Effectiveness of positron emission tomography for the detection of melanoma metastases

OBJECTIVE: The purpose of this study was to determine the sensitivity, specificity, and clinical utility of 18F 2-fluoro-2-deoxy-D-glucose (FDG) total-body positron emission tomography (PET) scanning for the detection of metastases in patients with malignant melanoma. SUMMARY BACKGROUND DATA: Recent preliminary reports suggest that PET using FDG may be more sensitive and specific for detection of metastatic melanoma than standard radiologic imaging studies using computed tomography (CT). PET technology is showing utility in the detection of metastatic tumors from multiple primary sites including breast, lung, lymphoma, and melanoma. However, little information is available concerning the general utility, sensitivity, and specificity of PET scanning of patients with metastatic melanoma. METHODS: One hundred three PET scans done on 76 nonrandomized patients having AJCC stage II to IV melanoma were prospectively evaluated. Patients were derived from two groups. Group 1 (63 patients) had PET, CT (chest and abdomen), and magnetic resonance imaging (MRI; brain) scans as a part of staging requirements for immunotherapy protocols. Group 2 (13 nonprotocol patients) had PET, CT, and MRI scans as in group 1, but for clinical evaluation only. PET scans were done using 12 to 20 mCi of FDG given intravenously. Results of PET scans were compared to CT scans and biopsy or cytology results. RESULTS: PET scanning for the detection of melanoma metastases had a sensitivity of 94.2% and a specificity of 83.3% compared to 55.3% and 84.4%, respectively, for CT scanning. Factors that produced false-positive PET scans were papillary carcinoma of the thyroid (1), bronchogenic carcinoma (1), inflamed epidermal cyst (1), Warthin's tumor of the parotid gland (1), surgical wound inflammation (2), leiomyoma of the uterus (1), suture granuloma (1), and endometriosis (1). The four false-negative scans were thought to be due to smaller (<0.3 to 0.5 cm) and diffuse areas of melanoma without a mass effect. CONCLUSIONS: PET scanning is extremely sensitive (94.2%) and very specific (83.3%) for identifying metastatic melanoma, particularly in soft tissues, lymph nodes, and the liver. A number of second primary or metastatic tumors and an inflammatory response can also be localized by PET. This observation mandates a close clinical correlation with positive PET and emphasizes the importance of establishing a tissue diagnosis. False-negative scans in the presence of metastases are rare (4% of scans). Metastases < or =5 mm in diameter may not image well. PET is superior to CT in detecting melanoma metastases and has a role as a primary strategy in the staging of melanoma.