Deep Learning Algorithm for Diagnose Endoscopic and Histological Images With Ulcerative Colitis

The goal of treatment for ulcerative colitis is to achieve histological and endoscopic remission. Aiming at the problem that the observer will be affected by subjective factors in the endoscopic evaluation of ulcerative colitis and the cumbersome diagnosis process of histological images, this paper aims to develop a computer-assisted diagnosis system for real-time, objective diagnosis of endoscopic images and use the trained CNN model to predict histological images of patients with ulcerative colitis. Diagnosing endoscopic remission of ulcerative colitis, the accuracy of the CNN is 97.04% (95% CI,96.26%:97.62%). Diagnosing the severity of endoscopic inflammation in patients with ulcerative colitis, the accuracy of the CNN is 90.15% (95% CI, 89.49%:90.82%). The accuracy of predicting histological remission was 91.28%. The kappa coefficient between the CNN model and the biopsy results was 82.56%. The proposed computer-aided diagnosis system can effectively evaluate the inflammation of endoscopic images of patients with ulcerative colitis and predict the remission of histological images with high accuracy and consistency.

Histologic Evaluation Using the Robarts Histopathology Index in Patients With Ulcerative Colitis in Deep Remission and the Association of Histologic Remission With Risk of Relapse

Background This study prospectively evaluated the risk of relapse according to the status of histologic activity in patients with ulcerative colitis (UC) who achieved deep remission. Methods Patients with UC in clinical remission (partial Mayo score ≤1) and endoscopic remission (ulcerative colitis endoscopic index of severity ≤1) were enrolled. Rectal biopsies were performed in patients, and histologic remission was defined as a Robarts histopathology index of ≤3. Receiver-operating characteristic curve analysis was conducted to determine fecal calprotectin cutoff values for histologic remission. The cumulative risk of relapse was evaluated using the Cox proportional hazards model. Results Among the 187 patients enrolled, 82 (43.9%) achieved histologic remission. The best cutoff value of fecal calprotectin for predicting histologic remission was 80 mg/kg (area under the curve of 0.646, sensitivity of 74%, and specificity of 61%). Among 142 patients who were followed up for >3 months, 56 (39.4%) showed clinical relapse during a median of 42 weeks. The risk of relapse was lower in patients with histologic remission than in those with histologic activity (P = .026). In multivariable analysis, histologic remission (hazard ratio [HR], 0.551; 95% confidence interval [CI], 0.316-0.958; P = .035), elevated C-reactive protein levels (HR, 3.652; 95% CI, 1.400-9.526; P = .008), and history of steroid use (HR, 2.398; 95% CI, 1.196-4.808; P = .014) were significantly associated with clinical relapse. Conclusions In patients with UC who achieved clinical and endoscopic remission, histologic remission was independently associated with a lower risk of clinical relapse.

Recommendations for Standardizing Clinical Trial Design and Endoscopic Assessment in Postoperative Crohn’s Disease

Background The lack of standardized methods for clinical trial design and disease activity assessment has contributed to an absence of approved medical therapies for the prevention of postoperative Crohn’s disease (CD). We developed recommendations for regulatory trial design for this indication and for endoscopic assessment of postoperative CD activity. Methods An international panel of 19 gastroenterologists was assembled. Modified Research and Development/University of California Los Angeles methodology was used to rate the appropriateness of 196 statements using a 9-point Likert scale in 2 rounds of voting. Results were reviewed and discussed between rounds. Results Inclusion of patients with a history of completely resected ileocolonic CD in regulatory clinical trials for the prevention of postoperative recurrence was appropriate. Given the absence of approved medical therapies, a placebo-controlled design with a primary end point of endoscopic remission at 52 weeks was appropriate for drug development for this indication; however, there was uncertainty regarding the appropriateness of a coprimary end point of symptomatic and endoscopic remission and the use of currently available patient-reported outcome measures. The modified Rutgeerts Score, endoscopic assessment of the anastomosis, and a minimum of 5cm of neoterminal ileum were also appropriate; although the appropriateness of other indices including the Simple Endoscopic Score for CD for endoscopic assessment of postoperative CD activity was uncertain. Conclusions A framework for regulatory trial design for the prevention of postoperative CD recurrence and endoscopic assessment of disease activity has been developed. Research to empirically validate end points for these trials is needed.

Translating Results from VARSITY to Real World: Adalimumab vs Vedolizumab as First-line Biological in Moderate to Severe IBD

Background Selecting a first-line therapy remains challenging in IBD. Adalimumab (ADM) and vedolizumab (VDZ) effectively lead to endoscopic remission in moderate to severe IBD. The VARSITY trial showed superiority of VDZ over ADM in ulcerative colitis (UC) regarding clinical remission and endoscopic improvement at week 52. We explored these results in a real-world setting of UC and Crohn’s disease (CD). Methods This retrospective study followed a cohort of biologic-naïve patients starting ADM or VDZ from 2015-2019. Patients had moderate to severe disease (endoscopic Mayo score ≥2 for UC, presence of ulcerations for CD) prior to therapy initiation. For UC, endoscopic remission (endoscopic Mayo score 0) and improvement (endoscopic Mayo score ≤1) at week 52 were assessed. For CD, endoscopic remission (absence of ulcerations) and improvement (markedly better endoscopy despite ulcerations) at weeks 26-52 were studied. Treatment persistence was also evaluated. Results In total, 195 biologic-naïve patients (109 UC; 86 CD) were included. In UC, VDZ was superior to ADM regarding endoscopic remission (29% vs 11%, P = .03), endoscopic improvement (51% vs 26%, P = .01) at week 52, and treatment persistence (P = .04). In CD, no differences in endoscopic remission (VDZ 48% vs ADM 60%; P=0.37), improvement (VDZ 76% vs ADM 77%; P = 1.00), or treatment persistence (P=0.43) at weeks 26-52 were seen. Safety profiles were similar in UC and CD. Conclusions This real-world cohort study of biologic-naïve IBD patients found VDZ to be superior to ADM as first-line treatment for patients with UC—but not CD—regarding endoscopic remission at week 52 and treatment persistence.

Effectiveness of treatment of moderate ulcerative colitis with prolonged mesalazine in real clinical practice

Introduction. Ulcerative colitis (UC) is one of the severe therapeutic diseases. High doses of oral granular mesalazine are required to maintain clinical and endoscopic remission of UC, which may be sufficient and supposedly more acceptable for patients, as some studies showed that adherence to topical therapy is significantly lower than to oral 5-ASA drugs.Objective of the study. To evaluate the efficacy of therapy of patients with moderate left-sided ulcerative colitis (UC) and pancolitis receiving prolonged-release ethylcellulose-coated mesalazine.Materials and methods. The evaluation of the outcomes of treatment of UC patients who received prolonged-release mesalazine was carried out. We examined 87 patients with UC who received granular ethylcellulose-coated mesalazine, of those 38 (43.7%) men and 49 (56.3%) women. The average age of the enrolled patients was 38.3 ± 12.6 years.Results and discussion. After 2 weeks from the beginning of therapy with prolonged-release mesalazine, the majority of patients – 71 (81.6%) responded to the therapy. After 12 weeks, 71 (81.6%) of 87 UC patients, who responded to therapy with prolongedrelease mesalazine, remained in clinical remission. On average, the Mayo score in the group decreased from 7.6 ± 0.99 to 2.6 ± 0.25 points. There was a significant decrease in CRP, ESR, leukocytosis, and fecal calprotectin. After 26 weeks, Mayo score in the group of patients remained on average at the level of 2.2–2.3 points. The number of UC patients with colon mucosal healing was 32 (36.8%) patients. A year after the start of therapy with prolonged-release mesalazine, 69 (79.3%) UC patients who responded to therapy had a clinical remission, of those 32 (36.8%) patients had a clinical and endoscopic remission. During the year of observation, no case of surgical intervention or re-hospitalization due to exacerbation of the disease was recorded in patients with UC who achieved remission.Conclusions. Treatment of moderate active UC should begin with oral mesalazine ≥ 3 g per day in combination with topical mesalazine. The prolonged-release mesalazines are the most preferred

Ustekinumab trough concentration affects clinical and endoscopic outcomes in patients with refractory Crohn’s disease: a Chinese real-world study

Background Ustekinumab (UST), a newly-used biologic targeting p40 subunit of IL12 and IL23 in China, exerts a confirmed therapeutic effect on the induction and maintenance therapies for refractory Crohn’s disease (CD). Therapeutic drug monitoring based on trough and antibody concentration is of core importance when treating patients who lose response to UST. We aimed to analyze the UST exposure–response relationship in CD treatment in the real-world setting. Methods We retrospectively enrolled patients with CD who received UST between March 1, 2020 and May 31, 2021, at the inflammatory bowel disease (IBD) center of the Sun Yat-Sun Affiliated Sixth Hospital. Baseline characteristic information, biomarker examination, clinical outcomes determined by the Crohn’s disease activity index (CDAI), and endoscopic outcomes evaluated using a simple endoscopic score for Crohn’s disease (SES-CD) at week 16/20 were collected. The optimal UST cut-off trough concentration was identified using receiver operating characteristic curve (ROC) analysis. Results Nineteen eligible patients were included in the study, the mean age was 29.1 ± 9.1 years and the mean disease duration was 5.5 ± 4.7 years. At the initiation of the study, 89.5% of the patients had been exposed to prior biologics, 42.1% had previous CD-related surgeries, and 52.6% had perianal diseases. At week 16/20 after the UST initiation, clinical response, clinical remission, endoscopic response, and endoscopic remission were 89.5%, 84.2%, 42.2%, and 73.7%, respectively. The cut-off optimal trough concentration for UST was 1.12 μg/mL, as determined by the ROC with an area under the curve (AUC) of 0.78, sensitivity of 87.5%, and specificity of 72.7%. Patients with a UST trough concentration > 1.12 μg/mL had a significantly higher rate of endoscopic remission than those without (70.0% vs. 11.1%, P = 0.02). Conclusions UST is an effective therapeutic option for refractory CD treatment. A UST trough concentration above 1.12 μg/mL was associated with endoscopic remission at week 16/20 after UST initiation. Trial registration This study was approved and retrospectively registered by the Ethics Committee of Sun Yat-Sen University (2021ZSLYEC-066, March 29, 2021) and the Clinical Trial Registry (NCT04923100, June 10, 2021).