To investigate interobserver agreement among endoscopists to grade the morphology and predict histopathology of colorectal lesions in patients with inflammatory bowel disease.

August 19, 2023

Patients diagnosed with inflammatory bowel disease (IBD) are at increased risk of colorectal cancer. To detect precancerous or cancerous polyps early in these patients, current guidelines recommend starting surveillance colonoscopy after 8 to 10 years. IBD Every 1-2 years after diagnosis.

Background mucosal inflammation in patients IBD Colorectal lesions can be difficult to identify, especially flat lesions. When evaluating colorectal polyps endoscopically, gastroenterologists use various scoring systems to describe the morphology of the lesion and predict its histopathology. The Surveillance International Consensus on the Detection and Management of Colorectal Endoscopic Neoplasms in Patients with Inflammatory Bowel Disease (SCENIC), published in 2015 in the journal Gastrointestinal Endoscopy, now requires endoscopists to It is recommended to evaluate polyp characteristics using the Paris classification system.

Nayanthara Coelho Prabhu, MBBS, a gastroenterologist at the Mayo Clinic in Rochester, Minn., explains how endoscopists accurately grade morphology and predict histology in patients with colorectal disease. He says he understands what he is doing. IBD It is important to use currently available classification systems.

“Using the Modified Paris Classification System for Evaluation” IBD The lesion has not yet been verified. “The results of this assessment directly impact real-time treatment decisions and patient outcomes,” explains Dr. Coelho-Prabhu. “The structure of the polyp determines the type of treatment performed, especially the type of excision technique that is best to ensure complete removal.” It may indicate a high possibility of precancerous changes. Conversely, certain characteristics may indicate a lower risk of malignancy, allowing for less invasive treatment options. ”

To analyze these issues, Dr. Coelho-Prabhu and colleagues conducted a study to analyze interobserver agreement (IOA) between endoscopists using the modified Paris classification system and We evaluated the accuracy of endoscopists’ pathology prediction in patients with visible colorectal lesions and: IBD. The results of this study were published in his journal Gastrointestinal Endoscopy in 2023, with Dr. Coelho-Prabhu as corresponding author.

method

In this study, 10 senior endoscopists and four residents from five tertiary care centers collected 100 unidentified endoscopic still images and 30 visible colorectal lesions found in patients with: I graded the video for the book. IBD. The endoscopist was asked to review each image and provide a single answer for each of the standard Paris classification, modified Paris classification, lesion border assessment, and histopathological prediction. The researchers then measured the agreement between these responses using Wright’s kappa coefficient and evaluated the agreement between ratings according to strict majority voting. Participating endoscopists had access to a teaching guide that included the Paris Classification and the Modified Paris Classification during the review.

result

Overall, the study data showed very low. IOA Accuracy is low for Paris and modified Paris classification; IOA For histopathological prediction of lesions.

• Wright’s kappa coefficients for all study endpoints (using both image and video datasets) ranged from 0.32 to 0.49. In a subgroup analysis comparing data from endoscopist trainees and senior endoscopists, Wright’s kappa coefficient was slightly higher, less than 0.6. IOA Among the trainees.
• Paris classification: overall IOA It was 0.41 for images and 0.42 for videos.
• Revised Paris Classification: Overall IOA It was 0.42 for images and 0.41 for videos.
• lowest lesion IOA or no consensus were mostly classified as Is, IIa and mixed using the Paris classification system and as sessile and superficial elevation using the modified Paris classification system. The difference between these categories is whether the lesion is less than or greater than 2.5 mm in elevation. The data show that this is clearly very difficult to confirm. A more important aspect is whether this difference is clinically relevant in terms of outcome, but there are no data to support a difference in outcome.
• Lesion border assessment: global IOA Boundary prediction was higher for videos (0.49) than for images (0.32). But he was still poor. This is important because only well-circumscribed lesions are candidates for endoscopic resection, whereas ill-circumscribed lesions must be removed by surgical resection.
• Histopathological prediction: overall IOA Still image histology was 0.39. When compared to ground truth pathology, the overall accuracy of histopathology prediction was 59% for senior endoscopists and 57% for residents.Overall IOA Video histology was 0.37. The overall accuracy of pathology prediction was 52% when compared to ground truth pathology. This evidence strongly supports the need for the development of artificial intelligence (AI)-based polyp detection and characterization tools in this patient population.
• The association between Paris classification and predicted histopathology was strong for both still and video images (P < 0.001).

“The key findings of our study suggest that there is an insufficient level of agreement among endoscopists when evaluating colorectal polyps using currently available scoring systems.” IBDDr. Coelho-Prabhu said: “Further research and research is needed to develop more accurate lesion characterization systems and to avoid missing precancerous lesions or leaving these lesions in patients without being biopsied or removed. We need innovative strategies.”and IBD. ”

Dr. Coelho-Prabhu said future research focused on colorectal polyp morphology and histopathological prediction will help clinicians refine surveillance protocols and follow-up colonoscopy to detect and identify high-risk lesions. It explains that it helps determine the optimal interval for endoscopy.

“Additional research is also needed to help better understand how a lesion’s characteristics influence its dysplastic potential to create a better internal histological scoring system.” says Dr. Coelho-Prabhu. “The ideal scoring system should be easy to use for community practitioners and have high interobserver agreement and reproducibility.”

Dr. Coelho-Prabhu and his co-authors also A.I. Tools for detecting and characterizing colorectal lesions in patients IBD It may help you tackle all of the challenges listed above.

“Our research team has also recently developed a new study. A.I. Tools for computer-assisted detection of colorectal lesions IBD” explains Dr. Coelho-Prabhu. “Automated characterization of these lesions is an ongoing research effort. Understanding the morphological changes associated with IBDAssociated lesions can help develop new models that leverage: A.I. Helps clinicians accurately recognize, interpret, treat, and monitor lesions.The result of our efforts to develop this A.I.The base tool was published on iGIE in 2023. ”

For more information

American Society of Gastrointestinal Endoscopy and American College of Gastroenterology. scenic International Consensus Statement on Surveillance and Management of Dysplasia in Inflammatory Bowel Disease. Gastrointestinal endoscopy. 2015;81.489.

Guerrero-Vinsard D et al. Interobserver agreement for the modified Paris classification and histological prediction of colorectal lesions in patients with inflammatory bowel disease. Gastrointestinal endoscopy. 2023;97:790.

Guerrero-Vinsard, D., et al. Development of artificial intelligence tools to detect colorectal lesions in inflammatory bowel disease. Iggy. 2023;2:91.

Refer the patient to Mayo Clinic.