AUTHOR=Wang Chen , Chen Jingjing , Zheng Nanxin , Zheng Kuo , Zhou Lu , Zhang Qianwen , Zhang Wei 

TITLE=Predicting the risk of distant metastasis in patients with locally advanced rectal cancer using model based on pre-treatment T2WI-based radiomic features plus postoperative pathological stage

JOURNAL=Frontiers in Oncology

VOLUME=Volume 13 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2023.1109588

DOI=10.3389/fonc.2023.1109588

ISSN=2234-943X

ABSTRACT=Objectives To measure the prognostic value of pre-treatment T2WI-based radiomics features plus postoperative pathological stage-based model in patients with locally advanced rectal cancer after neoadjuvant chemoradiotherapy.
Methods The T2WI based radiomics features were extracted, and a radiomic signature (RS) was constructed and validated to predict the risk of distant metastasis (DM). A nomogram model combining radiomic signature and postoperative pathological features was then built for further prediction of DM. The C-index and area under the curve (AUC) were calculated.
Results A total of 260 patients were included in this study, with 197 (75.8%) men and  a mean age of 57.2 (11.2) years. 15 radiomics features were selected and then incorporated to define the radiomic signature. Patients with high risk radiomic signature has significantly shorter distant metastasis free survival in both development cohort and a validation cohort. A nomogram consisted of radiomic signature, pathological T stage and N stage had AUC value of 0.72 (95%CI 0.60 to 0.83) in the development cohort and 0.83 (95%CI 0.73 to 0.92) in the validation cohort.
Conclusion The radiomic signature using T2WI-based radiomics features can differentiate patients with different risk of DM. A nomogram based radiomic signature and pathological stage is also a good predictor of distant metastasis free survival.