Evaluation of accuracy for 18 F-FDG positron emission tomography and computed tomography for detection of lymph node metastasis in canine oral malignant melanoma.

Abstract

Tumour stage has been demonstrated to have prognostic significance in canine oral malignant melanoma (OMM). Various evaluation techniques of positron emission tomography/computed tomography (PET/CT) have been reported for staging of head-and-neck tumours in people, but canine-specific data are limited, and reports for CT accuracy have been variable. In this prospective study, the head/neck of client-owned dogs with cytologically or histologically diagnosed OMM were imaged with 18 Fluorine-fluorodeoxyglucose (18 F-FDG) PET/ CT. Bilateral mandibular lymphadenectomy was performed for histopathologic assessment. Two evaluation techniques for CT and PET were applied by four independent observers. CT evaluation utilized both a standardized grading scheme and a subjective clinical interpretation. PET evaluation was first performed solely on 18 F-FDG-uptake in lymph nodes compared to background on a truncated scan excluding the oral cavity. Subsequently, the entire head/neck scan and standardized uptake value (SUV) measurements were available. Receiver operating characteristic analysis was performed with histopathology as gold standard. Twelve dogs completed the study and metastatic OMM was identified in six mandibular lymph nodes from five dogs. Of the CT-interpretation techniques, use of clinical grading performed best (sensitivity=83% and specificity=94%). Both PET techniques resulted in 100% sensitivity, but primary tumour site evaluation and use of SUV increased specificity from 78% to 94%. The SUVmax cut-point, 3.3, led to 100% sensitivity and 83% specificity. In this population of dogs, PET appeared to be highly sensitive but at risk of being less specific without use of appropriate parameters and thresholds.