Background: Whereas transcatheter aortic valve implantation (TAVI) has become the gold standard for aortic valve stenosis treatment in high-risk patients, it has recently been extended to include intermediate risk patients. However, the mortality rate at 5 years is still elevated. The aim of the present study was to develop a novel machine learning (ML) approach able to identify the best predictors of 5-year mortality after TAVI among several clinical and echocardiographic variables, which may improve the long-term prognosis. Methods: We retrospectively enrolled 471 patients undergoing TAVI. More than 80 pre-TAVI variables were collected and analyzed through different feature selection processes, which allowed for the identification of several variables with the highest predictive value of mortality. Different ML models were compared. Results: Multilayer perceptron resulted in the best performance in predicting mortality at 5 years after TAVI, with an area under the curve, positive predictive value, and sensitivity of 0.79, 0.73, and 0.71, respectively. Conclusions: We presented an ML approach for the assessment of risk factors for long-term mortality after TAVI to improve clinical prognosis. Fourteen potential predictors were identified with the organic mitral regurgitation (myxomatous or calcific degeneration of the leaflets and/or annulus) which showed the highest impact on 5 years mortality.

Despite expected improvement in left ventricular (LV) systolic and diastolic function after transcatheter aortic valve implantation (TAVI), the complex relationship between pre-existent LV systolic and diastolic function and changes in LV haemodynamics and clinical symptoms have been scarcely investigated. This study investigated the presence of pre-operative LV diastolic dysfunction and its improvement over time after TAVI alongside improvement in New York Heart Association (NYHA) class in high-risk patients with severe aortic stenosis. The study population (n = 358) was divided into two groups according to baseline LV ejection fraction (LVEF): LVEF < 50% (n = 96) and LVEF ≥ 50% (n = 262). We compared clinical and echocardiographic parameters between groups before TAVI, at 6 and 12 months follow-up. Grade III LV diastolic dysfunction was more frequent in patients with LVEF < 50% compared with patients with LVEF ≥ 50% (50.0 vs. 16.3%, P < 0.001). Systolic and diastolic echocardiographic parameters improved after TAVI together with improvement in NYHA class both in patients with LVEF < 50% (diastolic dysfunction grade ≥2: baseline 100% of patients; 12 months 58.8%, P < 0.001; NYHA III/IV: baseline, 93.8%; 12 months, 9.7%, P < 0.001) and with LVEF ≥ 50% (diastolic dysfunction grade ≥2: baseline, 87.1%; 12 months, 61.2%; NYHA III/IV: baseline, 74.5%; 12 months, 2.6%, P < 0.001). All-cause mortality was comparable between groups. TAVI exerts favourable effects on LV systolic and diastolic function with a remarkable improvement in LV diastolic function associated with improvement in NYHA functional class at follow-up. Prognosis at 1 year after TAVI was not influenced by baseline LV diastolic dysfunction both in patients with and without LV systolic dysfunction.

Objectives: Reduced cardiac output (CO) has been considered crucial in symptoms’ genesis in hypertrophic cardiomyopathy (HCM). Absolute value and temporal behaviour of O₂-pulse (oxygen uptake/heart rate (VO₂/HR)), and the VO₂/work relationship during exercise reflect closely stroke volume (SV) and CO changes, respectively. We hypothesise that adding O₂-pulse absolute value and kinetics, and VO₂/work relationship to standard cardiopulmonary exercise testing (CPET) could help identify more exercise-limited patients with HCM. Methods: CPETs were performed in 3 HCM dedicated clinical units. We retrospectively enrolled non-end-stage consecutive patients with HCM, grouped according to left ventricle outflow tract obstruction (LVOTO) at rest or during Valsalva manoeuvre (72% of patients with LVOTO <30; 10% between 30 and 49 and 18% ≥50 mm Hg). We evaluated the CPET response in HCM focusing on parameters strongly associated with SV and CO, such as O₂-pulse and VO₂, respectively, considering their absolute values and temporal behaviour during exercise. Results: We included 312 patients (70% males, age 49±18 years). Peak VO₂ (percentage of predicted), O₂-pulse and ventilation to carbon dioxide production (VE/VCO₂) slope did not change across LVOTO groups. Ninety-six (31%) patients with HCM presented an abnormal O₂-pulse temporal behaviour, irrespective of LVOTO values. These patients showed lower peak systolic pressure, workload (106±45 vs 130±49 W), VO₂ (21.3±6.6 vs 24.1±7.7 mL/min/kg; 74%±17% vs 80%±20%) and O₂-pulse (12 (9–14) vs 14 (11–17) mL/beat), with higher VE/VCO₂ slope (28 (25–31) vs 27 (24–31)) (p<0.005 for all). Only 2 patients had an abnormal VO₂/work slope. Conclusion: None of the frequently used CPET parameters, either as absolute values or dynamic relationships, were associated with LVOTO. Differently, an abnormal temporal behaviour of O₂-pulse during exercise, which is strongly related to inadequate SV increase, correlates with reduced functional capacity (peak and anaerobic threshold VO₂ and workload) and increased VE/VCO₂ slope, identifying more advanced disease irrespectively of LVOTO.