Quantification of the permissible liver resection rate using a new posthepatectomy liver failure prediction score

Abstract

Introduction: A novel strategy for hepatectomy is required since advances in image simulation systems, such as CT volumetry, have allowed quantitative assessment of liver volume. We present a prediction scoring system for posthepatectomy liver failure (PHLF). Methods: A total of 232 patients who underwent hepatectomy between 2013 and 2016 were enrolled. We retrospectively analyzed each characteristic data multiplied with resection or remnant liver rate as risk factors for severe, grade-B or -C, PHLF. We constructed a new PHLF prediction score from 154 cases by 2015, and investigated its utility using validation cohort of 78 patients in 2016. Results: Severe PHLF occurred in 21/154 patients (13.6%). The results of univariate and multivariate analyses identified three independent risk factors: ICGR-15 (%) × resection rate ≥ 3.11, Platelet (×103/μl) × remnant rate ≤ 130.3, and PT (%) × remnant rate ≤ 70.62 (P< 0.05). We decided cut-off values for each factor (3.0, 130, and 70.0, respectively) and constructed Volume-associated ICG-Platelet-PT score (VIPP score: 0-3), the sum of above conditions after each is converted to one point. The incidence of severe PHLF in patients with VIPP score of 0, 1, 2, and 3 was 0.0% (0/27), 1.8% (1/56), 16.7% (8/48), and 52.2% (12/23), respectively. The AUROC curve of VIPP score for severe PHLF was 0.871. In the validation cohort, the usefulness of VIPP score was confirmed (the AUROC curve=0.809). Conclusion: VIPP score can predict critical PHLF. We propose that this score is useful for procedure selection by quantitatively calculating the permissible liver resection rate.