Outcomes of surgical and transcatheter aortic valve replacement in patients with large and extra-large aortic annuli Authors: Xander Jacquemyn, Justin C Wang, Dustin Kliner, Derek Serna-Gallegos, Irsa Hasan, Takuya Ogami, Floyd Thoma, Yisi Wang, Catalin Toma, David West, Amber Makani, Ibrahim Sultan Published: 2026 ABSTRACT AbstractObjectiveThe study objective was to evaluate outcomes after aortic valve replacement in patients with large or extra-large annuli.MethodsWe conducted a retrospective cohort study of adults undergoing isolated transcatheter aortic valve replacement or surgical aortic valve replacement from 2021 to 2025 at a single quaternary care center. Eligible patients had large (annular area ≥575 mm2 or perimeter ≥85 mm) or extra-large (annular area ≥683 mm2 or perimeter ≥94.2 mm) trileaflet aortic annuli on preprocedural computed tomography.ResultsA total of 414 patients underwent aortic valve replacement, including 353 transcatheter aortic valve replacements and 61 surgical aortic valve replacements. The median age was 78.8 years (interquartile range, 73.4-83.8), and 20 patients (4.8%) were female. Extra-large annuli were present in 109 patients (26%). Transcatheter aortic valve replacement recipients were older (median 80.2 vs 68.0 years, P < .001) and had a higher burden of comorbidities, including heart failure, atrial fibrillation, diabetes, and prior myocardial infarction (all P < .001). Early 30-day mortality was low, occurring in 3 patients who received transcatheter aortic valve replacement (0.9%) and none of the patients who received surgical aortic valve replacement (0%, P = .470). Compared with an age- and sex-matched US population, patients who received surgical aortic valve replacement had numerically lower mortality (standardized mortality ratio, 0.29, 95% CI, 0.04-2.03), whereas patients who received transcatheter aortic valve replacement experienced higher-than-expected mortality (standardized mortality ratio, 1.69, 95% CI, 1.35-2.11). Heart failure rehospitalization was significantly more frequent after transcatheter aortic valve replacement even after adjustment for baseline characteristics (hazard ratio, 5.09, 95% CI, 1.88-13.80, P = .001).ConclusionsAortic valve replacement in patients with large or extra-large annuli is safe, with low early mortality. Differences in long-term outcomes primarily reflect differences in baseline patient risk profiles rather than treatment strategy.Key Words: aortic valve replacement, heart failure, large aortic annulus, surgical aortic valve replacement, transcatheter aortic valve replacement Standardized mortality ratios for SAVR versus TAVR in large aortic annuli. Central Message In patients with large aortic annuli, both TAVR and SAVR can be performed safely with low early mortality. Long-term outcome differences are primarily due to baseline patient risk profiles. SAVR achieves survival comparable to population norms in appropriately selected patients, whereas TAVR outcomes may reflect the higher comorbidity burden of those selected for TAVR. Perspective Optimal valve selection in patients with large annuli is uncertain. This study demonstrates durable outcomes with both TAVR and SAVR, while highlighting that SAVR achieves survival comparable to the general population. These findings emphasize patient selection, comorbidity burden, and timing of intervention in procedural decision-making. Surgical aortic valve replacement (SAVR) and transcatheter aortic valve replacement (TAVR) are established treatments for aortic valve disease.1 Although TAVR has demonstrated favorable outcomes across various patient populations, its safety and efficacy in patients with large or extra-large aortic annuli remain unclear. Large annuli, previously defined as an area 575 mm2 or more or perimeter 85 mm or more, are present in a clinically relevant subset of patients and pose unique procedural challenges, including difficulties with valve sizing, risk of device malposition, and higher rates of paravalvular leak due to suboptimal prosthesis-annulus apposition.2 Currently available transcatheter heart valves have defined upper annular size limits. According to device specifications, Evolut R or Evolut PRO+ 34 mm valves (Medtronic) can accommodate annuli up to a perimeter of 94.2 mm, and the 29-mm SAPIEN 3 (Edwards Lifesciences) can accommodate annuli up to 683 mm2. Therefore, patients with extra-large annuli may represent the anatomic limits of current transcatheter platforms. Nonetheless, several studies have reported use of these platforms in annuli that are outside these recommended size ranges.3,4 In contrast, surgical prostheses can accommodate a wide range of annular sizes, offering greater flexibility in the treatment of patients with large or extra-large anatomy. Although TAVR has demonstrated favorable outcomes across a range of patient populations and annular anatomies, the clinical profile of patients with large or extra-large trileaflet aortic annuli remains incompletely characterized, particularly in the context of both transcatheter and surgical strategies. We aimed to describe the clinical characteristics and outcomes of patients with large or extra-large trileaflet aortic valves undergoing isolated TAVR or SAVR. Material and Methods This retrospective cohort study included consecutive adult patients who underwent isolated TAVR or SAVR at a single quaternary care center between January 1, 2021, and January 1, 2025. Patients were identified from a prospectively maintained institutional valve registry (STUDY18120143, approved 4/17/2019). Eligible patients had trileaflet aortic valve morphology and large or extra-large aortic annuli as determined by preprocedural computed tomography angiography. Large annuli were defined as an annular area 575 mm2 or greater or perimeter 85 mm or greater, and extra-large annuli were defined as an annular area 683 mm2 or greater or perimeter 94.2 mm or greater, as consistent with the TAVR-LARGE Registry.2 Only patients undergoing isolated aortic valve replacement for primary aortic stenosis or mixed aortic stenosis and regurgitation were included. Individual treatment strategies were determined by a multidisciplinary heart team in all cases, based on surgical risk, suitability for TAVR, aortic root anatomy, vascular access, age, comorbidities, and patient preference in accordance with institutional practice and current guidelines. Exclusion criteria comprised bicuspid aortic valve anatomy, valve-in-valve TAVR, SAVR performed with concomitant cardiac surgical procedures, and absence of computed tomography angiography–derived annular measurements. Baseline demographic characteristics, comorbid conditions, and procedural details were abstracted from the institutional registry and electronic health records. Procedural outcomes, in-hospital events, and longitudinal follow-up were obtained through systematic chart review. All imaging measurements were performed according to standardized institutional protocols. Patients were categorized according to treatment strategy (TAVR or SAVR) for analysis. The primary outcome was all-cause mortality during follow-up, and secondary outcomes included heart failure hospitalization, stroke, and aortic valve reintervention. Statistical Analysis Continuous variables were summarized as mean ± SD or median with interquartile range and compared using Student t test or Mann–Whitney U test, as appropriate, whereas categorical variables were compared using chi-square or Fisher exact tests. Propensity scores were calculated to evaluate differences in baseline characteristics between groups and plotted to visualize the distribution. Survival analyses were conducted using Kaplan–Meier estimates with log-rank testing to compare SAVR and TAVR. Standardized expected mortality was estimated using age- and sex-matched survivals from the 2021 US Life Tables for the complete US population, as reported by the Centers for Disease Control and Prevention.5 Associations between treatment strategy and mortality were assessed using Cox proportional hazards regression. Given the low number of deaths in the SAVR cohort, Firth's penalized likelihood method was applied to obtain stable hazard ratio estimates. Variables included in the multivariable models were selected based on univariable significance and clinical judgment, taking into account clinically relevant predictors, with age and sex always retained in the fully adjusted models. For outcomes in which death acted as a competing risk, such as heart failure hospitalization, cumulative incidence curves were generated, differences between groups were assessed using Gray's test, and subdistribution hazard ratios were obtained from competing-risk regression models. All analyses were performed in the full cohort, using SAS version 9.4 (SAS Institute, Inc). Results After excluding 96 patients with bicuspid anatomy (of whom 65 [67.7%] underwent SAVR and 31 [32.3%] underwent TAVR) and 53 patients with valve-in-valve TAVR procedures, a total of 414 patients with large or extra-large annuli met the inclusion criteria, including 353 who underwent TAVR and 61 who underwent SAVR. Baseline characteristics are summarized in Table 1. Patients undergoing TAVR represented a significantly different clinical phenotype compared with those undergoing SAVR, as illustrated by the propensity score plot showing minimal overlap between groups (Figure 1). Patients treated with TAVR were significantly older than those undergoing SAVR (median 80.2 vs 68.0 years, P < .001) and had higher baseline morbidity across several domains (Table 1). The median STS risk score was 2.8 in the TAVR group compared with 0.9 in the SAVR group (P < .001). Severe aortic stenosis was present in 341 TAVR patients (96.6%) and 60 SAVR patients (98.4%), with preprocedural mean aortic gradients higher in SAVR than TAVR (median 47.2 mm Hg vs 43.0 mm Hg, P = .001). Congestive heart failure was more common in the TAVR cohort than in the SAVR cohort (87.5% vs 29.5%, P < .001), and correspondingly, baseline left ventricular ejection fraction was lower in the TAVR cohort (median 58.0% vs 62.0%, P < .001). Table 1Baseline characteristics of patients undergoing transcatheter aortic valve replacement versus surgical aortic valve replacement for large and extra-large aortic annuliVariableOverall (N = 414)SAVR (n = 61)TAVR (n = 353)P valueAge, y78.8 (73.4-83.8)68.0 (65.0-72.0)80.2 (75.4-85.2)<.001Female sex20 (4.8%)7 (11.5%)13 (3.7%).009White411 (99.3%)60 (98.4%)351 (99.4%).340BMI, kg/m229.4 (26.4-32.7)31.3 (28.2-34.8)29.2 (26.3-32.5).043STS predicted risk of mortality (%)2.6 (1.5-4.1)0.9 (0.8-1.3)2.8 (1.9-4.6)<.001Diabetes160 (38.6%)17 (27.9%)143 (40.5%).061COPD91 (22.0%)18 (29.5%)73 (20.7%).124Cerebrovascular disease90 (21.7%)10 (16.4%)80 (22.7%).273Hypertension380 (91.8%)52 (85.2%)328 (92.9%).044Prior myocardial infarction109 (26.3%)5 (8.2%)104 (29.5%).001Atrial fibrillation169 (40.8%)10 (16.4%)159 (45.0%)<.001Congestive heart failure327 (79.0%)18 (29.5%)309 (87.5%)<.001Prior revascularization55 (13.3%)6 (9.8%)49 (13.9%).045Prior pacemaker53 (12.8%)0 (0.0%)53 (15.0%)<.001Current dialysis6 (1.4%)0 (0.0%)6 (1.7%).305eGFR (CKD-EPI), mL/min/1.73 m270.4 (54.0-85.3)78.5 (70.2-91.1)66.8 (52.1-84.5)<.001Serum creatinine, mg/dL1.1 (0.9-1.3)1.0 (0.9-1.1)1.1 (0.9-1.4).003Hemoglobin, g/dL12.9 (11.6-14.1)14.1 (13.0-14.9)12.8 (11.5-13.8)<.001Echocardiography Aortic valve area, cm20.7 (0.6-0.9)0.8 (0.7-0.9)0.7 (0.6-0.8).016 Indexed aortic valve area, cm2/m20.3 (0.3-0.4)0.4 (0.3-0.4)0.3 (0.3-0.4).069 Mean aortic valve gradient, mm Hg44.0 (37.4-52.0)47.2 (41.4-59.0)43.0 (36.0-51.0).001 Left ventricular ejection fraction, %58.0 (48.0-63.0)62.0 (58.0-63.0)58.0 (45.0-60.0)<.001 Mitral regurgitation severity<.001 None44 (10.6%)11 (18.0%)33 (9.3%) Trivial/trace91 (22.0%)22 (36.1%)69 (19.5%) Mild165 (39.9%)23 (37.7%)142 (40.2%) Moderate114 (27.5%)5 (8.2%)109 (30.9%) Aortic insufficiency severity.076 None95 (22.9%)21 (34.4%)74 (21.0%) Trivial/trace86 (20.8%)11 (18.0%)75 (21.2%) Mild139 (33.6%)14 (23.0%)125 (35.4%) Moderate94 (22.7%)15 (24.6%)79 (22.4%)Computed tomography measurements Aortic annulus area, mm2626 (589-676)752 (653-829)618 (585-660)<.001 Aortic annulus perimeter, mm90.0 (87.2-94.4)101 (94.8-107)89.8 (87.0-92.7)<.001 Aortic annulus maximum diameter, mm32.0 (30.4-34.0)33.8 (32.5-35.5)31.5 (30.0-33.1)<.001SAVR, Surgical aortic valve replacement; TAVR, transcatheter aortic valve replacement; BMI, body mass index; STS, Society of Thoracic Surgeons; COPD, chronic obstructive pulmonary disease; eGFR, estimated glomerular filtration rate; CKD-EPI, Chronic Kidney Disease Epidemiology Collaboration. Baseline characteristics of patients undergoing transcatheter aortic valve replacement versus surgical aortic valve replacement for large and extra-large aortic annuli SAVR, Surgical aortic valve replacement; TAVR, transcatheter aortic valve replacement; BMI, body mass index; STS, Society of Thoracic Surgeons; COPD, chronic obstructive pulmonary disease; eGFR, estimated glomerular filtration rate; CKD-EPI, Chronic Kidney Disease Epidemiology Collaboration. Propensity score in the overall population. Propensity score density distribution between the TAVR and SAVR groups. Propensity scores were estimated using a logistic regression model including demographic, clinical, and echocardiographic variables. BMI, Body mass index; STS, Society of Thoracic Surgeons risk score; COPD, chronic obstructive pulmonary disease; CVD, cardiovascular disease; MI, myocardial infarction; AF, atrial fibrillation; CHF, congestive heart failure; LVEF, left ventricular ejection fraction; AVA, aortic valve area; AI, aortic insufficiency; MI, mitral insufficiency; eGFR, estimated glomerular filtration rate. Annular size distribution differed between treatment groups. Overall, 109 patients (25.7%) had extra-large annuli and 304 patients (71.7%) had large annuli. Extra-large annuli were more common among SAVR patients (47/61, 77%) compared with TAVR patients (62/353, 17.6%), whereas large annuli predominated in the TAVR cohort (291/353, 82.4% vs 13/61, 21.3% in SAVR). Consistent with this distribution, patients undergoing SAVR had larger annular dimensions on computed tomography, including greater annular perimeter (median 101 mm vs 89.8 mm, P < .001) and annular area (752 mm2 vs 618 mm2, P < .001). Among TAVR patients, transfemoral access was used in 335 cases (94.9%). Balloon-expandable valves were used in 250 patients (70.8%). In the SAVR cohort, the majority received bioprosthetic valves (56/61, 91.8%). Only 5 SAVR patients (8.2%) received mechanical valves. The median implanted aortic valve size differed between groups, with TAVR patients receiving larger valves (median 29 mm vs 25 mm, P < .001) compared with SAVR. Valve oversizing occurred in 221 TAVR cases (62.6%), with a median oversizing of approximately 10.3%. Procedural complications were rare in the TAVR cohort, with no cases of annular rupture, device embolization, or migration, and no conversions to surgery (0% for all). Postprocedural hemodynamics before hospital discharge differed between groups. Mean aortic valve gradients were lower in SAVR patients compared with TAVR patients (median 8.0 mm Hg [6.0-10.0] versus 10.0 mm Hg [7.0-13.0], P < .001). Paravalvular leak (PVL) rates differed between groups: Trace or trivial PVL occurred in 101 of 353 TAVR patients (28.6%) versus 4 of 61 SAVR patients (6.6%), mild PVL was observed in 55 TAVR patients (15.6%) and none of the SAVR patients (0%), and moderate regurgitation was seen in 2 TAVR patients (0.6%) and 0 SAVR patients (0%) (P < .001). New permanent pacemaker implantation occurred in 21 TAVR patients (5.9%) and none of the SAVR patients (P < .001). There were no periprocedural strokes in either the TAVR or SAVR cohorts (0% in both groups). Early 30-day mortality was uncommon, occurring in 3 TAVR patients (0.9%) and none of the SAVR patients (0%, P = .470). The observed-to-expected ratio for perioperative mortality in the entire population was 0.21 (95% CI, 0.07-0.65), with observed-to-expected ratio of 0.22 (95% CI, 0.07-0.68) in TAVR patients and 0.0 (95% CI 0.0-0.0) in SAVR patients, respectively. Patients were followed for up to 4 years. Mean follow-up duration was modestly longer in the SAVR cohort compared with the TAVR cohort (2.23 ± 1.17 vs 1.76 ± 1.23 years, P = .006). Unadjusted Kaplan–Meier analyses demonstrated significantly higher mortality after TAVR compared with SAVR (log-rank P < .001, Figure 2). When compared with an age- and sex-matched US reference population, SAVR patients experienced a numerically, but not statistically significant, lower mortality (standardized mortality ratio [SMR], 0.29, 95% CI, 0.04-2.03), whereas TAVR patients had higher-than-expected mortality (SMR, 1.69, 95% CI, 1.35-2.11). In univariable Cox proportional hazards analyses, treatment with TAVR was associated with significantly higher mortality compared with SAVR (hazard ratio [HR], 17.13, 95% CI, 2.38-123.14, P = .005). An increased risk of death was also observed with older age, atrial fibrillation, congestive heart failure, cerebrovascular disease, prior myocardial infarction, and higher preprocedural creatinine, whereas higher left ventricular ejection fraction and higher preprocedural mean gradient were associated with lower mortality (Table 2). In multivariable analysis using Firth's penalized Cox regression to account for the limited number of deaths in the SAVR cohort, treatment strategy was no longer independently associated with mortality. Instead, survival was largely influenced by baseline comorbidities (Table 2). During follow-up, stroke was uncommon, with only 1 documented event during follow-up and no significant difference between groups (0.3% vs 0%, P = .677). Two patients (0.6%) initially treated with TAVR subsequently required aortic valve reintervention: One underwent repeat TAVR at 1.8 months, and the other underwent SAVR at 42 months. Overall survival after aortic valve replacement. Overall survival in patients with large and extra-large aortic annuli. Shaded bands represent 95% CIs, and the dotted line indicates expected survival for an age- and sex-matched US population based on the 2021 US Life Tables.5SAVR, Surgical aortic valve replacement; TAVR, transcatheter aortic valve replacement. Table 2Univariable and multivariable models for all-cause mortalityVariableCrude HR (95% CI)P valueAdjusted HR (95% CI)P valueTAVR vs SAVR17.13 (2.38-123.14).0054.54 (0.82-25.26).084Age, per 1 y1.07 (1.03-1.10)<.0011.03 (1.00-1.07).082Female sex0.80 (0.25-2.52).7951.27 (0.43-3.76).668Atrial fibrillation2.27 (1.45-3.57)<.0011.61 (1.01-2.58).047Congestive heart failure3.47 (1.59-7.54).0021.70 (0.78-3.71).182Left ventricular ejection fraction, per 1%0.98 (0.96-1.00).0131.00 (0.98-1.02).804Preprocedural Doppler mean gradient, per 1 mm Hg0.98 (0.96-1.00).0221.00 (0.98-1.02).751Preprocedural creatinine, per 1 mg/dL1.37 (1.19-1.58)<.0011.37 (1.17-1.60)<.001BMI, kg/m20.96 (0.93-1.01).089--BSA, m20.46 (0.16-1.30).141--Diabetes1.16 (0.73-1.82).531--COPD0.80 (0.46-1.38).413--Cerebrovascular disease1.71 (1.07-2.72).025--Hypertension1.87 (0.68-5.10).225--History of liver disease1.37 (0.50-3.76).536--Prior myocardial infarction1.63 (1.03-2.57).038--Prior revascularization0.79 (0.43-1.47).462--Mild or greater mitral regurgitation1.32 (0.80-2.16).274--Moderate or greater aortic regurgitation0.92 (0.53-1.57).746--Indexed aortic valve area, cm2/m20.82 (0.09-7.60).861--Aortic valve area, cm20.64 (0.21-2.00).443--Implanted valve size, mm1.08 (1.00-1.17).045--TAVR, Transcatheter aortic valve replacement; SAVR, surgical aortic valve replacement; BMI, body mass index; BSA, body surface area; COPD, chronic obstructive pulmonary disease; CVD, cerebrovascular disease. Univariable and multivariable models for all-cause mortality TAVR, Transcatheter aortic valve replacement; SAVR, surgical aortic valve replacement; BMI, body mass index; BSA, body surface area; COPD, chronic obstructive pulmonary disease; CVD, cerebrovascular disease. Competing-risk analysis for heart failure rehospitalization, accounting for death as a competing event, demonstrated significantly higher cumulative incidence of heart failure rehospitalization after TAVR (Gray test P < .0001), as shown in Figure 2. In univariable analyses for heart failure hospitalization, treatment with TAVR was associated with a significantly higher risk compared with SAVR (HR, 7.05, 95% CI, 2.55-19.47, P < .001) (Figure 3). Additional factors associated with increased risk included diabetes, atrial fibrillation, congestive heart failure, and larger implanted valve size (Table 3). In multivariable analysis adjusting for clinically relevant covariates, TAVR remained independently associated with a higher risk of heart failure hospitalization (adjusted HR, 5.09, 95% CI, 1.88-13.80, P = .001). Among baseline comorbidities, congestive heart failure was the only variable independently associated with subsequent heart failure hospitalization (adjusted HR, 1.91, 95% CI, 1.09-3.38, P = .025). Other clinical and echocardiographic variables, including age, diabetes, atrial fibrillation, renal function, left ventricular ejection fraction, and preprocedural Doppler mean gradient, were not independently associated with heart failure hospitalization in the fully adjusted model. Heart failure readmission after aortic valve replacement. Cumulative incidence of heart failure readmission. Bands represent 95% CIs. SAVR, Surgical aortic valve replacement; TAVR, transcatheter aortic valve replacement. Table 3Univariable and multivariable models for heart failure rehospitalizationVariableCrude HR (95% CI)P valueAdjusted HR (95% CI)P valueTAVR vs SAVR7.05 (2.55-19.47)<.0015.09 (1.88-13.80).001Age, per 1 y1.02 (1.00-1.04).0690.99 (0.96-1.01).228Female sex0.96 (0.45-2.08).9181.20 (0.54-2.64).659Diabetes1.42 (1.01-2.00).0431.28 (0.90-1.82).163Cerebrovascular disease1.13 (0.77-1.65).5291.02 (0.69-1.50).929Prior myocardial infarction1.40 (0.98-2.01).0681.13 (0.77-1.68).533Atrial fibrillation1.63 (1.16-2.29).0051.35 (0.94-1.92).101Congestive heart failure3.10 (1.74-5.50)<.0011.91 (1.09-3.38).025Preprocedural creatinine, per 1 mg/dL1.11 (0.96-1.29).1581.05 (0.90-1.22).529Left ventricular ejection fraction, per 1%0.99 (0.98-1.00).1101.01 (0.99-1.02).456Preprocedural Doppler gradient, per 1 mm Hg0.98 (0.97-1.00).0240.99 (0.98-1.01).247BMI, kg/m21.00 (0.98-1.03).842--BSA, m20.62 (0.29-1.31).209--COPD0.91 (0.60-1.37).640--Hypertension1.59 (0.78-3.20).200--Prior revascularization1.22 (0.67-2.19).518--Mild or greater mitral regurgitation1.28 (0.87-1.87).206--Moderate or greater aortic regurgitation1.18 (0.79-1.75).423--Indexed aortic valve area, cm2/m25.12 (0.95-27.48).057--Aortic valve area, cm21.94 (0.81-4.66).136--Implanted valve size, mm1.09 (1.03-1.15).002--TAVR, Transcatheter aortic valve replacement; SAVR, surgical aortic valve replacement; BMI, body mass index; BSA, body surface area; COPD, chronic obstructive pulmonary disease; CVD, cerebrovascular disease. Univariable and multivariable models for heart failure rehospitalization TAVR, Transcatheter aortic valve replacement; SAVR, surgical aortic valve replacement; BMI, body mass index; BSA, body surface area; COPD, chronic obstructive pulmonary disease; CVD, cerebrovascular disease. Discussion In this contemporary cohort of patients with large and extra-large aortic annuli, we observed meaningful differences in outcomes between TAVR and SAVR. Although unadjusted analyses demonstrated higher mortality and heart failure rehospitalization after TAVR, these associations attenuated after multivariable adjustment, with survival primarily influenced by comorbid conditions. These findings highlight the complexity of outcome interpretation in patients with large annular anatomy, where treatment selection is strongly confounded by age, comorbidity burden, and perceived surgical risk. Our findings align with and expand on prior observations regarding the interaction among aortic annulus size, valve platform, and clinical outcomes. In subanalyses of the PARTNER and CoreValve High-Risk trials, TAVR demonstrated superior hemodynamics over SAVR in small (<18 mm) and medium (18-20 mm) aortic annuli, with larger effective orifice areas and lower rates of moderate-to-severe patient-prosthesis mismatch (small: 39.4% vs 62.5%, medium: 48.6% vs 69%, P ≤ .02), whereas in large annuli (≥20 mm) these advantages were attenuated (52.9% vs 56.3%, P = .69).6,7 Moreover, 1-year mortality was highest among TAVR patients with large annuli (24.8%) versus medium (18.7%) and small annuli (18.3%), with large annulus remaining an independent predictor of mortality (33% and 37% increased risk of mortality in large compared with medium and small annuli, respectively).7 Safety outcomes in TAVR are further supported by real-world evidence from the TAVR-LARGE Registry, which included patients with large and extra-large aortic annuli (area ≥575 mm2, perimeter ≥85 mm) treated with the 29-mm Sapien-3 or 34-mm Evolut-R.2 The registry demonstrated a low incidence of significant PVL (3%) and no cases of central aortic regurgitation. Thirty-day and 1-year mortality rates were 2.4% and 9.2%, respectively, with no significant differences between valve types.2 Furthermore, a multicenter German cohort of 144 patients examined the off-label use of third-generation TAVR prostheses in extra-large annuli (area ≥683 mm2, perimeter ≥94.2 mm).8 Device success remained high (82.3% vs 84.5% in propensity-matched controls with on-label annuli), and 30-day mortality was comparable between groups (2.5% vs 3.4%).8 Moderate PVL occurred more frequently in patients with extra-large annuli (8.9% vs 2.2%; P = .02), and postprocedural pacemaker implantation was lower (7% vs 15%; P = .05).8 A series of 74 patients with extremely large annuli (mean area 721 ± 38 mm2) undergoing TAVR with a 29-mm Sapien-3 demonstrated 100% procedural success.9 TAVR in this cohort was also associated with acceptable rates of moderate PVL (6.9%) and pacemaker implantation, although larger left ventricular outflow tracts and more eccentric annuli were linked to higher PVL.9 Long-term data from the PARTNER 2 and PARTNER 3 trials suggest that annular size (≥430 mm2) does not significantly impact outcomes or valve durability after TAVR with a balloon-expandable valve, although these studies did not include patients with extreme annular dimensions (mean 518.3 ± 58.0 mm2).10 In contrast, data in SAVR cohorts are more limited, a SURTAVI subanalysis showed that prosthesis size increased with larger indexed annuli in TAVR (P < .001) but not in SAVR (P = .74), suggesting a tendency toward smaller valve implantation in SAVR.11 Consequently, patient-prosthesis mismatch rates were lower with TAVR across all annular sizes (P < .001) and declined with increasing annulus size in both TAVR (P = .04) and SAVR (P = .03). This is important, because patient-prosthesis mismatch has been shown to be associated with adverse outcomes, including higher long-term mortality and worse postoperative hemodynamics.12,13 To further underscore the impact of aortic annular anatomy, a pooled analysis of 1519 SAVR patients from 3 trials demonstrated that a higher implanted valve-to-native annulus size ratio was independently associated with improved 5-year hemodynamics and survival.14 Recent real-world data from the largest reported experience with the Epic Supra 29 bioprosthesis showed that in patients with large annuli, 30-day mortality was low at 2.8%, with permanent pacemaker implantation in only 4.7%.15 Long-term follow-up of up to 10 years demonstrated excellent hemodynamic performance, with mean gradients of 11.3 mm Hg and cumulative mortality of 19.2% at 70 months.15 To put our findings in perspective, we must address the concept of lifetime management, because this is particularly relevant in patients with large annuli, who are increasingly younger or more anatomically complex. When selecting between TAVR and SAVR in patients where both are technically feasible, consideration must be given not only to the index procedure risk but also to the long-term durability of the implanted prosthesis and the feasibility of future reintervention.16 Bioprosthetic valves, whether transcatheter or surgical (91.8% of patients in our study), are subject to structural valve deterioration over time, and valve-in-valve TAVR has emerged as a preferred reintervention strategy for failed surgical bioprostheses.17 In patients with large annuli, the implanted surgical valve sizes (25 mm in our SAVR cohort) may limit valve-in-valve options due to concerns about residual gradients and patient-prosthesis mismatch within the existing surgical frame.12,13 Conversely, large-frame transcatheter prostheses may offer more favorable valve-in-valve landing zones. For younger patients with large annuli who are suitable for both approaches, a Heart Team discussion that explicitly incorporates lifetime management strategy, including projected valve durability and future reintervention options, is strongly recommended.16 Furthermore, although our data demonstrate that both approaches yield outcomes consistent with the underlying patient risk profile, with SAVR achieving population-level survival in appropriately selected patients and TAVR reflecting the higher comorbidity burden, the persistence of significantly higher heart failure hospitalization after TAVR even after multivariable adjustment argues for careful patient selection. Last, these findings complement and extend the TAVR-LARGE Registry,2 which enrolled only TAVR patients using earlier-generation devices (2014-2019) without a surgical comparator arm or without population-referenced survival benchmarking. Furthermore, our cohort spans a newer era 2021-2025 and includes both treatment strategies, enabling contextual survival comparison and competing-risk analyses. Prospective studies specifically enrolling patients with large and extra-large aortic annuli and comparing TAVR and SAVR outcomes are needed to confirm our findings and guide evidence-based decision-making in this anatomically distinct population. Limitations This study has several limitations. As a single-center, retrospective analysis, its findings may have limited generalizability. Definitions of large and extra-large annuli varied across the studies discussed, including our own, which may limit applicability to other settings. Moreover, patients with bicuspid aortic valve anatomy were excluded from our study to reduce anatomic and procedural heterogeneity; therefore, our findings may not be generalizable to this subgroup. We also did not perform subgroup analysis between large and extra-large annuli due to limited sample size and disproportionate distribution. The 2 treatment groups differ substantially in age, comorbidity burden, and anatomic profile, reflecting strong selection bias driven by clinical and anatomic factors, most notably the size constraints of currently available transcatheter platforms, which may have precluded TAVR for many patients in the SAVR cohort. Furthermore, although multivariable adjustment was performed, residual confounding cannot be excluded. As such, direct comparisons of TAVR and SAVR outcomes should be interpreted with caution. In extension, the SMR comparisons against the general US population should be interpreted as contextual, because observed survival differences may reflect the expected trajectory of a high-risk elderly population independent of the procedure performed. Then, we do not have detailed follow-up echocardiographic data, which limits the mechanistic interpretation of the observed heart failure outcome differences. The study also spans multiple generations of valve technologies, introducing potential era effects that may have influenced outcomes. Conclusions In patients with large and extra-large aortic annuli, SAVR was associated with survival comparable to the age- and sex-matched general population, whereas TAVR patients, likely reflecting higher baseline risk, experienced increased mortality. Conflict of Interest Statement I.S. receives institutional funding and consults for Abbott, Artivion, Atricure, Edwards Lifesciences, Medtronic, and Terumo Aortic. None related to the manuscript. All other authors reported no conflicts of interest. The Journal policy requires editors and reviewers to disclose conflicts of interest and to decline handling or reviewing manuscripts for which they may have a conflict of interest. The editors and reviewers of this article have no conflicts of interest. Footnotes The study protocol was approved by the Institutional Review Board, and the requirement for informed consent was waived due to the retrospective nature of the study (STUDY18120143, approved 4/17/2019). Dr Ibrahim Sultan, MD, is an Associate Editor. The peer review process for this paper was handled by Dr Leora B. Balsam, MD. References