Sigmoidoscopy is sufficient to assess endoscopic response to therapy in pediatric ulcerative colitis: A cohort study

One of the major treatment targets after initiating a new maintenance drug in ulcerative colitis (UC) is endoscopic healing.1 Sigmoidoscopy is a less invasive alternative to full colonoscopy, often using enema-based preparation and requiring shorter sedation. Despite these advantages and the typical UC pattern of maximal distal inflammation, full colonoscopy is being required by the US Food and Drug Administration (FDA) for assessing treatment response in licensing trials in pediatric UC (https://www.regulations.gov/document/FDA-2024-D-2682-0002). Several studies in adult patients have demonstrated a strong correlation between the degree of inflammation observed during sigmoidoscopy compared with colonoscopy.2-4 In contrast, one retrospective study reported that 40% of 154 patients with endoscopic healing in the rectosigmoid had inflamed mucosa in more proximal segments.5 There are only two similar studies in the pediatric population, one prospective with 50 patients and the other retrospective with 57 patients.6, 7 Both showed high agreement between endoscopic remission in the rectosigmoid region and the rest of the colon. However, neither study included this analysis as their primary aim, and no pediatric study to date has addressed whether sigmoidoscopy can assess the degree of colonic inflammation severity and not merely remission. We aimed to explore the rate of rectosigmoid healing in children who had active colitis in the more proximal colonic segments. We also examined whether rectosigmoid inflammation reflects maximal colonic severity. Ethical approval was obtained from the SZMC ethics committee, and all participants provided written informed consent at enrollment. This study utilized data from two prospectively enrolled cohorts. The first was an inception cohort enrolled as part of a biobank project at the Juliet Keidan Institute of Pediatric Gastroenterology, Shaare Zedek Medical Center between May 2015 and January 2024. The second was a prospective validation cohort of the TUMMY-UC patient-reported outcome score, with participants enrolled in Canada, the United States, England, and Israel between November 2020 and July 2021; details of the TUMMY-UC cohort have been previously published.8 Both cohorts enrolled children younger than 18 years with a confirmed diagnosis of UC or inflammatory bowel disease-unclassified (IBD-U), regardless of disease duration. In this study, we included only the UC patients, as IBD-U patients are typically excluded from regulatory trials. All children underwent full colonoscopy, during which the gastroenterologist prospectively completed the ulcerative colitis endoscopic index of severity (UCEIS) at each colonic segment.9 Endoscopic severity in individual colonic segments was defined by the UCEIS score for that segment: 0–1 for remission, 2–4 for mild disease, 5–6 for moderate disease, and 7–8 for severe disease.1 As the goal of this study focused on the accuracy of sigmoidoscopy in assessing treatment response, we excluded the initial diagnostic procedures, for which full colonoscopy is anyway indicated. Procedures in children with known primary sclerosing cholangitis (PSC) were excluded due to the distinct characteristics of this population and the documented tendency for right-sided colitis.10 Data collection was consistent across the cohorts including comprehensive clinical and demographic information, Pediatric UC activity index (PUCAI), and laboratory test results (e.g., erythrocyte sedimentation rate [ESR], C-reactive protein [CRP], albumin, white blood cells count [WBC], and hematocrit). Variable distribution was assessed using histograms and the Shapiro–Wilk test. Data were compared using unpaired Student's t-test, Wilcoxon rank-sum test, or the Kruskal–Wallis test, as appropriate. Correlations utilized the non-parametric Spearman's test. Agreement between maximal severity categories was calculated using Cohen's weighted kappa (κ) coefficient with scores of 0–0.19 regarded as absent, 0.20–0.39 as weak, 0.40–0.59 as moderate, 0.60–0.79 as strong, and 0.80–1.00 as excellent.11 Statistical analyses were conducted using R software, and a p-value of <0.05 was considered significant. Among 139 children with colonoscopic evaluations, 54 (38%) had their procedure performed only at the time of the initial diagnosis and one had PSC, resulting in 84 patients with posttreatment follow-up colonoscopy included in the study. Endoscopic remission throughout the colon was apparent in 21 (25%) children, while 34 (40%) had mild disease, and 29 (35%) had moderate–severe disease in at least one segment. (Table 1). In 13 children (15%), the entire colon was inflamed, 12 (14%) had disease limited to the rectosigmoid, and 12 (14%) had proctitis. In accordance with the typical nature of UC in which inflammation is most severe in the distal colon, the median UCEIS scores decreased progressively from the rectosigmoid region to the cecum (rectum 2.5 interquartile range [IQR]: 0.75–4, sigmoid 2 [IQR: 0–4], descending colon 1 [IQR: 0–3], and ascending and transverse colon 0 [IQR: 0–2]; p < 0.001). Of 23 children (27%) with endoscopic rectosigmoid healing, 21 (91.3%) had pancolonic healing; only two exhibited inflammation in more proximal segments (8.7% of the 23 with rectosigmoid remission, 2.4% of the entire cohort). In both cases, PUCAI, CRP, albumin, and hematocrit were normal. Sigmoidoscopy was highly accurate for assessing endoscopic remission, with a sensitivity of 96.8% and a negative predictive value (NPV) of 91%. Specificity and positive predictive value (PPV) were 100% as expected, since the presence of rectosigmoid inflammation by definition excludes pancolonic endoscopic healing and thus no false positive cases were present. There were eight (9.5%) patients with rectal sparing and inflammation in proximal colonic segments, of whom six (75%) had evidence of inflammation in the sigmoid (the other two were the ones described above). None of the patients with rectal or rectosigmoid sparing had documented use of topical rectal therapy. The correlation between the maximal UCEIS scores measured anywhere in the colon and in the rectosigmoid was excellent (r = 0.96, p < 0.001). The agreement between the maximal UCEIS 0–8 categories measured in the rectosigmoid and the entire colon was also excellent, with a kappa of 0.90 (95% confidence interval [CI]: 0.84–0.96; Figure 1). When categorizing inflammation by severity (remission, mild, moderate, or severe), sigmoidoscopy reflected the maximal degree of inflammation throughout the colon in 89% of all colonoscopies. In only seven patients (8% of all patients and 11% of those with any inflammation in the colon), there was worse inflammation in the proximal colon compared with the rectosigmoid. We found that in children with treated UC, the presence/absence of inflammation in the rectosigmoid region reflected the status of inflammation in the entire colon in 97.6% of the cohort. When focusing only the 23 children with rectosigmoid healing, only two (8.7%) had inflammation in more proximal segments. This suggests that despite current regulatory preferences, sigmoidoscopy is a valid assessment tool to evaluate treatment response in the research setting. Moreover, we also found that sigmoidoscopy not only reflected the presence of inflammation, but also its maximal severity in 89% of cases. It is noteworthy that proctoscopy alone may not be sufficient, as 75% of children with complete or relative rectal sparing did have inflammation in the sigmoid colon. Despite the disease tending to be more extensive in children,12 our findings align with most adult studies showing that sigmoidoscopy is a highly accurate monitoring tool in UC. In these studies, sigmoidoscopy alone reflected the presence of inflammation and its severity throughout the colon in 95%–100% and 91%–94% of patients, respectively.2-4 Moreover, strong agreement and high correlation have been observed between sigmoidoscopy and colonoscopy in assessing both endoscopic activity (κ = 0.92–0.97, r = 0.74–0.97) and histologic activity (κ = 0.67–1.0, r = 0.87–1.0).2 In the two aforementioned pediatric studies, endoscopic healing in the rectosigmoid predicted pan-colonic mucosal healing, with specificity of 83%–98%, sensitivity of 100%, PPV of 72%, and NPV 100%.6, 7 In one of the studies, rectosigmoid endoscopic scores also reflected histologic remission with specificity of 88% and sensitivity of 98%.7 The other found that adding a criterion of fecal calprotectin <250 to rectosigmoid healing improved the specificity to predict pancolonic healing to 98% and PPV to 93%, while reducing sensitivity from 100% to 78% and NPV from 100% to 91%.6 Ours is the largest pediatric study to date and the first to explore not only endoscopic healing but also degree of inflammation. A limitation of our study is that although the data were collected prospectively and the primary aim was to assess the role of sigmoidoscopy versus colonoscopy, patients were not enrolled specifically for this purpose. Additionally, some additional potentially relevant data, most notably calprotectin, were not systematically collected. Furthermore, the fact that clinicians scored the UCEIS score for each segment at the end of the procedure may have influenced their assessment of the rectosigmoid. The ability to reflect both the presence and severity of colonic inflammation from the rectosigmoid region suggests that sigmoidoscopy may serve as a sufficient and practical alternative for assessing treatment response in pediatric UC. This is especially relevant in the research setting where robustness and feasibility should be balanced and since ethical criteria mandate minimizing invasive testing in pediatric research unless absolutely necessary. While suitable for children with UC, in the routine clinical setup, however, the choice of sigmoidoscopy or full colonoscopy should be individualized, based on previous colonoscopic findings, disease course, symptoms, tolerability of the procedure and the presence of PSC or IBD-U. We thank Raffi Lev-Tzion, Oren Ledder, Eyal Shteyer, Amit Assa, Robert N. Baldassano, and Anne M. Griffiths for their valuable contributions to this study. This project has received no external funding or sponsorship and was funded from the internal resources of the Juliet Keidan Institute of Pediatric Gastroenterology Hepatology and Nutrition, Shaare Zedek Medical Center, The Hebrew University of Jerusalem, Jerusalem, Israel. Dan Turner: For the past 3 years, has received consultation fees, research grants, royalties, or honorarium from Janssen, Pfizer, Hospital for Sick Children (Canada), Ferring, AbbVie, Takeda, Atlantic Health System, Shire, Celgene, Lilly, Roche, ThermoFisher, BMS. Esther Orlanski-Meyer: During the past 3 years has received lecture fees from Abbott. The remaining authors declare no conflicts of interest. The data underlying this article will be shared on reasonable request to the corresponding author. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.