Predictive value of HIV-1 genotypic resistance test interpretation algorithms

Abstract

Background: Interpreting human immunodeficiency virus type 1 (HIV-1) genotypic drug-resistance test results is challenging for clinicians treating HIV-1-infected patients. Multiple drug-resistance interpretation algorithms have been developed, but their predictive value has rarely been evaluated using contemporary clinical data sets.

Methods: We examined the predictive value of 4 algorithms at predicting virologic response (VR) during 734 treatment-change episodes (TCEs). VR was defined as attaining plasma HIV-1 RNA levels below the limit of quantification. Drug-specific genotypic susceptibility scores (GSSs) were calculated by applying each algorithm to the baseline genotype. Weighted GSSs were calculated by multiplying drug-specific GSSs by antiretroviral (ARV) potency factors. Regimen-specific GSSs (rGSSs) were calculated by adding unweighted or weighted drug-specific GSSs for each salvage therapy ARV. The predictive value of rGSSs were estimated by use of multivariate logistic regression.

Results: Of 734 TCEs, 475 (65%) were associated with VR. The rGSSs for the 4 algorithms were the variables most strongly predictive of VR. The adjusted rGSS odds ratios ranged from 1.6 to 2.2 (P < .001). Using 10-fold cross-validation, the averaged area under the receiver operating characteristic curve for all algorithms increased from 0.76 with unweighted rGSSs to 0.80 with weighted rGSSs.

Conclusions: Unweighted and weighted rGSSs of 4 genotypic resistance algorithms were the strongest independent predictors of VR. Optimizing ARV weighting may further improve VR predictions.

Figure 1

Schematic definition of a treatment-change episode (TCE). Brackets indicate 3 TCE requirements: (1) a baseline genotypic drug-resistance test result obtained within 24 weeks before the treatment change, (2) a baseline plasma human immunodeficiency virus type 1 (HIV-1) RNA level obtained within 8 weeks before the treatment change, and (3) ≥2 plasma HIV-1 RNA levels obtained between 4 and 36 weeks after the treatment change. Additional requirements were a CD4 cell count obtained within 24 weeks before the treatment change and ≥4 weeks of salvage therapy.

Figure 2

Three-step method for calculating regimen-specific genotypic susceptibility scores (GSSs): (1) calculate drug-specific GSSs, applying a genotypic drug-resistance interpretation algorithm to the mutations present in the baseline genotype to estimate the activity of an antiretroviral (ARV) relative to its activity against a wild-type virus; (2) calculate the weighted drug-specific GSSs, multiplying drug-specific GSSs by a factor that accounts for differences in ARV activity between ARVs belonging to different ARV classes; and (3) calculate regimen-specific GSSs by adding the weighted drug-specific GSSs of the drugs used for salvage therapy.

Figure 3

Description of the complete patient population from which the treatment-change episodes (TCEs) were derived. Of 4654 patients who underwent genotypic drug-resistance testing between 1998 and 2007, 3600 patients had a known history of antiretroviral (ARV) treatment. Of these, 3029 had received ≥1 ARV drug treatment regimens, and 641 subsequently met study enrollment criteria. These 641 patients had 734 valid TCEs. HIV, human immunodeficiency virus.

Figure 4

Classification tree for the comprehensively weighted version of regimen-specific genotypic susceptibility scores (rGSSs) generated by the HIVdb algorithm. The root node contains all 734 treatment-change episodes (TCEs). The root node and all child nodes show the number of TCEs resulting in virologic failure (left side of node) or virologic success (right side of node). Using the R package Rpart (Recursive partitioning; version 3.1-39), an explanatory variable and a cutoff value were selected at each node to optimally separate virologic responders from nonresponders in the child nodes. Tree pruning was performed using the default parameters in Rpart for minimizing the complexity of the tree as a function of the improvement in classification accuracy. ARV, antiretroviral; HIV, human immunodeficiency virus.