National Cancer Institute
Twitter facebook linked in Email

2011 Research News


Math Shows Limits of Cancer Early Detection Strategies

One of the central tenets of oncology is that cancers are most susceptible to treatment when diagnosed early, and the earlier the better. For example, 90 percent of women diagnosed with stage I ovarian cancer will survive five years or more, while fewer than 30 percent of those diagnosed with stage III disease will be alive in five years. This harsh reality has prompted a concerted effort to find molecules in blood – biomarkers – that would signal the presence of a tumor before it is detectable by imaging and long before it begins to spread throughout the body.

That search might be more difficult than initially thought given the predictions from a mathematical model developed by Sanjiv Gambhir of Stanford University and post-doctoral fellow Sharon Hori. In a paper published in the journal Science Translational Medicine, the two researchers present work suggesting that current clinical biomarkers for ovarian cancer are unlikely to be detectable until a tumor has been growing for at least a decade and contained almost 2 billion cells. Such a tumor would be 25 millimeters in diameter, about the size of an olive. Gambhir is a member of the University of Southern California Physical Sciences-Oncology Center.

The researchers tackled the problem of biomarker detection sensitivity by starting with a single ovarian cancer cell and modeling its growth using well-known parameters that describe the development of ovarian tumors. The model then calculated how much of a generic biomarker the tumor cells would shed into the blood stream as the cells multiplied and the tumor grew in size and what the blood concentrations would be over time. For the shedding rate, Gambhir and Hari used data available for a known ovarian cancer biomarker called CA125.

According to these calculations, it would take at least 10 years of growth to detect biomarkers shed at this rate using detection technology available today in clinical laboratories. They noted that to be detectable at an early enough time to be clinically useful, a biomarker would have to shed at levels 10,000 greater than CA125 or other known cancer biomarkers. By varying the parameters in the model, the investigators calculated that a 10-fold increase in biomarker shedding rate or a 10-fold decrease in assay detection limit could allow for the detection of a tumor only five millimeters in diameter that had been growing for nearly eight years. The researchers noted that while it is important that biomarker discovery efforts continue, there must be parallel efforts to improve the sensitivity of biomarker detection technologies.

This work, which is detailed in a paper titled, “Mathematical model identifies blood biomarker-based early cancer detection strategies and limitations,” was supported by the National Cancer Institute's Physical Sciences in Oncology initiative, a program that aims to foster the development of innovative ideas and new fields of study based on knowledge of the biological and physical laws and principles that define both normal and tumor systems. An abstract of this paper is available at the journal's Web site.
View abstract