Hepatitis B virus (HBV) co-infection, common in HIV-infected patients in sub-Saharan Africa, is associated with impaired immunological recovery while on antiretroviral treatment (ART), and worse clinical outcomes, even in the context of effective ART. A clear association between HBV co-infection and early mortality has not been shown in African settings and the impact of tenofovir-containing ART on outcomes is unknown.
The prevalence of hepatitis B surface antigen (HBsAg) was determined in a cohort of patients enrolling in an ART programme in Kenya between 2003 and 2012. Clinical outcomes, immunological and virological responses to ART were compared between HIV mono-infected and HIV/HBV co-infected patients using Cox regression. The impact of tenofovir-containing ART,phased in over the observation period, on outcomes was determined in an analysis adjusting for confounders relating to time and indication.
7,155 patients were enrolled in the cohort and followed for 12,408 person years. HBsAg was detected in 451/7155 (6.3%, 95%CI 5.8-6.9%) of patients. HBsAg prevalence was higher in men than women (9.2% versus 5.0%, p<0.001) and increased with age.HBsAg positivity was associated with increased risk of death in crude analysis (HR 1.84, 95%CI 1.4-2.5, p<0.001). Among those initiating ART (n=6,214), HBsAg positive patients (n=419) had significantly impaired immunological recovery within the first year of ART compared to HBsAg negative patients (median CD4 cell count increase 110 cells/mLvs135 cells/mL, p=0.03), despite similar rates of virological suppression (90%vs88%, p=0.32). HBsAg positivity remained an independent predictor of mortality in adjusted analysis (aHR1.84, 95%CI 1.3-2.6, p=0.001). Among patients initiating tenofovir-containing regimens (n=3125), HBsAg positivity (n=350) was no longer significantly associated with increased risk of mortality (aHR 1.45, 95%CI 0.9-2.2, p=0.1) in contrast to the markedly increased risk in patients receiving non-tenofovir based regimens (aHR 3.32, 95%CI 1.8-6.2, p<0.001), p-value for interaction = 0.03.
Hepatitis B co-infection was associated with impaired immunological responses to ART and increased risk of mortality in this large cohort of Kenyan patients initiating ART despite adequate HIV virological suppression and no evidence for severe liver disease. Use of tenofovir-containing regimens significantly reduced mortality risk in HIV/HBV co-infected patients.