Randomized trials of multicancer screening tests: augmenting their ability to identify a genuine mortality benefit

Abstract

Randomized trials of the efficacy of multi-cancer early detection, by means of measurement of cell-free DNA and/or protein biomarkers in peripheral blood specimens, will attempt to document a difference in cancer mortality between persons assigned to intervention and control arms. Their ability to do so is limited by the relatively low rate of death from individual forms of cancer, the relatively low sensitivity of the tests currently being used, and the use of other cancer screening modalities among trial participants. However, if those same blood specimens also could be obtained from control arm participants in a given trial and then tested for the same markers, with results not known (or not made available) until the conclusion of follow-up for cancer mortality, it would be possible to compare mortality from given forms of cancer between test-positive individuals whose results were known and not known during the course of the trial. Such an analysis addresses the impact of a stimulus to offer targeted diagnostic testing, potentially leading to early treatment, against cancer mortality. Among persons who screen as positive, it should provide a relatively more sensitive means of gauging a possible mortality benefit resulting from multi-cancer screening.