F-18 fluorodeoxy-D-glucose positron emission tomography scan in the initial evaluation of patients with a primary melanoma thicker than 4 mm.

Metabolic imaging with F-18 fluorodeoxy-D-glucose positron emission tomography is one of the most sensitive and non-invasive techniques, and has proved useful in melanoma. We designed, in 2004, at the Institute Gustave Roussy, a prospective study to determine the value of F-18 fluorodeoxy-D-glucose positron emission tomography scanning in the detection of regional and/or distant metastasis in 25 new patients referred for the treatment of a primary melanoma thicker than 4 mm (tumor node metastases stage T4). The sentinel lymph node biopsy was proposed for all the patients without a palpable regional lymph node. Abnormal positron emission tomography scan findings were correlated to available histological data and to the course of the disease. The F-18 fluorodeoxy-D-glucose positron emission tomography scan identified 0/2 intact primary melanomas, 1/4 residual primary melanomas after limited excision, 0/6 lymph node basins with micrometastasis, 4/4 lymph node basins with enlarged palpable lymph nodes and 0 distant metastasis. The sensitivity and specificity of positron emission tomography scans for microscopic lymph node disease in basins were, respectively, 0 and 92%. A false-positive F-18 fluorodeoxy-D-glucose positron emission tomography result in a cervical basin led to a useless cervical lymph node dissection. In three patients, the positron emission tomography scan was positive in distant sites but none of these foci represented a true metastasis. In conclusion, it is not useful to include a positron emission tomography scan in the initial work-up of patients with primary melanoma, even in patients with thick primary melanomas (>4 mm). Sentinel lymph node biopsy remains the technique of choice for the most accurate initial staging.