Diagnostic performance of CT for differentiating peritoneal tuberculosis from peritoneal carcinomatosis: a systematic review and meta-analysis

Abstract

To undertake a systematic review and meta-analysis of the diagnostic performance of CT for differentiating peritoneal tuberculosis (PTB) from peritoneal carcinomatosis (PC). PubMed, Embase, the Web of Science, and the Cochrane Library were searched for papers published before 23 July 2019. The methodological quality of the studies was analysed. Overlapping descriptors used in different studies to denote the same image finding were subsumed under a single CT feature. Sensitivity, specificity, and positive and negative likelihood ratios were pooled. A summary receiver operating characteristic curve (sROC) was constructed and the area under the curve (AUC) of the included studies was calculated when possible. Six studies were included and 17 CT features were analysed. The pooled sensitivity and specificity of smooth peritoneal thickening were 59% (95% CI: 52-66%) and 84% (95% CI: 79-88%), respectively. The AUC of smooth peritoneal thickening was 0.83. Omentum line/rim, lymph node necrosis or calcification, and mesenteric macro nodules had a pooled specificity ranging from 95% to 100% and a pooled sensitivity ranging from 12% to 67%. The other 12 signs had a pooled sensitivity ranging from 21% to 79% and a pooled specificity ranging from 19% to 81%. Omentum involvement (cake-like pattern) showed a threshold-effect, so only the AUC (=0.70) was calculated. Smooth peritoneal thickening shows fairly good diagnostic accuracy, while omentum rim/line, lymph nodes necrosis or calcification, mesenteric macro nodules have good specificity but limited sensitivity. The informative features summarised in this study may aid clinical practice and future studies.