This Researcher Is Developing a Test for Early Detection of Many Cancers

Bert Vogelstein was just a med student in the 1970s when he encouraged his father to see a doctor about a weird lump on his cheek.

After a few tests, doctors diagnosed him with an aggressive salivary cancer. That lump that Vogelstein thought was a tumor? It was an enlarged salivary gland blocked by the cancer. If the inflammation hadn’t happened, his dad might have not have been diagnosed until it was too late for treatment. Instead, his father lived another 40 years. “He only survived because it was detected early.”

Vogelstein, a star cancer researcher at Johns Hopkins Kimmel Cancer Center, sees a future in which everyone can get a reliable early diagnosis and have a better chance of survival.

Vogelstein recently launched a company called Thrive Earlier Detection that will commercialize a simple blood test to screen otherwise healthy individuals for many forms of cancer—long before they display any symptoms. The test finds cancer by looking for genetic mutations and proteins using a combination of DNA sequencing, protein detection, and artificial intelligence.

The type of test, a liquid biopsy, is not new. Liquid biopsies are frequently used to identify the best course of treatment for cancer patients or as a way to check cancers in hard-to-reach places like the lungs.

But using liquid biopsies on seemingly healthy patients who show no symptoms of cancer is new. If successful, the tests hold the potential to become standard screening at annual exams—a non-invasive yet broad screening routine that could detect cancer long before the disease can spread.

Thrive’s first trial of 10,000 healthy patients is underway at Danville, Pa.-based Geisinger Health Systems. So far, Geisinger’s scientific team is encouraged by the data— but they say more tests are needed. “If it does succeed, it would be a huge breakthrough,” says David Ledbetter, Geisinger’s chief scientific officer. “Because we all know that the earlier you diagnose cancer, the better the outcomes.”

Cancer diagnostics, though, are riddled with cautionary tales of too much optimism and misdiagnoses. The PSA test, or the prostate specific antigen, was met with lots of excitement for its promise of detecting prostate cancer, but it wound up delivering both false positives and negatives. According to the National Cancer Institute, 25% of men receive a false-positive PSA test result that leads to a biopsy.

Debate has also emerged about mammograms—when to get them and whether the benefits outweigh the risks. Many screenings have led to unnecessary biopsies. According to the National Cancer Institute, 9.5% of tested women will have a false-positive exam, and at least half of U.S. women who are screened each year over 10 years will experience a false positive, with 7% to 17% of that group undergoing biopsies.

Launched last spring, Thrive is flush with $110 million from investors, led by Third Rock Ventures and has attracted an experienced executive team drawn from biotech firms like Foundation Medicine, which already offers liquid biopsy genomic tests for late stage cancer patients.

With fast-paced advancements in gene sequencing and artificial intelligence, it’s not surprising that Thrive is not the only company developing liquid biopsies in cancer detection.

In 2016 Illumina, a San Diego-based biotech company, spun off a startup called Grail to do the same. It’s pursuing its own blood test that could detect more than 20 kinds of cancers.

Backed by $1.5 billion from investors, Grail takes a very different approach to early detection of cancer. Rather than identifying genetic mutations in cells, Grail’s technology will use machine learning and data crunching to identify epigenetic methylation signals, or rather which parts are used by the DNA to be transcribed to a protein. “We found it’s these earlier signals that are more important,” says Alex Aravanis, Grail’s cofounder and chief scientific officer.

Grail worked with Harvard University and the Mayo Clinic on clinical studies and presented some of the initial findings to the American Society of Oncologists earlier this year. So far, its test returned false positive results less than 1% of the time.

Despite those milestones, plenty of issues still need to be sorted out for liquid biopsies to be widely used, says Dr. Richard Schilsky, an oncologist and chief medical officer for the American Society of Clinical Oncology. He says it’s not clear that the next generation of liquid biopsies would simply identify early-stage cancers. Rather, such blood abnormalities might instead indicate an aggressive cancer that’s already spread to the blood. Says Schilsky: “It’s going to take time and a lot of validation testing to sort out these issues.”

Liquid biopsies, Schilsky says, may be better suited for high-risk patients, such as those with the breast cancer BRCA1 gene, a smoker who has lung nodules indicated on CT scans, or cancer survivors at risk of new forms or reoccurrences. Broad testing of healthy populations naturally increases the odds for false positives, and those results can send patients down a rabbit hole of tests and anxiety.

Vogelstein says that while false positives are a legitimate concern, underdiagnosis is a bigger problem: 600,000 people died from cancer last year in the U.S. alone. “It would be a shame if fears about overdiagnosis limited the potential for reducing cancer deaths through early detection,” he says.

There’s plenty of work ahead. Both Grail and Thrive must work with the FDA on registrational trial design, which would drive clinical adoption, inclusion in guidelines, and reimbursement support from insurance companies and other payers. They must also train doctors on how to present results to patients and use the results to set up a course of treatment.

Vogelstein points out that these tests will not be 100% perfect. But the chance to identify cancer earlier would be an enormous win that could add years or decades to people’s lives.

“Virtually everyone who dies only dies because cancer is detected too late,” he says. “This is what I think about when I go to sleep and what I think about each morning.”

This Researcher Doesn’t Want Another Person to Die From a Cancer That Was Detected Too Late

More must-read stories from Fortune: