Baseline characteristics are presented in Table 1. His bundle pacing was attempted in 233 patients (mean age 74.6 ± 10.1 years, 48% female) of which 37% had pre-existing AF and 39% had chronic kidney disease. The ethnic mix of the patients was consistent with the population demographics of Singapore. Mean left ventricular ejection fraction (LVEF) was 54 ± 12% and 71% had normal LVEF equal to or exceeding 55%. The mean QRS duration (QRSd) was 110 ± 27 ms and 32% of patients had broad QRS complexes of more than 120 ms with right bundle branch block (RBBB) in 21%. The most common pacing indication was atrioventricular block (AVB) in 55.8% of patients and 83.7% received a dual-chamber pacemaker.
His Bundle Pacing Parameters at Implantation
Acutely, HBP was successful in 81.1% (Table 1 and Figure 1). Mean procedural and fluoroscopic times were 105.6 ± 36.8 and 13.9 ± 9.4 min, respectively. At implantation, the parameters achieved were: HBP threshold 1.3 ± 0.7 V at a pulse width of 1.0 ± 0.2 ms, impedance 566 ± 135 ohms, and sensed R-wave amplitude of 5.0 ± 3.9 mV. About 68.6% had an R wave of greater than 3 mV (Table 2). The mean programmed HBP output was 3.3 ± 0.4 V at a pulse width of 1.0 ± 0.2 ms, and mean paced QRSd was 111 ± 22 ms. Selective His bundle capture was achieved in 73% of patients and non-selective His bundle capture in 93.1%.
Patient flow diagram. HBP, His bundle pacing; LBBP, left bundle branch pacing; LV, left ventricular; RV, right ventricular septal; RVA, right ventricular apex; SND, sinus node dysfunction.
Of the 44 cases with unsuccessful HBP, 22 had RV septal capture only but with good capture thresholds. Therefore, the 3830 lead was left in the septum. Of the remaining 22 cases, the RVA pacing was performed in 11 patients, left bundle branch area pacing in 10 patients and a quadripolar LV led to perform BVP in 1 patient. The reasons for use of an alternative pacing location were due to high HBP thresholds in 12 patients; no mappable His bundle in 6 patients, inability to correct the bundle branch block in 2 patients, and unstable HBP lead positions in 2 patients.
Acute Complications and Loss of his Bundle Capture
Overall, complications occurred in 6 out of 233 (2.6%) patients in whom HBP was attempted. These included right atrial lead dislodgement requiring lead repositioning in 2 (0.9%) patients, pneumothoraxes in 2 (0.9%) patients, high RV septal capture threshold requiring lead repositioning in 1 (0.4%) patient, and poor wound healing requiring revision in 1 (0.4%) patient.
His Bundle Pacing Parameters at Last Follow-up and Late Complications
Mean follow-up was 259 ± 213 days during which loss of HBP occurred in 12.4% (29 of 233) patients (Figure 1). In 21 patients, only RV septal capture was demonstrated. There was loss of or high ventricular capture thresholds in eight patients, of whom four patients underwent lead revision with the lead repositioned at the RV apex. The remaining four patients had sinus node dysfunction (SND) and pacing mode was programmed from DDD (pacing in both atrium and ventricle, sensing in both atrium and ventricle, dual mode inhibition and triggering) to AAIR (pacing in atrium, sensing in atrium, inhibition mode with rate responsiveness).
Of the remaining 160 patients with continued HBP capture, the parameters were HBP threshold of 1.2 ± 0.7 V at a pulse width of 0.9 ± 0.2 ms, impedance of 444 ± 82 ohms, R wave of 5.1 ± 3.2 mV, output of 2.8 ± 0.6 V at a pulse width of 0.9 ± 0.2 ms. About 17.5% of patients had relatively high HBP thresholds of between 2 and 3.5 V (Table 2). Threshold increased by 1 V or more in 11.3% of patients and R-wave amplitude decreased by 50% or more in 8.6%. R wave remained above 3 mV in 70.2% of patients.
Clinical Predictors of Acute Success
Table 3 shows the regression analysis of clinical predictors of acute HBP success, with age, gender, LVEF, pacing indications, the presence of a broad QRS at baseline, and the number of procedures as variables. Widened QRSd independently predicted a lower likelihood of acute HBP success (adjusted odds ratio of 0.39, 95% CI 0.18–0.88, P = 0.02).
Learning Curves in his Bundle Pacing
Figure 2 shows regression analysis to investigate the relationship between HBP procedural volume (per operator) and procedural, fluoroscopic times, and HBP thresholds at last follow-up. Procedural and fluoroscopic times decreased and plateaued after 30–40 cases per operator. Better follow-up HBP thresholds were achieved after 20 cases.
Europace. 2022;24(4):606-613. © 2022 Oxford University Press
Copyright 2007 European Heart Rhythm Association of the European Society of Cardiology (ESC). Published by Oxford University Press. All rights reserved.