Host single nucleotide polymorphisms and biomarkers of neuronal damage and inflammation in people living with HIV.

Objectives: Blood-brain barrier impairment is frequent in people living with human immunodeficiency virus (PLWHIV), affecting the penetration of target cells and antiretrovirals into the central nervous system, through transporters (e.g. ABCB1), leading to neuroinflammation. This study aimed to identify variants of genes encoding transporters able to predict neuroinflammation biomarker levels.

Methods: Cerebrospinal fluid (CSF) and plasma samples were obtained from PLWHIV. The CSF biomarkers were quantified by commercial assays. Genetic variants were evaluated through real-time polymerase chain reaction (PCR).

Results: A total of 107 PLWHIV (163 samples) were included in the study: 79% were male, median age was 48.5 years, CD4% was 25%, and HIV-associated neurocognitive disorder (HAND) was observed in 17.8%. The ABCB1 2677G>T genetic variant showed a different allelic distribution according to the clinical group (P = 0.026). In linear regression analyses, HIV-related central nervous system disorders, ABCG2 1194+928CC genotype, log viral load, CSF-to-serum albumin ratio, β-1,42 levels, and CSF proteins were retained in the final model as factors independently associated with CSF neopterin levels; CSF proteins and integrase inhibitor use were associated with CSF tau level in the multivariate model. Phospho-tau regression analysis reported the ABCB1 2677GT/TT genotype and CSF proteins as predictors in the final model; sex, protease inhibitors, neopterin, and ABCB1 2677 GT/ TT genotype were predictors in the multivariate regression for β-1,42.

Conclusions: For the first time, pharmacogenetic and clinical features were found to be predictors of neuro-inflammation biomarkers.