Diffusion-weighted magnetic resonance imaging in ileocolonic Crohn's disease

Background: Diffusion-weighted magnetic resonance imaging (dw- MRI) utilizes differences in the motion of water molecules between tissues for image formation without administration of contrast materials. Inflammation in the bowel wall slows water transit resulting in lower apparent diffusions coefficients (ADC). Previous studies have shown that dw-MRI combined with conventional MR sequences can be useful for detection of Crohn's disease in the terminal ileum and colon. The present feasibility study examined the diagnostic performance of free-breathing dw-MRI without fasting, bowel preparation or contrast administration in ileocolonic Crohn's disease. Methods: A total of 10 patients with known Crohn's disease were included in this prospective and blinded study. dw-MRI was performed with a Philips Achieva 1.5T MR system and body coil (Philips Medical Systems, Eindhoven, The Netherlands). The MR protocol contained coronal e-thrive and SShT2 with free-breathing and a factor b fixed at 800 s/mm2. Patients were examined in the prone position before and after intravenous administration of 20 mg Hyoscin Butylbromide (Buscopan®, Boehringer Ingelheim, Basel, Switzerland). Ileocolonoscopy with Simple Endoscopic Score for Crohn's disease (SES-CD) served as gold standard. Active Crohn's disease was defined as a segmental score ≥ 1. Results: A total of 46 bowel segments were assessed with ileocolonoscopy and dw-MRI of which 22 (48%) were inflamed according to the gold standard (median SES-CD segmental score 4, range 2-8). ADC obtained with and without Buscopan correlated with a Spearman's rho of 0.64 (P < 0.001). Without Buscopan, there was a trend towards lower ADC in segments with Crohn's disease compared to segments without inflammation (1.43 x 10 − ³3 mm2/s vs. 1.48 x 10 − ³ mm2/s, P = 0.08, Table 1). However, this difference was not observed with Buscopan (P = 0.49). ROCanalysis revealed an area under the curve (AUC) of 0.56 and 0.64 with and without Buscopan, respectively (P = 0.3). In the transverse colon, dw-MRI significantly discriminated active from inactive Crohn's disease (1.58 x 10 − ³ mm2/s vs. 1.18 x 10 − ³ mm2/s, P = 0.01). Conclusions: The ability of dw-MRI to discriminate Crohn's disease from normal bowel segments is inadequate. Large variations of ADC in normal and diseased bowel segments emphasize the importance of optimal anatomical distinction for obtaining precise measurements.