Physiology-Based Electrocardiographic Criteria for Left Bundle Branch Capture

Authors, Journal, Affiliations, Type, DOI

• Authors: Marek Jastrzębski, MD, PhD; Grzegorz Kiełbasa, MD; Karol Curila, MD, PhD; Paweł Moskal, MD, PhD; Agnieszka Bednarek, MD, PhD; Marek Rajzer, MD, PhD; Pugazhendhi Vijayaraman, MD
• Journal: Heart Rhythm 2021;18(6):935–43 (June 2021). © 2021 Heart Rhythm Society. All rights reserved
• Affiliations: First Department of Cardiology, Interventional Electrocardiology and Hypertension, Jagiellonian University Medical College, Kraków, Poland; Department of Cardiology, Charles University, Prague (Curila); Geisinger Heart Institute (Vijayaraman)
• Type: Two-centre diagnostic-accuracy study — 124 patients, 357 ECGs. Observational, but with formal diagnostic performance metrics (sensitivity, specificity, ROC-derived cut-offs)
• DOI: 10.1016/j.hrthm.2021.02.021
• ⚠ Filename is misdated by five years. The raw file is LBBCapture-HRS-2026.md, but the article of record is 2021;18(6):935–43 — the citation line inside the PDF reads "(Heart Rhythm 2021;18:935–943)" and the running header "Vol 18, No 6, June 2021". The slug uses the correct year; raw_path still points at the 2026-named file
• Funding and disclosures: partly supported by the Charles University Research Program Q38, UNCE/MED/002 and 260530/SVV/2020. Dr. Vijayaraman has received research and fellowship support and speaker/consultant fees from Medtronic — a manufacturer of conduction-system pacing leads

Overview

Every previously proposed ECG criterion for left bundle capture — QRS duration, V6 R-wave peak time — shared a defect the authors name directly: they were arbitrary, never validated, and of unknown sensitivity and specificity. This paper replaces them with a principle. Because LBB pacing depolarises the left ventricle through the physiological conduction system, the patient's own native QRS becomes the reference standard for their own paced QRS. Two equalities follow and are tested here: the paced V6 RWPT measured from QRS onset equals the native V6 RWPT, and the paced V6 RWPT measured from the stimulus equals the LBB-potential-to-V6-R-wave-peak interval. In 124 patients and 357 ECGs both held, yielding criteria with sensitivity 88.2–98.0% and specificity 85.7–95.4% — and a fixed cut-off of V6 RWPT 83 ms (sensitivity 84.7%, specificity 96.3%), with <74 ms conferring 100% specificity. The practical inversion is the memorable part: if the V6 RWPT gets longer when you pace, you have not captured the left bundle.

Keywords

Electrocardiographic criteria; Left bundle branch pacing; Left ventricular activation time; Physiological pacing; R-wave peak time

Key Takeaways

Background

• LBB pacing is developing rapidly, with potential application in both conventional pacing and heart failure
• Confirming true LBB capture — rather than capture of only the adjacent septal myocardium — is the central diagnostic problem
• Existing ECG criteria (QRS duration; V6 R-wave peak time, i.e. time from QRS onset to the peak of the R wave in V6) to distinguish nonselective LBB pacing from LV septal pacing were "arbitrary, have never been validated, and have unknown sensitivity and specificity"
• The hypothesis: because LBB pacing results in physiological depolarization of the LV, the native QRS can serve as a reference for diagnosing LBB capture in the same patient
• Gold standard used: dynamic ECG manoeuvres with output-dependent and refractoriness-dependent QRS morphologic changes

Methods

• Only patients with evidence of LBB capture (QRS morphology transition) were included
• Several QRS characteristics were compared between the native rhythm and different types of LBB area capture
• Concordant versus discordant morphology matters: discordance is defined as R/Rs in V6 during pacing with rS during native rhythm

Results — Population Analysed

357 ECGs from 124 patients:

Results — The Two Validated Equalities

• Paced V6 RWPT (measured from QRS onset) = native V6 RWPT ✓ positively validated
• Paced V6 RWPT (measured from the stimulus) = LBB potential to V6 R-wave peak interval ✓ positively validated
• Criteria built on these rules achieved sensitivity 88.2%–98.0% and specificity 85.7%–95.4%
• The first criterion's performance in detail: "paced V6 RWPT (measured from QRS onset) − native V6 RWPT" correctly diagnosed nonselective LBB capture in 98 of 100 patients with ns-LBB capture, and misdiagnosed ns-LBB in 3 of 35 patients with LV septal pacing
• The interval relationships: average potential-to-QRS interval 22 ± 7.0 ms; in LBBB patients average TCT 63.0 ± 8.2 ms and average IDT 128 ± 16.8 ms; LV septal paced V6 RWPT was closely approximated by the difference IDT − TCT (71.4 ± 11.4 ms vs 66.0 ± 12.4 ms; P = .19)

Results — Fixed Cut-off Values

• Optimal V6 RWPT for differentiating LBB capture from LV septal capture: 83 ms (sensitivity 84.7%, specificity 96.3%)
• V6 RWPT <74 ms conferred 100% specificity for LBB capture diagnosis
• Discrimination depended on the underlying substrate: cut-offs determined for diseased LBB patients versus narrow QRS/RBBB patients differ, and the appropriate one must be selected

Clinical Translation and Conclusion

• Equivalency of LV activation times on ECG during native and paced LBB conduction was demonstrated
• Therefore: if V6 RWPT is longer during pacing, this finding is indicative of lack of LBB capture
• LBB pacing is physiological pacing and can be recognised on the ECG using physiology-based criteria rather than arbitrary thresholds

Limitations of the document

Authors' own stated limitations

• Two centres only, and the results "might not be universally applicable due to population and implantation technique differences"
• The individualized approach can fail when the onset of the nonselective-LBB paced QRS cannot be determined precisely — for instance with signal oversaturation after the pacing stimulus extending beyond the latency interval
• Morphology-dependent error: during rS morphology, V6 RWPT is shortened by cancellation forces and probably reflects the timing of the local depolarization wavefront less accurately. Consequently discordant morphology (R/Rs in V6 during pacing, rS during native rhythm) may produce a false negative
• The cut-off derived for diseased LBB may not apply to focal damage in an otherwise healthy LBB (e.g. mechanical or ablation-related injury during a cardiac intervention) — for those patients the narrow QRS/RBBB cut-off should be used instead
• The TCT measurement method was relatively complicated and not validated by precise endocardial mapping. The authors argue its accuracy indirectly: the TCT value obtained (~60 ms) and the good diagnostic properties of the criterion based on it suggest the method was adequate

Additional limitations evident in the study

• The gold standard is itself criterion-based. LBB capture was defined by output- and refractoriness-dependent QRS morphology transition — a consensus/manoeuvre-based reference, not an independent anatomic or electroanatomic confirmation (no endocardial mapping). A diagnostic-accuracy study is only as good as its reference standard, and here the reference and the index test are both ECG phenomena
• Patient selection was circular with respect to the finding. Only patients with evidence of LBB capture were included, so the cohort cannot speak to the harder clinical question — how the criteria perform in an unselected implant population where capture is genuinely uncertain
• No comparison against the criteria it criticises. The paper states that existing QRS-duration and V6-RWPT criteria were arbitrary and unvalidated, but does not report head-to-head accuracy for those criteria in this dataset, so the claimed improvement is asserted rather than demonstrated
• Sensitivity/specificity are quoted as ranges across criteria (88.2–98.0% and 85.7–95.4%), which obscures which criterion performs worst and where each should be preferred
• The individualized "native QRS as reference" method requires a reliable native QRS — unavailable in patients paced at baseline, in complete AV block with wide native QRS, or where the native rhythm is itself ambiguous. This limits the technique precisely in the pacing-dependent population
• Industry relationship: one co-author (Vijayaraman) has research/fellowship support and consulting fees from Medtronic, a manufacturer of the leads used for this technique; disclosed
• Rating rationale: rated high — deliberately above the three sibling LBBAP sources ingested alongside it. It is two-centre, the largest cohort in the set (124 patients, 357 ECGs), and uniquely provides formal diagnostic performance metrics and ROC-derived cut-offs rather than descriptive feasibility. The reference-standard limitation above is why it is not very high

Key Concepts Mentioned

• concepts/Left-Bundle-Branch-Area-Pacing — the technique whose capture this defines
• concepts/Conduction-System-Pacing — the umbrella category
• V6 R-wave peak time (V6 RWPT) — the central index; measured either from QRS onset or from the stimulus, and the two measurements serve different criteria
• Native QRS as its own reference — the paper's core principle: physiological pacing should reproduce the patient's own activation time
• IDT (intrinsic deflection time) and TCT — the interval components used to approximate LV septal paced V6 RWPT (IDT − TCT)
• LBB potential — the reference for the stimulus-to-V6-peak equality
• Concordant vs discordant morphology — discordance (R/Rs in V6 when paced, rS native) risks false negatives
• LV septal pacing (LVSP) vs nonselective vs selective LBB capture — the three entities the criteria must separate

Key Entities Mentioned

• concepts/Left-Bundle-Branch-Area-Pacing — primary concept page
• Jastrzębski et al. — the group whose work the COI-LBBAP source explicitly contrasts itself against ("2 distinct transseptal ECG transition patterns"); this is that paper, closing the loop between the two sources
• Jagiellonian University Medical College, Kraków — lead centre; Charles University, Prague; Geisinger Heart Institute
• Medtronic — disclosed support to one co-author
• Heart Rhythm Society; Charles University Research Program — funder

Wiki Pages Updated

• Created: wiki/sources/lbb-capture-criteria-hrs-2021.md
• Updated: wiki/concepts/Left-Bundle-Branch-Area-Pacing.md
• Updated: wiki/sourceindex.md
• Updated: wiki/wikiindex.md