Abstract
Background
Left atrial enlargement (LAE) is common in patients undergoing transcatheter aortic valve implantation (TAVI), but its prognostic significance may vary according to the clinical drivers of atrial remodeling. We examined whether prognostic factors differ by clinically attributed LAE subtype in patients undergoing TAVI with preserved left ventricular ejection fraction (LVEF).
Methods and Results
We retrospectively analyzed 317 consecutive patients with symptomatic severe aortic stenosis (AS) undergoing TAVI who had preserved LVEF and LAE. Patients were classified as AS-associated LAE (no atrial fibrillation [AF] and <moderate mitral regurgitation [MR]) or AF/MR-associated LAE (AF and/or ≥moderate MR). The primary endpoint was all-cause death. During a median follow-up of 2.7 years, 69 patients (21.8%) died. The all-cause mortality rate was higher in the AF/MR-associated group than in the AS-associated LAE group (30.7% vs. 19.0%; log-rank P<0.01), but this association was attenuated after adjustment for eGFR, log-BNP, and LAVI. In the AS-associated LAE group, higher log-BNP (hazard ratio (HR) 1.45, 95% confidence interval (CI) 1.06–1.98), higher E/e′ (HR 1.03, 95% CI 1.01–1.06), and lower TAPSE (HR 0.89, 95% CI 0.81–0.99) predicted death. In the AF/MR-associated LAE group, lower eGFR (HR 0.96, 95% CI 0.93–1.00) and higher log-BNP (HR 1.98, 95% CI 1.09–3.59) predicted death after adjustment for age. Interactions were significant for eGFR and E/e′.
Conclusions
Prognostic markers after TAVI differ by clinically attributed LAE subtype, supporting subtype-informed risk stratification.
Key Words: Aortic stenosis, Left atrial enlargement, Prognosis, Transcatheter aortic valve implantation
Central Figure.

Transcatheter aortic valve implantation (TAVI) is an established therapy for elderly patients with severe aortic stenosis (AS).1 Nevertheless, long-term outcomes after TAVI remain heterogeneous, and risk stratification based solely on valvular severity is insufficient, as recent data have shown that clinical presentation before TAVI, particularly heart failure or syncope symptoms, is associated with worse postprocedural outcomes.2 Cardiac damage staging has improved prognostic assessment by accounting for structural and functional consequences of AS beyond the valve, including a left atrial and mitral damage stage in which left atrial enlargement (LAE) is frequently observed.3,4 However, LAE in patients undergoing TAVI is not a single phenotype. In addition to pressure overload from AS, LA remodeling may be substantially influenced by atrial fibrillation (AF), mitral regurgitation (MR), and systemic comorbidities, which can each drive atrial enlargement through distinct pathophysiological pathways.5–7 LAE is also commonly interpreted as a marker of chronic diastolic dysfunction and elevated left ventricular filling pressures,8–10 and has been associated with adverse outcomes in AS and after TAVI, particularly when LA size is markedly increased or reverse remodeling is limited.11,12 This heterogeneity may explain why prognostic determinants linked to LAE differ across patients, and why conventional echocardiographic indices may not carry uniform prognostic meaning. Despite the high prevalence of LAE in recent TAVI populations, few studies have directly examined whether the prognostic implications of LAE differ according to its clinically attributed subtype. Therefore, we investigated whether the prognostic implications and determinants of LAE differ according to clinically attributed LAE subtypes in patients undergoing TAVI.
Methods
Study Population
This single-center observational study was a retrospective analysis at Osaka Metropolitan University Hospital. We screened 578 consecutive patients with symptomatic severe AS who underwent TAVI between January 2016 and April 2023. To reduce pathophysiological heterogeneity in accordance with the cardiac damage staging framework and the study hypothesis, exclusion criteria were applied sequentially in the order shown in Figure 1; patients meeting multiple exclusion criteria were counted only once under the first applicable criterion. We excluded: (1) advanced non-LA cardiac damage, including LVEF <50%, severe right ventricular systolic dysfunction defined as tricuspid annular plane systolic excursion (TAPSE) <10 mm based on prior TAVI literature,13 pulmonary hypertension ≥60 mmHg, or moderate-to-severe tricuspid regurgitation; (2) paradoxical low-flow/low-gradient AS, because the low-flow phenotype may independently affect LA remodeling, B-type natriuretic peptide (BNP) levels, diastolic indices, and prognosis, thereby confounding the clinical attribution of LAE; (3) conditions affecting LA structure or rhythm independent of AS, including prior permanent pacemaker implantation, congenital heart disease, or prior open-heart surgery; (4) normal LA size, defined as LA volume index (LAVI) ≤34 mL/m2; (5) competing non-cardiac prognosis, defined as active malignancy; and (6) in-hospital death or incomplete echocardiographic data. After exclusions, 317 patients with preserved LVEF and LAE (LAVI >34 mL/m2) were included. Patients were classified as AS-associated LAE (no AF and <moderate MR) or AF/MR-associated LAE (AF and/or ≥moderate MR) (Figure 1), because AF and MR can drive LA remodeling independently of AS severity. AF was identified by documented history and/or AF/atrial flutter on 12-lead ECG during the index hospitalization (paroxysmal, persistent, or permanent),14 and MR severity was graded according to guidelines.15 The study protocol conformed to the Declaration of Helsinki and was approved by the Ethics Committee of Osaka Metropolitan University Graduate School of Medicine (Approval No. 2021-064). Written informed consent was given by all participants.
Figure 1.

Patient selection flowchart. Exclusion criteria were predefined based on both the cardiac damage staging framework and the study hypothesis to reduce pathophysiological heterogeneity and to enable clinical subtyping of left atrial enlargement, rather than to reflect an unselected transcatheter aortic valve implantation population. Exclusion criteria were applied sequentially in the order shown; patients meeting multiple exclusion criteria were counted only once under the first applicable criterion. Advanced non-left atrial cardiac damage included LVEF <50%, severe right ventricular systolic dysfunction defined as TAPSE <10 mm, pulmonary hypertension ≥60 mmHg, and moderate-to-severe tricuspid regurgitation. Paradoxical low-flow/low-gradient AS was listed separately because the low-flow phenotype may independently affect LA remodeling, BNP levels, diastolic indices, and prognosis. Patients with LAE (LAVI >34 mL/m2) were classified into AS-associated LAE (no AF and <moderate MR) and AF/MR-associated LAE (AF and/or ≥moderate MR). AF, atrial fibrillation; AS, aortic stenosis; BNP, B-type natriuretic peptide; LAE, left atrial enlargement; LAVI, left atrial volume index; LVEF, left ventricular ejection fraction; MR, mitral regurgitation; TAPSE, tricuspid annular plane systolic excursion; TAVI, transcatheter aortic valve implantation.
Transthoracic Echocardiography
Comprehensive transthoracic echocardiography was performed before TAVI using commercially available systems. Standard LV dimensions and wall thickness were measured, and LV mass was calculated and indexed to body surface area. LVEF and LAV were assessed using the biplane modified Simpson method. Both measures were indexed to body surface area, and values were averaged over 3 cardiac cycles (5 cycles in AF). LAE was defined as LAVI >34 mL/m2 in accordance with guidelines.16 Transmitral E velocity was obtained by pulsed-wave Doppler. Tissue Doppler e′ velocity was measured at the septal mitral annulus, and septal E/e′ was used consistently in all patients. Measurements were averaged over 3 consecutive cardiac cycles in sinus rhythm, and 5 cardiac cycles in AF. Because lateral e′ was not uniformly available and because E/e′ may be less reliable in patients with AF and/or MR, E/e′ was interpreted as an exploratory Doppler marker rather than as a definitive estimate of LV filling pressure in the AF/MR-associated LAE group.8 Aortic valve area was calculated by the continuity equation, and severe AS was defined as peak velocity >4.0 m/s, mean gradient >40 mmHg, or aortic valve area <1.0 cm2.17 Stroke volume index was derived from LV outflow tract measurements. Systolic pulmonary artery pressure was estimated from the tricuspid regurgitant jet velocity and right atrial pressure. Right ventricular systolic function was assessed using TAPSE as the primary index, a guideline-endorsed measure of right ventricular systolic function, in accordance with guideline recommendations.18
Clinical Follow-up and Outcomes
After discharge, patients were followed at regular outpatient visits. For those who transferred care, follow-up information was obtained from referring physicians and/or medical records at the receiving institutions. Vital status was ascertained for all patients. The primary objective of this study was to determine whether the prognostic implications and determinants of LAE differed according to clinically attributed LAE subtype in patients undergoing TAVI. The primary endpoint was all-cause death after TAVI. Secondary objectives were to identify clinical, laboratory, and echocardiographic predictors of all-cause death in the overall cohort, and within each LAE subtype, and to evaluate whether these predictors differed by LAE subtype. No formal secondary endpoint was prespecified. Death was classified as cardiovascular, confirmed non-cardiovascular, or undetermined by chart review when possible. Deaths with insufficient information to determine cardiovascular or non-cardiovascular cause were classified as deaths of undetermined cause. Cause-specific outcomes were analyzed descriptively as exploratory outcomes because of the limited number of adjudicated cardiovascular deaths and the presence of deaths of undetermined cause. Outcomes were obtained from electronic medical records and confirmed by detailed chart review.
Statistical Analysis
Continuous variables are presented as mean±SD or median (interquartile range [IQR]), and categorical variables as n (%). Group comparisons used Student’s t-test or the Mann–Whitney U test for continuous variables and the chi-square or Fisher’s exact test for categorical variables, as appropriate. Survival was evaluated by Kaplan–Meier curves and compared using the log-rank test. Cox proportional hazards models were used to assess predictors of all-cause death after TAVI. Univariable analyses were performed first, followed by multivariable models with covariates prespecified a priori based on clinical relevance while limiting the number of covariates to avoid overfitting. To evaluate the association between LAE subtype and all-cause death in the entire LAE cohort, additional Cox proportional hazards models were constructed using AS-associated LAE as the reference category. Model 1 was unadjusted, Model 2 was adjusted for age and sex, and Model 3 was adjusted for age, sex, estimated glomerular filtration rate (eGFR), log-transformed BNP, and LAVI. AF and MR were not included as covariates because they were components of the LAE subtype definition. Cox models were additionally fitted separately within the AS-associated LAE and AF/MR-associated LAE groups using the same prespecified approach. Effect modification by LAE subtype was examined using interaction terms for eGFR and E/e′ based on the study hypothesis. These interaction analyses were hypothesis-driven and were interpreted cautiously because of the limited sample size. The proportional hazards assumption was assessed using Schoenfeld residuals. No clear violation was observed in the additional whole-cohort Cox models. In the AS-associated LAE subgroup model, age violated the assumption, so an age-stratified Cox model was used and the HR for age was not estimated. All tests were two-sided with P<0.05, and analyses were performed using R (version 4.3.1).
Results
Baseline Characteristics
Among 317 patients with preserved LVEF, 242 were classified as AS-associated LAE and 75 as AF/MR-associated LAE (Figure 1). Baseline clinical characteristics, AS severity, and procedural variables were broadly similar between groups (Table 1). In contrast, the AF/MR-associated group showed greater systemic vulnerability, with higher levels of frailty and BNP (285.9 [140.0–485.0] vs. 117.1 [56.0–226.2] pg/mL; P<0.001) and lower levels of albumin.
Table 1.
Baseline Characteristics in the Entire Study Population
| Overall (N=317) |
AS-associated LAE group (N=242) |
AF/MR-associated LAE group (N=75) |
P value | |
|---|---|---|---|---|
| Baseline clinical characteristics | ||||
| Age, years | 83.0 (80.0–87.0) | 83.0 (80.0–87.0) | 84.0 (81.5–87.0) | 0.668 |
| Male, n (%) | 89 (28.1) | 73 (30.2) | 16 (21.3) | 0.145 |
| BMI, kg/m2 | 22.1 (20.0–24.6) | 22.5 (20.6–24.9) | 20.5 (18.8–22.9) | <0.001 |
| CFS | 4.0 (3.0–4.0) | 4.0 (3.0–4.0) | 4.0 (4.0–5.0) | 0.022 |
| Logistic EuroSCORE | 10.7 (7.5–16.0) | 10.6 (7.5–15.6) | 12.1 (7.8–17.5) | 0.123 |
| Dyslipidemia, n (%) | 176 (55.5) | 149 (61.6) | 27 (36.0) | <0.001 |
| Diabetes mellitus, n (%) | 79 (24.9) | 63 (26.0) | 16 (21.3) | 0.449 |
| Hypertension, n (%) | 285 (89.9) | 217 (89.7) | 68 (90.7) | 1.000 |
| Smoking, n (%) | 96 (30.3) | 77 (31.8) | 19 (25.3) | 0.316 |
| Coronary artery disease, n (%) | 80 (25.2) | 63 (26.0) | 17 (22.7) | 0.649 |
| Peripheral artery disease, n (%) | 57 (18.0) | 44 (18.2) | 13 (17.3) | 1.000 |
| AF, n (%) | 53 (16.7) | 0 (0.0) | 53 (70.7) | <0.001 |
| Paroxysmal AF, n (%) | 25 (7.9) | 0 (0.0) | 25 (33.3) | <0.001 |
| Persistent AF, n (%) | 28 (8.8) | 0 (0.0) | 28 (37.3) | <0.001 |
| β-blockers, n (%) | 80 (25.2) | 54 (22.3) | 26 (34.7) | 0.034 |
| Mineralocorticoid-receptor antagonists, n (%) | 47 (14.8) | 26 (10.7) | 21 (28.0) | 0.001 |
| Angiotensin-converting enzyme inhibitors, n (%) | 178 (56.2) | 141 (58.3) | 37 (49.3) | 0.185 |
| Statins, n (%) | 146 (46.1) | 122 (50.4) | 24 (32.0) | 0.005 |
| Laboratory data | ||||
| Hemoglobin, g/dL | 11.4±1.6 | 11.5±1.6 | 11.0±1.5 | 0.016 |
| Albumin, g/dL | 3.8 (3.5–4.1) | 3.9 (3.6–4.1) | 3.6 (3.4–3.9) | <0.001 |
| eGFR, mL/min/1.73 m2 | 51.7 (40.2–65.1) | 52.9 (40.3–66.2) | 48.2 (40.1–60.1) | 0.206 |
| BNP, pg/mL | 142.4 (66.6–301.7) | 117.1 (56.0–226.2) | 285.9 (140.0–485.0) | <0.001 |
| Echocardiographic parameters | ||||
| LVEF, % | 63.0 (60.0–65.0) | 63.0 (60.0–65.0) | 60.0 (58.0–65.0) | 0.014 |
| LVDd, mm | 43.0 (40.0–46.0) | 42.0 (40.0–46.0) | 43.0 (39.0–46.5) | 0.684 |
| LVDs, mm | 24.0 (21.0–27.0) | 24.0 (21.0–27.0) | 25.0 (21.0–29.0) | 0.079 |
| Interventricular septum thickness, mm | 11.0 (10.0–11.0) | 10.4 (10.0–11.0) | 11.0 (10.0–11.5) | 0.671 |
| Posterior wall thickness, mm | 10.0 (9.0–11.0) | 10.0 (9.0–11.0) | 10.0 (9.0–11.0) | 0.515 |
| LVMI, g/m2 | 101.6 (87.3–119.9) | 100.2 (86.1–114.3) | 111.7 (94.5–131.1) | 0.007 |
| LAVI, mL/m2 | 48.4 (42.2–59.8) | 46.8 (40.5–54.9) | 62.7 (49.9–72.6) | <0.001 |
| E wave, cm/s | 81.0 (65.0–101.0) | 78.0 (63.0–94.5) | 99.0 (74.3–118.0) | <0.001 |
| e′ wave, cm/s | 3.8±1.4 | 3.6±1.4 | 4.3±1.4 | <0.001 |
| E/e′ | 24.1±10.3 | 24.2±10.6 | 24.0±9.2 | 0.864 |
| SVI, mL/m2 | 48.3 (43.1–53.3) | 48.4 (43.8–53.0) | 46.9 (39.8–56.7) | 0.351 |
| AoV area, cm2 | 0.66 (0.58–0.73) | 0.67 (0.58–0.74) | 0.64 (0.54–0.72) | 0.147 |
| Mean PG, mmHg | 48.0 (39.0–62.0) | 48.0 (39.0–62.0) | 49.0 (35.5–61.0) | 0.480 |
| Peak PG, mmHg | 84.0 (70.0–108.0) | 85.5 (71.0–108.8) | 83.0 (65.5–100.5) | 0.185 |
| Peak velocity, m/s | 4.6 (4.2–5.2) | 4.6 (4.2–5.2) | 4.6 (4.0–5.0) | 0.110 |
| TRPG, mmHg | 25.0 (20.2–30.0) | 24.0 (20.0–28.0) | 28.0 (22.2–34.5) | <0.001 |
| Systolic pulmonary artery pressure, mmHg | 28.0 (23.0–33.0) | 27.0 (23.0–31.0) | 31.0 (25.2–37.5) | <0.001 |
| TAPSE, mm | 20.7±4.1 | 21.0±3.1 | 19.6±6.2 | 0.013 |
| MR none-to-mild, n (%) | 286 (90.2) | 242 (100.0) | 44 (58.7) | <0.001 |
| MR moderate, n (%) | 23 (7.3) | 0 (0.0) | 23 (30.7) | <0.001 |
| MR moderate-to-severe, n (%) | 5 (1.6) | 0 (0.0) | 5 (6.7) | 0.001 |
| MR severe, n (%) | 3 (0.9) | 0 (0.0) | 3 (4.0) | 0.013 |
Data are presented as mean±SD for normally distributed variables and as median (interquartile range) for non-normally distributed variables. Categorical variables are presented as n (%). P values were calculated using Student’s t-test or the Mann–Whitney U test for continuous variables, and the chi-square test or Fisher’s exact test for categorical variables, as appropriate. AF, atrial fibrillation; AoV, aortic valve; AS, aortic stenosis; BMI, body mass index; BNP, B-type natriuretic peptide; CFS, Clinical Frailty Scale; eGFR, estimated glomerular filtration rate; LAE, left atrial enlargement; LAVI, left atrial volume index; LVDd, left ventricular end-diastolic diameter; LVDs, left ventricular end-systolic diameter; LVEF, left ventricular ejection fraction; LVMI, left ventricular mass index; MR, mitral regurgitation; PG, pressure gradient; SVI, stroke volume index; TAPSE, tricuspid annular plane systolic excursion; TRPG, tricuspid regurgitation pressure gradient.
Echocardiographic Findings
Echocardiography confirmed a larger LA phenotype in the AF/MR-associated group, with higher LAVI (62.7 [49.9–72.6] vs. 46.8 [40.5–54.9] mL/m2; P<0.001) and ≥moderate MR present exclusively in this subgroup (Table 1). Among patients with AF/MR-associated LAE, 44 had AF alone, 22 had ≥moderate MR alone, and 9 had both AF and ≥moderate MR. No patients in the final analytic cohort had moderate or severe mitral stenosis. Left ventricular size and aortic valve hemodynamics were comparable, whereas markers of right-sided load and function (including TR pressure gradient and TAPSE) differed between groups (Table 1).
Clinical Outcomes
During a median follow-up of 2.7 years (IQR 1.4–4.0), 69 patients (21.8%) died (Table 2). The all-cause mortality rate was higher in the AF/MR-associated group than in the AS-associated group (30.7% vs. 19.0%; P<0.01), and Kaplan–Meier curves showed significantly lower survival (log-rank P<0.01; Figure 2). Cardiovascular death was infrequent and similar between groups (2.7% vs. 2.5%; P=1.00). Among 69 deaths, 8 were adjudicated as cardiovascular, 36 as confirmed non-cardiovascular, and 25 were of undetermined cause. The causes of cardiovascular death were heart failure in 3 patients, stroke in 4 patients, and other cardiovascular cause in 1 patient. Among confirmed non-cardiovascular deaths, the most frequent causes were infection in 18 patients, malignancy in 6 patients, respiratory disease in 2 patients, senescence/frailty in 4 patients, and other non-cardiovascular causes in 6 patients. The details of cardiovascular, confirmed non-cardiovascular, and undetermined deaths are shown in Supplementary Table 1. Because the cause of death could not be determined in a substantial proportion of patients, cause-specific outcomes were considered exploratory and interpreted cautiously. In the entire LAE cohort, AF/MR-associated LAE was associated with a higher risk of all-cause death compared with AS-associated LAE in the unadjusted Cox model (HR, 1.93; 95% CI, 1.17–3.19; P=0.010). This association remained significant after adjustment for age and sex (HR, 1.98; 95% CI, 1.20–3.28; P=0.008). However, after further adjustment for age, sex, eGFR, log-BNP, and LAVI, the association was attenuated and no longer statistically significant (HR, 1.48; 95% CI, 0.85–2.58; P=0.171). In this fully adjusted model, log-BNP remained significantly associated with all-cause death. The proportional hazards assumption was assessed using Schoenfeld residuals, and no clear violation was observed in the fully adjusted model. These results are shown in Supplementary Table 2.
Table 2.
Clinical Outcomes After TAVI According to LAE Subtype
| Outcome | Overall (N=317) |
AS-associated LAE group (N=242) |
AF/MR-associated LAE group (N=75) |
P value |
|---|---|---|---|---|
| All-cause death, n (%) | 69 (21.8) | 46 (19.0) | 23 (30.7) | <0.01 |
| CV death, n (%) | 8 (2.5) | 6 (2.5) | 2 (2.7) | 1.000 |
| Follow-up duration, years | 2.7 (1.4–4.0) | 2.9 (1.5–4.1) | 2.0 (1.2–3.2) | – |
Values are n (%) or median (interquartile range). P values were calculated using the chi-square test or Fisher’s exact test, as appropriate. Cardiovascular (CV) death was adjudicated by chart review. Deaths with insufficient information to determine CV or non-CV cause were classified as deaths of undetermined cause and were not included as confirmed non-CV deaths. Cause-specific outcomes were considered exploratory because of the limited number of adjudicated CV deaths and the presence of deaths of undetermined cause. TAVI, transcatheter aortic valve implantation. Other abbreviations as in Table 1.
Figure 2.

Kaplan–Meier curves for all-cause mortality after transcatheter aortic valve implantation according to LAE subtype. Survival was compared between the AS-associated LAE group and the AF/MR-associated LAE group using the log-rank test. Abbreviations: AF, atrial fibrillation; AS, aortic stenosis; LAE, left atrial enlargement; MR, mitral regurgitation.
Predictors of All-Cause Death According to LAE Subtype
Cox models were fitted separately by LAE subtype (Table 3). In the AS-associated LAE group, higher log-BNP (adjusted HR 1.45, 95% CI 1.06–1.98; P=0.02), higher E/e′ (adjusted HR 1.03, 95% CI 1.01–1.06; P=0.008), and lower TAPSE (adjusted HR 0.89, 95% CI 0.81–0.99; P=0.032) were independently associated with death. In the AF/MR-associated LAE group, lower eGFR (adjusted HR 0.96, 95% CI 0.93–1.00; P=0.028) and higher log-BNP (adjusted HR 1.98, 95% CI 1.09–3.59; P=0.025) were independently associated with death after adjustment for age. Younger age was also associated with death in this subgroup (adjusted HR 0.90, 95% CI 0.84–0.97; P=0.008), although this finding was considered exploratory because of the limited sample size and potential selection bias. In the hypothesis-driven interaction analyses, the prognostic effects of eGFR and E/e′ differed by LAE subtype (P for interaction=0.049 and 0.037, respectively; Table 3), suggesting subtype-specific differences in risk interpretation.
Table 3.
Cox Regression Predictors of All-Cause Death by LAE Subtype
| Variable | AS-associated LAE group (N=242) | AF/MR-associated LAE group (N=75) | P for interaction by LAE subtype |
||||||
|---|---|---|---|---|---|---|---|---|---|
| Unadjusted HR (95% CI) |
P value | Adjusted HR (95% CI) |
P value | Unadjusted HR (95% CI) |
P value | Adjusted HR (95% CI) |
P value | ||
| Age | Stratified; HR not estimated |
– | Stratified; HR not estimated |
– | 0.93 (0.87–1.01) |
0.070 | 0.90 (0.84–0.97) |
0.008 | NA† |
| BMI | 0.98 (0.90–1.07) |
0.720 | 1.02 (0.92–1.14) |
0.710 | |||||
| CFS | 1.12 (0.80–1.58) |
0.510 | 1.04 (0.64–1.68) |
0.880 | |||||
| Logistic EuroSCORE | 1.01 (0.97–1.05) |
0.770 | 0.99 (0.95–1.03) |
0.640 | |||||
| eGFR | 1.00 (0.98–1.02) |
0.940 | 1.00 (0.99–1.02) |
0.65 | 0.96 (0.94–0.99) |
0.011 | 0.96 (0.93–1.00) |
0.028 | 0.049 |
| log-BNP* | 1.63 (1.21–2.19) |
0.001 | 1.45 (1.06–1.98) |
0.02 | 1.98 (1.13–3.44) |
0.016 | 1.98 (1.09–3.59) |
0.025 | 0.488 |
| LVEF | 0.94 (0.89–1.00) |
0.054 | 1.00 (0.91–1.10) |
0.980 | |||||
| LVDd | 0.98 (0.92–1.04) |
0.450 | 0.91 (0.84–0.99) |
0.019 | |||||
| LVDs | 1.01 (0.96–1.07) |
0.660 | 0.90 (0.82–0.98) |
0.021 | |||||
| LVMI | 1.01 (1.00–1.02) |
0.270 | 1.00 (0.99–1.01) |
0.700 | |||||
| LAVI | 1.01 (0.99–1.03) |
0.310 | 0.99 (0.98–1.01) |
0.530 | |||||
| E/e′ | 1.04 (1.02–1.07) |
<0.001 | 1.03 (1.01–1.06) |
0.008 | 0.98 (0.93–1.02) |
0.360 | 0.037 | ||
| Mean AoV PG | 1.00 (0.99–1.02) |
0.630 | 0.98 (0.96–1.00) |
0.091 | |||||
| Peak AoV V | 0.95 (0.65–1.39) |
0.800 | 0.82 (0.50–1.32) |
0.410 | |||||
| TRPG | 1.01 (0.98–1.05) |
0.450 | 1.01 (0.97–1.06) |
0.580 | |||||
| TAPSE | 0.87 (0.79–0.97) |
0.009 | 0.89 (0.81–0.99) |
0.032 | 0.95 (0.85–1.07) |
0.420 | 0.119 | ||
Data are hazard ratios (HR) with 95% confidence intervals (CI). Continuous variables are reported per 1-unit increase; units are indicated in the variable labels. Adjusted HRs are shown only for covariates included in the prespecified multivariable models; blank cells indicate variables not entered the multivariable model. Age was included as an adjustment covariate in the AF/MR-associated LAE model; its association with mortality was considered exploratory. †In the AS-associated LAE Cox model, age violated the proportional hazards assumption; therefore, an age-stratified Cox model was used and the hazard ratio for age was not estimated. AVA, aortic valve area. Other abbreviations as in Table 1.
Discussion
The present study results demonstrated that, among patients with severe AS undergoing TAVI with preserved LVEF, the prognostic implications of LAE differ according to its clinically attributed subtype. Because LAE is highly prevalent and commonly treated as a single phenotype, we intentionally restricted the analysis to patients with LAE to examine heterogeneity within this structural finding. In this setting, AS-associated LAE provides a relatively homogeneous context in which LAE is more likely to reflect chronic pressure overload and residual diastolic dysfunction, whereas AF/MR-associated LAE is more strongly influenced by AF-related atrial remodeling and/or MR-related volume overload.
An important consideration is that the observed differences in outcomes are unlikely to be explained solely by the presence or absence of AF or MR. Although both conditions are established predictors of adverse outcomes after TAVI,19,20 our approach compared patients sharing the same structural phenotype (LAE) while accounting for differences in the presumed drivers of atrial enlargement. This framework provided a clinically practical way to test whether LAE arising from different disease processes carries different prognostic meaning after TAVI. By analyzing only patients with LAE, we minimized confounding by the presence or absence of LAE itself and focused on heterogeneity within this common structural phenotype.
Patients with AF/MR-associated LAE had a higher all-cause mortality rate than those with AS-associated LAE, indicating that LAE is not prognostically equivalent across clinical contexts. Prior studies have associated marked LAE and limited LA reverse remodeling after TAVI with worse long-term outcomes.11,12,19 Our findings extend this evidence by indicating that LAE is not a uniform risk marker and that the clinical drivers of LAE, particularly AF and MR, identify a clinically vulnerable phenotype after TAVI. In the whole-cohort Cox analysis, AF/MR-associated LAE was associated with higher all-cause death in the unadjusted and age/sex-adjusted models; however, this association was attenuated after further adjustment for eGFR, log-BNP, and LAVI. These findings suggested that AF/MR-associated LAE should not be interpreted as a completely independent prognostic factor beyond established markers of systemic vulnerability, neurohumoral activation, and LA remodeling. Rather, AF/MR-associated LAE appears to represent a clinically relevant high-risk phenotype in which excess risk is partly explained by renal dysfunction, elevated BNP, and greater LAE. In analyses restricted to patients with LAE, we observed subtype-specific prognostic patterns, suggesting that the prognostic meaning of LAE depends on its clinical context and the downstream hemodynamic and systemic consequences associated with AF and MR rather than the presence of AF or MR alone. Consistent with this interpretation, in patients with AF/MR-associated LAE, reduced renal function (lower eGFR) and higher BNP were independently associated with death, whereas conventional diastolic indices (E/e′) were not. Although log-BNP was independently associated with all-cause death in both LAE subtypes, adjudicated cardiovascular deaths were infrequent and 25 deaths were classified having an undetermined cause. Therefore, this association should not be interpreted as indicating that BNP specifically predicted cardiovascular or non-cardiovascular death. Rather, BNP may have served as an integrated marker of overall clinical vulnerability, reflecting chronic hemodynamic stress, subclinical congestion, renal dysfunction, frailty, and systemic disease burden. Because of the limited number of adjudicated cardiovascular deaths and the presence of deaths of undetermined cause, we did not perform cause-specific Cox analyses, and the prognostic role of BNP for cause-specific death should be evaluated in larger studies. Although residual confounding and reverse causation cannot be excluded, these findings suggested that, in this subgroup, LAE predominantly reflects advanced atrial cardiomyopathy and systemic disease burden, potentially mediated by congestion-related end-organ dysfunction, rather than primarily reflecting AS-related ventricular pathology.21–23 The inverse association between age and death observed in the AF/MR-associated LAE subgroup should be interpreted cautiously, as it may reflect residual selection bias, frailty-related referral patterns or the limited sample size rather than a true biological effect. In contrast, in the AS-associated LAE group, higher E/e′ and reduced TAPSE remained independently associated with death, supporting a pathway linking elevated filling pressures to pulmonary hypertension and right ventricular dysfunction that may not fully reverse after valve replacement.
The observed interactions between LAE subtype and eGFR or E/e′ suggest that identical clinical markers may carry different prognostic meanings across subtypes. In AS-associated LAE patients, E/e′ likely captures residual pressure overload and diastolic dysfunction that may persist after TAVI,24,25 and thus remains prognostically relevant. Consistent with this concept, a recent study showed that E/e′ before and after TAVI was associated with early postprocedural BNP, supporting the clinical relevance of Doppler-derived diastolic indices for detecting residual LV burden after TAVI.26 In AF/MR-associated LAE patients, irregular rhythm, beat-to-beat variability, and MR-related volume overload may limit the ability of E/e′, particularly septal E/e′, to accurately represent LV filling pressures. Therefore, E/e′ should be interpreted as an exploratory Doppler marker rather than as a definitive estimate of filling pressure in this subgroup.27,28 This limitation may partly explain the weaker prognostic signal of E/e′ in the AF/MR-associated LAE group compared with the AS-associated LAE group. These findings support the interpretation of LAE and related markers in a disease-specific manner rather than equating LAE uniformly with LA/mitral damage.
Clinically, our results suggested that risk stratification in patients undergoing TAVI should not rely solely on the LAVI value, but should also consider the underlying drivers of LAE, such as AF and MR. In patients with AS-associated LAE (no AF and <moderate MR), detailed assessment of diastolic function and right ventricular performance appears particularly relevant for identifying residual hemodynamic burden. Conversely, in AF/MR-associated LAE patients, comprehensive evaluation emphasizing renal function, neurohumoral activation, and frailty may be more informative for long-term risk stratification and follow-up, and may help guide multidisciplinary management. Further multicenter studies incorporating post-TAVI remodeling and time-varying analyses are needed to confirm these subtype-specific interpretations.
Study Limitations
There are several to note. First, because this study was a single-center retrospective analysis and the sample size was not large; therefore, generalizability of the findings may be limited. Second, the extensive exclusion criteria may introduce selection bias; however, they were intentionally applied to reduce pathophysiological heterogeneity and to address a focused mechanistic question rather than to represent an unselected TAVI population. Third, LAE subtyping was based on clinical information (AF history/ECG and guideline-based MR grading) and the AF/MR-associated group included patients with AF only, MR only, and both conditions, leaving residual heterogeneity. In addition, right ventricular systolic function was primarily assessed using TAPSE, and other guideline-recommended indices (e.g., S′ or RV fractional area change) were not included due to missing data. Fourth, unmeasured confounding remains possible despite multivariable adjustment, and the interaction analyses may be underpowered. Finally, cause-specific death was considered exploratory because adjudicated cardiovascular deaths were few and a substantial number of deaths had an undetermined cause. Larger multicenter studies with more granular phenotyping and post-TAVI remodeling assessments are warranted.
Conclusions
In patients with severe AS and LVEF undergoing TAVI, the prognostic significance and interpretation of LAE vary according to its underlying clinically attributed subtype. AF/MR-associated LAE (AF and/or ≥moderate MR) may predominantly reflect advanced atrial cardiomyopathy and systemic disease burden, whereas AS-associated LAE (no AF and <moderate MR) may more closely reflect residual LV diastolic dysfunction and secondary right ventricular impairment. These findings highlight the importance of clinically attributed LAE subtyping for improving post-TAVI risk assessment and informing individualized clinical management.
Disclosures
T.S., A.I., Y.K., S.I., H.Y., A.O., N.N., T.Y., S.K., M.O., T.S., S.I., Y.H. and A.S. have no conflicts of interest to declare. K.O. received scholarships or donations from Abbott Medical Japan LLC. D.F. received honoraria for speakers’ bureaus from Nippon Boehringer Ingelheim Co., Ltd, Daiichi Sankyo Co., Ltd., Novartis Pharma K.K., Otsuka Pharmaceutical Co., Ltd., and Kowa Co., Ltd.
IRB Information
This study was approved by the Ethics Committee of Osaka Metropolitan University Graduate School of Medicine (Approval No. 2021-064).
Supplementary Files
Supplementary Table 1. Supplementary Table 2.
Acknowledgments
The authors thank all members of the Osaka Metropolitan University Heart Team. AI technology (ChatGPT, OpenAI) was used to help with the English language editing of this manuscript.
Funding Statement
Sources of Funding: This research did not receive any specific grant from funding agencies in the public, commercial, or non-profit sectors.
Data Availability
The deidentified participant data will not be shared.
References
- 1. Leon MB, Smith CR, Mack M, Miller DC, Moses JW, Svensson LG, et al.. Transcatheter aortic-valve implantation for aortic stenosis in patients who cannot undergo surgery. N Engl J Med 2010; 363: 1597–1607, doi:10.1056/NEJMoa1008232. [DOI] [PubMed] [Google Scholar]
- 2. Kemi Y, Yamashita E, Kario K, Kinoshita S, Sugano K, Yano H, et al.. Clinical impact of preoperative symptoms of aortic stenosis on prognosis after transcatheter aortic valve replacement. Circ Rep 2024; 6: 223–229, doi:10.1253/circrep.CR-24-0020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Généreux P, Pibarot P, Redfors B, Mack MJ, Makkar RR, Jaber WA, et al.. Staging classification of aortic stenosis based on the extent of cardiac damage. Eur Heart J 2017; 38: 3351–3358, doi:10.1093/eurheartj/ehx381. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Vollema EM, Amanullah MR, Ng ACT, van der Bijl P, Prevedello F, Sin YK, et al.. Staging cardiac damage in patients with symptomatic aortic valve stenosis. J Am Coll Cardiol 2019; 74: 538–549, doi:10.1016/j.jacc.2019.05.048. [DOI] [PubMed] [Google Scholar]
- 5. Dalsgaard M, Egstrup K, Wachtell K, Gerdts E, Cramariuc D, Kjaergaard J, et al.. Left atrial volume in patients with asymptomatic aortic valve stenosis (the Simvastatin and Ezetimibe in Aortic Stenosis study). Am J Cardiol 2008; 101: 1030–1034, doi:10.1016/j.amjcard.2007.11.048. [DOI] [PubMed] [Google Scholar]
- 6. Rusinaru D, Bohbot Y, Salaun E, Donal E, Calsata AC, Galli E, et al.. Determinants of left atrial volume index in patients with aortic stenosis: A multicentre pilot study. Arch Cardiovasc Dis 2017; 110: 525–533, doi:10.1016/j.acvd.2016.12.016. [DOI] [PubMed] [Google Scholar]
- 7. Abhayaratna WP, Seward JB, Appleton CP, Douglas PS, Oh JK, Tajik AJ, et al.. Left atrial size: Physiologic determinants and clinical applications. J Am Coll Cardiol 2006; 47: 2357–2363, doi:10.1016/j.jacc.2006.02.048. [DOI] [PubMed] [Google Scholar]
- 8. Nagueh SF, Sanborn DY, Oh JK, Anderson B, Billick K, Derumeaux G, et al.. Recommendations for the evaluation of left ventricular diastolic function by echocardiography and for heart failure with preserved ejection fraction diagnosis: An update from the American Society of Echocardiography. J Am Soc Echocardiogr 2025; 38: 537–569, doi:10.1016/j.echo.2025.03.011. [DOI] [PubMed] [Google Scholar]
- 9. Pritchett AM, Mahoney DW, Jacobsen SJ, Rodeheffer RJ, Karon BL, Redfield MM.. Diastolic dysfunction and left atrial volume: A population-based study. J Am Coll Cardiol 2005; 45: 87–92, doi:10.1016/j.jacc.2004.09.054. [DOI] [PubMed] [Google Scholar]
- 10. Tsang TSM, Barnes ME, Gersh BJ, Bailey KR, Seward JB.. Left atrial volume as a morphophysiologic expression of left ventricular diastolic dysfunction and relation to cardiovascular risk burden. Am J Cardiol 2002; 90: 1284–1289, doi:10.1016/s0002-9149(02)02864-3. [DOI] [PubMed] [Google Scholar]
- 11. Rusinaru D, Bohbot Y, Kowalski C, Ringle A, Maréchaux S, Tribouilloy C.. Left atrial volume and mortality in patients with aortic stenosis. J Am Heart Assoc 2017; 6: e006615, doi:10.1161/JAHA.117.006615. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Ito A, Izumiya Y, Iwata S, Ogawa M, Kim AT, Yahiro R, et al.. Left atrial volume index predicts future improvement of B-type natriuretic peptide levels after transcatheter aortic valve replacement. Int J Cardiovasc Imaging 2022; 38: 1463–1471, doi:10.1007/s10554-022-02538-7. [DOI] [PubMed] [Google Scholar]
- 13. Testa L, Latib A, De Marco F, De Carlo M, Fiorina C, Barbanti M, et al.. The failing right heart: Implications and evolution in high-risk patients undergoing transcatheter aortic valve implantation. EuroIntervention 2016; 12: 1542–1549, doi:10.4244/EIJD-15-00148. [DOI] [PubMed] [Google Scholar]
- 14. Joglar JA, Chung MK, Armbruster AL, Benjamin EJ, Chyou JY, Cronin EM, et al.. 2023 ACC/AHA/ACCP/HRS Guideline for the diagnosis and management of atrial fibrillation: A report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation 2024; 149: e1–e156, doi:10.1161/CIR.0000000000001193. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Zoghbi WA, Adams D, Bonow RO, Enriquez-Sarano M, Foster E, Grayburn PA, et al.. Recommendations for noninvasive evaluation of native valvular regurgitation: A report from the American Society of Echocardiography Developed in Collaboration with the Society for Cardiovascular Magnetic Resonance. J Am Soc Echocardiogr 2017; 30: 303–371, doi:10.1016/j.echo.2017.01.007. [DOI] [PubMed] [Google Scholar]
- 16. Lang RM, Badano LP, Mor-Avi V, Afilalo J, Armstrong A, Ernande L, et al.. Recommendations for cardiac chamber quantification by echocardiography in adults: An update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. J Am Soc Echocardiogr 2015; 28: 1–39.e14, doi:10.1016/j.echo.2014.10.003. [DOI] [PubMed] [Google Scholar]
- 17. Baumgartner H, Hung J, Bermejo J, Chambers JB, Edvardsen T, Goldstein S, et al.. Recommendations on the echocardiographic assessment of aortic valve stenosis: A Focused Update from the European Association of Cardiovascular Imaging and the American Society of Echocardiography. J Am Soc Echocardiogr 2017; 30: 372–392, doi:10.1016/j.echo.2017.02.009. [DOI] [PubMed] [Google Scholar]
- 18. Mukherjee M, Rudski LG, Addetia K, Afilalo J, D’Alto M, Freed BH, et al.. Guidelines for the echocardiographic assessment of the right heart in adults and special considerations in pulmonary hypertension: Recommendations from the American Society of Echocardiography. J Am Soc Echocardiogr 2025; 38: 141–186, doi:10.1016/j.echo.2025.01.006. [DOI] [PubMed] [Google Scholar]
- 19. Asami M, Dobner S, Stortecky S, Heg D, Praz F, Lanz J, et al.. Cardiovascular outcomes in patients with left atrial enlargement undergoing transcatheter aortic valve implantation. Catheter Cardiovasc Interv 2022; 99: 1908–1917, doi:10.1002/ccd.30132. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Miura M, Yamaji K, Shirai S, Hayashi M, Kawaguchi T, Arai Y, et al.. Clinical impact of preprocedural moderate or severe mitral regurgitation on outcomes after transcatheter aortic valve replacement. Can J Cardiol 2020; 36: 1112–1120, doi:10.1016/j.cjca.2019.12.022. [DOI] [PubMed] [Google Scholar]
- 21. Marrouche NF, Wilber D, Hindricks G, Jais P, Akoum N, Marchlinski F, et al.. Association of atrial tissue fibrosis identified by delayed enhancement MRI and atrial fibrillation catheter ablation: The DECAAF study. JAMA 2014; 311: 498–506, doi:10.1001/jama.2014.3. [DOI] [PubMed] [Google Scholar]
- 22. Narducci ML, Volpe M.. Atrial fibrosis detected by cardiac magnetic resonance in persistent atrial fibrillation: A useful risk stratifier but not ideal electrophysiological endpoint for catheter ablation. Eur Heart J 2022; 43: 3196–3197, doi:10.1093/eurheartj/ehac424. [DOI] [PubMed] [Google Scholar]
- 23. Coisne A, Lancellotti P, Vannan MA.. Mitral regurgitation, the left atrium and atrial fibrillation: Unlikely bedfellows or natural kindreds? Eur Heart J Cardiovasc Imaging 2024; 25: 587–588, doi:10.1093/ehjci/jeae036. [DOI] [PubMed] [Google Scholar]
- 24. Seo J, Kim AR, Cho I, Shim CY, Hong GR, Choi EY, et al.. Impact of left ventricular diastolic pressure changes on clinical outcomes after transcatheter aortic valve replacement. J Am Heart Assoc 2025; 14: e039372, doi:10.1161/JAHA.124.039372. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Nair RM, Chawla S, Verma B, Kumar S, Abou Hassan O, Ghimire B, et al.. Impact of elevated left ventricular filling pressure on long-term outcomes after transcatheter aortic valve replacement. Open Heart 2022; 9: e002015, doi:10.1136/openhrt-2022-002015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Ikenaga T, Kato Y, Kawahira Y, Miyazaki M, Hirata T, Teratani H, et al.. Early changes in left ventricular myocardial mechanics after transcatheter aortic valve replacement for severe aortic stenosis. Circ Rep 2025; 7: 365–371, doi:10.1253/circrep.CR-24-0139. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Olson JJ, Costa SP, Young CE, Palac RT.. Early mitral filling/diastolic mitral annular velocity ratio is not a reliable predictor of left ventricular filling pressure in the setting of severe mitral regurgitation. J Am Soc Echocardiogr 2006; 19: 83–87, doi:10.1016/j.echo.2005.07.005. [DOI] [PubMed] [Google Scholar]
- 28. Pak M, Kitai T, Kobori A, Sasaki Y, Okada T, Murai R, et al.. Diagnostic accuracy of the 2016 guideline-based echocardiographic algorithm to estimate invasively-measured left atrial pressure by direct atrial cannulation. JACC Cardiovasc Imaging 2022; 15: 1683–1691, doi:10.1016/j.jcmg.2022.03.022. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Table 1. Supplementary Table 2.
Data Availability Statement
The deidentified participant data will not be shared.
