59\u2005Pace and ablate for poorly controlled AF – how does it affect heart failure metrics?

Abstract

Background Atrioventricular nodal ablation (AVNA) is indicated in drug-refractory atrial fibrillation (AF)[1], and has been shown to improve cardiac resynchronization therapy (CRT) efficacy in this cohort[2]. There is concern over the potentially detrimental effect of long-term right ventricular (RV) apical pacing on heart failure metrics[3], but there is no consensus as to which pacing approach should be undertaken initially, especially when left ventricular ejection fraction (LVEF) is normal[4–6]. This observational study aims to inform as to the effect of AVNA on heart failure metrics and CRT efficacy. Methods Patients were identified retrospectively from records of patients undergoing AVNA for drug-refractory AF in a single centre over a 3-year period (January 2016 – December 2018). Baseline and post-procedure echocardiographic measurements of LVEF and NT-proBNP levels were obtained from electronic records. Those with a CRT device in-situ had pre- and post-AVNA biventricular pacing percentages recorded electronically, successful threshold defined as ≥95%. For initial analysis, patients were categorised as having impaired (<55%) or normal (≥55%) LVEF. Data represents mean ± SEM and were analysed using paired student t-test. Results 106 patients (mean age 72.1 years, 55% female) underwent AVNA during this time period. 18/106 (17.0%) had a normal LVEF at baseline. CRT was performed in 46/106 (43.4%), and 94.1% of this group had an impaired LVEF pre-AVNA compared with 6.9% with a normal LVEF pre-AVNA. Post-AVNA, LVEF significantly increased in patients with already impaired LVEF (31.4±3.2% to 41.2±3.8%, P<0.05) compared to those with normal LVEF (53.0±9.3% to 60.0±3.4%). NT-proBNP levels did not significantly change in either group. LVEF also significantly increased in patients post-AVNA who had a CRT device in-situ (30.1±3.1% to 49.8±6.5%, p<0.05) compared with those treated with RV pacing only (49.9±6.9% to 49.8±6.5%). In addition, NT-proBNP levels significantly decreased in patients who underwent CRT (3175±431.9 pg/ml to 1482.1±176.8 pg/ml, p<0.05) compared to those treated with RV pacing only (4135.1±911.6 pg/ml to 2722.5±671.5 pg/ml). The number of patients receiving successful biventricular pacing rose from 11/46 (23.9%) to 38/44 (86.4% (2 patients lost to follow-up)) post-AVNA.Abstract 59 Figure 1 Effects of AVNA on LVEF (A) and NT-proBNP (B) depending on prior normal vs impaired LVEF. Patients underwent echocardiogram and blood monitoring pre- and post-AVNA. LVEF, and NT-proBNP measured. Patients were categorized as normal (≥55%) or impaired (<55%) based on initial LVEF measurement. *P<0.05 compared to pre-AVNA. Data represents mean ± SEM. Effects of AVNA on LVEF (A) and NT-proBNP (B) depending on prior normal vs impaired LVEF. Patients underwent echocardiogram and blood monitoring pre- and post-AVNA. LVEF, and NT-proBNP measured. Patients were categorized as normal (≥55%) or impaired (<55%) based on initial LVEF measurement. *P<0.05 compared to pre-AVNA. Data represents mean ± SEM.Abstract 59 Figure 2 Effects of AVNA on LVEF (A) and NT-proBNP (B) depending on single/dual pacing vs CRT. Patients underwent echocardiogram and blood monitoring pre- and post- treatment. LVEF and NT-proBNP measured. *P<0.05 compared to pre-treatment. Data represents mean ± SEM. Effects of AVNA on LVEF (A) and NT-proBNP (B) depending on single/dual pacing vs CRT. Patients underwent echocardiogram and blood monitoring pre- and post- treatment. LVEF and NT-proBNP measured. *P<0.05 compared to pre-treatment. Data represents mean ± SEM. Conclusion AVNA was associated with a significant improvement in heart failure metrics in patients with an already impaired LVEF, and in those with a CRT device in-situ, we suggest this is through improved rate control and biventricular pacing percentage. A pace and ablate strategy was not associated with worsening in heart failure metrics in patients with a normal LVEF at baseline, suggesting up-front CRT in this group is not justifiable. Larger randomized controlled trials would be helpful to confirm these findings.