Clinical use conditions of lead deployment and simulated lead fracture rate in left bundle branch area pacing

Left bundle branch area pacing (LBBAP) is achieved by advancing the lead tip deep in the septum. Most LBBAP implants are performed using the Medtronic SelectSecure™ MRI SecureScan™ Model 3830 featuring a unique 4 Fr fixed helix lumenless design. Details of lead use conditions and long‐term reliability have not been reported. This study was designed to quantify the mechanical use conditions for the 3830 lead during and after LBBAP implant, and to evaluate reliability using bench testing and simulation.


| INTRODUCTION
Left bundle branch area pacing (LBBAP), which adopts the transventricular septal approach for the lead implantation, was introduced in late 2017. 1 Multiple clinical studies have found that LBBAP generates a relatively narrow QRS duration, fast left ventricular (LV) activation time, and a low and stable pacing capture threshold. [2][3][4][5][6] Thus, LBBAP has grown rapidly as a physiological pacing modality.  7 The long-term reliability of the 3830 lead in clinical use in atrial and ventricular locations has been well documented. [8][9][10] Implantation of the 3830 lead for LBBAP requires additional mechanical stress on the lead to advance the lead tip deep into the interventricular septum (IVS). Fluoroscopic observation of unique lead bending shapes near the ring electrode, that is, the fulcrum sign, has been suggested to help identify proper septal positioning. 11 The additional mechanical stress on an LBBAP lead could pose a risk for conductor fracture, a failure mode which may require lead replacement. Although short-to medium-term follow-up has not identified an increase in lead-related complications, 3,5,6 long-term data are not available due to the novelty of the LBBAP procedure.
Recent reports have described the importance of aligning invitro bench testing with clinical in-vivo use conditions to provide accurate assessments of implant durability. 12,13 Relevant use conditions for lead fracture reliability are handling during the implant procedure and the chronic bending environment. When sufficient in-vivo and in-vitro data are available, statistical modeling may be used to predict survival. The objective of this investigation was to forecast the long-term reliability of the 3830 lead in the LBBAP position. We did so by assessing the use conditions of the 3830 lead in the LBBAP procedure, conducting bench fatigue testing, and performing statistical modeling based on the in-vivo and in-vitro data.

| METHODS
This investigation consisted of two parts. Part 1 was the data collection of the lead use conditions at LBBAP implantation of a 3830 lead under the routine clinical setting and the computed tomography (CT) imaging that measured the mechanical performance of the 3830 lead. Part 2 was the evaluation of in-vivo lead curvature (inverse of radius), bench testing, and modeling prediction of the 3830 lead fracture reliability.

| Method Part 1: Real world data collection
The IMAGE-LBBP study was a multisite data collection study in patients who were indicated for pacing therapy and received LBBAP implantation, described in detail in the previous publication. 14

| Data collection
All data were collected at the 3-month follow-up visit, including patients' demographic info, lead use conditions (lead placement attempts and rotations), and multiphase electrocardiography (ECG)gated CT scan data, as described in the previous IMAGE-LBBP study publication. 14 The reasons for failed lead placement and lead repositioning were also documented at the implantation. Lead curvature amplitude and implant handling data were used to define limits for bench testing.

| Lead fatigue bench testing
Two primary differences from traditional ventricular implant were noted as most relevant to fatigue analysis: preconditioning from the higher number of turns to implant the lead deep into the F I G U R E 1 Mechanical measurements of the 3830 lead in a patient with left bundle branch area pacing (LBBAP). (A) The distal body of the LBBAP lead inside the interventricular septum (IVS). (B) Anterior oblique (left) and superior to inferior (right) view of the lead in diastolic (pink) and systolic (green) cardiac phases is presented. An example of the 3D lead bending throughout the cardiac cycle is shown in Supporting Information: Video S1. The bending along the 3830 lead was assessed throughout the cardiac cycle and the geometric curvature was determined. septum and the unique bending conditions due to the deep septal implant. "Preconditioning" refers to the stresses placed on the lead due to the torque that can build up during the implantation procedure, and the consequences of those stresses to the chronic fatigue performance of the lead. Leads were preconditioned before testing as follows: The lead was inserted into either a C315HIS or C304HIS (Medtronic Inc.) catheter such that approximately 2 mm of the distal tip of the lead exposed. The exposed portion of the tip was then mounted to a rigid fixture.
The proximal end of the lead was rotated by hand for a minimum of 20 continuous turns, then released. The process was repeated at least 5 times. This exposure represents an extreme scenario where a physician attempts to implant the lead at 5 different locations, each time applying 20 continuous rotations with no release of built-up torque. Preconditioning is illustrated in Figure 2A.
Following the preconditioning, two separate fatigue tests were performed, Test 1 was to identify the potential effects of implant handling, and Test 2 was to evaluate fatigue reliability in the most stressful in-vivo conditions. Test 1 was conducted at levels of bending substantially higher than observed in-vivo, intended to generate fractures on most samples. Half of the parts in this test were subjected to precondition to simulate implant handling as described above. The other half did not receive this exposure. Test 2 was performed at an upper percentile of the in-vivo observed bending, with simulated implant preconditioning performed beforehand. Test 1 performed bending in a uniform section of the lead and Test 2 fixtured the distal end of the lead to replicate a deep septal position, focusing the bending near the ring electrode. General fatigue test methodology follows. 15 Direct current resistance (DCR) of the tip and ring conductors was monitored throughout the test, and conductor fracture was identified by DCR exceeding 3000 Ω. Fatigue test conditions are illustrated in Figure 2 and Supporting Information: Video S2.

| Lead reliability modeling
Simulated fracture reliability was calculated following the approach described previously. 12 Briefly, individual patients were simulated to have a random lead curvature and lead fatigue strength, all from random distributions based on the CT imaging and bench testing. The simulated curvature was compared to the fatigue strength and the time to fracture was computed. This process was repeated many times to generate Kaplan-Meier survival curves for the leads.

| Rate of fracture of 3830 SelectSecure in standard pacing use
Estimate for fracture rate by anatomical zone was performed as described previously. 12 Returned product analysis (RPA) was used to determine the distribution of fracture location, and the actively monitored Medtronic Product Surveillance Registry (PSR) was used to estimate the overall rate of fracture.

| RESULTS
A total of 50 patients (age: 69.5 ± 10.4 years) with bradycardia indication for pacing therapy were enrolled from September 2019 to July 2020 and all received LBBAP. 14 At implant, physicians attempted LBBAP lead placement with 2.1 ± 1.3 attempts (range: 1-7; median: 2). Success after one attempt occurred in 22 patients (44%, Figure 3A). The number of lead deployments did not appear to be influenced by indications for pacemaker implantation ( Figure 3B) or implanting center ( Figure 3C). The number of lead body rotations at the final attempt was independent of the number of attempts ( Figure 3D), with 13 ± 6 rotations at the final attempt  Figure 4C). The angle between the distal lead body and the RV septum was 19.8 ± 10.5°at the cardiac diastolic period. Three cases (6%) required more than 20 lead rotations for successful lead deployment at the final target site and also showed the lead angled more than 20°from the perpendicular line ( Figure 4C). No lead tip (helix) was found in the LV blood pool in the CT images.
ECG characteristics of LBBAP are described in the previous publication. 14 The maximum curvature amplitudes of the 3830 lead measured from CT images were on average 0.23 cm −1 with a standard deviation of 0.09 cm −1 and a range of 0.55 cm −1 . Maximum curvature amplitude usually occurred within 15 mm of the ring electrode ( Figure 5). The length of the lead in the RV was 40 ± 12 mm. Figure 5 illustrates scenarios where high curvature amplitude occurs near the distal end of the lead, and where curvature amplitude at the distal end of the lead is very low.
A total of 66 leads were tested in Fatigue Test 1, and 36 were tested in Fatigue Test 2. Median curvature amplitude for Test 1 was 3.9 cm −1 , 700% higher than the maximum observed in-vivo condition.
The distribution of cycles to fracture in Test 1 was not affected by preconditioning (Supporting Information: Figure S1). Curvature levels in Test 2 ranged from the 95th percentile in-vivo condition to 300%

| DISCUSSION
The mechanical use conditions associated with LBBAP include the potential for a higher number of turns and more severe bending conditions compared to standard endocardial pacing. We found that there was no impact on fatigue performance after multiple applications of 20 turns, and that the level of bending curvature observed in the clinical portion of the study did not result in expected higher rates of long-term fractures for the 3830 lead. Bending stress is proportional to curvature, so we used curvature rather than angle as our measure of in-vivo stress. 15 The result of our model for intracardiac fracture rate (0.02% at 10 years) is within the confidence bound of an estimate from observed field performance. Other major clinical findings of this study were that implanters needed a median of 2 attempts and 13 lead rotations at the final attempt to achieve satisfactory LBBAP. The characteristic shape of the lead bending was a smooth curve and not a sharp hinge. The location of peak bending was usually more than 1 cm from the ring electrode and the surface of the septum.  The present study showed that a successful lead deployment for LBBAP required 2.0 ± 1.3 attempts with one-attempt success in 44% of the studied patients. This attempt number for LBBAP lead deployment appeared comparable to that for His bundle pacing (2.7 ± 2.2 attempts for successful HBP in the IMAGE-HBP study). 4 The failed lead deployments observed in the present study were mainly due to failed lead advancement in the septum (frequently with lead unwinding) and undesired paced ECG. The unwinding observed with failed lead advancement appeared to be due to the anchoring of the helix in the septal tissue, so that torque from clockwise rotations at the proximal end would build up until the lead was released. The torque release then produced counter-clockwise lead unwinding. The finding that there was no correlation between the number of rotations versus the septal thickness or the length of the lead inside the septum ( Figure 4A,B) implies that additional turns required may be due to the endocardial entanglement and endocardial barrier effects, as found in a human cadaver heart investigation of lead use conditions by Jastrzębski et al. 18 Although the observations of lead unwinding suggest more mechanical stress on the lead body, our testing of up to 5 attempts with 20 turns each did not reveal any impact to mechanical durability of the 3830 lead.
As described in the method, implanters maneuvered the delivery catheter in a way to make the distal catheter perpendicular to the ventricular septum. It is plausible to assume that the perpendicular angle of the distal catheter/lead would make the lead tip easier to penetrate the septal endocardium and travel the shortest distance to Our testing incorporated use conditions beyond the observed clinical experience, including continuous rotation with no torque release, extreme bending to generate fractures, and testing up to 400 million bending cycles to simulate a 10-year implant. The same approach used by Wilkoff et al. 12 showed the ability to distinguish between fracture rates across lead models and in different anatomical zones.
Our study also highlights the value of RPA combined with longterm, actively monitored clinical follow-up. Although there is no long-term data yet available for the 3830 lead in an LBBAP position, we were able to establish a baseline performance for fracture survival in standard pacing locations to provide a basis for comparison.

| Clinical perspectives
The lead deployment for LBBAP was achieved with a median of 2 attempts and 13 lead rotations at the final target site. While the fatigue bench testing and reliability modeling in the present study suggests a reliable long-term performance of the 3830 lead in LBBAP use conditions, long-term active monitoring of implanted leads will validate the prediction. However, it should also be noted that a regular practice is to feel the lead torque after several continuous turns to better control the penetration of the pacing lead and to prevent excess torque buildup, based on which multiple attempts of 20 continuous turns were applied as preconditioning with the worst case of use conditions. The use conditions of two or fewer attempts for achieving LBBAP occurred in 75% of enrolled patients, but seven attempts were also observed.

| LIMITATIONS
Our study has similar limitations as stated in the previous publication, 14 namely that the sample size is relatively small, only a single CT imaging session was conducted for each patient, and the CT images could be affected by metal-induced artifacts. Although the general bending fatigue test approach has been validated by others, it does not fully reproduce all aspects of in-vivo motion such as crushing or twisting (e.g., Twiddler's syndrome). Additionally, the previous fatigue testing work was for ICD leads, not pacing leads. The number of turns at implant was estimated in a clinical setting and not recorded by a calibrated instrument. Our accelerated fatigue test for conductor fracture used high speed to simulate 10 years of in-vivo lead bending in several months and did not accelerate lead insulation biodegradation. Because the study only evaluated the 4.1-French, lumenless, fixed helix 3830 SelectSecure design, the results may not apply to other lead designs. Additionally, we did not evaluate lead placement at the His bundle. Despite these limitations, our results can be interpreted together with medium-term clinical follow-up to suggest that the 3830 lead will have acceptable long-term mechanical performance when used in an LBBAP application.