They had me at the phrase “genetic superheroes.”
The phrase isn’t in the Nature Biotechnology study itself, which has a long, journal-ish title involving “analysis of 589,306 genomes” and “severe Mendelian childhood diseases.”
But the accompanying commentary by Daniel MacArthur of Massachusetts General Hospital and Cambridge’s Broad Institute is headlined “Superheroes of disease resistance.” It describes “genetic superheroes, people who remain healthy even though they carry genetic variants known to cause severe Mendelian diseases,” such as cystic fibrosis or sickle-cell anemia.
Calling Marvel and Universal Studios: There’s a new breed in town. The journal paper describes an epic search among more than a half-million genomes that turned up a tiny population of 13 people whose genes dictated that they should have been very sick or dead — but weren’t.
I turned to Dr. Isaac Kohane, chair of Harvard Medical School’s Department of Biomedical Informatics, for insight. His thoughts are transcribed below:
What does it mean that [Dr. Rong] Chen and colleagues studied a half-million individuals and found that 13 of them appeared to be quite resistant to a set of mutations that, in the textbook, are said to be absolutely disease-causing? Is this the new fountain of youth? The new disease barrier? Who are these superheroes? And what does it tell us about our medicine, as practiced today, and about our biology?
One optimistic perspective is that, just like the 80-year-old morbidly obese individual, or the 110-year-old happy smoker, we can learn by looking at the genomes of these individuals — or perhaps their lifestyle — what’s protected them from what is commonly thought of as a disease-causing insult, whether genetic or otherwise?
And indeed, there are some examples where we have found natural accidents: mutations in individuals that, for example, seem to protect them from cardiovascular disease and are now resulting in possibly useful cholesterol-lowering drugs.
From this perspective, finding these individuals — and more like them — is the tip of the spear of the new genomic medicine.
Nonetheless, I would argue that there is a less black-and-white perspective on these superheroes. And that is that they constitute a few points on a continuum of what we call the penetrance of genetic variants. By that I mean, the likelihood that a genetic variant, a mutation, will cause disease based on everything else that maintains your health. Let me get concrete:
It’s now been long known that mutations in the BRCA genes — BRCA 1 and 2 — are much more predictive of breast cancer in Ashkenazi-Jewish women with a family history of breast cancer than they are in non-Ashkenazi-Jewish women who have no family history.
So what is it about them that is different? Well, certainly there may be environmental differences, whether in diet, exercise and other exposures. And also, we know there are different genetic variants throughout your genome which counter-balance different assaults.
It’s actually a very old thought in medicine that goes back at least 200 years. It goes something like this: For critical life functions, we don’t have one system to maintain our physiological equilibrium, we have many failsafe, redundant systems so that if one system goes wrong, the other steps in.
It’s only when we’re very unlucky, in that we have a hit to multiple systems, that we start seeing these systems failing.
The evolutionary argument is that if we were so vulnerable to single hits, we’d be less likely to make it to reproductive age, whereas having multiple, redundant systems does in fact increase the chances you’ll have lots of healthy children.
This would suggest there are far more than just a few handfuls of individuals who are resilient for these mutations, and that it’s a matter of more complete data on even these half million that might illustrate that point.
Let me give you as an example: a study that happened almost 15 years ago in California. There’s a gene called HFE in which mutations have been found to cause a disease called hemochromatosis, where iron is deposited throughout your organs. This is a disease which can affect your heart and cause heart failure, cause sterility and premature death.
And if you go to a clinic — a hereditary hemochromatosis clinic — 80 percent of the individuals will have mutations in that gene. And what this study — which was done by investigators from the Scripps Institute on more than 40,000 consecutive patients at the Kaiser Permanente Clinics — showed was that indeed, there were 151 individuals who had two copies of the mutations that cause hemochromatosis in the textbook and in the hereditary hemochromatosis clinic.
And yet of those 151 individuals, only one had anything approaching the textbook clinical symptoms of hemochromatosis.
Why was this? Simply that unlike the poor individuals with the disease, these 151 individuals in the general population were resilient to the disease, because they had a different ancestry, and a different environment, and that’s just for one genetic disease. Increasingly, we are seeing that these textbook mutations, which putatively cause disease, don’t.
Which brings me to the immediate consequences of this study: It’s now common, almost trite, for a health reporter to say that of course, genetics and the environment play an important role in determining our health. But I don’t think you can overstate the effects of medical school education over the last 40 years, where we’ve been given brilliant examples of hard-won facts about diseases that are caused by genetic mutations. It’s hard to shake that pedagogical burden off, for a variety of good reasons that now are beginning to have real consequences.
That is, as it becomes apparent that different populations, different ethnicities, will have a greater or lesser penetrance or likelihood of disease given a mutation, some of these textbook cases — in fact, many — are likely to no longer pertain. And that will result — until we fully change our educational system and the decision support that goes into clinical decision-making — in a number of unwarranted costly and painful medical procedures.
What am I talking about? When it’s well-grounded in scientific fact, even something as draconian as bilateral mastectomy and oophorectomy — removal of the ovaries — may in fact be justified. But what if the certainty in some individuals is not that high? What if, for a whole variety of putative birth-defect mutations, cancer-causing mutations, neurodegenerative-disease-causing mutations, our knowledge base is fundamentally flawed because we have not had the large population studies with deeply clinically characterized studies of these individuals?
Well, we’ve already seen what that does. We saw it in the case of prostate-specific antigen, where putatively, having high levels of this protein was always diagnostic of a cancer that would kill you soon: prostate cancer.
Of course we’ve learned otherwise, and the diagnostic and therapeutic criteria have changed as a result, but only after tens of thousands of prostates were removed needlessly.
So the question arises: How many of these clinically irrelevant findings are there? Something that I called, 10 years ago, the ‘incidentalome’ — the biggest “ome” of them all, bigger than the genome, bigger than the proteome — the “ome” of all incidental findings.
Until we do the big pop studies and we understand clinically what’s going on with these individuals, I think we’re at increased risk now, by virtue of the ability to cheaply measure the genome, to actually subject all of us to some of the vicissitudes of the genomic information that we obtain through sequencing.
So what’s the antidote? Daniel MacArthur, in his accompanying editorial to the article by Chen and colleagues, I think points it out: It is doing larger genomic studies. But not just genomic studies — but rather genomic studies that are paired to rich clinical histories of each individual with their environmental exposures.
How is that going to happen? Well, a very good first step would be to support and accelerate the two initiatives that came out of the Obama administration — namely, the Blue Button Initiative, which would enable every electronic health record to discharge for you your entire clinical history, so that you could donate it for research or use it for clinical care.
And, as part of the Precision Medicine Initiative, Syncing for Science, the protocol which similarly allows you to extract data from multiple electronic health records to allow you to contribute your clinical data to match your genomic data.
This would allow researchers to really understand to what degree genomic variants actually do contribute to disease, and to what extent other factors, such as your environmental exposures or your lifestyle, may contribute.
And in that respect, this study, by pointing out the opportunity through this large population approach, may have made its largest contribution.