P153 Role of 3D endoanal ultrasound in perianal fistulising Crohns's disease. Preliminary results

Abstract

Background: Perianal fistulising Crohn's Disease is a challenging clinical situation; the development of appropriate diagnostic tools is crucial for correct patient's management. The role of 3D endoanal ultrasound is well established in the diagnosis of anal fistulas. In this report we investigate if these some specific endosonogaphic features (the Crohn's Ultrasound Fistula Sign -CUFS, a double track, the presence of debris in the fistula track or abscess, the maximum width of the fistula tract) may have a role in discriminating between cryptoglandular and Crohn's Disease related fistulas. Methods: 48 consecutive patients with anal fistulas were included in the study from July 2015 to January 2016. Each patient underwent a 3D endoanal ultrasound (B-K Medical, 2052 transducer) and subsequent surgery. 11 patients had an established diagnosis of CD. The abovementioned ultrasonographic features were searched for and compared between the cryptoglandular fistulas group (“crypt group”) and the CD related fistulas group (“CD group”). Cohen K Statistics was used to determine the agreement between ultrasound diagnosis (primary orifice, tract) and operative findings. Wilcoxon rank sum test has been used to compare the fistula width between cryptogenic and CD cases. Diagnostic accuracy of the Width of the fistula tract has also been evaluated with a ROC curve and the AUC. The role of all the abovementioned signs and ultrasonography features as diagnostic tools for perianal fistulising CD has also been investigated and Sensitivity, Specificity, Accuracy, Positive and Negative Predictive Value (PPV/NPV) and Positive and Negative Likelihood ratios (PLR/ NLR) have been calculated. Statistical analysis were performed using STATA 12 statistical software. Results: Preoperative ultrasound and surgical findings showed a very good agreement (k=0.96 for primary orifice and k=0.94 for fistula tract). Mean width of the fistula tract was 2.7 mm in the crypt group and 5.1 mm in the CD group (p<0.001, Wilcoxon rank sum Test). A width over 4 mm has been proposed as a cut-off for highly suspicious CD related fistulas. The frequency observed of the CUFS, double track, debris and width >4 mm is significantly higher in the CD group than in the crypt group (p<0.01, Fisher exact test). All of these signs show a high sensitivity, specificity, positive and negative predictive value and a high positive likelihood ratio for diagnosis of perianal Crohn's disease. Conclusions: Endoanal 3D ultrasound is a safe and reliable tool for diagnosis of perianal fistulas; some specific ultrasonographic features have been described to discriminate between cryptoglandular and perianal CD related fistulas, with a high level of diagnostic accuracy.