Abstract 3891: Early real-world experience with repeat multi-cancer early detection (MCED) testing

Abstract

Abstract Background: An MCED test (Galleri®, GRAIL, LLC, Menlo Park, CA) has been in clinical use since April 2021. We previously reported the results of screening the first ~53,000 individuals in the US with this test from real-world (RW) practice. Here, we report the RW experience with repeated MCED testing in ~5800 individuals. Methods: This MCED test analyzes targeted methylation in cell-free DNA and uses machine learning to detect a Cancer Signal (CS) and determine its Cancer Signal Origin (CSO). We analyzed results returned through August 2023 for those with a “No Cancer Signal Detected” (NCSD) result in Year 1 and an additional consecutive test. We excluded tests from clinical studies and sites that limited data sharing. Systematic collection of outcomes was attempted for all cases with a “CS detected” (CSD) result and continues via a quality assurance (QA) program. Results: In this RW cohort, a second test was ordered in 5794 individuals (58% male, average age=61.3 y, range: 24.0-89.0 and 42% female, average age=60.5 y, range: 23.0-89.0). Median interval between blood draws was 12.9 months, and 81% (4693/5794) were tested 10-18 months after the initial test. Among the 5794, 26 received a CSD result in the second test, of which 20 were within 10-18 months, corresponding to a CSD rate (CSDR) of 0.45% (95% CI 0.31%-0.66%; 26/5794). The CSDR was slightly higher in males (0.50% [0.32%-0.81%; 17/3367]) vs females (0.37% [0.2-0.7%; 9/2427]) and was lower than the Year 1 CSDR of 0.95% (95% CI 0.87-1.0; 510/53,744). The top CSOs reported during second tests were: Lymphoid, Head & Neck, Bladder and Urothelial, Colon & Rectum, Anus. Among the 26 CSD cases, the QA program confirmed 12 clinical outcomes (8 with a cancer diagnosis, 4 with no cancer diagnosis), and 14 cases are still under investigation or without outcomes at this time. Among 8 cases with a confirmed cancer diagnosis, accuracy of the 1st CSO prediction was 100%, and diagnoses reflected invasive cancers across multiple cancer types including anus, colon, head & neck, urothelial tract, ovary, and lymphomas. Staging information was available in 5 cases, including 4 stage I cancers (4/5, 80%; anus, head & neck, lymphoma, bladder) and one stage IV cancer (ovary). Conclusions: A subset of individuals who returned for MCED testing had cancers detected, including cancers without recommended screening, underscoring the potential value of annual screening with this MCED test. There may be systematic differences in this limited population of individuals who returned for repeat screening, reducing the signal detection rate through potential early adopter bias. Clinical data return was incomplete, and confirmed 8 cancers, including 4 stage I cancers. All cancers were found at the predicted CSO, demonstrating the clinical utility of this unique test feature that directs diagnostic work up. Citation Format: Richard Abrams, Rita Shaknovich, Jordan Lipton, Matthew McMillin, Adam Schneider, Matthew Margolis, Xinyi Hou, Yoobin Oh, Victoria Le, Earl Hubbell, Roger Jiang, Jeffrey M. Venstrom, Ora K. Gordon. Early real-world experience with repeat multi-cancer early detection (MCED) testing [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 3891.