Predicting pacemaker dependency after transcatheter aortic valve replacement: Analysis of ECG, clinical, and CT-imaging parameters.

OBJECTIVES

Conduction disturbances necessitating permanent pacemaker (PPM) implantation following transcatheter aortic valve replacement (TAVR) have been observed. However, limited data exist on ECG, clinical, and CT-imaging factors predicting PPM dependency after TAVR. This study aimed to identify predictors of pacemaker dependency in selected patients who required PPM implantation after TAVR with SAPIEN 3 prostheses.

MATERIALS AND

METHODS

This study included consecutive patients who underwent transfemoral TAVR with SAPIEN 3 prostheses at our institution between May 2012 and December 2019. Exclusion criteria were incomplete or non-diagnostic data, valve-in-valve procedures, TAVR in mitral position, previous surgical repair, and pre-implanted PPM. The primary endpoint was PPM dependency, defined as ventricular pacing percentage ≥ 1 % at the first outpatient follow-up after PPM implantation post-TAVR. Regression analysis was performed to compare a limited prediction model for PPM dependency using only selected variables to a full model with all available variables.

RESULTS

Out of 2105 patients who received TAVR, 350 (16.6 %) required pacemaker implantation post-TAVR. After exclusions, 301 patients remained, with 168 (55.8 %) PPM-dependent and 133 (44.2 %) non-dependent patients. Multivariate analysis identified prosthesis oversizing (OR: 1.09, p < 0.001), calcification below the left coronary cusp (LCC) (OR: 1.02, p < 0.001), and right bundle branch block (RBBB) prior to TAVR (OR: 2.20, p = 0.025) as significant predictors. A limited regression model predicted PPM dependency with an AUC of 0.752, significantly outperforming the full model (AUC: 0.660, p = 0.037).

CONCLUSION

RBBB prior to TAVR was the strongest predictor of PPM dependency post-TAVR, followed by prosthesis oversizing and calcification below the LCC. A limited prediction model with these variables demonstrated moderate predictive accuracy.