Incidence and predictors of new-onset atrioventricular block requiring pacemaker implantation after sutureless aortic valve replacement

Abstract

OBJECTIVES: In high-risk patients with severe aortic stenosis, aortic valve replacement (AVR) with a sutureless Perceval prosthesis (SU-AVR) can be performed instead of conventional AVR or transcatheter aortic valve implantation. Little data are available regarding postoperative conduction disorders after SU-AVR. We aimed to determine the incidence and predictors of new-onset complete atrioventricular block (NO-AVB) requiring permanent cardiac stimulation following SU-AVR. METHODS: We studied consecutive patients who underwent SU-AVR between 2013 and 2015. Early patients underwent partial aortic decalcification and subannular valve implantation (standard technique), while later patients underwent complete/symmetrical decalcification and intra-Annular valve deployment (modified technique). Predictive baseline and procedural variables and electrocardiographic parameters were identified using a logistic regression model. RESULTS: We included 140 patients (mean age, 78 ± 6.5 years; mean Log EuroSCORE II, 8.9 ± 10%; 28.6% concomitant myocardial revascularization). The most common postoperative conduction disturbances were LBBB (25%), NO-AVB (12.1%) and first-degree atrioventricular block (AVB) (7.9%). The incidence of NO-AVB was 61% lower with the modified versus the standard technique (P = 0.04). NO-AVB predominantly appeared within 24 h post-surgery, occurring >24 h post-surgery in only 2 patients (both with baseline conduction defects). Independent predictors of NO-AVB included baseline left QRS axis deviation (LaQD; P = 0.03), first-degree AVB (P < 0.01) and standard surgical technique (P = 0.02). CONCLUSIONS: NO-AVB is a frequent complication following SU-AVR, and its incidence strongly depends on the surgical technique. Baseline first-degree AVB and LaQD independently predict NO-AVB and should be considered when deciding the duration of postoperative electrocardiographic monitoring.