Multicancer screening tests: still the future?

Detecting cancers from simple blood tests was the future, but is it still? We think it is.

One of the great things about having a National Health Service is that we have enormous numbers of patients available on a single database on which to conduct coordinated clinical trials (with consent of course!) in a way most other countries can’t really do. The world has been watching with great interest as the latest of these large studies drew to a close recently, analysis of multi-cancer early detection test called Galleri by GRAIL. 

Galleri is a blood test that screens for signals of over 50 cancers by analyzing DNA fragments in the blood stream for specific chemical changes suggestive of cancer. It's commercially available in the US but not yet FDA-approved. It’s different to the Trucheck Intelli test offered commercially in the UK. 

Unfortunately the initial press release stated that the test "failed", causing GRAIL's stock to drop. However this conclusion may be premature and misleading: the test itself didn’t fail, the trial failed. The data haven’t been released yet and will be available for scrutiny later this year. However the trial asked a very specific question: could annual Galleri tests over three years reduce the number of people diagnosed with ‘advanced cancers’ (specifically stage III and IV cancers) compared to standard care? It appears the answer based on this trial is ‘no’. 

Within the data though there was a reported benefit in a high-risk group of 12 particularly lethal cancers, specifically showing a meaningful reduction in Stage IV diagnoses after the second and third annual screens. There was also an increase in early-stage (Stage I and II) cancers detected, which common sense would suggest averts the development of advanced cancers, So what if you do it for longer? What if you look at reducing just stage IV cancers? What about 10 year survival for those on the trial? There are many different ways of analysing these data and other ways to conduct meaningful trials, and there will likely be a much more detailed inquest into what actually was shown.

There is scepticism from some about these new screening tests - worry about overdiagnosis of cancers you would have died ‘with’ not ‘from’, false positives are possible (so indications of cancer being present when it really isn’t). However, many of the cancers for which they screen do not have existing screening programmes (particularly pancreatic and ovarian, two types of cancer that frequently present late and therefore have very poor survival statistics). Surely even a modest improvement is a major advance? 

I think these tests should not be declared a failure - the science is robust and barely questionable, we do need to have more data and more scrutiny though to determine the longer term benefit of screening with these tests.. The responsible move for now is to wait for the complete, peer-reviewed data. On a personal level, until a better solution appears on the market, I will continue to check my own Trucheck Intelli test every 3-5 years until I am age 50, and then every 1-2 years.