A Predictive Model of the HIV Infection Course

Abstract

The course of HIV infection depends on the degree of HIV suppression during ART.The objective: to evaluate feasibility of using three factors: viral load (VL), HIV reservoir and CD8 T cells before ART initiation to estimate future levels of “residual” viremia and develop a predictive model.Subjects and Methods. 74 naive patients with HIV infection were enrolled in the study. The HIV reservoir, VL and CD4, CD8 counts were tested before ART initiation and then every 6 months. The patients were followed up for 12 months. Discriminant analysis was used.Results. After 12 months of ART, 69/74 (93.2%) patients had undetectable VL including 14 patients with residual viremia (RV) and 55 in whom no virus was detected. The probability of persisting RV after 12 months of ART was assessed first by one factor – VL prior to ART, after that by three factors – VL, HIV reservoir (HIV DNA), CD8 count prior to ART. A predictive model formula was generated using those factors: D = -5.288-0.001 CD8+1.342VL+0.08 HIV DNA. With discriminant function D>0.065, the patient has a high probability of RV after 12 months of ART; with D ≤0.065, the probability of RV is low. This prognosis should be taken into account when choosing an ART regimen ensuring its maximum optimization.