Are staging F-18-FDG PET/MRI radiomic features associated with metastases in cancer of the gastro-esophageal junction?

Purpose: To identify quantitative imaging biomarkers at staging F-18-Fluorodeoxyglucose (FDG)-positron-emission-tomography/magnetic-resonance-imaging (PET/MRI) for predicting distant metastasis in patients with gastro-esophageal junction(GEJ) cancer. Materials and Methods: Following IRB approval and informed consent, 24 patients with histologically proven GEJ cancer were prospectively recruited; 4 patients were excluded for technical reasons. Finally, 19 male and 1 female (68.3±9.1 years) were considered. Patients were injected with 326±28 MBq FDG intravenously. Uptake time was 90 minutes. Two experienced radiologists and nuclear physicians reviewed the images in consensus. Maximum standardized uptake value (SUVmax) and tumor size were analyzed. First-order and second-order statistical texture features were computed on SUV values of the whole tumor volume. k-means clustering algorithm was used to assess the correlation of feature-pairs with the presence of distant metastases. Sensitivity (SE), specificity (SP), positive predictive value (PPV), negative predictive value (NPV) and accuracy (ACC) were calculated to quantify the discrimination ability of features. Results: Second-order entropy and maximum probability, linked to texture irregularity and homogeneity respectively, were the best feature-pair in discriminating patients with and without metastatic disease (SE=80%, SP=70%, PPV=73%, NPV=78%, ACC=75%). SUVmax (SE=80%, SP=30%, PPV=53%, NPV=60%, ACC=55%) and tumor size (SE=90%, SP=10%, PPV=50%, NPV=50%, ACC=50%) performed worse, particularly for specificity. Conclusions: These results confirm the common expectation that greater intra-tumor heterogeneity correlates with metastatic potential. The extraction of advanced quantitative PET imaging features from the primary lesion may help prognostication. Clinical Relevance statement: Radiomics may help in improving prognostication at staging.