A comparison between abacavir and efavirenz as the third drug used in combination with a background therapy regimen of 2 nucleoside reverse-transcriptase inhibitors in patients with initially suppressed viral loads

BACKGROUND: Our objective was to compare the rate of viral rebound and therapy failure in patients receiving abacavir or efavirenz as the third drug (in addition to 2 non-abacavir nucleosides) in combination antiretroviral therapy (cART) and to compare the rate of metabolic alteration associated with these regimens. METHODS: We conducted a multicohort prospective observational study of human immunodeficiency virus-infected patients who had attained viral loads < or = 80 copies/mL while receiving cART, without having previously received antiretrovirals. The rates of virological rebound, therapy failure, and lipid-level alteration during follow-up were calculated as the number of events divided by person-years of follow-up (PYFU). A multivariable analysis was performed using a Poisson regression model. RESULTS: We studied a total of 744 patients; the median age was 37 years, 27% of the patients were female, and 41% were heterosexual. There was a total of 854 PYFU spent receiving efavirenz and 285 spent receiving abacavir. The nucleoside reverse-transcriptase inhibitor pairs most frequently used were zidovudine/lamivudine (66% of PYFU), stavudine/lamivudine (17.6%), and stavudine/didanosine (5.4%). The adjusted relative rates of virological failure and therapy failure for abacavir, compared with those for efavirenz, were 2.17 (95% confidence interval [CI], 1.12-4.18; P = .02) and 1.41 (95% CI, 1.01-2.01; P = .05), respectively. CONCLUSIONS: Patients with virological suppression while receiving regimens containing abacavir appear more likely to experience virological and therapy failure than those receiving efavirenz as their third drug. Although this is a selected group of adherent patients, bias cannot be ruled out, because this is a nonrandomized comparison.