Prognostic significance of the hemoglobin–albumin–lymphocyte–platelet (HALP) score in rheumatic mitral stenosis: a long-term cohort study
K Ahmet Ferhat, G Ayhan, V Can, E Ozbek, L Dursun, K Omer, R SarikayaAbstract
Background
Rheumatic mitral stenosis (RMS) is a chronic immune-mediated disease associated with substantial long-term morbidity and mortality. Biomarkers reflecting systemic vulnerability may complement traditional hemodynamic risk assessment. The hemoglobin–albumin–lymphocyte–platelet (HALP) score integrates inflammatory and nutritional parameters and has emerged as a potential prognostic marker in cardiovascular disease.
Methods
In this retrospective cohort study, 127 patients with RMS were followed for a median of 124 months. The primary endpoint was major adverse cardiac and cerebrovascular events (MACCE). Discrimination was assessed using receiver operating characteristic (ROC) analysis. Survival was evaluated by Kaplan–Meier analysis and Cox regression. Incremental value was examined using nested models with bootstrap validation, calibration metrics, net reclassification improvement (NRI), and decision curve analysis.
Results
During follow-up, 21 patients (16.5%) experienced MACCE. The HALP score demonstrated modest discrimination (AUC 0.644; 95% CI 0.509–0.773). A cutoff value of 42.08 identified patients with significantly lower MACCE-free survival (log-rank p = 0.015). In multivariable Cox analysis adjusted for age and atrial fibrillation, low HALP (<42.08) remained independently associated with MACCE (HR 3.16; 95% CI 1.00–9.95; p = 0.049). Although addition of HALP to a clinical–hemodynamic model produced small improvements in apparent discrimination, optimism-corrected performance and reclassification analyses suggested limited incremental predictive gain.
Conclusions
In RMS, a low HALP score is independently associated with long-term adverse outcomes. While its incremental discrimination beyond hemodynamic severity appears modest, HALP may serve as an accessible adjunct marker of systemic risk.Roc curve