Abstract 1000 Poster Session IV, Tuesday, 5/4 (poster 105)

Background: Experience in adults has shown that combination therapy including HIV protease inhibitors (PI) can profoundly affect viral replication and slow progression of HIV-associated disease. Trials defining the influence of PI's and combination therapies on viral load (VL) and long-term outcome of HIV-infection in children have not yet been completed. Experience with infants and children who were receiving routine care in an HIV specialty clinic was reviewed to characterize the effect of changes involving one, two or three antiretrovirals. This review was aimed at estimating the potential impact of changing therapy and the character of changes associated with the greatest effect on HIV-infection in children.

Methods: Clinical and laboratory findings of children in whom antiretroviral therapy (ArT) was changed were reviewed. Successful response was defined as a reduction of VL of at least 0.5 log10 RNA copies/mL lasting for at least 3 months. Differences in characteristics and the character of the response associated with successful and unsuccessful changes were analyzed.

Results: 54 children were followed for a mean of 12.4 months (range 3-17 months) after a change in ArT. 10 of these children (18.5%) reduced their VL to undetectable levels. Of the 72 changes in ArT that were experienced by these 54 children, 35 resulted in a successful response. (Numbers in brackets in table indicate changes involving a PI).

Table 1 No caption
Full size table

A change involving three antiretrovirals was more likely to produce a successful response than a change involving one (8/9 v 8/25, p<0.01) or two agents (8/9 v 19/38, p<0.02). Reduction of VL by greater than 100-fold or to undetectable levels occurred more frequently in children who responded to a regimen containing a PI than in children who responded to reverse transcriptase inhibitors (RTIs;11/23 v 1/12, p=0.05). Furthermore, successful responses associated with addition of a PI were associated with a greater reduction in VL than those that involved RTIs (1.51 ± 0.68 v 0.86 ± 0.22 log10, p=0.003).

Conclusions: This experience suggests that changing ArT in HIV-infected children to regimens containing three drugs (one of which is a PI) is more likely to result in a successful virologic outcome than changes in ArT involving one or two drugs. Optimal outcome which is generally considered to be reduction of VL to undetectable levels, occurred less frequently in our experience compared to trials that have assessed the effects of ArT that included a PI. Failure to adhere to ArT and development of resistance may be possible explanations for this difference.