Skip Navigation
NIH

« Back to White Paper Page

Comments on "White Paper #1: Applying Genomics to Clinical Problems-Diagnostics, Preventive Medicine, Pharmacogenomics"

Nice set of questions, no glimpse of a plan. I am not sure what feedback you wanted with this posting. Questioning the questions you posted? Pulling it together into a coherent plan? Putting together a 'straw dog' of how we might get from here to the future will be valuable and then many of us can help shape the plan.

(59) Friday, December 19, 2008 9:48 PM


There are many good questions posed.
I think that genetic counselors already have developed methods of communicating with patients to minimize anxiety and stigmatization. Prohibiting direct to patient transmission of genetic results and requiring a genetic counselor to be part of the transmission of these results to patients should be part of the response to question #3. There should be no debate about this.
As a clinical researcher investigating a potentially life threatening genetic condition, and as a practicing physician, I know that most health care providers will want to have web based access to the same information that their patients have regarding their genetic condition and more detailed information about the implications of this data. I think it is extremely important that the variability in expression of genetic traits is emphasized.
A large problem is that many variants of unknown significance (VUS) are observed. It is often difficult to determine the implications of these VUS. Clinicians who can identify and report the clinical phenotypes associated with VUS in families should be involved in discussing possible health implications with the families.
Individuals negatively affected by the phenotype associated with a genetic condition are often motivated to support data collection that will help define the implications of both VUS and accepted mutations. How can this data be collected over years from throughout the USA? Can the government provide support to the small organizations that already do this work?
Data on the prevalence of a genetic condition is important for the application of a genetic screening test to a large population, but in many cases more efficient application of a test will be to members of a family in which a genetic condition has already been identified. In this case the prevalence of the genetic condition in the population at large is not important and should NOT limit the utility of the test. It is very important that further development of genetic testing address the question of what can be said to the family member who does NOT have the familial genotype that has been associated with the familial disease. When is enough known about the negative predictive value of the test and expression of the genetic condition that the 'negative' family member can be said to have NO increased risk of the disease.
Why should payers need any more information than the worst cost of the unrecognized, untreated condition, including malpractice claims, to motivate support for genetic testing and documentation of genetic counseling? Costs for long term care of disabled individuals are also relevant. Costs for identifying an individual with a known genetic risk can be kept low by focused testing for the familial variant exon.
Logically it is best to expand genetic testing through families, but small family size and the fact that many families are separated by divorce and other forms of discord, or simply geography, are limitations to this. Therefore genetic counselors should be able to develop databases that can be shared across state lines.
Question 6 is very important. It appears that at present it is very easy to obtain genetic data that has no valid interpretation. The sources of this data should be taxed to support the efforts needed to provide valid interpretation.

(61) Saturday, December 20, 2008 4:47 PM


Thank you for this thoughtful white paper. All white papers should have a section on ethics with appropriate references. This may have been written before but there is no reference in this paper. A section on ethics would provide clarification when ethical questiona rise.

(62) Sunday, December 21, 2008 3:54 PM


A major and important theme that appears to be missing in this white paper is regarding children.

Presumably, preventative medicine will need to be practiced on children. The need for genetic testing of children, and implementation of interventions in children, raises many questions that do not seem to be addressed in this white paper.

To provide an example, many GWAS studies have focused on type 2 diabetes and associated risk factors. If one considers an 60 yr old adult type 2 diabetic in a hospital setting for medical complications of their T2DM diagnosis, GWAS-informed genotyping for 'risk for T2DM' is unlikely to help either the patient or clinician.

On the other hand, 20% of males in University fulfill the diagnosis of metabolic syndrome (pre-diabetes), and 50% of 6th graders show abnormal threshold values for one or more metabolic syndrome features. How will testing (presumably population screening) be done in children, and interventions implemented?

(64) Monday, December 22, 2008 11:03 AM


This is an outstanding list of important questions. My major concern is that the word "trial" is never explicitly used. There is a great deal of concern within the cardiovascular community that high-technology diagnostic tests are being let loose on the market without adequate evidence that their use improves clinical outcomes or public health. Although this argument is typically targeted towards imaging tests, I think the same line of reasoning can be used when thinking about genomics. Therefore, I think would be perfectly reasonable to ask as an explicit question, "When is it necessary to perform randomized trials to determine whether genomic data improve patient outcomes or public health?"

(66) Monday, December 22, 2008 11:44 PM


Great dialogue to get going. There might be some discussion of the impact of epigenomics and the microbiome on diagnostics and predictions as well. Both of these are non-static and so unlike some aspects of germline sequencing or SNP/CNV typing may need to be done each time a patient presents. How will payors, patients, docs deal with the ongoing need to evaluate the role of the changing epigenome and microbiome in disease modification?

A second aspect is that of the genomes successes applied to not only diverse populations within the US but also overseas especially in low SES areas. While diseases like Malaria still present major international challenges how will the limited funding availalbe be reflected in the genome portfolio applied outside the US?

(68) Tuesday, December 23, 2008 8:42 AM


Many of the questions posed are common types in many fields other than medicine. Decision theory, micro economics, management science, and operations research all address these types of problems and have developed approaches and methods to assist in the effort. Involving these professions would be appropriate to the search to answers.

(69) Tuesday, December 23, 2008 11:20 AM


It might be useful to examine the impact of such genomic data on commom complex disorders. Since risk loci may account for less than 5% of the variance in disease presentation, while environmental effects alone and interacting with genotype might account for more substantial impact on disease liabilities, it is unclear how genotypic information will advance clinical treatment. On the other hand, pharmacogenetic data will have more practical and immediate impacts on clinical practice and patient care. I suspect even the insurance industry would view the inclusion of pharmacogenetic information as being "cost effective" since the use of such information will limit adverse side effects and aid in patient outcomes. My view is that this effort should lead with a pharmacogenomic initiative for pragmatic reasons.

(74) Monday, December 29, 2008 4:42 PM


The above questions are all reasonable ones that future research will need to provide answers (or in some cases partial answers to) I would also like to suggest that education of the public is extremely important, and this most start at an early age. I suggest that it is within the scope of NIH to ask, how are advances in genetic and epigenetics best presented to school children and young adults. I submit that the biggest issue that we face today is that we have a very large percentage of our population who do not understand the issues and complexities of these advances. How do we best educate our young people so that they are capable of sufficient understanding the science of genetics to make reasonable informed decisions for their own health?

(75) Tuesday, December 30, 2008 2:28 PM


The primary problem is engineering a paradigm shift from germ-theory medicine (single agent diseases model with the etiology discovered by reductionist approaches) to mathematical models of disease risk and targeted intervention (reverse engineering with system testing). The genetic information generated by genotyping cannot be understood independent of other genetic, metabolic and environmental factors ý framed within the context of the physiological functions and controls of relevant organ and systemic systems. To accomplish this paradigm shift, there must be extensive modeling, and experimental testing of these models in human disease. However, this is an engineering problem ý not an epidemiological problem. Since much is already known about the mechanistic properties of the biological systems under consideration, the testing can be compartmentalized and the outcomes predicted and accurately measured. This allows for human trials to be reduced from tens of thousands of individuals to a few (or one) markedl reducing the time and costs.

The question of marketing a new "product" to physicians and insurance companies is answered when you have a simple process of developing a "personalized prescription" for health, and accurate biomarkers to detect pathologic activity that trigger application of effective therapy. Financial modeling will affirm that "an ounce of prevention is worth a pound of cure". As models of nationalized health care are being developed, emphasis should be focused on disease prevention using this engineering-based preventative approach because expensive salvage treatments and operations will become increasingly expensive and must eventually be rationed.

(81) Thursday, January 1, 2009 6:53 AM


This document has a strong psycho-social bent, suggesting a strong emphasis on prospective behavioral research. As such, significant consideration for enhancing and elevating the role of the ELSI program and its budget would be warrented. The document is also strongly focused on disease. While this is important, it seems we know so little about the resilience of the human genome. Can we not begin to focus on "resilient genomes"? What genomic factors are in play when "at risk" individuals do NOT develop disease? What protective epigenetic events are tiggered with healthy behaviors/lifestyles? We have a huge section of our population moving developmentally into a phase of their life with concerns for generativity - many very healty - can we not engage these men and women in networks of research on the genetics/genomics of healthy aging?

(85) Saturday, January 3, 2009 1:49 PM


one of the key issues that i see as relevant to deployment of population based genetic is the lack of standards for software tools that will be used on the resulting libraries of sequence data. It seems that with the very short term possibility of actual full genome, mass population techniques, we will require a "gold standard" of tools so that apples can be compared to apples.
As these new rapid sequencing techniques take hold , during 2009/10 we will be faced with tens of thousands of full genome data sets. The compute requirement for such data bases needs to be addressed.
The key aspect is affordability so that rapid deployment can be achieved. The reality is that $1000 to $20,000 genomes will open a flood gate of highly relevant new critical data. Mass genomics has not been done before. This raises issues of storage, open access and the ability of the "system" to put the right data in the right hands. it is not reasonable to assume that key research personnel will be able to deal with the mass data. There job is to find new relationships which, until now, have eluded us.
We have seen not only the massive cost reduction in the cost of base sequence costs and now we also see an equal reduction in the compute side of things with such advances as Gpu based compute platforms. I would therefore like to suggest that a certain amount of time is spent thinking through the impact of new compute paradigms and the software they will use. it is nearly impossible to deliver cost effective solutions when there are no gold standard software tool that will address the demands of dealing with a data flow that will dwarf all other human information endeavors.
It is great to see a planning process in action. but it must consider the enormity of the coming problem.

(88) Sunday, January 4, 2009 6:21 PM


Additional Questions by main question number:

1) In the first and third paragraphs: how do we communicate these combinations more clearly and usefully to patients and clinicians?

2) What should and should not be patentable or be kept as proprietary information?

3) How do genetic test findings compare with family history, other lab findings, epidemiologic data and other current medical tools in changing patient behavior? What level of risk changes behavior?

How do we integrate the non-directive ideal of genetics with the very directive world of clinicians faced with 10-15 minute appointments? Do patients actually value this non-directive approach, or is it just the providers who value it (because the understand the misuses that have occurred in the past)?

How will clinicians consider cost in the testing equation? Are just economic costs considered (what ELSI issues can they identify?)? What ýmarketingý (how the recommendation is presented to them) will influence them one way or another? For example, would a recommendation from a colleague outweigh a recommendation from their EHR? Should recommendations from EHRýs be linked to outside sources? How will report formats and resources in those reports influence further testing?

How comfortable are clinicians with explaining statistical or chance data to patients? Does the comfort level change with patient socioeconomic or educational status? How effective is that statistical communication to a patient, and how can it be improved? How does that change with patient socioeconomic or educational status?

5) Are there effective ways to mitigate these barriers? Where has that succeeded before and what can we learn from those experiences?

(90) Tuesday, January 6, 2009 9:35 AM


for question#2: another issue regarding paying for laboratory testing is the area of laboratory quality, qualifications of testing personnel and individuals providing interpretation of test results. Genetic laboratory is considered high complexity testing, requiring substantial experience to provide quality diagnostic work for patients. How will patient test result quality be monitered and what oversight will be implemented? I'm thinking about the current situation with nutrigenomics and other "genomic testing" being provided for medical decision making, and also for the "push" for genetic testing to enter the general hospital laboratory, where there is no technical or doctoral level staff with experience in troubleshooting, test performance, test result interpretation or clinical knowledge in which to put results in context. We can easily do more harm than good for patients if this issue is not given some attention at the federal level. Related to this is the reimbursement and test cost issues, as the QA/QC and knowledge required for accurate genetic testing results costs money......

(93) Wednesday, January 7, 2009 4:11 PM


Intro first paragraph: Test availability, potential for reimbursement and cost-effectiveness are presented as separate from the benefits and risks. They depend to a large degree on the evidence of benefit and risk.
Intro second paragraph: No mention of the risk of tests emerging into the market (DTC and to clinicians) with little or no evidence of utility.
Q1 1st p. The talk of setting boundaries, seems to require a more fundamental consideration: what is the minimum set of information necessary to address these questions.
Q2 1st p. Most of the answers to the questions here are not different for genetic/genomic tests compared to other emerging tests and technologies that payers are asked to reimburse. One exception are plans that have a blanket genetic test exclusion (that is no genetic tests are covered irregardless of evidence). Another interesting question that may be out of scope is what defines a preventive genetic test as opposed to a diagnostic test? This is important as Medicare excludes coverage for all preventive services.
The last two paragraphs are really important.
Q3 First and fourth paragraphs are very important. The second paragraph is a bit 'exceptional' in that it generally takes 10-20 years for new practice changes to take effect genetic or not. Relevant to this in the third paragraph is the need to study new educational models including the role of point-of-care just in time tools.
Q4 Need to add questions about impact of this on provider workflow. There is a good body of evidence on this in the informatics literature, so partnering with informatics experts would be important to make sure all the relevant questions are on the table.
Q5 Another ethical issue relates to a couple of instances where law enforcement request samples obtained in the clinical setting in order to obtain DNA for use in a criminal investigation. In the third paragraph a balance of protection of employees with the good intent of some companies to use information to protect employees from harm, as opposed to discriminating against. The non-genetic example of attempting to keep pregnant women from being exposed to lead in a battery company could be instructive in identifying questions.
White paper is very well done and should be a good starting point for discussion.

(95) Wednesday, January 7, 2009 6:28 PM


Consider asking: how do we define the contexts in which genetic testing is most important, useful, and cost-effective? (for example, individuals with familial history of a disease phenotype with known SNPs of high penetrance, and for which effective prevention or early detection/treatment make measurable differences in outcome).

(96) Wednesday, January 7, 2009 6:59 PM


Thank you Dr's Valle and Manolio for starting an important discussion. As a (rare) solo non-university-based practicing clinical geneticist, I think it important to interject that we, as a nation, are not doing a very good job of providing uniform, available genetic services to the US population at present, let alone the envisioned future. Furthermore, the resources for doing so are increasingly scarce, at least in my two-large-state universe.
While the questions posed here are important in an ideally functioning care model, they are nearly moot in the current atmosphere of > 6 month waiting lists for those with rare disorders who manage to float above the primary care and insurerýs referral threshold. The key problem here is the financially unviable clinical genetics practice model, which is unfortunately out of the scope of NHGRIýs research. I suggest that research answering the questions posed here will hold mostly academic value until the underlying delivery issues are ameliorated.

(98) Wednesday, January 7, 2009 7:33 PM


The authors of the recently passed GINA legislation did a fantastic job, with one exception. The loophole allowing insurers to use genetic information they obtain ýincidentallyý is, in practice, HUGE. It must be closed. Until then, the protections afforded by the legislation are dependent on low-wage medical record workers segregating out genetic information - which they are unlikely to be able to identify - before responding to insurance information requests. That isnýt happening now and isnýt likely to in the future.

(99) Wednesday, January 7, 2009 7:37 PM


These are certainly thoughtful and thought-provoking questions... and they primarily come from the frame-of-mind that our inquiries into genetics is primarily to generate and provide information to people about 'risks,' 'problems,' 'needs for interventions,' and the like. Much of genetic counseling is the communication that risks are lower than feared by individuals (whom we automatically put into the role of 'patient'). In this context, how we think about the information we want to both generate and communicate significantly changes... as do the questions that fall under this heading.

From the meaning of 'individual differences' to the grand questions of 'what's the nature of human nature,' the future shape of genetic medicine, the framing of screening and diagnostics, and the other relevant tributaries to this white paper are, I think, ready to put into the mix.

(101) Wednesday, January 7, 2009 8:03 PM


The white paper is well thought out and tackles the major issues involved in personalized medicine. We probably need to address the ethical challenges that arise with genetic testing of individuals particularly with reference to who gets access to the genetic information besides the patient (or the guardian in case of children) and his care provider. This may become more important when genetic basis of susceptibility or resistance to infectious diseases becomes widespread.

Due to ethnic variations in SNP frequency and CNV it may be appropriate to address this issue as well. Since most SNPs associated with disease account for just a minor proportion of risk in particular populations both clinicians, patients and advocates of direct genetic testing need to be aware of that before causing undue alarm on the basis of a single test.

(102) Thursday, January 8, 2009 7:42 AM


Honestly? Genomic medicine seems quite irrelevant. People are getting tested for the thrill of it, and have no use for the outcome data. This, according to Duke's own Genome Center research. Why not put more energy/money/time into something real - like the new therapeutics coming out in post-translational control mechanisms?

(103) Thursday, January 8, 2009 8:40 AM


The importance of all questions considered in the paper written by Dr.D.Valle and Dr.T.Manolio is undisputable.
On my opinion, to achieve more fast success in resolving these questions it would be desirable to speed up the genomic research in populations with highly prevalent deleterious founder mutations to find out modifying risky and preventive alleles, which undoubtly exist. These results could demonstrate the significance of genetic testing in medical care and genetic consultation and make it more easy to explain the importance of such testing.
More information in media about success in one area will make genetic tests more acceptable by patients as well by clinicians.

(104) Thursday, January 8, 2009 10:07 AM


As it has historically, this discussion does differentiate between a test for a genetic disease and a test for genetic susceptibility, which are very different issues. The paper is very focussed on "the test" (a simple technical issue) and provides no discussion of the major issues involved in the impplimentation of programs to modify behavior, the only way that identifyong susceptibility will improve health and thus be cost effective.

(110) Monday, January 12, 2009 10:43 AM


The work in summarizing important questions is appreciated. These additional questions may cut across the white papers.

1. To what degree should NHGRI/NIH/HRSA/AHRQ, those with
patents on genetic tests/drugs, others fund:
-"public access"lab(s) for "rare" diseases (blood bank
model with near universal access when need occurs)
-extramural programs for clinical training in genetics
(pre/post doctoral to genetic counseling to nurse
specialists to care coordinators to specialized
certificates/programs in many areas)
-coordination of guidelines on genetic assessment/testing
with whatever emerges re a payment plan(s) for health
care (e.g.what would pay for performance look like in
genetics and who would care?)
2. If the genetic/genomic test leads to more specific therapy
paradigm is the economic driver for much of the research,
what determines balance of publicly available
services (e.g. newborn screening) and likely to be too
costly therapies for all of those with the need (given
patents or other high profit driver(eg price
increases in thalidomide?)
3. Does ELSI direction/percentage need to be examined regarding
trust, access, and clinical utility (TACU as a domain)?
4. If we were starting HGP/NHGRI from scratch, what would we do
differently?

(114) Tuesday, January 13, 2009 8:40 AM


Thank you for a well thought, comprehensive set of questions in a complex area.

Other issues that needs to be addressed:

Oversight of genetic testing: There are major gaps in oversight of genetic testing in the U.S. and these are well documented in the recent report of the Secretary's Advisory Committee on Genetics, Health and Society on this topic.

Analytical validity of these tests: As many of these tests are proprietary, the data on analytical validity is not available. Without this information, it is impossibile to determine the competence of the laboratory performing the test.

Clinical validity: What proportion of the common adult onset disorders are actually attributable to genetic variants of variable contribution? Clearly, other factors also contribute - lifestyle, environment, etc. - so it's important to clearly establish the significance of the genetic component which is likely to be quite different for different ethnic groups as pointed out in the White Paper questions.

Clinical utility: This is also alluded to in the White Paper but needs to be further expanded upon. What are these tests really going to add to the care of the patient? The answer is likely to be quite different for different tests or classes of test. And, it will vary depending how much effort someone has to put into changing their life depending on the results. Face it, we all know we should eat less fat, drink less alcohol and not smoke but how many take that advice until forced to? On the contrary, a pharmacogenetic test that clearly defines the better treatment option is likely to be of higher clinical utility because it requires little personal effort.

With respect to early adopters of DTC testing, I believe it is currently largely driven by curiousity. There is an enormous amount of media attention around this testing and those with sufficient financial resources may choose to have testing simply to see what all the fuss is about. However, with the price of such testing dropping to within the reach of many consumers (i.e. 23andMe's $399 genomic scan), more individuals may choose to have this type of testing rather than "legitimate" single gene tests because they believe they are getting more for their money without realizing the limitations of the tests. This increases the danger of such testing, in my opinion.

With respect to the best vehicles for education of both physicians and consumers, clearly the internet is the only workable option. The suggestion of a "GeneTests" like database is a good one but it is imperative that it be comprehensive (an onerous task given that there are currently >1600 inherited genetic tests available and this number increases by 25% each year) and that it be continuously updated with the most recent data regarding the test, recommended patient population, applications, analytical validity, clinical validity, clinical utility, ethical, legal and social implications, limitations of testing, professional society recommendations, regulatory status and any other relevant information. Such a database could also be linkable to EMR and have built-in algorithms when a physician orders a particular test.

(124) Tuesday, January 27, 2009 11:30 AM


Responses are made to individual sets of questions:
1. The question set seems to be based on an unproven assumption: genetic testing will supplant other existing clinical tests in the foreseeable future. Beyond the incredible cost of this conversion, the discovery of phenomena such as epigenetic modification, copy number variations, rare variants, and microRNAs, together suggest that DNA sequence information will provide limited diagnostic value by itself. This is even before considering the incremental effects and low frequency of many identified risk variants. Even when viewing genetic results so far, is it reasonable to suppose that a diagnostic test is likely? For instance, an analysis of combined genomic association studies for Crohnýs disease have revealed about 30 genetic variants which together account for about 20% of the diseaseýs heritability according to Dr. Jeffrey Barrett. Assuming that the other portion of variants will be even harder to find, does the trend suggest that genetic testing will really provide a significant diagnostic value in the foreseeable future (notwithstanding the research value in terms of understanding mechanisms of the disease)? Similar trends have been noted for a variety of mental disorders, CVD, and diabetes, to name a few. Moreover, clinical evidence indicates that family health history is a relatively robust predictor of certain disease risks at the present time. For colorectal cancer, family history is a more robust predictor (i.e. higher odds ratio) and familial forms of CRC are more common than hereditary forms. In fact, family history alone is sufficient to recommend an early intervention (colonoscopy before age 50). It seems ironic that a relatively simple ýgenetic testý is still not widely used in clinical practice, as the search for more sophisticated tests continues.
An alternative model for introducing genetics/family history into health care on a broad basis would focus on a transitional approach that requires better integration of genetic testing and family history assessment as part of ongoing clinical testing. This approach would require more research on the mechanisms that underlie familial forms of diseases such as colorectal and breast/ovarian cancer (which is more common than hereditary forms), a better cellular and mechanistic understanding of incomplete penetrance and variable expressivity, attempts to correlate traditional clinical test results with underlying genetic/familial predispositions, and a greater emphasis on detecting early indications of disease onset for those at elevated risk. Animal models such as Drosophila and mouse could provide useful insights about the mechanisms that underlie incomplete penetrance and variable expressivity: How is penetrance affected by different genetic backgrounds? Can the causes be attributed to individual variants, or alternatively, are other mechanisms involved (such as methylation)? How would varied or novel biological mechanisms associated with these genetic phenomena impact future plans for genetic testing in healthcare (this also relates to Questions in #6)? The questions that could be asked:
What is the best strategy for broadly introducing existing genetic tests of current value into healthcare? What does the scientific evidence suggest about the feasibility and scope of genetic testing as the primary basis for assessing disease risk? What biological mechanisms are responsible for the uncertainty of genetic test results?

2. Whatever is proposed concerning payers, it seems that focus groups, interviews, and surveys within the industry would register the obstacles and misgivings that exist about personalized medicine. Based on some anecdotal conversations, there may be less concern about the intrinsic value of genetic tests and preventive interventions, and more concern that the relatively high rate of client turnover (once every 3-5 years) tends to discourage an investment in preventive testing/intervention. In other words, it is important to be aware of what, if any, scientific concerns exist among payers, and also, what other concerns they have. So, the question that needs to be asked: What obstacles and barriers do payers identify for covering preventive interventions and testing?

3. With regard to educational models, it seems evident that patients and providers will first need to perceive the personal value of genetic and family history information before they learn about it and act on it. This is a cornerstone of adult learning theory. For patients, it will also be necessary to reassure them that the information can lead to action that improves their health and avoids potentially serious health consequences (Health Belief Model). For physicians, it will be necessary to convey first how they can use genetic and family history information in the clinic to recognize when a patient is at an elevated disease risk (this relates to Question set #4). They will also need specific recommendations, access to genetic counselors and specialists, and operational support to manage patients placed in above average disease risk categories. Physicians and other health professionals must also be comfortable about ELS issues and their responsibilities in the followup monitoring process.

One of the serious problems for the growth of genomic medicine is that its adoption will require that both provider and patient must be active participants at the outset. Again, focus groups and surveys administered by a variety of groups, to a variety of potential stakeholders, including health organization officials, hospital administrators, health care professionals, communities, physicians, and educators concerning current expectations and outlooks, would greatly facilitate the process of developing effective educational strategies. It seems plausible that there is no single educational model that works widely, but that education must be tailored at the local level. The questions that emerge are: What educational content and strategy is necessary to initiate a conversion to the practice of personalized medicine in primary care? What local factors need to be considered when implementing this strategy?

(126) Tuesday, January 27, 2009 11:36 AM


Great questions and good idea to have a plan. My observations:

In question #2 - What information will be needed by payers to convince them.... Suggest we think less in terms of "convincing." Instead, "what social and economic factors will influence payer and consumer choices regarding payment for genetically-based testing strategies?"

Question #3 - "how will patients and clinicians respond?" - this whole set of questions is infused with language such as "swaying" and "influencing" both patients and clinicians. I suggest that the focus shift to reflect the fact that both groups have the capacity to make appropriate decisions when provided with sufficient information. Language here would be more appropriatly "client centered" and reflect the desire to empower these groups to make choices, rather than convincing them.

Question #4 - same issues as #3. The tone is that clinicians and patients need to be manipulated to make responsible choices.

Finally, suggest that there the first question be "What ethical assumptions will guide future initiatives?"

An agreement regarding the guiding ethical assumptions will help to frame the discussion. For example, the suggestion that patients should be required to have genetic counseling in order to receive the results of genetic testing information is based on assumptions about the rights of patients.

Thank you for considering my comments.

(128) Wednesday, January 28, 2009 11:09 AM


In addition to the insightful questions already suggested, a few other issues should be queried in the discussing of applying genomics to clinical problems, including the following:

What is the role of epigenetic alterations, including DNA methylations and histone modification changes that may be reversed by pharmaceutical agents targeted to the specific pathways involved?

What central laboratory resources and strategies are needed for the downstream research to identify the functional biological mechanisms conferring susceptibility to disease and outcome?

For common variants that confer low risk, would it be useful to report aggregate results?

How can testing costs be kept low on genetic tests that are patented?

How do we deal with the fact that the majority of patients see only a primary care physician (who has no formal training in genetics)?

How do we go about training PCPs on how to convey risks and report genetic results?

(132) Thursday, January 29, 2009 9:04 AM


Our healthcare system is complex, with many types of providers, sources of health information, and ways to prevent/improve health. We'd suggest that the questions put forward in these white papers be phrased and considered in the context of this collective healthcare system.

We suggest examining the current and future health care delivery systems when considering how genomics will impact healthcare. One question not proposed is, 'What are future healthcare delivery models?' Thinking about the healthcare model for genetic services delivery will inform the type of questions to ask in these white papers. Will the system for utilizing genetic testing center around the clinician only? Are there circumstance when the healthcare provider should refer to a genetic specialist? If so, how will they know to refer? See the Genetic Services Policy Project for analyses of the current delivery system and thoughts concerning impacts to the future system.
current: http://depts.washington.edu/genpol/docs/Ch2.pdf
future: http://depts.washington.edu/genpol/docs/Ch7.pdf

We suggest thinking more broadly about genomic discoveries than the clinic. For example, in the introduction we suggest changing, ýApplying genomic discoveries to clinical problemsýý to, ýApplying genomic discoveries to health problemsý.ý Genomic discoveries will not be confined to the clinic, as DTC sales may not even involve a clinic encounter.

Thinking about the broader healthcare system may also inform what questions to ask or how to address them. For example, health promotion is a key component to public health. Involving this science or utilizing the approaches and lessons learned in this field will assist in efforts to educate patients. Additionally, NHGRI is well suited to examine the scientific questions about the evidence of effectiveness. Health economics/marketing professionals and representatives from the private sector might be more appropriate groups to address questions about reimbursement decisions.

Overall: We recognize the challenge in framing the conversation and partitioning issues into unique topics in order to discuss them. However, there is much overlap between the topics addressed in the 3 white papers. Questions asked in the context of one topic area also need to be asked in the context of the other paper topics. For example:

Increasingly, genomic data will be somewhat informative in the context of environmental factors and multiple variants that have incomplete penetrance, not just one or two gene mutations. Disease risks estimates and associated interventions will not always be clear cut, especially if newly identified variants that havenýt been well validated are incorporated into these risk estimates. This will have consequences for how we think about therapeutics (and associated side effects), patient/physician education, and shared decision making. All papers propose questions about education, but only one addresses the complexity of shared decision making and only one addresses the complexity of therapeutic choice. Somehow the concepts need to be integrated. Likewise, we suggest that questions about collaboration be included in all papers. For example in the first paper under question 1 ýHow can we efficiently identify genetic risk factors and non-genetic factors in tandem?ý

Overall: The language used implies an underlying assumption that all testing and genetic information is good and that providers, payers, and patients need to be convinced of this. It also implies a preset agenda rather than an honest inquiry that will be used to inform an NHGRI plan.

We suggest phrasing the questions to be inquisitive in a more neutral way. For example, under the second question delete ýconvince themý and change ýpersuasiveý to ýinformativeý or ýusefulý. Under the third question change ýmake patientsý and ýswayedý. Under the fourth question change ýconvincedý and ýencouragedý.

Overall: This process of open comment shows NHGRIýs commitment to community engagement. We suggest that NHGRI make the conversation accessible to more people by using plainer language where possible (e.g., where technical language is not required).

Overall: There are instances where we recognize the concept, but the language could be more concrete. For example, under question 2 itýs not clear what ýwithin reachý is and under question 4 ýthis informationý is used frequently, but itýs not always clear what is meant.

(139) Thursday, January 29, 2009 7:22 PM


Question 1: There are questions about the value/utility of identifying immutable diseases or variants with no disease interventions. What about the value/utility of identifying new variants with no known disease associations? Additionally, today, clinicians regularly diagnose findings of immutable disease (e.g., inoperable brain cancer) ý how is this different?

Question 1: The fourth paragraph assumes that family history is a part of standard clinical information. We question whether family history is utilized universally in current practice.

Question 1: What factors in addition to stigmatization should affect where this bar is placed? For example, cost and timeliness of testing and interpretation should likewise be considered.

Question 3: How does commercial marketing impact desire or uptake of services?

Question 3: More than just patient characteristics and desire should be taken into account to get an accurate understanding of how family history impacts behavior change and risk reduction. How does the ability of patients to reduce their risk contribute to the assessment of patient desire to identify and reduce risk? For example, what are the economic and social conditions under which people live and how do they impact patient ability to change risk when reduction of risk is desired. How can physical and social environments be taken into account? Also, the immediate and long-term impacts may differ. Both need to be examined to get an accurate understanding of how family history of disease affects behavior change and risk reduction.

Question 5: We suggest rephrasing the question so as to better incorporate the idea that the approach to ALL genetic testing should consider social stigmas, medical history, and personal context. Itýs not inherently clear which ýspecial populationsý would need ýspecial approaches,ý nor do we believe that it is necessary to identify and describe each at this time, but rather acknowledge the situation is likely to occur.

(140) Thursday, January 29, 2009 7:24 PM


Regarding question 2, "What information will be needed or required by payers to convince them to reimburse genetically-based testing strategies, so that financial barriers to their application can be minimized?":

Other questions to consider under this heading are: Will genetically-based testing strategies be made available through federally-funded health care programs (Medicare, Medicaid, the Indian Health Service, and the Veteran's Administration health care system)? Will the clinical criteria for using expensive genetic tests differ for federally-funded health care programs compared with private payers? How can equal access to new genetically-based tests be ensured for all patients regardless of type of health care coverage?

Regarding question 3, "How will patients and clinicians respond to information regarding individualized genetic risk and what strategies and resources will be most effective in educating them to maximize health benefits and minimize potentially negative aspects such as stigmatization and anxiety?":

The question included under this heading, "What types of educational models will work for the very different audiences of patients and clinicians?" is important. Other questions along the same lines include: How can educational tools be tailored so they are culturally appropriate for ethnically-diverse patient populations, including non-English speaking patients? What educational tools have already been developed that could serve as models?

Regarding question 5, "What special approaches to genetically-based diagnostic and prevention strategies may be needed in special populations (such as prostate cancer in African-Americans) or high-risk groups (such as workers with benzene exposure)?":

The first questions under this heading are important: "What impact might past misuses of genetic information and frank misinformation to stigmatize minority groups have on adoption or testing? How can potential skepticism among minority patients and clinicians about conclusions drawn from such data be addressed?" However, even before asking these questions, it may be useful to first ask other key questions about patient ethnicity and clinical use of genetic testing: What role will patient ethnicity play in cliniciansý decisions to conduct genetic tests? Will specific genetic tests be targeted at patients of particular ethnicities? How will patient ethnicity be determinedýwill clinicians make that judgment themselves or will patients be asked to describe their own ethnicities? Is there a risk of ethnic profiling in the application of new genetic tests in clinical settings? If so, how can this risk be mitigated?

Regarding question 6, "What information on prevalence, risk, modifiability, etc will be needed for clinicians and patients to understand and utilize genome sequence information, once it becomes widely available in diagnosis, prevention, and treatment?":

Other questions to consider are: How will patient "populations" be defined in order to determine prevalence, risk, etc.? What role will race and ethnicity play in defining patient "populations?" What are the potential benefits and risks of using race/ethnicity to define patient "populations" in determining the prevalence and risk of particular diseases? How might these benefits and risks differ for highly-penetrant diseases with Mendelian inheritance versus chronic diseases with multifactorial causes? How can the risk of stigmatization of particular racial/ethnic groups be minimized?

(147) Friday, January 30, 2009 1:14 PM


As a medical anthropologist who studies genetic risk, I'm especially excited to see that this white paper has behavioral components. However, I believe that this white paper lumps genetic diseases together too simply, since some diseases are late-onset, like Alzheimers, while others affect children, not the person who receives the genetic testing, like sickle cell disease. So, the decision to test and the impact of the results will differ depending on the disease. Thus, prevention efforts will have to be more tailored to the disease to be effective, rather than the same preventive procedure applied liberally as part of a larger framework of genomics.

(154) Friday, January 30, 2009 5:38 PM


This paper raises questions concerning the translation of SNP and similar association studies into clinical practise, especially with regard to reimbursement. I am certainly interested in the latter, but propose that many of these tests are already being offered without peer approval/review in the private sector, direct to consumer. DNA testing is available for nutrition advice, and many other lifestyle issues, often with little evidence to support the validity of this testing. Are you going to address this, and somehow alert the public to the possible misinformation they will receive from such unsupervised testing?? The issue is somewhat analogous to the FDA and non-regulated "supplements".

(162) Wednesday, February 11, 2009 7:33 PM


In section 2 and elsewhere, the phrasing of some of the questions implies a bias that genetic testing is beneficial and should be widely adopted, rather than an open-minded investigation of whether or when it might be beneficial, and at what cost (to society as well as individuals or payers). Instead of "What types of evidence are likely to be most persuasive to payers?", how about "What types of data influence payers decisions about whether to cover genetic testing?"

In section 5, it is not clear what the following refers to: "What impact might potential use of genetic information by law enforcement and similar agencies to target minorities have on testing, and how can this best be addressed in future clinical testing?"

A question you might add is whether individuals opting for genetic testing are adequately informed beforehand, given the risk of obtaining information that is uninterpretable at present, but potentially dismaying or stigmatizing when more information is available. Do individuals change their attitudes toward genetic testing after they digest the results, or over longer time periods?

(164) Thursday, February 12, 2009 12:47 PM


It is appropriate and important to see overlapping themes in this paper and the Education and Community Engagement paper, specifically questions regarding education and ultimate utility of the genomic products.

(165) Thursday, February 12, 2009 1:30 PM


Preamble: It seems all questions in the four white papers involve complex/common/multi-gene disorders (CCMD). No mention is made of single gene disorders as if their problems are solved. Thus it is assumed all will be found by relatively simple method and therapy, etc is suitable. Try eating the diet a teenager with PKU is insulted with, etc.

Whilst CCMD are fashionable/important and have vast sums spent on them, we seem to be getting little return. Little public institutional money appears to be spent on the 2,000-3,000 human inherited disease and the 17,000 to come.

Thus I believe the following three questions are important one of which is relevant to CCMD.

1. Is spending on single gene disorder disproportionately lower in relation to community costs than that on complex disease, in the diagnostic databasing, prevention and therapy areas.

This question is prompted by the fact that 71% of admissions to a major US hospital are genetically based (50% in Australia). Another study suggests 60% of all humans will be affected by mutations in a lifetime.

2. Are health professionals not able to obtain full information available worldwide in a timely manner on other patients worldwide to inform their advice to patients?

The fact that databases of mutations are needed is evidenced by the fact that they are used daily in diagnostic labs. Not only that, vast amounts of time are spent surfing the web outside databases to exclude the fact that a specific mutation has not been found. Missing a vital earlier paper could lead to legal action.

The fact that they are needed and their synthesis is not funded is evidenced by the fact that armies of individuals work voluntarily after hours to build these databases.

3. Are databases of mutations in genes causing single gene disorder, useful for those thousands of laboratories well funded to attempt to solve CCMD?

Single gene disorder gives very specific phenotypes and their analysis may give clues to candidate genes to test in CCMD.

4. Will databases of mutations in single gene disorder enhance the value of personalised sequencing soon to be available as many sequence changes in genes are of questionable importance but if they have been found to cause single gene disorder earlier this enhances the value of a sequence of an individual.

(169) Sunday, February 15, 2009 11:31 PM


One very useful resource would be a collection of DNA samples from 1,000 healthy adults of Northern European ancestry. These could be maintained at a site such as CIDR where they could be tested for putative pathogenic variants identified by investigators in patients with specific disorders. This resource would answer the question: is this a common variant? This central resource could be used over and over
and would be cost effective, relieving each investigator of the need to accumulate such a resource again. Of course, other ethnic groups should also be collected.

(171) Tuesday, February 17, 2009 6:24 PM


Regarding the last comment on the white paper, there already is a resource of CEPH/HapMap that are from individuals of Northern European descent and populations from all over the world. These are currently being utilized for genome-wide studies in relation to drug response. However, these are immortalized LCLs and may not truly reflect the tissue of interest. Thus, more repositories (from different types of samples), such as the HapMap will need to be available for researchers to explore disease/phenotype/genotype relationships for primary and follow-up studies.

(173) Wednesday, February 18, 2009 5:36 PM


An important theme that appears to be missing in this white paper is regarding the implication of genomics research in health outcomes among children, pregnant women, and individuals of reproductive age. This is particularly relevant in view of the increasing acknowledgment of the developmental origin of human adult disease and the earlier onset age of several important chronic disorders such as diabetes, obesity, and hypertension.

What should be done to effectively promote genomics, epigenomics, and pharmacogenomics research among children, pregnant women, and individuals of reproductive age in general?

Another theme that was discussed, but not explicitly, is efforts to improve modifiability of genomics.

One of the major powers of genomics research is that it will allow better targeting and tailoring of preventive efforts, by identifying gene-environment interaction/modifiability. In the past few years, despite considerable efforts, only moderate progress was achieved in this regard. Large scale observational studies, with well-quantified environmental and genetic factors, are helpful. Clinical trial/intervention studies may provide more conclusive findings. In order to improve the ýmodifiabilityý of genomics, should we promote and support clinical trials/intervention studies to incorporate initiatives of gene-environment interaction?

(174) Thursday, February 19, 2009 10:05 AM


Additional questions to consider:

Are there unique aspects of genetic technologies and their development that make the current reimbursement system unsustainable?

Does the type or severity of the condition impact an individualýs decision for testing? Can education efforts surrounding particular conditions serve as models for understanding this impact?

How can patients be involved in the crafting of guidelines for physicians in the appropriate method for delivering information to patients?

How will privacy and confidentiality be discussed in light of genome sequence information, and how will this dialogue engage all stakeholders?

(192) Friday, February 27, 2009 4:42 PM


Many very good points made in this summary. I do think that, based on experience with BRCA1/2, more recognition is needed of the psychological factors such as family history and family experience with cancer which govern perceived risk and adherence to recommendations for mutation carriers, the ultimate limiting step in the success of genetic testing. Education is not simply the provision of factual information, but, especially in emotion-laden areas such as genetic testing, understanding enough about the emotional factors which inhibit learning, family communication, and adoption of risk-reducing behaviors so that truly successful educational interventions can be developed.

We need also to look at multi-generational effects of genteic testing and personalized genomic medicine. Also, more work needed on cultural effects and especially impact of lower SES
on access, interest, uptake of testing, and utilization of results.

(196) Friday, February 27, 2009 9:29 PM


This is a nice set of questions, and seems particularly well thought out regarding the adoption and understanding of genetic test results by providers, payers, and patients.

One theme that could be better incorporated into this white paper under all three categories (prevention, diagnostics, and pharmacogenomics) is gene-environment interaction. When considering complex disease genetics in particular, it appears that the complexity underlying the mechanisms and processes must be accounted for. As such, gene-environment interactions and the biological, physiological, and behavioral mechanisms that underly gene variant-disease associations must be well understood before preventive strategies are prescribed or risk profiles are communicated. In addition, many factors influencing health are not static, that is, they affect risk and resilience in a variety of ways throughout the life course (e.g., in utero, during infancy and childhood, during pregnancy, or during old age). It is therefore anticipated that the actions of gene variants and gene-environment interactions may have critical or sensitive periods across the developmental trajectory.

The openness of your long-range planning process is to be lauded. This is a good model for other institutes/offices.

(199) Friday, February 27, 2009 10:26 PM


We strongly agree with the introduction to this White Paper, that ýDevelopment and clinical validation of genetic tests requires translation from basic science discoveries, assessment of what these advances add to currently available clinical resources and whether incorporation of any new test is justified by the strength of its benefits as compared to potential risks.ý In other words, genetic test results should direct management changes that, compared to standard care, result in net improvement in clinical outcomes that matter to the patient, the goal of any medical intervention. Evidence of such clinical utility may be collected directly in prospective clinical trials, and in some cases may be obtained less onerously by retrospective analysis of banked specimens from completed trials and/or creation of an indirect, linked chain of evidence. Tests without such evidence should not be accepted as routine practice.

We disagree, however, with the organization of the document into questions about what information will convince payers to reimburse genetic tests and about what information will convince providers to apply genetic testing strategies to their patients (as if providing such information is an unnecessary burden). First, the questions do not make it clear that not all available genetic tests are equally likely to result in useful medical information and improved clinical outcomes for the patient. Second, evidence of improved clinical outcomes as a result of genetic testing is as important to providers as it is to payersýboth groups have the goal of improving patient health efficiently and cost-effectivelyýand ultimately is of utmost importance to patients.

Thus, we suggest re-organizing this document such that the questions address the overall goals when considering adding genetic testing to standard clinical care, the evidence needed to show that those goals can be achieved for individual patient applications, and how tests with evidence of clinical utility can favorably impact payer policy, provider use, and patient acceptance. Additionally, questions related to strategies of translating new genomic developments into future genetic tests should be organized around the concepts of the evidence needed to initiate development (i.e. evidence of clinical validity) and the framework of evidence gathering and evaluation necessary to demonstrate improved patient outcomes.

We also find the use of the word ýriský to be unclear in many parts of the document. Risk can refer to risk of existing disease (diagnosis), risk of future disease (disease risk), risk of disease progression (prognosis), risk of adverse events from drug treatment (pharmacogenomics), etc. A general question is whether the risk predicted by the genetic test in question for any of those applications is sufficient to change medical management for at least some patients, such that net health outcome is improved. If not, then the question becomes, what is the value of knowledge? Does it somehow impact the health of the patient (e.g. quality of life) and can this be demonstrated? Or, is it useful personal information but not medical information? In the specific situation of using genetic tests to gain low-level risk information, e.g. on the likelihood of being diagnosed in the future with common multifactorial diseases, and to ask questions about effects of those obtained risks on patient and provider behavior, it might be helpful to describe that application more clearly.

With regard to single risk variants vs. the integration of several risk markers, an additional question is, ýHow do we know when enough variants and environmental factors have been identified to have confidence in the summary prediction?ý

Regarding strategies for providing genetic test information, we suggest questions that are better targeted torward an evaluative, evidence-based medicine format (e.g. ACCE, EGAPP). While GeneTests reviews are highly informative, they are focused on rare genetic conditions and do not employ a formal, evidence-based framework of evaluation.

(203) Monday, March 2, 2009 2:51 PM


The Association for Molecular Pathology (AMP) is an international medical professional association representing approximately 1,600 physicians, doctoral scientists, and medical technologists who perform laboratory testing based on knowledge derived from molecular biology, genetics, and genomics. Since the beginning of our organization we have dedicated ourselves to the development and implementation of molecular diagnostic testing, which includes genetic testing in all its definitions, in a manner consistent with the highest standards established by the Clinical Laboratory Improvement Act (CLIA), the College of American Pathologists (CAP), the American College of Medical Genetics (ACMG), and the United States Food and Drug Administration (FDA). Our members lead and work at the majority of clinical molecular diagnostic laboratories in the United States as well as in laboratories in many other countries. We are frequently involved in the development of novel molecular tests, and in the validation of laboratory developed or commercial assays.

Thank you for the opportunity to comment on the White Papers in Phase 1 of NHGRIýs Long-range Planning project. Our comments are as follows:

White paper 1:
Applying Genomics to Clinical Problems-Diagnostics, Preventive Medicine, Pharmacogenomics

Q1, additional questions:
- How do we integrate of the vast amount of information becoming available through genomic medicine, into the constraint practical environment of clinical practice?
- How do we sort actionable (vetted) information from early results and when should we consider findings at a genomic level ready for implementation, especially in the absence of an understanding of functional or epigenetic effects?

Q2, additional questions:
- It is important to ask how the cost of testing can be kept low, but also to determine how the cost of interpretation and education can be kept within reasonable limits (and what these are). The technology is expected to allow mass data generation, but QA and reliable interpretation of raw data is lagging behind.

Q3, additional questions:
These are excellent questions. What types of studies will be needed and what is the necessary scope to provide answers?
- Should we consider development of an educational track for a medical subspecialty under molecular pathology (ýapplied genomicsý, with many interdisciplinary components)?

Q4, additional questions:
- Where will clinicians find support for their questions about newly discovered information? Should this be centralized?

Q6, additional questions:
- How do we plan for clinician education and translation of identified risk factors into actionable information?

Q7, additional questions:
- How do clinicians ensure that information obtained through recreational genomics/DTC testing is analytically and clinically valid?

(207) Monday, March 2, 2009 2:56 PM




« Back to White Paper Page