Predicting post-hepatectomy liver failure after hemi-hepatectomy using a combined model based on age, ALBI score, and the future liver remnant-to-spleen volume ratio.

Post-hepatectomy liver failure (PHLF) remains a major cause of morbidity and mortality after major hepatectomy, and volumetric assessment based on future liver remnant (FLR) alone may not fully capture hepatic functional reserve or portal hemodynamic burden. This study aimed to evaluate a combined predictive model incorporating age, albumin-bilirubin (ALBI) score, and the future liver remnant-to-spleen volume (FLR/SV) ratio for predicting PHLF following hemi-hepatectomy. A retrospective multicenter study was conducted in 224 patients who underwent right or left hemi-hepatectomy between 2017 and 2024. Clinically significant PHLF was defined as ISGLS grade B or C. Liver and spleen volumes were quantified using deep learning-based CT volumetry. Independent predictors of PHLF were identified using multivariable logistic regression analysis. Model discrimination was assessed using receiver operating characteristic (ROC) curve analysis, and risk stratification was performed based on predicted probabilities derived from the final model. Clinically significant PHLF occurred in 32 patients (14.3%). Age (OR 1.03, p = 0.016), ALBI score (OR 4.13, p = 0.005), and FLR/SV ratio (OR 0.62, p = 0.008) were identified as independent predictors of PHLF. The combined model demonstrated improved discrimination (AUC 0.820) compared with individual predictors. Observed PHLF incidence increased stepwise across predefined risk categories. A combined model integrating age, ALBI score, and FLR/SV ratio improves prediction and risk stratification of PHLF following hemi-hepatectomy, supporting individualized preoperative risk assessment and surgical planning.