A predictive model incorporating inflammation markers for high-grade surgical complications following liver resection for hepatocellular carcinoma

Abstract

Background: Systemic inflammation and immune deficiency predispose surgical patients to infection and adversely affect postoperative recovery. We aimed to evaluate the prognostic ability of inflammation and immune-nutritional markers and to develop a predictive model for high-grade complications after resection of hepatocellular carcinoma (HCC). Methods: This study enrolled 1431 patients undergoing liver resection for primary HCC at a medical center. Preoperative neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio, prognostic nutritional index, Model for End-Stage Liver Disease score, Albumin-Bilirubin score, Fibrosis-4 score, and Aspartate Aminotransferase to Platelet Ratio Index score were assessed. Stepwise backward variable elimination was conducted to determine the factors associated with Clavien-Dindo grade III to V complications within 30-day postoperative period. The predictive model was internally validated for discrimination performance using area under the receiver operating characteristic curve (AUC). Results: A total of 106 (7.4%) patients developed high-grade complications. Four factors independently predicted a high-grade postoperative complication and were integrated into the predictive model, including NLR (adjusted odds ratio: 1.10, 95% confidence interval [CI], 1.02-1.19), diabetes mellitus, extent of hepatectomy, and intraoperative blood loss. The AUC of the model was 0.755 (95% CI, 0.678-0.832) in the validation dataset. Using the cutoff value based on Youden's index, the sensitivity and specificity of the risk score were 59.0% and 76.3%, respectively. Conclusion: Preoperative NLR independently predicted a high-grade complication after resection of HCC. The predictive model allows for identification of high-risk patients and appropriate modifications of perioperative care to improve postoperative outcomes.