Molecular Cancer Signal Localization in MCED Testing Minimizes Radiation and Imaging Burden Compared with Whole-body...

Efficient strategies for diagnostic evaluation of multicancer early detection (MCED) tests are needed to distinguish true-positive from apparent false-positive (FP) test results. Extending a previously published model, we compared 2 post-positive MCED test diagnostic evaluation strategies: a cancer signal origin (CSO)-guided approach with repeat MCED testing to resolve FPs (CSO-retest strategy) versus a whole-body imaging (WBI)-only strategy employing whole-body computed tomography (WBCT) followed by positron emission tomography (PET)/CT scans to resolve FPs. Analyses evaluated 2 scenarios with different test performance characteristics and specificities, with a number of imaging procedures required and radiation exposure for each strategy as outcomes measures. The CSO-retest strategy resulted in 10- to 20-fold reduction in the use of WBCT/PET/CT versus the WBI-only strategy across both scenarios. A 1% reduction in WBI-only test specificity from 99.5% to 98.5% led to a >3-fold increase in FPs and more than doubled WBI requirements. Estimated radiation exposure for initial diagnostic evaluations for the CSO-retest strategy were 0 to 26.1 millisieverts (mSv), whereas all initial diagnostic evaluations for the WBI-only strategy carried an obligatory exposure of 28 mSv. For the 50% of CSO predictions requiring initial imaging, dose exposure ranged from 0.28 mSv (mammography) to 26.1 mSv (triple-phase renal CT), or 1% to 93% of the dose required for the WBI-only strategy. The results suggest that CSO-guided diagnostic strategies incorporating targeted workups and MCED retesting markedly reduce radiation exposure compared with a WBI-only approach, and small changes in MCED test specificity lead to significantly more unnecessary workups.

Using MCED tests with molecular CSO prediction and MCED retesting to resolve apparent FPs enables targeted evaluations via existing diagnostic pathways, reducing radiation exposure compared with MCED tests without this feature that rely on WBI. These findings have important implications for MCED implementation including safety, cost, and accessibility.