Pancolonic Endoscopic and Histologic Evaluation for Relapse Prediction in Patients With Ulcerative Colitis in Clinical Remission

Miyuki Kaneshiro; Kento Takenaka; Kohei Suzuki; Toshimitsu Fujii; Shuji Hibiya; Ami Kawamoto; Maiko Motobayashi; Hiromichi Shimizu; Masakazu Nagahori; Eiko Saito; Ryuichi Okamoto; Kazuo Ohtsuka; Mamoru Watanabe


Aliment Pharmacol Ther. 2021;53(8):900-907. 

In This Article



From April 2018 to March 2019, consecutive patients with ulcerative colitis who had been in clinical and biochemical remission for at least 3 months prior were enrolled in this study conducted at Tokyo Medical and Dental University Hospital in Japan. Clinical remission was defined as a partial Mayo score of ≤2 and no individual subscore of >1.[14] The exclusion criteria were as follows: (a) proctitis, (b) change in medical treatment 3 months before enrollment; (c) prior colon surgery, unclassified inflammatory bowel disease, Crohn's disease, colorectal neoplasia, or concomitant infectious colitis examined during blood and stool testing; (d) colonoscopy contraindication; (e) biopsy contraindication because of hematopoietic disease or antithrombotic/anticoagulation therapy. On the day of colonoscopy, baseline data were recorded, and the disease extent was defined using the Montreal criteria.[15] The study was approved by the ethics committee of Tokyo Medical and Dental University. All eligible patients provided written informed consent.

Endoscopic and Histological Evaluations

All patients underwent standard bowel preparation with Niflec or MoviPrep (EA Pharma Co.), followed by endoscopic evaluation systematically. We performed standard colonoscopies (CF-HQ290ZI, CF-HQ290I, PCF-H290ZI, and CF-H260AZI; Olympus Medical Systems), captured endoscopic images from the most severe area of each colonic segment (ie, ascending, transverse, descending, sigmoid, and rectum), and subsequently obtained five mucosal biopsies from the centre of the corresponding area. Routinely, we perform chromoendoscopy after white-light evaluation; however, for this study, we only used white-light images.

Then, endoscopic severity was evaluated using the UCEIS.[6,7] In the UCEIS scoring system, the vascular pattern, bleeding, and erosions and ulcers correspond to 0–2, 0–3, and 0–3 points, respectively. We also used the UCEIS in scoring every captured endoscopic image. Next, three endoscopists, who had 11, 13, and 32 years of experience in IBD endoscopy, evaluated the UCEIS score of each image; any discrepancy was discussed amongst the endoscopists before determining the final score. The endoscopists were blinded to the clinical information and the results of other examinations.

Each of the five colonic segments mentioned above underwent biopsy; thus, five biopsy specimens were obtained. Using the Geboes score, three gastrointestinal pathologists with expert licences evaluated the histological severity of inflammation for every mucosal specimen;[16] any discrepancy in the final scores was discussed and resolved with a board-certified pathologist who had 13 years of experience in inflammatory bowel disease pathology. These pathologists were blinded to the clinical information and the results of the endoscopic studies.

Study Design and Endpoints

Originally, the UCEIS score was evaluated by sigmoidoscopy. From this score, the following three different modes of the endoscopic evaluation were compared (Figure 1): (a) "original UCEIS," which represented the score for the most severe area limited to the sigmoid colon and rectum (0–8 points); (b) "worst affected UCEIS," which depicted the score for the single most severely affected area throughout the entire colon (0–8 points); (c) "pancolonic UCEIS," which represented the cumulative score of the five colonic segments (0–40 points). Furthermore, we combined the endoscopic and histological evaluations and evaluated them together with the patients' outcomes.

Figure 1.

Three different endoscopic evaluations for the analysis. (A) Original UCEIS (Ulcerative Colitis Endoscopic Index of Severity); (B) Worst affected UCEIS; (C) Pancolonic UCEIS

All patients regularly visited the clinic every 1–2 months and were prospectively followed up for 1 year after colonoscopy. In this study, colonoscopies were not followed up for 1 year because doing such had no clinical benefit for those in remission. At every visit, physicians evaluated the partial Mayo score of each patient. A clinical relapse (combination of partial Mayo score ≥3 and C-reactive protein [CRP]/calprotectin positive) within 1 year served as the primary endpoint.[5] CRP >3 mg/L and calprotectin >250 μg/g were described as positive. We also evaluated disease-related complications such as therapy escalation, UC-related hospitalisation, and colectomy, and the relationships of the endoscopic and histological evaluations with the endpoints. Moreover, the colonoscopic strategy with the highest predictive value for patient prognosis was assessed.

Statistical Analysis

Our sample size was calculated with the assumption of relapse-free rates of 90% in the mucosal healing group and 80% in the active mucosal group,[17] using 80% power and 5% significance. Consequently, our target sample size was 278 patients. Between-group differences were assessed by Mann–Whitney U test for continuous variables and chi-square test for categorical variables. The agreement for UCEIS and Geboes scoring was calculated using an intraclass correlation coefficient (ICC). The relationship between the outcomes and mucosal findings was examined using receiver operating characteristic (ROC) curves and area under the curve (AUC). The cutoff value of each score was determined according to the highest sum of sensitivity and specificity. Moreover, cumulative endpoint-free rates were calculated by the Kaplan–Meier method and log-rank test. For single and multiple regression analyses, we used Cox's proportional hazards model, and the risks for endpoints were calculated with the hazard ratio (HR). We used the factors, which were statistical significance on single regression analysis, to multiple regression analyses. Contingency tables were also prepared to determine the accuracy of mucosal findings for predicting patient outcomes. Furthermore, P < 0.05 was considered statistically significant. All statistical data were analysed using the SPSS version 21.0 (IBM) and R version 3.3.0 (R Foundation for Statistical Computing). The article data will be shared at a reasonable request to the corresponding author.