Liver Disease Associated with Occupational Exposure ...

1 downloads 0 Views 2MB Size Report
Mark R. Cullen, M.D. is a Henry J. Kaiser Family. Foundation faculty ... KESTELL P, GILL MH, THREADGILL MD, GESCHER A, HOWARTH. OW, CURZON EH.
Liver Disease Associated with Occupational Exposure to the Solvent Dimethylformamide CARRIE A. REDLICH, M.D.; WILLIAM S. BECKETT, M.D., M.P.H.; JUDY SPARER, M.S.C.E.; KENNETH W. BARWICK, M.D.; CAROLINE A. RIELY, M.D.; HEIDI MILLER, P.A.; STEPHEN L SIGAL, M.D.; STUART L SHALAT, Sc.D.; and MARK R. CULLEN, M.D.; New Haven, Connecticut

Study Objective: To characterize an outbreak of liver disease among workers in a fabric coating factory; and to determine the outbreak's cause and natural history and strategies for clinical recognition, treatment, and prevention. Design: Clinical-epidemiological investigation. Setting: Academic medical center, Occupational Medicine Clinic, and worksite. Patients: Fifty-eight of sixty-six workers participated in the study. All had standard liver function tests at least once. Forty-six workers completed a questionnaire; 27 had more extensive clinical evaluation for recognized liver abnormalities. Results: A plant-wide outbreak of liver disease was recognized after a new employee presented with signs and symptoms of hepatitis. Evaluation of the worksite showed that dimethylformamide, a widely used industrial solvent and known hepatotoxin, was being used to coat fabric in poorly ventilated areas without appropriate skin protection. No other major hepatotoxic exposure was identified. Overall, 3 6 of 5 8 ( 6 2 % ) workers tested had elevations of either aspartate aminotransferase (AST) or alanine aminotransferase (ALT) levels. Enzyme abnormalities occurred almost exclusively in production workers ( 3 5 of 4 6 were abnormal), whereas only 1 of 12 nonproduction workers showed any elevations in enzyme levels ( P < 0.0001). Serologic tests excluded known infectious causes of hepatitis in all but 2 workers and changes characteristic of toxic liver injury were confirmed by histologic examinations of biopsy specimens from 4 workers. The ratio of AST to ALT levels was one or less in all but 1 worker. After modification of work practices and removal of workers most severely affected from exposure, improvement in liver enzyme abnormalities and symptoms in most patients were seen, although some patients showed persistent elevations of enzyme levels. Conclusions: An outbreak of toxic liver disease has been associated with exposure to dimethylformamide in the workplace. The diagnosis of toxic liver disease was established by the clinical histories, negative viral serologies, an enzyme pattern of ALT levels being greater than AST levels, epidemiologic data on coworkers, and liver biopsy specimens. The high prevalence of unsuspected liver enzyme abnormalities in these workers suggests that occupational liver disease may occur more frequently than is generally recognized.

LIVER DISEASE due to occupational exposure is believed

to be an uncommon event in modern industry ( 1 ) . Liver disease from exposure to well-recognized industrial hepatotoxins such as carbon tetrachloride, chlorinated pesticides including chlordecone (kepone), and yellow • From the Occupational Medicine Program and Liver Study Unit, Department of Internal Medicine, and Department of Pathology, Yale University School of Medicine; N e w Haven, Connecticut.

680

Annals of Internal Medicine. 1988;108:680-686.

phosphorus is also rarely seen. The incidence and natural history of occupationally induced toxic liver injury are not well defined, nor are strategies for differentiating it from other causes of hepatitis. These issues were studied during a recently recognized outbreak of liver disease in New Haven, Connecticut, associated with occupational exposure to the widely used industrial solvent dimethylformamide. Because of its excellent solvent properties, dimethylformamide is a versatile and widely used industrial chemical. It is used primarily in the manufacture of polyurethane products and acrylic fibers, as well as the production of pharmaceuticals, pesticides, and in other various applications. Dimethylformamide use is presumed to be widespread based on the estimated 33 million pounds produced in the United States in 1979, but there is no reliable estimate of the number of workers currently exposed ( 2 ) . Animal studies have shown that dimethylformamide is hepatotoxic in several species including cats, mice, rats, guinea pigs, and rabbits (3-10). Dimethylformamide is readily absorbed through the skin, respiratory system, and gastrointestinal tract. The primary organ of toxicity is the liver, which on histologic examination shows evidence of hepatic necrosis. However, the effects of dimethylformamide exposure in humans are not as well documented. Complaints of abdominal pain, nausea, vomiting, dizziness, headaches, loss of appetite, and alcohol intolerance have been reported in workers exposed to dimethylformamide (11-17). Although liver function has rarely been evaluated in such workers, abnormal liver function tests have been seen in a few instances (13, 14), primarily after large accidental exposures. Several studies (3, 18) have found no evidence of hepatotoxicity after more routine exposures to dimethylformamide. After the referral of a worker with suspected toxic hepatitis, we did a clinical-epidemiologic investigation of his coworkers and an environmental assessment of the workplace. Case Report

A previously healthy 40-year-old Hispanic man presented to the emergency room at Yale-New Haven Hospital with abdominal pain, nausea, and headache in October 1986. He had become ill after working for less than 2 weeks as an operator of a fabric-coating machine. He reported that other workers had © 1 9 8 8 American College of Physicians

Figure 1A. Liver biopsy specimens of an index case (Patient 1). A centrilobular and midzonal region shows lobular disarray and regenerative changes of irregular liver cell plates, binucleated hepatocytes, and variation in nuclear size and configuration. Enlarged Kupffer cells are present {arrows), but inconspicuous. (Hematoxylin and eosin; original magnification x 100.) Figure I B . Phagocytosed material in Kupffer cells stains positively, enabling them to be seen easily. (Periodic acid-Schiff; Original magnification x 100.)

similar symptoms. Standard liver function tests showed an elevated aspartate aminotransferase (AST) level of 15.8 u.kat/L (normal, 0.25 to 0.58 u.kat/L). Alkaline phosphatase and bilirubin levels were normal. Hepatitis A IgM antibody and hepatitis B surface antigen and antibody were negative, as was an abdominal ultrasound. The patient had no history of significant alcohol use, blood transfusions, or intravenous drug use. Based on a presumptive diagnosis of toxic hepatitis, he was removed from his workplace. His symptoms resolved and his transaminase levels improved. A liver biopsy was done 3.5 months later because of persistently elevated liver enzymes (alanine aminotransferase [ALT] levels, three times control). Microscopic examination showed changes consistent with a resolving toxic injury to the liver, including evidence of diffuse regeneration with irregular liver cell plates, binucleated hepatocytes, and variation in nuclear size and staining characteristics (Figure 1A). In addition there was spotty unicellular necrosis and, on staining with periodic acid-Schiff-diastase, enlarged Kupffer cells were seen in the sinusoids (Figure IB). A fat stain on a cryostat section of the biopsy specimen showed diffuse steatosis in a microvesicular pattern, which was not evident on conventionally prepared paraffin sections. There was no evidence of chronic disease. Six months after removal from work, transaminase levels had returned to normal. Methods WORKPLACE EVALUATION

The company produces polyurethane-coated fabrics. Large amounts of dimethylformamide and lesser quantities of other solvents including toluene, methylethylketone, 1,1,1 trichloroethane, and dichlorobenzene are mixed with polyurethane components, including diisocyanates, in a mixer. The polyurethanesolvent mixture is then poured into smaller containers and hand ladled onto moving fabric. Dimethylformamide and other solvents evaporate from the fabric as it dries. The finished product, rolls of polyurethane-coated fabric, is used to make bullet-proof vests, parachutes, and other items. Industrial hygiene evaluation of the workplace was limited to general observation because the company curtailed production shortly after the results of the screening liver function tests were obtained and before measurements of air concentrations of industrial solvents could be taken. The workforce consists of nonproduction workers such as office staff, management, inspectors, and supervisors; and productions workers such as machine operators, helpers, and mixers. Nonproduction employees work primarily in the office and inspection areas. Most of the production employees work on

one of six coating machines. A few persons work in a separate shed mixing solvents and polyurethane components. Historically, job turnover has been rapid with frequent switching of job categories. Precise information on the total number of employees newly hired or terminated was not available. SURVEY OF THE W O R K FORCE The company supplied a list of all sixty-six current employees, but information such as job category or length of employment was not made available. The employer had not obtained any medical information about his workers. As a result of the clinical findings in Patient 1, arrangements were made to do screening liver function tests on all employees. The initial screening liver function tests on 45 employees included AST, ALT, gamma-glutamyl transpeptidase, direct and total bilirubin, alkaline phosphatase, lactic dehydrogenase, total protein, and albumin levels. An additional 11 employees had liver function test drawn over the next 3 months, either because they missed the initial screening or joined the workforce after that time. Individual liver function test results and a questionnaire to determine demographic background, job history, and symptoms were sent to all workers. Those workers with A L T or AST levels greater than twice normal were referred to the Occupational Medicine Clinic; any other worker who wished to be evaluated was also seen. Thus, 27 workers were evaluated in the Occupational Medicine Clinic. Evaluation consisted of a history, physical examination, laboratory work including liver function tests, hepatitis A and B serologies, complete blood count, prothrombin time, and partial thromboplastin time. Additional blood tests including determination of blood urea nitrogen, creatinine, glucose, and electrolyte levels were done on many. Four had liver biopsies done; five had abdominal ultrasound examinations. For analysis of liver function test data, the highest enzyme elevations were used for all patients with two or more complete sets, unless otherwise specified. Results WORKPLACE EVALUATION

Large quantities o f d i m e t h y l f o r m a m i d e ( a p p r o x i m a t e ly 15 to 2 0 fifty-five gallon d r u m s per w e e k ) and lesser a m o u n t s o f other solvents including toluene ( 1 0 d r u m s per w e e k ) , m e t h y l e t h y l k e t o n e ( 5 to 10 d r u m s per w e e k ) , 1,1,1 trichloroethane ( 5 d r u m s per w e e k ) , and dichlorobenzene ( 0 . 5 d r u m s per w e e k ) were being used in poorly ventilated areas w i t h o u t appropriate skin protection. T h e Redlich etal. • Exposure to Dimethylformamide

6 8 1

Figure 2 . The workplace where employees were exposed to dimethyliformamide. One of the fabric-coating machines after local ventilation had been installed, and a nearby barrel containing a mixture of dimethylformamide, polyurethane and other solvents are shown {left). The previous inadequate ventilation system is also shown {right).

work areas where the coating machines and mixer were located had little ventilation (Figure 2). The workers operating the machines did not routinely wear protective clothing and had considerable gross skin contamination by the solvents. The drums and smaller cans containing the solvents and polyurethane mixture frequently were unlabeled and left uncovered. S U R V E Y OF OTHER W O R K E R S

Including the index case, data were available on 58 of 66 employees (Figure 3). Forty-six of these 58 either returned the questionnaire or were interviewed in the clinic. Average length of employment was 40 months, but 15 had worked for 3 months or less. Eighty-one percent of the employees were Hispanic, 14% were white. The mean age of the 58 workers tested was 33 years (range, 18 to 60). All but 2 workers were male. A worker was classified as a production worker if he had worked at one of the production jobs within the past 6 months. The job classification of the 12 workers who did not fill out questionnaires was based on the questioning of other workers and management. By these criteria, 46 employees were considered production workers, and 12 were nonproduction workers (Figure 3). All production workers were men who worked 12-hour shifts 5 days per week. Many also worked additional 8-hour shifts on weekends. All but 3 production workers were Hispanic; many did not speak English. One third of the nonproduction workers were Hispanic; two were female. Most of the 8 workers who did not participate appeared to be nonproduction personnel. Overall, 36 (62%) workers had elevated AST or ALT levels. Nineteen (33%) had elevations greater than twice normal (Figure 3). Nine of these nineteen had enzyme elevations greater than five times normal. At least 4 had sought medical attention for symptoms similar to Patient 1. Liver enzyme levels correlated with job classification (Figure 3). Of the 12 nonproduction workers, 11 had normal transaminase levels; only 1 had mild elevations. Of the 46 production workers, 35 had some elevation in liver enzyme levels and 19 had elevations greater than twice normal of one or more liver enzyme. Only 11 had 6 8 2

May 1988

• Annals of Internal Medicine • Volume 108



Numbers

normal liver enzyme levels. The mean ALT level for nonproduction workers was 0.50 i t 0.12 u,kat/L; the mean ALT level for production workers was 1.88 + 2.86 u,kat/L. The association between job classification and liver function test abnormalities was highly significant using the Fisher exact test on proportions (P < 0.0001) and the Student Mest on mean ALT levels (P < 0.01). The ratio of AST to ALT levels was one or less in all but one worker (Figure 4 ) . In most workers it was less than 0.6. This ratio is opposite of the ratio usually found in alcoholic liver disease, in which a value of two or more is typical (19). Interestingly, the relation between duration of exposure and transaminase levels was inverse, with short-term workers (less than 2 months) showing the greatest abnormalities (Figure 5). Given the highly transient nature of the workforce and the cross-sectional study design, this peculiarity could not be studied further. On the basis of the questionnaire or clinic interview obtained from 46 workers, the following symptoms were reported: gastrointestinal (anorexia, abdominal pain, or nausea) in 31; central nervous system solvent intoxication (headaches, dizziness) in 18; and alcohol intolerance characterized by a disulfiram type reaction (facial flush-

Figure 3. Flow chart showing data obtained f r o m workers exposed to dimethylformamide. Liver function tests were obtained f r o m 5 8 of 6 6 employees. Forty-six of these 5 8 were production workers; 1 2 were nonproduction workers. Of the 4 6 production workers, 1 1 had normal liver enzymes; 3 5 had one or more elevated transaminase level. Of the 1 2 nonproduction workers, 1 1 had normal liver enzymes and only one had mild elevations in transaminase levels. LFTS = liver function tests.

ing and palpitations after alcohol intake) in 11. Based on these findings, cessation of production was recommended pending the installation of appropriate protective engineering controls. Approximately 2 months after the initial liver screening tests were drawn, workers went on strike. MEDICAL EVALUATION AND FOLLOW-UP

After the initial screening, 27 workers, including all 19 with liver enzyme levels greater than twice normal, sought medical follow-up in our clinic. Twenty-six were production workers; one was a nonproduction worker (Figure 3). The symptoms noted above were frequently reported after the onset of employment at the plant. Viral hepatitis serologies showed that all 27 had negative hepatitis B surface antigen and negative hepatitis A IgM antibody. Hepatitis B surface antibody and core antibody were positive in 2. Only 2 workers gave a history of significant alcohol or intravenous drug use. None had received blood transfusions, and any use of medications was rare. Five patients with the most significant abnormalities are described in the appendix. All patients with greater than twofold elevations in their transaminase levels were transferred to a nonproduction area or removed entirely from work, depending on their symptoms and the availability of a nonproduction job. The symptoms of all these workers improved with removal from direct exposure to dimethylformamide. In most cases transaminase levels returned to normal or near normal in 1 to 5 months after the worker's removal from the workplace. However, transaminase levels of eight workers, all of whom have been on strike and removed from the production area, have remained mildly elevated or have risen slightly. Although the cause of these persistent transaminase abnormalities is unclear, it is possible that some of these workers have had continued exposure to dimethylformamide through contaminated shoes or clothing. Clinical histories have not indicated significantly increased use of alcohol among these work-

Figure 4 . Frequency distribution of the ratio of aspartate aminotransferase (AST) to alanine aminotransferase (ALT) levels in 3 6 production workers exposed to dimethylformamide with one or more elevated transaminase levels. If the patient had more than one set of liver enzyme tests, the set with the highest AST and ALT value was used. The ratio of AST to ALT levels was one or less in all but 1 patient, the opposite of the ratio normally seen in alcoholic liver disease.

Figure 5. Relationship between months of employment and alanine aminotransferase (ALT) levels in production workers exposed to dimethylformamide who completed a questionnaire (n = 3 8 ) . Months of employment were determined on the basis of questionnaire data and calculated f r o m month of initial hire to the date liver function tests were drawn. The ALT level used was f r o m the initial screening or f r o m the first clinic visit if the patient had continued exposure after the screening test.

ers or exposure to any other hepatotoxic agents during this period. Discussion

We have characterized an outbreak of liver disease among workers occupationally exposed to the solvent dimethylformamide. This study addresses several important issues about the prevalence, diagnosis, and prognosis of toxic liver disease in the occupational setting. Several factors may limit our ability to interpret the observed association as causal. Despite our inability to obtain liver function tests and questionnaires on all 66 employees, we did obtain liver function tests on 58 of 66 employees (88%), and questionnaires on 46 of these 58 (80%). Given the circumstances, including a largely transient non-English-speaking workforce and limitations imposed by the employer, we feel that we have sufficient information to characterize this population. The prevalence of abnormalities in transaminase levels in the general population is unknown. It is possible, but unlikely, that the elevated transaminase levels in these workers could be due to another cause such as alcohol, drugs, or viral hepatitis. Excessive alcohol use is probably the most common cause of elevated liver enzymes in a population such as this. However, we do not feel that alcohol can explain the current findings. Detailed histories of 27 workers showed little if any excessive alcohol consumption in most. This information was supported by the laboratory data. The ratio of AST to ALT was almost always less than one, a reversal of the ratio usually found Redlich etal. • Exposure to Dimethylformamide

6 8 3

in alcoholic hepatitis (18). In addition, none of the four liver biopsy specimens showed changes typical of alcoholic liver disease. However, alcohol use may potentiate the hepatotoxicity of dimethylformamide. Such an interaction has been shown in animals when carbon tetrachloride and ethanol or other aliphatic alcohols are combined (20,21). In most patients no other cause for the transaminase elevations could be ascertained after evaluation in the clinic, including a detailed history and physical examination, hepatitis A and B serologies, and liver biopsies in three workers. Although outbreaks of non-A, non-B hepatitis have been described (22) and would escape serologic detection by available techniques, this interpretation seems highly inconsistent with our epidemiologic data, clinical findings, and liver biopsy results. The possibility that these workers were exposed to a liver toxin other than dimethylformamide must also be considered. In addition to dimethylformamide these workers used several other solvents including methylethylketone, toluene, and 1,1,1 trichloroethane, and dichlorobenzene. These solvents have individually been shown to have little if any hepatotoxicity in animal and human studies; additionally, they were used in much smaller amounts than dimethylformamide. As with alcohol, it is possible that some of these solvents may have potentiated the hepatotoxicity of dimethylformamide. N o toxicologic data are available on the interaction of dimethylformamide with alcohol or other solvents. In establishing a causal link, it would have been desirable to obtain either air samples or some biologic marker quantifying the amount of dimethylformamide exposure. However, we were unable to take air samples before improvements in the ventilation system were made, and no good biologic marker of dimethylformamide exposure exists. It is probable that skin absorption of dimethylformamide is a more important route of exposure than inhalation, as most likely was the case with these workers. Unfortunately, no accurate means of quantifying skin exposure exists. Although a urine test that measures monomethylformamide, one of the urinary metabolites of dimethylformamide, does exist (23-28), there is no well established correlation between exposure to dimethylformamide and the amount of monomethylformamide in a spot urine sample. Also, the test is not widely available, and it was not feasible to obtain timed urine collections from these workers before their removal from the workplace. Because dimethylformamide hepatotoxicity is believed to be dose-dependent, a direct relation between length of exposure and elevated liver enzyme levels might be expected. Actually, the reverse was seen. One possible explanation for this paradoxical finding is that workers who became ill left the plant. The plant had a high turnover of employees, many of whom, like several of our patients, may have left after a few weeks because of symptoms described above. As noted, data on hiring and termination were insufficient to adequately test this hypothesized "healthy worker effect." Another possibility is that exposure to dimethylformamide induces the enzymes that me6 8 4

May 1988

• Annals of Internal Medicine • Volume 108

• Number 5

tabolize it, so that workers chronically exposed develop tolerance to its toxic effects. Several similarities and differences between the liver toxicity reported here and that found with carbon tetrachloride, the best-studied occupational hepatotoxin, should be noted. Hepatic necrosis and steatosis, which are typically seen with carbon tetrachloride poisoning, were also seen in three of four liver biopsy specimens, although findings were mild in two. These two biopsy specimens showed evidence of resolution and regeneration, however, and it is likely that the degree of hepatocellular injury would have been more pronounced had biopsies been done earlier in the patients' illnesses. The biopsy done on the worker with much longer exposure to dimethylformamide (Patient 6) showed evidence of moderately severe steatosis with spotty necrosis despite only mildly elevated transaminase levels. Resolution of the abnormalities of liver enzyme levels has taken longer (1 to 7 or more months) in these patients compared to that reported with carbon tetrachloride (less than 2 weeks) ( 1 ) . There are several possible explanations for this finding. Most human data on the natural history of carbon tetrachloride-induced liver injury is based on single or very short-term acute exposures, rather than weeks to years of exposure as reported here. Some of the workers may have had ongoing exposure to dimethylformamide after removal from work through contaminated clothing and shoes that they continued to wear. Alternately, dimethylformamide may have a different mechanism of hepatotoxicity and natural history than that of carbon tetrachloride. Finally, several aspects of this investigation should be emphasized for practitioners. An outbreak of liver disease among 62% of workers in a plant has been identified, although most workers had never sought medical attention. On the basis of an evaluation of the workplace and workers, an industrial solvent, dimethylformamide, was found to be the most likely cause of the problem. Both the clinical epidemiologic data presented here and previous evidence of dimethylformamide liver toxicity from animal and human studies were crucial factors in linking this outbreak to dimethylformamide. In addition, two clinical findings described in this study could help physicians recognize potential occupational liver disease and overexposure to widely used and potentially hazardous industrial solvents. First, the consistent finding of the ratio of AST to ALT levels less than one proved helpful in differentiating toxic from alcoholic hepatitis, and should alert the clinician to consider such a possibility. Accordingly, ALT measurements should be included in liver function tests where toxic injury is being considered. Second, the presence of acquired alcohol intolerance (degreaser's flush or disulfiram symptoms) should make clinicians consider solvent overexposure. Such symptoms are unusual, and because they are rarely volunteered by the patient, they should be inquired about specifically. Several aspects of this study suggest that toxic liver disease due to dimethylformamide or other agents may

be more common than previously recognized. Elevated liver enzyme levels were found in 76% of production workers exposed routinely to dimethylformamide, without accidental overexposure or unusual circumstances. Such a high incidence of elevated liver enzyme levels was not anticipated. Although some of these patients were symptomatic, most had minimal symptoms and had never sought medical evaluation. In addition, during the 15 years since the factory opened, many workers may have developed toxic hepatitis, but were never correctly diagnosed. Even though this factory is located within a few miles of a community health clinic and a major academic medical center, the problem remained unidentified until now. Review of the Occupational Medicine Clinic patient records indicated that a worker from this same plant had been evaluated 3 years ago. Although excess solvent exposure was suspected, and the patient had gastrointestinal symptoms and a mildly elevated AST level (ALT levels were not tested), the diagnosis of toxic hepatitis had been missed. This missed diagnosis raises the concern that liver injury due to dimethylformamide and possibly other industrial solvents or chemicals may frequently go undetected. The long-term effects of dimethylformamide exposure are not known. Although no evidence of chronic liver disease was found in liver biopsy specimens from four patients, three of these patients had relatively short exposures (several weeks to months). This cohort of workers will be followed to help ascertain the natural history and long-term effects of dimethylformamide hepatotoxicity. The mechanism whereby dimethylformamide exerts its toxic effects, the possible interactions with other solvents and alcohol, and the variability in individual susceptibility require further investigation.

Figure 6. Liver biopsy specimen of Patient 6 shows moderate steatosis in periportal hepatocytes in both microvesicular (arrow) and macrovesicular f o r m . Other hepatocytes show nuclear pleomorphism and binucleation, features of regenerative activity. (Hematoxylin and eosin; original magnification x 100.)

On physical examination the patient was jaundiced. There were no stigmata of chronic alcohol use. Abdominal examination showed a mildly enlarged nontender liver. Repeat laboratory values several days after the initial tests included AST level, 11.97 u,kat/L; and ALT level, 13.0 u,kat/L; an abdominal echo showed a diffusely echogenic liver; the size was within normal limits. Hepatitis A IgM antibody and hepatitis B surface antigen and antibody were negative. The patient was removed from his workplace and advised to abstain from alcohol; his symptoms resolved. Two weeks later his ALT level was 2.49 fikat/L, AST level was 1.51 /xkat/L; and total bilirubin level was 54.74 u,mol/L. By 2-month followup his transaminase and bilirubin levels had returned to normal. PATIENT 4

Appendix PATIENT 2

A 20-year-old previously healthy Hispanic man had worked at the same plant as Patient 1 for 2 weeks, applying dimethylformamide onto the fabric with a ladle. He complained of symptoms similar to symptoms of Patient 1, which developed within the first week of his new job. His AST level was 6.92 fikat/L and his ALT level was 15.57 u,kat/L (normal, less than 0.53 jutkat/L). Alkaline phosphatase and bilirubin levels were normal. Hepatitis A and B serologies were negative. Amylase, lipase, electrolytes, blood urea nitrogen, and creatinine levels were normal. He was removed from work and his symptoms improved. Ten days later his AST level was 1.03 jukat/L and ALT level was 4.37 jukat/L. The patient has been lost to further follow-up. PATIENT 3

A previously healthy 43-year-old Hispanic man was referred after we obtained the results of his screening liver function tests, which included AST level, 8.88 jukat/L; ALT level, 8.80 jukat/ L; total bilirubin level, 148 fimol/L (normal, less than 26 u,mol/L); direct bilirubin level, 92 /xmol/L (normal, less than 5.1 jumol/L); alkaline phosphatase level, 3.67 jukat/L (normal, 0.16 to 1.16 jmkat/L). He had worked at the plant for 4 months as a machine operator and helper. He had complained of abdominal pain, nausea, and jaundice for 1 to 2 weeks previously, but had not sought medical attention. He drank 6 to 12 cans of beer a night. He had no history of alcohol related medical problems, intravenous drug use, or blood transfusions.

A previously healthy 30-year-old Hispanic man was also referred based on the results of screening tests, which were AST level, 1.48 /xkat/L; ALT level, 2.62 u,kat/L; bilirubin and alkaline phosphatase levels were normal. He had worked at the company for 6 weeks before having his blood drawn. He complained of occasional abdominal pain and dizziness. He drank a few beers on weekends only and had no history of intravenous drug use, blood transfusions, or medications. The patient was transferred to a nonproduction job, packing the finished rolls of cloth, but on weekends he also worked on one of the coating machines. Two weeks later his AST level was 2.49 jukat/L and ALT level was 9.96 u,kat/L. He was removed from work. One month later his AST level was 1.00 jmkat/L and ALT level was 4.83 /ikat/L, and he had a liver biopsy done. The biopsy specimen showed evidence of a resolving mild toxic injury to the liver with mild centrilobular microvesicular steatosis and regenerative changes, identical to that of Patient 1. Five months after removal from work his AST and ALT levels had returned to normal. PATIENT 5

A previously healthy 18-year-old Hispanic man started work at the company after the screening liver function tests had been drawn. After 2 weeks at work he noted decreased appetite and abdominal pain. After 6 weeks at work laboratory values showed an AST level of 5.20 jmkat/L and ALT level of 9.66 u,kat/L; bilirubin and alkaline phosphatase levels were normal. Two weeks later he was evaluated in the clinic. The patient had previously worked at the same company for 4 months and had left because of these same symptoms, which resolved after he Redlichetal.

• Exposure to Dimethylformamide

685

left the company. The patient had no history of significant alcohol use, blood transfusions, or intravenous drug use. Repeat laboratory values included AST and ALT levels of 2.94 jukat/L and 11.89 jukat/L, respectively. Hepatitis A and B serologies were negative. Five weeks after removal from work his ALT level remained elevated at 10.91 jxkat/L, and his AST level was 3.15 jukat/L. He had a liver biopsy done that showed a chronic inflammatory infiltrate and nonspecific pattern of spotty lobular necrosis evidenced by scattered acidophilic bodies and aggregates of mononuclear inflammatory cells within sinusoids and replacing individual hepatocytes. An abdominal echo was unremarkable. Viral serologies for Epstein-Barr virus, cytomegalovirus, and toxoplasma were negative for any active infections. Seven months after removal from work his symptoms had improved and his ALT level had dropped to 2.31 jukat/L with an AST level of 0.81 /xkat/L.

toxicity of dimethylformamide in rats and mice. Drug Chem 1984;7:551-71.

Toxicol

5. LUNDBERG I, E K D A H L M, KRONEVI T, LIDUMS V, LUNDBERG S. Rel-

6. 7. 8.

9.

10.

ative hepatotoxicity of some industrial solvents after intraperitoneal injection or inhalation exposure in rats. Environ Res. 1986;40:411-20. LUNDBERG I, LUNDBERG S, KRONEVI T. Some observations on dimethylformamide hepatotoxicity. Toxicology. 1981;22:1-7. WILES JS, NARCISSE JK JR. The acute toxicity of dimethylamides in several animal species. Am Ind Hyg Assoc J. 1971;32:539-45. K E N N E D Y GL JR, SHERMAN H. Acute and subchronic toxicity of dimethylformamide and dimethylacetamide following various routes of administration. Drug Chem Toxicol. 1986;9:147-70. K E N N E D Y GL JR. Biological effects of acetamide, formamide, and their monomethyl and dimethyl derivatives. CRC Crit Rev Toxicol. 1986;17:129-82. MASSMANN W. Toxicological investigations on dimethylformamide. Br J Ind Med. 1956;13:51-4.

11. TOLOT F, ARCADIO F, LENGLET JP, ROCHE L. Intoxication par la

dimethylformamide. Arch Mai Prof. 1968;29:714-7.

PATIENT 6

A 37-year-old Hispanic man had worked for 10 years at the company primarily as a machine operator on one of the coating machines. He was referred based on the results of screening liver function test that showed a normal AST level and an ALT level of 1.41 jukat/L. He noted symptoms of alcohol intolerance and decreased appetite while at work. He had no history of blood transfusions, obesity, or significant use of alcohol or any medications. Three months later, during which time he had no further known exposure to dimethylformamide, his AST level was 0.747 /ikat/L and ALT level was 1.59 jukat/L. He had a liver biopsy done that showed moderately severe steatosis in a macrovesicular and microvesicular pattern with spotty necrosis and regeneration, consistent with a toxic injury to the liver (Figure 6). ACKNOWLEDGMENTS: The authors thank James Revkin, M.D. and Raul Perea-Henze, M.D. for their help with translation, and A. Brian West, M.B., M.C.R. Path, for his assistance with the photomicrographs. Grant support: in part by a National Institute of Environmental Health Sciences Clinical Investigator Award ES-00131. Carrie A. Redlich, M.D. was a Charles A. Dana Foundation fellow in Occupational Medicine during the conduct of this work. Mark R. Cullen, M.D. is a Henry J. Kaiser Family Foundation faculty scholar in general internal medicine. • Requests for reprints should be addressed to Carrie A. Redlich, M.D.; Department of Respiratory Diseases, University of Washington, BB-1253 Health Science Bldg., Rm. 12; Seattle, WA 98195.

12. TOLOT F, D R O I N M, GENEVOIS M. Intoxication par la dimethylfor-

mamide. Arch Mai Prof. 1958;19:602-6. 13. POTTER HP. Dimethylformamide-induced abdominal pain and liver injury. Arch Environ Health. 1973;27:340-1. 14. REINL W, U R B A N HJ. Erkrankungen Durch Dimethylformamide. Int Archiv Gewerbepathol Gewerbehyg. 1965;21:333-46. 15. L Y L E WH, SPENCE TW, M C K I N N E L E Y WM, DUCKERS K. Dimethyl-

formamide and alcohol intolerance. Br J Ind Med. 1979;36:63-6. 16. CHI VERS CP. Disulfiram effect from inhalation of dimethylformamide [Letter]. Lancet. 1978;1:331. 17. CHARY S. Dimethylformamide: a cause of acute pancreatitis? [Letter]. Lancet. 1974;2:356. 18. YONEMOTO J, SUZUKI S. Relation of exposure to dimethylformamide vapor and the metabolite, methylformamide, in urine of workers. Int Arch Occup Envir Health. 1980;46:159-65. 19. COHEN JA, KAPLAN MM. The SGOT/SGPT ratio—an indicator of alcoholic liver disease. Dig Dis Sci. 1979;24:835-8. 20. G A R D N E R GH, G R O V E RC, GUSTAFSON RK, et al. Studies on the

pathological histology of experimental carbon tetrachloride poisoning. Bull Johns Hopkins Hosp. 1925;36:107-33. 21. CORNISH HH, ADEFUIN MS. Potentiation of carbon tetrachloride toxicity by aliphatic alcohols. Arch Environ Health. 1967;14:447-9. 22. G R O V E S, A L I - S A L A D A, F A R A H MA, et al. Enterically transmitted

non-A, non-B hepatitis—East Africa. MMWR. 1987;36:241-4. 23. BARNES JR, R A N T A KE. The metabolism of dimethylformamide and dimethylacetamide. Toxicol Appl Pharmacol. 1972;23:271-6. 24. BARNES JR, HENRY N W 3RD. The determination of N-methylformamide and N-methylacetamide in urine. Am Ind Hyg Assoc J. 1974;35:84-7. 25. KESTELL P, G I L L MH, THREADGILL MD, GESCHER A, H O W A R T H

References 1. ZIMMERMAN HJ. Hepatotoxicity: The Adverse Effects of Drugs and Other Chemicals on the Liver. New York: Appleton-Century-Crofts; 1978:303-15. 2. SANTODONATO J. Monograph on human exposure to chemicals in the workplace: dimethylformamide. Prepared for Division of Cancer Etiology, National Cancer Institute. Center for Chemical Assessment, Syracuse Research Corp. Syracuse, New York; 1985. 3. CLAYTON JW, BARNES JR, H O O D DB, SCHEPERS GWH. The inhala-

tion toxicity of dimethylformamide ( D M F ) . Am Ind Hyg Assoc J. 1963;24:144-54. 4. CRAIG DK, WEIR RJ, W A G N E R W, GROTH D. Subchronic inhalation

6 8 6

May 1988

• Annals of Internal Medicine • Volume 108 • Number 5

OW, CURZON EH. Identification by proton NMR of N-(hydroxymethyl)-N-methylformamide as the major urinary metabolite of N,N-dimethylformamide in mice. Life Sci. 1986;38:719-24. 26. KRIVANEK N D , M C L A U G H L I N M, FAYERWEATHER WE. Monome-

thylformamide levels in human urine after repetitive exposure to dimethylformamide vapor. / Occup Med. 1978;20:179-82. 27. LAUWERYS RR, KIVITS A, LHOIR M, et al. Biological surveillance of

workers exposed to dimethylformamide and the influence of skin protection on its percutaneous absorption. Int Arch Occup Environ Health. 1980;45:189-203. 28. M A X F I E L D ME, BARNES JR, A Z A R A, TROCHIMOWICZ HT. Urinary

excretion of metabolite following experimental human exposures to DMF or to DMAC. J Occup Med. 1975;17:506-11.