J Magn Reson Imaging. 2026 Aug 13. doi: 10.1002/jmri.70499. Online ahead of print.
ABSTRACT
BACKGROUND: Brain development is altered in neonates with congenital heart disease (CHD), with well-characterised reductions in regional brain volumes and enlargement of CSF spaces. However, perioperative morphological changes remain incompletely understood.
PURPOSE: To identify brain regions showing spatial patterns of coordinated expansion and contraction that differ between neonates with CHD after cardiac intervention and controls, assess perioperative changes and effects of perioperative risk factors, using Structural Covariance Component (SCC) analysis.
STUDY TYPE: Prospective.
POPULATION: Forty-one neonates with CHD (male = 20) who underwent cardiac surgery or catheterization and 359 healthy neonates (male = 185).
FIELD STRENGTH AND SEQUENCE: 3T T2-weighted turbo-spin-echo sequence.
ASSESSMENT: Images were motion-corrected and reconstructed using a neonatal algorithm. Jacobian determinants calculated from non-linear image registration to a template were input into an Independent Component Analysis to identify SCCs (N = 40). SCC weightings were extracted, reflecting the degree to which the covariance pattern is expressed per neonate.
STATISTICAL TESTS: Postoperative SCC weightings were compared to those of controls using a general linear model or robust regression. Perioperative change was assessed using a linear mixed effects model, and associations between perioperative differences and age at surgery, cardiopulmonary bypass duration, and postoperative pediatric intensive care unit stay were assessed using partial Spearman's rank correlation. Analyses were adjusted for covariates and corrected for multiple comparisons using False Discovery Rate, significant at p < 0.05.
RESULTS: Sixteen of the 40 SCCs showed significant differences between neonates with CHD after surgery and controls, including white matter, cortical- and deep gray matter, brainstem, and CSF (f2 = 0.02-0.04), with seven also showing significant perioperative change (f2 < 0.01-0.27). An additional nine SCCs only showed significant perioperative change (f2 < 0.01-0.14). Perioperative risk factors were not associated with perioperative change (p > 0.463).
DATA CONCLUSION: This work highlights region-specific postoperative alterations and perioperative changes in brain morphology of neonates with CHD.
EVIDENCE LEVEL: 1.
TECHNICAL EFFICACY: Stage 3.
PMID:42596043 | DOI:10.1002/jmri.70499