Tolerability and efficacy of PI versus NNRTI-based regimens in subjects receiving HAART during acute or early HIV infection

BACKGROUND: Little is known about modifications to highly active antiretroviral therapy (HAART) initiated during acute or early HIV infection. METHODS: Reasons for first modifications of HAART regimens were recorded using the AIDS Clinical Trials Group form among 363 subjects who initiated HAART within 1 year of seroconversion from 2005 in the Acute Infection and Early Disease Research Program. Modifications recorded as due to "patient choice" or "physician choice" were clarified by query to the recording site. Times to events were analyzed by Kaplan-Meier methods; significance of differences was assessed by the log-rank test. RESULTS: Two hundred five of 363 (56%) subjects modified therapy, at a median of 425 days after initiation, by changing drugs, discontinuing treatment, or removing or adding drugs. Most modifications were attributed to toxicity (n ≤ 105, 51%), most of which was low grade; regimen simplification (n ≤ 18, 5%); and achievement of viral suppression (n ≤ 15, 7%). Time to first modification was shorter for those with shorter time from infection to initiation (P ≤ 0.005) and those having higher CD4 lymphocyte count at initiation (P ≤ 0.06). Modifications occurred sooner in subjects receiving regimens taken more than once daily (P < 0.001) or with more than 2 pills daily (P < 0.001). Most regimens were nonnucleoside reverse transcriptase inhibitor based or protease inhibitor based, and these did not differ significantly in rate and timing of modification. CONCLUSIONS: HAART initiated early in HIV infection was modified in the majority of cases, usually due to minor toxicities whose incidence was similar for protease inhibitor-based and nonnucleoside reverse transcriptase inhibitor-based regimens. Convenience of regimens (lower pill burden and dosing frequency) was associated with a lower rate of modification.