COVID-19 infection is associated with anti-M alloantibody development in pediatric patients: immunological characteristics and clinical implications for transfusion safety.

BACKGROUND

Anti-M antibodies are commonly observed as naturally occurring antibodies in the MNS blood group system, and their production has been linked to various microbial infections through molecular mimicry. However, the association between COVID-19 infection and anti-M antibody development in pediatric patients remains unexplored.

METHODS

This retrospective matched case-control study included 31 anti-M positive pediatric patients and 31 matched anti-M negative controls with NN phenotype from October 2022 to May 2025. Anti-M isotypes (IgM/IgG) were distinguished using saline and dithiothreitol-preserved plasma methods, followed by titer determination and IgG subclass analysis. Plasma cytokines were quantified by flow cytometry. COVID-19 infection history within 2 years was recorded.

RESULTS

Anti-M positive patients had significantly higher COVID-19 infection rates than controls (54.84% vs. 19.35%, OR = 5.06, P = 0.008); this association remained significant after adjustment for age and sex (adjusted OR = 5.35, 95% CI: 1.67–17.17, P = 0.005). Among those with IgG antibodies, anti-M IgG titers inversely correlated with the interval since COVID-19 infection (r = -0.63, P = 0.008). Reactive IgG (IgG1/IgG3) exhibited significantly higher titers than inactive IgG (median 64 vs. 4, P < 0.001), with all high-titer cases (≥ 64) exclusively in the reactive group (11/31, 35.48%). Anti-M positive patients showed elevated IL-8 (P = 0.006) and IL-12p70 (P = 0.029), and unexpectedly lower bilirubin levels (P < 0.01), suggesting metabolic adaptation rather than hemolysis.

CONCLUSION

COVID-19 infection is significantly associated with anti-M alloantibody development in NN-phenotype children, with over one-third reaching clinically significant titers. These findings suggest that enhanced antibody screening should be considered for pediatric transfusion candidates with recent viral infections, although the single-center design with modest sample size (31 pairs) warrants multi-center validation.