But scientists have faced challenges with the technology. Identifying where a cancer comes from is scientifically complicated, though at least one company is using machine learning to solve that. And although early research shows that some private companies are finding success, many tests still struggle with accuracy.
Last week, President Biden said he wants to foster research on these tests through his cancer “moonshot” initiative. He touted a government-funded clinical trial that will study the efficacy of multiple types of early-screening tests. The hope, Biden said, is that a tool comes out that can help halve cancer deaths in the United States within 25 years.
“Imagine a simple blood test during an annual physical that could detect cancer early,” he said.
In the United States, nearly 600,000 people die of cancer every year, according to the Centers for Disease Control and Prevention. Detecting cancer early is one of the best ways to save lives, experts said. But few cancers have tests that can do so, except for a few in places such as the breasts, prostate and lungs.
A handful of companies have entered the space to detect more cancers. Among them is Grail, a Silicon Valley-based biotechnology start-up that has developed the Galleri test, one of the furthest along, biotechnology experts said.
The company’s test works off the basic principle of finding DNA that tumor cells release into the bloodstream as they die and replicate. The Galleri test spots markers on DNA shed by tumor cells and feeds that data into a machine-learning algorithm that can detect whether cancer is present and in which organ, said Josh Ofman, Grail’s president.
The company says its test can identify more than 50 cancers early. The tests are not approved by the Food and Drug Administration, though they are actively seeking it, Ofman added. Most insurances don’t cover the test, but people can purchase it for a hefty $949 if they have a prescription.
Grail conducted an initial study in which it screened more than 6,600 people over age 50 for cancer with its test. It caught cancer in 35 people, and in 71 percent of those cases the cancers were not ones for which there is routine screening. Fifty-six healthy blood samples were incorrectly identified as cancerous.
The Biden administration’s initiative, run by the National Cancer Institute, plans to examine how effective blood tests are in identifying cancer early. It will look to enroll 24,000 patients ages 45 to 70 starting in 2024 for a four-year pilot study. This will lay the groundwork for a larger trial that aims to enroll 225,000 people, the White House said.
Which tests will be in the study have not been finalized, Ofman said, adding that Grail would be happy to collaborate.
Salil Garg, a clinical investigator at MIT’s Koch Institute for Integrative Cancer Research, said if tests can be perfected, they would become valuable as regular screening tools for patients. He noted that despite the attention being put into the space, there are several challenges to make this technology ready for the masses.
A positive result doesn’t necessarily mean there’s cancer present or there’s a mass that will turn into cancer, he added. Pinpointing where the cancer comes from is difficult because some cancers share similar DNA mutations or might not shed any DNA into the bloodstream at all, making it difficult for the tests to spot them.
“The open question is: In what context is this going to be useful information?” Garg said. “Very useful information versus not as helpful.”
