Abstract

BACKGROUND:

We explore the temporal and regimen-specific changes of HIV-1 drug resistance in a large cohort of antiretroviral-naive individuals starting highly active antiretroviral therapy (HAART).

METHODS:

Individuals (n = 2350) initiating first HAART between August 1996 and November 2004 were followed until November 2005 (median follow-up, 4.8 years; n = 6066 tests). A logistic regression model was used to predict the probability of the emergence of resistance, adjusting for baseline predictors.

RESULTS:

The cohort included 991 individuals initiating nonboosted protease inhibitor (PI)-based regimens, 475 initiating ritonavir-boosted PI-based regimens, and 884 initiating nonnucleoside reverse-transcriptase inhibitor (NNRTI)-based regimens. There was no difference in the development of resistance between nonboosted PI-based regimens (reference group) and NNRTI-based HAART regimens (odds ratio [OR], 1.09 [95% confidence interval {CI}, 0.84-1.42]), but there were greatly reduced odds for boosted PI-based regimens (OR, 0.42 [95% CI, 0.28-0.62]). Individuals initiating HAART more recently (2002-2004) were at a reduced risk of resistance, compared with those who started HAART between 1996 and 1998 (OR, 0.43 [95% CI, 0.30-0.61]).

CONCLUSIONS:

Individuals initiating first HAART with a boosted PI-based regimen had a 2.4-fold lower OR for developing HIV drug resistance than did those starting nonboosted PI-based or NNRTI-based HAART, at all adherence levels. The data demonstrate marked temporal improvement in the likelihood of the development of drug resistance for those initiating more recent HAART regimens.