Background:  The MRKAd5/HIV trivalent vaccine failed to reduce HIV incidence and viral load in participants who acquired infection, despite induction of HIV-specific CD8 T cell responses. In this study, we examined how the specificity of vaccine-induced T cell responses may have influenced the post-infection viral load outcome.

Methods:  We first identified CD8 T cell epitopes statistically associated with spontaneous HIV control, using optimally defined epitopes in interferon-g (IFN-g) ELISpot assays in untreated HIV-infected persons, including 74 elite controllers, 74 viremic controllers, and 102 progressors. In these infected individuals, targeting of specific epitopes including B*2705-KK10 (Gag p24), B*57/*5801-TW10 (Gag p24), and A*02-LV10 (Nef) was shown to be associated with reduced viral load; these we termed “good epitopes.” We then performed IFN-g ELISpot assays on post-vaccination, pre-infection samples from 19 STEP trial participants using 76 peptide minipools (consisting of 5 with 15 mers each) spanning Gag, Pol, and Nef. Targeted epitopes within positive minipools from vaccinees were predicted based on HLA class-I types. Targeting of the identified good epitopes after vaccination but prior to infection was then evaluated in relation to early set-point viral loads (sVL) after infection.

Results:  The total number of epitope responses per individual did not correlate with sVL in HIV-infected STEP trial cases. In contrast, we found that individuals responded to at most one good epitope, and those who did respond had lower sVL than those who did not (3.9 vs 4.8 log RNA copies/mL; p = 0.0003). Importantly, we found that among the 9 individuals in the analysis with A*02 but neither B*57 nor B*27 alleles, those who responded to the good A*02-restricted Nef LV10 epitope had lower sVL than those who did not (4.2 vs 5.0 log RNA copies/mL; p = 0.016).

Conclusions:  These results suggest that control of HIV is mediated by the targeting of specific T cell epitopes (not limited to Gag or “protective” alleles such as B*57 and B*27). As a consequence, the design of a successful vaccine immunogen may hinge on the inclusion of good epitopes and the exclusion of others that distract the immune system from targeting regions associated with control of viral replication.