Intrathecal production of anti-Epstein-Barr virus viral capsid antigen IgG is associated with neurocognition and tau proteins in people with HIV

Abstract

Objective: HIV and Epstein-Barr virus (EBV) co-infection has been linked to increased immune activation and larger HIV reservoir. We assessed whether anti-EBV humoral responses are associated with increased cerebrospinal fluid (CSF) inflammation and with neurocognitive impairment (NCI) in people with HIV (PWH).

Design: Cross-sectional analysis in 123 EBV-seropositive PWH either on antiretroviral therapy ( n = 70) or not.

Methods: Serum and CSF anti-EBV viral capsid antigen immunoglobulin G (anti-EVI) and CSF EBV DNA were measured by commercial immunoassay and RT-PCR. Seventy-eight participants without neurological confounding factors underwent neurocognitive assessment (Global Deficit Score, GDS). CSF total tau and 181-phosphorylated-tau (ptau) were measured by immunoassays together with biomarkers of blood-brain barrier (BBB) integrity, immune activation, astrocytosis, and intrathecal synthesis. Logistic and linear regressions and moderation analysis were used to investigate the relationships between CSF anti-EVI, GDS, and biomarkers.

Results: Twenty-one (17.1%) and 22 participants (17.9%) had detectable CSF anti-EVI (10.5-416.0 U/ml) and CSF EBV DNA (25-971 copies/ml). After adjusting for BBB integrity, age, and clinical factors, the presence of CSF anti-EVI was only associated with serum levels of anti-EVI, and not with CSF EBV DNA. CSF anti-EVI, tau and ptau showed reciprocal interactions affecting their associations with GDS. After adjusting for demographics and clinical parameters, higher CSF anti-EVI levels were associated with worse GDS (aβ 0.45, P < 0.001), and CSF levels of tau and ptau had a moderation effect on the strength of this association (models' P < 0.001).

Conclusion: Humoral immune responses against EBV within the central nervous system may contribute to NCI in PWH through mechanisms that involve neuronal injury.

Fig. 1.. Comparison of the proportions of participants ranking below the normative reference thresholds of neurocognitive tests between those with and without detectable cerebrospinal fluid anti-EBV viral capsid antigen IgG.

Neurocognitive tests: (1) delayed recall of the Rey–Osterrieth complex figure; (2) free and cued selective reminding test; (3) disyllabic words serial repetition test; (4) Corsi block-tapping test; (5) prose memory test; (6) digit span forward; (7) digit span backward; (8) digit symbol test; (9) trail making test A; (10) Stroop Color and Word Test interferences; (11) Stroop Color and Word Test interferences errors; (12) trail making test B; (13) verbal fluency test; (14) frontal assessment battery; (15) copy of the Rey–Osterrieth complex figure; (16) Groove pegboard test for the dominant hand; (17) Groove pegboard for the nondominant hand. * Mann–Whitney U test P value for comparison less than 0.05 after Yates’s correction.

Fig. 2.. Correlations between cerebrospinal fluid anti-Epstein–Barr virus viral capsid antigen IgG, global deficit score, and cerebrospinal fluid total tau and phosphorylated tau.

Panels a and c–f show Spearman’s rank correlations between the variables of interest by detectability of anti-EBV VCA IgG in CSF (blue dots=undetectable IgG; orange dots=detectable CSF IgG). Panel b shows the beta coefficients of univariable linear regressions among the variables of interest and for the interaction factors between CSF anti-EBV VCA IgG and CSF total tau or ptau for the outcome global deficit score (GDS).

Fig. 3.. Moderator effects of cerebrospinal fluid levels of total tau and 181-phosphorylated tau proteins upon the association between global deficit score and cerebrospinal fluid anti-Epstein–Barr virus viral capsid antigen IgG levels (n=78).

The graphs show the regression lines with the corresponding coefficients and significance [aβ (95% CI), P value] for the association between CSF anti-EBV VCA IgG levels and GDS when CSF total tau levels (panel a) or when CSF ptau levels (panel b) are at −1 SD (blue), at the mean (orange), or at +1 SD (red) of the distribution in the study population (moderation analysis performed by PROCESS v4.2). aβ, adjusted beta coefficient; anti-EBV VCA IgG, immunoglobulin G anti-EBV viral capsid antigen; CSF, cerebrospinal fluid; GDS, global deficit score; SD, standard deviation.