Dual time point 18F-FDG PET imaging for differentiating malignant from inflammatory processes.

UNLABELLED The aim of this study was to investigate the difference in the rates of FDG uptake between malignant and inflammatory cells and processes. METHODS IN VITRO STUDIES (18)F-FDG uptake by different tumor cell lines (human mesothelioma [REN]; rat mesothelioma [II45]; mice melanoma [B18F10]; mice mesothelioma [AB12]; human myeloma [GM1500]; and human ovarian cancer [SKOV3]) and peripheral blood mononuclear cells isolated from 8 healthy human volunteers was measured 20 and 60 min after FDG was added into growth medium. Animal studies: II45 cells were implanted into the left flank of rats (n = 5) and a focal inflammatory reaction (mechanical irritation) was generated in the right flank. PET images at 45 and 90 min after injection of FDG were obtained and standardized uptake values (SUVs) were determined. Patient studies: Seventy-six patients who had dual time FDG PET scans were retrospectively analyzed. All results were expressed as the percentage change in SUV of the later time image from that of the earlier time (mean +/- SD). RESULTS IN VITRO STUDIES Except for the SKOV3 cell line, which had only minimally increased FDG uptake (+10% +/- 26%; P > 0.3), all other tumor cell lines tested showed significantly increased FDG uptake over time (GM1500, +59% +/- 19%; B18F10, +81% +/- 15%; AB12, 93% +/- 21%; II45, +161% +/- 21%; REN, +198% +/- 48%; P < 0.01 for all). By contrast, FDG uptake in mononuclear cells was decreased in 7 of 8 donors. Animal studies: SUVs of tumors from 90-min images were significantly higher than those from 45-min images (+18% +/- 8%; P < 0.01), whereas the SUVs of inflammatory lesions decreased over time (-17% +/- 13% of the early images; P < 0.05). CLINICAL STUDIES The SUVs of delayed images from the known malignant lesions compared with those of earlier scans increased over time (+19.18% +/- 9.58%; n = 31; P < 0.001; 95% confidence interval, 15.8%-22.6%). By contrast, the SUVs of benign lung nodules decreased slightly over time (-6.3% +/- 8.1%; n = 12; P < 0.05; 95% confidence interval, -10.9% to -1.7%). The SUV of inflammatory lesions caused by radiation therapy (+1.16% +/- 7.23%; n = 8; P > 0.05; 95% confidence interval, -3.9%-6.2%) and the lesions of painful lower limb prostheses (+4.03% +/- 11.32%; n = 25; P > 0.05; 95% confidence interval, -0.4%-8.5%) remained stable over time. CONCLUSION These preliminary data show that dual time imaging appears to be useful in distinguishing malignant from benign lesions. Further research is necessary to confirm these results.