Atrophy of Ipsilesional Hippocampal Subfields Vary Over First Year After Ischemic Stroke

Background The structural integrity of hippocampal subfields has been investigated in many neurological disorders and was shown to be better associated with cognitive performance than whole hippocampus. In stroke, hippocampal atrophy is linked to cognitive impairment, but it is unknown whether the hippocampal subfields atrophy differently. Purpose To evaluate longitudinal hippocampal subfield atrophy in first year poststroke, in comparison with atrophy in healthy individuals. Study Type Cohort. Subjects A total of 92 ischemic stroke (age: 67 ± 12 years, 63 men) and 39 healthy participants (age: 69 ± 7 years, 24 men). Field Strength/Sequence A3 T/T1‐MPRAGE, T2‐SPACE, and T2‐FLAIR. Assessment FreeSurfer (6.0) was used to delineate 12 hippocampal subfields. Whole hippocampal volume was computed as sum of subfield volumes excluding hippocampal fissure volume. Separate assessments were completed for contralesional and ipsilesional hippocampi. Statistical Tests A mixed‐effect regression model was used to compare subfield volumes cross‐sectionally between healthy and stroke groups and longitudinally between 3‐month and 12‐month timepoints. False discovery rate at 0.05 significance level was used to correct for multiple comparisons. Also, a receiver operating characteristic (ROC) curve analysis was performed to assess differentiation between healthy and stroke participants based on subfield volumes. Results There were no volume differences between groups at 3 months, but there was a significant difference ( P = 0.027) in whole hippocampal volume reduction over time between control and stroke ipsilesionally. Thus, the ipsilesional whole hippocampal volume in stroke became significantly smaller ( P = 0.035) at 12 months. The hippocampal tail was the highest single‐region contributor (22.7%) to ipsilesional hippocampal atrophy (1.19%) over 9 months. The cornu ammonis areas (CA1) subfield volume reduction was minimal in controls and stroke contralesionally but significant ipsilesionally ( P = 0.007). CA1 volume significantly outperformed whole hippocampal volume ( P < 0.01) in discriminating between stroke participants and healthy controls in ROC curve analysis. Data Conclusion Greater stroke‐induced effects were observed in the ipsilesional hippocampus anteriorly in CA1 and posteriorly in the hippocampal tail. Atrophy of CA1 and hippocampal tail may provide a better link to cognitive impairment than whole hippocampal atrophy. Level of Evidence 2 Technical Efficacy Stage 3