Thursday 5 December 2013

23andMe vs Lisa Casey

Updated after reading Dale Yazuki's blog pointing to this post by Lukas Hartmann, which I've sumarised next to Shaheen Pasha's below.

Poor old 23andMe; first the FDA and now Lisa Casey, can they survive? And what would their failure mean for personal genomics?

courtesy of Genome.Gov
On November 22, the FDA issued 23andMe a “warning letter” suggesting they stop marketing their personal genome service. Want to find out more about what that warning letter might mean, then take a look at Genomics Law Report's coverage. This letter followed a previous one to several companies back in 2010, so this is a bit of an old news story. But just to spice things up Lisa Casey, a San Diego resident, has now launched a class action lawsuit on behalf of other 23andMe customers asking for $5M in damages.

Lisa Casey's class action lawsuit makes some big allegations about the lack of validation, and presents three claims:

1: that 23andMe falsely and misleadingly advertises their PGS services as providing "health reports ", "drug response" and "carrier status" when there is no analytical or clinical validation for the services.

2: that 23andMe uses information it collects from the DNA tests consumers pay to take to generate databases and statistical information that it then markets to other sources and scientific community in general, even though the test results are meaningless.

3: that they have marketed and advertised (even on TV) PGS tests in violation of FDA rules in particular the Federal Food, Drug and Cosmetic Act.

The FDA letter and this lawsuit will certainly make everyone sit-up and consider how personal genomics are being done. Do we need better regulation of tests; yes, but at the same time we need to make this information available to a population of interested users that is growing fast - far faster than regulators know how to keep up with. The knee-jerk is to deny tests to people, their response may be to send DNA samples outside the US for testing.

Who is Lisa Casey: I took a look at the lawsuit and scanned through, one thing jumped out of me immediately, Lisa Casey bought her first 23andMe test on September 19th, sent her saliva off for analysis on October 5th and received results on November 19th. The lawsuit was filed on November 27th, just eight days later!

Damages: The San Diego Union-Tribune mentions the $5 damages claim. Apparently they tried to contact her, and lawyer Mark Ankcorn and 23andMe yesterday but no-one could be reached for comment. Other coverage seems to confirm that Casey's story is how she took the test after seeing TV ads, and then decided to launch a class-action lawsuit after hearing about the FDA letter.

Plausible? I think not.

Unethical research: ABC news coverage quotes the San Diego lawyer who filed on behalf of Casey as saying "[23andMe PGS] seems to me to be a very thinly disguised way of getting people to pay them to build a DNA database".

What's thinly veiled, everyone who submits to 23andMe is asked to participate in research through 23andWe: 4 findings reported in Eriksson N et al 2010, Hypothyroidism in Eriksson N et al 2012, and Parkinsons in Do CB et al 2011, not to mention the many findings reported from their unpulished analyses. There are very clear opt outs; since 2010 purchasers of 23andMe kits "must explicitly choose to allow their genetic and survey data (in an aggregated and de-identified form) to be used in published research". 

Coverage in the news and the blogosphere: Genomics Law Report has covered this space (FDA’s regulation of lab developed tests) for longer than most and their blogger John Conley has a great post that goes into some detail. John finishes off by saying he's on the side of 23andMe but thinks the legal system is unlikely to find in their favour in a 23andMe:FDA showdown. I wonder what chance John reckons the class-action lawsuit has?
“F**k the FDA – who are they to tell me what I can and can not learn about my DNA” to “Personalized genomics is snake oil and it’s great that the FDA is stepping in to regulate it”. I fall somewhere in the middle - See more at:

GenomesUnzipped points readers to Michael Eisen's blog post (see below) from the day after the FDA letter went out, and shows screen shots from two websites that offer epidemiological risk analysis (one from the NIH the other from UCLA) and links to several others. There are some great comments on this post, mainly from detractors of the FDA point-of-view, but also from people in "support"; Lior Pachter makes the point someone shoulder be regulating this kind of test. Christopher Mason re-emphasises the thing most commentators are saying, that any results that might inform medical decision should be discussed with a Doctor and they are not going to take 23andMe, or any other results, in isolation. And Lisa Madlensky, a genetic counsellor, also points out that in her clinic she needs to witness the patient provided the saliva sample, anyone can send in saliva to 23andMe and claim it came from someone else and clinical labs may have better traceability than 23andMe. She'd "like to see more people engage with their genomic data, but to also seek the advice of providers trained in clinical genetics to help sort out the useful from the less useful".

Micahel Eisen (one of 23andMe's scientific advisory board) is somewhere in the middle of two opposing camps "F**k the FDA" and "Personalised genomics is snake oil". He presents a three step program to get to his point that personalised genomics will happen and companies like 23andMe have to be part of the solution (I agree). He makes the very valid point that the FDA's wanting to regulate 23andMe PGS as a medical device is unfair as a medical device is designed to do a very specific job and do it the same way every time. Interpretation of personal genomes is nowhere near this and inherently noisy. But stopping 23andMe might just turn off one of the avenues by which we get to the medical device the FDA want. There is a LOT of commentary on his post.

Patent-Docs points out that the lawsuit aims  "to provide "full restitution" to class members" and that the plaintiffs aim to go for punitive damages, that'll be  for the sleepless night thinking you're risk of male pattern baldness was higher than you'd hoped then! How much will a jury value that at?

GenomeWeb has coverage of the FDA letter, analysts impact assessments of the FDA letter, and the class action lawsuit.

Shaheen Pasha at Quartz discusses her personal story. She sent DNA off on an ancestry hunt (probably should have sent it somewhere better) but found her medical report compelling. Shaheen had been visiting doctors for six years with undiagnosable symptoms, this ended with a diagnosis of Hashimoto’s disease. Just a couple of weeks later the 23andMe results came back (after the usual 4-6 week wait) with an elevated risk of autoimmune disorders, including Hashimoto’s. She argues that this would have been a help had she taken the test six years ago.

Lukas Hartmann, got his 23andMe results in 2010 and was quite happy with the results. However he recieved an update that he had “two mutations linked to limb-girdle muscular dystrophy." Rather worryingly the 23andMe report stated that "a person with two of these mutations typically has limb-girdle muscular dystrophy." After some of his own research Lukas was able to determine that although he had two mutations these were in different genes, he did not have a homozygous disease causing mutation but separate heterozygous mutations in genes linked to LGMD. 23andMe had reported this completely erroneously. After sending in a bug-fix request his report from 23andMe now reads that he “has multiple mutations linked to limb-girdle muscle dystrophy, but they are in different genes. A person with such mutations typically does not have the condition, but can pass the mutations to offspring. May have other mutations linked to limb-girdle muscular dystrophy (not reported here).” Ammunition like this will be gold-dust to Lisa Casey and is more bad news for 23andMe.

What might be the impact of a "bad" 23andMe result: The FDA have suggested that an individual might undergoe surgery as a result of a 23andMe BRCA test, this seems nonsense given the need to have a surgeon agree to perform the procedure. The BBC has covered the 23andMe story too, but another recent article on their news website made me wonder if there might be other difficult to measure impacts from 23andMe results people were not expecting. The BBC ran a piece on the poverty facing the almost 20,000 people over 55 living with HIV in England today. Twenty to thirty years ago many of these people were effectively given a death sentence, so planning for the future was not high on their agendas. The BEEB points to a recent study by the Terence Higgins Trust that says many of those diagnosed with HIV in the 1980's became ill and had to give up work , some sold up, cashed in their pensions, went round the world and waited to die. They did not. Might this happen to someone who recieved bad results from 23andMe, read through discussion in chat-rooms and decided the end was in sight? It certainly sounds like a scenario more likely to occur than someone getting a double mastectomy on the basis of a single test, and convincing a doctor to perform the operation!
“F**k the FDA – who are they to tell me what I can and can not learn about my DNA” to “Personalized genomics is snake oil and it’s great that the FDA is stepping in to regulate it”. I fall somewhere in the middle - See more at:
“F**k the FDA – who are they to tell me what I can and can not learn about my DNA” to “Personalized genomics is snake oil and it’s great that the FDA is stepping in to regulate it”. I fall somewhere in the middle - See more at:
“F**k the FDA – who are they to tell me what I can and can not learn about my DNA” to “Personalized genomics is snake oil and it’s great that the FDA is stepping in to regulate it”. I fall somewhere in the middle - See more at:

23andMe and Me: I had my genome scanned a couple of years ago and found the whole experience to be quite fun, slightly enlightening and the data presented in a way that made it easy to explore and understand. I was unhappy about the subscription model and did not pay the monthly fee to get access to updated results. I did use Promethase to "analyse" my results and have looked at getting better genealogical analysis by uploading my data to other sites. I am still very intersted in getting the rest of my family screened too.

Lisa Casey is asking for $5M plus punitive damages for 23andMe customers, I'm assuming this does not cover me as I'm outside the US. This did make me think about how much I'm owed though. If 23andMe's own claims to have 500,000 users are true, and we accept everyone paid $99 then surely the claim's got to start at $50M. This lawsuit is not asking for enough cash, and if they get a payout how can I claim back my $99 plus $69 shipping? And what about monthly subscription packages? To be honest I'd be much more interested in an updated scan with the new Onmi5M chips (which have just come down in price), or my exome! SO if 23andMe need someone to stand up in court on their side I'm happy to be subpoenaed, business class is fine. But if they win 1st class home again would be great!

Oh, does anyone need a personalised cancer panel test? Another company shortly to receive similar letters from the FDA perhaps!


  1. It seems that you are not satisfied with all guidelines of FDA. And the experiences of patients with 23andMe is quite unusual.

  2. I think most 23andMe customers have similar opinions. I did an informal survey and found the vast majority of those surveyed were satisfied with their results (meaning they didn't want a refund). I think there is also some confusion about what the FDA warning represents.

    I put together a post describing this in greater detail (I also liked this post, BTW):


Note: only a member of this blog may post a comment.