Is the Endoscopic Remission Evaluation, Using the CREDO 1 Index / Score in CD Patients in Clinical Remission at Baseline, Predictive of Sustained Clinical Remission Using a 2-year Follow up

The CREDO 2 study follows CREDO 1 study, which aims to construct an objective evaluation of endoscopic remission in Crohn's Disease (CD). In addition to reproducibility and validation, the predictive value of this remission evaluation needs to be tested in different settings to valorise its usefulness in clinical practice and in clinical trials.

CREDO 2 aims to investigate whether the evaluation of endoscopic remission, as defined in CREDO 1, in patients in clinical remission is predictive of sustained clinical remission at 2 years. The design of CREDO2 is a multicentre longitudinal prospective cohort study. The screening period to include a patient is two weeks. Patients will be followed up to week 104. The primary endpoint is sustained clinical remission at week 104. The primary secondary endpoints are sustained clinical remission at weeks 26 and 52. The sustained clinical remission at 2 years is defined by the absence of relapse and complication of the disease. Relapse is defined as Crohn's Disease Activity Index (CDAI) > 220 or between 150 and 220 for 2 consecutive weeks with an increase of at least 70 points relative to the baseline CDAI, associated with an objective marker of inflammation: C-reactive protein (CRP) ≥5 mg/l and / or faecal calprotectin ≥250 μg/g. CD complication is defined by an intestinal resection surgery for CD, stricturoplasty, endoscopic dilatation, hospitalisation for intestinal strictures, abscess and / or fistula (including anoperineal disease) and/or therapeutic escalation. Therapeutic escalation is defined as an increase in the dosage of the treatment, shortening of the treatment interval, addition of a new treatment for CD (including corticosteroids, immunosuppressants, biologics, Janus kinase (JAK) inhibitors or any experimental treatment). The number of patients to be included is 320 when using a two-sided test with a type 1 error of 5% to detect with a power of 80% an association between a predictor and week 104 sustained remission failure, corresponding to a variation in proportions of patients who failed from 40% in the high-risk group to 20% in the low-risk group, and assuming a 12% loss due to treatment discontinuation or patient withdrawal. Patients will be recruited via two cohorts. First, patients included in CREDO 1 could be included in CREDO 2 if they agree and meet the criteria of non-exclusion. Assuming that 2/3 of the 15 patients included in each of the 16 centres involved in CREDO 1 could be included in CREDO 2, 160 patients will be included from CREDO 1. An additional cohort of 10 patients per centre will be included in CREDO 2 by the same local investigator within each centre using the same methods as for CREDO 1 cohort, except that recruitment criteria will be those of CREDO 2, to provide 160 additional patients to reach the targeted cohort size for CREDO 2. Recruitment will therefore be performed in 16 centres and each local investigator will have to register 20 videos, including those selected from CREDO 1, stratified by endoscopic remission status according to his/her global judgment: complete remission; almost complete remission; neither complete nor almost complete remission. The main inclusion criteria are adult patients, with established CD with ileal and / or colonic involvement, without significant clinical activity for more than 3 consecutive months with, at baseline, CDAI <150 and CRP <5 mg / l and faecal calprotectin <250 μg / g, stable maintenance treatment for more than 3 months, an ileocolonoscopy planned for CD and the decision to maintain or decrease the treatment (but not increase) after the colonoscopy. All maintenance treatments are authorized / immunosuppressive (azathioprine, 6-mercaptopurine, methotrexate), biologic (infliximab, adalimumab, certolizumab, golimumab, ustekinumab, vedolizumab) and JAK inhibitors.The main exclusion criteria are an incomplete record of the ileocolonoscopy or more than three resected ileocolonic segments (not counting ileocecal valve), taking NSAIDs in the two weeks prior to endoscopy, anoperineal MC without luminal involvement, suspicion of intestinal infection within 4 weeks prior to baseline endoscopic assessment. The expected period of recruitment is 12 months from the first patient included in the study and 20 patients will be included in each of the 16 centres in Belgium and France. At inclusion, demographic, phenotypic, medical history and treatment data will be collected and ileocolonoscopy will be performed and recorded by the local investigator using pre-specified standards, provided sufficient quality. Biological (albumin, haemoglobin, platelets, CRP and faecal calprotectin) assessments will be carried out at the inclusion and at each follow-up visit (week 26, 52 104 and unplanned). The clinical activity will be evaluated through CDAI. Within each centre, the local investigator, as local reader, will read the 20 ileocolonoscopy videos of the centre. In addition, each video will be read by central readers, selected among 12 central readers in four groups of 3 central readers. Each video will be read by 2 central readers and by a third one in case of disagreement between the first two. Each central reader will read a little more than 54 videos due to these disagreements. All these readers, local and central, will have undergone the training session validated by an examination on the evaluation of endoscopic remission. In a separate delayed session, the central readers will evaluate components of usual endoscopic severity indices. Agreement between data provided by local readers and central readers will be assessed through Kappa and intraclass correlation coefficient estimates. If agreement is satisfactory, endoscopic data provided by local readers will be used to investigate the association between baseline endoscopic evaluation and sustained clinical remission at week 104 using the logistic regression method, ROC curves, sensitivity and specificity. If agreement is poor, the association will be studied on central reader data using the same methods.

This study should provide a tool to evaluate the ability of endoscopic remission evaluation in patients in clinical remission to accurately predict sustained clinical remission. If it is the case, this tool could become the therapeutic objective for patients in clinical remission with ileal and/or colonic CD. If agreement is satisfactory between local and central reading, the tool could be used in therapeutic trials, but also in clinical practice. If this is not the case, the tool should be used in clinical trials using central readings.