A simple blood test can make a huge difference to cancer patients by evaluating whether the treatment to reduce tumours is working, say scientists.
British researchers have devised a way of closely monitoring the progress of tumours without the need for surgery or scans.
For the first time, they used blood analysis to detect deadly mutations in 20 of 38 patients studied with breast and ovarian cancer.
They even managed to build up a picture of how one woman's breast cancer responded to different treatments over 16 months.
The blood test, which could potentially save thousands of lives, could be available to patients within five years, according to the researchers.
Currently those with cancer often need to have tissue samples surgically removed for analysis.
The new test uses sequencing - a technique to read genetic code - to look for tiny pieces of DNA that cancer tumours shed into the bloodstream and are not picked up by current screening methods.
This could alert doctors when a treatment is not working, allowing them to try alternatives, as well as potentially sparing the patient unnecessary side effects.
Blood samples were taken from women either before, during or after chemotherapy at Addenbrooke's Hospital in Cambridge.
"This type of blood test has the potential to revolutionise the way we diagnose and treat cancer," the Daily Mail quoted Dr Nitzan Rosenfeld, who led the study at the Cancer Research UK Cambridge Research Institute, as saying.
"The great advantage is that it can be used to identify cancer mutations without surgery or a biopsy, making it safer and cheaper," he added.
DNA sequencing is widely used by scientists, but Dr Rosenfeld's team found a way to home in on key genes known to be 'cancer hot spots' and analyse them in detail.
This is the first time entire genes have been analysed from a blood test, which enabled the researchers to examine 20,000 possible mutations in six cancer-related genes.
Using a blood sample from a 56-year-old woman with breast cancer, they were able to see in retrospect that the first type of treatment she was given had not worked.
"We could see in a blood test what the doctors could only see using extensive imaging. It suggests that if we had looked at a different time point, we may have been able to spot it earlier," said Dr Rosenfeld.
He said more tests would be needed to see which other cancers could be detected using the same process, but there was no reason why it would not work if scientists targeted the right gene.
Their study has been published in the journal Science Translational Medicine.