CT features of subcutaneous, intermuscular, and intramuscular mast cell tumors in dogs.
Rebecca J M FarmerValerie J PoirierStephanie NykampMonica JensenRobert A FosterMichelle OblakRyan B ApplebyPublished in: Veterinary radiology & ultrasound : the official journal of the American College of Veterinary Radiology and the International Veterinary Radiology Association (2022)
Surgical removal is the treatment of choice for subcutaneous (SC), intermuscular (InterM), and intramuscular (IntraM) mast cell tumors (MCTs). Advanced imaging (CT or MRI) is frequently used for presurgical planning, but InterM and IntraM MCTs can be difficult to identify and delineate on CT. Aims of the current retrospective, diagnostic accuracy, observer agreement study were to describe the imaging features of SC, InterM, and IntraM MCTs on CT and to assess the limitation of CT to identify the full local extent of the MCT. Inclusion criteria for the study were dogs with a cytologically or histologically diagnosed MCTs determined to be SC, InterM, or IntraM MCT based on histology and/or a CT scan performed in the gross disease setting. Two board-certified veterinary radiologists reviewed the CT images and recorded location, contrast enhancement pattern, and delineation between the normal and abnormal tissue. Sensitivity and specificity of CT for determining location (SC/InterM versus IntraM) was 85.71% and 55.56%, respectively, when compared to consensus location based on surgical pathology report/CT/MRI review. There was a low inter-rater agreement for delineation (kappa: 0.150 (-0.070 to 0.370) and measurement had a low/moderate correlation (rho: 0.4667 to 0.5792). Upon review by a surgical oncologist, CT findings were deemed insufficient for curative surgical planning in 13 of 16 due to inadequate definition of tumor depth, compartment boundary (fascial plane) or MCT margins. The use of CT for presurgical planning of SC/InterM/IntraM MCT dogs has limitations, especially when differentiating MCT from the adjacent muscle.
Keyphrases
- contrast enhanced
- dual energy
- image quality
- computed tomography
- magnetic resonance imaging
- positron emission tomography
- magnetic resonance
- high resolution
- machine learning
- skeletal muscle
- deep learning
- high intensity
- artificial intelligence
- mass spectrometry
- rectal cancer
- prognostic factors
- toll like receptor
- inflammatory response
- pet ct