In general, this paper is sobering. I would have naïvely thought that monozygotic twins would be more similar in the diseases that afflict them.
Perhaps this may be seen as positive, in that our destiny is not in our genes. Rather, many other things that we can influence, such as our lifestyle and getting medical checkups, may be much more important for reducing our risk of prematurely being affected by diseases.
Even for Alzheimer's disease, one of the cases where genome sequencing has the potential to perform best, according to this study, only two twin pairs out of 10 in the study are "concordant," i.e., in only two out of 10 cases do both monozygous twins have the disease, whereas in the eight cases, only one twin is affected.
Food for thought is also that research approaches other than genomics, in particular, into physiology and epidemiology, may provide more economical routes to a better understanding into how we prevent many diseases.
The study by Vogelstein and colleagues constitutes a milestone for the field of human genetics, and has critical ramifications both medically and socially. By challenging the long-standing dogma that genetic testing would automatically provide absolute information on the future pathologies of an individual, this study resets the field and gives much more weight to the critical contribution that epigenetic modifications have in interpreting the history, physical status, and lifestyle of the patient.
This study is quite valuable in that it systematically attempts to ascertain the value of genetic predictions. It is expected that negative predictions are not useful. It was interesting that the authors indicated that they could predict a positive outcome for one of 24 major diseases on average. Such information would be useful to a person at risk for that disease.
However, I expect the real figure will be even higher than that, based on the clinical interpretation of genomes that have been analyzed thus far (the Quake genome, Ashley et al., 2010); the West family, Dewey et al., 2011) and now my genome just published in Cell (Chen et al., 2012). These all show increased risk for several important diseases from the genome sequence. It is true that in many cases a person will only die (or become severely affected) from one of them; when that happens, information from the other diseases is often lost or masked. Also, as noted by the authors, some of the diseases are likely to be related (i.e., not independent) and, thus, the genetic linkage is higher. I think a genome sequence will provide useful information that can alert people to many possible conditions, all of which can be followed.
References:
Ashley EA, Butte AJ, Wheeler MT, Chen R, Klein TE, Dewey FE, Dudley JT, Ormond KE, Pavlovic A, Morgan AA, Pushkarev D, Neff NF, Hudgins L, Gong L, Hodges LM, Berlin DS, Thorn CF, Sangkuhl K, Hebert JM, Woon M, Sagreiya H, Whaley R, Knowles JW, Chou MF, Thakuria JV, Rosenbaum AM, Zaranek AW, Church GM, Greely HT, Quake SR, Altman RB.
Clinical assessment incorporating a personal genome.
Lancet. 2010 May 1;375(9725):1525-35.
PubMed.
Dewey FE, Chen R, Cordero SP, Ormond KE, Caleshu C, Karczewski KJ, Whirl-Carrillo M, Wheeler MT, Dudley JT, Byrnes JK, Cornejo OE, Knowles JW, Woon M, Sangkuhl K, Gong L, Thorn CF, Hebert JM, Capriotti E, David SP, Pavlovic A, West A, Thakuria JV, Ball MP, Zaranek AW, Rehm HL, Church GM, West JS, Bustamante CD, Snyder M, Altman RB, Klein TE, Butte AJ, Ashley EA.
Phased whole-genome genetic risk in a family quartet using a major allele reference sequence.
PLoS Genet. 2011 Sep;7(9):e1002280.
PubMed.
Chen R, Mias GI, Li-Pook-Than J, Jiang L, Lam HY, Miriami E, Karczewski KJ, Hariharan M, Dewey FE, Cheng Y, Clark MJ, Im H, Habegger L, Balasubramanian S, O'Huallachain M, Dudley JT, Hillenmeyer S, Haraksingh R, Sharon D, Euskirchen G, Lacroute P, Bettinger K, Boyle AP, Kasowski M, Grubert F, Seki S, Garcia M, Whirl-Carrillo M, Gallardo M, Blasco MA, Greenberg PL, Snyder P, Klein TE, Altman RB, Butte AJ, Ashley EA, Gerstein M, Nadeau KC, Tang H, Snyder M.
Personal omics profiling reveals dynamic molecular and medical phenotypes.
Cell. 2012 Mar 16;148(6):1293-307.
PubMed.
Below, I'd like to offer some constructive criticism on the paper itself, along with thoughts on the engaging public discussion that it has prompted.
1. The work is conventional, and the findings not surprising.
I agree with Svante Paabo and and Paolo Sassone-Corsi that papers like this are important antidotes to any rash expectation that, for the average fairly healthy adult, precise and accurate genetic risk prediction will be easy. But I doubt that serious geneticists hold such expectations. As a meta-analysis of past twin studies, this paper reheats the old and simple observation that so-called "identical" twins don’t always get the same diseases.
Following in the long twin study tradition, the authors tackle the age-old question of how genetically heritable each disease is overall. And, beyond that, how does genetic risk vary among people? That is, for a given disease, do genomes contribute a portion of risk that varies widely, but smoothly, from person to person—or a portion of risk that varies more sharply from person to person—or some mix in between? If the paper says much that is new, it's in the last—in trying to sort particular diseases by estimating how well, on average, one may be able to forecast them from the genome alone.
This raises a key caveat: Many of the diseases the paper looks at are cancers, which are well known to be less genetically heritable than some other diseases. There in particular, the paper recasts longstanding knowledge as if it were a new grain of salt for the coming era of genomically personalized healthcare.
2. The work fits into a prevailing integrative view of genomically personalized healthcare.
Looking to the prospect of genomically personalized healthcare, serious thinkers have always understood that genomes will complement, not replace, conventional cornerstones of clinical care. Face-to-face doctor visits, family history, lab tests, and so forth will remain essential, but none on its own will tell us everything we need to know about disease risk. Lab tests often happen too late, and family history is limited in utility for exactly the reasons this paper highlights.
By rough analogy, note that weather forecasters use satellite photos every day for remarkably detailed insight into what is happening in the atmosphere; nonetheless, they would not try to predict today's high temperature in Leipzig or Palo Alto solely from such photos. Rather, they are using such modern, comprehensive views of the data alongside older ground-based instruments to predict the weather (and they still get it wrong sometimes).
Likewise, whole-genome sequencing will play a crucial role in everyday healthcare in future, but always together with other sources of medical insight. At Knome, we know that it’s crucial not to hype genomes as silver bullets of healthcare. To do so would do the public a disservice and, in raising false expectations of surefire returns, risk a backlash from funders of healthcare and research.
Yet it is important to note that discoveries from individual genomes have already helped many families. This has been true ever since the first genetic disease variant, for sickle cell anemia, was discovered in the late 1950s. As we continue to survey more people’s whole genomes while gathering data on their diseases and other traits, the average person's genome will indeed tell us more and more about what makes her unique, and about the distinctive health risks she faces.
3. The paper makes some optimistic and testable predictions.
It is worth stressing, as the article and Michael Snyder have, that the paper is cautiously upbeat, claiming to help us understand which diseases might be particularly amenable to genetic risk prediction. The basic observation on monozygotic twins should temper any undue expectation that WGS offers slam-dunk insight into the typical adult's long-term risk for all common diseases. At the same time, the authors do posit that whole-genome interpretation may nonetheless offer nearly all of us some significant hint of distinctive risk for at least one major disease.
Admirably, the authors make more specific predictions: that genome interpretation may eventually help clarify our risk for some autoimmune (such as type 1 diabetes and thyroid disease) and brain diseases (such as Alzheimer’s) particularly well. This includes the often overlooked question of identifying diseases for which someone may be at unusually low risk. In time, we'll see how these predictions fare, and whether readily genetically forecastable diseases tend to have particular physiological profiles.
Moreover, the paper acknowledges that whole-genome sequencing can help us spot strong risk for rare, serious diseases that may lurk in our genomes. This may be especially useful as we plan families with our spouses, so we can know what rare but potentially shared risk variants might be catastrophic if inherited together by one of our kids.
4. Whole-genome interpretation presumes it’s not all in your genes.
Responding to Paolo's "absolute information" comment, I think any notion that this paper debunks genetic determinism is a straw man. We have long known that genetic risk is not immutably deterministic. In fact, the endeavor of genomically personalized medicine is founded on that very point. That is, in trying to understand genetic risk, we specifically hope to learn to mitigate it, by controlling the environment of our habits: what we eat, what drugs and other treatments we take, and how we otherwise live our lives. Our genomes may be able to tell us a fair bit on those fronts.
5. Whole-genome sequencing is already helping greatly in cancers.
Some of the first clear examples of how useful genome sequencing can be are in familial cancers, such as BRCA-associated breast and ovarian cancers. Tumor sequencing, which the paper doesn’t address, is revolutionizing how cancer is treated, by finding key changes to the genomes of particular cells in the body that let them grow out of control. Such sequencing is reshaping how oncologists think of cancers from a simplistic tissue-specific view to one that highlights recurrent variants shared by tumors in different people and different tissues, which may nonetheless represent druggable targets.
6. Like all twin studies, this one is dogged by some minor concerns.
As the authors acknowledge, their work presumes that European monozygotic twins reasonably represent everyone, that is, that their assertions
will generalize to other ethnic groups;
aren’t confounded by ascertainment bias: "Doc, my twin has disease X. Do I?" This might tend to overestimate heritability if one twin's anchoring diagnosis means the other twin is more often or better diagnosed, or underestimate it if one twin, on seeing the other get sick, takes more preventive measures than (s)he otherwise would;
aren’t distorted by monozygotic twins’ distinctive health profiles. Factors include withstanding resource-constrained development in the womb, potentially lower-than-average birth weight, unusual profiles of maternal age or genotype, any subtle quirks of early embryonic cell division, the effects of lifelong social support from having a very similar sibling, etc.
The last few concerns are unlikely damning. After all, as noted, the basic findings of the paper boil down to the important and indisputable observation that monozygotic twins don't get the same diseases. But, like all the foregoing, they are worth keeping in mind as research into the causes of disease—genetic and otherwise—continues on all our behalf.
Of course DNA is not immutable destiny. The results of this study should not have surprised anyone, considering the complexities in the regulation of gene expression. The authors have done a great service by conducting a comprehensive study that emphasizes this.
Indeed, the influence of the environment on the epigenetic regulation of the expression of a disease phenotype has already been shown by us in a study of identical twins discordant for Alzheimer's disease (Mastroeni et al., 2009). Nevertheless, there do exist genetic sequences that are highly predictive of disease phenotype, and it is important to distinguish situations in which genes are destiny and those in which they are not.
References:
Mastroeni D, McKee A, Grover A, Rogers J, Coleman PD.
Epigenetic differences in cortical neurons from a pair of monozygotic twins discordant for Alzheimer's disease.
PLoS One. 2009;4(8):e6617.
PubMed.
The analysis by an esteemed group of genome and cancer scientists at
Johns Hopkins takes a novel approach to an important issue: Now that we
are on the verge of being able to sequence inexpensively any person's
entire DNA sequence, what will the information mean? Certainly, people
will learn about variations in their genome that might cause a rare
disease in them, or predispose an offspring to a rare disease. Genome
sequencing will also provide knowledge of which alleles a person has at
the apolipoprotein E (ApoE) gene, a common variant that, when present in one
copy, increases the risk of Alzheimer's disease around fourfold, and when
present in two copies, increases the risk several times more, depending
on ethnicity.
But the Johns Hopkins study asks a deeper question—about the
ability to predict risk for common diseases such as various cancers,
diabetes, and heart disease. Based on data that have been around for
decades on the likelihood that two identical (monozygotic) twins will
develop the same common disease, this new study calculates that simply
knowing the sequence will add little to what can be learned from the
family history alone.
This result certainly gives pause to those who
have predicted that whole-genome sequencing will revolutionize medical
care, and be the foundation of "personalized medicine." Hopefully,
people already have their healthcare "personalized" by their
physicians, and DNA analysis, selectively applied, can certainly have an
important role. But for the present, having your entire 3.2 billion-nucleotide DNA sequence on a flash drive is not something you need to have in your medical record.
Comments
Max-Planck Institute for Evolutionary Anthropology
In general, this paper is sobering. I would have naïvely thought that monozygotic twins would be more similar in the diseases that afflict them.
Perhaps this may be seen as positive, in that our destiny is not in our genes. Rather, many other things that we can influence, such as our lifestyle and getting medical checkups, may be much more important for reducing our risk of prematurely being affected by diseases.
Even for Alzheimer's disease, one of the cases where genome sequencing has the potential to perform best, according to this study, only two twin pairs out of 10 in the study are "concordant," i.e., in only two out of 10 cases do both monozygous twins have the disease, whereas in the eight cases, only one twin is affected.
Food for thought is also that research approaches other than genomics, in particular, into physiology and epidemiology, may provide more economical routes to a better understanding into how we prevent many diseases.
View all comments by Svante PaaboUniversity of California, Irvine
The study by Vogelstein and colleagues constitutes a milestone for the field of human genetics, and has critical ramifications both medically and socially. By challenging the long-standing dogma that genetic testing would automatically provide absolute information on the future pathologies of an individual, this study resets the field and gives much more weight to the critical contribution that epigenetic modifications have in interpreting the history, physical status, and lifestyle of the patient.
View all comments by Paolo Sassone-CorsiStanford University
This study is quite valuable in that it systematically attempts to ascertain the value of genetic predictions. It is expected that negative predictions are not useful. It was interesting that the authors indicated that they could predict a positive outcome for one of 24 major diseases on average. Such information would be useful to a person at risk for that disease.
However, I expect the real figure will be even higher than that, based on the clinical interpretation of genomes that have been analyzed thus far (the Quake genome, Ashley et al., 2010); the West family, Dewey et al., 2011) and now my genome just published in Cell (Chen et al., 2012). These all show increased risk for several important diseases from the genome sequence. It is true that in many cases a person will only die (or become severely affected) from one of them; when that happens, information from the other diseases is often lost or masked. Also, as noted by the authors, some of the diseases are likely to be related (i.e., not independent) and, thus, the genetic linkage is higher. I think a genome sequence will provide useful information that can alert people to many possible conditions, all of which can be followed.
References:
Ashley EA, Butte AJ, Wheeler MT, Chen R, Klein TE, Dewey FE, Dudley JT, Ormond KE, Pavlovic A, Morgan AA, Pushkarev D, Neff NF, Hudgins L, Gong L, Hodges LM, Berlin DS, Thorn CF, Sangkuhl K, Hebert JM, Woon M, Sagreiya H, Whaley R, Knowles JW, Chou MF, Thakuria JV, Rosenbaum AM, Zaranek AW, Church GM, Greely HT, Quake SR, Altman RB. Clinical assessment incorporating a personal genome. Lancet. 2010 May 1;375(9725):1525-35. PubMed.
Dewey FE, Chen R, Cordero SP, Ormond KE, Caleshu C, Karczewski KJ, Whirl-Carrillo M, Wheeler MT, Dudley JT, Byrnes JK, Cornejo OE, Knowles JW, Woon M, Sangkuhl K, Gong L, Thorn CF, Hebert JM, Capriotti E, David SP, Pavlovic A, West A, Thakuria JV, Ball MP, Zaranek AW, Rehm HL, Church GM, West JS, Bustamante CD, Snyder M, Altman RB, Klein TE, Butte AJ, Ashley EA. Phased whole-genome genetic risk in a family quartet using a major allele reference sequence. PLoS Genet. 2011 Sep;7(9):e1002280. PubMed.
Chen R, Mias GI, Li-Pook-Than J, Jiang L, Lam HY, Miriami E, Karczewski KJ, Hariharan M, Dewey FE, Cheng Y, Clark MJ, Im H, Habegger L, Balasubramanian S, O'Huallachain M, Dudley JT, Hillenmeyer S, Haraksingh R, Sharon D, Euskirchen G, Lacroute P, Bettinger K, Boyle AP, Kasowski M, Grubert F, Seki S, Garcia M, Whirl-Carrillo M, Gallardo M, Blasco MA, Greenberg PL, Snyder P, Klein TE, Altman RB, Butte AJ, Ashley EA, Gerstein M, Nadeau KC, Tang H, Snyder M. Personal omics profiling reveals dynamic molecular and medical phenotypes. Cell. 2012 Mar 16;148(6):1293-307. PubMed.
View all comments by Michael SnyderKnome, Inc.
Below, I'd like to offer some constructive criticism on the paper itself, along with thoughts on the engaging public discussion that it has prompted.
1. The work is conventional, and the findings not surprising.
I agree with Svante Paabo and and Paolo Sassone-Corsi that papers like this are important antidotes to any rash expectation that, for the average fairly healthy adult, precise and accurate genetic risk prediction will be easy. But I doubt that serious geneticists hold such expectations. As a meta-analysis of past twin studies, this paper reheats the old and simple observation that so-called "identical" twins don’t always get the same diseases.
Following in the long twin study tradition, the authors tackle the age-old question of how genetically heritable each disease is overall. And, beyond that, how does genetic risk vary among people? That is, for a given disease, do genomes contribute a portion of risk that varies widely, but smoothly, from person to person—or a portion of risk that varies more sharply from person to person—or some mix in between? If the paper says much that is new, it's in the last—in trying to sort particular diseases by estimating how well, on average, one may be able to forecast them from the genome alone.
This raises a key caveat: Many of the diseases the paper looks at are cancers, which are well known to be less genetically heritable than some other diseases. There in particular, the paper recasts longstanding knowledge as if it were a new grain of salt for the coming era of genomically personalized healthcare.
2. The work fits into a prevailing integrative view of genomically personalized healthcare.
Looking to the prospect of genomically personalized healthcare, serious thinkers have always understood that genomes will complement, not replace, conventional cornerstones of clinical care. Face-to-face doctor visits, family history, lab tests, and so forth will remain essential, but none on its own will tell us everything we need to know about disease risk. Lab tests often happen too late, and family history is limited in utility for exactly the reasons this paper highlights.
By rough analogy, note that weather forecasters use satellite photos every day for remarkably detailed insight into what is happening in the atmosphere; nonetheless, they would not try to predict today's high temperature in Leipzig or Palo Alto solely from such photos. Rather, they are using such modern, comprehensive views of the data alongside older ground-based instruments to predict the weather (and they still get it wrong sometimes).
Likewise, whole-genome sequencing will play a crucial role in everyday healthcare in future, but always together with other sources of medical insight. At Knome, we know that it’s crucial not to hype genomes as silver bullets of healthcare. To do so would do the public a disservice and, in raising false expectations of surefire returns, risk a backlash from funders of healthcare and research.
Yet it is important to note that discoveries from individual genomes have already helped many families. This has been true ever since the first genetic disease variant, for sickle cell anemia, was discovered in the late 1950s. As we continue to survey more people’s whole genomes while gathering data on their diseases and other traits, the average person's genome will indeed tell us more and more about what makes her unique, and about the distinctive health risks she faces.
3. The paper makes some optimistic and testable predictions.
It is worth stressing, as the article and Michael Snyder have, that the paper is cautiously upbeat, claiming to help us understand which diseases might be particularly amenable to genetic risk prediction. The basic observation on monozygotic twins should temper any undue expectation that WGS offers slam-dunk insight into the typical adult's long-term risk for all common diseases. At the same time, the authors do posit that whole-genome interpretation may nonetheless offer nearly all of us some significant hint of distinctive risk for at least one major disease.
Admirably, the authors make more specific predictions: that genome interpretation may eventually help clarify our risk for some autoimmune (such as type 1 diabetes and thyroid disease) and brain diseases (such as Alzheimer’s) particularly well. This includes the often overlooked question of identifying diseases for which someone may be at unusually low risk. In time, we'll see how these predictions fare, and whether readily genetically forecastable diseases tend to have particular physiological profiles.
Moreover, the paper acknowledges that whole-genome sequencing can help us spot strong risk for rare, serious diseases that may lurk in our genomes. This may be especially useful as we plan families with our spouses, so we can know what rare but potentially shared risk variants might be catastrophic if inherited together by one of our kids.
4. Whole-genome interpretation presumes it’s not all in your genes.
Responding to Paolo's "absolute information" comment, I think any notion that this paper debunks genetic determinism is a straw man. We have long known that genetic risk is not immutably deterministic. In fact, the endeavor of genomically personalized medicine is founded on that very point. That is, in trying to understand genetic risk, we specifically hope to learn to mitigate it, by controlling the environment of our habits: what we eat, what drugs and other treatments we take, and how we otherwise live our lives. Our genomes may be able to tell us a fair bit on those fronts.
5. Whole-genome sequencing is already helping greatly in cancers.
Some of the first clear examples of how useful genome sequencing can be are in familial cancers, such as BRCA-associated breast and ovarian cancers. Tumor sequencing, which the paper doesn’t address, is revolutionizing how cancer is treated, by finding key changes to the genomes of particular cells in the body that let them grow out of control. Such sequencing is reshaping how oncologists think of cancers from a simplistic tissue-specific view to one that highlights recurrent variants shared by tumors in different people and different tissues, which may nonetheless represent druggable targets.
6. Like all twin studies, this one is dogged by some minor concerns.
As the authors acknowledge, their work presumes that European monozygotic twins reasonably represent everyone, that is, that their assertions
The last few concerns are unlikely damning. After all, as noted, the basic findings of the paper boil down to the important and indisputable observation that monozygotic twins don't get the same diseases. But, like all the foregoing, they are worth keeping in mind as research into the causes of disease—genetic and otherwise—continues on all our behalf.
View all comments by Nathan PearsonBanner Research Institute
Of course DNA is not immutable destiny. The results of this study should not have surprised anyone, considering the complexities in the regulation of gene expression. The authors have done a great service by conducting a comprehensive study that emphasizes this.
Indeed, the influence of the environment on the epigenetic regulation of the expression of a disease phenotype has already been shown by us in a study of identical twins discordant for Alzheimer's disease (Mastroeni et al., 2009). Nevertheless, there do exist genetic sequences that are highly predictive of disease phenotype, and it is important to distinguish situations in which genes are destiny and those in which they are not.
References:
Mastroeni D, McKee A, Grover A, Rogers J, Coleman PD. Epigenetic differences in cortical neurons from a pair of monozygotic twins discordant for Alzheimer's disease. PLoS One. 2009;4(8):e6617. PubMed.
View all comments by Paul ColemanRaymond and Ruth Perelman School of Medicine at the University of Pennsylvania
The analysis by an esteemed group of genome and cancer scientists at
Johns Hopkins takes a novel approach to an important issue: Now that we
are on the verge of being able to sequence inexpensively any person's
entire DNA sequence, what will the information mean? Certainly, people
will learn about variations in their genome that might cause a rare
disease in them, or predispose an offspring to a rare disease. Genome
sequencing will also provide knowledge of which alleles a person has at
the apolipoprotein E (ApoE) gene, a common variant that, when present in one
copy, increases the risk of Alzheimer's disease around fourfold, and when
present in two copies, increases the risk several times more, depending
on ethnicity.
But the Johns Hopkins study asks a deeper question—about the
ability to predict risk for common diseases such as various cancers,
diabetes, and heart disease. Based on data that have been around for
decades on the likelihood that two identical (monozygotic) twins will
develop the same common disease, this new study calculates that simply
knowing the sequence will add little to what can be learned from the
family history alone.
This result certainly gives pause to those who
View all comments by Reed E. Pyeritzhave predicted that whole-genome sequencing will revolutionize medical
care, and be the foundation of "personalized medicine." Hopefully,
people already have their healthcare "personalized" by their
physicians, and DNA analysis, selectively applied, can certainly have an
important role. But for the present, having your entire 3.2 billion-nucleotide DNA sequence on a flash drive is not something you need to have in your medical record.
Make a Comment
To make a comment you must login or register.