Machine Learning Approach to Study the Interactions Between Environment and Intestinal Tissue Homeostasis in IBD

Not yet recruiting

• Conditions: Diabetes Mellitus, Type 1; Ulcerative Colitis

• Registration Number: NCT06120322
• Lead Sponsor: IRCCS San Raffaele

Brief Summary

The intestinal epithelial barrier is one of the most important security checkpoints of our body that constrains harmful factors from invading mucosal surfaces and facilitates the absorption of nutrients and water. Its correct functioning is essential for maintaining gut tissue homeostasis and proper immunity. However, such an equilibrium may be interrupted, resulting in an uncontrolled entrance of pathogenic stimuli that in turn activate a persistent gut immune response, with detrimental consequences for both local and systemic immunity. Alterations in the composition and functionality of the gut microbiome seem to be a central factor in affecting gut barrier integrity thus influencing intestinal permeability. The microbiome composition is impacted by dietary habits and environmental pollution and conditions, hygiene, genetic asset, and physical activity, which could interact in concert leading to dysbiosis, thereby influencing the immune response through the production of several metabolites. Chronic inflammatory diseases, including ulcerative colitis (UC) and type 1 diabetes (T1D), share microbiota dysbiosis, among pathologic characteristics, that may arise, be provoked, or be exacerbated because of barrier leakage. Therefore, these two chronic diseases may be considered prototype pathologies where the intrinsic connection between intestinal dysbiosis and the barrier leakage impact each other during the pathogenesis.

Detailed Description

This is an observational multicentre study performed on patients with an established diagnosis of UC (according to the standard classification) and patients with new-onset type 1 diabetes (T1D) which aims to identify environmental and genetic factors contributing to chronic inflammation within the intestine and in peripheral organs by taking advantage of Internet-Of things (IoT) technologies (web app) and machine learning approaches. During the colonoscopy procedure planned for patients with UC following the routine surveillance according to the normal clinical practice (0 and 12 months), the gastroenterologist will collect 8 additional biopsies; furthermore, blood samples and stools for UC patients and blood samples, stools, and urines for T1D patients will be collected at baseline and during the routine surveillance according to the normal clinical practice (0, 6, and 12 months) and stored for the following analysis. For T1D patients, blood, urine and stool sample collection are not planned for the normal clinical practice, but will be performed specifically for this research proposal at different points during the normal clinical practice clinical visit: baseline, after 6 months and after 12 months.

For UC patients, blood and biopsies are collected during the procedures already planned for normal clinical practice during clinical surveillance. The investigators will take advantage of this standard of-care procedures to collect an additional volume of blood (at baseline, after 6 months and after 12 months). Therefore, patients expressing their voluntary participation in the study will be asked to give fecal samples during the routine-surveillance visit (as per the normal clinical practice) at different: at baseline, after 6 months and after 12 months.

Ospedale San Raffaele (OSR - Operative Unit (UO)1 (UO1)) is the promoter of this study. The other centers participating in the study are:

* Ospedale Casa Sollievo della Sofferenza (CSS) - Foggia (UO2)

* Azienda Ospedaliera San Camillo Forlanini - Roma (UO3) 150 subjects in total (100 patients with UC and 50 patients with T1D) will be enrolled at the IBD Center (Department of Gastroenterology and Digestive Endoscopy) and at the Pediatric Unit at Ospedale San Raffaele.

Study Timeline

• Study First Submitted: 2023-09-29
• Status Verified: 2023-11
• Study Start: 2023-11-01
• Study First Submitted QC: 2023-11-02
• Last Update Submitted: 2023-11-02
• Study First Posted: 2023-11-07
• Last Update Posted: 2023-11-07
• Primary Completion: 2025-05-01
• Study Completion: 2025-11-01

Recruitment & Eligibility

• Status: NOT_YET_RECRUITING
• Sex: All
• Target Recruitment: 300

Inclusion Criteria

UC:

• adult patients (>18 years)
• established diagnosis of UC (any extent)
• the disease in remission as defined by normal bowel movements/day
• no rectal bleeding
• fecal calprotectin <250 ug/g)

T1D:

• age between 7 and 17 years
• clinical diagnosis of insulin-dependent type 1 diabetes
• positivity for at least one islet autoantibody (ICA, GADA, IA-2, IAA, ZnT8)
• no more than 3 months from first insulin injection

Exclusion Criteria

• unsigned informed consent
• celiac disease
• other intestinal inflammatory pathologies
• significant cardiac disease
• conditions associated with immune dysfunction or hematologic dyscrasia (including malignancy, lymphopenia, thrombocytopenia, or anemia)
• liver or renal dysfunction,
• tuberculosis,
• HBV, HCV, HIV, or active EBV or CMV infections

Study & Design

• Study Type: OBSERVATIONAL
• Study Design: Not specified

Primary Outcome Measures

Name	Time	Method
To identify the factors contributing to chronic inflammation within the intestine and in peripheral organs, in particular focusing on biopsies markers	1-18 months	RNA extraction and transcriptomics to identify molecular variation
To identify the factors contributing to chronic inflammation within the intestine and in peripheral organs, in particular focusing on feces markers	1-18 months	Evaluation of microbial composition of the gut: identification and quantification of different microbial species colonizing the gut
To identify the factors contributing to chronic inflammation within the intestine and in peripheral organs, in particular focusing on blood markers	1-18 months	Evaluation of the mmune response profile: measurement of the percentage of different immune cell populations

Secondary Outcome Measures

Name	Time	Method
Taking advantage of the measurements collected in Outcome 1, the investigators will use a machine learning-based multi-omics approach to easily recognize patient differences, stratification and characterization	19-24 months	Using MOFA, a machine learning based bioinformatics tool that comprehensively and simultaneously analyzes multiple omics and patient-specific data recorded during the follow up, the investigators could identify the origin of different clinical outcomes during the disease course, ultimately stratifying them based on environmental factors to which they were exposed and their molecular and genetic characteristics.

Trial Locations

Locations (1)

IRCCS Ospedale San Raffaele; 🇮🇹 Milan, Italy