In a recent paper, Wong and colleagues [1] concluded that the use of interferon in HBeAg-positive patients with chronic hepatitis B was cost-effective. This conclusion was based on two observations: 1) that interferon is effective in producing HBeAg loss and 2) that the natural history of HBeAg-negative chronic hepatitis is less ominous than that of HBeAg-positive disease.

The authors assumed that the natural history of disease in patients who did or did not seroconvert from HBeAg positivity with interferon was the same as that in patients who did or did not spontaneously seroconvert. However, certain factors predict interferon responses. Some of these factors (higher aminotransferase levels, lower hepatitis B viral DNA levels, and female sex) are also associated with spontaneous HBeAg or hepatitis B surface antigen clearance. Nonresponders tend not to have these factors, and they probably have a worse prognosis than untreated HBeAg-positive patients. An untreated group would contain more patients likely to seroconvert in the future.

In their appendix (a section some readers are less likely to read), the authors addressed the possibility that interferon merely accelerated the natural history and stated that according to this scenario, standard care was preferable. To date, no data prove that interferon actually prevents end-stage liver disease; a study lasting many years (or decades) is necessary to make that assessment. We also have no data to prove that altering the intermediate end point (a serologic event) translates into an improved clinical outcome. As we have seen, natural history studies can never answer this question. Studies using intermediate end points cannot prove the existence of clinical efficacy, and the calculations by Wong and coworkers cannot show that interferon treatment will allow us to reap economic benefit.