Septal scar as a barrier to left bundle branch area pacing

The use of left bundle branch area pacing (LBBAP) for bradycardia pacing and cardiac resynchronization is increasing, but implants are not always successful. We prospectively studied consecutive patients to determine whether septal scar contributes to implant failure.


INTRODUCTION
Left bundle branch area pacing (LBBAP) promises a revolution for treating bradycardia and achieving cardiac resynchronization, 1-4 but success rates are variable and lower when patients with ischemic cardiomyopathy are included. 2,3,5,6 The reasons for this have not been well established, but potentially include challenges in passing the lead through the septum either due to the presence of septal scar or due to lack of support from delivery sheaths due to cardiac dilatation.
The most widely used technique for LBBAP involves deploying the lead through the interventricular septum, which may be scarred in patients in patients with ischemic or non-ischemic cardiomyopathy.
Late gadolinium enhancement (LGE) on MRI is a reliable technique for non-invasively detecting, localizing and quantifying myocardial scar. 7,8 We conducted a systematic protocol of MRI scanning before attempting LBBAP, to test the hypothesis that late gadolinium enhancement in the basal and mid septum might be able to predict likelihood of success in achieving LBBAP.

Study population
This was a prospective study of patients where LBBAP was attempted by a single operator. Consecutive patients referred for device implantation for cardiac resynchronization or bradycardia indications were recruited. However, patients with a bradycardia indication were only included if they had a clinical indication for a cardiac MRI (suspected cardiomyopathy) and were hemodynamically stable, and therefore able to have a pre-procedure cardiac MRI scan. showing scar as a percentage of the myocardium in each segment ( Figure S1).
Using the short axis stack of late gadolinium enhancement slices, the basal septum was defined as the three most basal anteroseptal and inferoseptal segments. Incomplete slices (i.e., those too basal to include a complete ring of left ventricular myocardium) were excluded. The mid septum was defined as anteroseptal and inferoseptal segments in the next three slices ( Figure S2) beyond the three basal slices. The scar burden was expressed as the proportion of myocardium showing late gadolinium enhancement and was calculated as the average across segments.
To maintain consistency across all patients, bright blood imaging was used for scar quantification, but where dark blood imaging was available this was manually reviewed to ensure scar was not underestimated. During lead advancement, we monitored impedance and time to peak R wave (RWPT) (time from the stimulus to peak R wave in lead V5 or V6) with unipolar pacing.

Left bundle branch area pacing
At the time of carrying out the study there was no consensus criteria for confirming LBBAP, we used the following criteria during unipolar pacing at the time of device implantation: R prime in lead V1 and either: (1) A transition in QRS morphology during threshold testing or programmed stimulation, (2) Pacing stimulus to R wave Peak Time in lead V6 < 90 ms.
We retrospectively applied the criteria published by Burri et al. 11 and used these to confirm whether LBBAP had been achieved. 12 Lead depth was assessed by placing the end of the sheath onto the septum and measuring the distance to the tip of the lead or by contrast injection. Furthermore, in cases where the lead was not advanced or left bundle branch was not captured, the paced ECG morphology was consistent with right ventricular septal capture and the stimulus to R wave peak time in V6 remained prolonged.
We defined two modes of failure. Unsuccessful lead deployment was inability to advance the lead deep enough into the septum to reach the left bundle branch area. Unsuccessful left bundle branch capture was failure to capture the left bundle branch despite successful lead deployment.
Paced QRS duration during LBBAP was measured from the pacing stimulus to the end of the QRS.

Statistical analysis
Continuous variables were expressed as mean with standard deviation if normally distributed and mean with interquartile range otherwise.
Correlations were assessed by the Spearman rank correlation coefficient. Comparisons between groups were performed using the Mann Whitney U test. A p value < .05 was considered statistically significant.
We undertook an exploratory analysis of the likelihood of delivering a lead, predicted from the extent of basal septal scar and mid septal scar.
Statistical analysis was conducted in RStudio using the tidyverse package. 13

Patients
Thirty-five patients were enrolled in the study, demographic and clinical characteristics are shown in Table 1

Septal scar quantification
Late gadolinium enhancement was present in the basal septum or mid septum in 25 of the 35 patients. Mean septal scar burden was 12% (IQR 3.5%-21%). There was considerable variation in septal scar burden, with a range between 0% and 67%. shown in Tables S1 and S2.

Reproducibility
The reproducibility of scar quantification is as follows; for basal septal scar quantification the standard deviation of the intra-observer variability was 2.8% and the between scar measurement variability (after subtracting out the intra-observer variability) was 1.2%, and for mid septal scar quantification the standard deviation of the intraobserver variability was 3.3% and the between scar measurement variability (after subtracting out the intra-observer variability) was 0%.

LBBAP
The lead was successfully deployed to the left side of the interven-

Septal scar and lead deployment
There was significantly less septal scar in patients where left septal lead deployment was successful compared to those where lead deployment failed. Mean septal scar burden was 8% (IQR 2%-18%) in successful implants compared to 54% (IQR 53%-57%) in unsuccessful implants (p < .001, Figure 2). Mean Septal scar burden of <40% was associated with successful left septal lead deployment ( Figure 2). The relationship between left ventricular end diastolic volume and lead deployment is shown in the supplemental material and Figure S1.

Distribution of septal scar
We also analyzed the impact of basal septal scar and mid septal scar (Figures 3 and 4). We found that three patients had basal septal scar of between 40% and 45%. In two of these patients left septal lead implantation was successful in the mid septum (done with operator blinded to MRI result), in these two patients mid septal scar was found to be <40%. Whereas in the third patient lead implantation was unsuccessful and the MRI scan in this patient demonstrated extensive mid septal scar in addition the basal septal scar.

Septal scar and electrical parameters
We found no correlation between the extent of septal scar and Stimu-  Figures S3 and S4).

DISCUSSION
The findings from our study suggest that the presence of septal scar may be an important factor in failure to advance a lumenless pacing

Septal scar and lead deployment
In this study, lead deployment was unsuccessful in 14% of patients.
Our patient group was predominantly patients with a CRT indication and drawn from a tertiary center where advanced cardiomyopathy was frequent. 82% had cardiomyopathy and 34% had ischemic cardiomyopathy. 88% of our non-ischemic cardiomyopathy patients had scar, which is a higher rate than other published cohorts. 14 Therefore, the prevalence of septal scar in our population is likely to be higher than a general pacing population.
Early reports of LBBAP were largely in the context of bradycardia pacing in patients with structurally normal hearts. However, extending its application to heart failure including ischemic cardiomyopathy has inevitably brought higher failure rates. 3 Some authors have discouraged left bundle branch area pacing in ischemic cardiomyopathy for this reason, 15,16 suspecting that scar in the basal or mid septum may limit procedural success.
The findings from our study suggest that septal scar does appear be an important mechanism for failure to implant a lead deep into the ventricular septum and supports the suspicion that septal scar may impede lead progress through the septum. There was a clear association between extent of scar and success of left septal lead deployment, with a septal scar burden of < 40% predicting successful deployment.

Clinical implications
The findings of our study may be useful in clinical practice for the following reasons.
Firstly, being able to identify patients in whom lead implantation is likely to be more challenging, may help with procedural planning. If there is <40% scar burden, lead advancement is very likely to be successful, but if there is >50% septal scar burden lead implantation is likely to fail. Identifying more challenging cases pre-procedure may be helpful for procedure scheduling and when considering which tools to use and which pacing approach to adopt.
Secondly, the distribution of the scar within the septum may be important. Two patients in our series had >40% scar in the basal septum, we failed to implant the lead in the basal septum in these patients.
However, they had <40% scar in the mid septum and we were able to deploy the lead in a mid septal position in both of these patients.
In our study the operator was blinded to the septal scar information and therefore attempted to deploy the lead in the basal septum prior to moving to the mid septum. Having pre-procedural information could potentially help streamline the procedure, as the operator could avoid areas with extensive scar.
Thirdly, the recognition that sepal scar appears is likely an important mechanism for lead implantation failure, may help with the design and development of new implantation tools. It is possible that modifications in lead design or delivery systems could increase implantation success in patients where scar forms a barrier to lead penetration.
This study suggests that patients may benefit from the development of better tools to penetrate an extensively scarred septum. However, the implications of placing a lead in a region with extensive scar will need further study with long term clinical follow up.

STUDY LIMITATIONS
This was a single-center study with a relatively small number of patients (35 patients), studying larger groups will provide more information about procedural success in different patient populations.
The majority of patients had a CRT indication, which is a population of patients expected to have a higher scar burden. We did include con- To confirm left bundle branch capture we used conventional criteria based on current practice. There is still debate regarding the optimal criteria for distinguishing left conduction system capture from left ventricular septal capture. We confirmed left bundle area pacing (either or both left ventricular septal or left conduction system capture) in all patients where the lead was successfully deployed. We were therefore able to address our primary question of whether septal scar was associated with failure to progress the lead through the septum.
We only used a lumenless lead in this series and we therefore do not know whether using a stylet driven lead could have increased implant success rate.

CONCLUSIONS
Septal scar is associated with higher implant failures, when attempting to deploy a pacing lead to the left bundle branch area. Pre-procedure knowledge of septal scar burden and distribution could potentially help streamline implant procedures.

ACKNOWLEDGMENTS
The study was funded by the British heart foundation (grant reference FS/19/4/34013) and Imperial College Biomedical Research Center.

CONFLICT OF INTEREST STATEMENT
ZW Speaker fees, advisory board, consultancy Medtronic, advisory board Abbott, speaker fee Boston Scientific.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.