We thank Drs. Redlich and Cullen for mentioning the importance of obtaining a history of occupational and environmental exposure during evaluation of patients suspected of having liver disease. Some, but not all, of the features of NASH can be present in patients chronically exposed to organic solvents. At our institution, pathologists report liver biopsy specimens as being consistent with "toxic and/or metabolic injury" and do not necessarily use the term "NASH." This is a further reminder to clinicians that a thorough occupational history should be obtained in all these patients.

We read with interest Dr. Akpunonu's description of four obese patients with NASH who developed diabetes 5 to 18 months after NASH was diagnosed. Although this finding is interesting, data are insufficient to make any dogmatic statement that the finding of histologic features of NASH might predict an imminent diabetic state.

On the basis of the study by Van Ness and Diehl [1], Dr. Moseley suggests that the positive predictive value of the prebiopsy diagnosis of NASH is 81%. However, the data do not support this contention. In this study, patients with NASH (or steatohepatitis) were included in the subgroup of those with "fatty liver," and the positive predictive value of a prebiopsy diagnosis in these patients was only 56% [1]. The 81% mentioned by Dr. Moseley is the positive predictive value of a prebiopsy diagnosis in the subgroup of patients with "chronic necro-inflammatory activity." This subgroup included patients with chronic hepatitis B virus infection and non-A, non-B hepatitis.

Drs. Moseley, Walker, and Kirsch think that liver biopsy and extensive testing are not necessary in patients with NASH, especially because liver biopsy is a risky procedure and "no proven therapy for NASH exists." We concur that liver biopsy is an invasive procedure, but it is still the gold standard for diagnosing NASH. The decision to do a complete and thorough serologic workup in asymptomatic patients with elevated plasma alanine aminotransferase and aspartate aminotransferase values, including testing for rare disorders such as Wilson disease, must be individualized for each patient (clearly, we do not advocate testing for Wilson disease in elderly persons). We believe that liver biopsy may be useful in many patients for diagnostic and prognostic reasons because up to one sixth of patients with NASH later develop fibrosis and cirrhosis.