At present, oesophageal cancer has a terrible prognosis. Survival rates for people diagnosed with this disease – which affects the oesophagus, sometimes called the food pipe – are as low as 15% after five years in the UK. This means that 85% of those diagnosed with oesophageal cancer die within five years.
The main cause of this mortality is that this type of cancer is typically only detected at very late stages. People tend to have only vague symptoms to begin with, such as pain, nausea, weight loss or persistent heartburn, until they begin to have problems swallowing. At this point the cancer is likely to have spread to other sites and there is little hope of successful treatment.
A critical factor in any cancer survival is early detection and surgical removal (this is the only real “cure” for cancer, as there is nothing left to spread). But if we want to detect more cancers at an early stage in asymptomatic patients then we need a radical rethink, using non-invasive approaches, like blood tests.
My research group – and others around the world – have been looking into the potential of blood tests to detect different types of cancers, with some promising initial results. Also known as “liquid biopsies” it is hoped that these blood tests could be developed to identify several different types of cancers when they are in their early stages. And, as shown in our newly published pilot study, we have now used a blood test to correctly identify 72% of patients diagnosed with oesophageal cancer from a mixed cohort of 300 people with and without the disease.
We had previously focused on finding new ways to diagnose early oesophageal cancers using tissue obtained during an endoscopy. Also known as “a magic eye”, this procedure involves a long thin tube with a camera at the end being inserted into a person’s oesophagus via the mouth, allowing tissue (biopsies) to be retrieved. We had been examining the molecules in the collected cancer tissue to identify alterations that are typically seen in cancer, but not found in normal oesophageal tissue.
Certain molecular alterations in biopsy tissue, including DNA mutations and over-expression of certain proteins, among other indicators, are thought to be unique to the disease and so could be signs of cancer in its early stages. However, the major flaw with looking at these so-called cancer biomarkers is that it requires some oesophageal tissue to be collected in the first place. But as few people with this type of cancer will show symptoms, their tissue would not be collected nor investigated. So instead, we began developing a new blood test that has the potential to detect early oesophogeal cancer.
The test that we have been trialling measures DNA mutations in red blood cells which lead to changes in the surface proteins of the cells. While we (and other researchers) have shown that these mutated cells are present in the blood of healthy individuals at a level of about three mutant cells per million red blood cells, we have now found that oesophageal cancer patients carry about 10 mutant cells per million. This threefold rise in the number of mutant cells is the basis on which we work out which of the people tested have cancer.
Specifically, we look at how fluorescently tagged antibodies bind to red blood cells. A single mutation in a key gene – called PigA – removes the ability of the cell to present cell surface proteins (called CD55 and CD59). Mutated cells don’t bind to these antibodies so they can be distinguished from the abundant normal cells.
The mutations in the red blood cells do not cause the cancer but are a byproduct of it. We have likened this approach to a cancer smoke detector. Household smoke detectors warn of fire, but actually measure smoke. Our cancer test warns of cancer but actually measures blood cell mutations not the cancer itself.
While these results are promising, it must be noted that the test is still in its early days and has not been approved for use by doctors and health professionals. In the future, less-invasive tests such as the one we have developed may identify “at risk” cancer patients early, and help select patients to receive close specialist follow up. In this way we may be able to identify many early cancers before they spread and become inoperable. This could have a major impact on survival rates from a range of cancers in the future, including oesophageal.
Our cancer blood test proved to be fairly simple to administer, and is also far less expensive than some others. For example, one test developed in the US to detect a range of cancers costs around $500 to use, ours costs only £30. We hope this could herald a new chapter in cancer diagnosis and ultimately, as we used such small amounts of blood to confirm the cancer diagnosis, we may even see this analysis being performed with just a single drop of blood.