Short duration of lamivudine for the prevention of hepatitis B virus transmission in pregnancy: lack of potency and selection of resistance mutations.

This study sought to assess the antiviral efficacy of lamivudine (LMV) administered during third trimester to reduce maternal viraemia and to identify the emergence of LMV resistance. A prospective observational analysis was performed on 26 mothers with high viral load (>10(7) IU/mL). Twenty-one women received LMV (treated group) for an average of 53 days (range 22-88 days), and the remaining five formed the untreated control group. Serum samples from two time points were used to measure HBV DNA levels and antiviral drug resistance. The LMV-treated women achieved a median HBV DNA reduction of 2.6-log(10) IU/mL. Although end-of-treatment (EOT) HBV DNA in four (18%) LMV-treated women remained at >10(7) IU/mL (+/- 0.5 log IU/mL), no mother-to-baby transmission was observed. In contrast, a baby from the untreated mother was HBsAg positive at 9 months postpartum. Four technologies were used for drug resistance testing. Only ultra-deep pyrosequencing (UDPS) was sufficiently sensitive to detect minor viral variants down to <1%. UDPS showed that LMV therapy resulted in increased viral quasispecies diversity and positive selection of HBV variants with reverse transcriptase amino acid substitutions at sites associated with primary LMV resistance (rtM204I/V and rtA181T) in four (19%) women. These viral variants were detected mostly at low frequencies (0.63-5.92%) at EOT, but one LMV-treated mother had an rtA181T variant that increased from 2.2% pretherapy to 25.59% at EOT. This mother was also infected with the vaccine escape variant (sG145R), which was inhibited by LMV treatment. LMV therapy during late pregnancy only reduced maternal viraemia moderately, and drug-resistant viral variants emerged.