Multi-cancer blood tests, with the promise of detecting many cancer types from a single sample, have the potential to transform cancer screening.
However, evidence is lacking to support broad use of the tests in people who do not have symptoms, according to research commissioned by the Agency for Healthcare Research and Quality and published last week in Annals of Internal Medicine.
Researchers identified no completed, controlled studies reporting benefits in cancer detection, mortality or quality of life from screening with multi-cancer detection tests. They also found insufficient evidence on the accuracy and potential harm of the tests, mostly due to study limitations and unknown or inconsistent findings.
Researchers from the RTI International – University of North Carolina Evidence-based Practice Center and The Ohio State University College of Medicine conducted the review.
“These are pretty new technologies, and there's a lot of excitement, but it does take a long time to do all the right studies to figure out if these are actually beneficial,” Daniel Jonas, an author of the review and director of the division of general internal medicine at OSU’s Wexner Medical Center, said in an interview.
MCD tests are intended to identify cancer before clinical signs or symptoms. The tests use biomarkers, including cell-free DNA, proteins or small molecule metabolites, to detect positive signals for different types of cancer.
No MCD tests have Food and Drug Administration approval for multi-cancer screening, but some are available in the U.S. as laboratory developed tests, a regulatory pathway that does not require FDA approval, the researchers said. In addition, no insurance providers cover MCD tests for screening.
The tests are sometimes recommended to supplement standard-of-care screening methods such as mammography and colorectal cancer screening. An oncologist may use an MCD test to try to detect recurrence in a cancer patient who has already undergone treatment, Jonas said.
The researchers analyzed 20 studies, with more than 109,000 participants, that reported accuracy for 19 different MCD tests. They found the evidence from those studies was insufficient to evaluate accuracy and harm.
Thirteen of the studies estimated accuracy in confirmed cancer cases and cancer-free control participants. Jonas said it’s not clear how informative that approach is because a test would not be used that way in the real world.
Seven studies used a better methodology, said Jonas, following asymptomatic people over time to assess whether MCD tests could detect cancer before symptoms appeared. In that subset, accuracy varied widely, with the tests correctly identifying cancer at rates ranging from 7% to 71%.
The review also found that false positives could lead to unnecessary diagnostic procedures and radiation exposure.
“Why we need a big study that's controlled is that there's also actually a lot of potential harm we could do,” Jonas said. “When we do any test, there's some rate of false positives, and people go on to have additional testing that they didn't need to have, and sometimes there'll be invasive procedures that happen that didn't need to happen because it was a false positive result.”
An ideal study of multi-cancer blood tests would assess a large number of people randomized into groups who received the MCD test or the usual care, such as screening for specific cancers, said Jonas. Participants would be followed over time to determine how many deaths were prevented.
Jonas said that while MCD tests show some promise, there is no proof they will ultimately be beneficial for screening the population at large.
“We certainly want to find out if they are,” he said, “and we need to do good studies to find out.”
