A new blood test that could ‘help detect eight common cancers before they spread’ is big news today.

There’s lots of excitement around the study behind the headlines, published in Science. And rightly so. It marks an important next step for scientists working on blood tests for cancer.

But some important details were skipped over in the discussion in the news.

The first thing to know is that this experimental test is a long way off being used to diagnose cancer. As some experts have said, it’s “promising but with several caveats and a significant amount of further research is needed”.

And that’s because using a blood test to confirm that people with cancer have cancer is very different to using that same test to detect the earliest hints of cancer in otherwise healthy people.

What’s exciting about a cancer blood test?

Finding an accurate blood test for cancer is the ‘Holy Grail’ for many researchers. Taking blood is a lot quicker, cheaper and less invasive than some other tests.

Cancer cells are different to normal cells in many ways – they have changes in their DNA and make different molecules. And in some cases, cancer cells shed these fragments of errant DNA and faulty molecules into a patient’s bloodstream, theoretically leaving them floating for detection by a blood test.

But so far, finding the right chunks of DNA or molecules to look for has proven tricky. It’s hard to find a marker that is sensitive enough – meaning it finds cancer every time – and specific enough – meaning it doesn’t ‘find’ cancer when it’s not there.

Developing a test that worked would mean patients could be diagnosed quickly, potentially sparing them anxiety when they don’t have cancer and speeding up treatment when they do have cancer.

And it could help find cancers earlier when they’re likely to be easier to treat.

So while this research is an important step in the right direction, it’s not yet the finished article.

As Professor Paul Pharoah, a cancer expert at the University of Cambridge, said: “it remains a promising, but yet to be proven technology.”

What did the research find?

The study, from researchers at Johns Hopkins in the US, shows how their blood test can confirm the presence of 8 types of cancer (ovarian, liver, stomach, pancreatic, oesophageal, bowel, lung, and breast) in people already diagnosed.

Their test looks for faulty DNA and molecules from cancer cells that find their way into the blood. And they tested it on 1,005 patients who they already knew had cancer.

This is a key point. They were looking for cancer in patients who had already been diagnosed with cancer. To truly match the promise of a blood test for screening, as some news reports suggested this might be, the test would have to work in people who haven’t yet been diagnosed.

Across the 8 cancers the tests picked up cancer in an average of 70 in 100 cases. This is a measure of the test’s sensitivity. And that’s not bad for an experimental test.

But there was huge variation in sensitivity across the different types of cancer – it found 98 in 100 ovarian cancers, but only 33 in 100 breast cancers.

Interestingly, the test could also tell the researchers the organ or tissue where the cancer came from in 83 in 100 patients. This would be incredibly useful for doctors if this type of test were to one day be used to diagnose cancer.

Finding cancer when it’s there is one piece of the puzzle. Another is making sure you don’t ‘find’ it when it isn’t there – known as false positives.

So the US team ran the test on 812 healthy people. It came up positive in 7 of these people, which, for a test at this stage of development, is a pretty low false positive rate of around 1% and definitely a good sign for the specificity of this test.

What is the research missing?

Cancers are split into different stages – the lower the stage, the earlier the cancer. This means it’s likely to be smaller, hasn’t spread and is probably going to be easier to treat. Any test with the goal of being able to detect cancers at the earliest stages in healthy people must strike the right balance between sensitivity and specificity.

“The sensitivity of the [blood] test in stage I cancer is quite low, about 40%, and even with stage 1 and 2 combined it appears to be around 60%,” said Dr Mangesh Thorat, a cancer expert at Queen Mary University of London. “So the test will still miss a large proportion of cancers at the stage where we want to diagnose them.”

This is really important for a successful test – the earlier it can pick up cancer the better. If it’s only good at finding later stage cancers, these are harder to treat.

Also, the proportion of common cancers detected (such as breast, lung or bowel) isn’t as high as with other, rarer cancers included in the study.

This may mean that a screening programme has to test a very large number of individuals to detect one cancer,” said Thorat.

This increases the chance of false positives, potentially causing anxiety and further unnecessary testing.

Professor Nicholas Turner, a cancer blood test expert at The Institute of Cancer Research, London, and The Royal Marsden NHS Foundation Trust, said the 1% false positive could be a concern if a test like this were to be used to screen everyone, including healthy people.

There could be a lot of people who are told they have cancer, who may not have it,” he said. “We can only learn about how much this is, or is not, a problem through larger studies. Much larger studies will also be needed to address whether using these blood tests can help improve survival rates.”

Any new test also needs to be better than the ones we have already. For example, with bowel cancer, there is already an effective screening programme to help detect the disease at an early stage. It can also prevent bowel cancer developing in the first place.

And while the experimental blood test in today’s news looks like it does a good job of predicting where a cancer is growing, in about a quarter of cases it may not be able to do so.

“Lots of imaging may be required to further diagnose the cancer, and we may still fail to identify the site, which creates an issue of what to tell these otherwise healthy individuals,” said Thorat.

What should you take away from this?

This is really important research, deserving of bigger studies. We need to do more research like this.

Dr Richard Marais, director of the Cancer Research UK Manchester Institute, said: “Hopefully in the future we can develop a blood test such as this that can be used routinely in patients, perhaps once a year, to allow them to know earlier if they have cancer and to get them treated much sooner.”

But as Turner added: “The study does not support use of the blood test outside research studies, as it has not shown yet whether the blood test has the characteristics required for population screening.”

With further research, the hunt for this scientific ‘Holy Grail’ will continue. And larger studies will test these tests under the right conditions to know for sure if it can be used to detect cancers earlier.


Cohen, J.D., et al. (2018) Detection and localization of surgically resectable cancers with a multi-analyte blood test. Science