ctDNA-Based Pan-Cancer Screening Tests Achieve Over 99% Specificity but Sensitivity Remains Stage-Dependent
This review evaluates emerging multi-cancer early detection tests like CancerSEEK and GRAIL that use circulating tumour DNA to screen for multiple cancers from a single blood draw. While specificity exceeds 99%, sensitivity varies from approximately 40% in stage I to 80% in stage III disease. The author cautions that most studies to date have used diagnosed cancer patients rather than asymptomatic screening populations, and prospective randomised trials are needed to demonstrate clinical utility.
The original study
Circulating tumor DNA (ctDNA) as a pan-cancer screening test: is it finally on the horizon?
- Authors
- Duffy MJ, Diamandis EP, Crown J
- Journal
- Clinical chemistry and laboratory medicine
- Type
- Journal Article, Research Support, Non-U.S. Gov't, Review
- PMID
- 33856748
Original abstract
The detection of cancer at an early stage while it is curable by surgical resection is widely believed to be one of the most effective strategies for reducing cancer mortality. Hence, the intense interests in the development of a simple pan-cancer screening test. Lack of sensitivity and specificity when combined with the low prevalence of most types of cancer types in the general population limit the use of most of the existing protein biomarkers for this purpose. Like proteins, tumor DNA also can be released into the circulation. Such circulating tumor DNA (ctDNA) can be differentiated from normal cell DNA by the presence of specific genetic alteration such as mutations, copy number changes, altered methylation patterns or being present in different sized fragments. Emerging results with test such as CancerSEEK or GRAIL suggest that the use of ctDNA can detect cancer with specificities >99%. Sensitivity however, is cancer type and stage-dependent, varying from approximately 40% in stage I disease to approximately 80% in stage III disease. It is important to stress however, that most of the studies published to date have used patients with an established diagnosis of cancer while the control population were healthy individuals. Although the emerging results are promising, evidence of clinical utility will require demonstration of reduced mortality following evaluation in a prospective randomized screening trial.