Project 467201

Recent data suggests that stereotactic body radiation therapy (SBRT) may extend progression-free (PFS) and overall survival (OS) of patients with oligometastatic disease (OMD). However, SBRT is not effective for every patient, as some will develop early distant disease progression despite treatment, and others will develop severe side effects from radiation. Preliminary work on stratifying patients into risk groups based on clinical and pathological (CP) features has yielded modest results but predictive and prognostic models combining CP and radiomic imaging features for patients have not yet been explored. The primary objective of this project is to develop a radiomic enhanced prognostic model for predicting OS following SBRT treatment in patients with OMD. Secondary objectives are to create radiomic enhanced models for PFS, local control, and treatment related toxicity. It is hypothesized that the addition of radiomic features will improve accuracy of prognostic models compared to only considering CP features. A retrospective review will be performed to collect clinical, pathological, imaging, radiation planning and outcome data for patients who received SBRT treatment for extracranial oligometastases (1-10 lesions) between January 2015 and November 2021 at the London Regional Cancer Program (n~250). Radiomic features will be extracted from radiation planning tumour contours and combined with CP features to develop survival and classification models, which will be evaluated by reserving a holdout set. To combat overfitting, feature selection techniques will be used to narrow the model input features. Survival models will be evaluated by the Concordance index, and classification models will be evaluated by the Area Under Receiver Operating Characteristic curves.