(P-TS-9) Algorithm Driven Transfusion Management of Coagulopathy Associated with Percutaneous Hepatic Perfusion (PHP) Procedures

Background/Case Studies: Uveal melanoma is rare, and patients develop metastatic disease in about half of all cases with isolated liver metastases being most common ( >90% of cases). Hepatic metastases are typically refractory to systemic therapy and carry a poor prognosis. Percutaneous hepatic perfusion (PHP) is a minimally invasive treatment that delivers high-dose chemotherapy directly to the liver while limiting systemic exposure and was recently FDA approved as a treatment option for isolated metastatic melanoma to the liver. Due to high doses of anticoagulants, hemodilution, intraoperative blood loss (~700ml), and extra-corporeal filtration of the blood that occurs with PHP, patients are coagulopathic at procedure’s end with virtually all requiring transfusion support.

Study

Design/Methods: Patients at a single NCI/ NCCN institution with uveal melanoma metastatic to the liver (N=16), underwent PHP every 6 -8 weeks for up to 6 treatments cycles. Pre and post procedures labs (CBC, PT, PTT, INR, Fibrinogen) and TEG were performed in 50 consecutive PHP procedures (December 2023 through May 2025). An algorithm to guide transfusion strategy for correction of procedure related coagulopathy was designed using data from the first 25 procedures (Figure A). The subsequent 25 procedures were used to validate the data. Data analyses used GraphPad Prism v9.5.1.

Results/Findings: The mean Hgb pre/post procedure was (12.9 g/dl,9.2 g/dl). The mean PLT count pre/post procedure was (221,300/uL55,170/uL). The mean fibrinogen level pre/post procedure was (327 mg/dl,154 mg/dl). The mean INR pre/post procedure was (1, 1.9). Preprocedure TEGs were consistently normal except for 2 patients showing twice each low MA (mm), (mean 49.9, normal 52-69) associated with a mean PLT count of 142,500/uL. Post procedure TEG showed a prolonged K(min) in 70% (n=31), (mean 3.1, normal 0.8-2.1), reduced angle(deg) in 61% (N=27), (mean 59.6, normal 63-78), and low MA (mm) in 77% (N=34), (mean 46.6, normal 52-69) of cases. However, only 16% (N=8) of cases had a prolonged R-CK (min), (mean 10.9, normal 4.6-9.1). A mean of 1 RBC, 2 PLT, 1 FFP, and 1 pool of 10 units of cryoprecipitate were needed to correct coagulopathy and to reach the targets set for Hgb ( >10 g/dL), PLTs ( >75,000/uL) and INR (< 1.5) to remove the central vascular lines used for the procedure. No bleeding complications were reported.

Conclusions: Implementation of TEG together with conventional coagulation labs into an integrated transfusion algorithm allowed a personalized and evidence-based transfusion approach for efficient correction of coagulopathy while optimizing blood products.

Thromboelastography (TEG), National Cancer Institute (NCI), National Comprehensive Cancer Network (NCCN)