Article Text

Download PDFPDF

Reply: ‘More viral mutants, less HBsAg clearance? One size may not fit all'
  1. Alex Thompson1,
  2. Stephen Locarnini2,
  3. Peter Revill2
  1. 1 Department of Gastroenterology, St. Vincent's Hospital, Melbourne, Victoria, Australia
  2. 2 Victorian Infectious Diseases Reference Laboratory, Royal Melbourne Hospital at the Peter Doherty Institute of Infection and Immunity, Melbourne, Victoria, Australia
  1. Correspondence to Dr Peter Revill, Victorian Infectious Diseases Reference Laboratory, Royal Melbourne Hospital at the Peter Doherty Institute of Infection and Immunity, 792 Elizabeth St, Melbourne, 3000 VIC, Australia; Peter.revill{at}

Statistics from

We thank Dr Tseng and colleagues for their interest1 in our recent study that identified a negative association between the presence of hepatitis B virus (HBV) basal core promoter (BCP)/precore (PC) and precore (PC) variants mutants at baseline and lower likelihood of hepatitis B surface antigen (HBsAg) loss among 157 subjects with hepatitis B e antigen (HBeAg)-positive chronic hepatitis B (CHB) treated for 4 years with tenofovir therapy. 2–4 The likelihood of HBsAg loss was 41% in subjects with wild-type (WT) sequence at the BCP/PC loci versus 3% in subjects with detectable BCP/PC variants (next generation sequencing (NGS), Illumina MiSeq—threshold for detection >1%). The cohort included subjects with a broad range of HBV genotypes (genotype A, n=36 (23%); genotype B, n=24 (15%); genotype C, n=51 (32%) and genotype D, n=46 (29%)). HBsAg loss was more common in subjects with genotype A/D (24%) versus genotype B/C HBV (1%), and although genotype A was associated with HBsAg loss following univariate analysis, HBV genotype was not associated with HBsAg loss in multivariate analysis after adjustment for the presence of BCP/PC variants. We acknowledge that an effect of HBV genotype cannot be excluded in a cohort of this size. Further, BCP/PC variants were detected almost universally in subjects with genotype B (24/24 (100%)) or genotype C (46/51 (90%)), whereas HBsAg loss was very rare (genotype B 1/24 (4%) and genotype C 0/51 (0%)). The analysis was, therefore, underpowered to detect an association between WT versus BCP/PC variants and HBsAg loss in subjects with genotype B/C infection.

Similarly, it is difficult to compare findings from the correspondent's cited work,5 which focused on HBsAg loss over decades in the natural history of subjects enrolled in the ERADICATE-B study with HBeAg-negative CHB, in the absence of antiviral therapy. Although no association was observed between BCP/PC variants and HBsAg loss, a TaqMan-based qualitative PCR was assay was used to determine the frequency of the dominant sequences. This population-sequencing approach was less sensitive than the NGS method used in our study, and it is likely that NGS would have identified BCP/PC variants in the vast majority of this HBeAg-negative cohort, meaning that power to detect an effect was small. We are also familiar with the study in which immunotolerant subjects were treated with tenofovir±emtricitabine for 4 years.6 Although subjects enrolled in this study had high rates of genotype B/C HBV and were likely to be WT at the BCP/PC loci, no HBsAg loss was observed. HBsAg loss is likely to involve a host–viral interaction that requires immune pressure absent among individuals in the immunotolerant phase of CHB. Indeed, we hypothesise that the BCP/PC variants represent an immune evasion mechanism that aborts serological control among HBeAg-positive individuals who have entered the immune control phase of disease.

In conclusion, the observed association between the presence of BCP/PC variants and lower likelihood of HBsAg loss in individuals with HBeAg-positive CHB treated with long-term tenofovir therapy was robust, particularly in genotype A/D HBV. We agree that future studies should evaluate the association between BCP/PC sequence and HBsAg loss during nucleos(t)ide analogue therapy for HBeAg-positive CHB in people infected with genotype B/C HBV. These studies will need to be adequately powered to account for the high prevalence of BCP/PC variants in HBeAg-positive genotype B/C HBV observed using NGS.


View Abstract


  • Contributors All authors contributed equally to this letter.

  • Funding Gilead Sciences.

  • Competing interests PR and SL have received grant/research support from Gilead Sciences; SL is an advisor for Gilead Sciences and Bristol-Myers Squibb (BMS); AT is an advisor for and has received grant/research support from Merck, Janssen/Tibotec, Gilead Sciences and Roche and speaker's fees from Merck, Roche and BMS. PR is currently in receipt of a Royal Melbourne Hospital Keir Research Fellowship.

  • Patient consent Obtained.

  • Ethics approval Gilead Sciences.

  • Provenance and peer review Not commissioned; internally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.