Abstract

BACKGROUND:

Data comparing abacavir/lamivudine versus tenofovir/emtricitabine in antiretroviral-naive patients are controversial. We compared 48-week efficacy and safety of these combinations as substitutes of nucleosides in patients with virological suppression.

METHODS:

We randomly assigned 333 HIV-1-infected patients on lamivudine-containing triple regimens with <200 copies per milliliter for at least 6 months to switch their nucleosides to either abacavir/lamivudine (n = 167) or tenofovir/emtricitabine (n = 166). The primary outcome was treatment failure ["switching = failure" intention to treat (ITT) analysis, noninferiority margin 12.5%]. Secondary outcomes were time to treatment failure, virological failure, adverse events, and changes in CD4 count, fasting plasma lipids, lipodystrophy, body fat, bone mineral density, and renal function.

RESULTS:

Treatment failure occurred in 32 patients (19%) on abacavir/lamivudine and 22 patients (13%) on tenofovir/emtricitabine [difference 5.9%; (95% confidence interval -2.1% to 14.0%), P = 0.06]. Four patients in the abacavir/lamivudine group versus none in the tenofovir/emtricitabine group developed virological failure [difference 2.4; (95% confidence interval 0.05 to 6.0), P = 0.04]. Twenty-three patients (14%) assigned to abacavir/lamivudine and 10 (6%) to tenofovir/lamivudine experienced grade 3 or 4 adverse effects (P = 0.03). CD4 counts and plasma lipids showed higher increments in the abacavir/lamivudine group than in the tenofovir/emtricitabine group.

CONCLUSIONS:

In HIV-1-infected patients with virological suppression, abacavir/lamivudine did not meet the noninferiority outcome for treatment efficacy compared with tenofovir/emtricitabine.