n. INK-128 Almost all individuals were found to have HLA-associated polymorphisms in HIV-1 Gag early in the treatment interruption with a median of four polymorphisms per person. In the eight individuals with plasma available at ATI week 49, four were found to have accumulated additional Gag escape mutations. No significant differences were seen in the proportion of Gag polymorphisms between initial virologic suppressors and non-suppressors at either the first Initial Virologic Suppression and Immune Preservation was not Sustained Of the 10 participants with initial virologic suppression and a measured pVL at ATI week 49 off of ART, only 3 subjects continued to have a pVL,3.0 log10 copies/ml. Two of the individuals with sustained virologic control had protective HLA alleles: One participant was found to have HLA B27 and B57, while another had B27. For participants with initial virologic suppression, the median CD4+ cell decline between ATI weeks 16 and 49 was 82 cells/mm3 as ” compared to a decline of 99 cells/mm3 for initial non-suppressors. Viral Suppression after Therapeutic Vaccination PID 1 2 3 4 5 6 7 8 9 10 11 Study Arm Vaccine Vaccine Vaccine Vaccine Vaccine Vaccine Vaccine Vaccine Vaccine Placebo Placebo HLA group Protective Protective Protective Neutral Neutral Neutral Neutral Neutral Unfavorable Neutral Neutral Pre-ARV RNA copies/ml 3.3 ATI Set Point RNA copies/ml 2.9 2.4 ATI week 49 RNA copies/ml 2.7 4.8 2.6 3.5 1.7 3.9 3.6 3.7 4.1 4.8 5.1 3.7 2.7 2.4 1.7 4.0 3.8 4.8 2.9 2.6 2.99 2.8 5.7 2.6 2.1 1.7 PID, patient identification number. doi:10.1371/journal.pone.0034134.t002 available ATI time point or at ATI week 49. This finding suggests that the lack of sustained pVL suppression in the latter group could not be explained by the accumulation of HLA-associated escape mutations in Gag. The number of predicted amino acid mismatches was calculated between the vaccine gag sequence and the earliest available patient-derived HIV-1 sequence after ATI. These early sequences represent viral populations least likely to have been shaped ” by significant vaccine-driven immune responses. There was no significant difference in the number of Gag amino acid mismatches between the vaccine and patient HIV-1 sequences among initial virologic suppressors and non-suppressors who received the vaccine. The change in the number of vaccine-to-patient Gag amino acid mismatches between the first ATI time point and ATI week 49 was used as a potential reflection of vaccine-induced viral evolution. However, we detected little overall change in the number 
of mismatched amino acids between the two time points and no significant differences between initial virologic suppressors and non-suppressors. Immunologic Factors Associated with Initial Virologic Suppression There was no significant difference between initial virologic suppressors and non-suppressors in the CD4+ T-cell count at study entry. In the initial A5197 analysis, an inverse association was seen between the ATI set point viral load and the number of HIV-1 Gag-specific CD4+ IFN-c-producing CD4+ T cells at study weeks 8 and 38. 4 Viral Suppression after Therapeutic Vaccination There were no significant differences between initial virologic suppressors and non-suppressors in the number of HIV-1 Gagspecific IFN-c-producing CD4+ T-cells at week 8 or 38, or Gag-specific IFN-c-producing CD8+ T-cells. Similarly, no significant differences were detected between initial virologic suppressors and non-suppressors in the
