# Nomogram based on tumor burden score for prediction of prognosis of patients with hepatocellular carcinoma before hepatectomy

## TL;DR

This study creates a nomogram model to predict survival outcomes for hepatocellular carcinoma patients before surgery, using factors like tumor burden and AFP levels.

## Contribution

A novel nomogram model (OS-nomo) is developed to predict prognosis in HCC patients pre-hepatectomy using tumor burden score and other clinical factors.

## Key findings

- The OS-nomo model achieved a C-index of 0.71 for predicting overall survival.
- Patients with high tumor burden score, elevated AFP, and major hepatectomy had the worst survival outcomes.
- The model stratified patients into low-, medium-, and high-risk groups for personalized survival prediction.

## Abstract

To develop nomogram models predicting the prognosis for patients with hepatocellular carcinoma (HCC) before hepatectomy.

Patients treated at the Eastern Hepatobiliary Surgery Hospital and Zhongda Hospital, Southeast University, from January 2012 to July 2014, were retrospectively enrolled. Prediction models for overall survival (OS) and recurrence-free survival (RFS) were constructed.

A total of 1117 patients with HCC were enrolled in this study, and were divided into a training cohort (n=838) and a validation cohort (n=279). A prediction model for OS in the training cohort (OS-nomo, C-index=0.71), including alpha-fetoprotein (AFP), estimated hepatectomy extent, and tumor burden score (TBS) as independent factors (all P<0.05), was constructed. For clinical application, we stratified all patients into three distinct risk groups: low-, medium-, and high-risk group for OS, based on total points (TPs). Patients undergoing major hepatectomy, with AFP>20 ng/mL and high level of TBS had the worst OS.

When selecting patients with HCC for hepatectomy, factors including sex, CPS, AFP level, estimated hepatectomy extent, and TBS should be carefully considered. OS-nomo model could serve as important tool for personalized survival prediction.