Interpretation, accuracy and management implications of FDG PET/CT in cutaneous malignant melanoma

To investigate the accuracy of different interpretative approaches and to evaluate the management implications of fluorodeoxyglucose positron emission tomography/computed tomography (FDG PET/CT) in cutaneous malignant melanoma.

We retrospectively identified 60 consecutive patients who underwent 76 PET/CT scans for cutaneous malignant melanoma. PET/CT reports were classified as positive, negative, or equivocal for regional and distant disease. Scan indication (staging, restaging, surveillance, or therapeutic monitoring), tumour stage, presence or absence of regional or distant disease, and post-scan management changes were determined by review of all available medical records. Maximum standardized uptake values (SUV(max)) of all findings were noted. Diagnostic accuracy of PET/CT was compared using either a high or low threshold interpretation (i.e. subtle, but indeterminate findings coded negative or positive, respectively). The frequency of management changes was compared between patient subgroups (stratified by tumour stage or indication).

Using a high threshold interpretative approach, the overall accuracy of PET/CT for disease was 72.4% (55/76), which was significantly (P<0.05) greater than the accuracy of 53.9% (41/76) seen when using a low threshold approach. Per scan accuracy by staging site was 92.1% (70/76) for regional and 76.3% (58/76) for distant disease. PET/CT changed management in 21 of 76 studies (27.6%). When stratified by stage and indication, management changes occurred in all patient subgroups, except for stage I patients (0 of 5).

When interpreted with a high threshold approach, PET/CT demonstrates high accuracy for the diagnosis of both regional and distant disease in cutaneous malignant melanoma and frequently changes management in patients with stage II-IV disease referred for a variety of indications.