Improved HLM score: prognostic impact of adding NT-proBNP in HFpEF
J A Cedeno Valdiviezo, F E Cabello Montoya, Z D KobalavaAbstract
Background
Prognostic stratification in heart failure with preserved ejection fraction (HFpEF) remains limited. Although the HLM Score provides prognostic information, its discriminative performance is moderate, suggesting that incorporation of additional biomarkers such as NT-proBNP may improve risk prediction.
Methods
A total of 389 patients with HFpEF were included (mean age 73 ± 11 years, 60% male, LVEF 55 ± 4%), with a high burden of comorbidities. Optimal cut-off values for the HLM Score and NT-proBNP were determined using receiver operating characteristic (ROC) curve analysis within the study cohort. Patients were stratified into two groups: HLM > 7 and HLM > 7 combined with NT-proBNP > 876 pg/mL. The primary outcomes were all-cause mortality and heart failure rehospitalization. Prognostic performance was assessed using ROC analysis, Kaplan–Meier survival curves, and multivariable Cox proportional hazards models.
Results
Patients with HLM > 7 plus NT-proBNP > 876 exhibited a more advanced clinical phenotype, including greater pulmonary congestion (B-lines 31(14; 47), p < 0.001), worse right ventricular dysfunction (TAPSE/PASP 0.32 (0.26 ; 0.45) mm/mmHg, p < 0.001), increased hepatic stiffness (FibroScan, 13.6 (7.7 ; 24.2) KpA p < 0.001), larger left atrial size 41 (36 ; 48) cm2 (p < 0.01), and more pronounced renal impairment (creatinine 1.22 (0.96; 1.53) mg/dL and CKD – EPI by GFR 52 (34; 64) ml/min/1.73 m2 , p < 0.01). The HLM Score alone showed poor discrimination (AUC 0.564; p = 0.07), whereas the combined model significantly improved prognostic performance (AUC 0.638; p < 0.0001). Kaplan–Meier analysis demonstrated significant curve separation only for the combined model (χ² = 14.8; p = 0.0001). In multivariable Cox regression, the HLM Score alone was not an independent predictor, while its combination with NT-proBNP was associated with a nearly threefold increased risk of events (HR 2.70; 95% CI 1.60–4.58; p = 0.0002).
Conclusion
Incorporation of NT-proBNP into the HLM Score meaningfully improves risk stratification in HFpEF, enabling robust identification of a clinically vulnerable phenotype characterized by advanced congestion and markedly higher event risk. This simple, readily applicable approach may enhance identification of high-risk patients and support more precise clinical decision-making in HFpEF.AUC improvement with HLM + NT-proBNPFor image description, please refer to the figure legend and surrounding text.Kaplan–Meier survival by HLM + NT-proBNPFor image description, please refer to the figure legend and surrounding text.