We thank Dr. Lo for his insightful comments. We agree that the finding of greater hepatitis B surface antigen (HBsAg) loss among patients with high baseline viral load (HVL) than those with non-HVL is of interest.
In a previous report of this cohort, the 23 individuals who achieved HBsAg loss by week 192 of therapy were exclusively hepatitis B e antigen (HBeAg)+ at baseline. In addition, those individuals who achieved HBsAg loss had higher hepatitis B virus (HBV)-DNA levels at baseline, compared to HBeAg+ subjects who did not achieve HBsAg loss (median, 9.31 vs. 8.82 log10 copies/mL, respectively).
The viral genotype distribution among the 23 subjects with HBsAg loss was genotype (gt)-A: 14, gt-B: 1, gt-D: 7, and gt-F: 1, whereas among HBeAg+ individuals who did not experience HBsAg loss, distribution was gt-A: 46, gt-B: 32, gt-C: 69, gt-D: 79, gt-E: 5, and gt-F: 7.1 Among the 23 subjects who lost HBsAg, 7 were in the adefovir arm and all experienced HBsAg loss after the addition of tenofovir.
We also examined surface antigen levels and found that the median levels of HBsAg at baseline were 5.07 log10 IU/mL among those who lost HBsAg, compared to 4.51 among those who did not achieve HBsAg loss. Patients who experienced HBsAg loss had rapid, continuous declines of HBsAg, with the decline at week 12 >0.5 log10 IU/mL. No Asian patient achieved HBsAg loss.
In multivariable analysis, having had a diagnosis of HBV for ≤4 years was associated with HBsAg loss, compared to those who had their HBV diagnosed 5 years or longer (odds ratio: 6.9; P = 0.0375). Therefore, Dr. Lo raises several intriguing questions that will require additional studies to better understand, including potential differences between spontaneous versus therapy-mediated HBsAg loss.
Stuart C. Gordon, M.D.1Leland J. Yee, Ph.D., MPH2Eduardo Bruno Martins, M.D., D.Phil.2
1Henry Ford Hospital, Detroit, MI
2Gilead Sciences, Foster City, CA