Background:  It is unclear whether changes in immunologic markers that occur early after initiation of salvage ART will predict subsequent virologic response. We hypothesized that changes in immune activation could be used to predict those patients who will ultimately suppress viral replication following initiation of salvage therapy.

Methods:  A single-center prospective cohort trial was initiated in 34 patients about to initiate salvage therapy. Immunophenotyping data were collected at baseline, and at 2, 4, 8, and 24 weeks following initiation of a salvage antiretroviral regimen. Data were analyzed using both univariate and multivariate proportional hazards models of time to virologic suppression, defined as viral load <50 copies/mL at one visit.

Results:  All subjects were male, 84% white, 87% men who have sex with men (MSM). At baseline the median CD4⁺ count was 191 cells/mm³ and the median viral load was 4.3 log₁₀ copies/mL. The median duration of prior ART was 11 years with a median of 10 previous antiretroviral drugs. As part of their study salvage regimen, 81% of subjects received protease inhibitors (PI) and 59% received non-nucleoside reverse transcriptase inhibitors (NNRTI). Consistent with prior studies, history of an AIDS-defining illness, increased baseline viral load, and baseline CD4⁺ count <200 cells/mm³ were associated with a decreased likelihood of achieving virologic suppression (p ≤0.01 for each). Univariate proportional hazards models of time to virologic suppression with time-dependent covariates showed that increased CD38 and CD95 expression were significantly associated with reduced likelihood of achieving virologic suppression (RR = 0.66 [0.46 to 0.95], p = 0.02 and RR = 0.78 [0.63 to 0.97], p = 0.02, respectively). In a multivariate model controlling for baseline CD4⁺ <200 cells/mm³, increases in CD38 and CD95 expression as early as week 2 were predictive of reduced likelihood of achieving viral suppression (RR = 0.47 [0.27 to 0.81], p <0.01 and RR = 0.65 [0.48 to 0.89], p <0.01, respectively).

Conclusions:  Changes in CD38 and CD95 expression occur within 2 weeks after initiation of salvage ART. Increased expression of these activation markers predicts subsequent failure to achieve virologic suppression. Thus, when compared to measures of CD4⁺ T-cell count and viral load at much later time points, the prognostic value of changes in CD38 and CD95 expression may provide a distinct clinical advantage by allowing earlier assessment of the response to changes in ART.