Background:  Long-term spontaneous control of HIV-1 replication in HIV controllers (HIC) has been associated with protective HLA alleles and efficient T cell responses. We have observed such long-term viral control after treatment interruption (TI) in patients treated early at the time of primary HIV infection (PHI). We explored virologic and immunological mechanisms of such control.

Methods:  Ultra-sensitive HIV RNA and cell-associated HIV DNA assays were performed using real time PCR. HIV-specific CD8⁺ T cell responses were evaluated by IFNg ELISpot with optimal peptides and CD8⁺ T cell mediated HIV suppression assay. All patients were HLA typed. Mann-Whitney test was used.

Results:  We have included 10 HIV-1 patients who had controlled HIV RNA at <400copies/mL for a median of 6 years (range, 4 to 10) after HAART discontinuation. These patients had been treated within 10 weeks after the symptoms of PHI and for a median of 3 years (1 to 7.5). Before treatment, the median HIV RNA was at 5.2 log copies/mL (3 to 7.3) and CD4 was at 489/mm³ (371 to 955). At the time of interruption, the median CD4 increased at 910 (354 to 1636). At inclusion in this study, CD4 count had remained stable at 836 (388 to 1598) and HIV RNA was <40 copies/mL (<20 to 289); the median residual replication evaluated by ultra-sensitive assay in 5 individuals was 4 copies/mL (1 to 55); HIV DNA levels were low (1.8 log copies/10⁶ peripheral blood mononuclear cells (PBMC) (0.5 to 2.3), close to levels found in HIC (1.5 log). Interestingly, we observed that HIV control was associated with an HIV DNA decrease over the last 5 years, among 5/8 patients, even though ART had been interrupted. ELISpot revealed low frequencies of IFNg-secreting HIV-specific CD8⁺ T cells (618 ± 958 SFC/10⁶ PBMC). Accordingly, CD8⁺ T cells from 9/10 patients showed weak capacities to suppress HIV-1 superinfection of autologous CD4⁺ T cells (0.27 log p24 decrease CD4 vs CD4:CD8 [0 to 1.24]). One patient who consistently showed strong CD8⁺ suppressive capacity (2.56 log p24 decrease) has a protective HLA B*14 allele; no patients have the HLA B*57 allele and 1/10 has the HLA B*27 allele. Surprisingly, 5/10 patients carry the HLA B*35 allele.

Conclusions:  These patients may have achieved a significant reduction of the viral reservoir level through early treatment; after TI, they maintained strong capacities that control HIV reservoirs. In this context, the presence of HLA B*35, which can negatively affect the T cell response, did not seem to influence the durable viral control after discontinuing HAART.