Document k91kDL3Qgor2gy5OVErvn8LVO

RECENT EPIDEMIOLOGIC STUDIES OF PCBS VlUi*a R. Gaffey Honaanto Company 00 N. Lindbergh Blvd. t. Louia, MO 63167 Before the Yuaho incident in 19681 published atudiea f the health effecta of PCBs, with the exception of Meiga et al (1954), were eaaentially clinical atudiea of occd^ aional accidental aevere expoaurea. In the following decade ten epideniologic atudiea were publiahed (in addition to the any reporta on the Yuaho incident itself), all but one of utllrh dealt with occupationally exposed populations, and 1Y one of which studied the relationship between cancer and PCS exposure (Gaffey, 19S2). DISCLAIMER This document is a prelialaary draft. It has not been released formally by the Office of Drinking Water, B.S. Environmental Protection Agency, and should not at this stage be conatrued to represent Agency policy. It ia being Circulated for concents on its technical merit and policy lmplicationa. MONS 011524 1 2 KB EPIDEMIOLOGY Since 1978 there have been ten epidenielogic atudles f the health effects of FCBs. Seven of then are cross; sectional investigations of symptou and biochemical parameters. Of these, five are occupational, one is a Study of both occupational and non-occupational exposure, and one is non-occupational. The reaaining three reports are cohort aortality studies of occupationally exposed groups. In addition, there has been a report on cancer Mortality la the Yusho population. This paper briefly siiaaiarizes the status of the epidemiologic evidence on health effects as of 1978, and reviews and evaluates the subsequent studies in detail. IB EVIDENCE AS OF 1978 Ike Yusho incident of 1968 at first appeared to be a classic exaaple of the effects of a aassive ingestion of KBs. Over 1000 Japanese became ill after eating a cooking 11 contaninated with Kanechlor 600, a PCB coapound of Japanese Manufacture. The Most enuunn acute syaptoms observed were kyperpigaentatiou and acne-like lesions, central nervous system symptoms, discharge from the eyes, and vomiting and diarrhea. These symptoms were dose related, and some of them persisted for as long as six years. laboratory testa showed disturbances of liver function and fat netaholism, and there were clinical reports suggestive of some abnonalities la the children of Yusho mothers (U10SB, 1977). HONS 011525 3 KB EPIDEMIOIOCT As the sad of 1977, SI deaths ssonf Tusho patients kid beta identified, with an apparent excess of cancer deaths (Urabc et al, 1979). However, the data were not adjusted for age, and the coapleteness of ascertaiaaent of the deaths is not known. The elapsed tine iron ths original Incident to the reported deaths (one decade) was short enough to cast doubt on whether the cancers could in fact have arisen iron the Tusho*exposure. In any case, in the decade following the Tusho incident analytic techniques for identifying PCBs and aillied conpounds inproved 'considerably, and the cooking oil was reanalyzed. The new analyses showed that the Tusho patients had in fact ingested about the sane anount of polychlorinated dibenzofurans (PCDFs) as of PCBs, in addition to polychlorinated quater-phenyls (PCQs). Current deteminations of blood and tissue levels of these other conpounds in Tusho patients have shown total levels about equal to that of PCBs (Xinbrough, 1980). It therefore appears that the epidealologic findings in the Tusho incident should properly be attributed to the aore toxic PCDF rather than to PCBs. There were nine other published cross-sectional stadias of syaptoas and biocheaical paraaeters related to KB exposure, one of which was non-occupatlonal, aad one occupational study of cancer aorbidity and aortality. In suaaary, the occupational studies found chloracne or other deraatltls, and wild liver function abnoraalitiea HONS 011526 4 KB EPIDEMIOLOGY lft the absence of clinical illness, to be associated vith PCI exposure. Two studies of cholesterol levels gave contradictory results, and one study of triglyceride levels showed an increase. The non-occupatioual study shewed no association between blood PCB levels and any of the Yusho aynptons, but did not exaaine liver function or fat metabolism. The cancer study found an*apparent excess of nalignant melanoma based on three esses, but was later withdrawn because of concern about whether the exposed population had been correctly identified.. HCEHT STUDIES lhe seven cross-sectional atudies completed aince 1978 addresa a broader range of health effecta than did the pre-1978 studies, and in some cases used foraal statistical techniques to take account of confounding variables such as age, sex and weight. Two of the studies, Fischbein et al (1979) and Smith et al (1992) evaluated their findinsa in relation to higher chlorinated and lower chlorinated PCBa separately. The studies' findings can be clasaified under five general headings; dermatologic effecta, liver function, fat metabolism, other objective findings, and Reported symptoms and illnesses. The cohort mortality studies, although they collected data on all causes of death, .were concerned primarily with cancer. .. _ . HONS 011527 S PCI epidemiology Dtntttolailc Effect* Table 1 suoaarizes the results of the five studies that reported on densatologic effects. The negative study f Baker et si (1980) exaained 18 exposed workers, 19 cabers of their faailies, sad 89 coaasunity residents with exposure to fertilizer containing PCBs. The rest were occupational studies, two of which found chloracne and all of which found deraatitis. * Fischbein et al (1979) reported that SO percent of 326 capacitor oanufacturing workers reported a history of deraatological synptons, . the Boat coaaon being a rash. Those with sywptoas had higher blood levels of high chlorinated PCBs. Haroni et al (1981 II) reported ten cases of deraatitis (five diagnosed as active or past chloracne) in 80 exposed workers, but provided no further data on the blood PCB levels of those with deraatitis vs. those without. Chase et al (1982) found increased deraatitis, and sooe chloracne, in the aost exposed group out of 120 railroad aaintenance workera, but within the group found so significant association with blood or fat PCB levels. Saith et al (1982), in a study of 92 eaployees of two utility coapanies, found no consistent association between deraatitis and either high or low chlorinated blood PCB levels. Although the data are not consistent, they suggest that there aay ho a gross dose response relationship between deraatitis and blood levels of PCBs, possibly HONS 011528 8 FCB EPIDEMIOLOGY CMplicittd by variations in Individual susceptibility and in work habits that nay affect absorption of PCBs through the skin. Liver Function Table 2 shows the results of the six studies that investigated liver function. The two non-occupational studies showed no liver function anoaalies associated with blood PCB levels. The first one did not adjust for con-' founding variables (Baker al, I960). The second, a study of 458 residents of a counnity with high euviron- aental levels of PCBs, found an association which disappeared when age and alcohol consumption were taken into account (Kreiss et al, 1981). Of the four occupational studies, three showed various liver function anoaalies related to blood PCB levels. Chase et al (1982) adjusted the results for age. Saith et al (1982) adjuated the data for age and sex. The latter Study showed a significant association of liver function anoaalies with low chlorinated blood PCBs in one of the two plants studied but not in the other. Haroni et al (1981 II) found liver abnoraalities associated with elevated blood KB levels, but the liver abnormalities were defined as liver function anoaalies or liver related syaptoas or clinical findings of abnormality. They state that "only in a few cases was a well-defined liver failure (presence of syaptoas, hepatoaegaly, and abnormal liver findings) present. HONS 011529  rvs inuuuuuAii Fischbein et al (1979) found no liver function aaosulies asaociated with exposure, and in fact conaented oa the "paucity of abnoraal results" in their blocheaical studies. Vith the exception of Maroni et al, the occupational studies agree in finding few or no liver function abnoraalities, and no associated clinical illness. Fat Metabolism Table 3 suaaarizes the. results of the five studies that exaained fat aetabolisw. Fischbein et al (1979) found no association between cholesterol or triglycerides and blood PCBs. Ireiss et al (1981) found an increase in cholesterol with increasing blood PCB levels, bnt no relationship of triglycerides to blood PCBs vhen an adjustnent was nade for cholesterol level. Two other Studies, Baker et al (1980) and Chase et al (1982) agree that cholesterol is not associated with blood PCBs but that triglycerides are. Baith et al (1982) found cholesterol levels to be fosltlvely associated with low chlorinated PCBs in one of the two plants studied. Triglycerides were positively associated with high chlorinated PCBs in one plant and negatively associated in the other, where they were positively associated with low chlorinated PCBs. The preponderance of evidence is that cholesterol is not associated with blood PCBs. There is no obvious HONS 011530 explanation for the aaaodatioa of cholesterol and blood KBs found by Kreiss et al. However, their use of cholesterol as an adjustaent factor for triglycerides inpllea that they view cholesterol as an independent variable that predicts triglycerides. The results of Baker et al and Chase et al do not support this view, since they show that the putative predictor, cholesterol, is not associated with blood PCB levels while triglyceride levels are. The association between triglycerides and blood PCB levels appears to be aabivalent at the least. Other Objective findings Flschbeln et al (1979), Baker et al (1980) and Maroni et al (1981 II) exanined blood cbesristry and found no association with blood KB levels. Kreiss et al. (1981) found a statistically significant positive association between diastolic blood pressure and blood PCBs after adjusting for cholesterol, triglycerides, sacking and race. However, Saith et al (1982) reported no such association. < Varahav et al (1979) reported decreased vital capacity in 243 capacitor aaanfacturing workers compared with published reference standards. However, nost of the study papulation were current or fomer suokers, while the reference etandard is based an a oon-saokiag population. The effects of saoking could be sufficient to explsia the findings. HONS 011531  Kl EPIDEMIOLOGY teported Symptoms and Illnesses tlx of the studies investigated range of symptoms and llloessei. Two of thea reported positive finding* in occupationally exposed populations. Tlschbein ct si (1979) reported a history of gastrointestinal syaptoas la IS percent of 326 capacitor aanufacturing workers, a prevalence of froa 3.0 to 15.2 percent of various musculo-' skeletal syaptoas, and a prevalence of froa 4.8 to 27.8 of various neurological syaptoas. These were, however, unrelated to blood PCB level or duration of eaployaent. Haroni et al (1981 II) reported eight cases of gaatro-' intestinal complaints in 80 exposed workers with no Indication of whether there was a relationship to blood PCB level or duration of eaployaent. They also reported two bleeding haeaangioaas and one case of chronic ayelocytic leukeaia. ~ The other four studies reported no PCB related findings. Specifically Baker et al (1980) found no gtalatloaship with any of the following; fever, weight loss, hasrexia, fatigue, headache, eye irritation, cough, shortness of breath, hausen, vaulting, diarrhea, abdominal pain, arthralgia and 'persistent rash. * Xrelss et al (1981) teported the sane thing for prevalence of illness or weight tees * in' the `previous' year, hereof medication,' use ' of Medical care, history of heart disease,'and percentage of pregnancies ending in miscarriage, stillbirth or infant oath. HONS 0115^ 10 PCS EPIDENIOtOCY Chase ct al (1982) failed Co find any evidence of organ toxicity in a review of the aedical hiatoriea and physical findings in a group of exposed railroad Maintenance workers, and Snith et al (1982) reported no coneistent dose-dependent increase in either synptoas or past illnesses. Their inquiry covered digestive, respiratory, central and peripheral nervous spates, deraatologic, and esotional aynptons. The evidence shows no FCB related syaptoms or illnesses, since the two reports out of six that Mentioned ayaptons failed to show an association with level or duration of exposure. . Carcinogenicity Table 4 shows the Major findings of cancer Mortality is three cohort studies of PCB exposed workers. Brown et al (1981) studied a cohort of capacitor Manufacturing workers in two plants who had been exposed at least three Months between 1946 and 1975 inclusive in one plant or between 1940 and 1975 inclusive in the other. Follow-up wns More than 97 percent cooplete as of the end of 1975, and expected Mortality was calculated iron U.S. population rates, loa-statistically significant exceaaes were found (or rectal cancer, based on four deaths, and liver cancer, based oa three deaths. The rectal cancers showed a slight increase with an increase in the latency period, but the liver cancers showed no consistent trend. There were no HONS 011533 11 PCS EPIDEMIOLOGY UcroMi in nortality fro* these causes associated with increasing lengths of exposure. lertazxi et al (1981) also studied a cohort of capacitor aaaufacturing workers is a plaat near Mllaa. The cohort cooslsted of every person who had worked for at least six Booths between 1946 and 1970, inclusive, except for clerical workers. Follow-up was over 98 percent coaplete at of the end of 1978, and expected aortality was baaed on rates in the city where the plant was located. Data were analyzed separately for Bale (290) and feaule (1020) workers. i statistically significant excess Bortality froa all cancers, based on eight deaths, was found in Bale workers, and a statistically significant excesa for all causes was found in feaale workers, based on 15 deaths. There were non-signifleant excesses of lynphatic and heaatopoietic cancer in both sexes, based on four deaths, and of digestive cancer in Bales, based on three deaths. The sites involved in the digestive cancer deaths were stimach, pancreas and biliary tract. Do analysis by duration or latency of exposure was attespted because of the saall nuabera involved. The authors note that the cohort was very young, and that their follow-up will be continued. lack et al studied all hourly Bale workers who had keen eaployed la the aanufactnre of PCSs in a geaeral cheaical plaat for at least six aoaths between 1945 and 1965 inclusive. Follow-up was approxiaatcly 99 percent HONS 011534 12 KB EPIDEMIOLOGY complete (one person was lost) as of the end of 1977, and expected nortality was calculated froa U.S. population rates. There were eight cancer deaths. The only aoteP worthy finding was a non-statistlcally significant excess in lung cancer based on four deaths. Mo liver cancer deaths were found. DISCUSSION ' Doll (1981) has suggested criteria for establishing carcinogenicity froa epldenlologic evidence. They are Slallar to those proposed by the International Agency for Besearch on Cancer (1980) and are essentially as follows; (1) there is an excess in exposed groups beyond what can be expected by chance, (2) there is an appropriate relatiodrO ship with dose or duration of exposure, (3) there are no known biasing or confounding factors, (4) the association is observed repeatedly in different circuastances. The development of those criteria was stiaulated by a concern to provide generally accepted guidelines for establishing carcinogenicity, but they apply equally well to other health outcones. Criteria for establishing the absence of a health effect are inherently aorc difficult. The cliche that it is impossible to prove a negative is true but ulsleading, in the sense that it is also iapossible to prove a positive ssaociation by scans of epldeaiologic studies. A statist tically significant positive finding allows us to conclude HONS 011&33 > SC ghM-4AWtoW that an aasociatioa exiats, with a knows probability that the conclusion it wrong. Repeated positive studies reduce that probability in a Banner that can be calculated, but there remains a chance that the finding is false. On the ther hand, a negative study allows us to conclude that there is no association, with a probability of error that depends on how great an association really night exist. Repeated negative studies reduce this probability, but there is always a large probability that a very snail effect night not be detected? Therefore both positive and negative studies carry a risk of error, although in both cases the strength of the evidence nay be such as to nake that chance negligibly snail. Ctoas-sectional Studies There is a reasonable consensus that dematitis is associated with occupational exposure to PCBs, although the dose response relationship has not been established in all ef the studies. Two son-occupational studies agree that there is no dematitis associated with environnental, l.e., non-occupational exposure. The difference in the results ef the occupational and aon-occnpatioaal studies is consistent with the order-of-aagnitude differences in the exposures of the two groups. There is a similar preponderance of evidence that aili liver function aaonalles are associated with occupational but not with son-occupational exposure. One occupational HONS 011536 \ 1* KB EPIDEMIOLOGY study found associated clinical tyaptou, but the ress inlet three studies found no detectable clinical illness. By Ball's criteria of consistency, it la very unlikely that clinical illness is associated with the anoaalies found la occupationally exposed populations. findings concerning cholesterol levels again show a preponderance of evidence that cholesterol levels are not associated with blood PCB * levels, even though one non*occupational study did report an association. Ihe situation with respect to triglyceride levels is slightly aore ambiguous-. Two out of five studies, including one non-occupatioaal study, found a relationship with blood FCBs, so that the possibility that such a relationship really exists oust be considered. Studies of blood chesistry were uniformly negative. IWo studies of diastolic blood pressure gave contradictory results. One study of puluonary function suffered frou a failure to account for saoking so that its findings cannot be attributed to PCB exposure. None of these areas of Study appear to show any relationship to PCB exposure. None of the six studies of reported illnesses and ipptini constitute evidence by Doll's criteria. Four studies showed no relationship of a wide range of syuptoas to PCI exposure. Two occupational studies reported various symptoms, but did not show that they were related to blood PCB level or duration of eaployswnt. HONS 011537 15 KB triDEHlOLOGY Cohort Cancer Studio The aost noteworthy characteristic of the three nor-) tallty atudies la that they do not agree with each other, bctia cancer of the liver, rectus, stomach, pancreas, hillary tract and lung are each found in one of the atudies hut in none of the others. . Although the power of the studies is liaiited (the moat powerful one, Brown et al (1981) has a probability of about 0.33 of detecting a threefold excess in liver cancer S Mortality), the lack of a latency relationship in that study and the complete absence of liver cancer in the other two studies strengthens this power in a way that cannot be quantified precisely. ' The average duration of follow-up in each of the three studies was 15.3 years (Brown et al), 15.7 years (Haroni et al) and 20.2 years (Zack et al). Further follow-19, specially of the first two studies, would provide the longer latency and greater auaber of deaths that would increase the power of the studies, nevertheless, she data SUlUble at he present, tine do not provide the coasie-) Boat,.' exposure rolatad results that would Justify jEoaclsdisg that KBs are carcinogenic in huaun beings. HONS 011538 16 KB EPIDEMIOLOGY REFERENCES Meigs, J.H., J.J. Albon, B.L. Kartin. 1956. "Chloracne from aa unusual exposure to Aracblor." JAMA 156:1417. Caffey, W.l. 19S2. "The Epidemiology of KBs." (Paper presented at the International Symposium on PCBs in the Great Lakes, Michigan State University, E. Lansing, HI.) ElOSH. Sept. 1977. "Criteria for a recommended standard t occupational exposure to polychlorinated biphenyls (PCBs)." USDHEW, HIOSH Pub. Ho. 77-225. Urabe, H., H. Xoda, and M. Aaahi. 1979. "Present State of Yusho Patients." Ann. H.T. Acad. Sci. 320:273. Kimbrough, H. (ed.) 1980. Kalogenated Biphenyls, Tcr-X phenyls, naphthalenes. Dibenzodioxins and Related Products. Chapter 9 Bl, Elsevier/North Holland Bioaedical Press, Aanterdaa. fisebbein. A., M.S. Wolff, K. Lilis, J. Thornton, and I.J. Belikoff. 1979. "Clinical findings anong KB exposed capacitor nanufacturing vorkers." Ann. H.Y. Acad. Sci. 320: 203. Blnlth, A.B., J. Schloener, L.K. Lowry, A.W. Smallwood, R.N. Ligo, S. Tanaka, W. Stringer, M. Jones, K. Herein, and C.J. Olucck. 1982. "Metabolic and health consequences of occur) Rational exposure to polychlorinated biphenyls (PCBs)." Accepted for publication, Brit. Jour. Ind. Med. Baker, E.L., P.J. Landrigan, C.L. Clueck, M.W. Zack, J.A. Liddle, V.W. Burse, V.J. Houseworth, and L.L. Needhan. I960. "Metabolic consequences of exposure to polychlorinated biphenyls (PCB) in sewage sludge." Aner. J. Epld. 112:553. Harooi, M., A. Coloobi, C. Arbosti, S. Cantoni, Y. Poa. 1981. "Occupational exposure to polychlorinated biphenyls II. Health effects." Brit. J. Ind. Med. 38:55. Chase, K.H., 0. Wong, D. Thomas, B.W. Barney, and R.K. Simon. 1982. "Clinical and Metabolic Abnormalities Assn-') dated with Occupational Exposure to Polychlorinated Biphenyls (KBs)." Jour. Occ. Med. 24:109-116. Kreisa, K., H.V. Zack, R.D. Kimbrough, L.L. Needham, A.L. Shrek, and B.T. Jones. 1981. "Association of Blood Pressure and Polychlorinated Biphenyl Levels." JAMA 245. 2505. HONS 011539 17 PCS EPIDEMIOLOGY Varahav, X., A. Fischbeln, J. Thornton, A. Hiller, and I.J. Selikoff. 1979. "Decrease in vital capacity in PCB-exposed workers in a capacitor aanufacturing facility." Ann. H.T. Acad. Sci. 320:277. Brown, D.P. and H. Jonea. 1981. "Mortality and Induatrial Hygiene Study of Vorkera Exposed to Polychlorinated Biphenyls." Arch. Envir. Health 36:120. Bertaaxi, P.A., C. Zocchettl, S. Guerdlena, M. Della Foglla, A. Pesatori, and L. Xiboldi. April, 1981. "Mortal-^ ity Study of Hale and Feaale Workers Exposed to PCBs." Int. tvnpoaiua on Prev. of Occup. Cancer. Helsinki. Zack, J.A. and D.C. Huach. "Mortality of PCB Workers at the Monsanto Plant in Sauget, IL." In preparation. Boll, Jt. 1981. "Relevance of epideaiology to policies for the prevention of cancer.1*: Jour. Occ. Med. 23:601-609. IARC. 1980. "Monographs 'on the evaluation of the carcino genic risk of cheaicals to tumors." Vol. 19:18. M0NS 011540 general CB--otl -. ': . ' These documents are eoeevhet unobjectively written with ) unaciantific dlscuaalons auch an, ....... . on page 3-23 tt author atataa i 'fewnr, tha onrtuloai . uaa of sathenatical nodala and the aaauaptloos nada by tba . CM without oonparing thaaa results to reality an anally dsnonstrated." -......... ^ Tha report by tba tow Inaland Gaa Xaaoclatlon btjUi Tha discovery of nail anounts of PCBs, tha uaa of which hw baan bannad by tha OX, In natural gaa trannlaalon linaa la nalthar unazpactad nor eauaa for public alarm. Lator in dlacuaalng acuta KB toxicity tha report atataa i "Short tana effacta of coderata doaea In anlaala arc minimal. Tha toxicity is comparable to that expected fron auch aubataneaa conaldared ralatlwaly aafa auch aa table aalt, alcohol or caffeine." In discussing chronic affacta the report statesi "...KBa produce minimal changes in now borns. In fact they are no core toxic to tha fetus than moat dienicals, and less toxic with regard to birth defects than sons qnmmcn vitamins and horsonaa." Sunnary of tha Health effects of lOi Tha "Sunnary of tha Health effects of KBs" tends to sisguide tha reader as to what tha actual articles present. Tba owerall tone of tha fuanary la a refutation of tha technical presentation or an exaggeration of tha "negative" arguaents presented in tha six review articles. ' On pp. 1, 2-1, and 2-2 of tha Bunnary, tha authors cite and dlaeuas Neleburgtfr's and Wllliam'a concept of pronotsra and initiators, and epigenetic and genotoxle eachan1ana. Bowevar, they do not point out that tha concept and tha claaslflcetlon of carcinogens fay Nblsburgar and Nlllala are not ahared by many. On pp. 1^ 2, and 1-5 tha authors discuss tha large number of negative carcinogenicity studies for KBs and praaaat tba inadequacies of tha positive atadlas. They do net nantlen the Inadequacies of tha nayafelvs stadias, for sxanpla, tha reference papers show that only a taw of the negative studies are of sufficient duration to adequately characterize than as negative atadlas. Moat of the stadias qaotad la tha ala articles ate of omb year or lads duration (e.g., sea article by Crump, p. 24). for tha positive studies, they Criticise that tha length of exposure was tdo Short tor an adequate study design tp.` 2*3). This la not a relevant concern for a positive stedy. '1 HONS 011541 The author* also stated that th positive studies should be discounted btcauss of inadequate design. if this 1* tha ease, tha negative atudias should also ba dlscountsd as wall. It-is ! far uors difficult to proa* a nagatlas rssult than a positla* ona. Oftan. observation* cltad In tha raalaw papers ara naglaetad or oulttad In tha CHA Summary. For example, p. 3, paragraph 2. santanca 3 of Summary: *Ths preponderance of studies has not Identified a clinical disease associated with axpoaur* to PCS. nor has it provided parsuasiva evidence of health impairment" couparad to Paper Mo. 1. p. 22, paragraph 1, last santanca: . *...* single study which suggests that diastolic blood pressure ay ba related to blood levels of PCBs.* This study suggests seam cardiovascular toxicity. In addition,' deraatologie reactions is known to be associated with PCBs poisoning. The Sunnary contains too a>ch emphasis on the possible carcinogenicity of PCBs and does not Mention such on the possible chronic health effects. The six articles have sore details of the chronic effects. On* of the controversial issues in this document is that the CM contractor feels that PCBs are tumor promoters and not initiators. Since promoters may have threshold wd reversible effects, the authors feel that low level exposures do not represent a haxard for cancer. The evidence for this assertion is presented on page 2-7 through 2-8 and again on pages 3-17 through 3-19. These arguments can be summarized as follows. 1) PCBs increase the frequency of liver tumors in rodents: a tumor type which occurs spontaneously in these animals. 2) PCB induced tumors are not metastatic. 3) PCB tumors are not transplantable. 4) Promoter^ are not mutagens. Many of the mutagenicity test on PCBs were negative. 5) Continued presence of PCBs is required to sustain liver neoplasia and removal may cause tumor regression. There are several problems with this line of reasoning. First the mechanisms of action of initiators and proaiotsrs are becoming less dear. The dogma that initiators cause amtatlons and promoters have 'epigenetic* effects in cancer expression la no longer totally accepted. Cairns has recently proposed that initiators may not act through the induction of mutations by way of ORA damage as is usually thought. Further, the term epigenetic is not very useful: for example, what if a chemical caused a heritable change in chromatin structure by altering the action of DMA methylasss. This offsets the DMA and therefor* is 2 MOMS 011542 "genetic,* and yet neither the sequence ot buM nor the nanber of ehNMWM Because of neh difficulties it to better, perhaps, not to trtotyp chesleals. : a ^ Itm U on accept tbo contractors nuonlnf, It la laportant to onto that not all nutageniclty taata on KBa u* negative, fit tha Nonaanto dominant entitled _ *4 review and evaluation of carcinogenicity in nftoa and rata and nutagenlalty studIn with KB" ooooral poeitive nutageniclty taata van oltad aloof with -tha negative caaoltat Aroelor 1221 and 4-chlosoblphenyl warn positive in tha Anas taat. TCB and tha 3, 4 epoxide of TCB ' lnduead single atrand breaks In OKA of U-929 ealla. Kanaehlor 3M waa found to indue* ehronoaonal abarratlona _ln^ vitro ohil* Xanechlor 500 waa poeitiv* in also. Contlnuoua aspoaura of C3H Iff V4 ealla to Aroelor 1254 cauaad call tranafornatlen to typo XII fool. Thoao ealla foraad a aareona whan laplantad into irradiated nioa. Aroelor 1240 and 2,4,5,2',4a,5* - haracblorobipbanyl also cauaad aorpholofioal transforaation of C3B 10* Vi ealla. 4-Chlorobiph*nyl caused an iaeroaaa in unscheduled DBA synthesis in OK) ealla idiile Aroelor 1254 Induced ehronoeena dosage in ring downs fad a 10 ppn diet. Clearly, ' based on these data on* can argue that KBa nay hare Initiator aetlelty. - - Finally, in calculating the risk eatlaate for Ineraaoad hasan cancer eases eausad by exposure to KBa, the contractor nade (but did not stats) an anotaoua aaaunptlon. It was aasuned that because hapatonao wars lnduead in rodents, this waa the only tunor type for which banana war* at risk. Although tha . correlation between carcinogenicity in anlaala and nan is good, the issue of alts concordance is not clear, therefore, hunana nay be at risk for other types of tunoru as wall. Bines the incidence of hapatonao in tha ban population la low, tha risk sal;last* prodaeas a eery low nuabar of PCfe lnduead cancer. However, if these chonlcals ware also aoaoeiatad with lung eareinoaa, for sample, risk estlaatea would be such higher. rr*'* n- ; the latter payt of the Bean ary consists costly of a eoaplst* refutation of tha risk aasasanent dona by Cna^ and Hastarnan, tha last docunegt on . the "Aaoasaaent of Carcinogenic risks froa KSa la Food," with very little doeuaentatlen of tha reasons for not accepting Sr. Creep's aasasanent. In his aasasanent, certain nodels ace aped go- derive virtually safe doaea of KBa at apaelflad. risks. Be lacludoo tha probit nodal, the one-hit aodel, and gba nelti-otaga apdel...Be also usee tha one-hit and aultl-hlt aodelsj.to-derive ogtlnates of life tine estra risks to bunms for certain sebpopulotlonsat risk. Although the "tensity" is interesting reading, in east oases it is net of sufficient gclastifis validity to warrant a response. , in any ease, the Biiuaigy conclusions for risk assessnaat -ace ask supported by thg technical;review docunenta. . 7Za faet, enl(ka the other aeetlon such as epidaalology, tha risk aasasanent section 3 HONS 011543 dots not even attempt to sunaarise the Material In the risk assessaent technical document. Than ars several specific problems with tha Sumary. % ' 1) savaral places In tha Suaaary (pages 1-5 and 1-3) tha authors eoneluda that thara Is llttla reproductive risk from pels In animals. Tha Cruap and Nastaraan docuaent contains tha data iron tha raproductiva studlaa inaonkays which showad effects at lavals as low as 2.3 ppa. Tha KB data is unusual in showing raproductiva affacta at lavals lovar than thosa at which carcinogenic affacta aca saan. Tha fact that this la not avan mentioned in tha suaoaary docuaent, avan though ona of their own technical review docuaenta Include tha data, caats aoan suspicion in tha suaatasy docuaant. 2) Aside fro* tha biological arguments regarding aechanisa ef cancer which are used to refute Cruap's risk assasaaant, thara is another major arguaent proposed. It la that if tha risks are as large as Cruap's astlaatas indlcata, these risks would have bean detected by now in high risk populations. Tha review of tha apldaalology studias shew tha highest nuaber of cohort deaths studied for sufficient latency periods is less than 100. it la doubtful that relative risks of lass than 15-20 tlaas would be detected with this sire population (wa are- presently doing tha calculation). 3) Using the CPA Water criteria docuaenta, the authors in tha suaaary section calculated tha expected Increase over background cancer for certain subpopulations. There is a alstaka soamwhara In the conversion figure given In Table 3-5 equating 73 ng/1 to SO ug/dsy. The SO og/day doss not produce the expected increases over background shown in Table 3-S. A conversion to SO ng/day rather than SO ug/day would produce the eatinatss shewn in Table 3-S and on pages 4-10 the figure is given as SO ng/day. - 4) On pages 4-S, tha authors use the CAS docuaenta which show that If POBsjure not carcinogenic, tha allowable daily Intake would be 2X0 ug of KBs per day. This is calculated using raproductiva data in sinks (which wan similar to raproductiva data in nonkeys) and a safety factor of 100. This allowable dally Intake of 210 ug/day is not disputed by tha authors, yet the exposures shown in the table on page 4-11 for eapaeltor workers la 4 tines higher than this allowable dally Intake. Therefore, even without considering carcinogenicity, the allowable dally intake of 210 ug/day, which is used by the authors to question the such lower safe levels fro* cancar data, is already 4 tines lower than average exposure of eapaeltor workers. All the arquaents usad in tha Suaaary for negating certain aodels used for extrapolation of cancer risks by aeehanistie argument! would have no ralavance to reproductive data. HONS 011544 5) The aatbors of the aweary document hKm no attempt to Htiutt any tilts tecMlsa to their propoasd soM of cancer. This nay bo booonoo the prosit aodelf also used by Ctioy, ;1tm tba lN*t risks of almost any soM available at la* lavsla. ' Tbs oofs lmls shown an page 34 of Crump's report using this ` 1 bslo* tht exposure of average capacitor mothers, even for risks as high as - 1: . * tiia.'ii-w . V.;: tin Epidemiology of KBs Tbs Epidemiology of Klo by Saffsy was reviewed In greet detail. Our aosnsnts sra directed to answer tha question of whsthar his eonelnolons and ganarallsatlons ara justIflad basad on our cursory review of tha eltad studlas. We ehooa to bogln with tha review of tha oarclnoganlelty aactlon slnoa this la of neat ooncarn. - .. .. .. Carcinogenicityi In general, Oaffay's conclusions ooncarnlog tha strength of epidemiologic awldanca nacassary to support a causal association batwoan KB exposure and oarclnoganlelty at reasonable in light of tha erltaria established in Richard Boll's Taper. Howaaar, several of tha studies reviewed by Oaffay artassl at al, tack and Naschl suggest an overall lncreasa In risk fron death due to cancer and all studies reviewed by Oaffay do In fact reveal excesses for specific sitesj although aoet of these excesses ware not statistically significant or site concordant with one another, tt should be stated at the outset, that tha presence of consistency strongly supports causality, but Its absence does not rule it out. Oaffay deliberately onltted the prellnlnary uortallty findings of the Yusbo incident fron this section because this study was not an apldaalologle analysis dlraetad at cancar. Drabs at si reportad U deaths fron nallgnant neoplaans aneng 35 confined deaths of Yttsho victlns. ' The authors state that *thls rats (11/35) la substantially higher than the 31.lt that Is the Mortality rata ftpa neoplasaa In the sane prefecture* for that yer (1BT7) bat tbs'authors also state that *it would be prenature to eonoludo that this high nortallty bears association with fCi poisoning.* nevertheless, three of those deaths wart due to liver noncer sad tow (ns long cancer, which suggests earn ,with reports fron previous studios (lack and Nusch, ** r- Beveral of fadstudies reviewed by Oaffay (Balba at al. Back and Hueehr ara basad On so tow death*. especially workers with tsar s* non year* 'of exposure or IS ynm sines tint erpooure, tost tt would'Be prasistore to conclude that KBa ara dsflsltsly nan aarclnogeSid. Bar' fastened, fa' tha ana .study (Brown and Mnsl visr tbs greatest wunbor Of person yews observed," there KV-H1 deaths w*d it la not dear' how many of then deaths wore among workers with five or mors years of exposure or 15 years latency since first exposure to fCBe. & 8 ONS 011545 Furthermore, the power of detecting a significant elevated risk baa not been raportad for any of thasa stud las and thasa calculations should ba performed aa part of our ln-dapth review. In addition. It should ba notad that aost of tha studlas reviewed by Gaffay ara mortality studlas, and not Inddanca studlas, and ara tharafora baaad on daath cartlflcata Information which la known to have cartaln limitations for detecting eancsr risks duo to chaaieal axposuro. Tha principal limitation being that analysa* based on deaths alone may axeluda those workers who left employment {con a study plant and later developed cancer, but have not expired and are therefore not counted among tha cancer deaths, ' With tha exception of tha Tusho study (Drabs at al), no reliable data on past industrial exposure were reported. Two studies (Brown and Jonesi Bertaxxi at al)- reported relatively low airborne levels of VCBs as estimates of current exposure. Only Bahn at al reported that their study population ware 'believed to have been heavily exposed* to KBs. In the absence of high body burdan levels or clinical indications of chloracna, we would have to assume that occupational exposure to KBs was relatively low, and,therefore, would not be expected to produce a measurable differential effect on sociality. In summary, based on tha preliminary results reported in the literature, Gaffey's conclusions with respect to cancer risk seen reasonable, but he unjustifiably diamiases the significance of suggestive evidence. Bertaxxi at al, for Instance, recognising the limitations of their analysis based on a few nunber of deaths (27), stats that *the results so tar obtained are strongly suggestive of an excess of mortality from cancer among sale workers and an excess of mortality from all causes and from cancer among female workers.* Several of the studies reviewed by Gaffay (Brown and Jonas, Bertaxxi et al. Sack and Huech) determined the vital status of their cohorts as of 1977 and 197B. It is possible that formal requests to these Investigators to follow-up thefr investigations through 19B1 may resolve ambiguities that are due to insufficient numbers of deaths. Finally, Gaffay reports on the private communication of preliminary results from two investigators (houach and Sinclair) whose results should be secured as soon as possible so that we can evaluate their relevance to Gaffey's conclusions. It is our opinion that further rigorous evaluation by more qualified, experts is Inherently desirable and will add to our understanding of the significance and deficiency of these studies, but we doubt that there is sufficient epidemiological evidence at this time to support a causal association between KB exposure and cancer. Additional studlas among more highly expossd workers with sufficient numbers of deaths should be supported In ordar to resolve reasonable concerns about the potential carcinogenicity Of FCBa. e HONS 011546 *ccldntl Bear? Brpoaure i Gaffey reviews too Incidents In this section (Hales st si. Tusbo reports), bat Helps Is aore appropriately nritM In the next section.. _ . ... ^ " Its mmcws Studies <d>lch"\a*e been published In conjunction with the Itaske Incident represent a slsesble fraction of the total Health effect 11taxatare concerning Kha. Gaffey, and the aethors of the General Uectrle health effect report, attenpt to nlnlnlse the relevanco'of these findings tor two naln reaeonsi Cl) oJmpleo of the Kaaaenl Moo Oil were found to be contaminated with dloenaofprans. and (2) victims of this Incident Ingested large doses of KBs thick are Irrelevant to assessing the health effects fron lower occupational exposures found In the United States. Proper evaluation and Interpretation of these objections are crucial to ths assassaent of health offsets related to KB exposure. . - .. - Yusho disease is a Japanese tarn tdilch refers specifically to acuta ad sub-acuta poisoning tee to ingestion of Kanenl Bice Oil which was contaalnated with KBs. There aeene to be s general consensus that ths average amount of Xanechlor Ingested was satlnstsd to be about 2 91 while the alninua scant Ingested by s patient was about O.S ga. M of 1277, there were l.SCS victlns of TUsho, who exhibited s variety Of scuts synptoas ranging fron severe chloracne d byperplgnentation to less specific offsets such as vonitlng and diarrhea. . Ths noat notable finding fron this Incident seen to bs that patients seen three and six years after first coning down with yusho dlsaaaa still exhibited eynptone of TUsho, blocbenical abnormalities and persistent blood levels of KBs. Unless these effects can all reasonably ha attributable to ths dlbensoferan contamination, than I would say that ths severity, incidence and persistence of these effects all waarrant serious concern for KB exposure. Both the specific toxic effects (as determined from animal and human .studies) of PCDPs aid KBs must be compared, aa well as ths quantitative, evidence supporting ths extent of KDP contamination. ^ . ....... . With respect to the preliminary mortality reports, Geffey justifiably calls attention ho ths problem that ths popnlatlonat-riak, or denominator of the relative risk rates, cannot bo determined, and therefore Presents a problem for determining ths aotsal `attack rate* os well s the relevance oe elevated cancer deaths' among Yusho vietlas. This problem Is wltlgatad somewhat by the finding, that among p group of 14S known oners of oil, SO 'aoamad Isas that 72d nl, and St percent of thoao usere were *mffectod. Siopg thooe. spoused sore than 720 nl, 100 percent "were affected, t)il.<na)f suggest that Tusbo victims, os defined bp variooaoitnlcalcrlterih, dp in feet represent the total aoMfof people expoeed^tp.pontanlnated oil. This, of ooerse, 10;' iftn]eeiure and Ik subject to much criticism, hut until this problem la resolved, epideniologle analyeeo otllyxing the cohort approach ngy be..inappropriate and unacceptable. . . 7 HONS 011547  . '`ffay only briefly reviews the reproductive effects seen ***'1 children of Tusho victims end he neglect* entirely to on sore recent findings. The ETA Meter Qoellty Criteria reports that 'thirteen wen, 11 with Tueho end 3 .ij i ,t' but serried to sen with YUsho, delivered 10 live and 2 fillUjtn Infanta beween February IS and December 31, 19(0. Mine of the 10 had hypsrplgmentation, and S had graylah-dark Pigmentation of the gingiva and nails. Other findings of reduced Otowtti, prenaturely erupted teeth and lncreaeed eye discharge ' *tro also reported. Mine nore Infants with hyperplgnentation Tee? >ntn f Yusho aethers between 1969 and 1972 (Toshlaura The Water Quality Criteria document reports on several tudlus which found elevated levels of KBs in breast silk of `"n women. Again, the significance of these findings depends our Interpretation of the relevance of KDP contamination, f* ^11 as on the potential for lactatlng aethers with high body "Ulden levels to deliver PCBs to nursed infants. Environmental Levels aid Body Burdens: Oaf fey's conclusions *0u* Lessonable and justified. Be concludes that high exposers to I'Cha results In higher body burdens, all other things being There were a few Inconsistencies which can be adequately xpl.lined by Mrs rigorous Interpretation of the studies which *111 have to be addressed In a wore in--depth review. Spud rically, what Is the significance of high burden levels? Is thunj * threshold level for adverse health effects? What Is the i'l'il ricance of higher chlorinated compounds and what Is tha Potential for future bloaccumuletlon under current exposure Situations? Epidemiologic Studies of PCBs and Health i Dermatologic OLL`**tL.a Gaf fey justifiably concludes that *the data auggsat ti 'ii>|iy that when KB blood levels exceed shout IS0-200 ppb, cni'ii acne can occur,* but most studies have shown that the oci:i wnee of chloracna Is not further associated with blood KB uv"in.* Caffey suggests that this apparent inconsistency say ba * function of personal Idiosyncratic factors or that not all c.mi.u 0f high blood KB levels are associated with chloracna Pun.*,*, they do not have sufficient akin contact for chloracna to nuvt.i'jp. m any event:, dermatologic reactions have been cum latently associated with KB exposure and these reactions can nuciti under occupational exposure conditions as low as 0.1 mg/e3. i.lver funotiont It la difficult to oaks any Interpretation ol 1 livae findings without reviewing each of the nine studios. In '1"ii.,i ai, there does not seem to be consistent findings of uiiui.iiaai liver function or biochemical Indicators of ensyme *":l 'vity? nevertheless, several studies did report abnormalities 1,1 "nxyma levels which may have relevance beyond that suggested `V '.uffey. Kt metabolismi Gsffty's conclusions as atatad are latent with tne reposted findings. HONS 011548 8 "host studies, Including ons non-oceupatlonal study hare . associated Increased trlpyeerldes with KB exposure, the dots on J cholesterol see not consistent* an Increase, a decream ad no chance were found (one study each). BBC, cholesterol either decreased oc was unchanged (one study each). Bren If >9 exposure has sons effect on fat estabollsn. It appears ho be without any apparent clinical significance.* - fyeotoes. Illness and Other Conditions, All of the studies reviewed in this section should be aore thoroughly evaluated. Gaffey quotes hitk ot al. as saying 'that no studies to date hare shown that occupational exposure to KBs is associated with any adverse health outcone, to be distinguished frea deuonstrahle eubcllnlcal blochanical alterations." It is possible that these blocheaical ahnornalltles have not yet revealed any clinical sanltestations, and it is not clear whether-the eubcllnlcal alterations present serious health effect concerns. In general, nost of these studies are cross-sectional, and therefore oaks no effort to ascertain t^e true population-atrlak. In other words, the investigators did not identify their study populations frea historical e^loynent records - thereby ensuring that all workers bad sons nininal duration of exposure and furthernore, these studies aade no effort to identify workers who nay have left euployaaent for occupational health reasons. In aost cases, tbs study population represented thoee workers currently euployed and healthy. I .... -. _ 1. r HONS 0115**