The cancer blood test making waves – and what the numbers really show

Progress in cutting the global toll of cancer remains painfully slow, but a new blood test has sparked unusual levels of hope. Researchers say it could one day make routine screening far more effective by catching cancers earlier, when treatment has the best chance of saving lives.

The Galleri blood test, developed by US firm Grail, is the latest entrant to attract worldwide attention, after early trial results were described as “exciting” by researchers.

A press release claims the test, currently being trialled by the NHS, can detect signals from 50 cancers and correctly identify the disease in 62% of people who receive a positive result.

It also appears to be highly accurate at ruling cancer out, with a reported 99.6% success rate among those who were disease-free. At first glance, these headline figures appear to represent a significant step forward.

But before we reach for the champagne, it’s worth looking more closely at what these numbers really mean. Early promise does not always translate into real-world performance.

The Pathfinder 2 trial, involving 23,161 people aged over 50 from the US and Canada with no prior cancer diagnosis, produced the figures now circulating widely. Of the 216 participants who tested positive, 133 were later found to have cancer, giving the “positive predictive value” (PPV) of 62% that has been so widely reported.

That metric answers a crucial question: “If I test positive, what’s the chance I actually have cancer?” It also means, however, that 38% of positive results were false alarms.

Specificity – how often a test avoids falsely diagnosing cancer – is equally important, given the anxiety and medical follow-up triggered by an incorrect result. Here, the test performed well: 99.6% of people without cancer received a correct negative result.

Yet even this strong number has implications. If everyone aged over 50 in the UK were tested – more than 26 million people – the same rate would still generate over 100,000 false positives.

What has been less widely discussed is sensitivity, the measure of how many true cancer cases the test actually detects. On this measure, the result was 40.4%, meaning the test missed around three in every five cancers that appeared over the following year.

Galleri test results:

A chart showing how many cancer cases the Galleri test would correctly identify. — The Galleri cancer test in numbers. John Ferguson, CC BY-SA

The figure that’s been less widely reported

That shortfall may disappoint those hoping for a catch-all screening tool. It also raises the risk that patients could be falsely reassured by a negative result, potentially delaying a diagnosis.

Statisticians caution that the reported PPV, specificity and sensitivity are estimates rather than fixed values, and each comes with uncertainty. They also note that tests often perform less well outside carefully controlled trials, meaning real-world accuracy could be lower.

So, where does this leave the Galleri test? It may well become a useful addition to future screening programmes, provided that negative results are not viewed as definitive by patients or doctors.

But the low sensitivity means many cancers would still be missed in its current form. The test is also expensive – US$949 (£723) in the US – and no evidence yet shows that widely using it reduces cancer deaths.

The early data is encouraging, but perhaps the excitement deserves to be tempered. This technology may be a step forward, but it is not a solution on its own.

This article is republished from The Conversation, a nonprofit, independent news organization bringing you facts and trustworthy analysis to help you make sense of our complex world. It was written by: John Ferguson, University of Galway

John Ferguson does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.