Role of radiologic imaging at the time of initial diagnosis of stage T1b-T3b melanoma

BACKGROUND. In patients with T1b-T3b cutaneous melanoma the utility of radiologic imaging at the time of diagnosis is unclear. Whether initial imaging led to a change in stage or treatment plan was investigated. METHODS. The melanoma database was searched for patients with T1b-T3b primary lesions, clinically NO, and asymptomatic for metastatic disease. Radiologic studies conducted before wide local excision ± sentinel lymph node biopsy as well as all further imaging and investigations were analyzed. Outcome measures included upstaging, change in initial surgical management, true-positive, false-positive, true-negative, and false-negative rates of each imaging modality. RESULTS. In all, 344 preoperative imaging studies (chest x-ray [CXR], computed tomography [CT], positron emission tomography [PET]/CT) were performed on 158 patients, resulting in 49 findings suspicious for metastatic melanoma and 134 findings suggestive of nonmelanoma pathology. Only 1 of 344 (0.3%) studies, a PET/CT, correlated with confirmed metastatic melanoma. The false-positive rates were CXR 5 of 7 (71.4%), chest CT 21 of 24 (87.5%), abdomen/pelvis CT 10 of 11 (90.9%), head CT 2 of 2 (100.0%), PET/CT 3 of 5 (60.0%). No patient was upstaged or had a change in initial surgical management based on preoperative imaging. The cost of all initial imaging and imaging to follow-up abnormal findings was estimated as $555,308 for the 158 patients studied. CONCLUSIONS. Imaging at the time of initial diagnosis of T1b-T3b, clinically NO, M0 melanoma was of low yield with a high false-positive rate, and did not lead to upstaging or change in initial surgical management. These findings suggest that imaging of asymptomatic patients at the time of diagnosis may not be warranted.