In their excellent review article on nonalcoholic steatohepatitis (NASH) [1], Sheth and colleagues discussed many clinical settings and drugs associated with NASH but did not mention a frequently overlooked cause of this condition. Carbon tetrachloride, a chlorinated solvent and extensively studied hepatotoxin, is well known to cause steatosis and hepatic necrosis. Although carbon tetrachloride is no longer in wide use in the United States, several solvents still common used, such as dimethylformamide and perchloroethylene, have been associated with similar hepatotoxicity [2-4]. The clinical, laboratory, and histologic features of NASH described by the authors are also typical of long-term solvent exposure. Workers with chronic solvent hepatotoxicity usually present with asymptomatic elevation of plasma liver enzyme levels. As in NASH, alanine aminotransferase levels tend to be higher than aspartate aminotransferase levels, a finding that helps distinguish solvent hepatotoxicity from alcohol hepatotoxicity. The macrovesicular fatty changes, hepatocyte degeneration, and variable inflammation the authors describe are also typical of solvent hepatotoxicity. Whether solvent-related steatohepatitis can progress to fibrosis and cirrhosis, as NASH does in other clinical settings, is unclear.

We would like to remind readers of the importance of combined hepatotoxic effects. Human data are limited, but animal studies have shown that various solvents, ethanol, and drugs can potentiate each other's hepatotoxic effects [2, 5]. Similar interactions may occur in NASH. Identification of occupational and environmental hepatotoxic exposures is important for management of the individual patient and development of preventive strategies.

Because of the widespread use of solvents and other nondrug hepatotoxins at work and home, we remind clinicians of the importance of a careful history of occupational and environmental exposure in the evaluation of patients with suspected liver disease or injury, especially steatohepatitis. Clinicians would benefit from a reminder of this point, as the limited discussion of occupational exposures in Sheth and colleagues' otherwise thorough review shows.