AIMS: Transcatheter mitral valve repair (TMVR) is a safe and effective therapy for mitral valve (MV) regurgitation (MR) in patients at prohibitive surgical risk. To date, two different devices were used in the majority of cases. We aimed to quantify and compare the acute dynamic changes of MV geometry throughout the cardiac cycle after TMVR using MitraClip and PASCAL device. Furthermore, we aimed to identify predictors of post-procedural mean transmitral gradients in each treatment cohort and to analyse clinical and echocardiographic changes in 1-year follow-up.
METHODS: Peri-procedural transesophageal echocardiography including acquisition of pre- and post-interventional 3-dimensional MV imaging was performed in patients undergoing TMVR for secondary MR. TOMTEC Arena 4D MV-Assessment application was used to quantify the acute MV geometry changes in a saddle-shape MV model. Predictors of mean transmitral pressure gradients were identified in univariable and multivariable linear regression analysis. At follow-up physical examination and transthoracic echocardiography was performed.
RESULTS: A total of n=100 patients undergoing TMVR between 2019 and 2020 using MitraClip (n=50) and PASCAL device (n=50) were considered in this analysis. MR grade was sufficiently reduced ≤I° in PASCAL and MitraClip cohort (88% vs. 84%, p=0.19) at discharge. 4D-analysis of MV geometry revealed a consistent reduction of the anterior-posterior (AP) diameter, the annular area and circumference throughout the cardiac cycle in both device cohorts (p<0.01 for all). Interestingly, while lateral-medial (LM) diameter was left unchanged in PASCAL cohort in cardiac cycle (p>0.05), LM diameter was reduced in end-diastole in MitraClip cohort (p=0.033). Despite comparable preprocedural anatomical MV orifice area (AMVOA), postprocedural AMVOA was larger in PASCAL cohort in mid- (2.8 vs. 2.5 cm², p=0.04) and late diastole (2.7 vs. 2.3 cm², p=0.03). Mean transmitral pressure gradient at discharge was comparable in PASCAL and MitraClip cohort (3.0 vs. 3.3 mmHg, p=0.12). 
Stepwise multivariable linear regression model proofed the reduction of mid-diastolic AMVOA independently predictive for mean transmitral pressure gradients in PASCAL (β=-0.410, p<0.01) and MitraClip cohort (β=-0.318, p=0.02). Additionally, left ventricular (LV) end-systolic diameter (β=-0.448, p<0.01) was independently predictive in PASCAL cohort and in MitraClip cohort, the left atrial volume index (β=0.293, p=0.04) as well as an established marker of MV annular compliance (β=0.392, p<0.01) were independent predictors.
In 1-year follow-up, NYHA class improved in both cohorts (p<0.01). On echocardiography, evidence for reverse LV remodeling was found in PASCAL cohort with reduction of LV end-diastolic diameter (p=0.02).
CONCLUSION: PASCAL device and MitraClip showed comparable acute effects on MV geometry. However, the flexible nitinol scaffold of the PASCAL device might better preserve MV function and postinterventional AMVOA. In follow-up, both treatment cohorts demonstrated clinical benefits while evidence for reverse remodeling was found only in PASCAL-treated cohort.