Australasian Science Magazine

By Ainsley Newson

Liquid biopsies promise early detection of cancer, but some of their current limitations risk being overlooked.

Many of the difficulties in efforts to reduce cancer incidence and deaths arise because we don’t yet have reliable ways to effectively identify, pinpoint and treat it in its early stages. But what if there was a blood test that could pick up a range of early cancers, show where tumours might be and suggest the best treatment? Sounds amazing, doesn’t it?

The exciting thing is that researchers are making some headway in identifying cancer-specific DNA sequences and proteins from blood samples: a so-called “liquid biopsy”.

The technology is based on the knowledge that cancer tumours shed mutated DNA. If an over-abundance of certain DNA sequences is detected in our blood, this suggests there could be a tumour. The hope is that one day there will be a single blood test that could be used to routinely screen anyone in the population.

One study published in Science in January showed promising results, including a false-positive rate (where a healthy person’s sample erroneously suggests cancer) of less than 1%. But because we don’t yet know enough about the required DNA markers, it was better at detecting some cancers (such as ovarian cancer) than others (such as breast cancer). This work is now being extended to a bigger trial.

It’s a lucrative area with a potentially enormous market. There is significant start-up funding being invested, and even more media hype. A US company even took a test to market a couple of years ago, although this was swiftly met with a stern warning letter from the US Food and Drug Administration.

But let’s not get ahead of ourselves. As the authors of another review paper pointed out in March, this test is by no means ready for population-wide roll-out. And even if we get to the point of a fantastic test that correctly identifies and pinpoints tumours within all scientifically important parameters, these may not necessarily be cancers that need to be found. Liquid biopsy could suffer the same problem that befalls many current cancer-screening technologies: while we are getting much better at finding tumours, we are much less good at working out which ones need to be treated and which can be left alone.

Overdiagnosis is said to occur when a test or screen makes a true finding (i.e. not a false-positive) but whatever was identified never needed to be found because it would not have developed further. Conversations about overdiagnosis following cancer screening are not easy. If we hear the word “cancer” our (understandable) reaction is to want it gone immediately.

However, overdiagnosis causes problems because people are given treatments that they don’t need, including treatments that cause harm and cost the health care system money (which in turn impacts access to care down the line). But, it is often hard to tell in advance who needs treatment and who doesn’t.

Overdiagnosis is being talked about by some commentators on liquid biopsy, but it’s notable in its absence from exuberant media releases extolling progress in liquid biopsy research. Some also think it’s their right to demand access to unproven tests or treatments, and see any move to prevent this as medical paternalism or denial of treatment. Yet those responsibly for publicly funded health care are likely to want greater reassurance before committing resources. At the same time, it is terrible when a progressive cancer is missed in a patient when the prognosis would have been better if it was found and treated early.

As a DNA-based technology, and given the significant hype that currently surrounds all things genetic, liquid biopsy may give populations a false sense of security. Its perceived precision may make public conversations about overdiagnosis difficult. We are conditioned to seek information rather than not, we search for our own health information online, and are usually keener to do something rather than “watch and wait”.

How good should a test be before it is introduced to the wider population? Should we discuss not testing? For how long should we measure patient outcomes before rolling a test out to healthy people in a world increasingly less willing to wait? Should we discuss not intervening?

Liquid biopsy won’t give us these answers, but the questions are sure to come around soon.