Conduction System Pacing Compared With Biventricular Pacing for Cardiac Resynchronization Therapy in Patients With Heart Failure and Mildly Reduced Left Ventricular Ejection Fraction (I-CLAS)

Authors, Journal, Affiliations, Type, DOI

• Authors: Pugazhendhi Vijayaraman, Francesco Zanon, Shunmuga Sundaram Ponnusamy, Bengt Herweg, Parikshit Sharma, Manuel Molina-Lerma, Marek Jastrzębski, Zachary Whinnett, Kevin Vernooy, Rajeev K. Pathak, Roderick Tung, Gaurav Upadhyay, Karol Curila, and colleagues
• Journal: Heart Rhythm 2025;22(6):1512–1522
• Affiliations: 16 international centres — Geisinger Heart Institute (Wilkes-Barre, PA), Santa Maria della Misericordia Hospital (Rovigo), Velammal Medical College (Madurai), USF Morsani (Tampa), Rush University (Chicago), Hospital Universitario Virgen de las Nieves (Granada), Jagiellonian University (Kraków), Imperial College (London), Maastricht UMC, Australian National University, Banner Health (Phoenix), University of Chicago, Charles University (Prague), Baylor College of Medicine (Houston), VCU Health (Richmond), Hospital Universitari i Politècnic La Fe (Valencia)
• Type: Retrospective, multicentre, international, observational case-control study
• DOI: 10.1016/j.hrthm.2024.09.030

Overview

This I-CLAS analysis compares conduction system pacing (CSP — His bundle pacing or left bundle branch area pacing) with conventional biventricular pacing (BVP) in 1,004 patients with heart failure and mildly reduced ejection fraction (HFmrEF, LVEF 36%–50%) who underwent CRT at 16 centres between 2018 and 2023. CSP was associated with a 36% relative reduction in the composite of death or heart failure hospitalisation, driven almost entirely by fewer heart failure hospitalisations; all-cause mortality did not differ. In the one-third of patients with baseline LBBB the composite endpoint reduction was larger (HR 0.49). Paced QRS was narrower and capture thresholds were lower and more stable with CSP, while echocardiographic LVEF improvement was similar in both arms. The study is the HFmrEF component of the I-CLAS programme and supplies the mid-range ejection fraction stratum that the individual I-CLAS reports later combined with HFrEF data.

Keywords

Biventricular pacing; Cardiac resynchronization therapy; Conduction system pacing; His bundle pacing; Left bundle branch area pacing; HFmrEF; Heart failure hospitalization; Mortality

Key Takeaways

Study design and population

• Retrospective, multicentre, international, observational case-control study of consecutive patients undergoing BVP or CSP for CRT at 16 centres, January 2018–June 2023.
• Inclusion: NYHA II–IV, LVEF 36%–50%, and either LBBB or expected ventricular pacing >40%.
• Exclusions: age <18, pre-existing CRT device, failed CRT by either strategy, defibrillator indication, <6 months follow-up, and patients receiving HOT-CRT or LOT-CRT.
• 1,004 patients met criteria: BVP 178, CSP 826 (HBP 154, LBBAP 672).
• Mean age 73 ± 13 years; 34% women; mean LVEF 42% ± 5%; mean baseline QRS 145 ± 33 ms.
• CRT indication: AV block or AV node ablation 640 (64%), LBBB with dyssynchrony 214 (21%), pacing-induced cardiomyopathy 150 (15%).
• LBBB by Strauss criteria in 33% (335 patients).
• The BVP group had more men, more ischaemic cardiomyopathy, wider QRS and more LBBB; the CSP group had more women and non-ischaemic cardiomyopathy.

Procedural results

• Procedure duration shorter with CSP: 86 ± 43 vs 115 ± 48 minutes (P < .001).
• Fluoroscopy shorter with CSP: 14 ± 13 vs 19 ± 16 minutes.
• Paced QRS narrower with CSP: 129 ± 21 vs 144 ± 19 ms (P < .001).
• Implant capture threshold lower with CSP: 0.8 ± 0.5 V vs 1.2 ± 0.7 V at 0.5 ms (P < .001).
• LBBAP thresholds were lower than HBP both at implant (0.7 ± 0.3 V at 0.45 ms vs 1.1 ± 0.7 V at 0.7 ms) and at follow-up.
• Follow-up thresholds rose slightly in the BVP group (1.5 ± 1.0 V at 0.6 ms; P = .002) but remained stable with CSP (0.8 ± 0.5 V at 0.45 ms; P = .45).
• Procedural complications: 5.1% BVP vs 4.1% CSP (P = .57); lead revisions 3.9% vs 3.9%.
• Ventricular pacing burden was similar: 96% ± 10% vs 97% ± 6%.

Primary and secondary outcomes

• Primary composite of death or HFH: 22% CSP (180/826) vs 34% BVP (60/178); HR 0.64 (95% CI 0.43–0.94), P = .025.
• HFH: 13% CSP vs 23% BVP; HR 0.64 (95% CI 0.47–0.86), P = .003.
• All-cause mortality: 13% CSP vs 17% BVP; HR 0.88 (95% CI 0.59–1.32), P = .56 — not significant.
• In LBBB (n = 335): primary composite 18% CSP vs 32% BVP; HR 0.49 (95% CI 0.30–0.82), P = .006.
• In LBBB, HFH 11% vs 19% (HR 0.61, 95% CI 0.31–1.19) and death 10% vs 17% (HR 0.63, 95% CI 0.29–1.36) were individually non-significant.
• HBP vs LBBAP within CSP: no difference in the primary composite (31% vs 20%; HR 0.95, 95% CI 0.66–1.36, P = .77).

Echocardiographic outcomes

• Echo follow-up available in 785 patients (79%), mean 21 ± 15 months.
• LVEF improved in both arms: CSP 42% → 49% (P < .001); BVP 41% → 48% (P < .001).
• Change in LVEF was similar: +7.1% ± 11% vs +7.5% ± 10% (P = .86).
• In LBBB, LVEF change was +9% ± 10% vs +7% ± 10% (P = .31).
• LV end-diastolic diameter decreased modestly in both arms.

Arrhythmic events

• New-onset AF >30 seconds: 5.2% CSP vs 12.4% BVP (P < .001).
• Sustained VT/VF: 1.1% CSP vs 5.1% BVP (P < .001).
• Non-sustained VT: 5.7% vs 15.2% (P < .001).
• ICD upgrades similar: 2.2% vs 3.4% (P = .35).

Limitations of the document

• Non-randomised, on-treatment comparison. Treatment was chosen by operator/institution; patients who failed one strategy could cross to the other, and the true success rates of each approach were not analysed.
• Large baseline imbalance. Groups differed in size (826 vs 178), sex, ischaemic aetiology, baseline QRS and LBBB prevalence; no propensity matching was performed in this HFmrEF report, unlike the companion ≤50% analysis.
• Selection bias and unmeasured confounding are possible despite multivariable adjustment.
• Retrospective outcome ascertainment. Heart failure hospitalisations outside the implanting institutions may not have been captured.
• Follow-up is observational and relatively short (mean ~29 months) and all data were censored at a fixed date.
• HFmrEF is defined by a single baseline LVEF (36%–50%); LVEF fluctuates, and the group is biologically heterogeneous.
• No defibrillator-indication patients and no HOT-/LOT-CRT patients, limiting generalisability to the full CRT population.
• LVEF was measured locally, not in a core laboratory, and echocardiographic follow-up was incomplete (79%).
• Industry conflicts are extensive among the author list (Medtronic, Abbott, Boston Scientific, Biotronik, MicroPort honoraria and research support).

Key Concepts Mentioned

• concepts/Conduction-System-Pacing — the intervention being compared with BVP
• concepts/Left-Bundle-Branch-Area-Pacing — the dominant CSP modality (672/826)
• concepts/His-Bundle-Pacing — the other CSP modality (154 patients)
• concepts/Cardiac-Resynchronization-Therapy — the clinical context
• entities/Heart-Failure — the study population
• concepts/LBBB-Criteria — the subgroup with the largest treatment effect

Key Entities Mentioned

• International Collaborative LBBAP Study (I-CLAS) Group — multicentre consortium
• Geisinger Heart Institute — coordinating centre
• Medtronic, Abbott, Boston Scientific, Biotronik, MicroPort — device manufacturers (disclosed conflicts)

Wiki Pages Updated

• Created: wiki/sources/iclas-hfmref-hrs-2025.md
• Updated: wiki/concepts/Conduction-System-Pacing.md
• Updated: wiki/concepts/Left-Bundle-Branch-Area-Pacing.md
• Updated: wiki/entities/CRT.md
• Updated: wiki/sourceindex.md
• Updated: log.md