Purpose: To prospectively determine the accuracy of positron emission tomography (PET)/computed tomography (CT) with added CT morphologic information for depiction of metastases in patients with high-risk melanoma and negative findings for metastases at PET, by using histologic findings or additional imaging and/or follow-up findings as reference standard.
Materials and Methods: Institutional review board approval was obtained. Informed consent was obtained from patients. One hundred twenty-four consecutive high-risk melanoma patients (65 female, 59 male; mean age, 54.4 years; range, 15-82 years) were included. Fluorine 18 fluorodeoxyglucose (FDG) PET/CT was performed. First, PET/CT scans were evaluated for presence of metastases with increased FDG uptake; CT anatomic location was determined. Lesions were considered metastases if there was focal uptake higher than that of background tissue. Second, coregis-tered CT images of combined PET/CT scans were evaluated for presence of lesions without FDG uptake. Findings were compared with reference standard findings to determine the accuracy of each evaluation. McNemar test was used to assess statistical differences in accuracy.
Results: In 53 of 124 patients, metastases were found. In 46 of 53 patients with metastases, lesions had increased FDG uptake. In seven patients with metastatic disease, metastases did not have increased FDG uptake (maximum standard uptake value [SUV],
Conclusion: Dedicated analysis of coregistered CT images significantly improves the accuracy of integrated PET/CT for depiction of metastases in patients with high-risk melanoma.