Skip to main content
Premium Trial:

Request an Annual Quote

Oncology Community Sees Potential for Patient Harm in FDA OK of 23andMe BRCA Test


NEW YORK (GenomeWeb) – Since 2013, when 23andMe had to stop marketing health-related test reports due to regulatory action from the US Food and Drug Administration, the consumer genomics firm has detected but not reported one of three specific mutations in the BRCA1 and BRCA2 genes in nearly 3,000 of its customers.

These individuals can now decide if they wish to find out their status for these BRCA1/2 mutations after 23andme this week won FDA authorization to report the test results directly to consumers (DTC).

Individuals with BRCA1/2 mutations are at increased risk for breast, ovarian, and prostate cancers compared to the general population. However, the three mutations reported by 23andMe's test show up in 2.5 percent of Ashkenazi Jewish people but only up to 0.1 percent of individuals from other ethnic groups.

Despite FDA's blessing, many in the cancer community are concerned that DTC access for 23andMe's BRCA test has the potential to harm patients who don't understand its limitations, mainly that the SNP genotyping test detects only three mutations in two highly variable genes.

Though not all BRCA1/2 variants are cancer causing, some 5,000 mutations have been identified in these genes, according to the patient advocacy organization Facing Our Risk for Cancer Empowered. "We're not thrilled," said Lisa Schlager, VP of community affairs at FORCE. "We're concerned about confusion and misunderstanding about whether people are a true negative [for BRCA mutations]."

Since the three mutations 23andMe reports show up so rarely in the US population, Schlager and others in the cancer community worry that most individuals will get a negative result, and some will assume that they don't have a genetic risk for cancer when they do. Those who miss an opportunity to find out they have a genetic risk for cancer may not benefit from more frequent screenings or get the chance to have risk-reducing surgery. Their relatives, who might inherit the same mutation, may also not learn their risks in a timely manner.

"We're of the mind that if you're going to do BRCA testing, and you don't have a known mutation in the family, then the broader testing makes much more sense," Schlager said.

"I do think screening like this will have some consequences," said Alberto Gutierrez, who was director of the Office of In Vitro Diagnostics and Radiological Health within FDA's device division and who signed the 2013 warning letter to 23andMe that would end its ability to sell genetic health risk reports DTC. With this latest authorization, Gutierrez said the agency did the best it could to craft labeling language informing the public of the test's limitations and telling users to seek the advice from healthcare professionals to guide follow-on decisions.

23andMe, meanwhile, highlighted that its BRCA1/2 selected variants report is the first DTC cancer risk test to garner FDA's blessing. CEO Anne Wojcicki wrote in a blog post that the authorization gives individuals who don't know their Jewish ancestry, who lack family history of cancer, and who wouldn't have known to get tested for BRCA1/2, a way to learn their status.

Wojcicki highlighted breast cancer genetics pioneer Mary-Claire King's research showing that around 50 percent of individuals with BRCA1/2 gene mutations often lack the family history of cancer that would trigger testing according to current practice guidelines. Indeed, King has been advocating for population screening of all women for BRCA1/2 mutations starting at age 30, but she doesn't believe that 23andMe's latest FDA authorization will help achieve that goal in the US.

"SNP genotyping is no way to carry out mutation detection for genes like BRCA1 and BRCA2," said King, who in 1990 demonstrated that a single gene in chromosome 17, eventually named BRCA1, was causing breast and ovarian cancer in families. "SNP genotyping is not sequencing."

Doctors and genetic counselors commonly order genetic testing for hereditary cancer risk from labs that employ next-generation sequencing to analyze dozens of genes, including BRCA1/2. Myriad Genetics, which has tested the most number of women for BRCA1/2 mutations to date, offers a 28-gene panel that gauges the risk for eight hereditary cancers.

I don't want to see any more women die of ovarian and breast cancer ... [23andMe's DTC test] will increase the likelihood of that happening and not decrease it.

King's group at the University of Washington has also developed a panel test, called BROCA. She is also an advisor to newer genetic testing firms like Color and Veritas Genetics, which have ambitions to expand access to BRCA1/2 testing to more women in the US and use NGS to sequence the major hereditary cancer risk genes for around the same price as 23andMe's service.

"The 23andMe [genotyping] technology is last generation, and it's been that way for a while now," Veritas CEO Mirza Cifric said. "I understand the price pressure in terms of making it accessible to the consumer but … we don't have the technological and price point barriers that we did when 23andMe originally launched."

Moreover, unlike 23andMe's DTC service, a doctor has to order tests offered by Color and Veritas. "We encourage people to consult a healthcare professional or genetics expert to determine if genetic testing is right for them and which tests are appropriate," Schlager said. "We're really trying to encourage people to go the more conventional route." If people are concerned about not meeting requirements for insurance coverage or not being able to pay for testing, she noted, there are financial aid programs and low-cost providers like Color and Veritas that offer more comprehensive testing.

Both firms also offer genetic counseling to individuals with positive results, and facilitate testing for relatives who have a high chance of inheriting that same mutation. Color recently made an appeal for population screening for mutations in BRCA1/2, Lynch syndrome, and familial hypercholesterolemia genes, but highlighted access to genetic counselors and familial testing as critical to such an effort.

"I don't want to see any more women die of ovarian and breast cancer. My motive is very simple," King said. 23andMe's DTC test for select BRCA variants "will increase the likelihood of that happening and not decrease it."

Real risks

Kathy Hibbs, 23andMe's chief legal and regulatory officer, said that 23andMe is also concerned about the potential for patient harm and took the FDA route to mitigate those risks. "One thing that I do wonder often times with some anecdotes, … the hypothetical individual that people say they've heard about, is that actually a 23andMe customer?" she said, noting that the company has been restricted by the FDA from reporting on the three BRCA mutations for nearly five years.

Prior to the 2013 warning letter from the FDA, the company had hundreds of thousands of US customers who learned their BRCA status (positive and negative) for these three mutations. "The FDA has always had a database where people could report harms," Hibbs said. "There simply haven't been reported incidents of someone misinterpreting a negative result in such a way that prevented them from doing what we'd expect them to do, which is routine screening."

However, patient advocacy organizations, like FORCE, see very real risks in patients misunderstanding 23andMe's BRCA test report on selected variants. There are numerous social media groups for people who have BRCA or other hereditary cancer risk mutations, and within these communities, Schlager said there are posts almost daily from people discussing how they got testing through 23andMe, ran their raw data through a third-party company that analyzes that data, such as Promethease, and found out they had BRCA SNPs.

Some people on these social media groups have even posted about how based on these results they've scheduled consultations about a mastectomy, according to Schlager. "These people have never spoken to a genetics expert and some people are misinterpreting the results," she said. (Promethease does direct users of its service to discuss potentially significant clinical findings with a doctor and will connect people to genetic counselors.)

Genetic counselors recently surveyed about their experiences with consumers who had their raw data from DTC genetic testing companies, including 23andMe, analyzed by third-party interpretation services, said they had a hard time correcting people's mistaken understanding of the results. The study, recently published in Translational Behavioral Medicine, noted the experience of one counselor who tried her best to explain the difference between a polymorphism and a known pathogenic BRCA1 mutation, but this patient continued to believe she had a pathogenic mutation and was at increased cancer risk. Another counselor described how she had to spend a lot of time discussing results with a patient who was on route to having her ovaries and breasts removed based on a BRCA variant that didn't warrant such action.

When consumers find out they might harbor mutations associated with a serious illness through DTC channels and go to their doctors with the information, it's not uncommon for the physician to order confirmatory testing through a clinical genetics lab. Ambry Genetics has presented data showing that out of 49 patient samples in its internal database with alterations previously identified by DTC testing, it was able to confirm 60 percent of the results using Sanger sequencing or NGS, but found that 40 percent were false positives.

23andMe doesn't comment on what customers may be doing with their raw data with other companies, Hibbs said. However, the company will conduct education programs for physicians and outreach to patient advocacy organizations so they have an understanding of its test.

"We want to make sure that people know what they're getting and not getting, and they can make an informed choice about whether our over-the-counter product is the right product for them," Hibbs said, noting that 23andMe makes clear that people should not use its test if they have a personal or family history of cancer or if they have a BRCA mutation in their family. These types of warnings are no different from what consumers have to understand and act on with all over-the-counter products, she said.

If I were 23andMe, I'd be providing genetic counseling.

23andMe will also communicate to customers that their genetic results could have implications for family members, and provide customers information on the role of genetic counselors and how to access them. However, 23andMe doesn't include counseling for the $199 price of its service, which is concerning to Schlager, King, and others.

People who take 23andMe's test and find out they're positive for one of the three mutations will want to act on it, Gutierrez said, and that will require further testing, and discussions with medical professionals and genetic counselors. "For this test and for other tests, like [APOE testing] for Alzheimer's, if I were 23andMe, I'd be providing genetic counseling," said Gutierrez, who after leaving the FDA last year, joined the consulting firm NDA Partners as an expert consultant.

A regulatory disconnect

23andMe sees itself as pioneering a regulatory path for DTC genetic health risk testing, and the latest FDA action for its BRCA test is another step in that broader effort. FDA's authorization allowing 23andMe to market its BRCA test DTC is a distinction that no other cancer risk test has, Hibbs said, adding that the company's critics "may not have an FDA background and may not really appreciate what it means to go through FDA premarket review."

The FDA has approved other tests that can gauge BRCA mutations, for example, Myriad's BRACAnalysis CDx, but those have to be ordered by a doctor and are intended to guide treatment decisions. 23andMe's test is not indicated as a tool for making treatment decisions.

In fact, in its long list of caveats, the FDA stated that a negative test result doesn't mean that individuals don't have other BRCA1/2 mutations or deleterious alterations in other cancer-linked genes. The agency also warned that the test doesn't provide information on a person's overall risk for developing any cancer, is not a substitute for seeing a doctor for cancer screening, and should not guide decisions about treatment or surgery without confirmatory testing or genetic counseling.

"The use of the test carries significant risks if individuals use the test results without consulting a physician or genetic counselor," the FDA said in statement. In a special controls document, the agency outlines these risks and its expectations for how to mitigate them.

What FDA's authorization essentially amounts to, King said, is that 23andMe's BRCA test is analytically valid and can detect the three variants. "I don't argue that 23andMe [doesn't] know how to genotype these three SNPs," she said. "The problem arises when you don't have a mutation in those three SNPs."

Given the significant limitations flagged by FDA, King questioned if the agency has appropriately balanced the risks and benefits of providing 23andMe's test DTC. "I don't understand why the FDA authorized [23andMe's test] knowing that it was inadequate and put out their own notice indicating it was inadequate," said King. "I'm just a simple geneticist teaching undergraduates at the University of Washington. What do I know? But it does seem like a disconnect, doesn't it?"

Veritas' Cifric doesn't expect the present regulatory framework charted by 23andMe to help the labs who are using NGS to detect thousands of variants in BRCA1/2 and other cancer-linked genes. "This is not a step forward for anything other than other SNP-based approaches," he said. "This is pedestrian. It's just three variants."

An FDA spokesperson said in an email that after reviewing data submitted by 23andMe, the agency "determined that 23andMe has provided sufficient data to show that the test is accurate (i.e., can correctly identify the three genetic variants in saliva samples), and can provide reproducible results." 23andMe also had to conduct comprehension studies to show that the testing instructions, limitations, and reports were easy for customers to understand.

Although the details of the comprehension studies for its BRCA test are not yet posted by the FDA, Hibbs noted that 23andMe conducted similar studies in order to garner regulatory authorization for its other genetic health risk tests in April 2017. For that authorization, 23andMe enrolled approximately 700 individuals into seven study arms, five that tested understanding of hereditary thrombophilia reports, one on alpha-1 antitrypsin deficiency related to a genetic disorder impacting the lungs and liver, and another on G6PD deficiency to assess a red blood cell disorder.

Individuals in these studies had to watch an educational module and answer questions about a variety of concepts, for example, the purpose and limitations of the test, the relevance of ethnicity, and the role of non-genetic risk factors. Comprehension rates ranged from 73.3 percent to 100 percent, though the combined comprehension score for all concepts across all test reports was more than 90 percent. Based on the G6PD deficiency report scenario that had a score of 73.3 percent, 23andMe included a more in-depth explanation to clarify the test information to consumers.

These comprehension studies excluded individuals whom moderators found did not read through the report to answer questions. In the real world, while 23andMe will explain the test's limitations in labeling, on its website, and in test reports, Schlager noted there is still a risk that some people won't read that information or misunderstand it.

Whether people read them or not, one benefit of the FDA premarket review process is that it makes public the studies the agency used to approve a product. As a result, there is greater transparency regarding the product's strengths and limitations — information that is lacking for many currently marketed genetic tests performed within CLIA-certified labs without FDA approval or clearance.

Ultimately, this long-standing dual regulatory pathway for genetic tests — one through a CLIA certified lab and another via the FDA for companies that volunteer to submit to the agency or are required to after receiving a warning letter — somewhat dampens 23andMe's achievement. The FDA has authorized 23andMe's test as a DTC option, but the agency is also telling doctors and patients not to act on those results without confirmatory testing, which most likely will be done by tests performed in CLIA-certified labs that don't have FDA's blessing.

A few years ago the FDA tried to bring all LDTs under one regulatory framework, noting that some labs were making unsubstantiated claims that could harm public health. But the lab industry fought against this effort, and argued that tests with the agency's stamp of approval are not necessarily better than tests performed in CLIA-certified labs without FDA review, because these tests can be quickly improved based on the latest scientific advances. At the American Clinical Laboratory Association's annual meeting this week, Scott Gottlieb, who became FDA commissioner last year, noted that some of the strategies the agency's used to authorize DTC genetic health risk tests could also encourage more labs to take their tests through the agency.

For the time being, however, the dual regulatory pathway is the reality, Gutierrez said, acknowledging there are labs on the market that are performing robust genetic testing with the necessary expertise in place that don't have FDA approval or clearance. "The agency would prefer that they come through and be cleared and approved, but the LDT policy prevents them from requiring that."

Renee Crauthers contributed reporting from the American Clinical Laboratory Association's annual meeting.