Introduction: Left bundle branch block (LBBB) is associated with an increased risk of ventricular arrhythmia. Biventricular pacing (BVP) improves symptoms, systolic left ventricular function and mortality in heart failure with LBBB, but can be pro-arrhythmic. His bundle pacing (HBP) can overcome LBBB to produce more synchronous ventricular activation than BVP, but it is not known how ventricular repolarisation heterogeneity is affected, which is important in arrhythmogenesis. We set to out to measure the dispersion of repolarisation and activation recovery-interval (ARI, a surrogate for action potential duration) in narrow QRS, LBBB, BVP and HBP.
Methods: Patients were recruited into two groups. In the first group, patients with heart failure and LBBB scheduled to undergo clinically indicated BVP implant procedures were recruited. They underwent temporary HBP to attempt reverse LBBB during the BVP procedure. If HBP shortened activation time by ≥10 ms, patients were included (HBP cardiac resynchronisation therapy [CRT]). In the second group patients with normal, narrow QRS were recruited. Non-invasive electrocardiographic imaging was used to measure the following parameters in narrow QRS, LBBB, BVP and HBP: left ventricular activation time, left ventricular repolarisation time dispersion and left ventricular ARI dispersion.
Results: A total of 21 patients in whom HBP shortened LV activation time by >10 ms and an equal number of individuals with narrow intrinsic QRS were recruited. LV repolarisation dispersion was reduced by HBP-CRT (-42.0 ms, 95% confidence interval (CI) -52.3 to -31.7; p<0.001) but not by BVP (+11.9 ms, -6.24 to 30.1; p=0.182). The mean within-patient change in LV repolarisation dispersion from BVP to HBP-CRT was -56.5 ms (-70.5 to 42.5; p<0.001). LV repolarisation dispersion with HBP-CRT was not different from individuals with narrow intrinsic QRS (difference: 2.75 ms, -16.2 to 21.7; p=0.981). The magnitude of reduction in LV repolarisation dispersion with HBP-CRT from intrinsic LBBB appeared to be similar to the magnitude of LV activation time shortening (-45.9 ms, -59.3 to -32.4). However, LV activation recovery interval dispersion was also reduced by HBP-CRT (-56.5 ms, -70.5 to -42.5 ms; p<0.001). Repolarisation mapping demonstrated normalisation of repolarisation pattern by HBP-CRT.
Conclusions: HBP-CRT can normalise repolarisation dispersion, producing more physiological repolarisation compared with BVP, which does not resolve the repolarisation abnormality of LBBB. HBP-CRT improves repolarisation through both activation resynchronisation and modulation of action-potential duration. If these acute results translate to longer-term outcomes, HBP-CRT may reduce the risk of ventricular arrhythmias in heart failure with LBBB to a greater extent than BVP. ❑