Optimal antiretroviral drug levels improve outcomes

Most HIV clinicians are familiar with the results of the genotypic antiretroviral resistance testing (GART) trial, which indicated that genotyping improves virologic outcomes in patients failing a protease inhibitor (PI)-based highly active antiretroviral therapy (HAART) regimen, at least in the short term.

In a follow-up study, recently published in AIDS, the GART investigators found, not too surprisingly and consistent with at least one other study, that the antiretroviral drug levels achieved also were significant predictors of virologic outcomes.

In this retrospective follow-up to the GART trial, one objective was to determine the impact of antiretroviral drug levels (DL) during follow-up on virologic response to the antiretroviral regimen selected. Drug levels were assessed using untimed plasma specimens and the distribution of DL after 12 weeks of follow-up was classified as above (DLHigh) or below (DLLow) the median Inhibitory quotients [IQ = (DL at week 12)/(fold change in IC50 to wild-type)].

The investigators found that highly active antiretroviral therapy (HAART) regimens in which the antiretrovirals used were classified as DLHigh were associated with higher levels of viral suppression. Overall, the change in viral load was -0.18 if no drugs in the regimen had an IQ above the median, compared to -0.58, -1.06, -0.86, and -1.44 log10 copies/mL if one, two, three or four or more drugs, respectively, had an IQ above the median.

Further, for each active PI classified as DLHigh there was a viral load change of -0.85 log10 copies/mL change in HIV RNA levels at 12 weeks compared to a -0.55 log10 copies/mL change for each active PI classified as DLLow, and a -0.16 log10 copies/mL change for each inactive PI (i.e., a PI to which resistance testing indicated HIV resistance). For non-nucleoside reverse transcriptase inhibitors (NNRTI), DLHigh was associated with a -1.03 log10 copies/mL change in HIV RNA levels at 12 weeks compared to a -0.43 log10 copies/mL change for NNRTI drug levels below the median.

For nucleoside analogues (NAs), with at least one active NA with a DL above the median, the mean change in viral load was -0.80 log10 copies/mL compared to -0.68 log10 copies/mL with at least one active NA with a DL below the median, and -0.78 log10 copies/mL change with an inactive NA. Finally, the viral load reduction associated with each drug in the regimen with an IQ greater that the median was -0.27 (P=0.0001).

The authors conclude, “Previous analyses have shown that patients who receive three or more active drugs in their salvage regimens achieve the best responses. This analysis suggests that optimal drug levels are also important in achieving a successful response to subsequent regimens after virologic failure with HAART. Further study of the role of TDM in different patient populations should provide additional information on the potential use of antiretroviral drug levels in the management of HIV disease.”