HLA-B class I allele associations with neurological complications in pediatric SARS-CoV-2 infection: a retrospective observational study

Most severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infections in children are mild, yet neurological complications can occur. Host immune variability is partly determined by human leukocyte antigen (HLA) polymorphisms and may influence susceptibility. This study investigates whether specific HLA-B alleles are associated with neurological involvement in pediatric coronavirus disease 2019 (COVID-19). This retrospective study spanned over one year, including children with confirmed SARS-CoV-2 infection. Patients were classified into two groups: patients with COVID-19 with neurological disease (neuro-COVID-19), and those with COVID-19 without neurological disease (non-neuro COVID-19). A control group of 120 healthy children was included to represent baseline allele distribution. HLA-B class I allele typing was performed using polymerase chain reaction with sequence-specific oligonucleotide probes (PCR-SSOP). HLA-B49 represented the most common allele among children suffering from neurological COVID-19 (10.4%). None of the alleles reached statistical significance when compared in patients with and without neurological disease. A possible, but not statistically significant, lower frequency of the HLA-B52 allele was observed in cases with neurological manifestations ( p = 0.088). However, in exploratory unadjusted analyses, HLA-B53 showed higher odds of neurological involvement when compared with children without neurological disease and healthy controls (OR = 6.29; 95% CI: 1.23–32.13; nominal p = 0.027), while HLA-B45 demonstrated a positive trend (OR = 3.75; 95% CI: 0.87–16.22; nominal p = 0.077). These allele-level findings were not corrected for multiple comparisons and should therefore be interpreted as exploratory and hypothesis-generating. This study suggests a possible immunogenetic contribution to neurological complications in pediatric COVID-19, with a nominal unadjusted signal for HLA-B53. Given the small sample size, multiple-allele testing, and the absence of adjusted modelling, this finding should be considered exploratory and hypothesis-generating. Validation in larger multicenter cohorts with correction for multiple comparisons and adjustment for relevant confounders is required.