Document npe0kN50OdQooNQqBOGLG4Nba

jver Disease Associated With Exposure [o 1,1? 1-Trichloroethane ilichael J. Hodgson, MD, MPH; Alma E. Heyl, RT; David H. Van Thiel, MD 1,1,1-Trlchloroethane Is a halogenated hydrocarbonsolvent feommoniy used in industry because of its supposed lack of atotoxicity. Nonetheless, animal studies performed byseverIndependent groups have shown the solvent to Induce fat Reposition, vacuolar degeneration, and centrilobular necrosis, anges similar to those seen afterexposure to carbon tetrachloa, albeit of a much reduced magnitude, in animais exposed to agent. Four patients with fatty liver disease whose work ntailed substantial exposure to this agent were seen at the niversity of Pittsburgh (Pa). Based on this clinical experience, > believe that 1,1,1-trichioroethane should be reconsidered as i agent with potential hepatotoxicity in man. (Arch Intern Med. 1989;149:1793-1798) eath from cirrhosis and liver disease ranks 10th among the ?leading causes of death in the United States and 9th among the leading causes of lost years of life before the age of 65 years.1,2 Although fatty liver disease (FLD) need not neces sarily progress to cirrhosis, it is a well-known antecedent of this condition. The prevalence of FLD has been reported in several autopseries of accidental deaths. Kuller et al* described FLD as p big present in 111 of 428, or 25%, of sudden, nontraumatic ^deaths occurring between the ages of 25 and 64 years in Baltimore, Md. In that series, the proportion of cases of FLD attributable to alcohol was unknown. It is of some interest t Ground,4 after excluding alcohol-related fatalities, re nted FLD as being present in 21% of individuals between ie ages of 18 and 58 years who died in road or aircraft idents in Great Britain in 1984. Similarly, Alcastra5 re nted FLD in 32% of 1534 autopsies performed in Spain, of rhich less than half could be attributed to alcohol. Moreover, lartz and Komhuber6 reported FLD in 84% of men and 72% fwomen who died in accidents in Germany; rates one third as gh were recorded in nonaccidental deaths. Fatty liver disie has been described in 60% to 90% of morbidly obese idividuals and in an uncertain proportion of diabetics.'40 Nine Accepted for publication January 13,1989. From the Occupational and Environments Medicine Program, Division of *eral Medicine (Dr Hodgson), and the Division of Gastroenterology (Ms *yl and Dr Van Thiel), Department of Medicine, University of Pittsburgh School of Medicine. Reprint requests to Occupational and Environmental Medicine Program, tpnniviviersity ofPittsburgh SdiodofMedidne, LothropHall 149,190LothropSt, ;t*burgh PA 15261 (Dr Hodgson). ^ch Intern Med-Vol 149, August 1989 percent of persons undergoingjejunoileal bypass for therapy of obesitywho have FLD go on to develop cirrhosis.11 Similar ly, FLD associatedwith diabetes mellitus or obesity is known to progress to cirrhosis,10 Despite this sequence of events, many primary care physicians ignore the minor elevations in liver injury parameters frequently seen in patients, particu larly in those with fatty livers as defined by computed tomo graphic scanning techniques. This lack of interest exists de spite the fact FLD may be an early indicator of progressive liver disease. Clearly, most cases of FLD are attributed to alcohol abuse. However, only rarely are searches for additionalenvironmental or occupational exposures even undertaken when physi cians are faced with a nonalcoholic with a fatty liver. This is the case despite the fact that styrene,other aromatic com pounds,14"17 and trichloroethylene (inhaled during glue sniffing)&"have been implicated as important agents in the patho genesis of liver disease in humans, and these agents are thought to be rare contributors to the problem of cirrhosis on a population basis. On the other hand, 1,1,1-trichloroethane (1,1,1-TCE) is not considered to have great hepatotoxic potential primarily on the basis of several short-term human laboratory exposure studies, -- one cross-sectional human study in the work place, " and several animal studies. -- As a result of its per ceived low hepatotoxicpotential, it is widely used in industry. Nevertheless, Klaassen andPlaa28 have described hepatotox icity after 1,1,1-TCE and trichloroethylene exposure in dogs, which pathologically appeared to range in severity some where between that of tetrachloroethylene and dichloromethane. In rats, hepatotoxicity of 1,1,1-TCE has been seen at 200 times the dose required to produce hepatotoxicity with car bon tetrachloride. Mt^utt et al" described acute inflamma tion, centrilobar swelling of hepatocytes, vacuolar degenera tion and necrosis, and an increase in triglyceride deposits after 1,1,1-TCE exposure. These changes closely resembled those induced by carbon tetrachloride. Finally, Thiele et al" have reported a case of FLD that, although originally induced by trichloroethylene exposure, appeared to have its progres sion and improvement clearly related to an associated 1,1,1TCE exposure. Zimmerman11 has reviewed the factors associated with in creasing hepatotoxicity resulting from halogenated hydro carbon exposure. He states: "lbxicity appears inversely pro portional to the negative charge on the halogen atom, to Liver Disease--Hodgsonetal 1793 carbon-halogen bond energy, and to chain length and directly proportional to ease of homolytic cleavage, number of halo gens in the molecule, and to the atomic number of the halo gen." Based on this algorithm and on findings of the available animal studies, 1,1,1-TCE would be expected to have some hepatotoxicity potential. We report four cases of FLD associated with exposure to 1,1,1-TCE seen at the University of Pittsburgh (Pa) in the last 2 years. Three of the first 10 patients seen in our occupa tional medicine clinic in a study assessing the risk factors for FLD had very similarjobs entailing substantial exposure to this commonly used industrial solvent. The fourth patient, who required liver transplantation, had had intermittent sub stantial exposures. Although a case series such as this does not prove a causal association, the frequency of the associa tion was striking and raises the question of a causal associationbetween 1,1,1-TCE and FLD. REPORT OF CASES Case 1. --A 28-year-old white man was referred for the evaluation of FLD. An occupational history revealed that he began working in a powdered metal plant in 19& asa "quality control inspector." Hisjob included cleaningparts with 1,1,1-TCE in a cold bath approximately 1 to 2 hours per day. In September 1985, hisjob station was moved to within4.6 cm (15ft) of two hot tanks in which 1,1,1-TCE was used as a degreasing agent. No direct exposure to other hepatotoxins could be identified. The plant management at the patient's place of employ ment stated that perchloroethylene was used at times on rags to wipe parts clean. Industrial hygiene studies, obtained after diagnosis, showed 8-hour time-weighted average exposures of 1,1,1-TCE else where in the plant (after physical modification of the patient's work site) in the range of approximately one tenth of the permissible exposure level. No estimates were available for his former job. Before working for his current employer, he had worked in a textile mill as a spool changer, as a clerk in a grocery store, and as a moldingpress operator in a powdered metal plant, all without known expo sure to hepatotoxins. A history of present illness revealed that in the fall of 1985 the patient noted the gradual onset of anorexia, fevers, chills, and unin tentional weight loss, and suffered recurrent, episodic illnesses that were never diagnosed or treated specifically. He was removed from work for lOdays and told to abstain from alcoholbecause ofthe clinical diagnosis of hepatitis, with resulting improvement in the assessed liver injury parameters and improvement in his symptoms. When he returned to work, with recurrence cf exposure but with continuing abstinence from alcohol, his symptoms recurred and slowly worsened through the fall of 1986, when he was again removed from work. No pertinent medical, social, or family history of liver disease was obtained. The patient had consumed on average six to eight 360-mL (12-oz) cans of beer per weekend since the age of 18years. However, he failed to meet criteria fora diagnosis of alcoholabuse oralcoholism accordingto the Diagnosticand StatisticalManual cf Mental Disor ders (HxLrdEdition) and did not score excessivelyon the Alcohol Use Inventory," a standardized measure for detecting past as well as current alcohol abuse or addiction. The patient was shown to have FLD by liver biopsy. His clinical characteristics and laboratory results are listed in Table 1. Of particu lar interest is that his serum alanine aminotransferase levels were consistently greater than his aspartate aminotransferase levels, a finding quite atypical for alcohol-associated liver disease. He had never been exposed to blood or blood products and had no other risk factors for chronic viral hepatitis. Serologicaltests for viral hepatitis were negative. After the patient was no longer exposed to 1,1,1-TCE, his symp toms and liver injury parameters again improved. A subsequent liver biopsy showed markedly less steatosis. Case 2. -- A27-year-old white man was referred for the evaluation of toxic hepatitis. An occupational history revealed that in 1980 he had begun working as a process technician for a specialty steel manufacturer, with minimal exposure to all finishing processes. His work primarily involved exposure around a heated 1,1,1-TCE tank used for degreasing finished pieces. He was exposed between 5% and 1794 Arch Intern Med--Vol 149, August 1989 100%cf an 8-hourworkday. No separate fume hood was avahabij, the tank. He also had occupational exposures to perchloroettny and 1,1,1-TCE in cold form, as he used both substances dla weekly, for periods cf generally less than 1 hour, on a wipe rag s open cans used for brush cleaning. Industrial hygiene sampl3 1,1,1-TCE obtained over an 11-year period were reviewed. " ranged from 35 ppm (one tenth of the permissible exposure le 350 ppm) to 17718ppm (50-foldthe permissible exposure level),; Approximately 6 months (November 1983) before referral^ diagnosis, the patient developed progressive fatigue, and 4 mo before referral began to note intermittent but progressive: py The month before evaluation (March 1984), he put in a conside amount of overtime and noted that all of his symptoms worsen diagnosis cf "hepatitis" was made on the basis of elevated , enzyme levels (Table 1). No prior measures of liver injury available. An upper gastrointestinal series, barium enema, liver-spleen scan were performed and the findings were interpri as normal. The patient's medical, family, and socialhistories were noraeonlSJ utory, He consumed approximately one six-packofbeer per week had never been exposed to blood or blood products and had no < risk factors for chronic viral hepatitis. Serologicaltests for hepatitj were negative. The results cf physical examination were normal. Liver _ parameters and the findings of the liver biopsy are shown in Tab As in the previous case, the alanine aminotransferase levejs' always greater than the aspartate aminotransferase levels, and! level of y-glutamyltransferase was only minimally increased.. Case 3. --A 45-year-old white man was referred for evaluation abnormal results on liver function studies. An occupational 1 revealed that he had worked as a machinist for 19years in a tool-al die section. He cleaned parts by dipping them into cold 1 1, 1-TJ Although he occasionally used several other nonhalogenated hyd carbons and one halogenated paraffin in his work, each of 1 agents was considered to possess even less hepatotoxicity than l,f TCE. He had, on very rare occasions, used perchloroethylene, i ' wipe rags. A review of the patient's medical history revealed that in 1981 elevation in alanine aminotransferase level (twice normal) was ide fied. Serological tests performed at that time were negative! hepatitis A and B. In 1983 the patient reported recurrent episode*! abdominal- pain and cramping. In 1984 aspartate aminotransfer and y-glutamyltransferase levels were found to be just above i upper limit ofnormal. The patient's medical, social, and family I ies were noncontributory. The patient credibly denied all aleoli consumption and scored appropriately on the Alcohol Use Inv tory.32 He had never been exposed to blood or blood products and 1 no other risk factors for chronic viral hepatitis. Serological testsl hepatitis B were negative. Physical examination revealed a mildly enlarged firm liver wit! rounded edge. An upper gastrointestinal series, barium enema, anc{ liver-spleen scan were performed and the findings were interpret! as normal except for the presence of possible gastric varices. Upp gastrointestinal endoscopy confirmed the finding of esophageal i gastric varices; liver biopsy revealed fatty metamorphosis and i tive cirrhosis. Case 4.--A 50-year-old man was referred for evaluation for liv transplantation because of solvent-induced cirrhosis.' An occupatio al history revealed that he had worked for a public utility machinist beginning in 1975. In his first 5 years he rememb working with trichloroethylene on approximately five occasions, ginning in 1980, he worked with cold 1,1,1-TCE one to four 8-h shifts at a time. He estimated his exposure to that agent at appro mately 10% of his working hours. In addition, he was in the vidn while electricians sprayed 1,1,1-TCE on several occasions (simu tions revealed exposures in the range of 350 to 500 ppm). He r worked as a lathe operator in a variety of tool-and-die and mac shops from 1957 to 1975. Only between 1963 and 1965 was there even* remote possibility of solvent exposure, through the presence off vapor degreaser far away from his job station in his workplace. He presented with fatigue, nausea, and pain in the right upp quadrant of his abdomen in 1985. Physical examination revealed < obese man whose liver was firm with a rounded edge. Because ofliv function abnormalities in a never-in-a-lifetime drinker, a biopsy *P" imen was obtained and demonstrated cirrhosis. Moreover, the 1 tient had never been exposed to blood or blood products and hadl Liver Disease--Hodgson ffl othe heps  Age, y Height. em Weight kg Quetelet indexf Laboratory studies ALT IU/L (<37) AST, IU/L (<34) (Table 1.--Clinical Parameters and Uvier Biopsy Results* . .,, 1 28 170 98 34.2 2 27 175 90 29.4 Com No. 3 45 163 76 29.2 96*1 36 45-72 24-121 25-45 97-151 46-98 4 51 165 99 36.7 38-38 52-64 34-53 41 58 Alkalrie phosphatase. IU/L <<100) 79-81 85-90 30-97 115-129 Albumin, g/L {3.5*5.0) 4.1 3.3 4.6 2.4-3.6 Prothrombintime, s (11-13) 13.1 (11.8) 12.4 (11.8) 14.1 (11.8) Jver size (by CT scan), cm*3 *15* 0200-1500) 2761 1779 1289 Jver biopsy Histologic fndings Macrovesicular :rovesicular Macrovesicular Macronodular cirrhosis with dysplastic nuclei and piecemeal necrosis Degree of fat 4+ 1 + 3+ 3+ Zone III Mixed l<lll III Perivenular fibrosis + Sinusoidal fibrosis + SI an ientie for lesof erase e the ;storcohoi Periportalfibrosis Cellular infiltrate + Polynuclear cells Mononuclear and polynuclear cells, eosinophils + Mononuclear cells + Mononuclear cells Architecture Intact Intact Intact Intact ' indicates aanine aminotransferase; AST; aspartate aminotransferase; vGT y-glutamyttransferase; CT, computed tomographic; plus sign, present; an j 1 sian. absent. (ht in kilocyams divided by height in meters squared. nvend '".ad ts ?or risk factors for chronic viral hepatitis. Serological tests for at tis B were negative. TCE is reportedly a very frequently used solvent, we have seen no patients with such exposures with other forms of liver COMMENT disease among hospitalized patients at our institution and were therefore struck by the frequency ofthe association. , J1 four patients presented with FLD without other predisjng causes, although two can be classified as obese (Quetef index greater than 30). Alcohol consumption was not in tht The industrial and environmental use of 1,1,1-TCE as a solvent has increased substantially during the last several decades, because of its assumed lack of hepatotoxicity. This S- 5>ge associated with steatosis (maximum, 66 mL [2.2 oz] pet lyX The one patient whose pattern of liver injury parame- assumption, however, is based on the findings of a single cross-sectionalstudyand several laboratory exposurestudies y Ws was consistent with alcohol injury had never consumed > 5hol. None of the patients had received blood or blood performed on a few humans and ignores a body of data ob tained in animal studies that suggest that 1,1,1-TCE maybe | pxiucts, was diabetic, had sexual habits associated with ; RHl hepatitis, or demonstrated serological evidence of prior hepatotoxic. Because the usual criteria for liver toxicity, such as alter 1 ipatitis B or other viral infection. The primary exposure ir 'IBfburwas 1,1,1-TCE, although minimal exposures to other 1 hepatotoxins (perchloroethylene and ethanol) could be ations in the results of liver injury tests, are not readily translated into dose-response curves, Plaa et al" evaluated the toxicity of seven halogenated hydrocarbons in mice by . .. dented in three and two of the four cases, respectively. Nonalcoholic FLD is a well-recognized though poorly stud ' ^syndrome.11 Known risk factors include obesity, diabetes, measuring sleeping time, sulfobromophthalein retention time, and morphological characteristics of the liver. In these studies, 1,1,1-TCE demonstrated the least toxicity of the ; intestinal bypass surgery. Although between 40% and several agents investigated but nonetheless showed hepato 5 f morbidly obese uatients Dresentinv for theraDeutic ftervention have FLD on biopsy, only 2.7% of morbidly toxic potential. In a later study, Klaassen and Plaa compared the hepato mu!86 individuals in a working opulation had evidence of toxicity of seven different compounds by constructing dose- "normal results on liver tests! Hepatotoxjn exposure js VI erally considered a rare cause, although, to our knowl- response curves for aspartate aminotransferase and liver histology, They divided the hydrocarbons they studied into no reports of nonalcoholic FLD have systematically stigated the presence of liver toxins. " " Although 1,1,1- $. IS n e'fJc.Afch Intern Med-Vol 149, August 1989 two groups, those causing centrilobular necrosis at nearlethal doses (carbon tetrachloride, chloroform, and 1,1,2- Liver Disease -- Hodgson et al 1795 TCE) and those that did not (1,1,1-TCE, dichloromethane, trichloroethylene, and tetrachloroethylene). At 50% of the dose necessary to cause measurable organ dysfunction, only 1,1,1-TCE failed to cause vacuolization of centrilobular hepatocytes. The duration of induced alanine aminotransferase abnormalities, on the other hand, was similar for the five hydrocarbons other than chloroform and carbon tetra chloride. Prendergast et al" examined rats, guinea pigs, and several other types of animals under long-term (90 days) continuous and intermittent (5 days per week, 8 hours per day) exposure to 1,1,1-TCE. No evidence of liver injury was seen after continuous exposure at 370 or 135ppm, although at the latter dose a surprising number of animals died. McNutt et al" exposed mice over varying periods of time to 1,1,1-TCE and found changes closely resembling those in duced by carbon tetrachloride. They described acute inflam mation, centrilobar swelling of hepatocytes, vacuolar degen eration and necrosis, and anincrease in triglyceride deposits, all changes consistent with and suggestive of FLD. Further supporting studies, published earlier, were sum marized by Aviado et al." These authors reviewed the hepatotoxicity of 1,1,1-TCE in dogs, rabbits, guinea pigs, rats, and mice, as known before 1976. A series ofstudies performed at the Dow ChemicalToxicology Research Laboratories, Midland, Mich,20,28 demonstrated that a concentration of 500 ppm of 1,1,1-TCE was safe for animals, even after long-term exposures. These studies and two others21,28 also examined the effects of 1,1,1-TCE vapors on humans. The results are summarized in Table 2. Only after the induction of anesthesia were minimal functional abnor malities of the liver demonstrated. On the other hand, the longest duration of exposure was only 5 days, and no long term f"ol"low-up was perrformed'. Finally, one cross-sectional study has been published com paring workers exposed to 1,1,1-TCE with an unexposed population." Exposures over a 2-year period ranged from 100 to 350 ppm. Mean alanine aminotransferase levels were actu ally significantly lower in the exposed workers than in the unexposed workers. Recently, such phenomena in cross-sec tional studies have been interpreted as evidence for a "survi vor effect." If ill individuals leave the work force prefere: ly, the remaining members are "healthier," showing laboratory abnormalities and better survival charade: Interestingly, a greater proportion of exposed than _ posed workers (28 vs 21) experienced "abnormalities of. gastrointestinal system" by history, although the signify of this finding is uncertain. The excess symptom rate co interpreted as an early indication of liver disease not associated with abnormal results of liver tests. When d: then becomes severe enough to produce measurable abi malities, individuals would leave the work force. ) Thiele et al30 described a patient with liver disease that initially thought to have been caused by exposure to trii roethylene. After removal from exposure and subsequi improvement, exposure to 1,1,1-TCE was associated ' progressive deterioration of liver function. Cessation of e: sure was again temporally related to improvement. Se' earlier reports concerning acute toxicity in humans37-30 also been published. Two independent mechanisms are considered to contribl to the hepatotoxicity ofhalogenated hydrocarbons. First, blockage of nascent lipoprotein coupling with triglyce: leads to accumulation of triglyceride within cells. This mulation of fat can lead to progressive liver injury." Ho1 er, the exact cause or mechanism responsible for fat-indui hepatic necrosis is unclear. The second hypothesis is ti FLD may be related, at least in part, to the production' toxic-free radicals and/or peroxidation injury.11-15 In genef metabolites formed in phase lreactions (mixed-function o: dase reactions) are more toxic than are their parent con pounds."Thus, even xylene and other nonhalogenated hyi carbons, currently considered nontoxic, can give rise to to: intermediaries as a consequence of mixed-function oxi metabolism prior to their excretion. Alcohol, which is thoui to lead to hepatic damage, at least in part, as a result of' radical formation,17-" has been shown to potentiate the to: effects of organic solvents and drugs that can exert at 1 some of their toxicity through lipid peroxidation.31,50-,w Recei evidence indicates that short-cnain aliphatic hydrocarb without intrinsic hepatotoxicity may potentiate halogeenaiB hydrocarbon exposures; eg, acetone increases both the acui r w P bk k at P1 T at h> St nt le w dr it;. > --Summary of Hun____ ... Chamber Studies pf t,1,1*Trichloroethane Source, y Torkelson et al,201958 Stewart etal,21 1961 No. of Subjects ?1 ?1 ?1 ?1 ?1 6 6 3 2 3 7 Exposure Concentration, ppm 546 506 1000 920 1900 500 500 1000 1000 1000 To anesthesia {<2650) Duration, min 90 450 30 70-75 5 78 186 73 35 20 15 Rowe et al,22 1963 Stewart etal,22 1969 NM 2-5 (a total of 11 volunteers in 31 exposures) 520 500 420 6.5-7 *NM indicates not mentioned; AST, aspartate aminotransferase; and LDH, lactate dehydrogenase. fTwo of the seven subjects developed transient elevation of uroporphyrin excretion. Frequency (No. of Times Exposed) 1 1 1 1 1 1 1 1 1 1 1 1. 4. 5 Health Effects/ Liver Injury NM NM NM NM NM Normal Normal Normal Normal Normal Normalt None AST, LDH unch 1796 Arch InternMed-Vol 149, August 1989 Liver Disease--Hodgsons1* 19.- at: k: Dfthe Arc ffentia]. ? fewer 'ristics. 1 unex" f the Seance fjuld be ;3t was riehlo-quent 1 with fexpoeveral " have chronic hepatotoxicity of carbon tetrachloride in pate." be histologic characteristics of FLD are well described, ently, most investigators classify it by its zonal distribuwithin the liver and according to the size of the fat plets within hepatocytes.81 Although a considerable bunt of literature has been accumulated on drug-induced atitis, little information is available on the importance of upational and environmental risk factors for liver disease eneral and for steatosis specifically. In one recent study of 1 donors with elevated alanine aminotransferase levels," , were obese and 63% consumed alcohol "daily." No meni was made of occupational or environmental exposures to toxins. Moreover, the role of steatosis as a cause for or decessor of cirrhosis in nonalcoholics who are neither > nor diabetic and who show no evidence of viral disease i not been assessed. Exceptions are the recent epidemics ilting from exposures to kepone,59 trichloroethylene,18 and nethylformamide.80 However, only chloroform- and carbon achloride-induced liver disease (both steatosis and cirsis) have been well described and aregenerally accepted to 'causes of both acute hepatotoxicity and chronic liver se. uro of our four patients were clearly obese, ie, 30%over sight . The other two were not thin. Two mechanisms might |y a role involving the interaction of obesity and hydrocar- h exposure in the development of liver disease. Fust, it is l that the potentiation of hydrocarbon toxicity by alco- is dependent in part on the relative timing of alcohol ninistration.61 The presence of obesity may continuously jsent the liver with more substrate to form more epoxides, ge continuous low-level presence of free radicals may then additional exposure to the more potent halogenated ocarbons and thereby induce overt damage more readily. ond, fatty tissue may itself serve as a reservoir of haloge- ' hydrocarbons. These would then be continuously re ed over a much longer period of time than in individuals l less body fat, leading to an effectively much longer ation of exposure. A final additional contribution to toxic- Fin these cases may result from minimal exposures through contamination of commerciallyavailable 1,1,1-TCE or ground water with minimal amounts of solvents with substantially greater hepatotoxicity, such as trichloroethylene and perchloroethylene. Although we cannot absolutely discount such ''part per million" contamination, we think it unlikely to play a major role. In animal experiments, alcohol potentiates the effects of other hepatotoxins. Three of the four patients consumed minor quantities of alcohol, of a magnitude not generally associated with the development of FLD. We obviously can not exclude an interaction between the toxicities of alcohol and 1,1,1-TCE. Both obesity and alcoholmay therefore have played a role in the pathogenesis of liver disease in these cases. In summary, 1,1,1-TCE has been assumed to be free of hepatotoxicity in humans because the scanty evidence avail able has failed to demonstrate clear evidence of hepatotoxic-- ity rather than because it has been studied extensively--a type II error, or lack of power in documenting safety, rather than the certainty of safety. In addition to epidemiological evidence, animal bioassays and structure-activity relation ships are commonly used by toxicologists in interpreting possible causal associations. Although manyphysicians hesi tate to use animal studies to support causal associations in humans, in this case human hepatotoxicity might reasonably be expected, based on the evidence from animal studies. Human studies have not demonstrated lack of effects. Case series such as this cannot conclusively document the hazard ous nature ofan agent. Still, we are left with an agent (1)that has been shown to induce liver disease in animals, (2) that is associated with some degree of liver dysfunction in humans after substantial exposure, and (3)that has not been shown to be safe. On the basis ofthese four cases, we believe that 1,1,1TCE maybe apotential hepatotoxin in humans. This work was supported in part by grants from the National Institute for OccupationalSafety and Health, Centers forDisease Control, Atlarta,Ga (KOI OH00071-01), and the National Institutes of Health, Bethesda, Md (NIDDK 3266 and NIAAA 06601). The authors wish to thank Daniel H. Gregory, MD, John Pantalone, MD, Walter Ubinger, MD, and David Dunkert, MD, for the referral of their patients. References : Premature mortality-- United States, 1983. MMWR. 1986;35:357-366, Changes in premature mortality--United States, 1983-1984. MMWR. 6;35:29-31. ( Kuller LH, Perper JA, Cooper M, Fisher R An epidemic of deaths buted to fatty liver in Baltimore. Prev Med. 1974;3:61-79. , Ground KEU. Prevalence of fatty liverin healthy male adults accidentally id. Axial Space Environ Med. 1984;55:59-61. Alcastra MA. Incidencia de aapectos etiopalogenicos de la esteatosis: sion de 1534 autopsias. Rev Clin Esp. 1981;161:79-83. Hartz F, Komhuber HH. Haeufigkeit von Fettleber in Deutschland. JtMedWochcnschr. 1985:110:1232. E Kern WH, Heger AH, Payne JH, DeWind LT. Fatty metamorphosis of liver in morbid obesity. Arch Pathol Lab Med. 1973;96:342-346. Anderson T, Gluud C. Liver morphology in morbid obesity: a literature ly. Int JObes. 1984;8:97-106. . Nasrallah SM, Wills CE, Galambos JT. Hepatic morphology in obesity. jDisSci. 1981;26:325-327. . Van Thiel DH. Diabetes mellitus and hepatobiliary disease. Curr Con Gastroenterol. 1985;10:3-11. Schaffher F, Thaler H. Nonalcoholic fatty liver disease. In: Popper H, Tier F, eds. Progress in Liver Disease VIII. New York, N Y Crune & on; 1986:283-298. Harkonen H, Lehtniemi A, Aitio A. Styrene exposure and the liver. IJ Work Environ Health. 1984;10:59-61, Edling C, Tageson C. Raised serum bile acid concentrations after occuonal exposure to styrene: a possible sign of hepatotoxicity. Br J Ind Med. *;41:257-259. Id- Kurppa K, Husman K. Car painters' exposure to a mixture of organic nts: serum activities of liver enzymes. ScadJ Work Environ Health. 8:137-140. 1 InternMed--Vol 149, Augustl989 15. Axelson O. Solvents and the liver. Eur J Clin Invest. 1983;13:109-111. 16l Hansbrough JF, Zapata-Sirvent R, Dominic W, a ill tun ,T, Boswick J, WangX-W. Hydrocarbon contact injuries. J Trauma. 1985;25:250-252. 17. Dossing M, Arlien-Soborg P, Petersen LM, Ranek L. Liver damage associated with occupational exposure to organic solvents in house painters. Eur J Clin Invest. 1983;13:151-157, 18. Baerg RD, Kimberg DV. Centrilobular hepatic necrosis and acute renal failure in solventstuffers, Ann Intern Med. 1970;73:713-720. 19. Clearfield HR. Hepatorenal toxicity from sniffing spot-remover (trich loroethylene). DigDisSci. 1970;15:851-856. 20. Torkelson TR, Oyen F, McCollister DD, Rowe VK. Toxicity of 1,1,1trichloroethane as determined on laboratory animals and human subjects. Am Ind Hyg Assoc J. 1958;19:353-362. 21. Stewart RD, Gay HH, Erley DS, Hake CL, Schaffer AW. Human exposure to 1,1,1-trichloroethane vapor: relationship of expired air and blood concentrations to exposure and toxicity. Am Ind Hyg Assoc J, 1961;22:252-262. 22. Rowe VK, WujkowskiT, Wolf MA, Sadek SE, Stewart RD. Toxicity of a solvent mixture of 1,1,1 trichloroethane and trichloroethylene as determined by experiments on laboratory animals and humans. Am Ind Hyg Assoc J. 1963;24:541-554. 23. Stewart RD. Gay HH. Schaffer AW, Erley DS, RoweVK. Experimen ts, human exposure to methyl chloroform. Arch Environ Health. 1969; 19:467 24. Kramer CG, Ott MG, Fulkerson JE, Hicks N. Imbus HR. Health of workers exposed to 1,1,1-trichloroethane; amatched pair study. Arch Environ Health. 1978;38:331-342. 25. Klaassen CD. ElaaGL. Relative effects of various chlorinated hydrocarbo n-on liver and kidney function in dogs. Toxicol Appl Pharmacol. 1967; 10119-131. 26. Plaa GL, Evans EA, Hine CH. Relative hepatotoxicity of seven haloge- Liver Disease--Hodgsonet al 1797 nated hydrocarbons. J Pharmacol Exp Tker. 196$; 123:224-22$. 27. Prendergaat JA, Jones RA. Jenkins LJ, Siegel J. Effects on experimen tal animals of long-term inhalation of trichloroethylene, carbon tetrachloride, 1,1,1-trichloroethane, dichlorodifluoromethane, and 1,1-dichloroethylene. Tbxicol Appl Pharmacol. 1967:10:270-239. 28. Chloroethanes: Review of Tbxicity. National Institute for Occupational Safety and Health; Current Intelligence Bulletin 27; 1978. US Dept of Health, Education, and Welfare (NIOSH) publication78-181. 29. McNutt NS. AmsterRL, McConnell EE, Moms F. Hepaticlesionsafter continuous exposure to 1,1,1-trichloroethane. Lab Invest. 1975:32:642-664, 30. Thiele DL, Eigenbrodt EH, Ware AJ, Cirrhosis afterrepeated trichlor oethylene and 1,1,1-trichloroethane exposure. Gastroenterology. 1982;83:926- 31. Zimmerman H. Hepatotoxicity. New York, NY Appleton <& Lange; 1978. 32. Wanberg K, Horn J, Foster F. A differential assessment model for alcoholism. JStud Alcohol. 1977;33:512-643, 33. Hodgson MJ, VanThielDH, Lauschus K,KarpfM. Liver Injury parame ters in hazardous waste workers. JOccupMed. 1989;31:238-242, 34. Diehl AM, Goodman Z, Ishak K Alcohol-like liver disease in non alcoholics. Gastroenterology. 1988;95:1066-1062. 35. Hobbach RT. Nonalcoholic fatty liver. Cleve Clin J Med. 1988;55:136- 144. 36. Aviado DM. Zakhari S, Simaan JA, UUamer AG. Methyl Chloroform and Trichloroethylene in the Environment. Boca Raton, Fla: CRC Press Inc; 1976. 37. Stewart RD, Andrews JT. Acute intoxication with methyl chloroform JAMA. 1966;195:904-906. 38. Nathan AW, Tbseland PA. Goodpasture's syndrome and trichlorethane exposure. BrJ ClinPharmacol. 1979;8:284-286. 39. HaleyyJ, PitlikS, Rosen/eld J. 1,1,1-Trichlorethane intoxication: a case report with transient liver and renal damage. ClinJbancol. 1980; 16:467472, 40. Recknagel RO, Glende EA. Carbon tetrachloride hepatotoxicity: an example oflethal cleavage. CRCCtitfJei'Tbxt'cof. 1973;2:263-273, 41. Younes M, Siegers C-P. Interrelation between lipid peroxidation and other hepatotoxic events. Biochem Pharmacol. 1984;33:2001-2003. 42. Roubai WT, Tappel AL. Damage to proteins, enzymes, and amino acids by peroxidizing lipids, Arch Biochem Biophys. 1966;113:5-8, 43. Wills ED. Lipid peroxide formation in microsomes: general consider ations. Biochem J. 1969;113:315-324, 44. Kombrust DJ, Mavis RD. Relative susceptibility of irdcxosares fern lung, heart, liver, kidney, brain, and testes to lipid peroxidation: correlation withvitaminE content. Lipids. 1980;15:315-322, 45. DiLuzioNR, Hartman AD. Role of lipid peroxidation in the pathogene sis of the ethanol-induced fatty liver. FedProc. 1967;26:1436-1442. 46. Mitchell JR, Jollow DJ. Metabolic activation of drugs to toxicj stances. Gastroenterology. 1975:68:392410. 47. Reitz RC. A possible mechanism for the peroxidation of lipids c chronic ethanol ingestion. Biochim Biophys Acta. 1975;380:145-154. 48. Shaw S, Rubin KP, Lieber CS. Depressed hepatic glutathionej increased diene conjugates in alcoholic liver disease: evidence for lipid i dation. Dig Die Sci. 1983;28:585-589. 49. Anundi I, Hoegberg J, Stead AH. Glutathione depletion m isobl hepatocytes: its relation to lipid peroxidation and cell damage. Acta Phar, Tbxicol. 1979;45:45-5t, 50. Gurney R Tetrachloroethane intoxication: early recognition of d and prevention. Gastroenterology. 1943;1:1112-1126, 51. Slater TF, Hepatotoxic effects of alcohol. In: Lagundo TR, ed.' Radical Mechanisms inTissueInjury. London, England: Pion; 1972:171-fl 52. Di Luzio NR. Antioxidants, lipid peroxidation, and chemical-ind liver injury. Fed Proc. 1973;32:1875-1881. ' 53. Lachnit V, Pietschman H. Activity of serum glutamic-oxaloacetic-t aminase and aldolase in workers exposed to halogenated hydrocarbons'] Med Surg. 1960;29:523-525, 54. Cornish HH. Adefin J, Ethanol potentiation of halogenated alip solvent toxicity. Am IndHyg Assoc J. 1966;27:57-61. 55. Shaw S, Jayatilleke E, Ross WA Gordon ER. Lieber CS. Etl induced lipid peroxidation: potentiation by long-term alcohol feeding and it uationby methionine. J Lab CUn Med. 1981;98:417424, 56. Tagesson C. Liver damage and laboratory diagnostics. Scad J 1 Environ Health. 1985;ll(9uppl 1):X01-103. 57. Charbonneau M. Tuchweber B, Plaa GL. Acetone potentiation of c liverinjury induced by repetitive administration of carbon tetrachloride. Hi tology. 1986;6:694-700. 58. Friedman LS. DienstagJL, WatkinsE, etal. Evaluation of blood c with elevated serum alanine aminotransferase levels. Ann Intern 1987;107:137-144, _ 59. Guzelian PS, Vranian G, BoylanJJ, CohnWJ, Blanke RV. Liver ture&nd function in patients poisonedwith chlordecone(Kepone). Gastn ology. 1980;78:206-213. 60. Redlich C, Sparer J, Cowan D, et al. An outbreak of occupatiqj hepatitis -Connecticut. MMWR . 1987;36:101-102. ' 61. Traiger GJ, Plaa GL. Differences in the potentiation of carbon tetra ride in rata by ethanol and isopropanol pretreatment. 7bxicol Appl Phar ' 1971;20:105-112. 62. Brennan TA. Untangling causation issues in law and medicine: 1 ous substance litigation. Ann Intern Med. 1987;107:741-747, ta Sr