PET or PET/CT for detection of peritoneal carcinomatosis: a meta-analysis

The present study assessed the diagnostic performances of (18)F-FDG PET or PET/CT in detecting peritoneal carcinomatosis in patients with cancer.

Through a search of MEDLINE (January 1998 to September 2012), an overall weighted average for sensitivity and specificity was calculated using the weighted averages of the sample sizes in each relevant study. Pooled estimates of positive and negative likelihood ratios were calculated using fixed and random effects models, respectively, according to the heterogeneity among studies. A summary receiver operating characteristics (sROC) curve was constructed and the area under the sROC curve (AUC) was calculated. To explore heterogeneity, due to sources other than threshold effects, I-square was calculated.

The present study included analyses of patients (n = 513) from 7 studies. Results indicated a significant heterogeneity for sensitivity and specificity (I(2) > 50% and P < 0.05). The overall pooled estimates for sensitivity and specificity of FDG PET or PET/CT scans in the detection of peritoneal carcinomatosis were 72.4% (95% CI, 64.4%-79.5%) and 96.7% (95% CI, 94.4%-98.3%), respectively. The positive likelihood ratio was 10.414 (95% CI, 6.195-17.506) and the negative likelihood ratio 0.312 (95% CI, 0.159-0.612). The AUC was 0.9404. The overall diagnostic accuracy (Q* index) was 87.8%.

The high specificity may provide the reliability of a positive FDG PET or PET/CT to detect peritoneal carcinomatosis in patients with cancer. FDG PET or PET/CT has only weak power to exclude the presence of peritoneal carcinomatosis. By a good overall diagnostic accuracy, FDG PET or PET/CT may prove beneficial to surgeons when selecting appropriate patients on whom to perform laparoscopy or laparotomy.