Getting Serious About Personal Genomics’ Risks
After several months of public drama, the University of California, Berkeley’s ambitious program to introduce its incoming freshmen to personalized medicine reached its denouement in late August.
As part of its program, Berkeley offered students the option to participate in genetic testing for three common genetic variants relevant to the body’s ability to metabolize milk products, alcohol and folic acid. The University’s original plan was to allow students to elect to receive the results of their tests as part of the program. Two weeks ago, however, the California Department of Public Health (CDPH) ruled that if Berkeley wanted to return personalized genetic data to some of its freshmen, the testing must be conducted at the direction of a physician and performed by a licensed clinical laboratory. The significant logistical burden and cost of complying with the CDPH’s ruling forced Berkeley to modify its program. While some aspects of the program will go forward, no student will be able to access any personalized genetic information.
(CDPH’s ruling was unexpected. Berkeley’s Dean of Biological Sciences, Mark Schlissel, noted that the department’s ruling “relies on an interpretation of legal statutes that is entirely different from the interpretation of the same statutes by UC’s top lawyers.” The ruling itself has potentially significant implications for genetic research across the country, although that topic is the subject for a future post.)
The focus of this post is the rapid mobilization of critics of the Berkeley program and the power of public controversy to spur regulatory action and, ultimately, to force the University to adopt a fundamentally different approach to personal genomics education than originally intended. This in spite of a detailed internal review process that consumed substantial resources and required Berkeley’s Institutional Review Board (IRB) to approve the project. Examining how and why this happened is instructive for evaluating the future prospects of personal genomics research and innovation.
A Controversy Emerges. From the outset, a handful of bioethicists and public interest groups voiced hypothetical concerns about the risks of offering genetic testing to Berkeley’s freshmen. The Council for Responsible Genetics greeted the program’s launch with a letter to the University (pdf) that warned that genetic information “has the risk of being used out of context in ways that are contrary to the interests of the individual, perhaps even discriminatory and certainly privacy invasive.” Similarly, an article in The New York Times featured Boston University bioethicist George Annas, who posed the following hypothetical:
What if someone tests negative [for alcohol metabolization], and they don’t have the marker, so they think that means they can drink more? Like all genetic information, it’s potentially harmful.
Finally, the Center for Genetics and Society linked the Berkeley program to contemporaneous developments in direct-to-consumer (DTC) genetic testing, and warned that “students might think, ‘Berkeley gave it to us. It must be good. UC Berkeley would never be giving its incoming students anything bad or controversial.’”
In short order, what began as an innovative approach to introduce incoming students to genetics and personalized medicine by offering those students the opportunity to personalize their experience quickly became a controversy.
From Controversy to Regulation. Controversial educational initiatives are hardly new. Indeed, they are part of the mission of many institutions of higher education, including Berkeley. In responding to initial criticisms of the program (pdf), the University emphasized that “provoking a free and open discussion about issues surrounding genetic testing is an important aspect of educating our students to be informed citizens.”
Unquestionably, there is considerable value in subjecting all forms of innovation to close scrutiny. In fact, in any Common Rule-governed human subjects research, this is a requirement. Among the many criteria for IRB approval of a human subjects research project is the requirement that “risks to subjects are reasonable in relation to anticipated benefits.” The provision of informed consent is a separate, and similarly important, prerequisite to approval. Berkeley’s own IRB reviewed the University’s project, applied these and other statutory criteria, and ultimately approved the project.
Despite not being legally required to do so, Berkeley actively engaged with the program’s critics from the outset. A program that was vetted internally was now being vetted by the public, with the University’s active participation. In response to public feedback the University modified the project to clarify the project’s voluntary nature, the informed consent process and its separation from actual or perceived industry conflicts of interest.
Critics of the Berkeley program, however, were not satisfied. They continued to urge first the University and then California legislators to much more dramatically alter the program, or even to discontinue it entirely. The constant stream of criticism had an impact. Over the course of the summer, legislation was introduced that would have halted the program. That was followed by legislative hearings to debate the program’s merits and, ultimately, by the CDPH ruling that effectively ended the program in its originally-proposed form.
The rapid reaction of regulators to a debate that was largely driven, especially initially, by media reports, “expert” commentary and social media discourse was strikingly reminiscent of another mid-May personal genomics development.
The week before Berkeley’s program was announced, DTC genetic testing company Pathway Genomics and drugstore giant Walgreens announced a partnership that would have made Pathway’s consumer genetic test available through Walgreens’ stores. In Pathway’s case, the leap to controversy was even swifter: the initial story in The Washington Post describing the agreement warned of a “Pandora’s box of confusion, privacy violations, genetic discrimination and other issues.” Nonetheless, the end result was the same as regulators quickly stepped in and demanded changes. Rather than the CPDH demanding physician intervention and a clinical lab, in Pathway’s case it was the FDA declaring the product in question a medical device in need of a time-consuming and expensive medical device clearance or approval. In both cases, swift regulatory action effectively quashed the proposed activity.
Why Debating Personal Genomics is Difficult. Recent developments suggest that innovation in personal genomics is an increasingly difficult undertaking. In addition to the Berkeley and Pathway cases, examples include the FDA’s increased oversight of DTC genetic testing companies (in addition to Pathway), the GAO’s report on the perils of DTC genetic testing and, most recently, criticism of the University of Minnesota’s attempt to bring genetic research to the State Fair. Collectively, this suggests the emergence of a disturbing trend: developments in the area of personal genomics that deserve serious public debate are shaped from the outset by commentators, policymakers and lawmakers more concerned with making a point than with advancing the conversation.
Of course, it is hardly news that emerging areas of science and controversy generate controversy. In recent weeks, the safety and desirability of human embryonic stem cell research has sparked a heated public debate, just as it has at regular intervals for the past decade. The new dynamic facing personal genomics is the rapidity and ease with which any initiative may be branded as “controversial,” combined with the willingness of lawmakers and regulators to intervene directly and rapidly in such “controversial” activities. This may be as much a function of new paradigms in media, politics and public discourse as it is a function of personal genomics itself, but whatever the reason the concern is that it is having a chilling effect on innovation throughout the field.
On the commercial side, the effects of increasing regulatory uncertainty are evident, as businesses and investors are considering abandoning personal genomics or moving their operations – and attendant jobs and capital – overseas. On the research side, similar confusion – particularly in light of the Berkeley program’s fate – continues to discourage researchers from exploring innovative approaches that might help to accelerate our attempts to decipher genetic complexity and, ultimately, provide us all with more effective, less expensive health care.
Whatever the context, there can be no substitute for careful, public and reasoned debate when it comes to evaluating the appropriateness of a new personal genomics proposal. Similarly, there is no substitute for fully informed consent; for ensuring that all individuals – whether they are students, patients or consumers – understand the full extent of the risks attached to a decision to participate in a personal genomics activity. Both are critical in assuring that personal genomics is conducted in a responsible fashion.
But public debate and informed consent require more than an ability to enumerate hypothetical risks. When it comes to evaluating innovative personal genomics proposals, all of us – participants, funders (including taxpayers), media and commentators and, especially, policymakers and regulators – owe a duty to be thoughtful and balanced in assessing their merits. To be blunt, it requires all of us to do more than throw darts at the easiest targets.
This means understanding that it is not enough to simply enable public debate between those with opposing views on the merits of a particular project. It means recognizing that all innovation – scientific, technological, commercial, research, educational, etc. – carries with it a measure of uncertainty, but that uncertainty alone is an insufficient reason to slam on the brakes. It means acknowledging the difference between hypothetical or low-probability risks and actual, documented harms, and recognizing that the first step should be determining which is which. And most importantly of all, it means considering the benefits of innovation in personal genomics that accrue in addition to – and often because of – its risks.
This is not an easy task. Particularly in a field such as personal genomics, which is driven by new and often untested scientific knowledge and technology, it is trivial to examine a new idea and find something that could conceivably go wrong. Is it possible that a freshman tested for a genetic variant associated with alcohol flush reaction could interpret a negative result as a license to consume alcohol in excess? Of course it is possible, for the bar of “possibility” is exceptionally low. It is much more difficult to convert hypothetical risks into actual data on behavior (i.e., do individuals act to their detriment as a result of non-clinical genetic testing in general, and specifically in the case of the alcohol flush variant?), and more difficult still to balance such risks against the benefits of the same activity.
Keeping Our Heads. Realizing the promise of personal genomics will be impossible unless our society is willing to accept some measure of uncertainty and, yes, risk-taking. Our challenge is to figure out not only when the benefits of personal genomics outweigh its risks, but also who should be permitted to make that frequently difficult and personal risk-benefit decision, and in what contexts.
For those who would place that decision in the hands of individuals, there can be no question that we must first provide those individuals with the necessary information and perspective to make an informed decision. But the process of informing personal genomics participants – of informed consent – no matter how thoughtful and comprehensive, can only take us so far. The information will never be complete, the perspective will never be perfect, and the decision will never be without risk.
It is true, too, that there are many situations where society examines the risks associated with a particular activity and decides that they are simply too high – whether to the individual or to society as a whole – to be assumed, even knowingly and voluntarily, by the individual. We do not, for instance, let teenagers consume alcohol. We place restrictions on the acquisition or use of all manner of technologies, from automobiles to firearms. We require regulatory approval and a doctor’s prescription for most pharmaceuticals.
But as a society we also evince a deep respect for autonomy, leaving many risky decisions in the hands of individuals. The decision to drink alcohol or drive a car in the first place (assuming one is of legal age), to become pregnant (and even to terminate a pregnancy) and to provide informed consent to participate in scientific research: all of these decisions we leave in the hands of individuals.
We have not yet determined whether personal genomics is more like the decision to conceive a child– a personal decision free from state intrusion – or the decision to undergo chemotherapy – a personal decision highly regulated by the state. In a field with a landscape as diverse and rapidly-changing as personal genomics, the answer will frequently depend on context. Some aspects of personal genomics (e.g., genetic testing to determine a proper therapeutic treatment) warrant a greater degree of societal intervention than others (e.g., genetic testing to determine geographic ancestry).
The challenge is knowing where to draw that line. The risks posed by automobiles, firearms and pharmaceuticals are well-documented whereas, at least for the moment, the risks of personal genomics remain largely hypothetical. In the absence of clear data, the recent trend to deemphasize the benefits of personal genomics while focusing on its risks, and to use those risks as justifications to shift control away from the individual, should cause us all to question whether we are drawing that line in the proper place.
If personal genomics is ever to live up to its name, at some point we must allow individuals – including the future leaders of our society, as embodied by Berkeley’s incoming freshmen – to decide for themselves whether and how to participate. To do otherwise, and to continue to aggressively criticize and thereby discourage personal genomics innovation in our zeal to render it a riskless enterprise, would be a mistake.
I checked, it is Dan this time…good post! Hypothetical risks and potential responses is precisely one of the reasons why it is sad that the Berkeley study did not go ahead – each gene was associated with an environmental component, the students would have learnt about the importance of both and that genes are not so special, or deterministic on their own.
As for the hypotheticals, unless they are really probable and serious they are always tedious arguments. What if those zebra crossings (black and white stripey things on roads in Europe, I don’t remember seeing them in USA, for e.g. see Abbey Road ) encouraged people to cross the road with less care and attention? It’s serious, it has happened and people have been injured or killed because of zebra crossings (indeed in Italy I have seen more people knocked over on the crossings than on normal road – but that’s a different story…).
Put out that fire, it might burn someone…what do you call that, a wheel? Are you mad, think of the damage that could do…
This post uses the UC Berkeley fiasco to provoke debate about the regulation of DTC testing in general. Which is fine — as a critic of the Berkeley program (I frequently work with the Center for Genetics and Society), I am more than happy to see the discussion shift to larger and more important issues. But pointing out failures, whether technical or ethical, is not a “disturbing trend,” it’s part of the democratic process. If anyone is scaremongering here, it’s the author of this post, dredging up an alleged “chilling effect on innovation” as if he were more concerned with making a point than with advancing the conversation.
@Pete
It’s not scaremongering – I read this as chilling in the sense of cooling down rather than the horror film spine chilling interpretation. The former fits in with the reasonable tone of the rest of the article. As for cooling/chilling – it’s what is actually happening (it looks like there is a lot of heat in the sequencing investment market though). I’m sure the author is concerned with making a point (what’s the point otherwise…) and the point that comes across to me is a good analysis of the problem of overstating the risks – apart from the “…so they think that means they can drink more” stuff we have also recently seen the head of the FDA make up scare stories about unnecessary surgery etc. That’s where we could do with some chilling.
@Keith: Per the GAO, the DTC industry is providing “test results that are misleading and of little or no practical use.” Why blame the people pointing this out?
Unfortunately Mr. Vorhaus misses one of the most essential points that critics, including many Berkeley professors involved in teaching the societal, ethical, and legal implications of genetics, had of the “Bring Your Genes to CAL” program.
The “reasoned debate” that Vorhaus desired could only be possible once the students, i.e., the essential participants in the experiment, had received the benefit of a broad curriculum regarding genetics, personalized or otherwise, including questions regarding not only known benefits and risks, but potential ones as well. Asking students to agree to having their genetic information analyzed, stored, and controlled by a state university prior to the receipt of such knowledge is “risk taking,” inappropriately solicited to even “autonomous” incoming freshmen.
As one who teaches about these issues to students at UCLA in conjunction with their Center for Society and Genetics, I believe that the basic, simple, suggestion made by critics is precisely the conversation that the BYGC program lacked. The “hypothetical” risks of the collection of genetic information and genetic testing so blithely suggested by Vorhaus have developed into actual harms, the case law is there and Mr. Vorhaus should familiarize himself with examples such as Dan Greenberg and the Havasupai tribe. In short many of the critics simply requested that Berkeley fulfill its mission and Teach, then Test.
The project was conceived without consultation among scholars across the campus who have conducted clinical, scientific, anthropological, social scientific, bioethical and legal research on genomics and personalized medicine. The issues raised by concerned faculty with the Deans concerned the handing and retention of DNA samples, the imprecise language in the original announcement about the nature of the project as a ‘medical’, ‘scientific’, ‘educational’, ‘research’, ‘experiment’ , and/or all of the above. If the students were not research subjects, why were they being sent a ‘medical subjects bill of rights’ in their DNA test kit? If the purpose of the “On the Same Page’ project is for faculty and students to share an intellectual experience together why were the Letters and Science faculty not invited to “bring their genes” to the project? Why were 500 freshman who were also foreign students not included in the project? Soliciting consent prior to arrival, prior to discussion was also questioned. The concerned faculty also noted that quite a few of our incoming freshmen would be under 18 years of age which the Deans had not considered prior to our meeting. These concerns were not hostile to science but protective of it.
Dan — I think you should breath a sigh of relief that the “experiment” has been cancelled. This program was in a category with the infamous Ivy League nude photo scandal, and would have only left the personal genomics movement with egg on its face.
While you insist that the risks were hypothetical and overblown, the same could be said about the benefits. I think that Prof. John Hawks of U Wisconsin said it well:
Hawks may have said it well but he’s said it wrong. There are so many credible scientific studies on LCT, MTHFR and ALDH2 linking them to precise metabolic effects. They certainly “support the idea” that this genetic knowledge can improve health, of course they don’t prove it…they do “support” it. ALDH2 for example – approx 50% east Asians lack the protein and have a severely nasty reaction to even a small amount of alcohol (due to accumulation of acetaldehyde). The majority of these individuals tend not to drink, so no test required, it’s obvious. But individuals who are heterozygous for the particular genetic variation who won’t have the severe reaction and who drink have pretty high raised risks of some cancer. An example that I found in 5 secs at Huge Genopedia: http://bit.ly/b7qfta Note that A allele (risk allele) conferred an overall OR for esophageal cancer of 3.05, but ONLY in drinkers, and this OR went up to 11.93 in heavy drinkers. Genopedia is a good source of “credible science” for each of the genes (and many others) http://bit.ly/coFTYE
Apart from that it misses the point anyway – the study would have shown how gene-environment interactions work, how the effect of certain genetic variations depends also on diet, lifestyle and behaviour. i.e. that most genetic variation does not have a deterministic effect – it’s not just the genes.
It’s a pity that it didn’t go ahead, it would have been useful, the amount of mis-information and ignorance on genetics, environment and health demonstrate that
@Pete Shanks: Keith is correct. “…chilling effect…” is used to indicate the reality that uncertainty and hostility in this field are contributing to funding difficulties, and driving ideas and companies with potential to abandon plans or to pursue them elsewhere (i.e., overseas) or in a less innovative form.
Furthermore, I don’t think the GAO’s conclusions should be uncritically accepted by any means. See, e.g., this post on Genomes Unzipped.
@Debra Greenfield: You make an important point. Students – or any individual – must be encouraged to make informed decisions prior to engaging with personal genomics. The question, of course, is what is the threshold level of education and information required before participation is permitted?
That line is drawn in different places depending on the context. The Personal Genome Project – a true personal genomics research project – required that its first 10 participants hold a master’s in genetics or an equivalent degree. That requirement has since been reduced to inter alia the successful completion of an online entrance exam; a high bar, but not nearly as high as a post-graduate degree.
In the case of Berkeley’s program, the cohort comes in with a different – and arguably less sophisticated informational background – but the risk profile of participation is also substantially different. Could Berkeley freshmen make a decision on the basis of sufficient (in light of the risks) information to participate in the program? I think the answer is “yes”; recent experience indicates there is considerable value in integrating personal genomics services and education initiatives (see here).
The alternative you propose would lead to a sequential approach in which complete or nearly complete understanding is required before participation. Not only does that ignore the fact that participation itself may be integral to understanding, it also may not be feasible given the rapid rate of change in this field.
Finally, simply because informed consent has not succeeded in every instance (e.g., the Havasupai case) does not mean that the process itself is flawed particularly where, as here, you are drawing comparisons between examples that share relatively little in common.
@Neuro-conservative (and by extension, @John Hawks): The benefits of the program were never intended to be medical or clinical in nature; they were intended to be educational. This was not genetic testing designed to change lifestyles or improve health outcomes. It was a personalized approach to genetics education designed to drive learning and discussion. Take issue with the educational value if you must, but don’t confuse what Berkeley attempted to do with what clinical laboratories and medical geneticists do for their customers and patients.
Who funds the Center for Genetics and Society and why are they astroturfing the Genomics Law Report’s comment section? Take one look at their website. They are a radical organization which is implacably opposed to new technology. It’s hard to distinguish them from those on the far-right who talk about embryos as precious snowflakes, as even stem cells are too newfangled and “controversial” for Marcy Darnovsky.
The attack on Berkeley was coordinated by uber-Luddite Jesse Reynolds and they have not tired of crowing about it in their fundraising letters, meticulously listing every reporter they’ve ever duped into quoting them. Who watches this self-appointed “watchdog” and who is paying their bills?
Thanks, Dan, for this post. I think you’ve drawn a clear contrast.
I don’t agree with the premise that “risk-taking” is essential to progress in personal genomics.
Nancy Scheper-Hughes describes matters from what is — to me — the most relevant perspective: local specialists. Freshman education should not require an IRB approval.
Anthropologists have routinely involved introductory students in a number of genetic tests — blood typing, PTC testing, and sometimes other classical genetic markers. Some of these have mostly fallen out of use (HIV risk, expense of test kits, and greater concern for student privacy), but most are still used in classrooms at least somewhere. A few programs have introduced mtDNA or other ancestry tests in advanced courses. These tests have a substantial history and literature behind them, meaning that we have a well-developed curriculum to explain and contextualize genetic variation — including traits with high clinical relevance.
If the Berkeley program had been based in best practices, I doubt it would have attracted any attention. The average freshman is at much greater risk from sexting than from any unintended effects of a three-genotype screen.
So what’s the difference in this case? The program was universal (not part of an elective course), enrolled students in a clinical trial, and offered no privacy safeguards. Nominally optional, but what part of the university orientation process usually strikes incoming students as “optional”? Clearly the traits (lactase, alcohol flushing, and folate) were chosen because of their clinical relevance and their wide differences in frequency between continental populations. It’s not trivial to construct a correct lesson around traits like these in an anthropology course — I’d much rather have a random three SNPs from the 23andMe set, for example. Now imagine this “conversation” carried on campus-wide, without minimal involvement of people trained in human genetic variation. It’s a nightmare scenario, speaking purely from the educational standpoint.
And at the end, for all this effort, you’ve given students nothing of positive use for their own health. Yes, all three phenotypes have health effects. But genotyping provides nothing that can’t be summarized by general nutritional advice: “Eat leafy greens, drink alcohol and milk in moderation if you tolerate them well.”
I follow the comments with interest. I am a member of the general public who thinks the CA Public Health Dept is a menace and the genetics testing should have gone as designed.
All the drama against the testing amazes me. Where oh where are you people in the worst experiment in human genetics that goes unchecked and has for decades – sperm and egg donation. With nary a concern to the results, these donors just plop their genes out for someone else to use to create a tall, blond, brown eyed, athletic radiologist. With no way of finding out who the bio parent is or anything about the background. You worry about UC freshman? I worry about sperm and egg donor kids. They have been designed without their consent. This would never pass a human subjects review committee if presented in that context. This is not a plea for genetics testing in donor scenarios but a statement that here lies the real danger in genetics experiments, not the UC abandoned tests. They are nothing compared to donor kids. But since design a kid isn’t going away, of course there should be genetic testing prior and the kid should get the results.
My disgust with designing kids with only one bio parent in their lives has to do with human rights and not religion. I am an atheist, not right wing and I support gay marriage. I am tired of being pigeon holed so I mention this.