Drinking-water

Exposure can take place by drinking contaminated water as well as volatilization. In the United States, PCE is not detectable in drinking-water in most cases. The EPA Groundwater Supply Survey of 945 drinking-water systems nationwide reported PCE in 75 out of 945 systems in 1984; the median level of the positive samples was about 0.75 μg/litre, with a maximum of 69 μg/litre (2). In several studies in Canada, PCE was detectable in 6–60% of the tested samples. Reported mean values ranged from 0.1 to 0.9 μg/litre (6). In several other countries, including Germany and the United Kingdom, similar results have been observed (8,16). In some villages in Finland, concentrations of up to 180 μg/litre were found (17). In Japan, PCE levels in 99.7% of drinking-water (n = 5600) were less than 1 μg/litre with a maximum of 7 μg/litre in 2007 (18).

Food

Data on concentrations of PCE in food are scarce. In some studies from Germany and Switzerland reported in the early 1980s, relatively high total intakes of 87–170 μg/day were found (16). The results of market basket surveys in the United States, reported in 1987 and 1988, indicated lower levels. From the results of these surveys, total daily intake via food has been estimated at 0.12–65 μg/kg body weight (6). Several groups of researchers have reported elevated concentrations of PCE in fatty food products in residences and markets, owing to contamination from dry cleaning establishments nearby. In a supermarket near a dry cleaning shop in Germany, concentrations were 36 μg/kg and 110 μg/kg in cheese and margarine, respectively (8). In one instance in the United States, a very high PCE concentration was found in margarine (up to 50 mg/kg) in a shop next door to a dry cleaning establishment (2). Moreover, food grown on contaminated soil can contain PCE. In Japan, the daily maximum levels of PCE in the diet between 1990 and 1999 (n = 72∼81 in each year) were 4.4 μg/kg wet weight (14).

Species-dependent differences in biotransformation

Mice metabolize PCE to trichloroacetic acid to a greater extent than rats. In rats, saturation of the oxidative metabolism at exposure concentrations exceeding 678 mg/m³ (100 ppm) prevents the occurrence of the high trichloroacetic acid concentrations that are observed in mice at these dose levels. The available evidence indicates that in humans too, saturation of the oxidative metabolism of PCE occurs at ≥ 678 mg/m³ (≥ 100 ppm) (51). In a comparative study, the urinary levels of N-acetyl-S-(1,2,2-trichlorovinyl)-L-cysteine, one of the final metabolites in the glutathione biotransformation pathway, were shown to be higher in rats than in mice after inhalation exposure to PCE. In vitro tests in hepatocytes and kidney fractions of rats, mice and humans suggest that the potential for this biotransformation in the kidneys is also lower in humans than in rats (2,6,52). Experimental data show that rates of TCVG formation in rats and mice are higher in males than in females, the formation rates in mice being higher than in rats (48,53).

The overall kinetics of PCE oxidative metabolism differs significantly between humans and rodents. The excretion of trichloroacetic acid after inhalation of PCE has been shown to be slower in humans than in rats. The elimination half-time of trichloroacetic acid in the urine was approximately 4.1-fold longer in humans than in rats exposed to 10 or 40 ppm PCE. Maximal trichloroacetic acid concentrations in blood were 3- to 10-fold lower (depending on dose) in humans than in rats (48,54). It is assumed that saturation of PCE metabolism occurs at lower doses in humans (55,56).

While the target organs for toxicity in general are similar for different species, this is not the case for carcinogenicity.

Sex-dependent differences in biotransformation

Regarding renal toxicity, sex differences have been found in experiments on rats and mice. PCE is demonstrated to cause a significant release of lactate dehydrogenase in isolated kidney cells from male but not from female rats. TCVG causes much more lactate dehydrogenase to be released from male than from female rat kidney cells. Examination of effects on mitochondrial respiration in suspensions of isolated mitochondria showed that PCE and TCVG are more toxic in renal mitochondria from male rats than in those from female rats. Respiratory function in (renal) mitochondria from mice was significantly inhibited by PCE and TCVG, but showed little sex dependence (49). Experimental data showed that TCVG formation in kidney and liver subcellular fractions from male rats and mice were invariably higher than corresponding values in female rats and mice (53).

Toxicological effects

PCE has a low acute inhalation toxicity (LC₅₀ in rodents ≥ 16.6 mg/litre). Acute oral toxicity is also low (LD₅₀ in rodents ≥ 3000 mg/kg body weight) (6).

The liver, kidneys, blood and CNS are the target organs for systemic effects. Hepatic effects occur at lower dose levels in mice than in rats. In a series of experi ments in mice, continuous (24 hours/day) inhalation exposure for 30 days at concentrations of 61, 251, 508 or 1017 mg/m³ produced a dose-related increase in liver weight at all dose levels. Morphological changes in the liver were observed at 251 mg/m³ and, as stated but not fully reported, at all other dose levels as well. Plasma butyrylcholinesterase activity showed a slight increase at ≥ 251 mg/m³. The reversibility of the effects was exami ned at 1071 mg/m³ only, showing that at 120 days after cessation of treatment, only a slight increase in liver weight remained (83).

Within the United States National Toxicology Program (NTP) subchronic inhalation studies were carried out in rats and mice. In mice, liver effects were observed at 1350 mg/m³ (≥ 200 ppm) or higher. Karyomegaly of the renal tubule epithelial cells was observed at ≥ 1350 mg/m³ (≥ 200 ppm). At 680 mg/m³ (100 ppm), no effects were seen but at this level no histopathological examination of the liver was carried out. In rats, mild congestion of the liver was seen at ≥ 1350 mg/m³ (≥ 200 ppm) (lower concentrations not tested), with congestion of the lungs, decreased survival and growth retardation at higher dose levels (84). In chronic studies, renal tubular cell karyomegaly was observed in female rats and mice at both 1356 and 2712 mg/m³ (200 and 400 ppm) in rats and 678 and 1356 mg/m³ (100 and 200 ppm) in mice.

In addition, in rats, renal tubular hyperplasia (males), thrombosis and squamous metaplasia in the nasal cavity (males and females) and hyperlasia in the adrenal medullae (males) or cortices (females) were found. In mice only, liver cell degeneration and necrosis were seen at both concentrations (84). A study exposing mice to 2040 mg/m³ (300 ppm) perchloroethylene gas for 6 hours daily for 5 days revealed regeneration of olfactory mucosa. Two weeks and three months after exposure, the nasal mucosa was examined histopathologically. After two weeks, ciliated epithelium started to replace the olfactory epithelium. Damage was more persistent in the nasal mucosa of the olfactory region than in the respiratory region (85).

CNS effects have been observed at high dose levels. In the short-term NTP experiment in rats and mice, neurological symptoms were observed only at the highest concentration of 11 860 mg/m³ (1750 ppm) (84). Biochemical changes in brain tissues have been observed in rats and gerbils at concentrations of ≥ 407 mg/m³ (≥ 60 ppm). However, the toxicological significance of these changes is unclear (2). In a recent short-term study in rats, decreased weight of brain regions and reductions in neuronal marker proteins were seen at 4080 mg/m³ (600 ppm) but not at 2040 mg/m³ (300 ppm) (86). Haematotoxicity was observed in an inhalation study in mice at test concentrations of 915 and 1830 mg/m³ (no other concentrations tested) in which the animals were exposed for 6 hours/day, 5 days/week for 7.5 or 11.5 weeks (87). Rats that were acutely exposed to a high dose (500 mg/kg oral) of PCE and others subchronically to lower doses (5 and 50 mg/kg oral), showed significant effects on nociception and locomotor activity (88). An inhalation study assessed the effects of PCE on sustained attention in rats performing a visual signal detection task. It showed that inhaled PCE acutely impaired sustained attention in rats, and its potency weakened on repetition of the exposure. Accuracy was significantly reduced at 3390 mg/m³ (500 ppm), significantly elevated response time was seen at 6780 mg/m³ (1000 ppm) and significantly reduced trial completions were found at 10 170 mg/m³ (1500 ppm) (89).

Limited teratogenicity studies in rats, mice and rabbits suggest that PCE produces fetotoxicity and embryotoxicity at high dose levels (≥ 2034 mg/m³). In a behavioural teratogenicity study in rats, neuromuscular ability in pups was decreased at 6100 mg/m³ (900 ppm) but not at 2034 mg/m³ (300 ppm) (2). In a study on developmental toxicity, embryos from rats were explanted on gestational day 120 and cultured for 46 hours in the presence of PCE. Exposure to PCE ranging from 3.5 to 15 mM caused concentration-related effects on growth and development. The presence of hepatic microsomal fractions in the culture medium partially prevented the effects of PCE on the measured parameters of embryo growth and development (90). A study showed altered behaviour in adult mice orally exposed to PCE as neonates (5 and 320 mg/kg PCE/day between days 10 and 16 postnatally), indicating neonatal susceptibility of brain maturation in achieving long-lasting changes in adult behaviour (91). A developmental study in rats following inhalation exposure to PCE showed maternal responses that were of limited toxicological significance. Developmental effects at 600 ppm consisted of reduced gravid uterus, placental and fetal weights and decreased ossification of thoracic vertebral centra. Developmental effects at 1695 mg/m³ (250 ppm) were of minimal toxicological significance, being limited to minor decreases in fetal and placental weights. There were no developmental effects at 441 mg/m³ (65 ppm) (92).

PCE at 68–2043 mg/m³ (10–300 ppm) was found to increase, in a concentration-dependent manner, ACh- and high K+-induced muscle contraction in tracheal tissue from swine. Swine tissue was used because there are many similarities between swine and human airways. More research is necessary to be able to conclude that these responses play a role in airway impairment and hyperresponsiveness after PCE exposure (93). A study using renal cell suspensions of rats and mice showed that PCE caused acute renal cellular toxicity in rats and mitochondrial toxicity in rats and mice (49).

Carcinogenic effects

Data on the carcinogenicity of PCE have been evaluated in various health assessment documents (2,6,51). PCE-induced nephrotoxicity and nephrocarcinogenicity have been associated with metabolism by the glutathione conjugation pathway to form TCVG (48,53).

Oral and inhalation studies with NTP have been performed in mice and rats. In a gavage study in B6C3F1 mice (time-weighted average (TWA) daily dose levels of 536 and 1072 mg/kg body weight in males and 386 and 772 mg/kg in females), increased incidences of hepatocellular carcinomas were found in males and females (4/37, 32/49 and 27/48 in males and 2/40, 19/48 and 19/48 in females). In Osborne Mendel rats, no increase in tumour incidence was seen. This study has various limitations: numerous dose adjustments during the study, early mortality related to compound-induced toxic nephropathy, and pneumonia due to intercurrent infectious disease in both rats and mice (2,94). In the inhalation study in B6C3F1 mice (test concentrations 100 and 200 ppm, 6 hours/day, 5 days/week for 103 weeks, corresponding to 680 and 1360 mg/m³), incidences of hepatocellular carcinomas were increased (7/49, 25/49 and 26/50 males and 1/48, 13/50 and 36/50 females). In F344/N rats (test concentrations 200 and 400 ppm, 6 hours/day, 5 days/week for 103 weeks, corresponding to 1360 and 2720 mg/m³), PCE caused dose-related increases in the incidence of stage 3 mononuclear cell leukaemia in animals of both sexes: in males, 20/50 in controls, 24/50 at the low dose and 27/50 at the high dose; in females, 10/50, 18/50 and 21/50, respectively. The historical control for mononuclear cell leukaemia in rats in the same laboratory was 47% in males and 29% in females. Uncommonly occurring renal tubular cell adenomas or adenocarcinomas were found in male rats (adenomas,1/49 in controls, 3/49 at the low dose and 2/50 at the high dose; adenocarcinomas, 0/49, 0/49 and 2/50, respectively).

Although induction of peroxisome proliferation as a mechanism underlying the hepatocarcinogenic effect of PCE in mice appears attractive, a poor quantitative correlation was seen between peroxisome proliferation and tumour formation in the liver following administration by inhalation.

The induction of kidney tumours observed in male rats provides only weak evidence for (human) carcinogenicity. The observed increase in incidence is not statistically significant. In addition, two different mechanisms of action have been proposed for the induction of these tumours: alpha 2μ-globulin nephropathy, an effect specific for male rats, and formation of genotoxic metabolites in the kidneys as the final step of the glutathione biotransformation pathway. Given the low incidences observed, combined with the data on the mechanism of induction, it can be concluded that the result in male rats is equivocal evidence only for a risk of renal cancer in humans.

A review of six carcinogenicity studies of PCE in rats concluded that evidence of increases in mononuclear cell leukaemia has been seen in two studies in the F344 rat but not in other strains, and that as a rat-strain-specific effect it is not considered to be relevant to the evaluation of human cancer risk (95).

IARC (96) has concluded that the results of the available animal bioassays provide sufficient evidence for carcinogenicity in animals.

Mutagenicity

Comprehensive reviews are available of the data obtained from the many studies carried out with PCE (2,51,96). In vitro studies include tests for gene mutations in prokaryotes and eukaryotes and tests for chromosome aberrations in mammalian cell lines. In vivo studies include tests for chromosome aberrations in bone marrow in rats and mice and a dominant-lethal assay in rats. In addition, several studies have been carried out on DNA damage in vitro and in vivo. The body of results shows absence of mutagenicity of PCE in virtually all of the systems tested. The few weakly positive results may be due to the presence of mutagenic stabilizers in the test samples (2,51). Binding of PCE in vivo to unpurified DNA from several organs was noted in one study, although no covalent binding to purified hepatic DNA could be demonstrated in another. In the other available in vivo studies, it is not known whether the test compound (or its metabolites) reached the target tissues. This observation reduces the value of the negative results observed. Nevertheless, an in vivo study, in which male mice were administered PCE intraperitoneally, assessed the induction of micronuclei in hepatocytes and reticulocytes and showed that PCE might have a clastogenic effect in hepatocytes (97). For the metabolites of PCE that are formed by conjugation with glutathione (a minor biotransformation route demonstrated in rodents), positive results were obtained in in vitro studies in S. typhimurium TA 100 (reverse mutation), with metabolic activation with kidney microsomes, in the presence of glutathione and glutathione transferase (96).

Interactions with other chemicals

Results of studies on the influence of ethanol on the metabolism and toxicity of PCE in liver did not show synergism (2,98). Increased urinary excretion of metabolites of PCE and enhanced hepatotoxicity have been observed in rats following pretreatment with polychlorinated biphenyls (2). Simulation of interaction thresholds for human exposure to mixtures of TCE, PCE and 1,1,1-trichloroethane was done with PBPK models. Model simulations indicated that during combined exposures, certain toxicological effects may be expected to occur at lower exposure levels compared with PCE exposure alone (73,99).

Acute poisoning incidents

Acute exposure to PCE at air levels of 100–200 ppm can cause irritation of the skin, eyes and upper respiratory tract. Non-cardiogenic pulmonary oedema, nausea, vomiting and diarrhoea can occur. CNS effects have also been observed with acute inhalation exposure to PCE at 50–300 ppm. At these levels, neuromotor effects may be seen and results of certain coordination and behavioural tests may be abnormal. At higher air concentrations, loss of consciousness can occur (44).

Several acute PCE intoxications have been reported. In one case, up to 16 mg was ingested by a 6-year-old boy weighing 22 kg. He was admitted to hospital one hour after ingestion, where his condition deteriorated from somnolence to coma. The initial blood level was 21.5 μg/ml. He recovered completely without liver, renal or CNS injury and was discharged on day 9 (100). In another report, a 32-year-old man became semi-comatose and experienced oliguric acute renal failure after accidental ingestion of 75 g PCE. A renal biopsy performed on the 19th day after ingestion showed severe acute tubular necrosis. He regained normal renal function after five haemodialyses and conservative treatment. The amount of PCE in the blood and urine were not measured (101).

Three fatalities due to inhalation of PCE have been reported in the literature. A concentration of 4.5 mg/l was measured in the blood of a man aged 53 years weighing 70 kg, who died while he was cleaning and recycling PCE by distillation (102). A two-year-old boy died due to PCE that had been retained in the curtains of his bedroom. A concentration of 66 mg/l blood was measured (103,104). Finally, a 33-year-old man clearing a line in a dry cleaning establishment died with a blood concentration of 44 mg/l (105). A fatality due to poisoning by both PCE and TCE has been described, a 45-year-old woman being discovered unconscious and in cardiac arrest in a laundry area. Examination revealed that she was deeply comatose and had acute respiratory distress syndrome and severe metabolic acidosis. Cardiovascular instability and acute renal failure occurred and eventually the patient died; the results of the autopsy are described in the paper. The blood levels of PCE 2, 4, 5, 6 and 7 days after hospital admission were respectively 1319, 758, 787, 436 and 656 μg/l (106).

Effects on the CNS

Depending on the concentration, acute exposure can result in loss of coordination, reversible mood and behavioural changes, or potentially anaesthetic effects. People chronically exposed to PCE may experience ataxia, disorientation, irritability, peripheral neuropathy, short-term memory deficits and sleep disturbance. Reversibility depends on the degree of severity of the exposure and associated effects (44).

PCE has been associated with neurobehavioural dysfunction, including reduced attention in humans. In a series of controlled short-term studies in limited numbers of human volunteers, neurological symptoms (including dizziness, drowsiness and decreased functioning in motor coordination tests) and visual system dysfunction were observed at ≥ 678 mg/m³ (≥ 100 ppm) and 339 mg/m³ (50 ppm), respectively (2,16,107,108). Limited information on neurological effects following long-term exposure was obtained in occupational studies in dry cleaning workers. In a cross-sectional study with two groups of exposed workers (exposure concentrations 83 ± 53 mg/m³ and 364 ± 114 mg/m³, duration of exposure 127 or 141 days), small effects on scores in psychological tests were found. However, the response did not correlate with the exposure level in this study (2,6,109).

In another study in female dry cleaning workers with exposures of 6.8–408 mg/m³ (4-hour averages, median 102 mg/m³) and PCE concentrations in the blood of 12–864 mg/l (median 145 mg/l), performance test scores in a test battery for neuromotor functions were decreased. Neither duration of exposure nor blood concentration of PCE was significantly correlated with performance (110). In cross-sectional studies in dry cleaning workers, effects on blue–yellow colour vision were found at mean PCE concentrations of 42–102 mg/m³(111,112). In Italy, PCE-induced impairment of colour vision was studied in 33 dry cleaning workers. Exposure was evaluated with passive samplers and colour vision was assessed. Two years later, the workers were re-examined. The exposure to PCE had increased in subgroup A (median 1.7 ppm vs 4.3 ppm). In subgroup B, exposure was reduced (median 2.9 ppm vs 0.7 ppm). Colour vision worsened in subgroup A but no vision changes were noticed in subgroup B, indicating that an increase in exposure during a 2-year period can cause colour vision to deteriorate (113).

A case report described a 57-year-old woman exposed to supposedly high concentrations of PCE while working at a dry cleaning shop, who suffered from blindness for 9 days in the left eye and 11 days in the right eye with optic neuritis. Blood concentrations of PCE 48 and 80 hours after onset of the optic neuritis were respectively 1.08 mg/g and 0.65 mg/g (114).

In a prospective population-based cohort study on 88 829 children, offspring from dry cleaners (144 children) were followed from birth to age 21–33 years. Preliminary findings suggested an increased incidence of schizophrenia (4 cases) in these children (115). However, it should be mentioned that there was no exposure characterization, there were few cases and few details were provided. These findings need to be corroborated in follow-up studies.

Research from 2002 suggests that chronic, environmental exposure to airborne PCE adversely affects neurobehavioural function in healthy individuals. A significantly lower visual contrast sensitivity was reported in apartment residents exposed to PCE (daytime mean 620 μg/m³, nighttime mean < 100 μg/m³) compared to unexposed controls. The reliability of these data was later discussed in the literature. Methodological limitations preclude a definitive attribution of causation, and further research is necessary to draw reliable conclusions (99,116).

Effects on the liver

Case reports of human exposure to PCE show that it can cause hepatotoxic effects in humans, which include abnormal liver function tests (48), cirrhosis, hepatitis, hepatomegaly and liver cell necrosis (44). A dose–response relationship in humans for the effects on the liver is not completely known. In case studies of high accidental exposures, effects on the liver have been reported. In limited studies in dry cleaning workers exposed to a TWA concentration of 143 mg/m³ (21 ppm) over a 6-year period, serum enzymes indicative of liver function were not affected (62,112). A study in dry cleaners frequently exposed to PCE concentrations of 16 ppm (8-hour TWA) revealed mild to moderate hepatic parenchymal changes as determined by hepatic ultrasonography (117).

Renal effects

Nephrotoxic effects have been described in humans (48,118,119). Symptoms of renal dysfunction, including proteinuria and haematuria, have been associated with accidental exposure to anaesthetic concentrations of PCE vapour (2). In several cross-sectional studies in dry cleaning workers, the effect of PCE on renal function was examined. Female workers exposed for an average of 14 years to an estimated TWA concentration of 68 mg/m³ (10 ppm) had increased urinary levels of lysozyme and ß-glucuronidase suggestive of mild renal effects (120). No effects were observed in workers estimated to have been exposed to a TWA concentration of 142 mg/m³ (21 ppm) for 6 years (62). In a cross-sectional study, about 20 markers of early nephrotoxic effects were measured in workers in dry cleaning facilities (n = 50). Exposure was determined by analysing air samples collected during 4-hour periods randomly selected over the working week. The median exposure concentration was 102 mg/m³ (range, trace–580 mg/m³). Compared to the control population, the exposed group had significantly higher frequencies of abnormal values for a number of the markers in urine, including albumin, transferrin, tissue-nonspecific alkaline phosphatase and brush-border antigens. The significance of the findings cannot be easily assessed but may represent an early stage of clinically silent but potentially progressive renal disease (118). At an average airborne exposure of 8.4 mg/m³ (range 2.2–44.6), a PCE-related effect on the tubular reabsorption of retinol-binding protein in urine was observed in exposed workers from a dry cleaning shop (119).

Since there are sex- and species-dependent differences in bioactivation and detoxification pathways, more knowledge about the metabolism of PCE in several target organs and in different species is needed to improve human risk assessment (50).

Reproductive and developmental effects

A few epidemiological studies have reported reproductive or developmental abnormalities due to exposure to PCE in drinking-water. The 1998 Camp LeJeune study reported a weak association between PCE exposure due to contaminated drinking-water and birth weight outcomes for small-for-gestational age births. Stronger associations were observed between PCE exposure and birth weight for infants of mothers who were 35 years of age or older and for infants of mothers with a history of fetal death (121). Another study found oral cleft defects associated with PCE-contaminated drinking-water. The authors indicate that this study cannot resolve whether the drinking-water contaminants caused the adverse birth outcomes; these findings should therefore be followed up utilizing available drinking-water contamination databases (122).

In a comment on this paper, it is suggested that the oral cleft defects found are biologically plausible (123). According to these authors, a study of mothers with pregnancies affected by neural tube defects in Dublin (124) suggests a mechanism for the induction of neural tube defects related to a metabolic slowdown of the methionine synthase reaction. Recently, two cohort studies examined the effects of prenatal exposure to PCE-contaminated drinking-water on adverse birth outcome and risk of behavioural disorders. It was concluded that prenatal and early postnatal PCE exposure is not associated with disorders of attention, learning and behaviour (125). Further, it was suggested that prenatal PCE exposure does not have an adverse effect on birth weight or gestational duration (126).

The results of several studies have indicated that occupationally exposed women might suffer higher rates of spontaneous abortion (127–129). Exposure was classified by reported work history. A significant effect was found when the work tasks included dry cleaning for at least one hour daily or the women reported handling of PCE at least once a week (128,129). A significantly higher risk was found in operators compared to non-operators (127), but other studies have not found this association (130,131). Furthermore, a case-referent study performed within two cohorts did not give any substantial support to the hypothesis that exposure to PCE at work during pregnancy enhances the risk of adverse pregnancy outcome (132). A retrospective time-to-pregnancy study among Finnish women indicated a reduced ability to reproduce among 20 women exposed to PCE (concentration not clear) for 1–4 days a week or daily by inhalation. It should be noted that the studies have a number of limitations, primarily that exposure concentrations are not available (133).

The Camp Lejeune study of ATSDR (134) was not included in this guideline, since an error was made in the exposure classification; these data will be re-analysed by ATSDR at a later stage.

Mutagenic and carcinogenic effects

IARC concluded in 1995 that there is evidence for consistently positive associations between exposure to PCE and the risks for oesophageal and cervical cancer and non-Hodgkin's lymphoma. These associations appear unlikely to be due to chance, although confounding factors cannot be excluded and the total numbers in the cohort studies combined are relatively small. IARC therefore concluded that there is limited evidence for the carcinogenicity of PCE in humans (96).

PCE is reasonably anticipated to be a human carcinogen on the basis of limited evidence from studies in humans and sufficient evidence of carcinogenicity from studies in experimental animals. PCE has been studied by observing laundry and dry cleaning workers, who may also have been exposed to other solvents, especially TCE but also petroleum solvents. In several cohort and proportionate mortality studies, excesses have been reported of lymphosarcomas, leukaemias and cancers of the skin, colon, lung and urogenital tract. Some excess of lymphomas and of cancers of the larynx and urinary bladder was seen in a large cohort of dry cleaners. A familial cluster of chronic lymphocytic leukaemia has also been related to dry cleaning. Although these studies suggest a possible association between long-term occupational exposure to PCE and increased lymphatic malignancies and urogenital cancers, the evidence must be regarded as inconclusive because workers were exposed to petroleum solvents and other dry cleaning agents as well as PCE (135).

The results of a study in differently exposed dry cleaners and launderers indicated a reduction in oxidative DNA damage in exposed dry cleaners compared to launderers. The mean value for PCE in the blood of these dry cleaners was 0.078 mg/l, and the TWA was < 5 ppm in air for 17 out of 18 dry cleaners. However, PCE could not clearly be identified as the source of the effect (136). In two studies for genetic effects (sister chromatid exchanges and/or chromosome aberrations) in lymphocytes of occupationally exposed workers, no clear-cut effects were found (6). Another study showed modulation of the expression of some genes related to cancer induction in human cord blood cells (137).

A population-based case-control study was undertaken to investigate the association between breast cancer and PCE exposure from public drinking-water (drinking-water consumption as well as bathing habits were taken into account). The same author performed a more extensive case-control study a few years later. Both studies suggest that women with the highest PCE exposure levels have a slightly to moderately increased risk of breast cancer. Analyses of the same data using a newly constructed personally delivered dose model to make a more accurate estimation of the exposure did not significantly change the subjects' exposure classification (138–140). An association between PCE exposure from public drinking-water and lung cancer and possibly colorectal cancer was suggested, based on another population-based case-control study (141).

A paper from 2003 assessed (based on specified methodological and scientific quality criteria) 44 epidemiological papers that provided reasonable data on up to 17 cancer sites. The authors concluded that no evidence for an association between breast, prostate, skin or brain cancer and exposure to PCE was demonstrated. A relationship between PCE and cancer of the oral cavity, liver, pancreas, cervix and lung was considered unlikely by the authors. Scientific evidence was considered inadequate for laryngeal, kidney, oesophageal and bladder cancers (142).

In 2006, the epidemiological literature since 1990 on occupational chlorinated solvent exposure was reviewed. The paper reviewed 28 studies of PCE, all but 5 of which focused on workers in the dry cleaning industry with few or no other chemical exposures. Within the study populations with multiple exposures, three studies found some increased risk of biliary and liver or brain cancer; in the other studies, no increased risk for renal cell carcinoma or for any cancer was found. Census records for those working as a “dry cleaner or laundry worker” or from a database of those being monitored for PCE exposure were linked to cancer registries (three studies) and statistically significant increased risks for liver and pancreatic cancer, Hodgkin's lymphoma and leukaemia were found. In general, findings in occupational cohort studies have not been consistent, even within the same industries (143).

CNS effects and local irritation

The evidence is sufficient to conclude that PCE can cause irritation of mucous membranes and the respiratory tract. These effects generally occur at high concentrations. Acute exposure to PCE at air levels of 680–1360 mg/m³ (100–200 ppm) can cause irritation of the skin, eyes and upper respiratory tract in humans. Moreover non-cardiogenic pulmonary oedema, nausea, vomiting and diarrhoea can occur. CNS effects have been observed at high dose levels in rats and mice. In humans, CNS effects have also been observed with acute inhalation exposures of 340–2040 mg/m³ (50–300 ppm) of PCE. The evidence is sufficient to conclude that there is an association between PCE and CNS effects at sufficiently high concentrations. Since high concentrations typically do not occur in indoor environments for a long period of time, these end-points are not taken into account when setting the guideline value.

Liver and kidney effects

The evidence of an association between PCE exposure and effects on the liver is suggestive. Hepatic effects have been observed in rats and mice in subchronic studies. Case reports of accidental exposures of humans to PCE confirm that hepatotoxic effects can occur in humans, including abnormal liver function tests, cirrhosis, hepatitis, hepatomegaly and liver cell necrosis.

There is suggestive evidence of an association between PCE exposure and effects on the kidneys. In chronic studies, renal effects were observed at 1360 and 2710 mg/m³ (200 and 400 ppm) in rats and at 680 and 1360 mg/m³ (100 and 200 ppm) in mice. Reversible kidney damage in humans (e.g. proteinuria and haematuria) has been associated with accidental exposure to acutely toxic concentrations of PCE vapour. A long-term exposure study showed that renal effects may develop at lower concentrations, with the reported onset of renal damage occurring following exposure to a median concentration of 102 mg/m³ (range, trace–576 mg/m³).

Reproductive and developmental effects

A few epidemiological studies have reported developmental abnormalities due to PCE exposure in drinking-water (reduced birth weight and oral cleft defects). Other studies concluded that prenatal PCE exposure does not have an adverse effect on birth weight or gestational duration. Some studies have consistently indicated that occupationally exposed women might suffer higher rates of spontaneous abortion, while others found no association. All the above-mentioned studies have limitations, primarily that the exposure concentrations are not available. Therefore the evidence stays suggestive.

Cancer risk

IARC concluded that the results of the available animal bioassays provide sufficient evidence for carcinogenicity to animals. In these studies, an increased incidence of adenomas and carcinomas was observed in the livers of exposed mice. There is suggestive evidence from mechanistic studies that humans are likely to be less sensitive to the development of these tumours following PCE exposure. A low incidence of kidney tumours among male rats has been reported.

It can be concluded from this small and statistically non-significant increase, together with the data relating to the possible mechanism of induction, that the result in male rats is equivocal evidence only for a risk of renal cancer in humans.

The significance for humans of the increased incidence of mononuclear cell leukaemia, as observed in two studies in F344 rats, is unclear. This is due to the lack of understanding of the mechanism underlying the formation of this cancer type (which has a high background incidence) and the suggestion that it might be a rat-strain-specific effect.

IARC concluded in 1995 that there is evidence for consistently positive associations between exposure to PCE and the risks for oesophageal and cervical cancer and non-Hodgkin's lymphoma.

These associations appear unlikely to be due to chance, although confounding factors cannot be excluded and the total numbers in the cohort studies combined are relatively small.

A paper from 2003 assessed (based on specified methodological and scientific quality criteria) 44 epidemiological papers that provided reasonable data on up to 17 cancer sites. The authors concluded that no evidence for an association between breast, prostate, skin or brain cancer and exposure to PCE was demonstrated.

A relationship between PCE and cancer of the following sites was considered unlikely by the authors: oral cavity, liver, pancreas, cervix and lung. Scientific evidence was considered inadequate for laryngeal, kidney, oesophageal and bladder cancers. Since then, new studies have been published, but the evidence concerning the carcinogenicity of PCE is still not conclusive. The evidence for an association therefore remains suggestive.

From the weight of the evidence from mutagenicity studies, there are no indications that PCE is genotoxic. Although several in vitro studies indicate that conjugation of PCE with reduced glutathione (a minor biotransformation route demonstrated to occur in rodents) produces renal metabolites that are mutagenic in S. typhimurium TA 100, the significance of the latter results for humans is doubtful. In two studies investigating genetic effects (sister chromatid exchanges and/or chromosome aberrations) in lymphocytes of occupationally exposed workers, no clear-cut effects were found.

In conclusion, carcinogenicity is not used as an end-point, since there are no indications that PCE is genotoxic and there is some uncertainty about the epidemiological evidence as well as the relevance of the animal carcinogenicity data to humans. However, because of the remaining uncertainty about the carcinogenicity of PCE, it should be kept under review.