Page 10: Guidelines for Canadian Drinking Water Quality: Guideline Technical Document – Vinyl Chloride

Part II. Science and Technical Considerations (continued)

9.0 Health Effects

9.1 Effects in Humans

9.1.1 Acute Effects

Vinyl chloride is a narcotic agent, and loss of consciousness can occur from exposure to high concentrations (25 000 mg/m3). Acute exposure to high concentrations in air also causes central nervous system depression, with symptoms of dizziness, light-headedness, nausea, headache, irritability, poor memory, tingling sensations, weight loss, irritation of the respiratory tract and chronic bronchitis (IARC, 2008). Autopsies of several vinyl chloride workers who died following exposure to high levels revealed congestion of the liver, spleen and kidney (Cook et al., 1971).

9.1.2 Chronic Effects and Cancer Epidemiology

The association of vinyl chloride with the risk of death from liver angiosarcoma (ASL; cancer of the liver blood vessels) reached public attention after the discovery of three cases in a vinyl chloride plant (Creech and Johnson, 1974). Tumours of the respiratory tract, digestive system, cardiovascular system and other sites from vinyl chloride exposure have been reported in a few controversial studies; however, the risk of ASL has been demonstrated by many investigators to increase with the exposure length and intensity, the latency period, the year of employment (before 1970) and age at first exposure (Simonato et al., 1991; Mundt et al., 2000; Wong et al., 2002; Bosetti et al., 2003; Lewis et al., 2003). Other cancers, such as brain and lung cancer, lymphoma and HCC, have been sparsely observed (Monson et al., 1975; Waxweiler et al., 1976; Pirastu et al., 1990; Simonato et al., 1991; Wong et al., 1991) and their causal relationship with vinyl chloride remains controversial.

In the 1970s, the International Agency for Research on Cancer (IARC) coordinated a European multicentre study within Italy, Norway, Sweden and the United Kingdom with 14 351 individuals from 19 factories to investigate the relationship between exposure to vinyl chloride and liver cancer as well as cancer risks for sites other than the liver. IARC (1979) concluded that vinyl chloride could be associated with HCC, brain tumours, lung tumours and malignancies of lymphatic and haematopoietic tissues. The data, analysed by Simonato et al. (1991), revealed a nearly 3-fold increase in liver cancer deaths: 24 observed compared with 8.4 expected (standardized mortality ratio [SMR] = 286; 95% confidence interval [CI] = 186-425). The excess deaths from liver cancer were clearly associated with the elapsed time since the first exposure, duration of employment and quantitative exposure scenarios based on job titles (ranked level: ≤ 50, 50-499, ≥ 500 ppm; and cumulative estimates: 0-1999, 2000-5999, 6000-9999, ≥ 10 000 ppm-years). No statistically significant excess mortality was reported for the cancers at sites other than the liver due to exposure to vinyl chloride. A further analysis by Ward et al. (2001) followed up on an additional 8 years of mortality and cancer incidence and reported that mortality from all causes was lower than expected, whereas cancer mortality was close to expected. A total of 53 deaths from primary liver cancer (SMR = 2.40; 95% CI = 1.80-3.14) and 18 cases of liver cancer were reported which included 37 angiosarcomas (ASL), 10 HCCs, and 24 other liver cancers; a clear exposure response relationship was reported for all liver cancers, ASL, and HCC. Bosetti et al. (2003) confirmed the relationship between liver cancer and vinyl chloride exposure, however, showed that the increasing trend for HCC was appreciably less pronounced than for ASL, and that the relative risk for HCC was statistically significant only in the highest category of cumulative exposure. Again, no association was found with other cancer types.

The North American multicentre cohort of vinyl chloride workers has been followed by various investigators, with a focus on the plants in Calvert City and Louisville, Kentucky (Creech and Johnson, 1974; Monson et al., 1975; Waxweiler et al., 1976; Lewis and Rempala, 2003; Lewis et al., 2003). Monson et al. (1975) were the first to find increased deaths from all cancers, particularly of the liver, biliary tract (SMR = 11.0) and brain (SMR = 4.2), but also of the lung (SMR = 1.6) and lymphatic tissue (SMR = 1.5). In a follow-up based on cumulative estimates ranking workers within exposure groups ≤ 100, 100-400, 400-1000 and ≥ 1000 ppm multiplied by the number of months worked, Lewis and Rempala (2003) confirmed the high occurrence of liver and brain cancer among workers from the Louisville plant. Deaths from total liver cancer and ASL were strongly related to vinyl chloride exposure intensity (P < 0.001). Wong et al.(1991) analysed data, including from the Louisville plant, for 10 173 men exposed to vinyl chloride for at least 1 year before 1973 at 37 plants in the United States. This represents the largest group of vinyl chloride workers in North America. The study confirmed that workers exposed to vinyl chloride experienced a significant excess mortality due to ASL (15 deaths), cancer of the liver (unspecified type) and biliary tract (SMR = 641) and cancer of the brain and central nervous system (23 deaths compared with 12.8 expected; SMR = 180). However, half of the cases of brain cancer came from two plants that were manufacturing only PVC. Only the deaths from liver cancer showed increased occurrence by length of exposure and by latency period since first exposure and decreasing trends by age at first exposure. The study did not find any excess in respiratory cancer or lymphatic and haematopoietic cancers. Lewis et al. (2003) separated the Louisville cases from the large U.S. vinyl chloride workers cohort and found that liver cancer mortality remained high compared to the total U.S. cohort (SMR = 400 and 359, respectively), but brain cancer deaths increased dramatically (SMR = 229 and 142, respectively). Another follow-up to December 31, 1995, by Mundt et al. (2000) added the number of deaths after a latency period of 20-50 years. Again, the cancer risk was higher for workers employed earlier (SMR = 499 before 1950) and longer (SMR = 434 for 30 years or more). Liver and biliary tract cancer deaths were still in excess, although slightly decreased since the 1980s; brain cancer deaths were no longer different from the controls.

Data from nine United Kingdom vinyl chloride plants involving 7717 workers employed between 1940 and 1974 found only two cases of ASL (Fox and Collier, 1977). A follow-up including 5498 workers from this study added 10 years of latency and discerned exposure categories based on job titles and length of employment (Jones et al., 1988). A significant increase of deaths from liver cancer (SMR = 567) was reported among autoclave workers with the highest vinyl chloride exposure; out of 11 observed deaths, 2 were expected, and 7 of the 11 deaths were largely attributable to ASL (SMR = 1842).

Pirastu et al. (1990) conducted a mortality study on 5946 vinyl chloride and PVC workers from nine Italian plants. Analysis of the data from 253 deaths (death certificates, clinical and pathological information) confirmed the carcinogenic action of vinyl chloride, with 14 liver cancers, but no other cancers were reported in any of the other suggested target organs (i.e., lung, lymphopoietic tissues, brain). The cancers were observed in various job titles that exposed workers to different concentrations; however, these levels were not specified by the authors. Of the 14 liver cancers, seven were ASL and two were HCC. Later, Pirastu et al. (2003) updated the mortality and occupational history for an Italian plant in Porto Marghera until 1999. Cumulative exposures were determined on the basis of job- and time-specific exposure estimates, and classified into six exposure categories (0-735, 735-2379, 2379-5188, 5188-7531 and 7531-9400 ppm-years) with employment as an autoclave worker (ever/never) also considered in the analyses. Clinical and pathological data which gave the best diagnosis were used to identify cases of liver angiosarcoma and hepatocellular carcinoma. With regional rates as a reference, mortality in the cohort for all causes (SMR, 0.75; 90% CI, 0.68-0.83) and all cancers (SMR, 0.94; 90% CI, 0.81-1.09) was lower than expected. For all causes, the analysis by time since leaving employment and adjusted for latency showed that the SMR in the first year after leaving employment was 2.76 (90% CI, 1.94-3.91). Mortality rates for liver angiosarcoma (six cases) increased with latency and cumulative exposure, however, no cases were associated with duration of employment of less than 12 years, latency of less than 10 years or cumulative exposure of less than 2379 ppm-years. Mortality rates for hepatocellular carcinoma (12 cases) and liver cirrhosis (20 cases) showed a similar pattern. For this cohort, Mastrangelo et al. (2003, 2008) confirmed that an increase in deaths from liver cancer was associated with vinyl chloride inhalation exposure.

Boffetta et al. (2003) analysed the data from six studies of the U.S. and European cohorts. They found a low incremental average SMR from the studies for liver cancer other than ASL (SMR = 1.35; 95% CI = 1.04-1.77). Of the six studies, only four were still positive for liver cancer when ASL was excluded. Moreover, no association between brain cancer or lymphatic and haematopoietic neoplasms and vinyl chloride exposure was found, contrary to the findings of Theriault and Allard (1981) and Wong et al. (1991).

Laplanche et al. (1992) analysed data from a French prospective cohort study by the French National Institute of Health and Medical Research (INSERM) representing 1100 exposed and 1100 unexposed subjects. Diseases of the respiratory system did not differ between the two groups (relative risk [RR] = 1.1; 95% CI = 0.7-1.8); however, 3 cases of ASL and 14 cases of Raynaud's disease were found among the exposed group, and 1 case of Raynaud's disease was noted among non-exposed subjects.

Data from male workers employed for at least 1 year during the period 1950-1992 in Taiwan were analysed by Wong et al. (2002). Being employed when younger than 30 years of age increased the risk of death (SMR = 2.24). Those employed before 1970 were 5 times more likely to develop liver cancer (unspecified type) than the general population. No increased risk was found for longer employment duration or for development of brain tumours. The study confirmed an increased mortality from cancer (SMR = 1.30), with the liver as the main target organ (SMR = 1.78).

Based on five registry sources from the United Kingdom, including the national cancer registry and death certificates, no non-occupational cases of ASL among residents within 10 km of a vinyl chloride industrial site were found (Elliott and Kleinschmidt, 1997). Of the 52 cases of ASL registered during 1979-1986, only 11 were within 10 km of a vinyl chloride industrial site. From those 11 cases, 10 had been employed in a vinyl chloride plant, whereas the remaining case was thought to be a misclassified ASL diagnosis.

Infante et al. (2009) reported two cases of ASL in hairdressers and barbers who used hairsprays containing vinyl chloride as a propellant over a period of 4 to 5 years between 1966 and 1973. The peak exposure for Case 1 (hairdresser) was estimated to range from 129 to 1234 ppm, with a 8-hour time weighted- average (TWA) vinyl chloride exposure estimated to have ranged from 70 to 1037 ppm (median TWA estimated at 228 ppm); estimates for Case 1 were dependant on the products used (% vinyl chloride varied based on product used), the length of time of spraying the product, and the number of estimated air exchanges per hour. For Case 2 (barber), atmospheric levels of vinyl chloride from a 22 second spray in a room with 4 air exchanges were estimated to range from 129 ppm at the outset to approximately 40 ppm after 40 minutes had elapsed, with an average exposure of 70 ppm over the 40 minute period. The authors concluded that to their knowledge this represents the first literature reports of ASL cases identified among hairdressers and barbers who used hair sprays containing vinyl chloride as a propellant.

9.1.3 Non-cancer Endpoint Epidemiology

The liver remains the primary target organ of vinyl chloride for non-cancer endpoints (U.S. EPA, 2000b). Occupational exposure to high levels of vinyl chloride (2400 ppm-month) is an independent risk factor for the development of liver fibrosis (cirrhosis) (Hsiaos et al., 2004). Impaired liver functions with evidence of porphyria, liver enlargement, hepatocyte hypertrophy and fibrosis were observed by many authors in workers exposed to vinyl chloride for a variable, but considerable, amount of time (Gedigk et al., 1975; Lilis et al., 1975; Popper and Thomas, 1975; Berk et al., 1976; Doss et al., 1984).

Vinyl chloride tank cleaners were exposed to vinyl chloride at concentrations of hundreds of parts per million, producing a wide range of effects, including Raynaud-like syndrome, acroosteolysis (bone resorption in the fingers and toes), thrombocytopenia, scleroderma-like dermatitis, thyroid insufficiency, damage to the liver (steatohepatitis), spleen and lungs, as well as functional disturbances of the central nervous system (Cook et al., 1971; Lilis et al., 1975; ECETOC, 1988; U.S. EPA, 2000b). An effect of vinyl chloride on coronary heart disease was not found, but a direct relationship between vinyl chloride exposure and development of arterial hypertension was observed (Kotseva, 1996). Ho et al. (1991) revealed an increase of glutamic pyruvic transaminase concentrations before and after the occurrence of liver dysfunction in workers exposed to vinyl chloride levels ranging from 1 to 21 ppm.

9.1.4 Mutagenicity/Genotoxicity

Accidental and occupational exposures to vinyl chloride and its metabolites have been associated with chromosomal aberrations and gene mutations in humans. A single acute exposure to vinyl chloride in air after an environmental accident in Germany led to slight increases in chromatid breaks, acentric fragments, dicentrics and translocations in peripheral blood lymphocytes in 27 exposed individuals compared with those unexposed (Hüttner and Nikolova, 1998). However, the exposure did not produce any significant chromosomal abnormalities in the HPRT locus of lymphocytes, measured immediately or 2 years later, compared with 29 non-exposed individuals (Becker et al., 2001).

Mutations of both pro-carcinogenic genes p53 (A:T mis-sense mutations) and ras, leading to the expression of altered p53 and p21ras proteins (Trivers et al., 1995), were detected in tumour samples from workers who developed ASL and HCC after being occupationally exposed to vinyl chloride in the United States, France and Taiwan (Marion et al., 1991; De Vivo et al., 1994; Weihrauch et al., 2000, 2001, 2002). A dose-response relationship was observed between vinyl chloride exposure and the mutation of p53 and ras (Smith et al., 1998). Further, chronic occupational exposure to 1.3-16 ppm vinyl chloride resulted in sister chromatid exchange and chromosomal aberrations such as DNA breaks in a group of 52 workers (Sinués et al., 1991) and in another group of 57 workers exposed to an average of 2 ppm vinyl chloride with occasional excursions of up to 1000 ppm (Purchase et al., 1985).

Gene polymorphism has reportedly affected vinyl chloride workers in Taiwan and China. An increase in sister chromatid exchanges and micronuclei was observed in exposed workers and in individuals expressing specific alleles. For example, the homozygous individuals bearing the c2c2 allele of the CYP2E1 gene (P < 0.05) or the variant allele Arg399Gln of the DNA repair gene XRCC1 (P < 0.05) had more micronuclei (Wong et al., 2003; Ji et al., 2010). The variant allele c2c2 of CYP2E1 reportedly metabolizes vinyl chloride at a higher rate, thus generating more CEO per time unit (Wong et al., 2003). A significant increase in liver fibrosis incidence was reported among CYP2E1 c2c2 allele carriers in a group of 320 vinyl chloride workers employed in five different vinyl chloride factories in Taiwan (Hsieh et al., 2007). Fucic et al. (1996) demonstrated that the chromosomal damage was reversible after occupational exposure levels dropped below 1 ppm in vinyl chloride plants. However, sister chromatid exchanges were persistent and detected up to 10 years after the last occupational exposure to levels higher than 1 ppm (Fucic et al., 1996).

Chiang et al. (1997) reported that 16 µM of CEO or CAA for 24 hours induced similar mutation frequencies of the HPRT locus in human B-lymphocytes, but that CAA was more cytotoxic (cell proliferation assay). When plotted as a function of relative survival, CEO induced mutations at a rate similar to that of vinyl chloride, but 6 times higher than that of CAA. The authors concluded that CEO is responsible for larger deletions and for most of the mutations induced by vinyl chloride exposure (Chiang et al., 1997).

9.1.5 Reproductive and Developmental Toxicity

No adverse reproductive effects were reported in workers exposed to vinyl chloride (Hemminki et al., 1984), although an increase in total birth malformations in populations employed in the vinyl chloride industry has been suggested by a few investigators (Infante et al., 1976; Theriault et al., 1983). An association between parental occupational exposure to vinyl chloride and fetal loss has also been suggested (RR = 1.8) (Infante et al., 1976; Waxweiler et al., 1976), but exposures were not measured.

Theriault et al. (1983) investigated the incidence of birth defects in infants born between 1966 and 1979 to residents of Shawinigan, Quebec, a town where a vinyl chloride polymerization plant had been in operation since 1943. The incidence of birth defects (observed 159 vs. expected 107) was significantly higher in Shawinigan than in any or all of the matched control towns with no potential exposure to vinyl chloride. The incidence rate peaked in March and was lowest in September, which corresponded to variation in atmospheric levels of vinyl chloride at the time of the first 3 months of pregnancy (vinyl chloride monomer was not detected in air samples taken between December and February). No such variations were observed in the control communities. However, there was no excess of stillbirths in Shawinigan. As several other industries emitted pollutants into the atmosphere in Shawinigan, these observations remain inconclusive.

9.1.6 Children's Sensitivity

Evidence from animal studies suggests that very young children may be particularly sensitive to the carcinogenic effects of vinyl chloride (see Section 9.2.6). The rate of cell division in the brain, liver and kidney is much higher in the first 2 years of life than in later childhood, making the liver more sensitive to DNA adduction and hepatocarcinogenesis from vinyl chloride exposure during this time (Ginsberg, 2003). A second growth peak for the liver may occur around the age of 6, indicating that children this age may also be sensitive to vinyl chloride (Ginsberg, 2003).

9.2 Effects in Experimental Animals and in Vitro

9.2.1 Acute Effects

Prodan et al. (1975) reported that inhalation exposure to high concentrations of vinyl chloride for two hours has a narcotic effect described by phases of excitement, tranquillity and then death. The most sensitive experimental animals were mice, with 70% mortality at 107 mg/L, followed by rats, with 23% mortality at 375 mg/L, guinea pigs, with 75% mortality at 600 mg/L, and rabbits, with 50% mortality at 600 mg/L, corresponding to median lethal doses (LD50s) of 27 419, 47 640, 236 215 and 263 215 ppm for the four species, respectively.

9.2.2 Subchronic Exposure

Administration of vinyl chloride monomer dissolved in soybean oil by gavage to groups of Wistar rats (30 per group) at doses of 0, 30, 100 or 300 mg/kg bw per day, 6 days/week, for 13 weeks did not result in adverse effects at the 30 mg/kg bw per day level. A dose-related increase in relative liver weight was observed at the two highest dose levels; however, the increase was statistically significant only at 300 mg/kg bw per day (Feron et al., 1975).

Inhalation exposure of guinea pigs, rats, rabbits and dogs to 50 ppm (130 mg/m3), 100 ppm (260 mg/m3), 200 ppm (540 mg/m3) or 500 ppm (1300 mg/m3 vinyl chloride for 7 hours/day, 5 days/week, for 26 weeks did not induce any adverse effects at 50 ppm, in terms of appearance, growth, haematology, liver weight and mortality; however, rats exposed to 100 ppm had increased liver weights, one of the more sensitive indicators of hepatotoxicity (Torkelson et al., 1961). Exposure to 200 ppm resulted in increased relative liver weight in male and female rats, but there was no biochemical or microscopic evidence of liver damage; rabbits exhibited histological changes (characterized as granular degeneration and necrosis with some vacuolization and cellular infiltration) in the centrilobular area of the liver. No effects were observed in guinea pigs or dogs exposed to 200 ppm. Histopathological lesions of the liver (centrilobular granular degeneration) and increased organ weight occurred in rats exposed to 500 ppm.

9.2.3 Long-term Exposure and Carcinogenicity

Vinyl chloride administered by inhalation or ingestion induces neoplasms at multiple sites in several species of animals. In the rat, mouse and hamster, it has induced hepatic haemangiosarcomas (equivalent to ASL in humans), Zymbal gland tumours, nephroblastomas, pulmonary and mammary gland tumours and forestomach papillomas.

The minimum dose at which compound-related tumours were induced by inhalation (4 hours/day, 5 days/week) was 10 ppm (26 mg/m3) for 52 weeks for rats, 50 ppm (130 mg/m3) for 30 weeks for mice and 500 ppm (1300 mg/m3) for 30 weeks for hamsters (Maltoni et al., 1981). When vinyl chloride in PVC powder was administered orally to rats, the minimum effective dose of vinyl chloride (causing liver tumours) was 1.7 mg/kg bw per day (Til et al., 1991).

The first experimental data on the carcinogenic effects of vinyl chloride in rats were published by Viola et al. (1971). Male Wistar rats exposed by inhalation to very high doses (30 000 ppm) for 4 hours daily, 5 days/week, for 12 months, developed skin, lung and osteochondroma tumours. Moreover, general central nervous system degeneration, specifically degeneration of granular cells, Purkinje cells and the cerebellum was observed in 3-month-old rats.

The most comprehensive carcinogenesis bioassays relevant to the assessment of risk associated with the ingestion of vinyl chloride are those of Maltoni et al. (1981), Feron et al. (1981) and Til et al. (1983, 1991). Groups of 40 male and 40 female (80 per group) 13-week-old Sprague-Dawley rats received gastric intubations of 0, 3.33, 16.65 or 50 mg/kg bw per day of vinyl chloride dissolved in olive oil, 4-5 times per week for 52 weeks, followed by an extended observation period (Maltoni et al., 1981). At 136 weeks, no hepatic angiosarcomas were observed in low-dose and control rats. No dose-response relationship was observed in the induction of other tumours in these animals. Two nephroblastomas (one in each sex), one Zymbal gland tumour, one thymic angiosarcoma and one intraabdominal angiosarcoma were observed in the 50 mg/kg bw per day group; two Zymbal gland carcinomas (one in each sex) and three nephroblastomas (two in males and one in females) were observed in rats administered 16.65 mg/kg bw per day. Seventeen incidences of ASL (eight in males and nine in females) were found in the 50 mg/kg bw per day group; 10 ASLs (four in males and six in females) occurred in rats administered the 16.65 mg/kg bw per day dose (Maltoni et al., 1981).

In another experiment by Maltoni et al. (1981), vinyl chloride doses of 0.003, 0.3 or 1.0 mg/kg bw per day were administered by stomach tube to 10-week-old rats (75 of each sex per group), 5 days/week for 59 weeks. At 136 weeks, no exposure-related liver or kidney tumours were observed in low-dose or control animals. ASL was found in 1/74 males and 2/75 females administered 1.0 mg/kg bw per day and in 1/73 females administered 0.3 mg/kg bw per day, however, these increased incidences were not statistically significant. When Sprague-Dawley rats were exposed to high inhalation concentrations(above 10 000 ppm) for 4 hours daily, 5 days/week, for 52 weeks, ASL, Zymbal gland carcinoma, forestomach tumours and neuroblastomas increased significantly. Mammary gland tumours developed at 5 ppm and above in female rats, whereas nephroblastomas occurred in males at 100-2500 ppm. (Maltoni et al., 1981).

In a rat lifespan carcinogenicity study (Feron et al., 1981), Wistar rats were exposed to a mixture of vinyl chloride monomer and PVC powder via the diet. Groups of 60-80 rats of each sex were fed vinyl chloride at doses of 0, 1.7, 5.0 or 14.1 mg/kg bw per day; as a positive control, vinyl chloride dissolved in soybean oil (300 mg/kg bw) was also administered by gavage, 5 days/week. The experiment was terminated once 75% mortality was observed in the control group (135 weeks for males and 144 weeks for females). Hepatic and lung angiosarcomas were observed in males (27/55, 27/59, 6/56) and females (2/59, 9/57, 29/54) in the three highest dose groups (5.0, 14.1, 300 mg/kg bw per day), respectively, but not in the low-dose group or controls. Males at the 5.0 and 14.1 mg/kg bw per day doses developed 3 times more angiosarcomas than females. HCC and an increased incidence of foci of cellular alteration were observed at the lowest level (1.7 mg/kg bw per day). Also, centrilobular degeneration, necrosis and mitochondrial damage were noted in hepatic parenchyma.

Til et al. (1983) did a follow-up study, under the same conditions, except at lower doses and with 1% PVC instead of the 10% used by Feron et al. (1981); the results are described in Til et al. (1991). Oral vinyl chloride doses of 0.017, 0.17 and 1.7 mg/kg bw per day (corresponding to 0.014, 0.13 and 1.3 mg/kg bw per day when only absorbed vinyl chloride is considered) were administered to groups of 100 males and 100 females for 149 weeks except for the high-dose group, which consisted of 50 males and 50 females. An increased incidence of liver nodules was the only neoplastic response in rats administered 0.17 mg/kg bw per day, but both HCC (three per sex) and ASL (1/49 males, 2/49 females) were observed at the highest dose (1.7 mg/kg bw per day), with only the incidence of HCC in males being significantly different from that of the controls (Til et al., 1983, 1991). No ASL was observed at the 1.7 mg/kg bw per day level in the original study (Feron et al., 1981). Non-neoplastic results in the liver, such as increased cellular alteration, polymorphism and cyst formation, were observed in the highest exposure group for males (1.7 mg/kg bw per day). However, basophilic foci were observed in the lowest exposure groups (0.017 and 0.17 mg/kg bw per day) for females (Til et al., 1983, 1991); these are not considered as a precursor of hepatocellular tumour formation, as they arise at a different location (U.S. EPA, 2000b). The no-observed-adverse-effect level (NOAEL) for tumour induction in rats was estimated by the authors to be 0.13 mg/kg bw per day (Til et al., 1983, 1991).

The only carcinogenic bioassay in which vinyl chloride was administered to rats in drinking water is the unpublished work of Evans et al. (ECETOC, 1988). Groups of male and female (150 per group) Wistar rats received vinyl chloride in drinking water at concentrations of 0, 2.5, 25 or 250 mg/L (equivalent to daily intakes of approximately 0, 0.12, 1.2 or 12 mg/kg bw per day for males and 0, 0.22, 2.2 or 22 mg/kg bw per day for females) for up to 152 weeks, except for males and females of the highest dose group, which received the dose for 115 and 101 weeks, respectively. In rats receiving the highest dose, there was a significantly higher incidence of ASL (8/50 males, 8/49 females) and hepatomas (HCC) (3/50 males, 3/49 females). Only 1 of the 47 males developed hepatic angiosarcomas in the 25 mg/L dose group. There was no dose-response relationship in the development of kidney and brain tumours in these animals.

In a reproductive and developmental toxicity study on Sprague-Dawley rats (Thornton et al., 2002), an increase in liver weight was observed at 10, 100 and 1100 ppm in all F0 rats and at 100 and 1100 ppm in F1 male rats exposed by inhalation under the conditions described in Section 9.2.5. In maternal rats exposed to 100 ppm, the kidney relative to body weight ratio was statistically significantly increased, while in the 1100 ppm exposed group, the organ relative to body weight ratios for both kidney and liver were statistically significantly increased.

Female CD-1 and B6C3F1 mice, Fischer-344 rats and Golden Syrian hamsters were exposed to fixed concentrations of vinyl chloride (50, 100 and 200 ppm, respectively) for 6 hours daily, 5 days/week, for 6, 12, 18 and 24 months (Drew et al., 1983). When exposed after 8 months of age, all animals showed a decrease in their length of survival. Every species developed at least three types of cancer, including ASL, mammary gland adenocarcinomas, HCC, lung carcinomas and stomach adenomas. Vinyl chloride induced ASL in all three species (Maltoni et al., 1984).

Different authors have reported different cancer type incidences among mammalian species. Drew et al. (1983) found a 4- to 5-fold higher incidence of cancer than did Maltoni et al. (1981). Nevertheless, this could be explained by younger age of exposure and longer exposure period in the study by Drew et al. (1983). Accordingly, Bi et al. (1985) and Lee et al. (1978) found an increased occurrence of lung sarcoma in Wistar rats exposed to 100 and 3000 ppm vinyl chloride in air (6 hours daily, 6 days/week, for 18 months), although at a lower incidence than ASL; lung sarcoma was not found in mice, however, mammary gland carcinoma, adenocarcinoma or carcinoma has been reported in mice but not in rats (Lee et al., 1978; Drew et al., 1983).Animals exposed at a young age developed more cancers (Lee et al., 1978; Drew et al., 1983; Maltoni et al., 1984; Bi et al., 1985). Swiss mice appear to be more sensitive than B6C3F1 mice, as they reportedly develop lung carcinomas after 6 months of exposure, suggesting an interspecies variability. Both strains developed ASL and mammary gland carcinomas (Maltoni et al., 1984). In hamsters, the highest incidence of cancer was reported for those exposed for their first 12 months of age compared with those exposed after 12 months of age (Drew et al., 1983). Rats exposed after 12 months of age failed to develop vinyl chloride-related neoplasms, whereas a shorter latency of 6 months resulted in a significant increase in mammary gland adenocarcinomas, HCC and ASL. Rats were the only species that developed Zymbal gland carcinomas, nephroblastomas or neuroblastomas, and HCC; however, the HCC response did not seem to correlate with the dose. Hamsters did not develop mammary carcinomas, lung adenomas or hepatomas (Maltoni et al., 1984; Whysner et al., 1996); however, they were the only species to develop acoustic duct epithelial tumours (Maltoni et al., 1984). Apart from the latter result, mice and hamsters are not known to develop any additional tumours not reported in rats (Maltoni et al., 1984). Increasing the exposure for more than a year in rats or 6 months in mice and hamsters generally did not significantly increase the development of cancer if the exposure began early in life (Drew et al., 1983).

9.2.4 Mutagenicity/Genotoxicity

DNA etheno adducts, naturally occurring in bacteria and animals, can be formed and detected in mammals (mouse, rat, monkey, human, hamster) in vivo and in bacteria and yeast in vitro if vinyl chloride is pre-incubated with a mammalian microsomal system (Bartsch et al., 1975; Laib and Bolt, 1977; Swenberg et al., 1992, 2000; Dogliotti, 2006).

9.2.4.1 In Vitro

The Ames test indicated that vinyl chloride and its metabolites are mutagenic (Rannug et al., 1974; Bartsch et al., 1975). Mutations increase with dose and with increasing time of exposure (Malaveille et al., 1975; Bartsch et al., 1976, 1979; IARC, 1979; ECETOC, 1988). The addition of liver human and rat S9 activation system fractions (Rannug et al., 1974; Bartsch et al., 1975; Bartsch et al., 1975; Malaveille et al., 1975; Sabadie et al., 1980), either pre-incubated or not with P450 inducers (phenobarbital, 3-methylcholanthrene or Aroclor 1254), and a reduced nicotinamide adenine dinucleotide phosphate (NADPH) generating system stimulated the mutagenicity of vinyl chloride in Salmonella typhimurium strains TA1530, TA1535 and G-46 (Malaveille et al., 1975; Loprieno et al., 1976.

Rat hepatic microsomes co-exposed to CEO directly or generated after incubation with vinyl chloride and polyadenosine or polycytidylic acid (artificial cytidine) reportedly form ribonucleic acid (RNA) adducts (Laib and Bolt, 1977, 1978; Guengerich and Watanabe, 1979). CEO and CAA produce adducts that induce transversions (A↔T) but mainly transitions (GC→AT) in S. typhimurium and in key mammalian genes, such as p53 and k-ras (Matsuda et al., 1995; Barbin, 2000; Gros et al., 2003). CEO alone was found to be much more potent than CAA in inducing reverse mutations in S. typhimurium (Malaveille et al., 1975; Loprieno et al., 1976; Bartsch et al., 1979) and in Escherichia coli strains A3, A11, A23, A58, A88, A446 and A46 (Barbin et al., 1985; Perrard, 1985) at concentrations as low as 0.1 mM for 1 hour. The addition of epoxide hydrolase and glutathione S-transferase in vitro, which eliminates CEO, has been shown to inhibit the binding of CAA and CEO to DNA (Guengerich et al., 1979; Guengerich and Shimada, 1991).

Mutations in the yeast Saccharomyces pombe increased with time of exposure (5-60 minutes) to vinyl chloride at 16-48 mM concentrations only in the presence of an exogenous activation system from mouse liver microsomes (Loprieno et al., 1976).

9.2.4.2 In Vivo

RNA adducts with adenosine (Laib and Bolt, 1977; Guengerich et al., 1979) and cytidine (Laib and Bolt, 1978) were found in rats exposed by inhalation to vinyl chloride at 50 ppm for 5 hours. The RNA products were reported to decrease with time after exposure (Bolt et al., 1980). The more common adduct, albeit not mutagenic, is 7-(2-oxoethyl)-guanosine, with 98% occurrence (Bolt et al., 1986) in rats exposed to vinyl chloride (Swenberg et al., 1992). It is followed in lower proportion by the persistent miscoding inducers N2,3-ethenoguanosine and 3,N4-ethenocytosine, both representing ~1% each, and 1,N6-ethenoadenosine, which rarely occurs at less than 1% (Fedtke et al., 1990; Basu et al., 1993). When 10-day-old neonatal rats and their mothers were exposed for 4 days to 600 ppm vinyl chloride, high 7-(2-oxoethyl)-guanosine and N2,3-ethenoguanosine levels were revealed by GC and high-performance liquid chromatography, respectively (Fedtke et al., 1990). The N2,3-ethenoguanosine accumulation in liver of rats was linear between 10 and 1100 ppm (Morinello et al., 2002a) and increased with exposure time from 2 to 8 weeks when rats were exposed to 500 ppm (Guichard et al., 1996). No difference was noted between the amount of N2,3-ethenoguanosine in hepatocytes and non-parenchymal cells in the liver, but weanling rats had 2-3 times the amount of adducts compared with adults in the same conditions (Morinello et al., 2002a). Moreover, 3,N4-ethenocytosine and 1,N6-ethenoadenosine were found in livers of rats exposed for 2 years to 250 mg/L vinyl chloride in their drinking water, but no data were given in relation to non-exposed rats (Green and Hathway, 1978).

In rats, the AT→TA mutation was found at codon 61 of the ha-ras gene after exposure by inhalation to 500 ppm of vinyl chloride for 8 hours/day for 33 days, leading to liver tumours (Marion and Boivin-Angele, 1999; Boivin-Angele et al., 2000), and not at codon 13, as found in humans with ASL. Indeed, ha-ras and ki-ras are involved in rat hepatocyte and human sinusoidal cell growth activation, respectively. Moreover, none of the codons 12, 13 or 61 that are associated with human vinyl chloride carcinogenicity were mutated in rat ASL (Marion and Boivin-Angele, 1999; Barbin, 2000; Boivin-Angele et al., 2000). AT base pair transversion caused by 1,N6-ethenoadenosine in the p53 gene was associated with rat and human ASL after exposure to vinyl chloride (Boivin-Angele et al., 2000), revealing consistency between species. No difference was observed between the amount of 3,N4-ethenocytosine in rats exposed for 4 weeks and allowed to recover for 1 week compared with the ones killed immediately (Morinello et al., 2002a). Alderley Park rats exposed by inhalation to 1500 ppm vinyl chloride for 6 hours daily for 5 days had a bone marrow cell chromatid gap frequency similar to that of rats exposed to the same regime for 3 months (Anderson and Richardson, 1981). However, only two rats per dose group were used.

Drosophila exposed by inhalation to vinyl chloride at 30-10 000 ppm for 3-17 days had increased mature sperm recessive lethal mutations, but no chromosomal aberrations were observed (Verburgt and Vogel, 1977). However, no dominant lethal effects were observed up to gestational days 12-14 after exposure of male CD-1 mice to vinyl chloride at 10 000 ppm for 4 hours daily for 5 days or at 5000 ppm for 4 hours daily, 5 days/week, for 10 weeks (Himeno et al., 1983). Female NMR mice and female Chinese hamsters exposed by inhalation to vinyl chloride at 500 ppm for 5 days and at 12 500 ppm for 6 hours, respectively, had increased single strand breaks (Basler and Rohrborn, 1980; Walles and Holmberg, 1984).

9.2.5 Reproductive and Developmental Toxicity

No significant embryotoxic, fetotoxic or teratogenic effects were observed in groups of pregnant CF-1 mice, Sprague-Dawley rats or New Zealand White rabbits exposed 7 hours/day via inhalation to vinyl chloride at doses of 50 or 500 ppm on days 6-15 of gestation (mice) or at doses of 500 or 2500 ppm on days 6-15 (rats) or days 6-18 (rabbits) of gestation (John et al., 1981). Maternal toxicity, such as higher death rate, decreased feed consumption and increased liver weight, was observed in mice and rats exposed to 500 ppm and 2500 ppm, respectively. Fetal effects consisted of delayed skeletal development in mice and an increase in the incidence of ureter dilation in rats after exposure to a maternally toxic dose (John et al., 1981).

The dose-response changes in the reproductive system of rats exposed to vinyl chloride were evaluated by Biet al. (1985). Wistar rats were exposed chronically for 6 hours daily, 6 days/week, for 3, 6, 12 or 18 months to air concentrations of 0, 10, 100 and 1000 ppm. The authors observed a significant dose-response decrease in testis weight at 100 and 1000 ppm, accompanied by perturbations in the spermatogenesis process, including fusion of spermatids and swelling and necrosis of seminiferous tubule cells, with the severity of these effects directly related to the vinyl chloride concentration. According to Heywood and James (1978), approximately 20% of rats fed under normal conditions developed testicular atrophy. Sokal et al. (1980) exposed rats to vinyl chloride at 50, 500 and 20 000 ppm for 10 months. Degeneration of the seminiferous tubule cells and perturbations in spermatogenesis were reported; however, these effects did not increase in a dose-responsive fashion at higher concentrations. The fertility of the males, the capacity to generate offspring and the reversible potency of vinyl chloride were not examined. Based on these two studies, further research is needed to determine a clear relationship between vinyl chloride and reproductive effects.

A more recent vinyl chloride inhalation investigation on embryo-fetal development and reproduction in rats was carried out by Thornton et al. (2002). Four groups of F0 rats were subchronically exposed to vinyl chloride. Males and females (120 of each sex) were exposed for a 10-week pre-mating period and a 3-week mating period, followed by continued exposure for males and 20 days of exposure for gestational females to 0, 10, 100 or 1100 ppm for 6 hours/day. Thirty pups of each sex were selected from each of the four groups. The F1 generation was also exposed to the same conditions. F0 and F1 animals were examined twice daily for toxic effects. Physical evaluation of F1 and F2 pups was done on lactation days 0, 4, 7, 14, 21 or 25. Macroscopic and microscopic examinations of the reproductive organs were executed at weaning for 15 male and 15 female F1 and F2 pups. Then, sperm quality was analysed for F0 and F1 animals of each of the four groups. Supporting John et al. (1981), this study revealed no alteration of embryo-fetal development, teratogenicity or reproduction in the four groups of vinyl chloride-exposed rats. Body weight, pregnancy rates, postmortem observations, uterine status, feed consumption and duration of gestation were all unaffected during the pre-mating, mating and post-mating periods. The reproductive organs did not show any signs of adverse effects, including sperm motility, number and morphology (Thornton et al., 2002). In the F1 pups, the live birth index was lower for the 1100 ppm group (0.98), whereas the viability was lower for the 10 and 100 ppm groups (0.89 and 0.88, respectively). Still, the authors stated that historic data supported these results. For the F2 litter, the authors found a reduced pup number in the 1100 ppm group compared with the F2 controls; however, an increased viability was reported for the F2 litter exposed to 10, 100 and 1100 ppm. No dose-response relationship for birth rate was seen after vinyl chloride exposure (Thornton et al., 2002). The authors determined a no-observed-adverse-effect concentration (NOAEC) of 1100 ppm for embryo-fetal development and reproduction in rats.

9.2.6 Juvenile Sensitivity

Rodent studies have shown that animals less than 4 weeks old appear to be more susceptible to the formation and persistence of DNA adducts induced by vinyl chloride (Grosseet al., 2007). Morinello et al. (2002a) found that concentrations of N2,3-ethenoguanine were 2- to 3-fold higher in weanlings than in adults exposed for the same amount of time. The available data indicate that the cancer risk from a relatively short-term exposure in juveniles is approximately equal to the risk from long-term exposure in adults; several other genotoxic carcinogens have shown similar results (Ginsberg, 2003).

Laib et al. (1989) , using radiolabelled vinyl chloride, found an 8-fold increase in the amount of DNA alkylation in the liver of young rats compared with adult rats; this increase was attributed to increased levels of DNA synthesis per cell in younger rats.

Maltoni and Cotti (1988) examined the effect of age at the start of exposure on vinyl chloride toxicity by exposing rat breeders (13 weeks old at the start of the experiment) and offspring (12 days old at the start of the experiment) to vinyl chloride for 104 weeks. Another group of rat offspring was exposed for 15 weeks; all rats were observed until their death. The incidence of hepatocarcinomas was highest in the two groups of offspring; the incidence of ASL was highest in the offspring exposed for 104 weeks. Neuroblastomas, by contrast, occurred with the highest frequency in rats exposed for 104 weeks, regardless of starting age. The authors concluded that the incidence of hepatocarcinomas was influenced by the age at the start of exposure, ASL was influenced by both age at the start of exposure and duration of exposure and neuroblastomas were influenced only by the duration of exposure.

Cogliano et al. (1996) reviewed data on early-life exposures to vinyl chloride. They concluded that the cancer risk from a brief early-life exposure (weeks 1-5 for a rat) is approximately equal to that from chronic exposure to the same concentration in air beginning at maturity, but that the specific effects from early-life and full-life exposures may be different.

The U.S. EPA (2000b) reviewed data on early-life and adult cancer potencies based on studies with repeated exposures of juvenile (from birth to 5 weeks of age) and adult animals (from 13 to 52 weeks after birth) to genotoxic carcinogens, including vinyl chloride. Based primarily on the Maltoni and Cotti (1988) study, the median ratios of juvenile to adult cancer potencies for ASL from vinyl chloride exposure ranged from 9.8 to 29 (geometric mean = 14); the ratios for hepatoma ranged from 32 to 58 (geometric mean = 47). For other cancers resulting from vinyl chloride exposure, there was no clear trend towards higher cancer risks in juveniles compared with adults.

9.3 Mode of Action

Primary liver cancer or, more specifically, liver angiosarcoma (ASL; a cancer of the liver blood vessels) is the most severe endpoint that follows oral (food or water) or inhalation exposure to vinyl chloride, based on consistencies between epidemiological and experimental animal studies (Albertini et al., 2003).

The first key event in vinyl chloride-induced toxicity is the formation of the oxidized metabolites CEO and CAA by hepatocyte CYP2E1 (Bolt, 1986; Barbin and Bartsch, 1989; Albertini et al., 2003). Metabolites migrate by a cell to cell passage mechanism from hepatocyte P450 sites to the sinusoidal cells (Morinello et al., 2002b), where ASL originates.

The second key event is the generation of DNA adducts (IPCS, 1999; Holt et al., 2000). Highly reactive CEO is the major metabolite involved in their formation. The mode of action is supported by the fact that the toxicokinetics of the metabolite production, dependent on the activity level of CYP2E1 and the conjugation with glutathione catalyzed by glutathione S-transferase, are correlated to the adduct formation (IPCS, 1999; U.S. EPA, 2000b). This second key event is characterized by the persistence of the DNA adducts. Sinusoidal cells express DNA repair enzymes at lower levels than do hepatocytes (metabolism site), which may explain their higher amount of DNA etheno adducts (Swenberg et al., 1999; Holt et al., 2000; U.S. EPA, 2000b; Morinello et al., 2002a). After 4 weeks of exposure of rats to vinyl chloride at 1100 ppm, recovery for 5 days significantly lowered the amount of N2,3-ethenoguanosine in rat hepatocytes, but not in non-parenchymal cells, compared with the animals sacrificed immediately after exposure (Morinello et al., 2002a). Differences in the time of repair were observed between the three common DNA adducts in rats following exposure to vinyl chloride at 600 ppm for 4 hours daily over 5 days; a longer repair time was observed with 3,N4-ethenocytosine and 1,N6-ethenoadenosine (Swenberg et al., 1992; Dosanjh et al., 1994;OEHHA, 2000; Dogliotti, 2006).

The third key event is the formation of mutations specific to vinyl chloride (Bolt, 2005). The etheno adducts, which are the more persistent, but also the least numerous (~2%), lead to base pair transition and transversion mutations in cell growth regulator genes (Swenberg et al., 2000). In fact, rare A→T transversions in the p53 gene have been found in three out of six ASL cases in vinyl chloride workers (Hollstein et al., 1994), and a linear trend in mutant p53 expression as a function of workers' exposure intensity was found (Marion and Boivin-Angele, 1999; Wong et al., 2002; Luo et al., 2003). This mutation is chemical specific, as it is not found in Thorotrast (a radioactive thorium dioxide solution)-induced ASL and is uncommon in sporadic ASL (Barbin, 2000). Vinyl chloride workers with ASL have prominent ki-ras activating GC→AT transition mutations at codon 13 (Marion et al., 1991; De Vivo et al., 1994; Marion and Boivin-Angele, 1999). Vinyl chloride also produces sister chromatid exchanges, clastogenicity and micronuclei in humans (U.S. EPA, 2000b).

The fourth key event is the effect of genotoxicity on loss of cell cycle regulation and tumour progression (IARC, 2008). Worker exposure intensities were associated with p53 and Asp13-p21-ki-ras mutant protein expression levels (De Vivo et al., 1994; Trivers et al., 1995; Weihrauch et al., 2000, 2001; Luo et al., 2003). The biological plausibility is supported by the association of the p53 and ras mutations with ASL and HCC development in vinyl chloride workers (Weihrauch et al., 2000). DNA methylation of p23INK4A, a gene that inhibits cell cycle progression, was also observed in a cohort of vinyl chloride workers with HCC (Weihrauch et al., 2001).

The vinyl chloride non-cancerous mode of action is not as clearly defined as the cancerous process (U.S. EPA, 2000b).

Page details

Date modified: