Texture analysis of [18F]-fluorodeoxyglucose-positron emission tomography/computed tomography for predicting the treatment response of postoperative recurrent or metastatic oral squamous cell carcinoma treated with cetuximab

Abstract

To assess the value of the texture analysis of fluorine-18F-fluorodeoxyglucose-positron emission tomography/computed tomography (18F-FDG-PET/CT) in predicting the treatment response of postoperative recurrent or metastatic oral squamous cell carcinoma (POR/M-OSCC) treated with cetuximab. A total of 14 patients undergoing 18F-FDG-PET/CT with POR/M-OSCC were divided into the responder and non-responder groups according to cetuximab response by Response Evaluation Criteria in Solid Tumors (RECIST). The regions of interest (ROI) were set at the POR/M-OSCC lesions with the highest uptake of 18F-FDG, and the volumetric and texture features were analyzed. The features with correlation coefficient of 0.6 or more were further evaluated using the logistic regression analysis to create a model. The SHAPEVolume(vx), SHAPEVolume(mL), metabolic tumor volume (MTV), and gray-level run-length matrix run-length nonuniformity (GLRLMRLNU) were significantly different between the responder (n\u2009=\u20096) and non-responder (n\u2009=\u20098) groups (p\u2009=\u20090.044, 0.042, 0.047, and 0.012, respectively). The model’s area under the curve (AUC) was found to be 0.83, 0.83, 0.79, and 0.92, respectively. The heatmap with PET feature dendrogram showed four distinct clusters including them in patient’s responder and non-responder groups. Higher MTV, GLRLMRLNU, SHAPEVolume(vx), and SHAPEVolume(mL) in 18F-FDG-PET images may have the prediction values for treatment response with POR/M-OSCC treated with cetuximab.