Association of HIV diversity and time to virologic suppression and study endpoints.

The table shows results from age-stratified Cox proportional hazards regression models for the association of HIV diversity with time to virologic suppression (defined as two consecutive HIV viral load values <400 copies/mL) and study endpoints, including: composite outcome of virologic failure or death (VF/death); composite outcome of virologic failure or discontinuation of study treatment (VF/off-treatment); and virologic failure alone (VF alone). Note that higher POL region diversity is positively associated with a shorter time to virologic suppression (a good clinical outcome, HR: 1.57) and is negatively associated with a longer time to study endpoints that are bad clinical outcomes (VF/off-treatment, HR: 0.35; VF alone, HR: 0.19; VF/death, HR: 0.16). In other words, higher POL region diversity is associated with improved clinical outcome for all four variables. Analyses were adjusted for study treatment regimen, age (years), CD4 cell percentage (CD4%), and HIV viral load (<750,000 vs. ≥750,000 copies/mL). HRM scores were obtained for six regions of the HIV genome (see Methods); analyses were also performed for the mean (MEAN) and median (MED) of all six HRM scores. Hazard ratios (HR) are shown as per unit increase in HRM score. P values <0.05 are bolded. Abbreviations: HR: hazard ratio; CI: confidence intervals; VF: virologic failure.