Development and Validation of an MRI-Based Signature to Predict Distant Metastasis in Nasopharyngeal Carcinoma Before Initial Treatment: A Retrospective Cohort Study

Abstract

Background: We aimed to identify an MRI-based model to assess the risk of individual distant metastasis in nasopharyngeal carcinoma (NPC) patients before initial treatment. Methods: In this retrospective cohort analysis, 176 patients with nasopharyngeal carcinoma were included. By using the PyRadiomics platform, we extracted imaging textures of primary tumors in all NPC patients who did not exhibit distant metastasis before treatment. We then used mRMR and LASSO algorithms to select the strongest features and build a logistic model to predict distant metastasis in these patients. We tested the independent statistical significance of multiple clinical variables by using a multivariate logistic regression analysis. Findings: In total, 2780 radiomic features were extracted from the NPC patients. A distant metastasis MRI-based signature (DMMS), comprising seven features, was constructed to classify NPC patients into high- and low-risk groups in the training cohort, and validated in the independent validation cohort. Patients with high-risk scores had shorter overall survival than patients with low-risk scores in the training cohort (P < 0·001). A radiomics nomogram, which was based on radiomic features and clinical variables, was developed to assess the risk of distant metastasis in each patient; it showed a significant predictive ability in the training cohort [AUC, 0·84; 95% confidence interval (CI), 0·75-0·90] and the validation cohort [AUC 0·79; 95% CI, 0·63-0·95)]. Interpretation: The DMMS is a visual prognostic tool for predicting distant metastasis in NPC patients. It has the potential to improve treatment decision-support by distinguishing patients at high or low risk of distant metastasis. Funding: This research received financial support from the National Natural Science Foundation of China (81571664, 81871323, 81801665, 81771924, 81501616, 81671851, and 81527805); the National Natural Science Foundation of Guangdong Province (2018B030311024); the Science and Technology Planning Project of Guangdong Province (2016A020216020); the Scientific Research General Project of Guangzhou Science Technology and Innovation Commission (201707010328); the China Postdoctoral Science Foundation (2016M600145), National Key R&D Program of China (2017YFA0205200, 2017YFC1308700, and 2017YFC1309100). Declaration of Interest: There are no conflicts of interest to declare. Ethical Approval: The study design was approved by the appropriate ethics review board, which waived the requirement for informed consent due to the retrospective nature of the study.