Endoscopy and central reading in inflammatory bowel disease clinical trials: achievements, challenges and future developments

Abstract

Central reading, that is, independent, off-site, blinded review or reading of imaging endpoints, has been identified as a crucial component in the conduct and analysis of inflammatory bowel disease clinical trials. Central reading is the final step in a workflow that has many parts, all of which can be improved. Furthermore, the best reading algorithm and the most intensive central reader training cannot make up for deficiencies in the acquisition stage (clinical trial endoscopy) or improve on the limitations of the underlying score (outcome instrument). In this review, academic and industry experts review scoring systems, and propose a theoretical framework for central reading that predicts when improvements in statistical power, affecting trial size and chances of success, can be expected: Multireader models can be conceptualised as statistical or non-statistical (social). Important organisational and operational factors, such as training and retraining of readers, optimal bowel preparation for colonoscopy, video quality, optimal or at least acceptable read duration times and other quality control matters, are addressed as well. The theory and practice of central reading and the conduct of endoscopy in clinical trials are interdisciplinary topics that should be of interest to many, regulators, clinical trial experts, gastroenterology societies and those in the academic community who endeavour to develop new scoring systems using traditional and machine learning approaches.

Establishment of a validated central reading system for ileocolonoscopy in an academic setting

“…There are arguments in favour of using more than a single reader -either using statistical methods to combine reader scores and detect outliers or via methods of arbitration. [24,25] Whilst this may lead to further increases in confidence in scores, and is supported within our platform for future users who may require this level of confidence, this is not an absolute requirement of our system. Importantly, a minimum percentage of reads must be double-read to support ongoing quality assurance of the reader pool -a process that is implemented in our platform in a blinded manner.…”

“…We also demonstrate how the pool can be added to and monitored for quality assurance over time. There are arguments in favour of using more than a single reader—either using statistical methods to combine reader scores and detect outliers or via methods of arbitration 24 25. While this may lead to further increases in confidence in scores, and is supported within our platform for future users who may require this level of confidence, this is not an absolute requirement of our system.…”

Central Reading of Ulcerative Colitis Clinical Trial Videos Using Neural Networks

“…As described in our data preprocessing pipeline, a convolutional neural network developed and trained by Docbot on proprietary endoscopic video data to predict Boston bowel prep scores at the frame-level 19 was used to automatically remove frames from the endoscopy videos that were likely to be of poor prep quality. The Boston Bowel Prep Scale has 4 levels, poor (0), fair (1), good (2), and excellent (3).…”

“…2 Current practice in phase 3 clinical trials records colonoscopy or sigmoidoscopy videos at baseline and after the intervention submitted to central reading by experienced gastroenterologists. 3 Having only 1 central reader per read instance has largely been discontinued for phase 3 trials in favor of 2 þ 1 central reads to improve accuracy and reduce bias. [4][5][6] Relevant endoscopy scores include the endoscopic component of the Mayo score for ulcerative colitis disease activity (eMS) 7 and the Ulcerative Colitis Endoscopic Index of Severity (UCEIS), 8 the latter having been developed from the former.…”

UDCA in UC: A true disease modifier or symptoms reliever?

“…It should be noted that the lack of a central reading in this study could have potentially compromised the results because of the high rate of subjectivity and variability among the endoscopists when calculating the score. [ 2 ] Moreover, the Endoscopic Mayo score has its limitations for assessing the clinical outcomes and predicting prognosis in UC compared with other scores, for example, the Ulcerative Colitis Endoscopic Index of Severity (UCEIS). [ 3 4 ]…”