Procedural Endpoint in Left Bundle Branch Area Pacing: Conduction System Capture Versus LV Septal Capture (COMPARE-SYNC)

1. BACKGROUND AND RATIONALE

   Conventional right ventricular (RV) pacing produces a heterogeneous, left bundle branch block-like pattern of ventricular depolarization that results in electromechanical dyssynchrony. Over the long term this is closely associated with pacing-induced cardiomyopathy (PICM), heart failure, atrial fibrillation, and increased mortality. Conduction system pacing (CSP), which directly stimulates the native conduction system, was developed to overcome these limitations and comprises His bundle pacing (HBP) and left bundle branch area pacing (LBBAP).

   LBBAP, first described in 2017, has spread rapidly because it provides lower and more stable capture thresholds, superior sensing parameters, and a higher procedural success rate (approximately 91%) compared with HBP. LBBAP is an umbrella term for capture of the left ventricular subendocardial septal region and includes three subtypes:

   • Left bundle branch pacing (LBBP): capture of the pre-divisional left bundle, activating all fascicles simultaneously; LBB potential-to-QRS interval 34 to 25 ms; stimulus to V6 R-wave peak time (S-V6RWPT) <75 ms.
   • Left fascicular pacing (LFP): capture of an individual fascicle (left anterior, left septal, or left posterior) or its distal portion; potential-to-QRS interval <25 ms; paced QRS axis may change.
   • Left ventricular septal pacing (LVSP): capture of left ventricular septal myocardium without direct activation of the left conduction system; terminal R-wave present in lead V1; conduction system capture criteria not met.

   A substantial proportion of published LBBAP studies do not rigorously separate LBBP from LVSP, or define LBBAP success simply as reaching the septum, thereby mixing the two entities. In the Prague CSP trial (Heart Rhythm 2025), 12% of attempted LBBAP procedures resulted in deep septal pacing, and in the CSPOT study (Circ Arrhythm Electrophysiol 2024), 44% of procedures classified as successful LBBAP were in fact LVSP.

   Mechanistically, LBBP guarantees physiological ventricular depolarization by directly activating the left conduction system, whereas LVSP depends on myocardial conduction and may not provide the same degree of electrical synchrony. Conversely, LVSP offers advantages in procedural simplicity, threshold stability, and shorter procedure time, and several studies have found no significant difference in clinical outcomes between LVSP and LBBP. A prospective randomized controlled trial is therefore required to determine whether direct conduction system capture must be mandated during LBBAP, or whether achieving LVSP alone is clinically sufficient.

2. OBJECTIVE

   To compare the safety and clinical efficacy of a strategy that mandates direct left ventricular conduction system capture (LBBP or LFP) with a strategy in which left ventricular septal capture (LVSP) alone is accepted as the procedural endpoint, in patients undergoing LBBAP.

3. STUDY DESIGN

   Prospective, multicenter, randomized (1:1), parallel-group, single-blind (outcomes assessor blinded) controlled trial conducted at 5 or more centers in the Republic of Korea. Target enrollment is 300 participants (150 per group). All imaging and electrocardiographic endpoints are adjudicated by blinded central core laboratories.

4. RANDOMIZATION

   After written informed consent, participants are randomized 1:1 using a web-based central randomization system administered by the coordinating center (Severance Hospital). Stratification uses a single factor, baseline LVEF (<50% versus >=50%). Access to allocation information is maintained independently of the investigators and the sponsor.

5. STUDY DEVICE

   LBBAP is performed with the Ingevity Plus pacing lead (Boston Scientific Inc., Marlborough, MA, USA), an active-fixation helical lead with biomechanical properties suitable for LBBAP, implanted together with a commercially approved Boston Scientific pulse generator. Because the research question concerns the procedural endpoint strategy rather than a specific device, device-specific effects are acknowledged as a limitation.

6. PROCEDURE

   LBBAP is attempted in all enrolled participants. The procedural endpoint differs by allocation.

   LBBP mandatory group: the procedure continues until LBBP or LFP satisfying at least one conduction system capture criterion is confirmed. If conduction system capture cannot be obtained at the initial site, the lead is repositioned and the attempt repeated; more than three attempts are generally not recommended, although the final decision rests with the operator. Failure to achieve conduction system capture is recorded as a protocol deviation.

   LVSP enough group: once deep septal placement within the left ventricular subendocardial region is confirmed together with a terminal R-wave in lead V1 (LVSP), the operator may complete the procedure at his or her discretion irrespective of conduction system capture. Voluntary additional lead advancement in an attempt to obtain conduction system capture is permitted but not required.

   Conduction system capture criteria (at least one required):

   • Diagnostic QRS morphology transition during threshold testing: transition from non-selective to selective LBBP, or from LBBP to LVSP, as output is decreased, exploiting the higher conduction system capture threshold relative to the myocardial capture threshold.
   • Reproduction of the same transition pattern during programmed stimulation or stepwise output reduction.
   • S-V6RWPT <75 ms in patients with narrow QRS or simple right bundle branch block, or <80 ms in patients with advanced ventricular conduction disturbance such as left bundle branch block, nonspecific intraventricular conduction delay, or escape rhythm.
   • LBB potential-to-V6RWPT interval equal to the stimulus-to-V6RWPT interval (within 10 ms).
   • V6-V1 interpeak interval >44 ms.

7. CENTRAL CORE LABORATORIES

   Echocardiography core laboratory: The echocardiography laboratory of the Severance Cardiovascular Hospital serves as the central core laboratory, and all LVEF values used in analysis are core laboratory readings. Because the expected effect size (3% absolute) is comparable to the test-retest variability of Simpson biplane LVEF (3 to 5% even within a single reader) and inter-reader agreement can fall to an intraclass correlation coefficient of about 0.7 in a multicenter setting, central reading is essential to preserve statistical power. All sites follow a standard acquisition protocol requiring parasternal long-axis, parasternal short-axis, and apical 4-, 2-, and 3-chamber views, each stored as raw DICOM cine loops of at least three cardiac cycles at baseline and at 12 months. At the 12-month study, the sonographer is provided with the baseline images to reproduce identical views and angles. Baseline studies performed within 3 months before enrollment are acceptable, but the original DICOM data must be uploaded to the core laboratory within 2 weeks of enrollment. The core laboratory assigns an image quality grade (A, excellent; B, acceptable; C, inadequate) and requests repeat imaging when grade C is assigned. Core laboratory readers are blinded to treatment allocation, participating site, and acquisition time point; baseline and 12-month studies are read as pairs by the same reader with the time point order randomly presented. LVEF is measured by the Simpson biplane method, and LVESV and LVEDV are collected. If two readers differ by more than 5% in delta LVEF, a third reader adjudicates. Ten percent of the cohort is independently re-read for inter-reader intraclass correlation, and the same reader re-reads 10% for intra-reader intraclass correlation; both are reported. Optional substudies at capable centers include three-dimensional echocardiography (full-volume), global longitudinal strain by vendor-neutral speckle tracking, and myocardial work indices (global constructive work, global wasted work, global work efficiency) derived from noninvasive left ventricular pressure-strain loops.

   Electrocardiography core laboratory: Original 12-lead electrocardiograms obtained at baseline, immediately after implantation, and at 6 and 12 months are uploaded to a central ECG adjudication committee, which independently and blindly re-evaluates paced QRS duration, QRS morphology, and conduction system capture, operating independently of the investigators and the sponsor. The following intracardiac electrograms are mandatory and must be submitted: the electrogram at the moment of QRS morphology transition during threshold testing (including the LBBP to LVSP transition); real-time electrograms during threshold testing permitting measurement of S-V6RWPT and the V6-V1 interpeak interval; and electrograms documenting the presence and timing of the LBB potential. Surface electrocardiograms are recorded in 12 leads whenever possible and must include at least leads I, aVF, V1, V6, and aVL. All surface and intracardiac channels are recorded at 100 mm/s.

8. FOLLOW-UP SCHEDULE

   Screening (enrollment/procedure day): informed consent, demographics, physical examination, medical and surgical history, medications, blood tests, 12-lead ECG, echocardiography, and post-procedural device and lead measurements.

   Visit 1 (2 weeks to 1 month): medications, optional ECG, clinical follow-up. Visit 2 (6 months, window 3 months): medications, ECG, device and lead measurements, clinical follow-up.

   Visit 3 (12 months, window 3 months): medications, blood tests, ECG, echocardiography, device and lead measurements, clinical follow-up.

   Screening laboratory values obtained within the preceding 6 months are acceptable. No additional blood sampling is performed for research purposes beyond standard care.

   Collected variables include demographics (age, sex, weight, height, body mass index), vital signs, comorbidities (congestive heart failure, hypertension, diabetes mellitus, stroke or transient ischemic attack, coronary artery disease, peripheral vascular disease, atrial fibrillation or flutter or ventricular tachycardia, chronic kidney disease, liver disease, malignancy), surgical history (percutaneous coronary intervention, coronary artery bypass grafting, valve repair or replacement, ablation, transcatheter aortic valve implantation), laboratory tests (complete blood count, chemistry, NT-proBNP), echocardiographic measures (LVEF, LVESV, LVEDV, LVESD, LVEDD, left atrial volume index, E/E prime, mitral and tricuspid regurgitation severity), electrocardiographic measures (ventricular rate, PR interval, QRS morphology and duration, QT/QTc, rhythm), procedural variables (procedure date, success, procedure time, fluoroscopy time, final lead position and depth, QRS morphology transition, S-V6RWPT, V6-V1 interpeak interval, LBB potential, capture classification as LBBP/LFP/LVSP/DSP, intraprocedural complications), and post-procedural lead electrical measurements (capture threshold, sensing amplitude, impedance) with long-term maintenance of conduction system capture.

9. SAMPLE SIZE

   The sample size is based on the between-group comparison of change in LVEF at 12 months. With 80% power, a two-sided alpha of 0.05, an assumed absolute between-group LVEF difference of 3%, and a conservative standard deviation of 8.5% (published standard deviations of LVEF change in bradycardia pacing cohorts range from 7 to 9%; Abdelrahman et al. JACC 2018, Curtis et al. BLOCK-HF NEJM 2013, Curila et al. Prague CSP Heart Rhythm 2025), 130 completed participants per group are required. Allowing approximately 15% for dropout and missing data yields 150 per group, for a total of 300. The 3% minimal clinically important difference is supported by BLOCK-HF, in which an LVEF difference of approximately 3 to 4% between biventricular and RV pacing was associated with a significant reduction in death and heart failure hospitalization (hazard ratio 0.74, p=0.003).

10. STATISTICAL ANALYSIS

    Baseline characteristics are summarized descriptively; continuous variables are reported as mean with standard deviation when normally distributed and as median with interquartile range otherwise. Continuous variables are compared with the Student t test or the Wilcoxon rank-sum test as appropriate; categorical variables are reported as counts with percentages and compared with the Fisher exact test or the Pearson chi-square test.

    The primary analysis is an analysis of covariance with 12-month LVEF as the dependent variable and treatment group, baseline LVEF, and the stratification factor (baseline LVEF <50% versus >=50%) as covariates. Prespecified sensitivity analyses comprise (a) a mixed-effects linear model with site as a random effect (LVEF_12m ~ Group + LVEF_baseline + (1|Site)) to confirm robustness to multicenter clustering, (b) analysis restricted to the high-quality subset with completed core laboratory re-reading, (c) analysis restricted to image quality grade A, (d) stratification by equipment manufacturer, and (e) tipping point analysis for missing-not-at-random scenarios.

    Secondary clinical events (death, heart failure hospitalization, and others) are analyzed with Kaplan-Meier curves, the log-rank test, and Cox proportional hazards models. Learning curves for procedure and fluoroscopy time are modeled with cubic splines, and the change in slope with operator experience is tested by linear regression. All analyses are performed in both the intention-to-treat and per-protocol sets, with the intention-to-treat set used for the primary determination. Missing data are handled by multiple imputation. All tests are two-sided with p<0.05 considered statistically significant. Analyses use R version 4.0.3 or later (The R Foundation for Statistical Computing, Vienna, Austria).

11. SAFETY

    Pacemaker implantation as performed in this study constitutes standard care. Potential adverse events specific to LBBAP include interventricular septal perforation with pericardial effusion (incidence <1%), reintervention for lead dislodgement or malfunction (approximately 1.1% in LBBAP cohorts per meta-analysis and the 2025 ESC consensus), new moderate or greater tricuspid regurgitation (reported in up to 33% with basal lead positions; further study required), and upper extremity deep vein thrombosis, hemothorax or pneumothorax, and infection at rates equivalent to standard pacemaker implantation. In the LBBP mandatory group, repeated lead advancement attempts may prolong procedure and fluoroscopy time relative to the LVSP enough group, with a possible associated increase in procedural complication risk; this is monitored as a secondary safety outcome.

    Procedure-related major complications are defined as any death related to LBBAP; permanent loss of LBBAP lead function; need for lead reposition, replacement, explant, or revision; prolongation of hospitalization by 48 hours or more beyond the anticipated stay; hemothorax or pneumothorax confirmed by chest radiography or computed tomography; and interventricular septal perforation confirmed by echocardiography or fluoroscopy, whether intraprocedural or post-procedural, with concomitant pericardial effusion or tamponade recorded separately.

    Adverse events are monitored from the time of enrollment until completion of the final follow-up visit, death, study completion, or voluntary withdrawal. Serious adverse events, including deaths, are recorded in the electronic case report form and reported to the sponsor. An independent Data Safety Monitoring Board performs prespecified interim safety analyses and may recommend study termination.

12. WITHDRAWAL CRITERIA

Withdrawal of consent; a decision by the regulatory authority, a change in the opinion of the institutional review board, or a decision by the attending physician; a change in estimated life expectancy to less than 12 months; or an investigator determination that continued participation places the participant at risk.