Introduction

In Japan, the Ministry of Health, Labor, and Welfare (MHLW) registered all patients with ulcerative colitis (UC) from 2003 to 2014, and real-world data (RWD) are available in a national database. Deep learning models have been used to analyse RWD, but visualising the relationship between outcomes and explanatory variables is difficult.

Aims & Methods

Therefore, in this study, we used explainable artificial intelligence (AI) to calculate the regression coefficient for each parameter to determine the explanatory variables involved in steroid-resistant UC and constructed a model to predict steroid-resistant UC at the time of a patient's first visit. We obtained data on all patients with UC newly registered in the MHLW database between 2003 and 2011 (before the spread of biologics in Japan). Of the 79,096 patients identified, 4,003 who had three years of data, a Mayo score of 3 or higher at the first visit, and who had been using steroids from the time of the first visit were included in the analysis. For explainable AI, a point-wise linear model (PWL) was used to generate a custom-made logistic regression model for each case. Clustering analysis was performed based on the similarity of the coefficients (weights) of these logistic regression models. Finally, the association with outcomes was evaluated by analysing the weights of factors with a significant difference (p<0.01) between the clusters.

Results

The area under the curve (AUC) of the PWL model was 0.628 for the training set and 0.627 for the test set. The likelihood ratio was 1.36. By clustering, the analysed population was classified into two groups (S group, predicted to be in remission after three years: 2,040 cases; R group, predicted not to be in remission: 1,963 cases). In both groups, female sex (R group: median weight 0.0027), positive stool culture (R group: median weight 0.0041), and colonic mucosal abnormalities (R group: median weight 0.0113) had a strong relationship with steroid resistance. The R group tended to have significantly higher frequencies than the S group. In the R group, pseudopolyposis (median weight = 0.0351), caecal lesions (median weight = 0.024), CRP (median weight = 0.0163), ascending colon lesions (median weight = 0.0133), and transverse colon lesions (median weight = 0.0107) were observed. Conversely, the white blood cell count (S -0.0021 vs R -0.0037), frequency of defecation (S -0.0053 vs R -0.0055), bloody stool (S -0.0021 vs R -0.0016), stool form (S -0.0347 vs R -0.0367), and abdominal pain (S -0.0291 vs R -0.0383) were not significantly related to steroid resistance in both groups.

Conclusion

Explainable AI in a national database with three years of records allowed us to build a predictive model for refractory UC. Female sex, positive faecal bacterial cultures, colonic mucosal abnormalities, and lesions extending into the right colon were important predictors of steroid resistance. Understanding the predictors of steroid resistance and dependence at the time of steroid introduction, based on national data before the spread of biologics, can facilitate prompt and timely treatment plans for patients with UC.

References

Kanatani et al. Biology 11:951, 2022.

Disclosure

The authors, Hidekazu Suzuki, received the reserach grant from Biofermin Pharm. Co., and the lecture fee from Astellas Co., Otsuka Pharm. Co., EA Pharma Co., and Takeda Pharm. Inc.

Introduction

Cognitive Behavioral Therapy (CBT) is an evidence-based treatment for Irritable Bowel Syndrome (IBS), demonstrating long-term improvements in symptoms and quality of life (QOL). However, its implementation is hindered by the complexity of treatment protocols, high time and financial costs, and limited accessibility in Japan.

Aims & Methods

This study aimed to develop a smartphone-based CBT application and evaluate its effectiveness in alleviating IBS symptoms.
Adults aged 18 or older who met the Rome IV criteria for IBS and scored ≥110 on the Japanese version of the IBS Symptom Severity Index (IBSSI-J) were enrolled. Twenty-six participants were randomly assigned to either the intervention group (n = 14) or the waiting-list control group (n = 12). Outcome measures included the IBSSI-J, Japanese versions of the IBS-QOL (IBS-QOL-J), Pain Catastrophizing Scale (PCS-J), and Visceral Sensitivity Index (VSI-J). A two-way ANOVA was conducted with group (intervention vs. control) and time (pre, mid, post, follow-up 1) as independent variables.
The intervention group used the app for 5–10 minutes daily over two months. The app included three components: psychoeducation, cognitive restructuring, and exposure. Psychoeducation consisted of educational materials and follow-up quizzes. Cognitive restructuring used gamified multiple-choice content (135 items), replacing conventional thought records. Exposure involved users creating individual anxiety hierarchies and recording their thoughts before and after exercises. The app visualized anxiety changes and progress via graphs.

Table 1.
Intervention module of the application

Results

Eleven participants in the intervention group and eight in the control group completed the program. Significant group × time interactions were observed for all outcome measures. For the IBSSI-J, the interaction was significant (F(3, 51) = 4.41, p < .001), with the intervention group showing greater improvement at mid-point (F(1, 17) = 5.54, p = .03, η² = .25). For the IBS-QOL-J, a significant interaction was found (F(3, 51) = 8.00, p < .001), with improvement evident at follow-up 1 (F(1, 17) = 6.65, p = .02, η² = .28). For the PCS-J, the interaction was significant (F(3, 51) = 3.83, p = .02), with improvements at post (F(1, 17) = 9.26, p = .01, η² = .35) and follow-up 1 (F(1, 17) = 5.83, p = .03, η² = .26). For the VSI-J, the interaction also reached significance (F(3, 51) = 3.32, p = .05).

Conclusion

The CBT-based app was effective in reducing IBS symptom severity. Its effects persisted after the intervention and extended to improvements in QOL, visceral anxiety, and pain catastrophizing—psychological factors associated with the maintenance and exacerbation of IBS. It is assumed that the cognitive restructuring component of the app facilitated cognitive diversification among participants. By empirically challenging the plausibility of these diversified cognitions through behavioral experiments in the form of exposure tasks, participants were able to further develop more adaptive thoughts. This process is considered to have contributed to the improvement of symptoms and enhancement of QOL. This digital approach may offer a viable solution to current barriers in CBT access.

Introduction

Budesonide orodispersible tablet (BOT) is the only approved topical steroid available in Europe for the induction and maintenance of remission in patients with eosinophilic esophagitis (EoE). Available randomized controlled trials have demonstrated that, following induction of histological remission with BOT 1mg twice daily (bid), patients can be maintained in remission with BOT 0.5mg bid. It is currently unknown whether a dose of 1mg once daily (od) has comparable efficacy to 0.5mg bid.

Aims & Methods

In this prospective study conducted at two tertiary referral centers for EoE, we enrolled consecutive patients with EoE undergoing clinical, endoscopic, and histologic re-assessment while on maintenance treatment for EoE with BOT 0.5mg bid or BOT 1mg od. All included patients had achieved histological remission (i.e., <15 eosinophils/high-power field, eos/HPF) following induction of remission with BOT 1mg bid. At the end of the induction of remission period (T0), all patients were assessed using the dysphagia symptom questionnaire (DSQ), the EoE endoscopic reference score (EREFS), and peak eosinophil counts (PEC). At re-assessment while on maintenance treatment (T1), all patients underwent 1) assessment of dysphagia and adaptive behaviors with the DSQ and the Pisa EoE Adaptation Questionnaire v1.0 (PiEAQ). Clinical remission was defined as DSQ not increasing >30% compared to T0. 2) upper endoscopy (EGDS) with assessment of EREFS score and at least six esophageal biopsies, and 3) histological assessment using the EoE Histology Scoring System (EoEHSS). In addition, a comprehensive global assessment of clinical, endoscopic, and histologic findings was performed using the validated index of severity for EoE (I-SEE score). Kruskal-Wallis Rank Sum Test and Pearson’s Chi-squared were used for comparisons. Significance threshold was p<0.05.

Results

Fifty-six patients were included (46 males and 10 females). Following histological remission with BOT 1mg bid, 28 (50%) initiated maintenance treatment with BOT 1mg od and 28 (50%) with BOT 0.5 bid. At the end of induction of remission (T0), the two groups had achieved similar outcomes, with comparable median DSQ [0 (0-28) vs 0 (0-28)], EREFS [1 (0-2) vs 1 (0-1)], and PEC [1 (0-7) vs 1.5 (0-4)] (all p>0.3) (Table 1). At the end of the maintenance treatment (T1), the two groups had been taking maintenance treatment for a comparable duration [20 weeks (12 – 36) vs 20 weeks (12 – 32; p=0.7). Persistent histological and clinical remission at T1 was present in 86% of BOT 1mg od versus 93% of BOT 0.5 bid (p=0.7). In addition, EREFS scores were comparable between groups at T1 [1 (0-2) vs 1 (0-2), p=0.8]. Similarly, the EoEHSS stage and grade was comparable between groups, as was the PiEAQ score [1.5 (0.75-2) vs 0.5 (0-2); p=0.06], and the I-SEE score [ 3 (2-4) vs 3 (2-4); p=0.5] (Table 2).

Conclusion

Persistent clinical endoscopic, and histological remission was maintained in most patients treated either with BOT 1mg od or BOT 0.5mg bid for a median of 20 weeks, indicating that the total daily dose of BOT 1mg may be more important that the frequency of administration, at least in the short term.

Disclosure

Speaker or Consultancy for Apogee Therapeutics, Alfasigma, Dr Falk Pharma, Sanofi, Malesci, JB Pharmaceuticals

Introduction

The diagnosis of Coeliac Disease (CD) in adults still continues to rely heavily on duodenal histology, in spite of several studies demonstrating a positive predictive value (PPV) of 100% for Marsh 3 histology when anti-tissue transglutaminase (TTG) exceeds 10x upper limit of normal. The majority of these studies come from tertiary centres and the slow uptake of the non biopsy approach may be as a result of limited data on the histological outcome of positive serology originating in the primary care setting.

Aims & Methods

To determine the PPV of primary care-based TTG testing with concurrent anti-endomysial antibody (EMA) positivity for histological diagnosis of CD.
The immunology lab in our institution performs reflex EMA testing on all newly positive and newly borderline positive TTG samples from the West of Ireland (TTG≥5U/ml Celikey by Phadia). Over a 6 year period (2018-2023), 1,361 index TTG positive samples also proved positive when tested for EMA. Following electronic medical record review, we excluded all samples from patients who were <16 years old, had a previous diagnosis of CD made elsewhere, lacked biopsy data, or had biopsies that predated the serological test. Patients who had their coeliac serology testing requested by any hospital-based physician were also excluded to ensure that only patients identified by primary care-based screening were included in the outcome analysis. In this manner, 335 adult patients who had serology screening performed by their primary care physician were eligible for further analysis. These eligible patients were then stratified by TTG levels ( ≤20, 21-69, and ≥70 U/ml). PPV was calculated for each stratum. Logistic regression assessed the relationship between TTG values and biopsy results. Additional analyses examined influence of interval between serology and biopsy.

Results

Chi-square analysis confirmed significant association between TTG levels and biopsy results (p<0.00001). For each 10 U/ml increase in TTG, odds of positive biopsy increased by 18% (OR 1.18, 95% CI: 1.10-1.26). For TTG 21-69 U/ml, PPV declined with longer serology-to-biopsy intervals: 100% for ≤47 days versus 45.2% for >203 days. High TTG values maintained excellent PPV regardless of interval. Long-term follow-up of initially biopsy-negative patients revealed subsequent CD diagnosis in 19/40 (47.5%) in the low titre group, 11/30 (36.7%) in the intermediate group, and 7/11 (63.6%) in the high titre group.

Conclusion

TTG titre significantly predicts biopsy-confirmed CD, with TTG ≥70 U/ml having 90.4% PPV. The predictive value of intermediate TTG diminishes with increasing interval between serology and biopsy. A substantial portion of initially biopsy-negative patients across all TTG strata were subsequently diagnosed with CD during follow-up. These findings support TTG in risk-stratifying CD patients, suggest prompt biopsies for intermediate TTG values, and highlight the importance of long-term monitoring even in patients with negative initial biopsies.

References

1. Husby S, et al. European Society for Pediatric Gastroenterology, Hepatology, and Nutrition guidelines for the diagnosis of coeliac disease. J Pediatr Gastroenterol Nutr. 2012;54(1):136-60.
2. Penny HA, et al. Accuracy of a no-biopsy approach for the diagnosis of coeliac disease across different adult cohorts. Gut. 2021;70(5):876-883.

Clinical Case Summary

Introduction: Infliximab, an anti-TNF agent, has revolutionized the management of inflammatory bowel disease (IBD), promoting long-term remission and improving quality of life. However, paradoxical immune-mediated reactions, including autoimmune dermatologic conditions, have been increasingly recognized.
Aims & Methods: We report a rare case of paradoxical development of giant pyoderma gangrenosum (PG) in a patient with Crohn’s disease during Infliximab maintenance therapy.
Case Report: A 26-year-old female with Crohn’s disease (A2, L3, B1), in clinical remission under Infliximab, developed a pustular erythematous nodule (~5 cm) on the right popliteal region after the third maintenance dose. The lesion rapidly expanded to a painful ulcerative plaque with inflammatory borders, reaching 15 cm in diameter within 8 days. Histopathology revealed dense neutrophilic dermatosis, confirming pyoderma gangrenosum.
Results: This case exemplifies a paradoxical reaction induced by anti-TNF therapy. Although PG is more commonly associated with ulcerative colitis, it may occur in Crohn’s disease and be exacerbated by immunomodulation. The lesion was managed with corticosteroids and discontinuation of Infliximab, with partial improvement.
Conclusion: Paradoxical PG during anti-TNF therapy is rare but significant. Clinicians should be alert to unusual dermatologic manifestations in IBD patients on biologics. Early recognition and multidisciplinary management are essential to mitigate morbidity.

References

1) Paradoxical pyoderma gangrenosum induced by infliximab in a patient with Crohn’s disease.
• Autores: Llamas-Velasco M, García-Diez A, Daudén E.J Crohns Colitis. 2013;7(10):e479-81.
2) Paradoxical skin reactions to anti-TNFα agents: a study of 118 cases.
• Autores: Ramos-Casals M, Brito-Zerón P, Muñoz S, Soto MJ.Br J Dermatol. 2010;163(5):1159-65.
3) Paradoxical psoriasis and pyoderma gangrenosum induced by anti-TNF therapy in inflammatory bowel disease: a case series.
• Autores: Navarini AA, Trüeb RM, Imhof L, French LE.J Crohns Colitis. 2011;5(5):457-62.

Introduction

Gastric peroral endoscopic myotomy (G-POEM) is an emerging endoscopic technique for the treatment of refractory gastroparesis, particularly benefiting patients with pylorospasm. It is considered a safe procedure with a low risk of adverse events when performed by experienced endoscopists. Our aim was to perform a comprehensive analysis of all adverse events associated with G-POEM.

Aims & Methods

We performed a retrospective analysis of prospectively collected data from all patients who underwent G-POEM for gastroparesis at our centers (Prague, Brno) between November 2017 and April 2025. All procedures were performed by a single endoscopist (JM). The primary outcome was procedure-related adverse events rate and severity occurring during the perioperative and early postoperative period. Adverse events were categorized as follows: deviations from the expected peri and postoperative course were defined as "undesirable events." Undesirable events that met the criteria of the Clavien Dindo surgical classification (C-D) were considered "true adverse events".

Results

Out of 70 procedures, no true adverse or undesirable events occurred in 51 procedures (72.9%). Perioperatively, the following undesirable events occurred in 18 patients: : gastric (4; 5.7%) and duodenal mucosal injuries (2; 2.8%), major bleeding during the procedure treated endoscopically (1; 1.4%), serosal perforation (10; 14.3%), capnoperitoneum requiring puncture (3; 4.3%), and subcutaneous emphysema (1; 1.4%). In the early postoperative period, adverse events occurred in 3 patients: one patient complained of significant abdominal pain necessitating analgetics, subsequently, a small leak from the pyloric region to the subhepatic region was described on CT and gastroscopy confirmed ulceration at the tunnel site. The patient responded well to treatment with intravenous proton pump inhibitors. In two other cases, both involving patients with diabetic gastroparesis, decompensation of diabetes occurred during the hospital stay post-procedure. The following “undesirable” events were considered true adverse events according to the C-D classification: serosal perforation (10; C-D I), leak requiring prolonged hospitalization (1; C-D IIIa) and diabetes decompensation (2; C-D II). These complications occurred in 13 procedures (18.6%).
The mean duration of hospitalization after the procedure was 2 days (range 1-7). Three months after the procedure, 1 patient developed mild dumping syndrome, which was successfully treated conservatively.

Conclusion

G-POEM is a safe procedure. The rate of true adverse events according to Clavien-Dindo classification is low and severe AEs occurs rarely.