Insurer Services Blog

Unwrapping Multi-Cancer Detection Technology: A CRL Evaluation of Today’s Tests

Written by Dr. Heather Fehling Ph.D. | Dec 19, 2025 3:00:03 PM

Dr. Fehling explores the current landscape of multi-cancer detection tests (MCEDs), examining their technology, performance, and implications for insurance medicine through a hands-on evaluation and comparative analysis. 

 

’Tis the season for shopping—and while everyone else is hunting for deals online, we’ve been filling our carts with multi-cancer detection tests (MCEDs). Think of it as holiday shopping for medical directors: unwrapping MCED technology one box at a time.  We wanted to share our review of tests (GRAIL Galleri, Exact Sciences Cancerguard, 20/20 Genesystems OneTest, EpiSeek, Guardant Shield) to understand how they perform and the implications for the future of understanding and risk assessment.

Why MCEDs Matter to Insurance Medicine

MCEDs promise earlier cancer detection and potentially improved long-term outcomes. For life insurers, they raise key questions around screening feasibility in asymptomatic adults, false-positive burden and downstream risk, predictive value in mortality modeling, and the impact on underwriting and contestability.

Despite rising consumer enthusiasm, clinicians remain consistent: MCEDs should not yet be used for screening asymptomatic populations due to limited early-stage sensitivity and uncertain clinical benefit.  Still, understanding this technology is essential as consumer use grows.

My Cancerguard Experience

As part of our “secret shopper” initiative, I took the Exact Sciences Cancerguard test. The process was streamlined—easy ordering, four tubes of blood drawn by venipuncture, and a consumer-friendly report.  But from a medical director’s perspective, the technical transparency was limited. Cancerguard’s clinician materials simply stated the test included multiple protein biomarkers as measured by automated immunoassay, and a PCR-based methylation panel.  While no specific protein list was provided, Cancerguard was built on CancerSEEK, which incorporated CA-125, CA19-9, CEA, prolactin, HGF, TIMP-1, osteopontin, and myeloperoxidase.  Many of these proteins behave more like broad inflammatory signals, similar in function to hsCRP for systemic inflammation.

Cancerguard reports 50+ cancers and subtypes. However, validation data cover 21 cancers—and the commercial test notably excludes breast and prostate due to inadequate early performance.

How Cancerguard Compares with Other MCEDs

To put Cancerguard in context, below is a comparison of the major MCED platforms, including tests I have personally analyzed through published datasets.


 

 

 

 

Stage-Specific Performance: The Core Limitation

While assays differ in methodology, specificity across commercially available MCEDs remains uniformly very high (~99%), which should limit unnecessary follow-up.  Sensitivity, however, is the limiting factor.  Across all tests, Stage I sensitivity remains low (<20%), with moderate improvement for Stage 2 (40-60%), and strong performance for Stage III-IV (70-100%). This reflects tumor biology: early lesions shed very little DNA or protein into circulation.  These statistics underscore why clinicians currently oppose using MCEDs for screening asymptomatic individuals, and why the MCED providers emphasize the need to continue routine cancer screening using gold standard methods.

Below is a summary of the MCED tests from a regulatory, cost and operational perspective:



 Where Does This Leave Insurance Medicine?

  • MCEDs are currently not population screening tests.  Early-stage sensitivity is insufficient, and false positives could trigger cascade testing.  
  • A negative MCED is not reassurance. At Stage I, most cancers go undetected by all current platforms.
  • Mortality prediction and actuarial utility remain unproven.  No evidence yet supports improved long-term outcomes through MCED use.
  • MCEDs may become relevant for post-issue management, not underwriting, especially for individuals seeking diagnostic clarity after symptoms arise.
  • Multi-omics platforms (e.g., EpiSeek, Guardant Shield MCD) may outperform older single-modality MCEDs, however they are earlier in the validation process and require independent real-world data.

Looking Forward

We will continue monitoring and evaluating these tests—both through personal secret-shopper experiences and through structured analysis using new AI-powered MCED evaluation tools.  MCEDs remain promising, but technologically young. For now, they function more like broad biologic signal detectors for later stage cancers than true early-stage cancer screening tools.

As for our holiday shopping?  We’ll keep unwrapping the science—one test at a time—and return anything that doesn’t quite fit. 
  

 

About the Author

Dr. Fehling has been with CRL since 2008 and is a certified High Complexity Laboratory Director by the American Board of Bioanalysis (ABB). Dr. Fehling serves as the Laboratory Director for CRL’s Molecular Diagnostics/Esoteric Laboratory. More broadly, Dr. Fehling provides scientific leadership across all business units at CRL with her extensive experience in laboratory management, new assay development, and business development.