P120 Development and validation of a dedicated score to assess transmural response and transmural healing in patients with Crohn’s disease : final results of the DEVISE-CD project

Buisson, A.(1)*;Junda, J.(2);Vignette, J.(1);Lecoq, E.(1);Bouguen, G.(3);Pereira, B.(4);Hordonneau, C.(2);

(1)CHU Clermont-Ferrand, IBD Unit, Clermont-Ferrand, France;(2)CHU Clermont-Ferrand, Radiology department, Clermont-Ferrand, France;(3)CHU Rennes, IBD Unit, Rennes, France;(4)CHU Clermont-Ferrand, Biostatistics Unit, Clermont-Ferrand, France;


Transmural healing (TH) is associated with better outcomes that endoscopic remission in Crohn’s disease (CD) but is not yet considered as therapeutic target owing to the lack of  validated and consensual definition.

In the DEVISE-CD project, we aimed to 1) build and validate a dedicated score to assess transmural response (TR) and TH in CD patients and 2) to confirm that these definitions of TR and TH are associated with improved outcomes.


DEVISE-CD project included 3 parts: a monocenter cohort of development (N = 274 patients), a monocenter validation cohort (N = 224 patients) and a prospective multicenter validation study (N = 64 patients). Inclusion criteria are given in Figure 1.


In the development cohort, persistence of edema after treatment (HR = 4.78[1.74-13.13], p=0.002), contrast enhancement (HR=4.13[1.50-11.36], p=0.006), diffusion-weighted hyperintensity (HR=6.97 [2.50-19.41], p< 0.001), bowel thickening (HR = 3.73 [1.61- 8.66], p = 0.002), ulcers (HR = 1.90[1.21- 2.97], p=0.005), fat wrapping (HR=1.87 [1.22-2.87], p=0.004) and enlarged lymph nodes (HR=2.31 [1.48-3.63], p < 0.001) were associated with increased risk of bowel damage progression, contrary to the extent of the lesions. The study of collinearity/coexistence of these different lesions showed that edema, bowel thickening, contrast enhancement, and diffusion-weighted hyperintensity coexisted in > 95% of the cases, giving the same information. The bowel thickness (quantitative value) had the best sensitivity to change (mean difference standard = 0.30). C-score was obtained from a multivariable analysis of factors associated to bowel damage progression (Figure 2).

TH (C-score < 0.5), TR50 (> 50%-decrease of  C-score in all active segments) or TR25 (>25%-decrease) were associated with a reduced risk of bowel damage progression in the development cohort (HR=0.28[0.13-0.63], p=0.002; HR=0.30[0.15-0.63], p=0.001; HR=0.37[0.19-0.71], p=0.003, respectively) and in validation cohort (HR=0.15[0.04 -0.53], p=0.003; HR=0.36[0.14-0.88], p=0.025; HR=0.46[0.23-0.94], p=0.034; respectively).

In the multicenter study, TH, TR50 and TR25 rates were 28.3%, 41.3%, and 47.8%, respectively, after 12 weeks of anti-TNF and were associated with remission at W52(CDAI < 150, faecal calprotectin < 250 μg/g and no steroid) (OR = 4.6[1.3–15.6], p=0.016; OR=6.9[1.8–26.0], p=0.008; OR=6.0[1.6–22.3], p=0.008; respectively).


The C-score is a validated, reliable and easy-to-use tool that could be implemented in daily practice and clinical trials to assess therapeutic efficacy using news targets such as TH, TR50 and TR25.