This was a multinational, multicenter, observational nonprespecified retrospective analysis of prospectively collected data. A total of 4643 patients from 11 centers from Canada, France, Italy, and Spain who underwent TAVR between 2014 and 2022 were screened for inclusion in the study. Patients who died within the first 30 days after TAVR, those whose length of stay after the procedure was≥30 days, and those undergoing a transapical or transaortic approach were excluded from the analysis (435 patients, 9%). In 746 (16%) patients, NYHA class at 1 month of follow-up had not been recorded and they were therefore excluded from the analysis. Discharged patients rehospitalized for heart failure (HF) within the first month after TAVR were classified as NYHA class IV.

Each local heart team assessed the indications for TAVR, device type, and procedural approach based on an extensive clinical and anatomical preoperative assessment. The transfemoral approach was used by default, and alternative transarterial (transcarotid, transsubclavian/axillary) access was reserved for patients with unfavorable iliofemoral anatomy. This study was conducted according to the ethics committee of each participating center, and all patients provided signed informed consent for the procedures.

Baseline, procedural, and follow-up data were prospectively collected in each dedicated local TAVR database. The patients’ pre-TAVR work-up and subsequent follow-up were performed according to local protocols. Information on NYHA class at 30 days was required for inclusion in this study. The patients’ vital status was updated after each medical contact with recording of the date of the last contact for each patient.

Continuous variables are expressed as mean±standard deviation or median [Q1-Q3] according to the normality of data distribution assessed with the Shapiro-Wilk test. Categorical variables are expressed as number (%). The chi-square test or Fisher exact test were used to compare categorical variables. The Student t test or the Mann-Whitney U test were used to compare continuous variables. Factors associated with 1-month NYHA class III-IV were identified using logistic regression analysis. The variables to be analyzed were chosen based on previous causal knowledge and their association was assessed with univariable logistic regression analysis. Variables showing an association (P<.10) in the univariable analysis were included in the multivariable model.

For 1-year survival analysis, the Kaplan-Meier method was used to obtain event curves. The difference between the probability of event occurrence was assessed with the log-rank test. A Cox proportional hazards regression analysis was performed to evaluate the impact of 1-month NYHA class III-IV on mortality and HF hospitalization at 1-year follow-up. Models were adjusted for baseline confounders based on prior causal knowledge. A 2-sided P<.05 was considered significant for all statistical tests. All data were analyzed using the statistical package STATA version 15.0 (StataCorp LP, College Station, United States).

At 1 month of follow-up, 208 patients (6%) were classified as NYHA class III-IV. Of them, 160 (77%) showed advanced NYHA class (III or IV) before TAVR (4 of them required hospitalization due to HF within the first month after TAVR); 31 (15%) showed a deterioration in their clinical status without HF hospitalization in the first month; and 17 (8%) showed a decline with HF hospitalization within the first month of follow-up.

Patients with worse functional class after TAVR were similar in age and sex distribution to those in NYHA class I-II. However, patients in NYHA class III-IV showed a higher burden of comorbidities such as hypertension (93% vs 85%, P<.01), atrial fibrillation (43% vs 33%, P<.01), and COPD (38% vs 25%, P<.01). Consistently, these patients had a worse risk profile according to STS score (4.2 [2.8 -6.7]% vs 3.8 [2.5-5.8]%; P<.01) and EuroSCORE II (5.1 [2.6-8.7]% vs 4.0 [2.2-7.1]%; P<.01).

NYHA class III-IV patients had lower left ventricular ejection fraction (LVEF) (53.3±14.4% vs 55.9±12.9%; P<.01), with a higher proportion of patients with reduced LVEF (LVEF ≤ 40%: 23% vs 15%; P<.01), and lower peak and mean transaortic gradients (peak: 68.5±24.1 mmHg vs 76.7±24.4 mmHg; P<.01; mean: 41.7±15.6 mmHg vs 46.7±15.9mmHg; P<.01). Consequently, the rate of low-gradient AS was higher among patients with NYHA class III-IV at 30-day follow-up (42% vs 30%; P<.01).

Table 2 summarizes the procedural-related characteristics, in-hospital complications, and postprocedural echocardiogram results.

The transfemoral approach was used in 90% of patients, with a higher rate of a nontransfemoral approach in patients with 1-month NYHA class III-IV (19% vs 9%; P<.01). NYHA class III-IV patients less often had valve predilatation during the procedure (23% vs 36%; P<.01). The groups did not differ in any other procedural-related characteristic or the type and size of the implanted valve. The rate of procedural and in-hospital complications was similar between the 2 groups.

In the postprocedural echocardiogram, the mean indexed effective orifice area and mean aortic gradient were 1.0±0.3 cm2/m2 and 11.1±6.1 mmHg, respectively, with no differences between groups. Conversely, patients with 1-month NYHA class III-IV had a lower LVEF and a higher rate of LVEF ≤40% (19% vs 13%; P<.01); a higher rate of severe mitral regurgitation (21% vs 8%; P<.01); higher pulmonary artery systolic pressure (41.2±13.7 mmHg vs 37.1±12.2 mmHg; P≤.01); and a higher rate of severe pulmonary hypertension (15% vs 8%; P<.01) (figure 2, table 2).

Figure 2.

Comparison of echocardiographic parameters after procedure between patients showing NYHA class I-II vs NYHA class III-IV at 1 month of follow -up. A: indexed effective orifice area. B: transaortic mean gradient. C: left ventricular ejection fraction. D: pulmonary artery systolic pressure. E: severe mitral regurgitation. F: moderate or severe aortic regurgitation. AR, aortic regurgitation; iEOA, indexed effective orifice area; LVEF, left ventricle ejection fraction; MR, mitral regurgitation; PA, pulmonary artery.

(0.41MB).

The results of the univariable logistic regression analysis are described in the table 2 of the supplementary data. Figure 3 shows the results of the multivariable analysis performed to determine the predictors of 1-month impaired functional status. In the multivariable analysis, NYHA III-IV before TAVR was associated with a higher risk of having advanced NYHA class at 1-month follow-up (OR, 1.76; 95%CI, 1.08-2.89; P=.02). A history of COPD (OR, 1.80; 95%CI, 1.13-2.83; P=.01) and the presence of severe mitral regurgitation in the echo performed after the procedure (OR, 2.00; 95%CI, 1.21-3.31; P<.01) were also independent predictors of impaired NYHA class at 1-month follow-up.

Figure 3.

Multivariable logistic regression analysis representing the association of different covariates with the presence of NYHA class III-IV 1 month after TAVR. 95%CI, 95% confidence interval; LVEF, left ventricle ejection fraction; MR, mitral regurgitation; NYHA, New York Heart Association; PAPS, pulmonary artery pressure, systolic; PH, pulmonary hypertension.

(0.28MB).

A subanalysis of the factors associated with NYHA class III-IV at the 30-day follow-up was performed including only patients without previous COPD. In this subpopulation, postprocedural severe mitral regurgitation (OR, 2.24; 95%CI, 1.22-4.10); P<.01) and new-onset atrial fibrillation (OR, 3.08; 95%CI; 1.38-6.87; P<.01) independently predicted the presence of NYHA class III-IV at 30 days, while baseline NYHA class III-IV was no longer a predictor of 30-day NYHA class III-IV (table 3 of the supplementary data).

The median follow-up of the population was 12 [4-24] months. In 188 patients (5% of the final population), the follow-up was missing after the 1-month visit. During the first year after TAVR, 178 patients (5%) died, and 112 (3%) required rehospitalization due to an HF episode (table 3).

The presence of advanced NYHA class at 1-month follow-up was associated with a higher risk of all-cause mortality (HR, 3.68; 95%CI, 2.39-5.70; P<.01), cardiovascular mortality (HR, 3.22; 95%CI, 1.67-6.23; P<.01), and HF-related hospitalization (HR, 6.00; 95%CI, 3.76-9.60; P<.01).

In addition, worsening functional class (from baseline to 1-month post-TAVR) was associated with a higher risk of all-cause mortality (HR, 2.87; 95%CI, 1.50-5.53; P<.01), cardiovascular mortality (HR, 4.48; 95%CI, 2.02-9.96; P<.01), and HF hospitalization (HR, 6.38; 95%CI, 3.48-11.7; P<.01). Figure 4 shows a comparison of survival curves between patients with 1-month NYHA class I-II and III-IV; and between patients with worsening of their functional class and those who remained equal or improved.

Figure 4.

Kaplan-Meier estimates of 1-year all-cause mortality (A) cardiovascular mortality (B) and heart failure hospitalization (C) according to the presence of NYHA class III-IV at 1 month of follow-up. Kaplan-Meier estimates of 1-year all-cause mortality (D) cardiovascular mortality (E) and heart failure hospitalization (F) according to the change in NYHA functional class from baseline to 30 days of follow-up. TAVR, transcatheter aortic valve replacement.

(0.77MB).

Although TAVR has been consistently shown to improve NYHA class, a lack of improvement 1 month after TAVR has been reported in up to 16% of patients.7 In addition, persistent NYHA class III-IV at 1-year follow-up or HF hospitalization within the first year after TAVR was reported in about 7% of patients included in a multicenter Japanese registry.6 In our population, 20% of patients showed no improvement in NYHA class, and 6% were classified as NYHA class III or IV 1 month after the procedure.

Although TAVR has been demonstrated to be a cost-effective therapy in intermediate- and high-risk populations,9 these data raise concerns about the futility of TAVR in some patients, especially those with a higher comorbidity burden.8 Indeed, previous studies showed that up to one-third of patients have poor quality of life or die within the first year after TAVR. In addition, functional decline or lack of improvement is common in frail patients undergoing TAVR.10

COPD was an independent predictor of NYHA class III-IV at 1 month post-TAVR. COPD patients constitute a high-risk population, with an increased risk of respiratory-related complications after cardiac surgery.11 COPD patients undergoing TAVR have fewer respiratory complications and lower in-hospital mortality than those undergoing surgery.11 However, COPD is related to higher 1-year mortality and is more frequent in patients with no improvement of NYHA class in the first month after TAVR.6,7,12 Evaluating dyspnea in patients with COPD and severe AS may be challenging. The severity of intrinsic pulmonary disease may hamper the improvement of shortness of breath in COPD patients. Therefore, a correct assessment of pulmonary disease and its optimal treatment is desirable in these patients to improve patient-perceived outcomes. Regardless, patients with COPD should be informed about the possibility of a mild or no improvement in dyspnea symptoms after the procedure. A detailed explanation of the risk/benefit ratio of the intervention should be given, focusing on the potential survival improvement resulting from the procedure, even in the absence of significant symptomatic relief.

Interestingly, when we analyzed patients without COPD, the baseline NYHA class did not show a significant association with the presence of NYHA class III-IV at the 30-day follow-up. This finding highlights the impact of pulmonary disease on the symptoms of patients with valvular heart disease and the need for a systematic approach to patients’ comorbidities in order to improve TAVR outcomes.

Patients with impaired NYHA class 1 month after TAVR showed several signs of more advanced cardiac disease before the procedure, including higher rates of NYHA class III-IV and atrial fibrillation, and lower LVEF. In addition, these patients more frequently had low-flow low-gradient severe AS.

Patients undergoing TAVR in NYHA class III-IV have a worse clinical course, with a higher risk of mortality and HF hospitalization during the first year of follow-up.13,14 In addition, nonreversible pulmonary hypertension is related to poorer outcomes.15

The onset of AS-related symptoms is often accompanied by other cardiac injury,15 reducing the effectiveness of the aortic valve replacement in symptom relief and survival improvement.15–17

In the AVATAR trial, early surgery in asymptomatic AS patients reduced the composite outcome of all-cause death, acute myocardial infarction, stroke, or unplanned HF hospitalization.18 All this evidence would favor an early intervention in patients with severe AS before nonreversible cardiac injury appears, hampering the patient's symptom relief and survival. However, this enhanced proactive strategy would unavoidably lead to complications among asymptomatic patients and a substantial increase in health care expenditure. Thus, it is important to identify early markers of cardiac injury to improve the selection of asymptomatic patients who might benefit from early intervention. In addition, the ongoing EARLY-TAVR trial (NCT03042104) evaluating the effectiveness of TAVR (vs medical treatment) in reducing the composite endpoint of all-cause death, stroke, and unplanned cardiovascular hospitalization in asymptomatic AS patients will provide a definite response regarding the timing of the intervention. Meanwhile, once severe AS is diagnosed, patients should be closely followed up and undergo early intervention when symptoms appear, avoiding a delayed intervention at later stages of the disease.

Baseline moderate-to-severe MR is present in about one-fifth of patients undergoing TAVR and is associated with increased mortality. TAVR reduces the severity of MR in up to 60% of patients. However, residual MR after TAVR is associated with increased 1-year mortality.19

Data from the TRAMI registry showed that almost 9% of patients undergoing mitral transcatheter edge-to-edge repair (TEER) had a previous aortic valve replacement, either surgical or TAVR. Notably, patients with previous aortic valve replacement were older and had a higher rate of COPD, and 87% of them were in NYHA class III-IV at the time of TEER. Patients undergoing TEER after TAVR had an estimated 1-year survival below 50%.20 This study did not report the time from TAVR to TEER, so late interventions cannot be excluded. Of all the patients registered for TAVR, TEER, or transcutaneous mitral valve replacement at the United States Readmissions Database from 2014 to 2018, 110 patients underwent simultaneous TAVR and TEER. Simultaneous TAVR and TEER was associated with poorer outcomes compared with TAVR alone, reflecting the potential risk of a simultaneous procedure in an elderly population with a high comorbidity burden.21 In a multicenter study of TAVR patients with postprocedural moderate-to-severe MR and persistence of NYHA class III-IV in 90%, TEER at a median time of 164 days after TAVR was associated with significant improvement in MR and NYHA class. In addition, a trend toward improved survival was observed.22 Our data support the concept of the high risk associated with residual MR leading to impaired functional status early after TAVR. In addition, the increased mortality risk related to significant residual MR seems to start very early after the procedure, highlighting the major clinical relevance of close follow-up, medical treatment optimization, and earlier mitral intervention in such cases. The ongoing MITAVI (Treatment of Concomitant Mitral Regurgitation by Mitral Valve Clipping in Patients With Successful Transcatheter Aortic Valve Implantation) randomized trial (NCT04009434), assessing whether TAVR patients with moderate-to-severe MR can benefit from additional mitral valve clipping treatment, will provide important insights regarding the timing of intervention.

The adverse prognostic impact of advanced NYHA class at baseline or 1 year after TAVR has been previously described.6,12,13

In our population, NYHA class III-IV or an episode of HF hospitalization 1 month after TAVR was related to an approximately 3-fold higher mortality and up to 6-fold increased risk for HF hospitalization within the year following TAVR. Furthermore, most events occurred during the first 6 months of follow-up. This means that delaying interventions in these patients may significantly affect their prognosis.

We identified several factors at hospital discharge predicting advanced NYHA class at 1 month of follow-up. Besides an earlier intervention of patients requiring TAVR, an earlier evaluation of patients after TAVR is advisable to identify those with suboptimal symptoms improvement and detect possible causes for this outcome. Treatment should be optimized to improve comorbidity-related symptoms and prognosis (diuretic regimen, HF treatment, COPD treatment). The involvement of HF specialists should also be strongly considered in such cases. In addition, in patients with multivalvular disease, a planned staged procedure might help to improve symptom relief and prognosis. Further studies are necessary to address this topic.

This study has several limitations. This was a nonprespecified retrospective analysis of prospectively collected data. Consequently, some information relevant to the study topic may have not been consistently collected. Information on NYHA class at 30 days was missing in 18% of patients screened for inclusion, partially due to the loss of follow-up visits during the coronavirus pandemic from 2020 to 2021. The NYHA functional classification is semiquantitative, and its subjective assessment might be subject to interobserver variability and influenced by patients’ comorbidities and frailty. NYHA class was graded according to investigator criteria based on clinical interview but not based on standardized questionnaires. Due to its observational nature, the presence of unmeasured confounders that can influence study conclusions cannot be excluded. Detailed information on some echocardiographic signs of cardiac injury was not available. In addition, information on the completeness of the coronary revascularization before TAVR was available for less than one-third of the patients. Finally, specific information on COPD severity and the need for home oxygen therapy was not available.