Diagnostic Implications of Multi-Cancer Early Detection Testing in the Investigation of Cancer Symptoms: An Exploratory Retrospective Analysis of the SYMPLIFY Study

Background: The observational SYMPLIFY study reported the accuracy of a multi-cancer early detection (MCED) test in a referred symptomatic population. We explore how the MCED test may contribute to a faster or more efficient diagnosis if acted upon. Methods: We reviewed all cancers diagnosed in SYMPLIFY (ISRCTN10226380) using data collected at study sites plus two years of cancer registry data. All cancer diagnoses were classified based on the congruence between the participant's symptoms, diagnostic referral pathway, MCED test cancer signal origin (CSO) prediction, cancer site, and time to diagnosis. Findings: There were 533 cancers diagnosed among 5461 (9.8%) evaluable participants in SYMPLIFY during the 2-year follow-up period. Among the 79 participants with an apparent false positive test result in the original SYMPLIFY study, 28 (35%) were diagnosed with cancer based on cancer registry data, increasing the MCED PPV to 84.2% (80.1–87.6). In aggregate, only one of the 28 additional patients had a cancer diagnosed that was incongruent with a predicted MCED CSO. Among the 5014 patients with an apparent true negative MCED test result, 113 (2%) received a subsequent cancer diagnosis. In 101 (19%) of the 533 cancers diagnosed, the MCED test result might have contributed to a more efficient diagnosis had it been used to inform the clinical work-up. Conversely, 49 (9%) cancers might have taken longer to diagnose if the MCED test result alone had been used in the diagnostic process, directing investigations based upon an incorrect CSO prediction. Interpretation: These exploratory findings demonstrate a substantially higher rate of cancer diagnoses in symptomatic participants originally classified with a false positive MCED test result than those originally classified as true negative. We show how MCED tests have the potential to assist clinical decision making, which may in turn lead to a timelier cancer diagnosis for one fifth of cancers diagnosed. Funding: GRAIL Bio UK, Ltd. NIHR.