Abstract No. 315 Software-Based Quantitative Measurement of Drained Liver Volume after Percutaneous Biliary Drain Placement: A Personalized 3D Volumetry Model

Percutaneous biliary drainage (PBD) for hyperbilirubinemia correction is routinely applied on oncological patients for further administration of systemic oncotherapy, but only a minority of patients attain a normal serum bilirubin within 100 days. Currently, visual estimation of drained liver volume (DLV) is the conventional approach for decision-making. Quantitative analysis of DLV via PBD would potentially advance further standardization of clinical decision-making, which is currently an unmet need. The purpose of this study is to evaluate the feasibility and clinical utility of quantitative biliary duct-based liver volumetry using a novel semi-automatic segmentation methodology. This retrospective matched 1:1 case-control study for PBD clinical success included 10 patients with cancer (8 male, mean age, 64.5 y) with obstructive jaundice who underwent first-time PBD placement between 2010 to 2023. Clinical success was defined as total bilirubin of ≤1 mg/dL within 30 days of the procedure. Dilated biliary ducts, tumors, and liver were contoured semi-automatically on Raystation (Raysearch Laboratories, Sweden) and then imported into 3D-Slicer (www.slicer.org) for further processing. Centerlines of disconnected dilated biliary ducts were computed. The liver volume was then partitioned into biliary segments based on these centerlines. The DLV was then quantified based on the drain location and the corresponding biliary segments. The DLV percentage between the two groups was analyzed by the Mann-Whitney U test. Successful quantification of liver volume drained with PBD was achieved in all cases. The median of DLV was 70% (range 32-100%) and 48% (range 21-67%) of the total liver volume in the clinically successful and unsuccessful cohorts, respectively. The difference in DLV percentage didn't achieve a statistical difference between groups (P = 0.095). Accurately quantifying a patient-specific DLV by biliary segmentation-based three-dimensional liver volumetry is technically feasible. Further studies using this method would be of interest to determine if DLV is the primary driver for achieving serum bilirubin normalization.