Coronary CT assessment of left atrial remodeling in hypertrophic cardiomyopathy reveals selective lateral wall thickening
M Klis, N Bodagh, V Vigneswaran, A Von Kietzell, A Gharaviri, K Maciunas, S Niederer, R Rajani, S E Williams, M O'neillAbstract
Background
Although histological studies and surgical observations have noted atrial hypertrophy in hypertrophic cardiomyopathy (HCM), true left atrial (LA) wall thickness remains poorly characterised. Atrial remodelling in HCM may reflect not only hypertrophy but also complex interactions between elevated filling pressures, fibrotic deposition, and structural remodelling. Wall thickness is clinically relevant, as it may influence catheter ablation outcomes and correlate with arrhythmia susceptibility.
Methods
Adult patients who underwent clinically indicated coronary CT between 2013 and 2025 were screened for HCM. Forty patients with confirmed HCM were identified and compared to an age-, sex-, and BMI-matched control cohort. Automatic LA segmentation was performed using a U-Net convolutional neural network trained on SCOT-HEART CT angiograms. Blood and wall volumes were calculated directly from the segmented voxels, wall mass was estimated assuming myocardial density of 1.05 g/mL and mean LA wall thickness was calculated using a Laplace-based distance transform, which estimates local wall thickness at each voxel. Regional wall thickness was assessed in anterior, posterior, septal, and lateral segments by dividing wall voxels relative to median x- and y-coordinates of the LA wall.
Results
LA mass was significantly higher in the HCM group compared with control group (17.9±4.48g vs 13.3±6.58g, p<0.05. Because LA volume scales with the cube of radius while mass scales with the square of radius, we compared the ratio between the cube route of LA volume to the square route of LA mass was between the two groups to determine whether the increased LA mass can be explained solely by chamber dilatation. Although the HCM group showed a trend towards a lower ratio, consistent with disproportionate wall thickening relative to size, suggesting LA remodelling reflects both dilatation and hypertrophy. However, the difference did not reach statistical significance (1.19±0.01 vs 1.24±0.13, p=0.06). Linear regression confirmed significantly higher LA mass in HCM (estimate = 4.33 g, p = 0.002), whereas mean LA wall thickness was slightly higher but not statistically significant (estimate = 0.037 mm, p = 0.241). Regional analysis revealed no significant differences in anterior, posterior, or septal walls, but the lateral wall was significantly thicker in HCM group (1.17mm vs 1.09mm, p = 0.016). Age, sex, and BMI were not associated with LA mass or wall thickness.
Conclusion
HCM is associated with selective lateral LA wall thickening rather than uniform atrial hypertrophy, independent of age, sex, or BMI. These findings highlight the heterogeneous nature of atrial remodelling in HCM, which may have implications for arrhythmia risk stratification and procedural planning.