IMG06
Differential Hippocampal Subfield Injury Patterns and Memory Performance in Pediatric MS and Relapsing Mog-Associated Disorders
Objectives: To investigate memory deficit patterns and their relationship with hippocampal subfield changes in youth with pediatric-onset MS and MOGad.
Methods: We enrolled 53 healthy controls (HC), 48 youth with MS, and 11 with relapsing MOGad. At baseline, all participants completed 3T MRI and the Penn Computerized Neurocognitive Battery including tests of verbal, visuospatial, and facial memory. Participants with MS and MOGad were scanned longitudinally (MS: mean 3.29 scans over 1.73y; MOGad: 2.78, 1.75y). We compared changes of average magnetization transfer ratio (MTR), fractional anisotropy (FA), and mean diffusivity (MD) in each hippocampal subfield over time using multivariable generalized linear models accounting for age, disease duration, and a subject-specific random intercept. We compared accuracy for memory tests between groups using general linear models accounting for age, sex, and response reaction time. Finally, accounting for the same factors, we tested the association between baseline memory scores and baseline MRI metrics between groups for each subfield.
Results: Relative to HC or MOGad, verbal memory scores were lower in MS (-1.5, p 0.0074; and -2 units, p.0.0405, respectively). MOGad had lower facial memory scores vs HC (-3.1 units, p 0.065). Relative to HC, neither MS or MOGad showed significant volume differences for any subfields. However, lower FA and higher MD were found in the dentate gyrus/cornu ammonis (CA)4, CA2/3, subiculum, and pre-subiculum of youth with MS, and in the para- and pre-subiculum of youth with MOGad. None of the groups showed significant changes of MRI parameters over time. There were no significant associations between memory scores and subfield volume or microstructure.
Conclusions: Both pediatric MS and MOGad patients have memory impairments as well as structural integrity disruption in the hippocampus- although the patterns of cognitive and structural integrity in these two demyelinating disorders are distinct and do not correlate with hippocampal volumes or disruption to hippocampal subfield structural integrity. Future studies, including connectome network analysis, may help confirm and elucidate these findings.
