Key Points
Question
How does the performance of a new left bundle branch block (LBBB) criterion developed based on intracardiac recordings of patients compare with that of current strict criteria?
Findings
In this diagnostic study of 75 patients, LBBB pattern on surface electrocardiography (ECG) was categorized as being due to complete conduction block or intraventricular conduction delay based on assessment of intracardiac recordings. After correlation with surface ECGs, a time to notch in lead I longer than 75 milliseconds was proposed as a new ECG criterion for LBBB, and this criterion demonstrated improved specificity compared with conventional criteria in the derivation population as well as good performance in an independent validation population.
Meaning
Improving the specificity for LBBB criteria to identify patients with conduction block can aid in decision-making for corrective pacing procedures, including biventricular pacing or conduction system pacing.
Abstract
Importance
Current left bundle branch block (LBBB) criteria are based on animal experiments or mathematical models of cardiac tissue conduction and may misclassify patients. Improved criteria would impact referral decisions and device type for cardiac resynchronization therapy.
Objective
To develop a simple new criterion for LBBB based on electrophysiological studies of human patients, and then to validate this criterion in an independent population.
Design, Setting, and Participants
In this diagnostic study, the derivation cohort was from a single-center, prospective study of patients undergoing electrophysiological study from March 2016 through November 2019. The validation cohort was assembled by retrospectively reviewing medical records for patients from the same center who underwent transcatheter aortic valve replacement (TAVR) from October 2015 through May 2022.
Exposures
Patients were classified as having LBBB or intraventricular conduction delay (IVCD) as assessed by intracardiac recording.
Main Outcomes and Measures
Sensitivity and specificity of the electrocardiography (ECG) criteria assessed in patients with LBBB or IVCD.
Results
A total of 75 patients (median [IQR] age, 63 [53-70.5] years; 21 [28.0%] female) with baseline LBBB on 12-lead ECG underwent intracardiac recording of the left ventricular septum: 48 demonstrated complete conduction block (CCB) and 27 demonstrated intact Purkinje activation (IPA). Analysis of surface ECGs revealed that late notches in the QRS complexes of lateral leads were associated with CCB (40 of 48 patients [83.3%] with CCB vs 13 of 27 patients [48.1%] with IPA had a notch or slur in lead I; P = .003). Receiver operating characteristic curves for all septal and lateral leads were constructed, and lead I displayed the best performance with a time to notch longer than 75 milliseconds. Used in conjunction with the criteria for LBBB from the American College of Cardiology/American Heart Association/Heart Rhythm Society, this criterion had a sensitivity of 71% (95% CI, 56%-83%) and specificity of 74% (95% CI, 54%-89%) in the derivation population, contrasting with a sensitivity of 96% (95% CI, 86%-99%) and specificity of 33% (95% CI, 17%-54%) for the Strauss criteria. In an independent validation cohort of 46 patients (median [IQR] age, 78.5 [70-84] years; 21 [45.7%] female) undergoing TAVR with interval development of new LBBB, the time-to-notch criterion demonstrated a sensitivity of 87% (95% CI, 74%-95%). In the subset of 10 patients with preprocedural IVCD, the criterion correctly distinguished IVCD from LBBB in all cases. Application of the Strauss criteria performed similarly in the validation cohort.
Conclusions and Relevance
The findings suggest that time to notch longer than 75 milliseconds in lead I is a simple ECG criterion that, when used in conjunction with standard LBBB criteria, may improve specificity for identifying patients with LBBB from conduction block. This may help inform patient selection for cardiac resynchronization or conduction system pacing.
This diagnostic study aims to develop and validate a simple new criterion, time to notch longer than 75 milliseconds in lead I on electrocardiography, to identify left bundle branch block based on electrophysiological studies of human patients.
Introduction
Left bundle branch block (LBBB) was first identified as a clinical entity in canine models by Eppinger and Rothberger1 more than 100 years ago. Surface electrocardiography (ECG) criteria derived from this early canine work notably resulted in an accidental reversal of LBBB and right bundle branch block (RBBB) criteria in humans due to differences in anatomic orientation of the canine heart compared with the human heart, a misperception that persisted for decades. Perhaps surprisingly, these original criteria continued to inform multiple iterations of LBBB ECG criteria,2,3,4,5,6 culminating in the modern American College of Cardiology (ACC)/American Heart Association (AHA)/Heart Rhythm Society (HRS) ECG criteria for LBBB (QRS duration ≥120 milliseconds; a broad notched or slurred R wave in leads I, aVL, V5, and V6; absence of Q waves in leads I, V5, and V6; R-wave peak time >60 milliseconds in leads V5 and V6; and ST and T waves usually opposite in direction to the QRS complex).7,8 One limitation of these criteria is that they possess poor specificity for LBBB. Specifically, multiple studies have demonstrated that approximately one-third of patients who meet standard ECG criteria for LBBB in fact have intact His-Purkinje activation with intraventricular conduction delay (IVCD) or left ventricular (LV) hypertrophy masquerading as LBBB.9,10,11 This weakness was identified by Strauss et al12 when they formulated adjunctive criteria for LBBB based on mathematical modeling of conduction velocity in cardiac tissue (QRS duration ≥130 milliseconds in women or ≥140 milliseconds in men, a QS or rS pattern in leads V1 and V2, and mid-QRS notching or slurring in ≥2 of leads I, aVL, V1, V2, V5, and V6). However, the performance of the Strauss criteria was not evaluated against a gold-standard assessment of LBBB.
A second limitation of current criteria is the ambiguity of the individual components. For example, the ACC/AHA/HRS criteria for LBBB require that leads V5 and V6 should have an R-wave peak time longer than 60 milliseconds,7 but the criteria do not clearly elaborate on which peak should be used in a QRS complex with multiple R-wave peaks. Whether or not an individual criterion is met affects the final designation of LBBB or nonspecific IVCD, and this distinction is of growing clinical importance. In particular, multiple studies have shown that cardiac resynchronization therapy (CRT) is associated with greater benefit in patients with LBBB vs IVCD.13,14,15 Perhaps even more saliently, conduction system pacing (CSP) cannot restore electrical activation in patients with IVCD and intact His-Purkinje activation.16 This highlights the importance of accurately distinguishing between LBBB and IVCD for appropriate patient selection for CRT and CSP.
This study proposes a modification to the standard ECG criteria for LBBB based on observations from LV septal recordings of human patients and validates the performance of the modified criteria in an independent dataset of patients with complete conduction block (CCB). The aim was to develop a more reproducible method to allow for more accurate patient selection for CRT and CSP.
Methods
Study Design
In this diagnostic study, the criterion derivation cohort comprised patients who were prospectively enrolled in an electrophysiology registry between March 2016 and November 2019 after being referred to a single center (University of Chicago Medicine, Chicago, Illinois) for either ventricular tachycardia (VT) ablation or cardiac resynchronization device implant. These were subsequently compared with patients in an independent validation cohort (described later). All patients in this study met standard ACC/AHA/HRS criteria for LBBB at baseline on a preprocedural ECG aside from the criterion for R-wave peak time longer than 60 milliseconds in leads V5 and V6.7 This criterion was deemed not necessary for enrollment due to the issues with ambiguity and sensitivity discussed earlier. All patients provided written informed consent to undergo LV septal mapping during their procedure. Surface 12-lead ECG measurements and adjudication were performed by 2 physicians (J.S.T. and G.A.U.), and differences were resolved by consensus. Assessment of the derivation cohort was performed with concurrent analysis of intracardiac recordings as noted later. This study was approved by the University of Chicago Medicine institutional review board.
Intracardiac Recordings in the Derivation Cohort
LV septal mapping was performed as described previously.17 Briefly, access to the LV was obtained either in a transseptal approach in the case of VT ablation or in a retrograde fashion in the case of a device implant. Mapping was performed using a linear multipolar catheter positioned against the left side of the ventricular septum, and recordings were obtained and analyzed at 100-mm/s sweep speed using the CardioLab acquisition system (GE HealthCare). Left septal recordings were judged to have either CCB (defined as the sudden cessation of His-Purkinje potentials partway down the septum with a nonphysiologic basal-to-apical activation sequence of the ventricular myocardium) or intact Purkinje activation (IPA; defined as the presence of a complete activation sequence of His-Purkinje potentials down the length of the septum with subsequent physiologic apical-to-basal activation of the myocardium) as shown in Figure 1 (12-lead ECGs for the 2 patients in this figure are shown for reference in eFigure 1 in Supplement 1).
Figure 1. Mapping Left Ventricular (LV) Septal Activation Patterns.
Surface electrocardiography (ECG) recordings (top) and intracardiac mapping of LV septal activation (bottom) from patients who meet standard American College of Cardiology/American Heart Association/Heart Rhythm Society surface ECG criteria for left bundle branch block. A, Septal mapping in this patient shows His-Purkinje activation (highlighted in blue) with a site of interrupted conduction (red lines) as well as nonphysiologic activation of the septal myocardium from basal to apical. These findings are consistent with complete conduction block. B, This patient demonstrates intact basal to apical His-Purkinje conduction down the length of the LV septum (highlighted in blue) as well as physiologic apical to basal activation of the septal myocardium. These findings indicate intact Purkinje activation. Standard 12-lead ECGs for these 2 patients are shown in eFigure 1 in Supplement 1.
Criterion Development in the Derivation Cohort
To develop the new ECG criterion, surface ECGs of patients found to have CCB were compared with those found to have IPA. Prior work demonstrated that of all tested ECG criteria components for LBBB, the presence of QRS notching or slurring in the septal (V1, V2) or lateral (I, aVL, V5, V6) leads has the highest predictive value for CCB vs IPA.17 Within this dataset, patients with CCB often have notches or slurs late in the QRS complex compared with patients with IPA, who often have either no notch or a notch or slur early in the QRS complex. Therefore, the time was measured from QRS onset to the nadir of any notch present (or the midpoint of a slur), which is herein referred to as the notch time, for each of the septal and lateral leads in each patient (Figure 2). Receiver operating characteristic (ROC) curves for the time to notch vs the probability of CCB were then constructed. If multiple notches were present in a given lead, the time to the latest notch was found to have the best performance and thus was used. Absence of notch (time to notch = 0 milliseconds) and a global notch time criterion, wherein the time to notch was the period from QRS onset until the latest notch in any septal or lateral lead, were also considered. Calculations and curve construction were performed using MATLAB version R2020b (MathWorks). A Standards for Reporting of Diagnostic Accuracy (STARD) diagram showing patient flow and test results for this population was created (eFigure 2 in Supplement 1).
Figure 2. Measurement of Notch Times.
Time to notch is measured as the time from QRS onset to the nadir of the notch or midpoint of a slur. If multiple notches or slurs are present in the QRS complex, the latest one is used. The measurement shown demonstrates a time to notch of approximately 90 milliseconds in lead I.
Criterion Validation Cohort
Once the new criterion was established, an independent dataset was constructed to validate its performance. This was done by reviewing the preprocedural and postprocedural ECGs for every patient who underwent implant of a transcatheter aortic valve replacement (TAVR) between October 2015 and May 2022. As with the derivation cohort, all ECGs were adjudicated by 2 physicians (J.S.T. and G.A.U.) and differences were resolved by consensus. Mechanical impingement of the left bundle leading to LBBB is a known complication of TAVR,18 and there is no obvious mechanism by which a patient would acutely develop periprocedural IVCD. Accordingly, patients who developed marked QRS widening with LBBB morphology during TAVR implant were considered to have LBBB due to CCB following their TAVR (Figure 3). This served as a control group of patients with CCB and thus “true” LBBB for validation. A subset of patients with marked QRS widening after implant in fact met standard ECG criteria for LBBB prior to the procedure and before developing clear new QRS widening periprocedurally (eFigure 3 in Supplement 1). In this case, it was presumed that prior to the procedure they in fact had IVCD with IPA, then developed CCB on top of their IVCD following the procedure. Therefore, in this subset of patients, LBBB criteria could be applied to both the preprocedural IVCD ECGs as well as the postprocedural LBBB ECGs to assess criterion performance.
Figure 3. Periprocedural Left Bundle Branch Block During Transcatheter Aortic Valve Replacement Implant.
Example of preprocedural (A) and postprocedural (B) electrocardiograms (ECGs) for a patient who received a transcatheter aortic valve replacement implant. The postprocedural ECG shows a new marked widening of the QRS complex with new lateral QRS notching compared with the preprocedural ECG. This suggests that the patient developed a new left bundle branch block due to the transcatheter aortic valve replacement implant. This is a known complication of the procedure and is due to complete conduction block rather than intraventricular conduction delay.
Statistical Analysis
Continuous variables were assessed for normality using the Shapiro-Wilk test. Variables that were consistent with a normal distribution were reported as mean (SD), and statistical significance was assessed using a parametric t test. Continuous variables that were found to be inconsistent with a normal distribution were reported as medians with accompanying IQRs and analyzed for significance with the nonparametric Wilcoxon rank sum test. Categorical variables were reported as counts and percentages, and significance was determined using Fisher exact test for unpaired data and McNemar exact test for paired data. Tests were 2-tailed, and P < .05 was considered statistically significant. Sensitivity and specificity are reported with 95% CIs calculated using the Clopper-Pearson exact method. All calculations were performed using the MATLAB version R2020b software package.
Results
Study Population
Over the 3-year period, a total of 75 patients were enrolled in the derivation cohort: 37 for VT ablation and 38 for CRT. Baseline characteristics are shown in the Table. The median (IQR) age of this group was 63 (53-70.5) years, 21 (28.0%) were female, and the mean (SD) QRS duration was 164 (23) milliseconds. Within this population, 48 patients were found to have CCB, while 27 demonstrated IPA. The CCB and IPA subgroups were well matched overall, although the mean (SD) body mass index (calculated as weight in kilograms divided by height in meters squared) was found to be higher in the IPA subgroup than in the CCB group (32.1 [6.4] vs 28.9 [5.2], respectively; P = .03).
Table. Baseline Characteristics.
| Characteristic | Derivation cohort | Validation cohort with TAVR LBBB (n = 46) | |||
|---|---|---|---|---|---|
| Total (n = 75) | CCB (n = 48) | IPA (n = 27) | P value | ||
| Age, median (IQR), y | 63 (53-70.5) | 63 (50.8-69) | 62 (55.5-71) | .70 | 78.5 (70-84) |
| Sex, No. (%) | |||||
| Female | 21 (28.0) | 15 (31.3) | 6 (22.2) | .44 | 21 (45.7) |
| Male | 54 (72.0) | 33 (68.8) | 21 (77.8) | 25 (54.3) | |
| Race and ethnicity, No. (%)a | |||||
| Black | 24 (32.0) | 15 (31.3) | 9 (33.3) | >.99 | 17 (37.0) |
| Hispanic | 3 (4.0) | 2 (4.2) | 1 (3.7) | >.99 | 1 (2.2) |
| White | 46 (61.3) | 29 (60.4) | 17 (63.0) | >.99 | 27 (58.7) |
| Other | 2 (2.7) | 2 (4.2) | 0 | .53 | 1 (2.2) |
| Medical history, No. (%) | |||||
| Left ventricular ejection fraction, median (IQR), % | 27.5 (20.6-37.6) | 26.2 (21.0-32.5) | 29 (20.2-43.5) | .28 | 58.3 (43.2-63.5) |
| Nonischemic cardiomyopathy | 51 (68.0) | 31 (64.6) | 20 (74.1) | .45 | 10 (21.7) |
| Coronary artery disease | 28 (37.3) | 18 (37.5) | 10 (37.0) | >.99 | 26 (56.5) |
| History of myocardial infarction | 21 (28.0) | 14 (29.2) | 7 (25.9) | >.99 | 12 (26.1) |
| History of coronary artery bypass grafting | 9 (12.0) | 7 (14.6) | 2 (7.4) | .47 | 6 (13.0) |
| Diabetes | 26 (34.7) | 17 (35.4) | 9 (33.3) | >.99 | 21 (45.7) |
| Hypertension | 42 (56.0) | 26 (54.2) | 16 (59.3) | .81 | 45 (97.8) |
| Chronic kidney disease | 28 (37.3) | 21 (43.8) | 7 (25.9) | .14 | 23 (50.0) |
| Plasma creatinine, median (IQR), mg/dL | 1.1 (0.9-1.4) | 1.1 (0.9-1.4) | 1 (0.9-1.2) | .36 | 1 (0.7-1.3) |
| End-stage kidney disease | 2 (2.7) | 2 (4.2) | 0 | .53 | 3 (6.5) |
| BMI, mean (SD) | 30 (5.8) | 28.9 (5.2) | 32.1 (6.4) | .03 | 29.5 (7.0) |
Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); CCB, complete conduction block; IPA, intact Purkinje activation; LBBB, left bundle branch block; TAVR, transcatheter aortic valve replacement.
SI conversion factor: To convert creatinine to μmol/L, multiply by 88.4.
Data regarding race and ethnicity were determined based on self-report.
Formulating a New ECG Criterion for LBBB
Analysis of surface ECGs showed that patients with LBBB due to CCB tended to have more frequent notching or slurring in the QRS complexes of lateral leads and that these notches tended to occur later in the QRS complexes compared with patients with IPA. Figure 4A and eFigure 4 in Supplement 1 compare notch times in lead I for patients with CCB vs IPA. Patients with CCB were more likely to have a notch or slur in lead I (40 of 48 patients [83.3%]) than patients with IPA (13 of 27 patients [48.1%]) (P = .003), as shown in Figure 4B. By contrast, in the septal leads there was no increased likelihood of QRS notching in patients with CCB compared with patients with IPA. For example, lead V1 showed QRS notching in 16 of 48 patients (33.3%) with CCB and 7 of 27 patients (25.9%) with IPA (eFigure 5 in Supplement 1); this difference was not significant (P = .61). These data suggest that lateral leads may have more discriminatory power for LBBB than septal leads, and thus ECG criteria for LBBB may benefit from focusing on 1 or more of these leads.
Figure 4. Analyses of Lead I Notch Times in the Derivation and Validation Cohorts.
A, Box plots for lead I notch times in patients from the derivation group who were found to have complete conduction block (CCB) on electrophysiological study vs those with intact Purkinje activation (IPA). Notch times from patients in the transcatheter aortic valve replacement (TAVR) validation cohort with new left bundle branch block are also shown. The lower and upper borders of the boxes indicate the 25th and 75th percentiles, respectively; center horizontal line, median notch time; and whiskers, range. For this plot, patients without an identifiable notch or slur in lead I were assigned a notch time of 0 milliseconds. B, Percentage of patients who had an identifiable notch or slur in lead I on surface electrocardiography in the derivation CCB group (40 of 48 patients [83.3%]) compared with the derivation IPA group (13 of 27 patients [48.1%]) and the TAVR validation cohort (42 of 46 patients [91.3%]). The difference between the CCB and IPA groups was statistically significant. Derivation and validation cohorts were not statistically compared with each other as these represent 2 distinct populations.
To further investigate this hypothesis, ROC curves were constructed for each left-sided and septal lead as described in the Methods (eFigure 6 in Supplement 1). As expected, a longer time to notch (or time to latest notch, in the case of multiple notches) was associated with a greater probability of CCB vs IPA. For each ROC curve, a higher area under the curve (AUC) indicates that notch time in the associated ECG lead is a better overall discriminator of CCB vs IPA. Consistent with initial observations, these data show that the lateral leads offer better predictive value than the septal leads, which have curves lying near the nondiscrimination line and AUCs only slightly above 0.5. In particular, these data imply that notch time in lead I was the best overall predictor of CCB vs IPA, with an AUC of 0.77. Of note, a global notch time criterion was also considered, wherein the notch time would be taken to be the longest time from QRS onset to the last notch in any septal or lateral lead. However, this approach was found to have worse performance than using the notch times from lead I in isolation.
Having identified lead I as the lead with the greatest discriminatory power, various notch time thresholds for this lead were assessed. Based on these data, a notch time cutoff of longer than 75 milliseconds was felt to optimize the trade-off between sensitivity and specificity. This choice generated a criterion with a sensitivity of 71% (95% CI, 56%-83%) and a specificity of 74% (95% CI, 54%-89%) in this population. For comparison, the performance of the Strauss criteria was also assessed in this population and was found to have a sensitivity of 96% (95% CI, 86%-99%) but a specificity of only 33% (95% CI, 17%-54%). This lower specificity highlights the difficulty of distinguishing CCB from IPA in many patients even with application of the Strauss modifications to the current conventional criteria.
Validation of the Novel Criterion
This time-to-notch criterion was next validated on an independent dataset comprising patients who acutely developed periprocedural LBBB during TAVR implant. The records of 468 patients who underwent TAVR implant over the past 7 years were analyzed, and we identified 46 patients who developed acute periprocedural LBBB due to presumed CCB. Baseline characteristics for these patients are shown in the Table. The median (IQR) age was 78.5 (70-84) years, and 21 (45.7%) were female.
Surface ECGs for these patients were analyzed, and distributions of notch times in lead I are shown in Figure 4 and eFigure 4 in Supplement 1. It was found that 42 of these 46 postprocedural ECGs in the LBBB group (91%) had any notch or slur in lead I. This resulted in a sensitivity of 87% (95% CI, 74%-95%) for the new criterion in the validation population. The Strauss criteria performed similarly well in this population, correctly categorizing 41 of the 46 patients with postprocedural ECGs as having LBBB. This corresponds to a sensitivity of 89% (95% CI, 76%-96%).
Of the 46 patients in the validation cohort, 10 met standard criteria for LBBB prior to the procedure. As these patients all demonstrated marked acute periprocedural QRS widening consistent with periprocedural LBBB development, they were taken to have IVCD on their preprocedural ECGs as discussed in the Methods. Preprocedural and postprocedural ECGs for these 10 patients were analyzed for notching in lead I (eFigure 7 in Supplement 1). In this subset of patients, 4 of 10 patients (40.0%) had notching in lead I on the preprocedural ECG, compared with 10 of 10 patients (100%) with lead I notching on the postprocedural ECG (P = .03). Both the novel criterion and Strauss criteria correctly categorized all 20 ECGs in this limited dataset.
Discussion
The aim of this study was to develop a simple new surface ECG criterion to allow for more accurate discrimination between LBBB and IVCD. To our knowledge, this is the first proposed modification to the traditional ECG criteria for LBBB based on direct electrophysiological observations in human patients. It was found that late notching or slurring in lead I was the most predictive feature of LBBB compared with IVCD and that a threshold of time to notch longer than 75 milliseconds provided the optimal criterion for LBBB. Since the derivation cohort met the conventional ACC/AHA/HRS criteria for LBBB, the novel criterion was applied as a modification to these criteria (by replacing the R-wave peak time criteria in lead V5 or V6). In this population, the novel criterion offered substantially improved specificity compared with the Strauss criteria (74% vs 33%). As expected, this improved specificity comes at some cost to sensitivity, but sensitivity still remains greater than 70% in this set, which was considerably higher than that in an early report of sensitivity for the updated European Society of Cardiology criteria.19 An implication of this finding is that a shorter time-to-notch cutoff may improve sensitivity at the cost of specificity, and vice versa. Of note, in the derivation cohort, notching in the septal leads (eg, V1 through V2) had very little predictive value for discriminating CCB from IPA. Thus, it may be that some of the lesser specificity of current ECG criteria relates to incorporation of data from these leads in the criteria.
The new criterion was then tested in an independent validation cohort, and the sensitivity (87%) remained high in this cohort. On a small subset of patients with prior IVCD, the criterion had perfect accuracy in discriminating IVCD from LBBB. Overall, the performance of the new criterion was significantly better in the validation cohort than in the derivation cohort. A possible explanation is that the derivation and validation groups had different mechanisms of LBBB. The derivation group consisted of patients referred to the electrophysiology laboratory for both VT ablation and CRT implant. Many of these patients had diffuse cardiomyopathies of a variety of etiologies, and these would be expected to be able to cause conduction block at a variety of levels, including in the His bundle, left bundle proper, or distal conduction system. Additionally, many of these patients also had conditions, such as fibrosis or dilation and hypertrophy, that can mimic LBBB and make correct categorization more difficult.
By contrast, the TAVR validation population consisted of patients who all had essentially identical etiology of their conduction disease—mechanical impingement of the proximal conduction system—and overall had less diseased cardiac tissue. Thus, they represent a relatively homogeneous population who may tend to have a more archetypal LBBB surface ECG morphology compared with the patients in the derivation group. Additionally, IVCDs present in the validation cohort were overall less severe and less challenging to distinguish from LBBB than for patients in the derivation cohort. This hypothesis is supported by the fact that Strauss criteria also performed better in the validation cohort than in the derivation cohort. Similar findings were reported recently by Kawamura et al20 with respect to the performance of the Strauss criteria in a test set composed of patients who had received a TAVR and patients with CSP-correctable LBBB.
This work has clinical implications for patient selection for cardiac resynchronization. Prior work has demonstrated that wide QRS in patients with IPA cannot be corrected with CSP and patients may not clinically respond; thus, they should be given a traditional biventricular CRT system if cardiac resynchronization is attempted.17 Furthermore, when deciding whether to refer a patient for resynchronization, there is a substantial body of evidence suggesting that patients with LBBB pattern on ECG derive more benefit from biventricular CRT than patients with IVCD or RBBB patterns.12 The distinction may perhaps be even more relevant when contemplating CSP. Patients with wide QRS due to IVCD without underlying conduction block are less likely to benefit from CSP, and a traditional biventricular device or hybrid device (eg, His or left bundle pacing fused with CSP) may be more appropriate. The new criterion proposed here provides a simple yet specific assessment to assist in procedural planning. Put simply, patients without a time to notch longer than 75 milliseconds in lead I may be less likely to benefit from CSP.
Limitations
In prior work, several potential limitations regarding the intracardiac septal mapping procedure were discussed.17 These include preferential mapping of the posterior fascicle of the left bundle compared with the anterior fascicle when mapping with a retrograde aortic catheter, as well as difficulty distinguishing a septal branch of the left bundle when present. These concerns could potentially lead to incorrect categorization of CCB vs IPA. However, as discussed in the prior work, mapping correctly identified IPA in all tested control patients, suggesting adequacy of the mapping technique. Importantly, retrograde aortic or transseptal multielectrode mapping was not performed in the validation cohort of patients who had received a TAVR. Based on elegant work by Kawashima and Sato21 elucidating the anatomy of this location, mechanical compression leading to proximal conduction block and abrupt change in QRS with LBBB pattern appears most likely, but this is an assumption.
An additional potential limitation of this study is that the derivation and validation populations are quite distinct in terms of mechanism of conduction block as well as overall cardiac health. In particular, the validation cohort had a far lower percentage of patients with nonischemic cardiomyopathy, whereas the derivation cohort had a higher proportion of these patients. This limits the ability to directly compare the test characteristics in the 2 populations. However, the fact that the novel criterion performed well in a population markedly different from the population from which it was derived suggests that it may have broader application to a variety of cardiac patients. ECGs were reviewed with disagreements resolved by consensus, and interoperator variability could not be systematically assessed. Finally, this study is limited by the sample size of both the derivation and validation populations. This is especially relevant for the IVCD subgroup of the validation population, which precluded validation of the specificity of the criterion. This lessens the certainty regarding the values of the test characteristics found in this study.
Conclusions
In this work, intracardiac patterns of ventricular septal activation in human patients were analyzed to derive a simple new surface ECG criterion for LBBB. When used adjunctively with current ACC/AHA/HRS guidelines for LBBB, this criterion significantly improved specificity in the derivation cohort, with better testing performance than the Strauss criteria. In an independent validation population, the criterion continued to demonstrate high sensitivity. This work may have implications in choosing patients for CRT and CSP.
eFigure 1. 12-Lead ECGs for Patients Who Underwent Septal Mapping
eFigure 2. STARD Diagram
eFigure 3. Periprocedural LBBB in a TAVR Patient With Underlying IVCD
eFigure 4. Scatterplot of Lead I Notch Times in the Derivation and Validation Cohorts
eFigure 5. Analysis of Notching in Lead V1
eFigure 6. ROC Curves for Time-to-Notch Criterion
eFigure 7. Analyses of Lead I Notch Times in the Validation Cohort IVCD Subgroup
Data Sharing Statement
References
- 1.Eppinger H, Rothberger CJ. Zur analyse des elektrokardiogramms. Wien Klin Wochenschr. 1909;22:1091-1098. [Google Scholar]
- 2.Wilson FN. Concerning the form of the QRS deflections of the electrocardiogram in bundle branch block. J Mt Sinai Hosp N Y. 1941;8:1110. [Google Scholar]
- 3.Wilson FN, Johnston FD, Rosenbaum FF, et al. The precordial electrocardiogram. Am Heart J. 1944;27(1):19-85. doi: 10.1016/S0002-8703(44)90603-4 [DOI] [PubMed] [Google Scholar]
- 4.Rosenman RH, Pick A, Katz LN. Intraventricular block; review of the literature. Arch Intern Med (Chic). 1950;86(2):196-232. doi: 10.1001/archinte.1950.00230140032005 [DOI] [PubMed] [Google Scholar]
- 5.Scott RC. Left bundle branch block—a clinical assessment, I. Am Heart J. 1965;70(4):535-566. doi: 10.1016/0002-8703(65)90368-6 [DOI] [PubMed] [Google Scholar]
- 6.Willems JL, Robles de Medina EO, Bernard R, et al. ; World Health Organizational/International Society and Federation for Cardiology Task Force Ad Hoc . Criteria for intraventricular conduction disturbances and pre-excitation. J Am Coll Cardiol. 1985;5(6):1261-1275. doi: 10.1016/S0735-1097(85)80335-1 [DOI] [PubMed] [Google Scholar]
- 7.Surawicz B, Childers R, Deal BJ, et al. ; American Heart Association Electrocardiography and Arrythmias Committee, Council on Clinical Cardiology; American College of Cardiology Foundation; Heart Rhythm Society . AHA/ACCF/HRS recommendations for the standardization and interpretation of the electrocardiogram: part III: intraventricular conduction disturbances: a scientific statement from the American Heart Association Electrocardiography and Arrhythmias Committee, Council on Clinical Cardiology; the American College of Cardiology Foundation; and the Heart Rhythm Society, endorsed by the International Society for Computerized Electrocardiology. J Am Coll Cardiol. 2009;53(11):976-981. doi: 10.1016/j.jacc.2008.12.013 [DOI] [PubMed] [Google Scholar]
- 8.Kusumoto FM, Schoenfeld MH, Barrett C, et al. 2018 ACC/AHA/HRS guideline on the evaluation and management of patients with bradycardia and cardiac conduction delay: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines and the Heart Rhythm Society. J Am Coll Cardiol. 2019;74(7):e51-e156. doi: 10.1016/j.jacc.2018.10.044 [DOI] [PubMed] [Google Scholar]
- 9.Grant RP, Dodge HT. Mechanisms of QRS complex prolongation in man: left ventricular conduction disturbances. Am J Med. 1956;20(6):834-852. doi: 10.1016/0002-9343(56)90204-2 [DOI] [PubMed] [Google Scholar]
- 10.Vassallo JA, Cassidy DM, Marchlinski FE, et al. Endocardial activation of left bundle branch block. Circulation. 1984;69(5):914-923. doi: 10.1161/01.CIR.69.5.914 [DOI] [PubMed] [Google Scholar]
- 11.Auricchio A, Fantoni C, Regoli F, et al. Characterization of left ventricular activation in patients with heart failure and left bundle-branch block. Circulation. 2004;109(9):1133-1139. doi: 10.1161/01.CIR.0000118502.91105.F6 [DOI] [PubMed] [Google Scholar]
- 12.Strauss DG, Selvester RH, Wagner GS. Defining left bundle branch block in the era of cardiac resynchronization therapy. Am J Cardiol. 2011;107(6):927-934. doi: 10.1016/j.amjcard.2010.11.010 [DOI] [PubMed] [Google Scholar]
- 13.Bristow MR, Saxon LA, Boehmer J, et al. ; Comparison of Medical Therapy, Pacing, and Defibrillation in Heart Failure (COMPANION) Investigators . Cardiac-resynchronization therapy with or without an implantable defibrillator in advanced chronic heart failure. N Engl J Med. 2004;350(21):2140-2150. doi: 10.1056/NEJMoa032423 [DOI] [PubMed] [Google Scholar]
- 14.Bilchick KC, Kamath S, DiMarco JP, Stukenborg GJ. Bundle-branch block morphology and other predictors of outcome after cardiac resynchronization therapy in Medicare patients. Circulation. 2010;122(20):2022-2030. doi: 10.1161/CIRCULATIONAHA.110.956011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Zareba W, Klein H, Cygankiewicz I, et al. ; MADIT-CRT Investigators . Effectiveness of cardiac resynchronization therapy by QRS Morphology in the Multicenter Automatic Defibrillator Implantation Trial-Cardiac Resynchronization Therapy (MADIT-CRT). Circulation. 2011;123(10):1061-1072. doi: 10.1161/CIRCULATIONAHA.110.960898 [DOI] [PubMed] [Google Scholar]
- 16.Upadhyay GA, Vijayaraman P, Nayak HM, et al. ; His-SYNC Investigators . On-treatment comparison between corrective His bundle pacing and biventricular pacing for cardiac resynchronization: a secondary analysis of the His-SYNC pilot trial. Heart Rhythm. 2019;16(12):1797-1807. doi: 10.1016/j.hrthm.2019.05.009 [DOI] [PubMed] [Google Scholar]
- 17.Upadhyay GA, Cherian T, Shatz DY, et al. Intracardiac delineation of septal conduction in left bundle-branch block patterns. Circulation. 2019;139(16):1876-1888. doi: 10.1161/CIRCULATIONAHA.118.038648 [DOI] [PubMed] [Google Scholar]
- 18.Auffret V, Puri R, Urena M, et al. Conduction disturbances after transcatheter aortic valve replacement: current status and future perspectives. Circulation. 2017;136(11):1049-1069. doi: 10.1161/CIRCULATIONAHA.117.028352 [DOI] [PubMed] [Google Scholar]
- 19.Calle S, Timmermans F, De Pooter J. Defining left bundle branch block according to the new 2021 European Society of Cardiology criteria. Neth Heart J. 2022;30(11):495-498. doi: 10.1007/s12471-022-01697-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Kawamura I, Batul SA, Vijayaraman P, et al. ECG characteristics of “true” left bundle branch block: insights from transcatheter aortic valve-related LBBB and His-Purkinje conduction system pacing-correctable LBBB. Heart Rhythm. 2023;20(12):1659-1666. doi: 10.1016/j.hrthm.2023.09.004 [DOI] [PubMed] [Google Scholar]
- 21.Kawashima T, Sato F. Visualizing anatomical evidences on atrioventricular conduction system for TAVI. Int J Cardiol. 2014;174(1):1-6. doi: 10.1016/j.ijcard.2014.04.003 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eFigure 1. 12-Lead ECGs for Patients Who Underwent Septal Mapping
eFigure 2. STARD Diagram
eFigure 3. Periprocedural LBBB in a TAVR Patient With Underlying IVCD
eFigure 4. Scatterplot of Lead I Notch Times in the Derivation and Validation Cohorts
eFigure 5. Analysis of Notching in Lead V1
eFigure 6. ROC Curves for Time-to-Notch Criterion
eFigure 7. Analyses of Lead I Notch Times in the Validation Cohort IVCD Subgroup
Data Sharing Statement
