Letter: Should HBV Therapy Be Stopped Based on HBsAg Level Alone? Authors' Reply

Abstract

We thank Khan et al. for thorough and critical reading of our article “An open-label, randomized trial of different re-start strategies after treatment withdrawal in HBeAg negative chronic hepatitis B” [1]. Here, we would like to comment on some of the issues raised in their Letter [2].

As Khan et al. points out, two-thirds of the study participants in our Nuc-Stop Study had end-of-treatment HBsAg levels above 1000 IU/mL and none of these achieved the primary endpoint of HBsAg loss 36 months after nucleos(t)ide analogue (NA) withdrawal. This is in line with results from the RETRACT-B and CREATE studies, both of which were performed after our study was initiated [3, 4]. When we designed our study, the two available reports on frequency of HBsAg loss after NA withdrawal indicated approximately 20% chance of HBsAg loss after 3 years, but these studies did not study the impact of end-of-treatment HBsAg levels [5, 6]. More recent studies have found lower rates of HBsAg loss, ranging from 5% to 10% 3 years after NA withdrawal, and a clear correlation with end-of-treatment HBsAg levels [3, 4, 7]. Therefore, studies attempting to further elucidate the optimal re-treatment strategy after NA withdrawal should restrict enrollment to patients with HBsAg level below 1000 IU/mL, and even lower in patients of Asian ethnicity with HBV genotypes B/C.

Khan et al. correctly point out that NA withdrawal is not without risks. In our study, 25.4% of patients in the high-threshold arm and 14.1% in the low-threshold arm experienced severe medical events as defined by the study protocol; however, most of these episodes were considered unrelated to NA withdrawal. In total, 40.2% of the study participants experienced clinical relapse (HBV DNA > 2000 IU/mL and ALT > 80 U/L) but only 9.4% had a severe flare (ALT > 800 U/L), and none developed signs or symptoms of hepatic decompensation. More recently, we have shown that HBsAg level is a strong predictor of hepatic flares [8], consistent with findings from the RETRACT-B study [9].

We agree that to tailor the treatment strategy for the individual patient with chronic hepatitis B, better biomarkers are warranted, both to identify patients likely to achieve a functional cure after NA withdrawal, and also to pinpoint those at increased risk of severe hepatic flares. We are currently assessing HBV RNA and hepatitis B core related antigen (HBcrAg) in stored samples from patients in the Nuc-Stop Study to assess the diagnostic and prognostic performance of these two novel markers.

Stopping NA therapy is a simple and cheap intervention to boost functional cure in HBeAg negative chronic hepatitis B, but better biomarkers are needed to identify the subgroup most likely to benefit from it. In our study, we found signals that delaying re-treatment might increase the chance of HBsAg loss in those with end-of-treatment HBsAg below 1000 IU/mL, but the final verdict of an optimal NA stopping (and re-treatment) strategy in HBeAg negative chronic hepatitis B is yet to be decided.