SPC365

Development of a predictive model for pre-intervention adverse outcomes in patients with severe aortic stenosis awaiting transcatheter aortic valve implantation

Session: Sessão de Posters 26 - Implantação transcateter da válvula aórtica (TAVI): desfechos, complicações e biomarcadores
Speaker: Adriana Vazão

Congress: CPC 2026
Topic: F. Valvular, Myocardial, Pericardial, Pulmonary, Congenital Heart Disease
Theme: 15. Valvular Heart Disease
Subtheme: 15.2 Valvular Heart Disease – Epidemiology, Prognosis, Outcome
Session Type: Posters Eletrónicos
FP Number: ---
Authors: Adriana Vazão; André Martins; Mónica Amado; Joana Reis Pereira; Carolina Esteves; João Filipe Carvalho; Luís Graça Santos; David Durão

<p style="text-align:justify"><span style="font-size:12pt"><span style="font-family:'Times New Roman',serif"><span style="color:#000000"><strong>Background</strong></span></span></span><span style="font-size:12pt"><span style="font-family:'Times New Roman',serif"><span style="color:#000000">: Degenerative aortic stenosis is a leading cause of cardiovascular (CV) morbidity worldwide, and transcatheter aortic valve implantation (TAVI) increasingly replaces&nbsp;surgery, even in younger patients (pts). As TAVI availability remains limited and waiting times long, predictive models to identify pts at risk of clinical deterioration while awaiting intervention may be useful.</span></span></span></p>

<p style="text-align:justify"><span style="font-size:12pt"><span style="font-family:'Times New Roman',serif"><span style="color:#000000"><strong>Objective</strong></span></span></span><span style="font-size:12pt"><span style="font-family:'Times New Roman',serif"><span style="color:#000000">: To develop a predictive model for pre-intervention outcomes in pts with severe aortic stenosis (SAS) awaiting TAVI.</span></span></span></p>

<p style="text-align:justify"><span style="font-size:12pt"><span style="font-family:'Times New Roman',serif"><span style="color:#000000"><strong>Methods</strong></span></span></span><span style="font-size:12pt"><span style="font-family:'Times New Roman',serif"><span style="color:#000000">: Single-center retrospective cohort study including consecutive SAS pts who underwent&nbsp;pre-TAVI cardiac computed tomography (CCT) protocol (June 2022-October 2025). Clinical, transthoracic echocardiography (TTE) and CCT data were collected. Pts were followed from CCT until TAVI, death, or November 15, 2025 (median follow-up: 12 months).&nbsp; Pre-intervention outcomes included a composite of heart failure (HF) hospitalization and all-cause mortality. Pts with&nbsp;outcomes (group 1) were compared with those without (group 2). Independent predictors identified through multivariable logistic regression were used to construct a predictive model; discrimination was assessed by ROC&nbsp;curve analysis.&nbsp;</span></span></span></p>

<p style="text-align:justify"><span style="font-size:12pt"><span style="font-family:'Times New Roman',serif"><span style="color:#000000"><strong>Results</strong></span></span></span><span style="font-size:12pt"><span style="font-family:'Times New Roman',serif"><span style="color:#000000">: A total of 289 pts underwent pre-TAVI CCT (152 males, 52.6%; mean age 81&plusmn;5 years). During follow-up, 149 pts (51.6%) underwent TAVI (median time from CCT to TAVI: 10 [5&ndash;15] months). Pre-intervention outcomes occurred in 52 pts (18.0%), including HF hospitalization (n=31) and all-cause mortality (n=30). Group 1 pts more frequently had exertional angina (23.1 vs. 11.5%, p=0.028), atrial fibrillation (42.3 vs 20.3%, p&lt;0.001) and diuretic use (65.4 vs 41.4%, p=0.002). Baseline ECG showed more bundle branch block (42.6 vs 27.4%, p=0.041) and wider QRS (median 115 [46] vs 98 [39]ms, p=0.020). No significant differences were found in CCT parameters, including aortic valve calcium score (2297&plusmn;1133 AU vs 2560&plusmn;1332 AU, p=0.387). Echocardiography showed higher median left ventricular mass index (159 [42] vs 145 [59] g/m2, p=0.017), greater left atrial volume index (LAVi) (60&plusmn;20 vs 53&plusmn;15, p=0.002) and lower absolute global longitudinal strain (-10.3&plusmn;1.7% vs -16.9&plusmn;4.6%, p=0.028). NT-proBNP levels were also higher in group 1 (3630 [6430] vs 1130 [2235] pg/mL, p&lt;0.001). Multivariable logistic regression identified QRS duration (OR 1.010, CI 1.000-1.021, p=0.048) and LAVi (OR 1.024, CI 1.005-1.043, p=0.012) as independent predictors. The resulting model [p=1/(1+exp(-(-3.939 + 0.010xQRS + 0.024xLAVi))] showed acceptable discrimination (AUC = 0.636), with a sensitivity of 69% and specificity of 51% (Fig. 1).&nbsp;</span></span></span></p>

<p style="text-align:justify"><br />
<span style="font-size:12pt"><span style="font-family:'Times New Roman',serif"><span style="color:#000000"><strong>Conclusion</strong></span></span></span><span style="font-size:12pt"><span style="font-family:'Times New Roman',serif"><span style="color:#000000">: QRS prolongation and increased LAVi independently predicted pre-intervention outcomes in SAS pts awaiting TAVI, with modest performance. This model may help identify&nbsp;higher-risk pts who could benefit from prioritized intervention.</span></span></span></p>