Eur Radiol. 2025 Jan;35(1):117-126. doi: 10.1007/s00330-024-10880-8. Epub 2024 Jul 19.

Shankar Kumar ¹, Nikhil Rao ², Anisha Bhagwanani ³, Thomas Parry ¹, Maira Hameed ¹, Safi Rahman ³, Heather E Fitzke ¹, Judith Holmes ¹, Benjamin Barrow ⁴, Andrew Bard ⁴, Alex Menys ⁴, David Bennett ⁵, Sue Mallett ¹, Stuart A Taylor ⁶

Author information

¹Centre for Medical Imaging, University College London (UCL), London, UK.

²Department of Radiology, University Hospitals Coventry and Warwickshire NHS Trust, Coventry, United Kingdom.

³Department of Radiology, Frimley Health NHS Foundation Trust, Frimley, United Kingdom.

⁴Motilent Limited, London, United Kingdom.

⁵Takeda Pharmaceuticals Limited, Cambridge, MA, USA.

⁶Centre for Medical Imaging, University College London (UCL), London, UK. stuart.taylor@ucl.ac.uk.

Abstract

Objectives: Magnetic resonance enterography (MRE) interpretation of Crohn's disease (CD) is subjective and uses 2D analysis. We evaluated the feasibility of volumetric measurement of terminal ileal CD on MRE compared to endoscopy and sMARIA, and the responsiveness of volumetric changes to biologics.

Methods: CD patients with MRE and contemporaneous CD endoscopic index of severity-scored ileocolonoscopy were included. A centreline was placed through the terminal ileum (TI) lumen defining the diseased bowel length on the T2-weighted non-fat saturated sequence, used by two radiologists to independently segment the bowel wall to measure volume (phase 1). In phase 2, we measured disease volume in patients treated with biologics, who had undergone pre- and post-treatment MRE, with treatment response classified via global physician assessment.

Results: Phase 1 comprised 30 patients (median age 29 (IQR 24, 34) years). Phase 2 included 12 patients (25 years (22, 38)). In phase 1, the mean of the radiologist-measured volumes was used for analysis. The median disease volume in those with endoscopically active CD was 20.9 cm³ (IQR 11.3, 44.0) compared to 5.7 cm³ (2.9, 9.8) with normal endoscopy. The mean difference in disease volume between the radiologists was 3.0 cm³ (limits of agreement -21.8, 15.9). The median disease volume of patients with active CD by sMARIA was 15.0 cm³ (8.7, 44.0) compared to 2.85 cm³ (2.6, 3.1) for those with inactive CD. Pre- and post-treatment median disease volumes were 28.5 cm³(26.4, 31.2), 11 cm³ (4.8, 16.6), respectively in biological responders, vs 26.8 cm³ (12.3, 48.7), 40.1 cm³ (10, 56.7) in non-responders.

Conclusion: Volumetric measurement of terminal ileal CD by MRE is feasible, related to endoscopy and sMARIA activity, and responsive to biologics.

Clinical relevance statement: Measuring the whole volume of diseased bowel on MRE in CD is feasible, related to how biologically active the disease is when assessed by endoscopy and by existing MRE activity scores, and is sensitive to treatment response.

Key points: MRE reporting for CD is subjective and uses 2D images rather than assessing the full disease volume. Volumetric measurement of CD relates to endoscopic activity and shows reduced disease volumes in treatment responders. This technique is an objective biomarker that can assess disease activity and treatment response, warranting validation.