Response to highly active antiretroviral therapy according to duration of HIV infection

Objective: To evaluate whether duration of HIV-1 infection influences the r esponse to highly active antiretroviral therapy (HAART). Design: Prospective study of individuals (Italian Seroconversion Study coho rt) with well-estimated dates of HIV-1 seroconversion. Methods: This analysis included 277 participants who began HAART (defined a s three antiretroviral drugs used in combination). Cox regression models we re used to evaluate the association between duration of infection (as categ orical variable [less than or equal to3, 3-7.5, >7.5 years from seroconvers ion] or continuous variable) and an immunologic (rise in CD4 count >100 cel ls/mm(3)) and a virologic (decline in plasma HIV-RNA to unquantifiable leve ls) outcome. All analyses were stratified by center of recruitment and adju stment, when used, was for gender, age at inception of HAART, injection dru g use. previous antiretroviral therapy, lag-time between positive and negat ive HIV test result, year of starting HAART, clinical stage, CD4 count, and HIV-RNA at time of HAART. Results: HAART was initiated a median of 6.4 years after seroconversion, Th ere was a median follow-up of 1.6 years after starting HAART to the calenda r cut-off (November 1999). One-hundred-eighty-one (65.3%) patients experien ced a decline in viral load to below quantifiable levels and 184 (66.4%) ex perienced a rise in CD4 >100 cells/mm(3). In the Cox models, by 1-year incr ease in duration of infection, we estimated a lower crude hazard of achievi ng a CD4 count increase >100 cells (relative hazard [RH], 0.96: 95% confide nce interval [CI]. 0.92-1.01; p = .09), and a lower hazard of reaching an u nquantifiable level of plasma HIV-RNA (RH, 0.97; 95%CI, 0.93-1.02; p = .20) . After adjustment, these values became 0.99 (95%CI, 0.93-1.04; p = .62) an d 0.98 (95%CI, 0.93-1.03; p = .48), respectively. When duration of HIV infe ction was considered as a categorical variable, the results were consistent with those already described. Conclusions: These results suggest that the duration of HIV infection does not seem to play an important independent role in determining the virologic and immunologic responses to HAART.