All Times EDT
Keywords: real-world, real world, endpoints, oncology, response, progression, ORR, PFS
Tumor based endpoints like objective response rate and progression-free survival are considered important endpoints in oncology for evaluating tumor shrinkage/growth along with treatment efficacy. For solid tumors, they are often assessed using radiographic images combined with physical exams or other clinical information. Variability in assessing response is tightly controlled in clinical trials through the use of protocols that specify the assessment schedule, methods of assessment and standardized measurement criteria. However, there are no such requirements specified in routine clinical practice in the real-world setting and response data in retrospective real-world observational studies is often obtained using EHR-derived clinician notes, radiology reports or radiographic images. To inform the development of robust real-world response- and progression-based endpoints, we conducted a retrospective study in Non-small cell lung cancer (NSCLC) where we evaluated the feasibility of obtaining standard RECIST reads using images obtained in routine clinical practice and assessed concordance with response information abstracted from clinician notes. We will present the study design and results in this presentation along with analytical considerations for assessing the performance of real-world tumor based endpoints and utility for decision making.