Machine Learning Models for Predicting the Outcomes of Surgical Treatment of Colorectal Liver Metastases

Abstract

BACKGROUND: Surgical intervention remains the cornerstone of a multidisciplinary approach in the treatment of colorectal liver metastases (CLM). Nevertheless, patient outcomes vary greatly. While predictive tools can assist decision-making and patient counseling, decades of efforts have yet to result in generating a universally adopted tool in clinical practice. STUDY DESIGN: An international collaborative database of CLM patients who underwent surgical therapy between 2000 and 2018 was used to select 1,004 operations for this study. Two different machine learning methods were applied to construct 2 predictive models for recurrence and death, using 128 clinicopathologic variables: gradient-boosted trees (GBTs) and logistic regression with bootstrapping (LRB) in a leave-one-out cross-validation. RESULTS: Median survival after resection was 47.2 months, and disease-free survival was 19.0 months, with a median follow-up of 32.0 months in the cohort. Both models had good predictive power, with GBT demonstrating a superior performance in predicting overall survival (area under the receiver operating curve [AUC] 0.773, 95% CI 0.743 to 0.801 vs LRB: AUC 0.648, 95% CI 0.614 to 0.682) and recurrence (AUC 0.635, 95% CI 0.599 to 0.669 vs LRB: AUC 0.570, 95% CI 0.535 to 0.601). Similarly, better performances were observed predicting 3- and 5-year survival, as well as 3- and 5-year recurrence, with GBT methods generating higher AUCs. CONCLUSIONS: Machine learning provides powerful tools to create predictive models of survival and recurrence after surgery for CLM. The effectiveness of both machine learning models varies, but on most occasions, GBT outperforms LRB. Prospective validation of these models lays the groundwork to adopt them in clinical practice.

Clinical Prediction Models for Prognosis of Colorectal Liver Metastases: A Comprehensive Review of Regression-Based and Machine Learning Models

“…Ten studies developed prognostic models for prediction of recurrence [21,26,30,34,55,62,63,72,79,85]. Regression-based methods were used in seven studies [26,34,55,63,72,79,85], and the remaining three studies used ML techniques, including the least absolute shrinkage and selection operator (LASSO) [62], gradient-boosted trees (GBT) [30] and random forest (RF) with a globally optimal decision tree (OPT) analysis [21].…”

“…Ten studies developed prognostic models for prediction of recurrence [21,26,30,34,55,62,63,72,79,85]. Regression-based methods were used in seven studies [26,34,55,63,72,79,85], and the remaining three studies used ML techniques, including the least absolute shrinkage and selection operator (LASSO) [62], gradient-boosted trees (GBT) [30] and random forest (RF) with a globally optimal decision tree (OPT) analysis [21]. The latter was employed to identify the ideal margin width that minimizes the probability of intrahepatic recurrence within 5 years, and margins between 9 and 11 mm were proposed according to the diameter of the largest CRLM, the primary tumour nodal status and the primary tumour site [21].…”