997P - A practical risk classification of early recurrence and long term survival outcomes in hepatocellular carcinoma patients with microvascular invasion after curative hepatectomy: A decision tree analysis

Background

Microvascular invasion (MVI) is an independent poor factor for early recurrence (ER) and long-term outcomes in hepatocellular carcinoma (HCC) after hepatectomy. A practical, convenient, and accurate risk classifications is urgently needed to predict ER risk in HCC patients with MVI and to guide those patients to choose suitable adjuvant therapies to reduce postoperative recurrence risks and prolong long-term survival.

Methods

Data of 440 HCC patients with MVI after curative hepatectomy were retrospectively reviewed. They were randomly divided into a training set and a validation set. Kaplan-Meier curves and Cox regression analyses were used to identify independent prognostic factors of ER. Classification and regression tree (CART) analysis was used to develop ER risk classification in the training set and verified in the validation set.

Results

Multivariate analysis revealed that seven factors, namely, hepatitis B virus deoxyribonucleic acid load, tumor size, Edmondson-Steiner grade, MVI classification, satellite nodules, Ki67 positive index, and cytokeratin 19 expression were independent risk indexs for ER in HCC patients with MVI. The CART strategy showed a good concordance statistics of 0.77 in predicting ER risk, and had a better net benefit and a wider threshold probability range, as well as had a good agreement between the predicted value and actual observed value. The area under the time-dependent receiver operating characteristic curve of the CART strategy in predicting 1-, 3-, and 5-years recurrence-free survival (RFS) were 0.75, 0.78, and 0.84, respectively, and the corresponding overall survival (OS) were 0.74, 0.74, and 0.72, respectively, which were all significantly higher than other eight commonly classic HCC system stages (BCLC stage, Okuda stage, TNM stage, CNLC stage, French stage, CLIP score and JIS score). The ER, RFS and OS were greatly discrepancies between patients that had been stratified into three risk groups. Similar results were obtained in the validation set.

Conclusions

The CART strategy achieved optimal prediction for ER, RFS and OS for HCC patients with with MVI after curative hepatectomy.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

The authors.

Funding

This research was supported by the National Natural Science Foundation of China (NO. 82060427, 82103297), Guangxi Key Research and Development Plan (NO. GUIKEAB19245002), Guangxi Scholarship Fund of Guangxi Education Department, Guangxi Natural Science Foundation (NO. 2020GXNSFAA259080).

Disclosure

All authors have declared no conflicts of interest.