P309Aortic valve implantation-induced bundle branch block as a framework towards a more uniform electrocardiographic definition of left bundle branch block

INTRODUCTION The electrocardiographic (ECG) pattern of true left bundle branch block (LBBB) has not been fully clarified and various definitions of LBBB exist. New-onset LBBB after transcatheter (TAVR) or surgical (SAVR) aortic valve replacement implies a proximal pathogenesis of LBBB and thus may provide a reference to characterize and define true LBBB. PURPOSE This study compares ECG characteristics in aortic valve implantation-induced LBBB (AVI-LBBB) to a non-procedural-induced LBBB control group (co-LBBB) in order to set a more homogenous definition for true LBBB. METHODS The study enrolled all patients with new-onset TAVR- and SAVR-induced LBBB between 2013 and 2019. AVI-LBBB was defined as new-onset persistent LBBB occurring within 24h after TAVR or SAVR. Patients were matched for age, sex, ischemic heart disease and left ventricular systolic function to randomly selected co-LBBB patients in a 1:2 ratio. For inclusion in both groups, a non-strict LBBB definition was used (QRSD ≥120ms, QS or rS in lead V1, absence of Q wave in leads V5-6). ECG characteristics were digitally analysed by the MUSE algorithm and confirmed by two experts. All ECG recordings were classified according to 4 different LBBB definitions: MADIT, European Society of Cardiology (ESC), Strauss and American Heart Association (AHA). RESULTS 59 patients with AVI-LBBB (34 TAVR, 25 SAVR, median age 82 years, 42% male) were compared to 118 matched co-LBBB patients. All patients with AVI-LBBB presented with QRS notching/slurring in the lateral leads, whereas this was present in only 85% of the co-LBBB group (p = 0.001). QRS duration (148ms vs 145ms, p = 0.074) and R wave peak time (58ms vs 62ms, p = 0.065) were not significantly different among both groups. AVI-LBBB was characterized by a more rightward QRS axis (-15° vs -30°, p = 0.013). When comparing AVI-LBBB to LBBB controls with QRS notching/slurring, a comparable QRS axis was observed. Almost all AVI-LBBB patients met the MADIT (98%), ESC (100%) and Strauss (95%) definition. Only 18% of patients met the AHA definition, because of the low combined presence of QRS notching/slurring in all 4 lateral leads (54%) and because only 27% of patients had an R wave peak time >60ms in both leads V5-6. In the co-LBBB group, adherence to the different definitions was significantly lower compared to the AVI-LBBB group: MADIT 86% (p = 0.007), ESC 85% (p = 0.001), Strauss 68% (p < 0.001) and AHA 7% (p = 0.035). Lower presence of lateral notching/slurring and more patients with smaller QRS duration (QRS duration ≥130ms, 86% vs 98%, p = 0.007) in the co-LBBB group explain these results. CONCLUSIONS Discordance exists between various definitions in scoring AVI-LBBB. Our data show that presence of QRS notching/slurring in the lateral leads is a crucial feature of proximal LBBB, rather than QRS duration and R wave peak time. The AVI-LBBB population provides a framework towards a more uniform definition and criteria for assessing true, proximal LBBB.