Abstract Background The therapeutic options for Inflammatory Bowel Disease (IBD) have expanded with the introduction of JAK inhibitors. However, real-world data on upadacitinib (UPA) in patients with moderate to severe Crohn’s disease (CD) are scarce. Our aim was to assess the long-term effectiveness and safety of UPA in multi-refractory Belgian patients. Methods Data from all patients with active CD initiating UPA between September 2022 and January 2024 were retrospectively collected from 17 Belgian centres. Effectiveness endpoints were evaluated at week 24 and 52. Steroid-free clinical response and remission were defined using the two-component patient-reported outcome. Endoscopic response was defined as a decrease in baseline Simple Endoscopic Score (SES-CD) with ≥50%, and endoscopic remission as SES- CD ≤ 4 or absence of ulcers. Adverse events (AE), treatment discontinuation, CD-related hospitalization and surgery were assessed throughout follow-up. Non-response imputation was applied for clinical evaluation. Results A total of 140 patients were included (Table 1) with a median follow-up of 33 weeks (IQR: 3-97). The majority of patients (89%) were exposed to at least 3 biologics. At week 24, 53% and 40% of patients achieved steroid-free clinical response and remission, respectively (Figure 1). At week 52, these numbers were 37% and 29%. Endoscopic data were available in 48/140 patients at week 24, of whom 52% and 27% reached endoscopic response and remission. At week 52, these numbers were 63% and 37%. Of patients with active articular extra-intestinal manifestations at baseline 37% and 34% had a quiescent articular disease by week 24 and 52, respectively. 5 out of 9 patients with active perianal disease at baseline and on UPA by week 52, had persistent perianal active disease. Overall, 66 patients (47%) discontinued UPA during follow-up: 23 due to primary non response, 13 due to secondary loss of response, 16 due to AE, 3 due to patient’s choice and 11 due to CD-related surgery. 16 patients required CD-related hospitalization within 52 weeks. AE occurred in 60% of patients (84/140). Of these patients, 11% experienced serious AE of which 2 cardiovascular events: 1 transit ischemic attack and 1 deep venous thrombosis. The most common reported AE were herpes simplex reactivation (5/140), respiratory tract infection (7/140), cutaneous reactions (7/140) and acne (17/140). Only 3 patients had reactivation of herpes zoster infection. Conclusion In this real-world cohort of highly refractory CD patients, UPA effectively induced both steroid-free clinical remission and endoscopic remission by week 52. UPA was relatively well tolerated with respect to adverse events.

Abstract Background Achieving deep remission, encompassing clinical, endoscopic, and biological remission, is a long-term goal in Crohn's disease [CD] according to the STRIDE-2 guidelines. The role of histological remission in CD remains unclear. We evaluated the efficacy of vedolizumab [VDZ] for inducing histo-endoscopic remission in early versus late CD in the prospective, open-label LOVE-CD trial conducted in Belgium, the Netherlands and Hungary. Methods In the LOVE-CD trial, patients with moderate-to-severe disease (Crohn’s Disease Activity Index 220-450 and presence of ulcers at baseline endoscopy) received intravenous vedolizumab over a 52-week period. Early CD was defined as a diagnosis <24 months (treatment-naive or history of corticosteroid and/or immunomodulator use), and late CD as a diagnosis >24 months with prior anti-TNF exposure. Ileocolonoscopies were performed at three timepoints (baseline, week 26, week 52) during which biopsies were systematically collected in the terminal ileum and colon. Both the endoscopies and biopsies were centrally scored by experienced readers for the simple endoscopic score for Crohn's disease [SES-CD], Robarts’ histopathology index [RHI], and Geboes score [GS]. Endoscopic remission, histo-endoscopic mucosal improvement [HEMI], and histo-endoscopic mucosal remission [HEMR] were defined as SES-CD <4, GS ≤3B.1, and GS <2B.1 respectively. The association between histologic and endoscopic scores was studied with Spearman’s correlation coefficient. Intention-to-treat analysis with non-responder imputation was used to handle missing data. Results Of the 260 patients included in LOVE-CD (Table), 440 biopsies (336 colon, 114 ileum) from 179 patients at week 26, and 435 biopsies (323 colon, 112 ileum) from 177 patients at week 52 were analysed. There was a strong correlation between the endoscopic and histological assessment (colon r 0.69, ileum r 0.64, total r 0.64, p<0.001). Patients with early CD were more likely to achieve histological remission than patients with late CD at week 52 (38.4 % vs. 14.9%, p=0.00004). Endoscopic and histological healing rates at week 26 (endoscopic remission 36.9%, HEMI 30.4%, HEMR 24.2%) and week 52 (endoscopic remission 35%, HEMI 30%, HEMR 22.7%) were comparable. When looking at disease location, histological remission rates in the colon and ileum did not significantly differ (35.2% vs. 27.8%, p=0.15). Conclusion Vedolizumab was more efficient at inducing histo-endoscopic remission in early CD as compared to late CD, with most of the histo-endoscopic healing occurring during the first 26 weeks. Histo-endoscopic healing rates under vedolizumab were comparable in colonic and ileal disease. Histological scores (RHI, GS) are also applicable for the evaluation of ileal and colonic CD.

Abstract Background Non-invasive imaging methods, particularly intestinal ultrasound (IUS), are increasingly pivotal in guiding management in ulcerative colitis (UC). While long-term treatment goals focus on reducing IBD-related disability, the link between ultrasonographic findings and patient-reported outcome measures (PROMs), such as IBD-related disability, remains unclear. Methods Data on IBD-related disability (using the IBD disk) were prospectively gathered through a digital questionnaire from patients with UC undergoing routine IUS, and performed within maximal one week of each other. Patients with isolated proctitis were excluded. Bowel wall thickness (BWT) was averaged from two independent measurements >1 cm apart, alongside colour Doppler assessments, all blinded to IBD disk data. The Milan Ultrasound Criteria (MUC) score was calculated, with relevant ultrasound activity defined as MUC >6.2 (worst affected segment).1 A high IBD-related burden was defined as an IBD disk score >40. Correlation analysis was performed using Spearman’s rank coefficient. Results Thirty-one paired assessments (median interval: 0.0 [IQR 0.0–3.0] days) were conducted on unique UC patients (38.7% female, median age 32.6 [28.0–45.9] years, median disease duration 1.3 [0.3–6.6] years, 71.0% left-sided colitis). Ultrasound activity (MUC >6.2) was present in 51.6% of patients, while 29.0% reported a high IBD-related disability. Total IBD disk scores correlated significantly with maximal BWT (r=0.41, p=0.02), modified Limberg Doppler signal (r=0.35, p=0.05), and MUC scores (r=0.44, p=0.01). Patients without MUC-defined inflammation showed significantly lower IBD-related disability (median IBD disk score: 41.0 [26.0–52.0]) compared to those with inflammation (72.0 [58.0–78.0], p=0.001) (Figure 1A). Individual IBD disk components—abdominal pain, body image, education/work impact, emotions, energy, interpersonal interactions, defecation regulation, sexual function, and sleep—differed significantly between patients with and without ultrasonographic inflammation (Table 1). Quartile analysis revealed that normal ultrasonographic findings were most likely in patients with the lowest IBD disk scores (Q1 vs Q2–Q4: 87.5% vs 53.3%, p=0.02) (Figure 1B). Receiver Operating Characteristic (ROC) analysis suggested a stricter MUC cutoff of 5.3, achieving 71.0% [48.1–93.8%] accuracy in estimating IBD-related disability with a negative predictive value of 88.2%. Conclusion In patients with UC, IUS findings correlate with IBD-related disability and warrant further exploration as potential endpoint and/or treatment target. However, these metrics seem complementary rather than fully interchangeable. References 1.Allocca M et al. J Crohns Colitis 2023.

Abstract Background Inflammatory bowel diseases (IBD) are predominantly diagnosed during the second to third life decade. However, the disease may manifest itself at any age and current understanding of differences in clinical presentation and therapy use among age groups is still limited. We aimed to analyse age-related patterns in biological and surgical treatment among patients with IBD. Methods This study utilized data from the PANTHER cohort, a prospective Belgian inception cohort including 473 adult patients with IBD from 3 Belgian referral centres. Patient inclusion took place from 2015 to 2023. Patients were categorized into groups based on the age at diagnosis: 'young adult-onset' (18-39 years), 'adult-onset' (40-59 years), and ‘elderly-onset’ (³60 years). Baseline characteristics and treatments were analysed using Chi square, Mann-Whitney U tests, log-rank tests and/or Cox regression in SPSS. Results Baseline characteristics are shown in Table 1. No significant differences were found between age groups in terms of IBD diagnosis, gender, median follow-up duration, and smoking status. Use of biologics differed significantly across age groups with highest uptake in young-adult onset patients (65.8%, P<0.001). Significant differences were found in selection of first biologic with vedolizumab as most frequently selected option in the elderly compared to anti-TNF in the youngest age group (P=0.002). Time to initiation of a biological was earlier in the youngest cohort (P=0.002, Figure 1). Cox regression analysis revealed that older age at diagnosis (HR 0.987, 95%CI [0.98;0.996], P=0.006), UC (HR 0.495, 95%CI [0.50;0.39], P<0.001) and centre of follow-up (Brussels vs Leuven: HR 0.663, 95%CI [0.4;1], P=0.048) were associated with a lower risk of biological initiation. When analysing CD separately, independent risk factors associated with biological use were perianal disease (HR 2.26, 95%CI [1.6;3.2], P<0.001), L3 (HR 1.85, 95%CI [1.3;2.6], P<0.001,) and L4 (HR 12.7, 95%CI [1.6;98.7], P=0.015,) location (compared to L1 location). In UC patients, E2 (HR 7.9, 95%CI [2.8;22.4], P<0.001) and E3 (HR 6.9, 95%CI [2.4;19.3], P<0.001) were related to earlier use of biologics. During follow-up, 19% of CD and 5.1% of UC patients required IBD-related surgery. Univariate analysis showed a higher need for surgery in younger patients (P=0.042), however, this difference was no longer significant when analysing CD (P=0.272) and UC (P=0.09) patients as separate groups. Conclusion Analysis of the PANTHER Biobank reveals significant age-related variation in the administration of biological therapies among IBD patients.

Abstract Background Primary sclerosing cholangitis (PSC) is a chronic progressive cholestatic liver disease often necessitating liver transplantation (LTX). Approximately 70% of PSC patients have a concomitant diagnosis of inflammatory bowel disease (IBD) and could need treatment with biological therapy on top of the immunosuppression to prevent transplant rejection. Vedolizumab (VDZ) and ustekinumab (UST), both agents with a favorable safety profile, are often used in this population despite lack of data concerning safety and effectiveness in the post-LTX setting. Methods A retrospective multicenter case series was performed as a part of the European Crohn's and Colitis Organisation [ECCO] Collaborative Network of Exceptionally Rare case reports [CONFER] project. Primary endpoints were clinical and endoscopic remission at week 52, occurrence of infectious complications, occurrence of malignancy, hospitalizations, and death after liver transplantation. Results In this retrospective study, 58 patients (male n= 34 (59%), median age 42 (interquartile range (IQR) 32-52) were included across 16 participating centers of which 24 (38%) were treated with UST and 40 (63%) with VDZ. Twelve patients (20%) were diagnosed with Crohn’s disease (CD), 44 (76%) with ulcerative colitis (UC), 2 (3%)- with unclassified IBD (IBD-U) and in 12 (20%) patients had an ileal pouch anal anastomosis (IPAA). Median disease duration was 16 years (IQR 13-26) and 33 (56%) had received biological therapy prior to LTX (33% anti-TNF, 11% VDZ, 5% UST). Median disease duration for PSC was 15.5 years (IQR 11-25) and median time since LTX was 6 years (IQR 4-10). Clinical remission, assessed according to physician global assessment, at week 52 was achieved in 44% of VDZ compared to 38% of UST treated patients (p=0.17), while endoscopic remission was seen in 17% of patients in the VDZ group versus 33% in the UST treated patients (p=0.87). Clinical effectiveness was similar across CD (respectively 33% and 20%), UC (33% and 37%) and IPAA patients (36% vs 60%). Infectious complications occurred in 21 patients (29%; 27% VDZ vs 33% UST) post LTX on biological therapy (p=0.66), malignancy occurred in 10 patients (14.1%, 12.8% VDZ vs 16.7% UST, p=0.66), hospitalizations in 32 (45%; 51% VDZ vs 34% UST, p=0.15), and death in 2 patients (3.4%; 2.1% VDZ vs 4.2% UST, p=0.66) (see table). Conclusion In IBD-PSC patients who underwent LTX both UST and VDZ show similar effectiveness with clinical remission rates of respectively 44% and 38% after 1 year. Safety profiles are similar although infectious complications and occurrence of malignancy remains an important concern in this patient group.

Abstract Background Strictureplasty techniques were developed to limit the risk of short bowel syndrome in Crohn’s disease (CD) patients with affected small intestine. In the setting of extensive small bowel disease, long strictures might be treated with modified side-to-side isoperistaltic strictureplasty (SSIS). We studied long-term outcomes in patients who underwent SSIS and evaluated if radiological features on MR enterography (MRE) predict them. Methods This retrospective study included CD patients who underwent SSIS, for whom preoperative and six-month postoperative MRE were routinely performed. Simplified MARIA (MARIAs) score, considering wall thickness, edema, fat stranding and ulcers (score >1 active CD; maximum 5) was used. Improvement in maximal wall thickness was defined as decrease of at least 30% from baseline. Deep remission was defined as absence of symptoms and endoscopic remission (mRutgeerts score ≤i1), and clinical recurrence as occurrence of new symptoms confirmed by endoscopy (mRutgeerts score ≥i2b) or radiology (new strictures/inflammation) requiring treatment. Clinical data were collected from medical records. The correlation between recurrence and features within the MARIAs score was assessed. Results Thirty CD patients underwent SSIS and had pre- and six-month postoperative MRE (table). Immediately after surgery, 13 (43.3%) patients were continued or initiated on advanced therapy. Over a median [IQR] follow-up of 7.4 [3.3-9.3] years, 17 (56.7%) patients showed clinical recurrence of whom 6 (20.0%) needed surgical reintervention (figure). Deep remission was observed in 8 (26.7%) patients. Preoperatively, all patients had a maximum MARIAs score of 5. After surgery, the median [IQR] change in MARIAs score was 0.0 [0.0-0.8]. Cox regression analysis showed no association between decrease in overall MARIAs score, nor separate components of this score, with clinical or surgical recurrence free survival. Wall thickness decrease of >30%, observed in 15 (50%) patients, showed a trend for clinical recurrence free survival (p=0.098). Three patients showed normalization of MARIAs score and showed no recurrence thereafter. None of the clinical variables collected showed significant association with recurrence, except for immediate postoperative advanced therapy which had a 70% protection from clinical recurrence [p=0.036 / HR 0.3 (0.097-0.92)]. Conclusion Long-term follow-up of CD patients undergoing SSIS showed that 57% of patients had clinical recurrence and 20% needed surgical reintervention. The MARIAs score was not predictive of long-term postoperative clinical or surgical recurrence as most features of the score remained unaltered. These shortcomings of the MARIAs score should be addressed when designing improved scoring tools.

Abstract Background The growing number of advanced therapies has revolutionized the management of inflammatory bowel disease (IBD). Although early use of biological therapies is associated with better long-term outcomes, no data exist for the Belgian population. To this end, we evaluated treatment patterns in biological use and persistence in a Belgian inception cohort. Methods The PANTHER (Prognostic biobANk of paTients witH Early cRohn’s or colitis) cohort consists of adult IBD patients recruited in 3 Belgian IBD referral centres. Patients are included within 3 months after diagnosis and are naïve for immunosuppressives and biologicals, and without previous IBD-related surgery. Treatment use and outcomes are prospectively collected, and time trends for biological use were analysed using log-rank tests and Cox regression (R 4.3.2). Results Between 2015 and 2023, a total of 473 newly-diagnosed IBD patients were recruited (270 Crohn’s disease (CD) [57%]; 199 ulcerative colitis (UC) [42%]; 4 [1%] IBD type unclassified) (Table 1). During a median (IQR) follow-up of 2.6 (1.3-4.3) years, 64 patients (14%) required surgery (n=10 colectomy; n=54 ileocecal/small bowel resection); and 250 patients (53%) received biological therapy within the 1st year after diagnosis. Most patients were treated with anti-TNF (CD 67%; UC 55%) as first-line biological, followed by anti-integrins (CD 24%; UC 43%) and anti-IL12/23 (CD 9%; UC 2%). Time series analysis showed a significant increase in biological use within the 1st year after diagnosis when comparing patients diagnosed between 2015-2017 (44%) to those between 2018-2020 (57%), and to 2021-2023 (66%) (p=0.03) (Fig. 1A). Factors associated to this early biological use were younger age (HR=0.99 [95%CI: 0.98-0.99]), a diagnosis of CD (HR=2.2 [95%CI: 1.6-2.8]); and perianal disease in CD (HR=2.8 [95%CI: 1.8-12.8]). Within this early biological exposure group, 26 patients (10%) needed a resection later on. Therapy persistence over time was higher with early exposure rates in patients diagnosed in 2021-2023 (82%) and 2018-2021 (71%), as compared to 2015-2017 (63%) (p=0.08) (Fig.1B). The mode-of-action of first-line biological did not show any association with persistence (HR=1.0 [95%CI: 0.4-3.0]). Overall, only 26% of patients had to switch to a second-line, with a switch [anti-TNF >anti-IL12/23] being the most frequent in CD (50%); and from [anti-TNF >anti-integrins] (46%) or vice versa (40%) in UC. Conclusion In this Belgian inception cohort, two thirds of patients are currently initiated with biological therapy within the first year after diagnosis. This increased biological use is associated with high therapy persistence rates of >80% after a median follow-up of 1.5 years, and with low rates of surgical resections.

Abstract Background Aiming for histological remission on top of endoscopic remission, is an aspiring target in ulcerative colitis [UC]. We investigated the mucosal transcriptional profiles in UC patients with varying depths of histologic activity as compared to healthy controls. Methods UC patients with endoscopic improvement (Mayo endoscopic score [MES] 0 or 1) were recruited. Endoscopies were video-recorded and biopsies from the most affected segment were collected for histological assessment and bulk RNAseq (Illumina HiSeq 4000, single read). Geboes Score [GS] was determined by two pathologists blinded for endoscopic scores. The included patients were divided into endoscopic improvement with histologic activity [EIHA] (MES ≤1, GS ≥2B.1) and histo-endoscopic mucosal remission [HEMR] (MES ≤1, GS <2B.1). We distinguished a subpopulation within the HEMR group that we labelled as histo-endoscopic mucosal normalization [HEMN] (MES ≤1, GS ≤0.1). As comparators, 50 healthy individuals [HV] and 50 UC patients with active disease (MES 2-3, independent, matched by treatment) [AD] were included. Normalised RNA counts were analysed by principal component analysis [PCA] and differential expression analysis. Genes with a |log2 FC|>1 and a FDR <0.05 were considered as differentially expressed [DEG], and were selected for pathway analysis (R 4.3.2, DEseq2; QIAGEN IPA). Results In this study cohort 172 patients were included (Table). PCA analysis showed clustering patterns for each study group (Figure). In particular, separation along the PC1 axis tended to follow the different degrees of histo-endoscopic activity. Interestingly, besides inflammatory and tissue remodeling markers (e.g., DUOX2, MMP 3, CHI3L1) for PC1 and epithelial and metabolic markers (e.g., SLC9A3, LPIN3) for PC2, the gene with the highest variance for both PC1 and PC2 was Aquaporin-8 (AQP8). This gene encodes for a water transporter at the apical surface of the colon, and its mucosal expression level indeed follows a spectrum of histo-endoscopic activity (Figure). Differential expression analyses identified 1261 DEG in HEMN, 1684 DEG in HEMR (not HEMN), 2178 DEG in EIHA, all compared to HV. A total of 1014 overlapping DEG were identified for all these comparisons, being enriched for pathways related to inflammation, metabolic epithelial functions (lipid metabolism, bile acid regulation) and wound healing. Conclusion UC patients with histo-endoscopic remission have a mucosal transcriptional profile which significantly differs from that of healthy individuals, even in UC patients whose biopsies appear normal upon histological evaluation. The histo-endoscopic spectrum of the mucosal transcriptome in UC is nicely illustrated by the gene expression of Aquaporin-8 (AQP8).

Microscopic colitis is a chronic inflammatory condition of the colon. Firstline treatment consists of budesonide, with the consideration of biological agents in refractory cases. Celiac disease is a chronic immune mediated and gluten-induced enteropathy, with treatment consisting of a gluten-free diet. There is an association between microscopic colitis and instead of xand celiac disease, especially in refractory cases they can coincide. In this manuscript, we report for the first time the efficacy of tofacitinib, a pan Janus kinase inhibitor, in the treatment of concomitant microscopic colitis and celiac disease, resulting in persistent clinical and histological remission.

Abstract Background At present, the significance of histologic remission in Crohn’s disease [CD] is unclear. One of the main reasons is the heterogeneous distribution of the lesions. This contrasts with ulcerative colitis, where histologic scoring systems have been applied with success on limited numbers of biopsies. We assessed the variability of histologic activity in paired same-segment biopsies from CD patients. Methods All CD patients undergoing ileocolonoscopy between January 2020 and July 2022 at our tertiary referral center were eligible for inclusion. Clinical data were retrospectively collected from medical records. Endoscopies were video-recorded and the Simple Endoscopic Score for Crohn’s Disease [SES-CD] was used to evaluate endoscopic activity. Paired biopsies were taken within the most affected area per segment, in accordance with ECCO guidelines. The Global Histology Activity Score (GHAS), the Geboes Score [GS], the Robart’s Histopathology Index [RHI] and the Nancy Histological Index [NHI] were determined for each biopsy by an experienced IBD pathologist [GDH] blinded to the endoscopic scores. The cut-off values for histologic remission were set at GHAS ≤4, GS <2B.1, RHI ≤3 and NI ≤1 (absence of neutrophils for RHI and NI). Variability was assessed by intraclass correlation coefficient [ICC] and Kappa statistics. Association between histologic and endoscopic scores was studied with Spearman’s rank correlation coefficient. Results A total of 151 biopsy pairs from 128 patients were analysed. Patient characteristics are summarised in Table 1. Endoscopic activity was observed in 79% of segments. In patients with endoscopic remission, histologic activity (presence of neutrophils) was found in 29% of cases. There was a moderate association between the endoscopic and histologic assessment, with correlation levels ranging from 0.48 to 0.52 [p<0.001]. Histologic indices showed fair agreement between paired biopsies [Table 2]. Looking at the predefined cut-offs, moderate variability across the different histological scores was seen [Table 3]. When comparing biopsies from segments with endoscopic activity and those without, overall higher variability in active segments and substantial agreement in paired biopsies from inactive segments was observed [Table 4]. Conclusion The patchy nature of microscopic inflammation in CD is reflected in the histologic scores of paired same-segment biopsies, with only moderate agreement across histologic scoring indices, although lower variability was seen in biopsies from endoscopic inactive segments. Endoscopic activity only moderately reflects histological status. Multiple biopsies per segment are therefore needed for accurate histological evaluation, the optimal number necessary remains to be explored.

Abstract Background Most endoscopic scoring systems (including the Mayo endoscopic subscore, MES) do not consider the extent of inflammation and can therefore not pick up segmental endoscopic improvement. The modified Mayo endoscopic score (MMES) was developed to overcome this limitation. We examined the predictive value of the MMES in addition to the MES on long-term clinical outcome in ulcerative colitis (UC). Methods Between 2014 and 2017, patients initiating biologic therapy for active UC (baseline MES ≥2) were recruited. Patients without assessment of the upper margin of inflammation or with a clinical follow-up <12 months were excluded. We conducted a clinical and endoscopic assessment at baseline and week 8 (adalimumab, ADM) or 14 (golimumab, GOL; infliximab, IFX; vedolizumab, VDZ). Clinical response was defined as a decrease in the adapted Mayo score (excluding physician global assessment) with ≥2 points and ≥30%, plus a decrease in rectal bleeding score ≥1 or an absolute rectal bleeding score ≤1. We classified patients by evolution of endoscopic activity at week 8/14 in group A (endoscopic healing, MES ≤1), group B (segmental healing, MES >1 but decrease in MMES ≥30%) and group C (MES >1 and no drop in MMES ≥30%). Clinical relapse-free, discontinuation-free and colectomy-free survival was estimated by Kaplan-Meier analysis with log-rank test. Results A total of 150 UC patients were included (52% male, median age 42 years, median disease duration 7 years) with a median (IQR) follow-up of 61 (48–68) months. An anti-TNF was initiated in 74 (35 IFX, 23 ADM, 16 GOL), and VDZ in 76 patients. A significant reduction in MES and MMES was observed at week 8/14 (cf. table). In group A 67/69 (97%) patients achieved clinical response compared to 15/27 (55%) in group B and 11/54 (20%) in group C. During follow-up 60/93 (65%) patients maintained clinical response, 83/150 (55%) patients discontinued treatment due to primary non-response or loss of response and 33/150 (22%) patients underwent colectomy. The ΔMMES demonstrated to be of additional predictive value: there was a significant difference between groups B and C regarding clinical relapse, drug persistence and need for colectomy (cf. figure). Conclusion Although MES ≤1 remains the best predictor of long-term outcome, the MMES - identifying a subgroup with a segmental endoscopic response - demonstrated a clear additional value predicting long-term outcome in UC. These promising results merit inclusion of the MMES in future clinical trials. Table The MMES is calculated by multiplying the Mayo endoscopic subscore for the different colon segments with the maximal extent of inflammation (in decimeters) divided by the number of segments with active inflammation (Lobatón T, et al. JCC 2015; 9:846–52.) Figure