Posts Tagged ‘informatics’

DTC Genomic Testing: a Window on Society?

Tuesday, April 19th, 2011

I continue to believe that the discussion surrounding the Direct to Consumer (i.e. DTC) genomic testing is basically a healthy thing (pardon the pun).  It seems to me to be a window into the larger conversation about the role of health care in society and evidence that we do pretty well in letting information flow freely.  I think that’s a good sign of a free society.

The DTC genomics testing debate seems to have looped in a group of people that are: a) health conscious, b) sophisticated, and c) medical non-professionals.  These folks are interested in taking advantage of the health empowerment that information technology and genomic technology have provided.  In this endeavor, they are bumping up against the health care system, which is fairly conservative and has its own status quo.  This latter element is certainly something those sophisticates who have been empowered by genomic technology (via DTC genomics companies) are not necessarily interested in minding.  The good news is that the debate around genomics and health care empowerment has been civil so far.  The court in which the debate is being conducted is the halls of the US Food and Drug Administration.

The FDA had a meeting…

In between the summer of 2010 and the US FDA’s meeting in March 2011 to hear feedback regarding DTC testing, Amy McGuire and colleagues published an article in Science proposing a mechanism for regulating these new genetic tests.  Essentially, McGuire et al proposed a “risk-based stratification” of regulation of these tests.  This means that tests thought to carry higher risks of harm will faced tighter regulation.  For example, testing for breast cancer predisposition might be subject to significant oversight because women face the risk of surgery or other significant and potentially harmful healthcare consequences based on the outcome of the test.

On March 8-9, the FDA’s Molecular and Clinical Genetics Panel heard testimony from a variety of stakeholders of DTC testing.  On one side, as one would expect, were the DTC genomics testing providers arguing that these tests should not be regulated out of hand.  In fact, the representatives of these companies seem to be willing to accept the risk-based stratification approach that the FDA appears to be leaning toward.  Their backers further argue that individuals should have access to their own genetic information and that the FDA should not come down hard on a promising new industry.

On the other side were some significant groups, such as the American College of Pathologists, who argued that lax regulation could be harmful to consumers.  The significant risks of DTC genomic testing were pointed out, including privacy concerns and inadequate support for the inevitable health-related questions from the recipients.  These folks also raised other interesting new concerns that need to be vetted.  For example, inappropriate dissemination of genetic information could be harmful, both to the tested individual (through inappropriate use by employers and insurance companies) and to their relatives who share some of their genotype.

Others brought recent information to the FDA panel.  Subsequent to the initial actions by the FDA last summer, a number of studies have been performed to examine the behavioral consequences of DTC testing among consumers.  So far the results have not suggested any systematic negative consequences, such as anxiety, for those who undergo testing.

All told, my sense of the reports from the meetings was that it was a pretty fair exchange of information, which raised legitimate concerns both for and against DTC genomic testing.  Yet, there are a lot of questions still to answer about how these technologies will serve society.

What about prenatal sequencing and other ethical questions

Although I don’t have the answer to this question, the recent paper from Lo et al in which the genome of a fetus was sequenced, raises questions about the appropriateness of massive prenatal testing.  Prenatal testing for even a single gene faces tough scrutiny, so what do we do when we can test for thousands of gene variants?  In this example the testing was for a single gene mutation (beta thalessemia), but the result showed that an entire fetal genome can be sequenced using a blood sample from the mother.  One positive here: this testing approach is non-invasive.  However, one can imagine the potential demand for prenatal sequencing to determine not only disease susceptibility, but also “soft” traits, like presumed intelligence.

Another question that I think may be under appreciated by people in favor of DTC testing (as pointed out by CAP) is the impact of the results on those genetically related to the tested individual.  What if those people, brother and sisters, are not interested in their genotype or would rather not know?  Does one have the right to post their genotype on the internet?

An extension of that idea came to life with the West family, in which each of the four members of the family underwent whole genome sequencing.  Although the genomic sequences of the kids are not public, one of the parents did submit their genome sequence to the NCBI database of genome sequences, thus releasing half of the genetic info of the kids.  Is that a privacy breach?

Another twist to that theme revolves around research on parental choice with respect to DTC testing of their kids.  A report from Tercyak et al in the journal, Pediatrics, suggests that parents who themselves undergo DTC testing are more likely to have their kids tested.  Parents were also more likely to favor testing if they thought their child was at risk or if they had a positive risk-benefit view of DTC testing (duh!).  Given the variability in quality of follow-up with these tests, it seems fair to question the use of DTC genomic tests in children.

Social liberty?

Not to overdo it, but what I find interesting in the DTC genomics debate is the renegotiation of power between health care consumers and health care providers.  It’s a bigger trend than just DTC testing, as evidenced by broader trends in consumerism in medicine.

One of the key arguments heard from those in favor of DTC testing is that they have the right to know information about themselves.  It’s hard to deny the truth in that.  But, there is definitely a balance that must be struck between individual liberties and the welfare of society.  The discussion of what to do with DTC testing is an interesting place to listen to that discussion in a very personal way.

A Couple of Glimpses at the Genomic Future

Thursday, January 27th, 2011

There were a couple of stories in the news in the last few days regarding genomic/genetic testing that hit on recent controversies surround said testing.

In earlier posts, such as this one, I commented on the now world-wide discussion about the balance between regulation of DTC genetic testing and innovation.  A new browser plugin for viewing genomic data for data from Direct to Consumer testing company, 23andMe, was released by 5AM Solutions.  This plugin does some add-on processing of web pages as they are loading, such that single nucleotide polymorphisms (SNPs) mentioned by their common abbreviations are highlighted.  In addition, if one mouses over the highlighted SNP a balloon appears showing your genotype at that SNP, as well as links to SNP analysis resources.  This development is reminiscent of other internet technologies, such as Facebook and Linked In, where third party companies develop an overlay to platform software or datasets.  Only this time, it is personal genomics.  This little program represents the next step in mapping your genotype onto the mass of information available out there about genes, health, and life.  One can imagine where creativity may take personal genomics if it is not over regulated.

The second story could raise the specter of universal genomic testing in some people’s minds.  It certainly did touch that nerve for me.  The Department of Defense advisory group, Jason, produced a report that was released last week advising the DoD to move toward “’take a leading role’ in using personal genomics data”.  Among the goals that the report suggests is “eventual collection of complete human genome sequence data from all military personnel”, based on the notion that $100 complete genomic sequence will be available in the near future.  The supposition is that this information will give an advantage to the military in its missions.  If that is the thinking of the DoD advisory scientists with respect to genomic sequence data, then it is hard to imagine that the trend will go any other way than towards consideration of complete sequence data on individuals under other circumstances, too.  Given recent discussions on genome hacking we might want to keep an ear to the ground with respect to routine genome sequencing and its uses.

Recent Developments in DTC Genomic Testing

Wednesday, January 5th, 2011

It’s been relatively quiet on the DTC Genomic Testing front since the turmoil over the FDA’s decision to consider regulating DTC genetic/genomic testing last summer died down.  Nevertheless, there have been a few news items that yielded glimpses of what is going on in the industry.

In November 2010, 23andMe received additional investments from Google, New Enterprise Associates, and Johnson and Johnson Development Corporation.  In the last few days, the company announced that it would “experiment” with new pricing models.

In mid-December 2010, the Institute of Medicine (US) announced that it would be researching, then issuing a report on possible ways forward for and possible legitimate uses of predictive ‘omics-based tests.

Also in mid-December, DeCode Genetics published findings in Science Translational Medicine of a study of SNPs related to PSA expression in prostate cancer.  The DeCode group interprets their results as refining the utility of the PSA protein marker for prostate cancer.  In essence, the SNP markers described can be used to individualize PSA test results by predicting whether the individual’s range of PSA expression can be expected to be higher or lower than average.

Genome Hacking

Wednesday, January 5th, 2011

How genetic information will flow and be used is of interest to me in this environment of ever-cheaper and ever more rapid DNA sequencing.  What used to be fairly privileged, esoteric information is rapidly becoming a commodity item.  How will this affect our daily lives?

The prediction of heath states in the future, which is one of the uses of genetic information, could be of interest to not only the person whose DNA is being sequenced, but also of others who interact with this individual, for example, political opponents.  This was highlighted by stories from the Iceland Review and the Guardian, based on confidential cables leaked by Wikileaks.

The former story discusses a cable from the US embassy in Iceland regarding the possibility of Chinese spying on DeCode Genetics.  The latter story discusses a US State Department cable regarding collection of “biometric data”, including DNA, among other things, from individuals abroad.  Neither story reports on confirmed “DNA identity spying”, if you will, but together with stories, such as this one from the New Scientist, in which one reporter “hacked” the genome of another reporter by “surreptitiously testing” his DNA (with his knowledge in this case), these stories raise the specter of a flood of unauthorized DNA sequencing.

What could the response to this potential loss of privacy?  I wonder if at least some public figures would proactively have their genomes sequenced in anticipation of someone sequencing a sample of their DNA obtained in a clandestine manner.  I imagine most would find it better to control the circumstances of release of that type of information than to leave that control to some potentially unfriendly party.  In the extreme, would this ultimately lead to many individuals proactively sequencing their genomes to retain control over how to “spin” their health status?

So far, “genome hacking” does not appear to be a rising tide.  Since genomic information is really just another form of biometric information, like blood type or eye color, it may never create the huge stir that some fear.  As the conversation continues around how to use genetic information, this will be an interesting area to watch.

Genomic-level Testing: the Debate Continues

Friday, August 20th, 2010

This is an interesting debate unfolding here: what to do about Direct To Consumer (DTC) genetic/genomic testing.  The two sides seem to be

  • It’s my DNA and I want to know what’s there
  • You may not get correct information on your DNA and what will you do with it anyways?

I don’t think there is substantial objection to an individual knowing what the sequence of bases is in their DNA, however, the technical adequacy of DTC tests is an issue.  Will you receive the correct information about the sequence of your genes?  At this time, that is not assured.

The second big issue, and the one that has the greatest implications for health care delivery in the future, is how genomic information will be used.  I’m fine with people sequencing their own DNA, but what happens next has implications for those beyond the individual who had their DNA sequenced.  As pointed out by Annes and colleagues in NEJM (Risks of Presymptomatic Direct-to-Consumer Genetic Testing, 18AUG2010), there could be liability and other issues for the physician to consider when a patient brings in their DNA sequence information for consideration.  If the utility of that information is uncertain, then many unanswered questions will linger.  In their example, what if the 30-year old man presents sequence data with a “prostate cancer risk allele”?  The clinical utility of such information is currently unknown.  Does the physician order additional testing (and thus incur additional health care costs) or does the physician take no action.  What if the physician takes no action and some years later the patient is diagnosed with prostate cancer?  Is the physician liable?  If the decision to sequence one’s own DNA affected only the individual I think there would be little objection to the expenditure.  But it doesn’t affect just that individual.

The implications for our healthcare delivery system of the shifting of health care information access, as exemplified by DTC genomic testing, are also interesting.  In a certain sense this debate is the most visible current example of the ongoing negotiation between healthcare consumers and providers regarding ownership and cost.

As pointed out by Evans and colleagues, (Preparing for a Consumer-Driven Genomic Age, also in NEJM on 18AUG2010), patients increasingly know more than their physician about specific genetic topics.  This shift has been enabled by widespread internet access.  This empowerment of consumers by transfer of knowledge ownership from doctors/healthcare professionals/industry to patients should rebalance the value equation in favor of patients/consumers.  This in turn will reduce the cost of health care in the long run.

Cancer Genes in the News Again

Tuesday, March 30th, 2010

There seems to have been quite a bit of activity in the world of gene-based cancer diagnostics lately.  Just in the last 24 hours some of Myriad Genetics’ BRCA patents were ruled invalid in the United States ( and a recent article in the New England Journal of Medicine (Wacholder et al., NEJM 362: 11, 2010) suggested that breast cancer-associated SNPs add little to traditional models for assessing prognosis of being diagnosed with breast cancer.  These two developments illustrate two aspects of the ongoing debate about the genetic information in health care: how should genetic information be used and who owns it?

On the “who owns it?” front, the Myriad case is all about the patenting of genes.  This debate goes back to the early days of large scale gene sequencing when companies like Myriad Genetics, Human Genome Sciences, Incyte Pharmaceuticals, etc. were filing patent applications on gene sequences nearly as fast as they could sequence them.  The question then, as now, was “what is the invention here?”  Are naturally occurring gene sequences patentable?  This recent ruling against Myriad is of the opinion that naturally occurring gene sequences, even “in isolation”, are not inventions and hence not patentable.  Seems reasonable to me.  In my experience patents and the resulting restrictions on use of something as fundamental as a gene sequence do inhibit research, primarily by driving up the cost.  My feeling is that there are plenty of opportunities for inventiveness, and therefore intellectual property, a step further away from the gene sequence in the form of compounds, devices, and methods for diagnostic and therapeutic products.

The BRCA gene alleles in question appear to have quite a bit of information regarding the phenotype of the person in question.  This implies value in so far as that information can be used to assess the future risk of developing breast cancer.  Alleles like the BRCA 1 and BRCA 2 are fairly rare, though.  Clearly it would be good to have more broadly applicable genetic tests for breast cancer and other diseases.  A number of academic laboratories have published studies suggesting that “gene signatures” are useful in classifying breast cancers (see for example Sorlie et al., PNAS 100: 8418, 2003).

Quite a few companies and other organizations are developing gene signature-based tests in this area, based on the concept that combinations of more commonly occurring alleles will provide sufficient information to facilitate improved assessments of risk.  The first of these tests to reach market is the Oncotype Dx test from Genomic Health, Inc.  Several studies have been published on the Oncotype Dx test demonstrating the utility of the Recurrence Score in facilitating risk assessment.  The product is beginning to gain traction in the market.

The publication from Wacholder et al (above) suggests that combinations of several single nucleotide polymorphisms know to be associated with breast cancer adds little to traditional models for assessing risk of developing breast cancer.  While superficially it appears that the Wacholder study and other studies, such as Oncotype Dx studies, conflict I’m not sure that is true.  It may be that the question being asked in these two situations is different.  The Oncotype Dx test is looking at risk of metastatic disease, while the Wacholder study is looking at risk of primary breast cancer occurrence.  Further, the modeling methods used are different.

The upshot to me is that we are not done yet understanding how to use genetic information in making decisions about health care.  In a few cases it is fairly straight forward.  However, to reach the expectations set for genomic information a decade ago we will have to be able to use genetic information to assess health issues much more broadly than a handful of rare alleles.  This is proving harder to do than I think many expected.

The Future of Medicine?

Wednesday, March 10th, 2010

Leroy Hood proposes the “P4” view of medicine in an interview in MIT’s Technology Review ( The four P’s are Powerfully Predictive, Personalized, Preventative, and Participatory. These apparently will be the underlying concepts in, as the article puts it, a “revolution in medicine”. Really? I’m not so sure.

Don’t get me wrong. Leroy Hood is one of my science heroes: he was a rising star when I was a graduate student and post-doc. Subsequently, he has offered ideas and generated energy in areas of molecular biology that I really like, having helped invent automated sequencing and started an institute to promote systems thinking around cell and molecular biology. He is one of the smartest people that I know of.

However, in my mind there is a strong case that three of these “P”s, Predictive, Personalized, and Preventative are really not new. These concepts in medicine have been the subject of discussion and goal of improved medicine for many years. The impact of genomics and informatics on these three will be, in my opinion, an extension of work that predates DNA sequencing and computers. The one “P” that I think may be revolutionary is “Participation”.

Predictive describes what we all want our tests to be. If I take a glucose tolerance test, my physician is attempting to gather information on whether I have diabetes or not. With that information in hand and armed with knowledge of the natural history of people with or without the disease of diabetes, my physician can “predict” what my health state is likely to be at some point in the future (under certain assumptions). If you think about it, there is no test that can actually predict the future. They are all educated guesses with some probability of being wrong attached. The same will hold true for genomic based tests. Sorry, 23 and Me. More on this in a future post.

Personalized and preventative are likewise extensions of long-standing trends in medicine. Physicians have been working to personalize care for eons through practices such as prescribing different doses of drugs for different size people. And, as has been discussed for the last decade, there are some very powerful disincentives for personalizing therapeutic products too much: highly targeted market = small market.

Preventative even has its own cliché (I prefer to call it “distilled wisdom”): “An ounce of prevention is worth a pound of cure”. Whether the technologies mentioned in the article will have the impact on these long-standing trends in medicine that the hype has built up for them remains to be seen.

For my money, the one “P” that does have a case for being revolutionary is Participatory. For a long time patient participation in medical decision making was pretty much discouraged. Certain shifts in social structure, along with the huge increase in the availability of medical information in the last 30 years have opened the door to patient participation in health care decisions. Now, if I am a man, I have available to me more information about prostate disease than I can possibly absorb. Similarly, women can learn all that they want to know, and more, about breast cancer if they have access to a computer with an internet connection. Whereas one used to rely on one’s physician for the required information on whatever health problem, now the patient has the capability to acquire information on their condition and the available treatment options. As a result, each of us, in most of our health care situations, is empowered to participate in the required decision making. Is this where informatics will have its greatest impact on the revolution in medicine, rather than in terabytes of sequence data?

So, perhaps the real revolution in medicine is in patient education for its ability to facilitate participation in one’s own health care. As discussed in this recent article, nearly all women who received information on their breast cancer from the Oncotype Dx test from Genomic Health, Inc. were glad they took the test. However, only 2/3 to ¾ said they understood what the test was telling them, even when explained by their physician. So, to really participate in their health care decisions, these women need to understand what breast cancer is, what genomics is and what it can tell them about their disease, and how this information can support their decisions. I believe that from a collaboration of an informed, educated patient and their physician will come the best health care decisions of all.