Abstract
Background: Pulmonary embolism (PE) represents a major complication in patients with hematologic malignancies, yet existing risk assessment models such as the Khorana and ThroLy scores show limited applicability in this population. Novel tools incorporating routinely available clinical and laboratory markers are needed for accurate risk stratification. Objectives: To investigate the incidence and predictors of PE in patients with hematologic malignancies and to develop a new risk stratification model, the Hema-PE Score. Methods: This retrospective study included a total of 177 patients with various hematologic malignancies who were evaluated for, of whom 63 had pulmonary embolism (PE) and 114 served as controls. Clinical variables (immobility, central venous catheter) and laboratory markers (D-dimer/albumin ratio, hemoglobin, platelet count, CRP) were analyzed. Receiver operating characteristic (ROC) curve analyses were performed to assess predictive accuracy. A novel scoring system, the Hema-PE Score, was constructed and its performance compared with existing risk models. Results: PE was identified in 35% of patients. The D-dimer/albumin ratio showed strong discriminatory power for predicting PE (AUC = 0.82). Based on multivariable predictors, the Hema-PE Score was developed (range 0-7 points). At a threshold of ≥3, the score achieved 100% sensitivity and 76% specificity (AUC = 0.88). Compared with the Khorana and ThroLy scores, the Hema-PE Score demonstrated superior predictive performance across hematologic malignancy subtypes. Conclusions: The D-dimer/albumin ratio and the newly developed Hema-PE Score demonstrated strong predictive performance for pulmonary embolism in patients with hematologic malignancies. These findings suggest that the Hema-PE Score may serve as a practical and easily applicable risk stratification tool, supporting early diagnosis and guiding thromboprophylaxis decisions in clinical practice. Prospective multicenter validation studies are warranted to confirm its utility and to facilitate its integration into patient management strategies.