Feasibility and safety
The feasibility of using LBBP to deliver CRT was first reported by Huang et al. as a rescue pacing modality after failed coronary sinus lead placement [
1]. Since then, several groups have confirmed the feasibility and safety of LBBP using the Medtronic 3830 SelectSecure pacing lead in short-term studies [
1,
2,
8,
10]. In our study the implant success rate using LBBP was 78% and was comparable to data reported in recent multicentre studies evaluating the feasibility of LBBP for CRT indications [
8,
11]. Our implant success rate was limited mainly by: (1) inability to engage the septum, especially in patients with septal scar tissue after myocardial infarction (
n = 2) or severely enlarged cardiac chambers (
n = 4); and (2) inability to meet the predefined electrocardiographic criteria for LBBP (
n = 3). The current delivery sheath and lead were not specifically developed for LBBP and new developments in delivery sheath [e.g. increased sheath length and (steerable) sheaths offering more support] and possibly lead design (e.g. stylet driven leads) may improve the success rate. In addition, the predictive value of anteroseptal scar tissue as regards successful LBBP needs further investigation, as this could prevent futile attempts at LBBP.
LBBP can be performed swiftly, i.e. the total procedure time was approximately 30 min shorter when compared to BVP (and 20 min shorter when unsuccessful attempts were included in the analyses). Cannulation of the coronary sinus, venography and proper positioning of the coronary sinus lead can be challenging and time-consuming. In our opinion, placement of an LBB lead is a technically less complex and a more predictable procedure.
LBBP can be safely performed. No LBBP-related complications, such as lead dislodgement/perforation, a sudden increase in capture threshold or loss of LBB capture, were recorded during the 6 months of follow-up. It needs to be emphasised that the results should be interpreted with caution, since the sample size was relatively small and the follow-up period short.
With LBBP, sensing and pacing parameters were excellent when compared to BVP. At 6 months the capture threshold at 0.4 ms was 0.7 V for LBBP versus 1.5 V for BVP. Lead maturation, with subsequent low capture thresholds, may benefit from a deep intramyocardial lead position as compared to an epicardial position of coronary sinus leads.
Clinical outcomes
LBBP resulted in a greater reduction in QRS duration when compared to BVP. With BVP, activation of the ventricles utilises non-physiological, slow cell-to-cell conduction, whereas LBBP employs the intrinsic Purkinje conduction system, leading to a narrower QRS complex. In patients with conventional bradycardia pacing indications, Sharma et al. [
12] recently demonstrated that, in contrast to RV pacing, LBBP can prevent development of heart failure and mortality. RV pacing creates an activation pattern comparable to that of LBBB. Prior studies of CRT have demonstrated that QRS narrowing is associated with better clinical outcomes [
13,
14].
Recently, Huang et al. [
15] confirmed the feasibility and effectiveness of LBBP for CRT in preselected heart failure patients with LBBB and non-ischaemic cardiomyopathy. Implant success rates were extremely high (97%, 61 of 63 patients) with a stable capture threshold and R‑wave amplitude at 1‑year follow-up. The QRS duration narrowed from 169 ± 16 to 118 ± 12 ms during LBBP, which is comparable to our data in a mixed population with approximately 30% of patients having an ischaemic cardiomyopathy. The LVEF (33 ± 8% vs 55 ± 10%) and LV end-systolic volume (123 ± 61 ml vs 67 ± 39 ml) of their patients improved significantly compared to their baseline values. Their selection criteria—patients with typical LBBB and non-ischaemic cardiomyopathy—likely resulted in a high rate of super-responders in their study population. The largest retrospective multicentre study assessing the feasibility of LBBP for CRT was published recently by Vijayaraman et al. [
8]. LBBP was achieved in 85% of the patients (277/325). However, only 39% of their patients had baseline LBBB. The outcomes in this subgroup are comparable to our results, i.e. a reduction in QRS duration from 162 ± 24 ms to 133 ± 22 ms (
p < 0.01) with LVEF improvement from 30 ± 8% to 44 ± 11% (
p < 0.01) at 6 months of follow-up. NYHA class improved from 2.8 ± 0.6 to 1.7 ± 0.7.
Limitations, unknowns and future perspectives
The most important limitations of this study are its small sample size and its non-randomised design. Therefore, the results of this study must be interpreted with caution; in particular, comparisons between LBBP and BVP should be regarded as preliminary. The study was neither designed nor powered to determine differences in hard clinical endpoints. Large randomised controlled trials are needed to confirm the feasibility, long-term safety and clinical effectiveness of LBBP versus BVP for various CRT indications.
We conducted the study when LBBP was at a very early stage of clinical application, with limited criteria for LBB capture. In the current study LBB capture was confirmed by only two markers, i.e. (1) the paced QRS morphology in lead V1 demonstrating an RBB conduction delay or block pattern and (2) a stable and short LVAT. There were, however, no validated cut-off values for what the LVAT should be. In our cohort, the mean LVAT was 81 ms, which is similar to previously reported data [
16‐
18]. However, in patients exhibiting latency or diffuse peripheral conduction disease, LVAT values may be prolonged, even with LBB capture. LVAT may also be extended in patients with severely dilated hearts, where path length to the LV lateral wall is increased. Therefore, it cannot be excluded that in three patients LBBP attempts were incorrectly considered not successful due to an LVAT > 90 ms. On the other hand, LBBP attempts in patients in whom deep LV septal pacing (with secondary activation of the left-sided Purkinje system) was performed instead of direct capture of the LBB might have been classified as successful. Whether LV septal pacing is inferior to LBBP deserves further research, especially since Mafi-Rad et al. [
19] demonstrated that, in patients with sinus node dysfunction, LV septal pacing (probably with capture of distal arborisations of the left-sided Purkinje system) is able to preserve acute LV pump function (comparable to atrial pacing). In contrast, in heart failure patients the situation may be different. In the LOT-CRT study [
20] patients with evidence of LBB capture had better clinical outcomes than patients with LV septal pacing. Meanwhile, several groups have been working on the establishment of novel evidence-based, rather than arbitrary, criteria for differentiation between LBBP and LV septal pacing [
21‐
23].
Although LBBP seems a potential alternative to conventional CRT in patients with LBBB, it is still unknown which patients will benefit most from this novel technique. Heart failure patients with intraventricular conduction delay (IVCD) or RBBB might benefit less from LBBP, as LBBP may not resolve or resynchronise a delayed right ventricle. IVCD or LBBB may be the result of a diffuse, gradual conduction delay in the entire left bundle conduction system [
24], which may not be synchronised by pacing at the proximal left bundle. These patients may benefit most from classic BVP. In contrast, if the LBBB can be corrected by LBBP, the region of block is most likely situated proximal to the pacing site and with intact distal Purkinje conduction [
24].
Although analysis of all baseline and follow-up echocardiograms was performed in an anonymised manner, at random and after the last patient had completed follow-up, it cannot be fully excluded that outcomes may have been biased, since pacing leads in the interventricular septum were often easily visible.
In addition, delivering CRT by use of a dual-chamber pacemaker or ICD instead of a biventricular device may be helpful in limiting the continuously increasing health care costs.