Pre-treatment Magnetic resonance Enterography Findings Predict the Response to TNF-alpha Inhibitors in Crohn's Disease

Jordi Rimola; Agnès Fernàndez-Clotet; Nunzia Capozzi; Sònia Rojas-Farreras; Ignacio Alfaro; Sonia Rodríguez; Maria-Carme Masamunt; Elena Ricart; Ingrid Ordás; Julian Panés


Aliment Pharmacol Ther. 2020;52(10):1563-1573. 

In This Article

Abstract and Introduction


Background: Identifying predictors of therapeutic response is the cornerstone of personalised medicine.

Aim: To identify predictors of long-term healing of severe inflammatory lesions based on magnetic resonance enterography (MRE) findings in patients with Crohn's disease (CD) treated with tumour necrosis factor alpha (TNF-α) inhibitors.

Methods: This prospective longitudinal single-centre study included patients with clinically active CD requiring treatment with TNF-α inhibitors with at least one intestinal segment with a severe inflammatory lesion detected by MRE (segmental MaRIA ≥11). MRE data were obtained at baseline, and at weeks 14 and 46. The primary endpoint was healing of severe inflammatory lesions (MaRIA <11) in each segment. The secondary endpoint was healing of all severe inflammatory lesions on a per-patient analysis.

Results: We included 58 patients with 86 intestinal segments with severe inflammatory lesions. At week 46, healing of severe lesions was found in 51/86 (59.3%) segments, and complete healing of inflammatory lesions in all segments was found in 28/58 (48.6%) patients. Multivariable analysis found baseline-negative predictors of long-term healing of severe inflammation were ileal (as opposed to colonic) location (OR 0.00, [0.00–0.56] P = 0.002) and presence of creeping fat on MRE (OR 0.00 [0.00–0.57]; P = 0.001). Persistence of segmental MaRIA score >10.6 at week 14 was a negative predictor of healing at week 46 (OR 0.3 [0.04--0.38]; P < 0.001).

Conclusion: In patients with CD, the absence of creeping fat detected at baseline MRE and location of severe inflammatory lesions are clinically relevant predictors of long-term healing of severe inflammation under treatment with TNF-α inhibitors.


In the era of biological drugs, healing of severe endoscopic lesions has become a key therapeutic objective in Crohn's disease (CD) as achieving this goal is associated with improved long-term clinical outcomes, including higher rates of sustained clinical remission, decreased hospitalisation requirements and reduced need for surgery.[1]

Ileocolonoscopy requires bowel preparation and sedation or anaesthesia, and it also involves a small risk of complications, including bowel perforation. Moreover, in patients with active CD, the right colon can be reached in 90% of ileocolonoscopy examinations, but the terminal ileum is reached in only 72%-90%.[2–5] Finally, because CD is a transmural disease, endoscopic inspection of the mucosa alone cannot provide a complete assessment of all lesions. Both ileocolonoscopy and clinical assessment have a low sensitivity for the detection of penetrating complications (eg intra-abdominal fistulas and abscesses).[3,4,6–8]

Cross-sectional imaging enables the assessment of transmural lesions and extramural complications, providing additional information to complement ileocolonoscopy.

Recent studies exploring the role of magnetic resonance enterography (MRE) as a tool for detecting the response to treatment with TNF-α inhibitor agents and/or corticosteroids[9–13] found good agreement between ileocolonoscopy and MRE for assessing response after induction treatment. Moreover, recent evidence suggests that patients achieving mural healing on computed tomography enterography[14] or MRE have better long-term outcomes than patients achieving mucosal healing on endoscopy.[11,15,16] Although these findings need to be confirmed in further studies, they support the notion that transmural healing of severe intestinal inflammation on MRE might be a preferable therapeutic objective.

Among patients with CD, the response to tumor necrosis factor alpha (TNF-α) inhibitor therapy varies; less than 30% of those starting this treatment achieve long-term remission. Although various baseline clinical predictors of response have been reported,[17–19] the magnitude of their effect is small and insufficient to guide therapeutic decisions. Other studies focusing on endoscopic response found that detecting an early response may predict sustained remission.[20–22] However, predictive factors that can be identified before therapeutic decisions would be more useful, enabling individualised treatment. To our knowledge, no prospective observational studies have been specifically designed to identify MRE-based predictors of response and long-term remission to TNF inhibitors.[1]

The primary aim of this study was to identify predictors on pre-treatment MRE findings associated with long-term therapeutic success defined as absence of severe active inflammation on MRE at week-46 of TNF-α inhibitor treatment, at both segment-level and patient-level analysis. A secondary aim was to determine whether changes achieved during the induction period predict healing of severe inflammatory lesions at week 46.