Importance of RECIST 1.1 training program for improving performance of contract research organization (CRO) staff in solid tumor clinical trials.

Abstract

e21018 Background: Response Evaluation Criteria In Solid Tumours (RECIST) version 1.1 remains an essential part of the assessment of tumor burden in solid tumor clinical trials, being part of clinical trial objectives and endpoints. Use of RECIST 1.1 as a standard evaluation approach is associated with a number of challenges, including appropriate training of radiologists, clinical trial investigators and also CRO staff reviewing the study data. Methods: We developed a training program in RECIST 1.1 criteria which consists of a tutorial with illustrative examples from clinical trial practice, and a multi-level knowledge evaluation that includes a baseline test, post-tutorial test, and a follow-up assessment 3 months after the training via a multiple-choice quiz. In addition, we developed an anonymous online five-item survey to explore the satisfaction with the training program. This program was launched in June 2021 and is ongoing. We analyzed the results with SAS software calculating summary statistics (mean, median, SD, 95% CI) for the outcomes of all three tests and for changes in results between different tests. One-sample t-test was used to test if there was a significant improvement from baseline to post-tutorial test results. Results: As of December 31 st , 2024, 350 participants attended the training. There was a statistically significant improvement in post-training knowledge (median change from baseline to post-tutorial test = 20%; 95% CI 20.24 to 25.94, p<0.0001) and this level remained unchanged three months after completion of training (median change from post-tutorial to follow-up test = -0.54%; CI -4.93 to 3.85, p=0.8072). Most participants (80%) found the training informative and easy to understand and 75% would be able to immediately use the gained knowledge in their work. Follow up data revealed that the training significantly impacted individuals’ practice with 86 % reporting it had improved. Additionally, to inform our understanding of the most difficult parts of RECIST 1.1 criteria for learning, we categorized mistakes that trainees made in the baseline quiz and revealed that the majority of mistakes occurred in defining and measuring target lesions (44.2% of mistakes) and defining progression of disease (32.2%); this was followed by defining of non-target lesions (21.2%), and methods of assessment (2.4%). This data was used for re-structuring the training materials in order to focus on the topics that caused the most difficulty. Conclusions: To our knowledge, this is the first large-scale analysis of effectiveness of RECIST 1.1 training tools and the impact on trainees’ performance. Considering a significant improvement of knowledge gain and retention, this program has become an integral part of the training matrix at our CRO for positions involved in review of RECIST 1.1 data in solid tumor clinical trials.