Carbon Tetrachloride Toxicology and Liver Toxicity

RoseyFields evaluates the toxicology of carbon tetrachloride using the following evidence base, drawn from the peer-reviewed occupational, environmental, and experimental animal toxicology literature.

A note on sourcing: This summary cites the underlying peer-reviewed studies rather than the regulatory compilation (EPA IRIS Toxicological Review) in which many of these studies were originally identified and reviewed. Study details (species, dose, findings) are drawn from that publicly available compilation of the primary literature and have not been independently re-verified against the original journal articles by RoseyFields.

1. Chemical Identity and Physicochemical Properties Relevant to Toxicity

Carbon tetrachloride (CCl4; CAS No. 56-23-5) is a colorless, dense, nonflammable liquid halogenated aliphatic hydrocarbon, historically used as a solvent, dry-cleaning agent, fire extinguishant, grain fumigant, and chemical intermediate (notably in the production of chlorofluorocarbons) before its toxicity led to phase-out of most uses. Structurally it consists of a single carbon atom bonded to four chlorine atoms, a fully halogenated methane with no hydrogen available for the metabolic detoxification pathways (e.g., glutathione conjugation directly at carbon) available to less-substituted chlorinated solvents.

Carbon tetrachloride is volatile and highly lipophilic, properties that jointly determine much of its toxicological behavior: it is efficiently absorbed by inhalation, oral, and dermal routes, and once absorbed it partitions preferentially into fat and lipid-rich tissue. It is chemically unreactive under ambient conditions in its parent form — it does not react directly with cellular macromolecules — and its toxicity instead depends entirely on metabolic bioactivation to reactive free-radical intermediates (Section 3, Section 4). Because the parent compound itself is not directly reactive, carbon tetrachloride is not persistent or bioaccumulative in the environmental sense that metals or halogenated aromatic pollutants are; it does not accumulate irreversibly in tissue but is instead rapidly redistributed, metabolized, and eliminated, primarily via exhalation, following any single exposure (Section 3).

The compound’s high blood:air and tissue:blood partition coefficients (blood:air partition coefficients of 2.73–4.20 in human blood and 4.52 in rat blood, in vitro; Fisher et al., 1997; Gargas et al., 1989, 1986) favor rapid pulmonary uptake, while its affinity for fat (distribution volumes and tissue:blood ratios far higher in fat than in any other tissue; Sanzgiri et al., 1997; Dambrauskas and Cornish, 1970) create a slowly-clearing lipid reservoir that prolongs systemic exposure after nominal cessation of contact. This reservoir effect, rather than any covalent binding of the parent molecule, is the closest analog to “persistence” relevant to carbon tetrachloride toxicity, and it has direct toxicological consequences: daily occupational exposure schedules that place more of the absorbed dose into the poorly perfused fat compartment produce slower overall elimination and a different internal dose profile than the same cumulative exposure delivered continuously (Paustenbach et al., 1986a, b; Veng-Pedersen et al., 1987).

2. Toxicokinetics

2.1 Absorption

Carbon tetrachloride is readily absorbed by all three exposure routes. Oral absorption in rats is rapid: after gavage doses of approximately 18 or 180 mg/kg, peak liver concentrations were detected within 1 minute and peak blood concentrations within 10 minutes (Sanzgiri et al., 1995; Bruckner et al., 1990), and after intraduodenal injection of radiolabeled carbon tetrachloride in rats, at least 82% of the dose was absorbed based on recovery of label in exhaled air (Paul and Rubinstein, 1963). The dosing vehicle materially affects the rate and extent of gastrointestinal absorption: relative to neat carbon tetrachloride, absorption in male Sprague-Dawley rats was faster in aqueous solution or emulsion but slower, though ultimately more complete over 9 hours, when administered in corn oil (Kim et al., 1990a; Gillespie et al., 1990).

Inhalation absorption is also rapid and substantial: human volunteers exposed to 10 ppm vapor showed detectable carbon tetrachloride in exhaled air within 15 minutes (Stewart et al., 1961), and Lehmann and Schmidt-Kehl (1936) estimated approximately 63% pulmonary absorption in human subjects exposed to vapor concentrations of a few mg/L. In monkeys exposed at 46 ppm, average pulmonary absorption was approximately 30% of the inhaled amount (McCollister et al., 1951).

Dermal absorption, while measurable, is comparatively minor for whole-body vapor exposures: carbon tetrachloride was detectable in human alveolar air within 10 minutes of thumb immersion in the neat liquid (Stewart and Dodd, 1964), and in guinea pigs dermal application produced detectable blood levels within 5 minutes (Jakobson et al., 1982), but McCollister et al. (1951) concluded that the dermally absorbed fraction is negligible relative to inhalation for whole-body vapor exposure scenarios.

2.2 Distribution

Once absorbed, carbon tetrachloride distributes preferentially to lipid-rich tissue. In rats given an oral dose of approximately 3,200 mg/kg, peak tissue levels were similar in blood and muscle but twice as high in brain, five times higher in liver, and fifty times higher in fat (Marchand et al., 1970). A comparable pattern was seen by inhalation: rats exposed to 4,000 ppm for 6 hours showed the highest concentrations in fat, followed by brain, kidney, liver, and blood (Dambrauskas and Cornish, 1970). Whole-body autoradiography in mice demonstrated that immediately after inhalation exposure, volatile (parent-compound) radioactivity was highest in fat, bone marrow, and nervous tissue, while nonvolatile radioactivity (metabolites, including covalently bound material) was concentrated in the liver — with a distinct centrilobular pattern — as well as kidney cortex, lung, gastrointestinal mucosa, nasal mucosa, and, interstitially, the testis (Bergman, 1983). A comparative pharmacokinetic study across inhalation, oral bolus, and gastric infusion dosing found that fat consistently accumulated the highest tissue concentrations of carbon tetrachloride regardless of route, and that maximal levels were reached fastest by oral bolus dosing, followed by inhalation, then gastric infusion (Sanzgiri et al., 1997).

2.3 Metabolism

Carbon tetrachloride is metabolized primarily in the liver but also in the kidney, lung, and other tissues bearing cytochrome P450 (CYP450) content, in proportion to local CYP450 levels (Bergman, 1983). The initial and rate-determining step is reductive dehalogenation, catalyzed by an NADPH-dependent CYP450 enzyme — principally CYP2E1, with CYP3A contributing at higher exposure levels (Zangar et al., 2000; Raucy et al., 1993) — that cleaves one carbon-chlorine bond to yield chloride ion and the highly reactive trichloromethyl radical (•CCl3). The essential role of CYP2E1 in bioactivation has been demonstrated directly using CYP2E1 knockout mice: whereas wild-type mice given an intraperitoneal dose of 1.59 g/kg carbon tetrachloride showed a 422-fold increase in serum ALT, a 125-fold increase in serum AST, and marked centrilobular hepatic necrosis, CYP2E1-knockout mice given the identical dose showed no significant AST increase, only a slight (within-normal-range) ALT elevation, and no liver histopathology (Wong et al., 1998).

Under anaerobic conditions the trichloromethyl radical may dimerize to hexachloroethane or accept a proton and electron to form chloroform, both of which have been detected in exposed animal tissue (Reynolds et al., 1984; Dambrauskas and Cornish, 1970). Under aerobic conditions, the trichloromethyl radical reacts with molecular oxygen to form the considerably more reactive trichloromethyl peroxy radical (•OOCCl3), the species chiefly responsible for initiating lipid peroxidation (see Section 4). Both radical species can covalently bind cellular macromolecules — lipids, proteins, and nucleic acids — a process termed haloalkylation (Recknagel and Glende, 1973; Slater, 1981). Metabolism of carbon tetrachloride is also self-limiting at high exposure: the reactive intermediates generated destroy the CYP450 enzyme responsible for their own formation (“suicide inactivation”), a phenomenon documented directly by Manno et al. (1992, 1988), who found that although the kinetics of this suicide inactivation were similar between rat and human CYP450 in vitro, inactivation was substantially more severe in the rat (1 enzyme molecule lost per 26 substrate molecules metabolized) than in human enzyme preparations (1 enzyme molecule lost per 196 substrate molecules metabolized) — a quantitative species difference of direct relevance to interpreting rodent bioassay potency relative to expected human response (see Section 6).

In vitro comparison of hepatic microsomal metabolic rate constants across species (Zangar et al., 2000) found the human Michaelis constant (Km, 56.8 µM) and maximum velocity (Vmax, 2.26 nmol/min/mg protein) to be most similar to those of the rat (Km 59.1 µM; Vmax 3.1 nmol/min/mg protein) among the species tested, with mouse and hamster microsomes showing somewhat different kinetic parameters — evidence used to support the rat as a reasonably appropriate metabolic surrogate for low-dose human risk extrapolation, notwithstanding the CYP450 inactivation-rate difference noted above.

2.4 Excretion

The unmetabolized parent compound is excreted primarily in exhaled breath in both humans and animals. Using a four-compartment physiological model, Sato and Nakajima (1987) calculated that approximately 93% of an inhaled carbon tetrachloride dose is eliminated unchanged via the lungs in humans, with roughly 7% cleared metabolically. In rats given oral doses of 15 mg/kg, approximately 19% of the dose was exhaled as parent compound and 28% as carbon dioxide (accounting for the majority of the metabolized fraction), while only small amounts were excreted in urine (2.7%) and feces (11%) or remained bound in the liver (2.9%) (Reynolds et al., 1984). At higher oral doses (≥600 mg/kg), a substantially larger fraction (≥76%) was exhaled unchanged, reflecting saturation of the metabolic pathway at high dose. Elimination half-lives from individual tissues in the rat are broadly similar across tissue types (roughly 200–300 minutes for liver, kidney, lung, brain, heart, muscle, and spleen) but are markedly longer in fat (roughly 665–780 minutes), consistent with the poor perfusion of adipose tissue (Sanzgiri et al., 1997).

2.5 Species and Sex Differences Relevant to Interpreting Animal Data

Several species differences bear directly on the interpretation of the animal toxicology database. First, the CYP450 suicide-inactivation kinetics described above (Manno et al., 1992, 1988) indicate that rat hepatic CYP450 is more readily destroyed by carbon tetrachloride metabolites, on a per-molecule-metabolized basis, than human CYP450 — a difference with uncertain net effect on relative sensitivity (more inactivation could either self-limit toxicity or reflect a more reactive metabolic environment). Second, comparative in vitro metabolic rate constants (Zangar et al., 2000) show that mouse and hamster microsomes handle carbon tetrachloride somewhat differently than rat and human microsomes, consistent with the substantially greater sensitivity of mice relative to rats and hamsters observed directly in comparative subchronic inhalation bioassays (Benson and Springer, 1999; see Section 4.1). Third, comparative inhalation toxicokinetics across rats, mice, and hamsters exposed to 20 ppm carbon tetrachloride found that the percentage of initial body burden in major tissues immediately after exposure was higher in mice (40%) than in rats or hamsters (30% each), with elimination half-times for CO2 and for volatile organic compounds also differing measurably by species (Benson et al., 2001). Fourth, life-stage-related differences in CYP2E1 and CYP3A expression are relevant to interpreting extrapolation from adult animal bioassays to developmental risk: hepatic CYP2E1 does not reach adult levels in humans until sometime between 1 and 10 years of age, with a substantial postnatal increase between 1 and 3 months (Vieira et al., 1996), while CYP3A activity follows a different, non-monotonic developmental trajectory (Ginsberg et al., 2002).

3. Mechanism(s) of Action

Carbon tetrachloride is a metabolically activated, indirect-acting toxicant: the parent compound itself is essentially inert, and virtually all of its hepatotoxic, nephrotoxic, and carcinogenic activity depends on bioactivation by CYP450, predominantly CYP2E1, to reactive free-radical intermediates. This dependence has been demonstrated repeatedly and convergently: CYP450 inhibitors (e.g., SKF-525A, colchicine, silymarin, allylisopropylacetamide) block carbon tetrachloride-induced liver injury (Martinez et al., 1995; Letteron et al., 1990; Mourelle et al., 1988; Bechtold et al., 1982; Weddle et al., 1976); CYP2E1-null mice are protected from hepatotoxicity at doses that severely injure wild-type mice (Wong et al., 1998); and pretreatment of rats with a small, sublethal dose of carbon tetrachloride — which transiently depletes hepatic CYP450 — confers temporary protection against an otherwise lethal challenge dose, with the time course of protection closely paralleling the time course of CYP450 recovery (Glende, 1972).

3.1 Free Radical Formation and Covalent Binding

• CYP450-catalyzed reductive dehalogenation of carbon tetrachloride yields the trichloromethyl radical (
• CCl3), which under aerobic conditions is rapidly converted to the more reactive trichloromethyl peroxy radical (
• OOCCl3). Radical scavengers (e.g., N-acetylcysteine) and spin-trapping agents confer protection against carbon tetrachloride hepatotoxicity, supporting a causal role for these radicals in cellular injury (Brennan and Schiestl, 1998; Stoyanovsky and Cederbaum, 1996). Both radical species bind covalently to cellular lipids, proteins, and nucleic acids (haloalkylation); pulse radiolysis studies indicate the trichloromethyl peroxy radical reacts with macromolecules considerably faster than the trichloromethyl radical, although the trichloromethyl radical’s slower but more stable binding may account for the majority of persistent covalent adducts (Slater, 1981; Packer et al., 1978).

3.2 Lipid Peroxidation

The trichloromethyl peroxy radical abstracts hydrogen from polyunsaturated fatty acids in cellular membranes, initiating a self-propagating chain reaction of lipid peroxidation that degrades membrane integrity and generates reactive aldehyde byproducts, principally malondialdehyde (MDA) and 4-hydroxynonenal (4-HNE) (Klaassen, 1996; Recknagel and Glende, 1973). Lipid peroxidation following carbon tetrachloride exposure has been demonstrated by multiple convergent markers: conjugated dienes in liver lipid (Tribble et al., 1987; Rao and Recknagel, 1969), increased exhalation of ethane and pentane (degradation products of peroxidized omega-3 and omega-6 fatty acids, respectively) in treated rats (Younes and Siegers, 1985; Gee et al., 1981), and elevated thiobarbituric acid-reactive substances (TBARS) in liver and other tissues (Gasso et al., 1996; Fraga et al., 1987). Fraga et al. (1987) demonstrated that lipid peroxidation following carbon tetrachloride exposure is not confined to the liver: in rat tissue slices, TBARS increased approximately fourfold in liver, threefold in kidney, twofold in spleen, and fivefold in testis relative to controls, and Abraham et al. (1999) similarly reported significantly elevated lipid peroxide levels in the lung (65%), testis (200%), kidney (85%), and liver (200%) of rats exposed to carbon tetrachloride vapor over 12 weeks, together with elevated protein carbonyl content (a marker of oxidative protein damage) in liver, lung, and testis.

Reactive aldehydes generated by lipid peroxidation can diffuse from their site of origin and form promutagenic DNA adducts. Wacker et al. (2001) developed a sensitive detection method for 4-HNE-deoxyguanosine adducts and found that a single intraperitoneal dose of 500 mg/kg carbon tetrachloride elevated hepatic adduct levels 1.5- to 2-fold in F344 rats; Chung et al. (2000) reported a 37-fold increase in the same adduct type in rat liver DNA following a higher intraperitoneal dose (3.2 g/kg). These adducts, together with MDA-DNA adducts demonstrated in a dose-dependent manner in rat hepatocytes exposed to carbon tetrachloride in vitro (Beddowes et al., 2003), are mechanistically linked to the mutagenic potential of carbon tetrachloride discussed further in Section 5.

3.3 Glutathione Depletion

Reduced glutathione (GSH) quenches free-radical chain reactions and limits lipid peroxidation damage; its depletion is both a consequence and an amplifier of carbon tetrachloride toxicity. In rats developing carbon tetrachloride-induced cirrhosis, hepatic GSH levels declined significantly by week 5 of twice-weekly intraperitoneal dosing, with GSH peroxidase activity falling by week 7 as lipid peroxidation markers rose in parallel (Cabre et al., 2000). Pretreatment with GSH or its precursor cysteine partially protects against carbon tetrachloride-induced hepatic necrosis (Gorla et al., 1983; de Ferreyra et al., 1974), and depletion of GSH with buthionine sulphoximine increases basal oxidative DNA damage across mammalian cell lines generally, independent of carbon tetrachloride exposure (Will et al., 1999), underscoring the centrality of the GSH antioxidant system to limiting this class of oxidative injury.

3.4 Disruption of Calcium Homeostasis

Carbon tetrachloride exposure produces sustained, large increases (≥100-fold in some reports) in cytosolic calcium concentration (Agarwal and Mehendale, 1986, 1984; Long and Moore, 1986), driven by inactivation of the calcium-ATPase responsible for sequestering calcium in the endoplasmic reticulum (Moore et al., 1976; Lowrey et al., 1981) and by opening of ryanodine-sensitive calcium release channels (Stoyanovsky and Cederbaum, 1996). In vitro, carbon tetrachloride inhibited the plasma-membrane calcium transport system rapidly (within one minute) and profoundly (>90%) (Hemmings et al., 2002). The resulting rise in cytosolic calcium activates calcium-dependent cysteine proteases (notably calpain) and phospholipase A2, both of which degrade structural and membrane proteins critical to cell integrity. Limaye et al. (2003) demonstrated that pharmacological calpain inhibition administered one hour after a normally lethal dose of carbon tetrachloride improved survival in rats from 25% to 75%, directly implicating calpain-mediated proteolysis as a major contributor to lethal hepatocellular injury, distinct from the lipid peroxidation pathway.

3.5 Immunological and Inflammatory Contribution

A portion of carbon tetrachloride’s observed immunotoxicity is secondary to hepatotoxicity itself and the subsequent inflammatory hepatic repair process, including activation of resident Kupffer cells and hepatic stellate cells, rather than a direct, independent immunotoxic mechanism (mechanistic synthesis reflecting the studies summarized in Section 4.5).

3.6 Renal Mechanisms

Direct evidence in the kidney is more limited than in the liver but indicates that overlapping mechanisms operate. Dogukan et al. (2003) found that subcutaneous carbon tetrachloride dosing in rats produced renal histopathology (tubular necrosis, glomerular hypercellularity, interstitial fibrosis) accompanied by significantly increased renal malondialdehyde and decreased GSH peroxidase activity, consistent with oxidative-stress-mediated injury. Ozturk et al. (2003) similarly found elevated renal superoxide dismutase and catalase activity alongside reduced GSH peroxidase activity and severe cortical histopathology in rats given repeated subcutaneous doses, with co-administration of the methyl-donor betaine (which supports S-adenosylmethionine and, indirectly, GSH synthesis) substantially attenuating both the biochemical and histopathological renal findings — evidence directly parallel to the hepatoprotective effect of S-adenosylmethionine repletion described for the liver (Gasso et al., 1996).

4. Animal Toxicological Evidence by Organ System

The liver is, by a wide margin, the most extensively studied and most sensitive target organ for carbon tetrachloride in animals by both the oral and inhalation routes, followed by the kidney; effects on other organ systems are generally documented at higher doses or under specific study designs. Representative studies are summarized below by organ system.

4.1 Hepatic Effects

Across species and routes, hepatotoxicity follows a consistent dose-related progression from fatty change (steatosis) through inflammation, single-cell and centrilobular necrosis, fibrosis, and ultimately cirrhosis and bile-duct proliferation, generally accompanied by elevated serum ALT, AST, and sorbitol dehydrogenase (SDH). In a 12-week oral gavage study, Sprague-Dawley rats given a time-weighted average dose of 23.6 mg/kg-day showed body weight decrements, serum ALT elevated up to 34-fold and SDH up to 50-fold over control, and extensive hepatic lesions including lipid vacuolization and periportal fibrosis, while a NOAEL of 0.71 mg/kg-day was identified (Bruckner et al., 1986). A parallel recovery study in F344 rats (12 weeks of gavage dosing at 14.3 or 28.6 mg/kg-day) found that although serum enzymes and CYP450 content returned toward control levels within 8 days after exposure ended, cirrhosis identified histopathologically at the low dose persisted through the 22-day post-exposure observation period, indicating that some structural hepatic injury is not readily reversible even after biochemical markers normalize (Allis et al., 1990).

Subchronic mouse gavage studies confirm a comparable dose-response pattern: CD-1 mice given 86 mg/kg-day for 12 weeks showed markedly elevated liver enzymes and hepatocellular lesions including necrosis and portal bridging fibrosis, with vehicle-dependent differences in severity between corn oil and Tween-60 emulsion (Condie et al., 1986); a separate 90-day CD-1 mouse study using doses as low as 12 mg/kg-day found dose-related increases in serum enzymes and hepatic lesions at every dose tested, failing to identify a NOAEL (Hayes et al., 1986).

By the inhalation route, foundational subchronic studies across five species (rat, guinea pig, rabbit, dog, monkey) established a consistent hepatotoxic hierarchy of sensitivity, with monkeys the most resistant (adverse effects only at 100 ppm, the highest concentration tested) and guinea pigs and rats the most sensitive (NOAEL of 5 ppm; LOAEL of 10 ppm for mild liver changes) (Adams et al., 1952). A modern 13-week inhalation study in F344 rats found exposure-related increases in liver histopathology (fatty change, cytological alteration) beginning at the lowest tested concentration of 10 ppm, without a clear NOAEL (Nagano et al., 2007a). A comparative 12-week inhalation study directly contrasting rats, mice, and Syrian hamsters at matched concentrations (5, 20, or 100 ppm) found mice to be the most sensitive species, with significantly increased serum ALT/SDH, hepatic necrosis, and hepatocellular proliferation (measured by BrdU labeling) at 20 ppm, while rats and hamsters required 100 ppm to show comparable effects — a NOAEL of 5 ppm and LOAEL of 20 ppm in mice versus a NOAEL of 20 ppm and LOAEL of 100 ppm in rats and hamsters (Benson and Springer, 1999).

Chronic 2-year inhalation bioassays in F344 rats (Nagano et al., 2007b) identified a NOAEL of 5 ppm and LOAEL of 25 ppm for combined hepatic and renal effects, with fatty change, fibrosis, and cirrhosis increasing markedly in incidence and severity at 25 and 125 ppm in both sexes. The parallel 2-year mouse bioassay found comparable hepatic lesions (degeneration, cyst formation, ceroid deposition, bile duct proliferation) at 25 and 125 ppm, with a NOAEL of 5 ppm (Nagano et al., 2007b).

4.2 Renal Effects

Renal toxicity generally occurs at doses similar to or somewhat higher than those producing hepatotoxicity and is characterized by tubular degeneration, proteinuria, and, at chronic high exposure, glomerular damage. In the Adams et al. (1952) five-species inhalation study, slight renal tubular degeneration first appeared at 200 ppm and increased kidney weight at 400 ppm, concentrations well above the hepatic LOAELs in the same study. In the chronic 2-year rat inhalation bioassay, exposure-related increases in the severity of chronic progressive nephropathy (glomerulonephrosis) occurred at 25 and 125 ppm in both sexes, accompanied by increased blood urea nitrogen at ≥25 ppm and increased creatinine and inorganic phosphate at 125 ppm, indicating impaired glomerular filtration at the same concentrations producing histopathological renal lesions; a low-level increase in the severity (though not incidence) of proteinuria was also observed at the lowest tested concentration of 5 ppm, though its interpretation is complicated by the high spontaneous background incidence of age-related renal lesions characteristic of the F344 rat strain (Nagano et al., 2007b). Direct mechanistic studies using subcutaneous dosing (outside standard bioassay design) have demonstrated renal oxidative stress — elevated malondialdehyde, reduced GSH peroxidase activity — accompanying tubular and glomerular histopathology, findings that parallel the hepatic mechanism and were substantially mitigated by co-administration of the methyl-donor betaine (Dogukan et al., 2003; Ozturk et al., 2003).

4.3 Immune System Effects

Immunotoxicity of carbon tetrachloride has been demonstrated by parenteral, oral, and inhalation routes and consistently affects T-cell-dependent humoral immunity preferentially over T-cell-independent responses, generally at doses that are also hepatotoxic. Intraperitoneal dosing of female B6C3F1 mice (500–1,500 mg/kg-day for 7 days) suppressed the T-cell-dependent antibody response to sheep erythrocytes in a dose-related manner (36–53% suppression) while suppressing the T-cell-independent response only modestly at the highest dose (Kaminski et al., 1989, 1990). Oral gavage dosing of B6C3F1 mice at doses as low as 50 mg/kg-day for 14 days suppressed the IgM antibody-forming cell response, reduced numbers of specific T-cell subsets, and reduced host resistance to challenge with Listeria monocytogenes and Streptococcus pneumoniae, identifying 50 mg/kg-day as a LOAEL for oral immunotoxicity (Guo et al., 2000). By contrast, a 10-day oral gavage study in male F344 rats at doses up to 40 mg/kg-day found dose-related hepatotoxicity but no effect on any immunological parameter tested (natural killer cell activity, cytotoxic T-lymphocyte response, mitogen-stimulated lymphocyte proliferation, or T-cell-dependent antibody response), illustrating some inconsistency in immunotoxic sensitivity across species and study designs at comparable hepatotoxic dose levels (Smialowicz et al., 1991). A brief inhalation study in female BALB/c mice (100–300 ppm) found no effect on splenic (systemic) antibody response but a significant increase in local immune activity (plaque-forming cells, interferon-γ release) in lung-associated lymph nodes at 300 ppm, suggesting a portal-of-entry immunotoxic effect distinct from the hepatotoxicity-secondary mechanism seen with oral and parenteral dosing (Ban et al., 2003).

4.4 Neurological Effects

Neurotoxicity of carbon tetrachloride in animals is primarily an acute, high-dose phenomenon. Adams et al. (1952) observed signs of central nervous system depression in rats at inhalation concentrations ≥4,600 ppm. Frantik et al. (1994) quantified acute neurotoxic potency directly by measuring inhibition of electrically evoked seizure propagation in rats and mice following single 2–4-hour whole-body inhalation exposures, deriving isoeffective air concentrations of 611 ppm in rats and 1,370 ppm in mice — concentrations far above those producing hepatotoxicity in chronic studies, indicating that acute neurotoxicity is not the most sensitive endpoint for carbon tetrachloride under repeated low-level exposure conditions.

4.5 Cardiovascular Effects

No dedicated animal cardiovascular toxicology studies for carbon tetrachloride were identified in the reviewed literature independent of its role as a general marker of systemic toxicity (e.g., organ weight and serum chemistry changes referenced above); cardiovascular effects are not a well-characterized endpoint for this chemical in the animal database.

4.6 Reproductive Effects

No adequate reproductive toxicity studies of oral carbon tetrachloride exposure in animals were identified; available oral developmental studies (Section 4.7) focused on prenatal loss rather than parental reproductive function. By inhalation, testicular atrophy was observed in rats exposed to 200 or 400 ppm for 6 months (Adams et al., 1952), and reduced fertility was reported in rats exposed to 200 or 400 ppm for up to 10.5 months (Smyth et al., 1936); both findings occurred at concentrations well above those producing hepatotoxicity in the same studies. Bergman (1983) noted partly nonextractable (metabolite-bound) radiolabel in the interstitial testis of mice following inhalation exposure to radiolabeled carbon tetrachloride, consistent with local metabolic activation in testicular tissue, and Fraga et al. (1987) and Abraham et al. (1999) both documented substantial testicular lipid peroxidation (see Section 3.2) following carbon tetrachloride exposure.

4.7 Developmental Effects

The most consistent and best-characterized developmental finding for carbon tetrachloride in animals is early, maternally toxic-dose-associated full-litter resorption in rats, rather than structural teratogenicity. In F344 rats given 112.5 or 150 mg/kg-day by oral gavage on gestation days 6–19, full-litter resorption occurred in 44% and 71% of litters, respectively (versus 0% in controls), accompanied by maternal weight loss; no fetal malformations were associated with treatment, and among litters that were not fully resorbed there were no effects on postnatal survival or pup weight (Narotsky and Kavlock, 1995). Follow-up mechanistic work found that full-litter resorption was associated with significantly reduced serum luteinizing hormone (LH) in dams, and that co-administration of human chorionic gonadotropin (an LH surrogate) prevented the adverse developmental outcome, supporting a maternally mediated hormonal mechanism rather than direct fetotoxicity (Narotsky et al., 1997a, 1995). A vehicle-comparison study in F344 rats found that full-litter resorption was substantially more frequent with corn oil vehicle (67% at 75 mg/kg-day) than with an aqueous emulsion vehicle (8% at the same dose), while identifying a NOAEL of 25 mg/kg-day and LOAEL of 50 mg/kg-day for this endpoint with either vehicle (Narotsky et al., 1997b).

By the inhalation route, Schwetz et al. (1974) exposed pregnant Sprague-Dawley rats to 334 or 1,004 ppm on gestation days 6–15 and found significant reductions in fetal body weight (7–14%) and crown-rump length, with delayed sternebral ossification at the higher concentration; both concentrations also produced clear maternal hepatotoxicity (markedly elevated serum ALT, increased liver weight), identifying a LOAEL of 334 ppm for combined maternal and developmental toxicity, with no evidence of developmental effects occurring in the absence of maternal toxicity in this study. An earlier gavage study using very high, maternally lethal single or repeated doses (478 mg administered on 1–2 days of gestation) produced substantial total-litter loss in surviving dams, but because the dose also caused 21% maternal mortality, the study could not distinguish a direct fetotoxic effect from a consequence of severe maternal toxicity (Wilson, 1954).

5. Animal Carcinogenicity Bioassay Findings

Carbon tetrachloride is one of the more extensively studied rodent hepatocarcinogens, with positive findings replicated across at least eight independent bioassays spanning oral, inhalation, and parenteral routes in three species (rat, mouse, hamster).

Early National Cancer Institute-era studies established the qualitative finding: Eschenbrenner and Miller (1946) found liver tumors (hepatomas) in all strain A mice given average daily oral doses as low as 20 mg/kg-day, with no tumors at 10 mg/kg-day, establishing an early dose-response signal; comparable hepatocarcinogenicity was reported in additional mouse strains by oral gavage (Edwards, 1941; Edwards and Dalton, 1942; Andervont, 1958) and in hamsters (Della Porta et al., 1961). A formal NCI bioassay using carbon tetrachloride as a positive control for other chlorinated solvents (NCI, 1976a, b, 1977) found that oral gavage doses of 1,250 or 2,500 mg/kg in B6C3F1 mice for 78 weeks produced hepatocellular carcinoma in nearly all surviving treated mice (49/49 low-dose males; 47/48 high-dose males; 40/41 low-dose females; 43/45 high-dose females) versus 6% and 1% in pooled male and female controls, respectively, together with a significantly increased incidence of adrenal adenoma and pheochromocytoma. The concurrent rat bioassay (47–94 mg/kg in males, 80–159 mg/kg in females) found only a low incidence of liver tumors overall, though the increase in hepatocellular carcinoma was statistically significant in low-dose females relative to pooled controls; high early mortality in this study, particularly at the high dose, likely reduced its power to detect a carcinogenic effect.

The modern chronic inhalation bioassays (Nagano et al., 2007b; described further in Section 4) provide the most complete dose-response characterization. In F344 rats exposed to 5, 25, or 125 ppm for 104 weeks, hepatocellular adenoma and carcinoma incidence was significantly increased in both sexes at 125 ppm (e.g., adenoma or carcinoma combined: 40/50 males, 44/50 females, versus 1/50 in pooled male and 0/50 in pooled female controls), and although tumor incidence was not statistically elevated at 25 ppm, the carcinoma incidence in 25-ppm females (3/50) exceeded the historical control range from the same laboratory (0–2%) and was statistically significant relative to that historical control. In the parallel 2-year Crj:BDF1 mouse bioassay, hepatocellular adenoma and carcinoma incidence was significantly elevated in both sexes at 25 and 125 ppm, and — notably — the incidence of hepatocellular adenoma in female mice at the lowest tested concentration of 5 ppm (8/49, 16%) was also statistically significantly elevated compared to concurrent controls and exceeded the historical control range (2–10%), a finding of particular importance because this concentration was not associated with any hepatotoxic (cytotoxic) effect in the same study, raising the possibility of a tumorigenic response occurring in the absence of the cytotoxicity-driven regenerative-hyperplasia mechanism thought to dominate at higher doses. The same mouse bioassay also found a significantly increased incidence of adrenal pheochromocytoma in males at ≥25 ppm and in females at 125 ppm, with incidences (e.g., 32/50 in high-dose males) far exceeding the historical control range of 0–2%; nearly all of these tumors were histologically benign.

Limited parenteral-route data are also available: subcutaneous injection of carbon tetrachloride (average dose 0.29 mg/kg-day for 33–47 weeks) induced hepatocellular carcinoma in Osborne-Mendel, Japanese, and Wistar rat strains but not in Sprague-Dawley or black rat strains, indicating a strain-dependent susceptibility even within a single species and route (Reuber and Glover, 1970, 1967); intraperitoneal dosing induced hepatomas in C3H mice (Kiplinger and Kensler, 1963, as cited in the reviewed compilation).

Genotoxicity testing of carbon tetrachloride has been largely negative in standard bacterial and transgenic mammalian mutation assays, and the weight of evidence favors an indirect, secondary genotoxic mechanism operating chiefly through oxidative DNA damage and lipid-peroxidation-derived DNA adducts (Section 3.2) under cytotoxic conditions, rather than direct DNA-reactive mutagenesis (Beddowes et al., 2003; Wacker et al., 2001; Chung et al., 2000).

6. Comparison of Animal-Derived Findings to Human Epidemiological Findings

6.1 Hepatotoxicity

The qualitative pattern of human hepatotoxicity from occupational carbon tetrachloride exposure — elevated liver enzymes, hepatomegaly, and, at high exposure, jaundice and acute liver failure — is well conserved with the animal database, and case reports of acute human poisoning (fatal and nonfatal) closely mirror the centrilobular necrosis and steatosis documented experimentally. Quantitatively, however, a cross-sectional study of 135 carbon tetrachloride-exposed workers at three UK chemical plants (mean exposure categories ≤1 to 4–12 ppm) found only modest and inconsistent serum liver-enzyme changes relative to matched unexposed controls, with a statistically significant multivariate difference between exposed and unexposed groups but no clear dose-response relationship across exposure categories and no significant univariate differences for individual enzymes (Tomenson et al., 1995). This human finding of a comparatively subtle hepatic signal at exposure levels in the low parts-per-million range contrasts with the animal bioassay data, in which chronic inhalation exposure at similar or even somewhat lower adjusted concentrations (e.g., the 5 ppm NOAEL/25 ppm LOAEL bracket identified in the chronic rat and mouse inhalation bioassays; Nagano et al., 2007b) produced clear, dose-related hepatic pathology in rodents. Earlier, less well-controlled occupational case series (Kazantzis and Bomford, 1960; Elkins, 1942; Norwood et al., 1950) documented gastrointestinal and hepatic symptoms at somewhat higher, poorly quantified occupational exposures, consistent in direction though not precisely quantifiable against the animal dose-response data.

6.2 Carcinogenicity

Human epidemiological evidence for carbon tetrachloride carcinogenicity is considered inadequate to establish an association, in contrast to the robust and repeatedly replicated animal hepatocarcinogenicity findings. Several occupational cohort and case-control studies have reported associations between solvent-exposed occupations and non-Hodgkin lymphoma (Blair et al., 1998; Spirtas et al., 1991), lymphosarcoma and lymphatic leukemia (Checkoway et al., 1984; Wilcosky et al., 1984), esophageal and cervical cancer (Blair et al., 1990, 1979), breast cancer (Cantor et al., 1995), astrocytic brain cancer (Heineman et al., 1994), and rectal cancer (Dumas et al., 2000), but in each case exposure to carbon tetrachloride specifically was poorly characterized and confounded by concurrent exposure to other solvents (most of these studies were designed primarily to evaluate tetrachloroethylene or trichloroethylene). Critically, none of the human epidemiology studies reported an association with liver cancer — the principal tumor site identified in every positive animal bioassay — a discordance that may reflect the rarity of human liver cancer and correspondingly low statistical power in these occupational cohorts, rather than a true absence of hepatocarcinogenic hazard in humans.

6.3 Reasons for Divergence Between Animal and Human Data

• Statistical power and tumor rarity: liver cancer is comparatively rare in the general human population, and no epidemiological study of carbon tetrachloride-exposed workers has been sized or exposure-characterized adequately to detect a liver-cancer signal of the magnitude seen in rodent bioassays, even if a proportionally similar excess risk were present in humans.
• Exposure characterization and confounding: virtually all human carbon tetrachloride epidemiology comes from occupational cohorts with concurrent, often better-characterized exposure to other solvents (tetrachloroethylene, trichloroethylene), making it difficult to isolate a carbon tetrachloride-specific effect even where a chemical-class association with cancer exists.
• Toxicokinetic and metabolic differences: the species difference in CYP450 suicide-inactivation kinetics documented by Manno et al. (1992, 1988) — substantially more severe enzyme loss per unit substrate metabolized in rats than in human liver preparations in vitro — means that a given external exposure concentration does not necessarily translate to an equivalent internal dose of reactive metabolite across species, complicating direct extrapolation of rodent bioassay potency to human risk.
• Dose magnitude: many of the animal bioassays demonstrating clear carcinogenicity used doses or concentrations (e.g., 125 ppm by inhalation for 2 years; 1,250–2,500 mg/kg by oral gavage) that produced substantial concurrent cytotoxicity, hepatocellular necrosis, and regenerative hyperplasia — a well-supported mode of action for tumor promotion at high dose — whereas occupational human exposures, even historically, were generally lower and less likely to produce sustained cytotoxic injury of comparable magnitude, though the finding of increased liver adenoma incidence in female mice at a noncytotoxic inhalation concentration (5 ppm; Nagano et al., 2007b) indicates this mode of action does not fully explain tumor formation at the low end of the animal dose-response curve either.
• Mode-of-action uncertainty at low dose: because the hypothesized cytotoxicity-driven mode of action for carbon tetrachloride hepatocarcinogenesis is best supported at relatively high, clearly hepatotoxic exposure levels, and because data characterizing key mechanistic events at low, non-cytotoxic exposure levels are limited, the overall mode of action — and by extension the most appropriate method for low-dose human risk extrapolation — remains an area of substantial uncertainty rather than a fully resolved point of animal-human concordance or divergence.

7. Data Gaps and Areas of Ongoing Research

• The mode of action for carbon tetrachloride-induced liver tumors is reasonably well characterized at high, cytotoxic exposure levels (metabolic activation → radical-mediated cytotoxicity → sustained regenerative hyperplasia) but is not established at low, non-cytotoxic exposure levels, despite the finding of a statistically significant increase in liver adenomas in female mice at a concentration (5 ppm inhalation) that produced no detectable hepatotoxicity in the same study (Nagano et al., 2007b).
• No human epidemiological study has reported liver cancer findings specific to carbon tetrachloride exposure, the principal tumor site identified consistently across the animal bioassay database; whether this reflects a true species difference, inadequate statistical power, or historically poor human exposure characterization has not been resolved.
• The quantitative species difference in CYP450 suicide-inactivation kinetics between rat and human liver preparations in vitro (Manno et al., 1992, 1988) has not been fully translated into a quantitative cross-species internal-dose extrapolation framework for cancer or noncancer risk assessment.
• Renal mechanistic data remain considerably more limited than hepatic mechanistic data; the parallel oxidative-stress findings in kidney (Dogukan et al., 2003; Ozturk et al., 2003) are based on subcutaneous dosing regimens outside standard oral or inhalation study designs, and confirmatory work using standard routes and chronic-duration exposure would strengthen this comparison.
• The maternally mediated hormonal mechanism proposed for carbon tetrachloride-induced full-litter resorption in rats (reduced luteinizing hormone; Narotsky et al., 1997a, 1995) has not been extended to other species or confirmed as the operative mechanism for the developmental effects (reduced fetal weight, delayed ossification) observed by the inhalation route (Schwetz et al., 1974), which occurred alongside maternal hepatotoxicity rather than in a design that could isolate a hormonal pathway.
• Immunotoxicity findings are inconsistent across species and study designs at seemingly comparable hepatotoxic dose levels (compare the positive findings in mice at 50 mg/kg-day oral, Guo et al., 2000, with the entirely negative immunological findings in rats at up to 40 mg/kg-day oral despite clear hepatotoxicity, Smialowicz et al., 1991); the basis for this species difference in immune sensitivity has not been characterized.
• Lifestage-specific data on antioxidant (e.g., glutathione) capacity in children and juvenile animals remain sparse, limiting confident characterization of childhood susceptibility beyond inferences drawn from CYP2E1 and CYP3A developmental expression data alone.

References