The Information on birth weight, maternal age, parity, year of birth, race, sex, and data of last menstrual period vat ob tained from birth certificates, and where necessary hospital records and physi cian records were reviewed. From a file of all single live births In the surrounding two-county area, a tingle control was selected for each of these births occurring to a capaci tor worker matched on maternal age aod parity precisely; year of birth plus or minus a couple of years} and rsca. Th* controls were cross-checked against th* personnel files of tha CE records to assure ehat they had never worked ae GE. Tha.cohort was limited to person* who had worked ae least three months for GE and wmre divided into high and low exposure groups. High exposure areas were those parts of ehe plane la vhleh thers was direct contact with FCBs during or after the introduction of FOB* into th* capacitor. Th* areas of the plant with low exposure consisted of some women with clerical job* whleb presumably would have no contact with th* FCBs and women from th* manufacturing process prior to eh* Introduction of FCB*. to 1973, area sample* tekon In one jnrvay showed -thet th* value*, in tha high areej, following the impregnation of FCB* war* in eh* nalghborhood of 679 Vg/m and In tha lower areas ehsy ware 260. In 1977, the area* of high 202 HONS 026243 expotur* hd an avtrege value of 310 and the areas of low exposure of 22? Ug/a . Pereonal air templet vara only done during a NIOSH industrial hygiene survey In ehe high areas sod averaged 168 wg/m3 air. Air samples taken Imme diately outside tha plant averaged In the neighborhood of 6 yg/e vhlle ambient air levels are 0.1 ug/a3 or less. Uhen the birth weights of Infants from the low exposed group were com pared to county controls, It was determined that the 337 Infants of Bothers in the low exposure group had birth weights that were 66 grams higher than the birth weights of the Batched county controls. A comparison of gestational ages shoved chat the SI neonates froa high exposed woman had a mean decrease of 3.4 days in gestational age and the aesn difference of their birth weights vat alnua 93 grams when compered to county controls. This difference was not statistically significant. However, when SI neonates of high and 337 neonates of low exposed feaalet within the plant were coaparad, the neonates of the high exposed group weighed sn average of 133 graas lets then the neonates of the low exposed group. This difference was statistically significant. Uhen the data was adjusted for gestational age, it was found that the difference in weight woe largely due to differences in gestational age. Overall, 13.7Z of births occurred before 37 weeks of gestation in tha high exposure groups and S.6Z in the low exposure group. The point estlaate for the rate ratio was 2.6 with 90Z confldencs Intervals that did not cover one. The meaning of thasa results, s reduction in gestational age of approxlaately one week and a reduction In birth weight of about 133 grame, whan high and lew exposure groups were coaparad within the plant is open to Interpretation in term* of blologieal Implications. This study was tha only atteape to investigate reproductive outeoaes in PCB exposed females. All studies discussed by Dr. Caffay In the presentation that preceded Dr. Taylor's were either croae-eaetlonal studies In predominantly adult populations or mortality studies. Dr. Caffay took the liberty of comparing studlaa that were basically quite different. Two of thea were general population-type studies; the one by taker, at el. of CDC, wee dona In Blooalngton, Indiana. A general population that had been axpoaed to PCBe in toll wee compered to a control population. This population had no hlghor blood levalt of PCI* than the eoaperlaon group. It la therefore not surpris ing thst no otrlklng health affacts could ho associated with PCI exposure. The only axcaptlona were e oaall group of 18 workers or foxaar workers who had had somewhat higher exposures. Another study Dr. Caffay mentioned wee also a CDC study of a predomi nantly black population of tbo smell town, Trlana, Alabama. Thla population had bed axcaptlonal exposure to DOT rttlduts, but la addition, when terum templet ware analysed, PCBs wars also found In high concentrations in soma of tha peopla. A tottl of 638 people wars examined, and tha range of the PCte In that group in eerum wet 3.2 to 137,9 ug/1. There were 98 peraens In this group that had level* higher than 30 ug/I. There wee no extensive PCD contami nation in the town of Ttiene. The PCt blood levels la theta people could be associated vlth the consumption of fleh. Ae far as we could determine, the 203 HONS 026244 PCB concentrations In fish had usually not axcaadtd FDA guldalinat for tha past can yaars bafora tha study vaa dona. An additional laportant finding In thla study vaa that aa tha agt of tha paopla lncraaaad, thalr PCS blood lavala tatnad to lncrsase, and In faaalaa tha PCB blood lavala vara lovat than in salts. According to a recant paper by Mary WolIt, thara la alao an locraaaa In PCBa In adlpoaa elaaua vlth ago. Dr. Chats In a raeant publication, which vaa ona of the papara ravlavad by Dr. Caffay, found that tha vorkart vlth longar axpoaura had hlghar PCB body burdana. Tha incldanea of hypartanalon In thla black population froa Trlana studied by Kralee at al. vaa 301. At a aaan age of 21 years, thla la a hlghar Incldanea than vhat has bean rapoTted among blacks. After adjusting for other confounding variables for hypartanalon, such as valght, cigarette smoking, ago, alcohol consumption, thara vaa atlll a small contribution which had eo be attributed eo PCI axpoaura. Mora studies need to be dona to determine whether those findings ean ba verified. Thla population la quite different from tha popula tion that Dr. Blair Snleh examined where no association with high blood pres sure vas found. In thae population, the moan aga vaa 40, Expoaura vaa occupa tional to nature and moat of the workers wart white. In addition. Dr. Caffay discussed three mortality studies. All of thorn had In common that the number of available daath certificates vas vary small. Even though small sxcaaasa of cancsr vsrs found, this Information has to bs taken vlth eautlou because of the limitations of the studies. Mortality needs to ha examined further eo determine vhat eba significance of ebosa findings la. If the latency period for the Induction of cancer In humans la 20 to 40 years, a sufficient amount of tlma has not elapsed to determine whether exposure to PCBa is a rlak factor in the development of eancar. Although PCBa vara first manufactured in rhe aarly part of thla century, va are nov beginning eo eoe a larger population that hae bean exposed to PCBa for 10, 13, and 20 years or longer. Thla population will have to ba folloved. further bafora any decision on ebo Incidence of eancar in humans can ha made. Following ehese general comments, many different points wart raised by tha attendees. Dr. Taylor vaa asked vheeher tha children whose birth weight ha ravlavad vara followed after birth, lovmvar, since hla study was a record eaarch study, this vaa sot possible nor vas it possible to control for confounding variables. It vat, however, possible to establish that 11 children bad died .. up to the age of 9. Seven of those deaths occurred among tho 388 county controls and 4 occurred among tho infanta of tha 377 low exposure group of tha CP workers. In addition. It van again pointed out Chat tho lav exposed group and the high exposed group in tho CE plane vara not matched hue adjustments for varlabloa voro mado in tha analysis of tho data. The point vao made by Dr. Mary Wolff that Dr. Caffay'a etatamane about smoking among tho vorkars of tho Warsaw studies vaa irrelevant since tha find ings of raocrlctlvo pulmonary funeelon easts made in that study wars usually not associated with smoking. IDVavor, apoxlaa vara need in that plane. This aay ba an Important confounding variable. 204 HONS 026245 Additional attempt* to quantify expoaura la the two CE plant* etudied by Dr, Taylor wart udi by Mary Wolff. Tht lower ehlorlntttd blphtnylt In ptoplt wart itrongly corraltttd with air level* while the higher chlorinated biphenyl* ware not. Dr. Gaffey pointed out that the duration of followup In two mortality atudlta reviewed In hie preeentatlon (Brown and Bertaaxl) wat about 13-1/2 yeare and In the Zack etudy about 20 yeara. So, the reault la that the mortality rate, for example, In the Brown etudy for male* wee about 3.3 par 1000; In the Bertazil etudy about 2.79; but among aalee Brown'a rate wae 3.8; Sertazzl'e rate wee 0.48. It aeaaa to ba an extraordinarily low aortality rate and yet It vat almoac 100Z note than expected. It la vary difficult to axplaln. Zack'e rate for aalee wae 12.8 per 1000 which la underacandable bacauaa the cucoff data for adalaelon to the etudy vaa 12 yaara before the end of It, but It la puzzling. In ehe Bertazzl data Che cohort la admittedly young. Bit aalee have a lover rata than Brown'a, about half the rata, but hie feaules have about 1/8 the rata. It la vary puzzling to hia. Since in the Trlana etudy (Krelee, at al, 1981) expoture to FCle and DDT raalduet had occurred, the quaatIon vaa ralaad whether the affacta of DDT raelduee and PCS* could ba addltlva, and It vaa pointed one that the analyaie of the data did not aaaa to Indicate thla. Or. Blair Smith, Rational Inetltute for Occupational Safety and Health, indicated that In eapeeltor plane vorkara which ba had aendled tha mean aarum PCS levela of the lover chlorinated homologuea vara on eha order of 300 ug/1 with a background laval of 10 eo 15 ug/1. Tat, tha mortality data from tha Brown atudy of a plant which vaa almllar vara equivocal. Be latherefore lafe in a quandry about trying to impute eoma human toxicity to PCBa In eha occupa tional axpoeura altueelon. In chat regard Mr. David Brovn aakad him to tay that be vaa planning to follov up on that cohort which haa bean followed through 1973. Ha will ba following It through 1980. With additional yaara of follovup, eha etudy will atatlaeleally become more poverful and poaalbly aoma of tha bordarlln* exceai rieki for cancer may tun out to ba atatlaeleally algnlflcant or poaalbly noe. Dr. Smith vent on to polne out that in thoee eeudlee where eha magnitude of tha aaaoclatlon between TCB aarum levela and biochemical teeea vaa reported In a quantifiable vay that eha amount of variation vaa In general no more than about 10X. In regard to liver function ceaea and aarum lipid or plaema lipid abnormalltlaa, the aaaociationa appear vaak. Whan Dr. Smith analyzed hla data obtained from capacitor vorkara blood preeeura, nnadjuatod for aga, vaa correlated with aarum PCS laval; hue when It vaa aga adjuatad, the correlation want avay. At praaant, tc la not poaalbla to axplaln tha dlaerapancy between hla atudy and that by Kralaa, at al. (1980). Dt. Caffe?- then made tba polnt thee ehe Brown and Jon**' atudy haa aoma daea for longer latency period*, and tha cancer data euggaaca eo trend. Ha gave vinyl ehlorld* aa an example where, with eheee klnde of etudlee, aoma- 203 MQNS 026246 thing would heva bn teen. It vaa nottd by Dr. Kimbrough that in th Brown and Jona* atudy thara seemed to ba a trand for ractal eancar, and aioca vinyl chlorlda cauaaa (uch an unuaual tumor, it it much aora aaaily daeaccad. In addition, the polntad out that baiically analyzing daath carciflcatas la a vary inaanaltiva tool. Vheraupon Dr. Taylor objactad. Indicating that ralatlva to tha othar and point! praaantad hara, thay ara as aanaitlva, Ba also indicated that continu ing at a former vorlcar from CDC located in tha New York State Health Depart ment, ha would ba looking at and expected to have available information within tha next year on tha entire cohort of 7400 and soma odd workers of tha CE cohort. At this point, 750 deaths have been identified which will make the issue of arguing about thirties and forties reasonably moot for tha particular BCB compounds--that exposure was incurred at those particular facilities. Dr. Caffay in further commenting on the Brown study pointed out that there were only three liver cancer deaths; it is fairly difficult to show a trend with only three observations. Tha Brown atudy waa, in terms of . statistical power, a relatively weak study,. It will become more powerful with Increasing length of followup. Aa epidemiologists, we ere dealing with highly variable outcomes fraught with error, both due to lnhereot biological variability and variability due to analytical technique and mediocrity of the recording of information on death certificates. Clvan the inherent dirtiness of tha data, hew refined dost our measurement technique have to be in terms of separating out specific PCB isomers end homologues? In place of single number, the proliferetion ef additional numbers as predictors Is confusing. Whereupon Dr. Taylor stressed that whenever randomly sloppy data ere anelyted, the hiss is slways In the null direction. If e difference la demonstrated with sloppy data. It la mote likely that it la true than It isn't. Hovevar, if a difference is not daaonttratad and tha data ara sloppy, than It must ba determined whether tha inability to establish difference it related to poor measurement techniques. Mr. Bob Noonan with Batrak asked Dr, Caffey whether, in trying to explain the apparent contradictions end incon sistencies In findings in the eross-seccional studies, was an attempt made to compare PCB doaaga lavala and mode of exposure. Or. Caffey stated that he did not attempt to do that. Dr. Caffey, whan asked by Dr, Chase whether he felt that the studies he reviewed were not adequate stated that the crosa-sactloual studies vere quite adequate and that the mortality studies ara aora adequate than haa bean sug gested; and that the problems that have been raised about the power of the tests ars betide the point because a complex of things is evaluatsd. The power is examined to detect exeeaa mortality ot a trend. In the studies that do not show an excess, le is determined whether there is something that is almost there, but thara appear* to be nothing of this kind. Maybe as these studies are followed up further, the evidence will become even stronger, but the evidence does not indicate a carcinogenicity problem with PCS* or other problems beyond the dermatitis and the marginal liver abnormalities. According to Dr. Chess, when dealing with the issue ot cancer with a long latency period exceeding che latency periods allowed for In eny of the ehree eross-seetlonal studies, the studies ware not sufficiently powerful, either 206 MQNS 026247 collactivaly or Individually, to detect an lncraaiad rlak. Furthar, thara may not ba a trand In tha llvar eaneara, but thara la an axeaaa of llvar cancan In tha Brown itudy. Tha exeats la cooeadad by Brovm aa not being atatlttlcally significant. It la of Intaraat that thara la an axeaaa of llvar eanear In tha Yuaho population aa vail, again, not ataclatleally algnlfleanc, and tha ccnclu-- alon la that a aufflelanely povarful atudy la not yat avallabla to aay any thing daflnlta. IMl ha a baan aald by a eoupla of oebar tpcaktri. Dr. Caffay raapondad that ha did not station tha aortallty In tha Yuaho population bacauaa ha dlaeountad It aa a PCB-ralatad phanonanon. It la not known out of vhat population it aroaa bacauaa aa tin* haa gona on, tha daflnltlon of Yuaho haa changed, and tha denominator haa changed In vaya chat ara noc pradictabla. Secondly, tha aortallty data from tha Yuaho atudy hava not bean aga adjuatad. Thui It la not poaalbla to determine vhat percentage la too high alnca coaparlaona with an age-adjusted population cannot ba made. In other word*, thla la not an aga and tlse-adjuated proportional aortallty atudy; and third, tha alapaad time Interval, at leaac from tha Yuaho Incident until the first aortallty figures ara given vat tan year*. For the Yuaho deathe, the question is vhathar a latency period of let* than tan yaara, aay hava been too short to cause tha cancer. Aa far aa Brovn's atudy la concerned, avan If hia axeaaa In llvar cancar vara ataclatleally algnlfleanc, it ehould noc be accepted aa balng occupationally related unlaia It vara ralatad at tha vary sdnlaum to latency of axpoaura and poaalbly alao to duration, and neither of Cheat vara tha caaa. Dr. Cheat agreed on the ahortcomlngs of tha Yuaho data but the three atudlaa praaenttd by Dr. Caffay do have tone shortcomings aa vail. They ara noc adequate to draw a definitive conclusion aa to. tha carelnoganic rlak of FCB axpoaura yat. There la alao a qutatlon about Flaehbeln'a atudy. According to Table 2 of Dr. Caffay'a presentation, no dote relationship betvean llvar funcclon testa and PCB lavala vaa found. Tha Flsehbeln ertlele aaaaad to auggtst to Dr. Chase chat tha main llvar function rase* vara not significantly abnormal In the exposed group but that there vat a atatlatieally significant linear correlation between ae least one of the liver function tasca and PCB '.evali which has alto baan reported la ocher studies. Furthermore, In the 1979 Flachbtln study It vss statod chat analysis of data to determine vhathar an association between PCB serum lavala, choletcarol, and triglycerides txlscad would ba reported at a later data. Dr. John Brown of General Elaccrtc pointed out that chemically PCBs ara lipophilic agents vhleh tend to dlstrlbueo at oqual concentrations in all lipid pools la tha body within a fav days or voaka, and this means that tha levoL la lipid roaorrolrs, such os t)>* micellar lipids of tho lipoproteins in blood tend to approach thac of the PCB adipose tissue reservoir*. Aa a result. In any population of ladlvlduals having similar adipose tlaaua levels the blood lovol of PCB must of necessity parallal that of ehe level of blood lipid. So, thara la a confounding In thara. Tou predict that lo a population of such individuals you vould have to find a correlation unlaao tha second lav of thermodynamic* vora balng violated. Tho correlation may baeoaa weak and diffi cult to aao In a largo population reflecting a large apraad of adipose tlaaua PCB lavala. The face that that* la a eonfbunding between the lavala of blood lipid and blood PCB moan* that it la difficult to establish correlation* which infer that PCB baa caused a change In lipid aseabollsm or a correlate of 207 HONS 026248 elevated blood lipid auch hypertension. However, thi does not mean that uch hypothealied correlations cannot ba tested. Instead, what It aaaoa la that tha propar cotralata to look for In tasting althar hypertension or aarum lipid la not tha aarua PCS lcval but tha adlpoaa tlaiua PCB laval. In tha caaa of tha GE capacitor vorkar population, thla haa baan axplorad by GE. A significant correlation of tha aarua PCB laval* with both triglyceride* and cholaatarol vaa found. However, thla correlation vanlabaa vhan correlation* era a*de with adlpoaa tlaaua PCB laval*. Apparently Or. Chaaa ha* aada alallar obaarvatlon* In hi* population and Dr. Brown ralaad the quaaclon whether thla problaa of confounding had bean axamlned by other* who are looking for abnormalltlaa In lipid aatabollaa. It wa* pointed oue by Dr. Kimbrough chat to her knowledge, thla had not baan dona. Tha fact thae there era higher concentra tion* of llplda in aarua alght explain why eon* people hava higher PCB blood level*. They vara concerned about that In tha Trlana atudy baeauaa only blood level* war* dona. In a population that gat* older, cholaatarol laval* aaaa to lncraaaa In people that aat a weacern type diet and with It, PCB blood lavala. On* arguaanc agalnac chat vould ba the fact thae theta saema to ba a relatively conalatant ratio between (Cl blood lavala and PCB adlpoaa tlaaua laval*. Dr. Brovn Indicated that to the extant theaa ratio* have baan examined by CE, a batter correlation axlata between adlpoaa tlaaua PCB laval and aarua PCB level if a correction la aada for tha aarua lipid level. In raaponae to a quaaclon by Dr. Frlaaa, Dr. Kimbrough atatad that there are aone aclantlata who feel that It vould ba better to naa adlpoaa tlaaua for tha determination of body burden* and aa an axpeaura Index. It la of couree nora difficult to gat adlpoaa tlaaua than It la to obtain blood aaaplaa. In large atudlea, the tendency haa beeo to naa aarua and to deternine PCB* in aarua and aaauae that thla represent* a ana11 portion of what is In adipoea tlaaua. If result* from population* are revlowed where laboratory analyse* have bean adequate,' there seen* to be a constant ratio between PCB concentra tion* in adlpoaa tlaaua and in serum. This ratio varies vith tlaia to last exposure If exposure was recant and such problems as weight loss. Subsequently Dr. Ian Webber of ITS Corporation took Issue with Dr. Caffey'a statement that the Tusho oil Incident was not a PCB related incident. Dr. Wabbar stated that in tha case of tha Tusho oil tha PCBs had baan in uaa for soma considerable time and had degraded to give a "dirty" oil. It might be that tha heat exchanger application of PCBs results In a worst esse situation. Tranaformer oils on tbs other hand tend to degrade also. From a practical point of view, health studies thae Involve the largest amoune of exposure to people In terms of either those gaetlng rid of PCBs from contaminated oils or people Involved with retroffiling transformers, should ba done because these are the people who are likely to coma Into contact with dlbensofurana in PCB solutions. Rave any studies been done with used PCB*? Dr. Robert Bell from CE then Informed the audience that they were in the middle of a project evaluating AO-year-old tranaformer fluid. Their findings data Indicate that th levels of dlbensofurana In 40-y*ar-old fluid are at or below levels of the original PCB* that verb delivered to CE. Thus, transformers operating at ambient temperatures do not pose a significant haaard. 208 HONS 026249 Dr. Webber countered thet those transformers operating t ambient temperature probably do not contain slgnlfleant levala of dlbenzofurens. Thera are however, a lot of factor* lo the uae of PCB olla that com* Into play, the age, th* amount of oxygen Involved and hot epot* caused by faulting. Dr. Ball than mentioned th* Binghamton case. Dr, Kimbrough asked vhethar everybody was familiar vlth th* Binghamton situation. A transformer caught fir* and a lot of combustion products ware formed. Because eha transformer also contained chlorinated bansanas, chlorinated dlbenaodlexlna formed In addition to chlorinated dlbenzofurens, and because of the way the alrshaft was situated In relation to the transformer, all of ehls material was evenly dis tributed over a very large state office building. According to Dr. Ahmad, epidemiology Is a two-edged svord and one can pretty much argue one's ease depending upon how one interprets the data. Taking Dr. Gaffay's presentation and other Information as well, a correlation Is assumed betveen PCB blood concentration and actual exposure levels. In any kind of an epidemiologic study. It could be asked what th* actual exposure level of the Individual or cohort eo a given agent was. Obviously, surrogates ars used, and this could laad to all klnda of lntarpratatlona which may lead to poaltlve correlations or negative correlation* depending upon how le all comas out. Secondly, to address tha problem of confounding variables. Dr, Caffay In ona slide showed studies that ware done since 1978, Soma of the studies had been adjusted for confounding warlables, and aoma of than had not. Dr. Kimbrough's data showed that PCB blood levala Increase with age, at least In this particular group and that there vaa aaex difference In PCB blood lmval. This alona should warn ua that wa should try to adjust for variables In any of these studies. Ha have to adjust for age, ethnic differences, sax, and so on. Unless this la dona for all stadias, they cannot ha compared on tha same basis. Those studies that do taka smoking habits sad various other confounding variables Into account art obviously far batter. Unfortunately, many of tha studies presented here have not looked into those factors proparly. Dr. Caffey made a suggestion that statistical power could be Increased if studies vara combined. Issue must be taken with pooling different populations from different geographical areas, such as the Brown study and tha Bertazzl atudlaa, which vara dona In two different parts of th* world. The two population* obviously had ethnle differences and th* study protocols ustd were quit* different. Statistical power, even though v* should look at all of tha Information In a qualitative way, has to b* used in such a way that th* statistician will agra* gives a value of tb* sensitivity of eha study. Soma of eha atudlaa presented today vara occupational and others nouoceupstional. No information vaa given la tha nonoecupatioaal atudlaa about tha degree to which these individuals war* axposed. Problem* about latency periods exist. Since whan and how long vara Individuals exposed to PCIaT How did their blood concentration correlate vlth their actual exposure years ago? There ara so many problem* with each of that* studies, it cannot possibly ba concluded whether or not PCB* caua* th* kinds of problems presented. V* do have positive studies on dermatitis snd chloraeo*. then there ara azudlee that conflict with on* another. How ean It be concluded that this mesas that there are no problems associated with PCBsl 209 HONS 026250 Dr. Caffay raapondad that ha did not think that ha aald chart vara no problaaa, but that ha apaelflad tha problaaa that aaaaad to ba conalatautly turning up. In addition, thara la alaoat no way to gat around ualng aurrogaca* for axpoaura In cohort aortallty atudlaa bacauaa It la not poatlblt to naatura vith any praclalon tha axpoauraa that cook plaea many yaart ago. Furthermore, confounding varlablaa can frequently erase* spurious dlffarancaa. Thay can raraly vlpa out raal dlffarancaa bacauaa to eraata a dlffaranca vhara thara. In fact, la nona, any confounding variable vill do. But to vlpa out dlffarencea that axlat. a eonaplracy anong a number of confounding varlablaa la naadad vhleh la rather unlikely. Tha nonoecupatlonal atudlaa have In eonaaon the fact chae their axpoauraa ara lev, aod tha occupational atudlaa have In eoason tha fact that their axpoauraa ara high. It vaa not auggeated that aoaa alnple place of erlthaeclc can conblne all thaae atudlaa Into one overall nuaerical aatluata of povar, but any raaaonabla aaaaaamant of tha evidence haa to regard not the povar of an individual atudy, but tha meat of tha evldenct, not oaceaaarlly In any foraal vay, although this could ba dona. Dr. Klabrough pointed nut that individual auseeptlblllty mat ba conaldarad aa vail and that thara scans to ba an affect on reproduction In anlaala and that haa net baan atudlad to any groat axeane In huatana. An Inal data alao auggatt that fanalaa ara parhapa nora eenaltlve to the toxic affect* of PCBa than aalaa. and parhapa a greater effort ahould b* nad* to look for health effect* In feaalaa. Dr. Brovn raapondad to an earliar question ralaad by both Dr. Frlaaa and Dr. Ahaiad relating to tha paralatanca of PCBa la tha body and vhathar body burdana can ba naad a* Mature of lntagratod axpoaura. They ara attaapting to anavar thl* quaatIon through tholr atudlaa of the GE capacitor vorker group. The rtaults to date Indicate that aoat PCB lioaari of the Aroclor 1254 typo ara highly peralatant In tha body, and thay don't clear. Ones they have base deposited in tha body they ara there foravsr anlaas they can ba axeratad through lactation. This My bo trua for a few ltonari of the lover PCBa. Thara nay ba ona or tvo laonara that can bo uaed for traeklog integrated axpoaura. Thla la not certain yet. Moat of tha lover chlorinated honologuaa are cleared but certainly the higher laonara represent a paraanant record of axpoaura. 210 MQNS 026251 CHAPTER 6 HEALTH EFFECTS - LABORATORY STUDIES Th flrat papar In this chapter review* the daalgns and result* of inlul studies performed to assess tha toxicological affocto of PCBa. For tha aaka of thoroughness, thla ravlav lncludat aavaral itudlaa completed prior to 1978. Data ara ptaaantad and dlacuaaad on tha follovlng topic*: akin affaeta, reproductive dyafunctlon. teratogenicity, fatotoxlclty, liver structural and anzynatle alteration*, gaatrlc la*Iona, Immunosuppressive affaeta, porphyria Induction, carcinogenesis, and mutagenesis. Tha papar conclude* vlth a summary of knovn affaeta in anla*1a and a comparison of thaaa to finding* In humans. Tha aacond deal* apaclfleally vlth ona PCBa efface knovn to occur in both animal* and human* -- llvar anzyaa Induction. The data.demonstrate* In substantial detail hov substitution* In tha para, mats, and ortho positions can altar this activity. Tha antyma Induction by tha PCBa Is comparad to tha established strucma-actlvley relationship* of the phanobarbital and mathylcholanthrana type lnducars, as vail as to tha ora hasardou* polyehlorodlbanao-dloxlu compound*. Tha discussion auaaary begins vlth a detailed analysis of tha animal carclnoganlclty studies, both thoaa presented In tba papar and aavaral addi tional studlas. Tha discussion summary than axamlnas tha "initiator-promoter" theory and the Implications of rat llvar nodulas findings. 211 HONS 026252 POTENTIAL HEALTH EFFECTS FROM EXPOSORE TO POLYCHLORINATED BIPHENYLS; LABORATORY STUDIES S. L. FrlM, V. A. Drill, B. V. Hays, T. A. Looai*, C. B. Shaffer Frteae, Drill, Bay*, Local* and Shaffer, Inc. Consultant* In Toxicology Arlington, VA 22209 INTRODUCTION TMa presentation la axtractad from a raeant ravlav of potantlal haaleh affaec* In cha huaan froa axpoaura to polychlorinated biphenyls (PCBa) and ralatad lapurltlaa, a* praparad for cha National Elaecrlcal Manufaceurara Association and tha Edlion Elactrlc Institute nndar data of January 25, 1982. Two taaaa of apaclallata, rapraaantlng cha dlselpllnas of aadlelna, toxicology, pharaacologyblochaalicry and epidemiology, partlclpatad In tha study. Tha flrtt taaa revlaved tha PCI haalth affacta literature in datall through 1981, and ptiparad draft raporta on tha ralaclonahlp* batvaan PCS axpoaura and apaclflc haalth affacta In cast animals and In axposad paopla. Tha aaeond taan of aclantlsta lodapandantly rarlavad tha draft natarlal. ConeluaIon* and opinions raflact contribution* froa both taaaa. Tha praaont focus la on offsets In laboratory animal nodal* subjactad to controllad toxicological axparlaanta, vlth additional eoaaant on cha bearing of thasa result* on potantlal haalth problaas In axposad huaan populations. REPRODUCE I7E EFFECTS IK ANIMALS Reproductive Problems Numerous raporta have appeared In tha literature on eha affacta of poly chlorinated biphenyls on reproduction In animals. Cellart and Wilson (1979) shoved ehat female Sprague-Devley rats given 30 mg/kg Aroclor 1221, 1242, and 1260 by gavaga during days 14 through 20 of pragnaney had no affect on reproductive function as judged by (a) normal eserus cycles, (b) normal appearance of ovaries and utarl, and (e) fertility of males. Feeding 550 ppa of Aroclor 1254 for 67 days to Sherman rats rasultad In fever Uttar*, smaller litter site, and 100Z mortality by day three of tha FI pups. At 100 ppm aurvlval of both FI and PL offspring was raducad. Toe pups vara smaller than controls but appeared normal at veanlng. Aroclor 1260 fed at a dietary laval of 500 ppm (33.4 mg/kg) for 67 days prior to mating markedly raducad Uttar alie and survlval-to-vanning In the FI and PL generation. Dietary levels of 5 ppm Aroclor 1256 and 100 ppm Aroclor 1260had so effect on reproduc tion In rats exposed through tvo generation* (Linder at at. 1974). A dietary concentration of 2.5 ppm of Aroclor 1248 produced alteration in the menstrual cycle of adult female rhesus monkeys. Menses vara prolonged and menstrual bleeding vas Increased (Allen et el. 1979). Kapllngar -at al. (1971)-fad rats and dog* Aroelor 1242, 1254, and 1260 at doaoa of 1.0, 10.0, and 100 ppm. Bo advarse affect on reproduction vas observed In rats given 1.0 and 10 ppm Aroclor' 1242, but there vs* a decrease In survival of pups at 100 ppm of Aroclor 1242, 1254, and 1260 as vail as a daeraasa In mating lndlees. Tha affect of PCls on reproduction In baagla dogs 212 HONS 026253 and evin* ha* been studied by Earl t *1. (1974). Aroclor 1254 Interfered lgnifieaotly with reproduction Id dog* at doaaa above 2.5 eg/kg/day while la twine 10.0 og/kg/day lowered fertility and survivability of neonate*. The reproductive effect* la dogs are questionable becauit of unexplained change* in the reproduction pattern among control*. In a *tudy reported by Allen at al. (1974), tlx adult, female menu* monkey* were fed a diet containing 25 ppm of Aroclor 1248 for two month*. During thl* period, all animal* developed charaeterltclc alga* of toxicity, although they maintained a regular menatrual eycla. One animal died. At the beginning of the fifth month, or three month* after dlacontlnuance of exposure to the test lubitance, an attempt vaa made to breed the survivors. Three animals appeared to cooeaive, but only one succeeded In carrying her fteue to term. Although thla Infant vaa vail developed, it* body weight vaa considerably lover than that of the averaga rhetua Infant. Examination of the tissue* shoved no gross or microscopic lesion*. In a second study by theVlsconsln group (Barsotti *t al. 1976) involving 18 adult, female rhaeu* monkey*, nine vara fed a diet containing 2.5 ppm, and nln* vara fad a diet containing 5 ppm of Aroclor 1248, After seven months on the** diets, the eight surviving animals from the 2.5 ppm group and eight fromthe 5 ppm group vara mated with control malaa. All animals In tha 2.5 ppm group bacama pregnant, hut three of these resorhed their embryos; the remaining five gave birth to llva Infanta. In the 5 ppm group alx animal* bacama pregnant. Thea* impregnation* resulted la thraa abortions, one resorption, one stillbirth (suffocation during difficult delivery), and one uncomplicated birth. At birth, the six .slants vara small; hovever, other than their small stature and focal areas of dermal hyperpigmentatlon, chair ganaral appearance, hemograms, and osseous development as evaluated radiographically vara normal. Teratoganeala Mica PCBs were not taratoganle In mica at dosages up to 500 mg/kg given on days one through six or days savan through 11 of gestation (Toeruek 1973). Eovever, a recant study (Marks at al. 1981) vlth tha haxaehloroconganer 3,3', 4,4', S.S'-hexachloroblphenyl in mlee at dosages in tha range of 0.1-16 mg/kg/day during days six through 15 of gestation produced a significant increase in fatal malformations, a significant dacraasa in avarage fatal valght, and an lneraasa la the parcancaga resorptions. Soma mica exposed to S.A^'A'.-tecracbloroblphenyl vara reported to exhibit a "valtzlng syndroms" (TUson at al. 1979). Tha dosag* was 32 mg/kg administered by gavaga .to tha mothers on days 10 through 16 of gestation. The syndrome consisted of Increased motor activity, hyperreflexis, sod episodes of head bobbing and rotational movements that were present up to at least sight months of age. Roe all exposed mica vera affeecad similarly, but even those that did not exhibit the syndrome shoved diminishad performance In various Jtaurcbehavloral tests, Tllann ee al. (1979) refer to these observations as tha "behavioral teratology" of 4-CB. 213 MONS 026254 Thr art a plethora of studies In vhlch tht reproductive tfftcta of feeding varloua Aroelor* to rata havt bttn investigated (Calandra 1976, Ctlltrt and Wilaon 1979, Repllnger at al. 1971, KapUngtr at al. 1972, Linder at al, 1974, Villtotuve at al. 1971a), Collectively, thasa atudlaa raported fecotoxicity at tha dosage vat increased, but no teratogenicity vaa deaoottratad. It sap ba uaaful to nota tht ordar of aagnituda of tha dotagaa Involved. Vlllanauve at al. (1971a) found no affact froa tha adalnlatratlon of doaagtt of up to 100 ag/kg par dap orallp to pragnant faaaiaa froa tha sixth through tha fiftaanth dap of gaatatlon. Ooat In tha ttudp eonduetad bp Earl at al. (1974), pragnant bltchaa vara given dotagaa of 0.23, 1.0, or 5.0 ag/kg of Aroelor 1234 froa tha dap of braadlng to tht dap on vhlch thap vara nacroptltd. At 3.0 ag/kg, chert vac an lneraata In tha percentage of tetorptloni, a decreate In the nuaber of live pupa.par litter at birth, and a reduction in tha parcentaga of tboaa surviving to tvo vaakt. Tha tarata obtarvad contiattd of enlarged foncanellat, claft palataa, and tuperfluoua phalangat. Thla doaaga vaa tald to Halt diet contraption ttvaralp. The authort atata that one litter of the controla "nap have had a genetic defect that caused an unutuallp high incidence of tarata." Svlna Earl at al. (1974) included nlniatura tuint (Horne 1 strain) in the investigation datcrlbad inadlately above. Pregnant aovt vara given dally oral dotagaa of l ng/kg, 10 g/kg, or 30 ag/kg of Aroelor 1234. Doting began 21 days before braadlng and vaa continued until the day of necropap. Tha authort concluded that doee-ralated affacta vara seen at all treataent levels as evidenced bp dccraaaaa in tha nu&bar of pregnancies, nuaber of live pigs farroved par Uttar, and parcaataga of live offspring after tvo veeks of age. Syndactyly and cleft palates vara observed in tha young of aovs receiving 10 ag/kg, and patent fontanalles and cleft palates la tha offspring of those receiving 30 ag/kg, As in tha ease of tha dogs, tha higher dosages eausad a narked reduction in food consuaptlon. Fetotoalcltv Rabbits glvan 1.0 ag/kg of Aroelor 1254 dally for 28 daps of gestation shoved no affact on the developing fetus but doses of 12.3 to 30 ag/kg vara fetotoxle. Rats appear to ba more resistant than rabbits, as doses up to 100 ag/kg do not eause fatal .deaths or aalforaatlons (Vlllanauve at si. 1971a) Eabryotoxlc affects of Ranechlor 300 and 500 vara produced in Sprague-Davley JCL rats when fad at levels of 300 ppa la tha diet throughout gestation (Shlota 1976b). Dlete containing 3.0 ppa of Aroelor 1248 vara fetotoxle to rhesus aonkeps. Studies in ales on the transfer of PCBs to fetuses and offspring indicate that tha aaount transferred dapanda upon tha chemical . ttyuceure of the individual eoeoounds and the position of tha chlorine atoas vlthln tha chaalcal structure lHasuda at si. 1978, 1979). 214 HONS 026255 CAKINOCENICITT Rats Th three major studies pertaining to PCBs and hepatic carcinoma in the rat art those of Kimbrough t *1. (1975), Calandra (1975), and cha National Canetr Inatlcuta (1978). Kimbrough and covorkers, using tha Sherman aeraln of rat, found that 100 ppn of Aroelor 1260 la tha dlae for 21 months produced a aignlfleaot incraaaa la tba iaeldanea of hapatocallular carcinoma la faaala rata. Kala rata vara not atudlad. Tha admlnlatratlon of thraa dlffarant Aroclora at a deaa of 100 ppa to rata for 24 nontha did not induce hapatocallular carclaotaa (Calandra 1975). Tha author atataa that thla nagatlva finding vaa baaad on avaluatlon of tha llvar aactlona by thraa pathologists, who rssd tha slldas iodapandantly and aaparataly. Ha atatas furthar that Professor P. Pour raavaluatad tha Kimbrough alldaa and did not agraa vith tha raportod findings, Results of tba National Canesr Instltuta (1978) bloassay of Aroelor 1256 for posslbla eareinoganlcity differ fro* thoaa of Kimbrough vith Aroelor 1260. Fischer 366 rats vara usad and tha PCB mixture was admlnlstarad at throe does levels (2J, 50, and 100 ppa) In the diet for 104--105 weeks. Tha offset of the treatment on body weight and mortality at the higher dose levels demonstrated ehat maximum tolerated doses vara used, thus providing a satisfactory teat of carcinogenic potential. A sufficient number of raes of both sexes vat tvallabls for aasnlngful statistical analytes of the lneidanee of late-developing tumors. Treatment was associated vith only three hapatocallular carcinomas in mala rats, vhleh was not significantly differant from the controls, and it was concluded thac Aroelor 1256 vat not carcinogenic in the bloassay. leo at al. (1976) treated rats vith different doses of Ranechlor 500, 600, and 300 for periods up to 52 veekt; hapatocallular carcinomas vara not observed, but tha duration of treatment vat probably too short to rule out tha possibility of a positive response. Mice Ito at al. (1973) observed hepatocellular carcinomas in mice only in the high dote group given Ranechlor 500; lever dosat of Kaneeblor 500 or tha other Kaoechlors did not produce a carcinogenic response. The administration of 300 ppa of Aroelor 1256 In the diet for 11 monehs did not lnduca hepatocellular carcinomas in mica (Kimbrough and Linder 1976), Hepetocellular carcinoma did not occur in dogs (four male and four female par group) receiving 1, 10, or 100 ppa of Aroelor 1262, 1256, and 1260, ratpectively, In their diets tor two years (Calandra 1975). General Coamaht Animal studies do not provide convincing evidence that PCBs induce llvar eanear. Of tha major atudlaa in the rae, one hat been judged positive and two 215 HONS 026256 have been negative. Hepatic carcinoma vat not produced tn tha dog. Chronic administration of Aroelore In tha rat did not Induct bladder cancer, gastro intestinal. carcinoma, or cancer of tha thyroid gland, pituitary gland, adrenal gland, uttrua, lung, hematopoietic system. or other organa. Data from tha mouse provide only limited and restricted evidence for a carcinogenic affect of .the Japanese product Kanechlor $00. MPTACENIC EFFECTS Testa for mutagenic activity of BCBe in mammalian cell tyttama or Intact animals vara uniformly negative. Specific teats and their results are listed belov. Cytogenetic Analysts In Vitro Hooplngarner at el. (1962) activated cultured human lymphocytes with phytohemagglutinin and treated them vlth 10Q ppm of Aroclor 1254 at various stages of a division cycle. The calls then vara examined for chromosomal aberratlone for the different stage treatments. Aroclor 1254 had no apparent effect on chromosomal integrity as measured by cytologleal evidence. Dominant lethality In Rodents Green et el. (1979b) concluded that Aroclora 1242 and 1296 mere not mutagenic since they failed to Indues dominant lethal mutations In rats. Xn similar studies reporeed both by Kepllnger et al. (1972) and Calandra (1976), Aroclora 1242, 1254, and 1260 vara administered to male, albino mice In a single lntraperitoneal dosage of aleher 500 mg/kg or 1000 mg/kg. Subsequent matings of these animals to control females shoved no effect of treatment on the number of implantation or resorption sites, number of viable embryos, or preimplantation lots. Ames Test In Salmonella Aroclors 1221, 1254, and 1260 shoved little mucigenic activity la the Amea test (Wyndham at si. 1979) vlth Indications that Aroclors vlth lovar chlorine eonttnts art vaakly positive In the Salmonella teat systems. Bovever, these results must be discounted since they could not be replicated (McMahon et el. 1979, Safe 1960). At a genarsl finding, Hsddla and Iruct (1977), McMahon et al. (1979), Schoeny et el. (1979), end Safe (1980) mere unable to find evidence of mutagenicity of ?CSs in bacterial tasc aystans. Cvtoaenstlc Analysis In Vivo In tvo Investigations, PCBs vert Judgtd not to hava produead chromosomal abnormalities, Dlkshlth et el. (1975) examined seminiferous tubules of rses st different intervals after dally doeegee of 50 mg/kg of Aroclor 1254. Green et si. (1975) axsmlnsd both bone merrov and spermaeogoniel cells of rata after either e single dose of 5000 sg/kg of Aroclor 1242, or four dally dose* of 500 mg/kg. In all esses< itvss concluded that no significant chromosomal damage had occurred. 216 HONS 026257 Hlcroouclaus Tut 7Vi mleronueltus tt 1 Cast for agents that tend to break chromosomes (clsscogene). A olcromicleus la a fragment of chromatin chat haa brokan away fro* a ehro*oao*e during call dlvlalon and haa failed to be Included in either of the daughter nuclei. The phenooenon can be obaerved beat In newly formed erythrocytes alnee theae ataln differently fro* the Batura erythrocytes In that they exhibit polyehrovaala for a period of 2d hours or ao. Saddle and Bruce (1977) reported Aroclor 1254 as negative In a alcronucleus teat. OTHER HEALTH EFFECTS Effects on Skin Epithelial and follicular hyperplasia and hyperkeratosis have been reported to occur following the application of PCBs on the skin of rabbits (Vos and Beeas 1971). Skin lealons have also bean observed In rats and guinea pigs fro* the application of PCBs. Repeated application of Aroclor 1260 to the akin of rabbits caused thickening of the skin as a result of hyperplasia and hyperkeratosis of the eplthellua (Vos and Beeas 1971). Cutaneous effects have been elicited In aale rhesus aonkeys fed a diet containing 300 ppa Aroclor 1268. Within one month the animals lost consider able hair from the head, neck, and baek (Allen at al. 1973, 1975). Similar effects have been obaerved In female rhesua monkeys fed a diet containing 25 ppa Aroelor 1268 for two Booths (Allen et al. 1976). Within alx weeks the anInals began to lose hair and developed obvious signs of edema of the Ups and eyelids. Snail pustules involving hair follleles appeared about the nouth, cheeks, and neck. Abraheason and Allan (1973) showed that tha Infant sionkay was able to tolerate doses of PCBs that produce axtreae morbidity in adult aonkeys. They suggested that there nay be variations In absorption, distribution, metabolism, storage, and excretion In adule and Infant aonkeys chat nay account for these differences. Effects on Liver liver Weight The administration of PCBs aay induce an Increase In liver weight. The effect is sore prominent at tha higher dose levels; the lower doses of PCBs do not Increase liver weight. The detailed data of Kimbrough et al. (1972) Include findings thac Aroelor 1260, administered with the diet In doses of 500 and 1000 ppa for eight Booths, significantly Increased liver weight In aale and faaale ratal doses of 20 and 100 ppa were effective only In aale rats. Aroelor 1256 Increases liver weight la both aale and female .rats at doses of 20, 100, and 500 ppm. The Increase In liver weight is correlated with hepatic eell hypertrophy and an increase In smooth endoplasmic reticulum In the rat, rabbit, and aonkey (Vos at al. 1972, Allen and Abrahaaaon 1973, Bruckner ae al. 1976, Allen, Norbstk and Bsu T97*, Allen, Carsttus, and Barsottl 1976). 217 MONS 026258 General Hletology Tha admioiitration of PCB* to experimental animal* vill product hltcopithological change* la tho llvar. The aala affacta obcalnad ara cha produc tion of anlargad hapatocytaa, fat droplatt. and alight dagraa of nacroala; tha occurranca of thaaa chaogaa dapandt particularly on cha dota of tha PCS and to aoaa extant on Cha particular Aroclor. Nacroala aaana to ha nora aavara la tha rabbit than la tha rat. la ganaral, tha morphological changaa ohaarvad la tha llvar of PCB-traacad animala ara tlmllar to choaa found aftar traataanc with othar chlorlnatad hydrocarbon*. In a datallad atudy, Klabrough at al. (1972) found that Aroclor 1260 (0, 100, S00, and 1000 ppa) and Aroclor 1254 (20, 100, and 500 ppm), given In tha dlat of rata for alght aoncha, produced enlarged llvar calle and cytoplaaale ineluelone. At tha higher doaaa, there vae evidence of lipid accumulation, which waa aaaoelatad with a foamy cytoplaaa, and pigmanc accumulation la the Kupffar calla. Adanoflbroala Adanoflbroala (aynonyna: bile duct proliferation, bile duet adanomatoala, cholangloflbroala, fibroadenoma) haa been obeerved la aoaa raea receiving PCBa. Adanoflbroala la generally ragardad aa a benign laalon. Klabrough at al. (1972) obeerved adanoflbroala in rata treated with Aroclor 1260 and 1254; the affect vaa obeerved ae tha higher doea levale and particularly in anlmale receiving Aroclor 1254. Whan tha feeding of 500 ppa of Aroclor 1254 vaa dlacoatlnued and tha aalaala atudlad up to 10 additional aontha, tha fat and llvar concent of PCI raaalnad high and eha adanoflbroala paralatad (Klabrough at al. 1973). In a later paper involving treatment with Aroclor 1260 ae 100 ppa for 21 aontha, aantlon la only aade that a few raea ahowad araaa of adanoflbroala. Indicating a low dagraa of reaponea, but data ara not tabulate!, (Klabrough at al. 1975). In further atudlaa by tha ease group adanoflbroala vaa not obaarvad In rata fad Aroclor 1016 or 1242 (100 ppa) for up to 10 aontha (Buraa at al. 1974). Tha atudy of the National Cancer Inacltuta (1978) did not ravaal adanofl broala In rata receiving Aroclor 1254 at doaaa of 25, SO, and 100 ppa for 104-105 vaeka. Hvpemlaaclc Pod and Nodular Byperplaala Byperplaatlc foel or hyperplaaelc araaa repraaant minimal changaa In llvar hapatocytaa. Such foci or araaa of byperplaatlc changaa are uncommon In untreated young rata, but ineraaaa with age. The hyperplaaelc araaa or foci may coaxlet with nodular byperplaala and/or hepatocellular carclnoaa. Tha elgnlflceaee of byperplaatlc foel la that they nav bo part of a apaetrua eapabla of prograealng to a nodule (Squire and Iavltt 1975), Kapllnger at al. (1971) did not obcarve hapatle laalona In rata receiving different Aroclor* for 18 aontha; however, revaluation of tha elide* indicated a elgnlfleant Ineraaaa of nodular byperplaala in tha treated anlmala (** National Cancer Znatituta 1978). 218 MQNS 026259 Ito at *1. (1974) obaarvad nodular hyperplasia in rats receiving 100, S00, or 1000 ppa of Kanachlor $00; a lesser effect vaa observed vlth Kanechlor 400 and Chara vaa no significant 0f Kanachlor 300. Klnbrough at al. (197$) raportad a significant lneraaaa In tha lneidanca of hyperplastic foci or araaa and neoplastic nodulaa In female rata givan 100 ppa of Aroclor 1260, nixed vlth tha dint for 21 aontha. In cootraat, in anothar study, tha dietary adolnlatratlon of 100 ppa of Aroclor 1242, 1254, and 1260 for 24 aontha lncraaaad tha lneidanca of nodular hyperplasia; a laaaar affaet vaa obtalnad vlth 10 ppa, and 1 ppa did not produea a poaieiva raaponaa (Calandra 1976, Levlnskaa 1961). All thraa Aroelora producad an lncraaaad nuabar of hapatooaa at tha high doaa. Thara vaa no avldanca of nacaataala or lnvaslveneaa, and tha laalona vara ragardad bp tha pathologlata aa-banlgn tumora. Caatrlc lesions Tha oral adnlolatratlon of PCBa to nonhuaun primates haa baan ahovn bp Allan and eovorkara to lnduca hppartrophp and hpparplaaia of tha gaatrle ucoaa. Mucoaal epata map ba praaant within tha aplthallua of tha atooaeh, but In tha oucosal lining and tha auboucoaa, thara aap ba edaaa of tha aubaueoaa of tha atooaeh. In addleloo, thara la frequently Invasion of tha underlying oucoaa bp laolaead glandular alaaanta of tha oucoaal aplthallua. General offacta of this nature have baan produced in tha oonkap bp: 1. A alngle oral doaa of 1.3 g or 3 g of Aroclor 1248 (Allan, ttorback, and Hsu 1974). 2. Aroclor 1248, 300 ppa and Aroclor 5460 (a polychlorinated triphenyl), 5000 ppa la eha diet for 90 dapa (Allan, Abrahaaaon, and Horbaek 1973, Allan and Hotback 1973). 3. Aroclor 1248, 25 ppa In the dice for 2 aontha (Allan, Carotene, and Baraottl 1974). 4. Aroclor 1248, 100 ppa In the diet for 2-3 aontha (Allan 1975). 5. Aroclor 1246, 2.5 ppa and 3.0 ppa In the diet for up to one year (Baraotel and Allen 1973). (Caaerlc changea were not Hated for theaa doaaa in Allan'a ravlev.) 6. Tha hpperplaatlc gaatrltla aap pertlae for over one pear follovlng dlacontlnuance of KB expo euro (Allan 1973). A mixture of low chlorinated biphenyl* (Clophen A-30) vaa not obaarvad to produce gaatrle laalona In tha aonkap (latropouloa at al. 1977). Hppartrophp and hpparplaaia of tha gaatrle oucoaa have not baan raportad In tha rodent (Allan and Abrahaaaon 1973), rabble (Toe 1972), or In other ...species (Vqa and Koaraan 197D). 219 HONS 026260 Entvma Induction Various studies havt raportad on eba alfact of short taro azpoaura of rats to mixtures of chlorloatad biphenyls. On* such study by Ecoblchon at al. (1974) usad tha Aroclors Identified as Aroclor 1016, 1221, 1242, and 1234, and anothar study by s Prsnch lnvastl|stor, Narbonns (1980), usad a French prspsratlon known as Fhsnoclor DP6. In both studlas tha saapla matarlsl eonslstad of eonianars of tha chlorloatad blphanyls. Thaaa atudlas ahowad that whan rats wars admlniatersd 10 ppm of tha PCBs In thalr disc for only ooa day. tha livers froa tha animals ahowad Induction of P-430, anUlna hydroxy laaa, and aainopyrlna-!t-danathylasa. Tha Induction of thaaa anzyaas was maximum In about five days. If tha rats wars glvsn 10 ppa of tha PCBs in tha dlat for alght eonaacutlva days. Induction oecurrad In thraa to flva days and to a lassar exeent ovar tha remaining eight-day latarval. In tha Ecoblchon study, latraparltonsal Injection of tha Aroclors In dosaa of 30 mg/kg was made for thraa consacutlva days. Animals wars sacrificed 96 hours latar. Mlxsd function oxldatlva anzymot wars found to bo Induced In tha llvar, Tha greatest Induction was seen with the more highly chlorinated Aroclors. Tha enzyme Induction oecurrad rapidly on axpoaura to the Aroclors, and particularly to those Aroclors that are more highly chlorinated. The PCBs used 1m thasa studies ware commercial grade. Aroclors 1242 and 1234 are potent Inducers of hepetle microsomal enzymes. A single lotraperltoneal injection of 100 mg/hg of Aroclor 1242 to rats increased llvar weight, total microsomal activity, as measured by hydroxylatlou of acetanilide and N-demethylatlon of amlnopyrsne, and hepadc cytochrome P-430 (Bruckner at al. 1973). Iwunosuppress Ion Invlronmhntal chemical affects on lmmunocompatancs have been extensively ravlawad hy faith, luster and Vos (I960). A number of atudlas deed by them found PCBs to ba Immunosuppressive in various spades (from ducklings and chicks to mica and monkeys). In soma cases st dose levels that produce little offset other chan bapacocyta hypertrophy. Only one study (Street and Sheroe 1973), vhlch involved feeding low levels (0.18 to 6.3 mg/kg/dsy) of Aroclor 1234 to rabbits, showed no significant effect on humoral or call mediated rasponsa. Our review of tha Street and Sharma raporc Indicated that chare wsa a dose-related trend toward launoaupprasaion in thalr animal* and at tha higher doses (2.1 and 6.9 mg/kg/day) for four to tight waoka tha animals bowed significant Increase in llvar weight. Tha faith, Luatar, and Vos review Hats savera1 ganaral factors (nutritional status, hormonal lavals, and amounts of Imnunoragulatoty proteins such as slphs-fstoprotsln) that Influence lmeuno function. Therefore those atudlas la which lamuna function wts studied following exposure to lavala of tbo PCBs that raaultad In overt toxicity art meaningless from an lmmunotoxicologic viewpoint. In contrast, lamuaoauppraalon at dosage lavals that do not produce ganaral toxicity would ha significant. Porphyria Tha administration of PCBs to rats will produce a delayed type of porphyrls, t.a., deposition of porphyrins and thalr degradation products la tissues and excreta. Although tha mechanism of action of tha PCBs la unknown. 220 MONS 026261 It apparently differs from that of othr porphyrogenie compounds such as haxachlorobenzena or lsopropylacatamlde. It has bean thought that tha affaet of PCBa In animals may tndleata that thasa compounds can lnduea condltlona such as porphyria cutanaa tarda In man, but clinical data hava not shovn this to occur in PCB-exposed worker*. Tha facal contanc of coproporphyrin and protoporphyrin of rabblta vas lncraaaad by Aroclor 1260 and by hasachloroblphanyl, but tha diffaranca vas statistically algulfleant only for eoproporphytln (Vos at al. 1972). Bruckner ac al. (1974) observed an Increased ascratlon of urinary coproporpbyrln In rats given Aroclor 1242. 5 or 25 ppm for two, four, and six months. Goldstein at al. (1974) emphasised tha delayed onset of tha PCB-lnducad porphyria; rats fad 100 ppm of Aroclor 1254 became porphyrlc after two or seven months of treatment. Although tha azcraelon of coproporphyria and other porphyrins vas lncraaaad, tha largest elevation vas In tha uroporphyrin fraction. Thera vas also a marked accumulation of uroporphyrin In the liver. Tha studies of Coldsealn at al. (1974) suggest that PCS* may affect uroporphyrin formation or utilisation. Tha induction of delta-aminolevulinic synthetase (ALA synthetase, a rata-limiting ensyme In heme synthesis) does not appear tobe tha mechanism by vhlch porphyria la Induced by PCB, as It la for many porphyrogenie chemicals, in their eeudles the Increase In ALA synthetase activity vas probably secondary to the porphyria. Also, Aroclor acts to Increase liver cytochrome P-4S0 rather than to decrease It. Bexachlorobentane Is knovn to Induce a delayed type of hepacle porphyria similar to that produced by Aroclor 1254. lovever, the tvo responses apparently differ significantly as Goldstein at al. (1974) reported that tha rata fed PCBs did not exhibit tha nervous or eutaneous sips associated vlth hexachlorobensene poisoning. GENERAL COtCgNT, SPHMART AMD OPINIONS Body Burdens, Metabolism, and Kinetics Some general observations derive from consideration of the data on the storage, distribution, metabolism and excretion of commercial PCB mixtures (vlth associated impurities such as the PCDFs) In mammalian systems. Tha dynamics of tha pathvays by vhlch thaee processes occur In teat animal models hava bean explored intensively and have bean fairly veil vorked out. The variability of PCB affects vlth spades of taae animal models has been inves tigated for a number of Important effects, and some general structure versus activity rules have emerged from these studies. Thera thus exists a conceptual fraaavork, but not experimental data, for describing such phenomena In humans. Further, Important data on the biochemical interactions of PCBs, PCDFs and metabolites vlth tissues have been obtained that bear directly on Interpreta tion of toxic effects In theaa tissues. Cenerel Toxicity From acuta, suocnronlc end chronic studies of PCB effects In test animal models, s general summary of health effects encountered Includes! (a) a lev ordar of acuta toxicity; (b) skin lesions; (c) effects on liver tissue that 221 HONS 026262 art generally reversible at lovar doses but that trend toward irreversibility with lncraaaad dosing; (d) effects on tba gastric nucoaa; (a) effects on tba menstrual eye la and on reproduction; (f) porphyria; (g) affacta oo tha kidney; and alscallanaoua changaa Including haaatologlc altarationa, thymic atrophy and lymphatic changaa. Thara ia a eonaldarabla ranga of apaclaa auacaptlblllty to biological activity of tha FCBe, with ona rapraaaneativa comparison of ordar of daeressing activity baing mink, monkey, rat. It ia concludad that tbara it no baals for determining vfaleh. apaclaa aoat aecurataly aarvaa aa a aodol for prediction of ganaral aubehronlc/chronic toxicity in aan; rather, tha judgment of tha aoat auleabla aurrogata for aan auat ba aada Independently for aach major health affact. Skin and Other Cutaneous Tlteuea Commercial fCBa are capable of producing akin laalons in the aonkey aodal and in axpoaed humane. Chronic adalalatratlon of PCBs to tba aonkay producer chloraena. In huaana occupationally axpoaad to PCBa akin dlaordera including chloracaa have ocoaaionally baan obaarvad. There akin dlaordera have baan shown to ba reversible in aoaa of the animal and human studies. liver Effects FCB mixtures are capable of gaaarating important affects on liver tissue in animal models. These affects include; (a) increase ia liver weights; (b) histopathologic ehangas in livar including enlarged hepatocytes, deposition of fat droplets in tissue, and tissue necrosis and cell death; (c) enlargement of the liver; (d) the occurrence of adsnoflbrosls, a benign lesion; end (e) an Increase in proliferative lesions of the liver, including Increases in hyper plastic foci and nodular hyperplasia. These Increases in proliferative lesions appear to Increase in severity with increasing degree of chlorlnatloo of the FCB nlxture, as seen in the greater potency of Aroclor 12(0 ever Aroclor 125*. Tbs affects on liver tissues era generally reversible at lovar doses but trend toward Irreversibility with increased dosing. Gastric lesions When tested in monkeys end rodants, FCBa show species specificity with respect to a toxic effect on the gastric epithelium, or stomach lining. In the monkey, ulth Aroclor 1242, chore is a mucous conversion of the gastric epithelium that can best be described as a dysplascie growth pattern. This growth pattern does got constitute a neoplastic transformation. In rodants, PClo do not evoke this effect on the faatrlc mucosa. Cerclnotsnsaie A sliaabla volume of animal experimental work has been dlracted toward chronic studios in several speciest the mouse, tha rat and the dog. There are sone areas of interpretation that are still under debate by the investigators -concerned, but in the opinion ef the Category I team the following represents conclusions that can be drawn -from presently available data; (a) tha evidence on carcinogenicity is negative for gastroIntestInal carcinoma and bladder 222 HONS 026263 carcinoma In rati and hepatocellular carcinoma In tha dog; (b) aoma atadlaa hava raportad an lncraasa In hapacoealiular carcinoma In mica and rati axpoaad to commercial PCBi chronically, vharaai other atudlei have afforded negative raiulta. Reproductive Effect! Experimental work irlth commercial PCS mixture! and animal modala baa bean directed to tha reprodueclva procaaa itself, to taratoganasla in the offaprlng, and to poialble fetotoxlclty. In moat teat ipecles PCSa produce deleterious effect! on reproduction at high dosage*. In tha female monkey at relatively high dosages, difficulties are encountered In conception. In implantation of the fertilised ovum, and in carrying the fatus to term. Similarly doaed, tha male monkey does not transmit these difficulties to the reproductive process. The potential for PCB-lnduced teratogenesls, or production of defects in the embryo or fetus, has been investigated in several species of teat animals. Most animal teats have yielded negative results when PCBs vara administered to pregnant females during the critical periods for organogenesis. Several possible exceptions to this statement might be noted. In mice, a single congener, 3,4,3',*'-tetr*chlorobiphoyl, Induced a behavioral defect ("valtslng syndrome") that may be related to an anatomical defect in the inner ear. PCBs in mice may produce some delay in implantation. In rata, no groan teratological changes vare observed, but soma alterationa in thyroid structure or function produced by PCB administration might fall under the definition of t*Tata. In dogs end In evlne, no teretologlcal effects vara noted at lover PCB doses; however at the highest doses of PCB in the feed, at which the dam euffera marked reduction in food consumption and is therefore nutritionally deficient, there are dose-related teratologleal effects in the offspring. In monkeys dosed vlch PCBs, there are no doae-releted abnormalities in the offspring, but they are smaller in size. In test animals, l.e. rats, dogs and rabbits, PCBs are fatotoxlc whan administered ae relatively high doses to the pregnant female. Immunocompetents With respect to potential PCB effects on lsmunocompeeance In test animal systems, some observations are pertinent to the assessment of probable hasard in the human. It has bean observed In ehleks that if the PCB doeages are high enough, atrophy of tha splenic pulp and necrosis of tha lymphoid system can be demonstrated. Similar indicators of change in immunocompetents can also be produced in ducklings, mice, and monkeys as a consequence of sever* ehange in nutritional etatus. High doses of PCBs leading to marked reduction of food lntaka and utilisation can lead to such changes in nutritional status. These considerations load to two general conclusions relative to the PCB-exposed humansi (a) as projected from animal studies, if the PCB dosage is high enough to lead to general toxicity in the human (e.g., decreased food lneaka, fall in body weight, fall in weight gain), Immunosuppression nay be demonstrable secondary to malnutrition; and (b)>*t lower dose levels, there la no likelihood of significant Immunosuppression. 223 HONS 026264 Porphyria Porphyria, or dapoaitlon of porphyrin-darlvad plgnant* la lt*ar and urloa, baa baan obaarwad In axparlaantal anlaala auch aa cba rac and rabbit doaad with PCBa. Tha affact baa a dalayad onaat, and aapaclally in tha rat, aaaaa to taka plaea by naohanlaaa dlffarant from thoaa amployad by known chamlcal porphyria-producara In tha buaian. MONS 026265 REFERENCES Abrahansou LJ, Allan JR. 1973. Tha biological raaponaa of infant nonhunan prlnataa to a polychlorlnaead blpbanyl. Environ. Raalth Farapect. Exparlnental Iaaua Ho. 4181-86. Allan JR. 1979. Reproductive affacta of haloganatad aromatic hydroearbone on nonhuann prlaataa. Ann. H.T. Acad. Sciancaa 320:419-423 at al. Allan JR. 1973. Raaponaa of tha aonhuaan prlaata to polychlorinated blpbanyl axpoaura. Fad. Froc. 34:1675-1679. Allan JR, Abrahaaaon LJ. 1973. Morphological and biochemical changaa In tha liver of rata fad polychlorlnatad blphanyla. Arch. Environ. Contaa. Toxicol. 1:265-280. Allan JR, Abrahaaaon LJ, Rorback OB. 1973. Biological affacta of polychlorlnatad blphanyla and trlphanyla on tha aobhuaan prlaata. Environ. Raa. 6:344-354. Allan JR, Caratana LA, Baraottl DA. 1974. laaldual affacta of ahort-tara low-level axpoaura of aonhuaan prlaataa to polychlorlnatad blphanyla. Toxicol. Appl. Fhamacol. 30:440-451. Allan JR, Borback Dl. 1973. Folyehlorlnatad blpbanyl- and trlphanyl- lnduead gaacrlc aucoeel hyparplaala In prlaataa. Sclanca 179:498-499. Allan JR, Rorback DB, Ban IC. 1974. Tlaana aodiflcatlona in aonkaya aa ralatad to abaorptlon, dlatrlbutlon and axcratlon af polychlorlnatad blphanyla. Arch. Environ. Contaa. Toxicol. 2:86. Baraottl DA, Allan JR. 1975. Effaeta of polychlorlnatad blphanyla on raproductlon In tha prlaata. Fad. Froc, 34:338. Baraottl DA, Marlar RJ, Allan JR. 2976. Raproductlva dya/unction In rbaaua onkaya axpoaad to low lavala of polychlorlnatad blphanyla (Aroelor 1248). Fd. Coaaat. Toxicol. 14:99-103. Brucknor JT, Khanna KL, Corelah EH. 1973. Biological raaponaaa of tha rat to polychlorloatad blphanyla, Toxicol. Appl. Fhamacol. 24:434-448. Bruekaar JT, Ehanna EL, Coralah HH. 1974. Efface of prolonged lngeatlon of polychlorlnatad blphanyla on tha rat. Fd. Coaaat. Toxicol. 12:323-330. Brucknor JT, Ehanna KL, Cornlah HH. 1974. Folyehlorlnatad blpbanyl Induced alteration of biologic paraaatora In tha rat, Toxicol. Appl. Fhamacol. 28:189-199. Bnraa TV, Elabrough RD, Tlllaneuve EC, Jennlngo EV, Linder RE, Sovoeool GW. 19?A. Folyehlorlnatad blphanyla. Storage, dlatrlbutlon, excretion and recovery: liver oorpboiogy after prolonged dietary lngeatlon. Arch. Environ. Health 29:301-307. 225 MQNS 026266 Calaodra JC. 1976. Summary of toxicological studies on coaaerclal PCB. Proc. of tho National Conference on Polychlorinated Biphenyl*. EPA-560/6-75-004, pp. 35-42. Cl*ndr* JC. 1973. Summary of toxicological ttudlaa on comercial PCBa. Proc. of tba National Confaranca on Folychlorlnatad Biphenyl*. November 19-21. Chicago. IL. Dlkshlth TSS, Rockvood W, Abrahaa R, Coulaton F. 1975. Efface* of a polychlorinated biphenyl (Aroclor 1254) on rat taatl*. Exp. Mol. Pathol. 22:376-365. Earl FI, Couvllllon Jl, Fan loon EJ. 1974. Tha reproductive effects of PCB 1254 in baagla dogs and miniature *win*. (Abstract) Toxicol. Appll. Pharmacol. 29:104. Ecoblchon DJ Coasau AM. 1974. Comparative effects of comerclal Aroclor* on rat liver ensyme activities. Chen. Biol. Intetac. 9:341-350. Faith BE, Luster MI, Vo* JC. 1980. Effaces on laaunocoapetanca by chemicals of anvironaantal concern. Ievlevs In Blocheaical Toxicology. Hodgson, Band 4 Philpoe, ad. Heavier Publish, pp. 173-211. Cellert EJ, Wilson C. 1979. taproduetlva function In rats exposed pranetally to pesticide* and polychlorinated biphenyls (PCS*). Environ, lea. 18:437-443. Coldstaln JA at el. 1974. Ixperlaental hepatic porphyria Induced by polychlorinated biphenyls. Toxicol. Appl. Pharnacol. 27:437-446. Craan S. Carr JV, Falser KA, Oswald EJ. 1975. Lech of cytoganecic effaces in bone aarTov and spermatogonlal cells In rata traaead with polychlorinated biphenyls (Aroclor* 1242 and 1254). Bull. Environ. Contaa..Toxieol. 13:14-22. Craan S, Sauro FM, Friadaan L. 1975b. Lack of dominant lethality in rats treated with polychlorinated biphenyls (Aroclors 1242 and 1254). Fd. Cosaac. Toxieol. 13:507-510. Heddla JA, Bruce VR. 1977. Coaperlaon of tests for autagsnlclty or carcinogenicity using assays for spots abnormalities, formation of aleroouclel, and mutation in salmonella. Origins of human cancer. Hiatt HH at al., ads). Cold Iprlsg Harbor Laboratory, Cold Spring Harbor, NT. Vol. 4, p. 1549. Hooplngarner I, Samuel A. Rrause D. 1962. Folychlorlnatad biphenyl Interactions with tissue culture cells. Environ. Health Perspect. 1:155-158. Xatropouloe MJ, Felt I, Coulston P, Korte F. 1977. Histopsthology of low chlorinated biphenyls and haxachlorobenses* in feaele rhesus aonkays, Toxieol. Appl. Pharmacol. 41:173. 226 HONS 02626? Ito N, Nagasaki B, Aral M, Maklura S, Suglhara S, Hlrao K. 1973. Histopathologic studies on llvar tumorlgenasl* Induced In mica by technical polychlorinated blphanyl* and lea promoting aff*ct oa liver tumors lnduead by banzana haxachLorlda. J. Rae'l Cancer Inat. 31:1637. Ito H, Nagasaki H, Maklura S, Aral M. 1974. Hlaeopathological studlaa on Uvar tumoriganasis la rat* eraacad vlth polychlorinated biphenyls. Gann 65:343. Kapilngar KL, Faochar OE, Calandra JC. 1972. Toxicological studio* vlth polychlorinated blphanyla. Papar raad at PCS Confaranea, Quail Rooet Coofaraoca Cantar, Rougaaont, NC, December 20-21, 1971. Kapllngar ML, Faochar OE, Calandra JC. 1971. Toxicologic atudlaa vlth polychlorinated blphanyla. (Abttrace) Toxicol. Appll. Pharmacol. 19:402-403. Kimbrough RD. 1973. Induction of Uvar tumor* In Sharman atraln faaala rat* by polychlorlnatad blphanyl (Aroclor 1260). J. Rat'l. Canear Inat. 35:1433-1459. Kimbrough RD, Lindar RE. 1974. Induction of adanoflbroal* and hopatoma* of tha Uvar'In BALR/cJ mica by polychlorlnatad blphanyl* (Aroclor 1254), J. Nat'l. Canear Inn. 33(2):347-332. Kimbrough RD, Lindar RE, >ur*a W, Jaonlng* RW. 1973. Adanoflbroal* In tha rat Uvar -- vlth paraiatane* of polychlorlnatad blphanyla In adlpoaa tl**ua. Arch. Environ. Health. 27:390-395. Kimbrough RD, Lindar RE, Oalnaa TB. 1972. Morphological change* in Uvar of rat* fad polychlorlnatad blphanyla, light mleroacopy and uleraaervetur*. Arch. Environ. Health 23:334-364. Lavlnakaa GJ. 1911. Toxicity of Aroclor product* 1242, 1234 and 1260 to tha Uvar of albino rat*. Onpubllahad raaulta. October 14, 1981. Lindar RE, Galna* T3, Kimbrough RD. 1974. Tha affect of polychlorlnatad blphanyla on rat reproduction. Fd. Coamae. Toxicol. 12:63-67. Mark* TA, Klamel GL, Staple* RE. 1981. Influane* of *ymmatrlcal polychlorlnatad blphanyl laomar* on embryo and faeal development In mica. teratogenicity of 3,3',4.4',5,3'-haxachloroblphanyl. Toxicol. Appl. Pharmacol. 61:269-276. ' I. Maauda. 1979. Tranafar of varloua polychlorinated blphanyla to tha fetus** and offaprlng of mica. Pd. Coamae. Toxicol. 17:623-627. Maauda T, Kagsv* E, Kurokl H. 1978. Tranafar of polychlorlnatad blphanyla from mother* and fatuaa* and infanta. Fd. Coimat. Toxicol. 16:543-546. McMahon RE, CUna JC, Thompaon CZ. 1979. Aaaay of 835 t**t chemical* in tan taatar strain* using a nav modification of the Anas test for bacterial mutagen*. Canear R*- 39:692-693. ~ 227 MONS 026268 Narbonne JP. 1980. T1m eouria of induction of alcroaonal enryme* follovlng dietary adminietration of a polychlorinated biphenyl. Toxicol. Appl. Pharmacol. 56:1-7. National Car.car Inatltuta. 1978. Bloaaaay of Aroclor 1254 for poaalbla carcinogenicity. No. 38, nv. Safa S. 1980. Ambient vatar quality crltarla for polychlorinatad biphenyl*. Vaahlngcon, DC. O.S. Environmental Protection Agency. EPA 440/5-80-068, p. C-60. Schotny RS, Smith CC, Lopar JC. 1979. Non-autaganlclty for aalaonella of the chlorinated hydrocarbon* Aroclor 1254, 1,2,4-trlchlorobanaana, Mlrex and Repone. Mutation Rea. 68s125-132. Shlota R. 1976b. Eabryotoxlc affect* of polychlorioatad blphanyla (Ranechlor* 300 and 500) In rata. Okajlma* Pol. Anat. Jpn. 53:93-104. Squlra RA, Levitt MB. 1973. Clasilficatlon of (pacific hepatocellular laalona. Cancer tea. 35:3214-3223. Street JC, Sharma RP. 1973. Alteration of Induced cellular and humoral Immune reaponee by peeticldet and ehamleala of environmental concerns quantitative etudlea of lamunoeuppraealon by DDT, Aroclor 1234, carbaryl, carbofuran, and nethylparaehlon. Toxicol. Appl. Pharmacol. 32s387-602. Tllaon HA, Davie GJ, Mclachlao JA, Lucler GW. 1979. Tha effect* of polychlorinated biphanyl* given preoacally on the naurobehavloral development of mice. Environ. Rea. 18s466-474, Toeruek P. 1973. Effect of PCI on tha developing mouee. Chamoaphere 2s173. Tllleneuva DC, Grant DL, Ehera K, Clegg DJ, laer H, Phillip* VE. 1971a. The fetotoxlclty of a polychlorinated blphenyt mixture TAroclor 1234) in the rabble and In the rat. Environ. Fhyalol. 1:67-71. Vo* JG. 1972. Toxicology of PC8* for mammal* and for bird*. Environ. Health Parepect. 1:105-117. Voe JG, Beam* RB, 1971. Dermal eoxicley atudlee of technical polychlorinated blphanyla and fraction* thereof in rabble*. Toxicol. Appl. Pharmacol. 19:617-633. Vo* JG, Koeman JB. 1970. Comparative toxicologic atudy with polychlorinated biphenyls in chicken* with epeclal reference to porphyria, adama formation, llvar necro*l* end eiaaue realduee. Toxicol. Appl. Pharmacol. 17:636-668 Vo* JG, Hotanboom-Ram E, 1972. Comparative toxicity study of 2,4,5,2',4',5'-h*xacbloroblpheuyl and a polychlorinated biphanyl mixture in rabblte. Toxicol. Appl. Pharmacol. 23:363. Uyndham C, Davenish J, Safe 8. 1979. The in vitro metabolism, macromoleculer binding, and bacterial mutagenicity of 4-chloroblphenyl, a model PCB aubatrata. Re*. Commun. Chen. Pathol. Pharmacol. 15:563-570. 228 HONS 026269 PCBa 1 STROCmE-ACTlVITT RELATIONSHIPS S. Safa, L. Robartaou, T. Savyar, S. Bandlara, L. Safa Dapartaant of Phyaiology and Pharmacology Collaga of Vatarlnary Madlcinn Tcxaa AAM Unlvnralty Collaga Station, TX 77843 A. Parklnaon Dapartnaot of Blochaalatry and Dnig Matabollaa Hoffnano-La Rocha Inc Mutiny, NJ 07110 K. A. Caapball . Cualph-Vatarloo Contra for Graduatt Work in Chaalatry Oapartaant of Chaailatry Dnlvaralty of Cualph Gualph, Ontario R1C 2W1 M. Hullin Envlronaantal Protaetlon Aganey Croaa Ila Laboratory Groaa Ila, MI 48138 229 HONS 026270 tmODOCTIOM Polychlorinated biphenyls (PCJa) ara readily ayurhesiitd by the direct chlorination of biphenyl uelog iron aelta and/or iodine ee the catalyst. The degree of ehlorlsatloo of the reaultlng mixture la controlled by the relative amount of chlorine uaed In the reaction and aoac coaaarclal products ere graded and sold by their weight per cent chlorine concern (Butslnger at el. 1974). PCBs have been widely used in Industry as heat transfer fluids, hydraulic fluids, solvent extenders, fleas retardaocs,'organic diluents and dielectric fluids. The unusual Industrial versatility of PCBs Is directly related to their physical properties which Include: a) reslstanca to aclda and bases, b) compatibility with organic aacarlals, c) raslstanea to oxidation and reduction, d) excellent electrical lnaulaclng properties, a) thermal stability and f) non-flammability. Unfortunately, these physical properties coupled with their widespread use, relatively low acute toxicity and improper disposal haa resulted in the contamination by PCBs of every component of the global ecosystem (Butslnger at al. 1974, Landrigan 1980, Jensen and Sundetrom 1974, Hessarmann at al. 1979, Pujlware 1973). Moreover, the lipophilic nature and perslstance of PCBs also contributes to thslr high bioaccumulation potential and their blomagnlfleafIon In higher trophic levels of the food ehaln (Safe 1980). PCBs raslduaa ara routinely detected In fish, wildlife and human adipose tissue, blood and breast milk, C0HH0M 4MALTTICA1 AMP TOIICOIOCICAL MOSLEMS The analysis of KB residues In environmental samples present numerous problems which Include the following: 1. KB formulations and environmental isolates are highly complex mixtures of Isomers and congeners and 2. the composition of PCls obtained from diverse environmental matrices are highly variable due to the different rates of biodegradation, vaporisation, photodegradation and chemical degradation of the individ ual PCB isomers and congeners (Safe 1980, BallschsUter at al. 1978, larvey and Seelnhauer 1974). There era 209 possible PCB Isomers and congeners and most of the eonmerclal mixtures and environmental samples exhibit a multitude of peaks and this Is particularly evident la the capillary gas chromatogram of the commercial PCB, Aroclor 1280. Xt has baea common practice to quantitate environmental PCBa by comparing their peeked column gas chromatograms with the patterns exhibited by known amounts of individual coamerelal PCBs or nlxtures of these fomnlatloes (Sawyer 1978, Albro at el. 1981). This nethod relies on pattern matching using specific peaks for quautltaclon. It Is dear that if tha gas chromstogreme of the environmental PCB residues eannoe be "notched" with an appropriate eocktall containing known amounts of the coanerelal formulations than quantisation Is not possible. Comparable analytical problems have been encountered in the analysis of other complex halogenatsd aromatic pollutants, and polynuclear aromatic hydrocarbona (NBCC 1981). These prbblems experienced In PCB ahblysls by gas chromatography (GC) can be resolved in the following way: 230 HONS 026271 1) Th* eampla* aust b analyzed by high raaolutlon capillary gaa chromatography to anaura the maximum (operation of the laoaera and eonganara, 2) all 209 PCS laoaara and conganara muat ba available aa analytical acandarda for the Identification and quantitation of the Individual eoaponenca In the analyte. In collaboration with Dr. M. Hullln and aavaral covorkara, conalderable prograaa ha a bean mad* on the aynthaaia and characterization of all 209 PCI# (Tabla 1) and thle haa raaultad in the poaltlva identification of the Individual PCS component* praaant In commercial mixtures and In environmental samples. Baaed on the numbering schema propoaed by Bellachalttar and Zall (1980), Figure 1 llluatrataa ehe Identification of the Individual PCBa In Aroclor 1260 ualng the aynthatlc PCB atandarda (Kullin at al. 1981). Moat revleva dleeuaalng the biologic and toxic effecta of commercial PCBa note that the relative potanelea of thaae formulation# depend# on their degree of chlorination and alao on the apeclaa being uaed In the atudy (Ecoblehon and Coaeau 1976, Kimbrough 1976). The reaulta clearly auggeat that einco the toxlcltlea of ehe mixtures are variable, then ehe activltiee of the individual componenta of thaae alxturae may alao be different. Xeaaarch on the comparative biologic and toxic effecta of PCBa haa clearly illuatrated the marked differencea in the comparative toxlcltlea of PCI eongenera aa noted below. 1) 2,3',4,4',3-Pantaehlorobiphenyl la highly eabryotoxlc to chlckene where** 2,2',3.5'-t#tra and 2,2',4,4',5,5'-h*xachloroblph*iiyl are relatively non-toxic (Ax and Banaen 1975). 2) 3,3',4,4'-T#traehloroblphenyl la highly eoxle to thaaua macaque* where** the 2,2*,5,5'-ioa*r elicit* no ellnlcal effecta (KcHulty at al. 1980). 3) The toxic laalona eauaed by 3,3',4,4'.S.S'-haxachloroblphanyl In male mice ware much bot* aavara than thoaa obeervad after adainlatratlon of the 2,2',4,4',6,6'-, 2,2',4,4',5,5'- and 2,2',3,3',6,6'-bex*chloroblph*nyl* (Biocca et al. 1981). 4) A comparlaon of the toxicity of 2,3,3',4,4'-p*nta- and 2,3',4,4'tatrachloroblphenyl ahowad the former congener to be highly active and the latter compound to be inactive (Yamamoto et al. 1976). 3) Chronic administration of 2,2',3,3',4,4'- and 2,2',3,4,4',5'haxaehloroblphamyl eauaed an accumulation of liver porphyrin* where** adainlatratlon of tha 2,2',4,4',3,3'-hoxachloroblphenyl had no efface on hepatic porphyrin level* (Stonerd and Grieg 1976), .a 6) 3,3',4,4',3- and 2,3',4,4',3-pantachiorobiphanyl end 2,3,3',4,4',3haxachloroblpheuyl eauae thymic atrophy in rata vharaaa the 2,21,3*3',4,4',5- awL 2,2',3,4,4',3,3'-h*pt*chloroblph*nyl# were not toxic (Yoshlhara at al, 1979). 231 HONS 026272 Table 1. Prograaa la tha Sjmthaala of PCB laoaara and Conganara Cl Ccoupa 1 2 3 4 3 6 7 8 9 10 Total No. of laoaara 3 12 24 42 46 42 24 12 3 1 209 No. Synthaalaad or Avallabla 3 9 IB 40 42 40 24 12 3 1 192 232 MONS 026273 Fi|ur 1. High resolution glass capillary gaa chronatograa of Aroclor 1260. 233 MONS 026274 It If evident that aubtlf changes In tha structure of PCBs can ratult In marked dlffaranett In toxlcltlas. Thus, an assessment of tha potential long tana affects of PCBs on biological lystsas will dapand on tha concantratlon of tha Individual toxic conganars which ara praaant and which parslat In tha anlnal apaclas. This nacaaaltataa not only a high rasolutlon qusotitetlve and qualitative analysis of PCB residues, but prior knowledge concerning the potsotlsl or actual toxicides of the Individual PCBs. PCBs AS ISDOCERS Of HEPATIC MICROSOMAL DRUO-HCTABOLIZING EHZTKES One of tha Boat characteristic biologic affects of tha coanerclal PCBa Is thalr Induction of numerous hepatic microsomal drug-metabolising enzymes in rodents and particularly tha cytochroae P-450-dependent aonooxygenasas (Ecoblchon and Coaaau 1974, Alvaras and Kappas 1977a.b, Alvaras at al. 1973, Ryan at al, 1979), Administration of these mixtures results in increases in nicrosoaal aniline hydroxylase, banso(a)pyrane hydroxylase, biphenyl hydroxylasta, haloblphenyl hydroxylases as wall as numerous 0- and 0-dealkylases. The potent Induction of these cytochrome P-430 dependent aonooxygeneses by Aroclor 1254 has resulted In tha widespread use of PCB-lnduced alcrosoaes as the source of metabolic activation for chemicals In the Ames tact for their baetarlal muta genicity (BoUetein at al. 1979). Inducars of microsomal aonooxygenasas ara classically subdivided into two main catagorlas (Connay 1967, Lu et al. 1976, Sladek and Kannerlng 1976). One class la typified by phenobarbltone (PB) and the ocher by 3-methylcholanthrane (HC). PB-type Inducers anhanea microsomal N-damethylasa (2-3-fold), RADP8 cytochroae P-430 reductase (2-3-fold) biphenyl-4-hydroxylasa (2-4-fold) and aldrin apoxidase (3-3-fold) uharaaa MC-type Inducers enhance microsomal benso(a)pyrene hydroxylase (6-12-fold), 4-cbloroblphanyl hydroxylase (8-10-fold), athoxyrasorufin O-deethylase (up to 30-fold) and biphenyl 2'-hydroxylasa (8-12 fold). The preferential induction of specific mouooxygeneses by MC- and PB-eypa Inducars coupled with tha different electrophoretic and spectral properties of their Induced microsomal ensyma preparations elearly differentiates between the two classes of inducars. Tha ensyma induction properties of Aroclor 1254 and other conmerclai PCBs do not resemble that of either PB or HC but era similar to the pattern produced after eo-adalnletratlou of PB a MC (Alvaras and Xappas l977a,b, Alvaros at al. 1973), Mora recant seudias have shown that Aroclor 1234 induces cytochromes P-450b and ?-430c which are also the major lsosymes induced by PB and HC respectively (Kyan ae al. 1979). The mixed-type induction activity of Aroclor 1234 was presumed to be due to tha individual PB- and MC-type Inducars present in this mixture. PCBs AS MICROSOMAL HOHOOITCPtASB EM2THE IMPCCOtS: ETFECTS OF STRPCTURK OH ACtlVITT PB-Tme Inducars Initial studies suggested that PCBs which are PB-type Inducers must contain at least two sera end two ortho substituents| any further substitution changes only the quantitative nature of the induction activity (Poland and Clover 1977, Goldstein at ax. 1977). Thera is ample evidence for these correletlon aod some of the aoit active PB-type Inducers are 2,2',4,4'-tetra-, 2,2',4,4',5,5'haxa- and 2.2',3,3',4,4'-hexachloroblphenyl. However it has 234 HONS 026275 subsequently been ahown that PCS eooganar* which contain ona or no para subatituants, ara alao PB-typ* lnducara (Parkinson at al. 1980b). Soaa of thasa compound* Include 2.2' ,3,3*,5.3'-hax*ehloroblph*nyl, 3,3'-dlehlotoblphanyl, 2,3,3',3,3',6-haxachlorobiphanyl and 2,3,3',4',5,6-haxachloroblphanyl. Tha requirement for at laaat two ortho ehloro aubstltuanca eonflieea with tha Taportad activity of 4,4'-dlehloroblph*nyl and 2,3',4,4'-eatr*chloroblphanyl at P8-typa lnducara (Taaamoto at al. 1978, Parkinson at al. 1980b). Tabla 2 Illustrates tha structurally diverse (roup of PC>s which ara classified as PB-typa loducera and It la svldant that thara ara no apparant atructuraectlvity rulae for PCS* which anhancs this pattern of antyma activity. MC-Ttp* Inducers In 1977, two groups (Poland and Glover 1977, Goldstein et al. 1977) published papers outlining structure-activity rulas for PCBs as inducers of eytochrome P-448 dependent aonooxygensaes (or aryl hydrocarbon hydroxylase, ABB). Their data Indicated that PCBs which exhibit this type of Induction activity must be substituted In both para positions and at least one aeta position on both phenyl rings and not contain any orthochloto aubatltuenta. Tha three coapounda defined by tbaaa rulas, namely J,3k,4,4'-tetra-, 3,3',4,4',3-panta- and 3,3',4,4',3,3'-haxacbloroblphanyl ara all poeant ABB lnducara and ara approxlaate laoataraoaars of 7,3,7,8-tetracblorodibeniop-dioxin (TCDD), the aoat toxle haloganatad aroaatic chaalcal and aost potane Tnducar of alcroaoaal ABB (Poland at al. 1979). Ie was apparent that tha proposed structure-activity rules did not explain tha mixed-type Induction activity of tha coaaerdal mixtures sines tha three ABB lnducara ara only trace components of tha active conaarclal PCBs (Jentan and Sundatroa 1974, Bsllechalter and Ball 1910, Poland at.al. 1979). Por this raaaon, tha affect* of structure on tha activity of PCBs as ABB lnducara wars raavaluatad using a aarlaa of highly purified synthetic PCB congeners (Parkinson at al. I9fl0b, Parkinson and Safa 1981, Parkinson at al. 1980a,c, Parkinson at al. 1981a,b, Sawyer and Safa 1982). Kaoulraaants for Para Subatltuanta Tha requirements for para subatltuanta vara tasted using a aarlaa of PCB conganart with one, two and taro para aubseltuants and in all caaas PCBs which wars ABB lnducara all contained two para-chloro groups. (Figure 2). Beauirtasnts for Mata Bubstltuanta Tha requirement* for --ta subatltuanta vara evaluated by tasting all tha posalbl* aata-aubatltutad His derived troa 4,4'-dlchloroblph*nyl, a KB vhlch contains tha required two para-chloro groups. Plgure 2 denote* the flvs possible coapounda In this aarlaa and only tha l,J',4,4'*tatra-, 3,4,4',5-tatra, 3,3',4,4',5panta- and 3,3',4,4',S,5'-h*xachloroblphenyls vara ABB lnducara. Thu* on* new PCB congener, namely 3,4,4',3-tatrachloroblphenyl Is added to tha original list of ebraa ABB loducar* and It la apparsne that two mat* substituent* sre required for KBs vhleb induce ABB, however they ean be positioned on both or tha saaa phenyl ting. 235 MGNS 026276 Tabla 2. PCBs a* PB-Typa Inducers Conganar Nuabar of Cbloro Substituent Ortho Mata Para Reference 6,4'-dlchlorobiphetyl 2,3,3',4,4-pentachloroblpheojrl 7,2',4,4'-tatrachloroblphanyl 3,J'-dicblotobipheuyl 2,3,3',*,3,5'-bexachloroblphanyl 0 1 2 0 1 0 1 0 2 4 0 Goldataln t al. 1977 2 Parkinson t al. 1980b 2 Goldataln at al. 1977 0 Goldataln at al. 1977 l Parkinson c al. 1981a 236 MONS 026277 Fl|ur 2. Effects of para and --ca aubatituanta on cha activity of FCBa aa AHH inducaraj acciva (+), inactlva (-). 237 MONS 026278 Requirements for Ortho Substituents A logical approach for evaluating tht affactt of ortho-chloro substi tuents on eha activity of PCBa aa AHH Inductn la to aynchaslxa and taat all tht poaalbla aono- and dl-ortho-substltuted analogs of eha four MC-type Induetri. A suaaary of Chata compounds la given In Flgura 3. All the aono-ortho aubatltuead analogs vara mixed-type lndueara In tha laaaturt aala Wlatar rat and vara lndueara of AHH In rac hepatoma H-4-II-E calls In cultura. Moreover aany of thaaa coapounda, Including 2,3,3',4,*'-penta-, 2,3',4,4',3panta-, 2,3,3',4,4',5-hexa- and 2,3,3',4,4',3,5'-heptachloroblphenyl have bean Identified as coaponants of eha coaaarclal PCBa (Jansen and Sundstroa 1974, Ballachaltar and 2all 1980, Poland at al. 1979). Tha affacta of tvo ortho chloro aubatltuants alghe ba expected to have a aora pronounced efface on tha population of tha coplanar confornars and, hance, on thalr activity aa MC-type lndueara. Previous reports in eha literature suggested that PCBa substituted In at least one aaea and para position on each phenyl ring and at tvo ortho posltlona vara strictly fo-typa lndueara as avldancad by tha activity of tha 2,2',3,3',4,4'- and 2,2',4,4',3,5' haxachloroblphanyl lsoaers (Poland and Clover 1977, Goldstein ae al. 1977). Contradictory reports have suggested that thaaa lsoaers also poasaas MC-type activity (Seonard and Crlag 1978, Alvaras 1977). A subsequent ralnvaselgatlon of tha activity of 2,2',4,4',5,S'-hexacbloroblphanyl as a aleroeone1 antyaa Inducer has ahovn that tha MC-type actlvley of a 992 pure coaaarclal saapla (prepared by tha VlUunn coupling aethod) vaa due to contamination with eha highly active 2,3,7,8-tstrachlorodlbanaofuran (TCDP) (Goldstein at al. 1978). In attaapt to rasolva conflicting raporta In tha literature, tha affects of 2,2',3,3',4,4'-hexa-, 2,2',4,4',S,5'-hexa- and 2,2',3',4,4',5-hexachloroblphenyl- on tha hepatic alcrosoaal drug-netabollxlng enzymes vara evaluated In tha laaatura aala rae (Parkinson at al. 1980c). By comparison vlth eha affects of the classical anzyaa lndueara, PB and MC, 2,2',4,4*,5,5'-hexaehloroblphenyl vas classified as a pure PB-type inducer. In contrast, 2,2',3,3',4,4'-haxachloroblphanyl, irrespective of its synthetic route, exhibited PB-type and weak MC-cypa characteristics; tha aost prominent feature of the latter being a 7-fold Increase In 4-chleroblphenyl hydroxylase activity. P.l'.S'tl.A'tS-Baxachloroblphsnyl also resembled a alxad (PB * MC)-cypa lnducar although lta MC-type characteristics vara aora pronounced than those of 2,2',3,3',4,4'-hexachloroblphanyl. Since these three baxachloroblphanyls am dl-ortho substituted derivatives of tha pure MC-type lnduear, 3,3',4,4'-tatrachlorobiphenyl. It la clear that tha presence of tvo ortho chlorines does not necessarily abolish MC-type character. A detailed atudy of all the dl-ortho substituted analogs of the four AHH lndueara Indicated that, in addition to the tvo PCS lsoaers noted above, 2,3,4,4*,3,6hexa-, 2,3,3',4,4',6-hexa- and 2,2',3,3',4,4',3-hsptachloroblphenyl vara also mixed-type inducers (Parkloson at al. 1981b). Therefore the PCBa vhieh induce MC-type activity aust ba substituted at both para positions, at least tvo mats positions (but not necessarily on dif ferent phenyl tints) and can also contain one ortho substituent; the addition of a second ortho enioro substituent to the nucleus of ao MC-type Inducer does not necessarily ellalnate eha qualitative aspects of this activity If one of tha ortho substituents la part of a 2,3,4-trlehlorophenyl substitution pattern. 238 . MONS 026279 Cl Cl Cl Cl 2 3)' 4 4' 2 3' 4 4' 5 2 2' 4 4' 5 5' 2 2' 3 3' 4 4' 2 2' 3 4 4' 5* 2 3 3' 4 4' Cl 2 3 3' 4 4' S 2 3' 4 4' 5 J' 233' 44' J' 2 3 4 4' 5 2 2' 3 3* 4 4* 5 2 2' 3 4 4' 3 5' 2 3 3' 4 4* 5 6 2 3 3* 4 4' 5' 6 2 2' 3 3' 4 4' 3 5' 2 3 3' 4 4' 5 5' 6 2 3 4 4* 5 6 2 2' 3 4 4' 5 2 3' 4 4' 5' S Flgura 3. Mono .and di-ortho subatituttd analog* of 3,3,,4,4,-t#tra-3,4,4,,3-tatra-. J.J'.l.d'.S-panta- and 3,3' ,4,4*,5,S'-haxachloroblph#nyl. 239 HONS 026280 PCBs Which Do Hoe Induct Microsomal Monooxygenasea It has also bean shown that several PCBa containing oat to thrae chlorine toss or biphenyls chlorinated on only ona of cba phenyl rings ara vaak or lnaetlva aa Indue a r of eytochrona P-450 dependent monooxygsnases (Poland and Glovar 1977, Goldstein at al. 1977). Thalr lack of activity may ba dua to thalr more rapid rata of metabolism and claaranca. Soma of tha more highly chlorlnatad oeta-dacachloroblphanyla ara alao lnaetlva (Poland and Glovar 1977, Goldscain at al. 1977, Parkinson and Safa 1981). Anoehar group of PCBa which ara lnaetlva aa AHH lnduears contain 3,4 or 3,4,5 aubatltuelon on ona phanyl ring with a 2,4,6- or 2,3,3-trlchloro subatltutlon pattarn on tha aacond ring (Parklnaon at al. 1981a, Goldataln at al. 1981, Parkinson 1980). Thass eongannra do accumulate In tha livar and contain ona or tvo ortho-chloro groups, hovavar, for soaa raason thay ara lnaetlva as a nicrotonal nonooxyganaaa lnducar and ara ralatlvaly nontoxic. Tha ratlonala for thaaa raaults is not apparent. PCBa 8 COBBELATIOH BBTWEEH AHH tHDPCTIOW. Ah RECEPTOR BIHDIMO AND TOXIC ITT Poland and eoworkars hava atatad "Tha potency of haloganatad aromatic hydrocarbons to Induct hapatle AHH activity eorralataa vary eloaaly to thalr eoxlc potency" (1980) and "that tha toxlelty of thaaa chlorlnatad aromatic hydrocarbons la aadlatad through tha raceptor; that is tha Initial avant la tha toxic action la tha atarao-apaclflc recognition and binding of thta bjt tha cytosolic binding special" (Poland at al. 1979). Tha evidence for this hypoth esis la based on tha biologic and toxle affaeta of savaral eonganarlc poly chlorinated dlbanso-p-dloxlns (PCDDs) and haloganatad aromatle compounds which are approximate lsosFareomers of 2,3,7,8-TCDD (Poland at al. 1979). Tha raeulta which support the proposed correlations are noted below) 1) There Is an axcallant correlation between tha toxic poeancy of PCDO conganars, thalr AHH Induction activity and affinity for tha Ah cytosolic receptor protein.(33, for example 2,3,7,8-TCDD Is highly toxic, a potent AHH lnducar and avidly binds tha Ah raeaptor protein; 2,7,-dlchloTodlbanso--dloxlo Is ralatlvaly nontoxle, does not bind to tha receptor protein or Indues microsomal AHH), 2) 2,3,7,8-TCDD and ralatad approximate lsostaraomars ara ralatlvaly toxic to tbs genetically Inbred CS7BL/6J "responsive mica" and nontoxic to the aonraaponslvs DBA/2J mica (Poland at al. 1979, Goldataln at al. 1981), 3) 2,3,7,8-TCDO and approximate lsostaraomars Induce AHH in tha respon sive mica hovavar only 2,3,7,8-TCDD Is active (ae higher concentra tions) in the DBA/2J mica (Poland at al. 1979, Poland and Glover 1980, Poland at al. 1976, Poland at al. 1974). PCBa ara an Ideal group of haloganatad aromatic chemicals to test and confirm the correlations proposed by Poland and eoworkars. Table 3 illustrates tha relative AHH Inducing activities of 3,3',4,4,-tatra-, 3,4,4',5-tetre-. 3,3',4,4',3-oanta- and 3,3',4,4',S,5'-hexachloroblphenyl and thalr mono-ortho chloro analogs. The results Illustrate that with few exceptIons there la an axcallant correlation between the relative potencies of tha individual PCB conganars as AHH Inducers In Vlstar rats and rat hepatoma 240 MONS 026281 Table 3. PCB* - Effect* of Structure on Activity PCB Congener* (nuaber) Relative Per Cat Activity ABB Induction Cell Culture* Bat* Receptor Binding MC-Typ* Inducer* (3) Mixed-Type Inducer* (1) PB-Typ* Inducer* (4) Son-Inducer* (3) 2,3,7,8-TCDD 100-1 0.3-2.4*10"3 Inactive Inactive 400 l&tctlvc ioctiv , 100-35 6-1.5 <0.3 < 0.3 2,500 241 MONS 026282 H-4-II-E call* (Sawyer and Safe 1982). and thair avidity for tha Ah eytoaolie receptor (Bandlcra at al. 1982). Although tha toxicitlaa of thaaa mono-ortho chlorlnaead conganora haa not baan ayatematlcally ttudiad thara ara raporta in tha lltatatura (Ax and Banaan 1975# Tamamoto at al. 1976# Toahlhara at al. 1979) which confirm tha comparatively high toxicity of tha 2.3,3',4,4'-penta-, 2,3',4,4',3-panta- and 2,3,3',4,4',5-haxachlorobiphanyl. Currant raaaarch in ay laboratory (Parklnaon at al. Ho data, Safa at al., unpubllahad raaulta) with fanatically inbrad mica lndieataa that adminiatration of tha mixad-typa PCS inducara to tha raaponalva C57BL/6J mica raaulta in ABB induction and thymic atrophy wharaaa thaaa compounda do not lnduea ABB nor ara thay toxic in tha nonraaponalva DBA/2J mica. Tha raaulta obtainad for tha PCBa complaaanc tha data raportad for lavaral PCDD conganara and aupport tha corralatlona notad by Poland and covorkara (Poland at al. 1979, Parklnaon and Safa 1981). Horaovar, it la alao apparant that PCB conganara which lnduea ABB activity ara toxic and tha raaldua lavala of thaaa apaclfie conganara in flah wildlife and axpoaad human* may ba important indicator* of potential ahort and long tarm toxicity. PCBa IB HUMAWSl PKELIMIWAAT STUDIES Figure 3 auaaarlaee all tha PCBa expected to exhibit MC-cype activity baaed on tha above guidellnea and a number of thaaa iaomera and conganara have baan identified in commercial PCBa and in human* (blood, adipoaa tiaaua end braaat milk) (Kuwabara at al, 1979, Takuahijl at al. 1978, Takuahiji at al. 1979). Aj noted In Table 4 there la relatively high concentration of XC-typa Inducara which preferentially bloconcantrate in human braaat milk, A recant atudy In our laboratory (Parklnaon at al. 1980d) haa ahown that the doaa effactlug half-maximal (ED,.) Induction of ABB activity for a raconatltutad braaat milk PCB aample vaa *12 vmol'kg" wharaaa the ED,, for Xanechlor 500, a eommarclal PCB mixture, vaa 87 umot'kg" , Thua tb*3increaaed. biological potency of braaat milk PCBa reflect* the preferential bloconcentratloo of 2,3',4,4',5-penta-, 2,3,J',4,4'-penta-, 2,3,J',4,4',5-haxa- and 2,2\3',4,4'.3hexaehloroblphenyl which are mlxad-typa inducara and aliclt variou* eoxlc raaponaea in the rat, chicken and raaponalva C57BL/6J mica (Ax and Banaan 1975, Tamamoto at al. 1976, Toahlhara at al. 1979, Parklnaon at al. (1982), Safa at al., unpubllahad raaulta.) 242 HONS 026283 Table 4. Braaat Milk and Reconatltuted Brttst Milk PCBa* PCS Structure GC Purity <Z> PCB Concen tration in Japaneaa Braaat Milk (Z) PCB Coneentratlon la Kaconacltuead Braaat Milk Mixture (Z) 2,4,4'-Trichloreblpheny 1 2,2*,5.3'-Tecr*chlorobiphan7l 2,4,4',3-Tetrachloroblphenyl 2,2',4,3,5'-Pentaehioroblphenyl 2,3*,4,4'S-Peatacblorobiphenyl 2,2*,3,4'.S.S'-Haxachloroblphenyl 2,3,3',4,4'-Pantacbloroblphenyl 2,2',4,4'.S.S'-Haxaehloroblpheayl 2,2',3.4.4',3'-Hexaehloroblphenyl 2,2',3,4',3,3',6-Heptacblorobipheoyl 2,2',3,4,4',3',6-Haptacbloroblphany1 2,3,3',4,4',3-Bexacbloroblphenyl 2,2',3,4,4',3,3'-Bepcachloroblphenyl 2,2',3,3',4,4',5-Baptacbloroblphan7l Total 98.3 99.0 99.0 98.3 99.0 99.0 99.0 99.0 99.0 98.3 99.0 99.0 90.0 98.3 8.4 2.2 2.0 * 1.3 19.1 2.8 2.8 0.9 11.8 1.2 2.3 0.3 3.3 * 0.3 15.3 , 0.4 15.8 0.7 3.2 t 0.6 1.6 t 0.3 2.1 0.7 3.1 0.7 2.3 0.3 95.3 9.2 2.2 20.1 3.0 12.8 *eb 3.9 16.3 16.9 3.3 1.7 2.3 5.7 2.3 99.9 *Xuvabara at al. 1979, Yakuthljl at al. 1978, Yakuahijl at al. 1979, Parklaaoa .at al. 19804. HO not given. 20 HONS 026284 ACKNOWLEDGEMENTS Tha financial aaalatanca of The O.S, Environmental Protaction Agency, Health and Walfara Canada, tha National Inatitutaa of Baalth (1-R01-ES0279801), tha Cancar for Comparative Madielna, tha Natural Sciencea and Englnaarln| Raaaarch Council of Canada and tha Taxaa Agricultural Experiment Station (Crant Nunbar 6376) la gratefully acknowledged. 244 HONS 026285 REFERENCES Albro PV Corbett ST, Schroadar Jl.. 1981, Quantitative characterization of polychlorinated biphenyl mixture* (Aroclora 1248, 1254 and 1260) by gas chromatography tiling capillary column!. J. Chroa. 205: 103-111. Alvaras AP. 1977. Stimulatory affacts of polychlorlnatad blphanyl* (PCB) oo cytochromaa P-4S0 and P-448 mediated microsomal oxidations. Ini Microaomes and drug oxidations. Ullrich at al. ad, Pargaaon Press. Oxford, pp. 476-483. Alvaras AP, Bickers DR, Kappas A. 1973. Polychlorlnatad blphanyl*! a new type of inducer of cytochrome P-448 in tha liver. Proc. Natl. Acad. Sci. (USA) 70:1321-1325. Alvaras AP, Kappas A. 1977a. Heterogeneity of cytochrome P-450a Induced by polychlorlnatad biphenyls. J. Biol. Cham. 252:6373-6378. Alvaras AP, Kappas A. 1977b. The Inducing propertlae of polychlorinated biphenyls on hepatic mouooxygenases. Clin. Pharmacol. Thar. 22:809-816. Ax RL, Hansen 1C, 1975. Effect* of purified PCB analogs on chicken reproduction. Poultry Sci. 54:895-900. Ballscbmlter K, Zell K. 1980. Analysis of polychlorlnatad blphanyl* (PCB) by glass capillary gas chromatography. Composition of technical Aroclor- and Clophan-PCB mixtures. Fresanlus 2. Baal. Cham. 302:20-31. Ballachmlter K, 2*11 N, Nau IJ. 1978. Persistence of PCBa in the acospbere: will soma PCB-componants never degrade. Chamosphere 7:173-176. Bandlera S., Safa 8, Okay AB. 1982. Binding of polychlorlnaeed biphenyls classified as either PB-, MC- or mixed-typa lnducars to the cytosolic Ah receptor protein. Chen-Blol. Interact. 39:259-277. Blocea M, Cupta BN, Chue K, McKinney JD, Moore JA. 1981. Toxicity of selected hexachloroblphenyl isomers in tha mouse. Toxicol. Appl. Pharmacol. 58)461-474. Conney AH. 1967. Pharmacological Implications of mlerosonal ensyma Induction. Pharmacol, lev. 19:317-366. Ecoblehon DJ, Comaau AM. 1974. Comparative effects of commercial Aroclor* on rat llvar anxyme activities. Chaa.-Blol. Interact. 9:241-350. Pujlwara X. 1975. Environmental and food contamination with PCBa in Japan. Sci. Total. Environ, 4:219-247. Goldstein JA, Haas Jl, Llnko P, Barren DJ. 1978. 2,3,7,8-tetraehlorodlbenxo-p-dloxln in coamarei-'ir available 99Z pure polychlorlnatad biphenyl isoaer ideuclixed a* tna inducer of hepatic cytochrome P-448 and aryl hydro carbon hydroxylas* in tha rat. Drug Matab. Dlspos. 6:258-264. 245 HONS 026286 Coid*t*in JA, Hickman P. Bargman H, McKinney JD, Valk*r HP. 1977. Separation of pur* polychlorinated biphenyl iaooar* into two type* of induc*r on th* bad* of induction of cytochron* P-450, P-448. Chan.-Biol. Interact. 17:69-87. Coldataln JA, Linko P, McKinney JO, Albro PW. 1981. Kark*d in th* inductive ffact* of tvo eyamatrlcal hexachlorobiphenyl* and th* corr*pondlng uneyumetrlcal laomar on hapatlc monooxygenaaaa. Bloch*a. Pharmacol. 30:1008-1011. Barvey GR, Stalnhauar WC. 1974. Ataoapharlc transport of polychlorobiphenyla to th* North Atlantic. Atao*. Environ. 8:777-782. Holl*t*ln M, McCann J, Angeloaanto PA, Nlchola WW. 1979. Short-t*ra t**t* for carelnogan* and mutagen*. Hut, t*. 63:133-226. Huttlngar 0., Safa S, Zltko V. 1974. Th* ehemlatry of PCB*. CSC Pros*, Inc., Ohio, J*na*n S, Sundatroa 0. 1974, Structuraa and l*vla of moat cbloroblphanyla in two technical PCB product* and In human adlpoaa tlaau*. Ambio 3:70-76, Kimbrough ID. 1974. Th* toxicity of polychlorinated polycyclic compound* and related chemical*. CRCCrit. law. Toxicol. 2:443-448. Kuvabara K., Takuabljl T, Vatanaba I, Toabida S, Eoyama X. Xunita N. 1979. Ineraaaa In tha human blood KB law*la promptly followlni ingaatlon of flab containing PCBa. Bull. Euwiroum. Contam. Toxicol. 21:458-462. landrlgao PJ. 1980. Ganaral population axpoaur* to anwlronmantal concantratlona of haloganatad blphanyla. In: Baloganatad biphenyl*, tarphanyla, naphtbalanaa, dlbanxodioxina and ralatad product*. Kimbrough ID, ad. Blaawlar/lortb Holland, Amatardam, p. 267. Lu ATB, Kunttman R, Connay AH. 1976. Tba liwar nlcroaomal bydroxycomponanc*. Front. Caatrointaat. Baa. 2:1-31. McNulty VP, Backar CM, Coray IT. If80. Chronle toxicity of 3,3',4,4'-T*trachloroblpbanyla in rhaau* maeaquaa. Toxicol. Appl. Pharmacol. 36:182-190. Mullln M, lavka 0, fafa l, McCrindla S, Safa S. 1981. Synthaala of octa- and nonachloroblphanyi laomara and eonganara and thalr quantitation in cotmtarclal polythlorlnatad blphanyla and Identification in human braaat milk. J. Anal. Toxicol. 3:138-142. Rational Raaaareb Council Canada. 1981. Polychlorinated dlb*nio--dloxln*: limitation* to tha currant analytical technique*. Publication NXCC 18376. Parkineon A. 1980. Induction of multiple forma of cytochroma P-430 by polychlorinated blphanyla. Ph.D. thaala. Onlvaralty of Cualph. Gueloh On*-. 246 MQNS 026287 Parklnaon A, Cockarllna R, Safa S. 1980a. Induction of both 3-oathyleholanthrtna- and phanobarbltona-typa elcroeonal anayma activity by a aingla polychlorlnatad blphanyl laonar. Blochaa. Pharmacol. 29:259-262. Parklnaon A, Cockarllna R, Safa S. 1960b. Polychlorlnatad blphanyl laoaara and congenara aa lnducara of both 3-aathylcholaothraaa- and phanobarbltonatypa mlcrovomal aoxyaa acdvlcy. Cham. Biol. Intaracc. 29:277-289. Parklnaon A.. Robartaon L, Safa L, Safa S, 1981a. Polychlorlnatad blphanyla aa lnducara of hapatlc alcroaomal entymea: atructura-actlvlty rulea. Chaa. Biol. Interact. 30:271-285. Parklnaon A. Robert ion L, Safa L. Safa S. 1981b. Polychlorlnatad blphanyla aa lnducara of hepatic alcroaoaal eniyaea: affacta of dl-ortho aubatltutlon. Chaa. Biol. Interact. 31:1-12. Parklnaon A. Robartaon L, Safe S. 1980c. Bepacic alcroaoaal enzyaa Induction by 2,2'.3,3',4,4'- and 2,2* ,3',4,4',5-haxachloroblphanyl. Ufa Sclancaa 27:2333-2337. Parklnaon A, Robartaon IV, Safa S, 1960d. Raconatltuced braaat ailk PCBa aa potent lnducara of aryl hydrocarbon hydroxylaea. Blochaa. liophya. Raa. Coaaun. 96:682-889. Parklnaon A, Robartaon I, Uhllg L, Caapball HA, Safa S. 1982. 2,3,6,6',5-pantachloroblphanyl: differential affacta on tha C57BL/6J and DBA/23 alea. Blochaa. Phanacol. 31:2830-2833. Parklnaon A, Safa S. 1981. Aryl hydrocarbon hydroxylaaa Induction and ita relatlonehip to tha toxicity of haloganatad aryl hydrocarbona. Toxicol. Environ. Chaa. 4:1-46. Poland A. Clover B. 1977. Chlorinated blphanyl Induction of aryl hydrocarbon hydroxylaaa activity: a atudy of tha atrueture activity relatlonehip. Mol. Phanacol. 13:924-938. Poland A, Clover E, Rende AS. 1976. Staraoapaclflc, high affinity blodlng of 2,3,7,8-cecrachloTodlbeneo--dioxla by hapatlc cytoeol. Evidence that tha binding apaclaa la receptor for Induction of aryl hydrocarbon hydroxylaaa. J, Biol. Chan. 251:2936-4946. Poland A, Glover I, Boblnaon JR, Hebert DV. 1974. Genetic expreealon of aryl hydroxylaaa activity. Induction of aonooxyganaaa aetlvltiee and eytochrone P-*50 formation by 2,3,7,8-tatrachlorodlbanso-p-dloxln in nice genetically nonreaponaivo to other aryl hydrocarbona. J. Biol. chan. 249:5599-5606. Poland A, Craanlao VP, Kendo AS. 1979. Studlea on tha nochanlan of toxicity of tha chlorinated dlbanso-p-dloxiaa and related conpounda. Aon. H.T. Acad. Scl. 320:214-230. Boland E, Clover. E. 1980. 2.,3,7,8-catraehlorodibento-p-dioxin: aagragatlon of toxicity vlth tha Ah locua. Mol. Pharmacol. 17:66-94. 247 MQNS 026288 Ryan DE, Thonaa PE, Eorxtnlovakl 0, Lavlo V. 1979. Separation and eharaetarlaatlon of highly purified foraa of Uvar Bicroeotaal cytochrome P-430 fro# raci traactd with polychlorlnatad blphanyla, phenobarbltal, and 3-ethyleholanthrane. J. Biol. Chan. 254:1365-1374. Safa S. 1980. Metabolite, uptaka, atorage and bloaceunulaclon. In: Halogenated blphanyla, naphthalanaa, dlbenaodloxlna and ralatad producta. Elabrough K, ad. Elaevler/North-Hollend, p. 77. Safa S, Parklnaon A, Robertaon l, Bandlara S, Campbell HA. Onpubllahad raaulta. Sawyar LO. 1978. Quantitation of polychlorlnatad blphanyl raalduaa by alactron capture gee-llquld chromatography: collaborative atudy. J. Aaaoc. Offle. Anal. Chaa. 61i282--291. Sawyar T, Safa 3. 1982. PCS laoaara and conganara Induction of aryl hydrocarbon bydroxylaee and echoxyreeorufln O-daathylaaa actlvltlaa in rat hepatoaa calla. Toxicol, lettera 13:87-94. Sladak HE, Mannaring QJ. 1976. Evidence for a nav P-450 bnaoprotaln In hapatle alcroaonaa fro# #athylcholantbrana traatad rata, llochaa. Blophya. Raa. Coaaun. 24:668-674. Stonard MD, Crlag J*. 1976. Dlffarant pattama of hapatle alcroaoaal anaysa actlrlty produced bp adalnlatratlon of pura hexachlorobiphenyl laoaara and haxachlorobanaana. Chaa.-Biol. Intaract. 15:365-379. Waaaamann M, Vaaaamann (, Cveoa 8, Millar IJ. 1979. World PCBa aap: atoraga and affacta In Map and hla biologic environment In tha 1970a, Ann. H.Y, Acad. Sci. 320:69-124. Takuahljl T, Watanaba 1, Kuvabara E, Toahlda 8, Xopaaa K, lara I, Runlta M. 1978. Long-tara atudlaa of tha excretion of polychlorlnatad blphanyla (PCBa) through tha allk of an occupatlonallp-axpoaad vorkar. Arch. Environs. Contam. Toxicol. 7:493-504. Takuahljl T, Watanaba I, Ruyabere K, Toahlda 8, Eopama E, Runlta H. 1979. Lavala of polpeblorlnatad blphanyla (PCBa) and organo chlorlna paatlcldaa In huaan allk and blood collactad In Oaaka prafactura froa 1972 to 1977. lot. Arch. Occup. farlrona. laalth 43:1-15. Taaaaoto.l, Toohlaura 1. Pujlta K, Taaaaoto T. 1976. Matabollc and toxicologic evaluation of 2,3,4,3',4-pantachloroblphanyl In rata and`mica. Chaa. Phara. Boll. 24:2168-2174. Toahlhara 8. Ravano E. Toablaura H, Eurokl 1, Maauda T. 1979. Toxicological aaaaaaaant of highly chlorlnatad biphenyl conganara ratalnad In tha Tuaho patlanta. Chaaoaphara. 8:531-538. 248 MONS 026289 DISCUSSION SUMMART Renata D. Kimbrough, K.D. Centar for Disease Control Center for EnviroomanraL Health Atlanta, CA 30333 Tha discussion vaa started by Dr..Kimbrough. She stated that It vould not really ba poaalbla In tha tlaa allotted to comment on all tha Information prasantad by Dr. Prlaaa. Because of tha Tuaho outbreak In Japan and a number of outbreaks of chick adama dloaaaa In tha Onltad Stataa vhara tha types of compounda chat Dr. Safa vaa talking about had caured lilneta In chlckana bacauaa of accidental contamination of food, a number of atudlaa ware atartad with PCBa and related compounda more than a decade ago. It vaa eoon evident that although on a quantitative baala there vaa a tremendoua difference In tha toxic affacca batween PCBa, chlorinated dlbanxodloxlna, and chlorinated dlbanaofuraoa. that there vaa a almllar reaponee on a qualitative baala that dlffarant animal apaclaa gave to theae compounda. In humana, ehloracna haa been reported after ezpoaure to chlorinated naphthalenes, chlorinated biphenyls, the chlorinated dlbenaodioxlno, and the chlorinated dlbentofurana. This came type of reaction can alao ba produced In animals, particularly the subhuman primate, but alao on the rabbit ear and In nude mica. Before It vaa possible to detect chlorinated dlbensodloxlns chem ically, tha rabbit ear test vaa used. Dr. Alan Poland (University of Wisconsin) has now developed a call culture system vhara he can produce a ksratltlnlng reaction In the cells, formation of keratin in tha hair follicle la the primary lesion of ehloracna, vhleh has nothing to do with juvenile acne. The other point that vaa made In the preceding paper vaa that there have been a number of studies on the carcinogenic efface of PCBa. Some have been cagatlva and oehers have been positive. If these studies are examined more cicsely, it becomes obvious why chase discrepancies exlse. In adequate cancer studies, a sufficient number of animals should be tested over their lifespan. Oogs, for inatsnee, live somewhere between 10 and IS years. It Is therefore not surprising that no canear was found In a 2-year dog seudy. In their bloasoay studies, the 0. S. National Cancer Institute Is now using 30 animals par group. Usually high dosage lavela art used because tha tatting.ayscam la not vary sensitive. It mutt first be determined whether there la a carcinogenic affect bafera extensive studies are dona with larger numbers of animals and maybe lover dosage levels. Rovever, dosage lavals should not ba ao high to have the animals loss ao much weight that there la a lot of added toxicity. In tha study by Kimbrough, at al (1975). only a dietary level of 100 ppm (mg/kg) of Aroclor 1260 vaa used. Translated Into Intake by tha rata, this means that thay In early life consumed a dally does of 12 mg/kg (because thay aat more when thay are young). The daily doaa dropped later to 3 mg/kg. Only female rata vara used In the study. Tha reason- only farnaUe wart studied vaa bacauaa In an earlier etudy with only a fav rata, one bladder eanear had been noted In a fenala rat. No bladder 249 MONS 026290 cancer v* found in th* itcond study. However, 26 of the rats out of 184 developed hepecoeellular carcinoses, and in addition, othar pathological change* vara noted in th* liver*, nasal?, neoplaatlc nodule* alao rafarrad to as nodular hyperplasia, and areas of altaraclon. In the controls, ona hepatocellular cardnosa vas found in 173 rats and there was little other pathology. Ron* of the liver tusors macaataalred, Because of th* confusion about the claaslflcatlon of rat liver tumors, at Dr. Fries* sentloned earlier, ocher pathologists vara asked to review this study, auch aa Dr. Bobert Squire, who va* director of pathology for the blo*** atudias of th* National Cancer Institute, and tvo other pathologists, Dr. Strindberg and Dr. Montali (all authors of the paper). It vas realised at the els* chat tos* sclancists did not consider neoplastic nodules to be tusort. Because of this, a workshop vas convened to classify liver lesion* (Squire and Levitt 197)). It vas the conclusion of the workshop that neoplastic nodules (nodular hyperplasia) ver* in fact new growths (tusors) rather than sosething that vas just proliferation of tissue. There is still argusent about vhether neoplastic nodules will in else develop lneo hepatocellular carcinoses or vhether they Just occur concurrently with carclnosa. After tfi* seeting, th* Rational Acadtsy of Sciences decided to go on* step further. They convened a group of pathologists chaired by Dr. Harold Stewart to develop a docusent on the histologic typing of liver tusors of eh* rat (Institute of Laboratory Anlsal Baaourcea, 1980). Another study seneionad as being negative by Dr. Fries* vas the National Cancer Institute study conducted with the Aroclor 12)4 in sale and fesala rats. There ver* 24 antsals per group, which was a sosevhet smaller group than the 200 that Dr. Klsbrough used. They gave three different dosage levels, 23, SO, snd 100 pps Aroclor 1254 to eh* rats over their lifespan. Son* rats died early bue about 20 or 17 rata per group survived. These rats also developed neoplastle nodules (nodular hyperplasia). Ac the low dose, neoplastic nodules vere seen la S sales. Ac eh* aid dose they war* praaent in 8, and at the high dose ehey ver* presenc in 12.In eh* feaales at th* low dote, they vers present in 6, at the aid doe* in 9, and at th* high dose in 17. One hepatocellular carciaosa vaa seen at the sld dose andtvo at the higher dose in salea. In the fesalea, no hepatocellular carcinoses vara seen. Conparison of this study with th* study of Klsbrough at si. (1975), suggest* that if a larger nusber of rats had been used this study sighe have been positive. In th* KC1 atudy, a nusber of exposed raes had adenocarcinoma of th* stomach, and this again vas not statistically significantly different aa vaa the incidence of th* hepatocellular carcinoses. At the low dose and th* slddla dose, NCI reported one gastrointestinal carclnosa each. Morgan at al. (1981) reexaslnad a total of 191 stomachs of rats fros this study of vhich 144 raes had been exposed to Aroclor 12)4. With different staining techniques, they found six adenocarcinomaa in exposed rats, and observed a high incidence ,pf squamous.matgplatla in the eplthelius of the itoeaeh. Squamous meeaplasla is associated with the occurrence of adenocarcinoma in the stomach. Th* rat strain used vas F344. Adenocarcinoma of the atosach in this rat strain is extremely rare, 1 in 1754 (<0.0)S). Therefore, they concluded chae th* like- 250 MOMS 026291 lihood of observing 6 adenocarcinomas In 144 rati waa less than 0.05 (P<0.05). Preston at al. (1981), found that PCBs promoted the Induction of cancer Initi ated by diethyl nltrosamint. This study suggests what was also mentioned by Dr. Safe that PCBs may be promoters. Initiators are thought to cause a primary Insult In the cell which in time will cause this csll to become canceroua. Promoters will Increase the possibility of an Initiated call to be transformed loco a cancerous cell. At the moment there Is a lot of uncertainty on how promoters should be regulated. One problem with the particular compounds we are dealing with hare Is that these compounds seem eo be persistent In the body for extremely long periods of time. Other promoters are not persistent. Perhaps a difference betvean different groups of compounds based on their metabolic behavior should be made. Dr. Prlass also alluded eo the effect of PCBs on the Immune response in animals, and Dr. Safe mentioned thymic atrophy and atrophy of the spleen. To bring this Into context, there ere different types of Immune response*. The humoral immune response Is consistent with the development of antibodies to common Infections and to vaccinations. The cell-mediated Immune reaponae, which If impaired, will, far Instance, decrease the ability of rejecting foreign tissue grafts. This sytta* is affacesd when the thymus snd splsen sre atrophic, and is very sensitive in the early development of children and in young animals. Reproduction studies ware also mentioned end again there (earns to be some variation in rasponse for dlffarsot apecias. Dr. Frleas slao pointed out that (>'s< animal specias art more sensitive to the toxic affects of these compounds than others. If tha entire group of compounds Dr. Safa vss talking about is sxsmlaed, it saams that th# mink, the guinea pig. the monkey, and tha quail ot the chick embryo are particularly ssnsltlve while the rst and the mouae are less sensitive. That is slso born out in' reproduction atudlas if different mixtures sre tested, Doeege levels which produce effects vary widely. There sre Just a let of differences in some of the older data that was reviewed by Cr. Frleas. Now we are beginning to aea why this la so by doing work with Isomers which hopefully will eventually put all of this information Into place and make ic seam more logical. Dr. Safa recommended that wa should perform studlss with teomars In anlaala. For a llfstlma study a lot of material is nasdad. Will thsaa isomers really be available to do Chess types of studlss? Furthermore, Dr. Frleas emphasised that some subhuman prlaatas may ba nor# or laas suscep tible to thasa compounds than others. Sines subhuman primatas art vary expan sive, doing a lot of studies in them would ba difficult. Host of tha carcino genesis studies are dona in rats and mice because historically wa have axparltnca with that system and because rodents have a abort lifespan. Dr. Safa talked about atructura activity relationship of the stereo isomers. It seems to be vary important whether these chemicals are planar or not. Jin KcKinney (N1EHS) la also working In ehls area. Ba has recently published a paper in which ha poloes out that co-polarity is alto vary important 'McKinney and Singh 1961). Co-polarity may change in different solutions and with it th. abount of ehemleal that la planar or almost planar can vary. Parhapa Dr. Safa nay wish to comment on thla. 251 0262^2 mqns It was then elated by Or. Frless that Dr. Safa had a very Interesting cellular test system In which the induction of arylhydrocarbon hydroxylase could be messursd without using the Intact animal system end for the chemicals which Induced arylhydrocarbon hydroxylase a dote response curve which went upward with does could be developed. As the doae was Increased even further, the Induction did not Increase proportionally. Inhibition of excess substrate la a well known biological event in certain enzymes, and Dr. Frless raised ehe question to Dr. Safe whether If you Increase the dose even further, would a point be reached where the Induction would decraase to zero? Would an excess of the active Isomer, as It were, become Its own tool for negating Its own effect? Dr. Frlese also raised the question whether in localizing these high Isoners In tissues, can node! compound be devised which, In Itself, has little biological activity but could block the receptors and block the synthesis points which are triggered by these Mhot" isomers. Dr. Safe then responded to chase questions. He seated that some isoners are hard to get. They would have to be made. For some, It would be difficult to maka large enough quantities for any eort of In vivo studies, but there are a few that are amendable to synthesis. In response to Dr. Frless' question about lass Induction at high dosage levels. Dr. Safa stated that this also occurred la vivo vith FCSs because the animals are dying. This is probably not due to an Inhibition of the enzyme. The cell Is killed or the animal is dying. Unfortunately, an overdose would not be protective. Dr. Mary Wolff then added Information about planarity. Dr. McKinney's (NIZHS) thesis Is that the achievement of eo-planarlty Is In contradiction to an energy barrier to free rotation around the biphenyl bond. He has sbovn that there Is virtually no rotation. The barrier Is low, within 3,4, 3',4'-kinds of materials end goes up when you add ortho halogens but the degree of co-planarlty Is proportional to the percentage of the time that the two rings remain eo-planar. It Is Important to polnc out the much greater toxicity of the 2,3,4 and 3,4 substituted lsomera but In terms of vhac Is present In people, the Japanese work has suggested that the proportion of FC3s to FCDFs In either Tusho or in their general population people Is fairly consistent with whet they were exposed to. tn breast milk samples, the ratio of FCSs to th.e FCDFs may then be In the range of one million to one, and of those PCIs by far the larger proportion, certainly greater than SOX are the 2,3,4 substituted kinds of compounds. What does that mean In terms of relative toxicity? Do the 2,3,4 aubstltuted FCls completely outweigh the potentldl toxicity of the FCDFs? The chromatograms from breast milk very closely resemble those of the 4,2,4',5' tatrachloroblphenyl peak on the gas chromatogram. There la also a 2,4,3',*' tatrachloroblphenyl. That Is the only Isomer of that series that Dr. Safe failed to mention In his talk. Dr. Safe reemphasized that polarizability was the point Dr. McKinney vat trying to make. He then stated that 2,4, 3',4' la Inactive ae moderate high deeea. Tht 4,3,4'- if inactive In the ebloro series because! of lack of polar izability. When one ortho la chlorine added, It Is also Inactive. Therefore, It makes no contribution. The 2,3,4 substituted biphenyls do predominate In breast milk and In human tissues snd some of those analogues art active. 252 MONS 026293 Compared to Aroclor and Kanechlor thara la a fiva-to-evenfold increase in activity. So, "Taa," thay ara eha onaa that can exart toxicity, and thara seen to ba more of than in human tlaaua. Humana aaam to preferentially concentrata thoaa kiude of chamlcala. Jorenaeostrom and Jensen confirmed that a long tine ago. It la poaaibla that the FCBo which ara preferentially retained ara alao more toxic. Dr. Muriel Llppman, ERNACO, Inc., added eommenca to the evaluation of the carcinogenicity data by Dr. Frlaaa. Failure to obtain a etatietically aignlfleant lncraaaa in cancer incidence in created anlmala compared to their controla la called a negative atudy or negative teat raault. One la temptad to conclude from euch a atudy, or several such studies, that eha agent tested is not a carcinogen. This conclualon is not necessarily true. We know that there sre many reasons why a carcinogen may yield a negative east resule. One such reason is the sensitivity or the power of the experiment to detect the affect. In ordar to have a 90S chance of deeecelng a significant increase in cancer incidence In the treated animals at a F SX, which is the customary laval adopted, a sample else of 25 animals per group (thae is, 25 anlsMls not only in the group but alive at the earliest tlma to tumor) would require thae 9 out of 23 of the animals show eha lesion, (in .this esse the lesion to hepatocallular carcinoma) as compared with 0 out; of 25 controls. If the group site is 25 anlmala per one sax per group, treated and control, there must be a difference of 33S over background in order to be significant at the SS level to have a 90S chance of detecting this effect. When the background else la smaller, (e.g., there are 18 animals per group) a higher percentage increase over background la required in order to detect a significant offset. The Caiandra study that was discussed by Dr. Frlass had at moat Id animals per group. The HCI study had 2d animals par group of such sex or less and the Ito study hsd 12 in each treated group and only 4 controls. These were therefore extremely insensitive studies. A negative result vlth such insensitive studies doss not permit us to distinguish whether them vas no carcinogenic effect or whether the study vas simply too Insensitive to detect the affect. In contrast, a positive result on a very insenaltlva study tells us some thing about the carcinogenic potential of the compound, that it in fact can be detected at all in such an insensitive study. In the Ito study, there vas s 41.6X increase in hepatocellular carcinoma with Kanechlor 300} it is just such an insensitive test. That should carry a significant amount of weight. How is there anything thst we can do to help to distinguish between the truly negative case or s case of insensitivity! I think thara ara at least two things that va can do. One is to look at tbs actual incidence of the hepatocellular carcinoma. Is it very high! Is it, in the experience of pathologists, well above background, and secondly, look at morphological changes that may eonstltuta pre-nalignant change. Both of these options are available .and can bn considered in the overall weight of evidence which I would like to summarize. In the case of hepatocellular carcinoma, there was one very sensitive test in rats, that of Dr. Kimbrough's, where she indicated that sha startad vlth 200 animals and in which 1(X of the treated animals, or 253 MONS 026294 26 out of 184, shovad hepatocellular carcinoma. This vts highly significant. Thsrs vss also a vary lnaanaltlva taae, which shoved to Inertias In ale*. This vaa in alca, tha Ito study, with 41.6t lneldaoca. So you hava a study In rata and a study in alca that indicarad earclno|aoidty. Evan In tha NCI Study, a vary lnaanaltlva study, thara was sons lneraaaa (8X) at tha 100 ppm diatary laval. If all of tha othar data on hapatoaa, neoplastic nodulas, nodulaa of hyperplasia and hyperplastic foci era examined, almost all of tha experiments showed, especially all those Insensitive axparlmanta, not only a significant lneraaaa In thasa laalons, but a doae-raiaced response In these laalona. Not all of tha coanunley of pathologists agree that neoplastic nodulas and similar laalons are pre-aallgnant and wa vould not Ilka :o hava tha whola Industry wipe out FCBa on such data; but tha fact rams Ins that thara ara strong data Indicating hepatocellular carcinoma with an adequate test as wall aa supporting positive Information with Insensitive tests. This shows that tha weight of evidence vary strongly favors the conclusion that thara la a carcinogenic afface. Tha same thing can be said for tha othar small studies. Collectively these studies raprasane a body of avldanca that is not as fuxxy as Or, Frias* suggests, but allows ua to draw a conclusion until such time as Individual Isomers which Or. Safe will be able to provide can be actually taatad. If an Aroelor or a Kaoechlor preparation contains 40 to 60 components and only soma of chase ara carcinogenic, then tha doses of tha mixture are not high at all sloes tha administered dosa la not indicated by tha admlnlstared level but containing all tha compounds is ooly that fraction of tha adminiaearad dosa composed of active agents. Slnea soma of tha more toxic components ara present only in trace quantities, this vould suggsst that If thay ate active at all they must be vary potant. Tha affective does la not only rnducod by bnlng a mixture and active and Inactive components; it Is also rsducsd by tha factor of dlffsrantlal metabolism and differential pharmacokinetics of thaso components. So the valght of ovldoncs favors tha conclusion that thasa compounds, these mixtures that have been tasted, ere Indeed carcinogens. Tha data needs to be refined to determine their potency. Dr. Fries# than diacussed several point* about the procedures by which anlmsl studlas conceived and directed at shoving carcinogenicity in animal models than bacome surrogates for risk asssssmant in humans and tha crltsria that art required. Thara is no questioning the point that an adequately designed experiment Is nseassary in which the right numbers of animals erastad with sufficient number* of dona* followed over a sufficient period of time, approximating a considerable chunk of tha Ufatima, than followed toward tha end of thn lifespan to gat good dosa response numbers for malignancies is terribly important. Thara la no evading eh* point that this kind of Information should ha established for boeh aaxaa of animal*. Therefor*, on* would truly want an Internationally well-designed experiment with an agraad upon Aroclor, strain, dosing schedule and protocol to follow an Internation ally approved experiment for tha full course of ch* animal's Ufa, using boeh axes of animal*. Secondly, if poaltlvn result# vara.obtained which wars unequivocal, both statistically and biologically, chan a similar axperlmanc should ba conducted in a second apaclas. If tha flrat vaa a rodant speclas, preferably a nonrodene should ba used. Again, malignant tumors, uoe nodulas, 254 HONS 026295 not hyperplasia, not metaplasia, not potentially pracancaroua laslona but malignant tumors should b demonstratad. What Dr. Frlass racltad according to him la a kind of lntaroatlonal constnaua under tha umbrella of the International Agency for Research on Cancer In which powerful criteria bearing on tha sufficiency of evidence to brand the chemical as a carcinogen In an animal model, and then to use that Information for prediction or tranelatlon to the human really polnta to malignant tumors, controlled experiments, large chunks of lifetime and at least two species of animals. This information to this degree of rigidity does not exist with respect to the PCBs. That doesn't eay it shouldn't. The discussion of che meaningful Interpretation of potendally pracancaroua laslona becomes obviated when good experiments art available with or without cancerous tumors. Assuming that two species have shown malignant lesions In a significantly greater Incidence over the controls, then the actual exposure patterns of humans should be reviewed, adventitious usually, with occupational patterns as surrogates of Che greatest exposures. This la then translated, by use of modeling equations from the Incidence rates chat you saw in your most sensitive, positive animal model, to the prediction of risk in tha human, asking the regulatory accusation that the human la cskan as being as sensitive as tha most sensitive animal model, even If no effects are seen In the human, laving made that risk projection, using a number of empirical projecting equations, you only believe It when the loop le closed and caneer is observed In humans as a function of dose. There Is no denying that you want the perfect experiment, and none of the experiments which were listed as the three best are perfect. Dr. Kimbrough would agree that they are not perfect, and to the tune that they still disagree, and each has flaws end warts, che positive approach Is to design and do good, solid, perfect experiments, looking for caneer as the end point. Dr. Kimbrough then responded that these suggestions sre great. The problem Is that ve don't live In an Ideal world end doing these types of studies Inevo species Is very expensive, particularly If non-rodancs are used. For Instance, a dog study would take at least ten years and subhuman prlmats studies would lost oven longer than that. Fartlcularly with the FCBs, an additional question or problem exlets which may be quite uneurmounteble. As everybody hers has stated, there ere a lot of differences In the toxicity of the different lconers and congeners. Which mixture should be tested In these very expensive studies? At the Centers for Dlsesse Control, one of our goals nov Is to prevsnt disease. Should ve take note of any of the animal studios dons thus far end try to reduce exposure? Or should we wait until the evidence is there In people, that there actually has been an effect. That Is very often our dlleama, and also the dilemma of the regulator. Our foremost goal Is to protect the health of the people, and bow should that best be done? Ve arc tryingto develop nov at Che Centers for Disease Control, some vays to evaluate people for earlier Indicators of effects of exposure so It will no longer be necessary to welt until death certificates can be sorted. 255 MQNS 026296 Or. Frlaas stated chat ha agreed completely. People ihould ba protected on tha bails of present reality, and ona noda la to reduce all currant and futura axposura, absolutely, Dr. Edvard Burgar of Gaorgatovn Onlvaralty rasusad tha discussion on structura activity ralatlonshlpa, rafarrlng to tha alagant work by Dr. Safa vhlch ahovad soaa element of suceasa. In Dr. Safa's work, tha biological and points that vara usad to maasur* toxicity vara atrophy of parts of tha lymphoid aystam, tha thymus and tha splaan. To vhat axtant can this henaatly ba called a toxic affect, and vhat doaa It naan? Dr. Safa responded that ha did not knov. Splenic and thymic atrophy vara chosen because It ean ba Measured In a short elaa. Whan ona of your organs shrivel up, If you don't vant to call It toxicity, you don't have to. Whereupon Dr. Burgar pointed out that the thymus In the human being normally shrivels up with development. It doesn't In othar animals, but It doas In tha human being. If It doesn't. It 1> an abnormal avane. In fact. To understand vhat atrophy of tha thymua In animate may mean ha proposed that It may maan that tha Immune system la to soma axtant affected In a vay that ha nay not understand because the lymphocytes vblch are mainstays of tha Immune system ara derived to some axtant from tho thymus and from othar part* of tha lymphoid system. In that ease, should indices of dlseaso and points ba examined vhlch have nothing to do with cancer or othor things talkad about In so many of that# eastings--oantly, Immune-related diseases, such sa rheumatoid dlseaeaa and collagen disease*? Dr. Kimbrough pointed out that unfortunataly nona of tha members of tho panel vara Immunologists. Hovevar, in tho animal systems at least there la aom* evidence that if tha thymua of tho lnfsnt nous* Is irraditsd, thoo this mousa doas not develop properly but develops runt disease. Animals vlth runt dlsaasa may In tha and stags of their disease develop leukemia. This vas an affect of radlatloo on tho thymua and certain immuno suppressive drugs. In the thirties, tha thymus of children vaa irradiated because at that time the Idea existed that a large thymua vould somehov cause til health In children. These children Ister had a high Incidence of cancer of the thyroid bocauao of tho radiation exposure, but the other effects that were observed In rodanta have sot been reported In the children. The types of chlorinated hydrocarbon compounds discussed here, at least In rodents, affset the immune eystem the way radiation doss, but this has so far not been shovu In people for moat of these compounds. In order to evaluate the Immune system properly In population studies, base lines of normal variations In the general population need to bo developed first and laboratory methods noad to bo standardised. Dr. Intfir than ralterstsd that he vas bothered by the Jump that vas made by calling the association described ae one between structure and toxicity based upon an observation, the naanlng of vhlch he does not understand. Dr. Frlese then outlined vhat he felt vers neanlngful biological end points for Imnunocompscsncs, such as axposura to sndotoxlns. Invading viruses. Invading nlcroorganlens of the malarial parasite type, and tha aora subtle ehanga* In lamunoconpatenca, such as tha ability to reject grafts. In auaaary, neanlngful end point* ara the ability to withstand external challenge* of 256 HONS 026297 tnfectloua agents and tha ability to raspond differently to foralgn tissues. Dr. Burger addad that thara ara oma dlaaasa proeasaaa vhosa bases ara thoughc to ba, at laait In part, a darangad ltmuna sysesm. Dr. Zarvtna, Westloghouse 8&D, than voiced hla eottcama about tha follow ing quota: "Saetlon 761.30, eha CPA Administrator hereby finds that any axpoaura of human balnga or tha environment to PCBa as measured or datactad by any aclantlfteally aeeaptabla analytical sathod la a significant axpoaura." Should this ba eonatruad to mean that any axpoaura datactad by any analytical aathod would produca clinically datactabla symptoms In iumana? Ultimately this whola discussion ravolvas around what a human rasponsa la. In tha apldanlologlcal Investigations, snail that they ara, thara hava baan eartaln raaponsaa observed or not observed. Should It ba cencludad that any laval of PCS will produca clinical aynptoaa? Dr, Frlaas In hla rasponsa Indicated that tha quastlon about dona Tarsus rasponsa could not ba rasolvad frea apldanlologlcal studios which wars not designed In thse sophisticated a fashion, except perhaps for tha ehloracna affect. Cpldanlologlses hava looked at nany axpoaura situations, over tha last IS, 20, 2$ years and hava concluded that If tha folks over long periods of tins vara exposed to loss than 200 iig/m* eoneantrstlon In sir, the chances of developing tha stgns of ehloracna ara low. Whereas If tha exposure levels for reasonable periods of else, over years, rapstltlvsly, wars auch blghar than that, than their ehaaees of developing ehloracna ara hlghar. With rcspact to ehloracna, significant exposure really aaans aoe tha first datactabla aolecule that entered tha body, but that axpoaura over tlaa which gave a datactabla or raaaonabla lncldanca of tha afface, ehloracna. Vlth raspaet to tha Induction of ensynes, no Information la available. However, It certainly would ha greater than one nolecule or greater than tha currant sophisticated sensitive Hales of detection. Another worrlsona affect la tha lnfornatlon presented on tha infants who had low for torn valght. Although high and lew dosas wart discussed with the hlghar dose producing note of an afface, it Bust ba determined whether there la a does below which tha efface goes to taro. Perhaps tha efface on birth valght night ba linked co anryna Induction, and snzyna Induction aaana tha creation of catalysts for oxidation, and hormones and ttareldt could ba oxidised which would affect tha growth of tha Infant In utsro. It neads to ba determined whether this la a reversible affect, la eha ease of eha Tuaho infants, tha affect waa reversible. Dr. Kimbrough did net havo a definite answer either. However, for many years PCle hava beam prtaane In erace amounts In human adlpoee tissue, in serum, and other tisanes. Vs all carry low body burdens, snd we hare all Ingested trace amounts of FC1. A eartsln amount af environmental pollution exists that nothing ean ba dena about In tha short tarn. PCBa ara noe tha only thing in the anvlronnant. Va ara exposed to many other things, and to sort this out at eha vary lew eraea laval la impossible tn tbo general population. Dr. David Stalling from Flth and Vlldllfo Sarvlcn, wanted to know vhoeher Dr. Safa had made any rough completion between tha ortho>unsubstltutad PCIa in tarns of pstcene of total Arocior composition versus percent of total toxicity. He offered an alternative approach reported by him In tha Proceedings of tha Synposlua on PCBa published In eha Haw Tork Academy of Sciences which offers 257 HONS 026298 n alternative Co the large amount of work on pure Isomers, namely, that Aroclori can ba fractionated on carbon to give elution profile* which are Inversely ralitad to tha number of chlorine* in tha octho positions. Fraction* could readily be prepared, which are enriched If not totally containing tha l-chlorine PCS Isomers in eomperleon with the 2,3, and 4 substituted one* to evaluate Che toxic reaponte. It must be pointed out that chlorinated dibentofuraoe are preeent In fl*h, a large element In the human food ehalo, approximately at a ratio of one part of chlorinated dlbenaofuran to 100,000 pert* of PCB*. In tha published literature on chlorinated dlbenaofurane and PCBe, it 1* tafa to *ay that moat of the chlorinated dlbenzofuran* In Aroclor* are under 10 ppm, probably cloter to 1 ppm. Dr. Rapps'* group (Uaaa, Svedan) ha* done quite a bit of work on chl*, but there are about 35 laomar* of chlorinated dlbenaofuran* In PCI*. In the environment about four or five laomer* are found, all of which appear to be 2,3,7,8 lubaclcuted, and with the appropriate calculatlone, it can be determined that a 100-fold exeese of 2,3,7,. 8 subtltuced chlorinated dlbontofurana axlaca In tha flah that wa have analysed. Thl* ral*e* a lot of queselone, particularly elnce the finding* mad* In flah residue* point* to tba 2,3,7,8-aubatltutad dlbenaofuran*, exactly analogou* to the residue* la Tuaho oil. Dr, Safa responded that his group had experiments in progress to determine how much thl* "kind of new elaas" of toxle PC8* contribute, but no results are available. Nor is it known what the contribution of the polychlorinated dlbenaofuran* to the overall toxicity of PCS* is. However, the point must be made that the PCS* can contribute. They are toxle, and what proportion of this Is either In fish or In Tusho Is not known, but they are contributing and not everything can be blamed on polychlorinated dlbensofurans. Dr. Edward Paader, Southern California Edison Company, pointed out that Dr. Blair Smith from the National Institute of Occupational Safety and Health had only found equivocal changes In mortality patterns and no clear assocla* don with carcinogenic effects In his studies of workers exposed to PCBs In capacitor plants. Thess workers had PCB blood levels as high as 500 ppb (ug/1). Perhaps blood KB levels achieved In the animal studies could be compared to these extremely high human levelp and then conclusions could be drawn relevane to dose effect levels and responses or lack thereof In humans? Dr. Frias* in his response focused on the ehra* kay animal studies which went long enough and pointed out that this Information was not available. In the human epidemiological studies that were reviewed at the seating, exposure levels end seme blood levels are available but with so clear indicators of excess carcinogenesis ve night have essentially ehe negative information but we don't have the threshold for the positive Information. Dr. Kimbrough added that she did not analyse the tissue* for PCBe in her animal study and neither did anybody else In any of the other studies. However, It is very Important to actually find out what tha tissue level* are that cause an effect In target organs. An additional problan Is differences In species response where soma species may be mors susceptible at lover doeag* levels. 258 MONS 026299 Whereupon Dr. Feeder responded that that Information could ba developed from general single species pharmacokinetic studies. Is there any data for other effects which make these comparisons and determine blood levels so that tn fact, In looking for similar situations in hunans lnterspecles comparisons could ba made? Dr. Frless then pointed out thae the only example was the chloracne epidemiological Information where 200 wg/n3 in air wae associated with a higher Incidence of chloracne. This probably goes along with a blood level of about half the air level. Perhaps 100 ppb In blood or higher may be associated with a higher probability of developing chloracne', but It sura "as heck" hasn't been proved as csusatlon. Or. John Brown of general Electric Research and Development Center, emphasised that he was quite concerned about the Interpretation of chronic health effects obsarvad In animals that may or may not be chronically accumulating the agent being tested because In order to Interpret auch data Intelligently and make lnterspecles comparisons, ona would Ilka to know the level of the agent present at the time the health effects are being observed. In Dr. Kimbrough's first chart where she described tha feeding schedule to the rats that ware administered Aroclor 1260, It was Illustrated thee chosa rats received 3 or 4 g/kg of Aroclor 1260 se tha total dose. Matthews has reported that Aroclor 1260 is almost entirely retained In the rat which would Imply that those rats wars carrying on the order of 3 g/kg of Aroclor 1260 at the time of sacrifice, which Is approximately tha ID.. , It la hardly a trivial load. Is there any evidence of tissue levels present in the carcinogenicity tests that have been done to data or has anyone retained tissue so that It may be possible to go back and figure out whether tolerance levels were exceeded In the testing. Dr. Kimbrough responded that she would have to see whether she still had any tissues. In some studies where PCBs were fed for a period of tlma and then the rats were removed from exposure at tha end of the study, the*levels of PCBs were still very high in adipose tissue, as high as 1000 ppm, but since the study vas different, such extrapolations should perhaps not be made. Tha adipose tissue levels might have been higher, and of eoursa the amount that would be In the liver would be proportionally lower. With polybronlnated biphenyls this vas recently tested becausa It Is very Important. The problem always Is that tha person that does the cancer research works by himself or herself and the chemist does the same things and the two people never get together which is very unfortunate. Things should be dooa Hfferantly. Dr. Kenneth Chase of Washington Occupational Wealth Associates then expanded on clinical symptoms observed In humans. Traditionally In clinical medicine, we'distinguish betvean symptoms and metabolle abnormalities that are not necessarily accompanied by symptoms and certainly there are a number of studies, many of which have been discussed here, which have documented metabolic abnormalities. For symptoms resulting from PCB exposure, it would be useful to distinguish betvean acute, subacute, and chronic exposure. When vorkars ere exposed to FC> containing fluids, particularly In heated form, they can get Irritation of their skin, their eyes, the mucous membranes with complaints of throat Irritation and a cough. Those symptoms go away when the voxker la removed from exposure. Kot every single worker reports those symptoms, but 2S9 MO NS 026300 they are vary coneietent. Nobody ukii nuch laeue over thoac aympton*, but at loo| at that quaatlon vaa being raleed, It ahould ba pointad out that thara ara acuta aynptone ataodatad with acuta axpoaura to PCBa. Chloracna and othac dermaclcie nay ba aaaoctatad with acuta and longer axpoaura. Dr. friast made tha obeervation that thara aeema to ba a thraahold at vhlch chloracna la obaarvad in tana of work environment axpoaura. Tha figure of 200 ug/a waa tuggattad. Thara ara quotaa Ilka that alao in tha NIOSH crltarla docuaant, ate. In our axparlanca. moat of tha PCB axpoaura ia dermal and not Inhala tion, although that nyth kaapa gatting parpatuatad. PCBa aa tha cheuleta can tall ua don't have a high vapor prataura, but thay ara aaally abaorbad through tha akin. Thara ara atudiaa that vara citad yaatarday by Dr. Gaffay chat auggaat that tha likaiihood of developing chloracna probably haa an avful lot to do with work practlcaa and work habit*, and thoaa atataaanea ara alaboracad on in our papar. in Salth'a papar, and alao in Marool'a papar (Marooi at al. 1980), probably tha only lnvaatigator who triad to do aaaauraaant* of PCBa on tha akin ltaalf. Dr, Mary Jo Podienik, Madieal Collaga of Viaeoneln, than consented on tha raproductlva affacta aacribad to tha PCBa in lighe of Dr. Prlaaa' atataaant that induction of tha PASO ayataa nay raaule in an incraaaad aatabolian of aax atarolda, and could ba ralatad to tha lovar birth valghea. It haa baao auggaatad that toaa of tha raproductlva affacta which hava baan notad in "both rodanca and aubhuaan prlaataa, aa wall aa aaoatrual lrragularitiaa, ara dua to an incraaaa in tha aatabolian of aax atarolda, Vhlla it ia iaportant to axanina raproductlva affacta in sexually aatura aniaala, it ia going to ba vary iaportant to conaidar thaaa affacta in tha naonaca. It haa baan ahovu that offapring of aothara in rodanta pratraatad with polybroalnatad blphanyla ara raproductlvaly inconpataot latar in Ufa. It haa alao baan ahown that circulating aax atarolda in tha naonata are reaponelbla for normal raproductlva davalopaant, and it la probably via tha aachanlaa of thaaa aax atarolda acting on tha hypothalaaua to imprint it during tha neonatal period, Thle la probably alao true of the aax difference in tha hapatie aono-oxyganaaa ayataa aean in aaxually nature aniaala. Sinea PCBa fraaly enter braaat allk and alnca many of the conganera found in braaae niik ara potent indueara, tha affacta of PCBa on tha aonooxyganaaa ayataa in tha neonatal period ahould ba laveatigated nora cloaely. Dr. Blair Smith of NIOSB pointed out, having vitnaaaad knowledgeable darmacologlate argue over whether or not particular lealone are chloracna, that tha diagnosis la largely dependant upon tha clinician's pre-existing knowledge of axpoaura. Thla suggests that the dlagnoela of chloracna la not aa hard a algn of PCB axpoaura or toxlelty aa wa might Ilka to think. Thara la a lot of inter- and intra-obeorvor variability in tha dlagnoela. Tha number of eaaaa actually citad ara cufflelently email to raiaa tha question whathar thaaa obearvatloue ara not in aoma way biased by the lnvaatigator'a knowledge of tha axpoauraa. Tha other biological outcome which haa baan citad aa a hard or Harder Biological affect that xan ba obaarvad in hunana, ara tha birth waighta of infanta born to PCB expoaad aothara, tha only huaan data from the 0.S. vaa in Chapter 5. It vaa a coaparlaou of PCB high expoaad, low expoaad, and not expoaad (control) women in New Tork State. The raaulea compared to the 260 MONS 026301 control group ware Inconsistent. The assn birth weights vers lover for the high exposed than for tha controls, and they vara higher for the lov PCB exposed group then for the controls. Tha confldaoca Intervals cited by the investigator vers 901 eonfldenca Intervals. In effect, one has three comparisons at s P value that Is really 10X In looking at that data, and the results as presantsd could have been due Barely to chance. Dr. Kiabrough ended the dlscuselon by agreeing that as far as the coaaent on chloraene Is concerned, clinical observation ara an lapraelse science. Physicians thst have not had a lot of experience with chloraene Bight not even detect alld cases of chloraene. On the other hand, othsr lesions are nleclaaslfled as chloraene. Hovever, if a biopsy Is taken of a hair follicle. It should be possible to diagnose the lesion alcroscoplcally. As far aa the birth weights ara concerned, chare Is evidence in snivels chat reproduction Is affected, for that reason, and tha prellalnary results In huaans presented here, this problea should be explored further. A lot of other things affect birth weight, such as saoklng and alcohol consuaptlon by the Bother. In the huaan population. It vlll be difficult to sort these things ouc. Confounding variables ara one of the biggest problaas in these types of huaan scudiss. 261 HONS 026302 Addandun In an additional atudy with Sprapua Davlay rata fad 100 ppa (ng/kf) Aroclor 1260, In which thara wart 32 nala control#, 69 fanala control#, 46 nala axpoaad, and 47 fasala axpoaad raea axallabla for aieroacoplc axaninatlon, a high lacldanca of wall diffarantlacad hapatocallular carclnoua wara found in fanala rata and a tovar incidanca in nala# (Valtaan, R.H. and Horback, D.8., 1981). 262 MONS 026303 REFERENCES* Institute of Laboratory Animal Resources. 1980. Blitolo|le typing of Uvar tuner* of th rat. J. Natl. Cancer Inat. 64:179-206. Kimbrough RD, Squire RA, Linder RE, Strandberg JD, Nontail RJ, Buree TV. 197S. Induction of liver tuaora In Sherman strain rats by polychlorinated biphenyl Aroclor 1260. J. Natl. Cancer Inst. SSi1453-1459. Meronl M, Coloabl A. Antonlnl C, Foa V. 1980. Health affects of long tens exposure to polychlorinated biphenyls. Int Mechanism* of toxicity and hasard evaluation. Holmstedt R, lauvery I, Harder M. Roberfrold K, ads. Aasterdam/Nev Fork: Elsevler/Morth Holland Biomedical Pre**i p. 331. McKinney JD. Singh F. 1981. Structure activity relatlooshlpa In halogenated biphenyls, unifying hypothesis for structural specificity. Cham. Biol. Interactions 33.: 271-283. Miller JFAF. 1963. Role of the thymus In lsnunity. Ir. Med. J. 2:459-464. Morgen RW, Vard JM, Hartmann FI. 1981. Aroelor 1254 Induced Intestinal metsplasls and adenocarcinoma In the- glandular stomach of F344 rats. Cancer Res. 41:5052-5059. Squire RA, Levitt MB. 1975. Report of a workshop on classification of specific hepatocellular lesions of rata. Canear tea. 35:3214-3233. Veltman RB, Norback DB. 1981. Sequential light and electron microscopic analysis of hepatocellular earcinoma induced by Aroclor 1260. The Toxicologist, 1:65. *Addltlonsl reference*mentioned in the discussion are either cited In the paper by Frlsss or by Oaffey of these proceedings. 263 HONS 026304 CHAPTER 7 ENVIRONMENTAL EFFECTS Th paper In this chapter describes aeveral recant studies of th environmental affects of PCBa. Pra-1978 papara ara recognised, but cba tread toward favor acuta and more chronic atudlaa, and batear quality raporta la apparent, Tha author dlaeuaaaa PCSa dlatrlbutlon within tha aquatic environ ment, effects on phytoplankton production, dlatrlbutlon among different flah trophic levels (e.g., plauktlvoras, detrltlvoree, and carnivores), and the uae of aora sensitive 'Indicator' tpoclaa to aonltor overall PCBa levela and locata "hot apota". Alao tneludad la tha affacta of current FCla lavala on aenaltlva nonequacle apaelaa. aueh aa raproduetlon and lathallty In tha alnk, or harltabla aatabollc affacta In fish-eating blrda. Hypothaoliad aachanlaaa for tha partitioning of PCBa la water and lipid notarial ara dlacuaaed. aa are tha problana aaaoclatad with tha abaanca of a ataadard data -collaetlon and reporting format for thaaa typaa of atudlaa. Tha dlacueelou summary baglna with a preaantatlon denonatratlng tha ralatlonahlp between vary low lavala of PCBa contamination and aaaaurabla lncraaaaa in flah liver monooxyganaaa actlvltlaa. Thla deeerlbea tha rata of PCBa bloaceuaulatlon and anayma Induction In thaaa apaelaa. and tha disappearanca of tha affacta follovlag removal of tha contaminant. auggaaclng a poeaibie modal for monitoring cba time courae of PCBa contamination. Tha question la ralaad aa to bow reasonable or appropriate tha uaa of monooxygoneae actlvltlaa might be aa an Indicator of tonicity. Alao diacuaaad la a comparison of PCBa levels vhleh induce anayma and reproductive affects versus frank toxicity, and tha comparative toxic levels of dibentofurans In fish. 264 HONS 026305 ENVIRONMENTAL EFFECTS OF POLYCHLORINATED BIPHENYLS: IMPORTANT ECOLOCICAL ASPECTS Richard G. Hunter Corps of Engineers Tulaa, OK 74121 Tha paae ftv years have baen earkad by a prismatic grovth In atudlaa of the environmental affeeta of polychlorinated biphenyl (PCB). Although atudlta aay not have lncreaaed on a quantitative baala. the qualitative grovth of auch retearch efforta hae been lapraeaiva. Raporta of PCB realdue levele in Halted-populations from-email-areas have been replaced by aeudlea -of aeverat coaponanta of the ecoayataa over a large geographical reach. Atteapee are nov being aade to conetruct aophlatlcated aodela to prediet realduea In varloua environmental coaponanta baaed on input of PCB eo the ayatea. Acute tonicity atudlaa have given vay in aany eaaea to atudlaa of chronic toxicity which propose the aechanlaa producing the observed effect. Many of the advancea la the toxicological tachnlquea which are a part of thle prismatic grovth have been detailed by Mayer ee al. (I960). The purpoee of thle paper la to review recent (generally poet-1978) literature concerning the effacta of PCB on the environment. Original data concerning the PCB contone ef aeleeted fraebvator fieh apeciea la praaented aa an axaaple of area* requiring additional lnveatlgatlon. The inplleatlona of thane findInga tor establishing vater quality atanderda and designing more efficient PCB monitoring programs are dlscusaed briefly. RECENT ECOTOXICITT STUDIES Thera la lietla doubt avan ralativaly modeat lavala of PCB can advoraoly affect the vary base of eho aquatic erophie pyramid, Aroclor* 1248 baa baan ahovo to algnlfieantly doproaa the growth of a graan algao ae aquaoua concentra tion* of 11 to 111 ug/1 (Chrlatanatn and Zlalskl 1980) and 10 ttg/1 PCB lnhiblttd grovth of phytoplankton ae a vatlaey of call danalelaa (Klappol and McLaughlin 1980). In tha lattar atudy, a racovary of grovth occurred after the Initial Inhibition and thla vaa be1laved to ba partially dependant on population alee and phyelologlcal etaeue at tha time of expoaura. Only I ug/l of Aroctor 1254 reduced eha blomaaa and alaa of phytoplankton conmunltlaa (O'Connors 1978) and raducad call dlvlalon of tho alga Iaochryaia galbana (Harding and Phillipa 1978). Fifty ug/1 of KB algnlfieantly roducad eho uptake of C by tho merino diatom Thalaealoolra pssudonana through radueod photoaynehaala par eall and a decrease in viable cella (Kicliaala at al. 1982). Kaduead growth alao occurred in tha alga Eutlana gracilis upon axpoaura to 10,000 Ug/1 of Aroclor 1242 for eight daya (Bryan and blafaaon 1978). A yaaet erpoaed to aavaral apoclaa of Aroclor at concentration! of 25,000 ug/1 dleplayad reduced grovth after 160 hr (Tejedor 1978). In addition to decreased grovth, other effects Aroclor la e registered trademark of the Monaanto Chemical Company, Incorporated. 265 MQNS 026306 of PC8 have been observed. Ao alga exposed to 30 ug/l of PCB for only three hr ahovad raducad motility (Zullai and Beneeke 1978). Expoeure of tha graan alga Dunnallalla tartlolacta to 100 ug/l of Aroclor 1254 rasultad In Increased call division (Harding and Phillipa 1978), a phenoaenon staller to tha action of sons htrblcldts. Tha fact various apaclas of plankton vlthln a population say ba affactad at dlffarant lava la of PCB contaalnatlon has laportant raalfIcatlona for aquatic communities. Thoaa organiana most sanaitlva to PCB vlll ba eUmlnsttd from tha population at ralatlvaly low levels of PCB Input. This craatas voids vhlch ara filled by lass PCB sanaitlva apaclas, thus causing a shift in apaclas coaposltlon of tha population. Tha rasultlng alterations to aquatic food waba ara alaost Incoaprahanslbla and aay adversely affect organlsas otherwise quits tolerant of PCB contaalnatlon. Currant studlas have attributed chronic effects noted for other organochlorinas to PCB (Mehrle at al. 1982). Theae authors observed a decraasa In the aechaolcal properties of vertebras of striped bass (Morons saxatllls) collected froa tha Eastern 17.S. They believe this decrease to be a result of environmental contaminants, the most cotsaon of vhlch vie PCB. Whole body concent of PCB residues In these striped basa ranged froa 0.03 to 2.6 ug/g (vet weight). The mechanism Involved vas related to an Induced functional deficiency of VltamlsC, leading to a reduction In collagen formation, with the Increased ratio of minerals to collagen causing brittleness of the spine. This hypothesis vas originally advanced for tonaphsne (Mayer at al. 1978, Hamilton at al. 1981) and similar effeeta have been shown to result from kapona and mires (Mehrle at al. 1981). Tha Identification of this process represents a new direction In PCB ecotoxlcology research. Rot only does It Illustrate chronic processes perhaps eonon to a variety of toxle subetancas. It presents pathways for research Into countering these effects. The more traditional Investigations focusing on bioassay teats are lose veil represented in recant studies than in the past. Perhapa thla lndlcaees a shift of. emphasis to tha effects of tonic substances other than PCB or It nay be symptomatic of a general decline in research funding. Acute toxlcley of PCB to rainbow trout (Salmo lalrdnerl) in flow-through tests vas measured as 2 ug/l (Blrge 1979). These same investigators rsportad LCSOs for largemouth bass (Mlcroptarua salmoidaa) and redsar sunflsh (lepomls mlcrolophus) of 2.3 ug/l and 19ug/l respectively. Chronic effects on the 11fa cycle of brook trout (SalveIleus fontlnalls) occurred at concentrations of Aroclor 1234 ranging iron 0.) to 1.9 ug/l (Mauck 1978). Fathead minnows (Plnaphalas promslas) exhibited chronic effacta ae concentrations of 0.1 to 0.4 ug/l of Aroclor 1248 and 1,) to 4.0 vg/1 of Aroclor 1260 (SaPoa 1978). The 30-day LC30 obtained for fathead minnows exposed to Aroclor 1248 was 4.7 ug/l, the value for Aroclor 1260 vas 3.3 ug/l (DePoe 1978). Iloassaya conducted on shrimp (Cranton aeptanaplnoea) produced a 96-hour LC50 for Aroclor 1242 in saavster (20* C) of 13.0 ug/l while the same test 'Using Arochrr 1234 resulted to an LCS0 of 12.0 ug/l (McLeese end Metcalfe I960). Threshold levels vara 6.3 and 0.5 ug/l respectively. An attempt to determine the 96-hour LC50 for the two PCB species In sediments rasultad in no mortality of the ahrlmp at 780 ug/l of Aroclor 1242 end 3400 ug/l of Aroclor 1254 (McLeesa and Metcalfe 1980). Culf kllllflsh (Fundolus grandls) exposed 266 HONS 026307 for f4 hr to water containing 40 ug/l Arocior 1242 had highly significant (p0,00t) lncraasea in locomotor activity (Fingerman and Russell 1980). Aqueous concentrations of 8 ug/l Arocior 1242 inhibited aoiting in fiddler crabs (Oca pugilator) exposed for 38 days (Flngeraan and Fingerman 1979) while crabs exposed to 2000 ug/l Arocior 1242 for four days exhibited a greater dispersion of aelanln in their carapaces (Flngeraan and Fingerman 1978). The information obtained from these recent studies on the ecotoxlclty of PCS reinforce the wealth of information published during the early 1970s. Invertebrates, at a group, are among the organisms most sensitive to FOB, and salmon and trout are more readily affected than other fish species. In general, toxicity appears to Increase with the level of chlorination in a particular PCB species. Data for chronic toxicity studies on many common freshwater fish species is lacking and ehara is no bloassay information for entire groups of organisms. This Is particularly true for vlld mammals. Animals which are large, difficult to keep under laboratory conditions, or rare are noc represented in the literature on PCB toxicity. This aay be especially important since rarity or difficulty in surviving captivity might be indicators of hypersensitivity to environmental perturbations. As early caution by Johnson (1968) Is applicable to data on the highly Upld-soluble FCBs. He noted laboratory diets artificially high In fat might cause an apparent Increase In body content of organocblorlnes. Similarly, an Improper diet might cause utilisation of body fat containing PCB, thus increasing apparent toxicity. Methods of reporting the results of chronic toxicity studies of PCB are also subject to Improvement. A minority of the investiga tors calculated effective concentrations (EC) for a given affset, although this Is an excellent technique. Likewise, few studies presented threshold concentrations for the test organisms. IMPACT OP PCB OH 1EPR0DUCTI0H Pish Disruption of reproductive processes In fish exposed to PCB apparently is cot related to decreases in haechablllty of eggs. A comparison of eggs from wild populations of lake Michigan laka trout (Salvallnua namaycush) with aggs from hatchery stock found no affeet of PCBs on the early survival of the trout (Stauffer 1979). In that study, the eggs from the vlld populations contained PCS residues of 3.16 to 9.9 Ug/g, those from the hitchary fish ranged from 0.20 to 0.28 ug/g. Studies of Atlantic salmon (Salmo salar) also shoved no correlation between PCI levels in aggs and hatcha^lllty. the salmon aggs contalnsd 1.88 to 6.48 ug/g PCB on d lipid basis (Eltko and Saondars 1979). PCI also had no offset on tha hatehablllty of dried Brine ahrlmp (Artemis aallna) aggs whan hatching occurred In watar containing concentrations of 10 ng/l PCI (Kuvsbara at al. 1980). >lrda There ere few reports of PCB adversely affecting reproduction In birds. Japanese quell (Cocurnlx cotumlx Jeponlce) fed diet* containing 30, 100, and 450 ug/g PCB. far 3 trsaks during sexual maturity showed significant offsets. Theee included delayed laying and diminished laying capacity, lneraasad liver weight, testes changes, and reduced breaking strength of eggs (Bleesmsnn 267 MOWS 026308 1982). Exparinantal evidence obtained for Veetera grebes (Aechaophoru* occidental!*) Indicates PCS level* ere not directly correlated vith eggshell thinning although an Interaction batvaan Aroclor 1260 and DDE In reducing aigahall thlekaaaa vat suggested (Lindvail and Lev 1980). Seraach ovla (Otut aalo) fad a dlat containing 3 ug/g of Aroclor 1248 shoved no dacraaaa In aggahall thlclcnaaa, nusber of agga laid, young hatched, or young fladgad (McLana aod Hughe* 1980). The agga of thaaa ovla eontalnad PCS raalduaa ranging frou 3.9 to 17.8 Ug/g. Thara van no evidence of eggahell thinning In Croat skua (Cacharacta skua) eollaetad fro* aavaral locaclona. The average PCS concent of the adult akua uuacle tlaaua vaa 16.0 ug/g (wet weight); the egga averaged 20.0 ug/g (wet weight) (Purneaa and Buttoo 1979). A aurvey of ala bird apeclea; laughing gull (Lama atrlcllla). white lbla (Eudoclama albua), gloaay lbla (Plagadla falclnellut), American oystercatcher (Haaaatopu* palliatua), wlllat ((jatootrophorua aaalpalnatua). aod ruddy turaaton* (Aranaria tncarpraa), revealed no conaleeant trend of PCB In the egga (Blue and Lauoot 1978). Thaee lnveatlgatora found no obvloua probleaa with reproductive aucceaa In any of the apeclea. In a alallar tcudy conducted In Canada, tha lean PCB eoncanc of egga ranged fro* 0.32 ug/g In coason alder (Soaaterla noltleaUa) to 21.7 ug/g In rasorblll (Alca tarda) (Pearce at al. l5)9j. Those authora alao concluded thoa* apeclea were' unaffected by PCB lavela, Klne-*onth old Ballard ducka (Ana* platrrhvchua) fad a dlat that Included 23 ug/g of Aroclor 1256 for at loaac one vouch dlaplayed no ill affoeca on reproduction. Tha paraMtara *aaaurad Included the nuabar of bona laying, tha data od the flrat egg, elutch alto, hatching auccaaa, aurvlval to three veaka, nuabar of tlaea off tho neat par day, and total tlM off the neat (Cutter and Being 1980). Thera vaa a largo lacraaaa la fertility in tha mIo ducka fed tha PCB dlat which the authora were unable to explain. The effect of PCB on tho reproduction of hlrda aay be aora lueldloua than Indicated In tha pravloua review. Chlekene (Calua galua) receiving 5 ug/g of PCB In eheir diet ahowed no algnlfleant offacta on reproduction. Bowevor, feuala progeny of thaaa chlekana dlaplayed algnlfleant daeraaaee In the biological effectlvonaea of estradiol aa wall a* other alteratlone to their eecabollsa (Roautcky et al. 1979). Significant decreaaea In body taoperaeure at aabiont taaperaturea of 0 and -3* C occurred In *ournlng dovea (Zenaldura Meroura) fed a dlat with 60 ug/g Aroclor 1236 for 62 daya (Tori and Mayer 1981). Tha lnveatlgatora concluded FCBe could reduce winter aurvlvel of adult and Juvenile eournlng dovea. Thie lack of iufor*aeion on the effeeea of KB on progeny of hlrda rapreaauta a aaeded area of reaaareh and reeuphaaliea the well-known Med for atwdloa Maturing the affecta of toxic aubatancaa on aeveral goneratIona. It ia probable there are populatloua of placlvoroua hlrda 1* the 0.8. receiving 3 ug/g of PCB la their dlat. Kama la little Intonation la available concerning the effect* of PCB on vlld aumal* although aone intonation exlata on raaidue leva la In natural popula* tlona. Wild population* of hta brown bata (Eptaalcu* futcua) were found to contain elevatae lavela of Aroclor 1260 and to bave high atlllblrth rate*, but axperlMnt* failed to correlate the two (Clark 1978). Similar acudlee on little brown bat* Ofyotla luclfugua) Indicated high level* of KB (up to 23 268 MONS 026309 US/s) th* fetuses may have been responsible for tome stillbirths (Clark tod Kryultsky 1978). Mink (Mustela vlson) ftd dices eonttloln| Aroclor 1242 experienced total reproductive failure when PCB levels In their feed reechcd 5 ug/s tod all sink died vtien the disc cootaloed 20 us/s of PCB. Ferrate fad the seat levels of PCB showed no adverse effects of 5 ug/s but vara unable to successfully reproduce at dietary levels of 20 ug/s PCB (Blcavlns at el. 1980). Earlier literature Identified mink as being especially sensitive to FCB and It has been estimated 63S of the fish ssatplce collected In 1978 froa major watersheds near the Great Lakes would pose significant hazards to animals, such as alnk, feeding on the fish (Velth at al. 1981). MODES OF PCB ABSORPTION Conflicting Information Is available on tha aechanlsas by which the biota nay become contaminated by PCB end this la certainly a major area requiring further study. Bjerk and Brevlk (1980) provide three aechanlsas by which PCB may contaminate an anlaali 1) across the skin, 2) through respiration, and 3) Ingestion. In aquatic Investigations, It la difficult to separate PCB obtained from water and that received through respiration, so many Investigators have coablned the categories. Such a mechanism can act rapidly; larval striped bass exposed to squeous concentrations of Aroclor 125* obtained SOX of their final PCB concentration In the first 12 hr (Csllfano et al. 1980). Brevlk (1981) noted PCB content was not age-dapendent In several fish speclsa from a Norwegian lake. He believed this vss evidence of PCB partitioning between fish tissue lipids and surrounding water and stated this mechanism was ehe predominant factor In determining PCB levels In fish. A 5-year monitoring study of fish In a Finnish lake also revealed PCB concentrations In fish vers correlated with fat content (Hattula et al 1978). The same findings of a correlation between lipid levels and PCB content was repotted by van den Break (1979). Significant correlations between PCB and lipid content have been found for yellow perch (Perce flevescsns) and aalmonlds in Lake Onearlo (Armstrong and Sloan 19897! Indeed, that* appesrs to be a constant partition ing of PCB between the vetlous components of the environment. Partitioning of ?CB between the.vacer and particulate matter has been demonstrated (Biggs et al. 1980). Ae l*C-PCB concentrations of 25 mg/l, 19 to 22X of the PCB vac sorbed to particles and 70 to 72Z was in the water. When PCB concentrations vara increased four-fold, 88 to 89X was sorbed to particles end 22 to 23X vae la the voter. This sorption was shown to be reversible. Unfortunately, the concentration of PCB in an organism is not simply related to the lipid content of a particular tissue. One Investigator found no increase in PCB levels in liver tissues of exposed fish, although he was able to correlate sample lipid content and PCB level (van den Broek 1979). He viewed this as evidence of food chain accumulation. PCB content In grey mullets (Chelorn labtosus) was not dependent on age but was dependent on food Intake (Narionne 197$). A study by ehe State of Hew York found no significant eorTalatlon between PCB concentration and Upld content In striped bass (New York Dept. Environ. Conserv. Keport 1979) and Kahrla et al. (1982) also found no significant correlation-between PCB content end llpide in striped bass. Studies by Hunter et al, (1980a,b) correlated PCB content with feeding habits in several freshwater fish speelas. Levels of PCB lo seal bralna ware lower 289 HONS 026310 th*n other tissue of a similar fat contone, suggesting differential absorption of PCBs In fat (Rosavall at al. 1979). Mo uttar what tha mechanism, aoiula can concantrata extremely larga amounts of PCB from tha environment. In 14-day tlva car studies In tha Hudson Rlvar, fish concaneracad FOB to lavalt 13,000 to 26,000 tlaas thoaa In tha watar (Skea at al. 1979). Tha Importance of dapoaltlon of PCB abaorbad to auapandad partlclaa hat not baan overlooked. Two apaclat of marine banthoa vara abla to concantrata PCB from 3.8 to 10.8 tlaas tba amount In tha aadlaant In 32 days (McLaasa at al. 1980). Two apaclaa of polychaataa had 3-day bloconcantratlon factors of 236 to 373 (Courtnay ac al. 1978), and bioconcsneratlon factor* of PCB In fillats of 3 fish apaclas to PCB In aadlnants ranged from 0.319 to 10.319 (Huntar ae al. 1980b). ECOLOCICAL PROCESSES IM PCB COHTAMIHAIIOH Thara ara two aspects of PCB contamination which have bacons especially important In tha last few year*. They sra tha usa of Indicator organisms and tha ecological processes related to accumulation of PCB In the blots. Data recently collected from nine Oklahou lakes Illustrates tha importance of both areais of PCB investigation. All of the lakes examined ara Corps of Engineers impoundments locatad in tha eastern ona-half of Oklahou (Figure 1). Tha lakes vary greatly in morphometric factors, water quality, age, and vatarahad Industrialisation (Table 1). Sevan of tha lakes have not bean Identified at having FCB sources in the immediate watershed. Relatively severe FCB cootamlnetion was detected In the biota of Ft. Gibson Lake in 1979, An extensive investigation by en inter-agency task force led to the discovery of several sull, generally lov-level sources of FCB vithln an industrialised area at the upper and of the lake (State of Oklahou Report 1980). Vabbers Falls Lake has an lntarlm-NPDES permitted FCB discharge into eha upper lake. A paper recycling operation has discharged effluent containing a uximum of 2.3 pg/1 PCB alnea 1976. Methodology Three-hundred and aeven fish representing seven coomon species were eollsetsd from the nine lakes between 1979 and 1981. All fish were collected during tha summer months with the aid of gill nets and electro-fishing devices. Sample numbers were determined based on the surface eras of the lake, with approxlutely one fish per 330 to 430 surface ha of water at norul pool elevation. An exception to this rule occurred at Lake Eufaula, where due to budgetary constraints, only ona-half the required fish were collected. The samples ware prlurlly composed of about 631 datrltlvoras, 23Z carnivores, and 10X planktivoras. In some cases it vas not possible to collect all seven speclae. To examine the influence of fish age on PCB burden, daslrad else ranges vara established for each species (Table 2). These were based on langth-aga relationships (Lewis at al. 1968) and were typically 2-,3-, and 4-year old fish. Fish spsclss vara grouped into three trophic levels as follows! Planktivoras: Cisxard ehad (Porosou copldlanun) 270 HONS 026311 Nl-r. . Hep of th. .tudy .howln. |.k,, pled. MONS 026312 Table ). Selected Morphonotrie Dete for Nine Oklahoma l.akct Factor Sufaula Oologeh Ft. Clbaon Inpoundaent Texona Wabbera Falla litch Hu) ah Keyatone Few Age (year*) Surface Area (ha) Mean Deptl) (e) Heelnun Depth () Norphoadaphlc Index ltora|a Ratio Mean FCS In weter (pg/l) IF 41.179 4.9 24.7 10.5 0.50 <0.1 17 11.922 5.7 IS.) 11.7 0.51 <0.1 51 7.730 5.8 21.9 11.4 0.07 <0.1 57 35.41) 9.1 35.4 13.) 0.75 <0.1 II 4.411 5.0 13.3 34.) 0.20 <0,1 5 1.514 9,9 20.) 4.2 0. 34 <0.1 30 5,241 7.4 14.0 9.4 0.33 <0.1 17 22,388 II. 0 27.4 40.2 0.40 <0.1 4 15,379 11.8 21.7 42.3 0.S2 <0.1 HONS 026313 Tabla 2. Slit langaa of flah Collaccad from Nina Oklahoma Lakaa Spaelaa Hlvar carpauckar Carp Channal catfish Smallmoueh buffalo Hhlta baas Glaaard ahad. Hhlta erappla Minimum alia (am) 200 200 200 200 200 123 150 Maximum alia (am) 330 330 330 350 330 300 300 273 MONS 026314 Detritivores: Carp (Cyprlnus csrplo). channel catfish (Ictalurus punctstus), emellaouth buffalo (Ictlobus bubalus), rlvar esrpsucker (Csrplodss carplo) Carnivores: Whits crappla (Pomoxls annularis), whits baas (Morons chrysopa). Although tha feeding bablta of some of thaaa apaclaa ovarlap, thaaa groupings provlda ganoral trophic lavala, Tha tana dstrltivore nay ba criticized as too general, however the laportaot point la flah in thla group inhabit the bottoa aubatrate. Tha apaclaa in thla study vara chosao not only basad on trophic level, but also because, vltb tha exception of gizzard shad, they are prise contributors to tha eonnerclal or sport fishery. It should also ba noted the apaclaa are coamon-throughout tha study-era* and ara easily collected. A fillet was raaoved from each fish as prescribed by the,Pood and Drug Administration UJSFDA 1972). vrapped in aluminum foil, and frozen. The fllleta were then transported to a laboratory for determination of total PCS residues. Laboratory analysis followed methods raeomaended.by the PDA (1972). The samples vara analysed using s gas chromatograph vith a BJWi electroncapture detector. Du* to contractursl constraints. It vss necessary to utilize.four laboratories for sample analysis, for this reason, particularly strict quality control requirements vara maintained. These included sequential recovery and spiked recovery analysis for 5X of all samples. Sequential recovery yielded a minimum of 96X recovery from initial extractions while splkad recovery ranged from 89 to 108X, and tha results presented in this study were not eorreeted for recovery. Replicate analyses war* conducted for 5X of all samples and vara within 9X of the sample mesa, laagant/glasswara blanks were required with each 20 temple*. The minimum detection limit for total PCS* in fish flash was 30 Vg/kg, and all results vsre expressed on a vet-weight basis. Statistical analyses of tha data were conducted with the Statistical Applications System (SAS 1982). The analysis of variance functloo of the CIM procedure was used to compare PCS residues in fillets to fish species, trophic level, lake, fish length (within a species), and sample location (within selected lakas). When a significant difference (p0.0S) was indicated by tha analysis of variance, Duncan's multiple rang* test was used to locate the variance (p-Q.05). Semples found to contain nondetactable levels of PCS ware assumed to contain 13 Wf/kg for statistical purposes. Results Total PCB content in fish fro* tha nine lakes ranged from not detectable to 16,100 vg/kg and varied significantly (p-0.0001) between lakes. Two hundred forty-one (78.3X) of the fillets contained detectable levels of PCB residues and th* mean concentration for all sample* vaa 631 vg/kg. Tha greatest mean concentrations occurred in fish fro* the two lakes with known PCB sources; Ft. Clbson and Webbers Falls. Each of these lakes exhibited Significant variation in PCB residues in fish while the remaining lakes comprised a third group (Table 3). 274 HONS 026315 Table 3. Result* of Duncan's Multiple Range Test for PCB Content of Flah Froa Nine Oklahoma Lakes Lak* Ft. Clbaon Webbers Fells Birch Kjv Keystone Oologah Cufaula Texoma Hulah K 30 27 22 30 30 33 47 91 17 Mean PCB (ug/k|) 3244 1612 430 351 330 240 13* 86 13 Grouping* A B C C c c c c c `Means with the am* letter ere not elgnlfleantljr different (p - 0.05) 27) HONS 026316 When all samples vara grouped by trophic laval, dacrltlvoras had a naan PCS burdan of 781 ug/kg, vhlla carnivora* and planktlvoraa avaragad 312 and 136 ug/kg respectively (Tabla 6). Oatrltlvoraa had significantly graaear PCS raalduai (p*0.010) In thalr flllata than althar carnivora* or planktlvora*. Tha lnfluanca of trophic Laval on tha sample froai aacb Individual laka v* *l*o taatad (vlth tha exception of lulah Laka which did not hava dacaccabla PCS residua* in any aaapla). Datrltlvoraa In four of tha laka* contain*! algnlfleantly graatar (p-0.05) concentration* of PCS than tb* other trophic lavala. In eh* remaining four lake*, althar tbara va* no algnlflcant dlffarnca (p-0.05) batvaan trophic lavala, or datrltlvoraa and planktlvora* vara In a group with significantly graatar (p-0.03) PCS content than carnivora*. Data from the Individual flah apacla* coaprlalng tha three trophic lavala vara analysed to determine If any apaclaa within the lavala contained algnlf 1cantly higher PCB raalduaa. Both river carpauckar and carp flllata contained algnlflcantly graatar (p-0.0097) PCB raalduaa chan tha other apaclaa (Tabla 3). Ho algnlflcant difference (p-0.03) va* found baevaan tha raaalnlng apaclaa. Tha a*an concentration of PCB* In flllata ranged Iron 1099 ug/kg In rlvar carpauckar to 86 ug/kg In white crapple. Data froa aelactad atudy lake* vaa compared by geographical location of . tha aaapla to aacartaln the influence of aaapla location within a laka on the raaultlng PCB burden*. Tha four lakaa; Webber* Falla, Oologah, Eufaula, and Taxoma, each bad bean aaapltd la two or aora widely aaparatad location*, and tha aaapla* within tha laka vara alallar la number and composition. No algnlflcant difference (p-0.05) va* found batvaan lavala of PCB In flllata and gaographlcal location of tha aaapla for any of tha four lake*. Total PCB content va* alao compared to flah length (a* a aaaaura of aga) for each apacla*. Although tha alto rang* of tb* aaapla had bean kept rela tively small, no algnlflcant correlation (p-0.05) vaa found for any apacla*. Indicator Ortanlaaa It la probable that naan PCB content of flah from tha seven lake* vithouc known source* of PCB Input raproaant "background" lavala typical of many Southwestern D.S. lakaa. Similar lavala of PCB raalduaa In fl*h from lake* without known PCB source* have been found by tha Oklahoma State Department of Haalth (Oklahoma Stata Department of Health Kaport 1981). Tha result* of tha present study Indicate PCI residua In fish can b* related to PCB Input Into an aquatle ayatam and such Information may bo sufficiently sensitive to allow differentiation batvaan "normal" concentration* of PCB and alavaead lavala from either permitted nr non-permlttad discharge*. This represents a recant laau* of graae importance concerning PCB and the environment; tha usa of PCB eontant of organisms as Indicators of PCB pollution. Detecting, evaluating, and regulating PCB sources pcses a problem to environmental authorities for several raisons related to tha physical character istics of PCB. Whether eha PCB Input to an aquatic system la via landvaah or 'affluent discharge, the extremely low solubility of tha PCB nolaeula In water (DSEPA laport 1980, Sehoor 1975) and Its affinity for adsorption to suspended particles Insure PCB concentration* In nose waters will b* low. It may be almost lnpoaslbla to correlate any PCB found In tha water with tha magnitude 276 HONS 026317 Table 4. Result* of Duncan's Multiple Rang* Teat for PCI Burden In Fish of Three Trophic Levels Trophic Laval Datrltlvora* Carnivores Herbivores N 208 86 33 Mean PCB (ug/bg) 781 312 136 Grouping* A B B *Maans with the sane latter are not significantly different (p 0.05). 277 HONS 026318 Table 5. Results of Duncan's Multiple Rang* Tast for PCS Burden In Seven Fish Species Species River carpsucker Carp Channel cetfleh Smelloouch buffalo White baaa Clcsard shed White crapple N 41 52 47 64 28 33 42 Mean PCB (Ug/kg) 1099 1055 683 " 4U 401 136 86 Grouping* A A AB B B B B Mesne with the ease letter are not significantly different (p 0.05) 278 MGNS 026319 of the discharge, Ic li iaf to assume most PCS Is eventually deposited vlth sediments vhsn PCB In tha viesr and adaorbad to suapandad particles hava reached equilibrium, Deposition of PCS In eh* aadlaant occurs In rasponsa to a varlacy of factors which Includa thosa ralatad to aadlaant daposltlon as wall aa variations in tha attraction of PCB aolaeulss to particle*. Thus, data obtalnad from sadlaants Is difficult to incarprat and subjact to wlda variations; tha vatl-knovn pr.ar.ooanon of aadlaant "hot-spots" balng an example. At rt. Clbsoo Lake, tha aajority of tha contaainatad sadlaants vara confined to a snail tributary representing no nor* than *5 ha of sadinant. This la only about ona-half of IX of ch* total surface eras of sadlaants In the lake. Evan vlchln ch* contasilnacad cratk, PCB coneantratlons ranged fron 0.23 ug/g (dry weight) to 7.20 ug/g and vara significantly greater at areas of daposltlon suen as ch* jpttraam edge of bars and ch* dlta, as compared with tha tcraaa bottom and edge (Hunter at al. 1980b). Incroaslngl7 ov*r tha past f*v years, investigators have turned to tha biota as a sore of automatic, ela-lae*gratlng sampling davle*. Pish ar* aoae popular although a varlacy of organisms from starlings (Sturnus vulgaris) (Whit* 1979) to mollusks (Butler ee al. 1978) hav* bean used. An excellent review of the problems associated with this approach has bean prasantsd by Phillips (1978). In the Oklahoma data, the lack of algnlfleant difference batwaan sample location and PCB content of flllmts shows the fish warn ade quately distributed to provide a raprasantatlva view of conditions In tha lakes. Onfortucately, eha standard deviations of many of tha aamplas ara quit# large, which is a problem cobou to field observations. If one seeks to compere this data with data prasantsd by other Investi gators, two vary real problems arlsa. Tha first of these la what portion of tha body waa axanlnad. or wbathar tha aatlra organism was analysed, Schmitt at al. (1981) acknowledge criticism of the Hatlone1 Pesticide Monitoring Program for using whola fish tamp las and provides souod reasoning for ch# choice ef whola fish. These Includa tha faet that organisms eat other organ ises sntlrsly, and flilat standards designed to protect humane ara not partic ularly relevant in ecosystem studies. Arguments for using fillets, such at tats ef praparatlcn and comparison to established standard*, ara alao powerful. Ic was this Istttr polcc, tht comparison to established 7DA standard*, that resulted In fillets balng usad In tha present atudy. Tha second problem 1* vhether to arprast tha raaulta on a wac-clsaua weight baal* or a llpld-veighc basis. Although there Is soma Indication various lipid pools within an organism may differentially obtain PCB (RoseveU at al. 1979), reporting on a llpld-welght baals would eliminate much of tha variability reported to tha literature. Theta problems ara important enough to warrant tha aatabllahmant of recomoaoded procedure* by soma authoritative group. Tha atandardltatlou of raaulta to a whola organism (where sis* permit*), fat-weight baal* would be extremely useful for comparison of anvlrotuMntal aamplas and vould aid in any future effort towards setting criteria or standards for PCB In organism*. Assuming atap* art taken to collect comparable deta, Indicator organisms appear to be quit* useful In PCB monitoring programs. Careful planning and s knowledge f the ecological relationships of tb* organisms is rsqulrsd for auecatsiuT us*. The Oklanoma data show* two apaelas, carp and rivar carptuckar, contain significantly graatar (p0.0097) PCB residues than the other fish speclts. It appears certain organisms or groups of organisms within an 279 MONS 026320 environment will contain significantly greater concentration* of toxic sub stances than other organisms and this will occur even If biomegnlficeelon does not operate. Possible causative factors for this Increase are discussed later, but the existence of such species or groups has been demonstrated for PCB (Hunter at ai. 1980a,b) and arsenic (Hunter ec al. 1981). The Identifi cation of these organisms allows monitoring programs to be tailored to trophic levels or Individual species which will reduce the variability associated vlth biological sionltorlng. This technique has the further advantages of reducing costs and/or ailovlng Increased sampling of Che sensitive organism*. Such organisms will vary with geographical region and the toxic substance involved, although there Is reason to believe they would be functional equivalents. A study conducted during 1980 ae Webbers fall* Lake provide* an Ideal example of the usefulness of Indicator organisms. Eight secs of water samples vers collected at monthly Intervals from two depths at each of alx locations. No PCB was detected at a detection limit of 0.1 ug/l In any of these 96 samples. Input of PCB to the upper lake had been occurring at up to 2.5 ug/l for four year* prior to the study. Twenty-five sediment samples were collected from areas of deposition for IS miles dovnstream from the PCB source and analysed for PCB at a detection limit of 0.3 ug/g (dry weight). None of the samples contained detectable reeldues of PCB. A aeeond agency also made ample collections during the time of PCB Input. Concentration* of PCB In the dlaaolved end suspended phases of the water were last chan the datactlon limit of 0.1 ug/l In ill samples. Concentrations In sediments below the source ranged from lest than 1 ug/kg to 16 ug/kg, and avaragad $.2 ug/kg (Stoner 1980). Finally, tha group of 29 fish described earlier were collected from the vicinity of ths permitted discharge and at a point 13 tern downstrearn. There was no significant difference in PCB eontant between ths two sampla locations and PCB contact was significantly graatar (p0.05) In datritlvoraa than other trophic levels. Tha total PCB eontant of tha adlbla portion of tha 17 datritlvoraa rangad from 138 to 5940 ug/kg (vat valght). Thus, no informa tion was obtained from the analysis of tha vatar and sadlaent, but contamina tion of the biota at levala above aurroundlng lakes wee shown by PCB concent of the detrltlvoree. The Identification of groups or spaclaa with increased lavala of PCB suggests standards or criteria could ha aac using chase organisms an indicator of axcatalve PCB Input. Tha prasant data base Is not sufficient to allow recommendation of such a atandard, but le could be based on tha concen tration In the moat aanaltlva indigenous fish species. This Is analogous to terminology presently usad In many bloaiaav-baaad standards. This approach haa the advantages of bypassing tha insolubility problam of vatar standards, eliminating tha eonfualon of "het-apoea" in eadlmenc standards, maaaurlng actual lavala In tha biota, and representing actual process** of mixing and removal In tha syatam. Tha major disadvantages prssar.tly art ths previously mentioned lack of a data beta and the difficulty In comparing rtsults based on differing measurement and reporting systems. Oklahoma la praaantly moving In tha direction of diversified standards for PCS that racognlx* some of the .problem described ahpve, Ths.J.979 Oklahoma water quality standards daslgnata an aqueous standard and alao label concentrations of PCB In fish tissue exceeding 2000 ug/kg as a causa for concern and further investigation (Oklahoma Veter Resources Board Publ. 101 1979). Furthermore, the Oklahoma legislature la considering authorisation for the establishment of PCB standards based on 280 MQNS 026321 sediment concentration*. Such * combination approach say b a method for overcoming the llaltatlone of PCB regulation baaed oo Individual component* of the environment. Processes of Accumulation In the Blot* The ilgnlflcanc dlffereace In PCB level* beeveen the detrltlvore* and ocher trophic level* and between che carp and river carptucker and ocher ipecle* Indicate* the mechanism producing elevated PCB burden* 1* related to both lngesclon of tedlment* during feeding activltlee and adsorption of PCB across the skin. River carpiucker and carp have been found to ingest large amounts of bottom material during their feeding activities; various studies have found sand and detritus In a large pareancaga of the stomachs of bottom feeding fish (Wrenn 1968 and Mather 1970). Carp vnr* thought to feed primarily at the mud-vatar Interface by Sumaerfelt at ol. (1970). The smallest, least settleable pardcles would be found at this Interface and PCS* are asaoclatad most strongly with such particles (USEFA Report 440/3-80-068 1980). Thus, the opportunity for both PCB partitioning ineo body lipids and Ingestion of contaminated food and sediments would be maximised in this region. Whlla the size rang* of fish usad In this study was relatively small, tha lack of ' correlation batveen fish lsngth and PCB content of fillets within t species is further indication tha mechanism by which PCI is obtained is more complex than mare Ingestion. It is probable individual fiah coma into contact with and lngtst PCB laden material In a manner unrelated to aga or alta. One can hypothesis* fish obtaining a baaa concentration of PCB from water which is ralatad to tha lipid content of tha individual fiah. Another portion of tha total body burdsn of tha individual would be lagestad in the form of food, tedlmenca, cr suspended particlaa. The above hypothesis txplains the ecnalstant relationship of detritlvores, or detrltlvore* end planktivorss, containing algnifleantly greater PCB raalduaa than carnivorous spacias, but only it ona asaumas blomagnlfleatlon la not operating. As previously stetsd, PCI input to dstritivorss la maximised in their feeding ictlvltlss. Tha PCB Input to planktlveroua epeelss would also be Increased through ingestion of PCB-laden parcleles during filter feeding, however, If excretion is slew (there is conflicting evldanca on thle point, Callfano at al. 1980 and Malar 1982), or at least similar among fish apaclae, than the carnivorous apaclae would contain the greatest lavala of PCB if blomagnlficaelon occurred. This la especially cru* for tha carnivorous white base which la rather pray specific to tha plsnfctlvoroua glssard ahad la reservoirs. Difference* in lipid content between eh* trophle lavala do** not explain thle phenomenon. In tha subeample from Ft. Gibson Lake, where lipid concent vae specifically measured, Carnivorous specie* had a slightly higher aaan lipid content than tha other trophic lovols. Furthermore, tha lipid content of tbo samples vae not significantly correlated (p*0.03) with PCI content of carnivorous species, but vae correlated with PCB content la soma species of dotrltlvoros. This Is evidence blomagnlfleatlon of PCBs doa* not occur and apparane Increases with trophic level sr* related to interrelation ship* batveen levels in the food and environment, feeding habits, and lipid content of tha organisms A recent study (Melee 1982) llluseraed* the ability of organisms tj4 accumulate PCB. Midges (Chlronomus plumoeue) exposed to l and 10 Ug/g C-PCB 281 HONS 026322 contaminated sediment* developed significant body burdens of PCB from Ingestion of cooculaited ledlment, end there wet t close relationship betveen substrate concentration and body burden. Fathead minnows fed these aldges also accuaulatad PCB, reaching a maximum within 30 days. The study dsaoostrated organises can develop significant body burdens of PCB from the Inges tion of contaainated sediments and pass these to higher trophic levels. PCB IM WILD AW DIAL POPPLATIOHS There exist* a large data base concerning PCB levels In portions of wild anlaal populations. A study conducted on the Snake River shortly after the collapse of the Teton Daa found PCB concentrations In rainbow trout averaged 1010 ug/kg. Utah suckers (Catostoaus ardent) averaged 1800 ug/kg, and Rocky Mountain vhltsflsh (Prosoplua wllllaasonl) had a aean PCB content of 1400 ug/kg (Perry 1979). A sample of 58 fish froa major O.S. watersheds contained PCB residues ranging from not detectable (<300 ug/kg) to 140 mg/kg on a whole fish (vat weight) basis (Valtk at al. 1979). About 93X of the samples contained measurable levels of PCB. Commercially harvested mullet from the upper Great lakas contained PCB residues of 60 ug/kg to 790 ug/kg In adlbla portions (vet weight) (2ablk at al. 1978). A series ef whole body aggregates of six bottom feeding fish species from Pennsylvania waters contained PCB at 40 to 1630 ug/kg (vet weight). This was substantially gtester than similar composite* of carnivorous fish which had PCB residues of 30 to 860 ug/kg (Kurts 1978). Almost 75Z of Utah suckers collected from American Pall* Reservoir. Idaho. In 1973 had fungus-llke growths on the head and sides which the authors attributed to PCB (Kent and Johnson 1979). In that lake, Utah suckers over two-yesrs old were the only fish species with detectable levels of PCB. This was believed to result from exposure to contaminated sediments. Am Interesting study of the Budson River used female American shad (Aloes sapldlasima) on their spawning run to queotlfy PCB*. tinea the shed do net eat upon entering frashvetar. the difference between pre- and post-entry PCB burdens vat used as an Indication of PCI In the water. Shed from ehe lover station averaged 2000 ug/kg (whole fish, wee weight) while those from the upper river teatlon had a aean PCB eontsnt of 6100 ug/kg (Pastel *e el. 1980). All 206 rod-breasted mergansers (Mertua serrator) eggs collected In 1977 and 1978 from Lake Michigan contained PCB at a naan concentration of 22,800 ug/kg (Haeeltloe et al. 19S1). Theee levele wera balow concentration# which had an effect on mallard dock agge (Haealtlne and Prouty 1980). Studies of mall makers ef crows (Coruue hrechrrhrnchoe), coots (fuller aaerlcane), sterlings, end franklin's gulla**7tarua plptxcsn) from South Dakota found no PCI at e detection limit of 100 ug/kg (Grolchuo at al. 1970). In a sample of 330 piscivorous birds of 10 spoclas from s polluted Finnish Isks, assn muscle concentrations by species ranged from 220 ug/kg PCB ts 13.490 ug/kg. Mean PCB contsnc of tbo llvor tissue from tbesa birds was similar eo muscle concentra tion* (Ssrkka at al 1978). Residues of PCB found In leee eounonly examined portion* of the biota Include samples of six earthworms with no dotoetabla PCB (<S00 ug/kg). 22 snakao ranging from 1300 to 3800 ug/kg (wot weight), and four noetling birds containing 850 to 1300 U|/kg PCB (Bolnx ot al. 1980). Harp seels (Phegophllus 282 MONS 026323 groenlsndlcus) collected from the Gulf of St. tevrence In 1971 and 1973 contained 70 to 13,300 ug/kg (vet velghc) In various tissues (Rosavall at al. 1979) . A survey of liver and muscle tissue of mammals collected In southern Ontario found ?C3 concentrations In 138 fox (Vulpes ipp.) and 105 raccoon (Procvon lotot) ranged from 28 to 43 ug/kg. Carnivores such as 25 marten (Hartes amerlcana) and 20 mink had residues of 150 to 270 ug/kg, vhile 15 fisher (Martes pennant!) svengea 500 ug/kg PCS (Frank at al, 1979). The nine eggs of loggerhead turtles (Careers caratta) collected from Florida all, contained PCB at concentrations o( 32 to iOl Ug/kg. Neither of the tvo eggs from green turtles (CheIonia avdas) examined contained PCS above the 25 ug/kg detection level (Clark and Krynltsky 1980). Hose of the 19 smell cetaceans collected from southern California contained PCB at levels associated vlth reproductive impairment in pinnipeds (O'Shea at al. 1980), Spatial and Temporal Patterns The Identification of specific geographical areas vlth Increased PCB contamination It difficult due to relative Isolation of many sources, movement of the blocs, and cne Impact of Improved quantification techniques. Some species of fish from take Erie end Lake Seine Clair exceeded Canadian standards, but there vas no trend betveen 1968 and 1976 for the various fish apecles (Frank at al. 1978). Residues In fish from Lakes Buron and Superior also did not display recognisable temporal trends betveen 1968 and 1976 (Prank ee al. 1978). Concentrations of PCB in Cannete (Morue bsssanus) eggs decreased over six-year period from 7.7 ug/g (vet velghc) to 3.5 ug/g (Flmreite at al. 1980) . A similar study of vhlte-tallad aagla (Hallattua alblallla) tggs from Scandinavia shovad no decraase In PCB content over a (our-year period (Kolvueeerl at al. 1980). The National Pesticide Monitoring Program hat produced a veslth of Information relating to spatial and temporal trands of PCB contamina tion of fish (Velth at al. 1981. Velth at al. 1979, Butlar and Schutsmenn 1978) as veil ae ocher organisme (White 1979, Butler at el. 1978). Although a revlev of the Information obtained In chat etudy la beyond the scope of thla paper, tvo Important polnes can be made. The first Is that the majority of the biota of the O.S. Is aubjact to a relatively smell base-load of PCB contamination, and populations vhoee members are completely devoid of body burdens of ?C3 at modern detection levels ere a alnorley. The second point Is most long-term studies Indicate a general dovnvard trend in PCB contamination of the biota over the peat decade. CONCLUSIONS Baaed on the previous revlev of literature end original data, the follovlng conclusions and recommendations ere made: 1. Sloessey data on PCSa Is lacking for large groups of organisms. The use of "tendardlsed" species of bloasasy organisms such as fathead mlnnovs end reinbov trout facilitate comparisons betveen toxic substances but give llttls Indication of the effects of PCB on diverse organisms. Data is particularly needed for large mammals and various trophic levels of birds, Studies of the effects of PCB on progeny should be conducted. 283 MONS 026324 2. Thar* It constant partitioning of PCB batvacn various cosrpooants of the anvlronaant. Two neehanlsat, partitioning Into body lipids and Ingestion, control PCB residues In the aquatic biota. 3. PCBs are not aubjece to bionagnlflcatlon but readily bloconeantrata. 4. The majority of the biota In tha tJ.S. Is subject to a relatively small base-load of PCB contamination. Certain organisms, due to their habits, vlll concaln substantially elevated PCB burdens. 5. Small amounts of PCB can inhibit production of phytoplankton, adversely effecting entire aquatic scosystems. This raprtsents a major threat to tha environment and warrants cars In astabllshlng dlscharsa raoulramants. 6. It Is probable soma O.S. populations of sensitive mammals, sucn as mink, era being edvereely effected by exposure to PCB. Thaee affects would primarily be chronic and ralatad to reproduction but some acute lethal affect* era also likely. 7. It la doubtful PCB* ere edvereely effecting reproduction of flah in most U.S. waters. Tha more PCB sensitive fish, such as salmonlds, frequently have been eliminated by other human actions prior to the use of PCB. PCB levels might hinder reestablishment of such species. Other affects, such as a decrease In tha mechanical propertlea of vertebrae have been ehown to occur. 8. Most bird populations In tbe O.S. would not display acute effects from PCIa in the environment. Many species of piscivorous birds are exposed to dietary levels of PCB sufficient to cause metabolic effects In their offspring. 9. A lack of standardised reporting methodologies has reduced the ueefulnees of field data on PCBs. The establishment of tecommended sample types end reporting methods by an authoritative group< Is needed. These recommenda tions should be based on environmental aspects of PCB contamination nor strictly related to human health. 10. Determination and regulation of PCB In the environment is made difficult by physical characteristics of tha PCB molecule. The relative Insolu bility of PCB la water ead the affinity for adsorption to euopendod particles makes correlation of PCB end discharge magnitude difficult. Likewise, deposition la sediments Is not otrlctly related to PCB input. 11. The use of Indicator organisms hat become Increasingly refined over the pest few years. It may be possible in the near futur* to see PCB standard* based oa coaceatratlons In apecles which aost readily obtain PCB. The process of Identifying such species would be greaely facilitated by a central data bees such es BIO-STOBET. 12. Mott long-term, studies Indicate a downward trend In PCB contamination of tha biota over the past decade. 284 MONS 026325 REFERENCES Armstrong RV, Sloan RJ. I960. Trend* In levels of several knovn cbenical concaalnants In fish from New York Scats. New York Dept. Environ. Conaarv. Tech. Raport 80-2. Blasaaann, A. 1982. Effaces of PCB* on gonads, sax hormone balaoca and reproduction procassas of Japanese quail Coturnlx cocurntx japonlca aftar lngaatlon during saxual security. Environ*. Foil. (A) 27:15-30. Biggs DC, Povars CD, Rowland RC. O'Connors RB, Vurscer CF. 1980. Dpeake of polychlorinated biphanyli by.getural phytoplankton assaablagas: field and laboratory determination of C-FCB particle-vater lndax of sorption. Environ*. Poll. (A) 22:101-110. Blrga VJ. 1979. Toxicity of organic chaalcals to aabryo-larval scagas of flab. USEFA Raport $60/11-79-007. BJerk JE, Bravlk EM. I960. Organochlorlne eoapounda in aquaele environment*. Arch. Environ*. Cones*. Toxicol. 9:743-750, Blaavina MI, Auiarich RJ, Ringer U. 1960. Polychlorinated blphanyls (Aroclors 1016 and 1242): offsets on survival and raproduetlon in *lnk and ferrets. Arch. Environ*. Cones*. Toxicol. 9:661-665. Blus LJ, Laaont TC. 1978. Organochloriu residues in six species of estuarine birds. South Carolina, 1971-75. Paae. Monitor. Jour. 13:56-60. Bravlk EM, 1981. Organochlorlns residues in fish fro* lake MJosa la Honray. Bull, Environ*. Cones*. Toxicol, 26:679-680. Bryan AM, Olafsson ?G. 1978. The affect of polychlorobipbanyla (Aroelor 12*42) on blcarbonsta-14 uptake by Eutlana tracllls. Bull. Environ*. Contaa. Toxicol. 19:374. Butler PA, Kauoedy CD, Sehulecaann RL. 1978. Pesticide residues in estuarine olluaks, 1977 versus 1972 - Rational Pesticide Monitorial Progra*.. Past. Monit. Jour. 12:99-101. Buelar PA, Schutzmanu IL. 1978. Xaslduas of paatlcldas and PCBa in estuarine fish, 1972-76 - National Pesticide Monitoring Progra*. Past. Moolt. Jour. 12:51-59. Califano RJ, O'Connor JM. Paters LI. 1980. Uptake, retention and elimination of PCB (Aroelor 1254) by larval striped baas (Morons saxatllls). Bull. Savlron*. Contaa. Toxicol, 24:467-472. Christensen ER, Zieleki PA. 1980. Toxicity of arsanic and PCB to a green algaa (Chlaaydoaonas). Bull. Envlrooa. Contaa. Toxicol. 25:43-48. Clark DR. 1978. Uptake of dietary PCB by pregnant big brown bats (Eptaalcua fuacua) and tbalr fetuses. Bull. Environ*. Contaa. Toxicol. 19:707-714, 285 MONS 026326 Clark OR, Krynltaky A. 1978. Organochlorlna raalduaa nd reproduction in the little brovn bat, Laura 1, Maryland - June, 1976. Paat. Monitor. Jour. 12:113-116. Clark DR, Krynltaky AJ. 1980. Organochlorlna raalduaa in eggs of loggarhaad and graan aaa turtlaa naatlng at Karrltt Island, Florida - July and Auguat 1976. Paat. Monlt. Jour. 14:7-10. Courtuay VA, Langaton HJ. 1978. Optaka of polyehlorinatad blphanyl (Aroclor 1254) fros aadlsent and fros eeevater in tvo iatartidal polyehaataa. Environs. Poll. 15:303. Cuatar TV, Halnt GH. 1980. Reproductive aueeaaa and naat attantlveneee of mallard duck for Aroclor 1256. Environs. Poll. (A) 21:313-318. Dafoa DL. 1978. Effacta of Aroclor 1248 and 1260 oo tha fathead alnnov (Piaanhalla prose let). Jour. Flat). Ra. Board Can. 35:997. Fiaralta M. Kvaaath H, Bravlk EH. 1980. Mercury, DDE, and FCBa In agga froa a Norwegian Cannatt colony. Bull. Envlronai. Contaa. Toxicol. 24:142. Plngernen SV, PInganun M. 1979. Coaparlaon of tba affacta of fourtaan-day and chronic axpoauraa of a polyehlorinatad blphanyl, Aroclor 1242, on Bolting of the fiddler crab. Oca putllator. lull. Environs. Contas. Toxicol. 24:696-703. Plngaraan SV, Plngaraan M. 1978. Influancaa of tho polyehlorinatad blphanyl preparation Aroclor 1254 on eolor ehangaa of tha fiddler erab Oca nuillator. Mar, Biol. 50:37. Plngaraan SV, Ruaiell 1C. 1980. Effacta of tha polyehlorinatad blphanyl Aroclor 1242 on locosotar activity and on tha naurotranaslttara dopasina and noroplnaphrlna la tha brain of tha Culf killiflah, Pundulua grandia. Bull. Eovlrons. Concas. Toxicol. 25:682-687. Prank R, Braun BE, Holdrlnat M, Dodge DP, Nepaiy 8J. 1978. Raalduaa of organochlorlna inaacticidaa and polyehlorinatad blphenyla in flah fros Lakea Saint Clair and Erie, Canada - 1968-76. Paat. Monlt. Jour. 12:69-80. Trank I, loldrlnet M, Braun BE, Dodge DP, Spranglar CE. 1978. Raalduaa of organochlorlna lnaactleldaa and polyehlorinatad blphanyla In flah fros taken Buton and Superior, Canada - 1968-76. Paat. Monlt. Jour. 12:60-66. Prank B, Boldrinet KVB, Suda P. 1979. Organochlorlna and sarcury raalduaa In wild nassala la Southern Ontario, Canada, 1973-74. Bull. Environs. Contas. Toaleol. 22:500-507. Purnaaa R. Button M. 1979. Pollutant levala In tha great akua Catharacta ikua. Environs. Poll. 19:261-268. Craichua TA, Cuack BD, Aasan BD. 1978. Organochlorlna inaactlcida, poiychlori natad blphanyl, and saeal raalduaa In tone South Dakota blrda, 1975-76. Paat. Monlt. Jour. 12:4-7. 286 HONS 026327 Hamilton $J, Mehrle PM, Mayer FL, Jones JR. 1981. Mechanical properties of bona In channel catfish as affected by vitamin C and toxaphene. Trans. Amer. Fish. Soc. 110:718-724. Harding LW, Phillips JH, 1978. Polychlorinated biphenyl effects on marine phytoplankton photosynthesis and cell division. Mar, Biol. 49:93. Ksseltlne SD, Helm CH, Relchel W l, Moore JF. 1981. Organochlorlne and metal realdues In eggs of vaterfovl nesting on Islands In lake Michigan Off Door County, Wisconsin, 1977-78. Pest. Monlt. Jour. 13:90-97. Haseltlne SD, Prouty RM. 1980. Aroclor 1242 and raproductlva success of adult mallards (Anas olatyrhynchos). Envlrona. Res. 23:29-34. Hactula ML, Janatulnen J, Sarkka J, Paaslvlrta J. 1978. A flva-ytar monitor ing study of the chlorinated hydrocarbons In the fish of a Finnish lake eco system. Envlrona. Poll. 13:121-139. Heinz GH, Haseltlne SD, Hall RJ, Krynltsky AJ. 1980. Organochlorlne and mercury residues In makes fron Pilot and Spider Islands, lake Michigan 1978. Bull. Enlvonm. Contan. Toxicol. 23:738-743, Hunter RC, Carroll JH, Butler JS. 1981. The relationship of erophlc level to arsenic burden In fish of a Southern Great Plains Lake. Jour. Fresh. Scot. 1:121-127. Hunter RiJ, Carroll JH, Randolph JC. 1980a. Organochlorlne residues In fish of Lake Texoaa, October 1979. Pest. Monlt. Jour. 14:102-107. Hunter RG, Randolph JC, Carroll JH. 1980b. Polychlorinated biphenyls In the sedlnenta and fish of an OV.lshooa stream. Envlrona. Poll. (B) 1:233-240. Johnson DW. 19(8. Pesticides and fishes - a review of selected literature. Trans. Aaer. Fish. Soc. 97:398-424. Rent JC, Johnson DW. 1979. Organochlorlne residues In fish, water, end sediaent of Aaerlcan Falls Reservoir, Idaho, 1974. Peat. Monlt. Jour. 13:28-33. Rlappal CS, McLaughlin JJA. 1980. PCS toxicity to phytoplankton: affects of dose and density-dependent recovery responses. Bull. Envlrona. Contam. Toxicol. 24:696-703. Kolvuaaarl J, Nuuja 1, Pelokangas R, Flnnlund M. 1980. Relationships between productivity, eggshell thickness, and pollutant contents of addled eggs In the population of white-tailed eaglas Baileetua alblcllla l. In Finland during 1969-1978. Envlrona. Poll. (A) 2374l-ii. Kosutzky J, Adaaec 0, Bobokovs E, 1979. Effaces of polychlorinated biphenyls on poultry reproduction. Bull. Envlrona. Contaa. Toxicol. 21. Kurtz DA. 1978. Residues of nolychlorlnsted biphenyls, DDT, end DDT netabolltes In Pennsylvania streams, community water supplies, and reservoirs, 1974-1976. Pest. Monlt, Jour. 11:190-197. ' 287 MONS 026328 Kuvabar* K, Nakanura A, Kashlnoto T. 1980. Efface of pstroltua oil. paatietdsi, PC8s, and othar tnvlronstntal contanlnsnes on th# hatchsblllty of Artanla aalin* dry tgga. lull. Environs. Contas. Toxicol. 25:69-74. Lvt SA, Hopkins KD, Vhits TP. 1968. Avarags growth racaa and langch-valght ralatlomnlpa of aixtaan apaclaa of fish In Canton Raaarvoir. OK. Dapc. Ulldllfa Cooiarv. Raport. llndvail ML, Lov JB. 1980, Efface* of DDE, TDE, and PCS* on shall thickness of vastam graba aggs, Baar Rlvar Migratory Bird Rafuga, Utah - 1973-74. Paae. Monitor. Jour. 14:108-111. Mae HJ, Scaly* JO. 1981. Pattarna of PCB accunulielon by Pty of lake trout. Bull. Envlroos. Coots*. Toxicol. 27:368-373, Mathar D. 1970. Pood habits and faadlng chronology of channal catfish Iccalurua punccacu* (Raflnosqua) in Conovingo Raaarvoir. Proc. Southaaat. Aasoc. Can* Pish Con. 24:377-386. Mauck VI. 1978. Effaces of tha polychlorlnatad blphanyl Aroelor 1254 on growth, survival, and bon* davalopaant In brook trout (Salvallnua fontlnalla). Jour. Fish. Raa. Board Can. 35:1084. Mayor PI, Mahrla PM. Cmtchar Ft. 1978. Intaractlon* of toxaphaoe and vltaslo C In channal catfish. Trana. Aair. Plah. Boo. 107:326-333. Mayor Ft, Mahrla PM, fchoattgorl RA. 1980. Tranda In aquatic toxicology in the Dnltad Seat**: a parspaetlv*. In Procaadlngs of tb* 3rd 0SA-0SSR Sysposlua on eh* Bffaeta of Pollutanta Open Aquatic Ecoayatana: Thooratleal Aapoct* of Aquatic Toxicology. Duluth, MM: 0.8. Environmental Protactlon Agency, p. 241. McL**** DV, Metcalf* CD. 1980. Toxlcltlaa of aigbt organoeblorina compounds In aadlaant and aaawatar to Cranton aapcasaplnoaa. Bull. Environs. Contas. Toxicol. 25:921-928. Mciaasa DV, Maecalt* CD, Posaaek DS. 1980. Opeaka of PCBs fros sadlsaot by Haral* vlrana and Cranton aaptasaplnoaa. Arch. Portions. Contas. Toxicol. JsSOT-sTT McLano MAR, lughos Dl. 1980. laprodueelva auccaaa of acraach ovla fad Aroelor 1248. Arch. Environs. Contas. Toxicol. 9:661-665. Mahrla PM, lainaa TA, IasiIton 8, ludke JL, Mayor PL, Ribick MX. 1912. Relationship baevaan body eontaslnanta and bon* davolopsane in Eaat-Coaat tripod baas. Trana. Aaar. Plah. Soc. 111:231-241. Mahrla FH, Mayor Ft, Bucklar DR. 1981. Xapon* and slrax: offsets on bona davalopaant and avia bladder cospoaition in fatbaad alnnova. Trar.a. Aaar, Pish. Soc. !-K>i)8-6A3. Malar PC. 1982. Btoaccuautatlon of PCS'* by food vab concentration. Thlrtssnth Annual Moating of tha North American Banthological Society. 288 HONS 026329 Klchaals RA, Rev land RG, Hurstsr CF. 1982. Polychlorinated biphenyls (PCB) Inhibit photosynthesis par call In tha urlna dittos Thslsssloslra ptsudonsna. Environs. Poll. (A) 27:9-16, Harbonna JT, 1979. Polychlorinated blphanyl accutrulatloo In gray sullst* (Chalon labroaua); affacta of aga. Bull, Environs. Coneas. Toxicol. 22:65-68. Nsv York Dept. Environ. Conaarv. Rtporc. 1979, Analyala of tha fata of PCBa In tha acoayatas of tha Hudson actuary. O'Connara HB. 1978. Polychlorinated blphenyla say altar sarlna trophic pathvaya by reducing phytoplankton alaa and production. Science 201:737. Oklahosa State Dapartsane of Health Report. 1981. Draft report on sonltorlng toxic aubatancea In flah. Oklahosa Hater Resourcea Board. 1979. Oklahosa'a vatar quality atandarde 1979. Oklahosa Uatar Raaourcaa Board Publ. 101. O'Shea TJ, Brovnell RL, Clark OR, Halkar VA, Cay ML, Lasont TC. 1980. Orgsnochlorine pollutant* In nail catacaan* fros tha Pacific and South Atlantic oceans, November 1968 - June 1976. Past. Konlt. Jour. 16:35-66. Pastel M, Bush B, Kls JS. 1980. Accumulation of polychlorinated biphenyls In Asarlean shad during thalt migration In tha Kudson River, Spring 1977. Pact. MonIt. Jour. 16:11-22. Pearce PA, Peakall 08, Reynolds LM. 1979. Shall thinning and residua* of organochlotlnes and ssreury In ssablrd eggs, Eastern Canada, 1970-76. Peat. Monitor. Jour. 13:61-68. Parry JA. 1979. Pesticide and PCB residues In the Upper Soaka River acoayatas. Southeast Idaho, follovlng tha eollapaa of the Teton Das - 1976. Arch. Environs. Contaa. Toxicol, 8:139-159. Phillips DJH, 1978. Osa of biological Indicator orgaatiss to quantitate organochlorlna pollutants In aquatic anvlronsants-A review. Environs. Poll. 16:167-226. Rosawall ET, Muir DCC, laker BE. 1979. Organochlorlna residues in harp seal (Phasoohllus srosnlandlcus) Tissues, Gulf of St, Laurence, 1971-1973. Psst. Monlt. Jour. 12:119-192. Sangalang Cl, Prsasan 1C, Crovoll I. 1981. Testicular abnormalities In cod (Cadus sorhua) fad Aroclor 1256. Areh. Environs. Cootas. Toxicol. 10:617-626. Sarkka J, Bactula ML. Jasatulaan J, Paaalvlrta J, Palokaagaa R. 1978. Chlorinated hydrocarbon* and sarcury la birds of Lake Pa1Janna, Plnland 1972-76. Past. Monlt. Jour. 12:26-35. SAS Institute tac.. 1982. SAS Osar'* Guide: Statistic*. 1982 Edition. SAS Institute Inc., Cary, NC. 586 pp. 289 MGNS 026330 Schmitt CJ, Ludka JL, Walsh OF. 1981. Organochlorlne residues In fish: National Pastlclda Monitoring Program, 1970-74. Past. Monit. Jour. 14:136-206. Schoor WP. 1975. Problems assoclattd with lov-solubillty compounds in aquatic toxicity tests: theoretical model and solubility characteristics of Aroclor 1254 in water. Water See. 9:937. Skaa JC, Slmonln HA, Dean HJ, Colquhoun JR, Spagno JJ, Velth CD, 1979. Bioaccumulation of Aroclor 1016 In Budson River fish. Bull. Environs:. Contaa. Toxicol. 22:332-336. State of Oklahoma Report. I960. Reporc of ehe PCB Casio force to Governor George Nigh. Stauffer TM. 1979. Effects of DDT and PCBe on survival of lake trout eggs and fry in a hatchery and in lake Michigan, 1973-1976. Trane. Amer. Fish. Soc. 108:178-186. Stoner JD. 1980. Reconnaissance of polychlorinated biphenyls in ehe Arkansas River between Muskogee and Webbers Falls Lock end Dam, Oklahoma. USGS Open-File Report 80-216. Sunmerfelc RC, Mauck PE, Menalnger G. 1970. Food habits of ehe carp, Cyprinua carolo l.. in five Oklahoma reservoirs. Proe. Southeast. Assoc. Game Fish Comm. 24:332-377. Tejedor MC. 1978. Oxidative metabolism in Saccharoarces carevissiaa as affected by polychlorinated biphenyls. Bull. Environs. Contam. Toxicol. 22:439. Tori CM, Mayer LP. 1981. Effects of polychlorinated biphenyls on the metabolic rates of mourning doves exposed to low ambient temperatures. Bull. Environa. Contam. Toxicol. 27:678-682. .. L'SEPA Raporc 440/5-80-068. 1980. Ambient water quality criteria for poly chlorinated biphenyls. QSEPA Report. USFDA. 1972. Pesticide analytical manual. Pol. 1. van den Broak VLF. 1979. Seasonal levels of chlorinated hydrocarbons and heavy metals in fish and brown shrimps from the Medway estuary. Rent. Environs. Foil. 19:21-38. Velth GD, Ruehi DW, Leonard EM, Pugllsl FA, lemke A. 1979. Polychlorinated biphenyls and other organic chemical residues in fish from major watersheds of the United States, 1976. Peat. Monit. Jour. 13:1-12. Veith CD. Rushl DW, Leonard EN, Welch R, Pratt G. 1981. Polychlorinated biphenyls and ocher organic chemical raslduas in fish from major United States -watersheds tear the Great Lakes, 1978. Pest. Monit. Jour. 13:1-8. White DH. 1979. Nationwide residues of organochlorlne compounds in starlings (Sturnus vulgaris). 1976. Pest. Monit. Jour. 12:193-197. 290 HONS 026331 Urtnn WB. 1968. Llf hiatory aspects of amallaouth buffalo and freshwater druffl In Wbaelar Reservoir, Alabama. Proc. Southeast. Assoc. Gam* Fish Cobbs. 22:479-495. ZablV ME, Olson B, Johnson TM. 1970. Dlaldrin, DOT, PCBs, and aereury ltvala In freshwater aullat from the Upper Great Lakes, 1975-76. Past. Monlt. Jour. 12:36-39. Zltko V, Saundara RL. 1979. Efface* of PCBs and other organoehlortn* eompound* on tha hatchabiliey of Aelaneic aalaon (Salao aalar) eggs. Bull, Envlrona. Coneam. Toxicol. 21:125-130. Zullal H, Banacka G. 1978. Application of a new bloaasay to acraan eh* toxicity of poiychlorloatad blphanyls in blua-graen alga*. Bull. Envlrona. Contan. Toxicol. 20:786. 291 MONS 02633Z DISCUSSION SUMMARY M. J. Melaneon Department of Pharmacology and Toxicology Mad leal College of Wisconsin Milwaukee, Wt 33226 DR. MFLANCON! First, I will makt a few brlaf comments on some of tha Items dlacuaaad by Hr. Huntar. Hi* first allda. which ahovad aoaa of tha published toxicity data for PCBs vlth phytoplankton and to on, contalnad data on PCS affacta at watar lavala which ara in aoaa lnatancas such graacar than tha watar aolubllley of PCBs and in aost casta graatar thao lavala of FCBt aaan In tha anvlronaant. Rather chan ng/1 or vg/l concentrations of PCBs typical lavala in tha anvlronaant ara ng/l (Pavlov and Daxtar 1971), m In teraa of environmental lethality of PCBa to phytoplankton other than In Isolated lnatancas this la not likely to ba a problem. Tha second slide showed soma data on toxicity tests of PCBs In fish. Tha concentrations of PCBs which had observable affects ware typically higher than environmental lavala. One of tha probleme experienced by Mr. Hunter In hla studies was that tha assays for PCBa In watar available to him could not quantify PCB concentra tions lass than 0.1 pg/1. Thus In all of tha lakes la Mr. Banter's studies tha PCBa In tha vatar were too low to be quantified, and without tha ready availability of more sensltiva methods for assaying water for PCBs It will not ba poaelble to tall what role tha amount of PCBs In tha watar has on tha results of various studies. I would nov like to describe some work which I am doing and work which has been dona In soma other laboratories which ralata to PCBs In fish In the environment and to some of the Items discussed In earlier sessions. Dr. Frlass mentioned that one of the toxic affects of PCBs that was pratty universally found was Induction of entyme activity, and one example ha mantlooed was induction of monooxygenase activity. Dr. Safa then related cytochrome P448-type Inducers to tha 2,3,7,8-tetraehlorodlbentodloxln (TCDD) cytosolic receptor and toxicity. For those of you who ara not familiar with Induction In fish, fish respond In terms of cytochroms P-450 Induction and aonooxygenase activity only to P448-type (TCDD-type or 3-*ethylcholaothrene-(3-MC)type) inducers. So, in other words, of tha plethora of PCB eongenars only the ones that are the P448-, the TCDD-, the 3MC-typa lnducera can Induce monooxyganase activity In fish. Thus, if va study Induction of monooxyganase activity In fish and gat induction if has to have bean caused by ooa of this type of Inducer, one of those structures which Dr. Safa has shown as'being of the type chat bind* to tha TCDD receptor and praeumably somehow causes toxicity la higher organisms. 292 MQNS 026333 Another question thst vas raised this nomine vas whether anyone can detect any sort of effects resulting fron exposure to low levels of PCBs. I would like to present a couple of example* of this. First, I would like to mention a one work that vas done by the Department of the Interior Jointly by e group fro* the LaCrosse, WI laboratory and a group from the Colunbla, HO laboratory (Bills at al, 1981), They exposed rainbow trout In a flowing system to Aroclor 1254 at levels which were not environmental levels, but which approeched environmental levels more closely than most studies. Cne exposure vas at 10 ng/1 end the other wee at 100 ng/1. After 30 days of exposure the lower level exposure resulted in 0,28 mg/kg whole fish level PCBs, and the higher level gave 2.31. In each case over tvcnty chouaand-fold bloeccuoulation In a period of 30 days. They then tested these fish with nine fishery management chemicals and in three or four of those cases they found the toxicity was significantly changed by exposure to PCBs at those levels for one month. So, ve have an example of PCB exposure at moderately high levels affecting a biological system. 1 would like to mention now briefly some work which I have done thet also relatas to this. Remember, I mentioned that fish respond with Induced mono oxygenase activity only to the 3MC-type or P448-type lndueers. In these studies small rainbow trout were utilised for a dose response curve of hepatic microsomal monooxygenasa activity In resporae to a range of doses of Aroclor 1254 administered intraptritoneally (IP) In corn oil. I started dosing at .025 mgs per kg, ar.d Increased dosage until the response reached maximum and began decreaalng. The activity peaked at lata than 100 mge par kg, and.waa vail on lta way down than. The enzyme jsssy utilized wet ethoxyresorufin-0-deethylse activity, vhlch 2r. Safa referred to this morning. A significant elevation vat found at about 0.3 mgs per kg. This was a (ingle IP injtetlon of Aroclor 1254 at 0.3 mga par kg and the fish wars sacrificed five daye lacar and mlcroaomaa prepared by atandard methods and the assay done. We also looked at a different inxyae activity, ethoxyeoumarlnC'deethyitse activity, and hart again at 3.3 age par kg theta vas a sig nificant elevation la monooxygenete enzyme activity. A similar experiment vaa dona lr. carp. Sacrificing vai dona at three days, five days and nine days after IP injection of Aroclor 1254. Hart again v observed significant elevation of ethcxyreeorufin--deethyl*se activity afttr a tingle dole of 0.3 mga par kg. Nov, vhy am I doing thlal 1 am lntartitad In Induction In tht environment, what levele of PCBe and of other pollutant* will caue* environ mental Induction in tha fish. Nov, you remember X brought up th* point that the fish respond an.ly to the.JjMC-type inducers in Aroeler vhlch vs have heard ar* tha types of chemical* that vt have to worry about anyway. So. tf fiah art responding to these very senaitlvely, thle could give ue tome handle on vhat Is occurring. The point is thst in Lake Michigan, vhlch 293 MONS 026334 Milwaukee happens to b* right next to, body burden* of PCS* lo fl*h ere considerably beyond thl* 0.3 aigf per kg at vhich ve get algnifleant Induction. I also should mention son* other vork which has baen don* at tha Medical College of Wisconsin. Dr. Robert Binder, a postdoctoral fellow who la In Or. Lech's lab has exposed lake trout embryo* to PCB* shortly after hatching, tie find* that when he administers Aroclor 1254 via the water and gets a body burden of 0.3 mgs par kg he has about a 10-fold Induction of aonooxygenase activity In the liver* of the** embryo*. Again w* *** Induction of mooooxygenata activity at a fairly low level of PCB dotage. The other aid* of thl* research problaa, that of demonstrating environ mental Induction, la aomathlng t have started on. For this particular experiment ve maintained sosm carp In our dechlorlnatad Milwaukee City water and also puaped Milwaukee Inner harbor water Into another aet of tanks. We aalntalned tha tea^aratures aa similar aa possible and thay received tha same food. After 37 days of exposure the ethoxyrasorafln-O-deathylats activity In tha carp exposed to harbor water was Increased over l3-fold. At 33 days this substantial Increase was still present. At 60 days some of the fish were transferred from tha harbor water into fresh watsr, and at the next sampling tine at 78 days, the controls stayed about tha same, the harbor water exposed were still Induced over 10-fold, and the fish which war* switched to fresh water dropped to about twice control levels. The reason 1 pointed out that part Is there have been some papers lo tha paat In the U.S. on assaying fish as a means of monitoring for pollution, particularly by people looking for FAH-typ* Induction. In thesa studies however they never had controls or a situation where they could take tha Induced fish back to the lab and show they could get rid of the Induction, In other words that It wasn't a attain difference or something like this. There were two experiments don* In Yugoslavia (Kurelee et al. 1977, 1981), where thay hav* dor.* something similar to this, and obviously we have shovn It with the Lake Michigan Inner harbor water. What I would really Ilka to be able to do Is to follow this up and relate PCB levels In fish tissues to monooxyganas* ansym* activity and actually look for the specific isomers in the fish tissues to relate to this. May we nev have questions from the floor. OR. BROWN: John Brown, General Electric. I would like to ask a question about tha physiological significance to the fish of the nonooxygenaa* induction. 1 recall seeing accounts In the literature of goldfish having been put In tank* with PCB and absorbing up to 88,000 part* per million in their lipid tissue before they finally turned bally up, and this Is many orders of magnitude higher than the lavala that you have used in demonstrating ensym* induction. I also recall at least on* study reported with rainbow trout shoving a lack of toxicity vtth dlbaotofurans. This would suggest that the correlation between tntyme Induction and toxicity that la scan In warmblooded vertebrates may not hold In tha fish. Have you observed the correlation with toxicity? 294 MQNS 026335 DR. MELA.NCON: At the levels ve ere using obviously ve don't find lechsllcy or overt toxicity, and chat is one of the llaltatlont of Just assay ing for s single monooxygensse eetlvlty. It apparently Is very sensitive Indicator end night be of sone value in aonltorlng because of that, but at this point, no, I cannot relate Induction of nonooxygenase activity to a specific defect which la going to result In the death of the fish or a specific toxic synpton. One possibility la coexposure to PCBs and other chenlcals with subsequent activation of carcinogens and so on but I don't have any toxicity data on this sort of thing. DR. VOOICNIK: Mary Jo Vodlcnlk, Medical Collage of Wisconsin. Just to respond to that and to rtturn to reproduction, there are effects showing that PCB Injection In rainbow trout results In a decrease In circulating testosterone. This was not related, hovever, to monooxygenase activity. However, Custafson's group has shown that PCB adalnlstratlon to rainbow trout results In an In vitro stimulation of androstenedlone matabeliam (Hansson at al. I960). So, there Bay be an effect on reproduction, a physiological effect of Induction In fish. DR. STALLING: Dave Stalling, Fisheries Lab., Columbia, MO. I would like to respond to the comment about dlbentofuran and toxicity Insofar as 1 know, and 1 think the study you referred to was probably a feeding study by Zltko. The residue work accompanying that study shoved as much residue In the fish that were not bolstered as In those that ware fed. This study 1 think just points to the fact tnat -.-e have no good definitive toxicity data for aquatic ipeclas for elbanzcfurans; at least If there it, 1 would love to see It. To my knowledge we have no data at all for the furana, and there la probably only :r.< study relating to low levels of TCDD exposure. That vork was done by Heller anJ the concentration rang* chart vat from l to 100 ptrts per trillion. I thir.U In all caret at 100 parte per trillion they died. Growth and survival war* impaired In all treatment ranges, snd hlttopathology was also observed. So, just to at the record straight, I don't think ve have any valid data c: discuss for fursns. DR. MILANCOM: Does anyone with to make any comments on what the appropriate form of sampling thould be to sonitor what la going on in fish, whether it should be whole fish, fillet snd to on, lipid content or whatever? DR. STALLING: Dave Stalling, Fisheries lab. I am noe aura I can axactly answer that question. I think the vork that has been dona on the Craae Lakes shows clearly the need to taka your samples from a defined length or weight, or minimally from a year class and to keep track of sax. Now, In the National Festlelde Monitoring Program chare are 100 stations chat are sampled. Due to the preaa of the matter, in all cases composite samples of five fish each are used to as to raducs variability. It la not possible In that scope, at laaat with the level of funding associated with It, to do length, weight and age correlation and to cat good, valid data, year to year. 1 think it la imperative that at least the ssme age class be ussd shd that you kaap the growth factors t least well enough In hand. 295 HONS 026336 A far a* sampling whole bodltt, I think that la baaed entirely on vhat you want to do with the data. Regulatory agencies concerned with human health obvloualy are not too concerned with total body burden, but the predators that we might ba interested In assessing effects on don't really discriminate too much between edible portion for human consumption. So, I think that there la a dichotomy that will continue to exist, and It la only when funding or effort permits parallel studies between residues In fillets and whole bodies that ve are able to bridge that gap. I think, minimally In the biological sampling, you ahould never do specific studies on dynamics unless you have six samples. That Is Just from our laboratory experience. Three samples are not enough, and I think from a good, statistical design you vould probebly like to have considerably more. I believe the literature addresses that with regard to what lavel of residues you wish to discriminate against, and probably to get fsetors of 30 percent differences In residues you need 30 samples of comparable age, length and weight. DR. HELANCON: I would like to mention a study done by Dr. Richard Peterson at the University of Wisconsin, Madison In which he compared the rates of elimination of 2,5,2',5'-tetrachloroblphenyl from rainbow trout and from perch (Culney and Peterson 1980). Ve have previously done some work together on rainbow trout, and he found the lipid contents of both species were similar. He used hatchery flah, not fish caught out In the lake. The T ,,, S of elimination overall were similar, but the actual tissues vlthln the oodles of the fish which contained certain percentages of lipid vere different in the two species. So, If you look Just at fillets, the trout looked higher than the perch, but If you looked at the whole body, the differences tended to disappear. One comment I would make In regard to Mr. Hunter's presentation Is that he looked at several trophic levels with crapples and white bast as carnivores, but had no data on the small fish vhleh they could have been eating. Did you by any chance look at any of those? MX. HVNTER: In a reservoir white bass are fairly prey specific on glxsard shad. So, I think that In the site range we were looking at, I think that those two are directly comparable. Tou are correct on the white crapple. I did not look at the minnows. Funding just did not allow me to go Into that. DR. MELAXCON: Even In the glsxard shad, the ones you sampled were so big that those eould not be the ones they were feeding on. I vas wondering what the small onas would have been like? Vould you expect similar levels? MR. HUNTERt 1 think so, based on the study 1 mentioned where 80 percent of the burden obtained from the water occurred In the first 12 hours In fathead mlnnovs, 1 would assume that most of the glszard shad's burden vould come from this source and that It probably vould not Increase further with age. So, I vould suspect chat the very young fish would have similar levels to the older fish. 296 MQNS 026337 DR. MELANCON: In an experiment with PCB* whr 80 percent of final level vae reached In 12 hours, 1 would wonder whether it waa a atatlc syatam and the water level of PCBa wan juat baeoaing depleted. MB. HUNTER: That la a good point. DR. MELANCON: The other coaaient I want to Bake la that In aoaeplace like Lake Michigan where the carnlvorea are really the top of the food chain and taalltr flah like alewlvee eat eaaller organlaas which In turn have been down In the ttdlatnta, the pollutant vhlch vaa In tha aedlaent ultimately flnda lta way into the big carnlvorea In tha lake. Knowing the type of apeclea In a lake and tha Interrelatlonehlpa la probably aoaethlng that ehould be taken into account vhen one decldee what apeclea to aaaple and whae your ultimata bloaccuaulatlon factor eould-be-and-what-lt-coold-ba due to. DR. BROWN: John Brown, Ceneral Electric, again. 1 would Ilka to coaaent rather poaltlvely on Dr. Hunter'a propoaal to uaa the level* in flab a* an Indicator of environmental contamination and wonder If thla ehould not be puahed even a little bit further than ha did. A typical problea encountered In monitoring level* of xenoblotlc* In the general envlrottaent la the extraor dinary heterogeneity of the dlatrlbutlon of the agent in different coapooenta, particularly In different aedlaent* that aay have different dlatrlbutlon coefficient*. It aeeaa to ae It la poaalble for biota that are circulating in tha envlronaent to Integrate the actual level* over a fair rang*. Tou have Indicated relatively conalatent data for your biota. Tou have propoaed ualng thla a* an Indicator atandard. Thla 1* preauaably an Indicator of environmental eontaalnatlon. but we are only concerned with environmental contamination In the** caaaa becauae of lta effect on the biota. Perhapa, do you think that thla should b a primary atandard, rather than a aecondary one? MR. HUNTER: I think that there la going to have to be a lot of work don* to bridge the gap betvaen Input to the ayatea and level* In tha more PCB-aenal:lve apeclea. It can be done. Alto, It can be don* for toxic aubatance* other than PCBa. I have aeen the** relationship! in araenlc burdena In flah. They art very different. What you tee la very different. It la not In the bottea ftedera, but yes. I do, Indeed, think that thla doe* have aoae poaalbllitle*. DR. MELANCON: In a way flah have already been used a* monitor*. In fact, we know hov bird* that at* flah and *o on were on* of the flrat Indicators for aoae of thmae pollutants. We had a cat* In Wisconsin juat a couple of year* ago where routine DNR flah eaapllng procedure* led to finding aoae highly PCB-contaalnated carp In the Sheboygan River (Klelnert 1978). It vaa found that on the bank of the river waa residue left from a printing firm that had aoved out year* previously, and every year vhen the water vent up In the spring It leached PCB* out of there, deposited them In the lover part of .the river, and any flah that happenad to b* that* at that tlaa acquired large body burden* of PCB*. So, In othar word* In that case fish eerved a* monitor* to find a tourc* of tha PCB pollutant*. 297 MONS 026338 MR. HUNTER: 1 thick the cast at Fort Clbeon laka la another good axaaple of thia. The PCB contamination washed In from landflova and contaminated a creak channel, and at moat the creek channel held about AS hectares of sediment surface area. That 1* about 1/2 of 1 percent of the surface area of that lake, and we never did detect any PCB at the 0.1 mlcrograms per liter detection Halt In the water. So, ve could not find It in the water at our detection Halt, and going around looking for that 1/2 of 1 percent of the surface area of sediments Is kind of like a needle In a haystack and even within the contami nated creek channel the concentrations ranged from 0.23 to 7.2 parts per million on a dry weight for the sediment, and It was significantly correlated with areas of deposition, In other words, the area* at the upstream edge of mudbars and at the delta held more PCB contamination than, say, the stream bottom or the stream edge. So, again, there la so auch variability Inherent In sediment sampling that we almost hava to use the biota as an integrator. SUMMARY: It appears that there are advantages In using fish as biological monitors for pollutants such as PCBs. This could be done by using fish to accumulate materials that are vaterborne or present In the aquatic sediments with subsequent analysis of fish tissues for selected chemicals or possibly by examining hepatic monooxygenasa activity. The choice of apecles should reflect the monitoring gosls such as the types of pollutant* to be monitored. Addi tional studies would be needed co facilitate selection of appropriate fish species for each monitoring goal and additionally might help to provide informa tion on route, duration or magnitude of exposure rather chan Juec Indicating that exposure to a chemical of Interest (pollutant) has occurred. MGNS 026339 298 REFERENCES Bill* TP, Marking LL, Muck WL. 1981. Polychlorinated biphenyl (Aroelor 1254) residues in rainbow trout: effect* on *en*ltlvlty to nine fishery ehemlcsls. N, Aaer. J. Flh. M*n*g. 1:200. Gulney PD, Petarion RE, 1980. Olttrlbutlon and ellalnatlon of a polychlorinated biphenyl after acute dletery exposure In yellow perch and rainbow trout. Arch. Environ. Contam. Toxicol. 9:667. H*n**on T, Rafter J., Custafston J-A. 1980. Effect* of some common Inducer* on the hepatic microsomal metabolism of androetenedlon* In rainbow trout vlth pedal reference to cytochrome P-450-dependent entyae*. Bloehem. Pharaacol. 29:587. Klelnert FJ. 1978. Final report on the Investigation of FCBa In the Sheboygan river tyatea. Wisconsin DNR. Eurelee B, Brltvic S. Rljavec M. Muller VEG, Zahn RR. 1977. Benco(a)pyr*na nonooxygenese Induction In marine fleh-aolecular respona* to oil pollution. Mar. Biol. 44:211. Ruralec 8, Protlc M. Brltvic S, Retie H, Rljavec M, Zahn RR. 1981. Toxic effect* in fish and the mutagenic capacity of water from the Sava river of Tugoslavle. Bull. Environ. Contem. Toxicol. 26:179. Pavlov SP, B*xt*r RN. 1971. Physical and ehaalcal agents of the distribution of polychlorinated biphenyls In the Aquatic Environment. Aquatic Toxicology ASTM STP 667, Marking LL, Rlaerle RA, sd. Philadelphia: American Society for Testing and Materials, p. 195. 299 MONS 026340 CHAPTER 8 RISE ASSESSMENT Thl* paper begin* by carefully defining the tarn* toxicant, axpoaura, and hazard, and relating than mathematically to the tara rlak aaacaaoent. A chert overview of PCB* aa toxlcanta, (i.a., compound* which cauac acute and chronic toxicity, mutagenicity, fetotoxlclty, carcinogenicity, and teratogenlelty) follova. The aacend part of the paper deala with analyze* of the acute and chronic human epidemiology data currently available, and the procedurea typically uaed for aaeeaalng rlak. The rlak* aatoclated with PCS* and an eatlmaclon of the health rlaka aaaoelatad with the currant environmental level* ar* dlacuaaed. ASSESSMENT OP THE HUMAN RISKS TO PCB* ASSOCIATED WITH THE EXPECTED ENVIRONMENTAL EXPOSURES Robert C. Jam**, Ph.D, Ecology and Environment, Inc. Buffalo, NY 14225 Raymond D. Harblaon, Ph.D. Department of lnterdlaelpllnary Toxicology Unlveralty of Arkanaaa for Medical Science* Little Roek, AR 72205 HONS 026341 300 During the pate decades, a combination of "open" systems and sloppy or Illegal disposal practices have resulted In the release of large quantities of chanlcals Into the environment. In the late 1960s, vhen It became apparent that perslatent chlorinated hydrocarbon polychlorinated biphenyls (PCBs) vers ubiquitous environmental contaeir.anta which took years to biodegrade, attention was focused on their potential to blcaecumulate to vhet was believed to be toxic levels In animals and humans. As a result, questions srose regarding the chronic side effects caused by long-term exposure to these chemicals. Reflecting public concern for chemical contamination In general, Congress snsctsd tht Toxic Substances Control Act In 1976, vhlch mandstsd a phaseout of PCB manufacturing and usa. At Chat time, PCBs were considered a highly toxic and dangerous chemlcsl bscaust of their environmental psrslstsnce and their potential for causing permanent adveree health affacta, such as canetr. In recent years, however, a substantial lncrsass In scltntlflc data has reaultsd In s better understanding of both tht actual affacta of human axpoaurs to PCBs snd ths ehtmleilly-lnduesd processss Isadlng to cancer. While it was approprlata a fsv years ago to act on the prsmlsa that all potentially carcino genic chsmicsls represented similar hazards, rsesnt axperisnes hss shown thst there are important thaoraelcal and practical dlfferancaa In tha actual hazard each "carcinogen" represent* to public health. Similarly, there has been s substsntlsl lncrsass In tht smount of dsta reflecting the toxicity of PCBs In animals and humans. Considering the national concern expressed for environ mental contaminants, snd In particular for PCBs, It seams both appropriate and nccssssry to reevaluate the hazard and risks posed by PCBs now that we have more Informs den, ar.d ca continue this re-evsiuatlon aach time significant changes In our Information occurs. It Is che purpose of this paper to provide a framavork for Identifying tht Important toxicological considerations used In interpreting the risks inferred by toxicity dsta, at veil as to provide a discussion on the PCB dsta vltnin this trsn.ewcric to tt to arrive at an asstssaent of the risk to human health. DETlNltlCXS ASP BASIC PRINCIPLES OF ASSESSING RISKS A toxicant It a chemical agent that can product an adverse effect In a biological system. Such an adverts affect may be an alteration of normal function or destruction of Ilfs. This definition is broad since all chemicals art toxic at sotx loss, i.t., all chemicals are capable of altering soma function or producing death in soma biological organism. While this statement may team obvious, le serves to tmphatlze tha basis of risk aataaaaant vhlch la a determination of those circumstances and conditions under vhlch an advarat affect can be produced. At Call Mrak stated years ago,."there art no hsrmlass substenets; there art only harmlait vayt of using substances!" A chemical la toxic and products harm only vlthln prescribed conditions of usage. Risk la defined at tha probability that a substance vlll produce harm under specified conditions (Doull et el. 1980); thee la, It la a practical .considerscion to datarmin* vh<-her or not soma harm vlll be elicited from a specific chemlce. exposure. Safety Is tht 'reciprocal of'risk, or tht probability that a substance vlll not produce harm under the specified 301 MONS 026342 condition!. Thue, when determining the risk or safety of e cheated, the erltlcel lector 1* not necessarily the lntrlnelc toxicity of the chealeel per e, but the likelihood that the level of exposure to the chealcal is sufficient to express Its Intrinsic toxicity. In general teres, then, the risk Is approxlaated by the equation: RTxE where R Is risk, T Is toxicity, and E Is exposure. More accurately, the equation for risk la equal to the toxicity as a function (f) of the exposure, or: . R Tf(E) This better defines the extrapolation, since under certain conditions the risk Is not always linear over the entire dose-response curve; and since, depending on how one defines function (f), threshold or nonthreshold dose-response curves aay ba fitted to this equation. However, regardless of how one chooses to express the risks to ehealcal exposure. It la clear that: (a) the actual risk it dependent upon both the toxlelty (l.e., hazard) and the exposure (l.e., ssiount of ehealcal) end fb) to change either alters the risk. Therefore, to asks a risk assessment, one aust coapare the exposure probability with the various dose-response curves for each toxicity to determine whether or not a harmful response is likely to be Induced. In evaluating the risk associated with exposure to PCBs, we will atteapt to answer two questions. First, is the level of exposure of a sufficient magnitude such that a hazard exlstsT Second, If so, what Is the hazard or undesirable response expected for that level of exposure. To answer both questions It is neeetsary to first evaluate the hazards associated with PCBs; l.e., what are the toxic responses and at what levels of exposure will they occur? Many aspects aust be considered and weighed before the hazards to humans associated with PCB exposure can be evaluated. An evaluation of the human hazard and subsequent risk estimation for any chemical muat address the data base utilizing a thorough comparative toxicologic assessment. This means that the animal and human data available must be carefully considered In terms of: (e) the breadth and variety of the toxic responses manifested; (b) the degree of species variation or species consistency In the effects monitored; (c) the possible and/or proposed mechanisms of toxlelty; (d) the validity of the tests performed and their relevance for extrapolations to man; (e) the dosage used In animal tests versus the expected level of human exposures; end, finally, to the extent that the date are available, (f) the outcomes of serious poisonings and long-term oceupstlonal exposures as a guide to the expected human conse quences and as a teat of the extrapolations made from animal data. Only vhen a consistent pattern of toxlelty In animals coincides with human experiences can accurate end tale guidelines be promulgated. The following sections of this chapter will discus* the hazard eseoclated with PCB exposure *s Interpreted from the animal and fcunan data; tha choice end use of rltk assessment models to define the risk to PCBs from various environmental exposures; the accuracy with which the health risk* for e PCB 302 MQNS 026343 exposure can be estimated at the present tine; and the conclusions reached bp tM.a study. EVALUATING THE ACUTE HAZARD ASSOCIATED WITH PCB EXPOSURE FROM ANIMAL DATA Several reviews of the mammalian toxicity of PCBs have been published within the last decade (Flshbeln 1974, Kimbrough 1974, Feakall 1975, NIOSH 1977, IARC 1978, EPA 1980), Since much of this Information has already been given In the previous chapters, a cursory summary rsther than an exhaustive review la provided here. This summary presents the types of hazards to be expected end document* the extrapolation* on* might make to the high exposure human situation bated on the animal data. It must first be stated that PCS* have a lov order of acute toxicity a* meeeured by the median lethal doses (to 501 of ths population) (LD,Qs) for rodent species. As shown In Table 1, the median lethal dose* for various coaaaerelal PCB mixtures lie between 1.3 sod 19.2 g/V.g of body velght. Median lethal doses ss high as these classify PCB mixtures, especially the higher chlorinated mixtures, as only slightly toxic chenlcsls (Poult st al. 1960). PCB* produce a variety of physiologic changes In animal*. For example, in the liver. PCBs Increase the amount of endoplasmic reticulum, resulting In enlarged hepetocytes and thereby lncreaelng liver else end velght. Higher doses can cause liver damage end necroela. PCBs applied directly to the akin of rabblte have produced hyperkeratosis, erethyme, blisters, and desquamation. In prlmatat, orsl doss* of PCBs have lad to fecial edema, heir lost, and sene-llk* pustules, as well ss prolonged menstrual cycle* and Increased bleeding In female animal*. Additional symptoms reported in varlou* species Include gastric hyperplasia, decreases In red and white blood cells, atrophy of the thymus and spleen, and Increases In eerum levels of phospholipids, triglycerides, and cholesterol. However, as with lethality, th* doses required to cause these acute effects are generally high. The doses required to cauce subchronic effects can also be considered high when th* slow rat* of metabolism, and therefor* e tendency tc blciceusulete, Is taken into consideration. While a number of offsets have been reported for PCBe, those of the liver (l.e., enlargement end necrosis) sre probably the most consistently reported, as are those of th* akin chleracn* or dermatitis). These physiologic responses observed in the liver and sklr. are fairly predictable effects of msny chlorlnsted compounds. In sstesslng ths scuts and subchronic casts of animals, PCBs do not appear te be remarkably toxic chemleelo. The dermal lrrleatlon caused by PCBs Is not an unusual physiologic response for an organic, solvent-like chemical. The ehloracne observed In animals and humans ie also a frequently observed efface of other chlorinated chemicals and lo reversible In cases where PCBs sra the only suspect agent. Ths observed liver effects (1.*.. increased liver velght, proliferation of ondoplasmie reticulum, deposition of fat, cell death, and titan* necrosis) ar* eosmon responses to persistent, chlorinated hydrocar bons, which in general are enzyme Inducing agents sc lower dotes and hspttotoxlns at higher doses. Many other chemicals share one or both of thee* characteristics. While snzvme induction may inert*** th* toxicity of tom* chemicals snd lower th* dot* producing toxicity, this is not * general phenom enon. In msny instances, an increase In metabolism decreases the toxicity 303 MQNS 026344 Table 1. Median Lethal Doeee for Various Conaerelal PCB Mixtures Speelee Mouse (oral) Ret (oral) Rabbite (akin) Coanarclal Brand of PCBs Aroclor 1254 Renaehlor Aroclor 1221 Aroclor 1232 Aroclor 1242 Aroclor 1248 Aroclor 1254 Aroclor 1260 Aroclor 1262 Aroclor 1268 Areolor 4665 Arcelor 5460 Aroclor 1221 Aroclor 1232 Aroclor 1242 Aroclor 1248 Aroclor 1260 Aroclor 1262 Aroclor 1268 10 50 (ng/Vg body weight) 2,000 >1,875 4,000 4,500 8,700 11,000 4,000-10,000 10,000 11,300 10,900 16,000 19,200 3,200 2,000 1,300 1,300 1,300-2,000 1,300-2,000 2,500 304 MQNS 026345 of ocher ehsalcals. Similarly, inducing agents often lncreit* ocher route* of blotraoaformstlon that detoxify toxic chemical*, In fact, an true induction la lomatlmea used therapeutically. Therefore, Induction by ltialf cannot be considered a toxic response, nor can Its consequences be accurately predicted In human*. In monkeys, PCBs cause a conversion of the gastric epithelium that can be described as a dysplastlc grovth pattern. This doe* not constitute a neoplastic transformation (Drill et al. 1962). While FCBa have been reported aa inhibiting the immine system In mammals, this usually occurs at high doacs at vhleh general toxicity also leads to s reduced food intake and thus a reduced nutritional status. In some test species, PCBs provide changes in reproductive performance, but again usually at high or toxic dotes. For example, In monkey* and minks, this problem oceurs In the female animals at dosas clearly producing general toxicity and eigne and symptom* of intoxica tion. Reproductive dysfunction during chamlcsl Intoxication Is not a partic ularly unusual finding. Therefore, It Is concluded that the acute toxlelty of PCS* In animals Is not particularly remarkable, nor doe* the chemical represent an unusual hazard. chromic rmers or pcb* Mutagenicity A genotoxlc chemical Is any compound capable of Inducing * permanent, inheritable change In the genetic composition of a cell. Genotoxlc effects ere divided Into three type*, depending on the cellular level of the genetic lesion. Mutagenesis refer* to mutations or structural changes at the DMA level, called mlcroleslons (Doull at al. I960). Por eh* sake of discussion, ell tests are referred to in this paper a* mutagenicity test*. The other tvo types of gonctoxlc effect* art: clastogenssls, which refer* to structural chsngss at the chromosomal ltvtl, and aneuploldlsatlon, vhleh rsftrs to e change to chromosomal number. Both elascogenesla end aneuploldleetlon can alter the amount or expression of genatle metarial; both types of changes sre often refarrad to aa macrolealona (Doull at al. 1960). The "Arnes test," which use* mutant strains of the bsetsrl* Salmons 11s tvphlmurium thst heve lost the ability to synthesise their own histidine, Is probably the most popular and beat known test for mutagenic activity. Sines this amino add 1* Important to protein synthesis, and therefore grovth, these cells can grow only if histidine la added to the culture medium. A chemical mutagen een be Identified by e mutation in the locus of the hlstldloa defset vnlch reveres the bsetsrl* to the normal etete, thus allowing easily identi fiable bacterial eolonies to grow In a histidlne-deflelent medium. tfyndhea and eo-vorkars (1976), using the Ames essay, tsstad the several chlorinated biphenyl mixtures and reported that they vere weekly mutagenic. A review of their data Indicated that A-ehloroblphanyl had significant activity. Tat, Aroelor 1221, which has a chlorine eontent of 1.15 ehlorlne atoms ptr nolsevl* and there.ox. largely t mibnoehloroblphsnyl mixture, shoved a greatly reduced activity. This would teem to indicate that either 305 MQNS 026346 tha "heavier" PCB aolacules vtre inhibiting mutagenicity or chat the activity la graattat for the 6-ehloro-isoaer of tha eonochloroblphcny1 compounds, The teat reaulta for the polychlorinated biphenyl compounds 2,2',5,5'-t*tr*ehloroblphenyl and Aroelor 1268 aeea to indicate no aubatantlel autagenlc activity. Subsequent atteapta to raproduce the Wyndhaa at al. raaulta have failed (McMahon et al. 1979, Safa 1980, Raddle and Bruce 1977, Schoeny at al. 1979, Levlnska* 1981) (ate Table 2). The aajorlty of aaaaya reported dlacount any auggeatlon that PCBa are mutagenic In tha lata aaaay Including subsequent work by an author of the original obaervatlon. Several atudlea have deaonatrated that PCBa lack the ability to daaaga or alter chroaoeoaee. Green et al. (1975) looked for elgnlflcant ehroaoaoaal daaaga In the bone aarrov and epera cella of rate adalnittered the follovlng doaage reglaens: a alngla doae of Aroelor 1262 at 1250, 2500, and 5000 ag/kg; or 500 ag/kg per day for five daya. Even though aoae of the anlaala demonttrated definite eigne of Intoxication, there vaa no evidence of ehroaoaoaal abnoraalltlea. Theae finding* are conalatent vltb the reeulte reported by Dlkahlth et al. (1975) or Carthoff et al. (1981) In rat* and th* atudy reported by Hooplngamer et al. (1972) ualng cultured huaan lymphocyte*. Heddle and Bruce (1977) reported that Aroelor 1256 vaa negative In th* alcronuclaua teat, which aaaeure* claatogenle activity (ehroaoaoaal breakage) by aeaaurlng chromatin bodice In red blood cella. Creen at al. (1975), ualng the dominant lethal teat, which aeaaurea mutagenic event* In germ cell* by th* eabryotoxlelty It Induce*, reported that Aroelor 1262 and Aroelor 1256 gave negatlva reaulta In rata. Kapllnger et al. (1971) and Calandra (1975) alao reported a alallar lack of autaganlc activity In mica using tha dominant lethal test. levlnskaa (1981) hae racently reviewed eh* mutagenicity data for PCB*. Besides th* above-aentloned tests, Levlnskas discuss** many additional, though In several case* less often used, procedure*. In hi* review he refer* to a study by Odashlaa In whleh Kanechlor 300 and 500 were negative In th* standard strains of Salmonella typhlaurlua used In th* Aaea assay, as well a* several additional bacterial strains. However, Kaneehlor 300 was reported positive In one or nor* of four additional strain*, while Kanechlor 500 was listed a* negative. On th* basl* of aedlaantatlon rata*. Stadnlckl et al. (1981) have reported that the epoxide of tatrachloroblpbenyl caused ilngle-stranded ehroaoaoaal breaks In th* DHA of 1-929 cells sc eoneantratlons ranging froa 1 ug/al to 100 ug/al. A mixture of two hydroxy metabolites, and to a leaser extent tetraehloroblphenyl, caueed aoae daaaga at 20 ug/ml and significant damage at 100 ug/al. The significance of this test at th**a high In vitro concentration* 1* questionable. Using Prosonholla aelanotaster and Boabyx aorl a* the eukaryote test species, neither tha French PCB alxtures, ciophenJo and Clophen 50, or th* Japanese alxtures, Kanechlor 300 and 500, were autaganlc. Odashlaa (1960, 1961) studied ehroaoaoaal aberrations In Toshlda sarcoma cells at PCB con centrations camednf-SOX-growth inhibition: Kanechlor 300 gave'a positive response and Kanechlor 500 a negative on*. Odashlaa reported cytogenetic analysis of bon* aarrow cells In alee exposed to PCB* at a near lethal dose. This test produced the opposite result of the previous study. Odashlaa listed exposure to Kanechlor 500 as positive and exposure to Kanechlor 300 as negative (EPA 1960, Levlnska* 1981), 306 HONS 026347 Table 2, TCBas Tnlmlatlun of Microbial Mutagenicity Teata toe Teeter Mraln S, t ypliiaur luai TroSuct 007* 0)0)2 CD TASS TAIOO TA10000 TAI))J TAD)* TAD)7 TAI))S I. coll Wf 2 UP2uv Aroclor I2AS Aroclor K "" KinicHtor soo j.r.j-s*let taclilotol>i plieny) Kenechlor MO Aroclor 1221 4-cliloroblylianyl -- -- --- -- Ne|. 1**1 He(. Net. *|. -- -- -- Net. Nag. * Net. Net. ~ He*. -- -- Net. -Net. -- --- -Nag. N*|. Neg. So*. / Net. -- N*|. --Neg. Neg. Net. " -- -- " Nag. N*t. -- -- -- ---- -- -- "" Nag. -- Nag. -- Nag. -- N*|. Neg. " " Pot. *~ foa./ N*|. Source) tevinekee DS1. MONS 02634S Norback et al. (19B1) reported that Aroclor 1254 traneformed C3H10T11/2 cells to Type III foci after six veeks of continuous exposure to 10 ug/L of Aroclor 1254, while a concentration of 1 ug/mL vss negative. Norback and co-vorkers suggested that the affects of PCBs In culture Include proaotlon. In contrast to this study, Plenta (1981) reported Aroclor 1254 es negative In the Syrian hanater cell transformation test. Peakall (1981) reported chromosomal aberrations In doves fed a 10-ppm diet of PCBe. The frequency changed fro* 0.81 In controls to 1.8X In treated birds. The average rate of ehroaososul aberrations among treated eggs was higher than the highest control value In only t of 17 pretreated eabryos. In contrast, PCBs Injected into white Leghorn eggs up to concentrations of 20 ppa, while toxic, were not clastogenlc. A thorough review of the autagenlclty tests reveals that PCBe gave primarily negative responses. While some positive results were reported, these were largely In vitro tests and for the Kenechler brands, vhlch gave Inconsistent and mixed results for the 300 and 500 brand mixtures. The relevance of some of the In vitro teats that found positive acoree only at concentrations of 10 ug/ml or greater (vary high and unlikely in vivo blood concentrations) la questionable. This Interpretation Is aupporteTTy the fact that In many cases a similar experiment gave correspondingly opposite, l.a., negative, results. According to Levinskas, "considering the large number of teats conducted, It la not surprising that an occasional auspicious positive finding resulted, aimply on a statistical basis" (1981). A comparison of the number of Ames' tests unable to reproduce the results of wyndham at al. (1976) supports this contention. It la the consensus of this ravlaw that PCBe do not represent e signifi cant mutagenic risk to humans; a contention borne out or reinforced by the negative animal carcinogenicity studies for PCBt. Teratogenicity A teratogen le an agent that products congenital defect*. Although ttratogtnlc responses are generally Identified by daflnlt* anatomical defect* in eh* offapring, authorities also regard functional or biochemical changes at evidence of teratogenicity. An Important consideration In evaluating teratogenicity data, sometime* referred to ta "Ksrnofsky's rule," 1* that "any compound administered at the proper dleage, at the proper stage of'development, to embryo* of the proper epacles, will be affective In causing disturbances In embryonic development." What Karnofsky's rule calls attention to Is the fact that maternally toxic dotas capable of killing th* fatu* or producing teratogenic affect* In tha offapring sr* of much 1*** concern than those chemical* causing the same effaces at doses far below a maternally toxic level. If the pregnant female le made sick, than the delicate balance batvecn mother and fetus is likewise kffseted oi disrupted and an adverse fatal efface sight ba.expected. Depending on the (tag* of th* fetus, thl* disruptive pressure can be manifested In many way*. Tor example, disease, malnutrition, and stress hsv* 11 been shown to produce changes In the Dumber and quality of offspring 308 HONS 026349 produced. Even reversible acute effects on the Bother's central nervou* systea, though of short duration, say disrupt feeding enough to cause undesir able reaponaes In the unborn. Therefore, fetotoxlclty Is of Boat concern only when It occurs at doses far belov those eliciting toxicity In the Bother. The studies concerned with the teratogenicity of PCBs In mica are contradictory. Toerutk (1973) reported that PCBs are not teratogens In sice at dosages up to 500 ag/kg. However, note recent studies have revealed chemically-induced effects In the fetus or In newborns at lower doses. Vatanabe and Sugahsra (1981) Injected alee with Kanechlor 500 at dosages approxlaatlng 36 to 182 ag/kg/day on days 6 through IS of gestation. At the higher doaes, the pregnancy rate declined to one third of the control values, the nunber of resorbel fetuses doubled, and the rate of fetal cleft palates reached S.8Z. The authors stated that since there was no change In maternal or fetal velght gain, the FCBs used had a specific teratogenic effect not due to general aatsmal toxicity. Hovever, their own reaults do not support this contention. At the higher dosages, the asternal mortality was higher than the rate of cleft palates aaong the newborn (12X versus it). Also, the authors noted skin lesion#, alopecia, and enlarged livers In the daas. Indicating significant chealcal Intoxication. Marks et al. (1961) reported teratogenic affects for the specific chlorinated biphenyl compound 3,3',4,4',5,5'-haxaehlorobiphenyl. At dosages of ! to lo ag/kg/day given on days 6 through 15 of geatatlon there were significant Increases In the percentage of resorptions and fetal aalformatloos. Liver damage occurred In the dans and consisted of single-cell eentrolobulsr r.ecrosls and aicroabseeasea. The fatal livers had even a more pronounced injury, Including massive necrosis and the formation of a tubule-llke pattern of hepatocytes. Tilton ct al. (1979) reported that 32 ag/kg/day of the 3,3',4,4'-tatrachloroblphenyl, when ;lveo on days 10 through 16 of gestation, Induced signs of neurotoxicity in :he offspring. Some of the nevborn were labeled "spinners" because of thilr Increased motor activity and episodes of head bobbing and rotational roveaents. The spinners had decreased forellab grip strength and all troated anlaalt had difficulty In crossing vlra rods. Subsaguant studies (Agrswal 1961) rtvsslsd that thia axpotura laval of tatraehloroblphanyl significantly altered the dopaalne binding sites end the dopamine levels in the corpus strlstua of nevborn alee. Contrary to Marks et al. (1981), Kettason at al. (1981) reportad that doses of approilaataly 50 ag/kg/day of 2,2*,4,4',5,5'-hex*chlorobiphenyl did not increase the resorption frtgueney, and no fetal effects were noted other then enlarged livers. Likewise, Geller^ end Wilson (1979) reported that no reproductive effecti were eeeo In the offspring of alee exposed to Aroclors 1221, 1262, and 1260 at dosages of 30 ag/kg/day on days 14 through 20 of gestation. Rets have also been extensively studied using dosages up to 100 ag/kg per day. Although these etudles indicate that fetotoxlclty increases with dost, no tsratogtnlclty has been reported (Kepllnger ct al. 1971, Celandra 1976, Linder et al. 1974', Ttlleheuva et al. 1971, Ktpllngar 1972). 309 MQNS 026350 Malformations (Until have been reported in doge and plge In an abstract by Earl et al. (1974). Hovever, the dosages of PCB* used resulted in a severely reduced food consumption in both the dogs and pigs and the authors reported that there may have been an underlying genetic defect that contributed to the teratogenic response In dogs (Drill et al. 1982). Several studies have been published using monkeys as test species (Allan et al. 1974, Allen and Barsottl 1976, Barsottl et al. 1976, Allan at al. 1980). For six females fed 25 ppm of PCBs for tvo months, eoncaptlon vas Inhibited (Allen et al. 1974). Of the six, only one vas able to give birth; all female animals were clearly Intoxicated and one died. A second study (Allen and Barsottl 1974) tested 18 female animals fed diets of 2.5 or 5.0 ppm of Aroclor 1248. After seven months' exposure, the sight survivors from each doaage Xavel^vaxa brad.- All animals on ths lov-dose diet conceived. Thera vere three resorptions end five births. Of those animals given the higher dose diet, one resorption, three abortions, one difficult delivery and subsequent death, and one normal birth occurred. All newborns vere small In size end shoved hyperpigmentation, but vere otherwise normal (Allen and Barsottl 1976). In a number of species (e.g., mice, rabbits, rats, and monkeys), FCBs shoved no or contradictory teratogenic activity. The study conducted by Earl at al. (1974), still unpublished, provides unconvincing data since It Is possible that the-malformations observed In the offspring vere the result of severe maternal malnutrition (Drill et al. 1982), This conclusion vaa substantiated by Hansen at al. (1975), vho failed to find teratogenle effects in pigs when the mothers vere fed a diet of 20 ppm of Aroclor 1242 during gestation and nursing. PCBs have produced contradictory results In mice. Tests conducted by Marks et al. (1981) and Tlleon and co-vorkers (Tllsou et al. 1979, Agrsval 1981) Indicate that the recorded effects are seen only vlth ayimetrlcslly ehlorlnaced biphenyl compounds and that ehlorlnatlon in the three and four positions Is critical. This may explain vhy Toeruek (1973) and others (Mattson et al. 1981, Cellart and Wilson 1979) sev no or lesser effects vlth higher doses of mixtures containing small amounts of tha specific Isomers or with other hexachloroblphenyl Isomers at higher Coses. Hovever, studies by Blocca et al. (1981), like the negative data, greatly lessen the Impact of ths teratogenic findings in mice. These authors demonstrated that the 28-day U>.q for 3,3',4,4',5,5'-hexaehlorobiphenyl Is about 19 mg/kg/day (ranging baevaatr 11 through 33 mg/kg/day). Moreover, extending the observation Interval to 50 days, It vas found that the dosage levels 10 times lover caused 100S mortality. Disposition studies revealed that the toxicity of tho Isomers vas relsted to tissue accumulation and that for tho 28-day Interval the adipose tissue levels at the U>t0 dosage vere approxlutely 2,000 ppm. These data reveal that the teratogenle effect seen by Marks et al. (1981) occur at doses greater than those that allov for chronic survival. This Is conslatant vlth tha findings af Watantbe and Bugghara (19811 for vhlch tha taratogenlc effect* of PCBs in mica appear at dose* of significant mattrnal toxicity and lathality. In summary, then, the anlmel data do not Indicate that FCBs represent a signifi cant teratogenic risk at doeee not causing significant mattrnal toxicity. Furthermore, the date suggests that preventing subchronlc lechalley or chronic organ toxicity would prevent teratogenle effect*. Thue, tha risk to humans 1* 310 MONS 026351 xpseced-te be minimal, considering the much lover dote* that huotn* ere generally exposed to in the cnvlronaent end the epperent lack of toxicity expected at these dotes. Studies In animal* similarly demonstrate that PCBs represent a minimal reproductive and fetotoxlc hatard (for revlev ate Drill at al. 1982). While effects on reproduction or fetotoxlelty have been reported, the doses causing such affects are generally high. Rats arc fairly resistant, while minks seem to be particularly sensitive. Aulstich and Rlngar (1977) reported that minks fed 2 ppm Aroclor 1254 In their diets had decreased reproductive success. The effect at this level vas specific for Aroclor 1254 and was hot seen for Aroclors 1016, 1221 or 1262. The reproductive Impairment vas not permanent but consistently occurred at maternally lethal doses. Orbarg (1978) has demonstrated an Increased metabolism of progesterone, estradiol, and testos terone In animals precreated vlth PCBs. It may be proposed, then, that doses that greatly Increase the metabolism of Important hormones could alter the normal hormone homeoetesls and result In reproductive problems. The series of studies by Allen and co-workers (Allen et al. 1976, Allen and Barsottl 1976, Barsottl et al. 1976, Allen et al. 1980) have demonstrated that PCBs may disrupt menstrual cycles and significantly alter reproductive success in monkeys. However, the doses used vere clearly maternally toxic, fetotoxlc, and toxic to the offspring. Studies similar to chose In mlee Indicate that the body burdens In monkeys at the time of daath are quite high (Belley et al 1980). Since (a) high, maternally toxic or lethal doses vere generally required, (b) there are a number of negative studies, (c) the general human environmental exposure Is very lev, and (d) no reproductive problems in humans have been associated with exposures to PCBs only, PCBs do not appear to repreaent a algnlflcant reproductive rlek to humane at the lev expoeure levels generally occurring via tha environment (Drill et al. 1982). Carcinogenicity The carcinogenicity of PCBa hee recently been reviewed by aeveral authora (IARC 1978, CPA 1980, Levlntke* 198), Jamci et al. 1981, Drill et al. 1982). Tha following paragraphs provide a summary of tha data. Some of tha studlee reporting the tuaorlgtnielty of PCBa are llecad in Table 3. While many investigations have been conducted, only two studies hsvs reported s signifi cant lncreaie in hepatocellular carcinoma. Ito et al. (197A) reported hepato cellular carcinoma in mlcs after 12 months' sxposurs to a disc of 300 ppm of Kaotchlor 500. Kimbrough and eo-vorkert (1975) also reported hepatocellular carcinoma In fsmals rats fsd 100 ppm of Aroclor 1260 for 21 months. Hovavsr, contrary to thssa findings, many similar erudite have shown no ability of PCBs to product cancar. Kimbrough and lindtr (1976) found that 300 ppm of Aroclor 1256 failed to product htpacocallular carcinoma* In mlee after 11 month*. Studlas using Isst chlorinated Kanechlore were similarly unable to produce cancer (Ltvlntkae 1981). Likewise, Calandrt (1976), Levlntkat (1981), and the National Cancer Institute (NCI) (1978) hsvs reported tests on both Aroclor 1256, Aroclor 1260, and 8CB In rats at dots* up 100 to 200 ppm; non* vers found to b# carcinogenic, Cslsndrs (1976) has alto reported a two-yesr study of dogs give- *os*s of l, 10, end 100 ppm that vas negative. Studies of rhesus monkeys wars also negative (Drill st si. 1962, Allen and Norbsck 1973). 311 HONS 026352 TT.t TaHt* 1. TiiMri|nlc Effeett of ECI Author* Yaar Specie* Compound, Dotage tffact* klobroufh *t *1. 1971 Hagaaakl *t at. 1971 Allan l Ahrahanaoh 197) IM at nl. 197) Allan 4 Nnrback Rloura f lab* Aura* at al. Ito at *1. Klalirougli t al. Nak litre at al. 197) 197) 1974 1974 1974 1974 Kata Hie* kata Nlco Aroclora 1141, 1140, and I254i 100 ppa/52 waaka Ranaelilor* 700 , 400, and 5001 $00 ppa Aroclora 1248, 1254, and I242i 0.12/4 waaka Rancrhlor 5001 500 ppo/ )2 week* Rliatna uonkeya Arne lor 1240* )00 pp/ ) nontha lata Ranaelilor 400 Rati Aroclora 1242 and I014l 100 ppa/10 oonth* Rat* Ranaelilor* 500 , 400, and ]00| 1000 ppo NIc* Aroelor l2S4i )00 pp/ II itonth* Rat* Ranechlor 500i O.SX fCR* Hypertrophy Hapatooaa Induced by Ranechlor 500 only Hypertrophy; ragraaalv* change* Nodular hypcrplaata and wall differenti ated hapatoealtular careinooa* In the Itver] PCI* Ineraaaad hepatic neoplatata In olca and Inctaaaad the tuoora Induced by RHC Hyparplaala and dyaplaala of tattle oueota with Invaalon of adjacent tlaaue* Utopia*tic nodular Liver call ntcrotl* Nodular hyparplaala llapatooaa; adanolIbtoala Inhtbltad th* Induction of liver tuaot* by aavarat livar carcinogen* MONS 026353 Table I cantInued Author* Na|***ki at al. Klabrouth at al. klaura at al. Vaaaaraan at al. Lavlnakaa Y**r Spec!** Coapound, Do*|t Effect* 1974/7$ Mica 1*7$ kata 1976 kata 1976 kata 1*11 Rata Xanachlora 500, 400, and JOOi $00 ppa/)l waaka Arcelor IIAOi 100 ppa/ 21 aontha PCRa Aroclor 12541 200 ppa/ 26 aontha Aroclor* 1246, I2$4, and 1260i 100 ppa/24 aontha Hapatocallutar carelnoaa only for Kanachlor $00 Hapatocallutar carelnoaa* 24/164; hyper plaatlc nodutaa 146/164; adanoftbroala Antl-proaotara In aaparlaantal carcinoganaala llapatocallular adtnoaaa Modular hyparplaala; hapatoa* 6ourcei Jim* at al. 1961. etc HONS 026354 PCB mixtures present a distinct problem vhen atteaptlog to estimate the eaneer hazard, and therefore the cancer rlak, to humans baaad on data davalopcd In rodant studies. Tha crux of tha problem la the ralavaoce to placa on an lnereaaa In rodant liver tumors when attempting to extrapolate tha human rlak. This dilemma la actually tvofold. Fine, there hea been controversy over the relevance and nomenclature of rodent hepatic laalona. Some of thla controversy relatea to the criteria uaed to diagnose caocer. The NCI hat reconaended a scheme for tha claealflcatloo of hepatocellular tumora and ralatad laalona In rata (Squire and Levitt 1975). Thla recommandatloo ha* not been universally accepted, but la used In tome of the reports of PCB carclnogeneala. The NCI claaalflcatlon echeae la extensive and specific In defining neoplastic lesions. It vas concluded by NCI that benign hepatic call tuaora (l.e., tumors vlthout the potential for malignant behavior) could not be diagnosed consistently. Therefore, a term such as "adenoma," vhlch refers to a benign tumor, vas not recoesaended. It vat also determined that the term "hepatoma," used to denote a benign liver tumor, vaa Imprecise and thus vaa not recommended to deacrlba any of tha lesions under discussion. Contrary to tradition, detection of vascular Invasion or metaetasea vaa not considered by NCI to be essential for the diagnosis of hepatocellular carcinoma. Instaad, the use of Inclusive eytologleal criteria describing changes that may or may not have the potential to become the disease process ve call cancer, Including many responses that vlll never progress to cancer, vas accepted at the expense of the more definitive criteria of malignant or Invasive neoplasms. Thus, It seems that the criteria for establishing liver cancer may vary among pathologists. Only PCBs with a chlorine content of 60S produced cellular morphological changes In rat liver that vere classified as hepatocellular carcinoma using NCI guidelines. Hovever, this asms PCB mixture vaa tasted In another strain of rat and did not significantly Increase the Incidence of tumor formation. Several other studies also failed to shov any PCB-lnduced tumors, Including a study performed by NCI. Since there la only one report of PCB-lnduced hepatocellular earclnome In rats, ve must velgh this report against other studies vhlch vere negative. As previously stated, the use of Inclusive eytologleal criteria vlll define as cancer tome hepatocellular changes that vlll never progress to cancer. Furthermore, the liver changes associated vlth PCB exposure vere reported to regrets after the anlMl has been removed from the chemical exposure (Burse at al. 1971). While some Investigators have used the Inclu sive eytologleal criteria for classifying hepatocellular laalona, other Investigators feel this classification scheme lacks sufficient diagnostic discrimination. Thus, the single report of PCB-lnduced eaneer in rata may use a pathological description that differs from the terminology uaed for describ ing the benign tumors seen in the-other studies. The second problem stems from the need to determine the mechanism by vhlch a chemical Induces eaneer. Chemically-Induced carcinogenesis la a special aspect of toxicology, since chemical carcinogens may or say not obey 314 MGNS 026355 traditional toxicologic principle*. For example, although crclnogn* ihov dota-response relationships, may undergo blotraosformatlon to active or inactive metabolites, and demonstrate specific structure activity relationahlpa a* other toxicant* do, they may alao be unlike other toxicant*: they nay not denonstrate threahold* and there nay exlat a long and indefinite latency period between the critical blochenlcal effect and any cellular or phyalologlc expreaalon (tloull at al. 1980). In recent yeara, the scientific community ha* concluded that even anong ehenlcal carcinogen* there are vaat difference* In the nechanlaa* by whleh cancer 1* Induced. Chenleal carcinogen* can be teparated into two group*: on* vhlch obey* traditional toxicologic prlnclplea and one which doea not. Weisburger and Wllllana, In Casarett and Doullt' Toxicology: The Basic Science of Folsons (1980). separate carcinogen* a* either genotoxlc (initiator) or epigenetic (promoter), the dlttlnctlon being that genotoxlc carcinogen* induce cancer by Initiating a permanent change in DNA. Epigenetic ehenlcal* comprise tboi* ehenlcal* vhlch alter the nanlfeatatlon of cancer by other than genetic nean* and Include bonaonea, Immunosuppressants, cocareloogan*, and pronoter*. Thla dlatlnctlon between type* of earclnogena la extremely Important (Doull at al. 1980, Squire 1981, Valaburger and Ullllama 1981), *lnc* Initiating (genotoxlc) carcinogen* may cauae Irreversible damage and may not have threahclde. The concept vaa generally propoaed for all earelnogen* early In the laat dated* and vaa the bad* for developing certain nonthreahold riak aaaeaament model* for vhlch aome rlak la praaant at any level of exposure. On the other hand, chemical carcinogen* that promote cell* already predisposed to cancer do have demonstrable threshold levels. Therefor*, eh* risk assoc 1*ted with promoter* differ* from Initiators at lev dose* la that the risk exists only for eh* duration that the level of the promoter axeeeda the threshold. Rodents used in cancer test* can have a high spontaneous Incidence for liver tumor*. Therefore, If a statistically significant Increase 1* seen after chemical exposure, It Is Important to determine vhathar or not this phenomenon represents true Initiation ot merely promotion of the background Incidence. Such la the problem for FCBs. To make the distinction between Initiators and promoters, the mutagenicity data and other characteristics of the chemical or the carcinogenic response become critically determinant factors. Initiators are defined a* those agents producing damage to DNA vhlch permit cell survival but predispose tha cel^l for uncontrolled growth. There fore, Initiators must possess significant mutagenic activity. There ar* several other characteristics separating Initiating earelnogena and promoting agents; soma of these are listed In Table 4. The characteristics of FCls are also listed in Table 4. It can ba seen that the characteristics of FCBs resemble those of a promoting agent and sot those of an Initiator (James at al. 1981). In his review of the mutagenic and carcinogenic potential of FCBs, levlnskss (1981) likewise cites several authors and experiment* supporting tha contention that FCBs have promoting activity, Including tha NCI bloasaay (1978). 315 MONS 026356 Table 4. Characterlttlc Effacta of Cheaical Carcinogen* upon Tlaaua Efface Initiator* Proaotor* PCB* Cheaical produce* lncraaaaa of spontaneous tuaort at multiple sltaa. Chaalcal produce* a tuaor capable of craneplantation. Cheaical produce* a tumor which doe* not regret* once cheaical it raaovad. Cheaical au*t be autagenic. Cheaical i* affective at nontoxle do***. Chaalcal may cauaa the affect after a ingle dote. Cheaical thould increa** the effect of other carcinogen* acting on the target organ. Ptogresilon of neoplastic differentiation doe* not requira the continued preaenee of the chaalcal. A - i t - - - - - - t * generally * ioactlmaa * seldom or navar MONS 026357 316 In summary, vhcn assessing tha carcinogenic hazard of PCBa, one buic decide what velght to give each portion of the data. While one cannot Ignore the fact that posltiva carcinogenic activity haa been reported In rodenti, one suit take Into conaldcratlon the veakneaa of thla reaponae alnca It vaa not reproduced In aevercl ocher atudle*. Furthermore, there it a great deal of evidence to auggeat that If PCBe induce changes, It la.by vay of a proaiotlon mechanlam. Thus, on the beala of the animal data and considering the Inconsistency of the response, PCBe, aa veak promoting agents, at lov doses would likely represent a relatively small, If any, cancer hazard In man. Recognizing the problems associated vlth accurately Interpreting tha relevance of and risk inferred by animal carcinogenesis bloassaya, Robert Squire, former acting director of the Carcinogenesis Testing Program and head of NCI's Tumor Pathology Section, has proposed a ranking scheme for the animal evidence (Squire 1981). Squire proposes that the basis for determining the strength of a positive test resides in vhether the "suspect carcinogen" is positive In a number of species, gives rise to hlstogenetlcally different types of neoplasms In cne or more species. Induces malignant rather chan benign tumors, snd Is consistently positive In an appropriate battery of taste measuring genotoxlclty. The dose required to Induce tumors ae veil ae the spontaneous Incidence of neoplasms In tha control animals are also criteria. By ranking the test data for each criteria according to a point system developed by Squire, one gets an overall eeore to assess the strength of the evidence. Squire's approach Is appealing because It recognises several festures of an snlmsl bloassay that should be considered when attempting to estimate the likelihood that the human response Is similar. Since there are several cancer teste In animals for FCBs, the aggregate score that PCBe would receive depends on ehe velght assigned to the positive and negative data. Kovaver, the total score'for PCBs, regardless of the scoring, vould probably oe about 31 or lest. Soulre classifies a chemical receiving a score of less chan 41 ae a class V carcinogen, a claee for which he suggests that tha regulatory options are many and far less restrlotlve than for chemicals scoring higher In the ranking schema. Thus, according to this schema, even the "positive" evidence for PCBs does not rank them as a human cancer hazard that is as significant as tha hazard posed by many other ebemleals, some of which are still commonly used. . EPIDPilOLOCT Acute and Chronic Human Oats In recent etudles of occupational exposure, PCBs have not been proven to be a remarkably toxle chemical to persons In the work place, who have probably received the most significant exposure to PCBs. Ouv and eo-vorkers (1976) studied 34 vorkers from a single capacitor plant using Aroclor 1262. Of this group, 31 were exposed for more than one year and 16 for five years or more. Although Industrial hygiene at the plant had been generally poor prior to their study (e.g., air concentrations vere 1502 to 2222 of the allowable workroom air standard of 1.0 mg/m*), the clinical tests did not identify any significant health problems. One case of chloraene was reported as vail as iiv cases or an eczematous rash. Evan though tht avarags FCB blood lavtls of tha group wars ovtr 400 ppb comparad to nona dacactad In tha control study 317 HONS 026358 group, tht clinical tests measuring liver function revelled that bilirubin, ilkillne phoiphitue, terun protein, albumin, SCPT, and immunoglobulins for the group of factory vorktrs vere vlthln the normal limit* for each teat. In addition, bronoaulphaleln clearance did not correlate veil vlth either PCS blood levela or length of employment. A sore recent study of a far larger group of employees fro* a capacitor plant reported a similar lack of untoward findings (Fischbein at al. 1979). Based on this study of 326 employee*, it was possible to separate the employment records Into the following categories for years of exposure: 10Z had been exposed for 5 years or leas; 20.9X 5 to 10 years; 17.5X 10 to IS years; 11.6Z 15 to 20 years; 29.1Z 20 to 25 years; and I IS had been axposed for mors than 25 year*. A sumary of the akin complaint* revealed that 10,7Z of the employee* reported rathe* and 26.8X reported a burning aansatlon of the skin. Physical examination further revealed some redness, swelling, dryness, or thickening of the skin in approximately S0Z of the Individuals, while 2X experienced abnormal aeeretlooa from the eye*. The clinical chemistry of the study was generally unremarkable, and the researchera commented on the "paucity of abnormal result*." Of 321 parsons examined, the percentage of templet outside of the normal range vere: BUM 5.31; creatinine 1.7Z; SCOT 2.2Z; SCPT 7.2Z; LDH 2.5Z; alkaline phosphatase 1.2Z; SGGT 1.9Z; bilirubin 0.0Z; cholesterol 17.8Z; triglycerides 10.SX; and total lipid* 3.6Z. These data can be considered Indicative of the type of random test results to be expected when 321 persona are tested without simul taneously running a control group to compare against the standardised "normal" values used to indicate an abnormal finding. Likewise, the neurologic exami nations failed to produce any prevalence of abnormal findings. There was a decrease in the forced vital capacity of the lungs In 14Z of these workers compared to only 5.6S In the normal population (Varahav et al. 1979). While chis is an unusual finding, this change vaa not seen in one group, l.e., the nonsmoking women tested. The significance of this finding Is unknown at this time and require* verification from ocher studies. Maronl et el. (1981a, 1981b) have reported a study of Italian employees in a capacitor plant exposed to a foreign PCB mixture similar to Aroclor 1242. The mean age and employment of the 80 employe** examined va* 37 *ndjl2 years, respectively. PCB air concentrations ranged from .048 to .275 mg/m , and all tested surfaces vere heavily contaminated. Indicating that employee exposures were both dermal and Inhalation. The rasults shoved that PCB blood concentra tions ranged from 88 to 1319 ppb. Fourteen workers vert found to have skin disorders consisting of dermatitis, folliculitis, or chloracn*. Sixteen employees were reported to have abnormal liver effaces, consisting of hepa tomegaly and sometimes an lnersase in serum level of liver ensymes. The liver findings did not correlate well with either the duration of employment or the PCB blood levels, and the pattern for hepatic ensymt changes was a random finding inconsistent with hepatotoxle chemicals. It 1* evident from these data that the clinical changes vara unremarkable. Furthermore, control subjects were not included to determine the normal test variations among unaxpeeed person* of similar age*. Serum bilirubin and alkaline phosphatase levels were within the normal rang* of variation for all subjects. 3ie HONS 026359 Sob* of the conclusions reached In tvo racanc revlav* of tht vriou studies of PCS txpoaures to <Jat (Gaffey 1981, Btovn at *1. 1981) tr* note worthy, First, vhll* higher axpoauraa to PCS* tend to mean a higher body burden, chare was a general lack of correlation b*twain axpoaura duration and body burden. Second, although not all studies agree, theta was also a suggestion that the higher chlorinated PCts were note likely to aecuaulata In sdlpose tissue, suggesting a slower aetabolln rate, third, the general health of the occupationally exposed group vat considered good. In a number of studies revieved by Galley (1981) tvo ol the studies rtvleved failed to report clinical findings (other than stating that the workers were in good health) while one measured PCS blood level* of up to 1900 ppb m Finnish workers. 01 IS studies reporting clinical aeaeureeents, 11 reported dermatologic effects, 9 reported liver function tacts, 6 reported lipid aeatureaent*, end S reported on blood ehenlttry. Gaffey (1981) con cluded that vhll* there is re clear correlation betveen PCB blood levels snd chlorecne, the studies suggest that when PCB blood levels exceed ISO to 200 ppa, chloraen* can occur. He also concluded that dermatitis, like chlorecne, vaa a frequently observed effect, and that it aay be associated with the higher chlorinated PCB compounds. Of the nine reported studies of Uver function, only five found some mild change in liver function tests. Mo consistent pattern was identified, not we* an association betveen thee* change* end PCB blood levele Identified. Mo advert* health effect was associated with PCB expoeur* in any of the etudles. Of the tlx studies chat considered lipid metabolism, no cooeietent pattern associates vleh PCB exposure; on* study, by reporting a decree** In ell paraaettrt, eonttsdletad ell of the studies reporting an increase. Host etudles did associate increased triglycerides vlth the PCB expoeur*, hut the result* for cholesterol were inconsistent. Me clinical elgnlfietnce he* been attached to these findings. Of the five etudles of bleed chemistry, none reported any relationship betveen the test! snd PCB blood levels. Of tht tvo studies that assured blood pressure, on* found no association, while one reported an association betveen PCB levels and diastolic blood prastut* (Gaffey 1981). In (unary, a review of the human data from etudles of persona occupa tionally exposed or whose PCB levels were Increased via fish consumption reveals that tht combined result* of all studies, while demonstrating PCB* caused dermal problems, felled to identify any significant clinical disease essocltted with electrical grade FCBe. While some physiologic ehengss vara noced, no clinical Igalficanca could be attached to any of these, partic ularly the ehangee la liver function, which were minimal and generally not compared to an appropriate control group. It has alto baas noted that PCB* can indue* oxidative metabolism In persons occupationally sxpesed to thorn (Alvares ot el, 1977), Animal atuditt suggest that inducing agents can inersss* serum treneamlnae* level*, e.g., SCC7 (Whitfield at *1. 1972, Martin at al. 1975) and triglyceride* (Martin *r al. 1975). Thu*, the random, minimal increase* In eetum transaminase levels and Increased triglycerides found in parsons occupationally axposed to PCBs nay represent the predictable phenomena of nodoet entym* Induction. The National Institute of Occupational Safety and laalth (MIOSB) atudv likevia* indicate* that chronle liver, nervous, #t etrevlatory diseases tr* aot atsoclaesd vlth long-term exposure to laval* cf PCBs vhlcn were much higher than those encountered in the environment. 319 HONS 026360 Therefor*, if the environmental exposure to PCBt Is insufficient to Induce ehloracne (the most frequently observed effect), scute, subchronic, or chronic toxicity is probably not to be expected in hunans. In 1968 a mass outbreak of ehloracne and other tymptoma, later taraed "Tusho," vaa caused by the ingestion of cooking oil contaminated by PCBa and other chlorinated compounds. Hlguchi (1976) and Kurateuna (1976) reported that approximately half of the patients had abnormal lipid metabolism. Serum triglycerides vere elevated, while cholesterol ves not. Of 63 effected children examined, 23 boys showed transitory decresses in height and weight gain when compared to unaffected children, while 19 girls did not differ from the control group. Also, some of the infants born to women affected by Tuaho vere amall-for-date. Other symptom* la-newborns Included derk-brovn pigmentation, parchment-like skin, eruption of teeth, and larger than usual fontenelles. Indicating transplacental passage of PCBe to the fetus during gestation. Probably the most common symptoms and the ones largely responsible for the Identification of Yuaho's disease vere eye and skin problems. These consisted of acne-llka eruptions, follicular accentuation, availing of the eyellde with discharge fro* the eyes, end Increased pigmentation of the skin. Unfortunately, even though the ehloracne vae noe e permanent condition, it wee a discomforting disfigurement that lasted for a period of month* to years. Urabe at al. (1979) hava raportad on tha cauaa of tho daatha of 31 of tha 51 parsons vho died during tha first dacada after tha Yuaho Incident. Eleven death* rsaulted from varlou* neoplasms; a 35.4% caneer rata, which is higher than the 21.1Z rate generally reported for that area of Japan. No conclusion can be dravn from this at present, since tha Yuaho rata vaa not age-adjustad and further analysis of the Yuaho dltaas* must be don* at a later data. Of the reported cancera, the types of neoplasms varied. The number* of cancer for aach tissue vere: two stomach, one stomach and liver, two liver, three lung, one breast, and tvo malignant lymphoma. ' While the above effects described In Yuaho patients give some Idea of the possible consequence* of overexposurt to PCBa, eht exact cauaa of those symptoms is debatable. Tha type of PCB mixture contaminating tha rlet oil, Kaneehlor 600, his boan dlscovarad to contain polychlorinated dlbenzofurans (PCDFa) at concentration* aa high as 18 ppm (Hlguchi 1976, Kuratsuna et al. 1972). Measurement* of tha rlea oil used by Yuaho patlanti rovealsd that It contained PCBa ranging from 1 to 3000 ppm, 5 ppm of PCDFa, and approximately 1000 ppm of polychlorinated qutterphanyls (FCQs) (Kuratsuna 1976, Nagayama at al. 1976, Kuratauna 1980). Therefor#, it 1* not knovn which of tho Yuaho aids effect* can b* attributed solely to PCBs (Vos et al. 1970) and, If cancer la found In the future to hava bean significantly Increased in this group, Che source of cancer Induction will remain unknown, (Note: The PCDFs found In the Yuaho PCBa era about 1000 time* higher than those usually measured In American sourest af PCBa). Recently, Kuratauna (1980) reviewed the Yusho episode and suggested that tha levels of PCBs end contaminants were actually about half of the amount previously repertsd. He and others (Kashimoto at al. 1981) have also suggested that since e considerable number of victims did not racover from their symptoms after their level* of PCB* had dropped neer normal 320 MGNS 026361 levels, the high levels of contaminants still present nay be the more likely causal factors. Reports of high cancer rites among Mobil Oil eoployees exposed to PCBs (Aroclor 1254) at Mobil's Paulsboro, New Jersey, refinery have been Inter preted as Indicative of a possible link between PCB exposure and skin (mela noma) or pancreatic cancer (Bahn et el. 1976). The preliminary Mobil study reported eight cancers which developed between 1957 and 1975 among 92 research and development and refinery workers exposed to varying levels of Aroclor 1254 for five or six years In the late 1940a and early 1950a. Of the eight cancers, three vere malignant melanomas and two were cancers of the pancreas. Accord ing to N10SB, "This la significantly nore skin cancer (aclanona) and pancreatic cancer than would be expected In a population of this else, based on the Third National Cancer Survey." Hovever, It la difficult to drav any conclusions from this study because of the taall numbers of individuals exposed end the variety of other chealcals to vhlch they vere exposed. In any case, the study was vlthdrevn for revision and has not yet been re-relaaaad (Caffcy 1981). Of the studies reporting the effects of PCBs on the chronic Illnesses or the aortallty Incidence In humans, only two provide large population! of apparently single chemical exposure. Brown and Jones (1981) conducted a rttrospectlve cohort mortality study of 2,567 employees from two plants vhege PCBs were used to manufacture electrical capacitors. All workers Included -in the study were employed for at least three months In areas of PCB exposure." The vital status of 98X of the population was determined and 39,018 person years were accumulated. The types of PCBs used were Aroclor 1016, 1242, and 1254. The major causes of death for the 163 persons who had died by the time of the report are stperately listed In Table 5. The only statistically Signifi cant difference observed was for cancer of the rectum, but only If the female population of plant 1i la considered rather chan the total population. Hovever, the plants studied are located In an area vhere mortality from rectal cancer Is greater than ehe United States average (Gaffey 1981). As mentioned, this unusual Increase stened from a high number of deaths In the female population of plant 2: similar findings were not reflected in the female population cf plant 1 nor In the male population of either plant. Uhlle the liver cancer incidence was higher than expected. It was not significant. There vas no relationship between duration of employment involving PCB exposure and the risk of mortality due to cancer or cirrhosis of the liver. There vas no statistically significant relationship for any cause of death when the total exposed population was considered. Bertsssl et al. (1981) studied 27 deaths In 1310 workers exposed to PCBs for 20,565 person years. Mortality vas studied for the 25-yaar period between 1954 and 1978 for employment between 1946 and 1970, The study group had been employed for at least six months and the exposure vas largely to Aroclor 1254, For all eanears, the observed mortality was 14 versus 5.65 expected. This Increase, however, was significant only for the male population. Higher than expected incidences In cancer of the digestive organs and In cancer of the lymphatic cr hematopoietic systsms vere also observed, although these vere not significant, there Mere no reports of liver-cancer. 321 HONS 026362 Table 5. Major Cauae* of Death for Plant Worker* Expoaed to PCBe: Aroclor 1016, 1242, and 1234 Cauae of Death Ob eervcd/Expected All malignant neoplaea* tyataa Accident a All other cauaet All cauaet Cancer of atomach Intttclne Rectum Liver Pencreae Reaplratory Breaet Lymphatic Other All Cauaet 39/43.79 11/12.55 60/62.93 13/18.29 40/44.79 163/182.35 tUllcnant Neoplaeae 1/1.66 4/4.03 4/1.19 3/1.07 1/1.90 7/7.98 7/6.84 2/4.34 10/14.78 Source: Brown and Jonee 1981. 322 Standard Mortality Rate (89) (88) (95) (M) (80) (89) (60) (99) (336) (280) (53) (88) (102) (46) (68) HONS 026363 In assessing the epidemiologic evidence for carcinogenicity, It must b* remembered chit In sny epldeaiologlc study It may be lapoulbli to illalnati all of the unrelated but confounding variable* in the observed population*. Moreover, it 1* not uncoanon to ** eeaaurable Inert**** in on* type of cancer over what 1* mathematically expected, alnc* v cannot choo** a *ubpopul*tlon that reflect* exactly the baaellne cancer rate* of the toed population. In fact. It 1* well knovn that different are** of the United State* have dlfferine background rat** fot many of the *p*clflc type* of cancer. All thl* must be considered carefully, and no conclusion* should be made concerning any finding of excel* cancer which 1* not statistically significant, but only above expected value*, unlass thl* same exces* for that specific type of cancer 1* repeatedly demonstrated In subsequent studies. A positive correlation between exposure to a chemical and a resultant adverse health effect relies on the following conditions (Doll 1980): o A positive association must be seen In Individuals with known exposures; o The positive association cannot be explained by bias In recording, detection, or experimental design; o The positive association snist be statistically significant; o The positive association should show both do** and exposure period dependency; and o The positive association must be observed repeatedly In subsequent studies and cannot be a single, confounding,, variable observation. To date, the mortality studies concerned vlth PCS exposures present several problems of Interpretation, the most obvious and Important of which Is the different increase* In cancer type reported. Taken as a whole, these studies provide no support for assertions that PCBs sr* * eancsr-causlng cnsuictl. The variety of tumor sites found to be of most concern within tach study largely dlffara from the animal data and la Inconsistent with the selectivity of currently knovn promoting aganta or Initiating carclnogans. In summary, epidemiologic studies have demonstrated that coaaerclal PCBa art not remarkably toxic ehsmlcals after acute expoeur* end that, whan excess xpoaura doe* occur, the usual consequences are dermatologic and not of a serious or ptrmanant nature. Data from Investigation* involving chronic exposures yield similar conclusions. The majority of studies has not Identi fied a ellnleel disease associated vlth PCB exposure, nor has It provided persuasive evidence of health Impairment. There le no persuasive evidence of an excess In total mortality or In mortality due to eancar, cardiovascular dliesst, or nsrvoua ayataa disease associated vlth occupational expoeur* to PCS*. PCBa have not been linked to any human canear. future Investigations of occupationally exposed populations will no doubt clarify this issue as tha total mortality within the teat group increases vlth time and yields a larger mortality data bass. 323 MQNS 026364 PROCEDURES FOR ASSESSING RISK In attempting to extrapolate the human rlak from chemical expoeure using animal toxicity aa the baslt for the rlak extrapolation, there are eeveral model* to chooae from, the model choeen la primarily determined by the type of health haaard of moet concern. However, speaking In general term*, there are only two typee ueed. The flret type conslata of thoee method* uaed for extrapolating the human rlak directly from the dot* for which there wee no obtarvable animal toxicity. Thla method can be applied to moat toxicltle* (except cancer), tlnct chrechold* are aaeumed for thee* raeponttt. The aecond type of model le generally uaed to aeetai the rlek aaaoclated with carcino gen*. Since many eclentlata ataume no Identifiable threahold for thla toxicity, there la tom* rlek attached to any expoeure. Thla concept dictate* that mathematical model* be uaed to extrapolate to expoeure* far below thot* dosage* Inducing observable response* In the teat animal population. For noncancer toxlcltles (1.*., threahold toxicltle*), the models for extrapolating rlek are relatively simple and similar to the Suggested Ho Adverse Response Level (SHARL) proposed by the Hatlonal Research Council of the Hatlonal Academy of Science (1980), or the Lowest Observable Effect Limit (LOEL) uaed by the United States Environmental Protection Agency (EPA) (1980). Basically, this type of calculation la computed by assuming that humans are aa sensitive as the test spades used. Therefore, the amount Ingested by the test animal that gives no toxic response Is eh* sal* upper limit of exposure for humans, l.t., the human thrashold dose. The calculation essentially sukes the conversion on the bests of the els* differential between humans and the test species, usually a body velght com parison rather than surface area. It take* into consideration those factors that affect a truly comparative dose, e.g., absorption factors If the human exposure differs from the teat conditions or If the animal dotage regimen differs from eh* human expoeure interval. The calculation Is similar to the following: Safe Human Doe* H>o.o(ng/(kt) x 7Q(kx) x A.F. x E.R. x tl/2 - fmg/dey x S.P. TDo.o * threshold, or no observable effect dose In the test species; A.P. absorption factor, or ratio of animal absorption divided by the human absorption (absorption animal/ absorption human); E.R. exposure ratio (dosage regimen enlmel/exposure regimen human); tl/2 S.F. * ratio of accumulation (helf-llfi In animal/ half-life ln*humans); ted safety factor, which depends on the reliability of th* data uaed for extrapolation. 32* HONS 026365 Typically, th safety factor used v*rie fro* 10 to 1000, depending on the extent of animal data available and vhether or not there la any human data to aubatantlate the reliability of the number. Of course, the number calculated ahould uee chronic data vhere chronic expoaurea are expected. Thla type of model calculatee one value. Exposure at or below ehla value la conaldered aafe. the aecond type of calculation alto ha* aeveral method* or model* from vhlch to chooa*. However, all differ In their baale aeaumptloos or In the mathematical expreaeloo uaed. Hence, at the low expoaure* In question, their eatlmatlon of risk can vary dramatically (OTA 1981). Regulatory agenda* currently ua* the multistage model propoaed by Crump (EPA 1980, OTA 1981). Ualng this model, thm-riak-i* linearly proportional to the expoaure at lov doses. While each exposure Is considered to carry some risk, the acceptable rlalf or safe dose Is usually suggested as the exposure range for a 10*5 to 10 risk. This masna that dally exposurg to that dose vould Increase cancer by 1 person In 100,000 expossd (1.*., 10) or 1 in 10,000,000 parsons (10*') with a lifetime exposure. Risk to PCBS As statad earlier in this discussion, the determinant factor for assessing the risks sssodatsd with chemical exposure Is usually ths toxicity (acute or chronic) of most concern that can be expected to occur In human*. With PCBs, a* for aoy chemical, datervlnlng tha toxicity of moat concern la extremely Important to enttlng safe expoeure limits, likewise, determining the toxicity of moat concern elao determines tha public's psrceptlon of ths harm of ovsrsxposur* to PCBs, vhlch In turn affects risk assessment. As demonstrated by the types of analysis used, ch* difference of whether or not a chemical is carcinogenic ean translate Into magnitude* of differences In ths calculated dose conaldered to be safe. PCBs ar* s good axsmpl* of this dllesma. As seen In ths EPA's ambient vstsr quality criteria for PCBs (EPA 1980), ths safe doss calculated If PCBs ar* not considered to be carcinogenic vas 210 ug/dsy*, if PCBs are considered to bs carcinogenic, ths 10*3 risk dosage Is approximated at 180 ng/day. This 1* grsacsr than a 1000-fold difference in the "safe" exposure based on the chemical1a carcinogenic potential. Thus, determining ths carcinogenic potential of PCBs accurately has a large impact on Its perceived risk. Evan though many arguments may bs mad* regarding the carcinogenic potential of PCBs, ths answer to ths question regarding which lsval may b* safe ean b* found by s simpler, mors direct method. While ths preceding extrapolations might provide useful guideline* for safe exposures vhsn only animal data srs available, they can b* and should bs tempered by ths human experience whenever possible. While human data may not prove that the extrapolations from animal data ar* correct, they can bs utilised lu soma lnttsncas to determine if those extrapolations are Incorrect. Such Is ths cat* for PCBs. Ths model predicting the cancer risk from PCBs based on ths data from Kimbrough at al. study (197$) Is Uossr In ths dostg* rang* calculated for acceptable risk*. Therefore, linear extrapolation to that doa* approximating 325 MONS 026366 a lOOt cancer Incidence in humans using the slope for the curve in that portion of the dose response for lov risks provides an overestination of the actual dose required for a high Incidence of cancer (see Figure 1), Perform ing this extrapolation for PCBs (l.a., 10'5 risk is 160 ng/day, so a risk of 1 Is 103 x 160 ng/day equals 16 mg/day), At about a 16 ag/day-exposure of PCBs, we should expect a liver cancer incidence of 100X in the exposed population if the low exposure perception of the risk is correct. However, the Occupational Safety and Health Administration (OSRA) Halt , for occupational exposure to PCBs was 0.5 mg/m3 for Aroclor 1254 and 1.0 mg/m3 for Aroclor 1242 and other, lower chlorinated biphenyls until PCBs were banned. Assuming that the average voluae of air Inhaled per vorkday is approximately 10 a , then the allowable workday intake of PCBs could approx imate 5 to 10 ag. Granted that perhaps not every worker was exposed dally to these levels, reports such as Ouw et al. (1976) or Brown and Jones (1981) Illustrate that it is a reasonable assunptlon; therefore, we can assuae that the dally exposura approximates of our astuaptlon are a practical considera tion. Alto, the slope of the extrapolated line would be such that the ctncer risk above 1 ag/day of PCB exposure would be easy eo identify in huoans as it is about a 6X of the persons exposed. Since neither the study of Brovo and Jonas (1981) nor of Bertatsl et al. (1981) found any significant Increase in liver cancer, and since neither could demonstrate a consistent cancer risk, one aust conclude that the cancer risk extrapolated froa the animal data for low dose exposure is wrong and possibly by at least two to three orders of aegnltude or grester (depending on the actual relationship between lines A and B In the schcatatlc of Figure 1). By coaparing the rlak calculated using anlaal data to that found in 2,567 eapacltor workers (Brown end Jones 1981), we alght conclude that one of three possibilities or a combination of all three exists to explain this error in our calculations: (a) the aodel is wrong, (b) the anlaal species used does not represent aan, or (c) PCBs are not carcinogenic in human* at occupational exposura levels. The first possible explanation aay contribute to the problem, but is least likely to be the causal factor for, if reasons 2 and 3 were false, then an Increased incidence of cancer in capacitor vorkers is likely to have been aeaturable. Therefore, one can conclude that either 2 or 3 or tome combination of the three is the more likely reason for the discrepancy. In any caia. it can he demonstrated that, at the very laser, the reeulta of the enlael tests of Kimbrough et el. (1975) do not sppeer to represent c aodel or test data relevant for assessing the cancer risk in humans (Jaaes ec al. 1981). Interestingly, if the modeled cancer risk in the EPA report (1980) Is off by only three orders of magnitude, we reach a risk of cancer that la equivalent to the calculated LOEX in this document of 210 ug/day. It should also be mentioned that this saaa test cf the 10~* risk based upon animal data for DDT and r*>lordane suggests s lsrge ovcrsstlmatlcn of the risk for these chemicals as veil. Thus, this type of extrapolation teem* to be inappropriate for other chlorinated, persistent chemicals which enhance or product liver tumorlgenesis in rodents probably via epigenetic mechanisms, ESTIMATION OF THE HEALTH RISKS ASSOCIATED WITH ENVIRONMENTAL EXPOSURES Early in this paper it was stated that a thorough evaluation of the hazards assoclstsd with PCB sxposures must consider and weigh six different 326 HONS 026367 Cancer R isk loot A 8 Doae Laiand A Actual Rlak of Data or Modal B Extrapolated rlak from low doaa region 0A Aetual Doaa rcpreacntlng 100Z rlak of cancer DB Extrapolated doae rapreaantlng 100X rlak of cancar Figure 1. Extrapolation froa Low Cancer Rlak Expoaure to High Cancer Rlak Expoaure 327 MONS 026368 characteristics of the data, The first tvo factors deal vlth the breadth and variety of toxic responses produced and the degree of species variation associated vlth the effects monitored. Many of the PCB-lndueed adverse effect* are characteristic of a chlorinated organic compound of the PCS type, or at least these effects are conslstantly produced by many chlorinated organics. The effects include liver injury, liver enlargement, liver enzyme induction, irritation vhen applied to the skin, ehloracne, and (poaalbly) a disruption of the estrogenic balance in females due to Increased liver metabolism. Although these may indeed be considered adverse effects, all chemicals are capable of alter lng some physiologic function at a high enough dose. In the more relevant reproductive mutagenic, teratogenic, and carcinogenic tests, teats of chlori nated biphenyls were consistently negative or inconclusive. Some of the studies can be discounted because of inadequate study design, disputable interpretations, or use of unusual test methodology. It Is unnecessary to display a disproportionately high concern for those fev positive studies vhen their findings cannot be reproduced In e number of subsequent studies conducted under similar or identical conditions. Third, regarding mechanisms, sufficient evidence exists to Indicate that PCBs have the potential to promota cancar rather than lnltlata tumors. Similarly, the metabolism and elimination of PCI* In primate* 1* far lover than that of rodantt, suggesting that man la far last likely to genercte toxic metabolite* than rodents, l.e., man la probably less susceptible to llvar injury and any impact thla may have on eumorlgeneala. The fourth factor regards the relevance of the performed test. For the most pert, this was not factor vhen considering the results of most tests. There are, hovever, eome probleme in attaching the seme significance to all of the mutagenicity tests. The human relevance of single-stranded breaks in DNA at high In vitro ?CB concentrations or tha applicability of tha chromosomal easts In birds Is questionable vhen compared to the more often used and better eetabllehad teste for mutagenic or clastoganlc activity In mammalian and bacterial test syitema. Thus, ve have concluded that tha mutagenic, and tharafore Initiating, activity of PCBa In humans is minimal, l.e., It Is not a significant hazard to humans considering tha overwhelmingly negative test responses. Ths rtlavanca of extrapolating one positive carclnoganetls bloaasay In rati without balancing It against numerous negative once In this seme species also saams Inappropriate conalderlng the likelihood that PCBs art promoters; that tha organ response (lie., llvar cancer) has a high background incidence In rodanta (Blckarton 1971) and la sutcaptibla to promotion by such effects as Induction (Paralno at al. 1973, 1977) and recurrent llvar Injury (Data at al. 1976, Bareublum 1979, Schumann at si. 1980) (tvo known affects of PCBe); the species-strain preference of the positive effect; and the irreproduclhlllty of this effect. Considering tha promoting praaaures that high, chronic doses of PCBs could sxert, It is probably more unusual to find a number of responses .in rodenee negative. The fifth factor deals vlth the dosages required to produce animal toxleltle* ae compared to the expected human exposure. In animals, PCBs art acutely toxic only at high doses. Similarly, their chronic effects generally occur at relatively high dotes. Thus, taking into consideration the expected 323 MQNS 026369 lov environmental exposures to humans, PCS* do not appear to pose a significant risk. If akin disorders arc not Induced, other effects seen unlikely. The last factor deals with risks determined by human exposures. The epidemiologic studies have demonstrated that PCBe are not remarkably toxle chemicals sfter acute exposure and that, when excaes exposure does occur, the ususl consequences sre dermatologic end not of a permanent nature. In spite of over 50 years of use, PCB chronic exposures have added little or no additional adverse effects of note to this picture. The majority of studies has not Identified a clinical disease associated with PCB exposure other than chloracne, nor has It provided persuasive evidence of an unusual health Impairment. There Is no consistent evidence of an excess In total mortality or In mortality due to cancer, cardlovescular disease, or nervous system disease thet can be associated vlth occupational expoaure to PCBe. PCBs have not positively been linked to any specific human cancer. Therefore, It appears that PCBs are not a remarkable toxicant, but ehemleals that require high doses to produce harmful effects. In evaluating the risk associated vlth the environmental exposure to PCBs, ws should ansvsr two qusstlons. First, Is the level of expoeure of a ufflelent magnitude to constitute a hatard? Jellnek and Cornaliueeen (1975) astlmatad that tha dally PCB Intaka for a young mala adult dropped from approxlaataly 15 ug/day In 1971 to 8.7 ug/day In 1975. Thla aatlmata vaa baaad on tha assumption thet most of tha Intake vas from PCB-coneamlnated fish. Tha estlmata would not bo accurate for paopla who wero not oatlng auch fish, but nalthar Is It accurst* for person* who wore sxposod to wests sltas, contaminated v*ter, or other unusual sources. MacLeod (1981) more recently reported that Indoor air pollution with PCBs la greats^ than outdoor air pollution, and often ranges from 0.1 ug/aJ to 0.2 ug/m of air for many of tho locations tasted. Thla approximates to a 1.0 to 2.0 ug/day exposure to PCBs from indoor sir. Thla exposure elso approximates tha currantly recommended NI0SH expoeure level for Industry. Thus, It appears that persons within the United States sre environmentally exposed to microgram amount* of PCBs per day In what are probably tha higher expoaure instancas (excluding unusual condi tion*). This exposure le 1000 times lover than former occupational axpoauras, which, according to raeent evidence, produced no problems other than chloracn*. It would seem that a 1000-fold reduction In expoeuro would be sufficient margin of safety for any weak threshold tffacta that vara Identified, e.g., akin disorders and Induction of liver ensymes. while the NI0SH level is lover than the 0SHA standard, It would seem that the old limit reasonably protected the occupationally exposed persona. This bodaa veil for the rest of ua considering our far lover axpscttd axpotura from the environment. Furthermore, calculation* reveal that PCB exposures of mlcrograma per day art not Ukoly to significantly altar a person's stsady-stot* tlssus concentrations. For example, eht svarsg* 70-kg man la mad* up of 152 or greater body fat. In other vorde, he haa at least 10.5 kg of fat In which to store PCBs. Nonoccupatlonally exposed persons usually contain 2 ppm or loaa of PCBa in their body fat, thua, at 2 ppm tha average 70 kg man vould contain 21,000 ug or lata of PCBa. If on* assumed a vorat-ease estimate without considering metabolism, axeretloo, other tissue* etc.. It vould take 26.8 years at a dally exposure of 2 ut/day to double hi* fat levels of PCBs. At 10 ug/day, It vould take 5.8 yeata; at $0 ug/day, *c would taka' 1.2 years; and at 100 ug/day, It vould taka 0.6 years. Since the helf-llfe of PCBs le probably no longer than 329 HONS 026370 veral months, th* actual time to double the body burden la much longer than the above calculation* and would never be reached for the lover exposure*. Thus> th* exposure to PCBs faced by th* general population! estimated at about 5 ug/day or less for those exposed, Is not likely to have much of an impact upon the low body level* of contamination that have previously been measured In nonoccupatlonally exposed persons. Since these levels are far lover than the tissue concentration* experienced by capacitor worker*. It Is unlikely that any untoward consequences can be expected In th* general population. Th* calculations also deaonatrat* that occasional exposures to auch higher level* are not likely to Increase a person's adipose levels abova tha l to 2 ppa commonly measured. From th* above discussion, It can be seen that the current dally environ mental exposure most likely encountered from the environment la not of a sufficient magnitude to constitute an additional haxard because It 1* not likely to significantly altar our currant body burdens. Th* animal data ha* Identified certain hazards, but the negative data suggest that these hazards ar* not consistent or potent responses In every species. While va might speculate that very large does* of PCI* may causa various detrimental affects, th* highest exposure group, capacitor workers, appear* to have bean exposed at levels below those that would causa serloue adverse effects In human*. Thus, the current environmental exposure* to PCBs, which are much lower than previous occupational exposures, do not appear to be associated with an ldentlflabia hazard, and so the human risk seems to be minimal or Insignificant. In order that this paper la not misinterpreted. It Is emphasized that this conclusion does not spply to concentrated wastes or chemical vast* sites where the exposure may be higher or of a different composite nature. It la alto empha sized that higher than mlcrogram-per-day exposures to "pure" PCBa do not necessarily carry an added risk. Th* Federal Drug Administration (TDK) guidelines of 1 ug/kg/day, or a single exposure of 100 ug of PCBs, it still far below the previous OSMA regulation and occupational exposures or even the EPA's conservative estimate of tha 10CL. Until definitive evidence clearly demonstrate* a chronic toxicity in capacitor workers or other exposed persons, environmental exposures to coamMrclal grade PCBs (with low levels of PCDFs), which are far lass than tha former occupational exposure levels, appear to carry a minimal and Insignificant risk to humans. HONS 026371 330 REFERENCES Abrshamson LJ, Allen JR. 1973. Th biological response of infant noohuman primates to a polychlorinated biphenyl. Environ. Health Perapect. 4:11. Allen JR. 197$. Reaponte of the nonhunan primate to polychlorinated biphenyl exposure. Fed. Proe. 34:1675. Allan JR, Abrsheaeon LJ, Norback D.8. 1973. Biological effects of polychlorinated blphenyle and trlphenyla on tha subhuman primate. Environ. Rea. 6:344. ' Allan JR, Baraotel DA, Caratana LA. 1980. Residual affects of polychlorinated biphenyls on adult nonhuman primates and their offspring. J. Toxicol. Environ. Health 6:55. Allen JR, Norback DH. 1973, Polychlorinated biphenyl and triphenyl Induced gaatrlc mucosal hyperplasia In primates. Science 179:498. Alvarea AP, Flsehbein A, Anderson KX, Cappae A. 1977. Alteration In drug metabolism In workers exposed to polychlorinated biphenyls. Clin. Pharmacol. Ther. 22:140. Bahn AX, Roaenwalke X, Herrmsnn.H at ai. 1976. Malanoma after exposure to PCBs. Nev Engl. J. Med. 29S:450. Bailey J, Knauf V, Mueller W, Robson V. 1980. Transfer of hezachlorobeniena and polychlorinated biphenyls to nursing Infant rhesus monkeys: enhanced toxicity. Environ.* Res. 21:190. Barsottl DA, Allen JR. 1975. Effects of polychlorinated biphenyls on reproduction In the prlsute. Fed. Proe. 34:338. Barsottl DA, Marlar RJ, Allen JR. 1976. Reproductive dysfunction In rhesus monkeys exposed to lov levels of polychlorinated biphenyls (Aroclor 1248). Fd. Cosset. Toxicol. 14:99. Berenblum I. 1929. Tumor formation following freezing with carbon dioxide snow. Brit. J. Exper. Path. 10:179. Bertazzl PA, Eocchettl C, Cuercllena S, Foglla MS, Pasatorl A, Rlboldl l. 1981. Mortality study of male and fasale workers exposed to PCBs. Paper presented st the International Symposium on Prevention of Occupational Cancer, Balslnkl, Finland, April 21-24. Blekerton II, L. Goldberg, ed. 1973. Spontaneous tumors and related pathologic changes. In Carcinogenesis Testing of Chemicals, Cleveland, Ohio: CRC Press, p. 55. Brown DP, Jones M. 1981. Mortality and Industrial hyglena study of workers exposed to polychlorinated biphenyls. Ardh. Env. Health. (36):120, 331 HONS 026372 Brow JF, Coe JT, Poeock, Jr. HD. 1981. General Electric publication. Burae W, Kimbrough RD, Vlllaneuve EC, Jennings RW, Linder RE. Sovocobl CW. 1974. Polychlorinated biphenyle: atorage, diatribution, excretion and recovery: liver morphology after prolonged ingeation. Arch, Environ. Health 29:301. Calandra JC. 1976. Summary of toxicological atudlet on commercial PCBa. In Proceedinga of the National Conference on Polychlorinated Biphenyle. Chicago, IL. USEPA Report 360/6-75-004, p. 35. Date PK, Gothoakar SV, Bhlde SV, 1976. Effeeta of partial hepatactomy on tuaor incidence and metabolism of alee fed ehloacataalde. J. Natl. Cancer Inatlt. 56:493. Davie GJ, McLachlan JA, Lucler GU. 1979. Waltrlng ayndrome In alee follovlng prenatal expoaure to 3,4,3',3'-tatrachloroblphenyl. Teratology 20:79. Dlkahlth TSS, Rockvood V, Abrahaa R, Coulaton ?. 1975. Effect of polychlorinated biphenyl (Aroelor 1254) on rat teatla. Exp. Mol. Path. 22:376. Doll R. 1980. Relevance of epldealology to pollelea for the prevention of cancer. Aaarlcan Industrial Hygiene Association, San Francisco, CA. Doull J, Klaaseeo CD, Amdur MO. 1980. Casaratt's and Doull'a Toxicology: The baalc science of poisons New York: MacMillan Publishing Co. Drill VA, Frelss SL, Hays KV, Looals TA, Shaffer CB. 1982. Potential health effects in the huiaan from exposure to polychlorinated biphenyls (PCBs) and related lapurltlea. Report submitted to'(he'Edison Electric Institute. Earl FL, Couvilllon JL, Van Loon EJ. 1974. The reproductive effects of PCS 1254 In beagle dogs and alnlature svine. Toxicol. Appl. Pharmacol. 29:104. Environmental Protection Agency (EPA). 1980. Ambient water quality criteria for polychlorinated biphenyle. CSEPA Report 440/5-80D68. Flschbeln A, Wolff MS, Lllls R, Thornton J, Sellkoff I. 1979. findings among PCB exposed capacitor manufacturing workers. Ann. N.Y. Acad. Scl. 320:703. Clinical Flshbelm i. 1974. Toxicity of chlorinated biphenyls. Ann. Rev. Pharmacol. 14:139. Gaffey W. 1981. The epidemiology of PCBa. Monsanto publication. Cellert R>, VUson C. 1979-. Reproductive, function in rata exposed prenatally to pesticides and polychlorinated biphenyls (PCBs). Environ. Res. 18:437. Green S, Carr JV, Palmer RA, Oswald EJ. 1975. Lack of cytoganlc effects in bone marrow and spermatagonial cells in rats treated with polychlorinated biphenyls (Aroclors 1242 and 1254). Bull. Environ. Contam. Toxicol. 13:14. 332 HONS 026373 Green S, Seuro FH, Friedman L. 1975. lack of dominant lethality In rata treatad with polychlorinated biphenyls, Aroclora 1292 and 1254. Fd. Coamet. Toxicol. 13:507. Hanaen 1C, Byerly CS, Hetcalf RL, Bevlle RF. 1975. Effect of polychlorinated biphenyl mixture on twine reproduction and tissue residues. Am. J. Vet. Res. 36:23. Hlfuchl K. 1976. PCB Poisoning and Pollution, New York: Academic Press. Heddle JA, and Bruce WR. H.H. Blstt et si., ed. 1977. Comparison of tests for mutagenicity or carcinogenicity ualng aseaya for sperm abnormalities, formation of mlcronuclel, and mutation In salmonella. Origins of Buman Cancer. Cold Spring Harbor, KY: Cold Spring Harbor Laboratory. Vol. &, p. 15*9. Hooplngarner R, Samuel A, Krause 0. 1972. Polychlorinated biphenyl Interactions vlth tltaua culture cells. Environ. Health Perapect. 1:155. IARC. International Agency for Research on Canear. 1978. IARC monographs on the evaluation of the carcinogenic risk of chemicals to humane: polychlorinated blphenyli. World Health Organisation Publications. Ito K, Nagasaki H, Maklura S, Aral M. 1976. Hlstopathologlcal studlea on liver tumorlgenesls In rats treated with polychlorinated biphenyls. Csnn 65:5*5. James RC, Crammer HP, Harbison RO. 1981. Heslth effects of PCBs: summary and technical review. Report submitted to Chemical Manufacturers Association. Jelinek C, Cornellussen PE. 1975. levels of PCBs In the U.5. food supply. National Conference on Polychlorinated Biphenyls, Chicago, IL, p. 147. Kashlmoto T, Mlyata H, Kunlta S, Tung T, Hsu ST, Chang KJ, Tang ST, Ohl G, Nakagawa J, Yamamoto SX. 1981. Role of polychlorinated dlbenxofuran, Yuaho (PCB poisoning). Arch. Environ. Health, Vol. 36, p. 321. Kepllnger ML, Fsncher OE, Calandrs JC. 1972. Toxicologic studies with polychlorinated biphenyls. Toxicol. Appl. Pharmacol. 19:402, Kepllnger ML, Fancher OE, Calandrs JC. 1971. Toxicological studies vlth polychlorinated biphenyls. Paper presented at PCB Conference, Quell Roott Conference Center, Rougemont, NC. Kimbrough RO. 1974. The toxicity of polychlorinated polycyclic compounds sod related chemicals. CRC Crlt. Rsv. in Toxicol. 2:445. Kimbrough RO, llndst RE. 1974. Induction of sdenoflbrosls end hepatoma* of the liver In Belb/cd mice by polychlorinated biphenyl (Aroclor 1254), J. Met. Cancer Inst. 53:547. 333 MONS 026374 Kimbrough RD, Squirt RA, Linder RE, Standberg JD, Nontali RJ, Bursa W, 1975. Induction of livtr tumor* In Sherman (train female rata by polychlorinated blphanyl (Aroclor 1260). J. Nat. Cancar Inac. 55:1653. Kuratsune M. 1976. Some of tha racant find Inga concerning Tutho, National Conference on polychlorinated blphenyle. 0SE7A Report 560/6-75-006. p. 16. Kuratauna K. Yoahlaura T, Hatauaaha J. Yamaguchl A. 1972. Epldeolologlc atudy on Yuaho, a poleonlng caused by lngaetlon of rice oil contaalnated with a commercial brand of polychlorinated biphenyls. Environ. Health Farapact. 1:119. Kuratauna M, Kimbrough R, ad. 1960. Yuaho. In Balogenated blphanyla, triphenyl*, naphthaltnea, dlbenaodloxlna and related products. New York: Eaevlar/North Holland, p. 287. . ... levlnskse C. 1981. A revlav and evaluation of carcinogenicity atudlea In lea and rats and mutagenicity studies with polychlorinated biphenyls. Monsanto publication. Under HE, Calnas TB, Kimbrough RO. 1976. Tha offset of polychlorinated biphenyls on rat reproduction, fd. Coamat. Toxicol. 12:63. MacLeod RE. 1981. Polychlorinated biphenyls In Indoor air. Environ. Science Tech. 15:926. Marks TA, Klamcl CL, Staples RE. 1961. Influence of symmetrical polychlor inated biphenyl Isomers on embryo and fetal development of mice, 1. teratogenlcltv of 3,3',6,4',5.S'-haxacloroblphcnyl. Toxicol. Appl. Pharmacol. 61:269.' Maroni M, Colombl A, Cantonl S, fertoll E, Foe V. 1981. Occupational exposure to polychlorinated biphenyls In electrical workers, I. Environmental and blood polychlorinated biphenyls concentrations. Brit. J. Ind. Med. 38:69. Maroni M, Colombl A, Cantonl S, Farloll E, Foe V. 1981. Occupational exposure to polychlorinated biphenyls In electrical workers, II. Health Effects. Brit. J. Ind. Med. 38:53. Martin PJ, Martin JV, Goldberg DM. 1975. Cama-glutamyl transpeptidase, triglycerides and ensyme Induction. Brit. Med. J. 1:17. McMahon RE, Cline JC, Thompson CZ. 1979. Essay of 855 test chemicals In ten tester strains using a new modification of the Ames test'for bacterial mutagens. Cancer Res. 39:682. Nagayama J, Kuratauna M, Masuda Y. 1976. Determination of chlorinated -dlbenzofurana In kanachlors apd Yusho oil. Bull. Environ. Contam. Toxicol. 15:9. National Cancer Institute. 1978. Bloassay of Aroclor 1254 for possible carcinogenlcltv. Carcinogenesis Testing Program Division of Cancer Causa and Prevention. DHEV Pub. No. (NIH) 78. 336 HONS 026375 NIOSH. National Institute for Occupational Safety and Health. 1977. Criteria for recommended atandard. occupational expoaure to polychlorinated blphenyla (PCBa). OHEW Publ. No. #77. Office of Technology Aaat assent. 1981. Aaacaaaent of technologies for determining cencer rleke from the environment. Orberg J. 1978. Effect of pure chloroblphenyls on the reproductive capacity in female mica. Acta. Pharma. Toxicol. 42:323. Ouv HE, Simpson GR, Slyall DS. 1976. Use and Health effects of Aroclor 1242, a polychlorinated biphenyl. In the electrical Industry. Arch. Environ. Health 31:189. Paakall D8. 197S. PCBa and their environmental effects. CXC Crlt. Rev. Environ. Cont. 3:469. Paralno C, Fry RJM, Christopher JP. 1977. Enhancing effects of pbsnobsrbltone and butylatad hydroxytolueoe on 2-eeetylemlnofluorene Induced hepatic tumorlganeals In the rat. Fd. Cosmet. Toxicol. 13:93. Paralno C, Fry RJM, Staffaldt E. 1973. Enhancement of spontaneous hepatic tumerlgtnaslc In C.B mice by dietary phenobarbltal. J. Neel. Cancer Inst. 51:1349. 3 Safe Drinking Veter Committee. 1980. Drinking Veter and Health, Volume 3. Washington, DC: National Academy Press. Safe S. 1980. Ambient vater quality criteria for polychlorinated biphenyls. I'SEPA Report 440/5-80-068. Schoeny RS, Smith CC, Loper JC. 1979. Non-mutagenlclty for salmonella of the chlorinated hydrocarbons Aroclor 1254, 1,2,4-trlchlorobenzens. mlrex, and kepona. Mutation Res. 68:125. " Schumann AM, Quest JF, Vatanabc PC. 1980. The pharmacokinetics and macrobolecular interactions of perehloroethylene In mice and rats as related to oncogenicity. Toxicol. Appl. Pharmacol. 55:207. Squire R. 1981. Ranking animal carcinogens: a proposed regulatory approach. Science 214:877. Squire RA, Levitt MB. 1975. Classification of specific hepatocellular laslons. Cancer Rms. 35:3214. Tllson HA, Davis GJ, MeLachlsn JA, Lucler CV. 1979. The effects of polychlorinated biphenyls given prenacally on the neurobehavloral development of nice. Environ. Res. 18:466. Toeruek P. U>73., Effect nf PCB on developing mouse. Chemosphere 1:173. Torok P. 1976. Delayed pregnancy In NHRI mice created vith PCB 2,2'-dlchloroblphenyl, Bull. Environ. Contam. Toxicol. 16:33. 335 MOMS 026376 Urabe H, Had* H, Aaahl M. 1979. Present state of Tuaho patlanta. Ann. N.T. Acad. Scl. 320:273-276. Vllleneuve DC. Grant Dt, Khjera K. Clegg DJ. Baar H. Phillipe WJ. 1971. Th fetotoxlclty of a polychlorinated biphenyl mixture (Aroclor 1254) in tha rabbit and In tha rat. Environ. Phyalol. 1:67. Voa JC, Koeaan JR, Van dar Kaaa KL, Tan Hoavar da Brauv MC, da Vo* RA. 1970. Idantiflcatlon and toxicological evaluation of chlorinated dlbcntofuran and chlorinated naphthalene In tvo comerclal polychlorinated blphanyle. Ed. Coenet. Toxicol. 8:625. Varahav R, Fltchbeln A, Thornton J. Miller A, Sallkoff 1. 1979. Deereaaa in vital capacity In PCB-exposed workera In a capacitor aanufacturlng facility. Ann. N.T. Acad. Scl. 320:277. Hatanabe S, Sugahara T. 1981. Experlaental formation of cleft palate In alee with polychlorinated blphenyla. Toxicology 19:49. Welaburger JB, Vllllaae GM. 1981. Carcinogen teatlngi current problem* and new approachea. Science 214:401. Whitfield JB, Pounder RE, Heale G, Moa* DU. 1972. Scrua game-glutamyl tranapeptldaae activity In liver dlaaaaa. Cut. 13:702. Wyndhaa C, Davenlah J, Safa S. 1976. The In vitro aetabollia, aacroaolacular binding and bacterial autagenlcity of 4-chloroblphenyl, a nodal PCB substrate. Re* Com. Chen. Path. Pharmacol. 15:563. 336 HONS 026377 Raymond D. Harblson DISCUSSION SUMMARY Division of Toxicology University of Arkansas for Medical Sciences Llctla Rock, AN 72205 The eooaldaratlona of the human rlak aaaoclatad vlth exposure to poly chlorinated biphenyl* (PCBs) art varied and require a multidisciplinary evaluation. Rlak assessment demand a a rigorous evaluation of experimental animal data aa veil a* human txpoaurc data, to accompllah thla talk, all significant publlahed or readily available atudle* of the effect* of PCBa ver* reviewed. In addition to thla review, current thcorle* of carclno|*n**la and toxicity testing end data evaluation were uaed In evaluating and anthropomorphlalni the atudy reeulti. aa a reault of thla review and evaluation, the question wa* aakad, "Doe* any axpoaur* to PCBa repreaent a significant health rlak to humans?" There was no unanimous agreement aaong the participant* of the symposium a* to the anewer to thla question. The human rlaka aaaoclatad with exposure to PCBa may be considered In two categories, those that are reversible and those that nay be irreversible. The dleeuaalon began vlth a question from Me. Julia Belli (Cincinnati Cas and Electric) about the reversibility of three effects: chlorscne, enayme Induction and low birth velghts aaong offspring. In a discussion between Ms. Belli and Drs. Seymour Pries* (Drill, Pries*, Rays, Loomis & Schaffer, Inc.), James and Berblson, It was brought out that the usual human observable conse quence of exposure to excess concentrations of PCBs Is a dermatologic condition. The dermatologic condition sometimes produced Is chloracn* vhleh Is reversible and disappears on discontinued exposure. Thla condition la of an acuta nature and ehronlc exposures have added little or no additional adverse effects of consequence to this picture. Similarly, the capacity of PCBs to Increase liver blotransformlng entymes Is not unique to this class of compounds. Although mostly animal studies have been used to determine the effects of PCBs on liver blotransformlng enzymes, the effect has also been demonstrated In humans. However, increasing blotransformlng capacity of the liver enzymes can have beneficial effaces as well as possibly adverse effect*. Enzyme induction cannot slmpllstlcslly b* designated a toxic efface. Further, this effect Is reversible based upon current knowledge of the blotransformlng enzyme systems. Thus, both the PCB-inducsd chlorscne and possible enzyme Induction In human* are reversible biological responses. The Implication of any association of causation of low birth weight lo human offspring by exposure to PCBa Is not supportable at this time. The data presented at this meeting about low birth weight offspring and exposure to PCBs la premature and has not been subjected to a critical revlev. The low birth weight of some of the offspring exposed to PCBs and other contaminants at Yusho was revefslbl*. Therefore, there are no significant new leads ss to the potential for PCBs to produce low birth weight offspring sc this time, and 337 MONS 026378 the limited dti avtllabl* show the effect, if it occur*, Is transient end doee not occur In all offaprlng. The controversial but critical issue of assessing the potential risk of developing center as s result of FOB exposure, based on studies in experimental animals vs* addressed. Ms. Jacqueline Warren, of the National Resources Defense Council, expressed concern about two Issues: (1) the paper presented by Dr. James generally appeared to reviev studies that produced positive results more critically than it did similar studies that had negative results and (2) specifically, that a recent National Institute of Occupational Safety and Health (NIOSH) study shoved that one group of individuals exposed to FCBs developed a nonatatistieally significant three-fold excess of rectal cancer. Ns. Warren supported her first point by referring to en Environmental Protection Agency review of the study* on which the presentation by Dr. James vas based. This assessment, she said, stated that the study's reviev of the design of the various experiments tended to emphasize and exaggerata the negative affects, while discounting positive effects. Drs. James end Harbison responded by stating that numerous studies uslnf both laboratory nice and rats vara considered. Considering the number of studies which have indicated FCBs cause cancer end comparing this number to the number of etudlee which failed to demonstrate FCBs cause cancer, the conelution may be made based on thlt comparison that axposura to FCBs does not rtprsssnt t significant risk for increased cancer in humans. Further, after more chan JO yeari of FCBs use in industry, chars la no apparent causal relationship betveen any typa of cancer and axposura eo FCBs, and thara is no pparent increase in the incidence of esneer mortality among worker* occupa tionally axpoaad to high lev*la of FCBs. The largest study conducted by NIOSH, which involved over 2.S00 persona, did not detect any statistically significant axeass in the cancer mortality. Sine* this study failed to demonstrate an exceaa cancer rata in a highexposure population. It provide* soma reassurance that it la unlikely that future studies will ahov any increased risk of cancer from FCBs. The concern for the effects of FCBt on reproduction vet expressed by Ms. Warren, She supported that concern by the fact that Allen and eo-workarst had shown obearved adverse reproductive affects aftar exposures to concentrations of FCBs equaling 2.S and S.O ppm. She further stated that tha concentration of FCBs In human breast milk la In that concentration range or higher. *Ecology A Environment, Ine. 1981. Summery of the health effects of FCBs. lyffalo, NT> Ecology 4 Environment, Inc. Chemical Manufacturart Association contract no. FCB-3,0. `Barsottl, DA, KarlaT RJ, Allen JB. 1976. Reproductive dysfunction In rhesus monkeys exposed to low levels of polychlorinated biphenyls (Aroclor 1248). Fd. Cosmet. Toxicol. 14:99. 338 MONS 026379 Dr. James stated that the experimental enlmel studies which vert reviewed consistently demonstrated a lack of any teratogenic effect, and fetotoxlcity was observed only when maternally toxic dosea of PCBs were administered. The experimental animal teat results are consistent with the human experience. The Yusho Incident, a high PC8 exposure, failed to Identify any substantive concern for teratogenic effects. Considering the balance of the reproduction studies and the breadth of the mutagenic testa performed, PCBs do not appear to represent a human reproduction haxard. The overwhelming body of knowledge about PCBs suggests to Dra. James and Harbiton that there are tome dermatologic effeeta produced by PCBs, but these effects are reversible, and beyond them there are no clear Indicators of any other adverse health effects aaaoclatad with exposure to PCBs. The human exposure data and the experimental animal data are reassuring In that there have been no significant adverse health effects aaaoclatad with exposure to PCBe. Therefore, If there ere risks associated with exposure to PCBt, they ere minimal and unobservable. Epidemiological etudlte have shown that eoatmerclal PCBe are not remarkably toxic chemicals after acuta exposure and chat whan axcaaa exposure dots occur, the usual consequences are dermatologic and not of a serious or permanent nature. Chronic exposures have added little or no additional advaraa effeeta of note to this picture. The preponderance of atudlaa has not ldantlflad a clinical dlsaasa aaaoclatad with exposure to PCBs, nor has It provided pertuaalve evidence of health Impairment. Thera la no avldanca of an axcaaa In total mortality or mortality due to cancar, cardiovascular disease or nervous system disease associated with occupational expoaura to PCBs. PCBs have not baen linked to any human cancer, and studies to date indicate It la unlikely that future studies vlll establish such s link. It saems to this author that there was no sgreemtnt during tho risk assessment discussion as to the consequences of human exposure to PCBe. In feet, Me. Warren stated that, in her eplnloe, no meaningful conssnsus could be reached in the ebeence of more of the scientists who had performed the studies discussed sc chs symposium, but who wart unabla to attend It bacausa of Its tchedole. 339 HONS 026330 0. Huttingar Unlvanlty of Beyrouth Bayreuth, Germany CONCLUSIONS INTRODUCTION Thle lection li i luaairy of whet wee preeented end dlacueeed during the two daye of the PCS* Sympoalum. It le not Intended to be, end cennot be, e ont-nen riik attainment of the PCBe problea. Although I heve tried to preeent information in an accurate end unbiased aannar, ery profaeelonal background end personal blae will undoubtedly flavor these pages. The llaltatlon of such a summary will be further evident by realising that I wrote it during one day following the nesting, without having had the chance to see the papers beforehand. I did get advice froa speaker* and discussion leaders in areas outside ay own expertise, particularly toxi cology, but otherwise had no chance to discuss topics with the authors. Analytical Methodology In recent years It becaae evident that toxicological properties of individual coaponents In the PCBe alxtures show considerable differences. It is therefore necessary to use analytical techniques which allow detection and quantification of, discrete chloroblphenyls. High resolution capillary gas ehronotography (GC), after appropriate cleanup, la the method of choice to fulfill this need. Capillary CC Is no longer a research tool, It has now developed Into a viable technique that can be used routinely. Thare Is no apparent strong reason to analyse PCBs In environmental samples by packed column methods. In addition to isomer separation, capillary columns with their much higher resolving power also eliminate many Interfering compounds. Quantification of low chlorine ?CB, "early elutara," Is difficult if not impossible with ocher techniques. This problem Is also solved by the use of caplllaty columns. For proper Identification of all compounds In PCBs mixturss and environ mental eamplet, synthetic etandarda of all Z09 chloroblphenyls arc being synthaalsad, and fully characterized PCBa nlxturat are expected to be avail able aa standard! In the near future. Several round robin PCBa analyeaa are being earrled out or are planned. One auch comparative analysis schema will use the characterized etandarda. Por tome ilnple analyst* (a.g.. transformer fluids, whsra the chloroblphenyl pattern haa not bean changed by environmental or matabolic Influence*) utt of ilmple peered column* 1* *uffie lent, If proper standard* art being used. Cat ehromatography-aata apactromatry 1* (till the preferred aethod when Identification aspect* are Involved. In addition to th* tlaa-taatad electron 340 HONS 026381 lnpsct ionization node, negative ion chemical ionization it now available for PCBa analyaia. Thar* it atlll no standard nathod for quantification of PCBs. Coaparlton of tavaral peaks in tha aaapla vlth a matched PCBa ttandard la commonly utad, but coaparlton of fully charactarlaad paakt ldaally thou Id ba utad. Preferred dattetion net hodt teploy tha alactron capture datactor or aalactad ion aonitoring aatt spectrometry. Since dlffarant chloroblphenyIt vary in their rttponae, particularly in tha electron capture detector, work vlth authentic tttndardt it etaentlal. Tha coaplexitlts of envlronaentai eamplaa haa atlaulated the application of pattern reco|nltion tachniquat (SIMCA) to PCBa analyaia. Thit aethod attabllahea similarities and dlfferancaa betvaen aaapla raaldue profllat. Diatrlbutlon and Pate Tha raducad raleaie of PCBa during tha last tan yaara haa raaultad In lover PCBa eoncentratlone In tha envlronaent. This can ba aaan by comparing tha acaauraaant of PCBa in aevetal or|tnitat sada during this tlaa period. Tha reduction of PCBa content In aquatic organleae la mainly due to burial of PCBa In nonaoblle lake sediments and, to a laeaar degree, degradation. Of tha total U.S. production of PCBa since 1930, about 15S of tha PCBa are believed to have entarad tha aoblle envlronaent, with tha northern Atlantic Ocean being tha uln sink. An aatlaatad 1.7 to B.6Z of tha envlron aentai PCBa la in freahvatar aediaenta, asking this the second aoct laportane PCBa sink. Conversely, tha astlaaead aaouot of PCBa In freshwater Itself Is only 0.01-0.04Z of tha total. Thera la a fairly straightforward correlation between degree of chlorina tion and blodegradablllty: chloroblphenyla with a lov degree of chlorination era biodegradable, and aore highly chlorinated biphenyls are recalcitrant. In one study, aonochlorlnstad biphenyls ware shown to degrade vlth a half-life of days, whereas 2,2',4,4'-tatraehloroblphenyl showed no aaaeurabla degradation. In lake aediaenta half-lives ranged froa a few days for dlehloroblphanyls to 100 days for pentaehloro-lsoaera. Hatheaatlesl fate aodals now appear eo accurately predict tha fata, dis tribution and envlronaentai residues of chloroblphenyIs. Chloroblphenyls Incidentally Potaed In Process Screens Although aueh Information and know-how exists for the analysis for cosnterelal PCBa preparations and envlronaentai soap lea containing such residues, very little Is known on procedures for analysing Incidentally ganerated chloroblphenyls. In tha latter ease, the relative concentration of ehloroblphenyl isoaers la normally related to the Individual cbealcal processes forming ehea. The ehloroblphenyl coneent can range froa one to several Indlvldiipl component*^. In virtually, all cases, the pattern of laoaer and congener distribution is different froa that of the PCBs produced by chlorination of biphenyl. 341 HONS 026382 The absence of a characteristic fixed dlatrlbutlon of chloroblphenyl* In one pTocete streams Increase* analytical probleas since, theoretically, all 209 Isoaers and congeners could be present. The total generation of incidental chloroblphenyl* In the O.S., at concentrations belov SO ppa, Is astlaated to be 13.800 pounds, of which 95X la disposed of a* controlled waste. Little Is known on the actual exposure and the associated hazard fron chloroblphenyl* which are formed In process streaas and released at low concentrations. Studies on bloavallablllty are apparently lacking. This Information Is needed in eases such as the phthalocyanlne pigments where chloroblphenyl* seen strongly bound by the matrix. Bloaccuaulatlon and Pharmacokinetics PCts are highly lipophilic compound* that partition Into fat and lipidrich tissues In both fish and mammuls. Depending on the degree and position of chlorination, different chloroblphenyl* show different llpophllieltlee and, thus, different rate* of uptake, distribution and excretion. Uptake of PCBs by fish in their natural environment vas shown to be correlated to lipid content In some studies. However, It appears that the concentration of PCts In organisms Is not simply and always related to lipid content, although the studies demonstrating this oftan suffer from nonstanderdlted reporting methodologies. PCBs concentrstlons are reported either on s vetvelght basis or on a lipid basis. Standardisation Is urgently needed. Blomagnlflcatlon (Increased concentration In fish of higher trophic levels) Is not Inportent for PCts, but bloeoncentratlon (directly uptaka fron water) Is fast and efficient. Female relnbov trout and mice were shown to posses* alternative mechanisms of PCBs excretion during altered physiological states related to reproduction. Following water exposure of female relnbow trout to 2,2',5,3'-t*trschloroblphenyl, the half-life of elimination was 1,73 years. Upon Initiation of spavulng, the half-life of elimination for the same compound decreased to approximately 0.5 years. The half-life of elimination of 2,2',4,1*,5,3'-h*xtehloroblph*nyl fron female nice Is greater than 100 days. During lactation, however, the half-life from maternal adipose tissue decreased to two days. Vlthln two to five days the mothers had eliminated essentially their entire body burden. The hexsehloroblphenyl, In turn, vas stored quantitatively In the offspring. Because of the significantly lower fat content of human milk end comparatively much smaller milk quantities, this mechanism Is likely to be much less effective In humans. Ho significant amount of hexachloroblphenyl was transferred across the placenta during pragneney In the mouse experiment. From analysis of human milk samples and from samples of Tusho patients, it Is apparent that tha most planar chloroblphenyl* (1.*., those having the least orthochloro substitutions) are tha most persistent In the human body. From receptor binding studies It Is known that these are also likely to be the most toxic isomers. 342 MQNS 026303 Effct on Organism* In the Environment There ha* been considerable growth In studies on environmental affects of PCBs In rscsnt years, particularly from a qualitative point of view. Reporta of PCBs residua concantratlont In Halted populatlona froa aaall areas have been replaced by studlea on aevaral eoaponents of the ecosyetea over larger geographic areas. Acute toxicity studies have given way In many cases to studies of chronic toxicity and to Investigations Into the aechanlams of toxic action. Bloassay data on PCBs are still lacking for large groups of organisms. The use of standard species such as fathead alnnova and rainbow trout facilitate comparison between toxic substances, but give little Indication of the affect of PCBs on other organisms. Oats are particularly needed for large mammals and for various trophic levels of birds. Studies on the effects of PCBs on progeny should also be conducted. The majority of the biota In the D.S. la subject to a relatively small base-load contamination from PCBs, Some organisms, dua to their habits, will contain substantially elevated burdens. Small amounts of PCBs ean inhibit production of phytoplankton and could thus have significant ecological offsets. It Is probable that some sensitive mamaisls, such as mink, are being adversely affected by exposure to PCBs. These effects would primarily be reproduction-related, hut some acute lethal effects are also likely. In many cases, reproduction of fish In D.S. waters la probably not adversely affected by PCBs exposure, since the more sensitive species such as salmonolds have frequently been eliminated by other human actions. Host bird populations In the D.S. are not expected to display acute affects from PCBs In the environment. Many specie* of fish-eating birds, hovever, are exposed to dietary level* of PCBs sufficient to cause metabolic effect* In their offspring. In the pest few years, tha use of Indicator organisms has become refined. It nav be possible, in the near future, to establish PCBa standards based on concentrations found In those species which most readily accumulate PCBs. Human Health Effects The seut* toxicity of PCBs mixtures Is moderately low. Dasplta intense effort* during the last decade, chronic toxicity data for PCBa arc still Inconclusive. No significant new developments In animal tasting were reported during the last faw years. In animal studies, great variations are observed in toxicological responaee to PCBs and other related eompounda. The effects depend on the degree of chlorination In the chleroblpbenyl, the Isomer used end the species and sex of test animals. Mink, guinea pig, monkey, quail and thick ambry*-are among the more sensitive PCBs test organisms, vhlle the rat and mouse are among the less sensitive. Only a few characteristic responses ean consistently be observed In animal testing. These are chloraen* and ensyme Induction. Furthermore, these 343 MONS 026384 responses can also b observed In worker* who ire occupationally exposed to PCBs. Studies using anlnal aodsls and other test system have resulted In reported effects which include: (1) adverse effects on the reproductive process (teratogenicity or fetotoxlclty In animals), (2) adverse effects on liver tissue In anlaals with dose related Intensity and reversibility, (3) son* findings on liver cancer In rats and elce, but otherwise essentially negative results on carcinogenicity, (4) negative findings on autagenesls, (3) observation of gastric lesions In the monkey after administration of Aroclor 1242, (6) aone offsets on lmmuno-comp*t*nce In anlaals at dosages that cause changes In nutritional status, and (7) Induction of porphyria In the rat and rabbit. It Is clear that aore eaphasla smst be placed on the study of Individual Isomers, Already, structure-activity relationships In the chloroblphenyl series, related to the Induction of hepatic silcrosoml ensyae and binding to hepatic cytosol receptors, have yielded significant Information. In the polychlorodlbento-p-dloxln series, these paraaaters correlate well with toxic effects, and similar, If weaker, response* are observed,with the planar chloroblphenyl* (l.e., those having the laatt orthochloro-substltuents on the one hand, and a 3,3',4,4'-chloro-substltutlon pattern on the other). It was pointed out that entyme Induction cannot be considered a toxic effect and emphasis should be on meaningful biological end-points. However) apart from being useful In structure-activity studies, ensyme Induction could be Investigated as a prs-toxle effect marker, since early indicators are esaantlal to eoamunlty health care. An Internationally approved experimnt for eestlng the carcinogenicity potential of PCBs vas suggested. A good study design Including both sexes, acceptable species (both rodent and nonrodent) and a carefully vorked out protocol could yield significant results. It was also suggested that malignancy and various pra-cancerous lesions should be the indicator of carcinogenicity. Several cross-sectional epidemiological studies on the health effects of PCBs, and three cohort.mortality studies primarily concerned with the carcino genicity of PCBs, have been carried out since 1976. Evidence from the cross-sectional studies show that occupational exposure was associated with dermatitis and mild liver function abnormalities In the absence of clinical dlsea'se. These symptoms were not observed In nonoccupatlonsl exposures. The three mortality studies all disagreed on the cancers for which marked excesses were found. The toxic effects observed In Tusho patients Is believed to be due to mixed affects from PCBs, polychlorinated dlbenzofurane and poly chlorinated quaterphenyls. This case should, therefore, not be Included In evaluations of exposure effects of PCBs. Bisk Assessment Sessions on risk assessment are frequently the most difficult and least conclusive In this type of meeting, and the PCBs Symposium was no exception, 344 MQNS 026385 In the Chairman's opinion, no meaningful and scientifically sound discussion took place on this topic during the Symposium. The cost worrisome aspect of the PCBs problem Is the ubiquitous occurrence of this group of compounds In biological material. This la related to transport behavior, lack of easy biodegradation and the bloaccumulatlon potential of PCBe (thus, to physical-chemical properties tether then toslc effects). Still, the nagging feeling remains that, perhaps, afear 20 to 30 years some ae yet unknown or unstudied effect In men or other organisms may become apparent. On the other hand. It la evident that after 30 years of Industrial produc tion and use of PCBe, no significant clinical disease or clear epidemiological correlations have been observed In occupationally exposed persons. Consider ing the lover and declining exposure to PCBs via the environment. It seems unlikely that significant Injury will be observed from environmental exposure to PCBs. 345 HONS 026386 ATTENDEES Gil Addle Electric Povtr Reaearch Inatltute Jane Berglar Pacific Gaa 4 Electric Co. Karla Ahaed Natural kaaourcaa Defence Council Cheryl A. Blua Niagara Mohawk Power Cory. Toni K. Allan Wald, Uarkrader 4 Roee . Michael Bolger Food 4 Drug Adnlnatratlon Alvlto P. Alvaraa Dnlforaed Scrvlcaa Onlveralty Edward M. Braahler Aaerican Envlronaental Protection Cory. David C. Anderaon O.S, Envlronaantal Protection Agency Joaeph Breen O.S. Envlronaental Protection Agency Ruth Arlaaao Gtnaral Elaetrlc Co. Ronald Brooke General Electric Co. Karaha L. Barlow Dnlvaralty of Maryland Gloria Brooka-Ray CPC International, Inc. V. J, Barrow, Jr. Florida Power 4 Light Co. Melinda Bartholoaae Envlronaental Control Technology Cory. Frank S. Beal Weatlnghouae Electric Ccrp. Patricia J. Beattla Genaral Motora Cory. Robert Bell General Electric Co. Julia Lynn Belli Cincinnati Gaa and Electric Co. Craig Benedict O.S. Envlronaental Protection Agency John P. Brown, Jr. General Electric Reaearch 4 Develoyaent Jaaaa G. Bryant Standard Chlorine Edvard Burger Georgetown Oniverelty Kenneth L. Burgeee Dow Chealcal Coayany Laura Caaybell O.S. Envlronaental Protection Agency Ann E. Carey O.S. Envlronaental Protection Agency Jaaea C. Carnahan General Electric Reaearch 4 Develoyaent 346 MONS 026387 Nell Carrlker Tenneeee* Velley Authority Those* R. Catty Central Electric Co. Hortt Cetpera Stauffer Chtslcal Co. David Cetarao Bell Laboratorlta Kenneth H. Chaae Washington Occupational Health Aatoc. Margaret M. 1. Chu O.S. Envlronsental Protection Agency Jaros* T. Co* Central Electric Co. A. T. Colasarla Electric Mutual tnauranee Co. J, Patrick Conlou Xennecott Refining Corp. Robert A. Copeland Dow Cheslcal Coapany ' John H. Craddock Monaanto Induttrlal Cheslcal Co. Roger Crawford Martin, Craig, Cheater A Sonnenacheln - J. L. Crowell R.E. Dptegraff Manufacturing Co. Elbert Dag* U.S. Envlronsental Protection Agency Jasee P. Daley Wcatlnghouee Electric Corp. M. DaRoeha Sun Cheslcal Corp. Joan Davie* John* Hopklna Onlveralty Jaaea H. Davla Ward, Harkrader 4 Boat John DelUney Envlronsental Optlslaatlon, Inc. Paul DeeRoalera O.S. Envlronsental Protection Agency Jaaea R. Donnelley General Electric Co. Carol Dorg* Martin, Craig, Cheater A Sonneracheln Elizabeth A. Doyle Asarlcan Onlveralty Roderick T. Dwyer Asarlcan Mining Congraaa K. V. Eaaly Monaanto Co. Jan* C. Elekhoff National Food Proceaaor* Aaaoe. Mitchell D. Erlckaon Mldveat Reaearch lnetltut* Those* T. Evan* O.S. Envlronsental Protection Agency Edward J. Feeder Southern California Edlaon Co. Jasea W. Faleo O.S. Envlronsental Protection Agency Bruca Feetherttone Kirkland A Elll* Robert J. Feneterhels Cheslcal Manufacturer* Aaaoclatlon 347 HONS 026388 Joann* Ferrara John* Hopkina University John l. Fata American Paper Institute Charlta Fovlar Vcraar. Inc. Robert S. Frlcdaan General Electric Co. Killian Geratentant Stauffer Chealcal Co. Chrletlna Atkina Godack Edison Electric Inatituta John E. Colng Midwest Raaaarch Inatituta Vinca Griffin Public Service, Indiana Helen* Heduch U.S. Envlrotuaantal Protection Agency Stephan E. Hamilton General Electric Co. Karen Haoneratroa U.S. Envlronaental Protection Agency J. Harley Florida Power t Light Co. Stephen Harvey Algonquin Gaa Trinamiaaion Co. Tidal B. Henry, Jr. Electric Mutual liability Inaurance Co. E. J. Hobba Dow Corning Corp. Edvard Holland ENSCO Donald E, Hutchinson Anerlean Electric Power Service Corp. Stephanie R. Iren* U.S, Envlronaental Protection Agency Anthony W. Jackson Canadian Electrical Assoc. Caaey Jason U.S. Envlronaental Protection Agency Charles F. Jellnek Food and Drug Adslnlatratlon Ralph V. Jcnnlnga U.S. Envlronaental Protection Agency Denise Keener U.S. Envlronaental Protection Ageney Ronald Keener Roha and Haas Co. Edward Kern Envlronaental Optimisation, Inc. J. J. King, Sr. Florida Power ( Light Co. A. K. Klrksey DEPCO John J. Kish Berne* A Thornburg Ralph Y. Koval Elactrlc power Research Institute John Kcalnar Canada Center for Inland Hater David Kraa U.S. Envlronaental Protection Agency Charles Kueera AMAX. Inc. 348 MQNS 026389 Richard~V. lavton General Clactrle Co. Roberc E. Lea Energy Syetens Co. C. C. Lee U.S. Envlronnental Protection Agency Anthony Lee Technical Resource* George J. levlnakae Monsanto Co. Muriel llppaen ERKACO, Inc. - Mary Logan Martin, Craig, Chestar 4 Sonnarseheln Robert <5. MacDonald National Institute of Occupational Safety and Health Robert J. Maher Electric Mutual Liability Inauranca Co. Leonard C. Meier, Jr. General Electric Co. H. C. Manger Baltimore Ca* l Electrle Co, Richard Marcus ASAXCO, Inc. Ira H. Mark* JRB Assoc. Catherine A. Marshall U.S. Brevet* Assoc., Inc. Tlckl Lynn Martin Dov Corning Co. Antonia Msttla University of Maryland Bruce Mean* U.S. Environments! Protection Agency Peter Metaxotos American University Any Moll U.S. Envlronnental Protection Agency David Moor* Snlth, Moore, Snlth, Schnell A Hunter Michael D. Mullln U.S. Envlronnental Protection Agency J. Vincent Mabhols U.S. Envlronnental Protection Agency Joseph G. Massif Monsanto Co. Larry Masdhsn Ctntsr for Dlseast Control Robart T. Moonsn Aatrack Lsursncs J. O'Neill Monesnto Co. Robert Orchovskl Duquatne Light Co. Donald E. Park Ethyl Corporation Sally Paterson University of Toronto Carl Pfsffsnbsrger Unlversiey of Miami School of Medicine Stephen R. Powers, Jr. Dept, of Water Power, City of Los Angelas Susan P, Read Central Electrle Co. 349 HONS 026390 Matthev J. Reilly Environmental Research 4 Technology Thome* Robinson Vulcan Chemical* Richard Ronan Verier, Inc. Anthony Rutter U.S. Environmental Protection Agency Jack Savlckl International Association of Fire Fighter* M. P. Scarbel General Electrle Co. Deborah Schmall U.S. Department of Juatlca Alexander Smith NIOSH/Clnclnnatl John H. Smith U.S. Environmental Protection Agency Lee N. Smith Steptoe and Johneon Susan Sonchlk Vartar, Inc. Lloyd A. Spraglna National Electrical Manufacturer* A*aoc. David Stalling Columbia Nat'l Flah Reeaareh D.S. Flah 4 Wildlife Service Elizabeth Stain U.S. Department of Juatlce Harold Stein Ceneral Electric Co. Sharrell A. Stealing U.S. Environmental Protection Agency John J, Thorpe Private Consultant Jaaon Toth U.S. Environmental Protactlon Agency W. Corey Trench Edison Electric Institute William J. Trotter U.S. Food 4 Drug Administration Vivian Turner U.S. Environmental Protection Agency David Victor White Pine Copper Compeny Michael Walker Commonwealth Edison Co. Joseph C. Warnock Central Hudeon Gas 4 Electric W. H. Weaver Reliance Electrle Ian Webber MB Corp. Robert Welch Columbia Cas Co. Andrle Zervina Westlnghouse Research 4 Development Center 350 MQNS 026391 INDEX acceptability of data 134 accumulation, In flah eggs 99 acuta animals toxicology 259, 260, analyst*, biological madia 4 analysis capillary columns 183 alactron eaptur* CC 164, 182 mast spaetromatry 163, 182 quality assurance 1(3 sample tits 279 Aroelor fluids 138, 139, 142, 143 Aroclor 1242 130, 2(6, 267, 269 Aroelor 1248 130. 268 Aroclor 1254 130, 250, 2(6, 269 Aroelor 1260 130, 249, 259, 268 bass largemouth 266 strlpad 266, 269 vhlta 274, 276 bat 268 bantkot 270 binary liquid phase optimisation 161 bloaeeumulatlon 97, 130, 133, 293 bioasssy 266, 267, 280, 283 bloconetntratlon 270 blodlstrlbuelon mammal 105 flah 108 blomagnlflcatlon 280, 281 blo-radlstrlbutlon 128 blostablllty, chlorination 72 blood chaalstry 195 blood prsssura 195 braast milk 130 cancer biliary 196 in humans 189, 190, 196 livsr 196 lung 196 lymphatic and hematopoietic 196 pancreas 196 rsctal 196 stomach 196 capillary eoluama 6, 8 capillary columns versus packed columns 161 capillary GC phases 7 carcinogenesis 215, 222 HONS 026392 carnivora 270, 276 carp 276, 276, 280 catfish, channtl 276, 276, 280 characttrlttlci of analytical athoda 166 ehsaoaetrlcs 65 chicken 268 chloracna 190 chlorination stability 72 cholestarol 190, 192, 195 chronic toxicity 265, 267 cola, colorad bablas 129 co Hagan 266 controlled vastas 177 coot 282 erappla, white 276, 276, 280 erev 282 DDE 268 deposition 279 derul toxielty 217, 222 danatleia 189, 190 dottetora 8 decritivore 270, 276, 280, 281 dlbentofurans 296, 295 0C/MS/SIM 11, 12, 13, 151, 155, 161, 163 distribution anvlronnental 69 huaan hoaologs 83 dove, aournlng 268 duck, aallird 268 eagle, vhlte-tall 283 etTthvon 282 effective concentration 267 effluent dlacharge 276 elder, comoo 268 electron capture detectors 166, 169, 150 electroconduetlvlty (Ball) detector 150 electro-fishing 270 electron destruction 131 envlronaental Induction 296 enzyaaclc effects 220 encynetlc Induction general lnforaatlon 236 MC type Inducers 235, 238 nonlnduears 240 ortho substitution 238 pars substitution 235 PCB type Inducers 234 pre-1978 189 estradiol 168 352 HONS 026393 expoeure 186 Potential to Incidental PC8 177 fathead alnnov 266 ferratt 269 : fiddler crab 267 flah, Great lakea 132 Flaher 283 Pood and Drug Administration 276 fox 283 flow-through teat 266 Oasoett 263 gas chromatography/mass speetrowetry 162, 166, ISO, 1S1 gaatrlc toxicity 219, 222 gill net 270 Great lake* contamination 83 grebe, veatern 268 gull Franklin'e 282 laughing 268 hatard 302 hexacbloroblphenyl 183 high raeolutlon CC 61 human homolog distribution 83 hydroxy latlM- pare subetleutloa 108 adjacent unaubatltuted carbon 112 Ible gloeey 268 white 268 Identification of eonganera 162 lllneee 195, 196 lanunosuppreselon 220, 223 Incidental generation 139 distribution 30 ppm 177 manufacture below 50 ppm 177 processing below 50 ppm 177 use below 30 ppm 177 Indicator organism 270, 276 Induction 293 Industrial Analytical Chemistry snd Regulatory Surveillance 166 Ingestion 269 Isolation of lnddantal 162 kepone 266 kllllflsh, Gulf 266 landwash 276 LCS0 266 limits of Detection (100) 166 Limits of Quantification (10Q) 166, 165 HONS 026394 lipid 267, 269. 279, 281 liver function 189, 192 toxicity 221, 222 locomotor activity 267 Marten 283 matrix lntarftranca 183 aergeneer, rad-braaated 282 aetabollem 108 aldgee 281 allk, tranafer 118 ink 269, 283. 286 lrax 266 aixed GC phaaaa 1(1. 163 aonltorlng 297 aonochloroblphenyl 166, 167, 1SS, 159, 160, 162 aonooxygenaee activity 292 aullctt 269, 282 autaganlelty 216, 305 National Foeticide Monitoring Program 279, 283 packed coluan veraui eaplllary column 163 pattern recognition 5, 66, 163, 155, 182, 185 FCDFa 189 PCOe 190 perchlorlnatlon technique* 169, 150 perch, yellov 269 planktlvoree 270, 276, 280, 281 plankton 265, 266, 286 polar capillary coluana 163 polychaetea 270 porphyria induction 220, 226 preparative GC 162 production 67 pulaonary function 195 quantitative analyela of chloroblphenyla 166 quality aaauranca 163 quail, Japaneae 267 raccoon 283 raxorblll 268 receptor binding 260 redietribution 128, 136 regulatory Unit 138, 165, 166 Regulatory Surveillance and Induatrlal Analytical Chealatry 166 reliability of ehealcal data 138 reproduction 295 HONS 026395 reproductive effecte 195, 212, 223 reproductive proceeeee 267, 268 reeiduea 295 reeplratloo 269 retentloo behavior of chloroblphenyla U2 rlek aaaeaenent calculatlona 326 rlak Identification 300 river earpeucker 273, 276, 278 round robin enalyala of Incidental PC* 163 accuracy, recovery atudiea 163, 168 nethoda, calibration 165 participation 163 aaaple type* 163 auaueary 163, 166-166, 167, 169, 170, 172-177 tkua, great 266 anaka 282 aolubillty 276 atandarda 269 prlaary and aecondary 37 atarling 282 ateady-atate calculatlona 99 tructura-actlvlty relatlonahlpa 187 uckar, Utah 282 aunflah, Redear 266 yaiptoaa 195, 196 taratoganaaia 213, 308 thraahold eoncaneratlen 267 toxaphane 266 toxicant 301 toxicity, general anlnal 221 aaaple athodology 270 preparation 155 alee 295 aeal 269, 282 elactad ion aonlcoring 151, 156 had toxicology - 185, 186 trlglyearldaa 190, 192, i94 trophic level 276 tumetone, ruddy 268 turtle green 283 loggerhead 283 Aaarlcan 282 flttard -*70t 278, 278 ahriap 2(6, 267 validating aiethod criteria 163 variability, loterlaboratory 63 - 355 HONS 026396 vhltcfith, Rocky Mountain 282 vlllct 268 Tuaho I8, 190 saro aolaculaa 172, 177 3S6 HONS 026397