Background: Mitral regurgitation is the second most common valvular heart defect in industrialized countries with an estimated annual incidence of 2-3%. It is commonly treated by percutaneous edge-to-edge or surgical mitral valve repair. However, the geometric effects on the modified valves are less studied so far. The purpose of this study was to quantify and qualitatively visualize the changes in mitral valvular geometry over these two cohorts of patients.

Methods: A total of 33 patients with severe mitral regurgitation were included in the study conducted between March to December 2021. 23 of the patients underwent clip intervention and 10 of the patients underwent surgical mitral valve repair. The virtual valves were modelled from all phases (total 1631 time-steps) of 3D-transesophaegal echocardiography scans using an own software following which, the annulus and valve leaflet geometry were quantified both pre- and postoperatively in 6 different phases. This was then followed by the calculation of a novel visualization that allows the 3D models to be unrolled into a flat representation with color coding that indicates prolapsed areas (Fig.1) and allows for pre- and postoperative comparison. Subsequently, a user study was performed with 8 experts to evaluate the visualizations together with the expert’s confidence in drawing conclusions from the linked pre- and postoperative representations.

Results: The main findings were that percutaneous mitral valve reconstruction decreased anteroposterior diameter up to 3.9% (early systole, from 43.5 to 41.8 mm, p< 0.019). Annulus height showed a significant increase in late systole of 14.3% (7.8 to 9.1 mm, p< 0.013) and mitral valve orifice area significantly reduced at each time point with a maximum of 62.5% (late systole, from 0.8 to 0.3 cm², p< 0.001). In the surgical mitral valve repair group, there was a significant reduction at each time step of the cardiac cycle for every parameter studied, except the non-planarity angle. Most interestingly, the mitral valve orifice area substantially reduced by a maximum of 71.5% (mid-systole, from 0.7 to 0.2 cm², p<0.005) and the maximum reduction in annulus area was 55.5% (from 18.2 to 8.1 cm², p<0.001) in early diastole, indicating the significant downsizing effect of the implanted prosthetic ring. The results of the questionnaire revealed that all participants were able to recognize the conducted procedure and the patient-individual geometric effects could be studied (e.g. Fig. 1: smaller valve without prolapse after ring implantation). All correctly recognized the segment(s) of the prolapse and the clipped segment(s) after intervention with a high degree of certainty, trusting most a combined view of 3D and flatten representation (Fig.1). 

Conclusion: Significant changes in mitral valve geometry, related to the mitral annulus as well as the valve leaflets, were observed in the two cohorts after percutaneous edge-to-edge and especially after surgical mitral valve repair. Visualization on a patient-by-patient level allowed intensive examination of the respective pathology and therapeutic modifications. The digital, objective, and unified representation of the geometrical changes could be used, if linked to outcome, to better understand individual effects of treatment in the spirit of precision medicine.