P296 MAGNIFI-CD index is appropriate for treatment monitoring in perianal Crohn’s Disease

Abstract Background Perianal Crohn’s Disease (pCD) is a frequent and debilitating complication of (CD). Magnetic Resonance Imaging (MRI) the imaging modality for diagnosis, classification, treatment monitoring in pCD. To evaluate response, validated radiological index mandatory. Recently, MAGNIFI-CD has been developed. The aim our study was to validate index; evaluating reliability, responsiveness accuracy. Methods At two tertiary IBD referral centers, patients (≥16yrs) with complex pCD that had pelvic MRI before 3-24 months after medical and/or surgical were included. Two independent, blinded abdominal radiologists randomly scored total its separate items. clinicians clinical outcomes (remission, non-response) using “fistula drainage assessment” (FDA). Interobserver agreement (ICCs or weighted K), (according FDA; Wilcoxon signed rank test), test accuracy (ROC analyses follow-up (FU) MAGNIFI-CD) determined. Youden indices (YI) used choose optimal cut-off values FDA categories. Results Sixty-seven (median age 30 [IQR 23-47], 51% female) eligible mean CD duration 7.0 yrs (SD 8.7) 3.7 4.3), respectively. 44% biological naïve, 9% defunctioning ostomy 20% proctitis. Baseline 18 9–20]. an almost perfect ICC 0.88 (95%CI 0.78-0.92); reliability items moderate substantial. Between baseline FU significant decrease observed responders remitters (p<0.05), while non-significant (p=0.14) seen non-responders. AUROCFU 0.80 0.05) discriminating remission non-response/response, YImax 0.61 resulted value ≤10 sensitivity specificity 70% 80%. 0.75 0.07) response/remission non-response, 0.50 ≤14 73% 75%. Conclusion reliable supported by agreement. Given agreement, item “dominant feature” would need additional training better definitions. shows robust change does not improve significantly seems be acceptable excellent accurately determine response remission. We suggest at MAGNIFI-CD. These results support suitable trials.