Document QJ9xLJoo5XBkDGkg6ZK9X6Y7v

230 Annals New York Academy of Sciences [f 7\') Information solicited from relatives and fellow workers concerning the individu al's drinking history indicate only very modest alcohol consumption and suggest that this case, too, may be causally related to the individual's exposure to vinyl chloride. Among the other cancers resulting in death were a glioblastoma, a lymphosarcoma, and a reticulum cell sarcoma. Our observation of a brain tumor and two lymphomas conforms to observations in other vinyl chloride-exposed populations* and suggests the possibility that their origin may also be related to vinyl chloride exposures. In addition to the above, other malignancies were the underlying causes of death in this group of 24 workers. The details of the work histories of the 3 workers who died of hemangiosarcoma and the other possible occupationally related cancers appear in Table 6. All hut two worked extensively in the polymerization facility. All of the hcmangiosarcomas occurred in individuals who first worked prior to 1951. As the time from onset of exposure increases in cohort members with similar exposures, additional cases of occupational cancer may occur in this group of workers. Summary These data are derived from early follow-up of individuals exposed for 3 or more sears to vinyl chloride in e polymerization facility At least If percent of the deaths that occurred were causally related to exposure to vinyl chloride. Longer periods of observation are required to preside information concerning the full spectrum ol vinyl chloride-induced malignancies and their incidence among exposed workers. These data speak for the need to present exposure to vinyl chloride and for surveillance and early disease detection of those who have experienced vinyl chloride exposures in the past REFERENCES 1. Maltoni, C. A G. Lffemine. 1974. Carcinogenicity bioassays of vinyl chloride: current results. This monograph. 2. Lillis, R., H. Andi rson, W. J. Nicholson, S. Daum, A. S. Fischbein A I. I. Selikoff. 1974. Prevalence of disease among vinyl chloride and polyvinyl chloride workers. This monograph. 3. Veltman. G., C.-E. Lance, S. JOiie, G. Stein A U. Bachner. 1974. Clinical mani festations and course of vinyl chloride disease. This monograph. 4. Suctu, I., L. Prodan, E. Ilea, A. Paduraru A L. Pascu. 1974. Clinical manifesta tions in vinyl chloride poisoning. This monograph. J. Creech, J. L., Jr. A L. Mark. 1974. Liver disease among polyvinyl chloride pro duction workers. This monograph. 6. Heath, C. W,, H. Falk A J. L. Creech, Jr. 1974. Characteristics of cases of angio sarcoma of the liver among vinyl chloride workers in the United Slates. This monograph. 7. Thomas, LB.AH. Popper. 1974, Pathology of angiosarcoma of the liver among vinyl chloride-polyvinyl chloride workers. This monograph. 8. Tarershaw 1. R. A W. Gapfev. 1974. Morality study of workers In the manu facture of vinyl chloride and polyvinyl chloride. J. Occupational Med. In press. 725025 CHARACTERISTICS OF CASES OF ANGIOSARCOMA OF THE LIVER AMONG VINYL CHLORIDE WORKERS IN THE UNITED STATES Clark W. Heath, Jr. and Henry Falk Cancer and Blrtfi Defects Division Bureau of Epidemiology Center for Disease Control Atlanta, Georgia 30333 John L. Creech, Jr. B. F. Goodrich Company Louisville, Kentucky 4040} The relationship between vinyl chloride exposure and human angiosarcoma of the liver (ASL) first received serious attention in December 1973 when a case of this rare tumor was diagnosed at autopsy in a vinyl chloride (VCJ worker from the B. F. Goodrich chemical plant in Louisville, Kentucky. At that time, this case represented the third recognized among VC workers at this plant; prior cases had been diagnosed In May 1970 and March 1973.' Although at first the possibility was considered that case occurrence might be unique to the Louisville facility, the subsequent detection of cases at three other plants has shown the problem to be industry-wide. Ten additional cases have subsequently been identi fied among other VC polymerization workers in the United States (Table 1). Workers at a total of four different plants have been involved. The earliest case recorded to date occurred in 1961, the others being scattered over subsequent years with 5 in the past 2 years (1973-1974). In several in stances, the diagnosis of ASL was not specifically made until review of patho logic material during the early months of 1974. Death certificate diagnoses in such cases indicated "primary liver cancer," and in one instance merely "liver cirrhosis." The 3 most recent cases were diagnosed in the course of systematic medical screening, 2 at the Louisville plant in February 1974 and 1 at the Soulh Charleston plant in May 1974. These 3 patients are the only ones of. the 13 who are alive currently. All 13 cases have occurred in white males. About half have been in their 40s with an age range from 36-60 years and a mean age at time of diagnosis of 48.2 years. Length of time between first VC work and diagnosis has ranged from 1229 years (mean 20.3 years), mean total duration of VC work being 18.0 years. Comparison of Observed and Expected Inc/dence The reason on eliologic association between VC work and tumor could be inferred so readily on the basis initially of only 3 cases is (hat ASL is extraor dinarily rare. Based on data from the National Cancer Institute's Third Na tional Cancer Survey (1969-1971), expected annual incidence of this tumor is on the order of 0.0014 case/100,000, or about 25-30 cases a year in the entire United Stales population. Estimating the total VC worker population past and present in the United States at roughly 20,000 persons (an estimate which may hi* MnM nni" nn rn,rJx, -iL,a m . t .1 * -1,1 Annals New York Academy of Sciences Htf-ljj ?3R? $ <f X *e wn )\ 35Rp?Sn^8P5SS -=*!I ac?55^tf3tf ^>*N--O^Otote^ I;> ssssssszsssss w 3 KSSSSSSSSSSKR . . . d rt >>->->> ZZZ5* XXXXXXX*? n* w e e* **1^^*^ = = = oo0^ CA tT u u *Tj> _o ^Lu^- ll. jj J3 c --E rt *d l." s; O 10 .>3,*3>23>.2.3>2,3>2 3,>J23.>2f*5l*513SW,2^ 2fw\22- J333333Z52rf a 'uu'UC'OC U- OC UU_U *'n0*p0 OOo *OO0 TOO> "OOO*O0O*O0O*O0OjjMbb^tUrtlUO. uc OU..UO-UO..U.U....UO....UO..-.U.O.-"OSo".Oo^T?39i.qcr nc~ w2vi Heath et af.: Angiosarcoma of Liver in VC Workers 233 expected in such a population over a decade. Since 13 cases have thus far been identified (and it is not unreasonable to think that more cases may be found among other VC workers), one arrives at a risk ratio (a ratio of observed to expected cases) of at least 400:1. Needless to say, had the situation concerned a common tumor such as lung cancer, the picture would have been much less clear-cut and would have re quired many more cases and much more elaborate epidemiologic manipulation to associate tumor with industrial VC exposure. Clinical Features Medical and epidemiologic data have been examined In detail for cases at the Louisville plant in particular. The remainder of this presentation summarizes these and other data concerning VC workers at that plant. For each Louisville case, alt available medical records have been reviewed. Interviews have been held with each surviving patient and with dose relatives of all deceased patients to ob tain specific information concerning work history, toxicologic exposures, and family history. Pathologic material, particularly sections of liver and of tumor, have been reviewed both by local pathologists in Louisville and by Dr. Louis Thomas at the National Cancer Institute and Dr. Hans Popper at The Fogarty International Center. In most cases the diagnosis of ASL was renched only with difficulty; 3 of the 3 fatal cases were not diagnosed until autopsy. In only 1 of the 2 remaining fatal cases was ASL diagnosed unequivocally prior to death (4 months). In the other case, a diagnosis of "primary liver cancer" was made 5 months before death, but ASL was not clearly identified until recent review of autopsy material. Presenting signs and symptoms varied widely and were relatively nonspecific. Abdominal mass was a presenting feature in 4 cases, gastrointestinal bleeding in 2 (diagnosed initially in each instance as peptic ulcer), fatigue and weakness in 4, and noticeable weight loss in 1. In I case no signs or symptoms referable to cancer or liver disease were present and the diagnosis was made in the course of workup for another health problem. Duration of symptoms prior to diagnosis ranged from 1-t 1 months for 5 of the 6 cases displaying symptoms. In the 6th case, symptoms related to liver dis ease antedated diagnosis of tumor by 8-10 years. In that particular case, a diag nosis of liver fibrosis was made in May 1963 following two episodes of gastro intestinal bleeding in 2 years. A portocaval shunt operation was performed because of varices, splenomegaly, and increased portal vein pressure. Six years later (May 1971) the patient was reevaluated because of progressive symptoms, and at that time, liver scans suggested the presence of tumor. ASL, however, was not unequivocally diagnosed until autopsy in March 1973. At time of tumor diagnosis, physical findings among the 7 cases varied con siderably. In 2 cases, no abnormal findings at all were recorded. Three cases presented with liver enlargement, in each instance accompanied by splenomegaly. In a 4th case, splenomegaly alone was detected. One additional case presented with severe right upper quadrant tenderness in the absence of clcar-cut organomegaly. Laparotomy revealed free blood in the peritoneal cavity traced to what was tentatively diagnosed prior to death as "hemangioma of the dome of the liver." Abnormalities of one or more liver function tests were present in every case at or around the lime of diagnosis. In most instances, however, these abnormal!- 234 . Annals New York Academy of Sciences were at feast mildly increased in all 7 cases, total bilirubin in 6, and alkaline phosphatase in S. LDH was increased in 4 of 6 cases tested and BSP in I of 2. Diagnoses were made with the help of fiver-spleen scanning in 6 cases and hepatic angiogram in 2. Esophageal varices were demonstrated in 2 cases, one by x-ray, the other by endoscopy. Efforts at lissuc diagnosis were made by open liver biopsy in all 7 cases and by needle biopsy in 3; autopsies were performed in 4 of the 3 deceased cases. Open liver biopsy provided a diagnosis of tumor in 4 cases while in the remaining 3 cases, diagnosis was made at autopsy. Recent review of pathologic material in all 7 cases has shown evidence of portal fibrosis as well as of tumor. In addition, discrete subcapsnlar ffbrofic areas were described at time of open liver biopsy in one of the 2 cases diagnosed in 1974. In 2 cases, angiosarcoma was present in tissues other than liver. In one the picture was that ol local tumor spread. Emocmkjioc.ic Features The B. F. Goodrich Louisville plant emplnjs about 1200 person* of whom lhout 270 or 20 percent art engaged in VC pot>mentation activities. The plant first began potyvinjl chlortJe tl'VCt produtupn m 1942 after pilot operations dating from 1937 The present potymeri/jiion work force sue was reached in the late 1950'* All 7 patient* hid urtked in yob* duccily related to VC poly* mcriration, and in tbit capacity it can be assumed that all seven have had direct exposure fo vinyl chloride mon-unr! tVCMt initial involvement in VC wotk occurred from 12-24 years prior to tumor diagnosis (mean 20 0 years) (Tame I), Involvement wa* essent.atty continuous during these years tmean duration of VC work 17 0 years) Dates of initial industrial VC eaposure tanged from 19441943 soon alter the plant began production (2 cases), to at recently as 1937 and 1961. Ages at initial etposurc ranged from 21-lJ years (mean age 27.6 years). Although the data are far from conclusive, they do suggest o possible dose- response relationship when one tries to compare intensity of VC exposure with latent period. Virtually everyone entering VC polymerisation work at the Louis ville plant starts os a ``helper," a job which includes cleaning the interiors of polymerisation vats and hence involves maximum exposure to VCM. When duration of work os a helper fo rough measure of VCM exposure) is correlated with interval between first exposure and diagnosis, an inverse relationship is sug gested. For 3 coses who averaged only 5.3 months as helpers, exposure-to-diag- nosis intervals averaged 22.0 years; for the 4 remaining cases who averaged 36.5 months as helpers, this interval was 14.5 years. No evidence was found (hat exposure to other potentially hepatoloxic mate rials outside the work place played any consistent role in these particular cases. Only 2 patients had a history of increased alcohol intake, and in one the increase was borderline. One man gave a history of exposure to arsenic insecticides for several years as a teenager. In none was there any suggestion of thorium dioxide exposure. Acroosteolysis, past or present, was not observed in any case. Nonmauonant Liver Disease 725027 Because 2 of the 3 tumor cses which were initially recognized in Louisville had shown evidence of heoatic fibrosis antedating or coexisting with tumor, medi cal records at the Louisville plant were reviewed for cases of nonmalignant liver Heath et alt Angiosarcoma of Liver in VC Worker* 235 Table 2 Cases or Nohmauohant Liver Disease amono Louisville VC Workers A|t it DUjnMtt, Rite, Set 46 WM 28 WM 36 WM 16 WM Monih nd VMf 1 DU*U NV el Vein between rim vc Work end Dliffiotlt 12-68 1-72 9-73 9-73 24 6 24 29 Clinical Ftilurn En Ei* tire ed Uiffd Abnormal liver function tret* ipletn liver + ++ + -- AP, BSP, SGOT TB. SGOT TB, A1* TB, LDH, SGOT fttKelefle fciuree PorUt fibrotU Subnputu fitrnl, ++ + -f + 4* * AP, alkaline phosphatase; BSP, bromsulphatein clearance; SGOT, serum glu tamic oxaloacetic transaminase; TB, total bilirubin; LDH, tactic dehydrogenase. disease. A total of 4 such cases have been identified to date, again all among VC workers. The clinical and epidemiologic features of these 4 cases closely resemble those of the 7 tumor cases (Taute 2). All were white males, average age 46.5 yean at time of tint diagnosis. One was diagnosed in 1968, l in 1972, and 2 in I97J. VC work began on the average 20.8 years prior to diagnosis and in every Instance wai continuous. One patient, however, a 28-year-old min, had worked t the plant for only 6 years. Duration of illness prior to diagnosis ranged from 0-7 months. Although 3 of the 4 patients presented with signiffcant splenomegaly (one with hepatomegaly in addition), in 1 case physical examination was entirety negative. A variety of liver function tests showed abnormalities although none was severe and no consistent patterns were evident. In 2 cases esophageal varices were demonstrated, and in both instances, splenorenal shunt operations were performed. All 1 patients with splenomegaly underwent splenectomy. Open liver biopsies were performed in all 4 cases. In each instance, a picture of portal fibrosis was seen, and fn 3 cases it was accompanied by discrete nodutes of subcapsular fibrosis. The latter feature was indistinguishable pathologically from the subcapsular lesions observed in 1 of the 7 tumor cases. Discussion In light of recent animal experiments' and of clinical observations reported in European VC workers,* it seems likely that exposure to VCM in the course of VC polymerization work is responsible for malignant and nonmalignant liver disease in these cases. Epidemiologic features further suggest that either pro longed exposure and/or a long interval from initial exposure is required before liver disease becomes apparent. Although the length of lime involved seems quite variable (6-29 years), in all but I case the interval was 12 years or longer. We have no assurance, however, that the actual latency required for develop ment of liver disease (as opposed to latency in relation to clinical recognition) may not be considerably shorter. Unfortunately, if one can judge by the clinical features of the Louisville cases, manifestations of liver disease are quite variable, and in most instances were not delected unlit the process of hepatic fibrosis was so far advanced that splenomegaly and increased portal pressure were evident. Furthermore, if one envisions angiosarcoma as a late manifestation of the same pathologic process, then it seems even more likely that clinical detection of liver 236 Annals New York Academy ot Sciences disease lags considerably behind actual disease onsel. At prcscnl, therefore, a major concern is to develop mote sensitive means for detecting liver disease, not in its late malignant phase (if in fact a single pathologic spectrum is involved), but in the early stages when fibrotic reaction first sets in. Eventually, nonmalignanl liver disease, not ASL, may prove to be the most prominent disease problem associated with VCM exposure in terms of numbers of persons affected. One must also consider the possibility, however, that VCM may produce angiosarcoma at sites other than liver or that tumors of other kinds may also be induced. At present, animal experiments suggest that a wide spec trum of tumors may be involved,' although as yet there is no clear human evi dence to support the concept. In addition, no human eases of liver disease, whether fibrosis or frank angiosarcoma, have yet been strongly linked to VC exposure outside the industrial setting. Whereas this may mean that risk of dis ease is greatly reduced or even nonexistent at very low levels of VC exposure, i! is obviously too early to draw any conclusions In view of past experience with other oncogenic chemicals, it will be prudent, until many more data have been assembled, to view any exposure to VCM as potentially harmful. Summary A total of I3 eases of ASL have been documented to date among VC workers in four different plants in the United States. In this particular industrial popula tion. this number of eases represents at least a 400-fold increase over expected incidence for this eslremcly rare tumor, The first case occurred in 1961. Average age at diagnosis is 48.2 yean. Ascrage length of lime between initial VC work and diagnosis has been 20 3 years. A detailed review of 7 eases associated with one plant revealed that clinical features varied from little or no overt illness prior to diagnosis to advanced liver disease with portal hypertension and marked splenomegaly. Portal fibrosis was present in all 7 of these ASL eases as well as in 4 additional cases with non* malignant liver disease among VC workers at the same plant. These findings suggest that exposure to VCM in industrial sellings can produce hepatic fibrosis with angiosarcoma ns a late manifestation. Conventional tiver function tests may not be sensitive indicators of such liver impairment, at least in its early stages. References 1. CXEECtt, J. L., J*. & M. N. Johnson. 1974. Angiosarcoma of liver in the manufac ture of polyvinyl chloride. J. Occupational Med. 16: 150--151. 2. Maltoni, C. Unpublished observations. 3. Marsteller, H. 1.. W. K. Lelbacx. R. MOller. S. JOhe, C. E. Lanoe, H. O. Rohner & G. Veltwan. 1973. Chronisch-loxische Leberschkden bei Arbeitern In der PVC Produktion. Deut. Med. Woehschr. 98: 2311-2314. FROM MOUSE TO MAN--OR HOW TO GET FROM THE LABORATORY TO PARK AVENUE AND 59TH STREET* Marvin A. Schnelderman Associate Director [or Field Studies and Statistics, DCCP National Cancer Institute Bethesda, Maryland 20014 Nathan Mantel Biostatistics Center 7979 Old Georgetown Road Bethesda, Maryland 20014 Charles C. Brown Biometry Branch, FS&S National Cancer Institute Bethesda, Maryland 20014 Introduction If the title of this paper sounds like a guide to a road map, that is what intended to be, with information about detours, chuckholes, swamps, quagir and dead ends. In fact, our description of the problems may lead some peop conclude that you really cannot gel there from here. We think you can, how --perhaps not exactly there, but certainly Into the neighborhood. When we extrapolate moderate or high dose animal toxicity data to presu effects in man, we must do it like the animal trainer entering the lion's cage-- cautiously. The first caution fn this careful operation is to extrapolate from m to mouse from the higher dose levels where it is possibfe to see some results small or moderate-sized experiments to the very low doses, where even very 1 experiments are not likely to show much. We really do want to extrapolate an do not want to do these very large experiments because they are so full of j lems that they probably cannot be done well. We will discuss this later. Weinl in his paper on Trans-Science,' refen to this sort of difficulty that the meg periment poses as his fint kind of trans-science problem. Models for in-Mouse Extrapolation The usual way to extrapolate to responses at low doses Is to assert the t ence of some mathematical model of :dose-response and then compute th sponse at the desired dose or compute the dose for the desired response. V berg suggests this as a possible way out of the mega dilemma. Presently two classes of mathematical models are advocated for In-m extrapolation. Fint, there are the models that deal with yes-no data, the diet mous data models. These are concerned with whelher or not an effect occu The Working Group met at the Detmonlco Hotel, Park Avenue and J9lh '