We included consecutive patients undergoing LBBAP at Centro Hospitalar e Universitário de Coimbra, a tertiary high-volume CIED center, from January 2023 to January 2024. Indications for pacing included sinus node disease (single-chamber or dual-chamber), atrioventricular block (single-chamber or dual-chamber), slow atrial fibrillation (single-chamber unless paroxysmal or persistent atrial fibrillation, in these cases dual-chamber) and CRT [CRT-Pacemaker (CRT-P) or CRT-defibrillator (CRT-D)]. There were no specific criteria to select patients for LBBAP, RV pacing, or biventricular pacing. Before implantation, operators discussed the potentially longer procedure but with possible better long-term results in patients. All patients provided informed consent. The study complied with the Declaration of Helsinki and the study protocol was approved by the institutional ethics committee.

Each LBBAP procedure (successful or not) was registered in a database where we recorded implantation data such as the numbers of leads used, time to lead positioning, success of LBB area capture, electrical parameters (threshold, sensing, and impedance) and all measurements to confirm LBB capture. Relevant demographic variables and follow-up data were also registered in the database and were retrieved via hospital electronic medical records. Left ventricular dysfunction was classified as per the American Society of Echocardiography and the European Association of Cardiovascular Imaging.15

Implant success was reviewed by the operators involved in the procedure (at least two, J.F., P.A. or C.S.), as well as all the electrical measurements during implantation. We considered the duration of the procedure as the time from the first skin incision to the closure of the wound. Electrical parameters were analyzed at the end of the procedure to avoid hyperacute measurements during septal perforation. Threshold, sensing, impedance, and capture of LBB area were analyzed at hospital discharge, one month and six months after the procedure. We performed an ECG before every hospital discharge and at each patient's follow-up visit.

Complications from the procedure during follow-up were defined as loss of LBB area capture, macro-dislodgement (complete dislodgement from septum), infection (pocket, lead, endocarditis), atrial dislodgement, loss of R-wave sensing below 5 mV and increase in threshold above 1.5 V.

The endpoints of this study were the success of the procedure with capture of LBB area, visit to the emergency department or hospitalization for HF, loss of LVEF (defined as more than 10% reduction of LVEF) during follow-up and all-cause death.

The procedures were mainly performed by J.F. as the first operator, who also trained two additional operators (P.A. and C.S.). We used both styled-driven leads (SDL) and lumenless leads (LLL). The SDLs used in our registry were Solia S 60 (Biotronik, Berlin, Germany) and Ingevity (Boston Scientific, Marlborough, MA, USA) delivered using the Selectra 3D 55, 65 or 40 curve (39 or 42 cm catheter length) or Boston SSPC3 delivery system. The LLL used was the 3830-69 SelectSecure (Medtronic, MN, USA) delivered using the C-315 His sheath. The Solia S 60 lead was used in a few select cases if implantation with other SDLs or LLLs failed. Likewise, LLL was used in the few failed cases with Solia S 60 lead. Lead characteristics are shown in Table 1.

The implantation technique was performed as described in the 2023 European Heart Rhythm Association consensus document on CSP stimulation,13 with slight variations. First, subclavian access was secured with one (single-chamber device), two (dual-chamber devices) or three (CRT-D or left bundle branch-optimized CRT) guidewires according to the number of leads being implanted. We then advanced a lead to the RV apex for backup pacing, if the patient had LBBB, second or higher-degree atrioventricular (AV) block. We performed this step to have RV pacing ready if the patient developed asystole. Then, we advanced the SelectSecure 3D, C-315 or SSPC3 (curve is chosen according to right atrial and RV dimensions) into the RV, guided by a 0.035-inch J-tip guidewire positioned up to the RV apex, in a right anterior oblique (RAO) 20–30° projection. Using the technique described by Liu et al.,16 we then rotated the sheath with a slight counter-clockwise torque so that the end of the sheath was close to the septum and slowly retracted the sheath so that the tip was positioned below the tricuspid valve. We then visualized the tricuspid valve annulus (TVA) by injecting contrast medium (5–10 mL) through the sheath. Using the TVA summit as a reference, we then aimed to position the tip of the ventricular lead within a fan-shaped area drawn from the TVA summit considered the center of the circle, and the area considered between a radius of 15–35 mm and the angle ranging from +10° to −30°.16 This technique is also useful to avoid the entrapment of the septal tricuspid leaflet, which can be difficult to remove from the helix when a lead reposition is needed. Unipolar pace mapping was then performed to locate the optimal insertion site, aiming for discordance in lead II (positive or isoelectric) and lead III (isoelectric or negative), discordance in lead avR (negative) and lead aVL (positive) and, while not mandatory, a W pattern in V1. In the case of SDLs, we prepared the lead previously to insertion in the sheath by extending the screw, pre-tensioning and locking the lead using the funnel tool (Solia S) or lead end cap locked with stylet guide tool (Ingevity). The stylet was fully inserted and was connected to a cathodal clip to pace continuously while screwing. In the case of LLL, only intermittent pacing was performed. While screening by fluoroscopy in left anterior oblique (LAO) 30–40° projection, we rapidly rotated the lead in a clockwise motion, perpendicular to the septal plane while trying to maintain a stable and coaxial sheath position to guide the lead through the septum. We then continuously assessed lead depth by paced QRS morphology, fixation beat morphology, amplitude of sensed current of injury and unipolar pacing impedance. Perforation was diagnosed when the lead was easily moved into and off the left ventricle, rise in threshold or loss of capture during screwing and a marked fall in the current of injury amplitude and pacing impedance. In the case of perforation, the lead was pulled back, repositioned and screwed in a different septal position with similar initial pace mapping characteristics.

Criteria used in this study for confirmed or likely conduction system capture, as well as left ventricular septal pacing (LVSP) and deep septal pacing (DSP), not fulfilling criteria for LBBAP, were based on EHRA CSP consensus.13Confirmed LBB capture was diagnosed when we visualized a clear and sudden QRS transition with decreasing unipolar pacing output, V6 R-wave peak time (V6RWPT) <75 ms (or <80 ms in case of left bundle or bifascicular block, non-specific intra-ventricular conduction delay or idioventricular escape rhythm) or peak R-wave in V6 to terminal R-wave interpeak interval >44 ms. Likely LBB capture was diagnosed with a V6RWPT <85 ms (or <100 ms in case of left bundle or bifascicular block, non-specific intra-ventricular conduction delay or idioventricular escaper rhythm). If none of these criteria were fulfilled, but the electrocardiogram showed a paced terminal r/R-wave in V1, LVSP was diagnosed. If all criteria were not fulfilled, DSP was diagnosed and the procedure was considered a failure. All measurements at implantation were performed by the same operator, J.F., using LabSystem Pro (Boston Scientific, Marlborough, MA, USA) recording system at 100 mm/s sweep speed. We measured QRS duration after a pacing spike, from the onset of the rapid QRS upslope. Measurements during follow-up were performed by J.F. and D.F., and all measurements and capture of LBB area were reviewed by J.F. All electrocardiograms were performed during asynchronous pacing with the LBBAP lead to avoid confusion.

All analyses were performed using IBM® SPSS Statistics version 28 (IBM, Armonk, New York). Categorical variables were expressed in frequencies and percentages and continuous variables were expressed as mean±standard deviation or median and interquartile range for variables with or without a normal distribution, respectively. Continuous variables were analyzed by using unpaired Student's t-test after confirming for Gaussian distribution by Kolmogorov–Smirnov and Shapiro–Wilk tests, otherwise we used the non-parametrical Mann–Whitney test. For related samples, analysis was performed using paired t-test (parametric) or Wilcoxon test (non-parametric). Friedman Q test was used for related non-parametric samples or repeated measurements and ANOVA for parametric samples. Bonferroni correction was applied in the event multiple comparisons were made. Categorical variables were analyzed with Pearson χ2 test (or Fisher's test). The level of significance was accepted at <0.05, and testing was two-sided.

Mean procedure time was 81.7 min with a mean fluoroscopy time of 13.6 min. Mean LVAT was 79 ms. Mean peak R-wave in V6 to terminal R-wave in V1 interpeak interval was 41.3 ms. Conduction system capture was evaluated as confirmed in 79.1% of patients and likely in 11.8%. Paced QRS axis was positive in II and III in 21.6%, positive/iso in II and iso/negative in III in 56.8% and negative in II and III in 21.6% of patients. Procedural and electrocardiographic findings are presented in Table 3.

Mean QRS duration was significantly lower after LBBAP (p<0.001), irrespective of the presence of LBBB (p<0.001) or wide baseline QRS (p=0.004). To a lesser extent, we observed the same in patients with RBBB on the baseline (Table 4).

No severe intraoperative or per-operative complications occurred. The most frequent per-operative complication was intraprocedural helix damage (10.3%). During a mean follow-up time of 5.8±3.3 months, loss of conduction system capture occurred in 5 patients and macro-dislodgement in 2 patients (Table 5).

Parameters at discharge are reported in Table 3. At 1 month of follow-up, pacing threshold increased to 0.75 V (IQR 0.3; p<0.001) whereas both impedance (526 Ω, IQR 215) and sensing threshold (13.6 mV, IQR 8.3) remained stable (Figure 1). At 6 months of follow-up had further increased to 0.9 V (IQR 0.3; p<0.001), remaining within acceptable intervals. Impedance (468 Ω, IQR 234) and sensing thresholds (12.3 mV; IQR 10.8) remained stable. No infections or lead dysfunction were reported.

Figure 1.

Ventricular thresholds, sensing and impedance during follow-up.

(0.31MB).

A total of 24 patients underwent LBB pacing for CRT, 6 (25.0%) was LBB-optimized (LOT-CRT) and 6 (25.0%) as a bailout technique. Most were male (79.2%) with a mean age of 74.1±8.4 years. Almost half had ischemic cardiomyopathy (45.5%) and mean LVEF was 30.5±6.9%. Mean QRS width was 166.8±29.5 ms and 70.8% had an LBBB pattern. On average, patients had 2.2±0.8 leads implanted. Procedure duration was 114.2±26.9 min and the fluoroscopy time was 20.9±7.3 min. Paced QRS was significantly shorter (118.6±19.2 ms; p<0.001), with a mean difference of −47.7±28.7 ms. Mean LVAT was 81.1±10.5 ms and mean V6-V1 interpeak interval was 38.4±10.7 ms. Conduction system capture was confirmed in 68.4% of patients and considered likely in 10.5%. Positive/iso in II and iso/negative in III was the predominant morphology (61.5%) followed by positive in II and III (23.1%).

The acute pacing threshold was 0.7 V (IQR 0.4), sensing was 15.4 mV (IQR 10.8) and impedance was 439.9±107 Ω. During follow-up, pacing and sensing thresholds saw no significant changes (p=0.223 and p=0.368 respectively). Impedance also remained stable (p=0.488). Loss of conduction system capture occurred in 2 patients.

We acknowledge some limitations in our study. First, our study was nonrandomized, observational, and conducted at a single center. Second, most of the implantations were performed by the same operator. However, the same implantation technique was taught to other participating operators during the study and therefore the same technique was used in every procedure, with minor variations. Third, introducing new operators during the study led to outcomes such as higher procedure and fluoroscopy times. However, the same times were shortened as operators developed more experience. Fourth, we mostly used SDLs and, as such, had a small number of patients implanting LLL, therefore, the contribution of LLL to our results is underpowered. Fifth, our follow-up time was short. A longer follow-up time is needed to ascertain clinical outcomes, adverse events and long-term lead electrical performance.