Left vs. Normal QRS Axis During Left Bundle Branch Block Induced by Transcatheter Aortic Valve Replacement

Abstract

Background

The clinical impact of left QRS axis deviation (LAD) during new-onset left bundle branch block (LBBB) after transcatheter aortic valve replacement (TAVR) remains unclear.

Methods and Results

This single-center retrospective study analyzed 254 patients who developed new-onset LBBB during hospitalization after TAVR. Clinical and echocardiographic outcomes were compared between patients with LBBB and LAD (LBBB_LAD) and those with LBBB and a normal QRS axis (LBBB_NA). 96 patients (38%) had LBBB_LAD, defined as a QRS axis <−30°. A more leftward preprocedural QRS axis independently predicted LBBB_LAD (odds ratio 1.20 per 10° decrement; 95% confidence interval (CI) 1.09–1.33; P<0.01). At 3 years, there were no significant differences between groups in all-cause death (28% vs. 19%; P=0.14), cardiovascular death (6% vs. 5%; P=0.73), or heart failure rehospitalization (18% vs. 10%; P=0.07). However, LBBB_LAD was associated with a higher incidence of permanent pacemaker implantation (PPI) for atrioventricular conduction disorder (16% vs. 6%; P=0.02) and remained an independent predictor of PPI (Cox hazard ratio 2.46; 95% CI 1.06–5.73; P=0.04). Echocardiographic measures, including left ventricular ejection fraction, chamber size, and mitral regurgitation severity showed no significant longitudinal differences between groups.

Conclusions

Compared to post-TAVR LBBB_NA, post-TAVR LBBB_LAD is associated with an increased need for PPI, but not with adverse mortality or heart failure outcomes at 3-year follow-up. Closer and extended rhythm monitoring may be warranted in this subgroup.

Central Figure.

Several research groups have compared the clinical outcomes of patients with left bundle branch block (LBBB) with left QRS axis deviation (LBBB_LAD) with those of LBBB patients with a normal QRS axis (LBBB_NA).^1–4 Some of those studies described a poorer prognosis in LBBB_LAD, including a higher incidence of death, conduction disturbance, and/or the development of heart failure (HF),^1,2 whereas others found no significant difference in mortality rates between the LBBB_LAD and LBBB_NA groups.^3,4 However, limited data are available regarding QRS axis deviation during new-onset LBBB after transcatheter aortic valve replacement (TAVR), which is the most common post-procedural conduction disorder, despite extensive research on post-TAVR LBBB patients’ prognoses.⁵ Therefore, we conducted the present study to (1) investigate the incidence, clinical course, and echocardiographic characteristics of patients with LBBB_LAD after TAVR and (2) compare these data with those of patients with post-TAVR LBBB_NA.

Methods

Study Population

We enrolled a total of 766 patients with severe aortic stenosis who had undergone TAVR during the period from January 2016 to January 2022 at St. Marianna University Hospital (Kawasaki, Japan). Of them, 61 were excluded from the study for the following reasons: prior permanent pacemaker implantation (PPI) (n=44), pre-existing LBBB (n=16), and aborted procedure without valve implantation (n=1). The final study population thus consisted of 705 patients.

All patients who exhibited new-onset LBBB during the index hospitalization after TAVR were enrolled, and this cohort also comprised cases in which the conduction abnormality resolved within 24 h. According to the Valve Academic Research Consortium 3 (VARC-3) criteria,⁶ post-TAVR LBBB is classified into 3 subgroups: (1) transient (resolving within 7 days), (2) persistent (resolving within 1 month), and (3) permanent (persisting >30 days). However, in clinical practice, we occasionally encounter cases in which LBBB initially considered permanent resolves several months later, as well as cases in which LBBB classified as transient or persistent reappears after discharge. Because these categories are not always definitive, we chose to enroll all patients who developed new-onset LBBB during hospitalization after TAVR.

The study protocol was approved by the hospital’s ethics committee, and an opt-out system was used to obtain patients’ consent for the use of their clinical data for research purposes.

TAVR Procedure

The patients’ eligibility for TAVR was established based on the consensus of a multidisciplinary heart team that included cardiologists, cardiac surgeons, cardiac anesthesiologists, clinical engineers, and medical radiology technicians. Details of the TAVR procedure are provided elsewhere.⁷ Each patient underwent the implantation of a balloon-expandable SAPIEN XT/SAPIEN 3 valve (Edwards Lifesciences, Irvine, CA, USA) or a self-expandable CoreValve/Evolut R/Evolut PRO/Evolut PRO+ valve (Medtronic, Minneapolis, MN, USA). During the procedure, 100 UI/kg of unfractionated heparin was administered to achieve an activated clotting time >250 s. After the procedure, all patients were managed in a general high-care unit for ≥24 h. A temporary pacemaker was placed when necessary as a backup for third-degree or advanced atrioventricular block. Post-procedure, all patients were prescribed acetylsalicylic acid 100 mg/day to be taken indefinitely.

Electrocardiography, Echocardiography, and Computed Tomography

Standard 12-lead ECG and transthoracic echocardiography were performed before and during each patient’s hospitalization, at 1-, 6-, and 12-month follow-up examinations, and annually thereafter. The diagnosis of LBBB was based on the criteria recommended by the American Heart Association/American College of Cardiology Foundation/Heart Rhythm Society.⁸ Left QRS axis deviation was defined as a frontal plane QRS axis <−30° in accordance with these guidelines. ECG variables such as the PR interval, QRS duration, and QRS axis were automatically measured by a 12-channel ECG system (FCP-8800; Fukuda Denshi, Tokyo, Japan) and manually reviewed by the responsible investigating cardiologist.

Transthoracic echocardiography was performed in all patients to assess parameters such as the severity of aortic stenosis (peak velocity, mean pressure gradient, and area in cm²), left ventricular function, left ventricular dimensions, estimated pulmonary artery pressure, and valvular regurgitation. The valve orifice area was obtained with the use of the continuity equation. The left ventricular ejection fraction (LVEF) was determined with 2-dimensional echocardiography following the Simpson biplane method. The grading of the severity of regurgitation was performed according to the American Society of Echocardiography guidelines.⁹

On pre-TAVR computed tomography (CT) examination, aortic annular dimensions were measured using a standard protocol,¹⁰ and aortic valve calcification was quantified by the Agatston method. Left ventricular outflow tract calcification was evaluated and semiquantitatively classified as previously described.¹¹

Follow-up and Endpoints

Follow-up examinations were conducted at the outpatient clinic at 1, 3, 6, and 12 months and annually thereafter. Clinical information of patients who did not undergo regular checkups was obtained by telephone interviews or contact with the patient’s primary care physician. All clinical events were defined according to the VARC-3 criteria.⁶ Any death of unknown cause was considered as cardiovascular, as recommended by the criteria.

We determined the following endpoints and analyzed their prognostic significance: all-cause death, cardiovascular death, rehospitalization for HF, and need for PPI to treat atrioventricular conduction disorder (AVCD). Rehospitalization for HF was defined as any new unplanned overnight stay in a hospital because of worsening signs and/or symptoms of HF that required an intensification of medical or mechanical therapy, regardless of cardiac or non-cardiac triggers.

Statistical Analysis

Continuous variables are reported as the mean±standard deviation or median (interquartile range) and were compared using Student’s t-test or the Wilcoxon rank sum test depending on the variable distribution. Categorical variables are presented as the count and percentage, and were compared using Fisher’s exact test. Cumulative incidences of clinical events were estimated using the Kaplan-Meier method, and group differences were assessed with the log-rank test. Patients lost to follow-up were censored at the last known contact date. We performed multivariable Cox regression analyses to assess the relationships between clinical endpoints and variables that had clinical relevance and a P value <0.20 in univariate analyses. Multivariable logistic regression analyses were similarly conducted to identify independent predictors of LBBB_LAD by including variables with a P value <0.20 in the univariate analyses. We performed a two-way repeated-measures analysis of variance (ANOVA) to compare echocardiographic parameters (LVEF, LV end-diastolic/systolic diameters, and mitral regurgitation (MR) severity) at different time points between the LBBB_LAD and LBBB_NA groups. Receiver operating characteristic (ROC) curve analysis was used to assess the predictive value of the baseline QRS axis for the development of left QRS axis deviation at 1 month after TAVR. The optimal cutoff value was identified using the Youden index, and the sensitivity, specificity, and area under the curve were calculated. The statistical analyses were conducted using JMP^® 16 software (SAS Institute, Cary, NC, USA). A two-sided probability (P) value <0.05 was considered significant.

Results

Patients’ Characteristics

New-onset LBBB occurred in 254 (36%) patients after they underwent TAVR. Among them, 96 (38%) developed LBBB_LAD and 158 (62%) exhibited LBBB_NA. The results of our comparisons of baseline characteristics and procedural parameters between the LBBB_LAD and LBBB_NA groups are summarized in Table 1. The LBBB_LAD group had a significantly lower body mass index (21±3 vs. 23±3 kg/m², P<0.01) and a more leftward QRS axis (18±36° vs. 32±30°, P<0.01) than the LBBB_NA group. Other baseline clinical characteristics, as well as ECG and echocardiographic parameters, did not differ significantly between groups. During post-TAVR LBBB, the LBBB_LAD group exhibited longer QRS duration than the LBBB_NA group (158±16 vs. 153±14 ms, P=0.01).

Table 1.

Patients’ Baseline Clinical and Procedural Characteristics

	LBBBLAD (n=96)	LBBBNA (n=158)	P value
Clinical characteristics
Age, years	85±6	83±6	0.09
Women, n (%)	64 (67)	119 (75)	0.15
BMI, kg/m2	21±3	23±3	<0.01
NYHA ≥III	24 (25)	44 (28)	0.66
STS score, %	6.1±5.5	5.4±4.7	0.25
Clinical Frailty Scale, points	4.0±1.5	4.0±1.3	0.93
Hypertension, n (%)	86 (90)	142 (90)	1.00
Diabetes mellitus, n (%)	32 (33)	43 (27)	0.32
COPD, n (%)	7 (7)	11 (7)	1.00
Dialysis, n (%)	6 (6)	5 (3)	0.34
History of AF, n (%)	19 (19)	33 (21)	0.87
History of CAD, n (%)	13 (14)	17 (11)	0.55
History of CVD, n (%)	15 (16)	22 (14)	0.72
Hemoglobin, g/dL	11.4±1.6	11.6±1.6	0.38
eGFR, mL/min/1.73 m2	52±21	54±19	0.47
Serum albumin, g/dL	3.6±0.5	3.7±0.5	0.27
NT-proBNP, pg/mL	1,008 (378–3,647)	729 (328–1,978)	0.10
ECG characteristics at baseline
PR interval, ms	178±25	174±31	0.36
QRS duration, ms	99±11	99±10	0.71
QRS axis, °	18±36	32±30	<0.01
ECG characteristics during LBBB
PR interval, ms	192±31	191±38	0.89
QRS duration, ms	158±16	153±14	0.01
QRS axis, °	−48±12	3±29	<0.01
TTE parameters
Aortic valve area, cm2	0.59±0.16	0.62±0.18	0.17
Aortic valve peak velocity, m/s	4.3±0.9	4.2±0.9	0.40
Aortic valve mean-PG, mmHg	44±19	41±18	0.27
LVEF, %	63±10	64±10	0.91
LVDd, mm	42.8±6.6	41.9±5.6	0.25
LVDs, mm	27.5±7.4	26.8±6.0	0.38
IVST, mm	10.2±2.0	10.0±1.9	0.51
MR ≥moderate, n (%)	9 (9)	11 (7)	0.63
CT parameters
Aortic annulus diameter, mm	19.7±2.1	19.7±2.1	0.79
Aortic valve CA score, points	1,864 (1,439–2,781)	1,713 (1,144–2,454)	0.14
LVOT-CA ≥moderate, n (%)	9 (9)	14 (9)	1.00
Procedural parameters in TAVR
Approach, n (%)
Transfemoral	88 (92)	151 (95)	0.27
Transapical	4 (4)	3 (2)	0.43
Subclavian	4 (4)	4 (3)	0.48
Prosthesis size, n (%)
20 mm	0 (0)	6 (4)	0.09
23 mm	66 (69)	91 (58)	0.08
26 mm	24 (25)	49 (31)	0.32
29 mm	6 (6)	11 (7)	1.00
34 mm	0 (0)	1 (1)	1.00
Self-expandable valve, n (%)	17 (18)	32 (20)	0.74

Categorical variables are the absolute number and percentages of subgroup. Continuous variables are the mean±standard deviation or median (Q1–Q3). AF, atrial fibrillation; BMI, body mass index; CA, calcification; CAD, coronary artery disease; COPD, chronic obstructive pulmonary disease; CT, computed tomography; CVD, cerebrovascular disease; ECG, electrocardiography; eGFR, estimated glomerular filtration rate; IVST, interventricular septal thickness; LBBB_LAD, left bundle branch brock with left QRS axis deviation; LBBB_NA, left bundle branch brock with a normal QRS axis; LVDd, left ventricular end-diastolic diameter; LVDs, left ventricular end-systolic diameter; LVEF, left ventricular ejection fraction; LVOT, left ventricular outflow tract; MR, mitral regurgitation; NT-proBNP, N-terminal pro-B-type natriuretic peptide; NYHA, New York Heart Association; PG, pressure gradient; STS, Society of Thoracic Surgeons; TAVR, transcatheter aortic valve replacement; TTE, transthoracic echocardiography.

Overall and Cardiovascular Deaths

During the 3-year follow-up, a total of 47 patients died and 10 (21%) were cardiovascular deaths: decompensated HF (n=5), sudden death (n=4), and rupture of an abdominal aortic aneurysm (n=1). There were no significant differences between the LBBB_LAD and LBBB_NA groups in all-cause deaths (28% vs. 19%, log-rank P=0.14) or cardiovascular deaths (6% vs. 5%, log-rank P=0.73) (Figure 1A-a,b).

Figure 1.

(A) Time-to-event curves at 3-year follow-up. Kaplan-Meier survival curves for all-cause death (a), cardiovascular death (b), rehospitalization for heart failure (c), and permanent pacemaker implantation for atrioventricular conduction disorder (d). (B) Temporal changes in echocardiographic parameters: left ventricular ejection fraction (a), left ventricular end-diastolic diameter (b), left ventricular end-systolic diameter (c), and the severity of mitral regurgitation (d). Error bars: standard errors. AVCD, atrioventricular conduction disorder; LBBB_LAD, left bundle branch block with left QRS axis deviation; LBBB_NA, left bundle branch block with a normal QRS axis; LVDd, left ventricular end-diastolic diameter; LVDs, left ventricular end-systolic diameter; LVEF, left ventricular ejection fraction; MR, mitral regurgitation; PPI, permanent pacemaker implantation.

In the Cox multivariable regression analysis, the serum albumin concentration was identified as the only independent predictor of all-cause death (hazard ratio [HR] 0.16 per 1-g/dL increment, 95% confidence interval [CI]: 0.08–0.29, P<0.01) and cardiovascular death (HR 0.25 per 1-g/dL increment, 95% CI: 0.07–0.89, P=0.03) (Table 2).

Table 2.

Univariate and Multivariable Predictors of 3-Year Clinical Outcomes in Patients With LBBB_LAD and LBBB_NA Following TAVR

	Univariate		Multivariable
	Hazard ratio (95% CI)	P value	Hazard ratio (95% CI)	P value
All-cause death
Age, per 1-year increment	1.06 (1.00–1.12)	0.03	1.03 (0.97–1.09)	0.36
Female	0.91 (0.49–1.70)	0.77
History of AF	1.04 (0.52–2.10)	0.91
History of CAD	0.72 (0.26–2.02)	0.54
History of CVD	1.48 (0.72–3.06)	0.29
Serum albumin, per 1-g/dL increment	0.14 (0.08–0.26)	<0.01	0.16 (0.08–0.29)	<0.01
eGFR, per 10-mL/min/1.73 m2 increment	1.08 (0.93–1.26)	0.31
LBBBLAD	1.53 (0.86–2.72)	0.15	1.24 (0.69–2.24)	0.47
Cardiovascular death
Age, per 1-year increment	1.01 (0.91–1.13)	0.91
Female	0.92 (0.24–3.54)	0.90
History of AF	0.43 (0.05–3.42)	0.43
History of CAD	0.84 (0.11–6.65)	0.87
History of CVD	2.69 (0.70–10.42)	0.15	2.95 (0.76–11.48)	0.12
Serum albumin, per 1-g/dL increment	0.27 (0.07–0.94)	0.04	0.25 (0.07–0.89)	0.03
eGFR, per 10-mL/min/1.73 m2 increment	0.94 (0.69–1.31)	0.72
LBBBLAD	1.25 (0.35–4.42)	0.73
Heart failure rehospitalization
Age, per 1-year increment	1.05 (0.98–1.13)	0.18	1.05 (0.98–1.13)	0.19
History of AF	3.31 (1.55–7.08)	<0.01	2.97 (1.37–6.42)	0.01
eGFR, per 10-mL/min/1.73 m2 increment	0.77 (0.64–0.93)	0.01	0.79 (0.64–0.97)	0.02
LBBBLAD	1.97 (0.87–3.83)	0.08	1.83 (0.85–3.95)	0.12
LVEF at baseline, per 1-% increment	0.98 (0.95–1.01)	0.20
PPI for atrioventricular conduction disorder
Age, per 1-year increment	1.06 (0.98–1.14)	0.13	1.05 (0.97–1.13)	0.22
PR interval during LBBB, per 10-ms increment	1.03 (0.89–1.18)	0.66
QRS duration during LBBB, per 10-ms increment	0.87 (0.64–1.16)	0.36
LBBBLAD	2.62 (1.13–6.06)	0.02	2.46 (1.06–5.73)	0.04
Self-expandable valve	1.48 (0.58–3.74)	0.41
Aortic valve CA score, per 100-point increment	1.01 (0.96–1.04)	0.73

PPI, permanent pacemaker implantation. Other abbreviations as for Table 1.

Rehospitalization for HF

The Kaplan-Meier analysis revealed a similar rehospitalization rate for HF in the LBBB_LAD and LBBB_NA groups (18% vs. 10%, log-rank P=0.07) (Figure 1A-c). The median time until rehospitalization was 352 days (range: 13–1,058 days). The triggers for decompensated HF were infection (n=8), tachycardiac atrial fibrillation (n=3), valve dysfunction (n=2), drug withdrawal (n=2), uncontrolled blood pressure (n=1), third-degree atrioventricular block (n=1), and unknown cause (n=10). Of the 27 patients with decompensated HF, only 1 patient required the implantation of a cardiac resynchronization device. After multivariable adjustment, a history of atrial fibrillation and the estimated glomerular filtration rate were observed to be significant predictors of rehospitalization for HF (HR 2.97, 95% CI: 1.37–6.42, P=0.01, and HR 0.79 per 10-mL/min/1.73 m² increment, 95% CI: 0.64–0.97, P=0.02, respectively) (Table 2).

Permanent Pacemaker Implantation

Our analyses revealed a significant difference in the need for PPI for AVCD at 3 years between the LBBB_LAD and LBBB_NA groups (16% vs. 6%, log-rank P=0.02) (Figure 1A-d). The median time to PPI was 6 days (range, 1–820 days), and more than half of the procedures (13 of 23) were performed during the index hospitalization. After excluding in-hospital PPI events within 7 days of TAVR, LBBB_LAD remained significantly associated with post-discharge PPI (log-rank P=0.03) (Supplementary Figure 1). After multivariable adjustment. left QRS axis deviation during LBBB was independently associated with the need for PPI for AVCD (HR 2.46, 95% CI: 1.06–5.73, P=0.04) (Table 2).

Factors Predictive of Left QRS Axis Deviation During post-TAVR LBBB

The predictors of LBBB_LAD following TAVR in the univariate and multivariable logistic regression analyses are shown in Table 3. A more leftward preprocedural QRS axis was associated with the occurrence of LBBB_LAD after TAVR in the univariable analysis, and it remained an independent predictor in the multivariable model (odds ratio 1.20, 95% CI: 1.09–1.33 per 10-° decrement, P<0.01).

Table 3.

Univariate and Multivariable Predictors of LBBB_LAD After TAVR

	Univariate		Multivariable
	Odds ratio (95% CI)	P value	Odds ratio (95% CI)	P value
Age, per 1-year increment	1.04 (0.99–1.09)	0.10	1.00 (0.95–1.06)	0.94
Female	0.66 (0.38–1.14)	0.14	0.52 (0.24–1.13)	0.10
Baseline PR interval, per 10-ms increment	1.04 (0.95–1.15)	0.36
Baseline QRS duration, per 10-ms increment	1.05 (0.82–1.34)	0.71
Baseline QRS axis, per 10-° decrement	1.15 (1.06–1.24)	<0.01	1.20 (1.09–1.33)	<0.01
Baseline LVEF, per 1-% decrement	1.00 (0.98–1.03)	0.91
Baseline LVDd, per 1-mm increment	1.03 (0.98–1.07)	0.25
Baseline IVST, per 1-mm increment	1.05 (0.92–1.19)	0.51
Valve size/annulus dia., per 0.1 increment	0.91 (0.73–1.15)	0.43
Self-expandable valve	0.85 (0.44–1.63)	0.62
Aortic valve CA score, per 100-point increment	1.02 (1.00–1.05)	0.08	1.02 (0.99–1.04)	0.16

Abbreviations as for Table 1.

Post-TAVR Echocardiographic Findings

The chronologic changes in the LVEF, LV end-diastolic and end-systolic diameters, and the severity of MR according to the presence of LBBB_LAD are depicted in Figure 1B. The 2-way repeated-measures ANOVA revealed no significant differences in any echocardiographic parameters between the LBBB_LAD and LBBB_NA groups at the 3-year follow-up.

Discussion

There are 3 main findings of this study. (1) The incidence of LBBB_LAD among post-TAVR LBBB patients was 38% (96/254 patients), and all of the remaining cases displayed LBBB_NA. No cases of LBBB with a right or indefinite axis were observed in the study population. (2) The presence of LBBB_LAD did not increase the risks of overall or cardiovascular death, rehospitalization for HF, or LV dysfunction. However, it was an independent predictor of the need for PPI for AVCD. (3) The preprocedural QRS axis was significantly associated with the post-TAVR occurrence of LBBB_LAD.

To the best of our knowledge, this study is the first to report the long-term clinical impacts of the QRS axis during post-TAVR LBBB. In this study, despite it being well known that a significant proportion of new-onset LBBBs after TAVR resolve spontaneously before the 1-month follow-up,¹² we did not differentiate post-TAVR LBBB by its temporal course based on reasons described in the Methods. Nonetheless, whether LBBB resolves or persists remains clinically relevant when considering more tailored management strategies. Therefore, we categorized patients according to LBBB persistence beyond the 1-month follow-up: those with LBBB_LAD and LBBB_NA that persisted were classified as permanent LBBB_LAD (P-LBBB_LAD) and permanent LBBB_NA (P-LBBB_NA), respectively, whereas those with LBBB that resolved before 1 month were classified as resolved LBBB_LAD (R-LBBB_LAD) and resolved LBBB_NA (R-LBBB_NA).

Based on our resolution criterion, we classified the study population into 4 subgroups: (1) P-LBBB_LAD (n=25), (2) P-LBBB_NA (n=31), (3) R-LBBB_LAD (n=18), and (4) R-LBBB_NA (n=147) (Figure 2). Figure 3 presents the 4 groups’ clinical (Figure 3A) and echocardiographic (Figure 3B) data and the results of our comparison of the P-LBBB_LAD and P-LBBB_NA groups. The findings derived from this comparison were similar to those observed between the LBBB_LAD and LBBB_NA groups.

Figure 2.

Prevalence of left bundle branch block at 1-month follow-up examinations. *Excluded from the analysis due to ventricular pacing in 1-month ECG, loss to follow-up, or incomplete data. LBBB, left bundle branch block; P-LBBB_LAD, permanent left bundle branch block with left QRS axis deviation; P-LBBB_NA, permanent left bundle branch block with a normal QRS axis; R-LBBB_LAD, resolved left bundle branch block with left QRS axis deviation; R-LBBB_NA, resolved left bundle branch block with a normal QRS axis.

Figure 3.

(A) Time-to-event curves at 3-year follow-up. Kaplan-Meier survival curves for all-cause death (a), cardiovascular death (b), readmission for heart failure (c), and permanent pacemaker implantation for atrioventricular conduction disorder (d). (B) Changes in echocardiographic parameters. Temporal changes in the left ventricular ejection fraction (a), left ventricular end-diastolic diameter (b), left ventricular end-systolic diameter (c), and the severity of mitral regurgitation (d). Error bars: standard errors. Abbreviations as for Figures 1,2.

Impact of LBBB_LAD on Mortality Rate

Despite the confirmation of the negative effects of LBBB_LAD compared to LBBB_NA,^1,2 the present data show that the clinical outcomes in the LBBB_LAD group were comparable to those in the LBBB_NA in terms of overall and cardiovascular deaths, which was consistent with those observed for the P-LBBB_LAD and P-LBBB_NA groups. We speculate that the most plausible reason for the comparable outcomes despite a leftward QRS axis during LBBB might be the presence of prognostic factors other than LBBB_LAD. Almost 80% of the deaths among the present study’s patients had non-cardiac causes such as pneumonia, cancer, or advanced age. TAVR clearly resolved the patients’ severe aortic valve conditions, but an unfavorable preprocedural condition might have adversely affected their prognoses. The patients who underwent TAVR were already of advanced age and had many comorbidities. It can thus be postulated that the detrimental effect of LBBB on hard study endpoints might be masked by the impact of other negative predictors. This seems to have been demonstrated by our data indicating that the serum albumin concentration was strongly associated with death in the multivariable analysis.

Impact of LBBB_LAD on HF and Echocardiographic Parameters

It is well recognized that patients with LBBB experience a significant decrease in LVEF, which is followed by a higher rate of the development of HF compared to patients without LBBB.¹³ This outcome has also been observed in patients with post-TAVR LBBB,¹⁴ but there are limited data comparing LBBB_LAD and LBBB_NA. Our study found a longer QRS duration during LBBB in the patients with LBBB_LAD. Delayed conduction in the LV leads to dyssynchrony and may contribute to significant MR. However, there was neither a higher incidence of HF rehospitalization nor significant changes over time in any of the echocardiographic parameters examined in this study. This finding is consistent with the report by Das et al.,¹⁵ and further supported by the biomarker data (Supplementary Figure 2).

One of the possible reasons for this result is that the LBBB_LAD group had a preserved LVEF at baseline (mean LVEF 63%), and thus the deleterious impact of LBBB_LAD might have been attenuated. Nevertheless, the adverse effect of LBBB_LAD might exist, as indicated by the larger LV systolic/diastolic diameters in the P-LBBB_LAD patients (Figure 3B-b,c), although this difference was not significant. More samples or longer follow-up might clarify the deleterious effect(s) of LBBB_LAD on LV function. Another possible explanation is the preload dependence of echocardiographic dyssynchrony. Future studies using low-level bicycle stress echocardiography with speckle-tracking analysis may help determine whether patients with LBBB_LAD demonstrate greater stress-induced septal–lateral delay, MR changes, or strain abnormalities compared with those with LBBB_NA.

Impact of LBBB_LAD on the Need for PPI

Although prior studies have shown that patients who develop LBBB after TAVR have an increased risk of PPI,^14,16,17 few have examined this issue from the perspective of the frontal QRS axis. Conduction during LBBB relies mainly on the right bundle branch (RBB), so RBB impairment is considered as a key determinant of AVCD. In our study, however, LBBB_LAD, reflecting the left bundle-system condition, was associated with AVCD requiring PPI. This observation implies that patients with LBBB_LAD may have underlying RBB dysfunction. Given the anatomical remoteness of the RBB from the prosthesis, periprocedural mechanical effects on the RBB are minimal, so new-onset RBBB is rare after TAVR.¹⁸ Thus, the association between LBBB_LAD and higher risk of PPI likely reflects pre-existing RBB dysfunction rather than an intraprocedural adverse effect.

The risk of PPI appears to persist beyond discharge in patients with LBBB_LAD (Supplementary Figure 1). However, no risk stratification or pacing strategy has previously been guided by the QRS axis during post-TAVR LBBB.¹⁹ Because LBBB_LAD is a long-term predictor of PPI and may indicate pre-existing conduction system vulnerability, we present 2 proposals: (1) consideration of prolonged temporary pacing during the index hospitalization, and (2) a structured post-discharge rhythm-monitoring plan that includes periodic Holter recordings and, when feasible, extended 1- to 2-week ambulatory ECG monitoring to detect evolving conduction disturbances at an early stage.

As shown in Figure 3A-d, the presence of LAD at 1-month post-TAVR well predicted the risk of PPI irrespective of LBBB status. The patients with LAD at 1-month post-TAVR, namely P-LBBB_LAD and R-LBBB_LAD, had a more leftward preprocedural QRS axis (Figure 4A) compared to P-LBBB_NA and R-LBBB_NA. These findings suggested that TAVR patients with a more leftward baseline QRS axis may warrant closer rhythm surveillance for AVCD. ROC analysis identified 13° as the optimal cutoff (69% sensitivity and 73% specificity; Figure 4B).

Figure 4.

Association between baseline QRS axis and left QRS axis deviation at 1 month. Changes in the QRS axis (A) and the receiver operating characteristic curve (ROC) of preprocedural QRS axis for left QRS axis deviation at 1 month (B). ROC analysis demonstrated an optimal cutoff value of 13° with 69% sensitivity and 73% specificity. AUC, area under curve. Other abbreviations as for Figure 2.

Predictors of LBBB_LAD

Various predictive factors for new-onset LBBB after TAVR have been reported, including male sex, prior myocardial infarction, diabetes mellitus, and the type, size, or depth of the prosthesis,¹¹ but data regarding the development of LBBB_LAD after TAVR are limited. Our findings revealed that a more leftward QRS axis at baseline is significantly associated with LBBB_LAD following TAVR.

Anatomically, the LBB originates from the atrioventricular branching bundle beneath the membranous septum, which is located at the base of the interleaflet triangle between the right and noncoronary aortic leaflets.²⁰ We speculate that the left posterior fascicle (LPF) may be more susceptible to procedural injury than the left anterior fascicle (LAF), because the LPF branches more superficially and proximally. Thus, in patients with pre-existing LAF impairment, additional LPF injury during TAVR may readily produce complete LBBB, manifesting as LBBB_LAD. Conversely, when LAF dysfunction is minimal, the resulting LBBB tends to present as LBBB_NA. This proposed mechanism is supported by electroanatomical observations reported by Abu-Alrub et al.²¹ Preprocedural anatomical factors such as calcification burden and membranous septum length, as well as intraprocedural factors including prosthesis position, balloon pre-/post-dilation, and valve oversizing, may also influence post-TAVR LBBB phenotypes. A prospective study integrating pre- and post-TAVR CT with computational modeling is needed to clarify the anatomical substrate and prosthesis–septal interaction underlying axis deviation in post-TAVR LBBB and support the development of risk models to guide procedural adjustments, such as optimal implantation depth or valve selection, to reduce conduction disturbances.

Study Limitations

First, it was a retrospective observational study with loss to follow-up, and based on a single-center experience. The sample size was relatively small (n=254), reducing the statistical power. In addition, even when the values were manually reviewed by an experienced clinician, the measurements of the PR interval, QRS duration, and QRS axis were calculated primarily with the use of The ECG system’s proprietary algorithm, which may lead to variations among different systems. Moreover, certain potential predictors of LBBB_LAD after TAVR such as the depth of valve implantation and the membranous septum length on preoperative CT were not available in this study. Data on medications that could affect the cardiac conduction system were also unavailable. Finally, although LBBB_LAD was associated with a higher risk of PPI, this finding should be interpreted with caution given the limited number of events (n=23) and the wide CI. Larger studies will be required to validate the strength of this relationship.

Conclusions

During a 3-year follow-up after TAVR, the development of LBBB_LAD did not have a significant adverse effect on deaths, HF development, or LV function compared with LBBB_NA; however, it was associated with a higher rate of AVCD requiring PPI. In addition to examining the relationship between QRS axis during LBBB and clinical outcomes, our study provides further insight into the temporal evolution of post-TAVR LBBB by incorporating a longer follow-up and by stratifying post-TAVR LBBB into persistent and resolved phenotypes. In particular, patients with a more leftward preprocedural QRS axis were more likely to develop LBBB_LAD after TAVR and may therefore require closer rhythm surveillance both during hospitalization and after discharge.

Disclosures

None of the authors has any financial disclosures relevant to this article. This study did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Y.J.A. is a member of Circulation Reports’ Editorial Team but had no involvement in the peer review of this article and had no access to information regarding its peer review.

IRB Information

This study was approved by the St. Marianna University Ethics Committee (reference no. 6660).

Supplementary Files

Supplementary File 1

Supplementary Figure 1. Supplementary Figure 2.

Data Availability Statement

The data related to this article are not available because of ethical restrictions.