ABSTRACT
The nucleotide substitution G1896A on the precore (pc) region has been implicated in virological and serological responses during treatment in hepatitis B virus (HBV)-infected patients. Whether this mutation affects the therapeutic course of HIV-HBV co-infected patients, especially from Western Africa, is unknown. In this prospective cohort study, 86 antiretroviral (ARV)-naïve HIV-HBV co-infected patients from Côte d'Ivoire, initiating ARV-treatment containing lamivudine (n = 53) or tenofovir (n = 33), had available baseline pc sequences. Association of the pcG1896A mutation with time to undetectable HBV-DNA, hepatitis B "e" antigen (HBeAg) seroclearance (in HBeAg-positive patients), and hepatitis B surface antigen (HBsAg) seroclearance was evaluated using Cox proportional hazards regression. At ARV-initiation, median HBV-DNA was 6.04 log10 copies/mL (IQR = 3.70-7.93) with 97.7% harbouring HBV genotype E. Baseline pcG1896A mutation was identified in 51 (59.3%) patients, who were more commonly HBeAg-negative (P < .001) and had basal core promotor A1762T/G1764A mutations (P < .001). Patients were followed for a median 36 months (IQR = 24-36). Cumulative proportion of undetectable HBV-DNA was significantly higher in patients with baseline mutation (pcG1896A = 86.6% vs no pcG1896A = 66.9%, P = .04), but not after adjusting for baseline HBV-DNA levels and anti-HBV agent (P = .2). No difference in cumulative proportion of HBeAg seroclearance was observed between mutation groups (pcG1896A = 57.1% vs no pcG1896A = 54.3%, P = .7). Significantly higher cumulative proportion of HBsAg seroclearance was observed in patients without this mutation (pcG1896A = 0% vs no pcG1896A = 36.9%, P < .001), even after adjusting for baseline HBsAg quantification and anti-HBV agent (P < .001). In conclusion, lacking the pcG1896A mutation before ARV initiation appeared to increase HBsAg seroclearance rates during treatment. The therapeutic implications of this mutation need further exploration in this setting.
