Risk Assessment of Liver Metastasis in Pancreatic Cancer Patients Using Multiple Models Based on Machine Learning: A Large Population-Based Study

Abstract

Background. A more accurate prediction of liver metastasis (LM) in pancreatic cancer (PC) would help improve clinical therapeutic effects and follow-up strategies for the management of this disease. This study was to assess various prediction models to evaluate the risk of LM based on machine learning algorithms. Methods. We retrospectively reviewed clinicopathological characteristics of PC patients from the Surveillance, Epidemiology, and End Results database from 2010 to 2018. The logistic regression, extreme gradient boosting, support vector, random forest (RF), and deep neural network machine algorithms were used to establish models to predict the risk of LM in PC patients. Specificity, sensitivity, and receiver operating characteristic (ROC) curves were used to determine the discriminatory capacity of the prediction models. Results. A total of 47,919 PC patients were identified; 15,909 (33.2%) of which developed LM. After iterative filtering, a total of nine features were included to establish the risk model for LM based on machine learning. The RF showed the most promising results in the prediction of complications among the models (ROC 0.871 for training and 0.832 for test sets). In risk stratification analysis, the LM rate and 5-year cancer-specific survival (CSS) in the high-risk group were worse than those in the intermediate- and low-risk groups. Surgery, radiotherapy, and chemotherapy were found to significantly improve the CSS in the high- and intermediate-risk groups. Conclusion. In this study, the RF model constructed could accurately predict the risk of LM in PC patients, which has the potential to provide clinicians with more personalized clinical decision-making recommendations.

Construction and Validation of Risk Factors and Prognostic Model for Liver Metastasis of Pancreatic Cancer

“…Similar to previous ndings, we found that age, primary site, degree of differentiation, histological subtype, N stage, surgery, radiotherapy, lung metastasis, and bone metastasis are risk factors for PCLM (17,18). Furthermore, the prognostic factors for PCLM, including age, degree of differentiation, histological subtype, surgery, radiotherapy, chemotherapy, and lung metastases, were largely consistent with those for liver metastases of gastric cancer (19) and rectal cancer (20), bone metastases of pancreatic cancer (21), and lung metastases of PC (22).…”

Pancreatic ductal adenocarcinoma with synchronous and metachronous hepatic metastasis predicted by contrast-enhanced ultrasound

“…Mehdorn et al found significant differences in proteins involved in immune cell chemotaxis and migration and cell growth in the serum of patients with early- and late-stage hepatic metastases ( 27 ). Li et al used the clinicopathological characteristics of patients with PDAC to build artificial intelligence-based predictive models for the risk of hepatic metastasis ( 33 ). As for imaging, Zambirinis et al performed preoperative CT-enhanced scans in patients with PDAC and found that radiomics could predict early hepatic metastasis in PDAC ( 34 ).…”

“…As for imaging, Zambirinis et al performed preoperative CT-enhanced scans in patients with PDAC and found that radiomics could predict early hepatic metastasis in PDAC ( 34 ). Compared to previous studies ( 27 , 33 , 34 ), our study used a more convenient and radiation-free CEUS, which increases the use of available ultrasonographic information without adding additional burden to the patient. It thus provides a new way to assess the risk of hepatic metastasis from PDAC.…”