CRT Risk Stratified by MRI and Optimized by Left Bundle Branch Pacing: Long-Term Follow-Up of MADURAI LBBP

Authors, Journal, Affiliations, Type, DOI

• Authors: Shunmuga Sundaram Ponnusamy, Vidhya Ganesan, Vadivelu Ramalingam, Habibullah Moghal, Saravana Kumar, Ramvivek Ramamoorthy, Thabish Syed, Mariappan Murugan, Pugazhendhi Vijayaraman
• Journal: Heart Rhythm 2025;22(8):2018–2027
• Affiliations: Department of Cardiology, Velammal Medical College (Madurai); Department of Microbiology, Velammal Medical College; Department of Radiodiagnosis, Velammal Medical College; Geisinger Heart Institute and Geisinger Commonwealth School of Medicine (Wilkes-Barre)
• Type: Prospective, single-centre, non-randomised observational registry
• DOI: 10.1016/j.hrthm.2025.04.040

Overview

The MADURAI LBBP study risk-stratified 225 patients with left bundle branch block–associated non-ischaemic cardiomyopathy (LB-NICM) and LVEF ≤35% using late gadolinium enhancement cardiac magnetic resonance (LGE-CMR). Patients with <10% scar burden received LBBP without a defibrillator (LOT-DDD or LOT-CRT-P); those with ≥10% scar received LBBP plus an ICD (LOT-ICD or LOT-CRT-D). Over a mean follow-up of 30 months, the low-scar group had a much lower composite of death, heart failure hospitalisation or sustained VT/VF (6.9% vs 26.1%; HR 16.52), a far higher echocardiographic response rate (78.2% vs 15.7%), and progressive LVEF normalisation (82.2% vs 25% at 2 years). Presence of any scar predicted adverse events even within the low-scar group; absence of scar predicted normalisation. The study supports an MRI-guided, cost-conscious CRT strategy in which many LB-NICM patients can safely avoid a defibrillator.

Keywords

Left bundle branch pacing; Left bundle branch block; LOT-DDD; LOT-ICD; LOT CRT-D; Late gadolinium enhancement; Heart failure

Key Takeaways

Population and risk stratification

• Prospective, single-centre registry from 2019; 231 of 296 CRT-D–eligible patients met inclusion criteria; 225 analysed after excluding 6 (4 opted out, 2 lost to follow-up).
• Inclusion: typical LBBB, LVEF ≤35%, NYHA II–IV, no prior myocardial infarction, no other identified cardiomyopathy cause, unknown duration of LBBB/LV dysfunction.
• LGE-CMR before implant; scar burden quantified on a 17-segment model; <10% = low risk (group I, n = 202, 90%); ≥10% = high risk (group II, n = 23, 10%).
• Group I: 93% received LOT-DDD (n = 188), 7% LOT-CRT-P (n = 14).
• Group II: 35% received LOT-ICD (n = 8), 65% LOT-CRT-D (n = 15).
• Mean age ~60 years; group I had more women (58% vs 34.7%; P = .03).
• Baseline LVEF 30.1% ± 4.2% (group I) vs 27.2% ± 4.7% (group II) (P = .002).
• Baseline QRS 164.4 ± 19.7 vs 167.4 ± 24.3 ms (P = .50).
• LGE present in 44/225 (19.5%) overall; in group I, 21 patients (10.4%) had 1%–10% scar.

Procedural results

• LBBP with a lumenless 3830 lead in all patients.
• Fluoroscopy longer in group II: 22.7 ± 3.2 vs 11.4 ± 8.5 min (P < .0001).
• CS lead used in 7% of group I vs 65% of group II (P < .0001).
• Final paced QRS 117.6 ± 10.8 ms (group I) vs 130.9 ± 16.7 ms (group II) (P < .0001).
• V6 RWPT ≤80 ms in 71% of group I vs 43% of group II (P = .006).
• Nonselective-to-selective transition during implantation in 86% vs 39% (P < .0001).
• Capture threshold, impedance and R-wave amplitude did not differ.
• One pneumothorax in group I; no haematoma, effusion or pocket infection.

Echocardiographic outcomes

• LVEF group I: 30.1% ± 4.2% → 52.1% ± 9.9% at 6 months → 55.1% ± 10.1% at 1 year → 56.7% ± 9.1% at 2 years.
• LVEF group II: 27.2% ± 4.7% → 37.1% ± 9.2% → 41.5% ± 9.8% → 47.4% ± 11.2%.
• Echocardiographic response (≥15% absolute LVEF gain at 6 months): 78.2% vs 15.7% (P < .0001).
• LVEF normalisation (≥50%): 64.8% vs 5.2% at 6 months; 74.6% vs 25% at 1 year; 82.2% vs 25% at 2 years (all P < .0001).
• LVESV reduction greater than LVEDV reduction in both groups.

Clinical outcomes

• Primary composite (death, HFH, or sustained VT/VF): 6.9% group I vs 26.1% group II; HR 16.52 (95% CI 3.41–80.28), log-rank P = .0005.
• All-cause mortality: 2.9% group I (malignancy 2, renal 2, heart failure 1, sepsis 1) vs 8.7% group II (progressive heart failure 2).
• No syncope, sudden cardiac arrest or arrhythmic death in group I.
• Sustained VT/VF in 4 group II patients (17.4%), all treated with appropriate ICD discharges; one inappropriate detection due to T-wave oversensing.
• ICD indication (persistent LVEF <35% or sustained VT/VF) at 1 year: 4.4% group I vs 31.2% group II (P = .0001); at 2 years 2.8% vs 25% (P = .0009).
• Two group I patients lost LBB capture and required repositioning; one group II patient had a CS threshold rise >1 V.

Predictors

• Normalisation of LV function in group I: smaller LV volumes, V6 RWPT ≤80 ms (OR 4.47), baseline LVEF 30%–35% (OR 3.27), absence of scar (OR 5.21).
• Adverse clinical events in group I: larger LV dimensions, higher baseline LVEF? (the table shows baseline LVEF OR 0.49, i.e. lower LVEF associated with events), and presence of scar (OR 5.24).
• Across the whole cohort, presence of scar was the only independent predictor of ICD therapy (OR 13.19, 95% CI 2.807–62.005).

Interpretation

• LB-NICM is a progressive conduction disease that can reverse substantially with LBBP; many patients with low scar burden may not need a defibrillator.
• LGE-CMR identifies the subgroup in whom scar drives arrhythmic risk and non-response.
• The MADURAI approach reduced hardware and cost without an apparent safety penalty in low-risk patients, though the authors note the 10% cutoff was arbitrary and that any scar predicted worse outcomes.

Limitations of the document

• Single-centre, non-randomised, prospective registry with no BVP control group — the comparison is scar-stratified, not treatment-randomised.
• Small high-risk group (n = 23), limiting precision for the group II estimates.
• The 10% scar cutoff was arbitrary, derived from prior studies; adverse events occurred even with any scar in group I.
• Procedural preference bias — the centre uses LBBP as its primary CRT approach, and the high success rate may not be reproducible elsewhere.
• Arrhythmia ascertainment relied on device algorithms; asymptomatic slower non-sustained VT may have been missed.
• LVEF and scar measurements were local, without core-laboratory adjudication.
• No long-term randomised comparison of the MADURAI strategy versus conventional BVP-ICD.
• Industry relationships: two authors report Medtronic consultancy and research support; one holds a His-bundle pacing delivery-tool patent.

Key Concepts Mentioned

• concepts/Left-Bundle-Branch-Area-Pacing — the pacing strategy
• entities/Heart-Failure — the study population
• concepts/Late-Gadolinium-Enhancement — the risk-stratification tool
• concepts/LOT-CRT — LBBP plus coronary sinus pacing, used in a subset
• concepts/Pacing-Induced-Cardiomyopathy — related reversible cardiomyopathy concept
• entities/ICD — the device question
• concepts/Cardiac-Resynchronization-Therapy — the clinical context

Key Entities Mentioned

• Velammal Medical College (Madurai) — coordinating centre
• Medtronic — 3830 SelectSecure MRI SureScan lead, C315His catheter, Sprint Quattro ICD lead
• Geisinger Heart Institute — collaborating centre

Wiki Pages Updated

• Created: wiki/sources/mri-lbbap-hrs-2025.md
• Updated: wiki/concepts/Left-Bundle-Branch-Area-Pacing.md
• Updated: wiki/concepts/Pacing-Induced-Cardiomyopathy.md
• Updated: wiki/entities/ICD.md
• Updated: wiki/sourceindex.md
• Updated: log.md