Upfront Surgery versus Neoadjuvant Perioperative Chemotherapy for Resectable Colorectal Liver Metastases: A Machine-Learning Decision Tree to Identify the Best Potential Candidates under a Parenchyma-Sparing Policy

Simple Summary For patients with colorectal liver metastases (CLM), it is doubtful which treatment could be better between neoadjuvant chemotherapy followed by liver resection (NEOS) and upfront surgery (UPS). Our aim was to select the candidates who may benefit more from one or another treatment developing a machine-learning model. To do so, 448 patients were analyzed, and baseline differences were levelled out thanks to an inverse probability weighting analysis. Survival rates and risk factors were estimated for the two generated pseudo-populations. The best potential treatment (BPT) for each patient was determined thanks to a mortality risk model built by Random-Forest. BPT-upfront and BPT-neoadjuvant candidates were automatically selected with the development of a classification -and -regression tree (CART). At CART, planning R1vasc surgery, primitive tumor localization, number of metastases, sex, and pre-operative CEA were the factors addressing the candidates to BPT. Thanks to the decision tree algorithm, patients may be automatically assigned to the BPT based on their tailored risk of mortality. Addressing patients to neoadjuvant systemic chemotherapy followed by surgery rather than surgical resection upfront is controversial in the case of resectable colorectal -liver metastases (CLM). The aim of this study was to develop a machine-learning model to identify the best potential candidates for upfront surgery (UPS) versus neoadjuvant perioperative chemotherapy followed by surgery (NEOS). Patients at first liver resection for CLM were consecutively enrolled and collected into two groups, regardless of whether they had UPS or NEOS. An inverse -probability weighting (IPW) was performed to weight baseline differences; survival analyses; and risk predictions were estimated. A mortality risk model was built by Random-Forest (RF) to assess the best -potential treatment (BPT) for each patient. The characteristics of BPT-upfront and BPT-neoadjuvant candidates were automatically identified after developing a classification -and -regression tree (CART). A total of 448 patients were enrolled between 2008 and 2020: 95 UPS and 353 NEOS. After IPW, two balanced pseudo-populations were obtained: UPS = 432 and NEOS = 440. Neoadjuvant therapy did not significantly affect the risk of mortality (HR 1.44, 95% CI: 0.95-2.17, p = 0.07). A mortality prediction model was fitted by RF. The BPT was NEOS for 364 patients and UPS for 84. At CART, planning R1vasc surgery was the main factor determining the best candidates for NEOS and UPS, followed by primitive tumor localization, number of metastases, sex, and pre-operative CEA. Based on these results, a decision three was developed. The proposed treatment algorithm allows for better allocation according to the patient's tailored risk of mortality.

Automated summary

This study aims to develop a machine-learning model to determine whether upfront surgery (UPS) or neoadjuvant chemotherapy followed by liver resection (NEOS) is a better treatment option for colorectal liver metastases (CLM). By conducting an inverse probability weighting analysis, baseline differences among 448 patients were leveled out. A mortality risk model was built using random-forest to identify the best potential treatment (BPT) for each patient. The BPT-upfront and BPT-neoadjuvant candidates were automatically selected using a classification-and-regression tree (CART).