We agree that the information Dr. Stark describes should be fully explained to patients and health care providers. However, we believe that current antiviral therapies, because of their relatively low efficacy and effectiveness, are unlikely to cause major changes in the epidemic of HCV-related hepatocellular carcinoma in the near future.

Persons likely to have a demonstrable therapy-related reduction in risk for hepatocellular carcinoma are those with advanced fibrosis and cirrhosis. The validity and generalizability of the randomized, controlled trial cited by Dr. Stark (1) are questionable. That study was conducted in Japan, patients were treated with a high dosage of interferon- (6 MU 3 times weekly), and the reported risk reduction can mostly be attributed to the unusually low number of hepatocellular carcinoma cases in the control group. The retrospective studies cited by Dr. Stark are prone to inflated estimates of benefit because of selection bias favoring better outcomes among treated patients (2). It is hoped that ongoing multicenter randomized trials will provide more definitive answers.

For patients without advanced fibrosis or cirrhosis, the benefit of antiviral therapy in reducing risk for hepatocellular carcinoma is far less clear. Approximately 20% and 1% to 2% of all HCV-infected persons develop cirrhosis or hepatocellular carcinoma, respectively, and there are reliable predictors of those who will develop these complications (3). On the other hand, 20% to 50% of those treated achieve sustained viral response. Therefore, it is not at all clear whether those destined to develop cirrhosis and hepatocellular carcinoma will reap the benefits of reducing the risk for these disorders.

Last, and perhaps most important, there is a discrepancy between the efficacy of antiviral therapy in randomized, controlled trials, mediocre as it is, and the disappointing effectiveness obtained in real life, even in dedicated management settings. Comorbid psychiatric disorders, including alcohol and drug use (4); side effects of antiviral therapy (particularly in patients with cirrhosis); and preponderance of "unfavorable" genotype I lead to low participation, low adherence, and eventually low response rates (2, 5).

In summary, HCV infection is probably responsible for at least a proportion of the observed increase in hepatocellular carcinoma in the United States. While current antiviral therapy may benefit some patients, especially those with advanced disease, uncertainties about predictors of the clinical course of HCV infection coupled with relatively low efficacy and especially effectiveness of therapy prohibit a sweeping recommendation.