Left bundle branch pacing with and without anodal capture: impact on ventricular activation pattern and acute haemodynamics

Abstract

Aims

Left bundle branch pacing (LBBP) can deliver physiological left ventricular activation, but typically at the cost of delayed right ventricular (RV) activation. Right ventricular activation can be advanced through anodal capture, but there is uncertainty regarding the mechanism by which this is achieved, and it is not known whether this produces haemodynamic benefit.

Methods and results

We recruited patients with LBBP leads in whom anodal capture eliminated the terminal R-wave in lead V1. Ventricular activation pattern, timing, and high-precision acute haemodynamic response were studied during LBBP with and without anodal capture. We recruited 21 patients with a mean age of 67 years, of whom 14 were males. We measured electrocardiogram timings and haemodynamics in all patients, and in 16, we also performed non-invasive mapping. Ventricular epicardial propagation maps demonstrated that RV septal myocardial capture, rather than right bundle capture, was the mechanism for earlier RV activation. With anodal capture, QRS duration and total ventricular activation times were shorter (116 ± 12 vs. 129 ± 14 ms, P < 0.01 and 83 ± 18 vs. 90 ± 15 ms, P = 0.01). This required higher outputs (3.6 ± 1.9 vs. 0.6 ± 0.2 V, P < 0.01) but without additional haemodynamic benefit (mean difference −0.2 ± 3.8 mmHg compared with pacing without anodal capture, P = 0.2).

Conclusion

Left bundle branch pacing with anodal capture advances RV activation by stimulating the RV septal myocardium. However, this requires higher outputs and does not improve acute haemodynamics. Aiming for anodal capture may therefore not be necessary.

Graphical Abstract

Graphical Abstract.

What’s new?

• Left bundle branch pacing (LBBP) results in physiological activation of the left ventricle but is associated with delayed right ventricular (RV) activation.

• Delayed RV activation can be overcome by combining LBBP with anodal capture.

• In this study, we conducted a within patient comparison of electrical and haemodynamic response to non-selective LBBP with and without anodal capture.

• Our multi-electrode mapping shows that non-selective LBBP with anodal capture achieves earlier RV activation by stimulating the RV myocardium.

• Our results show that non-selective LBBP with anodal capture leads to shorter QRS durations and shorter ventricular activation times. However, this requires higher pacing outputs and does not offer any haemodynamic benefit.

• Our findings suggest that aiming for anodal capture when programming LBBP devices may not be necessary.

Introduction

Left bundle branch pacing (LBBP) may transform pacing therapy for both bradycardia and cardiac resynchronization.^1–6 Like His bundle pacing (HBP), it attains physiological left ventricular activation but offers several technical advantages:

• Pacing thresholds are lower and more stable.^3,7,8

• Operators may learn more quickly.⁹

• More distal conduction abnormalities can be treated.^7,8,10

His bundle pacing is still considered the most physiological ventricular pacing modality.¹¹ This is because HBP can produce physiological or near physiological biventricular activation.¹² During LBBP, right ventricular (RV) activation is typically delayed compared with HBP or normal intrinsic activation, and this is manifest on the 12-lead electrocardiogram (ECG) by a terminal R-wave (R-prime) in lead V1. The impact of this non-physiological RV activation on cardiac function and long-term outcomes is unknown. The observation of adverse outcomes in patients with right bundle branch block led to concerns that non-physiological RV activation during LBBP could adversely impact cardiac function.^13,14

There are two proposed methods to overcome this delayed RV activation.

First, atrio-ventricular (AV) delay can be lengthened to allow intrinsic right bundle conduction to fuse with left bundle area capture.⁴ However, such fusion is not always achievable (e.g. in patients who have a right bundle branch block or complete heart block) or desirable¹⁵ (e.g. if it requires a very long AV delay that impairs ventricular filling).

Second, the pacing output in a bipolar configuration can be increased so that the anode stimulates the RV. There is uncertainty regarding the mechanism through which anodal capture results in early RV activation. Some investigators have suggested that this occurs as a result of capture of the right bundle branch (bi-bundle capture when combined with LBBP), whereas others propose that it occurs through RV myocardial capture.^4,16–18 Crucially, it is not known whether programming the pacing output to obtain anodal capture offers any additional benefit with respect to haemodynamic function.

We investigated the mechanism of RV stimulation during non-selective LBBP with anodal capture and compared the electrical and acute haemodynamic effects of non-selective LBBP with and without anodal capture.

Methods

Study design and patient recruitment

We enrolled consecutive patients with an LBBP lead in whom it was possible to eliminate the R-prime in lead V1 with any pacing output during bipolar pacing. This involved bipolar pacing with sufficient outputs to capture the anode. To establish that the paced complexes resulted from a fusion of anodal and left bundle capture, we ensured that the ECG morphology was different from the pure anodal capture pattern of pacing between the ring and the pacemaker.

Left bundle branch pacing procedure

The trans-septal approach described by Huang et al.¹⁹ was used to implant the LBBP lead. We used simple criteria to confirm left bundle branch capture. The main criteria were change in QRS morphology with changing outputs or programmed stimulation. Other criteria may be helpful,²⁰ such as the presence of a short and constant R-wave peak time in lead V5/V6. Stimulus to R-wave peak time was measured with the 12-lead ECG from the stimulation artefact to the peak of the R-wave in lead V5 or V6.²¹ It was encouraging that the left ventricular activation pattern was consistent with conduction system activation in patients where multi-electrode mapping was available.

Electrical measurements

Twelve-lead electrocardiogram

We used the Bard electrophysiology system (Boston Scientific, Natick, MA, USA) to measure ECG parameters such as QRS duration and R-wave peak time in V5/V6. Threshold testing was carried out in VVI mode, during unipolar and bipolar configurations. The anodal capture threshold (loss of V1 R-prime) during bipolar pacing was noted.

In every patient, the paced morphology at different AV delays was also examined, during pacing at the lowest output demonstrating left bundle branch capture. We adjusted the AV delay in 40 ms increments, with the aim of identifying whether fusion with intrinsic conduction via the right bundle could be achieved (evidenced by a loss of R-prime in V1).

Multi-electrode mapping

In some patients, selected opportunistically, multi-electrode mapping was used to examine the ventricular activation pattern and measure ventricular activation times.

These patients wore a vest with 252 electrodes [electrocardiographic imaging (ECGi); CardioInsight, Medtronic, Minneapolis, MN, USA]. Electrode location relative to each patient’s unique anatomy was accomplished using a low-dose thoracic computed tomography (CT).

The patients’ unique heart anatomy was manually segmented from the CT scan and used to construct a three-dimensional (3D) model involving both ventricles with the mitral and tricuspid valves. The left anterior descending artery was also manually segmented and used to define the border between the left ventricle and RV.

The analysis was carried out using custom software in Python version 3.6 (Python Software Foundation, Wilmington, DE, USA). For every parameter, the average of five beats each was taken.

In order to study the ventricular (epicardial) activation patterns, the software was designed to produce wavefront propagation videos. These depicted wavefront propagation across the epicardium. For a single beat, the entire duration of the electromyogram for every electrode was shown on the 3D model. The voltage was depicted by cylinders moving outward from the model with the cylinder height proportional to the negative dV/dt. This enabled a visual interpretation of the activation wavefront rather than relying on automated annotations.

The software was also used to determine activation times in a semi-automated fashion. The left ventricle and RV were marked using the left anterior descending artery as an anatomical reference for the interventricular septum. Activation onset in an individual electrode was defined as the time of most negative dV/dt. Left ventricular activation time was the time from the earliest to the latest activation of the left ventricle. The right ventricular activation time was calculated as the time from the earliest to the latest activation of the RV.

Ventricular activation patterns, total, left ventricular and RV activation times were recorded and analysed during the following:

1. Non-selective LBBP (R-prime present in V1): unipolar or bipolar pacing at the lowest output that achieved LBBP.

2. Non-selective LBBP fused with intrinsic conduction via the right bundle: pacing at the lowest output for left bundle area capture and the shortest AV delay resulting in fusion.

3. Non-selective LBBP fused with anodal capture (no R-prime in lead V1): Bipolar configuration at the lowest output demonstrating fusion with anodal capture with a loss of R-prime in lead V1.

Haemodynamic response

Haemodynamic response was assessed during non-selective LBBP with and without anodal capture using a high-precision haemodynamic protocol.^22,23 In brief, this consisted of using beat-by-beat blood pressure measured either invasively or non-invasively. Each tested setting was compared with a reference setting. At least six alternations were made between the tested and the reference setting. The analysis was automated and the mean change in systolic blood pressure was calculated for each tested pacing configuration, relative to the reference setting, which was kept constant in an individual patient. The average of eight beats was calculated at each alternation. To avoid any interruption to pacing, we programmed a heart rate that was 10–15% above the intrinsic rate.

Ethics

All patients provided written informed consent. The study was approved by the health research authority (REC 19/YH/0174) and registered on ClinicalTrials.gov (NCT04221763).

Statistical analysis

Continuous variables are expressed as mean and standard deviation or median and interquartile range (IQR). Categorical variables are expressed as proportions. For normally distributed variables, a Student’s t-test was used for comparison and a paired Student’s t-test was used for dependent variables. A P-value of <0.05 was considered statistically significant. Statistical analysis was conducted in RStudio using the tidyverse package.

Results

Patient population

We recruited 21 patients with a mean age of 67 years, of whom 14 were male. The pacing indication was bradycardia in 5 and cardiac resynchronization in 16 patients. Patients received LBBP for both indications as the primary strategy. Multi-electrode mapping was available in 16/21 patients, and all were in the cardiac resynchronization subgroup.

Demographics are provided in Table 1.

Table 1.

Patient characteristics

Demographics
Male	14 (67%)
Age, years	67 ± 11
Device type
VVI	1 (5%)
DDD	4 (19%)
CRT-P	1 (5%)
CRT-D	15 (71%)
ECG morphology
Normal	2 (10%)
RBBB	3 (14%)
LBBB	14 (67%)
IVCD	1 (5%)
Paced	1 (5%)
ECG QRS duration, ms	157 ± 28
Cardiomyopathy	17 (81%)
Ischaemic	2 (12%)
Non-ischaemic	15 (88%)
LVEF, %	36 ± 10
RV TAPSE, cm	2.3 ± 0.5
PAP, mmHg	16 ± 11

CRT, cardiac resynchronization therapy; DDD, dual chamber pacemaker; ECG, electrocardiogram; IVCD, intra-ventriuclar conduction delay; LBBB, left bundle branch block; LVEF, left ventricular ejection fraction; PAP, pulmonary artery pressure; RBBB, right bundle branch block; RV, right ventricular; TAPSE, tricuspid annular plane systolic excursion; VVI, single chamber pacemaker.

Mechanism of loss of R-prime evaluated using multi-electrode epicardial propagation maps

We acquired multi-electrode epicardial propagation maps in 16 of the 21 patients. This allowed us to assess the ventricular activation pattern during the two ventricular pacing configurations.

During all forms of LBBP, left ventricular activation was rapid and physiological with the main ventricular activation wavefront proceeding from apex to base in 16/16 patients, which is consistent with activation via the left conduction system; in this case, most consistent with proximal left bundle branch or left anterior fascicular capture.²⁰

The pattern of RV activation differed between the two forms of LBBP (Figures 1 and 2).

Figure 1.

Activation during non-selective left bundle branch capture without anodal capture. Activation starts with local myocardial capture. The RV activates from base to apex (i.e. not consistent with right bundle activation), and the latest activated area is the basal free wall (which appears to activate from a wavefront propagating from the posterior wall). The left ventricle activates in a physiological fashion from apex to base. LAD, left anterior descending artery; LV, left ventricular; RV, right ventricular.

Figure 2.

Non-selective left bundle branch capture with anodal capture. The RV is activated early, compared with capture without anodal capture. Right ventriclar activation occurs from base to apex (i.e. different to intrinsic activation via the right bundle). Activation of the basal lateral wall of the RV is advanced compared with LBBP without anodal capture. Right ventriclar activation appears to occur via a single wavefront (rather than the two wavefronts observed during capture without anodal caputure). The left ventricle activates in a physiological fashion from apex to base. LAD, left anterior descending artery; LV, left ventricular; RV, right ventricular.

During LBBP without anodal capture, RV activation appeared to occur via activation from two wavefront. The first wavefront appeared to propagate from septal myocardial breakthrough and a second late wavefront propagating from the posterior wall (likely activation spreading from left conduction system capture). The RV was activated in a non-physiological pattern from base to apex, rather than apex to base, suggesting myocardial cell-to-cell conduction rather than right bundle capture. The basal RV free wall was the latest activated area in all patients (Figure 1).

Non-selective LBBP with anodal capture, resulted in the advancement of RV activation compared with activation without anodal capture in 16/16. The earliest area of RV activation was basal RV (due to RV myocardial activation), with activation spreading from base to apex of the RV, i.e. not consistent with activation via the right bundle. The activation of the basal lateral wall was advanced compared with pacing without anodal capture, which appeared to occur because of earlier RV myocardial activation.

Therefore, the propagation maps suggested that anodal capture advanced RV activation via myocardial capture (likely of the RV septum) rather than bi-bundle capture in all our patients (Figure 2).

During non-selective LBBP fused with intrinsic right bundle activation (achievable in 6/16), RV activation was rapid and consistent with physiological activation via the right bundle, with the activation wavefront propagating from RV apex to base.

Comparison of ventricular activation times: left bundle branch pacing with and without anodal capture

Non-selective LBBP with anodal capture produced a modest reduction in 12-lead ECG QRS duration (−12 ± 7.0 ms) compared with non-selective LBBP without anodal capture (mean QRS duration 129 ± 14 ms without anodal capture and 116 ± 12 ms with anodal capture, P < 0.01; Figure 3).

Figure 3.

Comparison of activation times with and without anodal capture. (Left) QRS duration (12-lead ECG) produced by non-selective LBBP without and with anodal capture, there was a modest reduction in QRS duration associated with anodal capture. (Middle) TVAT (measured with multi-electrode mapping) with and without anodal capture, there was a modest reduction in TVAT with anodal capture. (Right) LVAT (measured with multi-electrode mapping) with and without anodal capture, there was no difference in LVAT between the two configurations. ECG, electrocardiogram; LVAT, left ventricular activation time; NSLBBP, non-selective left bundle branch pacing; TVAT, total ventricular activation time.

In patients with cardiomyopathy, non-selective LBBP with anodal capture produced a modest reduction in 12-lead ECG QRS duration (−13.2 ± 7.0 ms) compared with non-selective LBBP without anodal capture (mean QRS duration 131 ± 14.6 ms without anodal capture and 117.9 ± 13.0 ms with anodal capture, P < 0.01).

In patients without cardiomyopathy, non-selective LBBP with anodal capture produced a modest reduction in 12-lead ECG QRS duration (−7.7 ± 7.2 ms) compared with non-selective LBBP without anodal capture (mean QRS duration 117.7 ± 7.5 ms without anodal capture and 110 ± 4.6 ms with anodal capture, P = 0.2).

In the patients in whom multi-electrode mapping was available (16/21), we observed a modest reduction in total ventricular activation time (TVAT) of (−7 ± 9 ms) with anodal capture compared with LBBP without anodal capture (TVAT 90 ± 15 ms without anodal capture and 83 ± 18 ms with anodal capture, P = 0.01; Figure 3). Left ventricular activation times did not differ between the two pacing configurations (68 ± 19 and 69 ± 18 ms, P = 0.6; Figure 3).

Prolonging AV delay resulted in fusion with intrinsic right bundle conduction in 6/16 patients, all had cardiomyopathy and multi-electrode mapping was available in all 6. We observed a modest reduction in QRS duration and TVAT, during fusion with intrinsic RV activation compared with LBBP with anodal capture without fusion [QRS duration: with anodal capture 115.7 ± 17.7 ms and fusion 107.7 ± 19.2 ms, confidence interval (CI) 1.4–14.6 ms, P = 0.03, TVAT: with anodal capture 79.3 ± 17.1 ms and fusion 77.7 ± 13 ms, CI −7.6 to 10.9 ms, P = 0.66].

Left ventricular activation times did not differ between the two pacing configurations (62 ± 3 and 59 ± 8 ms, P = 0.4; Figure 4).

Figure 4.

Comparison of activation time between non-selective LBBP without anodal capture and LBBP with AV delay programmed to allow fusion with intrinsic right bundle activation. (Left) TVAT (measured with multi-electrode mapping) with and without fusion with intrinsic right bundle conduction, there was a modest reduction in TVAT with fusion with right bundle branch conduction. (Right) LVAT with and without fusion with intrinsic right bundle conduction, there was no difference in LVAT between the two configurations. LVAT, left ventricular activation time; NSLBBP, non-selective left bundle branch pacing; RBB, right bundle branch; TVAT, total ventricular activation time.

Haemodynamic response

High-precision haemodynamic response was assessed in all 21 patients. There was no significant haemodynamic difference between non-selective LBBP capture only compared with non-selective LBBP with anodal capture: −0.2 ± 3.8 mmHg, P = 0.2 (Figure 5).

Figure 5.

Acute haemodynamic comparison of LBBP with and without anodal capture. We found no difference in acute haemodynamic response during left bundle pacing with and without anodal capture. NSLBBP, non-selective left bundle branch pacing.

In one patient, we observed a large improvement with LBBP with anodal capture compared with LBBP without anodal capture. This patient had intrinsic right bundle branch block with a QRS duration of 180 ms at baseline and raised pulmonary artery pressure (38 mmHg). Anodal capture resulted in a 14 ms reduction in total activation time, compared with the mean reduction of 7 ms across all patients.

Pacing thresholds

Left bundle branch pacing with anodal capture required higher pacing outputs, mean capture threshold was 3.6 ± 1.9 V at 0.9 ± 0.2 ms compared with 0.6 ± 0.2 V at 0.8 ± 0.3 ms for LBBP without anodal capture (P < 0.01; Figure 6).

Figure 6.

Pacing threshold for non-selective LBBP with and without anodal capture. Achieving anodal capture in addition to left bundle area capture required significantly higher pacing outputs. NSLBBP, non-selective left bundle branch pacing.

Discussion

This study confirms that achieving anodal capture during non-selective LBBP produces modest improvements in ventricular activation times, compared with non-selective LBBP without anodal capture. Our findings suggest that activation time was reduced by advancing RV myocardial capture, rather than direct capture of the right bundle (bi-bundle capture) in all our patients in whom we collected ventricular epicardial propagation maps.

Achieving anodal capture required higher pacing outputs and even with high-precision measurements, there was not even the slightest sign of better acute haemodynamics. These findings suggest that adding anodal capture improves the electrical appearance without any immediate benefit to cardiac performance.

Mechanism of early right ventricular activation

This is the first study to apply multi-electrode mapping to address the question of whether anodal capture during non-selective LBBP advances RV activation by capturing the right bundle branch or through RV septal myocardial capture.

Our data suggest that RV activation during non-selective LBBP without anodal capture occurs via two wavefronts, one propagating from septal breakthrough and the second a later wavefront which appears to initiated as a result of left conduction system activation. The wavefront propagation pattern in all 16 patients suggests that RV activation is non-physiological, proceeding from base to apex.

During non-selective LBBP with anodal capture, we observed an earlier RV activation by RV myocardial capture. The pattern remained non-physiological in all 16 patients (base to apex).

In our cohort, we did not observe direct capture of the right bundle branch or retrograde activation of the right bundle branch as the propagation pattern in the RV was not consistent with activation through the conduction system. A recent study using ultra-high frequency electrocardiography also suggested that advancement of RV activation with anodal capture was achieved through myocardial capture.¹⁸

The reason for this may be that the left and right bundles do not pass down the septum in mirror image positions, but rather the right bundle typically courses more anteriorly and superiorly than the left bundle. From the bundle of His, the left bundle branch branches into its fascicles and courses posteriorly (relative to the right bundle) just under the left ventricular endocardium. The right bundle travels as a continuation of the bundle of His and courses anteriorly relative to the left bundle branch. This anatomical separation of the right and left bundle fibres means that it is technically difficult to capture both with a single lead when it is implanted distal to the His bundle.

During non-selective LBBP without anodal capture, we observed two RV activation wavefronts, whereas with anodal capture, we only observed one wavefront. We think this difference occurs because the onset of RV activation is delayed, while the wavefront originating from left septal myocardial capture spreads through the septum to the RV. This delay in RV activation means that the RV is not fully activated when a second wavefront, occurring as a result of left conduction system activation, reaches the RV. Whereas with anodal capture, the septum is activated earlier and therefore, RV activation is complete prior to breakthrough of the wavefront originating from left conduction system capture.

Left ventricular activation

In the 16 patients where multi-electrode mapping was available, we confirmed physiological wavefront propagation in the left ventricle (apex to base) with short left ventricular activation times. This suggests that we obtained left conduction system capture. We did not observe any change in left ventricular activation pattern, stimulus to R-wave peak time in V6, or left ventricular activation time, during non-selective LBBP with and without anodal capture. This is what we would expect, as anodal capture is unlikely to influence left ventricular activation. It has been postulated that higher pacing outputs during LBBAP may advance left ventricular myocardial stimulation and therefore result in slower activation of the left ventricle,²⁴ we did not observe this in our cohort of patients.

The study was not designed to compare left septal only capture with left bundle capture, this interesting question requires investigation in future studies.

Haemodynamic response

Disappointingly, despite the clear advancement of RV activation, narrowing of the QRS and shortening of total activation time, there was no evidence of any haemodynamic advantage when anodal capture was delivered during non-selective LBBP. Indeed, the numerical values were a non-significant trend towards worse haemodynamics with anodal capture. The high-precision haemodynamic protocol we used can detect even small haemodynamic advantages using large numbers of repeated transitions, automated beat alignment and standardized statistics. Therefore, it seems very unlikely that adding anodal capture to LBBP will ever provide improvements in acute haemodynamics.

We speculate that the reason for this lack of haemodynamic benefit is that while anodal capture does advance RV activation compared with non-selective LBBP without anodal capture, this only produces a modest reduction in activation time by advancing the latest area of activation (RV basal septum).

It appears that left ventricular septal capture, during non-selective LBBP, is beneficial as it results in earlier RV activation. This assumption is supported by the findings from the study by Sun et al.²⁵ They observed a mean 25 ms reduction in QRS duration with non-selective LBBP compared with selective left bundle pacing [non-selective LBBP QRS 125 ms (IQR: 117–142 ms) vs. selective LBBP QRS 150 ms (IQR: 135–157 ms); P < 0.001]. During selective LBBP RV activation occurs due to breakthrough from the wavefront occurring as a result left conduction system activation, which occurs later than breakthrough resulting from left septal myocardial capture.

Implications for device programming

During biventricular pacing pacing physicians often aim to programme the device to achieve the narrowest possible QRS. However, the findings of our study suggest that for non-selective LBBP, while adding anodal capture shortens QRS duration, it does not improve acute haemodynamic function. This narrower QRS often comes at a substantial cost in terms of the pacing output required and therefore battery life.

It is possible that anodal capture thresholds could be lowered with dedicated leads, or with changes in the implant technique. We did not formally measure the anodal position in relation to the septum in our study. However, we performed anode-only pacing, and myocardial capture was achieved in all patients, suggesting that the anode was in contact with the myocardium.

Therefore, our findings suggest that outside of adequately powered randomized controlled trials of long-term effects, we should not routinely programme anodal capture for patients in whom this requires many-fold greater pacing output. This represented the majority of our patients.

Limitations

The multi-electrode mapping equipment was available only in 16/21 patients due to funding limitations. All 16 patients had an indication of cardiac resynchronization therapy, and in most of them, the baseline conduction pattern was left bundle branch block; this may not be a representative population. However, the activation patterns seen were consistent in all 16 patients, and apart from one outlier, we did not observe a large variability in the haemodynamic response.

The multi-electrode mapping technique, ECGi, can report only epicardial activation and not endocardial. Septal activation cannot be explored in detail. Invasive mapping would be required to address this, but this would have introduced an additional risk of complications. Nevertheless, we believe that the information from ECGi is persuasive and the activation occurs through myocardial stimulation, rather than through the right bundle.

The number of patients recruited was appropriate for a physiological study and delivered narrow CIs as planned. However, this is small compared with the sample size required for an event study where each patient contributes only one binary digit of information.

We did not address the difference in RV activation time or pattern between selective and non-selective LBBP. Everyone demonstrated non-selective capture at programmed outputs.

In this study, we used long programmed AV delays as one method of achieving fusion between paced left ventricular activation and native RV activation. However, this was purely an experimental method of temporarily achieving fusion and not a recommendation to programme this chronically in routine practice. Programming long AV delays in patients with a long intrinsic PR interval may adversely affect ventricular filling, which may offset the beneficial effects of intrinsic RV activation.¹⁵

In this study, we could not compare LBBP alone to LBBP combined with intrinsic right ventriuclar conduction, which should be performed at the same AV delay. Using our experimental set-up it is impossible to determine whether any additional haemodynamic benefit was acheived by combining LBBP with intrinsic RV conduction as the haemodynamic benefit may have resulted purely from AV delay optimisation and it's impact on ventricular filling. We cannot, therefore, answer the question whether acute haemodynamics resulting from RV activation via the intrinsic conduction system is superior to LBBP pacing–initiated RV activation. The aim of our study was to establish the impact of anodal capture.

This was an acute study, and we did not investigate the chronic impact of LBBP with and without anodal capture. Whether the addition of anodal capture results in any long-term benefits would need to be investigated in a separate study.

Conclusions

Adding anodal capture to LBBP does attenuate the delay in RV activation, and the mechanism is stimulation of the RV septal myocardium rather than capture of the right bundle branch. There is a clear but modest shortening of QRS duration and TVAT. However, this requires an approximately six-fold higher pacing output and shows not even the slightest sign of improving acute haemodynamics. Our findings do not routinely support programming of anodal capture during LBBP.

Funding

The study was funded by a grant from the British Heart Foundation (FS/19/4/34013) and Imperial College London Biomedical Research Center.

Data availability

Data are available from the authors upon request.