Hippocampal morphometry is altered in infants with congenital heart disease

Read full text
Authors: Barat Gal-Er, Alexandra F Bonthrone, Emily S. Nichols, Andrew T M Chew, Daniel Cromb, Chiara Casella, Alexia Egloff, Kuberan Pushparajah, J.K. Simpson, Mary A Rutherford, A David Edwards, Joseph V. Hajnal
+4 more Chiara Nosarti, Jonathan O’Muircheartaigh, Emma G. Duerden, Serena J Counsell
Cited by: 2
References: 43

Abstract

Abstract Congenital heart disease (CHD) is associated with impaired early brain development and an increased risk of adverse neurodevelopmental outcomes. Previous studies have reported smaller hippocampal volumes in infants, children and adolescents with CHD. However, it is unclear whether specific subfields are differentially sensitive as hippocampal subfield morphometry has not been assessed in this population. The aims of this study were to test the hypothesis that infants with CHD deviate from typical hippocampal morphometry, using a normative modelling approach with reference data from 217 typically developing infants, and characterize the relationship between hippocampal morphometric measures and both cerebral oxygen delivery and neurodevelopmental outcomes in infants with CHD. Infants with CHD [60 preoperative and 29 postoperative, gestational age at birth median (range) 38.43 (36.71–40.57) weeks, postmenstrual age at scan 39.86 (37.14–45.71) weeks] and typically developing infants underwent brain MRI on a 3T scanner, and T2-weighted and inversion recovery T1-weighted imaging were acquired. Phase-contrast angiography was acquired in 53 infants with CHD before surgery, and cerebral oxygen delivery was calculated. Cognitive and motor abilities were assessed at 22 months (N = 52) using the Bayley Scales of Infant and Toddler Development-Third Edition. Volumes of the whole hippocampus, subiculum, cornu ammonis 1–4, dentate gyrus, and stratum radiatum lacunosum and moleculare were obtained by segmenting T1-weighted MRI data using a U-net model trained on infant data using HippUnfold. Normative curves were generated from the typically developing infants using two models: one for relative hippocampal volumes and one for hippocampal gyrification. Models accounted for the infant’s postmenstrual age at scan, postnatal age at scan and sex. The hippocampal gyrification model also accounted for hippocampal subfield volume. Z-scores representing the degree of positive or negative deviation from the normative mean were generated for infants with CHD. Relative volume z-scores were reduced for the bilateral cornu ammonis 4 and dentate gyrus (median Z-scores −0.45 to −0.82) and were increased for the left subiculum (mean Z-scores: preoperative 0.44; postoperative 0.51) in infants with CHD. Hippocampal gyrification was reduced in the bilateral subiculum, cornu ammonis 1–4 and dentate gyrus both pre- and postoperatively (mean Z-scores, −0.50 to −1.33). There was no significant relationship between preoperative hippocampal morphometric Z-scores and cerebral oxygen delivery or neurodevelopmental outcomes. Our findings suggest that hippocampal morphometry is altered in infants with CHD; however these effects are not uniform across the hippocampus.

Fields of Study & Subjects

Congenital Heart Disease Studies Neonatal and fetal brain pathology Coronary Artery Anomalies

Publication details

Journal / source
Brain Communications
Publisher
Oxford University Press (OUP)
Volume
8
Issue
2
Pages
fcag060–fcag060
Publication date
1 January 2026
Language
EN