Blood test detects 8 common cancers

One in three of us will have a brush with cancer at some point in our lives. And, regrettably, in the vast majority of cases, by the time we discover that we’ve got the disease, it’ll already be at an advanced stage, and often too advanced for a simple curative operation to remove it. Now things might be about to change. Chris Smith spoke to scientist Nick Papadopoulos, at Johns Hopkins in Baltimore, who has come up with a way to test blood samples for proteins and DNA markers that can pinpoint the presence of early-stage cancer in 70% of cases, and even reveal where in the body the tumour is...

Nick - The problem was the early detection of cancer. We believe that detecting cancer early gives a better chance for survival of the individual that has that cancer. The standard now is to let people present with symptoms, go to the doctor, and then the doctor has to treat situations which are like metastatic cancers and cancers of late stages. We think that this approach to cancer has to change, and we need to start from detecting cancers early which is much easier to treat, sometimes only with surgery.

Chris - That sounds like a very laudable aim to try to pick this up as early as possible… but how?

Nick - What we elected to do was to develop technology that can detect pieces of DNA and proteins from cancer cells that are present in plasma. Our plasma or blood has a lot of DNA and proteins but most of them come from normal cells, so our task was to try to detect those traces of DNA and protein that come only from the cancer cells.

Chris - So you’re looking for DNA that spills out of damaged or diseased cells that are cancerous or potentially becoming cancerous, that goes into the blood and you can pick that up and you’re looking for  proteins which are also spilling over from cancerous cells? So this is a blood test for cancer anywhere in the body?

Nick - That is correct, yes. We called it a multi-analyte because we use more than one marker.

Chris - How did you choose what DNA to go after or what are the DNA signatures, if you like, that are the hallmarks that there is cancer somewhere in the body and it’s getting into the bloodstream?

Nick - There are a lot of studies for the sequencing of the DNA from a lot of human cancers and that information distills down to a number of genes that are mutated. Those mutations are specific for cancer cells and not from normal cells. So what we did is we took all that information, we distilled it down to 16 genes, and 61 regions of the genome that are commonly mutated in cancer cells and not in normal cells and we used that.

Chris - How did you know where the cancer is though because you taken a blood sample from, for instance, the persons arm? Blood goes all round the body so could be seeing cancer cells anywhere so how do you know where their tumour is just on the basis of a blood sample?

Nick - Because we use the information both from the DNA that comes from the cancer cells and the proteins that come from the cancer cells, that, actually, is a lot of information that an algorithm that we’ve developed analyses and narrows it down to two different locations of where in the body it is.

Chris - How sensitive is this? How good are you at finding the needle in the haystack?

Nick - In this particular study we tested 1,005 individuals; they had eight different tumour types. When we analysed the blood from these individuals on average we were able to identify about 70% of the individuals that had cancers, so the sensitivity, on average, is 70%. We were better in detecting certain tumour types like ovarian, and not as good as detecting other tumour types. Also we were better at detecting a little bit later stage cancers than early stage cancers.

Chris - What about the false positive rate, because the last thing anyone needs is to be told by the doctor you might have cancer and, in fact, they’re perfectly healthy?

Nick - Yeah, thats what was, and still is, our concern. We wanted the specifitivity to be at least 99%. We detected seven false positives out of 812 people presumably free of cancer. As you said, it’s very important not to worry a lot of people that they have cancer while they don’t.

Chris - Are we far away from actually having a realisable test here with what you’ve done?

Nick - Obviously, I don’t know how far away it is. I hope that within the next few years the test will start being used but we have to be successful with the follow-up studies.