Do New Blood Tests for Cancer Meet the Right Standards?

Biotech startups worldwide are rushing to market screening tests that they claim can detect various cancers in early stages with just a few drops of blood. The tests allegedly will simplify cancer care by eliminating tedious scans, scopes, and swabs at the doctor’s office. 

The promise of these early detection tests is truly “enticing,” Dr Hilary A. Robbins from the International Agency for Research on Cancer of the World Health Organization in Lyon, France, told the Medscape German edition.

In an opinion article in the New England Journal of Medicine, she emphasized that the new cancer tests are much less cumbersome than traditional screening strategies for individual cancers. Moreover, they could enable the early detection of dozens of cancer types for which no screening has been available so far.

Meeting the Criteria

The problem is that these tests have not met the strict criteria typically required for traditional cancer screening tests. To be considered for introduction as a screening procedure, a test usually needs to meet the following four minimum requirements:

• The disease that the test screens for must have a presymptomatic form.
• The screening test must be able to identify this presymptomatic disease.
• Treating the disease in the presymptomatic phase improves prognosis (specifically, it affects cancer-specific mortality in a randomized controlled trial).
• The screening test is feasible, and the benefits outweigh potential risks.

“The new blood tests for multiple cancers have so far only met the second criteria, showing they can detect presymptomatic cancer,” Robbins wrote.

The next step would be to demonstrate that they affect cancer-specific mortality. “But currently, commercial interests seem to be influencing the evidence standards for these cancer tests,” said Robbins.

Inappropriate Endpoints?

Some proponents of such tests argue that, unlike for previous cancer screening procedures, initial approval should not depend on the endpoint of cancer-specific mortality. It would take too long to gather sufficient outcome data, and in the meantime, people would die, they argue.

Dr Eric A. Klein from the Glickman Urological and Kidney Institute in Cleveland, and colleagues advocate for alternative endpoints such as the incidence of late-stage cancer in an article published in Cancer Epidemiology, Biomarkers & Prevention.

“The concept would be,” they write, “that a negative signal would not indicate a mortality benefit, leading to the study being stopped. A positive signal, on the other hand, could result in provisional approval until mortality data and real-world evidence of effectiveness are available. This would resemble the accelerated approval of new cancer drugs, which often is based on progression-free survival until there postmarketing data on overall survival emerge.”

Klein is also employed at the US biotech start-up Grail, which developed the Galleri test, which is one of the best-known and most advanced cancer screening tests. The Galleri test uses cell-free DNA and machine learning to detect a common cancer signal in more than 5src cancer types and predict the origin of the cancer signal. Consumers in the United States can already order and perform the test.

An NHS Study

Arguments for different endpoints apparently resonated with the United Kingdom’s National Health Service (NHS). Three years ago, they initiated the Galleri study, a large randomized controlled trial to assess the effectiveness of Grail’s cancer test. The primary endpoint was not cancer-specific mortality, but the incidence of stage III or IV cancer.

The results are expected in 2src26. But recruitment was stopped after 14src,srcsrcsrc participants were enrolled. The NHS reported that the initial results were not convincing enough to continue the trial. Exact numbers were not disclosed.

The Galleri study deviates from the standard randomized controlled trial design for cancer screening procedures not only in terms of the primary endpoint, but also in blinding. The only participants who were unblinded and informed of their test results are those in the intervention group with a positive cancer test.

False Security

This trial design encourages participants to undergo blood tests once per year. But according to Robbins, it prevents the exploration of the phenomenon of “false security,” which is a potential drawback of the new cancer tests.

“Women with a negative mammogram can reasonably assume that they probably do not have breast cancer. But individuals with a negative cancer blood test could mistakenly believe they cannot have any cancer at all. As a result, they may not undergo standard early detection screenings or seek medical help early enough for potential cancer symptoms,” said Robbins.

To assess the actual risk-benefit ratio of the Galleri test, participants must receive their test results, she said. “Under real-world conditions, benefits and risks can come from positive and negative results.” 

Upcoming Trial

More illuminating results may come from a large trial planned by the National Cancer Institute in the United States. Several new cancer tests will be evaluated for their ability to reduce cancer-specific mortality. A pilot phase will start later this year. “This study may be the only one with sufficient statistical power to determine whether an approach based on these cancer tests can reduce cancer-specific mortality,” said Robbins.

For the new blood tests for multiple cancers, it is crucial that health authorities “set a high bar for a benefit,” she said. This, according to her, also means that they must show an effect on cancer-specific mortality before being introduced. “This evidence must come from studies in which commercial interests do not influence the design, execution, data management, or data analysis.”

This story was translated from the Medscape German edition using several editorial t