HLA and T-Cell Receptor Investigations in Idiopathic and Paraneoplastic Opsoclonus-Myoclonus in Children.

BACKGROUND AND OBJECTIVES: Opsoclonus-myoclonus-ataxia syndrome (OMAS) is a rare pediatric autoimmune neurologic disorder, with 40%-50% of cases occurring as a paraneoplastic syndrome in the setting of neuroblastoma or other neuroblastic tumors. The strong clinical similarities between idiopathic and paraneoplastic OMAS have led to the hypothesis that neuroblastoma may initially be present in idiopathic cases but spontaneously regress before OMAS onset. Both B- and T-cell mechanisms appear to contribute to OMAS autoimmunity, although their precise roles remain unknown. The aim of this study was to investigate similarities and differences between idiopathic and paraneoplastic OMAS by analyzing HLA profiles and T-cell receptor (TCR) repertoires in a Spanish pediatric cohort. METHODS: Children with OMAS were enrolled in a nationwide prospective study (2013-2024). Neural autoantibody testing was performed in all patients. HLA genotyping was performed using next-generation sequencing (GenDx/Illumina) and compared with a reference Spanish population (n = 4,335). The Benjamini-Hochberg false discovery rate method was used to control for multiple comparisons across HLA antigens. TCR repertoires were analyzed in peripheral blood samples of 12 patients using the Adaptive Biotechnologies immunoSEQ assay. The bioinformatics tool Grouping of Lymphocyte Interactions by Paratope Hotspots (GLIPH2) was used to identify shared motifs within CDR3β-TCR sequences. RESULTS: Among 25 children with OMAS (median age 22 months [interquartile range (IQR): 15-38]), 12 (48%) had neuroblastoma and 3 (12%) had neuronal antibodies (one with anti-Hu, one with anti-myelin oligodendrocyte glycoprotein, one against unknown neuronal surface antigens). Some HLA class II antigens (DRB1*08, DRB1*10, and DQB1*04) were consistently overrepresented in patients with OMAS compared with Spanish controls (all p-adjusted <0.05). Distinct HLA antigens were associated with idiopathic OMAS (HLA-DRB1*08, HLA-DQB1*04), while paraneoplastic OMAS was linked to HLA-DRB1*10 (all p-adjusted <0.05), suggesting subgroup-specific genetic risks. TCR analysis revealed that patients with paraneoplastic OMAS showed greater TCR repertoire similarity, whereas shared sequences were rare in idiopathic OMAS. DISCUSSION: Despite similar clinical presentations, idiopathic and paraneoplastic OMASs appear to have distinct immunopathogenic mechanisms. In paraneoplastic OMAS, enrichment of specific HLA antigens and convergent TCR profiles supports antigen-driven T cell-mediated autoimmunity.