A few patients have recently asked me about “liquid biopsy” tests. These are a new frontier of testing in which a single blood sample is analysed for evidence of multiple types of cancer. The idea is to analyze a single blood sample and – ideally – receive reassurance. However if you got an abnormal result, you’d then be referred for further testing to either confirm or exclude a diagnosis of cancer. This is much the same process as our current prostate cancer screening program in which a man has a PSA test as part of his annual physical to indicate the possibility of prostate cancer and then, in the event of an abnormal result, is referred for further workup (e.g. a prostate MRI or biopsy) to determine if he in fact does have cancer, or if he simply has a “false positive.” However, unlike with a PSA – which checks only for prostate cancer – a “liquid biopsy” would be able to test for dozens of different types of cancer at once.
It’s wonderful to imagine a (not so distant) future in which a blood test at your annual physical can screen you for virtually every known type of cancer (including ones for which we currently have no good tools for early detection such as brain, pancreas, or ovarian) and either give you reassurance that you are cancer free, or conversely pick up a cancer in the very earliest stages and offer a much better chance for cure.
There are a number of such tests under development (the GALLERI test from Grail, and CANCERGUARD from the makers of Cologuard are two of the better known ones), and, as you can imagine, these are igniting great interest in both doctors and patients. And while none of these tests are yet to be FDA approved – nor to be covered by any type of insurance – a few are now commercially available for those who can pay out of pocket for them. Not surprisingly then, a few patients have asked me recently whether these are worth having.
To be honest, I don’t have a perfect answer to this question. Unfortunately these tests are too new, and the studies on them too small, to be able to confidently state just how much benefit one might get out of them. So I thought I’d take one study on one of the above tests (the GALLERI test) and break it down a bit in the interest of helping people make a more informed decision.
In the study (which you can read by clicking here) researchers took blood samples from 6,662 apparently healthy subjects and performed a GALLERI test on them. The results were as follows:
6,235 of those tested received a negative result and did not develop cancer over the next 12 months (“True Negatives”)
86 patients received a negative result, but within the next 12 months turned out to be diagnosed by other means with a cancer of some type (“False Negatives”)
35 patients received a positive result and, upon further workup, turned out to be diagnosed with cancer (“True Positives”)
57 patients received a positive result and, upon further workup, were found not to have cancer (“False Positives”)
A little further thought applied to the above reveals that of the 121 patients who developed cancer within the following 12 months, only 35 of were found to be positive on the GALLERI test. In other words, 75% of cancers were missed by the test.
Conversely, more than half of those patients who did get a positive result turned out not to have cancer. And while most probably suffered nothing other than some anxiety while awaiting their final diagnostic results, it’s hard to measure more subtle harms that some of these patients might have experienced as a result (radiation exposure from CT scans that might increase their risk for cancer later in life; biopsies that can lead to infection or other complications). In a study this small, it’s likely that nobody really got “hurt” from the false positives, but if we were to roll this test out to tens of millions of people tomorrow, it’s almost certain that somebody, somewhere, would get killed as a result of a false positive.
So while many hype the benefits of liquid biopsies, this study suggests that their performance isn’t in fact that awesome yet (I’m highlighting just one study with the GALLERI test here, but a recent review in the prestigious Annals of Internal Medicine found similar poor performance from many of GALLERI’s competitors).
Moreover, there’s a really important question – the only important question, really – that the above study doesn’t answer. Namely: does this test in fact save lives?
This might seem like a silly question to even ask, but consider the possibility that of those 35 patients who were diagnosed with cancer as a result of the test, how many actually had a better outcome as a result? How many died anyway either because we don’t have a good treatment for the type of cancer that they had, or because the test didn’t detect the cancer until it was big enough to be untreatable? Conversely, how many of those patients would have done just fine even if they had not received their cancer diagnosis until later (some cancers are very curable even when not caught until a late stage)?
What is really needed is a randomized trial that administers the GALLERI test to large number of people, follows them for several years, and gathers data on whether there are fewer cancer deaths in the group being tested. Fortunately, just such a study is being carried out right now in the UK. I am not alone in eagerly waiting for the results of that research.
To be clear, I am not arguing that nobody should get GALLERI (or one of the other similar tests). And in fact, I’m extremely hopeful that these kinds of tests will eventually revolutionize the detection of cancer and lead to fewer cancer deaths.
I do however want my patients to make informed choices. It’s so tempting to think “there’s a test that can help me find cancer early. All I have to do is shell out some money to have it. Why wouldn’t I do that?” And given how cancer strikes a special type of fear into most people’s hearts, this is very understandable. I myself have lost several loved ones to this horrible disease and would gladly spend a bit of extra money to improve the odds that myself or those I care about don’t suffer the same fate.
But for the time being, I remain skeptical that these tests are accurate enough to be worth the money. I sincerely hope – and am cautiously optimistic – that in a few years, after the tests have been better refined and more research has been done, I will be able to take a more favorable view.
