Antibodies to myelin oligodendrocyte glycoprotein in HIV-1 associated neurocognitive disorder: a cross-sectional cohort study

Abstract

Background: Neuroinflammation and demyelination have been suggested as mechanisms causing HIV-1 associated neurocognitive disorder (HAND). This cross-sectional cohort study explores the potential role of antibodies to myelin oligodendrocyte glycoprotein (MOG), a putative autoantigen in multiple sclerosis, in the pathogenesis of HAND.

Methods: IgG antibodies against MOG were measured by ELISA in sera and cerebrospinal fluid (CSF) of 65 HIV-positive patients with HAND (n = 14), cerebral opportunistic infections (HIVOI, n = 25), primary HIV infection (HIVM, n = 5) and asymptomatic patients (HIVasy, n = 21). As control group HIV-negative patients with bacterial or viral CNS infections (OIND, n = 18) and other neurological diseases (OND, n = 22) were included. In a subset of HAND patients MOG antibodies were determined before and during antiviral therapy.

Results: In serum, significantly higher MOG antibody titers were observed in HAND compared to OND patients. In CSF, significantly higher antibody titers were observed in HAND and HIVOI patients compared to HIVasy and OND patients and in OIND compared to OND patients. CSF anti-MOG antibodies showed a high sensitivity and specificity (85.7% and 76.2%) for discriminating patients with active HAND from asymptomatic HIV patients. MOG immunopositive HAND patients performed significantly worse on the HIV dementia scale and showed higher viral load in CSF. In longitudinally studied HAND patients, sustained antibody response was noted despite successful clearance of viral RNA.

Conclusions: Persistence of MOG antibodies despite viral clearance in a high percentage of HAND patients suggests ongoing neuroinflammation, possibly preventing recovery from HAND.

Figure 1

IgG antibody response to MOG in serum and CSF. Serum (A) and CSF (B) IgG antibody response to MOG in patients with HIV associated neurocognitive disorder (HAND), HIV patients with opportunistic infections of the CNS (HIVOI), HIV meningitis (HIVM), asymptomatic HIV infection (HIVasy), other infectious neurological diseases (OIND) and other neurological diseases (OND). Kruskal Wallis test (A: p < 0.01, B: p < 0.001). Dunn's post hoc test yielded a significant difference between HAND and OND patients (p < 0.01, A). The dotted line indicates a cut-off value of 0.6 OD, which is considered as serum immunopositivity for MOG IgG. In CSF samples Dunn's post hoc test yielded a significant difference of HAND and HIVOI patients compared to HIVasy (p < 0.05 and p < 0.01) and OND patients (both p < 0.001) and OIND compared to OND patients (p < 0.01). Data is shown on a logarithmic scale; the horizontal bar represents the median.*, p < 0.05; **, p < 0.01; ***, p < 0.001.

Figure 2

Viral load in CSF and plasma samples. Viral load in CSF (A) and plasma samples (B) (virus copies per milliliter) stratified for positive serum anti-MOG IgG antibodies. Data is shown on a logarithmic scale, samples below the detection limit are shown as 1 virus copy/ml. The horizontal bar represents the median. Repeated measures data was modeled by generalized estimation equations. ***, p < 0.001.

Figure 3

IgG antibody response to MOG in longitudinally studied patients. Serum anti-MOG IgG antibody responses in 7 patients with HIV associated neurocognitive disorder (HAND) documented follow-up visits and available CSF viral load data. The dotted line indicates a cut-off value of 0.6 OD, which is considered as serum immunopositivity for MOG IgG. Groupwise repeated measures data was modeled by generalized estimation equations and showed a highly significant effect (p < 0.001) of time of sampling on the concentration of anti-MOG antibodies.

Figure 4

Case report. Case report: increasing leukoencephalopathy under successful cART. MR scans and OD values for MOG IgG of one patient with HAND are shown.