P174. Prediction of response to anti-TNF treatment in Crohn's disease using magnetic resonance enterography (MRE) findings
D. Gibson1, A. Smyth1, D. Keegan1, K. Byrne1, E. Ryan1, D. Murphy2, G. Cullen1, H. Mulcahy1, D. Malone2, G. Doherty1, 1St Vincent's Hospital, Centre for Colorectal Disease, Dublin, Ireland, 2St Vincent's Hospital, Radiology, Dublin, Ireland
Anti tumour necrosis factor treatment (aTNFtx) is effective for moderate-severe Crohn's disease (CD). However, up to 1/3 of patients fail to respond to treatment (tx).
Evaluation of disease prior to tx is critical in small bowel CD. Magnetic resonance enterography (MRE) is as sensitive as CT in detecting inflammation, while having the appealing benefit of no radiation burden. MRE has also been validated when correlated with disease activity and severity in CD.
Aim: To determine if specific findings on MRE could help predict clinical response or early tx failure following commencement of aTNFtx.
From a prospectively maintained database of 3000 patients with IBD treated in a single centre, all patients with CD who were commenced on aTNFtx from June 2007 to 2012 were identified. Patients who had an MRE within 6 months prior to commencing aTNFtx were included.
Primary end-point: need for CD related surgery within 1 year of commencing aTNFtx.
Secondary end-points: time to surgery and time to treatment failure (either: 1. need for surgery or 2. aTNFtx stopped/switched).
Specific MRE findings; 1. penetrating complications e.g fistulae, 2. Small bowel stenosis (SBS) +/− dilatation, 3. Small bowel wall thickening (SBWT) and 4. Relative contrast enhancement (RCE) were then correlated to primary and secondary end points.
In total, 353 patients were commenced on aTNFtx for CD (182 adalimumab and 171 infliximab). Of these, 54 patients had an MRE prior to commencing aTNFtx (23 infliximab; 31 adalimumab). Median time from MRE to aTNFtx was 66.2 days (±58.4). 13/54 (24.1%) had surgery within the following year due to tx failure. The proportion of patients with SBS on MRE was significantly greater in the surgery within 1 year group; 10/13 (76.9%) vs 3/41 (7.3%) respectively (Chi-square p < 0.001). Frequency of other MRE findings was not significantly different in the group who required early surgery. Time to surgery was significantly shorter in patients with SBS on MRE (log rank p = 0.001). 34 patients were tx responders vs 20 who experienced tx failure. In multivariate Cox analysis, MRE variables independently associated with time to tx failure were SBS (p < 0.001), and penetrating complications (p = 0.027).
Presence of SBS on MRE can help predict need for early surgery and is associated with tx failure in patients who commence aTNFtx. These findings demonstrate the utility of MRE in planning tx and suggest that aTNFtx may be more effective when commenced early in disease course, prior to onset of fibrostenotic or penetrating complications.