Crohn's Disease Endoscopic REmission Definition in an Objective Way (CREDO1)

The CDEIS (Crohn's disease endoscopic index of severity) and its derivative, SES-CD (single endoscopic score for Crohn's disease) were constructed to evaluate the endoscopic severity of Crohn's disease (CD). The current therapeutic goal in CD is endoscopic remission, which is thought to be predictive of a favourable evolution of the disease. However, up to now there is no measure of endoscopic remission, objectively constructed and validated, not even a consensus for its definition. An objective measure of endoscopic remission is necessary, notably for the evaluation of new treatments, a request of the regulatory agencies (FDA and EMA), but also in clinical practice if one wishes to adapt the treatment of patient according to the level of endoscopic remission. The aim of the study is to construct a measure of endoscopic remission based on an objective assessment (depth or grade) of remission, that must be reproducible, little or not dependent on the reader interpreting the endoscopy. The CREDO 1 program is a cross-sectional study designed to construct objectively a measure of the endoscopic remission of CD through an index measuring its depth or a score measuring its grade and to evaluate its reproducibility. This study is composed of four main parts: Training and selection of readers: Selection and definition of all lesions to be studied at consensus meetings, first between the central readers, then with the participation of local investigators. Specific training of local investigators in the recognition of endoscopic lesions Checking the performance of the central readers Recruitment of patients by investigators - local readers: Inclusion criteria: CD diagnosed for more than 6 months, in clinical remission for at least 3 months, with a Crohn's disease activity index < 150 at the time of endoscopy. Each local investigator will have to record 15 videos divided according to his/her overall judgment into 3 status of remission: Complete Remission; Remission almost complete; Remission neither complete nor nearly complete. The colon will be divided into 7 segments: ileum, ileocecal valve, right colon, transverse colon, descending colon, sigmoid colon and rectum. Each segment will be identified by landmarks. Videos will be anonymized and then centralized at a healthcare provider (OVH - SAS). 16 French and Belgian centers will participate in the study leading to a total of 240 films. Analysis of videos by central readers: 12 central readers will independently read 80 out of the 240 videos Videos allocation to central readers will be performed through balanced incomplete block method to ensure that each central reader, as well as each central reader pair, has the same weight in achieving the results. A quantitative evaluation of the depth of endoscopic remission on a visual analogue scale and a qualitative evaluation of the grade of remission on a 5-level ordinal scale will be provided by each central reader, segment by segment and globally, in addition to the description of all lesions identified by segment. Analysis of the data recorded by the central readers: An index (depth of remission) or score (grade of remission) will be constructed from linear or multiple logistic regression models, considering the reader and endoscopy as random factors, to take into account the dependence between the readings performed on different endoscopies by the same central reader or on the same endoscopy by different readers. Variations in readings between central readers will be quantified through the Kappa coefficient for qualitative items and through the intraclass correlation coefficient for quantitative items. In an observational longitudinal multi-center cohort study (CREDO 2), it will be tested whether the index or score thus constructed is able to predict prolonged clinical remission for 2 years in patients in clinical remission at baseline. If this test is successful, we will have a primary objective and reproducible criterion for evaluating treatments in CD and a useful tool for the management of these patients in clinical practice.

Building an objective definition of the depth or grade of endoscopic remission in relation to what is observed during endoscopy in CD patients presenting with clinical remission

12 Central Readers, selected from the participating centers independent of the patient selection will be involved in the central reading process. They will have experience in IBD endoscopy and its evaluation and will be involved in the lesions' selection that need to be described during videos' readings. Videos allocation will be done by a randomized balanced incomplete block design. Each CR will review 80 videos and each video will be read by 4 CR (6 couples of 2 CR). The efficacy of this design is 0.90, meaning that this design has the same power than a complete design with a sample size reduced by 10% whereas the reading burden is reduced by 67% (80 instead of 240). Six videos, two of each remission status randomly selected, will be read a second time by the CRs to assess intra observer variability. The CR will provide the detection and quantification of the lesions in the endoscopic images, and a quantitative and qualitative remission evaluation

For each observed segment, the central reader will first evaluate the quality of the bowel preparation and the recording, secondly the CR documents all lesions, detected or not, and quantified those detected, then the CR will perform his/her segmental endoscopic remission evaluation, qualitative on a 5-grade scale and quantitative on a linear analogue scale. Finally, after the readings of all segments, the CR will provide his/her global endoscopic remission evaluations, qualitative on a 5-grade scale and quantitative on a linear analogue scale. Each CR will review 80 videos and each video will be read by 4 CR (6 couples of 2 CR). Based on the this the inter-observer variability will be assessed. Six videos, two of each remission status randomly selected, will be read a second time in a random order by each of the 12 CRs to assess intra observer variability.

For the quantitative evaluation, the CR had to: provide his/her evaluation of depth of remission per segment (ileum, ileocecal valve, ascending colon, transverse colon, descending colon, sigmoid colon and rectum) in marking a cross on a linear analog scale, with left extremity being absolutely no remission and right extremity being absolutely complete remission. The evaluation needs to be done directly after the viewing of the segment not at the end of the viewing of the complete recording. provide his/her evaluation of the depth of remission globally in marking a cross on a linear analog scale, with left extremity being absolutely no remission and right extremity being absolutely complete remission. The global evaluation needs to be done after the viewing of the complete recording. This global evaluation is not necessarily equal to the maximum or to the mean of the segmental evaluations.

Inclusion Criteria: ≥18 years of age Established CD for more than 6 months with histopathological confirmation available in the medical records of the patient Clinical remission as considered by the investigators global assessment ≥3 consecutive months Clinical remission defined as CDAI <150 at the moment of the endoscopy and CRP <5 mg/l and fecal calprotectin <250 µg/g Planned ileocolonoscopy for CD All treatments for CD are allowed The subject signs and dates a written, informed consent form and any required privacy authorization prior to the initiation of any study procedures. Exclusion Criteria: Pregnancy at time of endoscopy People unable to give their consent (because of their physical or mental state). Absence of written consent. Ulcerative colitis or IBD type unclassified Specific postsurgical settings: ileoanal anastomosis, ileostomy or colostomy Less than 3 evaluable ileocolonic segments (Incomplete recording of the endoscopy or more than three resected ileocolonic segments (not counting ileocecal valve)) in CREDO 2 Non-steroidal anti-inflammatory drugs intake within two weeks before baseline endoscopy Perianal fistulizing CD without luminal disease Contraindication for endoscopy Inaccessible ileocolonic segment even after attempt of endoscopic balloon dilation Suspicion of gastrointestinal infection within 4 weeks prior to baseline endoscopy Documented active or suspicion of intestinal tuberculosis Conditions which in the opinion of the investigator may interfere with the subject's ability to comply to the follow up with the study procedures. Exclusive CD of the upper gastrointestinal tract (Montreal classification L4) Montreal classification L1 without terminal ileal involvement Colon preparation with solutions other than macrogol or picosulphate solutions