Long-Term Safety and Feasibility of Left Bundle Branch Pacing in a Large Single-Center Study

Authors, Journal, Affiliations, Type, DOI

• Authors: Lan Su and Songjie Wang (joint first authors); Shengjie Wu; Lei Xu; Zhouqing Huang; Xiao Chen; Rujie Zheng; Limeng Jiang; Kenneth A. Ellenbogen; Zachary I. Whinnett; Weijian Huang (corresponding)
• Journal: Circulation: Arrhythmia and Electrophysiology 2021;14(2):e009261. Received 13 August 2020; accepted 14 December 2020. Open access (CC BY-NC)
• Affiliations: Department of Cardiology, The First Affiliated Hospital of Wenzhou Medical University; The Key Lab of Cardiovascular Disease of Wenzhou; Virginia Commonwealth School of Medicine, Richmond; National Heart and Lung Institute, Imperial College London
• Type: Prospective single-centre registry — 632 consecutive patients in whom LBBP was attempted (April 2017 – July 2019); 618 successful (97.8%); mean follow-up 18.6 ± 6.7 months, with 231 patients followed beyond 2 years
• DOI: 10.1161/CIRCEP.120.009261
• Funding: Key Research and Development Program of Zhejiang (2019C03012); Major Project of the Science and Technology of Wenzhou (ZS2017010). Disclosures: none

Overview

This is the largest long-term LBBP cohort published to date, and it is the study the field cites for the claim that LBBP thresholds are low and stable. In 618 consecutive successful implants followed for a mean of 18.6 months, unipolar capture threshold was 0.65 ± 0.27 V at implant and 0.69 ± 0.24 V at 2 years — essentially unchanged — while sensed R waves rose and impedance settled after the first month. In LBBB with heart failure (n=88), paced QRS fell from 167 to 124 ms (a 43 ms reduction, larger than the ~26 ms typical of biventricular pacing) and LVEF rose from 48.8% to 58.1% in patients with baseline QRS ≥120 ms. The safety picture is mixed but favourable: no procedure-related death, septal haematoma, coronary injury or LV thrombosis, two intraprocedural and one late septal perforation, two lead revisions for dislodgement, and a permanent RBB injury rate that the paper reports inconsistently (39 vs 55 patients). The most important caveat is that selective capture fell from 74.4% at implant to 30.9% at follow-up — the state the capture literature treats as the target is largely transient.

Keywords

Atrioventricular block; Bundle branch block; Bundle of His; Cardiac pacing; Cardiac resynchronization therapy; Complications; Heart failure

Key Takeaways

Cohort and success

• 632 consecutive patients attempted; 618 successful (97.8%). The 14 failures were attributed to septal scarring or hypertrophy
• The first 30 learning-curve patients were excluded — recruitment began after 30 LBBP procedures
• Indications: LBBB with HF 88 (14.2%); AVB or AF with AV node ablation 371 (60.0%) — AV nodal block 255, infranodal 116; SSS 159 (25.7%)
• Baseline: age 69.92 ± 11.19; male 56.0%; LVEF 57.12 ± 16.47%; LVEDD 53.67 ± 8.00 mm
• Procedure time 86.4 ± 43.5 min; fluoroscopy for lead placement 5.1 ± 4.6 min

Conduction and activation

• Overall paced QRS 112.94 ± 16.81 ms vs baseline 114.20 ± 32.40 ms (P = 0.305) — effectively unchanged
• LBBB: QRS 167.22 ± 18.99 → 124.02 ± 24.15 ms (P < 0.001) — a 43 ms reduction
• AVB/AF + AVN ablation: QRS 108.41 ± 26.69 → 111.35 ± 14.41 ms (P = 0.030)
• SSS: QRS 98.75 ± 19.13 → 110.59 ± 14.83 ms (P < 0.001) — the expected widening from delayed RV activation
• Sti-LVAT: all 73.87 ± 11.36 ms; LBBB 79.40 ± 11.34; AVB 73.78 ± 11.89; SSS 71.04 ± 8.80
• QRS axis was preserved overall (20.50° → 20.00°), and left-axis deviation in LBBB was corrected; the QRS transition zone rotated counter-clockwise
• LBB potential recorded in 476/618 (77.0%) — LBBB 54.5% (recorded during corrective His pacing in 37 two-lead cases and during RBBB-escape from LBB fascicles in 11), AVB 76.3%, SSS 92.5%
• ⚠ Selective capture at implant is reported as 460 (74.4%) in the Results and Table 2 but 466 (75.4%) in the Abstract. At follow-up, selective capture was present in only 191 (30.9%)
• Sti-LVAT shortened abruptly in 533 (86.2%) — the authors note this can be elicited "in almost all patients by early and multiple tests at the same site"

Pacing parameters over time

• The threshold curve is flat, which is the study's central claim and the main contrast with HBP
• ⚠ The 6-month R-wave SD of 65.35 mV is implausible (the other SDs are 5–6 mV) and is almost certainly a transcription error for 5.35; it undermines confidence in the table's transcription generally

Echocardiography

• At 1 year (n=560): LVEF 57.08 ± 16.60 → 62.36 ± 12.20% (P < 0.001); LVEDD 52.27 ± 7.51 → 50.73 ± 6.71 mm (P < 0.001)
• In baseline QRS ≥120 ms: LVEF 48.82 ± 17.78 → 58.12 ± 13.04% (P < 0.001)
• Tricuspid regurgitation: worsened by ≥1 grade in 62 (11.1%) and by ≥2 grades in 7 (1.3%); unchanged in 248 (44.3%); improved in 176 (31.4%)
• No echocardiographic evidence of mechanical perforation or valve leaflet damage

Complications

• No procedure-related death, cardiac arrest, septal haematoma, coronary artery injury, or LV thrombosis
• Septal perforation: 2 intraprocedural (detected by LBB/ventricular COI change plus sudden impedance drop; both repositioned without sequelae) and 1 during follow-up (required lead removal and reimplantation)
• Lead dislodgement with complete loss of capture: 2 — both to the RV the day after implantation, both revised
• Loss of LBB capture with preserved RV septal capture: 2, attributed to dislodgement to the RV septum (QS in V1, QRS >120 ms)
• Threshold increase >1 V: 11 patients; >3 V/0.5 ms: 6 patients, of whom local LV septal capture remained at 1.00 ± 0.16 V with Sti-LVAT 92.00 ± 6.72 vs 78.00 ± 7.77 ms and QRS 123.00 ± 10.10 vs 104.83 ± 7.86 ms compared with LBB capture
• Transient complete AV block during implantation: 42 (6.7%), of whom 26 had baseline LBBB; permanent complete AV block during follow-up: 16, of whom 12 had LBBB
• Pocket infection 2; haematoma 1; intravenous puncture arterial injury 2
• Deaths: 13, causes in the supplement

Right bundle branch injury — the inconsistency

• The Abstract states permanent RBB injury in 55 (8.9%)
• Table 4 states transient RBB injury 84 (13.6%) and permanent 39 (6.3%)
• The Discussion states transient injury in 20.4% and persistence in 8.9%
• Three different numbers for the same event. The likely reading is that 55 is the total with any RBB injury and 39 the permanent subset, or vice versa, but the paper does not reconcile them
• Mechanism: there is no clear anatomical demarcation between the distal His and the RBB on the right septum, and the RBB is thin as it crosses the tricuspid valve, so mapping near the distal His can injure it. Recommended avoidance: do not place the lead where a right bundle potential is recorded or where mapping caused transient RBB injury; in LBBB, place a temporary backup pacing lead before mapping
• RBB injury distribution: AVB 11.5% transient / 6.3% permanent; AF + AVN ablation 20.7% / 12.9%; SSS 20.1% / 5.7%; LBBB with HF — no transient RBBB cases recorded, but 29.5% transient and 13.6% permanent complete AV block

What the authors conclude

• LBBP is feasible (97.8%), preserves physiological LV activation in narrow QRS, normalises it in LBBB, has stable low thresholds, and has low long-term complication rates
• The LBBB QRS reduction (43 ms) compares favourably with biventricular pacing (~26 ms)
• The mechanism of the modest QRS widening in narrow-QRS patients is delayed RV activation, not loss of LV synchrony — the QRS axis is preserved
• Selective capture is lost over time because the local myocardial capture threshold falls after implantation, not because LBB capture is lost (the LBB threshold stays flat) — which is why routine pacing outputs later capture local myocardium
• The lead's proximity to septal myocardium provides a built-in backup: if LBB capture is lost, local LV septal capture at a lower output still gives a relatively narrow QRS

Limitations of the document

Authors' own stated limitations

• Single-centre prospective registry
• A relatively high rate of incomplete follow-up, attributed to COVID-19 and poor compliance
• Randomised multicentre trials are needed to verify long-term safety and clinical benefit
• The high success rate and proximal LBBP were achieved by implanters highly experienced in the technique

Additional limitations evident in the study

• No control group. All comparisons — to HBP thresholds, to biventricular pacing's QRS reduction, to RV pacing reintervention rates — are cross-study and historical, not randomised
• Follow-up is incomplete and selective. 560/618 (91%) at 1 year but only 231/618 (37%) at 2 years, so the "2-year stability" claim rests on a surviving, compliant subset. The authors acknowledge incomplete follow-up but not its potential to bias the threshold and echocardiographic trends
• Three internal inconsistencies undermine specific numbers: selective capture 460 vs 466; permanent RBB injury 39 vs 55; transient RBB injury 13.6% vs 20.4%. A 6-month R-wave SD of 65.35 mV is also implausible. These are transcription-level errors, but they matter in a paper that is the reference source for long-term safety
• The capture definition remains ECG-based. "Strict criteria" are applied, but there is no core laboratory, no direct electrophysiological confirmation, and 86.2% had an abrupt Sti-LVAT change — the authors themselves note the transition can be elicited in almost all patients with repeated testing, which weakens its specificity as a capture criterion
• The learning curve is excluded (first 30 cases), so the 97.8% success rate is a post-learning-curve figure and not what a new centre should expect
• The LBBB/CRT subgroup is small (88 patients, 14.2%) and uncontrolled; the LVEF improvement and QRS reduction are consistent with, but do not prove, a CRT effect. There is no comparison with biventricular pacing in the same cohort
• Echocardiographic changes are unadjusted. LVEF and LVEDD improvements could reflect regression to the mean, natural history, or concomitant medical therapy, and no core lab or blinded analysis is reported
• TR findings are observational and bidirectional — 11.1% worsened, 31.4% improved — and the improvement mechanism (resynchronisation, restored AV sequence) is speculative
• No lead extraction data, despite extraction being a recognised concern for lumenless leads in this position
• No cost, quality-of-life, heart-failure hospitalisation, or mortality analysis; the 13 deaths are listed but not analysed
• The "backup septal capture" reassurance is a double-edged finding: six patients lost LBB capture and continued with LV septal pacing, which is precisely the LVSP state whose clinical equivalence to LBB capture remains unproven
• Rating rationale: rated high. It is the largest prospective long-term LBBP cohort, with a clear, clinically important finding — flat thresholds to 2 years, normalisation of LBBB activation, and a low serious-complication rate — and it is the primary source for several claims the wiki carries. It is not very high because it is single-centre, uncontrolled, substantially incomplete at 2 years, ECG-defined for capture, and internally inconsistent on two of its headline safety numbers

Key Concepts Mentioned

• concepts/Left-Bundle-Branch-Area-Pacing — the technique; this is the largest long-term safety and feasibility cohort
• concepts/Conduction-System-Pacing — the umbrella category
• concepts/His-Bundle-Pacing — the comparator whose threshold instability LBBP is claimed to overcome
• entities/CRT — the LBBB-with-HF subgroup and the QRS-reduction comparison with biventricular pacing
• LBB capture threshold stability — 0.65 V at implant to 0.69 V at 2 years
• Selective vs nonselective capture — 74.4% selective at implant, 30.9% at follow-up, attributed to a fall in the local myocardial threshold
• Right bundle branch injury — the procedural hazard of mapping near the distal His; transient and permanent forms
• Septal perforation — 2 intraprocedural, 1 late; detected by COI change and impedance drop
• LV septal backup pacing — local capture at lower output when LBB capture is lost
• Tricuspid regurgitation — worsened in 11.1%, improved in 31.4%
• Sti-LVAT — 73.87 ms overall, longer in LBBB (79.40 ms)

Key Entities Mentioned

• concepts/Left-Bundle-Branch-Area-Pacing — primary concept page
• The First Affiliated Hospital of Wenzhou Medical University — the single investigating centre
• Weijian Huang; Lan Su; Shengjie Wu — the Wenzhou group
• Kenneth A. Ellenbogen (VCU); Zachary I. Whinnett (Imperial College London) — co-authors
• Medtronic SelectSecure 3830 and C315 His sheath — the implant system
• American Heart Association / Wolters Kluwer / Circulation: Arrhythmia and Electrophysiology — publisher

Wiki Pages Updated

• Created: wiki/sources/longterm-lbbap-circep-2021.md
• Updated: wiki/concepts/Left-Bundle-Branch-Area-Pacing.md
• Updated: wiki/concepts/Conduction-System-Pacing.md
• Updated: wiki/entities/CRT.md
• Updated: wiki/sourceindex.md
• Updated: wiki/wikiindex.md