Document p3Y5EMdGO4dbKm637x4QQGOa

Fd Chem. Toxic. Vol. 29, No. 10, pp. 713-718, 1991 Printed in Great Britain. All rights reserved 0278-691 S/91 S3.00 + 0.00 Copyright 1991 Pergamon Press pic LIFETIMEcfrt9-WEEK)~>fi4LCAKqiNOGENICITY STUD^pF VINYL CHLORIDE EN RATS* H. P. Til, V. J. Feron and H. R. Immel TNO Toxicology and Nutrition Institute, PO Box 360, 3700 AJ Zeist, The Netherlands (Accepted 25 June 1991) Abstract--A lifetime (149-wk) oral carcinogenicity study of vinyl chloride monomer (VCM) was carried out. Four groups of Wistar rats were used, each consisting of 100 males and 100 females, except for the high-dose group, which comprised 50 males and 50 females. VCM was administered by incorporating polyvinyl chloride powder with a high content of VCM into the diet. The actual exposure levels of VCM were 0 (control), 0.014, 0.13 and 1.3 mg VCM/kg body weight/day. Detailed histopathologjcal examin ation was restricted to the liver. In the final stage of the study, the mortality in the high-dose group was slightly higher than in controls. A variety of VCM-related liver lesions was found in the high-dose group. The lesions included increased incidences of liver-cell polymorphism, hepatic cysts, foci of cellular alteration, neoplastic nodules, hepatocellular carcinomas and angiosarcomas. Compared with controls, there were increased incidences of hepatic foci of cellular alteration in females of the mid-dose group and of basophilic foci of hepatocellular alteration in females of both the low- and mid-dose groups. There was no evidence that feeding of VCM affected the incidence of tumours in organs other than the liver. Thus, the present study showed that the feeding of VCM at a level of 1.3 mg/kg body weight/day can induce neoplastic and non-neoplastic changes in the livers of male as well as female rats. The feeding of 0.014 or 0.13 mg VCM/kg body weight/day may result in an increased incidence of (basophilic) foci of cellular alteration in the liver of female rats. It was concluded that 0.13 mg VCM/kg body weight/day is the no-observed-adverse-effect level with respect to the induction of tumours in rats. INTRODUCTION When, in the early 1970s, angiosarcomas were found in the livers of workers at vinyl chloride monomer (VCM) and polyvinyl chloride (PVC) pro duction plants, drastic measures were taken to reduce VCM exposure in the workplace. By changing pro duction processes and plant installations and by implementing new production processes, worldwide VCM exposure in the workplace could be reduced to 1 ppm. Great efforts were also made for consumers* safety, because it was feared that the residual VCM in PVC packaging materials could be indirectly ingested by humans by way of the PVC-packaged foodstuffs. Therefore, in addition to inhalation studies, mainly aimed at the protection of plant workers, feeding studies were performed. The aims of the latter studies were to examine whether the administration route would influence the effect of VCM, and to establish the oral no-adverse-effect level of VCM. A 28-day feeding study in rats with doses of 0, 90 and 300 mg VCM/kg body weight/day served as the preliminary experiment for a 90-day study in rats receiving oral doses of 0, 30, 100 and 300 mg VCM/kg body weight/day by gavage. A no-effect level of 30 mg/kg This study was sponsored by Verband Kunststofferzeugmde Industrie e.V. (Germany), tThis study was sponsored by the Dutch PVC producers, Dow Qtemical Europe S.A. and Verband Kunststofferzeugende Industrie e.V. (Germany). Abbreviations: PVC = polyvinyl chloride; VCM = vinyl chloride monomer. body weight/day was determined (Feron el al., 1975). These short-term studies were soon followed by a chronic oral toxicity/carcinogenicity studyf. The difficulty in incorporating such a volatile gas as VCM into the diet resulted in a higher actual VCM intake than intended; thus, in this study, 1.7 instead of I mg was used for the low dose, 5 instead of 3 mg for the mid dose, and 14.1 instead of 9 mg/kg body weight/day was used for the high dose. An extra group received 300 mg/kg body weight/day by gav age, 5 days/wk for 83 wk (Feron et al., 1981). From this study, it appeared that hepatic angiosarcomas occurred at doses of 5 mg/kg body weight/day and above, while hepatocellular tumours were found at all doses including the lowest level of 1.7 mg/kg body weight/day. Tumours attributable to VCM exposure and found at other sites included pulmonary angio sarcomas, extrahepatic abdominal angiosarcomas and Zymbal-gland tumours; these neoplasms oc curred at VCM levels of 5 mg/kg body weight and above. Moreover, some evidence was found that VCM enhanced the development of abdominal mesotheliomas and of mammary-gland adenocar cinomas. Thus, this study showed that VCM is a carcinogen in rats when administered orally, and that the no-observed-adverse-effect level of VCM in rats with respect to the induction of tumours was lower than 1.7 mg/kg body weight/day. It was noted that the tumour response in this study might have been influenced by the special route of administration of VCM (continuous slow release of VCM from the PVC granules in the gastro-intestinal tract) and by the limited period of time (4hr/day) that the rats had access to their diets. 713 ucc 110907 714 H. P. Til et al. For accurate extrapolation of experimental data to humans, the information obtained from such a test system should ideally include both a minimumtumour level and a no-tumour level. Therefore, in the early 1980s, a similar lifetime oral carcinogenicity study of VCM in rats was carried out, including lower dose levels of 0.014, 0.13 and 1.3 mg VCM/kg body weight/day. Based on the pathological find ings of the previous long-term study, histopathological examinations in this experiment were focused on the detection of liver lesions, Zymbal-gland tumours, mammary-gland carcinomas and abdomi nal mesotheliomas. This study is described in the present paper. MATERIALS AND METHODS Test materials, VCM was obtained from AK.ZO Zout Chemie (Rotterdam, The Netherlands) and PVC powder was supplied by Shell Nederland Chemie (Pemis, The Netherlands). Details of these materials were given in an earlier paper (Feron el al., 1981). A portion of this PVC powder was separated from residual VCM and was used for incorporation into the diet of the control animals. 50 kg PVC powder was mixed with liquid VCM up to a level of approximately 4600 ppm in a closed steel barrel on a roller-bank. This PVC powder was repacked and stored in tightly closed steel drums (each containing about 10 kg PVC) in a freezer at --20C until a few min before mixing into the diet. Diets. Preparation and administration of the diets were carried out as described previously (Feron et al., 1981). In the present study, however, the diets con tained only 1% PVC powder. For the treated groups, the PVC powder enriched with VCM was mixed into the stock diet to provide concentrations of 0.46, 4.6 and 46 ppm. PVC powder without VCM (lower than 0.2 ppm) was added to the diets such that the total amount of PVC powder in each diet was 1%. The diets were prepared daily, immediately prior to being offered to the rats. As a result of the evaporation of VCM, diets were available to the rats each day for a period of 4 consecutive hr only (generally between 10 am and 2 pm). The VCM content of freshly pre pared diets was determined once every fortnight; a total of 72 determinations for each dosage level was carried out during the study. From these determi nations the average VCM contents of the diets were calculated to be 0.49,4.49 and 44.1 ppm for the low-, mid- and high-dose groups, respectively. Since both the loss of VCM from the diets during the 4-hr period and the VCM content of the faeces were determined, as described previously by Feron et al. (1981), the actual exposure levels of VCM could be calculated. These levels were found to be 0.014, 0.13 and 1.3 mg VCM/kg body weight/day for the low-, mid- and high-dose groups, respectively (Table 1). Animals and housing. Newly weaned albino rats (Cpb; WU; Wistar random), obtained from the SPF colony of the Central Institute for the Breeding of Laboratory Animals (TNO, Zeist, The Netherlands), were used. Healthy rats were randomly allocated to four main groups. Three of the groups--the control, low- and mid-dose groups--each consisted of 100 males and 100 females. The high-dose group con sisted of 50 males and 50 females. After allocation, the rats were maintained on a basal diet for 4-6 hr each day, and given tap-water continuously for 7 days prior to the start of the study. The rats were boused under conventional conditions, in groups of five in suspended stainless-steel cages with wirescreen bottoms. Male rats of each group were housed in one room, and females were housed in a separate room. The rooms were ventilated with approximately 10 air cbanges/hr. The room temperature was kept at 23 1C and the humidity was 40-80%, with a 12-hr light/dark cycle. Rats in bad condition were housed individually in separate cages until they died or were killed in extremis. Experimental protocol. The rats were individually weighed, initially at wk 2 and 4, and once every 4 wk thereafter. Food consumption of 20 rats/sex/group was measured during wk 1-4 and 11-12, and then every 12 wk for periods of 2 wk (e.g. wk 24-26,36-38, etc.). All males still alive at wk 149 and all females surviving until wk 150 were killed in extremis. Samples of a large number of organs and tissues of each rat were fixed in 10% neutral buffered formalin. The organs to be examined microscopically were processed through paraffin wax, sectioned at 5 pm and stained with haematoxylin and eosin. Micro scopic examination was restricted to the liver and to all grossly visible tumours or suspected tumours in the abdominal cavity, Zymbal glands and mammary glands. For each rat, three pieces of liver, taken from three different lobes (the same lobes and the same sites of every rat), were examined microscopically. In addition, sections were prepared from liver tissue showing gross changes. Glutathione levels of the liver. Additional groups of 10 male and 10 female rats received 0,0.014,0.13 and 1.3 mg VCM/kg body weight/day. In wk 40 and 80, five rats/sex/group were killed by decapitation. The Table 1. Designed and actual dosage levels of VCM in rats maintained on diets containing PVC powder Designed VCM treatment Dietary level (ppm) Intake (mg/kg body weight/day) Actual initial dietary VCM level (ppm)' Oral intake of VCM Theoreiicalt (mg/kg body weight/day) Actual (mg/kg body weight/day) Actual oral exposure level of VCMJ (mg/kg body weight/day) 0 0.46 4.6 46 0 0.017 0.17 1.7 0 0.49 4.49 44.1 0 0.022 0.21 2.1 0 O.Olg 0.17 1.7 0 0.014 0.13 1.3 'Average dietary VCM content determined immediately after preparation of the diets, tAssuming no loss of VCM by evaporation from the diets. - __ JOral intake of VCM diminished by the faecal VCM. The VCM excreted in the faeces was considered to be still enclosed in the PVC granules, and thus not to have been in contact with the body. UCC - 110908 ermined, >81), the Iculated. ino rats the SPF ;ding of :rlands), cated to control, of 100 up con ocation, r 4-6 hr y for 7 its were groups ih wirehoused eparate imately kept at a 12-hr housed or were nrnary a from : same illy. In tissue ups of 13 and ad 80, l The auks. Oral carcinogenicity of vinyl chloride in rats Table 2. Cumulative mortality of rati orally exposed to VCM for up to 149 wk Treatment group (mg VCM/kg body weight/day) Initial no. of rats No. of dead rats by cod of wk: 16 36 52 72 88 104 116 128 140 149 0 0.014 0.13 1.3 Malta 100 1 1 2 4 6 13 21 38 59 80 100 1 1 1 4 8 15 25 43 63 80 100 0 0 1 4 7 13 23 38 59 82 so 0 0 0 i 2 7 18 24 35 42 0 0.014 0.13 1.3 Ftuaka 100 0 0 1 2 6 15 26 35 58 76 100 0 0 0 2 4 u 25 41 66 77 100 0 0 0 3 4 14 22 35 61 74 50 0 0 0 2 6 12 16 24 36 45* The value marked with an asterisk differs significantly (Fisher's exact probability test, one-tailed) from the corresponding control value (*P < 0.05). 715 livers were quickly removed for the determination of the glutathione content, according to an automated modification of the method of Sedlak and Lindsay (1968). RESULTS Clinical signs and survival No clinical signs attributable to VCM treatment were apparent in the low- and mid-dose groups. Nodules in the liver were palpated in a high number of high-dose females. Mortality was very low up to wk 72 (Table 2). Thereafter, mortality gradually increased, and at the end of the study the mortality rate was slightly higher in the high-dose group than in the other groups. Body weights andfood consumption Body weights (Fig. 1) and food intake were very similar for all groups. Glutathione levels of the liver There were no significant differences in liver gluta thione levels among the various groups, either after wk 40 or 80. Fig. 1. Average body weights of rats orally exposed to VCM. The body weight curves of the rats receiving 0,0.014, 0.13 or 1.3 mg VCM/kg body weight/day all lie within the shaded areas. Gross pathology The type and incidence of gross findings in the liver are given in Table 3. The incidence of liver nodules suspected of being tumours was higher in males and females of the high-dose groups than in controls or lower-dose rats. Many of these tumorous masses were small (up to 1 cm dia.), solid and pale, or had the same colour as the adjacent liver tissue; some were large (2-4 cm dia.), soft or firm, and occasionally slightly haemorrhagic. The appearance of cysts in the livers of old rats, particularly in females, is a common finding in the strain of rat used. However, the inci dence of such cysts was much higher in females of the high-dose group than in females of the other groups, including the control group. The cysts varied widely in size, were often multiple and generally contained a turbid, watery liquid. There was no evidence that any of the other gross lesions observed were related to the administration of VCM. Microscopy of the liver The type and incidence of treatment-related histopathologica! changes found in the liver are given in Table 4. Increased incidences of foci of cellular alteration, neoplastic nodules, hepatocellular carci nomas, liver-cell polymorphism and cysts occurred in the high-dose group. Moreover, in this group two females and one male had developed a hepatic angio sarcoma, whereas such tumours were not seen in any of the other groups. Since the morphology of the hepatocellular lesions was essentially the same as that of the liver lesions described in full detail in our previous publication on the chronic effects of oral administration of VCM in rats (Feron et al., 1981), no further description of these alterations is presented in this paper. The morphology of VCM-induced hepatic angiosarcomas has also been reported by Spit et al. (1981). Table 4 also shows that in females but not in males of both the low- and mid-dose groups, the incidence of basophilic foci of cellular alteration was significantly higher than in controls. In addition, in the mid-dose group the number of females showing foci of cellular alteration was significantly higher than in the control group. Other hepatic lesions observed included prominent sinusoidal cells, bile-duct proliferation, cholangiofibrosis, focal vacuolization of hepatocytes, accumu lation of RES cells, focal necrosis, individual cell necrosis and centrilobular liver-cell degeneration. ucc 110909 716 H. P. Tn. et a/. Table 3. Type and incidence of macroscopic changes in the liver of rats orally exposed to VCM Incidence of change Type of change Treatment group (mg VCM/kg body weight/day)... 0 Males 0.014 0.13 1.3 Females 0 0.014 0.13 1.3 Initial no. of rats Effective no. of rats Liver Tumour or suspected tumour Prominent lobular pattern Swollen/enlarged Discoloured (a) pale (b) dark Cyst(s) Small Small surface lesions (rosette, constriction) Spotted Granular surface 100 100 100 30 99 100 99 49 100 100 100 9* 100 96 50 49 1 1 2 7*** 2 1 3 8" 1 4 20 2 4 5l 0 1 31 3 2 00 3 2 01 1 3 12 3 3 00 0 001 2 1 54 17 14 22 33*** 01 10 1 0 t0 J 6 a2 0 0 20 3 0 00 2 1 62 2 0 02 0 1 20 Values marked with asterisks differ significantly (chi-square test) from the corresponding control value (**P < 0.01; ***P <0.001). There was no evidence that any of these lesions was related to the feeding of VCM. Microscopy of organs other than the liver There was no evidence that the incidence of mammary-gland or intra-abdominal tumours was related to the administration of VCM. No Zymbalgland tumours were found. DISCUSSION In this supplementary study no effects attributable to VCM were observed other than liver changes, and a slightly higher mortality in the high-dose group compared with the control group during the last 6-9 months of the experimental period. The increased mortality in the high-dose group might be due to the increased incidence of neoplastic and non-neoplastic liver changes in this group. In the previous long-term oral rat study with VCM (Feron et a/., 1981) a similar adverse effect on survival was observed in females receiving 1.7 mg VCM/kg body weight/day. Feron et al. (1981) reported that ingestion of VCM at a level of 1.7 mg/kg body weight/day resulted in a number of hepatic changes. These changes included an increased incidence of foci of cellular alteration and of neoplastic nodules, a few hepatocellular carci nomas, an increased incidence and degree of liver-cell Table 4. Type and incidence of treatment-related histopathological change* in the liven of ran orally exposed to VCM Incidence of change Type of chxnget Treatment group (mg VCM/kg body weight/day)... 0 Males 0.014 0.13 1.3 Females 0 0.014 0.13 1.3 Foci of cellular alteration Clear-cell foci One or a few Several Basophilic foci One or a few Several Mixed cell foci One or a few Eosinophilic foci One or a few All types of foci (total) No. of rats examined... 99 99 99 49 98 100 96 49 12 8 8 16** 4 53 0 0 0 3* 0 00 13*** 6*** 4 2 3 8* 0 0 00 9 20* 26*** 19*** 0 10 2 0 1 22 6 64 8* 1 I 21 17 12 15 30 0 0i 19 32 34 10*** 58 No. of fod-bearing rata... 16 12 15 23*** 19 27 31* 32*** Neoplastic nodules One Few 0 0 0 1* 0 1 1 0 0 02 0 00 9*** 1 Hepatocellular carcinoma 0 0 0 3* 1 0I 3 Angiosarcoma 0 0 01 0 00 2 Liver-cell polymorphism Slight Moderate Severe Cysts One Few Many 27 23 26 19 46 41 49 4 4 7 10** 14 13 8 i i 13 2 34 i 0 10 4 4 34 0 0 00 3 24 II 11 12 3 49 23 15* 9*** i 7 24*** (Specific hepatocellular lesions were classified according to Squire and Levin (1973). Values marked with asterisks differ significantly (Fisher's exact probability test, one-tailed) from the corresponding control value (*P < O.OS; *P <0.01; **)< 0.001). ucc 110910 49 I" 1 0 2 1 J3TM 0 4 2 0 .001). : last 6-9 ncrcased ue to the eoplastic jng-term a similar females ofVCM Ited in a included Iteration ar carciliver-cell A 13-- 6-- 19-- 2 8* lo ss 32-- 1 3 2 3 5* 1 7 4-- cflLOS; Oral carcinogenicity of vinyl chloride in rats 717 polymorphism, and an increase in the number hepatic cysts. In the present study the same type of treatmentrelated liver changes were found at 1.3 mg VCM/kg body weight/day. In addition, one male and two females of the high-dose group developed a hepatic angiosarcoma; in the previous experiment hepatic angiosarcomas were observed at VCM levels of 5 mg/kg body weight/day and above, but not at the lowest exposure level of 1.7 mg/kg body weight/day. Maltoni et al. (1981) also reported hepatic tumours in rats receiving VCM dissolved in olive oil, once daily by gavage, 4-5days/wk for 59 wk, and then kept untreated until wk 136; namely, three hepatic angiosarcomas, one extrahepatic angio sarcoma and one hepatoma in 149 rats receiving 1 mg VCM/kg body weight/day. No liver tumours were observed in rats receiving 0.03 mg VCM/kg body weight/day. Evans et al. (cited by ECETOC, 1988) observed one hepatic and one subcutaneous angio sarcoma in 54 male rats receiving 25 ppm VCM in their drinking-water for up to 152 wk, a level approximately equivalent to 2 mg VCM/kg body weight/day. Increased incidences of basophilic foci of hepato cellular alteration occurred in males of the high-dose group and in females of each of the test groups. In females, the incidence of this type of foci of cellular alteration in the liver increased with increasing dose level. Moreover, the number of rats showing foci of cellular alteration in the liver (all type of foci) was significantly higher in males of the high-dose group in females of the mid- and high-dose groups than in controls. These results suggest a relationship between the oral exposure to VCM at all dose levels and the increased occurrence of foci of cellular alteration in the liver of female rats. The incidence of basophilic foci of hepatocellular alteration in females of the control group (9%) was clearly lower than in female controls used in the previous long-term oral rat study with VCM (18%; Feron et al., 1981). Incidences of 12-17% were also found in females of control groups in three other long-term rat studies carried out at our Institute more or less simultaneously with the pre vious VCM study. These data cast doubt on the toxicological significance of the slight increase in incidence of this type of foci in the low- and mid-dose groups in the present study. However, the incidence of basophilic foci in females of control groups of four other long-term rat studies performed at our Institute, during the same period as the present VCM study, varied from 0-8%. Compared with these values, an incidence of 9% is not unusually low, and thus it supports the suggestion that the feeding of VCM at levels of 0.014 or 0.13 mg/kg body weight/day resulted in more basophilic fod of cellular alteration in the liver of females. At these levels no hepatic neoplasms or VCM-related tumours at other sites were observed, justifying the conclusion that 0.13 mg VCM/kg body weight/day was a noadverse-effect level with respect to the occurrence of tumours. A directive of the European Community (EEC, 1978) imposes a maximum permitted limit of 1 ppm residual VCM in finished plastic materials or articles intended for use in contact with foodstuffs. The same directive stipulates that any migration of VCM to food must not produce a concentration in the food greater than 10ppb, which is close to the detection limit of VCM in foodstuffs. Calculations based on a survey of residual levels ofVCM in PVC bottles, films and food in the period of 1974-1977, suggest that the human intake of VCM from the average diet was less than 0.1 pg/person/day (Crosby, 1982; MAFF, 1978). More recent data suggest maximum average daily intakes of 0.02 pg/person/day in the UK (MAFF, 1984) and 0.025 pg/person/day in the USA (FDA, 1986). Extrapolation of the results of the present rat study to humans using a linear nonthreshold extrapolation model, and taking into account an acceptable cancer risk from oral VCM exposure for humans of 10~`, leads to an acceptable oral intake of 0.4 pg VCM/person/day. Presently, the estimated maximum oral daily intake is only 0.025 /rg VCM/person, and the estimated cancer risk is 16 times lower. Moreover, the extrapolation model used is a conservative and prudent one, which implies that the actual cancer risk might be (considerably) lower than that calculated by means of the linear model. It may be emphasized that, on the basis of the results described in the present paper and the estimated intake of0.025 pg/person/day, the FDA (1986) calcu lated an individual lifetime risk of below 10~7, and stated that the linear proportional model used and the numerous conservatisms in the exposure esti mates exaggerate the risk (FDA, 1986). Comparison of the results of an epidemiological study of vinyl chloride workers with the predicted cancer risk in a cohort of workers (based on the results of VCM studies in rats) indicates that humans are less sensitive to the carcinogenic action of VCM than are rats (Gehring et al., 1979). More recent cancer risk assess ments of vinyl chloride using both rat and human data suggest similar sensitivity of both species (Chen and Blancato, 1989; Swaen et al., 1987). These data and considerations allow us to conclude that the cancer risk of oral daily intakes of 0.02 or even 0.1 pg VCM/person/day is small enough to be practically negligible. Indeed, such intakes can be considered virtually safe. Conclusions This study showed that the ingestion ofVCM at a level of 1.3 mg/kg body weight/day induces neoplastic and non-neoplastic changes in the livers of male and female rats, and that a level of 0.13 mg/kg body weight/day may result in an increased incidence of basophilic fod of cellular alteration in the livers of female but not of male rats. The no-observedadverse-effect level with respect to the induction of tumours in rats appeared to be 0.13 mg VCM/kg body weight/day. lire cancer risk of oral daily intakes of 0.02 or 0.1 pg VCM/person/day is estimated to be negligibly small. Acknowledgements--'We would like to thank Ms G. G. M. Fleer, Ms A. A. van Tuyl, Mr D. C. Veldhuysen and Mr J. M. Blom for conducting the animal experiments, Mr A. Schouten for the VCM determinations, Ms M. M. Andringa for typing the manuscript and Dr A. P. de Groot for critical review of this paper. 718 H. P. Til et al. REFERENCES Chen C. W. and Blancato J. N. (1989) Incorporation of biological information in cancer risk assessment: example--vinyl chloride. Cell Biology and Toxicology 5, 417-444. Crosby N. T. (1982) Analysis for residual monomer levels in plastics and in foods. Analytical Proceedings 19,428-430. ECETOC (1988) The mutagenicity and carcinogenicity of vinyl chloride: a historical review and assessment. Techni cal Report No. 31. p. 147. EEC (1978) Directive of the European Community. 30 January, 1978. 78/142/EEC. FDA (1986) Federal Register 51 (22), 4173-4188. Feron V. J., Hendriksen C. F. M., Speck A. J., Til H. P. and Spit B. J. (1981) Lifespan oral toxicity study of vinyl chloride in rats. Food and Cosmetics Toxicology 19, 317-333. Feron V. J., Speek A. J., Willems M. 1., Van Battum D. and de Groot A. P. (1973) Observations on the oral adminis tration and toxicity of vinyl chloride in rats. Food and Cosmetics Toxicology 13, 633-638. Gehring P. J., Watanabe P. G. and Park C. N. (1979) Risk of angiosarcoma in workers exposed to vinyl chloride as predicted from studies in rats. Toxicology and Applied Pharmacology 49, 15-21. MAFF (1978) The Surveillance of Food Contamination in the United Kingdom: Survey of Vinyl Chloride Content of Polyvinyl Chloride for Food Contact and of Foods. The Second Report on Vinyl Chloride. Food Surveillance Paper No. 2.6. pp. iv + 16. HMSO, London. MAFF (1984) Progress Report 1984. Fourteenth Report of the Steering Group on Food Surveillance. HMSO, London. Maltoni C., Lefemine G., Ciliberti A., Cotti G. and Carretti D. (1981) Carcinogenicity bioassays of vinyl chloride monomer a model of risk assessment on an experimental basis. Environmental Health Perspectives 41, 3-29. Sedlak J. and Lindsay R. H. (1968) Estimation of total, protein-bound, and nonprotein-bound sulfhydryl groups in tissue with EUman's reagent. Analytical Biochemistry 25, 192-205, Spit B. J., Feron V. J. and Hendriksen C. F. M. (1981) Ultrastructure of hepatic angiosarcoma in rats induced by vinyl chloride. Experimental and Molecular Pathology 35, 277-284. Squire R. A. and Levitt M. H. (1973) Report of a workshop on classification of specific hepatocellular lesions in rats. Cancer Research 35, 3214-3223. Swaen G. M., Hollander A. E. M., Krocs R., den Engelse L., Feron V. J., Mulder G. J., Verbeek A. L. M., Verschuuren H. G., Vogel E. W. and van der Wielen A. W. (1987) A scientific basis for the risk assessment ofvinyl chloride. Regulatory Toxicology and Pharmacology 7, 120-127. ucc 110912