Search in the Abstract Database

* = Presenting author

DOP013 Definition of therapeutic response criteria using MRI in Crohn's disease patients treated with anti-TNF therapy: a multicenter prospective study (the IRMA study)

Buisson A.*1,2, Messadeg L.3, Bouguen G.4, Goutorbe F.1,5, Reimund J.-M.6, Goutte M.1,2, Sautel C.7, Duron C.8, Scanzi J.1, Reymond M.1, Allimant C.1, Dapoigny M.1, Bommelaer G.1,2, Pereira B.9, Hordonneau C.3

1University Hospital Estaing, Gastroenterology Department, Clermont-Ferrand, France 2UMR 1071 Inserm/Université d'Auvergne; USC-INRA 2018, Microbes, Intestine, Inflammation and Susceptibility of the host, Clermont-Ferrand, France 3University Hospital Estaing, Radiology Department, Clermont-Ferrand, France 4University Hospital, Rennes, France 5Hospital of Bayonne, Gastroenterology Department, Bayonne, France 6University Hospital Hautepierre, Gastroenterology Department, Strasbourg, France 7Hospital of Issoire, Gastroenterology Department, Issoire, France 8Hospital of Montluçon, Department of Gastroenterology, Montluçon, France 9University Hospital, Biostatistics Unit, DRCI, Clermont-Ferrand, France


As Crohn's disease (CD) presents with transmural inflammation, mucosal healing remains an imperfect target. Magnetic resonance enterocolonography (MREC) has been showed as a reliable tool to assess CD activity. However, the definition of MRI therapeutic response is yet to be established in CD.

We aimed to identify the variations of MRI parameters between baseline and week 12 (W12), which are predictive of corticosteroids-free deep remission (CFREM) at week 52 (W52) in patients treated with anti-TNF, to define MRI criteria of therapeutic response.


CD adult patients needing anti-TNF therapy with CDAI >150 and at least one diseased-bowel segment on MRI were consecutively included in this multicenter prospective study. All the patients underwent MREC including diffusion-weighted sequences with no bowel cleansing and no rectal enema, before starting anti-TNF agents and at W12. CFREM was defined at W52 as CDAI <150 and CRP <5 g/L and faecal calprotectin <250 μg/g with no switch of anti-TNF agents and no bowel resection.


Overall 55 CD patients with 212 digestive segments were analyzed (Table 1).

Among diseased-segments at baseline, we observed significant improvement at W12 for ulcerations (54.2% vs 36.2%, p=0.037), oedema (95.8% vs 63.8%, p=0.001), comb sign (56.2% vs 36.2%, p=0.013), enlarged mesenteric lymph nodes (55.3% vs 31.1%, p=0.001), bowel thickness (8.3 vs 6.3mm, p<0.001), affected length (13.9 vs 11.9cm, p=0.015), apparent diffusion coefficient (ADC) (1.4 vs 1.57, p=0.001), Clermont score (26.8 vs 19.9, p<0.001) and MaRIA index (26.7 vs 19.8, p=0.001).

We assessed the variation of MRI parameters between baseline and W12 in patients achieving CFREM at W52 compared to those who did not among 55 CD patients. Using ROC curves, we identified that 25%-decrease of Clermont score (AUC=0.76, Se=85.5%, Spe=66%, NPV=75%, PPV=75%) or MaRIA index (AUC=0.68, Se=76%, Spe=60%, NPV=75%, PPV=75%) were highly predictive of CFREM at W52. In addition, five factors evaluated at W12 were predictors of CFREM at W52: disappearances of ulcerations (OR=6.4, p=0.05), of oedema (OR=3.3, p=0.05), of enlarged mesenteric lymph nodes (OR=5.9, 0.041), or of sclerolipomatosis (OR=12, p=0.03), and 10%-increase of ADC (OR=4.3, p=0.02). The likelihood of achieving CFREM at W52 was 17% when 0 or 1 factor was observed compared to 70% with at least 2 factors (p=0.001).


MREC is a reliable tool to monitor therapeutic response in CD. Our MRI response criteria are highly predictive of CFREM at W52, supporting their use in daily practice and clinical trials.