Results
Baseline characteristics are shown in Table 1. The median age of the patients was 79 (74 – 83) years, 33% of patients (n = 25) were female, and the STS SCORE was 11.0 (5.3 – 16.0)%. The mechanism of MR was secondary in 55 (71%) patients and primary in 22 (29%) patients. During a median of 221 (99 – 447) days of follow-up, 13% of patients in the entire cohort had CV events (event group). The event group, compared to the no-event group, had lower estimated glomerular filtration rate (eGFR) (25.2 (IQR: 19.9 – 31.5) mL/min/1.73m2 vs. 41.7 (IQR: 29.8 – 56.6) mL/min/1.73m2, p = 0.006), higher N-terminal pro-brain natriuretic peptide (4042 (IQR: 2794 – 7114) pg/mL vs. 1760 (IQR; 870 – 4955) pg/mL, p = 0.039), and lower TAPSE (12 (IQR: 7 – 14) mm vs. 16 (IQR: 13 – 22) mm, p = 0.009).
Procedural and clinical outcomes
One clip was implanted in 44 (57%) patients and two clips were implanted in the remaining patients (Table 2). All patients experienced technical success according to the Mitral Valve Academic Research Consortium (MVARC) criteria.(16) During a median of 221 (99 – 447) days of follow-up after TMVr, 5 patients died due to CV events and 8 patients had HF hospitalization. The event group, compared to the no-event group, had a higher prevalence of moderate residual MR at discharge (p = 0.012) and longer length of hospital stay post-TMVr (p = 0.003) (Table 2). On univariate Cox regression analysis (Table 3), CV events were associated with eGFR (HR: 0.960, 95%CI; 0.926 – 0.995, p = 0.027), TAPSE (HR: 0.874, 95%CI; 0.789 – 0.968, p = 0.010), and significant residual MR (HR: 11.652, 95%CI; 3.257 – 41.691, p <0.001). On multivariate Cox regression analysis, CV events were independently associated with TAPSE (HR: 0.788, 95%CI; 0.788 – 0.987, p = 0.029) and significant residual MR (HR: 9.373, 95%CI; 2.581 – 34.033, p = 0.001). TAPSE did not significantly change mean 4days after the procedure (15 (12 – 22) mm to 16 (12 – 21) mm, p = 0.708). Figure 2 shows the ROC analysis of TAPSE to predict CV events. TAPSE had the largest AUC compared to RV-FAC (0.793 vs. 0.681) and RV-S’ (0.793 vs. 0.626). TAPSE <11 mm was the best cut-off value for predicting CV events (Sensitivity 0.839; Specificity 0.556). Kaplan-Meier analysis showed that no RV functional parameters, except for TAPSE, had a significant prognostic power for predicting CV events (Figure 3).