The problems with the BRCA test from 23andMe

It’s been very interesting to read the range of responses to the recent FDA recent crackdown on the personal genetic service company 23andMe. Some have decried the FDA as too paternalisticwhile others have said the FDA is rightfully protecting the health of the public.

The evidence, I believe, shows the FDA is correct: 23andMe should not be doing genetic tests that have serious medical implications. Let’s look at an example to explain why.

A 35-year-old woman with no family history of cancer wants to know if she is at increased risk of breast cancer. She orders 23andMe online that, in part, includes testing of the BRCA gene which if positive could lead her to take significant breast cancer preventive steps such as prophylactic mastectomy.

Why this woman should not get the BRCA test from 23andMe

Inaccurate testing. Per the FDA letter to 23andMe this woman could have a false positive or negative BRCA test which could result in serious consequences either way.  Why? Because the only method shown to provide near 100% accurate testing is direct DNA sequencing, which Myriad Genetics (the original company which held the patent for BRCA testing) does. 23andMe, however, uses an assay with an unknown rate of success/failure.

Interpretation.  The results of the BRCA test are fairly complex. It is reasonable to be concerned about the average person misinterpreting the results.  Frankly, when I have received BRCA testing results I usually have patients see an experienced genetic counselor.  It’s not always a simple positive or negative. There can be categories such as “genetic variant, suspect deleterious” or “genetic variant, favor polymorphism.”

Not testing fully. 23andMe tests for only 3 different mutations in BRCA . The gold standard Myriad Genetic s BRCA test can detect up to 670. 23andMe does, however, acknowledge they don’t test for all BRCA mutations. But if their testing is not complete what’s the point in testing at all?

Who to test in the first place. In the general population mutations in BRCA are very uncommon (<5%) and depends on ethnicity. To limit incorrect testing results, the top U.S. healthcare task force recommends testing only in people with family history of ovarian or breast cancer. In fact, because there are so many BRCA mutations it is best to test for the mutation in relative who already has breast/ovarian cancer. That way if a mutation is identified the family member at risk can be tested for that specific mutation — limiting false negatives.

In summary, genetic tests such as BRCA are not straightforward in terms of who should get the test, the conducting the test and interpretation of the test. Testing requires a nuanced understanding of a patient’s medical and family history to determine risk; a testing company that uses rigorous methods; and trained professionals with a background in genetics to help interpret results.

I would not order the BRCA test from 23andMe and would hope this company stick to only testing genes which have no serious medical implications.

Nitin Roper is an internal medicine physician who blogs at The Roper Report and can be reached on Twitter @nitinroper.

email

Comments are moderated before they are published. Please read the comment policy.

  • Tumblingdi

    I did 23andme for me and my kids. No way would I not have gone and seen a geneticist and had formal testing and confirmation as well as consultation that included a thorough family history before acting on any aberrant results. To imply any credible surgeon would use a 23andme BRCA test result alone as a reason for prophylactic mastectomy in someone with no other risk factors seems quite hysterical to me as a physician. Our results were just what we expected from our family history and we used them as such. In private.

Most Popular