A Quantitative Framework to Study Potential Benefits and Harms of Multi-cancer Early Detection Testing

Background Multi-cancer tests permit identification of multiple cancers with one blood draw. The objective of this study was to quantify potential population impact of a multi-cancer test. Methods We formulate mathematical expressions for expected numbers of (1) individuals exposed to unnecessary confirmation tests ( UCT ), (2) cancers detected ( CD ), and (3) lives saved ( LS ) given disease prevalence and mortality and the test’s performance and expected mortality reduction. We consider additions of colorectal, liver, lung, ovary, and pancreatic cancer to a test for breast cancer using published performance characteristics of a multi-cancer test and prevalence of each cancer at ages 50, 60, or 70 based on 5-year incidence rates and corresponding 15-year probabilities of cancer death in the Surveillance, Epidemiology, and End Results registry, assuming 20% mortality reduction for each. Results UCT depends on screening age but is overwhelmingly determined by overall specificity of the test and is relatively insensitive to the types and number of cancers included. For a given overall specificity, UCT/CD is most favorable for higher-prevalence cancers (e.g., UCT / CD = 5.6 for breast+lung versus 6.5 for breast+liver at age 50). Under a common mortality reduction, UCT / LS is most favorable when the test includes higher-mortality cancers (e.g., UCT / LS = 48.5 for breast+lung versus 74.7 for breast+liver at age 50). Conclusions The harm-benefit tradeoffs of multi-cancer testing depend on the number and type of cancers included. Overall specificity is paramount for controlling unnecessary confirmation tests. For a given overall specificity, multi-cancer tests should prioritize prevalent and/or lethal cancers for which curative treatments exist.