Introduction

Both vedolizumab and ustekinumab can be considered for the treatment of ulcerative colitis (UC) but head-to-head trials are lacking. We aimed to compare the effectiveness of vedolizumab and ustekinumab after anti-TNF failure in UC patients.

Aims & Methods

All consecutive patients with UC treated with ustekinumab or vedolizumab after being exposed to at least one anti-TNF in 10 French centers between January 2019 and August 2022 were included retrospectively. The primary endpoint was clinical remission (Mayo score partial ≤ 2) without corticosteroid (SFCR) at week 16 (W16). Secondary endpoints were endoscopic remission (SFCR and endoscopic Mayo score ≤ 1 or fecal calprotectin < 150 µg/g), and histological healing (endoscopic remission and histological (Nancy index ≤ 1)). All the comparisons were performed using propensity score analyses (inverse probability of treatment weighting) adjusted for age, gender, smoking status, disease duration and extension, number of prior biologics or prior primary failure to biologics, the co-prescription of 5-ASA, corticosteroids or immunosuppressants, and disease severity.

Results

Among a total of 316 patients included, 165 and 151 received ustekinumab and vedolizumab, respectively. Baseline characteristics (gender, age, disease location, co-prescription of 5-ASA or corticosteroids) were similar between the two groups except for disease duration (ustekinumab, 9.3 vs vedolizumab 7.2 year; p=0.003), co-immunosuppressants (5.4% vs 20.5%, <0.001), and prior exposition to > 2 biologics (46.6% vs 5.9%, p<0.001). At W16, SFCR was respectively 40.3% and by 37.9% in the ustekinumab and vedolizumab groups (aOR = 1.10 [0.39-3.12], p=0.085). SFCR rates were 62.2% vs 53.2% (p=0.52) after one failure, 33.7% vs 44.3% (p=0.52) after two biologics and 19.1% vs 2.6% (p=0.03) after 3 biologics failure, in the ustekinumab and vedolizumab groups, respectively.
Endoscopic remission and histologic healing rates at week 16 were respectively of 17.5% vs 5.3% (aOR = 3.77 [1.25-11.36], p=0.018) and 11.1% vs 2.1% (aOR = 5.85 [1.47-23.30], p=0.012) in the groups ustekinumab and vedolizumab.

Conclusion

While no difference in terms of clinical remission was observed, ustekinumab appears to be more effective than vedolizumab to induce endoscopic and histological remission after failure of anti-TNFs in UC.

Disclosure

AbbVie, Ferring, Janssen, MSD, Takeda, Pfizer, Celgene, Biogen, Amgen, Boehringer Ingelheim, Hospira et Celltrion

Introduction

Eosinophilic esophagitis (EoE) is a chronic, condition linked to a type 2 inflammatory response and frequently associated to other atopic conditions. EoE may presents either sporadically, or with aggregation of cases within a same family. Limited information has been provided on the differences in clinical characteristics and disease features between sporadic and familiar EoE, largely due to the lack of large patient cohorts.

Aims & Methods

To find differences in sporadic and familiar cases of EoE which may support distinct clinical management, a cross-sectional analysis of the multicenter EoE CONNECT registry was performed. Demographic and clinical from patients recruited at 42 hospitals in 4 European countries were extracted on March 28th, 2025. Information extracted included 1) age at diagnosis and diagnostic delay; 2) symptoms leading to diagnosis; 3) disease endoscopic and histologic features at diagnosis; and 4) atopic concomitances.

Results

	S-EoE	F-EoE
Age at diagnosis (n= 3,179) <18 years old	18.5%	30.6%
Inflammatory Score (n= 2,723)	1.62 ± 1.3	1.93 ± 1.32
Fibrosis Score (n=2,723)	1.04 ± 1.04	0.97 ± 1.08
Fibrosis (n=1,701)	58.6 %	62.8 %
EoE-Phenotype (n=3,017)
Inflammatory	75.9 %	79.6 %
Mixed	12.6 %	8.6 %
Stricturing	11.5 %	11.8 %

Overall, 4,053 patients (75.4 % males, mean age 33.3 ±15.2 years) were analyzed, including 3028 (74.7 %) sporadic (S-EoE), 714 (17.6%) without familiar data and 311 (7.7 %) familiar (F-EoE) cases including 89 families. Among the family forms, n= 76 families showed 2 affected members, n= 11 presented 3 members, and n= 2 with 4 members. In 85/89 families we identified the index case consisting in sons (13), parents (23), siblings (38), or other distant kinships.
The F-EoE forms were diagnosed at younger mean ± SD ages than S-EoE (29.8±15.6 vs. 33.7±15.1; p<0.001) and included a higher proportion of pediatric patients (considered as those below 18 at diagnosis; 30.6% vs. 18.5%; p<0.001)
Symptoms leading to EoE diagnosis were similar across study groups, with slightly higher presences of abdominal pain (as they included more pediatric cases; 7.7% vs. 4.5%; p=0.01) and chocking episodes (7.7% vs. 4.5%; p=0.01) among F-EoE.
Endoscopy presented more severe features according to EREFS among F-EoE (mean ± SD being 2.86 ± 1.66 vs. 2.64 ± 1.68; p=0.054) at the expense of both fibrotic and inflammatory features, despite only the later reach statistical significance (1.93 ± 1.32 vs. 1.62 ± 1.3; p<0.001). Esophageal biopsies exhibited higher mean ± SD peak eosinophil counts in F-EoE cases (57.11 ± 40.7 vs. 52.87 ± 37.97; p=0.04).
F-EoE were more allergic patients, as they presented a higher mean ± SD number of concomitant atopies (1.44 ± 1.25 vs. 1.25 ± 1.22; p=0.009) and all were more common among them, with conjunctivitis and hives reaching statistical significance (p=0.01 and p=0.002 respectively). Overall, a higher proportion of F-EoE had at least one atopy compared to S-EoE (69.4 % vs. 63.8%).

Conclusion

F-EoE seems to present more severely, with earlier onset, more inflammatory load in endoscopy and biopsies and higher concomitant allergic manifestations.

Introduction

The Epi-IBD cohort is a prospective European population-based cohort of 1,390 patients diagnosed with inflammatory bowel disease (IBD) in 2010 and 2011 according to Copenhagen criteria. Of these, 482 were diagnosed with Crohn’s Disease (CD), 817 with ulcerative colitis (UC) and 91 with IBD unclassified (IBDU). The Epi-IBD cohort was established to examine differences in incidence, treatment strategies, disease course and prognosis among IBD patients in Eastern and Western Europe.

Aims & Methods

This study aims to describe the use of biological therapy as well as the treatment prior to biological therapy. IBD patients were followed prospectively from the time of diagnosis until the 31^st December 2020, death, emigration or loss of follow-up. Clinical data on surgery, hospitalizations, and medical treatment, were captured throughout the follow-up period and entered into a validated web-database, www.epi-ibd.org.

Results

During a median follow-up period of 10.7 (interquartile range (IQR), 10.0-11.4) years, 34% (N=165/482) of CD and 15% (N=122/817) of UC patients were started on biological therapy. Biological therapy was initiated within the first year after diagnosis in 16.8% (N=81/482) of all CD and 3.8% (N=31/817) of all UC patients. In median biological therapy was started at 1.0 year (IQR, 0.3-3.1) and 2.7 years (IQR, 1.6-4.0) after CD and UC diagnosis, respectively. First-line biological therapy was infliximab in 74% (N=213/287) and adalimumab in 22% (N=64/287) of the patients. The remaing (N=10/287) received either Golimumab, Vedolizumab, Ustekinumab or were part of a trial/unknown. The sequences of biological therapies are displayed in Sankey diagrams, Figures 1 and 2. Herein illustrated is that 34% (N=56/165) CD and 34% (N=41/122) UC received at least two different biological drugs. Immunomodulators were initiated simultaneously to the first biological therapy in 19% (N=31/165) of CD patients and in 9% (N=11/122) of UC patients. Overall 63% (N=104/165) CD and 51% (N=62/122) UC patients were receiving concomittant immunomodulators with the first biological therapy.
The sequences of therapy leading up to biological therapy are displayed in Sankey diagrams, Figures 3 and 4. Herein illustrated is that the most common sequence of treatment before the first biological therapy for CD was 1) Steroids, 2) Immunomodulators (21%, N=34/165), and for UC was 1) 5-aminosalicylic-acid (5ASA), 2) Steroids, 3) Immunomodulators (39%, N=48/122). Prior to the first biological therapy one, two and three or more courses of systemic steroid were used in 41%, 28% and 17% (N=67, 47 and 28) of CD patients and 43%, 24% and 26% (N=52, 29 and 32) of UC patients. In 14% (N=23/165) and 7% (N=9/122) of CD and UC patients, systemic steroids were not used prior to first biological therapy. In patients having immunomodulators prior to biological therapy, these were administered for a median of 0.5 (IQR, 0.1-1.9) years and 1.1 (IQR, 0.2-3,2) years in CD and UC, respectively. For 5ASA this was 0.8 (IQR, 0.2-1.5) years for CD and 2.2 (IQR, 0.8-4.3) years for UC.

Conclusion

In this multicenter European population-based cohort, different prescribing patterns for biologics emerge: 34% of CD and 15% of UC patients received biologics within 10 years after diagnosis. Infliximab was the most common first-line biological therapy. Only about half of the patients received concomitant immunomodulators. Prior to the first biological therapy, most patients were exposed to systemic steroids and many to repeated courses.

Introduction

New-onset diabetes (NOD) is associated with the development of pancreatic ductal adenocarcinoma (PDAC) and is suggested as potential biomarker to detect PDAC at an earlier stage. International guidelines recommend routine glucose measurements in pancreatic surveillance of high-risk individuals (HRIs), however, little is known about its added value in addition to imaging. Therefore, this study evaluates the added value of longitudinal glucose measurements as a diagnostic biomarker for PDAC in high-risk surveillance cohorts.

Aims & Methods

HRIs with a CDKN2A germline pathogenic variant participating in pancreatic cancer surveillance who had ≥ 1 fasting blood glucose (FBG) and/or HbA1c measurements were included in this study. Pancreatic surveillance consists of annual magnetic resonance imaging (MRI) and endoscopic ultrasound (EUS), and blood sampling including blood glucose measurements. Individuals with diabetes mellitus at baseline, less than 3 years of follow-up and all follow-up after pancreatectomy were excluded from the study. Data was collected on demographics, FBG- and HbA1c measurements, MRI- and EUS examinations, and pathology reports. NOD was defined as two consecutive FBG values of ≥ 7 mmol/L or one HbA1c value of ≥ 48 mmol/mol. Univariate multivariable logistic regression was performed to assess the relationship between NOD and PDAC, with adjustment for sex, age and smoking history. To quantify the diagnostic performance of NOD as a marker for PDAC, receiver operating characteristic (ROC) curve and area under the curve (AUC) were computed.

Results

In total, 220 HRIs were included in the analysis. Median age was 61 (IQR 53-71) years, 62.7% of participants were female, and the mean BMI was 26.2 (4.12 SD). More than half (54.1%) of the studied cohort had a positive smoking history and the median amount of glucose measurements per person was 7 (IQR 5-12). During the study period, 26 (11.8%) HRIs developed NOD, of whom 5 (19.2%) later developed PDAC. The other 23 (82.1%) PDAC cases remained NOD-free. Multivariable analysis showed no statistically significant relationship between NOD and PDAC (OR 1.21; 95% CI, 0.39-3.78). A statistically significant association was exclusively observed between age and development of PDAC (OR 1.05; 95% CI, 1.01-1.1). Furthermore, NOD did not differentiate between HRIs with- and without PDAC (AUC 0.54; 95% CI, 0.46-0.61).

Conclusion

This study demonstrated no added value for longitudinal glucose monitoring in HRI participating in an imaging-based pancreatic cancer surveillance program.