HIV-1 Provirus DNA load and genotype in patients with long-term successful HAART

I Pellegrin1, I Garrigue1, P Merel1, ML Chaix3, C Rouzioux3, MH Schrive1, JL Pellegrin2, and H Fleury1. Department of Virology1 and Department of Internal Medicine and Infectious Diseases2, Bordeaux University Hospital, 3, Virology cohorte PRIMO, Necker Hospital, Paris (France).

Background: During potent and prolonged therapy, HIV-1 DNA can remain detectable in PBMC and lymphoid tissues. Drug-resistant viruses selected by non-suppressive pre-HAART regimens can enter and persist in this reservoir, as well as archival wild-type HIV-1, both forms potentially released. We characterized the residual circulating reservoir in patients on prolonged successful HAART. Material and Methods: Proviral DNA extracted from patients PBMC was quantified by real time PCR and analyzed for reverse transcriptase (RT) and protease (PR) genotypes according to the IAS list 2002. Results: HIV-1 proviral DNA levels and RT and PR mutations were assessed in PBMC from 139 patients on suppressive HAART (<50 copies/ml) since 40 [30; 48] months. Median number of previous antiretroviral therapy (ART) was 2 [1; 3.3]; 38 patients (27%) started ART with dual therapy. At initiation of first ART, CD4+ count and plasma viral load were 245 [113; 410] cells/mm3 and 4.8 [4.1; 5.3] log10 copies/ml, respectively. Considering the entire follow-up, nadir of CD4 count was 192 [98; 308] cells/mm3. 104 of 139 patients (74.8%) exhibited an undetectable plasma load since their first (ART), of whom 52 (50%) were still on their first line suppressive regimen; 30 of 139 patients (21.6%) experienced viral rebound on-treatment before achieving suppression of viremia for prolonged periods of time and 5 of 139 (3.6%) were not documented. Detectable provirus loads were measured in 81.6% patients with a median of 2.7 [2.3; 3] log10 copies/106 PBMC. PBMC genotyping showed 34% (45/133) RT with 1-5 NRTI or NNRTI-related mutations when 57.7% (75/130) PR had 1-8 PI-related mutations (without considering L63P, found alone in 17% PR), including 6 PR with 1 to 3 primary-mutations. Importantly, mutated DNA sequences were obtained in 26.9% (28/52) RT and 45.2% (47/52) PR from patients with suppression of viral replication since their first ART, including RT (15/28) and PR (22/47) from patients on first-line suppressive regimen. PBMC genotype was associated with number of previous lines of treatment (p=0.018), start of treatment with mono or dual therapy (p=0.001) and plasma load undetectability since the first treatment (p=0.02). Peripheral provirus load level was associated only with plasma load undetectability since the first treatment (p=0.029). Conclusion: We quantitated HIV-1 proviral DNA in most patients on prolonged successful HAART and characterized viral population present in PBMC: viruses harbored drug resistance mutations clearly attributable to selection under suppressive HAART, even in patients on first line therapy.