Identifying Optimal HIV Viral Load (VL) Thresholds for Predicting Antiretroviral Treatment Failure (TF) Using ROC Curve Analysis
Methods: Longitudinal patient specimens were collected through the VERxVE study, a randomized, double-blind (DB) study in treatment-naïve patients treated with nevirapine immediate-release or extended-release (VIRAMUNE IR or XR), plus emtricitabine and tenofovir DF. Stored plasma samples from up to 24 timepoints (pre-treatment through 144 weeks of DB treatment), were tested with the Roche COBAS® AmpliPrep/COBAS® TaqMan® HIV-1 Test, v2.0. The BRAVO patient endpoints of TF (which differed from the primary VERxVE endpoints) were determined by two independent HIV physicians based on VL trajectories. A simple logistic regression model and ROC curves were created to determine the optimal VL thresholds for predicting TF at each study timepoint.
Results: Of 526 evaluable patient-series, 71 patients (13%) were TF by BRAVO criteria. By ROC curve analysis, a VL threshold of 95 copies/ml at week 24 maximized the sensitivity (88%, 95% CI: 84-90%) and specificity (56%, 95% CI:41-70%) of predicting eventual TF [AUC (Area Under Curve):0.75]. Overall sensitivity and specificity were similar (overlapping 95% CIs for the sum of sensitivity and specificity) when compared to ROC curves for commonly used VL thresholds of 50 or 200 copies/ml.
Conclusion: ROC curve analysis identified a VL threshold that maximized sensitivity and specificity in predicting TF. Viral load thresholds of 50 or 200 copies/ml can also be used without a significant decrease in overall sensitivity and specificity. As VL assays may differ in sensitivity, ROC curve analysis can be help optimize the clinical utility of VL monitoring.
Roche Molecular Systems: Employee, Salary
P. Robinson, Boehringer Ingelheim: Employee, Salary
A. M. Quinson, Boehringer Ingelheim: Employee, Salary
T. Do, Roche Molecular Systems: Employee, Salary
See more of: Poster Abstract Session