Establishment and Validation of a Nomogram Based on Inflammation-Immunity-Nutrition Biomarker Scores to Predict Postoperative Early Recurrence in Patients with Hepatocellular Carcinoma: A Multicenter Study

Yuhan Zhang,1,* Jin Tang,2,* Yan Liu,3,* Limei Tao,1 Zhiying Liu,1 Kexi Liao,4 Qiaoying Yuan1 1Department of Nutrition, The First Affiliated Hospital of Army Medical University, Chongqing, People’s Republic of China; 2Department of Nutrition, Guang’an People’s Hospital, Guang’an, People’s Republic of China; 3Prevention and Health Section, 958th Hospital of the Chinese People ‘s Liberation Army, Chongqing, People’s Republic of China; 4Department of Hepatobiliary Surgery, The First Affiliated Hospital of Army Medical University, Chongqing, People’s Republic of China*These authors contributed equally to this workCorrespondence: Qiaoying Yuan, Department of Nutrition, The First Affiliated Hospital of Army Medical University, Gaotanyan Road 30, Shapingba District, Chongqing, 400038, People’s Republic of China, Email qiaoyingyuan@tmmu.edu.cn Kexi Liao, Department of Hepatobiliary Surgery, The First Affiliated Hospital of Army Medical University, Gaotanyan Road 30, Shapingba District, Chongqing, 400038, People’s Republic of China, Email lkx@tmmu.edu.cnPurpose: Early postoperative recurrence of hepatocellular carcinoma (HCC) significantly impairs patient quality of life and shortens survival. However, existing models rely on single-center or single-dimensional data, making accurate detection of early postoperative HCC recurrence challenging. Thus, designing/evaluating a reliable, non-invasive, comprehensive tool to predict HCC recurrence risk is crucial for guiding postoperative individualized antitumor treatment and improving prognosis.Patients and methods: We retrospectively enrolled patients with HCC (n=1424) receiving curative-intent hepatectomy at the First Affiliated Hospital of Army Medical University of China between December 2012 and December 2022. Patients were randomly stratified into training and testing cohorts in a 7:3 ratio. Using least absolute shrinkage and selection operator (LASSO) logistic and multivariate logistic regression, we screened optimal predictors and subsequently developed a nomogram alongside an online calculator. The prediction model was externally validated at two other medical institutions (n = 218). The area under the curve (AUC) of the receiver operating characteristic, calibration, and decision curves were used to evaluate model performance.Results: The nomogram intuitively showed nine independent risk factors in the prediction model for short-term recurrence in patients with HCC: Edmondson Steiner III–IV, tumor satellite nodules, vascular invasion, largest tumor > 5 cm, alpha-fetoprotein (AFP) level ≥ 400 μg/L, DeRitis ratio ≥ 1.49, gamma-glutamyl transferase (GGT) level ≥ 63.5 U/L, prognostic nutritional index (PNI) < 46.18, and neutrophil-to-lymphocyte ratio (NLR) ≥ 1.91. The AUCs of the training, testing, and validation cohorts were 0.760 (95% CI: 0.731– 0.790), 0.784 (95% CI: 0.741– 0.828), and 0.787 (95% CI: 0.728– 0.846), respectively, indicating good predictive performance. The calibration and decision curves indicated that the model could be translated into tangible clinical benefits.Conclusion: We constructed and evaluated a nomogram based on inflammation-immunity-nutrition biomarker scores to predict early postoperative recurrence of HCC, offering a free, user-friendly online calculator for quick access to results. This calculator empowers clinicians to convert complex clinical data into actionable insights, enabling the design of risk-stratified postoperative management strategies.Keywords: HCC, postoperative early recurrence, prediction model, nomogram, inflammation-immunity-nutrition biomarker scores