Ovarian cancer 'is not detectable' in the blood for years

In collaboration with the Press Association

A mathematical model designed by scientists in the US indicates that ovarian cancer may not be detectable in the blood using current technology until 10 years after the disease first starts to develop.

The researchers say their work could make future efforts to develop diagnostic blood tests more efficient.

The study by experts from Stanford University School of Medicine used models originally developed to predict the concentration of drugs injected into the blood as they move in and out of the bloodstream. The investigators linked these to other models of tumour cell growth.

They looked at a known ovarian cancer 'biomarker' called CA-125, and modelled how large a tumour needs to grow before it begins to shed CA-125 into the bloodstream.

A biomarker is a biological molecule found in the blood or other tissues that can be used to measure the activity of a cancer. Similar to how looking out the window for people with umbrellas is an indirect way of detecting rain, looking for cancer biomarkers can help doctors detect disease and plan treatment.

Lead researchers Sharon Hori and Sanjiv Gambhir believe the model used to calculate this 10-year gap from when ovarian cancer starts to develop and when it is detected is broadly applicable across all solid tumour types.

It has been known for some time that cancer patients have a better chance of survival the earlier a tumour is detected. But this research, published in the Science Translational Medicine journal, is the first study to mathematically model the size of a tumour and its related blood biomarker levels.

Many biomarkers for different cancers have been identified, but testing their accuracy as an early diagnostic tool can be a long and costly process. The new model could be used to speed up this process, and prompt new and accurate ways to detect ovarian and other cancers before they have a chance to spread to other parts of the body.

Dr Gambhir said: "This model could take some of the guesswork out of it. It can be applied to all kinds of solid cancers and prospective biomarkers as long as we have enough data on, for instance, how much of it a tumour cell secretes per hour, how long the biomarker can circulate before it's degraded and how quickly tumour cells divide."

He adds that identifying biomarkers that are made exclusively by tumour cells will also speed up the detection process.

Dr Laura McCallum, Cancer Research UK's science communications officer, said: "Detecting cancer at an early stage when treatment is more likely to be successful is one of the most promising ways to reduce deaths from the disease. Biomarkers have the potential to offer a simple, non-invasive way to detect cancer early and scientists, including our own, are working hard to find ones that can do this reliably.

"Mathematical models like this, designed to predict the most effective biomarkers, could help improve the bench to bedside success of such tests in the future."


  • Hori, S., & Gambhir, S. (2011). Mathematical Model Identifies Blood Biomarker-Based Early Cancer Detection Strategies and Limitations Science Translational Medicine, 3 (109), 109-109 DOI: 10.1126/scitranslmed.3003110