Background
Improvement in cognitive function is described in persons with HIV commencing ART for the first time. We assessed the impact of initiating either dual or triple ART on longitudinal changes in biomarkers of neuronal inflammation and injury, and their associations with cognitive function.
Methods
Biomarker concentrations were assessed at baseline (W0) and after 96 weeks (W96) in individuals randomised to commence darunavir/r and either tenofovir-DF/emtricitabine (triple-ART, n=119) or raltegravir (dual-ART, n=119) in a previously reported cognitive sub-study of NEAT-001/ANRS143. Plasma biomarkers assessed included neurofilament light protein (NfL), glial fibrillary acidic protein (GFAP), sCD14, CXCL10, neopterin and IL-6. Multivariable linear regression models, adjusted for age and sex, examined the association between W0 and W96 changes in biomarker concentration with baseline characteristics (CD4 count, composite cognitive test score (NPZ) across 7 domains, duration of known HIV, and smoking status) and randomised treatment arm (longitudinal analysis only).
Results
In 238 individuals median age was 38 (IQR 31, 46) years, 87% male and 83% of Caucasian ethnicity. Baseline median log10 HIV RNA was 4.73 (IQR 4.23, 5.11) copies/mL and CD4 count 350 (IQR 285, 412) cells/µL. At W96, most individuals had plasma HIV RNA <50 copies/mL (96% triple-ART, 91% dual-ART, p=0.06) and NPZ score increased by 0.15 (triple-ART) vs 0.27 (dual-ART) (p=0.80). At W0, greater concentrations of age-adjusted NfL, GFAP and CXCL10 were statistically significantly associated with lower CD4+ count (all p<0.01) but no other baseline characteristics. At W96 biomarker concentrations decreased (mean change log10 GFAP -0.07 pg/mL, IL-6 -0.15 pg/mL, CXCL10 -0.38 pg/mL, sCD14 -0.03 ng/mL, neopterin -0.28 nmol/L (all p<0.01), NfL -0.02 pg/mL (p=0.08)). A greater decline in neopterin concentration was observed in those randomised to dual-ART compared to triple-ART (mean change log10 -0.30 vs -0.25 nmol/L, respectively. p=0.022; Table). Biomarker changes were statistically significantly associated with CD4 increase (GFAP, CXCL10, sCD14, neopterin, NfL; p<0.002) but not NPZ-score.
Conclusions
Improvements in biomarkers of neuronal health following 2 years of ART were associated predominantly with improvements in CD4 count and partly by randomised drug arm, but not with cognitive function changes.
Table or Figure
![]()