Research ArticleCancer Modeling

Mathematical Model Identifies Blood Biomarker–Based Early Cancer Detection Strategies and Limitations

Science Translational Medicine  16 Nov 2011:
Vol. 3, Issue 109, pp. 109ra116
DOI: 10.1126/scitranslmed.3003110

You are currently viewing the editor's summary.

View Full Text
As a service to the community, AAAS/Science has made this article free with registration.

The Early Bird Doesn’t Always Catch the Worm

The oft-quoted dictum “timing is everything” has special meaning in the realm of cancer diagnosis. Early detection can greatly improve patient prognosis: The longer the cancer goes unnoticed, the bigger it may get and the greater chance it has to spread, making treatment much more difficult. Because patients may not notice any symptoms until the cancer has already progressed, there’s been a huge push to identify cancer biomarkers. Ideally, these biological hallmarks may indicate the presence of disease well before symptoms appear.

Hori et al. now develop a mathematical model to predict how early doctors can catch disease by screening blood for cancer biomarkers. Starting with a single ovarian tumor cell, the authors modeled the release of cancer-related molecules (putative biomarkers) into the bloodstream (“shedding”) on the basis of well-characterized tumor growth and shedding rates for ovarian cancer. Hori et al. found that the tumor could grow for more than 10 years—to about the size of an olive—before it was detected by current clinical blood-based biomarker assays. These data suggest that, although identifying new biomarkers is critical for early cancer diagnosis, the sensitivity of biomarker detection must be optimized further if it is to have an effect on disease outcome in the clinic.