Imaging in patients with severe mitral annular calcification: insights from a multicentre experience using...

To investigate valve sizing and the haemodynamic relevance of the predicted left ventricular outflow tract (LVOT) in patients with mitral annular calcification (MAC) undergoing transatrial transcatheter valve implantation (THV).

In total, 21 patients undergoing transatrial THV, multiplanar reconstruction (MPR), maximum intensity projection (MIP), and cubic spline interpolation (CSI) were compared for MA sizing during diastole. In addition, predicted neo-LVOT areas were measured in 18 patients and correlated with the post-procedural haemodynamic dimensions. The procedure was successful in all patients (100%). Concomitant aortic valve replacement was performed in eight patients (43%) (AVR group). Sizing using MPR and MIP yielded comparable results in terms of area, perimeter, and diameter, whereas the dimensions obtained with CSI were systematically smaller. The simulated mean systolic neo-LVOT area was 133.4 ± 64.2 mm2 with an anticipated relative LVOT area reduction (neo-LVOT area/LVOT area × 100) of 59.3 ± 14.7%. The systolic relative LVOT area reduction, but not the absolute neo-LVOT area, was found to predict the peak (r = 0.69; P = 0.002) and mean (r = 0.65; P = 0.004) post-operative aortic gradient in the overall population as well as separately in the AVR (peak: r = 0.91; P = 0.002/mean: r = 0.85; P = 0.002) and no-AVR (peak: r = 0.89; P = 0.003/mean: r = 0.72; P = 0.008) groups.

In patients with severe MAC undergoing transatrial transcatheter valve implantation, MPR, and MIP yielded comparable annular dimensions, while values obtained with CSI tended to be systematically smaller. Mitral annular area and the average annular diameter appear to be reliable parameters for valve selection. Simulated relative LVOT reduction was found to predict the post-procedural aortic gradients.