Inefficacy of F-18 fluorodeoxy-D-glucose-positron emission tomography scans for initial evaluation in early-stage cutaneous melanoma

Abstract

Background: The purpose of the current study was to determine the sensitivity and specificity of initial F-18 fluorodeoxy-D-glucose-positron emission tomography (FDG-PET) scanning for detection of occult lymph node and distant metastases in patients with early-stage cutaneous melanoma.

Methods: The authors conducted a prospective nonrandomized clinical trial. Inclusion criteria were patients with cutaneous melanoma tumors > 1.0 mm Breslow thickness, local disease recurrence, or solitary intransit metastases without regional lymph or distant metastases by standard clinical evaluation. All patients underwent whole-body FDG-PET scanning before surgical therapy. Abnormal PET findings were studied by targeted conventional imaging and/or biopsy. FDG-PET scans were interpreted in a blinded fashion. Regional lymph node basins were staged by sentinel lymph node biopsy (SLNB). PET scan findings in regional lymph nodes were compared with histology of SLNB specimens. Abnormal distant PET scan findings were studied with repeat conventional scan imaging at 3-6 months and were correlated with the first site(s) of clinical disease recurrence. Blinded PET scan findings were correlated with all information to determine sensitivity and specificity.

Results: There were 144 assessable patients with a mean tumor depth of 2.8 mm. The median follow-up for these patients was 41.4 months. Blinded interpretations of FDG-PET scan images showed that 31 patients (21%) had signs of metastatic disease, 13 patients had probable regional lymph node metastases, and 18 patients had 23 sites of possible distant metastases. SLNB and/or follow-up demonstrated regional lymph node metastases in 43 of 184 lymph node basins in 40 patients (27.8%). Compared with all clinical information, FDG-PET scan sensitivity for detection of regional lymph node metastases was 0.21 (95% confidence [CI], 0.10-0.36) and specificity was 0.97 (95% CI, 0.93-0.99). No distant sites were confirmed to be true positive by targeted conventional imaging/biopsy at the time of presentation. Thirty-four patients (23.6%) presented with 54 foci of metastatic disease at initial disease recurrence. FDG-PET scan sensitivity for prediction of the first site(s) of clinical disease recurrence was 0.11 (95% CI, 0.04-0.23). Excluding patients with brain metastases, FDG-PET scan sensitivity for detection of occult Stage IV disease in patients was 0.04 (95% CI, 0.001-0.20) and specificity was 0.86 (95% CI, 0.79-0.92).

Conclusions: FDG-PET scanning did not impact the care of patients with early-stage melanoma already staged by standard techniques. Routine FDG-PET scanning was not recommended for the initial staging evaluation in this population.