umor shrinkage or progression, time to tumor progression, and progression-free survival are surrogate imaging end points that are frequently used to rate the efficacy of anticancer treatment. The Response Evaluation Criteria in Solid Tumors (RECIST) guidelines have been introduced to standardize such measurements and to provide cutoff values for response classification (1-3).Since the introduction of RECIST guidelines (1-3), a multitude of studies have been performed to assess withinand between-reader agreement of analyses with RECIST versions 1.0 and 1.1. The focus of these published studies was on the effect of manual versus automated size measurements and of uni-versus bidimensional or volumetric measurements of target lesions (4-16). To avoid between-reader variability from these causes, the RECIST consortium recommends using dedicated oncology software tools that help automatize and thus standardize the measurement of target lesion diameters (2).However, with or without use of these software systems, RECIST requires readers to select target lesions. Depending on the type of response classification system used (ie, the maximum number of reportable target lesions), readers must base their response analysis on a more or less small sample of the overall tumor load. It is the implicit assumption of all current response assessment systems that size changes of these target lesions are sufficiently representative of size changes of the entire bodily tumor load. Under the World Health Organization, readers were free to select and measure as many target lesions as deemed appropriate; under the first version of RECIST (RECIST 1.0), readers could select up to
2528 Background: Response-classification-systems, e.g. RECIST1.1, and dedicated oncology software-tools (DOST) are used to standardize response assessment. Expectation is that different readers should yield the same response-classification for any given patient. We investigated real-life variability between readers who, as in clinical practice, were free to select target-lesions (TL). Methods: Prospective study on 316 patients with metastatic disease who underwent 932 CT-studies, yielding a total 616 follow-up occasions (baseline vs. follow-up) for analysis. All CT-studies were independently evaluated by 3 radiologists who used state-of-the-art DOST (MintLesion). Readers were free to select TL in the respective baseline study, and did so independently. Kappa-statistics were used to analyse agreement for RECIST1.1 response-class-assignment depending on whether readers had selected the same or different TL.To investigate possible impact on treatment decisions, agreement was also determined after aggregating response classes into progressive (PD) vs. non-progressive (CR/PR/SD). Results: Readers used the same TL in 38.6 (238/616), different in 61.4% (378/616). Where readers happened to select the same TL, agreement was “almost perfect” (κ = 0.966 [96%-CI: 0.912–1.00] for assignment of individual response-classes, and 0.977 [0.898–1.0] for the distinction progressive-vs-non-progressive). Where readers had selected different TL, agreement was only “moderate” (0.583 [0.541–0.624] for individual response-class-assignment, and 0.644 [0.587 to 0.701] for distinction progressive-vs.-non-progressive). Choice of the same TL was associated with agreement for distinction between progressive-vs.-non-progressive-disease in 97.7 % [95.4%–100.0%] of patients; choice of different TL was associated with disagreement in 44.7% [37.6%–51.8%]. Conclusions: If different radiologists use RECIST1.1 and DOST for response assessment, they will select different TL more often than not. Just depending on whether TL selection was concordant or not, radiologists will exhibit perfect agreement, or substantial disagreement, even for distinguishing progressive vs non-progressive disease.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.