2.2.1. Inhalation Exposure

2.2.1.1. Death

Case studies of methylene chloride poisoning during paint stripping operations have demonstrated that inhalation exposure can be fatal to humans (Bonventre et al. 1977; Hall and Rumack 1990; Stewart and Hake 1976). Although quantitative estimates of exposure levels were not reported for these cases, levels of methylene chloride in various tissues were reported: liver (14.4 mg/dL), blood (51 mg/dL), serum (29 μg/mL), and brain (24.8 mg/100 g) (Bonventre et al. 1977; Hall and Rumack 1990). The cause of death in these cases was uncertain; however, myocardial infarction was reported in one case (Stewart and Hake 1976). Death also occurred in two workers involved in oleoresin extraction processes and liquid cleaning operations (Moskowitz and Shapiro 1952; Winek et al. 1981). Exposure reportedly occurred from less than 1 hour up to 3 hours, but the concentration of methylene chloride was not reported. The compound was detected in the lung (0.1 mL/500 g wet tissue), brain (0.27 g/L), and blood (29.8 mg%) (Moskowitz and Shapiro 1952; Winek et al. 1981). Two cases of lethal poisoning following acute inhalation of extremely high concentrations of methylene chloride in air (estimated as up to 168,000 ppm) occurred in two workers burying barrels containing mixed solvents and solid chemical waste in a well about 2 meters below ground level (Manno et al. 1992). Methylene chloride concentrations in blood of the two workers were 572 and 601 mg/L, respectively. Blood carboxyhemoglobin (COHb) concentrations were about 30% higher than normal. Death appears to have been caused by narcosis and respiratory depression due to the acute effects of high concentration methylene chloride on the central nervous system. One death in the United Kingdom resulting from acute inhalation occupational exposure to methylene chloride (concentration not provided) was attributed to acute narcosis resulting in respiratory depression (Bakinson and Jones 1985); necropsy revealed evidence of liver and spleen congestion. One case of fatal gassing occurred during paint-stripping of a chemical tank (Tay et al. 1995). The paint stripper contained 74% w/w (weight by weight) of methylene chloride; the concentration of methylene chloride vapor within the tank was later estimated to have been well above 100,000 ppm. The worker did not wear a respirator and did not engage a forced ventilation system into the confined space of the tank. He was found unconscious and died 4 days later. His methylene chloride blood concentration was 281 mg/L.

Another case of fatal poisoning was reported in Korea: the body of the chief executive of a painting and coating factory was discovered in an underground trichloroethylene tank in which methylene chloride was being used to removed rust from iron sheets (Kim et al. 1996a). Autopsy revealed pronounced organ congestion in the brain, kidneys, liver, and lungs; blood tissue samples contained 3% COHb and 252 mg/L of methylene chloride. This methylene chloride blood concentration is similar to that reported for other fatalities (Manno et al. 1992; Tay et al. 1995). Concentrations of methylene chloride in other tissues ranged from 26 mg/kg in the lungs to 75 mg/kg in the brain (Kim et al. 1996a). The data were sufficient to confirm that the death was due to methylene chloride poisoning.

In another study, no increase in deaths in methylene chloride workers, assessed by life table analysis, was found after exposure to 30–120 ppm (time weighted averages) for over 30 years (Friedlander et al. 1978). Fiber production workers exposed to methylene chloride (140–475 ppm) for at least 3 months did not have a significant increase in mortality (Lanes et al. 1993; Ott et al. 1983b).

Studies in animals confirm that methylene chloride may be lethal after inhalation exposure at high concentrations. Acute exposure to 16,000–19,000 ppm of methylene chloride for 4–8 hours caused death in rats and mice (NTP 1986; Svirbely et al. 1947). Also, one of four female monkeys died after 10 days of continuous exposure to 5,000 ppm methylene chloride (MacEwen et al. 1972). Data suggest there is a narrow margin between concentrations causing anesthesia and death. An LC₅₀ of 16,189 ppm was reported in mice acutely exposed to methylene chloride (Svirbely et al. 1947). No deaths were found in mice exposed for 4 hours to 16,800 ppm, but 70% of the mice exposed to 17,250 ppm died (NTP 1986). Repeated exposure from intermediate to lifetime duration at levels ranging from 1,000 to 16,000 ppm can cause increased deaths in rats, mice, guinea pigs, rabbits, and dogs (Burek et al. 1984; Heppel et al. 1944; NTP 1986). Results of the different inhalation studies described in the NTP (1986) report illustrate that with increasing duration, the lethal exposure level decreases. A 19-day intermittent exposure to 6,500 ppm of methylene chloride or a 13-week exposure to 4,200 ppm were not lethal to rats or mice, but exposure for 2 years at 4,000 ppm reduced survival in female rats and in mice of both sexes (NTP 1986). Although the same target organs (central nervous system, lungs, liver) are affected in mammals, the available mortality data suggest differences in sensitivity among species, with dogs being more sensitive than mice and rats.

An LC₅₀ value and all reliable LOAEL values for lethality in each species and duration category are recorded in Table 2-1 and plotted in Figure 2-1.

Table 2-1

Levels of Significant Exposure to Methylene Chloride - Inhalation.

Figure 2-1

Levels of Significant Exposure to Methylene Chloride - Inhalation.

2.2.2. Oral Exposure

2.2.2.2. Systemic Effects

No studies were located regarding respiratory, cardiovascular, musculoskeletal, or dermal effects in humans or animals following oral exposure to methylene chloride. Gastrointestinal, hematological, hepatic, renal, endocrine, and metabolic effects after oral exposure to methylene chloride are discussed below.

The highest NOAEL values and all reliable LOAEL values for systemic effects in each species and duration category are recorded in Table 2-2 and plotted in Figure 2-2.

Gastrointestinal Effects. A fatal single oral dose of Nitromors, a paint solvent containing 75–80% methylene chloride resulted in severe corrosion of the gastrointestinal tract, perforation, peritonitis, septicemia, and death (Hughes and Tracey 1993). A man who ingested 1–2 pints of Nitromors in a suicide attempt, and started treatment with diuresis and hydrocortisone 1.5 hours later, had episodic gastrointestinal hemorrhage and duodenojejunal ulceration that developed into diverticula 6 months later (Roberts and Marshall 1976). No studies were located regarding gastrointestinal effects in animals after oral exposure to methylene chloride.

Table 2-2

Levels of Significant Exposure to Methylene Chloride - Oral.

Figure 2-2

Levels of Significant Exposure to Methylene Chloride - Oral.

Hematological Effects. COHb levels were elevated to 9% in a woman following lethal ingestion of 300 mL of Nitromors, a paint remover solvent containing 75–80% methylene chloride (Hughes and Tracey 1993). The authors reported that this case study was the first to reveal that ingestion of methylene chloride results in the formation of COHb, as occurs with methylene chloride inhalation. A man who ingested 1–2 pints of Nitromors in a suicide attempt exhibited gross hemoglobinuria, a symptom of intravascular hemolysis (Roberts and Marshall 1976).

In an acute rat study, hemolysis developed within 2 days following a single gavage dose of 1,325 mg/kg of methylene chloride (Marzotko and Pankow 1987); no such effect was seen at 798 mg/kg. In rats that were given 420 mg/kg/day of methylene chloride in the drinking water for 3 months, mean hemoglobin concentrations were elevated in males, and erythrocyte counts were elevated, although mean corpuscular hemoglobin was reduced, in females (Kirschman et al. 1986); because no quantitative data were presented, this information is not presented in Table 2-2. In rats of both sexes that were given 55–249 mg/kg/day of methylene chloride in drinking water for 2 years, red blood cell counts and hematocrit and hemoglobin levels were increased over concurrent control levels (Serota et al. 1986a); however, Serota et al. (1986a) indicate that half of these increases were statistically significant without specifying which ones. A dose level of 6 mg/kg/day was a NOAEL. In mice exposed similarly to 60–236 mg/kg/day no significant hematological effects were observed after 104 weeks of treatment (Serota et al. 1986b).

Hepatic Effects. No studies were located regarding hepatic effects in humans after oral exposure to methylene chloride.

In rats given two doses (39–1,275 mg/kg each) of methylene chloride by gavage, there were no significant effects (within 24 hours) in the liver on levels of glutathione, cytochrome P-450, or serum ALT (Kitchin and Brown 1989). However, at the highest dose level, the activity of ornithine decarboxylase, an enzyme involved in cell growth, was significantly increased, and DNA damage was detected in the livers of rats. In another rat study, acute exposure to doses of 1,095 mg/kg/day of methylene chloride by gavage resulted in liver necrosis (Ugazio et al. 1973). After ingestion of methylene chloride in drinking water at or above doses of 1,200 mg/kg/day for 3 months, hepatic effects in rats included centrilobular necrosis, granulomatous foci, accumulation of ceroid or lipofuscin, and dose-related hepatocytic vacuolation (Kirschman et al. 1986). Liver damage was indicated by increases in serum ALT in males (166 mg/kg/day) and females (1,469 mg/kg/day), and by an increase in serum AST in the latter group. In a parallel 3-month study in mice, there were dose-related hepatic centrilobular fatty changes over the dose range between 226 and 2,030 mg/kg/day (Kirschman et al. 1986). Chronic ingestion of methylene chloride in drinking water has been associated with histological alterations of the liver (cellular foci and areas of cellular alterations) of rats exposed to dose levels of 55 mg/kg/day or greater (Serota et al. 1986a) and fatty changes in mice exposed to levels of 236 mg/kg/day (Serota et al. 1986b); the NOAELs were 6 and 175 mg/kg/day in rats and mice, respectively. F344 rats and B6C3F₁ mice were exposed for 104 weeks to methylene chloride in deionized drinking water at target doses of 0, 5, 50, 125, or 250 mg/kg/day (for rats) and 0, 60, 125, 185, or 250 mg/kg/day (for mice) (Serota et al. 1986a, 1986b). Based on the calculated intake of 6 mg/kg/day for rats (Serota et al. 1986a), a chronic oral MRL of 0.06 mg/kg/day was calculated as described in the footnote of Table 2-2.

Renal Effects. Hemoglobinuria as a result of hemolysis was noted in the case of a suicide attempt by ingestion of the paint remover Nitromors (Roberts and Marshall 1976); the authors indicated that the renal damage that could have been a consequence of hemolysis was averted by treatment with diuresis and hydrocortisone. In an acute animal study, administration of a single oral dose of 1,325 mg/kg of methylene chloride inhibited diuresis in rats (Marzotko and Pankow 1987). In a 3-month study, the pH of the urine was lowered in all rats that ingested methylene chloride at levels 166 mg/kg/day, and increased kidney weights were observed in female rats receiving 1,469 mg/kg/day of methylene chloride, but not 607 mg/kg/day (Kirschman et al. 1986).

Endocrine Effects. No studies were located regarding endocrine effects in humans following oral exposure to methylene chloride. In the only animal study that mentions endocrine effects, administration of a single oral dose of methylene chloride (526 mg/kg) to male rats resulted in angiectasis (dilatation of capillaries) of the adrenal medulla and statistically significant increases in secretion of catecholamines (epinephrine and norepinephrine) (Marzotko and Pankow 1987). The authors suggested that these findings appeared to be stress-related.

Ocular Effects. No studies were located regarding ocular effects in humans after oral exposure to methylene chloride. No treatment-related ophthalmologic findings in rats administered up to 249 mg methylene chloride/kg/day in the drinking water for 104 weeks (Serota et al. 1986a). No further relevant information was located.

Metabolic Effects. Metabolic acidosis was detected in a man who attempted suicide by drinking 1–2 pints of Nitromors paint remover that contained methylene chloride as the active ingredient (Roberts and Marshall 1976); he recovered following treatment with diuresis and hydrocortisone.

2.2.3. Dermal Exposure

2.3.1. Absorption

There is no information to determine whether absorption of methylene chloride in children is different than in adults.

2.3.2. Distribution

There is no information to ascertain whether distribution of methylene chloride would be different in children than in adults.

2.3.3. Metabolism

Methylene chloride metabolism is not known to be qualitatively different in children than adults. Information regarding the developmental expression of enzymes that metabolize methylene chloride is discussed in Section 2.7, Children’s Susceptibility.

Inhalation Exposure

Available data suggest that there are two pathways by which methylene chloride is metabolized. One pathway utilizes the mixed function oxidase (MFO) enzymes and produces carbon monoxide (CO) (Figure 2-3, Pathway 1). The other pathway involves the glutathione transferase (GST) and produces carbon dioxide (CO₂) (Figure 2-3, Pathway 2). It has been postulated that CO₂ can also be produced by the MFO pathway if the reactive intermediate in this pathway (postulated to be formyl chloride) reacts with a nucleophile prior to elimination of the chloride ion and formation of CO (Figure 2-3, Pathway 3) (Gargas et al. 1986).

The MFO pathway seems to be the preferred pathway for methylene chloride metabolism following inhalation exposures. Human subjects exposed by inhalation to 500 ppm or greater for 1 or 2 hours experienced elevated COHb concentrations indicating that methylene chloride was metabolized to CO by the MFO pathway (Stewart et al. 1972). The COHb concentrations rose to an average of 10.1% saturation 1 hour after the exposure of 3 subjects to 986 ppm of methylene chloride for 2 hours. The mean COHb concentration at 17 hours-post exposure remained elevated (3.9% saturation) above the pre-exposure baseline value (1–1.5% saturation). The exposure of 8 subjects to 515 ppm of methylene chloride for 1 hour increased the COHb level, which remained elevated above baseline for more than 21 hours.

In human subjects exposed by inhalation to 50–500 ppm of methylene chloride for up to 5 weeks, COHb concentrations could be predicted from methylene chloride exposure parameters (Peterson 1978). However, the exhaled breath concentrations of methylene chloride correlated better with exposure parameters than did COHb concentrations. No differences in methylene chloride metabolism between male and female subjects were detectable and there was no induction of metabolism to CO during 5-weeks exposure to concentrations ranging from 100 to 500 ppm of methylene chloride (Peterson 1978).

Metabolism of methylene chloride in animals has been shown to be similar to that in humans in an experiment by Fodor et al. (1973). Albino rats exposed to methylene chloride showed COHb formation, confirming the observation that methylene chloride is metabolized to CO following inhalation exposures. Among rats exposed to methylene chloride for 4 hours, concentrations of 500 or 2,500 ppm resulted in similar maximal blood levels of COHb (Wirkner et al. 1997). A concentration-response relationship between inhaled methylene chloride and the maximum changes in COHb values was observed when 4 rabbits were exposed to 1,270–11,520 ppm of methylene chloride for a 4-week period (Roth et al. 1975). The larger the exposure concentration, the longer it took for the changes in blood COHb values to reach their maximum values. The maximum COHb concentration was observed 1.5 hours after exposure to 1,270 ppm and 3.5 hours after exposure to 11,520 ppm.

Figure 2-3

Proposed Pathways for Methylene Chloride Metabolism.

In rats, inhalation of 50 ppm for 6 hours resulted in 26 and 27% of the body burden being recovered as expired CO and CO₂, respectively, during the first 48 hours after the exposure period (McKenna et al. 1982). At 1,500 ppm, 14 and 10% of the body burden were recovered as CO and CO₂, respectively. These values are consistent with the concept that the enzymes responsible for the metabolic conversion of methylene chloride to CO and CO₂ become saturated at high-exposure concentrations.

Anders and Sunram (1982) monitored the fetal and maternal blood concentrations of methylene chloride and carbon monoxide following inhalation exposure of pregnant rats to methylene chloride at 500 ppm for 1 hour. Whereas the level of methylene chloride in fetal blood was significantly lower than in maternal blood, the levels of carbon monoxide were about the same. The authors suggested that the maternal liver has higher biotransforming activity than the fetal liver and that the metabolically generated carbon monoxide equilibrates between the fetal and maternal circulations.

Thier et al. (1991) investigated whether the human metabolism of methylene chloride was similar to monohalogenated methanes; these compounds are metabolized via a glutathione-dependent pathway in human erythrocytes in a subgroup of the human population (called “conjugators”), whereas another subgroup does not exhibit erythrocyte glutathione-mediated metabolism (called “nonconjugators”). Blood samples were taken from 10 volunteers who had previously been determined to be either “conjugators” (subgroup B) or “nonconjugators” (subgroup A) of monohalogenated methanes. The samples were exposed to radiolabeled methylene chloride, incubated, centrifuged to separate blood plasma from cellular fraction, and the distribution of radioactivity between the different blood compartments measured. For individuals from subgroup B, radioactivity in blood plasma increased over time, reaching 30% in the low-molecular weight fraction and 5% in the high-molecular weight fraction after 9 hours. For individuals in subgroup A, almost no radioactivity was found in either blood plasma molecular fraction. In all samples from both groups, no radioactivity was found in either erythrocyte cytoplasm or membranes. Thus, although dihalogenated methylene chloride does not appear to undergo metabolism in erythrocytes, some metabolic transformation of methylene chloride appears to have occurred in the plasma of all individuals classified as “conjugators” (subgroup B), whereas this metabolism did not occur in “nonconjugators” (subgroup A). The authors concluded that these data provide evidence for enzyme polymorphism in humans with respect to methylene chloride metabolism.

The contributions of the MFO and GST pathways to the metabolism of methylene chloride were studied in male rats by Gargas et al. (1986). Based on the data from these studies, the MFO pathway is a high affinity-low capacity pathway with a metabolic rate of 47 μmol/kg/hour. The GST pathway, on the other hand, has a lower affinity than the MFO pathway but a higher capacity. As part of this study, rats were pretreated with an inhibitor of the MFO pathway or a glutathione (GSH) depleting agent. Concentrations of COHb were measured to assess the relative contribution of each pathway to the total metabolism of methylene chloride. Pretreatment with the MFO inhibitor essentially abolished CO production by the high-affinity saturable MFO pathway and limited metabolism to the GSH-transferase pathway. Concentrations of COHb following treatment with the GSH-depleting agent were increased 20–30% compared to untreated controls, indicating increased activity of the MFO pathway. Some CO₂ was produced, indicating that despite some depletion of GSH, some methylene chloride was still metabolized by the GST pathway or by conversion of CO to CO₂ via the MFO pathway.

There appear to be species differences in pathway preference. The GST pathway is more active in the mouse than the rat, based on studies of metabolic end products in both species during and immediately after 6 hours of exposure to 500, 1,000, 2,000, and 4,000 ppm of methylene chloride (Green et al. 1986c). The cytochrome MFO pathway leading to CO and COHb was shown to be saturated at the 500 ppm-exposure level. After saturation of this pathway had occurred, the blood levels of methylene chloride in the rat increased almost linearly with concentration, indicating little further metabolism in this species. In contrast, there was evidence for significant metabolism of methylene chloride in the mouse at high-concentration levels by the GST pathway, which leads to CO₂. Comparison of expired CO₂ levels after 4,000-ppm exposure for 6 hours showed almost an order of magnitude more CO₂ produced per kilogram of body weight in the mouse than in the rat. A marked difference was seen in the rate of clearance of methylene chloride from tissues, as measured by its elimination in expired air. Although cleared from blood rapidly in both species, the rate of clearance from rat tissues was markedly slower than in the mouse. This slow release in the rat sustained metabolism for up to 8 hours after the end of exposure and, consequently, had a marked effect on the overall body burden of metabolites. Methylene chloride was cleared from tissues in the mouse in less than 2 hours. Overall, saturation of the cytochrome P-450 MFO pathway occurred at similar levels in both species, but significantly more methylene chloride was metabolized by the GST pathway in the mouse when assessed either from the blood levels of methylene chloride or by CO₂ formation at high-concentration levels.

The isoenzymes involved in the metabolism and biotransformation of methylene chloride have recently been identified for each major pathway. The MFO pathway involves cytochrome P-450 2E1 and the glutathione-mediated pathway involves a Θ (theta) class glutathione S-transferase, GSTT1-1 (Blocki et al. 1994; Mainwaring et al. 1996b; Meyer et al. 1991). GSTT1-1 is expressed in a number of human organs in a tissue-specific manner which is different from the pattern of expression of other glutathione S-transferases, specifically those from the α, μ, and π classes (Sheratt et al. 1997). A recent in vitro study by Sheratt et al. (1997) detected very low levels of GSTT1-1 in human lung cells, suggesting that the human lung is likely to have a low capacity for activating methylene chloride into reactive metabolites. Although Mainwaring et al. (1996b) found that the overall distribution of mRNA and protein for GSTT1-1 and GSTT2-2 in the liver and lungs of humans was lower than in mice and rats, they found localized high concentrations of GSTT2-2 enzyme in human bile-duct epithelial cells. However, since the enzyme did not localize to the nucleus in these cells, the risk of genotoxic effects from methylene chloride metabolism would appear to be small. Mainwaring et al. (1996b) also found locally high concentrations of GSTT1-1 mRNA in a small number of Clara cells and ciliated cells of the alveolar/bronchiolar junction in one human lung sample out of four. Mainwaring et al. (1996b) also found that rates of metabolism of methylene chloride were low in human tissue, and lower in the lung than in the liver. Thus, it is possible that, in some individuals, specific cell types may be vulnerable to genotoxic effects from reactive intermediates of methylene chloride metabolism, although the overall risk is likely to be low.

The oxidative cytochrome P-450 2E1 is presumed to metabolize methylene chloride to carbon dioxide via a reactive intermediate, formyl chloride; this metabolic pathway also produces carbon monoxide (Green 1997; Reitz 1990). The glutathione pathway metabolizes methylene chloride to carbon dioxide following the formation of both formaldehyde and a glutathione conjugate, putatively chloromethyl glutathione. No carbon monoxide is produced during glutathione-mediated metabolism (Green 1997; Reitz 1990). Neither the formyl chloride, nor the chloromethyl glutathione, nor any other glutathione conjugate of methylene chloride, has been isolated and characterized. However, according to Green (1997), the formation of these reactive intermediates is consistent with the end products formed, and with the enzymes and cofactors available.

An extremely detailed analysis of these two metabolic pathways was conducted in mice, rats, hamsters, and humans, both in vivo and in vitro. These studies provide evidence for the concentration-dependent behaviors of the two pathways and compare the metabolic rates by each pathway in the four species (Bogaards et al. 1993; Green 1991, 1997; Reitz et al. 1989). In vivo, the cytochrome P-450 MFO pathway was the major route of metabolism at low-concentration inhalation exposures to methylene chloride. In both rats and mice, saturation of the MFO pathway occurred at concentration levels of 500 ppm and above, with maximum COHb levels reported to be 12–15%. In vitro studies verified that this pathway was similar in all four species. In contrast, the glutathione S-transferase pathway was the major metabolic pathway at exposure concentrations used in the rodent cancer bioassays, and showed a linear concentration response. Furthermore, metabolic activity in mouse tissue was more than 10-fold greater than metabolic activity in rat tissue. The glutathione-mediated metabolic rates in hamster and human tissues were even lower than those observed in the rat (Green 1997). In human tissues, metabolic rates in the lung were about 10-fold lower than those in the liver (Green 1997). Although the Θ (theta) class glutathione S-transferase (enzyme 5-5 in rat) has a high specific activity for metabolizing methylene chloride (Meyer et al. 1991; Sheratt et al. 1997), the μ-class GSTs (enzymes 3-3, 3-4, and 4-4) are, as a group, 800-fold more abundant in rat liver (Blocki et al. 1994); in rat liver cytosol, the Θ-class enzyme (5-5) and the μ-class group of enzymes (3-3, 3-4, and 4-4) are each responsible for half of the metabolism of methylene chloride to formaldehyde (Blocki et al. 1994). In addition, another Θ-class enzyme (12-12) is present in rat liver, but its lability during isolation has prevented analysis of its specific activity (Meyer et al. 1991). Although Schroder et al. (1996) demonstrated that the human erythrocyte GSTT1-1 is polymorphic, from N-terminal modification, and differs from liver and lung GSTT1-1, it is not yet known whether liver and lung human GSTT1-1 are polymorphic. Thus, the enzymatic basis for methylene chloride metabolism in different tissues in different species is not completely elucidated.

Nelson et al. (1995) examined the distribution of erythrocyte GSTT1-1 polymorphisms among five different ethnic groups: North American Caucasians, African-Americans, Mexican-Americans, Chinese, and Koreans. Polymerase chain reaction (PCR)-based genotyping of erythrocyte GSTT1-1 demonstrated that significant ethnic variations occur. The prevalence of the nonfunctional genotype (i.e., the one lacking the ability to metabolize methylene chloride) was highest among Chinese (64%), followed by Koreans (60%), African-Americans (22%), Caucasians (20%), and Mexican-Americans (10%). These data suggest that there are ethnic differences in metabolizing capacity; additionally, substantial variations in GSTT1-1 polymorphisms also occur within ethnic groups (Nelson et al. 1995)

Kim et al. (1996b) conducted a series of experiments using the vial equilibration technique on freshly isolated liver tissue of male Sprague Dawley rats in order to characterize the metabolism of methylene chloride. Several different hepatic microsomal preparations were used to compare and contrast the MFOand GSH-mediated pathways. To produce glutathione depletion in one study, rats were injected intraperitoneally with 250 mg phorone/kg body weight in corn oil prior to sacrifice and preparation of liver homogenate for in vitro testing; controls received corn oil only. A glycerol buffer preparation significantly inhibited methylene chloride metabolism while a sucrose buffer containing EDTA and KCl did not. The use of substrates for P-450 2E1 (i.e., ethanol, pyrazole) completely inhibited methylene chloride metabolism, indicating that the methylene chloride was metabolized preferentially by the MFO pathway under the conditions of this study. Pretreatment with phorone to produce hepatic glutathione depletion had little effect on the metabolic rate of methylene chloride, demonstrating that the glutathione-mediated metabolism was not a major pathway under the conditions of the study. The authors concluded that these results demonstrate that at low exposure levels, little methylene chloride is metabolized by the GSH pathway, and thus these results confirm the results of other investigators.

Several PBPK models have been developed that can be used to predict tissue-specific exposures to methylene chloride, taking into account absorption, distribution, and metabolism. A PBPK model was used to provide quantitative estimates of the levels of methylene chloride in various organs of four mammalian species (rats, mice, hamsters, and humans) following inhalation exposure (Andersen et al. 1987). The model, which incorporates a variety of variables representing the blood and tissue concentrations of methylene chloride, exhaled methylene chloride, and instantaneous rates of metabolism by each pathway, was validated by comparing predictions of concentrations of methylene chloride in blood with time-course data obtained with Fischer-344 rats, Syrian Golden hamsters, B6C3F₁ mice, and human volunteers. The predicted values for each of the four species were in agreement with the experimental data. The model was also shown to predict the appearance and elimination of methylene chloride metabolites reasonably well.

Other PBPK models are discussed in Section 2.3.5.

Oral Exposure In a lethal poisoning case, a 56 year-old woman ingested approximately 300 mL of Nitromors, a paint remover solvent whose major ingredient is 75–80% methylene chloride (Hughes and Tracey 1993). COHb levels in the patient were increased to about 9% in blood samples taken several hours later. According to the authors, this case demonstrated that conversion of ingested methylene chloride to COHb occurs in humans; previously, the conversion of ingested methylene chloride to COHb had only been reported in rats.

No other studies were located regarding metabolism in humans after oral exposure to methylene chloride.

Animal data on metabolism indicate that the process appears to be similar for inhalation and oral exposures. When rats were gavaged with a single oral dose of 526 mg/kg of methylene chloride, COHb levels in blood rose to nearly 10% (Wirkner et al. 1997). When methylene chloride was administered to rats and mice by gavage at daily doses of 50 and 200 mg/kg in the mice and 50 mg/kg in the rats, there was a dose-dependent biotransformation of methylene chloride to CO₂ and CO (Angelo et al. 1986a, 1986b). Pulmonary excretion of methylene chloride, CO, and CO₂ could be detected within 30 minutes of administration. Initially, exhaled CO₂ levels from the methylene chloride subjects exceeded CO levels in the rats. However, the amount of exhalant did increase with time. A similar profile of metabolism was apparent in mice based on exhaled methylene chloride, CO, and CO₂ values except that the exhaled CO₂ values support studies which indicated that the GST pathway is more important in mice than in rats (Green et al. 1986c, 1988).

Immediately following administration of doses of 500 or 1,000 mg/kg of methylene chloride in corn oil, the values for pulmonary excretion of methylene chloride, CO₂, and CO were all lower than when methylene chloride was administered in aqueous solution, reflecting the slower absorption of methylene chloride from the corn oil vehicle (Angelo et al. 1986a). Three hours after administration the corn oil, pulmonary excretory patterns were similar, but still had lower values than those for the methylene chloride in aqueous solution.

Pankow and Jagielki (1993) studied the in vivo metabolism of methylene chloride to carbon monoxide as measured by the COHb level in blood under the following conditions: pretreatment with methanol; simultaneous administration of methanol; and pretreatment with glutathione-depleting chemicals. Male Wistar rats were administered 6.2 mmol (=0.4 mL)/kg body mass methylene chloride via gavage and >148 mmol/kg of methanol. Six hours following methylene chloride administration, blood samples were taken for COHb determination. The animals were then sacrificed and glutathione levels were measured in the liver. The authors concluded that the cytochrome P-450 2E1 oxidative pathway is responsible for the formation of COHb and the metabolism of methylene chloride to carbon monoxide, and suggested that the two metabolic pathways of methylene chloride (i.e, oxidation cytochrome P-450 2E1 and conjugation via glutathione/glutathione-S-transferase) may be independent at study doses in rats. However, the rationale for this suggestion was unclear.

A comparison of the rates of pulmonary elimination of methylene chloride, CO₂, and CO for oral and intravenous exposures of rats indicated that biotransformation rather than absorption was the rate-limiting factor that controlled the CO and CO₂ production (Angelo et al. 1986b). The same situation was true in mice when the pulmonary elimination patterns of the methylene chloride in water were compared to those from intravenous injection (Angelo et al. 1986a). This was not the case when the methylene chloride was given in corn oil.

These data suggest that both the MFO and GST pathways can participate in the metabolism of methylene chloride. Factors which influence the metabolism of methylene chloride by the oral exposure route are the rate of absorption and the distribution of the MFO and GST enzyme systems in the various tissues and the transportation of methylene chloride via the aqueous blood components to the liver or via the chylomicrons to systemic circulation.

Dermal Exposure No studies were located on the metabolism of methylene chloride after dermal exposure to humans or animals.

2.3.4. Elimination and Excretion

Elimination and excretion of methylene chloride in children are not expected to be different from adults; however, this has not been investigated.

2.3.5. Physiologically Based Pharmacokinetic (PBPK)/Pharmacodynamic (PD) Models

Physiologically based pharmacokinetic (PBPK) models use mathematical descriptions of the uptake and disposition of chemical substances to quantitatively describe the relationships among critical biological processes (Krishnan et al. 1994). PBPK models are also called biologically based tissue dosimetry models. PBPK models are increasingly used in risk assessments, primarily to predict the concentration of potentially toxic moieties of a chemical that will be delivered to any given target tissue following various combinations of route, dose level, and test species (Clewell and Andersen 1985). Physiologically based pharmacodynamic (PBPD) models use mathematical descriptions of the dose-response function to quantitatively describe the relationship between target tissue dose and toxic end points.

PBPK/PD models refine our understanding of complex quantitative dose behaviors by helping to delineate and characterize the relationships between: (1) the external/exposure concentration and target tissue dose of the toxic moiety, and (2) the target tissue dose and observed responses (Andersen et al. 1987; Andersen and Krishnan 1994). These models are biologically and mechanistically based and can be used to extrapolate the pharmacokinetic behavior of chemical substances from high to low dose, from route to route, between species, and between subpopulations within a species. The biological basis of PBPK models results in more meaningful extrapolations than those generated with the more conventional use of uncertainty factors.

The PBPK model for a chemical substance is developed in four interconnected steps: (1) model representation, (2) model parametrization, (3) model simulation, and (4) model validation (Krishnan and Andersen 1994). In the early 1990s, validated PBPK models were developed for a number of toxicologically important chemical substances, both volatile and nonvolatile (Krishnan and Andersen 1994; Leung 1993). PBPK models for a particular substance require estimates of the chemical substance-specific physicochemical parameters, and species-specific physiological and biological parameters. The numerical estimates of these model parameters are incorporated within a set of differential and algebraic equations that describe the pharmacokinetic processes. Solving these differential and algebraic equations provides the predictions of tissue dose. Computers then provide process simulations based on these solutions.

The structure and mathematical expressions used in PBPK models significantly simplify the true complexities of biological systems. If the uptake and disposition of the chemical substance(s) is adequately described, however, this simplification is desirable because data are often unavailable for many biological processes. A simplified scheme reduces the magnitude of cumulative uncertainty. The adequacy of the model is, therefore, of great importance, and model validation is essential to the use of PBPK models in risk assessment.

PBPK models improve the pharmacokinetic extrapolations used in risk assessments that identify the maximal (i.e., the safe) levels for human exposure to chemical substances (Andersen and Krishnan 1994). PBPK models provide a scientifically sound means to predict the target tissue dose of chemicals in humans who are exposed to environmental levels (for example, levels that might occur at hazardous waste sites) based on the results of studies where doses were higher or were administered in different species. Figure 2-4 shows a conceptualized representation of a PBPK model.

PBPK models for methylene chloride are discussed in this section in terms of their use in risk assessment, tissue dosimetry, and dose, route, and species extrapolations.

2.3.5.1. Methylene Chloride PBPK Model Comparison

The first methylene chloride-specific pharmacokinetic model was developed by Andersen et al. (1987) for use in methylene chloride human risk assessment, formulated by integrating information on mouse physiology, methylene chloride solubility characteristics, and metabolic rate constants to describe the disposition of methylene chloride in target tissues. This model was actually a modification of earlier models developed by Ramsey and Andersen (1984) for generic PBPK analysis of volatile chemicals and by Gargas et al. (1986) for study of the kinetics and metabolism of dihalomethanes. The model could predict the time-course of the parent chemical and the production of metabolites by both GSH and MFO pathways. A more comprehensive model developed by Andersen et al. (1991) is also capable of describing the production of carbon monoxide during oxidative metabolism by the MFO pathway and the production of COHb by CO binding to blood hemoglobin. This model (Andersen et al. 1991) provided a coherent description of experimental data from both rodents and humans. In an earlier draft of this toxicological profile, the Andersen et al. (1991) model was used to conduct route-to-route extrapolation of the inhalation data in humans from the Putz et al. (1979) study to develop an equivalent oral dose that was evaluated as the basis for an acute oral MRL. (However, a different critical study has been selected and has been modeled as described below). The model calculated the peak level (6.37%) of carbon monoxide in the blood that produced a neurological effect (decreased visual and psychomotor performance and auditory function) following exposure to 200 ppm of methylene chloride (Putz et al. 1979). The equivalent concentration of methylene chloride in drinking water that will produce the same neurological effect was 865 mg/L. Using a daily drinking water consumption value of 2 L and an average human body weight of 70 kg, the LOAEL was calculated to be 25 mg/kg/day. Reitz et al. (1988, 1989), Reitz (1990, 1991), and Andersen and Krishnan (1994) used the Andersen et al. (1987) model to make predictions about GSH-mediated target tissue doses associated with tumorigenesis, or the lack thereof, in the long-term inhalation and drinking water bioassays. Using target tissue dosimetry as a concentration surrogate for atmospheric concentrations of methylene chloride, these investigators calculated a unit inhalation risk for methylene chloride that was two orders of magnitude lower than EPA’s, using the same linearized multistage methodology.

Figure 2-4

Conceptual Representation of a Physiologically Based Pharmacokinetic (PBPK) Model for a Hypothetical Chemical Substance. Note: This is a conceptual representation of a physiologically based pharmacokinetic (PBPK) model for a hypothetical chemical substance. (more...)

Dankovic and Bailer (1994) used the models developed by Andersen et al. (1987) and Reitz et al. (1988, 1989; Reitz 1990, 1991) to examine the effects of exercise and intersubject variability on estimation of human dose levels of methylene chloride. Casanova et al. (1992) developed an extended PBPK model for DNA-protein crosslink (DPX) formation in mouse liver, based on the model developed by Andersen et al. (1987).

Reitz et al. (1997) and DeJongh et al. (1998) have expanded earlier models (Andersen et al. 1987, DeJongh and Blaauboer 1996, Ramsey and Andersen 1984) to include simulations of brain concentrations of methylene chloride. Reitz et al. (1997) used their model to make route-to-route dose extrapolations based on brain methylene chloride concentrations. Winneke (1974) had concluded that neurological effects (decreased visual and auditory functions) of acute exposure to methylene chloride were based primarily on the properties of the parent compound, and not on the accumulation of COHb, which had been the conclusion of Putz et al. (1979). Based on the study of Winneke (1974) in humans, an inhalation exposure to 300 ppm methylene chloride for 4 hours was predicted to result in a similar peak brain methylene chloride concentration as an exposure to 565 mg methylene chloride/L drinking water, both exposures correspond to approximately 4 mg/L of brain tissue. Using a daily drinking water consumption value of 2 L and an average human body weight of 70 kg, the LOAEL was calculated to be 16 mg/kg/day. This LOAEL was used as the basis for the acute oral MRL (see Section 2.5). DeJongh et al. (1998) used a rat PBPK model to estimate brain methylene chloride concentrations in rats corresponding to inhalation exposures to the rat 15-minute or 6-hour LC₅₀.

Earlier models have also been extended to simulate humans during pregnancy and lactation. The Reitz et al. (1997) model is based on an earlier model of trichloroethylene pharmacokinetics in the pregnant rat (Fisher et al. 1989) and includes compartments representing mammary tissue, placenta, and the fetus. The Reitz et al. (1997) model has been used to simulate fetal doses of methylene chloride in rats that were exposed to methylene chloride by the inhalation route. Exposure to 1,250 ppm methylene chloride (4,342 mg/m³) for 7 hours/day during gestation days 6–15 was predicted to result in a dose to the fetus of 3.67 mg/L of fetus. The corresponding drinking water intake in the human that would result in the same fetal dose was estimated from the human PBPK model to be 5,000 mg/L drinking water. Fisher et al. (1997) described an extension of the PBPK model of Ramsey and Andersen (1984) which simulates amounts and concentrations of methylene chloride in human breast milk that would result from inhalation exposures. The human lactation model includes a breast milk compartment and parameters for simulating breast milk production and intake of breast milk by nursing infants. Simulations of a workday maternal exposure to methylene chloride at an air concentration of 50 ppm (174 mg/m³) yielded a predicted intake in the infant of 0.213 mg/24 hours.

Advances in modeling methylene chloride pharmacokinetics have also been achieved by linking methylene chloride models to other PBPK models or to other types of simulation or prediction models. For example, OSHA (1997) developed a PBPK model that utilizes a Monte Carlo approach to simulate variability in methylene chloride pharmacokinetics. The model outputs a probability distribution of the amount of methylene chloride metabolized in the lung through the glutathione-S-transferase pathway. Poulin and Krishnan (1999) developed a structure-based PBPK model that uses information on chemical structure to estimate tissue:blood and tissue:air partition coefficients needed to run the model. This approach can be used to model volatile organic chemicals for which empirically-based estimates are not available (Poulin and Krishnan 1999). A PBPK model of methylene chloride has been linked to a PBPK model of toluene (Pelekis and Krishnan 1997). The composite model has been used to simulate the pharmacokinetic and toxicodynamic outcomes (e.g., COHb concentrations) that might result from interactions between methylene chloride and toluene at the level of the mixed function oxidase pathway.

2.4.1. Pharmacokinetic Mechanisms

The physical properties of methylene chloride, particularly its lipophilic nature, high vapor pressure, and high serum/air partition coefficient, suggest that it is likely to be absorbed across the alveolar membranes of the lung, mucosal membranes of the gastrointestinal tract, and the skin by passive diffusion. Once in the body, it is widely distributed, with the greatest amounts accumulating in the more lipophilic tissues; this probably also occurs by passive diffusion.

Andersen et al. (1987), Reitz (1990, 1991), Reitz et al. (1988), and Andersen and Krishnan (1994) have used a PBPK model to predict amounts of methylene chloride metabolism produced (mg equivalents of methylene chloride metabolized per L of liver tissue per day) produced by the two major pathways of methylene chloride metabolism: the MFO pathway and the GSH-mediated pathway. This model identifies the pathway which appears to activate methylene chloride to reactive intermediates but neither characterizes the putative metabolite nor describes the mechanisms of action. In essence, the MFO pathway is oxidative and appears to yield carbon monoxide as well as considerable amounts of carbon dioxide. The glutathione-dependent pathway produces formaldehyde and carbon dioxide, but no carbon monoxide. Potentially reactive intermediates are formed in each of the metabolic pathways for methylene chloride: formyl chloride via the MFO pathway and methylchloroglutathione via the GSH pathway. Distribution of methylene chloride metabolism between these pathways is dose dependent. The MFO pathway is a high-affinity, limited-capacity pathway which saturates at relatively low airborne concentrations (about 200–500 ppm). In contrast, the GSH pathway has a lower affinity for methylene chloride, and does not appear to saturate at experimental concentrations (<5,000 ppm). Thus, at low concentrations, most of the methylene chloride is metabolized by the MFO pathway. As exposure concentrations increase and the MFO pathway saturates, metabolism by the secondary GSH pathway is observed.

Predicted amounts of methylene chloride metabolism produced by each pathway (mg equivalents of methylene chloride metabolized per L of liver tissue per day) are presented in Reitz’s (1990, 1991) analyses. The predicted amounts of MFO metabolites formed in lung and liver tissue are nearly identical for the 2 concentration levels in the mouse inhalation cancer bioassay (2,000 and 4,000 ppm) because saturation of the MFO pathway occurs at administered airborne concentrations of approximately 200–500 ppm. However, mouse liver and lung tumors are consistently higher in the 4,000 ppm exposure group than in the 2,000 ppm group in this bioassay. Therefore, there is no concentration-response relationship between rodent tumor incidence and MFO metabolite levels. The PBPK model also predicts that levels of MFO metabolites produced by long-term drinking water administration of 250 mg/kg/day methylene chloride to B6C3F₁ mice should be similar to the levels of MFO metabolites produced in the inhalation bioassay. However, no statistically significant increases in tumors in the drinking water bioassay were observed; therefore, MFO metabolites would not be predicted to be involved in rodent tumorigenicity by either route of administration. Predicted rates of metabolism of methylene chloride by the GSH pathway correlate more clearly with the induction of lung and liver tumors, or lack thereof, in the two chronic studies. In the inhalation study, predicted levels of GSH metabolites at 4,000 ppm are higher than predicted levels of GSH metabolites at 2,000 ppm. In contrast, predicted levels of GSH metabolites formed in target tissues during drinking water administration of methylene chloride are very low. Thus, the pattern of tumor induction, or lack thereof, in both studies shows a good correlation with the rates of metabolism of methylene chloride by the GSH pathway. Because the dose dependency of GSH metabolite formation is nonlinear at low concentrations, the delivered dose arriving at the target sites cannot be directly extrapolated from very high inhalation concentrations (4,000 ppm) to the very low concentrations (<1 ppm) typical of human exposure.

Casanova et al. (1996) developed an extended PBPK model for DNA-protein crosslink (DPX) formation in mouse liver, based on the model originally developed by Andersen et al. (1987). The extended PBPK model estimated area under the curve for methylene chloride in mouse liver as the independent variable. Tissue-specific yields of DPX were used as the dose surrogate. Estimates were made of the amount of DPX formed in the mouse liver at methylene chloride inhalation concentrations used in the bioassay (i.e., 2,000 and 4,000 ppm) and plotted against liver tumor yields in the mouse. DPX thus served as a concentration surrogate for airborne methylene chloride concentrations. The model assumes that DPX formation is associated with methylene chloride mouse liver tumorigenicity. Because formaldehyde produces DPX, and GSH-mediated metabolism of methylene chloride produces formaldehyde as a reactive intermediate, the authors suggest that formaldehyde is involved in the genotoxic mechanism(s) of action associated with mouse liver tumorigenicity. When excess lifetime cancer risk was estimated using the mouse liver tumor data and two alternate dose measures, DPX and airborne vapor concentrations, the maximum likelihood estimates were similar, but the upper-bound estimate using DPX was two orders of magnitude lower than that using airborne concentrations. Thus, DPX appears to be a reasonable dose surrogate for methylene chloride inhalation exposure. However, most other investigators do not consider DPX formed by the weakly mutagenic activity of formaldehyde to be the putative mechanism of action of methylene chloride-induced liver tumorigenicity (see Section 2.4.2, Mechanisms of Toxicity). Furthermore, human hepatocytes do not appear to form DPX in measurable amounts, as do mouse hepatocytes (Casanova et al. 1996).

2.4.2. Mechanisms of Toxicity

The lung, the blood system, and the nervous system are the major target organs of toxicity associated with exposure to methylene chloride.

Non-neoplastic Mechanisms. In humans, Snyder et al. (1992a, 1992b) have reported headache, chest discomfort, cough, and the presence of alveolar and interstitial infiltrates in the lung as a result of short-term high-concentration vapor exposure to methylene chloride in confined, unventilated rooms or basements. In B6C3F₁ mice exposed to 4,000 ppm of methylene chloride vapors for 6 hours (Foster et al. 1992), the major initial morphological effect observed in mouse lung was acute Clara cell damage. However, the damage appeared to resolve after five consecutive daily exposures to methylene chloride. The appearance and disappearance of the lesion in the Clara cell correlated well with the activity of cytochrome P-450 monooxygenase in the Clara cell, as assessed immunocytochemically (CYP2B1 and CYP2B2) in the whole lung and biochemically in the freshly isolated Clara cell (as determined by ethoxycoumarin O-dealkylation and aldrin epoxidation).

Over 13 weeks (5 days/week) of exposure, the acute Clara cell damage, which developed after a 1-day exposure but resolved after 5 consecutive exposures, reappeared on reexposure after a 2-day weekly break. The severity of the lesion diminished as the study progressed. The authors suggest that the reason for the decrease or disappearance of the lesion was due to an adaptation/tolerance in the Clara cell to methylene chloride that was linked to a marked decrease of methylene chloride metabolism by cytochrome P-450 pathways. Glutathione (GST) activity in the Clara cell either remained unchanged or increased following methylene chloride exposure.

Inhalation and ingestion exposures to methylene chloride result in the production of carbon monoxide associated mainly with metabolism via the MFO pathway. CO binds to hemoglobin, and can cause carboxyhemoglobinemia. In two fatal human cases of methylene chloride poisoning, COHb was elevated to approximately 30% (Manno et al. 1992). Other reports on human and animals show that COHb increases from baselines of 0–2 to 4–15%, under varying regimes of methylene chloride inhalation exposure.

Neurotoxicity resulting from exposure to methylene chloride is believed to be associated with the lipophilic properties of methylene chloride; however, the precise mechanisms of neurotoxicity are not known. Presumably, the methylene chloride enters cell membranes, which in the case of neurons, interferes with signal transmission, in a manner similar to general anesthetics (De Jongh et al. 1998; Sikkema et al. 1995). Neurotoxicity is also assumed to be caused by the hypoxia that results from the formation of COHb.

Cancer. With regard to tumor induction in the rodent lung and liver, methylene chloride is postulated to be activated to an unknown reactive intermediate via metabolism. There are two major metabolic pathways: the MFO pathway, specifically cytochrome P-450 2E1 and glutathione-glutathione S-transferase-mediated (GSH-GST) pathway. The isoenzyme involved in the GSH-GST pathway has been identified as a Θ (theta) class glutathione S-transferase, GSTT1-1, which is present in moderate quantities in the mouse lung, but has been detected only at very low levels in human lung and liver tissue samples (Mainwaring et al. 1996b; Sheratt et al. 1997). These findings suggest that in humans the lung and liver are likely to have little capacity to activate methylene chloride into its reactive metabolites. However, higher than background levels of GSTT1-1 mRNA were detected in a small number of Clara cells and alveolar/bronchiolar ciliated epithelial cells of one human lung sample (out of four) and of GSTT2-2 enzyme in the biliary epithelium of the human liver (Mainwaring et al. 1996b). Thus, it is possible that, in some individuals, these specific cell types may be vulnerable to genotoxic effects from reactive intermediates of methylene chloride metabolism, although the overall risk is likely to be low.

The MFO pathway is oxidative and appears to yield carbon monoxide as well as considerable amounts of carbon dioxide. The glutathione-dependent pathway produces formaldehyde and carbon dioxide, but no carbon monoxide. Potentially reactive intermediates are formed in each of the metabolic pathways for methylene chloride: formyl chloride in the oxidative pathway, and formaldehyde and chloromethyl glutathione in the conjugative pathway. Neither formyl chloride nor the glutathione conjugate of methylene chloride has been isolated or characterized, although Green (1997) reports that their formation is entirely consistent with available information on glutathione-mediated metabolism. Distribution of methylene chloride metabolism between these pathways is dose dependent. The MFO pathway is a highaffinity, limited-capacity pathway which saturates at relatively low atmospheric concentrations (approximately 200–500 ppm). The GSH pathway, in contrast, has a lower affinity for methylene chloride, but does not appear to saturate at experimentally produced concentrations (<5,000 ppm). Thus, the MFO pathway accounts for most of the metabolized methylene chloride at concentrations less than 500 ppm, but as exposure concentrations increase above the MFO saturation level, increases in the amount of methylene chloride metabolized by the secondary GSH pathway are seen (Reitz 1990). The concentration dependency of these two metabolic pathways is consistent with the tumor results obtained in long-term rodent inhalation and drinking water cancer bioassays of methylene chloride and supports the assertion that GSH-mediated metabolism is responsible for methylene chloride-induced tumorigenicity in B6CF₁ mice.

There is no evidence to suggest that methylene chloride is a direct acting carcinogen; the marked species differences in carcinogenicity induced by methylene chloride are not typical behavior of direct-acting compounds. Methylene chloride also does not exhibit the chemical reactivity towards nucleophiles normally associated with direct action (Green 1997). Therefore, metabolic activation is required which interacts in some way with mouse tissues to cause tumors.

A series of bacterial mutagenicity tests has demonstrated that: methylene chloride induction of bacterial mutagenicity is expressed more strongly in Salmonella typhimurium TA 1535 modified to express a mammalian GST Θ class enzyme (NM5004 strain) than in the original strain (Oda et al. 1996); methylene chloride induction of bacterial mutagenicity S. typhimurium strain TA 100 is unaffected by the presence of GST α or π classes (Simula et al. 1993); methylene chloride is less mutagenic in a S. typhimurium GSH-deficient strain (TA100/NG11) as compared to TA 100 (Graves et al. 1994a); and bacterial testing with 3 K12 strains of Escherichia coli showed that methylene chloride (activated by S9 mouse liver fraction) and formaldehyde were mutagenic only in the wild-type E. coli, a characteristic shared with crosslinking agents; these data initially suggested a mutagenic role for metabolically-derived formaldehyde in E. coli (Graves et al. 1994a).

These bacterial assays demonstrated that in in vitro tests, methylene chloride was activated by a Θ class GST enzyme to a bacterial mutagen in S. typhimurium and behaved similarly to formaldehyde in E. coli tester strains. However, in the Chinese Hamster ovary (CHO) assay involving the hypoxanthine-guanine phosphoribosyl transferase (HPRT) gene assay, studies of DNA single strand breaks and DNA-protein crosslinks at mutagenic concentrations of methylene chloride and formaldehyde showed that both these compounds induced DNA single-strand breaks; only formaldehyde induced significant DNA-protein crosslinking (Graves et al. 1996). Similar findings were observed in cultured, freshly isolated mouse hepatocytes (Graves and Green 1996), but not in rat hepatocytes (Graves et al. 1994b, 1995). The authors concluded that, although formaldehyde might play a role in methylene chloride genotoxicity, its weak mutagenicity and the absence of methylene chloride-induced DNA-protein crosslinking in the CHO/HPRT assay suggested that methylene chloride-induced DNA damage and resulting mutations are likely produced by its glutathione conjugate, putatively chloromethylglutathione. Graves and Green (1996) also concluded that these results suggested that the mechanism for methylene chloride tumorigenicity in the mouse liver was likely to be genotoxic and mediated by the GSH pathway. Observed species differences in liver tumorogenicity between the mouse and the rat might result from species differences in the amount of GSH-mediated metabolism induced by methylene chloride exposure.

A series of studies have been conducted to elucidate the precise genetic mechanisms of methylene chloride carcinogenicity. Female B6C3F₁ mice were exposed to vapor concentrations of 2,000–8,000 ppm for 2 years and sacrificed at various intervals to evaluate a number of genotoxic endpoints.

Replicative DNA synthesis in the bronchiolar epithelium, examined by the use of the labeling index (LI), indicated that mice exposed to 2,000 ppm of methylene chloride for 2–26 weeks decreased to 40–60% of controls (Kanno et al. 1993). Mice exposed to 8,000 ppm of methylene chloride have a less dramatic decrease in LI. No pathological changes were found in the exposed lungs. Thus, high-concentration exposure to methylene chloride for up to 26 weeks reduces the cell turnover of bronchiolar cells in these mice; therefore, increased cell proliferation does not appear to be involved in mouse lung tumorogenesis. DNA single-strand breaks were detected in the livers of B6C3F₁ male mice, but not Syrian Golden male hamsters, immediately following inhalation exposure to 2,000–8,000 ppm for 6 hours, but not 2 hours after exposure, suggesting active DNA repair (Graves et al. 1995). The DNA of mouse Clara cells incubated in vitro with methylene chloride was also damaged at high concentrations. Pretreatment of mice with a glutathione depletor prior to inhalation exposure caused a decrease in the amount of DNA damage detected, suggesting a GST-mediated mechanism; similar findings were observed in Clara cells incubated in vitro with methylene chloride and a glutathione depletor.

Devereux et al. (1993) analyzed liver and lung tumor, induced in female B6C3F₁ female mice by inhalation of 2,000 ppm of methylene chloride for 6 hours/day, 5 days/week exposure for up to 104 weeks, for the presence of activated ras proto-oncogenes. In methylene chloride-induced liver tumors, mutations, mainly transversions or transitions in base 1 or base 2, were detected and were similar to those observed for the H-ras gene in spontaneous liver tumors. Mutations were also identified in the lung. The K-ras activation profile in the methylene chloride-induced tumors was not significantly different from the profile in spontaneously-occurring tumors. No other transforming genes were found in the nude mouse tumorogenicity assay. The authors concluded that at present, no transforming genes other than ras genes could be identified in either mouse liver or lung tumors. Based on liver tumor data, they also suggested that methylene chloride may affect the liver by promoting cells with spontaneous lesions.

Hegi et al. (1993) studied allelotypes of 38 methylene chloride-induced lung carcinomas from female B6C3F₁ mice exposed 6 hours/day, 5 days/week for 2 years to 2,000 ppm. The allelotypes were examined for various genotoxic endpoints, and the results were compared with genotoxicity findings in two other reciprocal-cross mouse strains. Throughout the genome, allelic losses occurred infrequently, except for markers on chromosome 4, which were lost in approximately half of the carcinomas. In lung adenomas, chromosome 4 losses were associated with malignant conversion. Methylene chloride-induced liver tumors did not demonstrate chromosome 4 loss, which indicated that this finding was specific for lung carcinomas. Preferential loss of the maternal chromosome 4 was also observed in carcinomas in B6C3F₁ mice. On chromosome 6, an association between K-ras gene activation and allelic imbalances was also found in B6C3F₁ mouse lung tumors. When allelotypes of tumors in mice from two reciprocal cross strains, AC3F₁ and C3AF₁, were examined and compared to the findings in B6C3F₁ mice, one allele of the putative chromosome 4 tumor suppressor gene was shown to be inactivated. Whereas the results in B6C3F₁ mice suggested that nondisjunction events were responsible for the chromosome 4 losses, tumors from both reciprocal-cross mouse strains appeared to show small interstitial deletions in a chromosomal region homologous with a region in human chromosome which is often lost in a variety of tumors, including lung cancers. In human chromosomes, a candidate tumor suppressor gene, MTS1, is located in this region.

In another genotoxic analysis with the same cohort (Hegi et al. 1994), loss of heterozygosity at markers near the p53 gene on chromosome 11 and within the retinoblastoma tumor suppressor gene were examined in methylene chloride-induced liver and lung tumors and compared to spontaneous tumors in control mice. The authors concluded that inactivations of p53 and the retinoblastoma tumor suppressor gene were infrequent events in lung and liver tumorigenesis in mice exposed to methylene chloride.

Replicative DNA synthesis was examined by Kanno et al. (1993) to evaluate the potential role of treatment-induced lung cell proliferation on pulmonary carcinogenicity in female B6C3F₁ mice exposed to 2,000 or 8,000 ppm of methylene chloride for 6 hours/day, 5 days/week for 2 years. By the end of the study, there was a statistically significant increase in lung tumors in exposed animals when compared to controls. Cell proliferation was assessed in the lung after 1, 2, 3, or 4 weeks of inhalation exposure to 2,000 or 8,000 ppm, and after 13 and 26 weeks exposure to 2,000 ppm, as measured by changes in labeling indices (LI). The LI of both bronchiolar epithelium and terminal bronchioles were substantially decreased in mice exposed to 2,000 ppm of methylene chloride for 2–26 weeks. Similar findings, but not as severe, were observed in mice exposed to 8,000 ppm. The decreases in LI were not accompanied by cytotoxicity. The authors concluded that high-concentration exposure to methylene chloride for up to 26 weeks reduces cell proliferation in lung epithelial cells in female B6C3F₁ mice.

Maronpot et al. (1995b) assessed replicative DNA synthesis after 13, 26, 52, and 78 weeks of inhalation exposure by female B6C3F₁ mice to 2,000 ppm of methylene chloride for 6 hours/day, 5 days/week. A statistically significant decrease in the hepatocyte LI was only observed at 13 weeks. In lung epithelial cells, the results were similar to those observed by Kanno et al. (1993). No increases in replicative DNA synthesis were found in liver foci cells or lung parenchymal cells. K-ras gene activation in liver tumors and H-ras gene activation in lung tumors did not differ between methylene chloride-induced tumors and those observed in control animals. The authors concluded that these oncogenes were not involved in mouse tumorogenesis.

DNA-protein crosslinks (DPX) in lung and liver were examined by Casanova et al. (1992, 1996) in male B6C3F₁ mice exposed to 2,000 and 4,000 ppm of methylene chloride for 6 hours/day for 2 days and in male Syrian Golden hamsters exposed to 3,500 ppm. The authors suggested that formaldehyde derived from GSH-mediated methylene chloride metabolism might be forming DPX in mouse liver. Although DPX were detected in mouse liver, there was no evidence of DPX formation in mouse lung, hamster liver, or hamster lung. Additionally, DPX are not formed in measurable amounts in human liver tissue (Casanova et al. 1996). Therefore, the induction of DPX by formaldehyde in mouse liver might be a species-specific, tissue-specific response. A subsequent in vitro study by this laboratory confirmed the absence of DPX formation in response to methylene chloride in human, rat, and hamster hepatocytes, whereas a dose-response was observed in mouse hepatocytes (Casanova et al. 1997). In a different experiment, a dose-response in the formation of RNA-formaldehyde adducts was observed in hepatocytes of all four species (mouse>human>rat>hamster). RNA adduct production was related to the expression of the GSTT1-1 enzyme; the human liver sample lacking GSTT1-1 did not produce RNA adducts.

In vitro tests using two different strains of Salmonella, one with and one without GST expression, revealed that methylene chloride may be mutagenic by at least two pathways (DeMarini et al. 1997). The bacterial strains employed were TA100 and RSJ100; the latter is a derivative of the strain TA1535 (not normally mutagenized by methylene chloride) that contains recombinant rat GSTT1-1. In RSJ100, methylene chloride was mutagenic at moderate doses, and produced a single class of mutation (GC AT transversions) only at the middle C of the target CCC. Although methylene chloride was mutagenic in TA100, it required a much higher dose to match the mutation frequency in RSJ100. Furthermore, it produced a variety of lesions and mutations (predominantly GC TA transversions) at the first and second positions of the CCC target. The implication of these results is that genotoxic effects of methylene chloride will be different, depending on the GSTT1 phenotype. Those who lack the functional gene would only be susceptible to genetic damage from methylene chloride under high exposure levels, whereas carriers might be vulnerable to genetic damage at much lower exposure levels.

According to Maronpot et al. (1995b), the precise mode of action of methylene chloride-induced mouse tumorigenicity appears to be elusive and has not yet been confirmed.

2.4.3. Animal-to-Human Extrapolations

The two major difficulties in applying the results of rodent cancer bioassays to humans involve extrapolation from the high-concentration rodent exposures to the lower concentrations typical of human exposure conditions; and the interspecies extrapolation. The predominant tumors of interest with regard to methylene chloride-induced tumorigenicity are mouse lung and mouse liver. However, species differences among rodent tumorigenic, genotoxic, and morphological responses to methylene chloride, as well as differences between mice and humans, appear to limit the applicability of mouse tumor data to humans, according to some authors. Foster et al. (1992) have suggested that the Clara cell may have a role in mouse lung tumor induction; there is a substantially higher number of Clara cells in the mouse than in other rodent species or in humans. GSH-mediated activation of the proximate carcinogenic agent of methylene chloride in mice has been associated with a specific isoenzyme, the Θ (theta) class GSTs; Sheratt et al. (1997) have shown that this isozyme is expressed at very low levels in human pulmonary cells in vitro, suggesting that in humans, the lung has little capacity to activate methylene chloride into biologically reactive intermediates. In vitro studies (e.g., Graves et al. 1995) have shown species differences in the ability to induce DNA single-strand breaks in mouse, rat, hamster, and human cells that are compatible with the known rodent carcinogenicity, or lack thereof, in chronic cancer bioassays. These in vitro studies suggested that humans are unlikely to be more susceptible than rodents to methylene chloride-induced liver cancer.

In a comparison of rates of methylene chloride metabolism by each of the two major pathways in four species, the mouse, rat, hamster, and human, Green (1997) presented in vivo and in vitro evidence that the MFO pathway metabolic rates are similar among all four species and saturated at concentrations of 500 ppm or above. In contrast, the GSH-mediated metabolism is linear over the concentration range studied; it is also the major metabolic pathway for methylene chloride in mice at concentration levels used in cancer inhalation bioassays. Furthermore, the activity in mouse tissues is more than an order of magnitude greater than the activity in rat tissues. Hamster and human tissues show metabolic rates for this pathway that are even lower than those found in the rat. These findings are corroborated by the localization study of Mainwaring et al. (1996b), which found higher levels of GSTT1-1 mRNA and protein in the liver and lung of mice than in rats or humans. In mechanistic terms, Green (1997) concludes that these studies demonstrate that differences in glutathione-mediated metabolism among mice, rats, hamsters, and humans are correlated with differences in carcinogenicity of methylene chloride. This is supported by the levels of glutathione that have been detected in the livers of the different species. The hepatic concentrations of GSH (in nmol/10⁶ cells) were approximately 129 in B6C3F₁ mice (Ruch et al. 1989), 50 in Sprague-Dawley rats (Jones et al. 1978), 22 in Syrian Golden hamsters, and 21 in humans (Steinmetz et al. 1988). Thus, the mouse is the most sensitive species to metabolic activation of methylene chloride by glutathione metabolism, whereas humans appear to be the least sensitive.

However, other evidence suggests that methylene chloride may be potentially carcinogenic in humans. Mainwaring et al. (1996b) analysis of the distribution of mRNA and protein for GSTT1-1 and GSTT2-2 in the analysis of the liver and lungs of mice, rats, and humans corroborated the finding of Sheratt et al. (1997) that the overall levels in human tissues are much lower than in those of mice. However, the immunodetection of localized high concentrations of GSTT2-2 enzyme in human bile-duct epithelial cells is potentially significant, considering the increased incidence of biliary cancer following chronic exposure to methylene chloride as reported by Lanes et al. (1990). On the other hand, the GSTT2-2 antibody did not localize to the nucleus of human biliary epithelial cells, which would tend to reduce the potential genotoxic effect in humans. Although Mainwaring et al. (1996b) found that rates of metabolism of methylene chloride were very low in human lung, they also detected higher than background amounts of GSTT1-1 mRNA in a few Clara cells and ciliated cells of the alveolar/ bronchiolar junction of the lung in one human sample out of four. Therefore, despite the general low level of GSTT1-1and GSTT2-2 in human tissue, it is possible that in some individuals, specific cell types within the human liver (bile duct) and lung might produce genotoxic reactive intermediates as a result of methylene chloride metabolism. Based on GSTT enzyme distributions and concentrations, the carcinogenic risk from methylene chloride in humans appears to be low as in rats rather than high as in mice.

The mouse model has been employed in assessment of excess lifetime cancer risks in humans from inhalation exposure to methylene chloride by EPA and others. Recent data, both in vitro and in vivo, strongly indicate that the higher sensitivity of the mouse, relative to humans, must be taken into consideration if mouse data are to be used to estimate potential human health risks.

Minimal Risk Levels for Methylene Chloride

Systemic Effects

Respiratory Effects. In one workplace accident, acute inhalation exposure to methylene chloride (probably at a high concentration) resulted in bilateral pulmonary congestion with focal hemorrhage (Winek et al. 1981). Less severe respiratory symptoms (cough, breathlessness, chest tightness) were reported in occupational exposure incidents (concentrations unknown; Bakinson and Jones 1985). Exposure to 18–1,200 mg/m³ (5–340 ppm; 8-hour TWA) resulted in irritation of the respiratory tract in one occupational study (Anundi et al. 1993). However, a study in humans found no effect on pulmonary function following repeated exposures to methylene chloride vapors (up to 500 ppm) (Stewart et al. 1972). Studies in animals corroborated the severe pulmonary effects (congestion, edema, inflammation) of acute or intermediate exposures at high concentrations (Heppel et al. 1944; NTP 1986). Resolution of acute Clara cell damage in mice exposed to 4,000 ppm of methylene chloride was correlated with cytochrome P-450 activity in the lung (Foster et al. 1992). No information was found on the respiratory effects of low levels of methylene chloride in humans near hazardous waste sites or industrial urban areas or in animals.

Cardiovascular Effects. Myocardial infarction occurred in one case of acute inhalation occupational exposure (Stewart and Hake 1976). However, occupational studies in humans did not find any association between exposure to methylene chloride at 75–475 ppm and cardiac abnormalities (Cherry et al. 1981) or excess mortality due to ischemic heart disease (Hearne et al. 1990; Ott et al. 1983b). Further, a cross-sectional study of workers showed no excess of electrocardiographic abnormalities among those exposed to methylene chloride (Ott et al. 1983c). Another study in humans did not reveal effects on cardiovascular functions at concentrations up to 500 ppm (NIOSH 1974). These findings suggest the cardiovascular system is not a sensitive target for exposure to methylene chloride in humans. One study in mice reported atrioventricular block following acute inhalation exposure to very high levels of methylene chloride (>200,000 ppm) (Aviado and Belej 1974). The relevance of this finding is limited since the exposure level was so high.

In an experiment with rats, using an ischemia-reperfusion model, intravenous methylene chloride infusion (leading to blood concentrations of <0.1 mg/mL) markedly increased the atrioventricular block during the reperfusion phase (Scholz et al. 1991). From these results, the authors concluded that the initial coma resulting from methylene chloride-induced poisoning is likely to result not only from anesthetic effects, but also from sudden onset of cardiac arrhythmias.

Gastrointestinal Effects. Nausea and vomiting were reported in 13 out of 33 occupational cases of acute inhalation exposure to methylene chloride in the United Kingdom (Bakinson and Jones 1985); exposure levels were not reported in these cases. In mice exposed to 4,000 ppm of methylene chloride for 2 years, dilatation of the stomach was reported (NTP 1986). In humans attempting suicide, ingestion of a single oral dose of Nitromors, a paint remover solvent containing 75–80% methylene chloride, resulted in severe corrosion and ulceration of the gastrointestinal tract (Hughes and Tracey 1993), peritonitis and septicemia occurred in the fatal case, whereas intestinal diverticuli developed during recovery in the another case (Roberts and Marshall 1976). No other studies were located regarding gastrointestinal effects in humans or animals after exposure to methylene chloride.

Hematological Effects. Metabolism of methylene chloride results in excess carbon monoxide and increases in COHb, which contributes to hypoxia (Tomaszewski 1998). Blood COHb concentrations were about 30% higher than normal in 2 lethal cases in which workers were estimated to be exposed to extremely high concentrations (up to 168,000 ppm) of methylene chloride in a confined work space (Manno et al. 1992). Several hours after an adult woman ingested a fatal oral dose of Nitromors, a paint remover containing 75–80% methylene chloride, her COHb level was 9%. In all three fatal exposures, cause of death was not associated with elevated COHb. In autoworkers who were exposed to methylene chloride dermally and by inhalation (3–154 ppm), blood COHb measurements taken within 24 hours of exposure were 1.2–11% for nonsmokers and 7.3 and 17.3% for two smokers (Kelly 1988). One-day occupational exposures to methylene chloride at levels below ACGIH standard (50 ppm, 8-hour TWA) produced small increases in COHb levels in both nonsmoking and smoking adults (Soden et al. 1996); additional daily cumulative exposure to methylene chloride did not further increase COHb levels. Increases in blood COHb levels between 5 and 6.8% were measured in nonsmoking volunteers exposed to methylene chloride at concentrations up to 200 ppm for 4 to 7.5 hours (DiVincenzo and Kaplan 1981; Putz et al. 1979).

Other studies have reported increases in red cell count, hemoglobin, and hematocrit in women, but not in men, occupationally exposed to concentrations of up to 475 ppm during an 8-hour workday (Ott et al. 1983d); these effects were judged by the authors to be suggestive of compensatory hematopoietic effects. Similar findings were not observed in rodents chronically exposed to methylene chloride by inhalation at concentrations up to 3,500 ppm (Burek et al. 1984), but were observed in rodents exposed orally for 3 months at 480 mg/kg/day (Kirschman et al. 1986), or for 2 years at 55–249 mg/kg/day (Serota et al. 1986a, 1986b). Intravascular hemolysis was reported in the case of a man who attempted suicide by ingesting the paint remover, Nitromors (Roberts and Marshall 1976) and in a high-dose acute gavage study in rats (Marzotko and Pankow 1987).

Hepatic Effects. Human data are limited on the effects of methylene chloride on the liver. A slight exposure-related increase in serum bilirubin (but not at levels of clinical significance) was observed in workers with exposure up to an average of 475 ppm of methylene chloride, but serum levels of hepatic enzymes (e.g., aspartate aminotransferase, alanine aminotransferase, lactate dehydrogenase, and alkaline phosphatase) were not elevated (Ott et al. 1983a); another occupational study found no exposure-related changes in hepatic enzymes (Anundi et al. 1993). In another study, methylene chloride vapors (up to 500 ppm) did not affect comparable serum enzyme activity in volunteers (Stewart et al. 1972). However, the liver appears to be a major target organ following methylene chloride exposure in animals, particularly at high exposure levels (>5,000 ppm; Heppel et al. 1944). Histomorphological and biochemical changes of the liver occur following acute inhalation (6 hours to 7 days) at high concentration levels (5,200 ppm) (Morris et al. 1979), while fatty changes and biochemical alterations (altered cytochrome P-450 levels) were also observed at lower concentrations (100 ppm) for continuous, 24-hour intermediate-duration exposure (100 days) (Haun et al. 1972; Kjellstrand et al. 1986; Weinstein and Diamond 1972). Cytoplasmic vacuolization was observed in rats at 25 ppm (Haun et al. 1972). Using these data, ATSDR derived an intermediate inhalation MRL of 0.3 ppm, as calculated in Table 2-1. Exposure to 1,000–4,000 ppm for 2 years resulted in an increased incidence of hemosiderosis and focal hepatic necrosis in rats (NTP 1986). An increased incidence of fatty changes occurred following chronic exposure at 500 ppm, but not at 200 ppm (Nitschke et al. 1988a); fatty changes were reversible when exposure ceased. Using data from the Nitschke et al. (1988a) study, ATSDR derived a chronic inhalation MRL of 0.3 ppm, as shown in Table 2-1. Necrosis or fatty changes in the liver have been observed in rats given high oral doses (>1,000 mg/kg/day) of methylene chloride (Kirschmann et al. 1986; Ugazio et al. 1973). Chronic ingestion of methylene chloride in drinking water has been associated with fatty changes in rats at 55 mg/kg/day or greater and in mice at 175 mg/kg/day or greater, but not at 6 mg/kg/day (Serota et al. 1986a, 1986b). Based on this value (6 mg/kg/day), a chronic oral MRL of 0.06 mg/kg/day was calculated as described in Table 2-2.

Endocrine Effects. No relevant information was located regarding endocrine effects in human or animals associated with exposure to methylene chloride.

Renal Effects. Kidney function was not altered in humans repeatedly exposed to methylene chloride vapors (up to 500 ppm) for 6 weeks (Stewart et al. 1972) and no alterations in urinary microglobulins or N-acetyl-beta-glucosaminidase were detected in workers chronically exposed to methylene chloride (Anundi et al. 1993). In rats, nonspecific renal tubular and degenerative changes occurred after continuous intermediate-duration exposure to methylene chloride vapors (100–5,000 ppm) (Haun et al. 1972; MacEwen et al. 1972) or chronic exposure to 4,000 ppm (NTP 1986). Similar renal changes occurred in dogs exposed to 1,000 ppm for 4 weeks (MacEwen et al. 1972). There were no studies reporting renal effects following oral exposure to methylene chloride in humans. In rat oral studies using methylene chloride at doses >1,300 mg/kg/day, a single dose inhibited diuresis (Marzotko and Pankow 1987) and treatment for 3 months increased kidney weights in females (Kirschman et al. 1986). At $166 mg/kg/day, methylene chloride lowered the pH of urine in rats of both sexes (Kirschman et al. 1986). There are no data that would enable an assessment of the potential for renal effects in humans living near hazardous waste sites.

Dermal Effects. No studies were located regarding dermal effects in human or animals associated with inhalation or oral exposure to methylene chloride. However, in some occupational accidents, direct contact with methylene chloride has resulted in second or third degree chemical burns within 30 minutes (Hall and Rumack 1990; Wells and Waldron 1984; Winek et al. 1981).

Ocular Effects. No studies were located on ocular effects in humans by oral or dermal exposure. However, repeated direct exposure to methylene chloride vapors (up to 500 ppm) caused mild irritation to the eyes of volunteers (Stewart et al. 1972). In one occupational accident, direct contact with the liquid (duration unspecified) resulted in severe corneal burns (Hall and Rumack 1990). In animals, small increases in corneal thickness and intraocular tension were reported after exposure to vapors of 490 ppm of methylene chloride or greater, but effects were reversible within 2 days after exposure ceased (Ballantyne et al. 1976). Inflammation of the conjunctivas and eyelids as well as increases in corneal thickness and intraocular tension were observed following direct contact of methylene chloride (0.1 mL) with the eyes of rabbits. Effects were reversible within 3–9 days (Ballantyne et al. 1976).

Immunological and Lymphoreticular Effects. No studies were located regarding immunological effects in humans after inhalation, oral, or dermal exposure. Splenic atrophy was evident in dogs that died following continuous intermediate-duration exposure to vapors of methylene chloride (1,000 ppm) (MacEwen et al. 1972), but the incidence did not increase over control levels in rats and mice chronically, but discontinuously, exposed at a level of 4,000 ppm or less (NTP 1986). Splenic fibrosis was observed in rats after chronic inhalation exposure at $1,000 ppm methylene chloride (Mennear et al. 1988), but was not observed in a recent intermediate-duration study in rats exposed to 5,187 ppm methylene choride (Halogenated Solvent Industry Alliance, Inc. (2000). This latter study also found that the IgM antibody response to SRBC was not significantly altered in exposed rats relative to controls. Due to the very limited and inconsistent nature of the database, further studies are necessary before conclusions can be drawn about the relevance of these findings to human health.

Neurological Effects. Studies in humans and animals indicate the central nervous system is an important target for methylene chloride; in the case of acute exposures, anesthetic responses, which subsided once exposure ceased, have been reported in humans and animals (Bakinson and Jones 1985; Hall and Rumack 1990; Heppel et al. 1944; Snyder et al. 1992a, 1992b). Neurological effects in humans have included headache, dizziness, confusion, memory loss, intoxication, incoordination, paresthesia, and in severe cases, unconsciousness and seizures (Bakinson and Jones 1985; Hall and Rumack 1990; Kelly 1988). Degraded performance in various psychomotor tasks was reported in humans acutely exposed to methylene chloride (200 ppm or greater) in experimental studies (Fodor and Winneke 1971; Putz et al. 1979; Stewart et al. 1972; Winneke 1974). Impaired performance during exposure to methylene chloride was associated with a rise in blood levels COHb to about 5%, which occurred after 3 hours of exposure at 200 ppm (Putz et al. 1979). In more specific measures of cortical function, Winneke (1974) attributed impaired performance in volunteers following 3–4 hours of exposure to 300 ppm of methylene chloride to the properties of the parent compound; exposure to 50–100 ppm of carbon monoxide did not generate the same adverse effects. Based on this value (300 ppm), an acute inhalation MRL of 0.6 ppm was calculated as described in Table 2-1 (Winneke 1974). Studies in factory workers chronically exposed to methylene chloride revealed no evidence of neurological or behavioral impairment at exposure levels of 75–100 ppm (Cherry et al. 1981). There was a reduction in test scores pertaining to mood changes in workers in 1 of 3 rapid rotation shifts exposed to vapors of methylene chloride (28–173 ppm), but no effects were observed on performance as determined by digit symbol substitution scores (Cherry et al. 1983). Mood changes are very subjective and these results may reflect other causes.

Neurological effects have been noted in animals following exposure to methylene chloride independent of the route. Narcotic effects of methylene chloride (incoordination, gait disturbances, reduced activity, somnolence) were observed in monkeys, rabbits, rats and guinea pigs following acute exposure at 10,000 ppm (Heppel and Neal 1944; Heppel et al. 1944); dogs exposed at this level became uncoordinated, then excited and hyperactive (Heppel et al. 1944). These effects wore off when exposure ceased. Studies in rats and gerbils suggest that methylene chloride induces regional alterations in DNA concentrations, amino acids levels, and enzyme activities in the brain. There was a decrease in succinic dehydrogenase activity in the cerebellum in rats exposed to vapors of methylene chloride at concentrations of 500 ppm or greater and signs of increased protein breakdown in the cerebrum at 1,000 ppm (Savolainen et al. 1981). The DNA concentration decreased in the hippocampus and cerebellum in gerbils exposed to 210 ppm of methylene chloride, indicating decreased cell density in these brain regions, probably due to cell loss (Karlsson et al. 1987; Rosengren et al. 1986). Levels of aminobutyric acid increased in the posterior cerebellar vermis of gerbils exposed to 210 ppm of methylene chloride; however, the significance of this finding is uncertain (Briving et al. 1986). On the other hand, studies in rats exposed to methylene chloride (2,000 ppm) for 13 weeks revealed that the compound did not cause clinical, postural, sensory, locomotor, evoked potential, or pathological effects (Mattsson et al. 1990). However, when rats were given an intraperitoneal injection of methylene chloride, 115 mg/kg was sufficient to alter the pattern of flash evoked potentials (Herr and Boyes 1997); in this study, the responses generated in methylene chloride-treated rats were different from those generated by other solvents, leading the authors to conclude that lipid solubility alone was insufficient to predict the neurotoxic effects. The mechanism by which methylene chloride exerts its effects on the central nervous system is not clear. Anaesthetic effects of the parent compound and hypoxic effects of its metabolite carbon monoxide may both contribute to the observed symptoms.

Reproductive Effects. Data on reproductive toxicity in humans are limited; there are no studies involving oral exposure. One group of cases reported genital pain, testicular atrophy, recent infertility, and low/abnormal sperm counts in workers who inhaled vapors of methylene chloride and who had direct contact with the liquid on the job for more than a year (Kelly 1988). Exposure to methylene chloride in this group was confirmed by blood COHb analysis and the presence of typical neurological deficits. A retrospective study of pregnancy outcomes among Finnish pharmaceutical workers reported a slightly increased risk of spontaneous abortion (not quite statistically significant) associated with inhalation exposure to methylene chloride (Taskinen et al. 1986). No animal reproductive studies were conducted for oral or dermal exposures to methylene chloride, and the inhalation studies were negative. Methylene chloride did not adversely affect fertility and litter size in rats that inhaled methylene chloride vapors at concentrations up to 1,500 ppm or less for two generations (Nitschke et al. 1988b). After intermediateduration exposure (6 weeks) to vapors of methylene chloride (200 ppm or less), there were no microscopic lesions in testes of rats (Raje et al. 1988). Uterine, ovarian, and testicular atrophy has been noted in rats and mice chronically exposed for 2 years to 4,000 ppm of methylene chloride, but this was reported to be secondary to malignant neoplasms (NTP 1986). Based on these data, methylene chloride does not appear to pose a hazard to human reproduction, except at very high exposure levels.

Developmental Effects. There are few studies addressing developmental effects in humans. A retrospective study of pregnancy outcomes among Finnish pharmaceutical workers reported a slightly increased risk of spontaneous abortion associated with inhalation exposure to methylene chloride (Taskinen et al. 1986). A study examining the effect of low environmental concentrations of methylene chloride (<0.01 ppm) on over 90,000 births found no significant effect on birth weight (Bell et al. 1991). Studies in rats demonstrated that methylene chloride crosses the placenta and that metabolism of methylene chloride by the maternal liver elevates the blood levels of carbon monoxide in the fetus (Anders and Sunram 1982). Animal studies demonstrated that inhalation of methylene chloride vapors at concentrations of 1,250 ppm produced minor skeletal variants: delayed ossification of sternebrae in rats and extra center of ossification in the sternum of mice and rats (Schwetz et al. 1975). Fetal weight was reduced and behavioral changes occurred in rat pups following exposure of dams to 4,500 ppm of methylene chloride (Bornschein et al. 1980; Hardin and Manson 1980). The significance of these observations is uncertain since each of the three studies used only one concentration level and the observed effects occurred at maternally toxic concentrations. Growth (yolk sac blood vessel growth, body length, protein concentration) and development (somite addition) were significantly retarded in rat embryos cultured in the presence of methylene chloride at $0.5 mg/mL, but not at 0.2 mg/mL (Brown-Woodman et al. 1998); the effect on the yolk sac blood vessels may have interfered with the ability of the embryos to take up nutrients. Methylene chloride had no effect on heart function in these embryos. Although fetal body weights were decreased in these animal studies, the absence of other fetotoxic effects, embryolethality, or major malformations, suggests that methylene chloride is not likely to cause developmental effects or behavioral changes at levels normally encountered in the environment; with current standards and procedures, workplace exposure of pregnant women is unlikely to be hazardous to the fetus. The reduction in the rate of spontaneous abortions observed during the later years of the Finnish study was attributed partly to improved industrial hygiene (Taskinen et al. 1986).

Genotoxic Effects. No studies were located regarding the genotoxic effects of methylene chloride in humans after inhalation, oral, or dermal exposure. In vitro results were mixed in bacterial assays and in tests employing mammalian cells (Table 2-3). Methylene chloride has caused chromosomal aberrations in some studies, but not in others. Given the evidence of in vitro clastogenicity and its negative results in unscheduled DNA-synthesis and DNA-binding studies, methylene chloride may be a weak mutagen in mammalian systems. The chemical has been evaluated in several in vivo assay systems in animals to assess its potential to induce gene mutation and cause chromosomal aberrations or DNA damage and repair. Many studies were negative and some were positive (Table 2-4). Tissue-specific genetic damage in mice following exposure to methylene chloride, suggests that the ability of tissues to metabolize methylene chloride may determine its genotoxicity (Sasaki et al. 1998). Specific polymorphisms in the metabolizing enzymes are associated with genotoxicity. In vitro, low concentrations of methylene chloride were more mutagenic when a functional GSTT1-1 gene was present (DeMarini et al. 1997).

Human erythrocytes expressing the GSTT1-null phenotype had a higher incidence of sister chromatid exchange than normal following exposure to methylene chloride (Hallier et al. 1994). Similarly, expression of the DraI DD mutant of CYP2E1 was associated with an increase in bleomycin-induced single-strand breaks in human lymphocyte DNA (El-Zein et al. 1997a); possibly methylene chloride exposure would induce similar genetic damage in individuals with that genotype. Additional details are presented in Section 2.10, Populations that are Unusually Susceptible.

Table 2-3

Genotoxicity of Methylene Chloride In Vitro.

Table 2-4

Genotoxicity of Methylene Chloride In Vivo.

Cancer. A significant increase in bile-duct cancer was observed within a cohort of workers who had been exposed to methylene chloride (at 1,700 ppm, 8-hour TWA) for up to 28 years (Lanes et al. 1990). Epidemiology studies have not revealed a causal relationship between deaths due to cancer and occupational exposure to methylene chloride at lower levels (475 ppm or less) (Friedlander et al. 1978; Hearne et al. 1987, 1990; Ott et al. 1983b). It should be noted that these latter studies had limited power to detect very small increases in cancer and are not sufficient to rule out a carcinogenic potential of methylene chloride. Studies in animals exposed via inhalation have demonstrated that methylene chloride can increase the incidence of naturally-occurring tumors. When administered by inhalation, methylene chloride (2,000 ppm or greater) increased the incidence of alveolar/bronchiolar neoplasms in mice of both sexes (NTP 1986). Concentrations of 500 ppm or greater of methylene chloride increased the incidence of benign mammary gland tumors per animal in females and male rats (Burek et al. 1984; Nitschke et al. 1988a; NTP 1986). The incidence of liver tumors increased over concurrent control levels in male mice and female rats administered methylene chloride (50–250 mg/kg/day) in drinking water; however, the incidence of lesions in treated groups were within the historical range of control values and showed no dose response (Serota et al. 1986a, 1986b). The results of recent toxicokinetics studies suggest that the parent compound and/or reactive metabolites produced by the GST pathway are the source of methylene chloride-induced tumor increases. Based on these findings, the EPA has ranked methylene chloride as a Group B2 carcinogen (probable human carcinogen). The EPA has calculated that an upper limit 10⁻⁶ risk level corresponds to 0.0006 ppm and 0.001 mg/kg/day for inhalation and oral 70-year continuous exposures, respectively. OSHA (1997) determined that methylene chloride is a potential occupational carcinogen, based on rodent inhalation studies and PBPK modeling, and established an inhalation unit risk of 3.62 x 10⁻³ (μg/m³)⁻¹ for occupational exposures.

The Department of Health and Human Services (NTP 1999) has determined that methylene chloride may reasonably be anticipated to be a human carcinogen. IARC (1987) has classified methylene chloride in Group 2B (possibly carcinogenic to humans). EPA (IRIS 1999) has determined that methylene chloride is a probable human carcinogen. Section 2.10 discusses polymorphisms in genes that metabolize methylene chloride that may be associated with an increased risk of cancer.

2.8.1. Biomarkers Used to Identify or Quantify Exposure to Methylene Chloride

Measurements of parent methylene chloride and its metabolites in expired air, blood, and urine have been used as indicators of exposure. Elimination of methylene chloride from the body occurs primarily through pulmonary excretion; approximately 70–75% is excreted unchanged at concentrations from 50 to 200 ppm (DiVincenzo and Kaplan 1981). Pulmonary excretion was rapid during the first hour, then began to decline as steady state was approached. By 7 hours postexposure, expired air contained less than 1 ppm of methylene chloride for exposures between 50 and 150 ppm. The 7-hour values for an exposure of 200 ppm were twice that for the other concentrations. At 16 hours, negligible levels of methylene chloride were detected in all concentration groups. Therefore, measurements of methylene chloride in expired air as a biomarker are useful only if they occur within 6–8 hours of the most recent exposure (DiVincenzo and Kaplan 1981).

Methylene chloride can be detected in blood. Because it is cleared from blood very rapidly, this method is only useful for monitoring recent exposures. A plasma half-life of inhaled methylene chloride in humans is estimated to be 40 minutes (DiVincenzo et al. 1972).

Methylene chloride has been detected in adipose tissue in humans, but animal studies suggest that methylene chloride is cleared so rapidly (90% decrease in 2 hours), that it is useful for measuring only recent exposures (Carlsson and Hultengren 1975). Methylene chloride was detected in the breast milk of mothers living in urban industrial areas, but the route of exposure was not analyzed (EPA 1980e; Pellizzari et al. 1982).

Methylene chloride is also excreted in urine. Humans exposed to 100 ppm of methylene chloride vapor for 2 hours averaged 26.6 μg methylene chloride in the urine within 24 hours after exposure and 85.5 μg at exposure levels of 200 ppm (DiVincenzo et al. 1972).

Levels of COHb in the blood may also be used as an indicator of exposure to methylene chloride. Levels of COHb are concentration- and time-dependent (Stewart et al. 1972). Human subjects exposed to concentrations of 500 ppm or less for 1 hour experienced elevation in COHb levels (1–4%). These levels rose to an average of 10% saturation within 1 hour after exposure to higher concentrations (1,000 ppm for 2 hours) (Stewart et al. 1972). Exposure to concentrations as low as 100 ppm for 7.5 hours or 200 ppm for 4 hours resulted in COHb elevation above 5% in nonsmokers (Putz et al. 1979; Stewart et al. 1972). Levels of COHb can remain elevated above preexposure levels for more than 21 hours post exposure (Stewart et al. 1972). DiVincenzo and Kaplan (1981) showed recovery to baseline each day for 5 consecutive days. COHb saturation levels were lower than in Stewart et al. (1972). Different methods of COHb detection were used by the authors. Although COHb levels can be used as an indicator of exposure to methylene chloride, this biomarker is not specific. Exposure to other sources of carbon monoxide, such as tobacco smoke, incomplete organic fuel combustion, and automobile exhaust, will also increase COHb levels.

Ghittori et al. (1993) examined several methods for conducting biological monitoring of workers exposed occupationally to methylene chloride. Twenty males (12 smokers and 8 nonsmokers), employed at different jobs in a pharmaceutical factory where methylene chloride was used to wash gelatin capsules, wore a personal passive dosimeter in the respiratory zone during half of work days (4 hours) in order to measure the weighted mean-inspired environmental concentration of methylene chloride. Tetrachloroethylene was also present in the work environment. Immediately after the end of the exposure, a urine sample was collected using gas-tight samplers. Carbon monoxide (CO) was determined at the end of the shift using a portable instrument.

Ambient exposures ranged from 49 to 168 ppm. No significant correlation was observed among the CO of all subjects and the concentration of methylene chloride in ambient air. When those workers who smoked were removed from analysis, a correlation between the methylene chloride concentration in air and the CO concentration in alveolar air was found ® = 0.87). Significant linear correlation was found between the environmental concentration of methylene chloride in the breathing zone and methylene chloride concentration in urine ® = 0.9 uncorrected, 0.72 corrected for creatinine). The authors suggest that measuring methylene chloride in the urine was a more useful measure of exposure, since unlike COHb in the blood or CO in exhaled breath, it was unaffected by smoking.

Concentrations of CO in exhaled air studied in other groups of volunteers were shown to reach a maximum at 1–2 hours after exposure and were also directly proportional to the magnitude of exposure both during and after exposure. These findings are consistent with those studies.

The authors concluded that the results indicated that methylene chloride urinary concentration could be used as a possible biological index to evaluate methylene chloride air exposure, according to what has been observed for other solvents. However, at high ambient exposures (i.e., those exceeding 5–7 times the threshold limit value-time weighted average [TLV-TWA] of 49 ppm), the use of methylene chloride as a biological index might be problematic because exposure to very high concentrations gives disproportionately less COHb in the blood and more unmetabolized methylene chloride is produced.

2.8.2. Biomarkers Used to Characterize Effects Caused by Methylene Chloride

As discussed in Section 2.2, the effects that are most often observed in humans exposed to methylene chloride vapors are central nervous system depression and behavioral effects. Clinical signs and symptoms which may be monitored include irritability, narcosis, and fatigue. Impairment of visual, auditory, and psychomotor functions can also be evaluated to detect early effects on the central nervous system. Since these effects also occur following exposure to numerous other chemicals, they are not specific for methylene chloride exposure and evaluation is often subjective.

Honma and Suda (1997) have investigated the effects of a single intraperitoneal administration of several short-chain chlorinated compounds, including methylene chloride, on lipoproteins in plasma and liver in male Fischer F344 rats. Changes in lipoproteins caused by these solvents were compared with hepatotoxicity markers such as GPT (ALT). Following administration of methylene chloride in olive oil, concentrations of lipoproteins (VLDL, LDL, HDL), triglyceride, cholesterol, and GPT in plasma were determined, as were changes in liver weight and amounts of triglyceride and glutathione in liver tissue. For methylene chloride, peaks of changes in these endpoints were observed at 8 or 19 hours following compound administration. The dose dependency of these changes were investigated after intraperitoneal dosing of 300 and 1,000 mg/kg methylene chloride. HDL decreased significantly at a dose of 300 mg/kg, whereas a marked increase in LDL occurred at 1,000 mg/kg. GPT also increased significantly at a dose of 1,000 mg/kg. The authors conclude that changes in some lipoproteins in plasma occurred at lower doses than those causing elevation of GPT activity; therefore, these lipoproteins may be able to serve as sensitive and simple markers of adverse liver effects (i.e., biomarkers of effects). However, further investigation is necessary to determine whether these endpoints can serve as biomarkers of effects in humans when exposure is via inhalation and confounding variables are be numerous.

No other biomarkers have been identified to characterize effects associated with exposure to methylene chloride.

For more information on biomarkers for renal and hepatic effects of chemicals see ATSDR/CDC Subcommittee Report on Biological Indicators of Organ Damage (1990) and for information on biomarkers for neurological effects see OTA (1990).

2.11.1. Reducing Peak Absorption Following Exposure

Human exposure to methylene chloride may occur by inhalation, ingestion, or by dermal contact. Mitigation approaches to reduce absorption of methylene chloride have included general recommendations of separating contaminated food, water, air, or clothing from the exposed individual. Externally, exposed eyes and skin are flushed with a clean neutral solution such as water or normal saline. If the victim is alert and oriented and has an intact gag reflex, water or milk may be administered after ingestion of small amount of methylene chloride to wash residual chemical through the esophagus and dilute the contents in the stomach. In cases where persons have ingested more than several swallows, emesis can be induced with ipecac syrup after it has been determined that the victim is alert and oriented and precautions have been taken to protect the respiratory tract from aspiration of gastric contents. Once the individual is placed in the care of a health professional, gastric lavage can be performed within 1 hour of the exposure. Activated charcoal and cathartics are frequently recommended; however, no data exist to support their efficacy (Ellenhorn 1997).

Once methylene chloride has been inhaled or ingested, it is readily absorbed through the lungs or gastrointestinal tract. Dermal exposure also results in absorption, although at a slower rate than the other exposure routes and some references question whether the amount absorbed would be sufficient to cause systemic toxicity (Ellenhorn 1997; Stewart and Dodd 1964). Once absorbed, methylene chloride is rapidly metabolized in the liver in part to carbon monoxide. Because of the affinity of hemoglobin for carbon monoxide, COHb levels in the blood will increase, but free circulating carbon monoxide will not increase.

2.11.2. Reducing Body Burden

Inhalation data demonstrate that a portion of the absorbed methylene chloride is stored in fat tissue as evidenced by continued increase in COHb levels. It is likely to occur only with long-term exposure and at high concentrations (DiVincenzo et al. 1972). Investigations in humans following oral exposure to methylene chloride have failed to detect significant retention in fat or other tissue stores (Angelo et al. 1986a, 1986b; Ellenhorn 1997).

Following absorption, methylene chloride is distributed mainly to the liver, brain and subcutaneous adipose tissue (Carlsson and Hultengren 1975). The liver is the primary site of metabolism, although additional transformation occurs in the lungs and kidneys. In the liver, methylene chloride may undergo metabolism by two pathways. The first pathway produces CO and CO₂ and is saturable at a few hundred ppm. The second pathway yields formaldehyde and formic acid and shows no indication of saturation at inhaled concentrations up to 10,000 ppm. Acute toxic effects (central nervous system depression) may persist for hours after removal from the source of exposure because of continued metabolism of methylene chloride released from tissue storage (ATSDR 1990). COHb levels can continue to rise, peaking 5–6 hours after exposure. Peak levels of 12 (NIOSH 1974) and 50% (Ellenhorn 1997) have been reported. Most of the effects seen following acute high-concentration exposure are due to the anesthetic properties of the parent compound. Fatalities have occurred with COHb levels less than 10%; thus, it is not likely that carboxyhemoglobinea was the cause of death. A study by Scholz et al. (1991) suggests that, in addition to anesthetic effects, the sudden onset of cardiac arrhythmia induced by acute exposure to methylene chloride may contribute to its lethality.

The metabolic contribution of each pathway appears to vary in humans, particularly with the exposure level, and therefore toxicity extrapolation between high and low doses is complex (Gargas et al. 1986). Furthermore, recent studies suggest that the second pathway is considerably more active in certain animal species, particularly mice, a finding that complicates interspecies comparisons (ATSDR 1990; Green et al. 1986b, 1986c).

The body eliminates methylene chloride primarily through the lungs. A small amount of unchanged methylene chloride is also eliminated in the urine and feces. At low doses, a large percentage of methylene chloride is metabolized to form COHb and eliminated as carbon monoxide, while at higher doses more of the unchanged parent compound is exhaled (ATSDR 1990). The administration of 100% oxygen is efficacious in reducing the half-life of carbon monoxide and should be continued until COHb is less than 5%.

The prompt application of hemoperfusion and diuretic therapy eliminated metabolic acidosis and hemoglobinuria in the case of an acute oral exposure to paint remover containing methylene chloride (Roberts and Marshall 1976). This treatment was thought to have prevented renal damage, but did not address the problem of gastrointestinal ulceration.

2.11.3. Interfering with the Mechanism of Action for Toxic Effects

The primary effects of exposure to methylene chloride appear to be hepatotoxicity and neurotoxicity. However, mechanisms of action for these toxic effects are not well characterized and it is difficult to speculate concerning specific methods for preventing these effects.

Because one consequence of metabolism of methylene chloride is production of COHb, methylene chloride exposure victims are sometimes given supplemental oxygen. The administration of oxygen increases the dissociation of carbon monoxide from hemoglobin and hastens the reduction of COHb. The half-life of COHb is normally approximately 5.3 hours but this can be reduced to 60–90 minutes with inhalation of 100% oxygen. Hyperbaric oxygen as used in carbon monoxide poisoning from direct carbon monoxide inhalation may be useful when high COHb levels are present (Haddad and Winchester 1990). At high COHb levels, hyperbaric oxygen will reduce the half-life to 20–40 minutes (ATSDR 1990). Because carbon monoxide is generated metabolically, this often necessitates a longer duration of oxygen therapy after methylene chloride poisoning than with carbon monoxide poisoning.

Specific mechanisms of action relating to neurological effects are not well understood. Steroids and mannitol have been used to decrease cerebral edema caused by methylene chloride toxicity, but their value in preventing later neurologic sequelae remains unproven (ATSDR 1990).

Victims with cardiopulmonary disease or workers who are exposed to carbon monoxide sources could have an elevated risk from methylene chloride metabolism and should be monitored closely. One reference suggests that physical exercise and smoking be avoided because these activities may have an additive effect on the COHb level (Ellenhorn 1997).

2.12.1. Existing Information on Health Effects of Methylene Chloride

The existing data on health effects of inhalation, oral, and dermal exposure of humans and animals to methylene chloride are summarized in Figure 2-5. The purpose of this figure is to illustrate the existing information concerning the health effects of methylene chloride. Each dot in the figure indicates that one or more studies provide information associated with that particular effect. The dot does not necessarily imply anything about the quality of the study or studies, nor should missing information in this figure be interpreted as a “data need”. A data need, as defined in ATSDR’s Decision Guide for Identifying Substance-Specific Data Needs Related to Toxicological Profiles (ATSDR 1989), is substance-specific information necessary to conduct comprehensive public health assessments. Generally, ATSDR defines a data gap more broadly as any substance-specific information missing from the scientific literature.

As shown in Figure 2-5, studies of humans exposed to methylene chloride by the inhalation route are available. These have focused mainly on neurological, cancer, and systemic effects (e.g., cardiovascular). One report focused on possible reproductive effects following inhalation (and possible dermal) exposure. Other end points (immunological, developmental, and genotoxic effects) have not been evaluated. There are no studies on the effects of methylene chloride in humans after ingestion. Other than one study evaluating potential reproductive effects following inhalation and possibly dermal exposure, no other end points have been evaluated following dermal exposure.

Studies in animals have also focused mainly on inhalation exposure and several end points have been evaluated. Effects of oral exposure have focused on death, systemic effects after intermediate and chronic exposure, genotoxicity, and cancer. Other end points have not been evaluated. There are limited reports on the effects of methylene chloride after direct ocular exposure of animals.

Figure 2-5

Existing Information of Health Effects of Methylene Chloride.

2.12.2. Identification of Data Needs

Acute-Duration Exposure. Available data indicate that the central nervous system is the primary target of inhaled methylene chloride in humans (Bakinson and Jones 1985; Fodor and Winneke 1971; Hall and Rumack 1990; Putz et al. 1979; Stewart et al. 1972; Winneke 1974), rats (Heppel and Neal 1944; Rebert et al. 1989; Savolainen et al. 1981), guinea pigs, rabbits, dogs, and monkeys (Heppel et al. 1944). An acute inhalation MRL was derived for methylene chloride based on a LOAEL of 300 ppm for neurological effects in humans (Winneke 1974). Respiratory effects (cough, breathlessness, and in one fatal case, pulmonary congestion with focal hemorrhage) were observed in several cases of acute occupational exposure to methylene chloride in humans (Bakinson and Jones 1985; Snyder et al. 1992a, 1992b; Winek et al. 1981). No cardiovascular effects were seen in humans acutely exposed to concentrations between 100 and 475 ppm (Cherry et al. 1981; Ott et al. 1983c). Gastrointestinal effects (nausea and vomiting) were reported in several cases of acute occupational exposure to methylene chloride (Bakinson and Jones 1985). Acute inhalation exposure to methylene chloride increased the blood COHb level in humans (DiVincenzo and Kaplan 1981; Manno et al. 1992; Putz et al. 1979; Soden et al. 1996). An acute inhalation study in mice (Aviado and Belej 1974) suggests that the cardiovascular system may be a target for methylene chloride toxicity; however, the effects were observed at lethal concentrations. An acute study in guinea pigs demonstrated an increase in hepatic triglycerides, but no histopathological effects on the liver following exposure to methylene chloride at 5,200 ppm (Morris et al. 1979). Since the most sensitive acute effects in humans are neurological, additional animal studies that evaluate changes in analogous functions (visual, auditory discrimination tasks) at low levels of exposure might be needed. These experiments should monitor CO/COHb levels in the animals to correlate exposure and effect. Furthermore, the animals should be genotyped with respect to for GSTT1 and CYP2E1 to evaluate the mechanism of toxicity.

The only studies in humans on the effects of acute oral exposure to methylene chloride are reports of suicide attempts by ingestion of paint removers. One of these studies reported suppression of the central nervous system and metabolic acidosis (Roberts and Marshall 1976). Corrosion of the gastrointestinal tract was also reported (Hughes and Tracey 1993; Roberts and Marshall 1976). Hematological effects included elevated COHb (Hughes and Tracey 1993) and intravascular hemolysis leading to hemoglobinuria (Roberts and Marshall 1976). No acute respiratory, cardiovascular, hepatic, endocrine, or immunological effects have been reported in humans following ingestion of methylene chloride, and there are no acute oral studies on the effect of low doses in humans. Limited studies in animals involved exposure of rats to high doses of methylene chloride via gavage, resulting in increased mortality (Kimura et al. 1971; Ugazio et al. 1973); as in suicide cases, respiratory failure as a result of suppression of the central nervous system was the cause of death in these animal studies. Hemolysis occurred and diuresis was inhibited in rats gavaged with 1,325 mg/kg of methylene chloride (Marzotko and Pankow 1987); in rats given single oral doses $526 mg/kg, the only endocrine effects noted were dilatation of capillaries of the adrenal medulla and an increase in the secretion of catecholamines (Marzotko and Pankow 1987). In rats gavaged twice with 1,275 mg/kg, DNA damage and an increase in the activity of ornithine decarboxylase were detected in the liver (Kitchin and Brown 1989). Genotoxicity (DNA breakage) was detected in nuclei from selected rat organs following a single dose of methylene chloride (1,720 mg/kg) (Sasaki et al. 1998). Since there is a potential for oral exposure for humans near hazardous waste sites (through contaminated groundwater sources where volatization is restricted), additional animal data on the effects of acute oral exposure to lower doses of methylene chloride would strengthen the database. An acute oral MRL was derived by using a PBPK model to extrapolate human inhalation data from Winneke (1974) to estimate the concentration of methylene chloride in drinking water needed to produce a tissue specific concentration equivalent to that produced by exposure to 300 ppm of methylene chloride (Reitz et al. 1997).

Case reports indicate that methylene chloride can cause eye and skin damage following direct contact with the liquid (Hall and Rumack 1990; Wells and Waldron 1984; Winek et al. 1981). An acute study in rabbits demonstrated adverse ocular effects (Ballantyne et al. 1976). Dermal absorption of methylene chloride has been demonstrated in animals (McDougal et al. 1986), and therefore, is possible in humans. Additional dermal studies in animals are needed to provide further insight into the potential risks of living near hazardous waste sites.

Intermediate-Duration Exposure. Suggestive evidence for liver effects in humans exists, as workers exposed to methylene chloride (up to 475 ppm) via inhalation have shown increased serum bilirubin levels, but not at clinically significant levels (Ott et al. 1983a). Other liver injury parameters were normal. Repeated exposures to methylene chloride vapors (up to 500 ppm) did not alter serum enzyme activity, pulmonary function, or urine chemistry significantly in exposed and control groups (NIOSH 1974). However, irritation of the respiratory tract, but no changes in urinary chemistry were noted after exposure to a TWA of 5–340 ppm in an occupational study (Anundi et al. 1993).

Available data for rats, dogs, monkeys, rabbits, guinea pigs, gerbils, and mice indicate that repeated exposure to methylene chloride via inhalation produces central nervous system effects (Briving et al. 1986; Karlsson et al. 1987; Rosengren et al. 1986), respiratory effects (Heppel et al. 1944; NTP 1986) hepatic effects (Haun et al. 1972; Heppel et al. 1994; Kjellstrand et al. 1986; MacEwen et al. 1972; Norpoth et al. 1974; NTP 1986; Weinstein and Diamond 1972), and kidney effects (Haun et al. 1972; MacEwen et al. 1972). An intermediate inhalation MRL was based on a LOAEL of 25 ppm for hepatic and renal effects in rats (Haun et al. 1972).

No quantitative data are available for oral exposures of intermediate duration in humans. Hepatic damage and lowered urinary pH in rats was reported following oral exposure 166 mg/kg/day for 3 months (Kirschman et al. 1986), but the reporting of results was incomplete in this study. Furthermore, since the lowest dose was a LOAEL, additional intermediate oral studies are needed in animals to evaluate specific tissue and organ effects at lower doses and to determine threshold levels and dose-response relationships. Reitz et al. (ATSDR 1997) developed a PBPK model to extrapolate inhalation data for an oral MRL, based on the inhalation study of Haun et al. (1972), but the reporting of results in this study was incomplete; a new intermediate oral animal study would be preferable. No intermediate oral MRL was derived because of lack of adequate data.

There are no quantitative intermediate dermal exposure data available for humans or animals. Given the potential for human exposure to methylene chloride through paint strippers, adhesives, glues, paint thinners, wood stain, varnishes, spray paint, and automobile spray primers, information on the effects of dermal exposures are needed to estimate more accurately the risks to human health from methylene chloride.

Chronic-Duration Exposure and Cancer. No data are available on the non-neoplastic effects in humans after chronic exposure to methylene chloride via the oral or dermal routes. Neurological and reproductive effects have been reported in workers following combined inhalation/dermal exposures for up to 3 years (Kelly 1988). Studies in animals suggest that the liver is a target organ following chronic inhalation (Burek et al. 1984; Nitschke et al. 1988a; NTP 1986) and oral (Serota 1986a, 1986b) exposure. Both a chronic inhalation MRL and a chronic oral MRL were derived using hepatotoxicity as the critical endpoint (Nitschke et al. 1988a; Serota et al. 1986a). Renal effects were also observed following chronic inhalation of methylene chloride in rodents (NTP 1986). Dermal data in animals after chronic exposure to methylene chloride are not available. Additional dermal studies in animals are needed to enhance our understanding of potential risk to people living near hazardous waste sites.

Epidemiological studies of human occupational cohorts show no increase in cancer of the lung, liver, or any other organs from occupational inhalation exposures (Friedlander et al. 1978; Hearne et al. 1987, 1990; Ott et al. 1983a). There are no human oral or dermal exposure data for the cancer endpoint. Inhalation studies in animals show a concentration-dependent, statistically significant increase in liver and lung adenomas and carcinomas in mice exposed to high concentration of methylene chloride (Mennear et al. 1988; NTP 1986) and benign mammary gland tumors in rats (Mennear et al. 1988; NTP 1986) following 2 years of exposure to methylene chloride. The evidence for carcinogenicity in animals from oral exposures (Serota et al. 1986a, 1986b) is inconclusive, and there are no dermal data available. Therefore, additional chronic oral and dermal studies are needed to clarify the cancer risk of ingested methylene chloride. The carcinogenic mechanism of inhaled methylene chloride in animals is not yet understood despite an extensive database on toxicokinetics and potential mode(s) of action. Additional information on the mechanisms of carcinogenesis is needed to provide further insight regarding current findings that (1) the concentration-response may be nonlinear at low exposure concentrations and there is a concentration range below which carcinogenicity is unlikely to occur and (2) the mouse is not an appropriate model for investigation of potential human carcinogenicity because it is much more sensitive than any other species.

Genotoxicity. Genotoxicity data show mixed results. Generally, mutagenesis assays are negative in mammalian in vivo studies (Burek et al. 1984; Sheldon et al. 1987). Methylene chloride is mutagenic in bacteria (Gocke et al. 1981), but results for Drosophila (Gocke et al. 1981) are contradictory. This compound has shown a dose-response relationship for clastogenesis in mammalian cells in vitro (Thilagar et al. 1984a) but rats exposed in vivo (Burek et al. 1984) have shown no effects. In mice exposed in vivo, tissue-specific DNA breaks were observed in mice (Sasaki et al. 1998), which suggests that genotoxicity of methylene chloride may be dependent on the expression of metabolizing enzymes. In view of the unknown carcinogenic mechanisms, additional in vivo studies of clastogenesis are needed to provide useful markers, both for dosimetry and specific mechanisms. A recent in vitro test using strains of Salmonella with or without recombinant GSTT1 has demonstrated that methylene chloride may have more than one mechanism of genotoxicity (De Marini et al. 1997). In bacteria expressing mammalian GSTT1, methylene chloride, at moderate doses, caused a single type of genetic lesion. In bacteria without GSTT1, higher doses were required to induce genetic lesions and the results were more varied. These results may explain the variability observed in previous studies. Therefore, additional genotoxicity studies should be carried out on mammalian cells for which the expression of GSTT1 and/or CYP2E1 isoenzymes is defined.

Reproductive Toxicity. Data on reproductive toxicity in humans are limited. One case study reported genital pain, and low sperm counts (testicular atrophy in some cases) in workers who inhaled vapors of methylene chloride and had direct contact with the liquid on the job for up to 3 years (Kelly 1988); they were also exposed to styrene at low levels. Exposure to methylene chloride was confirmed by blood COHb levels. In another study, workers exposed to methylene chloride for a shorter time showed no reproductive effects (Wells et al. 1989b). A case-control occupational study reported a nearly significant association of methylene chloride exposure and the incidence of spontaneous abortion (Taskinen et al. 1986). There are no oral or dermal exposure data that pertain to reproductive effects in humans. A two-generation inhalation study in rats reported no effects on fertility and litter size (Nitschke et al. 1988b). In addition, no effects on the testes were observed in dominant lethal tests involving male mice that inhaled concentrations of methylene chloride up to 200 ppm for up to 6 weeks (Raje et al. 1988). Uterine, ovarian, and testicular atrophy were observed in rodents following chronic exposure to methylene chloride at 4,000 ppm (NTP 1986). There are no data on reproductive effects in animals following oral or dermal exposure. Intermediate-duration oral and dermal studies that incorporate histopathological analysis of the reproductive organs are needed to address this data need. Part of the analysis should include a determination of whether the reproductive organs express GSTT1 and/or CYP2E1, as a first step in evaluating their possible role in the reproductive toxicity of methylene chloride.

Developmental Toxicity. No data are available on developmental toxicity of methylene chloride following inhalation, oral, or dermal exposure in humans, aside from the case-control study relating exposure and the rate of spontaneous abortion mentioned above (Taskinen et al. 1986). Exposure to 0.01 ppm of methylene chloride had no effect on birth outcome in a study of over 90,000 pregnancies (Bell et al. 1991). Several inhalation studies in animals indicate that methylene chloride can cross the placenta (Anders and Sunram 1982). Some of these studies showed statistically nonsignificant malformations in rats and mice or decreased fetal weight at maternally toxic concentrations (Bornschein et al. 1980; Hardin and Manson 1980; Schwetz et al. 1975). However, there was a statistically significant increase in the incidence of delayed ossification of sternebrae (Schwetz et al. 1975). The use of only one concentration in these studies precludes any evaluation of concentration-response relationships.

However, Reitz et al. (1997) developed an inhalation route-to-oral route extrapolation and rodent-to-human species extrapolation using PBPK modeling of the developmental toxicity data in Schwetz et al. (1975). The resulting LOAEL was an intermediate oral dose of 142 mg/kg/day. Additional studies for inhalation and dermal exposures in two species would be useful in clarifying the developmental toxicity potential of this chemical. Conducting studies on animals with known genotypes with respect to metabolizing enzymes GSTT1 and CYP2E1 are needed to evaluate the risk of exposure to methylene chloride.

Immunotoxicity. No human data are available on immunotoxicity of methylene chloride. Animal data are limited on the immunotoxicity of methylene chloride. One study was located that indicated that methylene chloride caused splenic atrophy in dogs following continuous intermediate-duration inhalation exposure (MacEwen et al. 1972). Splenic fibrosis was observed in a chronic rat study (Mennear et al. 1988), but occurred only at the highest concentration tested (1,000 ppm) and only in one sex. A recent intermediate-duration study, also in rats, found no evidence of gross or microscopical alterations in the spleen at a concentration of 5,187 ppm methylene chloride (Halogenated Solvent Industry Alliance, Inc. (2000). This study also found no effect on IgM response to SRBC. Additional immunotoxicity studies including evaluation of humoral- and cell-mediated immunity are needed to determine whether this system is susceptible to methylene chloride, as some chlorinated hydrocarbons do affect the immune system.

Neurotoxicity. The central nervous system is a target for both short- and long-term inhalation exposures in humans. These data are derived from experimental (Fodor and Winneke 1971; Putz et al. 1979; Stewart et al. 1972; Winneke 1974;) and occupational studies (Lash et al. 1991; White et al. 1995) that reported alterations in behavioral performance and various psychomotor tasks following exposure to methylene chloride. Other neurotoxic effects noted in occupational studies included dizziness, headaches, nausea, memory loss, paresthesia, tingling in hands and feet, and loss of consciousness (Bakinson and Jones 1985; Hall and Rumack 1990; Kelly 1988). It should be noted that some of these workers were exposed to other unspecified solvents at the same time. Winneke (1974) attributed the neurological effects in volunteers following a 3–4 hour inhalation exposure to 300–800 ppm of methylene chloride to the anaesthetic properties of the parent compound since exposures to 50–100 ppm of carbon monoxide alone did not produce these effects. The specific parameters in the Winneke (1974) study, critical flicker fusion frequency, auditory vigilance, and other psychomotor tasks, were considered to be specific measures of ‘cortical alertness’ or central nervous system depression. Putz et al. (1979) correlated neurological deficits following inhalation exposure to methylene chloride (200 ppm) with the accumulation of COHb in the blood and carbon monoxide in the exhaled breath; performance did not deteriorate until 3 hours of exposure had elapsed and blood COHb levels exceeded 5%. Although the Putz et al. (1979) study seems to implicate the metabolite CO as the neurotoxic agent, there is a need for studies to determine the relative contributions of parent compound and metabolite to the neurotoxic effects of methylene chloride. The Winneke (1974) study, which was deemed to have used more specific measures of central nervous system depression than the Putz et al. (1979) study, was used to derive an acute inhalation MRL and, by route-to-route extrapolation, also to derive an acute oral MRL (Reitz et al. 1997).

Subtle neurological effects were reported in rats after acute exposure to methylene chloride. Changes in somatosensory-evoked potentials were observed after 1 hour exposure to 5,000 ppm of methylene chloride (Rebert et al. 1989), and changes in cerebellar enzymes were detected following 2 weeks of exposure at 500 ppm (Savolainen et al. 1981). Inhalation studies in rats did not reveal neurobehavioral effects after intermediate-duration exposure at 4,500 ppm of methylene chloride vapors (Bornschein et al. 1980). In other studies, neurochemical changes were reported in gerbils at 210 ppm (Briving et al. 1986; Karlsson et al. 1987; Rosengren et al. 1986). There were no treatment-related neurophysiological or neuropathological effects in rats exposed to concentrations of 2,000 ppm for 13 weeks (Mattsson et al. 1990).

There are no oral or dermal studies in either humans or animals with regard to neurotoxicity of methylene chloride. Data on electrophysiology, functional observational batteries, and neuropathology from 90-day oral or dermal exposures in at least two different species would be useful to further evaluate neurotoxic effects from these routes of exposure, since human exposure can occur from all three routes at hazardous waste sites.

Epidemiological and Human Dosimetry Studies. Information on the effects of methylene chloride in humans comes from occupational studies, case reports of acute high exposure, and studies with volunteers. Asphyxia and eventually death occurred in a subject acutely exposure to a high but undetermined concentration of methylene chloride in the air (Winek et al. 1981). Exposure to lower concentrations affects primarily the central nervous system; signs and symptoms observed include dizziness, incoordination, loss of balance, unconsciousness, and decreased performance in tests of sensory and motor functions (Bakinson and Jones 1985; Fodor and Winneke 1971; Hall and Rumack 1990; Putz et al. 1979; Stewart et al. 1972; Winneke 1974). These effects are likely to be caused by a combination of the anaesthetic properties of the parent compound and accumulation of COHb which forms as a result of methylene chloride metabolism. There is a potential for low-level general population exposure from ambient methylene chloride emissions from paint removal, aerosol use, metal degreasing, electronics and pharmaceutical manufacturing, and food processing (Callahan 1981; CPSC 1987, 1990; NAS 1978). Such exposure levels are unlikely to produce adverse neurological effects. Hazardous waste sites release methylene chloride into the air, groundwater, surface water, and soil. There are very few monitoring data for methylene chloride outside the occupational setting (Singh et al. 1981). Because methylene chloride evaporates readily from water and soil, inhalation is the main route of potential exposure. In the unlikely event that long-term exposure of the general population (in the past or present) to low levels of primarily methylene chloride is identified, individuals should be monitored for hematological and hepatic effects, as these have been observed in studies in animals.

Biomarkers of Exposure and Effect.

Exposure. The presence of methylene chloride in the postexposure expired breath is the most commonly used biomarker of exposure. Methylene chloride is easily detected in expired air for 24 hours following a vapor exposure (NIOSH 1974). Methylene chloride can be detected in blood, but because clearance is so rapid, this method is only useful for monitoring recent exposures (DiVincenzo and Kaplan 1980). The half-life of COHb following methylene chloride exposure is twice that following exposure to carbon monoxide (NIOSH 1974; Stewart and Hake 1976). COHb in blood may be monitored; however, COHb is cleared so rapidly (plasma half-life of 40 minutes) that it is unlikely to be useful for monitoring environmental exposures (DiVincenzo et al. 1972). Urinary excretion of methylene chloride is measurable for several hours postexposure (DiVincenzo et al. 1972).

Effect. Changes in the nervous system are sensitive, but not specific, biomarkers of methylene chloride effects. Clinical signs and symptoms which may be monitored include narcosis, fatigue, and analgesia (Bakinson and Jones 1985; Hall and Rumack 1990; Putz et al. 1979; Winneke 1974). Impairment of neurobehavioral functions and electromyographic measurements of nerve conduction velocity and amplitude can be monitored to detect early signs of neurotoxicity in people exposed to methylene chloride (Glatzel et al. 1987; White et al. 1995). Additional studies that couple measurement of methylene chloride with tests for determining nervous system effects are needed to correlate exposure with adverse health effects of this chemical.

While neurobehavioral tests are not specific for methylene chloride-induced toxicity, they do identify potential health impairment. Studies to develop more specific biomarkers of methylene chloride-induced effects are needed to assess the potential health risk of methylene chloride exposure near hazardous waste sites.

Absorption, Distribution, Metabolism, and Excretion. Methylene chloride is a volatile liquid with high lipid solubility and modest solubility in water. In humans, it is rapidly absorbed by the inhalation (DiVincenzo and Kaplan 1981; McKenna et al. 1980) and ingestion routes of exposure (Roberts and Marshall 1976). Dermal absorption is suspected to occur in humans, but quantitative data are lacking. Methylene chloride is also readily absorbed by animals following inhalation (DeVincenzo et al. 1972; MacEwen et al. 1972; McKenna et al. 1982) and oral exposure (Angelo et al. 1986a); dermal absorption data in animals are also lacking. Once absorbed, methylene chloride is quickly distributed to a wide range of tissues and body fluids. Absorption and distribution of methylene chloride can be affected by a number of factors, including dose level, vehicle, physical activity, duration of exposure, and amount of body fat (Astrand et al. 1975; DiVincenzo et al. 1972; Engstrom and Bjorstrom 1977).

Distribution data in humans are lacking, but it has been found in human breast milk (EPA 1980e; Pellizzari et al. 1982). Methylene chloride is widely distributed in animal tissues after inhalation exposure (Carlsson and Hultengren 1975; McKenna et al. 1982). The highest concentrations are found in adipose tissue and liver. Methylene chloride has been found in blood from rats’ fetuses (Anders and Sunram 1982). After acute exposure, methylene chloride disappears rapidly from fat (Carlsson and Hultengren 1975). No studies were located regarding distribution of methylene chloride following dermal exposure in animals. Distribution of methylene chloride does not seem to be route-dependent and it does not bioaccumulate in tissues.

Methylene chloride is metabolized via two pathways, the MFO pathway, which produces CO and CO₂, and the GST pathway which yields only CO₂ (Gargas et al. 1986). Methylene chloride is metabolized almost exclusively by the MFO pathway at low exposures. The only data available for humans on the toxicokinetics parameters are for inhalation exposures. These data show that the GST activity is low in humans relative to other animal species (Andersen et al. 1987). It has been postulated that the activity of the GST pathway in rats is less than that of mice and might only become significant when the P-450 pathway (MFO) has been saturated (Green et al. 1986b, 1986c). Toxicokinetic data for oral and dermal exposures would be useful because some exposures to humans are expected to be from contaminated drinking water and soils. Toxicokinetic models (Andersen and Krishnan 1994; Andersen et al. 1987, 1991; Dankovic and Bailer 1994; Ramsey and Andersen 1984; Reitz 1990) have been developed that account for the nonlinearities in the internal dose across exposure levels that arise from concentrationdependent changes in absorption, distribution, excretion, and saturation of metabolism following inhalation exposure to methylene chloride. Reitz et al. (1997) have developed an inhalation route-to-oral route extrapolation for methylene chloride. The metabolism of methylene chloride in humans appears to be qualitatively similar to that in animals (Fodor et al. 1973; Wirkner et al. 1997).

Methylene chloride is excreted via expired air primarily as CO and CO₂ at low concentrations (McKenna et al. 1982). As concentration increases, more unchanged parent compound is exhaled. A small fraction of absorbed methylene chloride or metabolites has been detected in urine of humans occupationally exposed (DiVincenzo et al. 1972) and in the urine and feces of experimental animals (McKenna at al. 1982). Additional data on excretion parameters as functions of concentration are needed to understand the toxicokinetics of methylene chloride.

Comparative Toxicokinetics. Generally, animal data on toxicokinetics parameters substantiate those from humans. There is rapid absorption through the lung (DiVincenzo and Kaplan 1981; DiVincenzo et al. 1972; MacEwen et al. 1972; McKenna et al. 1982). The metabolites of methylene chloride in different species indicate that they are metabolized by the same pathways to CO or to CO₂ (Fodor et al. 1973; Gargas et al. 1986; Wirnker et al. 1997). The data from both human and animal studies indicate that the target organs of methylene chloride toxicity are the same across species. Interesting differences do exist, however, in some metabolic parameters. The rate of clearance from rat tissues was markedly slower than in the mouse following inhalation exposures. In contrast to findings of a direct proportional relationship between inspired air concentration of methylene chloride and blood level in man and other animals, there is a report that the ratio at steady state between blood levels of methylene chloride and exposure increases as the concentration increases in rats (Green et al. 1986b, 1986c). Distribution of methylene chloride across tissues in rats and humans has been shown to be lipophilic except for one rat study that showed less of the chemical in adipose tissue (McKenna and Zempel 1981). Species differences in placental anatomy and physiology may contribute to variations in fetal exposure to methylene chloride or its metabolites following maternal exposure. Additional data on the impact of differences in species sensitivity to specific internal doses by different routes of exposure are needed to resolve questions about extrapolation of predicted effects from different routes in different species.

Methods for Reducing Toxic Effects. There is no specific treatment or antidote for methylene chloride intoxication. Available methods and treatments (e.g., gastric lavage with the airway protected, or the use of oxygen as in carbon monoxide poisoning) for reducing peak absorption have been shown to be beneficial (Shih 1998; Tomaszewski 1998). There is no evidence that activated charcoal or cathartics are effective (Shih 1998). Because methylene chloride is metabolized to carbon monoxide, mitigation strategies (e.g., hyperbaric oxygen) used for carbon monoxide intoxication have been effective in increasing the oxygen carrying capacity of the blood (Tomaszewski 1998).

Children’s Susceptibility. Data needs relating to both prenatal and childhood exposures, and developmental effects expressed either prenatally or during childhood, are discussed in detail in the Developmental Toxicity subsection above.

There are no studies that specifically addressed effects of exposure to methylene chloride in children. There is no evidence from human toxicity studies that children are, or are likely to be, more susceptible to health effects of exposure to methylene chloride than adults. No cases of accidental poisoning in children due to methylene chloride exposure have been reported. The nervous system is a sensitive target for acute exposure to methylene chloride in adults, and the response in children is likely to be similar.

Although developmental studies in animals indicate that methylene chloride is not a teratogen (Bornschein et al. 1980; Hardin and Manson 1980), there is a need to evaluate neurological/neuro-behavioral effects in animals exposed in utero. Subtle neurological effects could result from hypoxia (CO-mediated) or from reactive intermediates of metabolism that would only be revealed by appropriate behavioral testing of the offspring. It is not clear that the “wheel running activity” and “avoidance learning” tests that Bornschein et al. (1980) employed in rats exposed in utero were adequate to reveal neurological deficits. Acute effects observed in an adult human study involved degraded performance on visual and auditory discrimination tasks (Putz et al. 1979). If neurological effects were detected in mice, it would be useful to conduct additional developmental studies using mice in which functional genes for GSTT1 and/or CYP2E1 have been knocked-out, to discover which metabolic pathway is implicated in developmental neurological effects.

There is no direct evidence regarding pharmacokinetics or the mechanism of toxicity of methylene chloride in children. However, considering what is known about the expression of metabolizing enzymes (see Section 2.7 Children’s Susceptibility), there is no reason to expect that children over the age of 6 months will be qualitatively different from adults; by 6 months, the last of the fetal hemoglobin has been replaced by the adult isoforms. At fetal and early postnatal stages, however, children will be more vulnerable to methylene-chloride-generated CO because of its effect on fetal hemoglobin (Longo 1977). CO binding to Hb causes oxygen to be held more tightly, so that tissues receive less oxygen; the net effect is that the CO clears more slowly from the circulation. It is not known to what extent methylene chloride (or metabolites other than CO or CO₂) can cross the placenta in humans or accumulate in breast milk. A study in animals showed that methylene chloride crosses the placental barrier (Anders and Sunram 1982), but quantitative data are lacking. There are no animal studies testing whether methylene chloride can pass into breast milk. There are no studies quantifying maternal exposure and output into breast milk. The Fisher et al. (1997) PBPK model predicts a relatively low transfer into breast milk following maternal inhalation exposure.

In adults, the biomarker of exposure to methylene chloride is the parent compound, and the biomarkers of effect are CO in the exhaled breath or COHb in the blood. There is no reason to expect that biomarkers in children would be any different.

It is expected that any chemical interactions involving methylene chloride would have the same effects in children as in adults. The exception would be the fetal period, during which concentrations of metabolizing enzymes are significantly lower than in the adult. A fetus exposed to methylene chloride and another chemical that requires CYP2E1 for metabolism, might be at higher risk than an adult.

Treatment with 100% oxygen has been used to enhance the rate of clearance of CO from the bloodstream after acute methylene chloride exposure. Although there is no specific information regarding pediatric treatments for methylene chloride exposure, 100% oxygen has been used for treating pregnant women exposed to CO (Longo 1977). In these cases, it is recommended that the oxygen treatment be extended 3.5–5 times longer than the time required to bring the maternal COHb level down to 3%, to ensure that the fetus is adequately treated. If an exposed mother has neurological symptoms or a COHb level above 15%, a single hyperbaric oxygen treatment is indicated (Tomaszewski 1998). It does not appear that any pediatric-specific methods need to be developed.

Since methylene chloride may be genotoxic, depending on the genotype with respect to GSTT1 and CYP2E1 (see Section 2.10), it is possible that parental exposure could affect the fetus. At issue would be whether the gonads express alleles that are associated with genotoxicity, since the effect is intracellular.

Child health data needs relating to exposure are discussed in 5.8.1 Identification of Data Needs: Exposures of Children.

2.12.3. Ongoing Studies

A number of research projects are in progress investigating the health effects and mechanism of action of methylene chloride. These projects have been identified from FEDRIP (1999).

Dr. J.B. Wheeler, at Vanderbilt University, Nashville, Tennessee, is comparing the ability of rat and human glutathione S-transferases (GST) to form mutagenic DNA adducts with glutathione in the presence of different haloalkanes. He will examine the mechanism of activation by GST, the stability of the glutathione conjugates, and the DNA adducts produced. This research is sponsored by the National Cancer Institute.

Dr. T.R. Devereux, at the NIEHS, is identifying critical target genes and genetic alterations that may be important in chemical carcinogenesis. He is characterizing genetic alterations in oncogenes (e.g., ras) and tumor suppressor genes (e.g., p53 and p16) from rodent tumors and human cancers generated by exposure to chemical agents. He is also examining polymorphisms in the human p16 gene that may predispose to lung cancer.

Dr. D.A. Bell, of the Genetic Risk Group at the NIEHS, is conducting case-control studies of 5,000 genotyped individuals to test the impact of cancer susceptibility genes on cancer of the bladder, lung, liver, colon, stomach, prostate, and breast. He and his colleagues have identified ethnic differences in the frequency of at-risk genotypes for glutathione transferase (M1, theta 1, and Pi), and N-acetyltransferase (1 and 2).