Objective: To assess the added benefit of scanning lower extremities and skull in addition to 'skull base to upper thigh' images in PET/CT evaluation of metastatic melanoma.
Subjects/methods: Reports of consecutive whole-body PET/CT scans from January 2003 to March 2006 in patients with melanoma were retrospectively reviewed. PET abnormalities in the brain/scalp and lower extremities were tabulated by location and whether they were 'anticipated' or 'unanticipated' based on previously available data. Findings were correlated with pathology, other imaging studies, and clinical follow-up.
Results: Two hundred and ninety-six PET/CT examinations in 173 patients with melanoma were included. Twenty-five of the 296 (8.4%) scans showed brain/scalp abnormalities. Of these, only four (1.4% of all scans) showed unanticipated abnormalities: two were false positive findings, and two (0.7% of all scans) represented metastases in addition to multiple other metastases in the usual field of view. Fifty-nine of the 296 (19.9%) scans showed lower extremity abnormalities. Of these, 13 (4.4% of all scans) showed unanticipated abnormalities which were equivocal or suggestive of malignancy: eight (2.7% of all scans) represented metastases in addition to multiple other metastases in the usual field of view, and five represented false positive findings. In no case was an unanticipated isolated malignant lesion identified in the brain/scalp or lower extremities.
Conclusions: In patients with no known or suspected primary or metastatic melanoma involving the head or extremities, inclusion of these regions on PET/CT is of low yield and appears to offer little significant additional benefit, as detection of additional metastases in these patients is unlikely to change clinical management. Routine skull base to upper thigh images may be adequate for this subset of patients with melanoma.