The question Dr. Thomas and colleagues raise is: What is the course of non-A, non-B (mostly C) hepatitis after the second decade? Among those in whom earlier nonhepatic deaths might be less common, the ultimate threat of HCV may be greater. However, the populations described have other health risks (related to drug or sexual practices or to a lower socioeconomic environment), which may adversely affect their life expectancies.

Indeed, data from the study by Seeff and colleagues [1] (which included 32 of our patients in the group of 568 with post-transfusion hepatitis) suggested that, in older patients, no difference exists in overall survival between those who did and those who did not have hepatitis [1]. An increased number of hepatic deaths, however, occurred in the former group [1]. The life-Table analysis from our study indicated that, 16 years after the onset of the hepatitis, 15% to 20% of those who survived had evidence of liver failure. During that time, approximately 40% of our cohort died of nonhepatic causes, a number large enough to overshadow additional mortality related to hepatitis (an explanation for the observations by Seeff and colleagues). End-stage liver disease may even be a problem in a few older patients.

The natural history of chronic viral hepatitis in an older patient receiving a blood transfusion may not be the same as that in a younger patient infected by a sexual, intravenous, or unknown route. In addition to the age differences, the immunologic state and load of infecting virus may differ.

If, however, the natural histories are similar, one would still expect 80% to 85% of those younger persons to be free of any manifestation of liver failure 16 years later. Certainly, more persons may ultimately develop symptoms of hepatic failure, but how many and over what time span?

I urge caution in interpreting the Japanese data [2]. That study did not show that all patients infected with HCV developed hepatomas 29 years later. Rather, Kiyosawa and colleagues [2] identified 21 patients with hepatomas who had previously been transfused and calculated that, on average, 29 years (range, <15 years to >60 years) had passed before the cancer was diagnosed. Even if we accept the transfusion date as the time of infection, we have no information regarding the denominator, that is, how many others were at risk but in whom cancer has not yet developed. If the probability of cancer development (or the appearance of liver failure in general) is low, intensive follow-up and expensive therapy may not be worthwhile.

I agree that we should not diminish our efforts to "prevent" HCV infections, but I wonder about "treating" (that is, using interferon) in these infections after they occur. Chronic viral hepatitis takes many years or decades to progress to clinically important liver disease. As at least one patient showed, histologic cirrhosis does not always mean that an imminent clinical problem exists. Likewise, an apparent biochemical remission did not prevent liver failure in our patient. Interferon therapy is expensive, cumbersome, and dangerous. The randomized trials only studied patients for 6 to 12 months, after which most of the untreated controls were crossed over to interferon, obviating any possibility of assessing the effect of the treatment on the development of liver failure [3]. Interferon may only make tests better. My perspective is that at least one randomized trial showing clinically meaningful efficacy (that is, the prevention of liver failure) is necessary before interferon use can be advocated.