Wednesday, March 7, 2018

FDA Allows DTC Access to 23andMe's 3-mutation BRCA test; Trends and Controversies

FDA has approved 23andMe to provide non prescription, direct to consumer marketing of a 3-mutation BRCA risk test.   The 3 mutations are founder mutations in Ashkenazi populations.

According to readily available press, about 1 in 400 people have a BRCA mutation, while the founder mutations may occur in 2% of Ashkenazi women.   Highly aware that a negative test does not rule out other BRCA mutations (or mutations in other breast cancer risk genes), the FDA released the test with "special controls" (which can include warnings and education).   The 13 page letter to 23andMe is already online and lists a description of the special controls.

Shorter and longer articles at Genomeweb here and here.   A deeper dive additional article by Turna Ray here [2800 words].  New York Times here.  WSJ here.   23andMe's own blog about the authorization here.

Manyu Different Genetic Regulatory Innovations from FDA 

This new authorization follows a trend of rapidly released genomic test regulatory innovations at FDA.  The following form a moving wavefront of FDA innovation:
  1. Allowing easier release via new pathways for FDA endorsed DTC genetic testing in early 2017.
  2. Approval of the OncoMine Dx Target 23-gene panel in lung cancer, with 3 PMA approved genes and also with 20 clinically reportable genes in June 2017.  
  3. Clearance of the MSKCC large gene panel tumor test in mid November 2017 based on a pathway using New York State approval has the linchpin.
  4. Approval of the Foundation One CDx test in late November 2017 with two novel features:
    1. Most of the PMA genes were based on analytical concordance or bridging studies.  For these genes, the results are analytically similar to results with at least one previously approved kit (and not necessarily on clinical trial or clinical data specimens).   
    2. Release to clinicians and patients of "signatures" that are not FDA approved, such as MSI and TMB. 
  5. Release of BRCA testing with very small DTC panels (this article).

Mutation Prevalence in 4-Grandparent Ashkenazi Populations

Last summer, Walsh et al [UW-Seattle] ran a 23-gene test breast cancer risk test in 1007 women with breast cancer and 4 reported Ashkenazi grandparents.  74% of all breast cancer risk mutations were a BRCA founder mutation, 5% were a different BRCA mutation, and the remaining 22% had a mutation in an entirely different breast cancer risk gene.   Altogether, a mutation in some gene was found in 14% (142 of 1007) patients in this ethnic cohort.   Presumably, as soon as you move to less than 4 Ashkenazi grandparents, results will diverge.  In a different area of genetics, Haque et al. (2017) reported widely different recessive risk profiles across ethnic groups in over 300,000 persons tested.  


As the FDA becomes more flexible and less paternalistic, fewer will complain the agency is too paranoid.   

But the flip side is that more stakeholders will be concerned about unintended risks or adverse events.  While the FDA allows this DTC test, the FDA press release states that: "The use of the test carries significant risks if individuals use the test results without consulting a physician."



While CEO Anne Wojcicki noted that "being the first and only DTC genetics company to receive authorization to test for cancer risk without a prescription is a major milestone" - and that's true - the authorization is based on technical accuracy and controls such as patient warnings, which other labs should find easy to replicate if they want to.

The FDA cited a 1997 NEJM study as its source of statistics.

Note that the test is described as "authorized" rather than cleared or approved.

The approval is described as de novo, but the detailed documentation for the authorization and pathway may not appear on the FDA website for weeks or months.   An example of a 77-page earlier FDA scientific review and decision on 23andMe testing is here.   [Update: At least a 13-page letter to 23andMe is already online; in other de novo approvals, full FDA paperwork can lag a long time before public display.]

BioPharmaDive describes the test as "tentatively OK'd by FDA" but I can't tell what's "tentative."

In her long-form article, Turna Ray notes the paradoxical positioning of FDA regulation.   FDA authorizes accurate DTC testing by 23andMe, but requires confirmation by another lab for clinical action, and the second lab is likely to be a "pure LDT" and not be FDA reviewed at all.  

FDA does not generally regulate LDT genetics.  However, it does proactively regulate DTC genetics with clinical implications; (ancestry or paternity is non regulated).