Vinyl Chloride Epidemiology and Liver Cancer

CAS No. 75-01-4 RoseyFields — Chemical-Specific Epidemiology Page

RoseyFields evaluates the human epidemiological evidence for vinyl chloride as follows, drawing on the peer-reviewed occupational, community, and molecular-epidemiological literature to characterize what is known — and not known — about health outcomes in humans exposed to vinyl chloride (VC).

1. Study Populations and Data Sources

The VC epidemiological literature is built from several distinct types of study populations, each with different strengths, and RoseyFields treats the convergence of findings across these different populations as a key source of confidence in causal inference.

1.1 Occupational Cohorts

The great majority of the evidence base derives from retrospective mortality cohort studies of workers in VC monomer (VCM) production and polyvinyl chloride (PVC) polymerization facilities, assembled independently across multiple countries and time periods: Great Britain (Fox and Collier, 1977; Jones et al., 1988), the United States (Tabershaw and Gaffey, 1974; Cooper, 1981; Monson et al., 1975; Waxweiler et al., 1976; Wu et al., 1989; Wong et al., 1991; Buffler et al., 1979; Ott et al., 1975), Germany and Austria (Weber et al., 1981), Sweden (Byren et al., 1976), Italy (Belli et al., 1987; Pirastu et al., 1990, 1998), the former Soviet Union (Smulevich et al., 1988), Taiwan (Du and Wang, 1998), and a large 12,706-worker multicentric pooled cohort spanning several European countries (Simonato et al., 1991). Because these cohorts were assembled independently, in different countries, industries, and time periods, by different research groups using different methods, a confounder or bias common to all of them (e.g., a shared unmeasured co-exposure, a shared analytic error) is far less plausible than it would be if the evidence rested on a single cohort — this cross-population replication is central to the strength of the causal inference for VC’s hepatic effects in particular.

1.2 Community and Contamination-Site Studies

A smaller but important body of evidence addresses non-occupational, environmental exposure. Early community studies examined birth defect incidence among populations residing near VC/PVC manufacturing plants (Edmonds et al., 1978; Theriault et al., 1983), and a national case-finding study assessed whether any angiosarcoma cases in Great Britain could be attributed to non-occupational, community-level exposure near VC sites (Elliot and Kleinschmidt, 1997). More recently, a series of studies of Marine Corps and Navy personnel, and civilian workers, exposed to VC together with trichloroethylene, tetrachloroethylene, and benzene in contaminated drinking water at Marine Corps Base Camp Lejeune, North Carolina, have examined mortality (Bove et al., 2014), cancer incidence (Bove et al., 2024), male breast cancer (Ruckart et al., 2015), and birth defects and childhood cancer following in-utero exposure (Ruckart et al., 2013). These contamination-site studies are valuable precisely because they capture a different exposure route (drinking-water ingestion and inhalation/dermal exposure during bathing) and a different, non-occupational population (including women, children, and in-utero exposure) than the occupational cohorts, extending the generalizability of VC hazard findings beyond the historically male, adult, industrial-worker evidence base — while also illustrating a recurring limitation, discussed in Section 4, that environmental VC exposure rarely occurs in isolation from co-contaminants.

1.3 Molecular Epidemiology and Biomarker Studies

A distinct set of studies examines biomarkers of genotoxic effect and early biological response in exposed workers, rather than clinical disease endpoints: sister chromatid exchange and micronuclei frequency (Sinues et al., 1991), chromosome aberrations (Purchase et al., 1978; Hansteen et al., 1978), and p53 tumor-suppressor gene mutation biomarkers (Smith et al., 1998; Luo et al., 1999; Chen et al., 1999; Trivers et al., 1995; Marion et al., 1996; DeVivo et al., 1994). These studies serve a different evidentiary function than the cancer/mortality cohorts: rather than establishing disease risk directly, they demonstrate a biologically plausible, dose-responsive genotoxic mechanism operating in exposed humans, which strengthens the causal interpretation of the cohort mortality findings by supplying the mechanistic link between exposure and the downstream cancer outcome.

1.4 A Notable Gap: General-Population Biomonitoring

Unlike some other high-production-volume industrial chemicals, VC has not been the subject of large-scale general-population biomonitoring surveillance comparable to programs such as NHANES, reflecting both VC’s rapid metabolism and elimination (which limits the utility of blood or urine VC measurement as a biomarker of past exposure; see RoseyFields’ companion toxicology summary) and the fact that general-population exposure levels are several orders of magnitude below occupational historical exposures. RoseyFields treats the absence of such data as a genuine evidentiary gap (Section 8 of RoseyFields’ companion toxicology summary) rather than as evidence of safety at low, general-population exposure levels, and relies instead on the convergence of occupational and contamination-site data, together with the toxicological dose-response relationships derived from PBPK-adjusted animal data, to characterize risk at lower exposure levels.

2. Health Outcomes Investigated in Humans, by Body System

The table below organizes every health outcome investigated in the VC human epidemiological literature by body system, and classifies the strength of evidence for each into one of three tiers: strong/consistent and quantitatively modeled; suggestive but inconsistent; or null/inadequate. Section 3 discusses the strong-evidence outcomes in further detail.

2.1 Hepatic / Hepatobiliary • STRONG — Hepatic angiosarcoma. Consistently and strongly elevated across essentially every occupational cohort with adequate follow-up and case ascertainment; the strongest and best-replicated finding in the entire VC literature (Section 3.1).

• STRONG — Hepatocellular carcinoma. Consistently elevated liver/biliary cancer mortality after excluding or accounting for angiosarcoma, in multiple independent cohorts (Section 3.1).

• STRONG — Noncancer hepatic disease (fibrosis, hepatocellular/sinusoidal hyperplasia, portal hypertension). Well-documented, dose- and duration-related in clinical case series across multiple countries (Section 3.2).

• SUGGESTIVE — Liver cirrhosis. Elevated in some cohorts (Du and Wang, 1998), plausible as a precursor lesion to hepatocellular carcinoma, but not universally assessed or replicated across cohorts.

• SUGGESTIVE — Nonspecific liver function test abnormalities at low exposure levels. Reported in some cohorts (Ho et al., 1991; Du et al., 1995) but confounded by alcohol use and lacking test specificity (Section 4).

2.2 Hematologic / Lymphatic

• SUGGESTIVE — Lymphoma / leukemia / lymphatic-hematopoietic cancers. Elevated in several cohorts (Weber et al., 1981; Simonato et al., 1991; Smulevich et al., 1988; Tabershaw and Gaffey, 1974) but with small case numbers, inconsistent statistical significance, and no clear exposure-duration gradient in most analyses.

2.3 Nervous System

• SUGGESTIVE — Brain and CNS cancer (predominantly glioblastoma multiforme). Elevated in multiple cohorts (Monson et al., 1975; Byren et al., 1976; Waxweiler et al., 1976; Weber et al., 1981; Cooper, 1981; Wong et al., 1991) but generally without a consistent duration- or dose-response gradient, and IARC’s independent review characterizes this evidence as weak (Section 5).

• STRONG (acute, high-exposure only) — Acute CNS depression (dizziness, headache, ataxia, loss of consciousness). Well documented at high, narcotic-range acute exposures (Lilis et al., 1975; Langauer-Lewowicka et al., 1983), but not established as a chronic, low-level exposure effect.

2.4 Respiratory

• NULL/CONFOUNDED — Lung cancer. Elevated in some early cohorts (Monson et al., 1975; Waxweiler et al., 1976; Buffler et al., 1979) but attributed on further analysis to co-exposure to PVC dust rather than VC itself (Waxweiler et al., 1981), and IARC’s independent review found no association between VC and lung cancer (Section 5).
• NULL — Pulmonary function decrements. No evidence of shift-related pulmonary function change in exposed workers.

2.5 Vascular / Dermal / Musculoskeletal

• STRONG (high occupational exposure) — Raynaud’s phenomenon and acroosteolysis. Well-documented in highly exposed workers, particularly reactor/tank cleaners (Lilis et al., 1975; Fontana et al., 1995), though these findings reflect a historical, high-dermal/high-inhalation exposure scenario largely eliminated by modern industrial hygiene practice.

2.6 Reproductive and Developmental

• NULL/INADEQUATE — Birth defects and fetal loss from parental occupational exposure. No solid association found across several studies (Hatch et al., 1981; Infante et al., 1976).

• NULL — Birth defects from community/residential proximity to VC plants. No clear association found in either community study conducted (Edmonds et al., 1978; Theriault et al., 1983).

• SUGGESTIVE (contamination-site, in-utero exposure) — Neural tube defects and childhood hematopoietic cancers. A modern case-control study of in-utero exposure to VC-contaminated drinking water found elevated but statistically imprecise odds ratios for childhood cancers with first-trimester VC exposure (Ruckart et al., 2013); this is a different population and exposure route than the occupational studies above, and the finding, while suggestive, has wide confidence intervals and has not yet been independently replicated.

2.7 Other Cancer Sites

• SUGGESTIVE — Soft tissue / connective tissue sarcoma. Elevated in updated analyses of large occupational cohorts, though with smaller relative risks than for liver cancer, and characterized as suggestive by IARC’s independent review (Section 5).

• SUGGESTIVE, unreplicated — Malignant melanoma. Reported in one cohort (Heldaas et al., 1984) but not substantiated in other studies, and characterized as weak evidence by IARC.

• SUGGESTIVE, contamination-site — Male breast cancer. Elevated in a case-control study of Camp Lejeune Marines with cumulative VC (and co-contaminant) exposure (Ruckart et al., 2015), a rare cancer where numbers are necessarily small; not previously assessable in the occupational cohorts, which included almost no female workers and few reported male breast cancer cases.

• NULL/INADEQUATE — Female breast cancer. No cases reported in the one cohort with substantial female representation (Smulevich et al., 1988); data are too sparse for any conclusion.

• NULL — Kidney/renal cancer (nephroblastoma). No human evidence of excess kidney cancer, despite this tumor type being observed in one animal bioassay.
• SUGGESTIVE, limited — Respiratory/laryngeal cancer at specific plants. Elevated in one Italian cohort (Belli et al., 1987; Pirastu et al., 1998), plausibly confounded by concurrent PVC dust or other workplace exposures at those specific facilities.

2.8 Genotoxicity / Molecular Biomarkers (Mechanistic Evidence)

• STRONG — Chromosomal damage (SCE, micronuclei, chromosome aberrations). Consistently elevated and exposure-correlated across independent worker cohorts (Sinues et al., 1991; Purchase et al., 1978; Hansteen et al., 1978), and shown to be at least partly reversible after exposure ends (Hansteen et al., 1978).

• STRONG — p53 tumor-suppressor mutation biomarkers. A clear, statistically significant, monotonic dose-response relationship between cumulative VC exposure and mutant p53 biomarker prevalence has been independently replicated in French (Smith et al., 1998) and Taiwanese (Luo et al., 1999) worker cohorts, with the specific mutation pattern also distinguishing VC-associated from non-VC-associated angiosarcoma (Chen et al., 1999; Soini et al., 1995).

3. Strong-Evidence Outcomes in Detail

3.1 Hepatic Angiosarcoma and Hepatocellular Carcinoma

Hepatic angiosarcoma is the signature VC-associated malignancy and the outcome with by far the strongest, most consistent, and most extensively quantitatively modeled evidence in the entire literature. Angiosarcoma is an extremely rare tumor in the general population (historically about 20–30 cases per year in the entire United States, and roughly 1.4 cases per 10 million population per year in Great Britain (Elliot and Kleinschmidt, 1997)), which gives even small numbers of cases exceptional epidemiological significance: a rare-disease signal is far less likely to arise from confounding or chance than a signal in a common disease. The first published association — three cases in PVC-plant workers — was reported by Creech and Johnson (1974). Since then, statistically significant excess liver cancer mortality, with angiosarcoma histologically confirmed as the predominant or a major contributing tumor type, has been independently reported in essentially every cohort with adequate case ascertainment: British workers (4 cases, including 2 angiosarcomas, in an early cohort of 7,717 (Fox and Collier, 1977); 11 liver deaths, 7 angiosarcomas, in a follow-up cohort of 5,498 (Jones et al., 1988)); American workers (Monson et al., 1975; Tabershaw and Gaffey, 1974; Cooper, 1981; Wu et al., 1989; Wong et al., 1991, the last identifying 15 angiosarcoma deaths in a large industry-wide cohort); Swedish workers (Byren et al., 1976); German/Austrian workers (12 observed vs. 0.79 expected liver cancer deaths (Weber et al., 1981)); Italian workers across three separate plants (Pirastu et al., 1990, 1998); and Taiwanese workers (elevated morbidity odds ratios of 4.5–6.5 for primary liver cancer hospitalization (Du and Wang, 1998)). The largest and most statistically robust single study, a pooled multicentric European cohort of 12,706 VC/PVC workers, found 24 liver cancer deaths against a strongly elevated standardized mortality ratio, with histopathologic confirmation of angiosarcoma in 16 of 17 cases with available tissue, and a clear positive relationship between excess risk and time since first exposure, exposure duration, and estimated cumulative exposure (Simonato et al., 1991).

Exposure metrics across these studies have relied primarily on job-category/maximum-exposure classification schemes rather than direct biomonitoring, since VC’s rapid metabolism precludes reliable biomarker-based reconstruction of historical cumulative exposure (Section 4.2); reconstructed average exposure levels for the classic British and European cohorts have been estimated in the range of roughly 10–300+ ppm depending on job category and plant (Fox and Collier, 1977; Simonato et al., 1991). Clinical case series of confirmed angiosarcoma patients corroborate the cohort findings, documenting latency periods on the order of 9–35 years from first exposure to tumor diagnosis, with younger age at first exposure associated with shorter latency in at least one series (Lelbach, 1996; Lee et al., 1996). Hepatocellular carcinoma, a distinct but related tumor type, is also consistently elevated in the same cohorts once angiosarcoma is accounted for or excluded (Wong et al., 1991; Du and Wang, 1998; Lelbach, 1996), consistent with the toxicological evidence (RoseyFields’ companion toxicology summary) that VC’s liver-carcinogenic mode of action operates through at least two overlapping but molecularly distinguishable pathways.

3.2 Noncancer Hepatic Disease

Independent of the cancer findings, VC-exposed workers show a well-characterized spectrum of noncancer hepatic pathology: subcapsular, portal, and perisinusoidal fibrosis; hepatocellular and sinusoidal-cell hyperplasia; and portal hypertension, documented in clinical case series across multiple countries (Gedigk et al., 1975; Lilis et al., 1975; Marsteller et al., 1975; Popper and Thomas, 1975; Tamburro et al., 1984; Buchancova et al., 1985). Focal hepatocellular and mixed hyperplasia are identified as the earliest anatomic lesions (Popper and Thomas, 1975), and severity of degenerative change correlates with cumulative exposure duration, with at least partial reversibility documented after exposure ceases (Gedigk et al., 1975; Lilis et al., 1975). Because this evidence rests on clinical case series and cross-sectional worker examinations rather than on formally modeled cohort dose-response analyses, RoseyFields treats it as strong but qualitatively (rather than quantitatively) characterized — adequate to establish that noncancer hepatic injury is a genuine VC effect, but not adequate to derive a human-epidemiology-based noncancer dose-response value independent of the animal-based reference dose/concentration described in RoseyFields’ companion risk-assessment methodology.

3.3 Genotoxic Biomarkers

The molecular epidemiology literature provides some of the most quantitatively precise dose-response data in the VC evidence base, notwithstanding that these are biomarker rather than disease-endpoint studies. A study of French VC workers found adjusted odds ratios for mutant p53 biomarker positivity that rose in a clear, monotonic, and statistically significant fashion with estimated cumulative exposure (ppm-years) — from an odds ratio of roughly 4 at the lowest exposure category to roughly 13 at the highest — with non-overlapping confidence intervals across categories (Smith et al., 1998); a similar dose-response relationship was independently replicated in Taiwanese workers, where roughly 13% tested positive for p53 biomarkers overall (Luo et al., 1999). Chromosomal-damage biomarkers (sister chromatid exchange, micronuclei, chromosome aberrations) likewise correlate with cumulative or current exposure level across independent worker cohorts (Sinues et al., 1991; Purchase et al., 1978), with at least one study demonstrating that elevated chromosome aberrations present during active exposure resolved after 2–2.5 years without further exposure (Hansteen et al., 1978), directly supporting a genotoxic mechanism that is exposure-dependent rather than a fixed, pre-existing susceptibility marker.

4. Study Quality Considerations

RoseyFields evaluates every epidemiological finding discussed above against the following study-quality considerations, which materially affect the weight properly given to each outcome.

4.1 Confounding

• Smoking. Not consistently controlled for in the older occupational cohorts; this is a plausible confounder for the lung cancer signal, and the demonstration that VC-cohort lung cancers were not enriched for smoking-associated histological subtypes (Monson et al., 1975) together with the isolation of the lung cancer signal to PVC-dust-exposed (rather than VC-exposed) subgroups (Waxweiler et al., 1981) substantially weakens — rather than strengthens — a causal VC-lung cancer interpretation.

• Alcohol use. A plausible and specifically identified confounder for the low-level liver-function-abnormality findings, where 8 of 12 affected workers in one study were current or former drinkers (Ho et al., 1991), substantially limiting the weight RoseyFields places on that specific finding.

• Hepatitis B infection. A significant co-morbidity in the Taiwanese cohort, where 10 of 11 liver cancer cases with available medical information were hepatitis B carriers (Du and Wang, 1998); because hepatitis B is itself a potent independent liver carcinogen, studies that do not separately analyze or adjust for viral hepatitis status in populations with high endemic hepatitis B prevalence require more cautious interpretation, though the angiosarcoma-specific findings (a tumor type not associated with viral hepatitis) are not subject to this confounder.

• Co-exposure to other workplace chemicals. Several cohorts involved concurrent exposure to other known or suspected carcinogens (e.g., the Cooper (1981)/Tabershaw and Gaffey (1974) cohort noted concurrent exposure to at least 19 other potentially carcinogenic compounds), complicating attribution of non-liver findings (brain, lung, lymphatic/hematopoietic cancers) specifically to VC, though the liver/angiosarcoma signal is not similarly confounded because angiosarcoma has no other well-established occupational or chemical cause besides VC and, to a lesser extent, arsenic and thorotrast.

• Co-exposure to other solvents in contamination-site studies. The Camp Lejeune studies (Ruckart et al., 2013, 2015; Bove et al., 2014, 2024) involve concurrent drinking-water contamination with trichloroethylene, tetrachloroethylene, and benzene, making it difficult to attribute any specific finding uniquely to VC rather than to the co-contaminant mixture; RoseyFields treats these studies as informative about the mixture’s hazard and about VC’s plausible contribution, but not as freestanding, VC-specific dose-response evidence.

4.2 Exposure Misclassification

Nearly all occupational cohort exposure estimates rely on job-category or “maximum exposure ever experienced” classification schemes rather than direct measurement, because contemporaneous industrial hygiene monitoring was sparse or absent for most of the relevant historical exposure period. This tends to bias risk estimates conservatively (toward the null) when workers are classified by maximum rather than cumulative exposure (Fox and Collier, 1977), but can also introduce non-differential misclassification that attenuates true dose-response gradients, and, in pooled multi-plant analyses, differential misclassification across facilities with heterogeneous industrial hygiene practices (Simonato et al., 1991). RoseyFields treats reported point-estimate potencies from any single cohort as approximate, and instead relies on the consistency of direction and approximate magnitude of effect across the full set of independently reconstructed cohorts.

4.3 Reverse Causation

Reverse causation is not a plausible explanation for the cancer mortality findings, given the long, well-documented latency between first exposure and tumor diagnosis (commonly two to three decades) and the biological implausibility of incipient angiosarcoma causing increased VC exposure. It is a more legitimate concern for some of the low-level noncancer liver-function-test findings, where subclinical liver dysfunction from an unrelated cause could in principle affect a worker’s job assignment or duration of employment in ways that correlate with reported exposure category; RoseyFields accordingly gives less evidentiary weight to these specific findings than to the cancer mortality data.

4.4 Small Numbers and Statistical Fragility

Because angiosarcoma is such a rare tumor, several of the individual cohort studies rest on very small absolute case counts (as few as 2–4 cases in some early cohorts), meaning that individual point estimates and confidence intervals are correspondingly unstable even where relative risks are very large. RoseyFields addresses this by emphasizing the pattern of replication across independently assembled cohorts (Section 1.1) rather than relying on any single cohort’s point estimate, consistent with standard epidemiological practice for rare-outcome causal inference.

4.5 Healthy-Worker Effect and Selection

Like most occupational cohort mortality studies, the VC literature is potentially subject to a healthy-worker effect (currently employed workers tend to be healthier than the general population, which can bias overall mortality comparisons toward the null for non-target outcomes) and to selection related to which workers remained employed long enough to be captured in higher-cumulative-exposure categories. Because the angiosarcoma/liver cancer signal is so large in magnitude relative to any plausible healthy-worker bias, and because that bias would if anything work against detecting an effect, RoseyFields does not consider it a material threat to the hepatic cancer findings, though it may contribute to under-detection of milder, non-fatal noncancer outcomes.

5. Independent Expert Panel and Agency Weight-of-Evidence Conclusions

Two independent bodies have separately reviewed this literature and reached weight-of-evidence classifications, providing an important cross-check on RoseyFields’ own evaluation.

5.1 U.S. EPA

Under EPA’s carcinogen risk assessment guidelines, VC has been classified as a Group A “known human carcinogen,” the highest weight-of-evidence category, based on the epidemiological evidence discussed above together with corroborating animal bioassay data (discussed in RoseyFields’ companion toxicology summary). EPA further concluded that VC is a known human carcinogen by both the inhalation and oral routes and highly likely to be carcinogenic by the dermal route.

5.2 IARC

The International Agency for Research on Cancer (IARC) independently classifies VC as a Group 1 agent, “carcinogenic to humans” — its highest classification, reserved for agents with sufficient evidence of human carcinogenicity. IARC’s most recent detailed reassessment concluded that VC causes both hepatic angiosarcoma and hepatocellular carcinoma, and, together with evidence that VC increases the risk of liver cirrhosis (itself a risk factor for hepatocellular carcinoma), found this evidence sufficient for the Group 1 classification (IARC, 2008). Notably, IARC’s independent review reached specific conclusions on several of the “suggestive” outcomes identified in Section 2 that RoseyFields treats as authoritative cross-checks on its own evidence grading: IARC found suggestive evidence of increased risk of soft- and connective-tissue malignancy; found no association between VC and lung cancer; and characterized the evidence for increased risk of brain cancer, lymphatic/hematopoietic cancers, and melanoma as weak (IARC, 2008). This independent classification is materially consistent with RoseyFields’ own tiered evidence assessment in Section 2, and RoseyFields treats the concordance between the two independent reviews as a further source of confidence in the overall weight-of-evidence conclusion for hepatic angiosarcoma and hepatocellular carcinoma specifically.

6. Use of Epidemiological Evidence to Support Causation Opinions in

Litigation

6.1 General Causation

For general causation — whether VC exposure is capable of causing a given disease in humans at all — RoseyFields relies primarily on the strong, consistently replicated, cross-national occupational cohort evidence for hepatic angiosarcoma and hepatocellular carcinoma (Section 3.1), corroborated by the independent EPA and IARC weight-of-evidence classifications (Section 5) and by the toxicological mode-of-action evidence in RoseyFields’ companion toxicology summary showing a shared, mechanistically plausible pathway (genotoxic metabolite → DNA adduct formation → p53/ras mutation → tumor) operating in both exposed animals and exposed humans (Chen et al., 1999; Marion et al., 1996; Smith et al., 1998). This triangulation across independent human cohorts, independent expert bodies, and independent (animal and human biomarker) mechanistic evidence is the strongest form of general causation support available for any chemical exposure, and RoseyFields’ reports rely on exactly this structure of evidence — not any single study — to support a general causation opinion for VC-induced hepatic angiosarcoma and hepatocellular carcinoma. For outcomes in the suggestive or null tiers (Section 2), RoseyFields does not offer an affirmative general causation opinion, and states plainly where the evidence is inconsistent, weak, or confounded (e.g., lung cancer, birth defects from parental occupational exposure), consistent with the independent IARC and EPA conclusions on those specific outcomes.

6.2 Specific Causation

For specific causation — whether a given claimant’s disease was more likely than not caused by their particular VC exposure — RoseyFields integrates the general-causation evidence base with case-specific exposure reconstruction (drawing on the dose-reconstruction methodology in RoseyFields’ companion risk-assessment methodology document) and with differential etiology: for hepatic angiosarcoma specifically, the tumor’s extreme rarity in the unexposed population and the absence of other well-established occupational or environmental causes (with narrow exceptions such as arsenic and historical thorotrast exposure) makes VC exposure history itself highly probative of specific causation once general causation and adequate exposure are established — a very different specific-causation posture than for a common tumor type with many competing causes. For hepatocellular carcinoma and noncancer hepatic disease, RoseyFields’ specific-causation analysis additionally addresses competing or contributing causes such as viral hepatitis status, alcohol use, and other hepatotoxic exposures (Section 4.1), consistent with the confounding considerations identified in the general-causation literature itself, and quantifies claimant-specific exposure and latency against the ranges established in the key cohort and case-series studies (Section 3.1) to assess biological plausibility of the claimed exposure-disease relationship.

6.3 Application to Contamination-Site Litigation

For matters involving environmental or drinking-water VC contamination rather than occupational exposure, RoseyFields draws specifically on the Camp Lejeune contamination-site literature (Section 1.2) as the most directly analogous non-occupational human evidence available, while explicitly noting the co-contaminant confounding inherent in that evidence (Section 4.1) and supplementing it with the PBPK-based, route-adjusted dose-response values derived from the occupational and animal evidence (RoseyFields’ companion documents) to support quantitative risk and causation opinions specific to the exposure route and population at issue in a given matter.

References