Objective(s):
The treatment response to neoadjuvant chemoradiation (nCRT) differs largely in individuals treated for rectal cancer. In this study we investigated the role of radiomics to predict the pathological response in locally advanced rectal cancers at different treatment time points. 1) prior to start of any treatment using baseline T2-weighted MRI (T2W-MR) 2) at the start of radiation treatment using planning CT.
Methods
Patients who received nCRT followed by surgery between June 2017 to December 2019 were included in the study. Based on the histopathological tumour response grading (TRG) criteria, 58 patients with TRG 1 were classified as complete responders (pCR) and the rest as incomplete responders (IR). The gross tumour volume was defined by the radiation oncologists. Following resampling, 100 and 103 Pyradiomic features were extracted from T2W-MR and planning CT images, respectively. Synthetic minority oversampling technique (SMOTE) was applied to correct for class imbalance. Four Machine Learning classifiers were used to build clinical, radiomics and merged models. Model performance was evaluated on a held-out validation dataset following repeated stratified 3-fold cross validation using area-under-the-receiver-operator-characteristic curves (AUC) with bootstrap 95% confidence intervals.
Results
150 patients were included. Clinical models performed better (AUC = 0.68) than the radiomics models (AUC = 0.62). Overall, the clinical + T2W-MR merged model performed the best (AUC = 0.72) in predicting the pathological response prior to therapy. Clinical + Planning CT merged models could only achieve highest AUC = 0.66.
Conclusion
Combining clinical and radiomics from baseline T2W-MR improves the prediction of pathological response in rectal cancer. Validation in larger cohorts is warranted before they can guide clinical decisions; especially in watch and wait strategies.