1005 STROKE VOLUME INDEX AND TRANSVALVULAR FLOW RATE TRAJECTORIES IN SEVERE AORTIC STENOSIS TREATED WITH TAVR

Aims

The prognostic impact of flow trajectories according to stroke volume index (SVi) and transvalvular flow rate (FR) in patients with severe aortic stenosis (AS) undergoing transcatheter aortic valve replacement (TAVR) remains poorly assessed. We evaluated and compared SVi and FR prior and after TAVR for severe AS.

Methods and results

Patients were categorized according to SVi (<35 ml/m2) and FR (<200 ml/sec). The association of pre- and post-TAVR SVi and FR with all-cause mortality up to 3 years was assessed with multivariable Cox regression models. Among 980 patients with pre-TAVR flow assessment, SVi was reduced in 41.3% and FR in 48.1%. Baseline flow status was not an independent mortality predictor (SVi: HR 1.39, 95%CI 0.81-2.40, FR: HR 0.86, 95%CI 0.51-1.46). Among 731 patients undergoing early (5 days, IQR 2-29) post-TAVR flow assessment, SVi recovered in 40.1% and FR in 49.0% patients with baseline low-flow. Reduced FR following TAVR was an independent predictor of mortality (HR 2.08, 95%CI 1.07-4.04) while SVi was not (HR 0.68, 95%CI 0.34-1.36). Three-year estimated mortality in patients with recovered FR was lower as compared to patients with reduced FR (13.3% vs 37.7% vs, p=0.003) and similar to patients with normal baseline FR (p=0.317).

Conclusions

Baseline flow status was not an independent predictor of mid-term mortality among all-comers with severe AS undergoing TAVR. Flow recovery early after TAVR was frequent. Post-TAVR FR, but not SVi, was independently associated with mid-term all-cause mortality. By impacting flow status, AV replacement modifies the association of flow status with outcomes.

951 PROGNOSTIC IMPACT OF DISPROPORTIONATE FUNCTIONAL MITRAL INSUFFICIENCY IN PATIENTS UNDERGOING PERCUTANEOUS “EDGE-TO-EDGE“ VALVE REPAIR

Aims

The ideal phenotype of patients with functional mitral regurgitation (FMR) who benefit most of transcatheter edge-to-edge repair (TEER) is still unclear. Some studies have suggested that patients with disproportionate FMR may have a better outcome. The purpose of this study is to evaluate the prognostic value of proportionate versus disproportionate FMR, in patients undergoing TEER with MitraClip system.

Methods and Results

The multicenter observational MITRA-CTV registry includes 200 patients with moderate-severe (3+) to severe (4+) FMR undergoing MitraClip, between March 2013 and June 2021, at three European institutions such as Magna Graecia University of Catanzaro, University of Turin (Italy) and University of Vigo (Spain). Patients were defined as having proportionate or disproportionate FMR if their EROA/LVEDV (left ventricular end diastolic volume) ratio was ≤ or > from the median value (0.15), respectively. The primary endpoint was the composite of death from all causes and rehospitalizations for HF, at 1-year follow-up. The secondary endpoint was composed of the individual components of the primary endpoint and cardiovascular death. Patients with disproportionate FMR had higher EROA (0.47±0.2 cm2 vs 0.27±0.1 cm2) and smaller ventricles (LVEDV: 207 ± ml vs 239 ± 83 ml) than those with proportionate FMR. Procedural success was achieved in 95% of patients. Notably, 30-day residual MR was comparable in patients with disproportionate versus proportionate FMR. There were 4 (2%) deaths during hospitalization, and the median hospital stay after the procedure was 9 days (IQR 7- 9 days). At Kaplan-Meier analysis, an EROA/LVEDV ratio both higher and lower than the median value (0.15) was not associated with an increased incidence of the primary endpoint of death and rehospitalization for HF (HR 1.17 CI 95%[0.72;1.90],p=0.50). The only independent predictors of clinical outcomes at 1 year were: the presence of CKD (HR 4.11 95% CI[1.73;9.75], p=0.0014) and a post TMVR hospital stay >10 days (HR 2.53 95% CI[1.24;5.18],p=0.0111).

Conclusion

In our study, there were no significant differences in outcome in patients with proportionate versus proportionate FMR undergoing TEER with Mitraclip system.

471 IMPACT OF COMPLETE REVASCULARIZATION ON DEVELOPMENT OF HEART FAILURE IN PATIENTS WITH ACUTE CORONARY SYNDROME AND MULTIVESSEL DISEASE

Background

The impact of complete revascularization (CR) on survival and occurrence of heart failure (HF) after ACS is still unsettled. Goal of this study was to evaluate the impact of CR on HF hospitalization and adverse outcomes in patients with ACS and multivessel coronary artery disease undergoing PCI.

Methods

Consecutive ACS patients with multivessel disease from the CORALYS registry were included. First hospitalization for HF or cardiovascular (CV) death was the primary endpoint. Patients were stratified according to CR.

Results

Of 14699 patients in the CORALYS registry, 5054 had multivessel disease. 1473 (29.2%) underwent CR, while 3581 (70.8%) did not. Over 5 years follow-up, CR was associated with a reduced incidence of the primary endpoint (adjusted HR 0.66, 95% CI 0.51-0.85), first HF hospitalization (adj HR 0.67, 95% CI 0.49-0.90), CV death (adj HR 0.56, 95% CI 0.38-0.84) and all-cause death (adj HR 0.74, 95% CI 0.56-0.97). The results were consistent in the matched population and in the IPTW analysis. The benefit of CR was consistent across ACS presentations (HR 0.59, 95% CI 0.39-0.89 for STEMI and HR 0.71, 95% CI 0.50-0.99 for NSTE-ACS) and in patients with LVEF>40% (HR 0.52; 95% CI 0.37-0.72), while no significant benefit was observed in patients with LVEF≤40% (HR 0.77; 95% CI 0.37-1.10, p for interaction 0.04).

Conclusions

In patients with ACS and multivessel disease, CR reduced the risk of first hospitalization for HF and CV death, as well as first HF hospitalization, CV and overall death. When feasible, CR should be performed in all patients with ACS to reduce the incidence of HF and death. Future studies are needed to assess the evidence of CR in patients with depressed LVEF.

146 AORTIC ACCELERATION TIME/EJECTION TIME RATIO AND BI-VENTRICULAR PERFORMANCE IN SEVERE AORTIC STENOSIS

Introduction

Echocardiographic evaluation of severe aortic stenosis (SAS) is is important to guide the therapeutic approach but often challenging. Recent studies have demonstrated that the ratio of acceleration time/ejection time (AT/ET) is a simple and reproducible echocardiographic parameter that integrates aortic stenosis severity evaluation and adds information on patient's prognosis.

Aim

The aim of the study is to investigate the role of the ratio of acceleration time (AT) and ejection time (ET) and its major determinants in severe aortic stenosis .

Methods

Consecutive echocardiograms of patients with severe AS referred to our center were analyzed offline using Tomtec Arena (Tomtec, Untershlei heim, Germany). AT was measured from the start of the CW Doppler aortic wave, to the peak of the aortic jet. ET was calculated from the same starting point, to the end of the CW Doppler aortic wave.

Results

A total of 135 patients with severe aortic stenosis formed the study cohort: patients with AT/ET below the median value of 0.35 (vs. higher) presented lower LVEDV (60 vs. 71 ml/mq; p 0.014), left ventricle mass index (116 vs 130 g/m2; p 0.035) and higher LVEF (58 vs 50%; p 0.001), GLS (- 14 vs - 12%; p 0.025), FAC (46 vs 41%; p 0.01), SBP (141 vs 131 mmHg; p 0.003).

At multivariable analysis the major AT/ET determinants were systolic arterial pressure and bi-ventricular performance parameters. The following nested regression were created: the first inclusive of systolic arterial pressure (PAS), fractional area change (FAC), left ventricular mass indexed (LVMI), global longitudinal strain (GLS) (R2=0.48 p<0.001), the second inclusive of PAS, FAC, LVMI, GLS, AVA (R2=0.57, p<0.001), the third inclusive of PAS, FAC, LVMI, LVEF, AVA (R2=0.64, p<0.001).

Conclusion

Our study demonstrated that AT/ET ratio relates quite well with LV performance in the context of SAS. An high ACT/ET ratio tends to be associated with a poor bi-ventricular performance and LV negative remodeling. It is possible that this simple parameter in the next future could help in staging the disease among SAS patients.

56 ANALYSIS OF A 23,270-PATIENT STUDY: UNSUPERVISED MACHINE LEARNING WITH CLUSTER ANALYSIS IN PATIENTS DISCHARGED AFTER AN ACUTE CORONARY SYNDROME

Background

Characterization and management of patients with acute coronary syndromes (ACS) remain challenging, and it is unclear whether currently available clinical and procedural features can suffice to inform adequate decision making.

Methods

Details on patients discharged after an ACS were obtained by querying an extensive multicenter registry, detailing patient features as well as management details. Clinical outcomes included fatal and non-fatal cardiovascular events at 1-year follow-up. After missing data imputation, two unsupervised machine learning approaches (k-means and Clustering Large Applications [CLARA]) were used to generate separate clusters with different features. Bivariate and multivariable-adjusted analyses were performed to compare different clusters for clinical outcomes.

Findings: 23,270 patients discharged after ACS were included. Two clusters were identified by k-mean algorithm (k1 and k2), and two clusters by CLARA algorithm (C1 and C2). Differences in 2-years outcomes between k1 and k2 and between C1 and C2 were substantial. K2 cluster (N=21,988) in comparison with k1 cluster (N=1,282) had significantly higher occurrence of death (9.5% vs 3.8%, p<0.001), reinfarction (7.2% vs 3.7%, p<0.001), and major bleeding (6.0% vs 3.0%, p<0.001). Similarly, C1 cluster (N=11,268) showed a worse prognosis than C2 cluster (N=12,002): death (4.8% vs 3.5%, p<0.001), reinfarction (4.5% vs 3.4%, p<0.001), and major bleeding (3.6% vs 2.8%, p=0.001). Most associations did not hold at multivariable analysis based on supervised learning techniques, with the exception of major bleeding (odds ratio=1.37 [95% confidence interval 1.02-1.83] for the k1/C1 subcluster vs k1/C2 subcluster, p=0.039.

Conclusions

A machine-learning based clustering approach is effective at face value to inform on the prognosis of patients with ACS managed invasively. These findings can be leveraged to inform decision-making in this setting, but also highlight the potential role of cluster analysis in first-in-man, and preapproval studies of medical devices.

943 THE RIGHT VENTRICULAR-ARTERIAL COUPLING AS AN INDICATOR OF PROGNOSIS IN PATIENTS WITH FUNCTIONAL MITRAL REGURGITATION UNDERGOING TRANSCATHETER “EDGE-TO-EDGE“ VALVE REPAIR

Aims

Patients affected by severe functional Mitral Regurgitation (MR) complaining symptoms despite optimal medical therapy should undergo intervention. When the surgery, the gold standard, is not indicated due to high surgical risk, the transcatheter edge-to-edge repair (TEER) should be considered, if feasible. In patients undergoing TEER, the clinical outcome is not always optimal and strongly correlates to the patient's clinical conditions, so a correct selection of the patients is essential. In this regard, some studies have evaluated the RV-PA coupling as an important predictor of outcome in patients with Heart Failure (HF). In clinical practice, RV-pulmonary artery (PA) coupling could be estimated in a non-invasive way through the relationship between TAPSE (systolic excursion of the annular plane of the tricuspid valve) /PAPs (systolic pressure of the pulmonary artery) ratio that gives information about the state of contractility and adaptability to the load of the RV. In this study, we sought to evaluate how the TAPSE /PAPs ratio at baseline may improve prognostic stratification in patients undergoing TEER with the MitraClip system.

Methods and Results

Data from 236 patients with symptomatic, moderate to severe functional MR, subjected to implantation of MitraClip between March 2012 and June 2021, were obtained from the University's MITRA-CTV, multicenter observational register comprising data from the Magna Graecia University of Catanzaro (Italy), the University of Turin (Italy) and the University of Vigo (Spain). The median follow-up was 686 days (IQR 393-1131 days), with a 1-year follow-up in 224 of 236 (95%) patients. We divided the population into two groups based on the median value of the ratio TAPSE / PAPs ≤ 0.35 and TAPSE / PAPs> 0.35. The primary endpoint of this study includes Re-hospitalization for HF and Death from all causes at one-year follow-up. At Cox regression analysis, Hospital stay> 10 days (HR 1.67, 95% CI [1.03-2.77], p = 0.039) and the TAPSE / PAPs ratio ≤ 0.35 (HR1.58, 95% CI [1, 01-2.48], p = 0.0488) independently increased the risk of the primary endpoint at long-term follow-up. At Kaplan Meier analysis, a TAPSE / PAPs ratio of ≤ 0.35 was related to an increased incidence of the primary endpoint of Rehospitalization for HF and Death (HR 1.54, 95% CI [ 1-2.41], p = 0.0464).

Conclusion

In our study, the right ventricular-arterial coupling, estimated through TAPSE/PAPs Ratio, was identified as a predictor of outcome in patients with severe functional MR undergoing TEER.

891 EARLY EVOLUTION OF CARDIAC DAMAGE STAGING FOLLOWING TRANSCATHETER AORTIC VALVE REPLACEMENT: PREVALENCE, CLINICAL PATTERNS AND PROGNOSTIC SIGNIFICANCE

Background

The early evolution of extravalvular cardiac damage following transcatheter aortic valve replacement (TAVR) as assessed by a previous validated score system remains unstudied. We sought to assess the patterns of early cardiac damage change among patients with severe aortic stenosis (AS) undergoing TAVR and its prognostic implications.

Methods

The RECOVERY-TAVR is a multi-center, international retrospective registry including all consecutive patients undergoing TAVR in thirteen high-volume centers. All the enrolled patients with available paired echocardiography assessment pre- and post TAVR were included in this sub-analysis. Patients were categorized according to the extension of cardiac damage based on a previous published and validated classification (stage 0, no damage; stage 1, left ventricular damage; stage 2, left atrial or mitral valve damage; stage 3, pulmonary vasculature or tricuspid valve damage; and stage 4, right ventricular damage). The primary endpoint was a composite of all-cause mortality or first heart failure hospitalization at 1 year. The association of cardiac damage stage evaluated prior and following TAVR along with the staging evolution was assessed with multivariate Cox regression model (that include hemoglobin, NYHA class and max aortic valve gradient) for the primary outcome.

Results

Of 1331 Patients included in the RECOVERY-TAVR registry with a full echocardiographic pre-TAVR assessment, 892 patients with available paired echocardiography exams were finally included in this analysis (pre-TAVR assessment: median 8 days prior to TAVR; post-TAVR assessment: median 7 days post-TAVR). 63 (7.1%) had stage 0/1, 433 (48.2%) had stage 2, 235 (26.3%) had stage 3 and 161 (18%) had stage 4 myocardial damage. Pre-TAVR myocardial damage staging was associated with the primary outcome (Adj-HR for myocardial stage increase: HR 1.40, 95% CI 1.01–1.93). Following TAVR 274 (30.7%) patients experienced myocardial damage improvement and 161 (18.1%) myocardial damage worsening. Post-TAVR myocardial damage staging was more strongly associated with the primary outcome (HR 1.55, 95%CI 1.14–2.10) as compared to pre-TAVR assessment. Male Sex (p = 0.044) and post-procedural permanent pacemaker implantation (p = 0.044) was associated with myocardial damage worsening, while the use of a balloon-expandable valve (p = 0.011) was associated with myocardial damage improvement. Early myocardial damage worsening (HR 1.89, 95%CI 1.12–3.21), but not early myocardial damage improvement (HR 0.86, 95%CI 0.54–1.37) was associated with the primary outcome.

Conclusion

In patients undergoing TAVR, the extent of extravalvular cardiac damage prior to and early after TAVR has an independent prognostic value while early myocardial damage worsening following TAVR portends a poor prognosis. Whether strategies to improve procedural success and treatments addressing extravalvular myocardial damage early following TAVR may improve outcomes has to be prospectively assessed.

1149 MACHINE-LEARNING BASED PREDICTION OF IN-HOSPITAL DEATH FOR PATIENTS WITH TAKOTSUBO SYNDROME: THE INTERTAK-ML MODEL

Background

Takotsubo syndrome (TTS) is burdened by a not negligible rate of an impaired short-term prognosis. Current existing models, based on classical statistical methods, showed only moderate accuracy to predict the risk of in-hospital adverse events following admission for TTS. We sought to design a machine-learning (ML) based model to predict the risk of in-hospital death among patients admitted for TTS, and to provide clusters of TTS patients associated with different risks of adverse short-term prognosis.

Methods

A Penalized Logistic Regression-based ML model for predicting in-hospital death was trained and tested on a cohort of 3482 patients with TTS from the international, multicenter, InterTAK Registry. 33 clinically relevant variables were selected to be included in the prediction model. Model performance was assessed according to area under the receiver operating characteristic curve (AUC). A K-Means clustering algorithm was designed to stratify patients into phenotypic groups based on the most relevant features emerging from the main model.

Results

The overall incidence of in-hospital death was 5.2%. The InterTAK-ML model showed an AUC of 0.88 (95%CI 0.87-0.90) and 0.87 (95%CI 0.83-0.91) with respect to in-hospital death prediction in the train and test cohorts, respectively. By exploiting the 5 variables showing the highest feature importance (use of catecholamines, type of triggering factor, left ventricular ejection fraction, white blood cell count, heart rate), TTS patients were clustered into five groups associated with different risks of in-hospital death (29.4% vs 3.9% vs 1.6% vs 1.3% vs 0.7%).

Conclusion

A ML-based approach for the identification of TTS patients at risk of adverse short-term prognosis is feasible and effective. The InterTAK-ML model showed accurate discriminative capability for the prediction of in-hospital death. To support clinical decision-making, TTS patients can be clustered into groups entailing different risks of death based on routinely collected variables.