Efficacy of 1998 vs 2006 first-line antiretroviral regimens for HIV infection: an ordinary clinics retrospective investigation

Purpose: The evidence suggesting increased HAART efficacy over time comes from randomized trials or cohort studies. This retrospective multicenter survey aimed to assess the variation over time in the efficacy and tolerability of first-line HAART regimens in unselected patients treated in ordinary clinical settings. Methods: Retrospective analysis of data of all patients starting first-line HAART regimens in 1998 and 2006 at adhering centers in the Italian CISAI group. Results: For the 543 patients included, mean age was 39.1 ± 9.8y in 1998 and 41.0 ± 10.7y in 2006 (p=0.03), with a similar proportion of males. Baseline mean log10 HIV-RNA was 4.56 ± 0.97 copies/mL in 1998 vs 4.91 ± 0.96 copies/mL in 2006 (p<0.001); baseline mean CD4 T-cell counts were 343 ± 314/mm3 in 1998 vs 244 ± 174/mm3 in 2006 (p<0.001). The following outcomes were significantly improved at 48w in 2006: proportion with undetectable HIV-RNA (86.3% vs 58.0%; p<0.001); mean increase in CD4 T-cells count (252 ± 225 vs 173 ± 246; p<0.001); HAART modification (20.1% vs 29.2%; p=0.02); HAART interruption (7.3% vs 14.6%; p=0.01); proportion reporting optimal adherence (92.2% vs 82.7%, p=0.03). No differences were observed in the prevalence of grade 3-4 WHO toxicities (26.4% vs 26.6%; p=0.9). Multivariate logistic regression showed that being treated in 1998 remained an independent predictor of virological failure after several adjustments, including adherence. Conclusions: Our data from patients not included in clinical trials or cohort studies provide an additional line of evidence that the effectiveness of HAART significantly improved in 2006. Treated patients, however, were significantly older and more frequently late HIV presenters in 2006 than in 1998.