Integration of echocardiographic and biomarker parameters for functional stratification and risk prediction in HFpEF patients

Abstract Background Heart failure with preserved ejection fraction (HFpEF) presents diagnostic and prognostic challenges due to its heterogeneous pathophysiology. A multimodal strategy combining echocardiographic and biomarker data may enhance clinical risk stratification. Purpose To assess the combined value of echocardiographic and circulating biomarker parameters in predicting symptom severity and HF hospitalizations in patients with HFpEF. Methods We prospectively evaluated 88 patients with chronic HFpEF (mean age 73.7 ± 8.2 years; 60.2% female). Baseline data included clinical and echocardiographic variables alongside serum levels of NT-proBNP, ST2, IGFBP-7, and CA-125. Logistic regression and ROC analysis were used to assess predictive performance. Results During a median follow-up of 15.9 months, 17 out of 88 patients were hospitalized for HF. Among echocardiographic parameters, the strongest predictors of hospitalization were left ventricular mass index (LVMi ≥108 g/m²; AUC 0.805, sensitivity 88.9%, specificity 73%) and septal E/e′ ratio ≥15.6 (AUC 0.793, sensitivity 88.9%, specificity 81.4%). From the biomarker panel, ST2 had the highest predictive performance (AUC 0.78; cut-off 18.3 ng/mL), followed by NT-proBNP (AUC 0.699; cut-off 967 pg/mL) and CA-125 (AUC 0.665; cut-off 5.9 U/mL). In multivariate analysis, ST2 (HR 1.085; p=0.007), NT-proBNP (HR 1.001; p=0.032) and LVMi (HR 1.021; p=0.041) remained independent predictors of hospitalization. Moreover, a high-risk subgroup defined by LVMi ≥108 g/m², NT-proBNP ≥967 pg/mL, and ST2 ≥18.3 ng/mL exhibited a six-fold increased risk of hospitalization. A combined logistic regression model incorporating LVMi, septal E/e′, NT-proBNP, ST2, and CA-125 achieved excellent discriminative performance for predicting HF hospitalization, with an AUC of 0.861 (95% CI: 0.770–0.953; p