Document e70jj7JOB75gpDBqkN4wJbezE

Y?iK-'' 'V;-' ' DEPARTMENT O F HEALTH & HUMAN S E R V IC I H& pyt Ape'll 26, 1985 from Chief, Superfund Implementation, Group rtiDiic naaun aarvic* Cantare for DiMat* Control Memorandum EXHIBIT 5 Subj$iReview of EPA, Region V, Evaluation of the Midland, Michigan Soil Data ' Midland, Michigan - . S 'y S S , j. ' "' To Louise A. Fabinski Public Health Advisor EPA Region V - - ' I#. The material submitted on the subject site has been reviewed by the Chronic Disease Division, Center for Environmental Healthy Centers tot Disease // /.V; Control. v I hope that y o u .find their comments'useful.i V1' ;; 'S S - j:/:' COMMBMIS :In followup to previous reviews *>f the environmental soil sampling results from Midland, Michigan, we were asked t o :review a draft Risk Assessment evaluation of these data which had. been prepared by EPA. The essential conclusion of this Risk Assessment is that, because of a history of ` occupational exposures to TCDD and the documented potential for exposure to contaminated fish from the Tittabawassee River, exposure should be limited to soil contaminated to 50 parts per trillion (ppt) of lower. This level, approximately l/20th of the C D C 's previous site~spacific recommendation, is generally consistent with levels seen in other industrialized areas of the country where the manufacturing of TCDD containing products is known not to occurred. .V;' ' S S S S " S - S :-',-.' mm.I9pS-sHS v.. ,Vg ^ /S4#, lim r ' ; ; ' f e , .3 0i-rVSS. m j . : L ;- V - . He find no major fault with the methodology used in preparing this risk assassment; the methodologies and approaches are similar to previous work done at CDC except that different assumptions have been used in several reas such as: a greater amount of dermal contact with soil, using only upper level confidence limits for the estimates of a virtually safe dose, (VSD), the lower average'lifetime body weight measure, etc. It i s 'to be e j e c t e d that, given ib lve! of uncertainity regarding many of these parameter^ differenjbf risk issessmeht using slightly different methodologies and/or different sets of assumptions will -derive somewhat divergent end-*-results. (It /should he noted that the CDC estimat was reviawed extensively by a peer review panel of outside experts before being formally adopted.) However; even in this situation in Which estimates of excess risks were different by more than one ordi&of magnitude, the ^ absolute risks of developing cancer during a putatively exposed individual's lifetime are essentially unchanged when one considers that the "background" Cif "normal" lifetime risk'of cancer in the United States is between 25 and 30 percent. All of the approaches discussed within this document are generally perceived to be conservative approaches in an attempt to develop prudent and respohsiblereeoflmendationsfortheprptection of thepublic's .health. : / -1". 'fr-:V -''' v:- ^ -S - S ' - S 1"S''-':*','' ' S / S S ' - ' v :v/'-. 3 0 c>5' JW3S"*?*}<- m ,: W s m m - . - : v? ; P M 0 .......... K*.V 'i vi: 'i?. -'; .'.; :{i.3; i.j;:., r b'.r -!,:- Page 2 - Louise A' Fabinski -,.v'1 V^V:v,SWi:-.:;.'*! '/K, s^stherefore, we reiterate our previous recommendations concerning access restriction or cleanup/stabalization of onsite, contaminated area to .prevent current, as well a s t future exposures. The logistics of '^implementing these public health protective measures do not appear to be .exceedingly difficult. The more severe problem of exposure to contaminated vfish from local rivers is best addressed directly and not through compensatory decreases in allowable exposures to .contaminated soils. 4^v#' : i-1 kk-Vk 1mfriCise P- To this end, we also strongly suggest a concentrated public health education campaign to lim it, or preferably eliminate, consumption of the highly contaminated local -fish as well as the development of reasonable- ~ Alternatives thereto. . %;r .;vV;Cv* `ii.; A M > - * * '.'''1 1 . orgi A. Jones r: _b-. v f ; ) i aasfc; 3-C> I Co Vq-ZrCcrH 35 Per Copy EPA says dioxin levels in local fish to persist By KATHY GRAY Daily News staff writer The levels of dioxin in the Tittabawassee River sediments and fish downstream from Dow Chemical Co. will not go down significantly in at j least the next several years, accord ing to a report released today by the U.S. Environm ental Protection Agency. Even though Dow has sig nificantly decreased the levels of di oxin in its discharge to the river, the EPA stated in its report that low levels in both river sediments and native fish will persist for at least several years. River sediments containing toxic contaminants attributable to Dow were found 19.5 miles downstream from Midland, the EPA reported. The EPA did not include health i risk assessments offish consumption in this study,ribut will release that data separately at a future date. The fish data with regards to di oxin didn't show any significant changes in in the levels in carp over time, but the EPA did note that the levels in catfish seem to have de creased since 1978. The 1985 levels in walleye and smallmouth bass (averaging 4.4 and 5.0 parts per trill ion respectively) were virtually un changed from 1983 levels. , Dioxin is an unwanted byproduct zenes also were consistently present of herbicide production and in its in samples tested by the EPA. most toxic form-- 2,3,7,8 TCDD--is "This data suggest that scattered a known anim al carcinogen. Its or diffuse sources throughout the effects on humans is less certain. plant (sewer sludges, contaminated "Because of the distribution of soils, pond sediments) may be con2,3,7,8 TCDD in Midland area soils tributing residual discharge loadand the persistence of 2,3,7,8 TCDD ings to the outfall,"the report stated, in the environment, bottom feeding When Dow's current wastewater fish and othergame fish fromthe Tit- permit expires in 1988, th EPA will tabawassee River may not exhibit work with the Michigan Department significantly lower levels of 2,3,7,8 of Natural Resources to discharge; TCDD in the near future,"the report limitations for several volatile and stated. "Runoff from the city, the semi-volatile organic pollutants, w astew ater discharge from Dow The report is one part ofthe EPA'e Chemical, and atmospheric deposi- dioxin initiative in Midland. This tion from Dow Chemical operations particular study examined Dow's w ill c o n tin u e to c o n tr ib u te wastewater, Tittabawassee River 2,3,7,8-TCDD to the river system." sedim ents and native fish. Rep- The EPA said th e w astew ater resentatives from the EPA will be in treatm ent system at Dow is com- Midland July 29 to review the report parable if not superior to best avail- with interested Midland residents, able technology, (BAT) but "in- Appointm ents are availab le beprocess controls for several organic tween 2:30 to 5:30 p.m. and from 7 to chemical process are not equivalent 8 p.m. on that date. To make apto BAT, even when considered with pointaient, call John Perrecone at superior end-of-pipe treatment." (800)621-8431.or(800)572-2515. The report noted that several vol- The EPA already has released re* atile organic chemicals, including ports on the levels of dioxins in Mid carbon tetrachloride, m ethylene land surface soils and the quality of chloride and styrene were among public and private water supplies in those not equivalent to BAT. Resi- the Midland area and Dow's brine dual levels of toxic semi-volatile or- operations. Reports still are schedg a n ic p o llu ta n ts, such aa p en - uled to be released based on emis* tachlorophenol and chlorinateben- sions from Dow's inciiierator. ' | J-( 3 ' 1 UNITED STATES ENVIRONMENTAL PROTECTION AGENCY REGION V 230 SOUTH DEARBORN ST. CHICAGO. ILLINOIS 80604 J u ly 15, 1986 REPLY TO ATTENTION OF 5RA Enclosed is a copy of a report t i t l e d , "Dow Chemical Wastewater C h a ra c te ri zation Study, Tittabawassee R iv e r Sediments and Native F is h ." This report presents the re s u lts of USEPA's recent study of d io xin s and other to x ic p o llu ta n ts at the Dow Chemical-Midiand P la n t. The report also presents summaries of p rio r USEPA and Michigan Department of Natural Resources stu d ie s at the Dow Chemical-Midiand P la n t , in clu d in g stu d ie s of Tittabawassee R iv e r n ativ e f is h and sediments conducted during the period 1978 to 1985. F i n a l l y , the report includ es an overview o f the Clean Water Act requirements fo r National P o llu ta n t Discharge Elim in atio n System (NPDES) permit conditions fo r best a v a ila b le technology as they p e rta in to the Dow Chemical-Midland P la n t. The inform ation and data contained in t h is report are being considered by se ve ral program o ffic e s at the Michigan Department of Natural Resources and USEPA Region V fo r processing of environmental control permits fo r Dow Chem ical. The major fin d in g s are h ig h lig h ted below: The discharge of 2 ,3 ,7 ,8 -te tra c h lo ro d ib e n z o -p -d io x in (2378-TCDD, or d io x in ) from Dow Chemical to the Tittabaw assee R iv e r , and bioaccumulation of 2378-TCDD in caged f is h and n a tiv e f is h have been confirmed. 0 Dow Chemical began operating an e fflu e n t f ilt r a t io n system in November 1985 to remove 2378-TCDD from i t s discharge to the Tittabaw assee R iv e r . Since that time the company has achieved a 67% reduction in the discharge of 2378-TCDD. Current d ischarg e le v e ls have c o n s is te n tly been le s s than 10 p arts per q u a d rillio n (p p q ), the in te rim discharge lim it e sta b lish e d by the Michigan Water Resources Commission and approved by USEPA. The estim ated current annual d isch arg e loading of 2378-TCDD from Dow Chemical to the Tittabawassee R iv e r i s 0.00012 kg (0,00026 lb s ) . Fu rther reduction o f the discharge may r e s u lt from a d d itio n al in -p la n t control measures c u r re n tly being in s t a lle d by Dow Chem ical. 0 Although s ig n ific a n t progress has been made in reducing the d ischarge le v e ls of 2378-TCDD from Dow Chem ical, i t is not li k e ly that the improvement in 2378-TCDD le v e ls found in n a tiv e f is h from the Tittabawassee R iv e r w ill be as ra p id . 2378-TCDD is p e rs is te n t in the environment. Continued low -2 - le v e l re le a se s from Dow Chem ical, su rfa ce ru no ff from the Midland a re a , and the cu rren t burden of d io xin s in r iv e r sediments may re s u lt in a gradual d e clin e in 2378-TCDD le v e ls in f is h over a number of y e a rs . Although 2378-TCDD was not detected in Tittabaw assee R iv e r sediments at d etection le v e ls ranging from 10 to 30 p arts per t r i l l i o n (p p t ), i t is most li k e l y present at lower le v e ls . The d is trib u tio n of the other te tra c h lo rin a te d d1benzo-p-diox1ns (TCDDs) in Dow Chemical treatment pond sedim ents, e fflu e n t s o lid s , and Tittabaw assee R iv e r sediments is c o n s is te n t. Tittabaw assee R ive r sediments contaminated by chlo rin ated d io x in s and furans a ttrib u ta b le to the Dow Chemical-Midiand Plant extend from Midland to the G ra tio t Road-Center Road reach of the r i v e r , or about 17.1 to 19.5 m iles downstream from Midland. Dow Chem ical's discharges of to ta l suspended s o lid s , ammonia-N, to ta l phosphorus, and to x ic organic p o llu ta n ts have been reduced from le v e ls determined by USEPA in 1981. Discharge le v e ls of to ta l d isso lved s o lid s have remained about the same and lim ite d data suggest an increased discharge of to x ic m etals (cnromium). The o v e ra ll Improvement in the discharge le v e ls can be a ttrib u te d to changes in production operations at Dow Chem ical, reduction in the discharge flo w r a t e , and in s t a lla t io n of the efflu e n t f i l t e r system. For the most p a r t , wastewater treatment f a c i l i t i e s in s t a lle d by Dow Chemical are co n siste n t w ith model wastewater treatment f a c i l i t i e s considered by USEPA fo r development of national e fflu e n t lim it a tio n s g u id elin es fo r organic ch em icals, p e s tic id e s , and re la te d chemical manufacturing pro c e s s e s . Best a v a ila b le technology e fflu e n t lim ita tio n s and best management p ra c tic e s programs w ill be developed fo r several v o la t ile and s e m i- v o la tile organic p o llu ta n ts found in Dow Chemical untreated and treated w astew aters. These BAT e fflu e n t lim ita tio n s and control programs w il l be considered w ith water q uality-based e fflu e n t lim it a tio n s developed by the Michigan Department of Natural Resources as part of the next NPDES permit fo r the Midland P la n t. This report is one of a s e r ie s that present r e s u lts of d iffe re n t aspects of USEPA's stu d ie s of d io xin s and other to x ic p o llu ta n ts in the Midland, M ichigan, a re a . To d ate , a report has been d istrib u te d regarding le v e ls of d io xin s in Midland su rface s o ils and p o ten tial p u b lic h ealth r is k s associated with exposure to those le v e ls , and a report was issued regarding the q u a lity of public and p riv a te water su p p lies in the Midland area and the Dow Chemical brine o p e ra tio n s. We are planning to issu e a report concerning Dow Chemical in c in e ra to r em issions te s tin g and ambient a i r monitoring la t e r t h is y e a r. At th a t time an o v e ra ll review of p u b lic health r is k s w ill be presented. AO 0 (9 -3 USEPA re p re se n ta tive s w ill be a v a ila b le on J u ly 2 9 , 1986, a t the Grace A. Dow Memorial P u b lic L ib ra ry in Midland to review t h is rep o rt w ith in te re ste d members of the p u b lic during the afternoon (2 :3 0 -5 :3 0 p .m .) and evening (7 :0 0 -9 :0 0 p .m .). Appointments may be arranged by c a llin g John Perrecone, O ffic e o f P u b lic A f f a ir s a t 1-800-621-8431 or 1-800-572-2515 ( I l l i n o i s ) . Any questions or comments you may have in the in te rim should be d ire cte d to Gary Amendola at the fo llo w in g address: U .S . Environmental Pro tectio n Agency Region V Eastern D is tric t O ffice 25089 Center Ridge Road W estlake, OH 44145 <2o C ^ O DOW CHEMICAL WASTEWATER CHARACTERIZATION STUDY TITTABAWASSEE RIVER SEDIMENTS AND NATIVE FISH JULY 1986 GARY A. AMENDOLA DAVID R. BARNA U .S . ENVIRONMENTAL PROTECTION AGENCY REGION V ENVIRONMENTAL SERVICES DIVISION EASTERN DISTRICT OFFICE WESTLAKE, OHIO 2. 05k( I I , OBJECTIVES The primary o b je ctive s of t h is work are to q u a n tify the co n ve n tio n al, nonconventional, to x ic organic and to x ic inorganic p o llu ta n t discharges from the Dow Chemical - Midland Plant and to assess the need fo r a d d itio n al wastewater treatment and best management p ra c tic e s necessary to achieve Best A va ila b le Treatment Econom ically Achievable (BAT) as defined by the Clean Water A ct. The inform ation and data contained in t h is report are being used by the Michigan Department of Natural Resources and USEPA-Region V to develop a proposed BAT NPDES permit fo r Dow Chemical. Secondary o b jectives in clu d e : (1) ch a racte riza tio n of untreated wastewaters and in -p la n t sludges and sedim ents; ( 2) determ ination of the types and the extent of bioaccumulation of p o llu ta n ts discharged by Dow Chemical in f is h ; (3) sub-part per t r i l l i o n analyses of e fflu e n t samples fo r PCDDs and PCDFs; and (4 ) development of inform ation about contamination of n ative f is h and sediments in the Tittabawassee R iv e r. I I I . SCOPE OF WORK Major f ie ld surveys were conducted by Region V and MDNR in 1981 and by Region V in 1984. A multi-phased f ie ld program was planned in the sp ring and summer of 1981 and executed in la t e summer and e a rly f a l l of 1981. F ie ld programs included the fo llo w in g : (1 ) a s ediment su rve y of the Tittabaw assee R iver to determine whether s ig n ific a n t to xic p o llu tan t contamination of the sediments has o ccurred; (2 ) four 24-hour composit e samples of Dow Chemical water in t a kes and e f f l u ent dis c h a rges to determine p o llu ta n t discharges at the low parts per b illio n range; (3 ) one large-volume 24-hour composite sample of Dow Chemical water in ta k e s , c e rta in e fflu e n t d isc h a rg e s, and the re ce ivin g water to determine discharge ra te s of PCDDs and PCDFs in the sub-part per t r i l l i o n range; (4 ) a s t a t ic daohnia bioassay and an a lg a l assay to determine whether or not the Dow Chemical main process wastewater e fflu e n t e x h ib its acute to x ic e ffe c t s or stim u lato ry e ffe c t s on alg al growth; (5 ) an Ames tg s t of the main process wastewater discharge to determine whether the e fflu e n t e x h ib its mutagen ic pro p e rti e s ; ( 6) a f is h bioaccumulation study to determine the le v e l and rate of bioaccumulation of p o llu ta n ts discharged by Dow Chem ical; and (7) analyses of n ative f is h from the Grand R iv e r fo r organic compounds. As part of USEPA's comprehensive study of d io xin s and other to x ic p o llu ta n ts , sampling was conducted in 1984 at the major process wastewater sewers at Dow Chemical; at other nonprocess wastew ater s . in clud in g i n cin e ra to r wastew aters . ground water c o lle c tio n system s, and la n d f il l dewatering s y s t ems; at the treated discharge to the Tittabawassee R iv e r ; and fo r T ittabaw assee R ive r s ediments . Data obtained from USEPA's 1981 and 1984 surveys are compared with Dow ChimTcal monitoring data and other a v a ila b le data from p re lim in a ry d io xin in v e s tig a tio n s conducted by Region V in 1978. 2 IV. MAJOR FINDINGS AND CONCLUSIONS A. Dow Chemical - Midland Plant 1. Untreated Wastewaters and Sewer Sludges Untreated wastewaters from process and nonprocess operations at the Dow Chemical - Midland Plan t contain high le v e ls of numerous chemical compounds. Raw waste loadings of v o la t ile compounds determined during the 1584 USlfPA survey fcarbon te tra c h lo rid e (940 lb s/d a y l : me th y le ne c h lo ride ( 920 lb s/d a y ) : styrene (570 lb s / d a y ); chloromethane (410 lb s/d a y ); toluene (350 lb s /d a y ); benzene (160 lb s / d a y ); and ethylbenzene (122 lb s /d a y )] were g reater than raw waste loadings of s e m i- v o la tile compounds [phenol (520 lb s / d a y ); 2 ,4 -d ic h lo ro phenol (45 lb s / d a y ); 1,2-dichlorobenzene (20 lb s / d a y ); pentachlorophenol (16 lb s/d av) : 2 ,4 ,6 -tric h lo ro p h e n o l (13 lb s / d a y ); and naphthalene (13 lb s / d a y )]. ^ T h e high le v e ls of v o la t ile compounds are s i g n if ic ant from an a ir p o llu tio n stand point. Emission of o n e-sixth of the v o la t ile compounds from the sewerage and wastewater treatm ent systems would be s u f f ic ie n t to c la s s if y the plant as a major source of v o la t ile organic carbon (V 0C). The fin d in g s of ch lo rin ate d benzenes and pentachlorophenol in untreated wastewaters long a f t e r term ination of production of these compounds suggests continued leaching of the compounds from sewer system sludges and plant s o i ls . Most of the untreated wastewater loading of PCDDs and PCDFs can be a ttrib u te d to co n trib u tio n s from vario us process sewers and the hazardous waste in c in e ra to r. The raw waste loading of TCDDs was estimated to be about 6 .9 x 10"4 lbs/day (3 .1 x 10-4 kg/day) and about 1.3 x 10"2 lbs/day (6 .1 x 10~3 kg/day) fo r TCDFs. Although 2378-TCDD was not detected in untreated wastewaters from the process sewers or the hazardous waste in c in e r a to r , other te tra - o c ta CDDs and CDFs were found in the 1984 USEPA stu d y. 2. T e r tia ry Pond Sediments Sediments from the tertiary pond system were found to be contaminated with several organic ch em icals. Surface sediments from the prim ary (pentagonal) and secondary (re c ta n g u la r) ponds were found to contain la rg e r numbers and higher le v e ls of p o llu ta n ts than found in t e r t ia r y pond sedim ents. These data suggest the pond system has been at le a s t p a r t ia lly e ffe c t iv e in removing s e ttle a b le p o llu tan ts not removed in the b io lo g ic a l treatment f a c i l i t y . Chlorinated benzenes were found at r e la t iv e ly high le v e ls in primary and secondary pond sediments (13-67 ppm) compared to t e r t ia r y pond sediments (ND-1.5 mg/1). Surface sediments in the ponds were g e n e ra lly found to be more h e a v ily contaminated than bottom pond sedim ents. The gradient of PCDDs and PCDFs across the pond system was s u b s ta n tia lly le s s than fo r other p o llu ta n ts . These data suggest that PCDDs and PCDFs entering the pond system are attached to f in e r p a r tic le s that tend to s e t t le over a wider area than other s e m i- v o la tile p o llu ta n ts which may be asso ciated with 3 =2 o o 2~3> heavier p a r t ic le s . 2378-TCDD was detected at 1.7 ppb in prim ary pond su rface sedim ents, 3 .8 ppb in secondary pond su rfa ce sedim ents, and from 0.10 to 0.93 ppb in t e r t ia r y pond su rface sedim ents. B . Dow Chemical - O u tfa ll 031 Discharge 1. Wastewater C h aracterizatio n Process changes at the Midland p lant and water conservation measures have re su lted in a gradual reduction in the average discharge flow from o u t fa ll 031 from over 50 MGD in the mid 1970s to le s s than 20 MGD today. Recent monitoring by USEPA and Dow Chemical suggest th a t discharges o f to x ic organic p o llu ta n ts and c e rta in nonconventional p o llu ta n ts have been reduced sin ce 1981. The apparent in crease in the d ischarg e of to x ic metals is a ttrib u te d to chromium discharges which are higher than measured by USEPA in 1981. A summary of annualized e fflu e n t discharge loadings is presented below: ___________Estim ated Annualized Discharge in Tons________ 1981 1984 1984-:1985 USEPA Survey USEPA Survey Dow Chemical Monitoring Total dissolved so lid s Total suspended so lid s Total kjeldahl nitrogen Ammonia-N Total phosphorus Toxic organic pollutants Toxic metal p o llu tan ts 148,000 680 330 270 46 15.5 5.7 150,000 1,040 [180] 87 27 14 [5 ] 1.9 6.9 129,000 (n e t) 420 (n e t) -- 23 (n e t) 11 4.9 16.7 Note: [ ] Estimated current annual fu ll- s c a le discharge loading based upon p ilo t plant f i l t e r d ata . E fflu e n t phosphorus data from the f u ll- s c a le f i l t e r system in s t a lle d in November 1985 are not a v a ila b le at this w riting . The estim ated annual d ischarge of phosphorus from o u t fa ll 031 is 5 ton s/year based upon lim ite d p ilo t plant s t u d ie s , about 90% le s s than loadings determined in 1981. 2. PCDDs and PCDFs Based upon s ix months of f u ll- s c a le operation of the e fflu e n t f i l t r a t i o n system, Dow Chemical has achieved a 67% reduction in the discharge loading of 2378-TCDD (9 .9 x 10~7 kg/day to 3 .3 x 10" ? kg /day). The cu rren t estim ated annual discharge is 1.20 x 10"4 |<g/year. TCDD analyses by Dow Chemical in d ic a te that 2378-TCDD comprises le s s than 3% of the to ta l TCDDs p re se n t. The predominant TCDD isom ers, both before and a f t e r p ilo t f i l t r a t i o n , are 1368-TCDD, 1379-TCDD, and 1237+1238-TCDD. Based upon lim ite d d ata , the u n filte re d o u t fa ll 031 discharge appears to contain higher le v e ls of TCDFs than TCDDs and higher le v e ls of other PCDFs than corresponding PCDDs. P ilo t plant f i l t e r data suggest the fu ll- s and HxCOOs, the fu ll- s c a ,.. _r may be achieving more than 90% removal of TCDDs, 2378-TCDF, and OCDD. Only data fo r 2378-TCDD have been reported fo r te r system at t h is w ritin g . 3. Biomonitoring S ta t ic bioassays (Daphnia magna) conducted by USEPA in 1981 fo r the o u t fa ll 031 discharge in d icated the d ischarge e xh ib ite d no acute t o x ic it y to te s t organisms. The discharge exh ib ite d a stim u la to ry e ffe c t on alg al growth and caused no m utagenicity in the Ames t e s t (d ir e c t and ra t l i v e r enzyme a c tiv a te d t e s t procedures). The 1981 USEPA stu d ie s were completed at a time when the average e fflu e n t discharge was about 34 MGD. Biomonitoring conducted by Dow Chemical in 1985 as required by NPDES permit MI0000868 yie ld e d the fo llo w in g re su lts fo r flow-through stu d ie s: Acute t o x ic ity 48-Hour LC50 Daphnia Magna 40% e fflu e n t Pimephelas Promelas " T f at head mi nnow) ~~ No t o x ic it y Chronic t o x ic ity MATC (geometric mean) 35.8% e fflu e n t 21.7% e fflu e n t* *embryo-larval test At the time of the Dow Chemical s tu d ie s , the discharge flow was about 20 MGD. Dow Chemical a ttrib u te d acute t o x ic it y to daphnia to the s a l in i t y of the e f f lu e n t . The mass discharge of s a lt s was about the same as that encountered during the 1981 USEPA s tu d ie s . However, the concentration o f d isso lved s o lid s was about 40% higher due to the reduction in discharge flo w . Dow Chemical a lso reports that fo r the minnow study (em bryo-larval t e s t ) , there were no observed concentration re lated e ffe c ts at hatch and a normal hatch o ccurred, ye t no org anisms su rv ived beyond 13 days. _ Np cause fo r the chronic t o x ic it y observed was suggested by Dow Chemical. Test water fo r the Dow Chemical bioassays was p re fT lte re d through a 25-micron so ck. S in c e , chemical analyses of the wastewater before and a fte r f i l t r a t i o n were not reported, the e f f e c ts of t h is procedure are_not known. 4. Bioaccumulation Studies Fin a l re s u lts from the 1981 USEPA-MDNR bioaccumulation study confirmed p re lim in a ry re s u lts with respect to the discharge of 2378-TCDD from o u t fa ll 031 and the accumulation of 2378-TCDD and other TCDDs in caged c a tfis h exposed to the plume of o u tfa ll 031 in the Tittabaw assee R iv e r . The p re lim in a ry c o n tract lab o rato ry re s u lts _for PCDFs could not be confirm ed. A unique fin d in g is that 1368-TCDD accumulated in caged f is h exposed to the o u t fa l1 / r iv e r water m ixture at higher le v e ls than 2378-TCDD. A fte r 28 days of exposure, 2378-TCDD reached n e arly 40 ppt and 1368-TCDD to about 160 pp t. Penta-CDDs (140 p p t), hexa-CDDs (43 p p t), and TCDFs (454 ppt) were found in these f is h by USEPA a n a ly s t s . There was no in d ica tio n that an e q u ilib riu m le v e l of 2378-TCDD had been reached a fte r 28 days o f exposure. The f is h exposed to the plume of the discharge accumulated g reater numbers and higher le v e ls of other organic ch e m icals, in clud in g poly nuclear arom atic compounds, ch lo rin a te d phenols, and p e s tic id e s , than did fis h exposed a t control s i t e s . A recent biouptake study conducted by Dow Chemical did not demonstrate s ig n if ic a n t uptake of 2378-TCDD or 2378-TCDF in c a t f is h exposed fo r 28 days to a m ixture of 15% e fflu e n t and 85% r iv e r w ate r. Hexachlorobenzene uptake reached 3.7 ppm (whole f is h sample) in the Dow Chemical stu d y. Most other organic chem icals included in the study protocol did not e x h ib it s ig n if ic a n t accumu la tio n over the te s t period . As with other biomonitoring by Dow Chem ical, the te s t water was p r e filt e r e d using a 25-micron sock p rio r to exposing the organisms. The e ffe c t of th at procedure on the t e s t r e s u lts is not known. 5. P o llu ta n ts of Concern From a wastewater treatment technology stan d p o in t, the p rin c ip a l to x ic p o llu ta n ts of concern are lis t e d below. The evalu atio n of appropriate Best A v a ila b le Technology e fflu e n t lim ita tio n s and Best Management P ra c tic e s programs w ill focus p rim a rily on these to x ic p o llu ta n ts . V o la tile Organic Po llutants Benzene Carbon te tra c h lo rid e Ethylbenzene Methylene ch lo rid e Toluene Styrene Sem i-Volatile Organic Po llutants 2,4-Dichlorophenol 2 .4 .5 - Trichlorophenol 2 .4 .6 - Trichlorophenol Pentachlorophenol 1.2 - Di chiorobenzene 1 .3- Di chi orobenzene 1 .4 - Dichi orobenzene 1 .2 .4- Trichiorobenzene 1 .2 .4 .5 - Tetrachlorobenzene Hexachlorobenzene 2 .4- D 2 .6- D 2 .4 .5 - T Dinoseb B is(ch lo ro b u tyl) ether 2,3,7,8-TCDD (PCDDs and PCDFs) To xic Metal P o llu ta n ts " Antimony Chromium Nickel Zinc 6 UNITED STATES ENVIRONMENTAL PROTECTION AGENCY REGION V EXHIBIT D ate: To: From: ub ject: February 2 8 , 1985 David Stringham , Deputy D ire c to r Waste Management D iv is io n J . M ilton C lark Toxic Substances, Section Evalu atio n of Data C o llected during U .S . EPA's 984 F ie ld Study of M idland, Michigan Attached i s a d r a f t o f my re p o rt. A rig o ro u s e v a lu a tio n of. data re c e n tly gathered in M idland has been made w ith emphasis on human h e a lth r i s k B a s e d upon the r i s k a n a ly s is performed, recommendations have a ls o been developed. T h is document should fu r th e r d isc u s sio n to guide d e cisio n making regarding the need fo r rem edial a c t io n s , a d d itio n a l m onitoring and human h ealth stu d ie s in M idland. I t has o ffered to provide one posslb l a n a ly s is o f the M idland f i e l d study and should not be viewed a s re p re se n tin g any p o s itio n the Agency may u ltim a te ly chose. I recommend th a t the document be forwarded to the C h lo rin a te d D io xin Work Group in EPA/HQ and the C en ters fo r D isease Control fo r Internal per re v ie w .. I t I s hoped th a t th document w i l l be u sefu l in the e f f o r t to develop the interagency evalu atio n o f the a v a ila b le d ata. 'K ,9 S S L rSo il Data FOR INTERNAL R E flE w !' l. Evaluatio n of Data C o llected during U .S . EPA's 1984 F ie ld Study o f the M idland, Michigan Area ' P e s tic id e s and t o x ic Substances B ran d i T ~ T" RgTn"`VHITS'.'"EPA -----FO R IP 2 8 FEB 1985 ^ 'REVIEW ik 'S Dow Chemical In -P la n t Samples ;" 2 ,3 ,7 ,8 -T C D D VV 7 7 "T A ' A n a ly t ic a l r e s u l t s , found in Table 1 (a tta c h e d ), o f a n alyse s o f 15 % su rface s o il samples (0 -1 ") taken in sid e the Dow p la n t grounds revealed - 2,3,7,8-TCD D c o n ce n tra tio n s ranging from 10 p a rts per t r l l l o n (p p t) t , ; t o 30 p a rts per b i l l i o n (p p p ). The high value o f 30 p a rts per b i l l i o n , station/sam ple #15, found south and e a st of b u ild in g 874 and immediately north o f the ra ilro a d t r a c k s ,' re p re se n ts an average o f d u p lic a te a n a ly s is on the sam ple.' The average s o il co ncentratio n o f the 15 samples was found to be 2 .4 ppb, w ith three o f the 15 samples having more than 1 ppb 2 ,3 ,7 ,8 -T C D D . The median value was 220 pp t. Sample # 5 , c o lle c te d southwest o f b u ild in g 956 and immediately south e a st o f the Dow in c in e r a t o r , was found to contain 3 .5 0 ppb of 2,3,7,8-TC D D . r Sample #7, c o lle c te d about 1/2 m ile northeast o f the Dow in c in e ra to r a t th northwest corner o f 11th and J s tre e ts contained 4*60 ppb o f ' 2,3,7,8-TCDD. i / Samples #11, 12, 13, and 15 were a l l c o lle c te d about 1/2 m ile n orth east o f the in c in e ra to r. Concentration of 2,3,7,8-TCDD in samples #11, 12, and 13 were reasonably c o n siste n t ranging from 220 to 440 p p t; however, sample 15, a t 30 ppb 2,3,7,8-TCD D i s d isp o rp o rtlo n a te ly h ig h . Samples # 4 ,7 ,8 , and 9 , c o lle c te d in a li n e moving northwest from the in c in e ra to r to an approximate d istan ce o f one m ile re v e a ls an In c o n siste n t p a tte rn . Sample #4 was 20 p p t; sample # 7, (1/3 m ile ) 4,600 p p t; sample #8 , (3/4 m ile ) 150 p p t; and sample #9, (1 m ile ) 10 pp t. . 2 . To tal PCDDs and PCDFs / !T PCDD and PCDF co n ce n tra tio n s were determined in one s o il sample, c o lle c t ed west o f b u ild in g 934 (sta tio n /sam p le # 4 ). The co ncentratio n o f 2,3,7,8-TCD D (0 .2 7 ppb) was found to be 84% o f the to ta l TCDOs detected (0 .3 2 ppb). C on cen tration s o f penta CDDs and hexa CDDs were 0 .2 4 ppb and 4 .0 ppb, r e s p e c t iv e ly . The r a t io o f hexa CDDs to 2,3,7,8-TCD D was 1 4 .8 . Hepta CDDs were found a t a concentratio n o f 75 ppb. Several PCDFs were a ls o detected 1n t h is sample. 2,3,7,8-TC D F was present a t a con c e n tra tio n o f 27 p p t, w h ile to ta l TCDFs amounted to 900 p p t. Hexa CDFs and hepta CDFs were 3.1 and 15.4 ppb re s p e c tiv e ly . B . R e sid e n tia l S o il Samples 1 . 2,3,7,8-TCDD 7 7 7 7 7 /A ' S u p e r f ic ia l s o il samples (0 - 1 ") were c o lle c te d and analyzed from M idland, M ichigan; an in d u s t r ia l area in M iddleton, O hio; and fo u r n atu ral are as in M innesota. T h is sampling design was to determine i f 2,3,7,8-TCDD , D bb FOR INTERNAL -2- REVIEW ONLY A f --f ru n f i r '. . & contam ination was more or le s s unique to a lo c a tio n , such as Midland, where 2 ,4 , 5-Trich io ro p hen o l (2 ,4 ,5 -T C P ) and a v a r ie t y o f 2 ,4 ,5 - T d e riv a t iv e s had beep m anufactured. For more inform ation on design and s a p lin g see the September 2 2 , 1983 EPA document, " S o il Screening Survey a t , Four Midwestern S ite s " ( 1 ) . Data from t h is sampling e f f o r t 1s attached and found 1n Tab les 2-5A. Yard composite s o il samples from three lo c a tio n s , upwind (w est/southw est) p f Dow Chem ical, see Table 2 , were found to contain non-detectable le v e ls o f 2 ,3 , 7 , 8-TCDD. D etection lim it s ranged from 2 to 4 p p t. At two of these lo c a tio n s s o i ls c o lle c te d near downspouts or under ro o f d rip l i n e s , were found to have low le v e ls o f 2 ,3 ,7 , 8-TCDD; 6 and 9 ppt. Of fo u r lo c a tio n s sampled from the perim eter o f the Armco p la n t In M iddleton, Ohio only one s o il sample o f s ix c o lle c te d and analyzed was found to be p o s it iv e fo r 2 ,3 ,7 , 8-TCDD and a t a le v e l o f # p p t. ( fa b le 4 ) , In c o n tra s t, a ll nine samples c o lle c te d along the perimet* o f,t h e Dow-M1dland p la n t were found to be p o s itiv e fo r 2 ,3 ,7 , 8-TCDD. C oncentrations o f 2 ,3 ,7 , 8-TCDD ranged from a low o f 10 ppt to a high Of 2.03 ppb, the average co ncentratio n being 327 ppt and the median v a lu e , 69 pp t. (See Tab le 2 ) . . 'V Of seven p u b lic use lo c a tio n s sampled in M iddleton, Ohio only one was found to contain 2 ,3 , 7 , 8-TCDD a t a concentration o f 4 p p t. A ll nine p u b lic use areas sampled In Midland were found to have 2 ,3 ,7 , 8-TCDD 1n s o i l s w ith va lu e s ranging from j^ ^g 170 pp t . The a rith m e tic mean was 56 pp t; the median v a lu e , 19 ppt*. The h ig h est co n cen tratio n s o f 2 .3 . 7 . 8- TCDD observed were those c o lle c te d from lo c a tio n s most near the Dow p la n t or lo cated from e a st n o rth east to north northeast o f the f a c i l i t y . As p re v a ilin g wind d ire c tio n in M idland, see Fig u re 1 , In "S o il Screening Survey a t Four Midwestern S it e s ," i s from the southwest or w e s te rly d ir e c t io n , d eposition o f 2 ,3 ,7 , 8-TCDD by a i r 1s suggested. The h ig h e st 2 ,3 ,7 , 8-TCDD concentration of 110 p p t, however, was found a t B u llo c k School , lo ca te d about 1/2 m ile southwest o f the Dow f a c i l i t y . Of the 19 re s id e n t ia l s u p e r f ic ia l s o il samples (0 -1 ") analyzed in M idland, 18 were p o s it iv e fo r 2 ,3 , 7 , 8-TCDD* Concentrations ranged from non-detectable to 22Q p a rts p er t r i l l i o n . The a rith m e tic mean o f t h is data was 54 p p t; tne meilTan, 22 p p t. Two o f seven s o il samples from re s id e n t ia l areas in Middletown, Ohio were found to be p o s itiv e fo r 2 .3 . 7 . 8- TCDD. These sam ples, both found a t one lo c a tio n , contained 4 and 5 ppt o f 2 ,3 ,7 ,8 -T C D D . The median value o f a l l the Ohio samples was "non-detectable" (< 3 p p t). An average valu e o f 2 .2 ppt was reported by Dow Chemical in s o ils from 20 in d u s t r ia liz e d areas in the U.'S. (November 5 , 1984 re p o rt) ( 2 ) . None o f the three n a tu ra l areas sampled in Minnesota contained d etectab le le v e ls o f 2 ,3 ,7 ,8 -T C D D . (See Table 5 ) , The data from our stu d ie s o f the fo u r n a tu ra l a re a s , and M iddleton, O hio, when combined w ith the r e s u lt s on 20 I n d u s t r ia l iz e d areas reported by Dow, re v e a ls higher le v e ls o f 2,3,7,8-TCD D in M idland, than 1n o th er areas stu d ie d . I f f is h data I s Included fo r com parison, d iscu ssed l a t e r , t h is fin d in g becomes even more apparent. f o r in t e r n a l -2-^:'^ ` R E V IE W 0 ; \ L * V'"-v'- '& Concentrations o f 2,3,7,8-TCD D found in s o i ls from yard s tended to de crease w ith d ista n c e from Dow ch e m ica l. (See Tab le 2 ) . Fo r in sta n c e , a l l yard sam ples, B -3 , D-3, and E -3 , c o lle c te d about 3700 y a rd s from the Dow in c in e r a t o r , were found to have lower co ncentratio ns o f 2 ,3 , 7 , 8-TCDD than sam ples, B - l , C - l , and E - l ' which were a l l taken about 2100 yards from the in c in e r a t o r . T h is p attern suggests 2 ,3 ,7 , 8-TCDD d eposition by a i r . An a i r d ep o sitio n hypothesis i s fu rth e r supported by comparing the 2 ,3 , 7 , 8-tCDD le v e ls observed in s o ils c o lle c te d near downspouts or under d r ip lin e s w ith those found in y a r d s . Downspout and d r ip lin e s o il concen t r a t io n s o f 2,3,7,8-TCD D were approxim ately two to ten tim es higher than those found in y a rd s . 2 . T o ta l PCDDs and PCDFs Concentrations o f PCDDs and PDCFs in su rfa c e s o il samples are presented i n Tab le 6 (a t ta c h e d ). Samples c o lle c te d upwind o f Dow Chemical (#13401 and #13395) contained only low le v e ls o f hepta and octa CDDs (1340 ppt) and no d e te ctab le le v e ls o f PCDFs.. A ll s ix p u b lic access are as sampled i n Midland had,measurable co n cen tratio n s o f hexa CDDs, ranging from 63 to 441 ppt and a l l samples contained PCDFs, three w ith 2 ,3 ,7 ,8 -T C D F and hexa CDFs (sample #s 13375, 13393, and 13394). Of f iv e samples analyzed from Middletown, Ohio only one contained d etectab le le v e ls o f hexa CDDs , and o nly one w ith d etectab le le v e ls o f PCDFs. No hexa CDFs were found In the s o il sam ples. S u rface s o il samples from three w ild e rn e ss areas ini1Minnesota containad low le v e ls o f hepta and octa CDDs (200 ppt) and rto d e te ctab le le v e ls o f PCDFs. ! 3 . Computation o f 2,3,7,8-TCDD Eq u ivalen t R isk s Methods are c u r re n tly being developed to assess the r is k from exposure tp c h lo r in a te d d io x in s and dlbehzofurans o th er than fo r 2 * 3 ,7 , 8-TCDD. These a re d iscu ssed in the September 2 6 , 1984 d r a ft EPA document en t i t l e d "H ealth Hazard Assessment, fo r C h lo rinated D io xins and -D ibenzofurans o th er than 2,3,7,8-TC D D " ( 3 ) . Of the e ig h t methods c ite d in the document two, the Sw iss enzyme and the EPA methods, appear to have the b est s c i e n t i f i c foundations as they are based upon enzyme in d u c tiv e p ro p e rtie s and ca rcin o g e n ic assessm ents. These methods are . shown below. Where Isomer s p e c if ic a n a lyse s were u n a v a ila b le , the t o x i c i t y o f the 2 , 7 , 8- substituted;hom ologubs i s u t iliz e d . e A: E stim a te s o f R e la tiv e T o x1 c1 tie s For PCDDs and PCDFs f '...... Exp ressed 'as E q u iv a le n ts o f 2 ,3 ,7 , 8-TCDD ' ` Compound' EPA Method (1984) Sw iss Method 2 ,3 ,7 ,8 -T C D D o th er TCDDs 2 ,3 , 7 , 8-PeCDDs o th er PeCDDs 2 ,3 ,7 , 8-HxCDDs O ther HxCDDs 2,3,7,8-HpCDDs o th e r HpCDDs OCDDs , ; ; ; 1 ' 0.01 o .2 - v 0.002 0.04 v U r ' i r- 0 .004 ; ;o "Hv . 1 , 0.01 0.1 0 .1 0.1 0 .1 0.01 0.01 :5 - ?H Table A: Continued -4- FOB INT A,, h e V i e w JJ iVJL V F* 'TSVSii.'it*v-5k*A Compound EPA Method (1984) 2 .3 .7 .8 - TCDFs o th er TCDFs '2,3,7,8-PeCD Fs o th er PeCDFs 2 .3 . 7 . 8- HxCDFs o th e r HxCDFs HpCDFs, OCDF i T ; 0.001 ,, 0.1 * 0.001 ?. ; 0.1 ; " 0.001 0 Sw iss Method 0.1 0.1 0.1 0.1 0.1 0.1 PCDD and PDCF data from M idland, converted to 2 ,3 ,7 , 8-TCDD eq u iva le n ts! i s presented in the follow ing se ctio n . v Table B: 2 ,3 ,7 ,8 - T C DD E q u iva le n ts o f Other K d D s and PcDFs Found in Midland S o ils L o c a tio n : Bu llo ck School V irg in ia Park Method i ' Method PCDDs/PCDFs Level w 2 ,3 ,7 , 8-TCDD no TCDDs 220 PeCDDs 120 HxCDDs 410 flpCDDs 2,400 2 ,3 ,7 .8 -T C D F 15 ; TCDFs -- ' FeCDFs no HxCDFs 170 : E q u iv a le n ts ( p p t)1: EPA Sw iss no no 2 -2 l 24 12 16 41 0 V , 24T--TT 2 ? 2 - 11 n 17 17 nre? - ~ rr? ; Level (PPt) EPA Sw iss 76 290 no 240' 410 13 - 40 64 T T ' 76 76 r2 2 22 11 10 24 04 1 . 1: - .- : li n 17 17 ; " T3? - - t o Vvj'iv- m W" Longview School 2 3,7 , 8-TCDD 78 tCDDs 170 PeCDDs TOO* HxCDDs 340 HpCDDs 2,300 2 ,3 ,7 ,8 -T C D F 13 TCDFs PeCDFs ND HxCDFs 260 Equivalents (p p t): * value estimated 78 78 11 20* 10* 14 * 34 0 23 1 1' 00 26 26 m - f~ m Mapelton School 15 15 15 40 0 0 ND 0 0 63 2 6 380 0 4 ND :.,Q 0 ND 0 0 ND 0 0 o 'l SO 0 3 1 V tw B"i.,.;;.*4 - wvv. ` -5 - m ' ' tabi e'B: Continued ;^ ^ f f | c a t | o n : v `: 1 ' .": j : - ?' .7-:. i ijf-. ;J|J;r::2 3 ,7 ,8 -T C D D ip '1'.v: T tn iU ' ..:-1'" } " PCDDs ? HxCDDs $ .:v HpCDDs 2 ,3 ,7 ,8 - TCDF ;- . 'PPjei CDDFFSs - :::K*DFs V. Central School \ Bal 1 diamond) 221 EPA Sw iss 12 40 ND 86 350 ND ND ND ? p t): 12 0 0 3 0 .0 0 0 IF ' 1 2 ; 0 0 9 4 0 0 0 DRAFTF C B ip J N j J.H NAL REVIEW Oi^LY County L in e Road PPt EPA Sw iss 33 ND : 0 V ND 0 67 3 350 0 ND 0 ND 0 ND . /' 0 . 3 0 0 7 3 0 0 0* 13 | -$ 4$?<vS-i! .vw^i :aijfcttt.i*: ','^i, . f.' ' ' m: The PCDD and PCDF co n cen tratio n s expressed as 2,3,7,8-TCD D e q u ivale n ts range from a low o f 6-13 ppt to a high o f 182-219 p p t. R e la t iv e ly small d iffe re n c e s e x is t between the Sw iss and EPA methods to d e riv e 2 ,3 ,7 ,8TCDD e q u iv a le n ts fo r o th er PCDDs and PCDFs. Three of the lo c a tio n s sampled; B u llo ck Sch o o l, V ir g in ia P a rk , and Longview School have 2 ,3 , 7 ,8-TCDD q u iv a la n t s o f 7 tP .2 0 tim es J h e equiyja,e.Hts found-ID s o ils a t County Lin e Roaa, C entral Scho o l, and Mapelton School. C o n sisten t r a t io s o f 2 ,3 ,7 ,8-TCDD e q u iv a le n ts to 2 ,3 ,7 ,8-TCDD measured s o il le v e ls can be observed. At B u llo ck School t h is r a t io i s 1 .8 3 ; 1.88 fo r V ir g in ia Park and 1.99 fo r Longview School. The average o f these r a t io s i s 1 .9 0 . The co n siste n cy 1n the r a t io suggests a s in g le prim ary source fo r the PCDDs and PCDFs observed In these r e s id e n t ia l, p u b lic use s o i l s . : J p ; . . ....... x;Wlu-vK^,- Assuming th a t a r a t io o f 1.90 2 ,3 ,7 ,8-TCDD e q u iva le n ts to 2,3,7,8-TCD D e x is t s fo r a l l other lo c a tio n s sampled in M idland, i t i s p o ssib le to d e riv e estim ated 2 3 ,7 ,8-TGDD e q u iv a le n ts fo r r e s id e n tia l s o il samples from measured 2,3,7,8-TCD D s o il d a ta . An average yard s o il co ncentratio n o f 2 ,3,7,8-TC D D can be estim ated assuming th a t co n ce n tratio n s o f 2 ,3 ,7 ,8 TCDD in s o ils near downspouts rep resent 10% o f the area o f the yard contaminated a t t h is le v e l and 2,3,7,8-TCD D co n cen tratio n s 1n yard com p o s ite s o il samples away from downspouts, re p resen t 90% o f the contami nated s o i l s . These c a lc u la t io n s are shown below In Tab le C . o'1'- .v Field l Humber W :f ' : ^T'" B *l - ' f , B.l-l^ r; \ b-3-i Tab le C: Average 2 .3 .7 ,8-TCDD E q u iva le n t . , Levels In Residential Surface.. S o il Samples in M idland, Michigan Measured 275X51CPP (ppt) 76 160 20 270 Average 2 ,3 .7 l8 -fC D D $011 .Concentrati on 1RBL. ; 84 ;; 45.' Estim ated '2,3,7',8lCDD Equivalents ' Tn~Sltt~2 ' . IME 160 ' i '- 032-- FOR INTERNAL ;.'7- u-6^ vV U` iv t-V Estim ated Measured Average 2 ,3 ,7 ,8-TCbD , 3 , 7 ,8-TCDO 2,3,7,8-TCDD E q u iv a le n ts A;A' (PPt) ^ il Conentttton in S o ils 2 ; tppt) /;" ( p p t T 26 54 r:V ,', .;: 28 12 35 240 18 19 31 26 28 49 9 10 20 53 67 36 53 19 '-i!S : V--` 11,7/i/t,;*8 Yard composite s o il sample, 0-1" 3 1' * Downspout composite s o il sample, 0-1" 1 Sw iss Model '' >: ,V:;v'3 A A ''' \ A > A A A ' ; :' 2 E q u iv a le n ts using Sw iss Model, EPA model y ie ld s estim ates approxim ately if|p f;lq s s f As shown in Table C , s u p e r f ic ia l s o il co ncentratio ns o f PCDD and PCDFs, expressed ds 2,3,7,8-TCD D e q u iv a le n ts , would range from 19 to 160 ppt. I f co rre cted fo r re c o v e rie s o f in te rn a l standards some valu es would be s lig h t ly higher (10-25% ). F iv e o f the seven lo c a tio n s have 2 ,3 ,7 ,8 - TCDD e q u iv a le n ts oyer 50 p p t. the average 1s 68 p p t; the median, 53 p p t. In c o n tra s t, s o il samples from Middletown, Ohio would have 2,3,7,8-TC D D e q u iv a le n ts ranging frm 1 to 13 pp t, w ith ah average o f 5 ppt and a median o f 3 p p t, using the Sw iss Model. Fo r two o f the s i t e s , B -l and C -3, co n ce n tra tio n s o f 2 3 ,7 ,8-TCDD in downspout s o ils `I.a-iI1,?;-.h V :-m"ake a la rg e c o n trib u tio n to the o v e ra ll yard average ,3 ,7 ,8 -T C D D . : M Mi C7 Risk Assessment from Exposure to 2,3,7,8-TCDO and other PCDDs and PCDFs A : . ::I.n . MidTan1d ttesi"eI.Mnl)O'WT0n;.`v;n3(' tfl s V^.:---w/;: ,T;; v 1.^ .L - A t-'V'AA- ' . V. Background A A : A '- A A A a A A ;^ Both EPA and Centers fo r D isease Control (CDC) have prepared r i s k a sse ss ments f o r exposure to 2,3,7,8-TCD D 1n s o i l s . EPA*s f i r s t exposure and r i s k assessm ents were performed by the C h lo rinated D ioxin Work Group In response to the fin d in g o f 2,3,7,8-TCD D contam ination in Hyde P a rk , Love C a n a l, and a t the Syntex s i t e . Based upon these assessm ents, EPA entered In to consent decrees and a d m in is tra tiv e orders to re q u ire cleanup and capping o f contaminated areas to the detection lim it o f 2,3,7,8-TCDD (from 1-50 ppt) ( 4 ) . In September 1982, EPA prepared an exposure and r i s k assessm ent fo r contaminated s it e s found in M issouri ("Region 7 Exposure/R1sk Assessm ent"* D r a ft , September 2 8, 1982) ( 5 ) . The cancer r1 ? k s from exposure to 2 ,3 , 7 ,8-TCDD In re s id e n tia l s o i l s were determined the Car^hogftrti Assessment Group Model , v ih e ;h ig h e s t cancer r i s k ; o <j 3 ,0 03 3 :m y- 1 N 1 :- R N A L EvV o i\ L , Ws found to be fo r young c h ild re n who could in g e s t s o il by the behavior 'fap*'* :,. .rr.-. . ' , , J l c a l1ed p ic a * ' I n ; the Region 7 assessm ent, it was assumed th a t c h ild re n would consume 0 .5 grams o f s o il per day fo r 7 y e a r s . Under these*condi- tlo n s a c h ild would have a cancer r i s k o f approxim ately one chance in / alOOO (1 0 " 3 ). using a upper bound l i m i t o f r i s k o f 95%, i f exposed to '' s o ils h a v in g ,! ppb 2 ,3 ,7 ,8 -T C D D . Since 1982, as discussed 1n EPA's < 'Ambient Water Q u a lity Document, February 1984, the e xtra p o late d slope n\ / l i n e o f 2 ,3 ,7 ,8 - fCDD animal cancer data has now been changed from 4 .2 5 x K J , . 1 (05 (mg/kg-bw/day) " 1 to 1 .56 x 105 (mg/kg-bw/day) _1 ( 6 ) . T h e re fo re , 1V using t h is nw slope d e riv a tio n and EPA's exposure assessment o f `1982, a le v e l o f 1 ppb o f 2,3,7,8-TCD D in s o i ls would pose a s lig h t ly lower lif e t im e cancer r i s k o f 0 .3 7 - x; TO*3 . In h a la tio n o f 2 ,3 , 7 , 8-TCDD contam- . f . inatqd dusts was found tp pose much low er r i s k s than from in g e stio n and i dermal absorption o f 2 ,3 , 7 , 8-TCDD from s o i ls was considered so u n lik e ly as to not to pose a r i s k . `ifc.' 'v' M V . ./' \` . . - ' ' / I n e a r ly , 1983 EPA requested CDC to determine i f a 1 ppb le v e l o f 2 ,3 ,7 ,8 - ; TCDD in s o i ls represented an a p p ro p riate le v e l o f concern. T h is a c tio n , in p a r t , was prompted by the widespread u t iliz a t io n o f 1 ppb d etectio n 1 ,, l i m it s by Region 7 f o r t h e ir analyse s o f 2 ,3 , 7 , 8-TCDD in s o i l s . In ; Ja n u a ry 1984, CDC published an abbreviated a n a ly s is , o f t h e ir r is k a ssess- : ment f o r exposure to 2,3,7,8-TCDD which concluded th a t a le v e l 1 ppb o f ,4 2 ;,3 ,7 ,8 t TCDD. :in re s id e n t ia l so i 1s was ah approprlate le v e l o f concern * f o r human h e a lth ( 7 ) . A more complete 2,3,7,8-TCD D exposure and r is k assessment document was re le a se d by CDC in J u ly 1984- (" H e a lth /Im p lic a tio n s o f 2,3,7,8-TCDD con tam ination o f R e sid e n tia l S o i l ," Kimbrough, e t a l . ) (8 )* Th authors concluded th a t so1,l le v e ls g re a te r than 1 ppb, pose a le v e l Of concern, however, 1 t was recommended th a t management d e cisio n s should a ls o con s id e r an e va lu a tio n of s p e c ific circum stances a t each contaminated s i t e . CDC's r is k assessment considered 2,3,7,8-TCDD v ia exposure to s o ils in clu d in g th a t from dermal co n tact and by in h a la tio n , although , th e-d u st co n cen tratio n used in c a lc u la tio n s was not pro vid ed . U n like EPA's exposure model, CDC u t il iz e s a 10-12 y e a r s o il h a l f - l i f e fo r 2,3,7,8-TC D D , A ir and w ater contamination o f 2,3,7,8-TCD D were b elieved not to make a s ig n if ic a n t c o n trib u tio n to TCDD in ta k e In most s it u a t io n s , and th e re fo re was not considered in the model. However, CDC did sta te th a t TCDD cpuld e v e n tu a lly end up 1n the food c h a in , p a r t i c u l a r l y in f i s h . Under these co n d itio n s CDC s ta te d , " i f TCDD e n te rs a food c h a in , th e re i s an unknown a d d itio n a l source o f exposure which must be added to the r is k o f those in d iv id u a ls exposed to contaminated s o il and o f a l a r g e r ; undefined p o p u la tio n ". CDC a lso commented th a t i f contaminated s o i ls wer near waterways and Could contaminate these w aterw ays, accept ab le le v e ls o f 2,3,7,8-TCD D may need to be lower (than 1 ppb). The CDC r i s k assessm ent i s based upon a lin e i^ m u lti stage e x tra p o la tio n o f the Kociba animal cancer data fo r 2,3,7,8-TCDD as Interp reted by Squire ( 8 ) . CDC'S e x tra p o la tio n d if f e r s from EPA in th re e ways ( 6 ) . F i r s t , the curve f i t was performed not on adm inistered dose but on l i v e r concentra tio n a t term inal s a c r if ic e * Second, CDC's extra p o latio n does not a d ju st 2568b A O o 3H ""tWREV-rV V.sSSV^.;W' ^ ''if' W Sv `rtf : & rI : ' U-, ,ti. - p p -y-y. ......... .. i ',f FOR n ; - 8- R ^ v i, - v . DRAFT .f o r high e a rly m o rta lity of te s t anim als. F in a lly ,'C D C uses a ra t- to man e x tra p o la tio n on the b a s is o f body weight ra th e r than su rfa ce a re a . These m o d ific a tio n s .re s u lt in a 95% u p p e r-lim it potency value e stim a te , when co rrected back to adm inistered dose, y ie ld s a slope fa c to r o f 3 .6 x ., - y w . f\ mm yn/ / bn ni j ^- u Kn u/ / uHuajw# \ " l n uu hn i r h -J Yj v coo fryiWAW+C.Mn tw .hK7u a faVirvtbnVpI n f A 7 t h a n t h t d erived by EPA. In CDC's exposure assumptions a 70 kg in d iv id u a l (male) was used as a model. A p o t e n t ia lly higher r is k In d iv id u a l, one o f lower body w e ig h t, such as a 50. kg person (fem ale) i s often used when ^performing r i s k assessm ents. : -; U$1rig CDC's exposure and r is k assessment assum ptions, exposure to 1 ppb Of 2,3,7,8-TCD D in s o il over life t im e has an upper 95% p ro b a b ility of causing cancer o f 2 .3 x 10" 5 . Expressed s lig h t ly d if f e r e n t ly , a t a con c e n tra tio n of 1 ppb, exposure to 2,3,7,8-TCDD in s o ils could accumulate to a s u f f ic ie n t dose In ju s t three y e a rs to in cre a se an in d iv id u a l's r is k o f cancer by one chance 1n a m illio n ( 8 ) . I f no more than 10" 6 upper proba b i l i t y of causing cancer i s d e s ire d , using upper 95% confidence li m it s , then a s o il co ncentratio n o f 4*4 j>pt. would be an acceptable le v e l o f 2 ,3 ,7 ,8 -T C D D .' Fo r a 10" upper 95% p ro b a b ility o f cancer a s o il le v e l o f 440 ppt would be re q u ire d . I f EPA 's e xtra p o la tio n o f animal cancer 1s employed in the CDC model these valu es decrease to 10 ppt fo r a 10" 6 r i s k and to 102 ppt fo r a 10"5 r i s k . These va lu e s decrease fu rth e r to approxim ately 7 ppt and 73 p p t, re s p e c tiv e ly , 1 f a 50 kg in d iv id u a l 1s used as a model. krv-'V. ww>p 2* R isk From Exposure to R e sid e n tia l S o ils 'm*fe i l f t ' S'ii". ` s i . t ''^3 ' -,.i , 1`'" : Tab le D, below, presen ts the upper 95% life t im e cancer r is k s from exposure to contaminated s o ils observed in M idland, Michigan using (1) EP A 's 1982 exposure and r is k model m odified by the new, 1984 cancer e x tra p o la tio n (2 ) CDC's exposure and r is k model and (3 ) CDC's exposure and r is k model using EPA's 1984 animal e xtra p o la tio n (4 ) a 50 kg in d iv id u a l as a model and (5 ) estim ated to ta l PODDs and PCDFs. (See d lscu s- $1on, page 5). > V WWW - '; Table D: * Estim atio n o f Cancer R is k s from Exposures to S o ils in Midland, Michigan using R l s l f r V ~ and Exposure Mode)s o f EPA and CDC* y location c#r..-. iCField #) u^;/;Vv;-N? . . ' '>'I1-V it.i*').t,'ii:.'' e - i- L >1;.' ; 8 -3 -L :Sfe:v .-.i e - i- L C-3-L 0-3-L .-* E - I- L -3-L fi't ,*V' j/fclXi,. " Upwind o f i Midland m i - ' " (2 samples) Estim ated Average 2 ,3 ,7 ,8-TCDQ Equivalents jn Soil . (ppt) Cancer R isk Models T O -------- EPA 19822 CDC Modified 160 5.9x10-5 3.7x10-6 2.2x10*5 86 3.2x10-5 2.0x10-6 1.2x10*5 53 1.9x10-5 1.2 x lO a6 7 .2 x1 0 -6 67 2.5x10-5 1.5x10-6 9 .0 x l0 - 6 36 1.3x10-5 8.3x10-7 5.0x10-6 53 1.9x10-5 1.2x10-6 7.2x10-6 19 7.0x10-6 -4.4x10-7 2.6x10-6 ' 2 ; y-k c: .7 .4 x l0 - 7 4.6x10-5 2.8x10" ' : V ' k ' " ; yy / - H 3 3 3 .y . . y -! t-j m W ^KW'&..*....., b :w ^ - : Continued :^T u rv t N "1 Z. I x-! -9- H E v i c.y v ; O i\i- fc lie illr b b 'i Average 2,3,7,8-TCDD E q u iv a le n ts in S o il ..... (ppt) "\i'- ..y. - Cancer R isk Models "T c ~ ~ EPA 1982? CDC Modified ? i; 'V-, :v \ -i' ^ yb'W ddliow n;'' bybb-Bb; | * ; y ( f c i :o b- y . I l dCiDl eS) 1.8x10-6 >1.1x10-7 6.9x10-7 . v-.y y .y.' " y .$ 2 , Minnesota f c ::'ciS?saip1s) / y y ik l;' CIO-7 <10" 8 ' "`bi'-'1-.:, .-:b.jn^&;b::bb.b;V'b b: ' <10-7 vbb.. ... ^ - airvb I*U sin g S w iss Model ; EPA model y ie ld s e q u iv a le n ts about 20% le s s .liS ld p e fa c to r o f t .56x105 (mg/kg-bw/ditylrl u t iliz e d | t y ''\ EPA. slo p e fa c t o r o f .56x10.5 (mg-kg^bw/day)" ) | 50 kg*b1w in d iv id u a l. v f \ y (M odified model, i s 6 . Ox CDC's o rig in a l ,model) fQ'y.i "b' f e b v ;b/ W eM v ^ . th ree caneen r is k models y ie ld roughly e q u iva le n t r i s k s . The EPA . model y ie ld in g the h ig h e st r i s k s . For 2,3,7,8-TCDD e q u iv a le n ts g re a te r than 50 ppt al.1 models produce cancer r i s k s exceeding 1 0 -6 . Above 75 ppt the CDC,:modified model and the EPA model in d ic a te cancer r i s k s of 10-5 or g re a te r, a t the 95% upper confidence 1im it . iteri-. ; : ; W V' V jff I t should be noted th a t EPA 's exposure .amd r is k assessment f o r 2 ,3 ,7 ,8 TCDD may not adequately co n sid e r 1ntake o f s o il from vegetables or gardens grown in contaminated s o i l s . While 2,3,7,8-TCDD bloacCumulatio n does not o c cu r, co atin g o f ;p la n t m a te ria l by 2,3,7,8-TCD D contami nated s o i ls has been documented - (6 ). CDC assumes in t h e ir e xp o su re ? and r i s k assessment a s o il in g e sti n o f 100 mg/day, although 1 t i s ' mmm not c le a r i f s o il in g estio n from garden work or from consumption of mm foods grown in gardens has ben includ ed in t h is f ig u r e . Some authors have e stim a te d so il in ta k e by geophagla to be as g reat as 3g/day ( 9 ) . v::i- I f t h is v a lue w e re ;s u b s titu te d in to CDC's assessment model the r is k s from 2 ,3 , 7 ,8-TCDD would be d ra m a tic a lly in cre a se d . A range o f assump t io n s can a lso be made to estim ate dermal Exposure to s o i l s CDC estim ated d a ily dermal exposure a f t e r age 15 as 100 mg/day. According to CDC, 10 grams o f s o il le s s than 1 mm th ic k can be spread over an area o f about 15cm? Or about 6 in ch es? ( 8 ) . T h is i s le s s o f a su rfa c e area which e x is t s fo r oji sid e o f a hand. For a in d iv id u a l 'i . working in 'a garden, an exposed sk in su rfa c e area o f 30 inches? and c o n ta ct w ith 50g o f s o il may be more r e a l i s t i c . While such d ir e c t m tm y '-V ?V>sbv&>- exposure would o nly e x is t f o r a p o rtio n o f the y e a r , 5r6 months, one or more days per week, i t coul& be a fo r instance very sig n ifica n t s - v . - , ro u te o f exposure i f human dermal absorption o f 2,3,7,8-TCD D e x i s t s . Lb In animal s t u d ie s , Po ig er and S c h la t te r determined th a t 0,1-6% o f drm ally ap p lied 2,3,7,8-TCD D in ap p lied s o ils was absorbed ( 1 0 ) . I f these s o il exposure assum ptions, are placed In to the CDC model, r is k s from exposure to 2,3,7,8-TCD D in s o i ls would be In c re a se d . Converting PCDDs and PCDFs to 2,3,7,8-TC D D e q u iv a le n ts when Isom eric s p e c ific analyses are unavailable also increases calculated cancer r is k . hoi LC>03<o jW / Given the u n c e rta in tie s 1n the exposure to s o i l s , the cancer potency o f 2 ,3 ,7 ,8 -T C D D , and supportive human epidem iological s tu d ie s , a*' prudent p o lic y would be one e lim in a tin g exposure to contaminated o ils which have 2 ,3 , 7 ,8-TCDD e q u iv a le n t co n ce n tratio n s exceeding 71C3 p p t. In M idland, exposure to PCDDs and PCDFs could be reduced by removing the top 3" o f s o il from ya rd s where co n ce n tra tio n s exceed 75 ppt as 2,3,7,8-TCD D e q u iv a le n ts . In many In s ta n c e s , r i s k s could tie reduced to acceptable le v e ls by e lim in a tin g exposure to downspout s o i l s . A t play a re as a t schools i t would be d e sira b le to l i m it exposure to s o ils having 150 ppt or more 2,3,7,8-TCDD e q u iv a le n ts . P lay a c t iv it y 1n such a re a s may occur fo r a number o f y e a r s , exposure o ccu rrin g p rim a rily by dermal co n tact and by in h a la tio n . A higher le v e l Of concern fo r these a re a s , as compared to contaminated s o ils a t p riv a te re sid e n c e s, probably i s acceptable as a la rg e r amount of exposure would be expected to occur 1n pre-school aged c h ild re n le s s than 5 y e a rs o f age p layin g in p riv a te re s id e n tia l y a rd s . These recommendations have a ls o been made in the co ntext o f o th er ro u tes of exposure to 2,3,7,8-TCD D in M idland, p rim a rily by in h a la tio n o f ambient a i r and f is h consumption, dlscussed 1n sub sequent s e c tio n s , and suggestive h e alth data from M idland. Rates o f connec t iv e and s o ft t1ssue cancer among Midland women, have been s t a t i s t i c a l l y s ig n if ic a n t fo r over two decades (1 1 ). No other Michigan County and only 13 co u n tie s in the nation have had s t a t i s t i c a l l y s ig n if ic a n t ra te s f o r two su cce ssive decades, 1950-1979, where there have been two o r more ca se s recorded per decade (1 2 ). Connective and s o ft t is s u e can ce rs have been asso ciated w ith exposure to 2 ,3 ,7 ,8 TCDD ( 6 ) . While exposure tO 2,3,7,8-TCDD or o ther PCDDs and PCDFs i s y e t to be determined in these c a se s, animal stu d ie s show th a t fem ales exposed %p 2,3,7,8-TCD D are more s e n s itiv e than males ( 6 ) . B ir t h d e fe cts In'M id lan d county have a ls o been elevated compared to other Michigan co un ties (1 3 ). In the e a rly 1970s the ra te s of oral f a c ia l and u ro g en ital d e fe c ts rose p r e c ip it o u s ly , then d eclined (1 4 ,1 5 ). Sugh d e fe c ts are c o n s is te n t w ith those observed in anim als exposed to 2,3,7,8-TCD D ( 6 ,1 4 ) . EPA's Health Review D iv is io n 1n 1979 conducted p re lim in a ry a n a lyse s of Midland b ir t h d e fe ct data and beleived s u f f ic ie n t evidence e x iste d to w arrant a case control study o f these b irth d e fe cts In cooperation w ith the Michigan Department o f Health and CDC, but the study has not been in it ia t e d (1 4 ) . CDC's a n a ly s is o f Dow C h em ical's repro d uctive study o f w ives o f Midland workers exposed to PCDD's revealed a rate of o ra l f a c ia l c le f t s tw ice twice that expected (1 5 ,1 6 ). While such h ealth inform ation demonstrating a re la tio n s h ip w ith expo sure to PCDDs/PCDFs and adverse human h e a lth - e ffe c ts 1n Midland i s f a r from c o n c lu s iv e , 1 t may be s u f f ic ie n t ly suggestive to w arrant reducing fu r th e r exposure to these substances. Fo r t h is and the fo llo w in g reasons a clean-up le v e l fo r 2,3,7,8-TCDD in t h is instance of le s s than 1 ppb i s j u s t l f i b l e In M idland: (1 ) Cancer r i s k s equal or exceed 10" o r 10" fo r life t im e exposure to s o il co n ce n tratio n s o f PCDDs and PCDFs of 75 ppt or g reater expressed as 2,3,7,8-TCDD eq u ivale n ts. Y ififg: v MM:, cf.-*';, , :ixff:lk : 0 B iN - n - R E V I E v. (2) Large u n c e rta in tie s e x is t in 2,3,7,8-TCDD exposure analyses developed to d a te , w ith some not being s u f f ic ie n t ly conser v a t iv e .' / (3 ) H is t o r ic exposure to PCDDs and PCDFs 1n Midland may have occurred fo r a time period of 40 ye ars or more, given chemical manufacturing h is to ry . (4 ) Other routes o f PCDD and PCDF exposure, by ambient a i r and f is h consumption, s t i l l e x is t in Midland and h is t o r ic concentra tio n s in these media may have been f a r g re a te r, having a lre a d y co n trib u ted to high body burdens o f PCDDs and PCDFs. j, v - '(> u- (5) A portion of M idland's population, perhaps a larg e p o rtio n , has < had occupational exposure to PCDDs, w ith some sm a lle r groups p o s sib ly exp erien cin g adverse h e a lth e ffe c t s from t h is exposure, 8 -;'p f i :V 1 I . v At the present time in s u f f ic ie n t data e x is t s to id e n tify areas in Midland needing remedial a c tio n s . F u rth e r sampling e f f o r t s would be necessary , to id e n t ify those areas where 2,3,7,8-TCD D concentrations exceed a chosen action level , ' $&*:. !* m 'W'.'Y*; -vi- R ,' Ambient A ir Data Data provided by Dow Chemical on 2,3,7,8-TCD D co ncentratio ns in ambient a i r r e presented below in Table E ( 2 ) . A to ta l o f 10 measurements were taken: two w ith in the Dow P la n t , s ix a t the Dow P la n t fence lin e and two in the c i t y o f M idland. A ll samples were found to be p o s itiv e fo r 2 ,3 ,7,8-TC D D . The data generated In d ic a te ra th e r la rg e flu c tu a tio n s in a i r co ncen tratio n s o f 2,3,7,8-TCDD. Table E : Ambient A ir Concentrations (pg/m3) o f 2,3,7,8-TCDD in M idland,l^ fcH 'lgan' (Narch V if e e b r u a r y Location Number o f Measurements Range Arithm etic Mean Dow P la n t 2 0.019-0.022 0.021 Dow P la n t Fence Lin e 6 0 .010- 0.21 0.064 C it y o f Midland 2 0.019-0.16 0.090 2iSv* 1 . Data supp lied by Dow Chemical (Reference #2) 2 . Data o f EPA u n a v a ila b le a t t h is time R isk assessm ents from in h a la tio n o f 2,3,7,8-TCD D have been made by EPA in regards to em issions from m unicipal waste in c in e r a t o r s . Exposure* b io - a v a i l a b i l i t y and o th er assumptions are discussed 1n more d e ta il In the document, "Assessment o f Em issions from a Recent Municipal Waste Combustor," December 16, 1983 ( 1 7 ) . 61ven these assumptions and in co rp o ratin g the most re cent change In the cancer slope determination fo r 2,3,7,8-TCD D , the life t im e cancer r i s k s from exposure to va rio u s a1r le v e ls o f 2 ,3 ,7,8-T C D D , measured in Midland are shown below in Table F . Using a r a t io o f 1.90 to estim ate t o t a l r i s k s from PCDDs and PCDFs as done fo r s o i ls ; r is k s fo r estim ated a i r le v e ls o f PCDDs and PCDFs,/has5)been d e riv e d . 25690 ^ ^ oo3| '^;. y-^V ' ,V.T!*'V'VvA'.}*1, v !i$`:.;:i ' ' .-"V. ,, v ';:'(( " Table F : mA^a:'-,... , . ;* ; Concentration -; -. -c . ,(pq/nr*) '.ji'.. -fir \\ i/;-/'%;/// 412- F Q R I.N ",R.N'A*R E V l L vv w . , i L ife tim e Cancer R is k s from Exposure to Ambient A ir L e v e ls o f 2,3,7,8-TCDD and Estim ated PCDDS and PCDFs in Midland Cancer Ri sk 2 ,3 ,7 ,8 -T C D D Cancer Ri sk PCPD s/PCDF s iO-21 v(high value ^ a t Dow P la n t Fence lin e ) i; I p . 16 (high value ii in Midland) Fence lin e ) .09 (average value in Midland .6 .9 x TO-6 5 .3 ` x 103 ,0 .x 10" 1.3 x TO" 5 1.0 x 10" 5 5 .7 x 10-6 1 As upper 95% confidence l i m it ` Estim ated from the r a t io of PCDD and PCDF to 2,3,7,8-TCDD a s 2,3,7,8-TCDD q u ivalan ts found in s o il m% >m M Jte.'.' V W<l : As shown from data presented in the Table F , cancer r is k s from life t im e exposure to 2,3,7,8-TCD D and from the estim ated co n cen tratio n s o f to ta l PCDDs and PCDFs are a l l 1n the 10" ; to 10" range. No inform ation i s a v a ila b le on h is t o r ic a i r le v e ls o f PCDDs and PCDFs in M idland, although p a r t ic u la t e m atter from Dow' s ro ta ry k iln in c in e ra to r in 1978 operated w ith o u t supplementary fu e l had reported p a r tic u la te co n cen tratio n s o f 'fc ^ /.S -T C D D as high as 13,200 ppb and hexa-CDDs of 65,000 ppb (1 7 ). With supplementary f u e l, the co ncentratio n o f PCDDs in p a r tic u la te m atter was reduced to low ppb le v e ls , o r by a fa c t o r o f 1000 or more. H is t o r ic oper a tio n s o f these and other in c in e r a t o r s , a s w ell a s fu g it iv e PCDD em issions from the f a c i l i t y when 2 ,4 ,5 - T manufacture was o c cu rrin g , could have given r is e to f a r higher a i r le v e ls o f PCDDs and PCDFs in p reviou s y e a rs than those observed c u r r e n t ly . I f so, exposure to PCDDs and accumulated cancer r i s k s may have reached unacceptable le v e ls fo r long term re s id e n ts o f M idland. I n Dow Chem ical' s November 5 , ISB ^ 'hepbrt o f point sources o f 2,3,7,8-TCD D in M idland, h is t o r ic waste in c in e ra to r p ra c tic e s were id e n tifie d as the p rin c ip a l sources of 2,3,7,8-TCDD s o il contamination, as cu rre n t emission r a t e s from the ro ta ry k i 1n in c in e ra to r and three other smal1 er 1n cin e ra to r sources could not account fo r le v e ls o f 2,3,7,8-TCDD which have accumulated .in s o i ls in Midland or a t the p la n t s it e ( 2 ) . Dow reported th a t in c in e r a t o r ..Combustion gases were d ir e c t ly vented to the atmosphere u n t il 1968 and high ' temperature in c in e r a tio n , needed to destroy d io xin was not f u l l y implemented u n t il 1978. Although h is t o r ic a i r co n cen tratio n s of PCDDs and PCDFs are unknown a t t h is tim e , i t should be p o ssib le to produce models using reco rds o f p a st disp osal p r a c tic e s and in c in e r a to r operation in order to produce e stim a te s o f p revio u s exposure and asso ciate d r i s k s . Adding to the r is k s c a lc u la te d fo r PCDDs and PCDFs are a i r co ncen tratio n s o f o ther carcinogens or suspect carcinogens found p re se n tly o r h is t o r ic a lly in M idland. The M ichigan Department of Natural Resources has accumulated a i r data in Midland 'which should be evaluated in o rd er to a sse ss a d d itio n al r i s k s . In one n atio n a l survey o f a i r em issions from in d u s t ria l f a c i l i t i e s , using estim ated \ em ission d ata , the Dow f a c i l i t y in Midland was ranked as the number one a i r source fo r to x ic substances 1n the S ta te o f Michigan (1 9 ). As life t im e cancer r i s k s from ambient a i r exposure to ju s t PCDDs and PCDFs are in t h e r p n ; 10" to 10" range, more a i r m onitoring o f these and other hazardous -, chem icals i s recommended. FOR IN E h N A L DRAFT- 13- R E V IE W & U iML Y Cancer R is k s from Exposure to S o ils and Ambient A ir L e v e ls o f PCPDs n In C D t's exposure and r i s k assessment i t was assumed th a t co n ce n tratio n s o f 2 .3 .7 .8 - TCDD in airb o rn e d usts would equal those found in s o i l s . F u rth e r more, i t was assumed th a t Indoor a i r le v e ls o f 2 ,3 ,7 ,8 - TCDD on p a r tic u la te m atter would equal those found in ambient a i r . T h is assumption was viewed by some commenters as o v e rly p r o te c tiv e . Assuming th a t Midland re s id e n t ia l s o ils average 50 ppt o f 2,3,7,8-TCD D and to ta l p a r t ic u la t e m atter averages 70 ug/m3 , the expected maximum concentra tio n o f 2,3,7,8-TCD D in a i r would be 3 .5 fg/m3 . A ir le v e ls in Midland exceed t h is valu e by as g re a t as 50 f o ld , suggesting th a t r e s id e n t ia l s o il alone could not be re sp o n sib le fo r the observed co n ce n tra tio n s o f 2 .3 .7 .8 - TCDD in ambient a i r in M idland. A value o f 70 ug/m3 i s the prim ary a i r standard fo r to ta l suspended p a r t ic u la t e s . ` I t i s u n c le a r what p a r t ic u la t e m atter co ncentratio ns were used in CDC's In h a la tio n exposure assessm ents* T h e re fo re , w ith the CDC r i s k models i t Tl may not be v a lid to add cancer r is k s from measured a i r co n cen tratio n s o f 2 .3 .7 .8 - TCDD and o th er PCDDs to those r is k s c a lc u la te d from s o i l s . I f CDC assumed dust le v e ls o f 140 ug/m3 , as mentioned In t h e ir re p o rt, then a s o il co n cen tratio n o f 1 ppb, would r e s u lt in a i r le v e ls o f 2,3,7,8-TCDD approxim ately equal to those measured In M idland. T h e re fo re , i f the CDC s o il exposure r is k model was d erived w ith these assumptions 1t is in a p p ro p ria te to add r i s k s from ambient a i r exposure to 2,3,7,8-TCD D to those p re v io u s ly found fo r s o i l s . I t i s In te re s tin g to note th a t the average s o il co ncentratio n observed on the Dow p la n t s it e was 2.38 ppb and 0 .7 4 ppb along i t s p e rim e te r. Given the very la rg e area o f the Dow , p la n t s i t e , 1500 a c re s , a irb o rn e dust and s o ils from the s it e may be a prim ary c o n trlb u te r to c u rre n t a i r le v e ls o f 2,3 ,7,8-TC D D . Modeling should be conducted to determine c o n trib u tio n s from e x is tin g a i r sources 1n Midland to those from p la n t and r e s id e n t ia l s o i l s . Using the EPA model, cancer r i s k s from in h a la tio n could be added. For s o ils co n tain in g 5 ppt 2,3,7.8-TCD D e q u iv a le n ts , life t im e cancer r is k s would In c re a se frOm 2 .8 x 10_ 5 to 3 .8 x 10" 5 I f the cancer r is k from lif e t im e exposure to the peak va lu e o f 0 .1 6 pg/m3 2,3,7,8-TCD D converted t;o 2,3,7,8-TC D D e q u iv a le n ts f o r estim ated a i r le v e ls o f other PCDDs and PCDFs i s In clu d e d . v IV . Fish Analysis 1 . Overview o f 2,3,7,8-TCDD F ish Data F is h from 13 Michigan R iv e rs were analyzed 1n 1983 to determine fle s h co n ce n tratio n s o f 2 ,3 ,7 ,8 -T C D D . A to ta l o f 74 f is h were an alyze d . Whole f is h carp samples ranged from non-detectable to 190 ppt o f 2 ,3 ,7 ,8 -T C D D , w ith the high value found in carp c o lle c te d from the Tittab aw assee R iv e r $N in M idland. T h is data 1s found in Table 11 (a tta c h e d ). The average o f 2 .3 .7 .8 - TCDD found in whole carp in r iv e r s o th er than the Tittabaw assee averaged 3 p p t, or 1.6% o f the co ncentratio n observed in whole f i s h from the Tittab aw asse e R iv e r . No f is h analyzed exceeded 10 p p t. A s t a t i s t i c a l t - t e s t o f t h is data re v e a ls th a t carp from the Tittabaw assee r i v e r are s ig n if ic a n t ly d iffe re n t from f is h from other Michigan r iv e r s a t a confidence l i m i t g re a tly exceeding 0 .0 0 1 . I t th e re fo re can be sta te d w ith v ir t u a l c e rta n ty th a t Tittabaw assee R iv e r f is h are uniquely contam 256S2 in ated and not the r e s u lt o f chance. FO R i n 1> -14- REV IEW -IN. In d iv id u a l carp f i l l e t samples from the Tittabaw assee R iv e r ranged from 12 t a 530 ppt 2 ,3 ,7 ,8-TCDD, w ith an a rith m e tic mean of 50 ppt and a median* o f 25 p p t. A s k in le s s f i l l e t composite o f f iv e c a t f is h averaged 75 ppt of 2 ,3 ,7 ,8-TCDD w h ile a composite o f skin-on f i l l e t s o f f iv e smallmouth bass contained 5.1 ppt o f 2 ,3 ,7 ,8-TCDD. F iv e in d ivid u a l skin-on f i l l e t s o f w a ll eye averaged 3 .9 ppt o f 2 ,3 ,7 ,8-TCDD. v /, 2 'O verview o f PCDD and PCDF F is h Data - W 'k A composite Of f iv e whole carp samples analyzed in 1981, downstream o f Midland were found to contain 81 ppt TCDD, 31 ppt penta CDD, 44 ppt hexa CDD, 53 ppt hepta CDD. and 14 ppt beta CDD. PCDFs were found to be as fo llo w s : 37 ppt TCDF, 73 ppt penta CDF, 145 ppt ,hexa CDF, 31 ppt hepta CDF, and 4 ppt octa CDF. About 90% o f the to ta l TCDD observe^ was 2 ,3 ,7 ,8TCDD,(18). / m# ' -' 't ' i f e .'lit 3 . Exposure and R isk A n a ly s is From Consumption o f Tittabaw assee R iv e r F is h Although a f is h a d v iso ry a g a in st consumption o f f is h from the Tittabaw assee R iv e r has been in place sin ce 1978, there have been re p o rts o f In d iv id u a ls consuming f is h from the r i v e r ( 2 0 ) . T h is 1s not to be unexpected as many people tend to ignore f is h a d v is o r ie s , p a r t ic u la r ly those in poorer economic groups. Fo r people consuming even small amounts o f bottom f is h from the Tittabaw assee R iv e r over t h e ir lif e t im e s , the cancer r is k s are f a i r l y d ram atic. To provide a p e rsp e ctive o f such r i s k s , Table G has been developed which g ive s frequency o f f is h consumption and asso ciate d cancer r i s k s ,, fo r d if f e r e n t s p e c ie s , and f b r both 2 ,3 , 7 ,8-TCDD and estim ated to ta l PCDDs and PCDFs as 2 ,3 ,7 ,8-TCDD e q u iv a le n ts ; PCDDs and PCDFs other than 2 ,3 ,7 ,8-TCDD have been estim ated from 1981 composite carp a n a ly s is by the r a t io s o f congeners to to ta l TCDD, assuming th a t 90% o f the measured TCDD i s 2 ,3 ,7 ,8-TCDD. f e : Tab le G: L ife tim e Cancer R isk s From Consumption o f T i t tabawassee k lv e r Ff$h in Midland , . M't ;; Frequency o f Cart> .'riV\ . Consumption* 2 ,3 ,7 ,8 - 1fotal TCDD2 PCDDs/PCDFs^ Iv'' Once per week 7.0x10-4 1.1x10-3 C a tfish '2,'3,778 Total TCDD4 PCDDs/PCDFs5 1.1x10-3 1.6x10-3 S p o rtf1sh 2 ,3 ,7 ,8 Total 7 TCDD6 PCDD/PCDFs7 5.7x10*5 8.6x10-5 Sjpnce* per 0|sonth; 0^-V!'` 0 Once per y r . 1.8x10-4 1.5x10-5 2.6x10-4 2.2x10-5 2.7x10-4 2.2x10-5 4.0x10-4 3.2x10-5 1.4x10-5 1.2x10-6 2.1x10-5 1.8x10-6 : * f l:\ K. J ' ` Ten times over Lifetim e 2.2x10-5 3.2x10-5 3.2x10-6 4.7x10-6 1 .7 x l0 -7 2.6x10-7 iri' ----- ' '!'*!--..... ....' r Assumes 1/2 pound f i s h per meal 2 Average va lu e o f 50 ppt 3 ' C alcu late d from r a t io s o f PCDDs/PCDFs observed in composite sample o f carp analyzed from the Tittab aw assee R iv e r 1n 1981. Both Sw iss and EPA models y ie ld 2 ,3 ,7 ,8-TCDD e q u iv a le n ts o f 22 ppt fo r PCDDs and PCDFs o th er than fo r 2 ,3 ,7 ,8-TCDD V , ` -' -15- FO R IN T E R N A L 1 'R EV IEW uKL'n 75 ppt See footnote #3 above. Both Sw iss and EPA models y ie ld 2 ,3 ,7 ,8-TCDD e q u ivalen ts o f approxim ately 35 ppt fo r PCDDs and PCDFs o th er than 2 ,3 ,7 ,8-TCDD. * 4 p p t-;; .- ; See footnote #3 above. -PCDDs and PCDFs oth er than 2,3,7,8-TCDD have e q u ivalen ts ......... O f ;2 p p t. .. " : t fl !Cancer r is k s fo r the in freq u en t consumer of H ttabaw assee R iv e r f i s h , fo r in s ta n c e ten tim es over lif e t im e , Ore reasonably small being in the range o f 1(H>. For more freq uent consumers o f carp and c a t f is h , even a 1/2 pound meal once per y e a r , cancer r is k s exceed 1 0 "5 . I f consumption i s more fre q u e n t, once a month, r is k s exceed 10-4 . Consumers o f only s p o rtfis h would have e xce ssive r i s k s , g re ate r than 10" 5 , fo r consumption equaling or exceeding 1/2 pound meal m onthly. A once a month consumption o f 1/2 pound o f f is h i s equal to 7 .3 grams o f f is h d a ily or s lig h t ly higher than EPA' s average f is h c,oii|umption^value o f 6 .5 g/day used In ambient w ater q u a lity c r i t e r i a docu ments. I t i s l i k e l y th a t a s m a lIn fra c tio n , o f Midi andl s population would consume bottoms f is h o f s p o rtfis h a s "fre q u e n tly as once per y e a r or once per io iit h , adding s ig n if ic a n t ly to cancer r i s k s re ce ive d from exposure to PCDDs and PCDFs in s o ils and ambient a i r . Probably many fisherm en o f the Tittabaw assee R iv e r would tend to d isca rd ~7?;w;'.. ' -Oa^jl being le s s the most r e a lis Consumer would de tic eat s ira b le way to equal fo r eating generate an amounts of c than c a tfis h average ris k arp, ca tfish , or s p o rtfis h . Perhaps i s to assume th a t a and s p o r t fis h . Fo r an w i f : - S ,:':' f % in d iv id u a l consuming 1/2 pound o f f is h per y e a r the cancer r is k would be 5 .6 x10" 5 ; on a monthly b a s is the r i s k would be 6 .8 x10 " ^ fo r to ta l PCDDs and PCDFs expressed as 2 ,3 ,7 ,8-TCDD e q u iv a le n ts . Other ch e m ica ls, such as ;PBBs and pentachlorophenol, have a lso been dectected in Tittabaw assee f i s h , I p t i f e so the r i s k estim ates c a lc u la te d h e re , may underestim ate the can cer r i s k . r M `:: women consuming a 1/2 pound meal o f f i s h , w ith an average o f 63 ppt 2 ,3 , 7 ,8-TCDD e q u iv a le n ts , in ta k e would be over 2,000 pg/day or about 40 ng/kg-bw/day fo r a 50 kg w o m a n .S h o rt term sub-human primate stu d ie s hve demonstrated rep ro d uctive e f f e c t s , In clu d in g conception f a ilu r e s and m isc a rria g e s a t a*dosages as low os 1 .8 ng/kg-bw/day ( 6 , 8 ) . C le a r ly woman consuming meals o f f is h from the Tittab aw assee In Midland would have a high r i s k o f adverse re p ro d u ctive outcomes. I f . a s a fe ty fa c to r o f 1000 i s u t i liz e d f o r humans, as by CDC, then no consumption by women o f any q u a n tity o f Tittab aw asse e R iver, f is h would be recommended. Cum ulative R is k s from Exposure to S o i l , Ambient A i r , and F is h in Midland /and Recommendations fo r Remedial, Adtions ' To tal r i s k s from exposure to PCDDs and PCDFs in a ll media are presented 1n Tab le H. The assessm ent does not in clu d e exposure to other chem icals Vfhlch may be in ambient a i r and in f i s h . R e a lis t ic and w orst case sce n ario s have been c a lc u la t e d . A ll r i s k s re p re se n t lif e t im e exposure and the upper 95% p r o b a b ilit y o f c o n tra c tin g c a n c e rs . Using ty p ic a l (average) concentra t io n s o f 2 , 3 , 7 . 8*TCDD e q u iv a le n ts found in re s id e n tia l s o il (50 ppt) and in a i r (0 .0 9 pg/m^) cancer r i s k s would range from 6 .9 x 10" 6 to 2 .4 x 1 0"5 , using the CDC m odified model and the EPA model, r e s p e c tiv e ly . Using the w o ^ t observed s o il arid a i r le v e ls measured in M idland, cancer r i s k s would ;; 25694 ; S'Rp'O 'rfw. ; # X : 'V FOR INFERNAL -16- R E V I E W O is L 'V -''-r''-\W;' -;' '!-., g$r'' <V> be'1n the range o f 2 .2 x 10-5 to 6 ,9 x 1 0 -5 . i f f 1sh consumption o f te n , r ' " t \ v 1/2 pound meals in a lif e t im e ar added to those r is k s found from exposure o average le v e ls o f PCDDs and PCDFs, then r is k s in cre a se to 1 .5 x 10-5 - 3 .3 x 10` 5 . For w orst case s it u a t io n s , more frequent consumers o f Tittabaw assee R iv e r f i s h , would have cancer r is k s in the range o f 10~4 to - mmp 'IfcfcM x' fp tx W: Exposure Table H: Combined L ife tim e Cancer R isk s From Exposure to PCbDs and PCbFs in R e sid e n tia l S o ils , AmFlent Air and F is h in M idland, Michigan . ' ; sriMr- E W ;-: si m - mMms. ! 'm & S c e n- r rr iorrr Soi l l 1 . t y p ic a l 50 ppt: (Average) ,2. Worst 160 pp t; Observed i ; v X '-X Soil + A ir A ir1 6 .9 x10-6(CDC)2 Ovl7pg/m3: 1.8x10-5(EPA) ,, 2 .2 x l0 ' 5(CDC)* 0.30pg/m3; 5 .9 x10" 5(EPA) 5.7x10-5 1.0x10-5 F is h 1 lO x: 8 .2 x l0 -6 1 . Mo: 6.8x10-4 2 . Wk: 2.7x10-3 Typ ical : S o il + A ir = 6.9x10-5 (CDC m odified) '1'a;. (Average) 2.4x10*5 ( EPA) Worst Observed: S o il + A i r > 2.2x10-5 (CDC m odified) 6.9x10-5; (EPA) .. Soil + A ir + Fish '' 1 T y p ic a l: S o il (CDC m odified) + A ir + F is h (10x In lif e t im e ) 3 (Average) 1.5x10-5 S o il (EPA) + A ir + F ish ( IQ x in life tim e )* "Worst'Observed:' Xv,-x ' 1 S o il (CDC) m odified) + A ir + F is h (Once per month)3 :: 7.0x10-4 ' S o il (CDC m odified) + A1r + F is h (Once per week)3 2.7x10-3 S o il (EPA) + A ir + F is h (Once per month)3 7.5x10-4 S o il (EPA) + A ir + F is h (Once per week)3 2.7x10-3 1 Using estim ated v a lu e s o f some PCDDs and PCDFs expressed as 2 ,3 ,7 ,8-TCDD e q u iv a le n ts . '2 CDC model as m odified v 25695 x O O H3 : : /'j: V;:'.;',v ; - V ". ?} r.-* -17- FOR - INTERN: R E V U .. iff a I f only 2,3,7,8-TCDD i s included in the r is k analyses, the calcu lated c r i s k s decrease by about h a l f , s t i l l leavin g r is k s higher than accep table. U T h is i s p a rtic u la rly true fo r the worst-observed case situ a tio n of residen t i a l s o il and g ir contam ination. Where Cancer r i s k s have exceeded 10" or approached 10- 3 , U .S . EPA has g e n e ra lly required remedial a c tio n s . P r io r it o the 2,3,7,8-TCDD contamination in M is s o u ri,- s ite s were'capped or cleaned up where 2,3,7,8-TCD D exceeded d e te ctio n lim it s (1-50 ppt) ( 4 ) . Although th e r i s k s 1n Midland do not w arrant emergency a c tio n , i f only s o il and a i r * r i s k s are co n sid ered , long term reduction 1n r is k 1s In d ica te d and would 1 be c o n s is te n t w ith p re v io u s ,a c tio n s o f the Agency. I t i s a lso im portant, as d iscu ssed by CDC, th a t each 2,3,7,8-TCD D s it e should be evaluated as to i t s own unique c h a r a c t e r is t ic s . While a 1 ppb 2,3,7,8-TCD D a ctio n le v e l f o r clean-up of re s id e n tia l s o i ls and re lo ca tio n o f in d iv id u a ls can be defended, i t I s more a p p lic a b le to s itu a tio n s where p r io r exposure has been of a sh o rte r duration arid o th er exposure ro u tes such a*s a i r , f is h and occupation do not e x is t ( i . e . , Times B e a ch ). I n M idland, chlo ro pheno lic .m anufacture and dlsposal o p eratio n s began in the 1930s ( 2 ) . H is to r ic a i r ( arid f ls h co n cen tratio n s o f PCDDs and PCDFs and human exposure and re s u lta n t body burdens i s unknown. Fo r these reasons and due to the suggestive heal.th data which e x is t s in Midland county i t 1s recommended th a t lower a c tio n le v e ls fo r 2,3,7,8-TCD D 1n s o ils be employed. Recommendations . W w fii C o n sid eratio n should be given to covering or removing s o i ls on the Dow P la n t s it e and around I t s p erim eter exceeding 1 ppb 2 ,3,7,8TCDD e q u iv a le n ts. T h is a ctio n would reduce the p o ten tia l fo r o f f - s it e d is p e rs a l, help decrease ; e x ls t in g a i r le v e ls in M idland, and reduce worker exposure to contaminated s. . R isk reduction from exposure to p riv a te re s id e n tia l s o ils having 2 ,3 ,7 ,8 - ,.JCPD e q u iv a le n ts exceeding 75 ppt or where 2,3,7,8-TCDD exceeds 40 PPt j Should be co nsid ered . Emphasis should be piaced on those are as where .c h ild r e n p r in c ip a lly p lay ( i . e . , back y a rd s ) or where gardens e x i s t . In -many in s ta n c e s , perhaps in the m a jo rity o f c a se s, lim it in g exposure to v sol i s in the v i c i n it y o f downspouts may be s u f f ic ie n t to reduce r i s k to ^ tce p ta b le le v e ls . iAdditipnaH sem plIng w i l l be required to id e n t if y , re sid e n c e s having s o il 2,3,7,B^TCDD co ncen tratio n s o f 40 ppt or g re a te r. ^ 3I 'V, . S fW - 'lim it in g exposure to s o ils in re c re a tio n a l areas of p u b lic use s it e s and 'a t schools having 150 ppt o r g re a te r as 2,3,7,8-TCDD e q u iv a le n ts o r about 180 ppt 2,3,7,8-TCD D should be co nsid ered . &J i y-' lip#:: 4 . E stim a te s o f r i s k s from h is t o r ic a i r le v e ls o f PCDDs and PCDFs; a s w ell a s other* t o x ic substances, should be made using appropriate modeling and a i r data. v:' '"T ' j; Ip-?"' A d d itio n al ambient a i r and source m onitoring fo r PCDDs and PCDFs and other hazardous chem icals should be implemented on a re g u la r b a s is to determine e x is t in g r i s k s and to co n sid e r remedial a c tio n s . A- J ac> ; v !| ism flte F O R IN "M A -18* eR E V I L vv $7 p r e lim in a r y h ealth stu d ie s should be considered fo r Midland re sid e n ts a t aVvm-Va.1 high p o te n tia l h e alth r i s k . I d e a ll y , these would be in d iv id u a ls liv in g M i l near the Dow f a c i l i t y fo r over 20 y e a rs and w ith average o r higher* jR$-. s m ? i ,:than average, s o il co n ce n tratio n s (2 ) employees and t h e ir f a m ilie s of Dow pbem iCai exposed to PCDDs as id e n t if ie d in the Townsend repro d uctive ;,p tw d y and (3 ) consumers o f Tlttabaw assee R iv e r f i s h . The in cid en ce ra te s y p f p i s e a s e , rep ro d uctive outcomes*- and blood chem istry in these groups ..should be compared to matched groups outside of M idland. Smith and 'f 'l Thompson have re c e n tly Id e n tifie d and Interview ed 128 fisherm en o f the ` Tlttabaw assee and have p re lim in a ry data regarding f is h consumption (2 0 ). Standard epidem ological approaches to gather and evalu ate such inform ation `" have been developed and should not be d i f f i c u l t to Implement in t h is L' 1 s it u a t io n . To a sse ss rep ro d uctive outcomes o f Dow w orkers, the townsend , ?' data s e t , could be re-evalu ated arid compared w ith co ntro l populations ' o u tsid e o f M idland. The H ealth Review D iv is io n o f EPA has been reported . ; to have a computer p rin to u t o f the Townsend study (1 5 ). Human m ilk -,v samples could also be c o lle c te d fo r re sid e n ts of 20 y e a rs o r more to , ' estim ate PCDD and PCDF body burdens. A unique opportunity e x is t s in 7 `- . ,, Midland fo r a ssessin g human h ealth impacts from long term exposure to yy, .,7; * .fCDDs, p a r t ic u la r ly from f is h consumption, Performing such stu d ie s would W r / . `V , seW e not only in thei assessment o f p u b lic h ealth in Midland but y#ould con- f Urthe r-^o 'the understanding ; o f h ealth impacts from PCDDs and PCDF s . Disclaim er document in no manner should be viewed as representing U .S . EPA 's f in a l p o sitio n regarding actio n s in Midland, Michigan. , . . i H f e t i i . . 1: A . : J l . * ,- I * ! , 1 .v ,,,-r ~ t- H iiio n C la rk , Pli.D . * y . , ' 'V . m m > !a^y-z'-yy., mn Mi 25697 2 0 o<=f ;<T hw ; ri-'.t;:1;-. -19References c: v> o k i- Y f*- ,C- ;c l . U.S, EPA. Region V "Soi 1 Screen!ng Survey at Four Midwestern Sites", September 22, 1983. Cr^' '"V; - K-;-c . ';;V -.- -: , .' ^ v ?;Dow. Chem ical, "P o in t Sources and Environmental L e v e ls o f 2 ,3 ,7 ,8-TCDD on f - i ; M idland P la n t S it e o f the Dow Chemical Company and in the C ity o f Midland, 1 Kovmber 5 , 1984. . ! ' 3 . B e ll i n , J . and D. B a rn e s, "H ealth Hazard Assessment fo r C h lo rin ated i;yri :'A>.; 'V - ,' " D io x in s and - Dibenzofurans o th er than 2 ,3 ,7 ,8-TCDD, D r a f t , U .S . EPA, ' September 26, 1984. -- i; ..'v. ' ^ - : 4 . K u re n t, E "Enforcement Concerns In the Process and Outcomes o f Clean-up , 'M r L e v e ls fo r D io xin in S o i l " , memorandum to C . Kleveno , U .S . EPA, 'k September 17, 1982. . , t ; 5 . U .S . EPA, "O utline For Region 7 Exposure/Risk Assessm ehi, " d r a f t ;#1, " K September 28, 1982. 6 . U .S . EPA, Ambient Water Q u a lity C r it e r ia For 2 ,3 ,7 ,8-TCCD, February 1984. 7 . Centers fo r Disease C o ntro l, "H ealth-Risk Estim ates fo r 2 ,3 ,7 ,8 -T e tra Chlorodibenzo d io xin in S o i l ," M orbidity and M o rta lity Weekly R epo rt, V o l. 3 3 , No. 3 , Jan uary 2 7 , 1984. :&.-< 'c.'Jiv,. .*i *' 8 .Kimbrough, R .D , e t a l . , "H ealth Implementations of 2 ,3 ,7 ,8 - T e tr a c h lo r o - dibenzodioxin (TCDD) contam ination o f R e sid e n tia l S o i l " , Centers Fo r Disease C ontrol, J u ly 1984. 'I!! `hjk'v- .. Wedeen, R . P . e t a l . , "Geophagic Lead Nephropathy"; Case R ep o rt, Environmental Research, V o l. 17, pg. 409; 1978. TO. P o ig e r, H. and C . ` S c h la t t e r , "In flu e n c e o f S o lven ts and Absorbance on Dermal and In t e s t in a l Absorption o f TCDD", Food Cosmet. T o x ic o l. , V o l. 18, : pg. 4 77 , 1981 . ..... '"/T v ^ 11. i! ;llS W 12. } im'fvM Michigan Department o f P u b lic H e a lth , "E va lu a tio n o f S o ft and Connective T issu e Cancer M o rta lity Rates in Midland and o th er Selected Michigan Counties Compared N a tio n a lly and S ta te w id e ," May 4 , 1983. Riggan, W. e t a l . , U .S . Cancer M o rta lity Rates and Tre n d s, 1950-1979, Volume I I , EPA-600/T^5T-OT5b, September 19837" ; , 1 3 . H arg er, J . R . E . , " F i r s t Cut A n a ly s is o f B ir t h D efect D a ta ", memorandum to l ' D. I s l e i b , T o xic Substance co ntro l Commission/ S ta te o f Michi gan , ilf I October 29, 1980. 1i ,1 4 , P o o le , C . , "C ase-control study o f B irth D efects in Midland County* \ M ichigan", H ealth Review D iv is io n , U .S . EPA, October 4 , 1979. If ' - v . - r .'' -i'-#*. 3 P 'M - *. v-V' V v - ^ 'y . t, 'V;V.V '" 25S93: I H(q - ! r t ',; 'V "; ' ; : ; : sr . : ' <''- ;ii'v :; `. ill ;-/3 it!!i:[7 - FOR IN Y iR N A L -ZQ- R E V IE W O N L Y l;i $ . EPA Over-sight on D ioxin Contam ination, Hearing before the Subcommittee on N atural R esources, A g ric u ltu ra l Research and Environment o f the '*" Committee on Science and Technology U.S.\House o f R e p re se n ta tiv e s, F i r s t . ieS iS io n N u m b er 8 9 , Msfcl); ,2 3 1 9 3 3 ^ USj Qcverriment.P r i n t i nc^.Office, Washington, ` - 16. Townsend, e t a l " S u r v e y o f Reproductive Events o f Wives o f Employees 1 Exposed to Chorinated Dioxins" , Amer. J o u r , o f Epidem ., V o l. 115, No.- 5 , pg. 695, 1982. . ' , .V : - - i:?i;V-; ^ I T . ; t p o k .M . " TCDD Em issio ns fo r Municipal Waste Combustors1' , memorandum, U ,S . EPA, December 1 6 , 1983. ..vi- vy yifv y '7 w- ly - . I- '`I.' r:,;i-i 4 " J, 16. U ..s;, EPA, Region V , "A Report on P o ly ch lo rin a te d D1benzo-p-Droxins (PCDDs) ,J* '* apd TJolychlojrfnated Dibenzofurans (P C D Fs): A Summary of. S tu d ie s conducted '<*%[' 1p the G reat Lakes A re a, J u iy 1 9 8 1 .' 'W" : .'!" :J: y I "; 19. National Clean A ir C o a litio n ," Major In d u stria l Sources of Potent!al Toxic A ir P o llu ta n ts", A p ril 16, 1982. .V' ,:y& ; z u . Ssmmiixthn, Fr . anda Ww. Tihnommpson, "Fisherman o f the T itta b a w a sse e ", Environment, * V61. 2 6 , No. 5 , 1984. 7v...xryf' ">. y y.y^Y!' --v - .-V'-i;V-? , f - S r y fm ,w m k y --I/-'!. :;V" ' S Y /.'v ` S'. l fy.7y 'v : .Miy .. - . I . . .. 25699 ,2 oC 4 7 . + T3?-f 'iriSSi ' ..- - > : 2378-TCDO ` . ,V.- -- ;.- : ^ DOW CHEMICAL- D LA N D aANT -- '' - . ' . IN*PLANT SURFACE SO IL SAMPLES ; Field Id en tificatio n . Location ' 2378-TCDD (DL) _ j (ppb) Station 1 Station 2 Station 3 Station 4 S ta tio n s Station 6 Station 7 "'Sta tio n 8 Station 9 Statio n 10 S ta tio n 11 Statio n 12 S ta tio n 13 Statio n 14 S tatio n 15 South o f 492 B u ild in g ' 0. I8 (0.003) South o f 1005 B u ild in g ; Southwest o f 703 B u ild in g 0*074 (0.005) South o f 703 B u ild in g 0 .4 2 (0 .0 1 3 ) West o f 703 B u ild in g 0.020 (0.003) Southwest o f 956 B u ild in g ; East o f 703 B u ild in g 3 .5 0 (0 .0 3 9 ) Northwest o f 1159 B u ild in g ; North o f Shot Pond 0*13 (0 .0 1 3 ) and J S tre e ts - Northwest Corner 4.60 (0.083) 8 th and 6 S tre e ts - Northwest Corner a t Steam P ip e !in e ; 0 .1 5 f . (0.01 ) " N o r t h w e s t o f 1050 B u ild in g at F S tre e t ' o .o io Co.003) South o f 543 B u ild in g ; West o f 14th S tre e t =0.045 ( 0.005) 16th and G S tr e e ts - Southwest Corner 0.44 (0.016) 16th and G S tre e ts - Northwest Corner ;; . . . 0.46. (0 .01 2) 17th and G S tr e e ts - Northwest Corner . :- 0 .2 2 ` (0.007) West o f 934 B u ild in g . 0,27 (0.007) South and East o f 874 B u ild in g North o f RR Tracks - 25.0 (0.50) [36.0JR (0.28) ^ V i. ik t Recovery X S o lid a . 71 92.9 52 96.2 90 96.1 63 95.0 100 96.4 103 ' * 98.5 101 96.3 '65 _ ; 99.1 61 90.2 80 : 99.3 ; - 119 99.6 94 . - .9 6 .0 \ ; 51 99.3 .100 100.0 % f 66 '- ; 85.5 (1 ) 2378-TCDD co n cen tratio n s and d e te c tio n le v e ls (01) reported in parts per b i ll ion (ppb). (2) .1.Recovery - Recovery of internal standard.:-.1.^;^;2378-TCDD or 2378-TCDO) expressed as percent. (3 ) I Sol id s - S o lid s content o f sample determined a f t e r sample homogenization, expressed as p ercen t. A n a ly tic a l re s u lts not adjusted fo r moisture co ntent. (4 ) [ ]R Repeat a n a ly s is o f same sam ple. ^ % ; TABLE 2 ^ S W J - ' ^ . 'V J$yK-: : 2378-TCD S IT E n -M IDLAND, MICHIGAN AREA SURFACE SOIL SAMPLES .Sample . F ie ld i!/Number Id n ti f i c a tio n -- *'" ' * . I k I -- -- -- . "Upw ind Areas L o c a tio n 2378-TCDD (D L) (PPb) % Recovery X Solid s 13354 UPW-l-L 113343 UPW-1-1 >*13401 :i UPW-2-L 7 1 3 3 9 5 . UPW-4-L .`" Y -' * > * M . f. UPW-4-D ND .0.005 NO ND 0.009 0 .0 0 2 ) 0 .0 02 K 0 .0 0 4 ) 0 .0 0 4 ) 0 .0 0 1 ): 82% 76% 90S 84% 84% iV|t r a c ie Oui and P e rim e te r Samples - Dow Chem ical-M idi and P la n t ) v , * -, II3 5 ? ':/ ' T0-4-G ; T0-5-G West - RR a t P o s e y v llle Rd. 0.011 (0 .0 0 3 ) North - A u s tin S t . and 0.25 (0 .0 1 8 )* Jefferson Ave, * T0-9-G E a s t - Saginaw Rd near Gate 25 0.014 (0 .0 0 2 ) : PER-2-L 0.31 (0 .0 1 4 ), 41?7* PER-2-3 PER-5-L 0.069 (0 .0 0 3 ) 0.21 (0 .0 0 8 ) j,g 3 4 $ r/.-. PER-8-L North - Washington S t . and Bay C it y Road 0.010 (0 .0 0 5 ) ;^ 3 3 9 ;':: o PER-9-G # 4 1 8 V ' PER-10-L Ea st - South Saginaw S t . 2.03 (0 .0 4 2 ) 0.040 (0 .0 0 2 ) 52% 21% :y 68% : ' 84% 55% 62% 104% 72% 73% 94.97! 98.63; 94. OX 98.1% 8 4 . 2% 99.3% 99.0% 99.0%, 82.8% 99.0% 99.2% 99.0% 95.6% 97.1% P u b lic Use Areas 13362 '* 1 3 3 5 4 13374;: i ^13392 : 713393 ^13340 ;it3 3 7 5 # 1339)7 P -l-L P-2-L P-5-L P-6-L P-7-L p-a-L P-9-L P**10-L '1 1 3 3 9 4 , : P -ll-L ! ^ "R e sid e n tial Areas \ I #13007 ; * 13003 13101 13102 : A -l-L A -l-1 A-3-L A-3-1 Sugnet School Wallen Park County L in e Roads Mapleton School Longview School Parkwood Park V irg in ia Park Central School (b a ll diamond) B u llo c k Schoool , 0.019 (0 .0 0 2 ) 0.028 (0 .0 0 2 ) 0.003 (0 .0 0 1 ) 0.015 (0 .0 0 3 ) 0.078 (0 .0 0 3 ) 0.17 (0 .0 0 6 ) 0*076 (0 .0 0 3 ) 0 .0 2 (0.003) a \io , . Grfr (0 .0 0 2 ) . 64% 64% 104% 86% 88% 94% . 58% 92% . 81% 99.5% : 89.7% 99.7% 80.2% 98.8% 96.5% 89.2% 96.6% 100.0% ' A *v- 25701 a o cWTf m rr TABLE 2 `(co n tin u e d ) vs*:^ 2378-TCDD S IT E g l - MIDLAND, MICHIGAN AREA SURFACE SOIL SAMPLES F ie ld r 'ij!.I/ld ieWni!"t}?i*f*9iIc" a"t io' 'nT* L o c a tio n lO e f t 'd e n t i V f A reas (c o n tin u e d ) 'ftJ-: -----------`-1............................. ' I II mm* B -l-L IIIS; B -l-1 f l l B-3-L B-3-1 B t-4-L 0*4-1 2378-TCDD (D L) (PPb) 0.076 0.16 0.020 0.27 (0 .0 0 7 ) (0 .0 0 6 ) (0 .0 0 1 ) (0 .0 1 3 ) ; % Recovery X Sol id s 100% 63X 71X X- ' ' 74% . ' 93.2% 97*0%97.8% 87.1% C -l-L C -l-1 C-3-L C-3-1 Cr,4-L C-4-1 0.026 (0 .00 1) 0.054 (0 .0 0 3 ) 0.012 (0 .0 0 1 ) 0.24 (0 .0 1 5 ) 86X 64% 100% 71% 78.7% 84.4% 93.6% 98.1% m $ m A: fn $ 2 t : ! T O 3 l ' 1: -J.l33f)4 113 Y133& D -l-L D -l-D D-2-6 D-3-L Dr3-1 '' E - l- L E -l-l E-3-L E--3-1 F - l- L F - l- 1 HD (0 .0 0 1 ) 94% 0.024 (0 .0 0 1 ) 100% 0.028 (0 .0 0 1 ) 92% 0.018 (0 .0 0 1 ) 100% 0.031 (0 .0 0 4 ) : : ,/ 86% . 84.0% 99.5% 94.9% 90.9% 99.8% 0.026 (0 .0 0 3 ) 0.049 (0 .0 0 1 ) 0.009 0.Q01) 0.020 (0 .0 0 1 ) ; 96% ' 80% . ! ' 75% 70% 97.5% 97.0%. 97.6% 99.1% 0.031 TO. 001) 0.013 (0.001)3 ; 73% " 92% 99.0% 96.8% aneous Midland 0.021 (0 .0 0 8 )* 49% 88.7% - :; fe'wAw;.,'. If S ' 2378-TCDD c o n c e n tra tio n s and d e te c tio n le v e ls (D L) repo rted in p a rts per b i l l i o n (p p b ). Y- X1Recovery r Recovery o f in te r n a l standard (Cl37 2378-TCDD o r 13c 2378-TCDD) expressed as. percent.. X S o lid s - S o lid s co ntent o f sample determined a f t e r sample hom ogenization expressed as percent. A n a ly tic a l r e s u lt ; not adjusted fo r inoisture content. F ie ld id e n tific a tio n o f sam ples: L o c a tio n Type UPW - Upwind TO - Track Out PER - P e rim e te r P - P u b lic Use L - Y ard , la w n , o r open area com posite 1 o r D - pownspout o r d r ip lin e co m p o site: r m o G - Open area grab sample `-tybS^O y4h$r-IT`,^',li"' -\ A- \v " ' TABLE 4 '. IV: CJ/O-ILUU S IT E #3 - MIDDLETOWN, OHIO AREA . SURFACE SOIL-SAMPLES Sa m p l e 7' F ie ld :; .; ^Number ; I d e n t if ic a t io n L o c a tio n 2378-TCDD (D L) - 2 (PPb) Area" >. Ham11ton , Ohio UPW-5-L UPW-5-L 0.004 (0 .0 0 2 ) 0.003 (0 .0 0 1 ) <sfr % Recovery. % Solids 98% 106% 79.3% 79.9%' W^er^lirt^ter - ARMCO, I n c . , Middletown P Ia h t :PS:1,........ PER-12-L PER-1Z-1 PER-13-L SBM P $2 '/If; 1376 PER-13-D PER44-L PER-15-L :StS *pS-' f t -lf:;i3 3 3 8 T" 13339 ' fv-!13$90 " v.<. 1*' Areas G -l-L P -ll-L P-12-L P - 1 3 - L P-14-L P--15--L P--15--1 HD 0 .0 0 2 ) ND 0 .0 0 1 ) 0.004 0.003) ND 0 .0 0 2 ) ND 0 .0 0 2 ) ND (0 .0 0 2 ) 70%. 88% . 94% 7fir. > 106% 76% 87.4% 88.4% 86.4% 87.4% 84.9% " 88.4% HD (0 .0 0 1 ) HD (00.0( 0 2 ) 0.004 (0.001 HD (0 .00 1 ND (0 .0 0 1 ) HD (0 .0 0 2 ) HD (0 .0 0 2 ) 94% 72% 82% 96% 96% 106% 82% 77.5% 87.5% 87.5% 83:2% 82.0% 78.8% 67.3% i$f R e s id e n t ia l1 A reas ^ `I13'S36 ' * 13348 #13355 ;# li4 i)3 1 # `133$1 iM 13351 i #;-$1. 3 3 $ 0 " I: G-2-L G-2-1 H -l-L H -l-1 K -l-L K -l-I K-2-L I# " !I Wm;;-.>:.. iscellaneo us 14189 SIudge 0.004 (0 .0 0 1 ) 0.005 (0 .0 0 2 ) HD (0 .0 0 2 ) ND (0 .0 0 2 ) >. ND (0 .0 0 2 ) HD (0 .0 0 3 ) HD (0 .0 0 2 ) 92% 84% 78% 70% 86% 54% 86% 87.7% 86.7% 95.9% 94.7% 81.3% 88.3% 86.4% Middletown Sewage SIudge ND (0 .0 3 1 )* 6% . v 92.3% 2570-3, j o 1? I TABLE 4 (continued) 2378-TCDD SITE #3 - MIDDLETOWN, OHIO AREA SURFACE SOIL SAMPLES draft (1 ) 2378-TCDD co n cen tratio n s and d e te ctio n le v e ls (DL) reported in parts per ' b i l l ion ( ppb) (2 ) % Recovery - Recovery o f in te rn a l standard (Cl37 2378-TCDD o r 2378-TCDD) expressed as percent. (3 ) X S o lid s - S o lid s content o f sample determined a f t e r Sample homogenization, ! expressed as percent. Analytical results not adjusted for moisture content. (4 ) Fie ld id e n tific a tio n o f sam ples: lo c a tio n Type UPW - Upwind TO - Track Out : PER - Perim eter P P u b lic Use A-K - R esid en tia l L - Yard, law n , or* open area composite 1 or D - Downspout o r d r ip lin e composite 6 - Open area grab sample ;: * Data not v a l i d . Q u a lity assurance o b je c tiv e not a c h ie v e d . i ".'V <) <KDr o": j a DRAFT Sf7^ .'.vy^y;- '\1 . .;., .= -"'fv va .'. ; v^t1?^^V.'A:'.,'i!,::A.i '.:;. #X'1.>;V-v':.V*.s.A. '"''' ' TABLE 5 0LV07/0U- TIr\sr\xJnU SITE #4 - MINNESOTA NATURAL AREAS . SURFACE SOIL.SAMPLES is! . Sample F ie ld 'It li Number Id e n t if ic a t io n | f e ' ' V A a --' 13373 IW D-l-L W f ' . w - "- : .a p ^ 3 f 8 ; ^ :V BSP-4-L A! a ,A % M iS .A 'IV Location . Itasca Wilderness Bluestem P r a ir ie K e ttle R iv e r $ 2378-TCDD (DL) .. . .(PPb) .. - ND (0.003) ND (0.001) * ND (0.002) % Recovery 642 982 1062 2 Solids 94.72 98.02 79.52 .......P...e...m....b...i.n...a......T....r..a....i..l......P...r..e...s...e...r..v...e..........................N...D......(.0 .0 0 2 ) 802 a, &';.,?"V:- ' .-f i>v.-VAJ'>-.;'.:..;-,\r. : " . ^ . ' A , , .. A , ' r ; . \ V . ..* - A . ' ,` . S i A M A :A ; A V , ; . . ' : ' . ' ' A ' A . , V . . \ v ,,: 1' : / v A : . .A ' ..' 53.52 :^A-A'v|i' i, :- ..yA :'. . - ,; ' . . ; '?|;^ )i?378-tCD D concentrationsand. -'le ve ls'(D L) reported In parts per v,>; Ai.^, . b illio n (ppb). '-X L . i-3: 1sti{ 2 Recovery - Recovery o f in te rn a l standard ( C l3? 2378-TCDD or * 3C 2378-JCDD) expressed as percent. %Solids - Solids content of sample determined after sample homogenization, 'Mv. - , expressed as percent. Analytical results not adjusted for moisture content. Field identification of samples: /, /i (i "' i ; 1 iVv ; r::; -My-]' .. W'-/ . ' ` a. -: - , A3--: n ,-" 'ft'-.s.; 'M' ' Location UPW - Upwind 1 TO Track Out : PER Perim eter P Pub! ic Use A-K>j R esid ential 'V 'It- .V- %-w ` Type . ' Y ard , law n , o r open area composite , 1 or D.V Downspout or d r ip lin e composite /- Open area grab sample *'?r ft ffisBit.-'-';: v ! * v i - ' .' 1r iB R i- S l' -,.. :1'a;i, '; :fy v\ i\ . . . . . ' ^ ' A- jr. 's V i ,!i ': }.<>' i1; ; f:,i 1 2 5 7 0 ij o s3> \*t+ 6 t U .7 5 .5 ,, '.'-r TABLE 5-A :' . 2378-TCDD Control and Blank Samples draft ' <* .5 - - T . Sampi e Number 14196 14184 13368 13308 ; 13320 t 13333 13344 ' ; 13344 ; 13357 13380 ; 13396 : 396 13409 ; 13385 14183 14195 > 13309 ! 13321V 13332 13345 13356 13369 13381 13397 13408 13414 13415 13416 - 13417 13384 Field Id en tificatio n 2378-TCDD (DL) Control Soil Control Soil Control Soil Control Soil Control Soil Control Soil Control Soil Control Soil Control Soil Control Soil Control Soil Control Soil Control Soil Control Soil ND (0 .0 0 6 )* 0.013 (0 .00 4) 0.006 (0 .00 2) 0.032 (0 .00 3) 0,012 (0.00 3) b ND (0 .0 1 0 )* ' ND (0 .0 0 7 )* 0.015 (0 .0 0 5 )* ND (0 .0 0 1 )* 0.005 (0 .00 3) 1 f ND (0 .0 1 9 )* - ND (0 .0 0 7 )* ND (0 .0 5 4 )* ND (0 .0 4 2 )* Blank Soil ND (0 .0 0 2 ) Blank Soil ND (0 .0 0 3 ) Blank Soil ND (0 .0 0 1 ) - Blank Soil 0.CC2 (0 ,0 0 1 ). Blank Soil , ND (0*002) Blank Soil ND (0 .0 0 2 ) B1 ank S o il : ND (0 .0 0 1 ) Blank Soil ND (0 .0 0 2 ) Blank Soil ND (0 .0 0 1 ) Blank Soil ND (0*001) Blank So il ND (0 .0 0 1 ) Can Blank (P ER -5) ; ND (0 .0 2 7 ) Can Blank (In - P la n t ) ND (0 .0 2 4 ) Can Blank ND (0 .0 2 5 ) (Middletown) Can Blank #6 ND (0 .0 3 0 ) Blank Soil ND (0 .0 0 3 ) % Recovery % Sol ids 46% 59% 66%i 58% 50% -,< 40% r 34% r 22% 23% 66% ; 10% , 19% 22% 13% 97% 81% 83% : 80% 7 73% 96% ; 100% 70% 100% 106% 84% 102% 100% 96% ' i: - ' 7' 98.4% 98.4% 98.4% 98.4% 98.4% 98.4% 98.4% 98.*% : 98.4% 98.4% 98.4% 98.4% 98.4% 98.4% 53.7% 53.7% 53.7% 53.7% 53.7% 53.7% 53.7% 53.7% NR 53.7% 53.7% ' A-'.; ' .v ' -7". :7 : v7.-,.-v-, -'1 \[ ' 'Z"- 7--v; '7'.: , V'' 80% 103% 53.7% Notes : (1 ) 2378-TCDD co n ce n tra tio n s and d etectio n le v e ls (D L) reported in p arts per b ill ion (p pb). (2 ) % Recovery - Recovery o f in te rn a l standard ( C l 37 2378-TCDD or 33c 2378-TCDD) expressed as p ercent. (3 ) % S o lid s - Sol id s content o f sample determined a f t e r sample homogenization, expressed as percent. A n a ly tic a l re s u lts not .^ adjusted fo r m oisture content. (4 ) NR - Not re p o rte d . ,v rrirV ... 2.570b- * Data not v a l i d . Q u a lity assurance o b je c tiv e not a c h ie v e d . 5.0 0 5 ^ Sample N o.: Field ID .: 13401 - HJPU-2-L s i t e h *M m M t t i J 'M c m m i a rea SURFACE 5 0 I t SAMPLES l Upwind : * -- .? ; Dow Chemical - -In-Plant 13395 UPW-4-1 13406 ' Statio n 5 1341E ; " T S ta tio n 14 ; lo ca tio n ; PCOOs (OU 2378-TCDD To tal Is o TCDDs T o ta l penta CDDs . To tal hexa CDDs ' To tal hepta CDDs OCDD . ND (0 .0 0 4 ) NO (0 .0 0 4 ) NO (0 .0 2 4 ) ND (0 .0 2 4 ) 0.15 ( 0.024) 0.34 (0.026) -Incinerator - West bf 934 Pullding ND (0 .0 0 4 ) ND (0 .0 2 3 ) ND (0 .0 2 3 ) 0.17 (0.034) - 0.33 (0 .03 4) 3 .5 (0.039) . ; ; 0 .2 7 0.32 0 .2 4 4 .0 7 5 .0 3 7 5 .0 (0 .0 0 7 ) (0 .0 6 7 ) (0 .0 6 7 ) (0 .9 ) (1 .3 ) PCDFs (P L ) ; 2378-TCDF To tal TDFS Total penta CDFs Total hexa CDFs Total hepta CDFs OCDF ND (0 .0 0 4 ) ND (0 .0 0 8 ) ND (0 .0 2 2 ) ND (0 .0 3 1 ) ND (0.0511 ' ND (0 .0 0 4 ) ND (0 .0 0 8 ) ND (0 .0 2 3 ) ND (0 .0 2 8 ) ND (0 .0 4 5 ) 0 .4 5 (0 .0 6 ) Notefh (T ) 4-i . ic i O C o n cen tratio n s o f PCDDs , P C D F s , and d e t e c t i o n le v e ls (DU) reported in p a r t s p er b i l l i o n ( p p b ) . - . 0.027 (0 .00 7) 0 .9 0 3 .1 .1 5 .4 8 .6 (0 .1 4 ) (0 .1 3 ) (0 .38 ) (0 .48 ) :;ig ip Samp! e Ho. : Fiel d ID .: 13374 P-5-L . v" . ,*%- "-. TAnBtLrE 6 jV"'.^...^ _v ~ > ' : ;V -'rT^ S '^ - 'ttD D s and PCDFs ;' . S IT E S i - MIDLAND, MICHIGAN AREA SURFACE SOIL SAMPLES '..\-` 13392 P-6-L P u b lic Use Areas 13393 P-7-L 1337.5 P-9-1 13391 P-10-L 'v-".ii-MV I 13394 P -il-L Lo ca tio n :: DPs (P L ) .; 78-TCDD t a l i s o TCDDs t a l penta CDDs t a l hexa CDDs t a l hepta CDDs DD 0.003 (0 .0 0 1 ) -=. NO (.0.001) *=' :K B t {0 .0 1 4 ) i O . 067 (0 .0 0 7 } vO . 35 (0 *01 3 ) 3 ;1 (Q;.;096) 0*015 (0*003) cr* 040 ^ ND {0*035> 0,063 (0 .0 3 5 ) 0.38 (0 ,0 2 8 ) 0.86 (0 .0 2 7 ) 0,078 (0.003) oar ; . in te rfe re n c e . 0.34 ( 0.02) 2 .3 (0*055) 7iO (0.068) 0.076 (0 .0 0 3 ) 0 .2 9 -:;V -0.10 (0 .0 1 8 ) 0.24 (0 .0 1 8 ) 0.41 (0 .0 9 3 ) 12.0 (1 .5 ) 0.012 (0 .003) 0.040 - ND (0 .0 3 4 ) 0.086 (0 .01 4) 0.35 (0 .031) 0.68 (0.031) 0,11 0.22 0,12 0.41 2 .4 7 .0 (0 .0 0 2 ) (0 .0 2 2 ) (0 .0 2 2 ) (0 .0 4 2 ) (# 052) OFs (P L ) . ;W 78-TCOF : ; tal-TCDFs tal penta CDFs tal hexa COFs, tal hepta CDFs of ^ '- ND (0 .0 0 2 ) NO (0 .0 0 2 ) HD (0 .0 1 ) -: ND (0 .0 1 ) .65 ( .0 2 ) 0.044 (0 .0 2 3 ) ND (0 .0 0 5 ) 0.013 (0.007) NO (0 .0 0 8 ) -D ( Q .024) 0.14 (0. 43) 0.10 (0 .0 7 1 ) - :r NO (0 ,0 2 5 ) M lp (# & ): 0.72 (0 .0 2 f) ' 0.64 (0.037) te s: (1) 48-Sp Concentrations o f PCODs, PCDFs and detection Is { DL) re p o rte d in p a rts per b i l l i o n ' (p p b ). 0.013 (0 .0 0 2 ) .s 0.040 (0 .0 1 ) 0.064 ( 0 .0 1 ) 0.50 (.34) 0.37 (0 .0 4 9 )' ND (0 .0 0 5 ) ' 0.015 (0 .0 0 3 ) ND (0 .0 0 7 ) NO (0 .0 2 9 ) 0.16 (0.062) 0.11 (0 .07 0) 0.11 0.17 0.82 0.66 (0 .0 1 7 ) (0 .0 3 7 ) (0 .0 4 5 ) (0 .0 4 5 ) Sample No.: Field 10,r Location*,: PCDDs (DL) " 2378-TCDD " Total iso TGDDs Total penta CDDs Total hexa CDDs Total hepta CDOs GCOD Upwind 13377 13400 0.003 (0.001) NO 0.001) ND (0 .0 1 0 ) NO (0.010) 0.023 (0.007) 0.20 (0.009) ifr:' r/0.004 (0.002) ND (0.002) 8 0 (0.044) 0.072 (0.044) 0.20 (0.17) 10.6 (0;23) PER-13-L 0.004 (0.003) NO (0.003) ND (0.020) ND (0.020) 0,15 (0.014) 0.28 (0.01*) PCDFs (DL) 2378-TCDF Total TCDFs Total penta CDFs Total hexa CDFs Total hepta CDFs OCDF 0.002 (0. 001) NO (0.008) NO (0.008 ND (0.018 ND (0.025) NO (0.004) NO (0.016) NO (0.036) ND (0.023) ND (0.056) 0-006 (0.004) . ND (0.010) ' : ND- (0.1... ND (O.l - ND (0. 046) Notes: (1) 07 . on *7^ t o .o S ' vrv Concentrations o f K D D s , PCDFs^ and detection le v e ls (D L) reported in p a rts per b i l l ion (ppb)< Perim eter 13372 P ER rl4-l . J!D (0 .0 0 2 ) ND (0.002) ND (0.014) ND (0.009) 0.073 (0.018) 0.17 (0.026) NO (0.001) ND (0.012) ND (0*012) ND (0.017) ND (0.024) 13375 PEJU15-L l :ND (0.002) ND (0.002) ' NO (0.039) .... NO (0.039) 0.12 (0.009) 0.84 -(0.071) , m (0.005) NO (0.015) v NO (0.015) 0.043 (0.023) 0.050 (0.026) _ '- 7 f " v 7 :7 V :7 7 7 . . ; : 4 TABLE 9 . : 7~ ..,7, -- ' - 1-7 : . ' ' - v ' , . . 7 .. . ... ' ; -7 0 -7 7 -7 '>; : f ' ' - ' PCDDs and PCDFs S IT E |4 - MINNESOTA NATURAL AREAS 77 7 SURFACE SOIL SAMPLES Sample No. : Field ID .: L o c a tio n : 13373 IW D -l-L v Itasca Wilderness : ! 13378 ! BSP-4-L ; : ;; Bluestern P ra irie 13379 K k-l-L -, Kettle R iver PCDDs (D L) ' 2378-TCDO T o ta l fso TCDDs Total penta CDOs To tal hexa CDDs Total hepta CDDs OCOD . ... ' : U . L ': . 7 NO (0 .0 0 3 ) ND (0 .0 0 3 ) ND (0 .0 1 0 ) ND (0 .0 0 5 ) 0.047 (0 .0 0 9 ) 0,13 {0 .0 1 9 } ND (0 .0 0 1 ) - * - 7-- NO (0 .0 0 1 ) 7 ; - . " ND (0 .0 1 2 ) , !i : ND (0 .0 1 2 ) ,, 0.091 (0 .0 1 0 ) p . 20 (0 .01 2). .- - T ND (0 .0 0 2 ) ' ND (Q .002) ND (0 .0 0 4 ) ND ( 0.0 11 ) 0.025 (0 .0 0 9 ) 0.092 (0 .0 1 1 ) PCDFs ( D U . 7 7 7 7 7 7 7 , . 2378-TCDF To tal TCOFs : " Total penta CDFs To tal hexa CDFs To tal hepta CDFs OCOF v - VF - . - . - - t * . : ' 7 V - f r t : ' . 7 - - 7 ' 7 .7 ' V , . 7 ; ;7; v > ,77: i % - 7- . 7 7 y . ' ' 7 . ; 7 ND (0 .0 0 2 ) ND (0 .0 0 2 ) ND (0 .0 1 0 > ND (0 .0 1 0 ) ND CO.017) ND (0 .0 2 4 } ND (0 .0 0 2 ) . " - 7 7 : V - ' "77 ' V . : v 7 :.,.; - - ND (0 .0 0 9 ) *' ND (0 .0 0 9 ) . 7 . . ND (0*009) ; ' : ND (0 .0 1 0 ) i--_ ; . ND (0 .0 0 1 ) - ND (0 .0 1 3 ) ND (0 .0 1 3 ) ND (0 .0 0 9 ) ND (0 .0 1 4 ) N otes: (1 ) C o n cen ^ itio n s o f PCDDsy PCDFs , and d etectio n 1e v e l$ { DL) reported 1n p a rts per b i llio n (p p b ). to ui -O: fiSo Q 'C Sample N o.: Field ID .: L o c a tio n : V:^ 13330 Control S o il :ODs (D L) 378-TCDO r 3ta l is o TCODs . 3ta l penta COOs tal hexa COOs ta l hepta COOs ;DD 0.005 (0.003 NO 0. 003; 0.081 0.033; 0.64 (0.033] ; NA r-N A * -OFs (P L ) s78-TC0F t a l TCOFs >tal penta COFs stal hexa COFs ta l hepta COFs :o r' # ; 0 . 017 ( 0 .0 0 5 ) NA NA NA NA # if S s S li^ 8 f e .- : " '' ' ` ' WS< S-ii*'jrV--v";;::- v..... . '........... . .. ., -v* . . , r: ... . . .. r . / '. ' . V - . ` . ^ - v :.. .................. -v - i:^ P C D f ^ - CONTROL AND BLANK SAMPLES - Blank and Control .-Tv- : -;.,,;^Aivc;. '-,-.--' . 1 ' : r . - . ' - . ' * - ---- 13381 * Blank S o il 13414 Can B1ank PER* 5 13415 Can Blank In -P la n t Blank Can Blank ^ ~.' >' i ;' ' ; Blank Middletown 'i*\* ; Can Blank * Blank #6 ND (0 .0 0 1 ) :N0 [0.027) ; ND [0 .0 2 4 ) ND .(0 .0 2 5 ) ,, NO (0 .8 3 ] . . N" Q* ' 0;-QDl) ; ;, r - : N D [0 .0 2 7 ) - i i> 0 .0 1 5 ) - : , o 0 .0 9 7 ) NO 0 .0 1 5 ) ' . -no [0 .0 9 7 ) j , ^ NO [0 .0 2 4 ) ' " ND 0 . 1 0 ) - ' , NO Q. 10) NO (0 .0 2 5 ) , , NO (0 ,0 9 8 ) / NO (0 .0 9 8 ) - NO ( 0. 03; NO ( o . n ; ND (0 .1 1 ] 0.031 0 .0 05 ) ND .0 .1 7 ) ~NB [0.15) ND ( 0 .1 7 ) ND ( 0 .1 9 ] 0.27 0.007) 0 .8 3 .0 .2 2 H c*. . 0.74 ;o.!9> y"%:;VX\`'- r v 0.89 (0 .2 3 ) 0 . AA (0 .2 4 ; " x -t*V......... - - NO (0 .0 0 1 ) NO o . o n ; ND ( o - o ii m ( 0. 012] NO 0,-013] 0 .0 3 9 ) 0.039) 0.021) 0 .3 5 ) NO (0 .0 2 2 ] no ( 0. 022; NO [0.032; ND NO (0 .0 2 3 ) (O J NO (0 .1 5 ) V.-M. 027) 0.074) 0. 12) 0 .1 6 ) 0 .1 6 ) ? te s : ( l ) Concentrations o f PCODs, PCDFs, and d e te ctio n 1evels (OL) reported in parts per M ilto n (ppb) NA * Not a n a ly z e d . Peak b ro ad e n in g fn G C /N S V cn system prevented quanti ta tio n . -rD k -OD L - o c- W -r- . ; * *.-. -.V m . 'j "W}^ ~ ' 1 9 8 3 f ISH COLLECTION > - - : - RESULTS FOR 2,3,7,8-TCD D ; ,v . - 'V ; ; x . Sample Id e n t if ic a t io n EPA ? MDNR Speciesr fish / Sample Sample Type S t , Joseph R ive r a t B e rrin Springs 13218 SJR-1W Carp . .v 5 WHL Huron R iv e r a t F la ir o c k 13281 Carp R iv e r R a is in a t Monroe - i - i- v' ;; ;> . -6 WML j ;t;7 ^ ^ 13282 RR-4W Carp Cl In to n R iv e r a t Ht. CTemeiis . 3 WHL T-r/a'-.v -`'.T 13283 CR-6W Carp ' 1 r WHL Kalamazoo R iv e r a t A ll egIn r 13284 KR-4W Carp _ WHL Pi-ne R iv e r a t Alma ..............- . 13286 TR-1W Carp Muskegon R iv e r a t Rogers Dam ` * ' 5 . WHL _i?. '. r >>~~-7-v-r%<r,?r->L 13287 UMR-1W Carp ; . -5 'v i WHL S t , C la ir R ive r at Alqonac 13288 SCR--1W O - a W-A O /(OT*, :COTv v;r-. Carp ' >; 5 WHL Length; W : *r- Height . / ._ J E l_ 35-37,5 .... 460-700 32-38 ' 400-780 59-64,5 - 3300-5020 65 4400 35-41 680-1000 54-59 2000-3580 40-43.5 51-54 940-1100 1920-2480 L ip id Content (<) > -'r ^ 2,3*7,8-TCDD ' . z4 ppt ( o u '.._ r : . 3. 6 - .3,1 (2 .5 ) 5; a v 0,6 (0 .2 ) '8 .7 5.7 (0 .2 ) 6.8 2.4 (0 .1 ) 2.1 2 .8 (0 .7 ) 7,3 ; 8.6 (0 .4 ) KB (< 0,8 ) 4.3 (0 .3 ) Ausable R iv e r a t Hio 13289 Carp ...' .. ... . 4 WHL 57-72,5 2680-5370 \ NO 0 . 8 } Shiawassee R iv e r a t Chesanlng 13290 ; >5 52-68 2360-4840 ND (1 .2 ) Muskegon Lake a t Muskegon 13420 13422 13421 ML-2A HL-5W ML-5A Carp Pike ' Pike % - r ~ \ . 8 ^ \ SF-R -3 - , : #HL I: 54-66 57-63 55-76 f : 2040-4420 1140-1720 920-3080 ` i ' 5 .2 ( 2 . ) 5,3 3.9 (1 .6 ) - NO (2 .4 ) Grand R iv e r a t Grand Ledge 13245 GR-1R 13219 GR-1 13220 GR-2 13221 GR-3 13222 GR-4 13223 GR-5 13224 GR-8 13225 GR-7 13226 GR-8 13227 GR-9 13228 GR-10 13229 GR-11 13230 ^ GR-12 13231 ^ GR-13 13232 z i " GR-14 13233 GR-15 Carp 5 - VHL 71'Carp 1 SF-L Carp V- 1 SF-L Carp ' 1' SF-L Carp 1 SF-L Carp 1 SF-L . Carp 7 SF-L ^ Carp l . SF-L . Carp , . SF-L Carp 1 SF-L Carp 1 SF-L Carp ' 1 SF-L v.-; Carp 1 ' SF-L Carp 1 . SF-L : Carp - .- v i SF-L - Carp 1 SF-L 3 4 .5 - 3 9 .5 .3 0 31 31.5 31.5 27 32 29 - 29 40 40 42.5 41 38.5 36.5 37 680-1040 400 420 510 540 300 510 380 410 1010 1020 1140 1140 880 770 830 5 .4 1 .9 --. -- w tm mm * m m . W 2 .9 1.9 2.-2 3.7 1.6 . ;-- i ' 0 .6 6.0 (0 .6 ) ND (2 .5 ) 6.9 (2 .5 ) : ; NO (2 .5 ) ND ( 2 .5 ) ND (2 .5 ) , ND (2 .5 ) NO (2 .5 ) ND (< 0 .4 ) 3.7 (0 .3 ) 0.4 (0 .2 ) 7>2 (0 .2 ) 3.5 (0 .2 ) 1.5 (0 .5 ) ND ( 1 .2 ) 0.6 (0 .3 ) ki-L. Sample Id e n t if ic a t io n EPA MDNR Species f i s h / . ^Sample Sample Type * Length (cm) .Weight (gm) Content - 2*3,7,8-TC 00 ... ( ? ) - - - ?-' ppt (DL) ^ Grand R iv e r a t Grand Ledge {con *t ) 13234 13236 13237 13238 13239 13240 13241 1324Z 13243 13244 13246 13247 GR--I 6 GR-17" GR-18 GR-19 GR-20 GR-21 GR-22 GR-23 GR-24 GR-25 GR-5A GR-6A Carp 1 SF-L Carp - -1 SF-i. Carp 1 SF-L Carp Carp Carp 1 - 1 - -1 - SF-L SF-L ' SF-L Carp Carp Carp . - ,A l ' -l SF-L "'-SF-L SF-L' Carp l SF-L Bullhead . 5 . SF-R Bass - ;-5 . .SOF-R 36 - 53 . .57 55 54 <. 40 45.5 I -46 " 24.5-29 ; 28-36 , _ 700 2190 ' 1600 2940 2900 2940 1100 ,1400 , '1540 660 240-340 : ; 280-690 wtmm . ~' 8.1 ` -1 ,2 --- ' 1 .3 1.7 ..... 1 . ND ( 0 . 7 ) NO , ( 0 .5 ) - ' l l (0 .9 ) 1,8 (0 .3 ) ND ( U S ) ND ( 0 .4 ) - ND 2 .7 ) ' ND ( 1 .5 ) ;1 .2 (0 .5 ) - 12 ( 0 .3 ) ND 0 . 2 ) ND 0 . 2 ) - v;:-\ ,-;r . T1 ttabaw assee R iv e r a t Midland 13273 13248 13249 13250 13251 13252 13253 13254 13255 13256 13257 13258 13259 13260 13261 - 7R-1W TR-1 TR-2 TR-3 TR-4 TR-5 TR-6 TR-7 < TR-8 TR-9 TR-10 TR-11 TR-12 TR-13 TR-14 "'iw.-/"**C>- t - . uo V 3 Carp 5 WHL Carp 1 SF-L Carp 1 SF-L Carp 1 SF-L Carp 1 SF-L Carp 1 SF-L Carp 1 . : SF-L Carp 1 - SF-L Carp 1 . SF-L Carp 1 SF-L Carp 1 SF-L Carp 1 SF-L Carp 1 SF-L Carp 1 SF-L Carp 1 SF-L v; 34-58 37 39 37 -- 40 42 -.37 35 ' 36,5 43 .. 45 41 43 4L 5 43 500-2860 ' 700 840 - 780 860 850 720 ' 560 640 1040 1240 880 1030 1060 1070 10? : 4.5 1.7 2. 7 5 .3 -3.-5 1 .2 1 .8 0 .9 6.. 6 1 .2 3.8 4 .3 6 .4 4 .9 i 190 38 18 28 27 , 16 . 16 12 18 45 38* 13 14 26 22 (0 .6 ) (0 .4) (0. 3) (0 .2 ) (U 5) (0 .3) (U 0) (0 .6) (0 .2 ) (0 .3) 0 .6 ) (0 .3 ) (0 .6 ) (0 .4 ) (0 .6) s' -- 'r . ~-. i V* -j M B rz$* U W lflK - ^ ( c o n t i n u e d ) <- Sample Id e n t if ic a t io n EPA MDNR Species F ish / Sample Sample > Type *: ^Length (cm) Tlttabaw assee R iv e r a t Midland (con' t ) 13262 13263 13264 13265 , 13266 13267 13268 13269 13270 13271 13272 13274 13275 13276 13277 13278 13279 13280 ' TR-15' , r TR-16 TR-17 . TR-18 TR-i 9 TR-20 TR-21 TR-22 TR-23 TR-24 TR-25 TR-4A TR-6A TR-66A TR-67A TR-68A TR-69A TR-70A Carp 1 SF-L Carp :1 Carp 1 SF-L Carp Carp i; : . .^1 SF-L SF-L , Carp ; c SF-L Carp 1 ' - ::*'SF-L Carp - " 4 ^ 1 ~ ' SF-L Carp ,1 SF-L Carp 1 ; SF-L Carp 1 SF-L Ch. C a tfish 5 SF-R Sm. Bass 5 SOF-R Walleye 1 SOF-R W alleye 1 SOF-R. Walleye 1 SOF-R Wal1eye 1 SOF-R Walleye 1 SOF-R -. '' 41 ' 39,5 63,5 64,5 5 1 ,5 48 47 44 51,5 49 46,5 45-55 24-38 56 52 49 39 38 Sample Types SF-R - S kin less F i lle t - Right Side SF-L Skinless F ille t - L e ft Side SOF-R * Skin on F i l l e t - R ig h t Side SOF-L * Skin on F i l l e t - L e f t Side WHL - Whole F ish o Of frytv iODL cn CT ... * D etection L im it Weight (<p) L ip id Content (*) 2,3,7,8-TCOO PPt (DL) '" S im 1000 820 3300 3450 ... 1820 - 1560 1500 1320 ' 1680 1540 1380 800-1760 220-860 , 1580 1320 1070 570 500 - 7.9 46 (0 .6 ) 3.4 (0 , 7) - v 8 .0 : : 34 (1 .2 ) 2.3 12 (1 .6 ) , 4 .0 . 24 1 .6 ) 3.3 25 (1 .4 ) " . 4.6 . 90 (0 .3 ) : ,1 1 . .. 530 (1 .6 )- 3 .6 41 (1 .3 ) 3 ,6 70 ' (1 .4 ) : 1.4 17 (1 .6 ) 7.3 V 75 (0 .4 ) - i 1 .0 5.1 (0 .2 ) a 1.4 5.1 ( 0 .2 f / :) 2 .0 3.6 (0 .4 ) V! 1.3 3 .9 (0 .2 ) : 0 .6 2.8 (0 .3 ) j 0.6 V 4.1 (0 .3 ) i '&s-.'-'-vf>.^>- -fr-flr*:" ft^ ^ ^ 0 0 0 ***'Vi.'X/i-tT-;'-'-/vy\ '.' MICHIGAN IOXI^STtoe Y^v AVIAN SAMPLES RESULTS FOR 2,3,7,8-TCD O . -> -- : Sample Id e n t if ic a t io n EPA FWS Species I - : - ... . . -": : ' - Sample D e scrip tio n :.' L ip id Content (i) - : 2 ,3 ,7 ,8 -T C D D PPt (DL) SaginawBay near Bay C ity 13291 13297 13292 13293 13294 13295 13296 13298 13299 FWS S8C3-032 FWS S8C2*011,012 ' " FWS S3C3-G12,013,014 ` ~ -v - FWS S8C3-015 ,0 1 8 ,0 1 9 ,0 2 0 ' FWS SBC3-021,024,025,028*028 FWS SBC3-001,00 6,0 07 ,00 8 : FWS S8C3-003,004,005 r r T W S $802-001,002 :^ ^v^FWS S8C2-004 Common Tern - Common Tern ` Common te rn Common Tern Common Tern Common fe rn Common Tern .' . Common Tern ' Common Tern Who!e Body Composite Composite Composite Composite C o m po site. Composite Composite Composite 0 ) Adult (2 ) Chicks - (3 ) Chicks (4 ) Chicks 5 Chicks (4 ) Eggs (3 ) Eggs (2 ) Eggs (4 ) Eggs : 3.1 3..8 3 .3 ; 4.3 4 .4 ` 6.7 7 ,7 * 7 ,6 ; =. 25 ( 3 .1 ) 23 (3 .5 ) 35 ( 2 .4 ) 38 (2 .2 ) 34 (2 .9 ) 40 (1 .5 ) 51 ( 1 .6 ) 63 ( 1 .5 ) 37 ( 1 .2 ) <. * I n s u f f ic ie n t sample to analyze f o r l i p i d c o n te n t. o rc CJT < T jm -.. -* .................................... . .... SSiM I i "'' '`*-r:3V't^-*'?*;- -1*. .-!*>i'y"^;f-.*iv>53 ' 2378-TC00 SUMMARY OF RECENT ENVIRONMENTAL MEASUREMENTS MIDLAND; .MICHIGAN AREA - \ . , -f - -- : :.; '.v ; - - - - - ..... . ..... Ambient A ir * (March 1983-February 1984) *=" Total - a l l measurements Dow P lan t Dow P la n t Fence Line C ity o f Midland Number of Measurements - \ . - - ' Range .5'i.v - T v ' . , ` "A rithm etic r r Arithm etic Standard i;v ^ Mean D e v ia tio n . r ' *- - ' :' <" ;i Geometric > Mean ^ 7 ' - - . -.V.3* 10 2 6. 2 0.010-0.21 P9/n3 0. 19-0.022' pg/m^ O .U lO -0 .21 i.Pg/m3 0 .019-0.16 pg/m3 : . 0.061 0.021 0.064 0.090 0.069 &/M' . : y y y y y y t r} . : : t t y : 0.036 0.075 v 0-038. * Data provided by Dow Chemical Company - . : - v i S - .. .r.. - ; - V ' V - Tittabawassee R ive r Fish (August-September 1983) y > < ~ >. ' Carp - Whole f is h composite (5 f i s h ) Carp - in d iv id u a l s k in le s s f i l l e t s 25 : : C atfish - s k in le s s f i l l e t composite (5 f is h ) i;\ r ym ' \ V . . ' : '-; J iV S i.- Smallmouth Bass - skih-on f i l l e t .composite.~{5-. f is h ) 1 Walleye - in d ivid u a l skin-on f i l l e t s 5 190 ppt 12-530 ppt 75 r ppt 5.1 ppt 2 .8 - 5 .1 ppt . - - . . ^ 50; - ' V ,; 102 . 3.9 0.8 r 5. ' .V 28.6 - ' V" . V ' . :x i 3.8 Surface S o ils (October-December 1983) Dow Plant i Dow Plant Perim eter Public Use and R e sid e n tia l Areas Residential Areas* R e sid e n tia l Downspouts* - / . 0 15 . V 9.,, 17 8 > 8.- 0 .0 1 -2 5 .0 PPb 0 .0 1-, 2.3 ppb 0.003-0.17 ppb 0.009-0.076 PPb 0.013-0.27 ppb " , - y :: 2.38 .0 .7 4 0.037 0.025 0.104 ' ; .. . -= . 6.41 . 1,43 0 .0 4 2 0.022 -0.103 0.24 o.i7 0.026 0.020 ^0.062 r'- - r :~r$ i :' > . * Data fo r fo u r a d d itio n a l $ b _ - " ^ c>4> f c o -'a -; -''-W ' s ite s a v a ila b le 12/15/84. vt - . : - > " fr*v7`^...>^^V,.'.>':?;'- ;%/, - .-.r\ :: K. - ' v JUL 2 1 1980 MEMORANDUM SU B JE C T : Ep id emiologic In ve stigation o f ;Birth Defects in Oregon FROM: - M a r t i n P. Halper, D i r e c t o r Survey and Analysis Division (TS-793) ' /` TO : Edwin Johnson Deputy Assistant Administrator for Office of Pesticides Programs THRU: M a r i l y n C. Brackerj/jr-! C'/J* Deputy Assistant Administrator for Office of Program Integration and Information (TS-793) The Survey and Analysis Division (SAD) recently received the attached proposal from the Center for Disease Control (CDC)t C e n t e r is p r o p o s i n g that O P TS s u p p o r t an e p i d e m i o l o g i c retrospective matched case control investigation in w e st er n Oregon. The object of this endeavor would be to determine w h e t h e r p a r e n t a l e x p o s u r e to a e r i a l h e r b i c i d e a p p l i c a t i o n is associated with the birth of infants having central nervous system (CNS) defects. Th' This proposal was the result of a preliminary assessment of the situation by CDC personnel after they were contacted by Oregon p h y s i c i a n s and o t h e r c o n c e r n e d c i t i z e n s in the area. C D C 1s review of data amassed through their National Birth Defects Monitoring Program indicated that there were increases in rates of certain CNS birth defects between 1970 and 1979. Specifically, 8 cases of anencephaly were, observed with 4 e x p e c t e d ; .16 cases' of h y d r o c e p h a l y w e r e o b s e r v e d w i t h 10 expected. I I - < ,v ...S r e * - * , ^ /-* T. 25188 T0bte> 2 iAccordingly, SAD is forwarding a copy of the CDC proposal for review and comments. We would appreciate receiving all comments b y J u l y 29 so t h a t w e m a y r e s p o n d to. C D C in a t i m e l y fashion. Attachment cc: Dave Menotti, OGC Jim Reisa, OTS 25189 <510 i 'V P DEPARTMENT OF HEALTH, EDUCATION, AND W ELFARE PUBLIC HEALTH SERV ICE CEN TER FO R D ISEASE CO N TRO L A T LA N T A , G E O R G IA 30333 T E L E P H O N E : (404) 633-331 1 404-452-4036 June 18, 1980 Susan Strassman-Sundy Survey & A nalysis D ivision (TS-793) U .S ., E.P.A. 401 M S t r e e t , S.W. Washington, D.C. 20460 Dear Ms. Strassm an-Sundy: As fo llo w -u p to our telep h o n e c o n v e r sa tio n regard in g a proposed stu d y o f h e r b ic id e exposure and b ir t h d e f e c t s , I am sen d in g th e r ep o rt we prepared f o r th e Oregon P e s t ic id e A n a ly tic R esearch Center (PARC) and the S ta te H ealth Department. We f e e l th e proposed stu d y , i f PARC and th e Oregon H ealth Department d e c id e to fo r m a lly i n v i t e CDC to h elp ' conduct i t , i s very much th e type o f coop erative e f f o r t th at would be covered by th e pending interagency agreement betw een SAD/EPA and th e Chronic D is e a s e s D iv is io n , CDC. Our problem as m entioned in our c o n v e r sa tio n , i s th a t we c u r r e n tly do not have th e manpower or th e funds to conduct such a la r g e stu d y . Because o f our mutual i n t e r e s t in h e r b ic id e exposure and h e a lth e f f e c t s , th e p ro p o sa l sh ou ld be o f i n t e r e s t to th e E .P .A . I f EPA fin d s th e p ro p o sa l o f in t e r e s t and cu rren t funding i s a v a ila b le , we sh ou ld e x p lo r e our mutual in te r e s t as soon as p o ssib le in a n tic ip a tio n o f th e request fo r a s s is t a n c e from th e PARC d and th e Oregon H ealth D ept. Sincerely yours, Larry Edmonds, M.S.P.H Epidem iologis t B irth D efects Branch Bureau of Epidemiology 25190 REPORT TO THE PESTICIDE ANALYTIC RESEARCH CENTER (PARC) THROUGH JAMES GOOGINS, M.D. STATE EPIDEMIOLOGIST OREGON HEALTH DEPARTMENT June 10, 1980 L arry D. Edmonds B irth D efects Branch Chronic D isease D iv isio n Center for D isease Control A tla n ta , Ga. 30333 25191 During th e F a ll o f 1979, a group o f p h y sic ia n s and o th e r concerned c i t i z e n s from L in co ln County, Oregon r ep o rted t h e ir b e l i e f th a t an e x c e s s iv e number o f b a b ie s in th a t county w ere b e in g b o m w ith b ir th d e f e c t s . They lin k e d th e se data w ith h e r b ic id e exposure (2 ,4 -D ) and c a lle d fo r a ban on th e u se o f 2,4 -D u n t il more data are a v a ila b le . B ecause o f th e se c o n c e r n s , and the r e s u l t s o f th e EPA Study o f a b o r tio n s and 2 ,4 ,5 - T in A ls e a , Oregon, the B irth D e fec ts Branch review ed th e data a v a ila b le through the N ation al B irth D efects M onitoring Program. That review in d ica ted th a t there were in c r e a se s in the r a te s o f anencephaly in Benton County (8 observed c a se s, 4 e x p e c te d ), and o f hyd rocephaly in Lane County (16 ob served c a s e s , 10 expected ) between 1970 and 1977. The BDMP f i n d i n g s , a lo n g w ith th e c o n c er n s o f th e L in c o ln County p h y sic ia n s le d us -to d is c u s s th e s e d a ta w ith th e Oregon H ealth D epartm ent. John G oogin, M.D, S t a te E p id e m io lo g is t , i n v i t e d CDC to m eet w ith th e Oregon P e s t i c i d e A n a ly s is R esearch C e n te r (PARC) Board on November 2 1 , 1979. The Board was formed to e v a lu a te p o s s i b l e a d v e r se h e a lt h e f f e c t s o f c h e m ic a ls used in a g r ic u lt u r e . The PARC Board ask ed th e B ir th D e f e c t s Branch t o rev iew a v a ila b le d a ta and make a recommendation- about what kind o f ep id em io lo g ic study should be done to e v a lu a te th e r e l a t i o n s h i p betw een m a tern a l ex p o su re to h e r b ic id e sp r a y in g and b irth d efects. Study Approach A r e tr o s p e c tiv e case^ con trol- stud y appeared to be th e most ap p rop riate type o f study for sev era l reasons. F ir st because b irth d efects are rare even ts, i t would be d i f f i c u l t oth erw ise to ob tain enough c a ses fo r a study w ith s u ffic ie n t s t a t is t ic a l power. Second, the ban of 2 ,4 ,5 -T in 1978 elim inated th e source o f exposure. In a d d itio n , a r e tr o s p e c tiv e stud y would a ls o allow th e op p ortu n ity to e v a lu a te p o s s ib le in c r e a se s in m alform ation r a te s from so u r ce s o th er than h e r b ic id e s . The d e f in it io n o f c a ses and c o n tr o ls would be r e l a t i v e l y s t r a ig h t forw ard. Cases would be id e n t if ie d a s Oregon r e s id e n t b ir th s id e n t if ie d w ith c e n tr a l nervous system d e fe c ts and co n tro ls as normal Oregon in f a n t s matched, on m aternal r a c e , p la c e and date o f b ir t h . A ll c a se s would be id e n t if ie d , even th ose from urban a r ea s, because the p relim in ary BDMP d ata found h ig h CNS d e f e c t r a t e s in c o u n t ie s p red om in an tly urban a s w e ll as th o se predom inantly r u r a l. A r e l i a b l e , unbiased e stim a te o f exposure would be needed fo r both c a se s and c o n tr o ls . D eterm ining th is r a te p o ses some d i f f i c u l t y . Because o f obviou s b ia s r e s u ltin g from q u estio n in g in d iv id u a ls about spraying exposure and the u n a v a ila b ility o f a r e tr o s p e c tiv e b ioch em ical t e s t , we need ed some o t h e r method o f .d e te r m in in g e x p o su r e . We un dertook a p i l o t stu d y to lea rn i f I t was f e a s ib le to o b ta in an unbiased e stim a te o f ex p o su re, and i f i t w as, to lea rn th e r a te o f exposure. The la t t e r fig u r e i s n ecessa ry to d eterm in e an adequate sample s iz e o r to determ ine th e s t a t i s t i c a l power o f any sm aller sample s iz e . 25192 Pilot Study The purpose o f th e p i l o t stu d y was to e stim a te th e r a te o f h erb icid e ex p o su re in B en to n , P o lk and L in c o ln C o u n tie s a s a measure o f ex p o su re in w e ste r n O regon. T hese th r e e c o u n t ie s are p r im a r ily f o r e s t e d and L in c o ln and Benton c o u n tie s are where th e i n i t i a l concerns about h ir t h d e fe c ts were r a is e d . For th e p i l o t s tu d y , a random sam ple o f 72 b ir t h s was s e le c t e d from each cou n try u s in g v i t a l r e c o r d s . An e q u a l number o f b ir t h c e r t i f i c a t e s 0-2) w ere chosen fo r each y e a r , 1973-1977, and one was randomly s e le c t e d from each month. The r e s id e n c e s o f th e 72 b ir th s w ere determ ined, and fo r th ose w ith r u r a l a d d r e s se s, th e ow ners o f t r a c t s o f la n d w ith in two m ile s o f th e resid en ce o f th e ca se w ere so u g h t. I d e n t if ie d landow ners were co n ta cted and asked to give sp ray h is t o r y o c c u r r in g from 1 month b e fo r e co n c ep tio n o f t h e c a se t o 3 months a fte r conception. R esults o f P ilo t " _- x . T h ir ty r u r a l c a se s w ere i d e n t i f i e d . The o r ig in a l approach to id e n t if y ex p o su r e s was n o t f e a s i b l e ; T here w ere to o many owners w it h in 2 .0 m ile s and th e owners had n o, o r few w r itte n reco rd s, so they had to depend upon r e c o lle c tio n o f any s p r a y in g , when i t o ccu rred , o r what was sprayed . I t was o b v io u s, from t h is p il o t stu d y , th a t we could n o t r e lia b ly document h e r b ic id e usage by q u estion ing in d iv id u a ls or land owners. Extension o f P ilo t The o n ly d e f i n i t i o n o f h e r b ic id e exposure th a t appeared to be w orkable was fo r a e r ia l h e r b ic id e sp ra y in g n ear a r e s id e n c e th a t i s documented in some o f f i c i a l form. With t h is d e f in itio n in mind, we p lo tte d a l l th e sample b ir th s on county maps. We p l o t t e d th e 30 r u r a l b i r t h s as b e f o r e , and a l l th e remain ing b irth s lis t e d w ith a c ity address were p lo tted as resid en ts w ith in the c i t y l i m i t s . We a ls o u se d com puter p r in t o u t s , g en era ted f o r th e p i l o t stu d y by th e Department o f F o r e stry , to id e n t if y any fo r e st land th a t was sprayed betw een Jan. 1974 and Jan. 1980. Under Oregon law , a l l land owners (ex clu d in g th e U .S . F o r e s t S e r v ic e and th e bu reau o f Land Management h a v e to f i l e a pro posed spraying plan w ith th e S ta te Department o f F orestry. The m ajority o f the b ir th s p lo tte d were n ot near U .S. F orest S erv ice land so the s ta te clearinghouse was a good s o u r c e o f in fo r m a tio n . We assumed .the U .S . F o r e s t S e r v ic e w ould furnish data i f requested. We n e x t rev ie w e d th e sp ray d a ta to i d e n t i f y any a e r i a l h e r b ic id e sp r a y in g th a t occu rred w ith in two m iles o f any sam ple b ir t h . U sing t h is a d m itted ly crude e stim a te , we id e n t if ie d 8 b ir th s th a t had a e r ia l h e r b ic id e sp rayin g w ith in 2 m iles o f th e ir resid en ce during th e c r i t i c a l p eriod around concep tion. There w ere 2 a d d it io n a l b a b ie s th a t had p o s s i b l e sp r a y in g w it h in two m ile s o f t h e ir r e s id e n c e . The ex a ct lo c a tio n o f sp rayin g was unknown, but sp rayin g o ccu rred i n th e a r e a . Thus, we i d e n t i f i e d a maximum o f 10 ''exposed" i n f a n t s , fo r a r a te o f 4.6%. T his "exposure" r a te should be a p p lic a b le to th e w estern areas o f Oregon where th e v a st m ajority o f b ir th s occu r. 25193 5200 7/ ft Page Three I t i s im p o r ta n t to n o te t h a t t h i s ''exp osu re" r a te i s o n ly a rough e s t im a te o f a e r i a l h e r b ic id e s p r a y in g w it h in 2 m ile s o f a r e s id e n c e , and not a measure o f actu al exposure to h e r b ic id e s. I t appears to be im practicable, i f not im p o ssib le, to determ ine r e tr o s p e c tiv e ly what the tru e exposure of p regn an t women w as. We can o n ly e v a lu a t e i f a e r i a l s p r a y in g o ccu rred in th e g e n e r a l p r o x im ity o f i n f a n t s w ith CNS d e f e c t s more o f t e n than in. a group o f norm al i n f a n t s . The m is c l a s s i f i c a t i o n o f women in t o ex p o sed and n o t ex posed c a te g o r ie s would have the e f f e c t o f low ering th e power o f the study to detect p ossib le teratogen icity. Samole S iz e E stim ate The sam ple s i z e needed to d e te c t an exposure e f f e c t depends on th e ra te o f " e x p o su r e" , th e in c r e a s e i n r i s k among t h e s e e x p o se d , and th e s t a t i s t i c a l power d e s ir e d . U sin g th e e s t im a te o f 4.6% o v e r a ll "exposure" from th e p i l o t , i f we w ant to h ave a t l e a s t an 80% ch an ceo f fin d in g a 2 - f o l d in c r e a s e in r i s k , we w ould need 435 c a se s and 435 c o n t r o ls . I f th e a tta c k r a te i s 1 -1 /2 tim es g r e a t e r i n e x p o sed than in u n ex p o sed , we w ould need 1 ,5 3 3 each o f c a s e s and c o n t r o ls . A s t a t i s t i c a l power o f 80% was th e minimum fig u r e w e had hoped f o r . Un fo r tu n a te ly , we can't fin d enough ca ses in w estern Oregon w ith documented exposure s t a t u s to a c h ie v e t h i s pow er. T ab le 1 p r e s e n t s an e s t im a t e o f th e number o f c a se s th a t co u ld b e lo c a t e d , in te r v ie w e d , and th e h e r b ic id e exposure documented. T h is number o f. c a se s (264) and an eq u a l number o f c o n tr o ls would produce a s tu d y w ith a 62% chance o f d e t e c t in g a 2 - f o l d d if f e r e n c e i n a tta c k r a t e s , u sin g a 1 -ta ile d te s t at the 0.05 le v e l o f sig n ifica n ce . I f the attack rate is 1-1/2 tim es g r e a t e r i n c a s e s th an c o n t r o l s , th e n we h ave a 27% chance o f fin d in g a differen ce. Proposed E pidem iologic Study H y p o th e sis: The h y p o th e s is to be t e s t e d i s w hether m aternal exposure to a e r ia l spraying o f h erb icid e i s a sso c ia te d with, the b ir th o f in fa n ts w ith c e n t r a l n erv o u s sy ste m s (CNS) d e f e c t s . Study D esig n : A r e tr o s p e c tiv e matched c a s e -c o n tr o l stud y in w estern Oregon. In d ep th p e r s o n a l in t e r v ie w s would b e con d u cted o f 264 CNS c a s e s and 264 normal c o n tr o ls to e v a lu a te v a r io u s mdte'rnal and p a ter n a l r is k f a c t o r s . C o st: To con d u ct t h i s proposed stu d y w ould c o s t a p p ro x im a tely $192,000 (See attached bu dget). R equired Time: With, adequate s t a f f th e stu d y could b e com pleted in 9 to 18 months a f t e r approval to conduct the study has been g iv e n . Summary: With c o m p le tio n o f th e p i l o t , some ju dgem ents can be made about th e m e r its o f th e proposed e p id e m io lo g ic stu d y o f " h e r b ic id e exposure" and b ir t h d e f e c t s . The measurement o f ex p o su re i s tro u b leso m e. L ack ing some b e t t e r way to id e n t if y exposed m oth ers, we can o n ly in v e s t ig a t e th e r e la t io n s h ip between p roxim ate a e r i a l h e r b ic id e s p r a y in g and CNS b ir t h d e f e c t s . Such a stu d y would compare th e s e exposure r a t e s in in f a n t s w ith CNS d e f e c t s and normal in f a n t s . 25194. STO G ISI Page Four W ith th e a v a i l a b l e sam ple o f 264 c a s e s and 264 c o n t r o l s , we would have a 62% ch an ce o f d e t e c t i n g a 2 - f o l d r e l a t i v e r i s k . I f th e r e i s a 50% in c r e a s e in th e a t t a c k r a t e among th e e x p o s e d , we w ould have a 27% chance o f d e t e c t in g t h is d if fe r e n c e . These power fig u r e s are n ot p a r tic u la r ly encouran^ing. In a d d it io n , i t may v e i l b e th a t th e s e r e l a t i v e r is k s a p p ly o n ly to women w ith a s p e c ific le v e l or ty p e .o f exposure. Sin ce the a v a ila b le measure o f ex p o su re i s so c o a r s e , th e r e l a t i v e r i s k s o f "exposed" women b ased on a e r i a l sp r a y in g d a ta m igh t b e much lo w e r . In t h a t c a s e , th e s t a t i s t i c a l power would be even low er. lb is im portant th at the lim its o f such an epidem iologic study are i d e n t i f i e d i n t h e b e g in n in g . We f e e l t h a t I f a s tu d y i s u n d erta k en , th e c a s e co n tro l approach i s th e b e st approach. I f you decide to go ahead w ith such n stu d y , th e B ir th D e fec ts Branch would be w illin g to a s s i s t th e S ta te Health Departm ent and th e PARC Board in p la n n in g and im p lem en tin g th e s tu d y . Table I E stim ated number o f c e n t r a l nervous sy stem c a ses in Oregon in the years (1974 to 1979) for which exposure can be documented. (Estim ate based on Oregon r a te from th e B ir th D e fe c ts M onitoring Program d a ta .) 380 <V E stim ated number o f c a se s in e a s te r n Oregon. (The exp osu re r a te h e r e may ba...low er th an i n p i l o t a r e a .) 50 E stim a ted number o f CMS c a s e s i n w e ste r n Oregon, 1 9 7 4 -1 9 7 9 . 330 -E stim ated number o f c a s e s i n w e ste r n Oregon th a t we w i l l be a b le to l o c a t e and in t e r v ie w . (E stim a te o f 20% l o s t to fo llo w u p b ased on A tlanta experience.) * 264 25196 Budget o f Proposed C ase-C ontrol Study In Oregon Personnel - S alary and B e n e fits Person Years Category & grade Salary & Benefits 2.0 2 . O' 1 .0 1 .0 .2 .50 H ealth T echnician (GS-7) (ca se fin d in g - 8 persons @ 12 wks) H ealth T echn icians (GS-7) in terv iew in g - 8 persons@ 12 wks) Coder (GS-5) P u b lic H ealth A dvisor (GS-11) S t a t is t ic ia n (GS-13) E pidem iologist (0-5) $31,000 31,000 13,000 24,000 7,000 14,000 Travel Equipment & S u p p lies Rent and Communications & U t i l i t i e s P rinting Subtotal P lu s CDC overhead @ 20% T otal $120 .35,0 1,5 2,0 1,5 160, 32, $192,' 25197 ^ 10 075- 2 & tJT 7 Defect encephaly drocephaly ina Bifida Total CNS i t I Li]l lieft Palate esia enal Agenesis eduction Deformity C'.-.'ns Births in EDM? .'ioni tored COC 45 65 56 125 329 /o 393 450 433 626 W^ /tO V w <* f 3 P m P d/v^fl -- Oregon, 1970-1978 /<.,/, O. \ \ Lincoln No. Cases Rate '. B e n t o n ., NoV-^ Cases Nv/Rate Polk No. . Cases Lane No. Cases Rate O SS Total Lincoln, Benton, Po-lk, Lane Counties No. Cases Rate SU :A / / < v / Y P - / c v7 / Z7/ ^ / 4-- 70-77 All Oregon U.S. No. Cases Rate Rate 2 15.3 --- 9 9.9* \ 4 4.4 -- -- 12 4.6 17 6.6* 23 6.2 21 5.6 85 4.9 4.6 39 4.0 4.4. 1 7. 7 10 11.0 -- 9 3.5 20 5.4 72 4.1 6.2 48 18.6 23 8.9 79 21. 2 28 7.5 307 (z j D 102 5.8 5.2 33 12.8* 41 11.0 167 9.6 9.6 1 1.1 11 4.3 12 3.2 46 2.6 1.6 4 4.4 11 4.3 15 4.0 61 3.5 3.4 --- 1 1.1 4 1.5 5 1.3 23 1.3 1.0 1 7.7 3 3.3 17 6 . 6 ^ 21 5.6 59 3.4 3.3 1 7.7 3 3.3 27 10.4 31 8.3 155 8.9 8.1 1,307 9,123 1,051 25,360 37,341 174,659 7,903,54 100% 100% 89% 65% 3 3 / 0 '0-77 significant increases Benton - anencephaly (observed 8, expected 3.7) p = .05 Total CNS (observed 23, expected 16.68) p = .05 Lane - hydrocephaly (observed 16, expected 9.9) p = .05 cleft lip (observed' 23, expected 11.6) p a .01 reduction deformity (observed 16, expected 7.3) p = .01 -0 ' S) / / o o - f r fr-fcrl ..Regional A g ricu ltu ra l F o re st W eekly ';r V 1* - , ' - . ' /J \ V-; ' '`i -, /'.!"' > ,!. , . i . " -i. > / / ? V ' . v \ H '- V '- .* d i SALEM, OREGON FRIDAY, DECEMBER 12, 1980 (USPS O89-A0O) V o lu m 5 1 , N u m b e r .50 Look at `highly exposed5segment proposed . . . PARC shifts study emphasis By CLAUDE STEUSLOFF pesticide study he outlined (Capital P ra\i barm hdilnr would require $100.000 In fund SALEM, Ore - The Pesti ing from the state of Oregon. cide Analytical T tw irch Cen He said the CDC Is not a fund ter (PARC) lias rejected a ing group, the U.S. Environ '`retrospective type <>f study" mental Protection A g e n c y of the pesticide pmblein In would not contribute, but the coastal counties of Oregon as U.S. Forest Service could have outlined in a re|>ort made by some funds. Edmunds esti the U.S. ('enter for Disease mated tolal cost of the pro G'lilrol, Atlanta, Ga posed study at $192,000. In a change of direction, PARC members decided to study a proposal to be made by Oregon Slate University on the effect of live pesticide 2,4-D ``.on highly exposed peo ple in Eastern Oregon and Jie Willamette Valhy." That proposal Is due at the next meeting of the group, Jan. 30, at which time they will also consider a report of pesticide studies ma ie by Dr. James Googin of the Univer sity of Oregon Health Sciences Center, Portland. Marla Gillam, F tgene, of the Northwest Coalit on for Al ternative: to Pestii 'dfs, sug gested PARC study "the rise Edmunds said birth defects among children horn in Ore gon runs lower than the na tional average but two coun ties have defects significantly higher than tlie U.S. average. A pilot study by the Atlanta CDC was done in Lincoln, Ben ton and Polk counties on use of 2,4-D in primarily forested areas. Based on the Oregon rate of central nervous sys tem birth defects from the Birth Defects Monitoring Pro gram data, it was estimated 264 cases could be located and Interviewed In Western Ore gon. Specifically, the study de sign rejected by PARC was and distribution of the Use of "A retrospective m a t c h e d chemicals in Oregon." case-control study, in-depth The action came after a personal Interviews would be morning of technical discus conducted of 264 central ner sion on pesticides by a group vous system cases and 264 of 15 scientists last Thursday normal controls tn evaluate at the Oregon Department of various maternal and paternal Agriculture, under direction of risk factors. With adequate Bill Kosesan. administrator of staff, the work could have been the department plant division completed In nine to 18 and chairman of PARC. monllis. Warren Weslgarth of the De The report states that a re partment o f Environmental liable, unbiased estimate of Quality, made the original mo . exposure to 2,4-D would be tion. He said he did so "after needed for both cases and con looking at all the data in the trols. Because birth defects CDC report" and with the hope are rare events. It would he something could be going from difficult to obtain enough rases the center before the summer for a sludy with sufficient applications of pesticides. statistical power, other than Larry Edmunds, epidemiolo In a retrospective study. gist with the disease control Edmunds said the CDC re center at Atlanta, said the port Is "only an outline of an approach to the problem." During the discussion, he said the proposed further study has some advantages but there may be more disadvantages and there is no need to make a study for the sake of a study. Mike Newton, professes' of forest ecology at OSU, con* lite T w w t ^ that the CDC pilot study h y p o thesis blaming 2,4-D in coastal county birth defect problems is erroneous. Hr said use of 2,4 D la a national problem and hi many farming areas the use is much greater than in the timbered coastal community. He cited widespread roadside weed and brush u se of 2,4-D and oilier herbicides. Newton questioned whether the incidence of coastal de fects is extraordinary and said birth defects lu the region may be due to oilier causes such as high phosphate levels in soils and excessive rales of mercury and copier in whter supplies at tile coast. He said a study of residents In urban areas at the coast -- well removed from timber -- showed that 17 percent of the population had high residues of 2,4-D in their urine. " A study, if it is positive, would show defects but they would be subject to olher vari able causes than herbicides. We don't know what causes de fects. There is only a 50 per cent chance of picking up herbicide e f f e c t s with a study," said Newton. He said there is a 10,000-to-l safety factor in herbicide ap plications. "Drugs, alcohol, etc., show birth defects In users, such confounding fac tors are large and will confuse the study." Newton continued. Gillam said "lots of citizens re angry" about 2,4-D appli cations and warned that "soon er or later you must look at use of herbicides." James Wilt, agricultural Ex tension specialist in chemis try, OSU, proposed looking at positive controls for birth de fects due to 2,4-D. Virgil Freed, head of the agricultural chemistry depart ment at OSU, called for a s t u d y of possible defects among husbands and wives who have used pesticides on family farms in Morrow and Umatilla counties, for many continuous years. Dr. Stringham. family phy sician In Lincoln County, sug gested both fathers and moth ers should be studies for birth defects. The Pesticide Analytical Re search Center consists of rep resentatives from the state de partments o f Agriculture, Health Division, Environmen tal Quality, Forestry and Fish & Game. One member from the public at large is appoint ed by the governor. a o 7 7  38. R. T. Sauer, R. R. Yocum, R. F. Doolittle, M. Lewis, C. 0 . Pabo, Nature (London) 298, 447 (1982). 39. B. von Wilken-Bergmann and B. Mller-Hill, Proc. Natl. Acad. Sei. U.S.A. 79, 2427 (1982). 40. H. Nick et al., Proc. Natl. Acad. Sei. U.S.A. 79, 218 (1982); H. Nick et al., J. Mol. Biol. 161, 417 (1982); K. Arndt, H. Nick, F. Boschelli, P. Lu, J. Sadler, ibid., p. 439; R. Kaptein and K. Wuthrich, personal communication. 41. T. A. Steitz, D. H. Ohlendorf, D. B. McKay, W. F. Anderson, B. W. Matthews, Proc Natl. Acad. Sei. U.S.A. 79, 3097 (1982). 42. D. H. Ohlendorf and B. W. Matthews, unpub lished data. 43. I. T. Weber, D. B. McKay, T. A. Steitz, Nucleic Acids Res. 10, 5085 (1982). 44. D. H. Ohlendorf, W. F. Anderson, B. W. Mat thews, J. Mol. Evol. 19, 109 (1983). 45. D. H. Ohlendorf, W. F. Anderson, M. Lewis, C. O. Pabo, B. W. Matthews, J. Mol. Biol., in press. 46. J. C. Wang, M. D. Barkley, S. Bourgeois, Nature (London) 251, 247 (1974). 47. T. Maniatis and M. Ptashne, Proc. Natl. Acad. Sei. U.S.A. 70, 1531 (1973). 48. N. C. Seeman, J. M. Rosenberg, A. Rich, ibid. 73, 804 (1979); C. Hlene, FEBS Lett. 74, 10 (1977). 49. P. H. von Hippel, in Biological Regulation and Develooment, R. F. Goldberger, Ed. (Plenum, New York, 1979), p. 279. 50. M. H. Caruthers, Acc. Chem. Res. 13, 155 (1980). 51. O. G. Berg, R. B. Winter, P. H. von Hippel, Biochemistry 20, 6929 (1981); R. B. Winter and P, H. von Hippel, ibid., p. 6948. R. B. Winter, O. G. Berg, P. H. von Hippel, ibid., p. 6961. 52. L. Guarente, J. S. Nye, A. Hochschild, M. Ptashne, Proc. Natl. Acad. Sci. U.S.A. 79, 274 (1982). 53. A. Hochschild, N. Irwin, M. Ptashne, Cell 32, 319 (1983). 54. D. K. Hawley and W. R. McClure, J. Mol. Biol. 157, 493 (1982); Cell 32, 327 (1983). 55. We thank R. T. Sauer for communicating results in advance of publication and C. O. Pabo and M. Ptashne for helpful comments on the manu script. Supported in part by NIH grants GM28138 and GM30894 (Y.T.) and GM20066 (B.W.M.), a grant from the M, J. Murdock Charitable Trust, NSF grant PCM-8014311 (B.W.M.), a grant from the Canadian Medical Research Council through the MRC Group on Protein Structure and Function (W.F.A.), and an NIH postdoctoral fellowship (D.H.O.). * Address correspondence to B. W. Matthews. ner laid down by its statutes. That man ner is to set standards and enforce them, and our enforcement powers are strong and pervasive. But the standards we set, whether technology- or health-related, Science, Risk, and Public Policy must have a sound scientific base. Science and the law are thus partners William D. Ruckelshaus at EPA, but uneasy partners. The main reason for the uneasiness liqs, I think, in the conflict between the way science really works and the public's thirst for certitude that is written into EPA's laws. We are now in a troubled and emotion confidence. The polls show that scien Science thrives on uncertainty. The best al period for pollution control; many tists have more credibility than lawyers young scientists flock into fields where communities are gripped by something or businessmen or politicians, and I am great questions have been asked but approaching panic and the public discus all three of those. I need the help. of nothing is known. The greatest triumph sion is dominated by personalities rather scientists. of a scientist is the crucial experiment than substance. It is not important to This is not a naive plea for science to that shatters the certainties of the past assign blame for this. I appreciate that save us from ourselves. Somehow, our and opens up rich new pastures of igno rance. But EPA's laws often assume, indeed Summary. A climateoffearnow dominates thediscussionofenvironmental issues. demand, a certainty of protection greater The scientific community can help alleviate this fear by making a greater effort to than science can provide with the cur explain tothe publicthe uncertainties involved inestimates ofrisk.Current statutory rent state of knowledge. The laws do no mandates designed to protect public health both demand levels of protection that more than reflect what the public be technology cannot achieve and are uncoordinated across government agencies. A lieves and what it often hears from peo common statutory framework for dealing with environmental risks is needed. In ple with scientific credentials on the 6 addition, care must be taken toseparate the scientificprocess ofassessing riskfrom o'clock news. The public thinks we the use of such assessments, together with economic and policy considerations, iii know what all the bad pollutants are, the management of risksthrough regulatory action. precisely what adverse health or envi ronmental effects they cause, how to measure them exactly and control them people are worried about public health democratic technological society must absolutely. Of course, the public and and about economic survival, and legiti resolve the dissonance between science sometimes the law are wrong, but not all mately so, but we must all reject the and the creation of publje policy. No wrong. We do know a great deal about emotionalism that surrounds the current where is this more troublesome than in some pollutants and we have controlled discourse and rescue ourselves from the the formal assessment of risk--the esti them effectively by using the tools of the paralysis of honest public policy that it mation of the association between expo Clean Air Act and the Clean Water Act. breeds. sure to a substance and the incidence of These are the pollutants for which the I believe that part of the solution to some disease, based on scientific data. scientific community can set safe leveis our distress lies with the idea that disci and margins of safety for sensitive popu plined minds can grapple with ignorance lations. If this were the case for all and sometimes win; the idea of science. Science and the Law at EPA pollutants, we could breathe more easily We will not recover our equilibrium (in both senses of the phrase); but it is without a concerted effort to more effec Here is how the problem emerges at not so. tively engage the scientific community. Frankly, we are not going to be able to emerge from our current troubles with out a much improved level of public the Environmental Protection Agency. EPA is an instrument of public policy, whose mission is to protect the public health and the environment in the man William D. Ruckelshaus is Administrator of the U.S. Environmental Protection Agency, Washing ton, D.C. 20460. This article is based on a talk he gave at the National Academy of Sciences, Wash ington, D.C., on 22 June 1983. 1026 SCIENCE, VOL. 22 XOC7^ More than 10 years ago, EPA had the we find a confusing picture. In assessing is risk management. The second proce Clean Air Act, the Clean Water Act, a a suspected carcinogen, for example, dure involves a much broader array of solid waste law, a pesticide law, and there are uncertainties at every point disciplines and is aimed toward a deci laws to control radiation and noise. Yet where an assumption must be made: in sion about control. to come were the myriad of laws to calculating exposure; in extrapolating In risk management it is assumed that control toxic substances from their man from high doses where we have seen an we have assessed the health risks of a ufacture to their disposal--but that they effect to the low doses typical of environ suspect chemical. We must then factor in would be passed was obvious even then. mental pollution; in what we may expect its benefits, the costs of the various When I departed EPA a decade ago, when humans are subjected to much methods available for its control, and the the struggle over whether the federal lower doses of a substance that, when statutory framework for decision. The government was to have a major role in given in high doses, caused tumors in NAS report recommends that these two protecting our health, safety, and envi laboratory animals; and finally, in the functions--risk assessment and risk ronment was ended. The American peo very mechanisms by which we suppose management--be separated as much as ple had spoken. The laws had been the disease to work. possible within a regulatory agency. This passed; the regulations were being writ One thing we clearly need to do is is what we now do at EPA and it makes ten. The only remaining question was ensure that our laws reflect these scien sense. whether the statutory framework we had tific realities. The administrator of EPA created made sense or whether, over should not be forced to represent that a time, we would adjust it. margin of safety exists for a specific Risk Assessment substance at a specific level of exposure where none can be scientifically estab We also need to strengthen our risk Scientific Realities lished. This is particularly true where the assessment capabilities. We need more inability to so represent forces the cessa research on the health effects of the Ten years ago I thought I knew the tion of all use of a substance without any substances we regulate. I intend to do answer to that question as well. I be further evaluation. everything in my power to make clear lieved it would become apparent to all the importance of this scientific analysis that we could virtually eliminate the at EPA. Given the necessity of acting in risks we call pollution if we wanted to Functions of Regulatory Agencies the face of enormous scientific uncer spend enough money. When it also be tainties, it is more important than ever came apparent that enough money for all It is my strong belief that where EPA, that our scientific analysis be rigorous the pollutants was a lot of money, I came OSHA (the Occupational Safety and and the quality of our data be high. We to believe that we would begin examin Health Administration), or any other so must take great pains not to mislead ing the risks very carefully and structure cial regulatory agency is charged with people about the risks to their health. We a system that would force us to balance protecting public health, safety, or the can help to avoid confusion be ensuring our desire to eliminate pollution against environment, we should be given, to the both the quality of our science and the the costs of its control. This would entail extent possible, a common statutory for clarity of our language in explaining haz some adjustment of the laws, but not all mula for accomplishing our tasks. This ards. that much, and it would happen by about statutory formula may well weigh public I intend to allocate some of EPA's 1976. I was wrong. health very heavily, as the American increased resources to pursuing these This time around as administrator of people certainly do. ends. Our 1984 request contains signifi EPA, I am determined to improve our The formula should be as precise as cant increases for risk assessment and country's ability to cope with the risk of possible and should include a responsi associated work. We have requested $31 pollutants over where I left it 10 years bility for assessing the risk and weighing million in supplemental appropriations ago. It will not be easy, because we must it, not only against the benefits of contin for research and development, and I now deal with a class of pollutants for ued use of the substance under examina expect that risk assessment will be more which it is difficult, if not impossible, to tion, but against thb risks associated with strongly supported as a result of this establish a safe level. These pollutants substitute substances and the risks asso increase as well. interfere with genetic processes and are ciated with the transfer of the substance I would also like to revitalize our long associated with the diseases we fear from one environmental medium to an term research program to develop a base most: cancer and reproductive disor other through pollution control prac for more adequately protecting the pub ders, including birth defects. The scien tices. I recognize that legislative change lic health from toxic pollutants. I will be tific consensus is that any exposure, in the current climate is difficult. It is up asking the outside scientific community however small, to a genetically active to those of us who seek change to make for advice on how best to focus those substance embodies some risk of an ef the case for its advisability. research efforts. fect. Since these substances are wide But my purpose here is not to plead for In the future, this being an imperfect spread in the environment, and since we statutory change; it is to speak of risk world, the rigor and thoroughness of our can detect them down to very low levels, assessment and risk management and the risk analyses will undoubtedly be affect we must assume that life now takes place role of science in both. It is important to ed by many factors, including the toxici in a minefield of risks from hundreds, distinguish these two essential functions, ty of the substances examined, the popu perhaps thousands, of substances. We and I rely here on a recent National lations exposed, the pressure of the regu can no longer tell the public that they Academy of Sciences report on the man latory timetable, and the resources avail have an adequate margin of safety. agement of risk in the federal govern able. Despite these often conflicting This worries all of us, and it should. ment. Scientists assess a risk to find out pressures, risk assessment at EPA must But when we examine the premises on what the problems are. The process of be based only on scientific evidence and which such estimates of risk are based, deciding what to do about the problems scientific consensus. Nothing will erode 9 SEPTEMBER 1983 1027 ________________ S 0 0 77 -t public confidence faster than the suspi cion that policy considerations have been allowed to influence the assessment of .risk. Risk Management Although there is an objective way to assess risk, there is, of course, no purely What we need to hear more of from scientists is science. I am going to try to provide avenues at EPA for scientists to become more involved in the public dia log in which scientific problems are de scribed. Lest anyone misunderstand, I am not suggesting that all the elements of man aging risk can be reduced to a neat mathematical formula. Going through a agement strategies that flow v assessment may indeed differ, do on each agency's statutory mar, the judgment of the ultimate dt maker. But even at the management stat there is no reason why the approache cannot be coordinated to achieve tlk goal of risk avoidance or minimizatioi with the least societal disruption possi objective way to manage it, nor can we ignore the subjective perception of risk in the ultimate management of a particu lar substance. To do so would be to place disciplined approach can help to orga nize our thoughts so that we include all the elements that should be weighed. We will build up a set of precedents that will ble. I have been exploring with the Whit, House and the Office of Managemen and Budget the possibility of effecting better intragovemmental coordination o too much credence in our objective data be useful for later decision-making and the way in which we assess and manage and ignore the possibility that occasion will provide more predictable outcomes risk. ally one's intuition is right. No amount of for any social regulatory programs we To push this one step further, I believ- data is a substitute for judgment. adopt. it is in our nation's best interest to shar Further, we must search for ways to In a society in which democratic prin our knowledge of risks and our approac describe risk in terms that the average ciples dominate, the perceptions of the to managing them with the other devei citizen can comprehend. Telling a family public must be weighed. Instead of ob oped nations of the world. The environ that lives close to a manufacturing facili jective and subjective risks, the experts mental movement has taught us the in ty that no further controls on the plant's sometimes refer to " real" and " imagi terdependence of the world's ecosys emissions are needed because, according nary" risks. There is a certain arrogance terns. In coping with the legitimate con to our linear model, their risk is only in this--an elitism that has ill served us cerns raised by environmentalists, we 10~6, is not very reassuring. We need to in the past. Rather than decry the igno must not forget that we cope in a worlc describe the suspect substances as clear rance of the public and seek to ignore with interdependent economies. If our ly as possible, tell people what the their concerns, our governmental pro approach to the management of risk i? known or suspected health problems are, cesses must accommodate the will of the not sufficiently in harmony with those 01 and help them compare that risk to those people and recognize its occasional wis the other developed nations, we coulc with which they are more familiar. dom. As Thomas Jefferson observed, " If save our health and risk our economy. ; To effectively manage the risk, we we think [the people] not enlightened do not believe we need to abandon ei must seek new ways to involve the pub enough to exercise their control with a ther, but to ensure that it does not hap lic in the decision-making process. wholesome discretion, the remedy is hot pen, we need to work hard to share Whether we believe in participatory de to take it from them, but to inform their scientific data and understand how tc mocracy or not, it is a part of our social discretion." harmonize our management technique? regulatory fabric. Rather than praise or with those of our sister nations. lament it, we should seek more imagina In sum, my goal is a government-wide tive ways to involve the various seg Interagency and International process for assessing and managing envi ments of the public affected by the sub Coordination stance at issue. They need to become ronmental risks. Achieving this will take cooperation and goodwill within EPA. involved early, and they need to be in Up to this point I have been suggesting among Executive Branch agencies, and formed if their participation is to be how risks should be assessed and man between Congress and the Administra meaningful. We will be searching for aged in EPA. Much needs to be done to tion, a state of affairs that may partake of ways to make our participatory process coordinate the various EPA programs to the miraculous. Still, it is worth trying, work better. ensure a consistent approach. I have and the effort is worth the wholehearted For this to happen, scientists must be established a task force with that char support of the scientific community. I willing to take a larger role in explaining ter. believe such an effort touches on the the risks to the public--including the I further believe we should make uni maintenance of our current society, in uncertainties inherent in any risk assess form the way in which we manage risk which a democratic polity is grounded in ment. Shouldering this burden is the across the federal regulatory agencies. a high-technology industrial civilization. responsibility of all scientists, not just The public interest is not served by two Without a much more successful way of those with a particular policy end in federal agencies taking diametrically op handling the risks associated with the mind. In fact, all scientists should make posed positions on the health risks of a creations of science, I fear we will have clear when they are speaking as scien toxic substance and then arguing about it set up for ourselves a grim and unneces tists, ex cathedra, and when they are in the press. We should be able to coor sary choice between the fruits of ad recommending policy they believe dinate our risk assessment procedures vanced technology and the blessings of should flow from scientific information. across all federal agencies. The risk man democracy. ? 1028 SCIENCE, VOL. 221 ^ 030 vTONAL NICE l Tl I Office1 of Cancer Communications January 1991 Room 10A24 M;iryl;ind 2089? Occupational Risk of Cancer from Pesticides: Farmer Studies Studies from the United States and other countries have shown that farmers have a higher risk for certain cancers, particularly cancers of the t bLood and immune systems {Leukemia, non-Hodgkin*s Lymphoma, and multiple myeLoma). The reasons for the increased risks are not d e a r , and scientists are Looking at chemicaL and other occupational exposures common to farming to identify the possibLe cause or causes. Farmers are exposed to a variety of potentiaLLy harmful substances during their work day, including: pesticides (fungicides, herbicides, insecticides, rodenticides, and others), chemical solvents, fueLs and oils, animal viruses, and other microbes. Some of these agents are known or suspected carcinogens. In the L980s, NationaL Cancer Institute (MCI) researchers began severaL studies comparing pesticide use by farmers with cancer to those without the disease (case-controL studies). The first of these investigations was a scudy of Kansas farmers with soft-tissue sarcomas, Hodgkin's disease, and nonHodgkin's Lymphoma (MHL). Published in 1986, this study showed that farmers who used herbicides, especiaLLy 2 ,4-dichLorophenoxyacetic acid (2,4-D), had an excess of NHL as compared to farmers and nonfarmers who did not use these chemicals. The risk of deveLopingyNHL /increased with frequency of herbicide use. Farmers who used herbicides on their farms 20 or more days per year had a six times greater risk of developing NHL than nonfarmers. Within the high- <R O O 2 exposure group, farmers who mixed and applied the herbicides themselves had an gight-fold greater risk.. Soft-tissue sarcomas and Hodgkin's disease were not linked with use of herbicides. MCI scientists have now completed case-control studies of NHL among farmers in eastern Nebraska and leukemia among farmers in Iowa and Minnesota. Each study included all cases of these cancers occurring in a specific geographic area. Hospital records and pathology slides from the cases were carefuLly reviewed by pathologists to verify the cancer diagnoses. Study subjects, or their next-of-kin, were interviewed for details about the subjects' use of agricultural pesticides and other factors that might be associated with these cancers. Pesticide exposure reported by the cancer patients was compared to exposure reported by randomly selected men without those cancers who lived in the same area. Non-Hodgkin's Lymphoma Shelia Hoar Zahm, Sc.D., and collaborators at NCI and the University of Nebraska Medical Center studied 201 men living in 66 counties in eastern Nebraska who developed NHL between 1983 and 1986. The researchers compared the pesticide exposures of these men to exposures of 725 men from the general, population who did not have this cancer. Although there was no overall excess of NHL among farmers, farmers who mixed or applied 2,4-D had a 50 percent increased risk of the disease. The risk of NHL also increased with frequency of 2,4-D use. Farmers who used the pesticide for 20 or more days per year had a three-fold higher risk than those not exposed to the chemical. The longer farmers waited to change into clean work clothes after pesticide application, the higher their risk of NHL. The risk of NHL among farmers who changed out of work clothes immediately after completing the 1 3 ap.plicacion was similar to that seen among nonfarmers. Farmers who continued to use these work clothes the following day or longer had nearly five times the NHL risk of nonfarmers. The investigators also looked at the possible effects of pesticides ocher than 2,4-D, and ocher factors chat might be associated with risk of NHL, such as other agricultural exposures, medical conditions, exposure to radiation, and tobacco use. In other reports, non-2,4-D pesticides, particularly organophosphate insecticides, have also been shown co increase the risk of NHL. The NCI study results suggest chat 2,4-D and organophosphate insecticides have independent influences on the risk of cancer. Further evaluation of organophosphate insecticides is necessary to quantify any possible risks associated with their use. The association of 2,4-D and NHL in Nebraska farmers is consistent with che previous NCI study conducted in Kansas. Pescicide exposures occurring many years in the past are difficult to assess accurately, and tend to cause -- an underestimation of the cancer risk.? Dr. Zahm and her collaborators believe the evidence suggests that the use of 2,4-D in the agricultural setting increases the risk of NHL among persons handling this pesticide frequently. Leukemia Linda Morris Brown, M.P.H., and colleagues at NCI, che University of Iowa, and the University of Minnesota studied a total of 578 cases of leukemia diagnosed among men in Iowa and Minnesota between 1981 and 1984 (293 cases :n Iowa and 285 cases in Minnesota).^ To compare the effects of pesticide exposure, 1,245 men without the disease were selected at random from che generaL population. The- researchers found a slight excess of leukemia, especially chronic lymphocytic leukemia, among farmers. Exposure to specific fungicides, 4 herbicides (including 2,4-D ), or crop in s e c tic id e s did not increase the risk.. The risk for Leukemia was significantly elevated among farmers who used certain pesticides to inhibit insects on animals rather than on crops. The insecticides were: the organophosphates crotoxyphos (11 times greater risk), dichlorvos (2 times), and famphur (2 times); the natural product pyrethrins (4 times); and the chlorinated hydrocarbon methoxychlor (2 times). Risk of leukemia did not increase consistently with how often any pesticide was used. However, the greatest risk among farmers using c,,_h_L_o_r1d__a_n_e, DOT, dichLorvos, and maLathion occurred among those using the chemicals on aniraaLs 10 or more days a year. The risk of leukemia was further increased among chose farmers who had used insecticides at least 20 years before the study. The investigators improved the accuracy of the risk estimate from individual pesticides by accounting for the cancer risk from smoking, nonfarming occupational and chemical exposures, family history of cancer, and other factors. Larger studies are stiLL needed to fully evaLuate the effects of any single pesticide on Leukemia risk. The increased cancer risk from animal insecticides versus crop pesticides may result from closer proximity to animals on a regular basis. Animal insecticides need to be carefully scrutinized in future studies. ### For additional information about cancer, write to the Office of Cancer Communications, National Cancer Institute, Bethesda, MD 20892, or call the Cancer Information Service toLl free at 1-800-4-CANCER. Spanish-speaking staff members are also available during daycirae hours. ) TO N A L NC l Tl T Office of Cancer Communications Maryland 0892 September 1991 Dogs Exposed to Herbicide Develop More Lymphomas Dogs whose owners used a common weed killer on their lawns have, an increased rate of an ijnmune system cancer (lymphoma )_(closeLy-related-- to--human-''-' ,^non-Hodgkin1s lymphoma_>>accorVing to a National Cancer Institute (NCI) study. The findings are consistent with previous studies showing that humans have an increased risk of non-Hodgkin's lymphoma due to frequent exposure to the herbicide (weed killer) 2,4-dichlorophenoxyacetic acid (2,4-D).* Published in the September 4 issue of the Journal of the National Cancer Institute,* the study showed that owners of dogs diagnosed with malignant lymphoma applied the herbicide 2,4-D to their lawns and/or employed professional lawn care companies to treat their lawns about a third more often than owners of dogs that did not have the disease. When the owners alone applied four or more treatments of liquid or granular 2,4-D to the lawn each year, the risk of lymphoma among their dogs doubled. Dogs and humans are known to have similar reactions to some cancercausing substances, and dogs may signal cancer risk in people exposed to the \ same substances.^- "This study supports the idea that exposure to 2,4-D, as used for lawn care, pLays a role in causing lymphomas in dogs," said Howard M. Hayes, D.V.M., of NCI's Environmental Epidemiology Branch, the principal *The study is titled "A case-control study of canine malignant lymphoma: positive association with dog owner's use of 2,4-dichlorophenoxyacetic acid herbicides ." A o o *S ? 2 investigator Eor the study. "The study aLso suggests that the potential, health hazards of human exposure to 2,4-D at home warrant further study." Dogs exposed to 2,4-D consistently showed an increased risk, of lymphoma compared to dogs not exposed to the herbicide, regardless of the sex or breed r of the dog. In general, maLe dogs are normally at a slightly higher risk than female dogs. Boxer, bull mastiff, and Scottish terrier breeds are aLso more frequently diagnosed with this disease. Separate anaLyses showed no increase in Lymphomas due to exposure to other chemicaLs including: flea- and tick-controL products, household insecticides, tobacco smoke, commercial cleaners (on draperies and carpeting), and outdoor insecticides and/or fertilizers. No List of chemicaL products was provided to the owners during the written or telephone interviews with the researchers. "We did not want to 'prompt' the owners to identify certain products or chemicaLs," explained Hayes. Hayes and his colleagues interviewed the owners of dogs treated for Lymphomas (491 dogs), other cancers (479 dogs), or other illnesses (466 dogs) at one of three veterinary teaching hospitaLs. Information was gathered about the amount of tobacco smoked in the house, the dog's Life history, use of chemicaLs (in the house, on the dog, and on the Lawn), and the dog's access to the yard. Dogs with no access to their owners' yards were considered to be unexposed. Dogs with illnesses known or thought to be reLated to chemicaL exposures (Lower urinary tract cancer, nonspecific skin inflammations, and nerve disorders) were not incLuded in the study. Other authors of the study are Robert E. Tarone, Ph.D., and Kenneth P. Cantor, Ph.D., from NCI; Carl R. Jessen, D.V.M., from the CoLLege of Veterinary Medicine, University of Minnesota, St. Paul, Minn.; Dennis M. 3 McCurnin, D.V.M., from Che CoLLege of Veterinary Medicine, CoLorado Scate Universicy, Fort CoLLins, Colo.; and RaLph C. Richardson, D.V.M., from the School, of Veterinary Medicine, Purdue University, West Lafayette, Ind. References 1. Blair A: Herbicides and non-Hodgkin's lymphoma: New evidence from a study of Saskatchewan farmers. J NatL Cancer Inst 82:544-545, 1990. 2. Hoar SK, Blair A, Holmes FF, et al: Agricultural herbicide use and risk of lymphoma and soft-tissue sarcoma. J Amer Med Assoc 256:1141-1147, 1986. 3. Zahra SH, Weisenburger DQ, Babbitt PA, et al: A case-control study of non-Hodgkin's lymphoma and the herbicide 2,4-dichLorophenoxyacetic acid (2,4-D) in eastern Nebraska. Epidemiol 1:349-356, 1990. 4. Hueper WC, Wiley FH, Wolfe HD: Experimental production of bladder tumors in dogs by administration of beta-napthylamine. J Indust Hyg Toxicol 20:46-84, 1938. 5. Clickman LT, Domanski LM, Maguire TC, et al: Mesothelioma in pet dogs associated with exposure of their owners to asbestos. Environ Res 32:305-313, 1983. 6. Hayes HM, Tarone RE, Casey WW, et al: Excess of seminomas observed in Vietnam service U.S. military working dogs. J NatL Cancer Inst 82:10421046, 1990. 7. Tarone RE, Hayes HM, Hoover RN, et aL: Service in Vietnam and risk of testicuLar cancer. J Natl Cancer Inst (in press, 1991). 8. Hayes HM, Hoover RN, Tarone RE: Bladder cancer in pet dogs: A sentinel for environmental cancer? Am J EpidemioL 114:229-233, 1981. Attached is a List of questions and answers about the study. 4 Questions and Answers Study of Herbicide Exposure in Dogs with Lymphoma 1. What is Lymphoma? Lymphoma is a general term for cancers of the immune system, the complex network of specialized organs and cells that help defend the body against disease and infection. The two main types of lymphoma found in humans are called Hodgkin's disease and non-Hodgkin's lymphoma. Malignant lymphoma in dogs is considered the canine equivalent of non-Hodgkin's lymphoma in humans. 2. How common is malignant Lymphoma in dogs? What causes the disease? Malignant lymphoma is a.caibsnon cancer in dogs. About one out of seven dogs diagnosed with canceir at university veterinary teaching facilities in the United States has lymphoma. The causes of lymphoma in dogs are not fully understood. The disease occurs at aLL ages but more frequently in older dogs. Slightly more male dogs than female dogs get lymphoma. Boxers, bull mastiffs, and Scottish terriers are more Likely to develop this cancer than other breeds of dogs. Although Lymphoma in other laboratory and domestic animals has been associated with a certain type of virus (retrovirus), no such association has been observed in dogs. 3. How common is Lymphoma in humans? What causes the disease? In the United States, over 37,000 people wiLL be diagnosed with nonHodg^Trr^s~tymphoma' and~ another f74Pu with Hodgkin' s dTsease~~ih~T99T. Between 1973 and 1988, the incidence rate for non-Hodgkin* s lymphoma --increased over 50 percent; one jof the largest increases for any cancer. As with dogs, scientists cannot yet explain the causes of lymphoma in humans. A number of possible causes are being studied, including immune system damage, exposure to certain chemicals, viruses, heredity, and radiation. 4. Can humans "catch" lymphoma from their pets? There is no evidence that any form of cancer can be spread from pets to their owners. ALthough a retrovirus is known to cause Lymphoma in cats, humans are not susceptible to this virus. 5. What did this study show? ( The study showed that dogs ng in households in which the herbicide ! 2,4-dichLorophenoxyacetic at id (2,4-D) was applied to the lawn, by owners ( or commercial lawn care companies, developed malignant lymphoma more CLooSS' 5 j frequently than dogs in households where this was not the case. The study uss not designed to look at the causes of other cancers. 6. Why is the study important? The study points out the need to conduct additional studies on the human health implications of 2,4-D exposure in the home environment. Dogs and humans may react similarly when exposed to a cancer-causing substance. 7. What is 2,4-D and what products contain it? The chemical 2,4-D is used to kill broadleaf weeds (not crab grass) on lawns and in gardens. It is one of the most frequently used herbicides in the United States. Many companies manufacture and use it in lawn and garden weed killers. This herbicide is listed on the label as 2 .4-D. 2,4-D acid, or 2,4-dichlorophenoxyacetIc~acid. 8. Is the risk the same fro Liquid and granuLar herbicides? The researchers found no difference in the risk of Lymphoma in dogs between owners who used products containing Liquid 2,4-D and owners who ^used products with granuLar (dry) 2,4-D. 9. How do dogs come into contact with 2,4-D? Dogs can'be exposed to 2,4-D by walking, roLLing, and lying on a treated lawn; Licking their fur; and eating treated grass. Researchers do not beLieve that dogs are harmfully exposed by inhaling the herbicide. 10. What are pesticides and herbicides? A pesticide is a poison used to destroy "pests." The term incLudes herbicide (a substance that is destructive to pLants), insecticide (an agent that kills insects), rodenticide (a poison for rodents), and fungicide (a compound that destroys moLds). 11. Could exposure to other substances have caused the increase in canine Lymphome seen in this study? In this study, the researchers aLso looked at use of flea and tick controL products, use of household insecticides, cigarette smoking in the home, frequency of commercial cleaning of draperies and carpets, and the use of insecticides and/or fertilizers in the yard. Exposure to these substances was not shown to increase the risk of Lymphoma. The possibility exists that exposure to other unstudied substances may have caused the increase in the incidence of lymphoma. 6 la Cher evidence Chet exposure Co 2,4-D increeses Che risk of Lymphoma i in humans? Several studies of pesticide use among farmers have Led scientists Co beLieve that the use of phenoxyacetic acid herbicides in agriculture, in particular 2,4-D, increases the risk of non-Hodgkin's Lymphoma among people handling this herbicide frequently. 13. Has exposure to 2,4-0 been Linked to other types of cancer in humans? At this time, there is no evidence linking exposure to 2,4-D to increased risk of other types of cancer in humans. NCI continues to study the potential hazards of exposure in specific populations, including lawn care workers, herbicide applicators, and farmers. 14. What shouLd be done to reduce exposure to 2,4-D? The herbicide 2,4-D is absorbed by plants and broken down by sunlight within a few days of application. Exposure can be minimized by Limiting the amount of time spent in treated areas for several days. 15. Are children at risk if they play on lawns treated with this chemical? This study showed that dogs whose owners used 2,4-D on the lawn were more likely to develop malignant lymphoma. While dogs and humans exposed to the same cancer-causing agent may develop the same disease, the resuLts of this study cannot prove that using herbicides on the lawn is dangerous to children. Parents who are concerned should Limit their children's exposure to treated areas. 16. Is 2,4-D the same chemical as Agent Orange? Agent Orange, a mix of herbicides used to strip leaves from trees in Vietnam during the war, was a combination o^2 , 4 - D and 2,4,5-T (2,4,5trichLoro-phenoxyacetic acid). The chemical commonly known as "dioxin" (2,3,7,8-tetrachlorodibenzo-p-dioxin) often forms when 2,4,5-T is manufactured and can be found as an impurity in that herbicide. 17. Is further research pLanned? NCI is currently conducting several studies of the cancer risk of frequent exposure to 2,4-D including studies of cancer among lawn care workers, cancer in county herbicide applicators, and brain and stomach cancer among farmers. These studies will be completed in two to four years. NCI and the NCI-funded Children's Cancer Study Group are also conducting a study of the potential causes of childhood Leukemia, including exposure 20 090 7 Co herbicides Like 2,4-D, and ocher pescicides. The scudy began in 1989 and is expecCed Co concLude in 1994. An NCI sCudy of Che risk, of cescicuLar cancer in men who served in Vietnam, and chus had possible exposure Co Agent Orange and ocher chemicals, wiLL be published in OcCober. A previous NCI scudy of military guard dogs in Vietnam shovedTMan increased risk of seminomas (testicuLar cancers). U-- L8. Where can more information about environmental hazards from exposure to herbicides and other pesticides be obtained? Questions about the potential environmental and heaLth hazards of pesticides can be answered by calling the National Pesticide Tele communications Network (NPTN) at 1-800-858-PEST (1-800-858-7378). NPTN \ is sponsored by the U.S. Environmental Protection Agency. 19. Where can a dog be treated for cancer? For information on treatment for dogs with cancer, contact any college of veterinary medicine, the local office of the Veterinary MedicaL Society, or a private veterinarian. ### For additional information about cancer, write to the Office of Cancer Coamunications, National Cancer Institute, Bethesda, MD 20892, or caLl the Cancer Information Service toil free at 1-800-4--CANCER. Spanish-speaking staff members are available. 2 .0 0 ? / .' V**' 2<oassL I uv'M---. .-**j 15 REPORTS ment should receive further investigation. (J Natl Cancer Inst 83:1226-1231, 1991) Case-Control Study of Canine Malignant Lymphoma: Positive Association With Dog Owner's Use of 2,4-Dichlorophenoxyacetic Acid Herbicides Non-Hodgkin's lymphoma has had the second fastest increase in cancer in cidence rates of all human cancers in the United States over the last 15 years (/). The etiology of non-Hodgkin's lym phoma* i^ unclear, although there arc Howard M. Hayes* Robert E. Tarone. Kenneth P. Cantor. Carl several .well-established risk factors. Known risk factors include age. male gender (2). congenital immunodeficiency R. Jessen, Dennis M. McCurnin, Ralph C. Richardson stales, infection with the human im munodeficiency virus (J). other immunealtering conditions |c.g.. transplant recipients (4)J, chemotherapy for Hodg kin's disease (5), and being a member of A hospital-lnised case-control study of a lymphoma-prone family ((>). com panion dogs examined the risk of Evidence continues to accumulate im developing canine malignant lymphoma plicating phcnoxyacetic acid herbicides, associated with (he use of chemicals in in particular 2.4-dichlorophenoxyacetic and about (tie home. Inform ation from acid (2.4-D). with risk of non-Hodgkin's a self-administered owner questionnaire lymphoma in farmers (7). A population- and/or a telephone interview of about based case-control study of non491 cases, 4 6 6 nontum or controls, and Hodgkin's lymphoma in Kansas farmers 479 tum or controls indicated that reported a 2.2-fold excess risk for those owners in households with dogs that who frequently mixed or applied her developed malignant lymphoma applied bicides (specifically 2.4-D) that rose to 2.4- dichlorophenoxyacetic acid (2,4-D) over sixfold among frequent users (<V). A herbicides to ttieir lawn and/or employed mortality study of 70 000 Saskatchewan commercial lawn care companies to farmers, age 35 years and older, showed a treat their y ard significantly more fre significant positive association between quently than control owners (odds ratio = 1.3). In addition, the risk of canine malignant lymphoma rose to a twofold excess with four or more year ly owner applications of 2,4-D. The non-Hodgkin's lymphoma and the nuntber of acres sprayed with herbicides. predominanllyifi.e.. 75%-90% by weight) 2.4D (9). A recent case-control study of non-Hodgkin's lymphoma in Nebraska findings in this study are consistent found a significant threefold excess in with occupational studies in humans, farmers who mixed or applied 2.4-D which have reported modest associa more than 20 days per year ( 10). With the tions between agricultural exposure to extensive use of phcnoxyacetic acid her 2.4- D and increased risk of non-Hodgkin's bicides in the United States since the midlymphoma, the histology and epidemiol 1960s ( / / ) and the frequent use of 2,4-D ogy of which are similar to those of canine on public park lands, golf courses, and lalignant lymphoma. The present study suggests that human health implications of 2,4-D exposure in the home environ- private lawns ( 1 2 ). particularly by com mercial lawn care companies (/./). the potential exposure opportunities to people and farm and companion animals con tinue to lie substantial. Malignant lymphoma, the canine equivalent of non-Hodgkin's lymphoma, is the most frequently diagnosed liemalolymphopoictic cancer in dogs (14). Ap proximately one in seven companion dogs brought to medical attention at U.S. university veterinary teaching facilities and diagnosed with any type of malignancy will have malignant lymphoma. Presently, seven new cases of malignant lymphoma are diagnosed per 1000 canine patient years-at-risk at U.S. university veterinary medical teaching hospitals (Hayes HM: unpublished results). Although canine malignant lymphoma has been extensively studied, its etiology is obscure. The disease occurs at all ages, but it is more frequently diagnosed as the dog grows older (15)'. as in humans, there is a male excess. Several breeds have been identified at increased risk ( 14). in cluding the lx>xcr (16). bull mastiff ( / T). and Scottish terrier (/S'). Unlike m a lig nant lymphoma in other laboratory and domestic animal species, no canine retrovirus has been associated unequivo cally with the disease, although virus-like Received December 19. 1990: revised M ay ' 1991: accepted M ay 28. 1991. H . M . Mayes. R . E . Tarone. K . P. Cantor. Epidem iology and Biostatistics Program . D ivision o f Cancer Etiology. National Cancer Institute. Bctbesda. Md. C . R . Jessen. College of Veterinary Medicine. University o f Minnesota. St. Paul. D. M . M cCum in. Veterinary Teaching Hospital. College o f Veterinary M edicine. Colorado Stale University. Ft. Collins. R . C . Richardson. Veterinary C lin ical Sciences. School o f Veterinary M edicine. Purdue U n iv e rsity. West Lafayette. Ind. We thank M . A . Tucker. A . B lair. P. S. Ilarlg c. L. A . D rinlon. and G . K . O g ilvie for review ol the manuscript and S . M. Xalim for helplul comments. to: It. M .. Mayes. D .V .M .. F-U" vironmental Epidemiology Branch. National In stitutes o f Health. Execu tive I 'la/a North. K m . -t-D. Itelhesda. M D 20892. ^ oo?3 I:iinc<cs in allecled tissue have Ixxn reported ( 1 0- 21 ). Naturally occurring canine malignant lymphomas are similar to human non-Hodgkin's lymphoma in histologic type, biological behavior, and jell' responses to the same chemother;i|>eutic agents (22 -2 4) . These malig nancies have occurred in companion pets and their owners in the same household (25.26). The pel dog is normally fastidious about its grooming habits. It is common for a dog to clean itself by licking its paws, thereby ingesting any substances it has walked in or that have been applied to its body accidentally or intentionally as therapy or prevention. The companion dog is not generally subjected to direct occupational exposures experienced by its owners, although it may have contact with particulate matter or residues brought- into the home on its owner's clothing and shoes ( 27 ). Thus, environ mentally related cancer in the pel dog may be considered primarily the result of local exposures in and around the owner's home. To investigate canine malignant lym phoma and its relationship to chemical exposures, particularly the herbicide 2.4- . we conducted a case-control study of dogs diagnosed with malignant lym phoma at three university veterinary medical teaching hospitals. We used a questionnaire to obtain information from dog owners on demographics, lifestyle, and household and lawn chemical ex posure. Materials and Methods Subjects Dogs with histopathologically con firmed malignant lymphoma, newly diag nosed from January 1984 through February 1988, were identified using computerized medical record abstracts from three veterinary medical teaching hospitals that participate in the Veterinary Medical Data Program (25!). The col laborating veterinary medical teaching hospitals were the University of Min nesota at St. Paul. Purdue University at West Lafayette. Ind.. and Colorado State University at Ft. Collins. <u'o comparison groups of hospital i s e d control animals, matched by age group, hut not by sex. were chosen from dogs seen at the same veterinary medical teaching hospital in the same year as the identified case, with one-to-one matching for each control group. Age was reported in range values on the Veterinary Medical Data Program medical record abstract (i.e.. I. 2-3. 4-6, 7-9. 10-14. and 15+ years). The first control group was selected from all other tumor cases diag nosed at the veterinary medical teaching hospital, excluding transitional cell car cinomas of the lower urinary tract be cause of a potential etiology related to chemical exposures. Twenty percent of these tumor controls had bone cancer, 19% skin tumors (excluding melanoma). 13% oral and/or nasal tumors. 7% mast cell tumors, and 41% other neoplasms. The second control group was selected from dogs ceil for any other diagnostic reason, excluding those with conditions possibly related to chemical exposures (e.g.. nonspecific dermatitis, neuro pathies). Twenty-six percent had degen erative diseases. 18%- broken bones or soft tissue trauma. 14% diagnostic ex aminations or elective surgery. 14% in fections. and 28%- other conditions. Written Questionnaire/Tclephone Interview Owners of selected case and control animals were identified by the participat ing veterinary medical teaching hospital. A standardized questionnaire was sent to the last known address of each owner. The cover letter to the questionnaire noted that the dog in question had been seen at the veterinary medical teaching hospital on a particular date and was selected to be part of a study being con ducted by the National Institutes of Health. No mention was made of the animal's diagnosis at that time or of the disease of interest in the study. The ques tionnaire requested information about demographic characteristics of all people living in their home, basic information about the pet's life history, household use of chemicals (in and about the home) in cluding those directly applied to the pet. and personal use of and/or the employ ment of commercial companies applying chemicals of whatever kind on the lawn and garden. In addition, owners were asked about opportunities for exposure of their pels to lawn and garden chemicals, including frequency and seasonality of access to the yard. The questionnaire did not provide a list rtf chemicals which the owner could consult in responding to the various questions regarding home. lawn, and gardening chemicals. After 2 weeks, owners who did not respond were sent another letter request ing participation. U.S. Postal Service torwarding addresses were used when available. If there was no subsequent response, an attempt was made to conduct a telephone interview using the stand ardized questionnaire. Trained inter viewers were not told the name of the disease or exposures of interest or which households had case subjects. Data Analysis The relative risk, as estimated by the maximum likelihood estimate of the odds ratio (OR) (29). was used as the measure of association between malignant lym phoma and the variables of interest. Sum mary ORs and 95% confidence intervals (CIs) were calculated, stratifying to con trol for potential confounding factors as judged by a change in the effect estimate. The Haldane-Smith-Mantcl procedure was used to test for trends in proportions (50). Logistic regressions evaluated the effects of several factors simultaneously (3/). Results Malignant lymphoma was diagnosed in 588 canine patients that met eligibility criteria. Questionnaires were returned or telephone interviews were completed by 491 (84%) households with dogs having malignant lymphoma (case households). Of the 583 nontumor control households. 466 (80%) responded: of the 586 tumor control households. 479 (82%) responded. Forty-five percent of the total completed responses came from telephone inter view's. Most nonresponses were due to the inability to contact the owner. Only four owners (owners of one case animal and three tumor controls) who were con tacted refused to participate. There were no discernible differences between study groups in demographic variables, although median income was slightly higher for case households versus control households, as was the number o! household members supported by that in come (Table I). Slightly fewer case than ^ " I No. 17. September 4. Idol '2-00'/^ R l: PORTS 1227 r T a b le L Demographies l . i k s i \ U* and environmental characteristics ol reporting households expressed a1. .1 percentage (1 the eligible studv croup the employment of commercial lawn ser vice was 1.9 (95% Cl = 0.XX-4.I4) (Table Characteristic M ali!:nant lymphoma <n = -M l) Nontumor control tn = 46b) Tum or control 3). Adjustment for age in these analyses tn = 479) did not change the ORs. Liquid 2,4-D was sprayed by fewer owners of case Median income* S3 3 500 S-10 500 S31 MKI animals than owners of control animals Mean household members supported by income* M ean No. ot' years dog lived with owner 1Ionic location In rural area W ithin 2 miles o f factory o f plant that emitted gases, fumes, or particles o f some kind P et's diet Com m ercial dog food Drinking water from community source 2.77 9.32 29.5 12.4 97.6 69.2 2.67 2.68 applying 2,4-D. a difference that was ot significant. 9.05 9.75 A positive trend (/*<.02) was evident for the yearly number of owner applica 32.8 32.6 tions of 2.4-D. This trend was significant 13.5 16.9 (/'<.05) for both long-term use (>5 years' duration prior to diagnosis) and short term use (<5 years' duration). No sig 97.0 97.7 nificant association was observed 72.5 70.4 between malignant lymphoma risk and duration of owner application of 2,4-D Potential for in-home chem ical exposures Dog allowed inside home Home had carpet/area rugs Carpets or rugs ever com m ercially cleaned Drapes present in home Drapes ever com m ercially cleaned Cigarettes smoked in home Insecticides used in home E v e r applied fle a and tick preparations 95.7 93.9 12.1 88.8 45.9 47.5 37.9 6.3 Potential for exposures outside o f home O w n er had yard used by dog Chem icals ever used on yard during dog's lifetim e Dog ever in A m erican Kennel C lu b s I k i w or hunting trial Doc ever com m ercially groomed 95.7 67.2 10.6 22 95.3 93.3 12.6 86.7 46.3 46.4 38.0 7.1 94.9 67.6 11.6 3.9 (OR = 1.5 for >5 years of use; OR = 1.3 for <5 years of use). We could not distin 93.3 92.5 guish possible latency effects from dura 1l . l tion effects, as the questionnaire did not ask for the date on which the owner 87.5 48.0 began 2,4-D use or the employment of 48.2 professional lawn care services. No dis 42.2 cernible trend was observed for yearly 7.6 number of commercial lawn treatments (Table 3). 94.8 Multivariate analyses considered the 64.9 cver-never variables of Ilea and tick con 10.6 trol exposures (i.c.. collars, powder, spray, dip), in-home use of insecticides, 2.5 cigarette smoking, the presence of carpet and window drapes (and whether they * Percent not answ ering question was X .P * o f case households, 8.4^ of nontumor control households, and b.X'/ir o f tumor control households. * Percent not answ ering question was 3 .V o f case households. 4.3** o f nontumor control households, and 5.2 '* o f tumor control houseliolds. were commercially cleaned), use of yard insecticides and/or fertilizers, dog breed risk group and sex. and application of 2.4- D by owner and/or commercial lawn care service. Separate analyses con control households were located in rural areas. However, case households were less frequently located within 2 miles of industrial plants or facilities reported as emitting into the atmosphere. Dogs studied in this investigation usually ate commercial dog food, and most were provided drinking water from a community source. They were usually allowed inside their owner's home, and most of these homes had carpel and/or area rugs and window drapes. Most dogs were allowed access to their owner's yard. Lifestyle experiences and exposures inside and outside the home were quite milar between the two control groups, ijccausc of a lack of discernible differen ces between tumor and nontumor control households, subsequent analyses utilized pooled controls (Table I). Using mixed breed dogs as the standard {OR = l), sidered 1) number of applications per canine breeds previously identified as year of flea- and tick-control products. 2) having a low risk of malignant lymphoma number of applications of household in (14) were found to be underrepresented in secticides per year, number of cigarette the case series (OR = 0.3); breeds ex packs smoked per year by household pected to be at high risk had a twofold ex members, and the number of limes car cess representation (Table 2). Approx pets and drapes were commercially imately two thirds of all households sur cleaned per year for dogs allowed inside veyed had used lawn chemicals of some the owner's home, and 3) number of ap type on their yards (OR = 1.1). When all plications per year of yard insecticides dogs who did not have access to their and fertilizers for those yards available to owner's yard were classified as uncx- the dog. The only factors related to case posed. the OR for owner application of households in the multivariate analyses 2.4- D and/or the employment of commer were breed risk group and owner applica cial lawn care service was 1.3 (93% Cl = tion of 2.4-D and/or use of a commercial 1.04- 1.67). No difference in OR was lawn care service. detected for owner application of 2,4-D When the OR for owner application ol only (OR = 1.3) and sole use of commer 2.4- D and/or commercial lawn treatments cial lawn treatments (OR = 1.3); how was evaluated for each breed risk group, ever. the OR for owner application plus an inverse relationship with breed risk l V)  able 2. O K s u l canine malign.ml l>m|>homa by host Characteristic No. o f eases No. o f controls OR Breed* M ixed breed Lo w -risk breeds combined Average-risk breeds combined ! ligh-risk breeds combined Sex* i'e m ale Male I2X 16 22X m 237 234 243 1.0 94 0.3 496 0.9 112 2.0 50X 1.0 437 1.2 93% Cl 0.1 7-0..39 0.66-1.14 1.40-2X3 0.97-1.34 Adjusted for sex in calculating bree d risk group O R s and breed risk group in calculating the effect o f sex. T a b le 3. OR-; o f canine malignant lymphoma by exposure Characteristic No. o f cases No. of controls OR Owner application o f any yard chemical and/or commercial lawn treatment* No Yes O w ner application o f 2.-4-D and/or commercial lawn treatment* None or dog never allowed in yard Yes Com m ereial la w n treatment on ly* Owner application o f 2.4-D only* O sm er application o f 2.4-D and commercial lawn treatments* No. o f com m ercial lasso chemical treatment s/y t 0 1 2 >4 No. o f owner applications of 2 .4 -D /vt 0 or dog never allow ed in yard i 2 3 >4 161 330 300 191 113 60 16 300 16 20 19 76 300 20 2X 11 17 312 1.0 62 l.l 641 1.0 304 1.3 1X9 1.3 w 1.3 16 1.9 M l 1.0 23 1.4 23 1.3 33 l . l 122 1.3 M l 1.0 34 1.3 47 1.3 17 1.3 17 2.0 Trend test 95%. C l 0 .X 4 -I.3 7 1.04-1.67 0.96-1.70 0.90-1 .`10 0 .X X -4 .14 0.70-2.X4 0.75-2.X6 0.60-2.14 0 .9 5 -I.X7 0.72-2.50 0.74-2.13 0.57-3.13 0.92-4.15 /*<.()2 ` Adjusted lor sex and breed risk group. ` Adjusted for sex and breed risk group: includes 16 cases and 16 control households where the ow ner ap plied 2 .4 -0 and also had com m ercial lawn treatments. level was found, the OR being highest for low-risk breeds (Table 4). Male dogs, known to be at elevated risk for malig nant lymphoma, exhibited the same OR 0 .3 ) as females for owner application of 2.4-D and/or commercial lawn treatment (Table 4). To test recall, owners were asked if their dog had died: if so. they were asked d it had cancer and the cancer type was requested. Among the owners of case animals who reported never using 2.4-D md/or commercial lawn care services or never allowing their pet access to their yard. s|jnhily more knew that their pet died of malignant lymphoma compared with the case households that applied 2,4D or employed commercial lawn services (Table 5). Discussion We found a modest association (OR = 1.3: 95rf CI = 1.04-1.67) between malig nant lymphoma in companion dogs and their owners' applying 2.4-D on their lawn and/or using commercial lawn care services, with the risk rising with increas ing number of owner applications per year. These findings are consistent with studies suggesting that agricultural 2.4-D use increases the risk of non-Hodgkin's lymphoma among farmers (S-I O). Also consistent with human studies was the lack of a significant association with dur ation of 2.4-D application by the owner. Findings of case-control interview studies may be biased because of dif ferential recall of cases and referents. In the last several years, results from several studies of this design have suggested links between use of phcnoxyacctic acid herbicides and development of nonHodgkin's lymphoma in humans. Some have suggested problems in recalling may explain these associations ( 13). In the current study, recall bias could have oc curred if owners of case animals were more likely to remember using such her bicides on their yards and, therefore, ex posing their pets. The lack of any association, among cases, of owner's knowledge of their pet's type of cancer at death and opportunity for exposure to 2,4-D strongly argues against such a bias in this study (Table 5). Any recall bias related to socio economic status probably did not affect this study because case and control households were strikingly similar in median income, number of household members supported by such income, use of commercial dog food, home lifestyle (service by municipal water supply, presence of carpct/arca rugs and drapes, access inside the home by the dog), or the presence of yards. Bias may also arise from an unwillingness of certain selected subjects to participate. The percentages of responses from case and control household groups, however, were nearly the same. The major limiting factor was the ability to locate owners, not an un willingness to participate. Only four of 1440 successfully contacted owners refused to participate. Recall may be hampered by the lapsed time between the time period of interest and the questionnaire/interview date. The response data analyzed in this study were acquired 10-58 months after the selected animals were seen at the veterinary medi cal teaching hospital: however, control dogs were matched to case dogs by age and year of medical attention. Therefore, the lapsed time between the period of in terest (the years the dog lived with its owner until the date of university treat - -- ' si ^ is '. N o. 17. S e p le n iP c r 4 . 1991 R lil'OKTS 1229 1 ab le 4 . O K s for e xp o su re to o w n e r-ap p lied 2 .4 1 ) atul/or co m m e rcial la w n treatm ents fo r m alicn au t Ivin plunna hy canine breed risk croups ami hy sex C h a ra c te ris tic N o. <*1 cases l\ \ p o > e d Unexposod No. o 1controls Imposed Unoxposed OK* '15*. C l R isk group M ix e d breed ! (!w-n*A breeds Average-risk breeds 1Iigh-n>k breeds Sex Tem ale M ale 45 X.4 SO 10.4 1.1 0.0*1- 1.7S V V' yo ( >ft 4,(i(r*u,07 (M 1.47 155 .441 1.4 1.0.4-2.04 47 7 ; 40 72 1.2 0 .0 7 -2 .1 0 SS 14*) 152 .450 I..4 (1.4)4-1.87 10.4 151 i s : 2X5 I..4 0.4)4-1.84 * Adjusted lo r sex in calculating breed risk group O R s and breed risk group in calculating the effect o f sex. 1 able 5 . O w n e r's recall o f pet's m alignancy at death by yard chem ical use among malignant lymphoma case households Yard ch em ical use No, * O w ner's knowledge of m alignant lymphoma. ck O w ner applied 2,4-L) only Com m ercial law n treatment only O w ner applied 2 .4 -D and also used com m ercial lawn care treatments O w ner neither applied 2.4-D nor used a com m ercial lawn care com pany or the pel was not allowed in the yard 00 115 16 .400 01..4 00.8 58.8 04.8 inert) and {lie date of the questionnaire or interview was approximately the same for owners of case and control dogs. It is possible that unmeasured eonfounders could explain the risk, but we collected data on the most likely con comitant exposures in the household en vironment. Household insecticides, contact insecticides for Ilea and lick con trol. and outdoor insecticides played no discernible role in the association be tween 2.4-f) exposure and development of malignant lymphoma. We controlled for sex and breed association in all analyses: these two factors play a role in the occurrence of the disease, but they did not strongly affect the risk from 2.4-D ex posure. Although viruses have been as sociated with lymphoma development in many animal species (52). they have not been linked to lymphoma development in the dog. While viral infection of some type may he an important factor, it is un likely to be responsible for the significant association between canine malignant lymphoma development and owner use of 2.4-D and/or commercial lawn care ser vices. The type of 2.4-D applied hy the owner, iiquid spray or granular, had little influence on risk. The consistency be tween the ORs for male and female case animals with 2.4-D exposure also argues against inhalation of herbicide spray as a likely route of exposure, because the male dog's sniff-scent marking tendencies might place males at greater risk than females (33). Occupational studies of her bicide exposures by agricultural workers generally agree that inhalation is a minor source of human exposure (34). A major weakness of the current study is the lack of precise exposure data for herbicides. Frequency of application was the primary measure of exposure used in this study. The lack of a clear pattern of association with total number of yearly commercial lawn treatments contrasts with the consistent application-response relationship (/J><.()2 for trend) exhibited for owner application of 2.4-D (Table 3). The yearly number of applications of 2.4D by the owner is more likely to reflect the actual 2,4-D exposure opportunity for pel dogs than the number of lawn treat ments by commercial lawn care com panies because some commercial treatments during each year may contain only fertilizer and/or insecticides and. thus, not lead to herbicide exposure. In a laboratory carcinogenicity study, 12 male and 12 female beagles were ex posed to 2.4-D in their diet lor 2 years t.'3). No malignant tumors were reported in any of the exposed dogs or in the three male or three female control dogs. A re. lated 2-vear feeding study ol 2.4-D in f)s- horne-Mcndcl rats (35) resuited tn an increased incidence of lymphosarcomas in exposed male rats (10 of 125 exposed male rats developed lymphosarcomas compared with 0 of 25 control rats; Fisher's exact test yields P = .15) but no m evidence of increased tumors in female rats (1 of 125 exposed female rats com pared with 0 of 25 control rats). Studies involving IS male and 18 female mice from each of two strains gave no evidence of carcinogenicity for 2.4- D fed for 18 months (36). The isopropyl, butyl, and isooctyl esters of 2.4- D were also tested, and no significant increases in tumor incidence were ob served (36). In companion studies. 2.4-D and its esters were administered by sub cutaneous injection in IS-month car cinogenicity experiments. Although the results of the injection studies were not given in the initial report ( 3 6 ). a sig nificant increase in reticulum cell sar comas was reported lor the isooetyl ester of 2.4-D in one female mouse strain (37). No reticulum cell sarcomas were ob served in the IS male mice ol the auie strain similarly exposed. Thus, although Ivmphorcticular tumors have been found in excess in two rodent studies, laboratory animal experiments provide only weak sup|x>ri for a role of 2.4-D in lymphoreticular carcinogenicity. The current investigation began with an a priori hypothesis that exposure to 2.4-D might be associated with development ol malignant lymphoma in the pet dog. Our findings support this proposition, with a modest statistically significant OR ol 1.3 for owner application of 2.4-D and/or commercial lawn treatments. Further sup port derives from the higher risk observed for pets of owners who applied their own 2.4- D in addition to exposure to commer cial applications (OR = 1.9; 9 5 % Cl ~ 0.88-4.14) and the increase in risk to a twofold excess with four or more owner applications per year. In view ol a sug gested link between non-Hodgkin's lym phoma in humans and farm exposures to 2.4- D. and because malignant lymphoma \2M ) <200^ 7 Journal o f the N ational C ancer hiMiiui.' mi ilk (log is considered die cquiviilcni ol non-Hodgkin's lymphoma in humans, our linding ol an association lieiween 2.4-1) ex[X)sure and development of malignant lymphoma in companion dogs suggests that the human health implications of 2.4-0 exposure in the home environment warrant further investigation. References (/> N.v m o v u . C l n i i r io k H i-.a i. im S i a m s i k s : Cancer S ciiisfics Review I973-X7. DHHS Puhl No. (N IH 90-27X9. Beihesda. M d: Na(011*1i Center for Health Statisties. I y VO (2 ) S i k v i j i .lan ctu E p id c m io i.o g v . an d E nd R c s u .Ts : Incidence and M o n a liiy D ata. 197377. Natl C a n ce r Inst Monogr 57:1-10 X 2. 19X1 (. i E k in .k C E : Acquired immunodeficiency syndrome-associated lymphomas. In H ie Lym phomas: Current Concepts in Patho-gencsis and Management. J Natl Cancer Inst Monogr 10:45-54. IVVO (-/) Hu u v i-k R . Ek.m m i-s i J E J k : R isk o f cancer in renal Transplant recipients. Lancet 2:55-57. 1973 (5 ) T u k i .k M A . C o l i.m a s C N . C o x R S . t:T ai R isk ol second cancers alter treatment for Hodgkin's disease. N Engl J Med 3 lX :7 iv-X I. I OSS - <r>) (ik tJ.M : M il : N on-H tklykm ' s lymphoma and mycosis fungoidcs. ht Cancer Epidemiology and Prevention (Schoitcnfcld I ) , Traumem JC Jr. cds). Philadelphia: Saunders 19X2.pp 75477X (7) Ih.Aik A : Herbicides ami non-Hodgkin's l\m phonia: N ew evidence from a study o f SaskatchewaiH'aniXTs. J Natl Cancer Insi K2:544-545. 1990 I I o ak S K . lit MR A . lloi.M t s E E . i t \t : Aericulttir.il herbicide use and risk o f lymphoma and soft-tissue sarcoma. JA M A 250:1141-1147. 19X6 (V) W u a.i: DT. S i;mi \ n \ v R M . W n.k in s K . i t a i.: M onaliiy studs o f Canadian male farm operators: N on -lh 'dg kin 's lymphoma mortality and agricultural practices in Saskatchewan. J Nall Cancer Inst S2:575-5X2. 1990 (10) Z aIIM S I L W f lSLNBl ROCK D D . liA H ItllT P A . i t a l : A case-control study o f non-Hodgkin's lymphoma and the herbicide 2.4-dichlorophenoxyacctic acid <2.4-D) in eastern Nebras ka. Epidem iology 1:349-356. I WO a h E io h -k s T O . A n d r iu -n as P A . A n d lkso n T W : banners' U se ol Pesticides in 1976. A g ric u l tural Econom ic Report 4 IX . Econom ics. Statis tics. and Cooperatives S e rvice . Washington. DC: L'S D A . IV7S I.Mi k s vnoNAt. A g i-n c y io k R i:si:.\k( n on C vNctiK: Som e fum igants, the herbicides 2.4D and 2 .4 .5 -T . chlorinated diben/odioxins and miscellaneous industrial chem icals. IA R C Monogr E va l Carcinog R isk Chem Man 15:1354.1977 ( : ` 1 ('Il M il. M)R At M CI LIVRA I. Sot M-: AND `ft .CM n u io u y: Comments from C A S T . Perspectives on the safety o f 2.4-D . A m es. Kss;t: C A S T , December I9S7 ( Pun s n K W A , M c K w P W : The occurrence of tumors in dom estic anim als. Natl Cancer Inst Monogr 5 4 :1 -2 1 0 . p*SM h >) S c u m i i h k R : C o m p a riso n o f aec- .nul \ c \ spccilie incidence rate patterns ol the leukem ia n .m p k -. in IIn- cat and ||,L- ,| ,, f . j \ C I 70:971 977. I 9X3 ( / ( ij P r i i m i r W A : Canine lyinphnrn.'i: Kelm ive nvk in lhe G ive r breed. JN C I 39:K33-K45. 1967 ( 1 7 \ O mons 1)1:: A pro.|>eelive survey o f fam ilial canine lym phosarcom a. J Nail Cancer Inst 72:909 -912. 19X4 (/.S') llv y i.s IIM : llo sp ila l prevalence o f cancer in die Seoiiish terrier. In Scoilish Terrier C lu b o f Am erica Hamlltook (Cooke C M . Mach S . ed si. srd ed. I.yn ch h u r . V a : Scoilish Terrier C lu b of A m erica. 19X6. pp I5 X - I6 7 t / 9 i C m si-m a s A L . Boi-l- W J. B k k .m iw ii.i . A S . IT si.: Prelim in ary report on virus-like panicles in canine leukem ia and derived cell cultures. Cancer Res 2 7 :IX - 2 5 . lJf.7 ( 2 0 1 C o n e s I I . O ia i'Ma s A L . L in e r JW . i :t a i.: C e llu la r transm ission o f canine lymphoma and leukemia in beagles. J Nail Cancer Inst 4 5 :1 0 0 - 1 (1 2 3 .1 9 7 0 (2/1 K i m e rit R : V irus-like siruciures in canine Ivmphatic leukosis tumor cells. Berl Munch T ie ra r/ll W ochcnschr K 4 :6X -70 . 1971 (22) M s c L s s c s E G . B rown N O . P a is a ik A K . i :t a l : C y c lic conihinalion cheinoiherapy of canine'iymphosarcoma. J A m Vet Med Assoc 1 7 8 .I4 7 X -I I X I . 19X1 (2.11 W c i.i .c k R E . T iic ii.i .n G M . M a d i.w c i.i . B R : Cheinoiherapeutic responses in does with lym phosarcoma and hypercalcemia. J Ant Vet Med A ssoc IX I:X 9 |"-X 9 3 . 19X2 (2 4 ) M s c E w i.s E G : S|x>ntancous tumors in dogs and cals: M odels for the study o f cancer b io l ogy and treatment. Cancer Metastasis Rev 9 :125-13(1.1990 (3 5 ) S<_h ir i R A . M i .c iiv E C : Naturally occurring canine and feline lymphomas as biomedical models. In A nim al Models for Biomedical Re search. National Academy ol Sciences Kcpon. I'ubl No. 1594. Washington. D C : N A S . I96X (361 Mm m iii.ro C A . W i i .son E D : Lym phoid neoplasia in the canine. In The Canine as a Biomedical Research Model. Immunological. Hematological and Oncological Aspects iSlufrine M . W ilson ED . cdsi. Washington. D C : Technical Information Center. US IX-partmciit o f Energ y. 19X0. pp 290-313 (2 7 | G i .'Ckm a n L T . Do x iv n s k i L M . M a o c ir i; T G . ill \L: .Mesothelioma in | v l dogs associated iih exposure o f their owners to asbestos. E n viron Res 32:3 0 5 -3 1 3 . 19X3 (2 A'l H vv i s MM. W i i .son G P . Mo r a ii- M: The veterinary m edical data program (V M D P l: Pa-t. present, and future. In Proceedings o f In ternational Sym posium on A nim al Health and Disease Data Banks. December 4-6. 1978. A P H IS M isc I' ubl No. 13X I. Washington. D C : L'S D A . 1979. pp 127-132 (2 9 ) G vri J J : Point and interval estimation o f the common od Js ratio in the combination o f 2 x 2 tables with fixed m arginals. Biom etrika 57:471-475. 1970 t.!()| T hom as D G . G a r t J J : Stratified trend and honioeeneity analyses o f propoilions and life table data. Coinpnt Biom ed Res 16:116-126. I9S5 ( M | E noi i .m.vn L : Slepvsise logistic regression. In BM D P Statistical Software (Dixon W J. ed). Berkeley: L n iv C a lif' Press. 19X5. pp 330-344 153) lu t i'.is GM . M v d iw iii B R : l.eukemia-sareonia disease com plex. In Veterinary Cancer Medicine i lheilen GM . Mudcwell B K . etlsl. I'lul.'.delphia: Lea \ l ebiger. |9 M . pp 704- 20X S i s u a i K l l . A s i s m i J I : I : I itiun.ttiv'ii patterns in the lahoraiorv Ivag le . Behaviour 47:257 3t~. 197 1 (.(4 ) V i a r y K A : U n u a ry e xcie lio n of 2.4-1) in commercial lawn specialists. Appl Ind lly g 1:119-121. 19X6 (3 5 ) M a n s i.n AVI I . (J t Ait t: M L . M a u i.rm vnn K T . i t a i.: Chronic toxicity o f 2.4-dichloropheno xvacclic acid in rats and dogs. To xico l Appl Pharmacol 2 0 :12 2 - 129. 1971 (36) I s m s JR M . lit.i.A S ii B M . V a i.i .rio M G . i t At.: Bioassay o f |K` sticidcs and industrial chem icals for m m origcn icily in m ice: A prelim itiarv note. J Natl Cancer Inst 42:1 ! ( ) ! 1114. 1969 (37) Rl:t tu R M D : C arcin og enicity and toxicity o f 2.4-dithloroplieiioxyacctic acid. Sci Total E n viron 3 1-.2(13-2 IX . 19X3 -Grade, Latent Prostate :er Volume: Predictor of Clinical Cancer Incidence? Alice SyWhittcmore* jjoseph B. Keller. itehecca Betenfikx VVc hypollicsiVe Ih a t^ a c h cell in lowgrade ((leasoii grade 1-3) prostate cancer tissue isMt rjsk of transform a tion into a cell wnicit produces a highgrade (Clcason ^rade 4-5) clinical cancer after a short period of growth. As a consequence/, the volume of lowgrade, latent cancer tissue in the p ro s tate glands of! men at any age determines their incidence rate for high-grade, clinical cancer a few years later. Autopsy and incidence data for both white men,and black men support this conclusion/ with a tum or growth period of about 7 years. The transfor mation rate is sim ilar for black men Received January 3. I W I : revised A pril -X. |V 9 |; accepted June 5, 1991. A . S . W hiitem oiv is supported bv an Outstanding Investigator A w ard (C A -4 7 4 4 X ) from the National Cancer Institute. ' National Institutes of Health. Department o f Health and Human Services. A . S. Whittemore. D ivision o f Epidemiology. Department o f Health Research and Poliev. Stanford University School o f M edicine. Stanford. C alif. J. B . K e lle r (Departments o f Mathematics and M echanical Eng in eering ). R . Betensky (Department of Statistics). Stanford U niversity. Stanford. C alif. We thank John M cN e al. Pelayo C o n v a. and R vu ich i Yatani for providing data on latent cancer volume and prevalence and Dee W est and Thom as A. Stanley for helpful comments. ^CtnivspimJi'ncc ht: A lice S. Whittemore. P h.D .. Department o f Health Research and P o lic\. Stanford U n iversiiv Sclu>ol o f M ed icin e. H R P Build in g . Room 113. Stanford. C A 94305-5092. \ ^ ' S.v No. I 7. S o p ic m lv r 4 . I9 9 | REPORTS 1231