PURPOSE

We aimed to differentiate tuberculous peritonitis (TBP) from peritoneal carcinomatosis (PC) using a visual positron-emission tomography/computed tomography (PET/CT) scoring system based on mesenteric fluorodeoxyglucose (FDG) uptake.

METHODS

PET/CT scans from 31 patients with TBP and 92 patients with PC were retrospectively reviewed. A visual PET/CT scoring system for mesenteric FDG uptake was used according to the following characteristics: FDG uptake intensity (low = 0, moderate = 1, high = 2), FDG uptake deposits (uniform = 0, irregular = 1, ascitic = 2), FDG uptake focality (diffuse = 0, segmental = 1, focal = 2), nodularity on the corresponding CT (nonnodular = 0, micronodular = 1, macronodular = 2) and mesenteric lymphadenopathy (absent = 0, lymphadenopathy without FDG uptake = 1, lymphadenopathy with FDG uptake = 2). The FDG uptake intensity, deposits, focality, nodularity and mesenteric lymphadenopathy scores between TBP and PC were compared using chi-square tests. The diagnostic performance of this scoring system for differentiating TBP from PC was analyzed using a receiver operating characteristic (ROC) curve. P < 0.05 was considered statistically significant.

RESULTS

Twenty-four patients with TBP (77.4%) and 56 patients with PC (60.9%) had mesenteric FDG uptake (P = 0.095) and were included for evaluation with the visual PET/CT scoring system. PC lesions scored higher than TBP lesions in FDG uptake deposits (P < 0.001), focality (P < 0.001) and nodularity (P < 0.001). No significant differences were observed between PC and TBP lesions in FDG uptake intensity (P = 0.396) and lymphadenopathy (P = 0.074). The total score that combined deposits, focality and nodularity had significant value for differentiating TBP from PC (area under the curve (AUC) = 0.869, P < 0.001), and a cutoff > 1 had a sensitivity (the accuracy for diagnosis of PC) of 80.4% and a specificity (the accuracy for diagnosis of TBP) of 75.0%.

CONCLUSION

A visual PET/CT scoring system based on mesenteric FDG uptake performed well in differentiating between TBP and PC.