Cerebrospinal fluid (CSF) biomarkers of iron status are associated with CSF viral load, antiretroviral therapy, and demographic factors in HIV-infected adults

BACKGROUND: HIV-associated neurocognitive disorder (HAND) remains common, despite antiretroviral therapy (ART). HIV dysregulates iron metabolism, but cerebrospinal fluid (CSF) levels of iron and iron-transport proteins in HIV-infected (HIV+) persons are largely unknown. The objectives of this study were to characterize CSF iron-related biomarkers in HIV+ adults and explore their relationships to known predictors of HAND. METHODS: We quantified total iron, transferrin and heavy-chain (H)-ferritin by immunoassay in CSF sampled by lumbar puncture in 403 HIV+ participants in a multi-center, observational study and evaluated biomarker associations with demographic and HIV-related correlates of HAND [e.g., age, sex, self-reported race/ethnicity, ART, and detectable plasma virus and CSF viral load (VL)] by multivariable regression. In a subset (N = 110) with existing CSF: serum albumin (Q(Alb)) measurements, Q(Alb) and comorbidity severity were also included as covariates to account for variability in the blood–CSF-barrier. RESULTS: Among 403 individuals (median age 43 years, 19% women, 56% non-Whites, median nadir CD4+ T cell count 180 cells/µL, 46% with undetectable plasma virus), men had 25% higher CSF transferrin (median 18.1 vs. 14.5 µg/mL), and 71% higher H-ferritin (median 2.9 vs. 1.7 ng/mL) than women (both p-values ≤0.01). CSF iron was 41% higher in self-reported Hispanics and 27% higher in (non-Hispanic) Whites than in (non-Hispanic) Blacks (median 5.2 and 4.7 µg/dL in Hispanics and Whites, respectively, vs. 3.7 µg/dL in Blacks, both p ≤ 0.01); these findings persisted after adjustment for age, sex, and HIV-specific factors. Median H-ferritin was 25% higher (p < 0.05), and transferrin 14% higher (p = 0.06), in Whites than Blacks. Transferrin and H-ferritin were 33 and 50% higher, respectively, in older (age > 50 years) than in younger persons (age ≤ 35 years; both p < 0.01), but these findings lost statistical significance in subset analyses that adjusted for Q(Alb) and comorbidity. After these additional adjustments, associations were observed for CSF iron and transferrin with race/ethnicity as well as CSF VL, for transferrin with sex and ART, and for H-ferritin with plasma virus detectability and significant comorbidity (all p < 0.05). CONCLUSIONS: CSF iron biomarkers are associated with demographic factors, ART, and CSF VL in HIV+ adults. Future studies should investigate a role for CNS iron dysregulation, to which an altered blood-CSF barrier may contribute, in HAND. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1186/s12987-017-0058-1) contains supplementary material, which is available to authorized users.