The impact of initial highly active antiretroviral therapy on future treatment sequences in HIV infection

Objectives: To determine whether the initial use of non-nucleoside reverse transcriptase inhibitors (NNRTI) or protease inhibitors (PI) differentially influences subsequent HIV therapy. Design: A cohort study using a prospective clinical database in a university-based HIV clinic. Subjects: A total of 440 HIV-seropositive patients, naive or nucleoside experienced, initiating therapy with either an NNRTI or PI between January 1998 and July 2003 and followed to December 2003. Main outcome measures: Time until stopping the first regimen and until exposure to all antiretroviral classes (excluding tenofovir and enfuvirtide) according to the type of initial regimen. Results: A total of 291 subjects initiated HAART with PI and 149 with NNRTI; median follow-up 3.1 and 2.3 years, respectively. Subjects starting NNRTI remained on their initial regimens longer (median time to change 2.1 versus 1.6 years; log rank P = 0.03). Overall, subjects initiating NNRTI-based regimens were less likely to alter their therapy. Previous nucleoside exposure was an important predictor of treatment modification. Subjects initiating NNRTI-based HAART were also less likely to experience virological failure than those initiating PI-based HAART. Individuals starting with NNRTI were exposed to fewer regimens (15 versus 25% received three or fewer regimens), and showed a trend towards lower rates of three-class exposure (7 versus 12%). Conclusion: There is a high rate of treatment modification among patients initiating HAART. The initial use of NNRTI-based HAART was associated with more durable treatment and lower rates of virological failure, which may translate into a reduced need for multiple salvage therapies. (C) 2004 Lippincott Williams Wilkins.