The breast cancer genes BRCA1 and BRCA2 are the two of the famous sequences of DNA in the world. In the 90s, their discovery upended cancer research, kicking off a frenzy to find other genes linked to cancer. That in turn kicked off another frenzy: of companies selling kits that test for ever-more newly discovered cancer genes. Today, the cancer genetics startup Color Genomics is announcing that it will expand its test for breast and ovarian cancer genes to include a total of 30, including genes related to pancreatic, stomach, colon, and prostate cancer.
Color stands out among the field because of its price: $249. That’s cheap enough to leapfrog insurance companies, the traditional gatekeepers of genetic tests. Insurance typically only covers genetic tests for cancer—which can run to over a thousand dollars among Color’s competitors—if a patient has risk factors like family history of cancer at a young age. Color also helpfully refers would-be customers to doctors to order the test. “Historically the test has been very expensive, and the process to get testing meant jumping through lots of loops,” says president and cofounder Othman Laraki. “We want for access to no longer be a barrier.”
If that sounds like Silicon Valley disruptspeak, consider the world Color comes from. Before founding Color, Laraki worked for Google and Twitter. So did another cofounder and CEO Elad Gil, who is also an angel investor. (Gil also has a PhD in genetics.) On top of that, the Bay Area-based company has partnered with hot startups like Slack, Medium, and Stripe to offer subsidized Color tests as an employee benefit.
It’s easy to nod along: More access, more genes, more tests all sound like much-needed defenses in the interminable war against cancer. But doctors caution that our ability to sequence DNA—to build machines that can manipulate DNA samples at scale and write code that can assemble gene sequences—has far outpaced our ability to understand how those genes cause cancer. Most of the 30 genes on Color’s panel or the 25 genes on Myriad’s panel or 32 genes on GeneDx’s panel? They don’t matter enough to be useful.
“If you talk to docs, they say, ‘BRCA, that’s the only thing I’m interested in because I don’t know what to do with the other information,’” says Timothy Hamill, former director of UCSF Clinical Laboratories. “Doctors don’t know want to do with it. Patients don’t know what to do with it.”
The Murky World of Cancer Genetics
BRCA1 and BRCA2 turned out to be outliers among cancer genes. Their effect on risk is dramatic: 45 to 65 percent of women with a BRCA mutation will develop breast cancer in their lifetime, compared to 12 percent of women in the general population. But these two genes account for only 5 to 10 percent of all breast cancers.
The vast majority of breast cancers—and indeed, the vast majority of all cancers—are the result of some combination of environment, lifestyle, and genes. Take the example of the gene ATM, says James Evans, a cancer geneticist at the University of North Carolina. ATM might elevate a woman’s lifetime risk of breast cancer from 12 to 24 percent, and it’s included in many cancer panels, including Color’s. So if a woman with a family history of breast cancer finds out she carries an ATM mutation, does that completely explain her family history? Should her sisters and daughters get tested, too? Should she be offered a preemptive mastectomy? The answer with BRCA is usually yes. But with the ATM’s modest increase in risk, it’s not so obvious.
When I asked Laraki how his company selected genes for the panel, he said, “We work very closely with researchers who specialized in each of these major cancers”—indeed, BRCA1 pioneer Mary-Claire King is an advisor to Color—”to select genes where a) the science is very well established and b) the impact of the mutations in the genes is significant.” But the devil is in the details. Genes are thousands of letters of DNA long, and a mutation in one letter can have no effect while a mutation in another can render the gene useless. Even with the most well-studied genes like BRCA1 and BRCA2, tests can turn up what geneticists call “variants of unknown significance.” Color, to its credit, requires customers to get their test results through a doctor or genetic counselor, who can explain the results—but you can only explain so much about a variant nobody knows anything about.
Evans is also skeptical of the prostate cancer genes in the expanded panel. “The risk associated with those genes are not really clear,” he says. In fact, a footnote in Color’s own press release says, “Note that research on genes associated with hereditary prostate cancer is still in its early stages.”
So why include prostate cancer at all? Well, here’s something to consider: Color’s original breast and ovarian cancer was primarily geared toward women. It can double its potential market by selling to men.
Tests for All
Color is hardly the only company selling customers on a large panel of genes with unclear clinical value. “That’s true of all companies, except Color is marketing more aggressively to the population” says William Foulkes, a cancer geneticist at McGill. In fact, it is marketing to a different population than the one that usually gets tested for cancer genes. By setting a price low enough to cut out insurance companies that restrict who gets these tests, Color can sell to anyone—whether or not they have a family history of cancer.
This is an important distinction, because cancer genes have typically been studied in families with a history of cancer. Color touts stories of women who found BRCA mutations despite no known family history. But does a new BRCA variant confer the same amount of risk as one that has a documented history of causing cancer? “I think the jury’s still out on the what the risk is,” says Evans. In 2013, the US Preventative Services Task recommended against BRCA testing for women with no family history of it.
Things could change. Evans says he’s actually “enthusiastic” about the idea of testing the general population for certain genes, but he wants to see the evidence first that it’s a net good. “It sounds like I’m an academic. I want to say, ‘Study it, study it, study it,’” he continues. “But it’s true, we need to study it.”
Academics have reason to be wary. The history of cancer is littered with overzealous interventions—frequent mammograms, prostate cancer screening—that had to be walked back once researchers realized they were causing more harm than good. Those tests often picked up breast and prostate cancers that would have never harmed the patients, who nevertheless endured extra tests and extra surgeries, not to mention extra stress, confusion, and fear.
Color in fact has its fingers in research on whether broad cancer gene screening can lead to better outcomes. At the University of California, San Francisco, Laura van’t Veer uses Color’s test to screen women in the 100,000-patient ATHENA Breast Health Network. She’s optimistic about adding another tool to the toolkit. “If healthy women get this test provided as part of their screening,” she says, “then we can optimize mammography screening tools and maybe reduce it for everyone who has very low risk.” But it will take years before that data can lead to clear changes in prevention and treatment.
That’s not so compatible with startup time scales. The same argument about going slow, waiting out the uncertainty was being made about BRCA1 and BRCA2 just a few years ago, says Laraki. Color is moving relentlessly forward. Even if it can’t offer clarity on some cancer gene mutations, it promises to update customers as the research becomes available. Eventually, it can get there—though it could be decades rather than years or months. For now, it all depends on your appetite for uncertainty.
Barbara Jacoby is an award winning blogger that has contributed her writings to multiple online publications that have touched readers worldwide.