Enhancing prediction of thrombosis associated with breast cancer using prechemotherapy hematologic and coagulation characteristics

INTRODUCTION: The applicability of venous thromboembolism (VTE) risk assessment models (RAMs), to breast cancer (BC) populations remains unclear. We aimed to compare the efficacy of current RAMs and examine the potential of additional hematologic parameters and thromboelastography (TEG); a point of care test, in improving VTE prediction in breast cancer (BC) patients. METHODS: In this pilot study, female BC patients were recruited before chemotherapy and followed for 6-12 months for VTE. VTE risk was assessed using Khorana score, Vienna CATS, PROTECHT, COMPASS-CAT, New Vienna CATSCORE, MDACC CAT, and hypercoagulability status. TEG and hematologic parameters were analyzed, and a modified RAM was developed. RESULTS: Among 47 patients, 5 (10.6%) developed VTE. PROTECHT was the strongest predictor [area under the curve (AUC) = 0.844], followed by Vienna CATS (AUC = 0.781). Adding immature granulocytes and red blood cell count to PROTECHT optimized prediction (AUC = 0.856). CONCLUSION: Incorporating hematologic parameters into PROTECHT may improve VTE risk prediction in BC patients, warranting further evaluation in larger studies.