Type O blood group associates with higher anti‐JC polyomavirus antibody levels

Abstract Background Patients with multiple sclerosis (MS) and high anti‐JC polyomavirus (JCPyV) antibodies in blood have an increased risk for the development of progressive multifocal leukoencephalopathy (PML) when treated for MS. To test the hypothesis that type O blood group associates with anti‐JCPyV antibody levels and the risk of developing PML, we characterized ABO blood group antigen on blood samples of 62 patients with PML, and 64 MS controls without PML. Methods Monocentric retrospective cohort study. Anti‐JCPyV antibody levels in arbitrary units (AU) were determined in sera using an ELISA‐based method, and blood group specific antibodies using standardised test erythrocytes. Results Anti‐JCPyV antibody levels were higher in individuals with blood group O compared with all other blood groups (O: median AU: 129; not O: median AU: 53; p = .005). This association was not observed for the closely related BK virus. Of the 62 patients with PML, 29 (47%, 95% confidence interval (CI) 35%–59%) were of blood group O, which showed a nonsignificant trend to differ from the expected distribution in the German population (41%), and the MS controls studied (36%, 95% CI 25%–48%). Conclusion The ABO blood group O antigen associates with higher anti‐JCPyV antibody levels and may impact the risk of the later development of PML. The overrepresentation of blood group O in cases with PML was in line with a previous publication. Larger studies are warranted to assess a potential value of host genetic markers, such as the ABO status, for PML risk prediction during immunotherapy.

for the treatment of MS (Warnke et al., 2015), or efalizumab for psoriasis (Schwab et al., 2012), an observation that also hinders the development of several biologicals for immune mediated conditions. Detection of anti-JCPyV antibodies and the level of the anti-JCPyV antibody response (so-called serum index values) (Plavina et al., 2014;Warnke, Ramanujam, et al., 2013) are accepted risk factors for the later development of PML. The reasons why higher antibody levels to JCPyV associate with higher risk of developing PML are not well understood.
Host genetic factors may influence the anti-JCPyV antibody levels, and thus may be the underlying cause for such link. This has been shown for several human leukocyte antigen (HLA) class II variants presenting antigens to CD4+ T cells important in the defense against infections, but studies thus far were also underpowered to display associations outside the HLA region (Sundqvist et al., 2014).
PML is characterized by a lytic infection of oligodendrocytes and astrocytes and is often localized in the subcortical area between the gray and white matter. In this area, there are hemodynamic factors such as the reduction of vessel calibre that might be contributing factors facilitating the evasion of JCPyV. Early studies to detect anti-JCPyV antibodies were based on the fact that JCPyV can aggregate type O erythrocytes (hemagglutination inhibition assay). Since JCPyV can be present on the surface of B-lymphocytes, this may induce a clotting of lymphocytes and type O erythrocytes, which could associate with a higher likelihood of viral transmission to the brain in the small arteries in the subcortical area between the gray and the white matter in patients with type O erythrocytes. As such it has been hypothesized that patients with type O erythrocytes may be at higher risk of the later development of PML (Khoury et al., 2013). In this monocentric cohort study, we assessed if type O blood group was associated with higher levels of anti-JCPyV antibodies. Age, sex, and the underlying condition that predisposed for PML were registered with the blood samples as available from the treating physicians.

Stored
Individuals with a positive blood/CSF antibody index toward measles, rubella and zoster (so-called "MRZ reaction") and without the suspicion of PML were classified as non-PML MS patients (Jarius et al., 2017) and served as control group.
ABO blood types of all patients were determined via "ID-DiaCell ABO" (Product ID 50520; DiaMed GmbH) reverse grouping meaning determining the blood group via detecting the blood group specific antibodies in the patient's sera (anti-A, anti-B) using standard-ized test erythrocytes (A1, B, O) and patient sera. Sera were tested in an enzyme-linked immunosorbent assay using a JCPyV-VP1 protein fused to glutathione S-transferase (GST) as antigen, and anti-JCPyV antibody levels and anti-BKPyV antibody levels in arbitrary units (AU) were determined as previously published (Warnke, Pawlita, et al., 2013;Warnke et al., 2017;Warnke et al., 2014

Standard protocol approvals, registrations, and patient consents
The study was approved by the local ethics committee (No. 5683R), waiving the requirement for written informed consent for the retrospective analysis of the stored, anonymized samples at the Institute for Virology, Duesseldorf, for the purpose of this study.

RESULTS
62 patients with PML and 64 non-PML MS patients were studied. PML patients were older and non-PML MS patients had a higher female to male ratio (Table 1)

DISCUSSION
Our study supports the view that, besides HLA associations (Sundqvist et al., 2014), non-HLA associations may exist that influence the risk of  (Jarius et al., 2017).