Neurol Neuroimmunol Neuroinflamm. 2026 Jan;13(1):e200449. doi: 10.1212/NXI.0000000000200449. Epub 2025 Nov 17.
ABSTRACT
BACKGROUND AND OBJECTIVES: Spinal cord leptomeningeal enhancement (LME) can be observed in children with myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD) and with seronegative myelitis. We investigated whether the presence of spinal cord LME in MOGAD and seronegative myelitis is associated with distinct clinical, CSF, and MRI findings.
METHODS: Study participants were identified among the 490 children and adolescents recruited to the Canadian Pediatric Demyelinating Disease study following an incident attack of CNS demyelination. Inclusion criteria for this study were: (1) evidence of spinal cord lesions on MRI, (2) available postgadolinium MRI sequences, and (3) available MOG and aquaporin-4 (AQP4) antibody results. None of the AQP4 antibody-positive participants met our inclusion criteria and only 1 participant with multiple sclerosis exhibited LME. We therefore focused the study on children with MOGAD and seronegative myelitis and compared the clinical, CSF, and MRI features between participants with and without LME.
RESULTS: Our cohort included 33 participants with MOGAD (median age 5.9 years, 55% women) and 45 with seronegative myelitis (median age 11.9 years, 33% women). Spinal cord LME was detected in 20/33 (61%) participants with MOGAD and 14/45 (31%) with seronegative myelitis. Among children with MOGAD, those with LME were more likely than those without LME to have longitudinally extensive myelitis ([LETM], 19/20 vs 8/13, p = 0.024); H-sign (15/20 vs 5/13, p = 0.036), tumefactive cord lesions (10/20 vs 1/13, p = 0.021); complete cross-sectional involvement (16/20 vs 5/13, p = 0.026); nodular lesional enhancement (7/20 vs 0/13, p = 0.026); and more spinal cord lesions (p = 0.036). LME in MOGAD was not associated with greater CSF protein content or cell count nor predicted relapse rate or clinical recovery. Children with seronegative myelitis and LME were more likely than those without LME to have tumefactive lesions (6/14 vs 4/31, p = 0.048) and complete cross-section involvement (11/14 vs 13/31, p = 0.028) but did not differ in terms of H-sign, LETM, lesional enhancement, or number of lesions.
DISCUSSION: The presence of spinal cord LME is associated with more extensive spinal cord abnormalities on MRI in children with MOGAD and to a lesser extent in those with seronegative myelitis. The biological underpinnings of this finding and its clinical implications should be assessed in further studies.
PMID:41252657 | DOI:10.1212/NXI.0000000000200449
