Background: HIV-infected patients who fail to reconstitute their CD4 T-cell counts during suppressive antiretroviral therapy (ART) have increased risk of both AIDS-related and non-AIDS-related morbidity and mortality. Improved understanding of immunological nonresponse (INR) is necessary to enable earlier clinical intervention.
Methods: In a cohort of 112 HIV-infected patients starting ART, we performed a serial analysis of 32 plasma-soluble markers, assessed by multiplex cytokine and enzyme immunoassay. Samples were drawn pre-ART and during the first 3 years of treatment, with a final observation time of 8.4 years (interquartile range, 7.0-10.7 years) on ART. Long-term INR (LT-INR) was defined as failure to reach a CD4 T-cell count >350 cells per microliter. Marker stability was evaluated by parallel analysis of samples from ART-naïve and HIV-seronegative controls.
Results: Baseline CD4 T-cell counts predicted subsequent LT-INR (n = 15) [odds ratio, 1.10 (95% confidence interval: 1.01 to 1.19) pr. 10 cells/μL reduction in CD4 count, P = 0.030] in the cohort as a whole, but not in patients with baseline CD4 counts <200 cells per microliter (n = 78). LT-INR was best characterized by elevated plasma levels of the CC chemokine macrophage inflammatory protein 1β (MIP-1β), both at baseline (pre-ART) and during ART. In patients with baseline CD4 counts <200 cells per microliter, baseline MIP-1β predicted LT-INR [odds ratio 1.23 (95% confidence interval: 1.02 to 1.47) per 10 pg/mL increase in MIP-1β, P = 0.029].
Conclusions: Elevated pre-ART levels of MIP-1β identified LT-INR patients who started ART at CD4 counts <200. INR was characterized by persistently high MIP-1β during suppressive ART. Thus, MIP-1β may be of use for early identification of LT-INR.