To the Editor:
The recent article by Berg1 reviews the efficacy of pegylated interferon (PEG-IFN) for the treatment of chronic hepatitis B (CHB) based on the results of recent trials.2, 3 The author suggests that nucleos(t)ide analogs (NAs) may be as effective as (PEG)IFN-based treatments when considering HBeAg response rates achieved over longer treatment durations. We believe this conclusion to be unproven; long-term data on the efficacy of NAs are scarce and may be subject to bias.
In our opinion, the HBeAg seroconversion rates reported after long-term continuous adefovir therapy should be interpreted with caution.4 These rates were determined using Kaplan-Meier estimation, which could overestimate response rates. A basic assumption of this approach is that a patient will retain the measured outcome whether or not they remain in the study. Because seroreversion is known to occur in CHB, the suitability of this approach for determining efficacy is questionable. Response rates thus generated should be interpreted carefully, particularly if relapse rates (or seroreversion) are not provided. It is also unclear how many patients remained in the adefovir study during long-term evaluation and whether patients in the long-term study differed from nonparticipants; selection bias might influence the results. Similarly, the long-term rates of response to lamivudine are from a cohort study.5 More recent data from a large retrospective study showed HBeAg seroconversion rates of 35% after 5 years of lamivudine treatment.6 After the first year of lamivudine treatment, HBeAg seroconversion rates increased by only 1% to 7% per year. In comparison, the HBeAg seroconversion rates achieved with PEG-IFNα-2b (29%) and PEG-IFNα-2a (32%) after a finite 1-year treatment duration and only 6 months of follow-up look particularly favorable.2, 7
NAs have considerably broadened our treatment options for CHB. Nevertheless, current treatment strategies for CHB could be differentiated into those that provide sustained off-therapy responses with a finite treatment duration (IFN-based) and those that provide therapy-maintained responses (NAs). This reflects the different modes of action of the treatments: while NAs have a primarily antiviral mode of action, IFNs have dual immunomodulatory and antiviral effects. Although NAs are potent suppressors of viral replication, the relatively low rates of HBeAg seroconversion achieved after 1 year (approximately 20% with lamivudine, 12% with adefovir, and 21% with entecavir), suggest that suppression of HBV DNA is not always sufficient to tip the balance in favor of the host's immune response. HBeAg seroreversion rates as high as 50% have been reported following NA-based treatment,8 and the combination of PEG-IFNα-2b and lamivudine has been shown to provide significantly higher rates of sustained response than lamivudine monotherapy 3 years after the end of treatment (29% vs. 9%, respectively).9 As far as sustained HBeAg seroconversion is concerned, IFN-based therapy seems to have the edge over NAs, even if the mechanisms behind this remain undefined. Knowledge gained in this area may help improve rates of HBeAg seroconversion and, ultimately, HBsAg seroconversion. With IFN-based therapies, HBsAg seroconversion seems to be an achievable goal, in both the short and long terms.2, 3, 10