P522 A multicenter study to validate Magnetic Resonance Enterography against histological assessments of stenotic disease in patients with Crohn's disease
Coimbra A.*1, Rimola J.2, Cuatrecasas M.2, De Hertogh G.3, Van Assche G.3, Vanslembrouck R.3, Glerup H.4, Hedemann Nielsen A.4, Hagemann-Madsen R.4, Bouhnik Y.5, Zappa M.5, Cazals-Hatem D.5, D'Haens G.6, Stoker J.6, Meijer S.6, Rogler G.7, Boss A.7, Weber A.7, Zhao R.1, Keir M.1, Scherl A.1, deCrespigny A.1, Lu T.1, Panes J.2
1Genentech, Inc., South San Francisco, United States 2Hospital Clinic, Barcelona, Spain 3University Hospitals Leuven, Leuven, Belgium 4Regional Hospital Silkeborg, Silkeborg, Denmark 5Hospital Beaujon, Paris, France 6Academic Medical Center (AMC), Amsterdam, Netherlands 7University Hospital Zurich, Zurich, Switzerland
Assessment of stenosis is important in the management of Crohn's Disease (CD). Magnetic Resonance Enterography (MRE) is a useful diagnostic tool for detecting bowel strictures, and parameters such as gadolinium contrast “delayed gain of enhancement” (DGE) and “magnetization transfer ratio” (MTR) have been associated with fibrosis in resected specimens.   In an attempt to develop a CD MRE imaging protocol for clinical trial use, the purpose of this study was to identify MRE techniques that show association with a predefined histological score of fibrosis, independent of inflammation, in CD. The primary objective was to verify concordance between DGE and a histopathology fibrosis score as a reference standard. MTR was evaluated as secondary and apparent diffusion coefficient (ADC) as exploratory objective.
Informed consent was obtained from 51 stenotic CD patients eligible for resection surgery across 6 European centers in a prospective trial evaluating MRE for the detection of fibrosis in CD. Imaging was performed a median of two weeks (>90% <8 weeks) prior to surgery with no treatment changes in between. Local pathologists and radiologists annotated the location of 56 histological samples on the MRE scans, which was identified in the most stenosed region of corresponding resected specimens. An experienced pathologist centrally read the 56 samples using the Chiorean score . A single radiologist measured DGE, MTR, T2, ADC, and MaRIA index [1,3]. Correlation of histology and MRE metrics was performed using Pearson's R.
No association of fibrosis or inflammation with either DGE or MTR was found. ADC was associated with fibrosis (R=−0.36, p=0.011). MaRIA score was correlated with inflammation (0.31, 0.031). Monitoring of image data quality by central reader resulted in >95% evaluable scans.
In this prospective, multi-center study, DGE and MTR were not found to be in concordance with histological measure of fibrosis, in contrast to previously published findings. This may have been due to a spatial mismatch between histological tissue sample and imaging measurement region of interest, inconsistencies in the interval between imaging and surgery, the effects of histological preparation, or lack of granularity in histological scoring. However MaRIA scores were in good agreement with inflammation scoring, so it is possible that MRE metrics assist in the assessment of CD stenosis. Further development of histological scores and MRE metrics will inform upon MRE protocol development in CD.
 Jordi Rimola et al. (2015), Characterization of Inflammation and Fibrosis in Crohn's Disease Lesions by Magnetic Resonance Imaging, Am J Gastroenterol
 Sheila Pazahr et al. (2013), Magnetization transfer for the assessment of bowel fibrosis in patients with Crohn's disease: initial experience, Magn Reson Mater Phy
 Jordi Rimola et al. (2009), Magnetic resonance for assessment of disease activity and severity in ileocolonic Crohn's disease, Gut