바로가기  >

Abstract

Purpose: To develop and validate a radiomics model for evaluating pathologic complete response (pCR) to neoadjuvant chemoradiotherapyin patients with locally advanced rectal cancer (LARC).

Experimental Design: We enrolled 222 patients (152 in the primary cohort and 70 in the validation cohort) with clinicopathologically confirmed LARC who received chemoradiotherapy before surgery. All patients underwent T2-weighted and diffusion-weighted imaging before and after chemoradiotherapy; 2,252 radiomic features were extracted from each patient before and after treatment imaging. The two-sample t test and the least absolute shrinkage and selection operator regression were used for feature selection, whereupon a radiomics signature was built with support vector machines. Multivariable logistic regression analysis was then used to develop a radiomics model incorporating the radiomics signature and independent clinicopathologic risk factors. The performance of the radiomics model was assessed by its calibration, discrimination, and clinical usefulness with independent validation.

Results: The radiomics signature comprised 30 selected features and showed good discrimination performance in both the primary and validation cohorts. The individualized radiomics model, which incorporated the radiomics signature and tumor length, also showed good discrimination, with an area under the receiver operating characteristic curve of 0.9756 (95% confidence interval, 0.9185-0.9711) in the validation cohort, and good calibration. Decision curve analysis confirmed the clinical utility of the radiomicsmodel.

Conclusions: Using pre- and posttreatment MRI data, we developed a radiomics model with excellent performance for individualized, noninvasive prediction of pCR. This model may be used to identify LARC patients who can omit surgery after chemoradiotherapy. 

Author information

Liu Z1,2, Zhang XY2, Shi YJ2, Wang L3, Zhu HT2, Tang Z1, Wang S1, Li XT2, Tian J4,5, Sun YS6.
1
CAS Key Laboratory of Molecular Imaging, Institute of Automation, Beijing, China.
2
Key Laboratory of Carcinogenesis and Translational Research (Ministry of Education/Beijing), Department of Radiology, Peking University Cancer Hospital and Institute, Beijing, China.
3
Key Laboratory of Carcinogenesis and Translational Research (Ministry of Education/Beijing), Department of Gastrointestinal Surgery, Peking University Cancer Hospital and Institute, Beijing, China.
4
CAS Key Laboratory of Molecular Imaging, Institute of Automation, Beijing, China. sys27@163.com jie.tian@ia.ac.cn.
5
University of Chinese Academy of Sciences, Beijing, China.
6
Key Laboratory of Carcinogenesis and Translational Research (Ministry of Education/Beijing), Department of Radiology, Peking University Cancer Hospital and Institute, Beijing, China. sys27@163.com jie.tian@ia.ac.cn.