Background:  We analyzed whether pre-treatment interruption proliferative capacity and cell surface markers predicted CD4 count change during treatment interruption in chronically HIV-infected subjects.

Methods:  HIV-infected subjects (n = 27) with various viral load outcomes on treatment with combination ART (cART) who underwent a treatment interruption lasting more than 2 months were examined in this retrospective study. Peripheral blood mononuclear cells (PBMC) from a time point (designated as the baseline time point) within 6 months before the treatment interruption were tested using the CFSE dilution assay to detect proliferation to a Gag p55 peptide pool, CMV lysate, staphylococcus enterotoxin B, and anti-CD3. Percentage and median fluorescent intensity (MFI) of CD28 and CD57 expression on CD4⁺ and CD8⁺ T cells were measured.

Results:  The CD4 count change from baseline to CD4 nadir during treatment interruption was a median (interquartile range [IQR]) of –100cells/mm³ (–25, –154). CD4⁺ and CD8⁺ T cells proliferation to CMV lysate, and per cent CD4⁺CD57⁺ T cells correlated negatively with CD4 decline during treatment interruption (r = –0.575, r = –0.397, and r = –0.493; p <0.05 for all analyses; Spearman correlation).  Upon dividing the group dichotomously based on the median change in CD4 count into good and poor CD4 responders, good responders had a median (range) CD4 change of –26 cells/mm³ (–17, –83) and poor responders a median CD4 change of –154 cells/mm³ (–117, –313). The good responders had less CD4⁺ and CD8⁺ proliferative potential to CMV lysate and a lower percentage and MFI of CD4⁺CD57⁺ T cells versus poor responders (median 1.19 vs 8.79%, 0.54 vs 6.86%, 2.05 vs 4.43%, and 56.74 vs 76.21 MFI, p <0.05 for all analyses; Mann-Whitney test).

Conclusions:  In treated HIV-infected subjects in the chronic phase of infection having a spectrum of pre-treatment interruption viral load before undergoing treatment interruption, the lack of proliferation to CMV lysate and lower expression of CD57 on CD4⁺ T cells at baseline were associated with a smaller drop in CD4 count during treatment interruption. Pre-treatment interruption immune parameters may have predictive value for CD4 count decline during treatment interruption.