Objective: The objective of our study was to characterize benign lesions showing increased 18F-FDG uptake and to determine their incidence on whole-body FDG PET/CT performed in oncologic patients. In addition, the performance of PET alone and PET/CT in characterizing lesions as benign was compared.
Materials and methods: A retrospective review of 1,134 consecutive reports of PET/CT studies performed in patients with proven or suspected malignancy over a 6-month period yielded 289 patients with 313 lesions that showed increased FDG uptake but were suspected to be benign (nonphysiologic) or indeterminate. Lesions were subjectively categorized on the basis of the intensity of FDG uptake (mild, moderate, or marked) as compared with background activity. For each lesion, a decision was made as to whether a benign diagnosis could be obtained by the CT part of the study, the PET pattern, or clinical correlation, or whether histologic sampling was necessary. The performance of PET alone and PET/CT for characterizing lesions as benign was compared. Two hundred twenty-nine of the lesions were assessed further: 210 were benign and 19, malignant. The final diagnosis was determined by pathology (n = 67), PET/CT follow-up (n = 58), correlative imaging (n = 59), clinical correlation (n = 32), or typical benign pattern on PET/CT (n = 13).
Results: The causes for benign uptake of FDG were inflammatory processes (n = 154, 73.3%), benign tumors (n = 23, 11%), hematoma or seroma (n = 17, 8.1%), fracture (n = 7, 3.3%), fat necrosis (n = 3, 1.4%), and others (n = 6, 2.9%). For lesions with moderate or marked uptake of FDG (n = 117, 55.7%), a benign diagnosis could have been suggested on either PET or CT (e.g., a "hot" osteophyte) in 33 lesions (28.2%), on CT alone (e.g., peritoneal fat necrosis) in 38 lesions (32.5%), on PET alone (e.g., sialadenitis) in 10 lesions (8.5%), or by clinical correlation (e.g., dental abscess) in four lesions (3.4%). A benign diagnosis could not be established without histology (e.g., colonic polyp) in 32 lesions (27.4%). The performance of PET/CT was superior to that of PET alone in characterizing lesions as benign (p < 0.001).
Conclusion: Benign lesions with increased FDG uptake are found in more than 25% of the PET/CT studies performed in patients with proven or suspected malignancy, with inflammation being the most common cause. Lesion characterization on the CT portion of the PET/CT study increases the specificity of PET/CT reporting, especially for lesions with moderate or marked FDG uptake.