Background:  Continuous HAART is standard of care for HIV-infected patients but lifelong adherence and tolerance are important concerns. We assessed the safety of a 50% drug-sparing strategy.

Methods:  We randomly assigned to 2 groups 403 adults with a nadir CD4 >100/µL, plasma HIV RNA viral load <200 copies/mL, and CD4 ≥450/µL for ≥6 months while on HAART. One group switched to an 8-weeks-off/8-weeks-on intermittent therapy; the other maintained continuous therapy for 96 weeks. The primary outcome was the cumulative incidence of immunologic failures (confirmed CD4 <300/µL). Non-inferiority of the intermittent vs continuous therapy arm was obtained if the upper bound of the 95% confidence interval of the difference (UBCID) in proportion of immunologic failures remained £7%.

Results:  Of 403 patients, 12 withdrew consent. At baseline, 80% of patients were male, with a median age of 41.6 years, and 8% had AIDS. Median CD4 was 741/µL with a nadir at 280/µL; 43% and 46% of patients were on a non-nucleoside reverse transcriptase inhibitor (NNRTI)- and protease inhibitor (PI)-based regimen, respectively; 7 (4%) and 12 (6%) patients in the intermittent and continuous arm were lost to follow-up. In the intermittent are, 2 patients died (deaths were not related to HIV). No AIDS-defining event was recorded. Median (IQR) proportion of time on therapy, intermittent vs continuous, was 51.5% (49.5, 55.7) and 100% (100, 100). The proportion who reached the primary immunologic endpoint was:  in intent-to-treat analysis 7 (3.6%) vs 3 (1.5%) (UBCID: 5.6%); in per patient analysis 6 (3.5%) vs 1 (0.6%) (UBCID: 6.5%). All these patients had a nadir CD4 <300/µL. At week 96, the proportion of patients, intermittent vs continuous, with CD4 >450 /µL and with viral load ≤400 copies/mL was 75% vs 92% (p <0.0001) and 81% vs 90% (p = 0.02). The first 200 patients were enrolled in a virologic sub-study. The proportions of patients with a viral load ≥1000 copies/mL after ≥6 weeks of therapy (defined as virologic failure) were similar in the intermittent therapy (17 of 98) and continuous therapy arms (14of 99, Log rank p = 0.51). Resistance genotype available for 24 patients (14 intermittent therapy, 10 continuous therapy) was wild type in 9 (7, 2, respectively) and with main resistance mutations in 15 patients (7, 8, respectively). Median number of mutations per patient was 4 in the intermittent therapy and 5 in the continuous therapy arm (p = 0.07). Resistance mutations to protease inhibitors were detected in only 3 patients (1, 2, respectively), to NRTI in 13 patients (7, 6, respectively), and to NNRTI in 7 patients (1, 6, respectively).

Conclusions:  In this population of patients, a fixed structured treatment interruption strategy of 8-weeks-off/8-weeks-on appeared clinically and immunologically safe over 96 weeks while sparing 48.5% of drug exposure. Patterns of drug resistance mutations in patients with virologic failure appeared similar.