MPNforum Magazine

Between SN(i)Ps and Spit
This Is Getting Personal

By Zhenya Senyak

Two weeks ago 23andMe announced it would be launching a large-scale MPN genomic database to help turbocharge research. This dramatic initiative requires a tissue sample. They want my living cells, floating in a solution of saliva. In return, they promise to tell all.

They’re not going to tell me about my phenotype, that face that looks back at me from the mirror, these long bones, sparse hair, jug ears, hazel eyes, and million other traits that make up the me that negotiates this brief passage on earth including, most likely, a predilection for myeloproliferative neoplasm.

23andMe gets a lot more personal than that. They promise to reveal the secrets of my genotype, the underlying blueprint that created this creaking phenotype that comes before you with this story. Within those cells harvested from my mouth rests the cell nucleus. And in the nucleus is a tightly wound strand of information, my DNA, a code based on only four letters but three billion characters long.

The promise 23andMe makes is to read that code – after it is appropriately processed by an independent laboratory in San Diego– and deliver to me nothing less than the tea leaves of my future, a reading, for example, of the likelihood of me contracting one or more of several dozen calamitous diseases. In a way, they propose to crack open the fortune cookie of fate and interpret it for me to read, privately and personally. Securely.

Not only that, 23andMe will map my antecedents whose tracks are clearly visible in that same unwound strand of DNA, draw back the curtain on their prehistoric sex lives, and report back to me, the recipient of those long reproductive efforts.

Participate and you get something more..

Once you actually get your kit and register you have the option to fill out a questionnaire about your life, your diseases, meds, symptoms, etc. and, if you give your consent, and return your DNA-bearing spit you’ll become part of a DNA database to be used in research. The MPN database is a separate initiative to further research into the disease that plagues all of us in this Forum and several hundred thousand others worldwide, born and unborn. You and your DNA will be enrolled in that MPN database once the cohort of 1000 participants is recruited.

And it’s all absolutely free. Free?

Having been stiffed by the Tooth Fairy in my youth, this kind of largess raises a few questions. Why? Who are these selfless souls offering to part the dark veil of our future? What’s in it for them? How real is this offer? What are the risks? How sure can I be this information won’t fall into the hands of foreign agents…or my insurance agent for that matter?

Getting to a clear understanding of this extraordinary offer to our MPN community turned out not to be as simple as one would think.

Who are these people behind 23andMe? That has to be the first question . If genomics teaches us anything, it’s little differences create big differences, . Looking at our biochemical structures, we’re almost identical. Our basic DNA building blocks – called nucleotides – are 99.9% identical, a difference of 1 in 1000 nucleotides in two randomly selected humans. That diversity, a shift in letter combinations – a single nucleotide polymorphisms, or SNP on a strand of DNA – is central to both quick scanning of the human genome and the MPN DNA project at hand..

Single nucleotide polymorphisms

See Wikipedia, SNPs for source

Single nucleotide polymorphisms)

Although we share 99.9% of our genome with any random person, we know through bitter experience we can’t rely on genetic similarity alone to leave our car doors unlocked at night. Or blindly trust our pension funds to corporate managers. So in considering a matter as vital to our health and security as the code for our personal DNA, we need something more discriminating than SNPs and genetic markers. Luckily, there is a guideline, the surprisingly robust and accurate Duck Theory.

If something looks like a duck, waddles like a duck, lays eggs like a duck, and quacks like a duck, then chances are it is a duck.

23andMe presents its MPN DNA story as a crusade originating in a daughter’s desire to save her father’s life. There’s no doubt this is true. The father is one of our own, an MF patient enrolled in the Cytopia clinical trial at Stanford. The daughter is an executive and legal counsel at 23andMe. They originally called their MPN DNA project cohort a “community” — now an initiative — and their literature talks of sharing this online database to enable faster, more extensive, and cost effective research. The fervor and sincerity of folks we talked to at the company and research associates at Stanford is palpable. 23andMe quacks like a True Believer.

23andMe also waddles like a corporation, definitely not a charitable non-profit. In many ways 23andMe is similar to other California biotech operation amply funded by heavy duty venture capitalists.

Seasoned Silicon Valley investors sit on the Board. Its CEO and president, Anne Wojcicki is married to Sergey Brin, one of the founders of Google, who plunked millions of his own money into the corporate till. Google’s Venture Capital group invested another $6.5 million. By the time co-founder Linda Avey left the company, in 2009, to make way “for new leadership,” 23andMe had already amassed a war chest of $23 million from investors like Genentech, and News Enterprise Associates, as well as Google. Last year, in a new round of financing they were joined by MPM Capital and Johnson & Johnson.

Considering the company’s high impact marketing, tight-lipped PR, strong financing and presumed stock IPO trajectory you’d have to say 23andMe quacks and waddles just like a Corporate Duck.

It’s easy to be confused about who 23andMe is. The company hasn’t broken substantial new ground in any areas of genetics, has so far contributed little to biological science. Dipping into public databases for the raw materials of its direct to consumer (DTC) genetic testing business, it essentially acts as a broker, between SNP array – gene chip manufacturers like Illumina and processing labs and the consumer. The essential value it adds appears to be statistical analysis, contributions to custom gene probe designs, processing of data, packaging and marketing . Definitely marketing.

One look at the bright and cheerful crayon colored packaging, slick website, comic book style endorsements, (“Saved my life, ” C.K. Kansas), its past mass sales through Amazon and you know you are looking at the work of a sophisticated consumer marketing organization.

Another of the company’s businesses is leveraging of its databases for research by external organization, researchers, foundations. For that 23andMe has both its general member database of over 100,000 (“82% of whom agree to participate in research projects” ) and the company’s disease-specific on-line research cohorts – in Parkinson’s Disease, Sarcoma, and now MPNs.

Todd Sherer, CEO of the Michael J. Fox Parkinson’s Research Foundation, points to economics as a benefit of working with 23andMe ‘Using traditional methods, getting 4000 study participants would cost millions of dollars. 23andMe has done it for a small fraction of what it would have cost otherwise.” According to 23andMe, “published studies show recruiting costs of approximately $2000 per enrolled subject for a typical clinical study compared to a quarter of the cost using 23andme for an entire study from recruitment to analysis.”

That would seem to place the price tag for using a 23andMe online research cohort of 1000 participants at $500K, from recruitment to analysis.

However, as Ashley Gould, 23andMe vice president and general counsel, points out, although the company has provided research services for a fee ” We have no current plans for any paid projectrelated to MPNs. If any evolve they would only proceed if 23andMe customers choose to participate, choose to take surveys and provide individual consents as an informed, personal choice on their part. Specific to this MPN research

Ashley Gould

, 23anMe is sharing aggregate MPN data with our unpaid Stanford research collaborators. We will be thrilled to share aggregated data with additional MPN researchers who join our project as scientific advisors or collaborators… Our goal with this project is to publish on what we find and inspire others (whether academic or pharma) to build upon our work and translate it into improved treatment and patient care options.:

Seeing a Corporate Duck in the medical services business is normal. Most of the important medical developments of the past century have come about because entrepreneurs risked their money backing scientists. . Cognitive dissonance arises from seeing a committed group of young scientists, executives, and statisticians launching an expensive research initiative to benefit a small MPN community without a clear revenue return in sight

These could simply be conjoined objectives. 23andMe is simply developing an asset that could be used in collaborative or third-party research projects down the line.AND they are on a personally motivated MPN crusade. The professionals at 23andMe seem genuinely committed to use this initiative to explore genomic paths to help discover the causes of myeloproliferative

Creator: Carl Banks

disease. Whatever the mechanism, through some genomic mendelian mutation 23andMe seems to have created a hybrid duck, an odd duck, a Corporate True Believer duck might just be on our side.

>>MPNforum Conclusion>> To the question, Who are these people? we can now unequivocally respond, ” They’re 99.99% just like us. “

How real is this offer?

23andMe clearly has the capabilities to convoke a cohort of MPN patients for its on-line DNA database; assess the SNP array analyses they receive; and generate a limited but statistically defensible personal DNA report based on historical data. As one renowned university geneticist told us, “This is not state of the art. This is not leading edge science.” But it doesn’t pretend to be. This is genetics for the masses, the Human Genome Project brought home for practical use, a kind of pop biology report.

Understanding the elements of the product 23andMe delivers means taking a broad look at the genetic science in play. Both the regulatory criticism of the company and the risks we undertake by participating in this initiative are embedded in the applied science. Perhaps the most damning conclusion was reached by the Government Accounting Office in its report last year on Direct-to-Consumer Genetic Tests, a study of four companies, including 23andMe. “The test results we received are misleading and of little or no practical use to consumers.”

The GAO documents its findings in a 33 page report focused on four DTC genetic testing companies, including 23andMe. The limitations of applicable genetic science and the absence of standards for adopting risk markers are at the root of contradictory risk predictions reached by these companies. Using publicly available studies of genetic markers presumed to contribute to a donor’s risk for disease, “all four companies acknowledged that, in general, DTC genetic test companies test for different risk markers and this could result in companies having different results for identical DNA.”

Remember that 99.9%? A major source of human variation is found among the 10 million commonly known SNPs, those strips of DNA with changes in a single base. Some of those changes won’t matter much, some will create new polypeptide chains and make proteins that change things. On average there are about 50 SNPs per gene pair base. So mapping the SNPs permits correlating changes in DNA sequences with disease states or predispositions to disease. Like other DTC genomic testing companies, 23andMe looks at only a fraction, maybe 1/10 of 1% of the human genome. The rest is the magic of SNP arrays and statistical analyses projecting from haplotype data..

The issue is further complicated since analysis of SNP can lead to statistical assumption of gene sequences but cannot determine the state of the gene, whether or not it is expressed — on and active — which may be far more significant in determining likelihood of disease risk. SNP analysis is a more powerful tool in the 23andMe MPN DNA initiative since there it can play a significant role in identifying regions of interest for further deep genomic testing via, for example, full gene sequencing.

The MPN initiative

Except for rare instances, we are at a very early stage in understanding correlations between inherited genetic anomalies and disease states. Certainly in myeloproliferative neoplasms neither the triggering nor proliferating mechanisms are well understood nor are the inherited predispositions mapped

While 23andMe doesn’t appear to have the in-house capability to evaluate the complex pathogenesis of myeloproliferative neoplasms, it does have a powerful ally.

Jason Gotlib, M.D.

The company has a research collaborative relationship with Dr. Jason Gotlib and his associates at Stanford University. Gotlib, currently lead investigator in the Cytopia trials at Stanford was a primary investigator in Incyte’s now completed COMFORT trials at Stanford. He is a board-certified hematologist, part of the new generation of clinical investigators specializing in myeloproliferative neoplasms. And he believes in this research project. “The intent of this research initiative,” says Gotlib, ” is to attempt to cull genotype data from large numbers of MPN patients to better understand whether there is an inherited predisposition to these diseases and to see if there is any correlation between SNP data and aspects of disease diagnosis and treatment.”

The prospective benefits to the MPN community of this offer carry certain risks that we need to assess individually before we ship our DNA off to California. There are two main areas of concern: health and privacy.

The health risk cuts both ways. Results from direct to consumer DNA tests are unmediated. Often there is a warning sticker recommending the consumer bring these results to his physician for evaluation. On the face of it, this had to have been written tongue in cheek since the likelihood of physicians having a deeper knowledge of genetic sources of disease than patients is low.

Along with valid warning signs of potential genetic weaknesses, reports can contain indications of high risk for certain conditions that are unwarranted. If the net result of reading such reports is to create some degree of stress and a visit to the doctor’s office for a checkup, no great harm is done. The incidence of patients changing diet or exercise habits based on these reports is similarly very low. .

The real danger arises in us believing a low risk ranking is true when in fact we might carry a very high genetic risk for a disease. A case in point is breast cancer. Most of the SNPs analyzed lie outside actual gene locations. Since less than 1/10 of 1% of the total human genome is sampled via high throughput SNP arrays, chance of missing deadly genes like the BRCAs implicated in breast cancer can be considerable. (23andMe does test for the three most common BRCA gene mutations.)

Speaking to this issue, here’s the way Dr. Jeff Shuren, director of the FDA Center for Devices put it in July, 2010:” Failure to validate the accuracy, reliability, and clinical implications of a test can result in patient harm from misdiagnosis, failure to treat, delay in treatment, inappropriate treatment, or avoidable adverse events.”

In 2010, the FDA concluded that “23andme’s Personal Genome Service is a diagnostic device subject to requirements of the Federal Food, Drug and Cosmetic Act . 23andMe claims the service is educational. Early this year, Ashley Gould, vice president and corporate counsel, appeared before the FDA DTC Genetics Advisory Panel to present the 23andMe case. According to the FDA, as of this date discussions are still in process. The 23andMe slide presentation made at that meeting is here: http://www.fda.gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/MedicalDevices/MedicalDevicesAdvisoryCommittee/MolecularandClinicalGeneticsPanel/UCM248565.pdf

>>MPNforum Conclusion>> The health risk appears minimal IF we take 23andMe at its word and use these Personal Genome reports as educational tools only and don’t take drastic health actions based on risk rankings or refrain from taking action based on presumed genetic health. As ever, like mom always said, eat lots of vegetables and fresh fruit, in a low-fat balanced diet, avoid sugars and hydrogenated vegetable oils, brush your teeth often, exercise daily and play nice.

Security – for some the deal breaker, for others no big deal

Maintaining the security of your DNA data is a primary goal of 23andMe “Even I,” said Jason Gotlib “have no access to personal data.” According to Kimberly Barnholt, 23andMe Project Manager of the MPN initiative,

Kim Barnholt

“23andMe has a number of safeguards in place to ensure confidentiality of data…Investigators who have access to any sensitive data (phenotype, genotype) do not have any access to patient account or contact information. All data in the research computing environment is coded with individuals’s unique research ID (anonymized).

“Conversely, 23andMe project managers who have access to patient contact information do not have any access to sensitive data. Thus individually identifying information cannot be linked to genetic and phenotype data. All research is performed in a secure computing environment with access restricted to investigators and system administrators.”

Maybe. There’s little doubt that 23andMe has taken security precautions worthy of NASA…which has been hacked. There are just inherently weak points in the chain of security that can be breached. Deanonymization is a science that may be as advanced as genomics in its own way. Individual SNP data that exists on hard drives might be accessed and linked to larger data bases to unlock personal identities. Customers have codes to access their data, another weak point in the chain. Hackers say ” Data can be either anonymous or useful it can’t be both.”.

At issue is how critical is DNA security is to someone. Privacy of personal health records is a major concern worldwide, particularly now with the increase in electronic medical recordkeeping and electronic sharing of diagnostic data. It’s a personal issue,

>>MPNforum Conclusion>> 23andMe has done as good a job as possible in securing personal genomic data and keeping it separate from personal identification. It’s an imperfect world and there is a possibility your DNA information will be revealed. The risk seems small.

And that’s why I’m in. The health risk to informed consumers is minimal, the privacy breach risk is inconsequential to me, and there’s some chance adding my DNA to the pool of other MPN patients can help some investigator down the road uncover significant correlations among genetic anomalies and MPNs.

There’s many a SN(i)P twixt cup and lip.

>>Contact>>
For more information about the MPN research program,
mpn-help@23andme.com or www.23andme.com/mpn
__________________________

© Zhenya Senyak and MPNforum.com, 2011. Unauthorized use and/or duplication of this material without express and written permission from this blog’s author and/or owner is strictly prohibited. Excerpts and links may be used, provided that full and clear credit is given to Zhenya Senyak and MPNforum.com with appropriate and specific direction to the original content.