In real-world oncology studies of solid tumors, response to treatment can be evaluated based on electronic health record (EHR) documentation, called clinician assessed response (CAR) [referred to in previous publications as ‘real-world response’]. This approach differs from standardized radiologist-measured response (i.e., RECIST). This study evaluated feasibility and concordance between CAR and imaging response based on RECIST (IRb-RECIST), and investigated discordance causes.
This study is one of the first to systematically collect routine care solid tumor imaging from real-world data sources to implement and show the feasibility of a RECIST adaptation using independent centralized imaging review. The imaging response based on RECIST (IRb-RECIST) was then used to benchmark a clinician assessed response variable derived from clinician-documented response in the EHR.These two approaches had moderate concordance overall in the patient cohort studied, which appeared to be affected by factors such as treatment type or availability of scans.
Why this matters
The analysis of EHR-derived data requires a distinct toolbox of metrics and methods. However, investigators need to understand the interrelation between that analytic armamentarium and the standardized methodology of prospective research, if the full potential of RWE as a complement and supplement of clinical trials is to be reached. This study represents a significant step in that direction in the field of oncology. This study characterized the concordance between EHR-based criteria for the assessment of treatment response, previously described by our researchers, and a real-world adaptation of imaging-based RECIST with centralized independent reading. Not only do the results indicate that EHR-based data variables (in this case treatment response) have validity, but the study approach paves the way towards establishing analytic standards in RWE research.