MRI and Blood-based Biomarkers Are Associated With Surgery in Children and Adults With Ileal Crohn's Disease

Abstract

Background: Despite advances in medical therapy, many children and adults with ileal Crohn's disease (CD) progress to fibrostenosis requiring surgery. We aimed to identify MRI and circulating biomarkers associated with the need for surgical management.

Methods: This prospective, multicenter study included pediatric and adult CD cases undergoing ileal resection and CD controls receiving medical therapy. Noncontrast research MRI examinations measured bowel wall 3-dimensional magnetization transfer ratio normalized to skeletal muscle (normalized 3D MTR), modified Look-Locker inversion recovery (MOLLI) T1 relaxation, intravoxel incoherent motion (IVIM) diffusion-weighted imaging metrics, and the simplified magnetic resonance index of activity (sMaRIA). Circulating biomarkers were measured on the same day as the research MRI and included CD64, extracellular matrix protein 1 (ECM1), and granulocyte-macrophage colony-stimulating factor (GM-CSF) autoantibodies (Ab). Associations between MRI and circulating biomarkers and need for ileal resection were tested using univariate and multivariable LASSO regression.

Results: Our study sample included 50 patients with CD undergoing ileal resection and 83 patients with CD receiving medical therapy; mean participant age was 23.9 ± 13.1 years. Disease duration and treatment exposures did not vary between the groups. Univariate biomarker associations with ileal resection included log GM-CSF Ab (odds ratio [OR], 2.87; P = .0009), normalized 3D MTR (OR, 1.05; P = .002), log MOLLI T1 (OR, 0.01; P = .02), log IVIM perfusion fraction (f; OR, 0.38; P = .04), and IVIM apparent diffusion coefficient (ADC; OR, 0.3; P = .001). The multivariable model for surgery based upon corrected Akaike information criterion included age (OR, 1.03; P = .29), BMI (OR, 0.91; P = .09), log GM-CSF Ab (OR, 3.37; P = .01), normalized 3D MTR (OR, 1.07; P = .007), sMaRIA (OR, 1.14; P = .61), luminal narrowing (OR, 10.19; P = .003), log C-reactive protein (normalized; OR, 2.75; P = .10), and hematocrit (OR, 0.90; P = .13).

Conclusion: After accounting for clinical and MRI measures of severity, normalized 3D MTR and GM-CSF Ab are associated with the need for surgery in ileal CD.

Keywords: Crohn’s disease; biomarkers; granulocyte-macrophage colony-stimulating factor autoantibodies; magnetic resonance imaging; surgery.

Figure 1.

Research MRI images from a surgical patient with Crohn’s disease with stricturing behavior involving the terminal ileum. A, Coronal T2-weighted single-shot fast spin-echo (SSFSE) image shows wall thickening and luminal narrowing of the terminal ileum (arrow) with more proximal bowel dilation measuring greater than 3 cm. B, Axial T2-weighted SSFSE image with fat-saturation again shows wall thickening and luminal narrowing of the terminal ileum (arrow) with intramural edema and perienteric inflammation. C, Axial 3D magnetization transfer images—MT pulse on (bottom image, arrow), MT pulse off (top image, arrow). D, Axial modified Look-Locker inversion recovery (MOLLI) T1 relaxation image (arrow). E, Representative axial intravoxel incoherent motion diffusion-weighted images with increasing b-values (left to right, arrows).

Figure 2.

Tukey box plots showing circulating and imaging biomarkers in patients with Crohn’s disease with medical vs surgical management. A-C, Fecal calprotectin, circulating soluble CD64, and circulating GM-CSF autoantibody (Ab) measurements determined by ELISA. D-F, Simplified magnetic resonance index of activity (sMaRIA) score, T1 relaxation estimates (MOLLI T1), and normalized 3D magnetization transfer ratio (MTR) determined by MRI. Ctrl: non-IBD healthy controls (n = 42); Med: Patients with CD with nonsurgical medical management (n = 83; n = 81 for MRI biomarkers); Surg: Patients with CD with surgical management (n = 50).