S1 Raw data -

Background
Tumor necrosis is a significant risk factor affecting patients’ prognosis after liver resection (LR) for hepatocellular carcinoma (HCC). We aimed to develop a model with tumor necrosis as a variable to predict early tumor recurrence in HCC patients undergoing LR.
Materials and methods
Patients who underwent LR between 2010 and 2018 for newly diagnosed HCC but did not receive neoadjuvant therapy were enrolled in this retrospective study. Six predictive factors based on pathological features—tumor size > 5 cm, multiple tumors, high-grade tumor differentiation, tumor necrosis, microvascular invasion, and cirrhosis—were chosen a priori based on clinical relevance to construct a multivariate logistic regression model. The variables were always retained in the model. The impact of each variable on early tumor recurrence within one year of LR was estimated and visualized using a nomogram. The nomogram’s performance was evaluated using calibration plots with bootstrapping.
Results
Early tumor recurrence was observed in 161 (21.3%) patients. The concordance index of the proposed nomogram was 0.722. The calibration plots showed good agreement between nomogram predictions and actual observations of early recurrence.
Conclusion
We developed a nomogram incorporating tumor necrosis to predict early recurrence of HCC after LR. Its predictive accuracy is satisfactory.