Development of a New Simplified Tool to Predict LNPCPs Histology and Assess the Risk of Submucosal Invasive Cancer. The Colorectal Regular-Irregular Score

NCT07584486 | Not Yet Recruiting

• 1 other identifier: ONZ-2024-0023
• Study Type: interventional
• Target: 764
• Locations: 1 country
• Sites: 1

Brief Summary

Colorectal cancer (CRC) is the third most common malignancy worldwide and the second leading cause of cancer related death. It can be prevented by endoscopic detection and complete resection of colorectal polyps. The JNET (Japanese NBI Expert Team) classification is clinically useful to predict the histology of large non-pedunculated colorectal polyps (LNPCPs) using narrow-band imaging at endoscopy. Japanese experts can reliably predict histology including the presence and depth of submucosal invasive cancer (SMI) using JNET with accuracy \>87%. On the other hand, the International Evaluation of Endoscopy classification-JNET (IEE-JNET) group demonstrated that ESGE and JGES endoscopists had sufficient accuracy for JNET 1 (93.0%) but insufficient accuracy for JNET 2A/B and 3 (respectively 62.1%, 55.1% and 85.1%). Reliably distinguishing between JNET 2A, 2B and 3 has a profound clinical relevance, since JNET 2A lesions can safely be resected using pEMR whereas JNET 2B lesions should be resected en-bloc (EMR or ESD) due to the increased risk of cancer and JNET 3 lesions are preferably treated with surgery due to the high risk of deeply invasive carcinoma and the necessity of lymph node resection. This study aims to validate a new simplified score, the Colorectal Regular-Irregular Score (CRIS) to fulfill the urgent need for a more effective and easier to use tool to predict LNPCPs histology. CRIS is a simplification of the JNET score which is mainly used by Japanese endoscopists or experts, recent evidence suggests its accuracy when used in everyday endoscopy in the Western world is insufficient. The investigators aim to compare JNET with CRIS for LNPCPs histology prediction amongst Western endoscopists using both original JNET interpretation and a clinically relevant approach. The study consists of three work packages (WPs): Work package one involves an expert online study where twelve expert endoscopists will evaluate 32 high-quality images of colorectal polyps using both JNET and CRIS classifications. Work package two involves an image/video-based online study where non-expert participants will be randomly assigned to rate images and videos using either JNET or CRIS, with performance re-evaluated after three months. Work package three involves a clinical study in a live endoscopy environment where non-expert endoscopists will participate in a randomized controlled trial assessing 10 colorectal polyps using either JNET or CRIS.

Conditions

Colorectal Cancer

Keywords

Colorectal cancer; Endoscopic imaging; colorectal polyps; endoscopist

Outcome Measures

Primary Outcomes (1)

• Diagnostic Accuracy of CRIS Classification for Submucosal Invasive Carcinoma (Sensitivity, Specificity, and Overall Accuracy)

  Sensitivity, specificity, and overall diagnostic accuracy of the CRIS classification for predicting submucosal invasive carcinoma compared with histopathological evaluation (reference standard). Accuracy is calculated as the proportion of correct classifications (true positives + true negatives) divided by total assessments. Results will be reported with 95% confidence intervals.

  Immediately after 5-minute training and rating of 32 images (approximately 1 hour per participant)

Secondary Outcomes (6)

• Diagnostic Accuracy of CRIS Versus JNET Among Expert Endoscopists (Sensitivity, Specificity, Overall Accuracy)

  At completion of expert image rating (approximately 30 minutes per expert)

• Diagnostic Accuracy of CRIS Versus JNET Across All Participant Categories (Sensitivity, Specificity, Overall Accuracy)

  Through study completion, an average of 3 years

• Change in Diagnostic Accuracy From Baseline to 3-Month Follow-up for CRIS Versus JNET (Sensitivity, Specificity, Overall Accuracy)

  3 months after initial assessment

• Inter-observer Agreement for Polyp Classification Using CRIS Versus JNET (Fleiss' Kappa Coefficient)

  At completion of baseline image assessment (approximately 1 hour per participant)

• Correlation Between Proposed Endoscopic Treatment and CRIS/JNET Classification (Percentage Agreement)

  At completion of baseline image/video assessment (approximately 1 hour per participant)

Study Arms (2)

JNET Classification Training (Active Comparator)

ACTIVE COMPARATOR

Other: Learning tool (JNET/CRIS)

CRIS Classification Training (Experimental)

EXPERIMENTAL

Other: Learning tool (CRIS/JNET)

Interventions

Learning tool (CRIS/JNET) OTHER

Participants will follow a 5-minute learning video (intervention) on CRIS and later for JNET.

CRIS Classification Training (Experimental)

Learning tool (JNET/CRIS) OTHER

Participants will follow a 5-minute learning video (intervention) on JNET and later for CRIS.

JNET Classification Training (Active Comparator)

Eligibility Criteria

• Age: 18 Years+
• Sex: all
• Healthy Volunteers: No
• Age Groups: Adult (18-64), Older Adult (65+)

You may qualify if:

• Consenting Endoscopists of varying abilities and grades (endoscopist)
• Endoscopists who did not previously encounter the score (endoscopist)
• LNPCPs detected or referred for resection (patient)

You may not qualify if:

• Video of inadequate quality as per opinion of the principal investigator
• Endoscopist does not undergo learning intervention (endoscopist)
• Patient does not consent to data collection for the study (patient)

Sponsors & Collaborators

• University Hospital, Ghent: lead

Study Sites (1)

UZ Gent

Ghent, 9000, Belgium

Location

MeSH Terms

Conditions

Colorectal Neoplasms

Condition Hierarchy (Ancestors)

Intestinal Neoplasms; Gastrointestinal Neoplasms; Digestive System Neoplasms; Neoplasms by Site; Neoplasms; Digestive System Diseases; Gastrointestinal Diseases; Colonic Diseases; Intestinal Diseases; Rectal Diseases

Study Officials

• David Tate

  University Hospital, Ghent

  PRINCIPAL INVESTIGATOR

Central Study Contacts

David J Tate

CONTACT

+3293321063; StudiesTissue.Resectie@uzgent.be

Study Design

• Study Type: interventional
• Phase: not applicable
• Allocation: RANDOMIZED
• Masking: NONE
• Purpose: DIAGNOSTIC
• Intervention Model: CROSSOVER
• Sponsor Type: OTHER
• Responsible Party: SPONSOR

Study Record Dates

• First Submitted: August 14, 2025
• First Posted: May 13, 2026
• Study Start (Estimated): July 1, 2026
• Primary Completion (Estimated): December 31, 2028
• Study Completion (Estimated): December 31, 2028
• Last Updated: May 13, 2026

Record last verified: 2026-05

Locations

BE (1)