Abstract Body

ACTG A5241 (OPTIONS Study) randomized individuals experiencing treatment failure to omit or add nucleoside reverse transcriptase inhibitors (NRTIs) to a regimen that had a cumulative activity of 2 or more active antiretroviral agents. There were more deaths and clinical events observed in those randomized to add NRTIs. We hypothesized that clinical events were associated with markers of inflammation and mitochondrial dysfunction.

Cohort study of 174 participants enrolled in OPTIONS (N=413) selected randomly and enriched to include those with clinical events (death, AIDS defining opportunistic infections and non-AIDS clinical events). Protein levels relating to inflammation (IL-6, TNFr1, TNFr2, sCD14, CRP, Insulin) or mitochondrial dysfunction (NADH dehydrogenase [C1], FGF21 and GDF15) were measured by Luminex and ELISA, respectively at baseline, weeks 24 and 48 and evaluated for their association with the composite endpoint of clinical events. At baseline sampling, all participants were taking a failing regimen of NRTIs plus protease inhibitors. The statistical analysis included univariate parametric (t-tests) and non-parametric tests (Wilcoxon test) with selected variables analyzed using linear and generalized linear models.

174 participants were evaluated with a median age of 47 years, 40% women; 43% Black, 20% Hispanic/Latino, 36% white. There were 58 participants with clinical events and 116 participants without clinical events of whom 35% vs. 36% were randomized to omit NRTIs, respectively. At baseline, sCD14 (555,263 vs 448,584 pg/mL, P=0.03); CD4 count (148 vs. 209 cells/mm3, P=0.03); CD4:CD8 ratio (0.16 vs 0.22, P=0.02) and VACS Index (46 vs. 33, P=0.02) were significantly different in those who subsequently experienced a clinical event. At baseline, there were no significant differences in the two groups NADH dehydrogenase activity, FGF-21, GDF-15, IL-6, TNFr1, TNFr2, insulin or HIV RNA levels. Censoring for those with clinical events before weeks 24 or 48, FGF-21, sCD14, CD4, CD4:CD8 ratio and VACS index were significantly different at weeks 24 and 48 (Table). Analyses were similar when adjusted for randomization to omit or add NRTIs. Only sCD14 remained significant on multivariate analyses at baseline, week 24 or week 48 (Odds ratio, 1.0).

Severity of illness, biomarkers of inflammation and mitochondrial dysfunction were associated with clinical events. Randomization to omit or add NRTIs was not associated with clinical events. sCD14 identifies a group at higher risk of developing clinical events.