Background: Several studies of children with human immunodeficiency virus (HIV) type 1 infection have demonstrated sustained increases in CD4+ cell count, even when virological failure has occurred after receipt of highly active antiretroviral therapy (HAART), but these studies were of limited duration. Moreover, the CD4+ cell count threshold at which antiretroviral treatment should be initiated is still unsettled. The aim of this study was to define the long-term impact of HAART on CD4+ cell percentage and viral load according to CD4+ cell percentages before HAART was initiated.
Methods: We conducted a retrospective study of 113 pretreated HIV-1-infected children stratified by pre-HAART CD4+ cell percentage (<5%, 5%-15%, 15%-25%, and >25%). The inclusion criteria were as follows: initiating HAART with a protease inhibitor, having 6 years of follow-up after starting HAART, having a CD4+ cell count or viral load recorded before initiation of HAART, and having received mono- or dual-nucleoside therapy before starting HAART.
Results: During the first 2 years of HAART, HIV-1-infected children experienced a significant increase in CD4+ cell percentage and a decrease in viral load (P<.05). During their last 4 years of receiving HAART, we found a significant decrease in viral load but not an increase in CD4+ cell percentage, because the CD4+ cell percentage reached a plateau after the second year of HAART. Moreover, children with CD4+ cell percentages of <5% at baseline did not achieve CD4+ cell percentages of >25% after 6 years of HAART. Children with CD4+ cell percentages of 5%-25% at baseline had a strong negative association with achieving CD4+ cell percentages of >30% for at least 6 and 12 months but not with achieving CD4+ cell percentages of >30% for at least 24 months.
Conclusions: Long-term HAART allowed for restoration of CD4+ cell counts and control of viral loads in HIV-1-infected children. However, initiating HAART after severe immunosuppression has occurred is detrimental for the restoration of the CD4+ cell count.