A multi-cancer early detection blood test using machine learning detects early-stage cancers lacking USPSTF-recommended screening

Janet Vittone; David Gill; Alex Goldsmith; Eric A Klein; Jordan J Karlitz

doi:10.1038/s41698-024-00568-z

A multi-cancer early detection blood test using machine learning detects early-stage cancers lacking USPSTF-recommended screening

NPJ Precis Oncol. 2024 Apr 17;8(1):91. doi: 10.1038/s41698-024-00568-z.

Authors

Janet Vittone¹, David Gill², Alex Goldsmith³, Eric A Klein⁴, Jordan J Karlitz⁵

Affiliations

¹ Mayo Clinic, Rochester, Minnesota, USA.
² Intermountain Healthcare, Salt Lake City, UT, USA.
³ Colorado Center of Medical Excellence, Denver, CO, USA.
⁴ GRAIL, LLC, Menlo Park, CA, USA.
⁵ GRAIL, LLC, Menlo Park, CA, USA. [email protected].

Abstract

US Preventive Services Task Force (USPSTF) guidelines recommend single-cancer screening for select cancers (e.g., breast, cervical, colorectal, lung). Advances in genome sequencing and machine learning have facilitated the development of blood-based multi-cancer early detection (MCED) tests intended to complement single-cancer screening. MCED tests can interrogate circulating cell-free DNA to detect a shared cancer signal across multiple tumor types. We report real-world experience with an MCED test that detected cancer signals in three individuals subsequently diagnosed with cancers of the ovary, kidney, and head/neck that lack USPSTF-recommended screening. These cases illustrate the potential of MCED tests to detect early-stage cancers amenable to cure.