Purpose
This study examined whether radiomics and deep learning are predictive of tumour response in patients with oesophageal squamous cell carcinoma (OSCC) treated by neoadjuvant chemo-radiotherapy (nCRT) and surgery.
nCRT followed by surgery is considered the standard of care for patients with locally advanced OSCC, although surgery, in the form of oesophagectomy, is associated with significant mortality and morbidity[1,2]. With up to a third of patients achieving pCR after nCRT, these patients may potentially be spared surgery [3, 4]
Methods and materials
Ninety-five patients who had undergone pre-treatment 18F-FDG-PET studies were included and classified as those who achieved complete pathological response (pCR) and those who did not, based on the histologic Mandard-TRG system for tumor regression grading.
Primary tumours were segmented on the PET images using a fixed threshold approach.
Radiomics features were extracted usingthe contoured volume of interest (VOI). The entire LIFEx featureset was used.
Eight deep features were extracted from a convolutional neural network (CNN) for volumetric medical imaging (VNet) trained on the VOI contours...
Results
pCR was achieved in 36 (37.9%) patients.
The HCR signature achieved an area under the receiver operating characteristic curve (AUC) of 0.787.
A single deep feature was shown to be predictive of pCR with an AUC of 0.741.
The model was improved when both the HCR and deep features were combined (AUC 0.839).
Conclusion
Our results are comparable with the published literature, providing further support for the use of radiomics in prognostication in OC [5, 6].
Our model with traditional radiomicfeatures provided similar performace to the deep feature models for predicting pCR in OSCC patients undergoing nCRT. We found only minor gains by combining the features in a hybrid model.
Personal information and conflict of interest
J. van Lunenburg; Hong Kong/HK - nothing to disclose W. H. K. Chiu; Hong Kong/HK - nothing to disclose
References
Shapiro J, van Lanschot JJB, Hulshof MCCM, van Hagen P, van Berge Henegouwen MI, Wijnhoven BPL, et al. Neoadjuvant chemoradiotherapy plus surgery versus surgery alone for oesophageal or junctional cancer (CROSS): long-term results of a randomised controlled trial. The Lancet Oncology. 2015;16:1090-8. doi:10.1016/s1470-2045(15)00040-6.
Yang H, Liu H, Chen Y, Zhu C, Fang W, Yu Z, et al. Neoadjuvant Chemoradiotherapy Followed by Surgery Versus Surgery Alone for Locally Advanced Squamous Cell Carcinoma of the Esophagus (NEOCRTEC5010): A Phase III Multicenter, Randomized, Open-Label Clinical Trial. Journal of...