About the SA Blog Network



Opinion, arguments & analyses from the editors of Scientific American
Observations HomeAboutContact

Why Total Reporting of Genetic Results Is a Bad Idea

The views expressed are those of the author and are not necessarily those of Scientific American.

Email   PrintPrint

A model of the DNA double helix

Laboratories that sequence an individual’s entire genome should limit the results they report to clinicians and their patients based on certain usability criteria, according to the first set of guidelines on the subject from the American College of Medical Genetics (ACMG). Geneticists simply don’t know what to make of most of the information that they find and it’s much too easy to misinterpret many of the results.

The ACMG’s advice is not as contradictory as it sounds—and could probably apply to a lot of other medical situations. Whenever a doctor asks me or a loved one to undergo any kind of diagnostic test, my first response is usually “What are we going to do with the results?” Mostly, I just want to get an idea of what kind of treatment the physician is likely to recommend based on the test findings. After all, if the recommended course of action is something that I don’t intend to follow, then what is the point of getting the test in the first place?

For example, if a pregnant woman has already decided she could never get an abortion, I don’t think it makes a lot of sense in most cases for her to agree to an amniocentesis, in which a needle is inserted into her womb to withdraw some amniotic fluid for genetic testing. The procedure itself has a small risk of triggering a miscarriage, so why take chances to get information about a possible abnormality? She’ll find out soon enough—after the baby is born—if the child is healthy or needs medical attention.

But there’s another reason to ask “What are we doing with the results?” whenever a clinician wants “just to run a few tests.” Sometimes, in the search for one thing, you may discover something else. Such so-called incidental findings can be quite momentous—such as when an x ray for a cracked rib reveals a shadow on the lung that’s suggestive of cancer. Or they can lead to a lot of anxiety for no good reason.

Years ago, someone in a biology lab at my school unexpectedly discovered that he had an extra X chromosome while completing one of the class assignments. (This was in the early days of do-it-yourself chromosome testing and the news spread pretty quickly among the other students.) The young man in question exhibited no clear signs of his unusual genetic makeup and probably could have spent the rest of his life happily unaware—at least until he tried to have children. (Men with the XXY condition often have fertility problems.) As it was, he was profoundly embarrassed—although no one teased him, as far as I knew. The school later took measures to make sure students were better informed about what they might discover when they took part in this lab project, but the incident made a big impression on me.

Robert C. Green, a genetics researcher and the co-chair of the ACMG working group that came up with the new guidelines, knows first hand about how common incidental genetic findings are. He recently had the protein-coding part of his genetic code (the so-called exome, which represents about 1 percent of any given individual’s entire genetic makeup) sequenced by a commercial gene-typing company. They reported that he had a genetic variant for Treacher-Collins syndrome, an inherited birth defect that often leads to an underdevelopment of the bones in the head, resulting cleft palate or in its most severe manifestations, severe disfiguration of the jaws and cheek bones.

The thing is, Green does not exhibit any of the outward signs of Treacher-Collins syndrome. He apparently has what’s now being called a non-pathogenic variant of the gene. As Green’s own example makes clear, there is no reason to report an incidental finding of a variant for Treacher-Collins syndrome taken from an adult. (The new ACMG guidelines do not apply to genetic screening tests of newborns, which have their own problems.)

Unless you are specifically looking for a Treacher-Collins gene, the vast majority of its genetic variants have no clinical bearing. For whatever reason, the variants that are most likely to show up as result of a search for something else—the definition of an incidental finding—are not going to give you any medically meaningful information, and they could do a lot of harm.

Update: You can read the full text of the ACMG’s guidelines, with my annotations, below. See also what Ricki Lewis has to say about the new recommendations and the tension between “genetic empiricists” like her and “genetic libertarians.”

About the Author: Christine Gorman is the editor in charge of health and medicine features for SCIENTIFIC AMERICAN. Follow on Twitter @cgorman.

The views expressed are those of the author and are not necessarily those of Scientific American.

Rights & Permissions

Comments 11 Comments

Add Comment
  1. 1. T. E. Hieatt 3:19 am 03/22/2013

    I don’t agree. If I have testing done, I want all knowledge available. Instead of withholding information, simply explain the information better. Incidental findings should be packaged with information about level of severity or lack thereof. Patient education. Perhaps it’s time for genetic companies to employ genetic counselors. And then there’s the issue of liability for not reporting an incidental finding that turns out to be a serious or deadly condition.

    Link to this
  2. 2. EvolvingApe 4:19 am 03/22/2013

    What a pompous piece!

    While you may not know what to do with the result, or how to interpret it, someone else may have the specific knowledge.

    The test is already done, all the results should be made available to the patient and the professional who requested the tests, with appropriate disclaimers.

    The pregnant woman example above is also presumptuous: even people with strong religious convictions have been known to change their minds when faced with reality (if rarely).

    The bottom line is: it is not for you to decide.

    Link to this
  3. 3. Christine Gorman 6:55 am 03/22/2013

    T.E.H wrote: “simply explain the information better.” That’s the whole point. The genetic information can’t be better explained. No one knows what it means.

    The new guidelines assume there will be updates in the future, as new information becomes available. But this is the best advice from a range of genetic experts for the situation as it exists right now.

    Link to this
  4. 4. Christine Gorman 7:04 am 03/22/2013

    EvolvingApe wrote “While you may not know what to do with the result, or how to interpret it, someone else may have the specific knowledge.”

    Again, that’s the whole point, no one else has the specific knowledge at present. Whole gene scans are producing more data than ANYONE knows what to do with.

    As for the pregnant woman example, I agree it’s not for me to decide, it is her choice. But by saying that she might change her mind after the amniocentesis, you’re assuming that the amniocentesis shows a genetic defect.

    What about the more common situation, where the amniocentesis shows no genetic problems but the 1% risk of miscarriage kicks in and she loses the baby anyway? Given her believes, how will she feel about having caused the miscarriage of a perfectly healthy baby?

    Link to this
  5. 5. Christine Gorman 8:14 am 03/22/2013

    that’s beliefs, not believes.

    Link to this
  6. 6. vittles 9:11 am 03/22/2013

    Hi Christine,

    Not to ‘pile on’ here; but, your comments regarding amniocentesis are a bit misleading. First, the risk of miscarriage is considerably lower than 1%. It may be 0.5% at the high end and is likely closer to 0.25% or less in most facilities where the procedure is performed regularly (the Mayo clinic has stats on this). As there is still some risk associated with the procedure, I don’t believe it is usually offered unless the chances of finding some abnormality outweigh those risks.

    Unfortunately, I think this is just a poor example of a good point. The amniocentesis example is more similar to a situation where one may be diagnosed with a disease (or strong pre-disposition to develop a disease) for which there is no known treatment or cure (such as Alzheimer’s). The difference in the pregnant mother example used in the article is that there are known options for coping with the diagnosed ailment, of which abortion is only one of the choices (mothers may choose to look for adoptive parents, or seek help in learning to raise a special needs child such that she is ready to provide the best possible care from the moment of delivery).

    As you mention, the true reason for the recommendation is different from these examples. The reason is simply an inability to interpret the implications of the test results given our current knowledge (or lack thereof) of human genetics and epigenetics. This inability can easily lead to misinterpretation and either undue worry (or undue comfort) about a particular disease or condition.

    A better example might be if I were to ask for the blueprints to the Large Hadron Collider and someone actually gave them to me. I would have obtained information; but, I would have no way of making sense of that information. I might know that all of the information necessary to build a super collider is present in the data; but, I couldn’t tell you what any individual piece of that information means. Of course in this case, there are people who know how to interpret and use those blueprints, while in the case of genetic testing, the human species does not currently have enough knowledge to make sense of the information we are able to obtain.

    Link to this
  7. 7. neuroimagine 9:26 am 03/22/2013

    But there are a few mutations where information is available and there are prevention strategies.

    Link to this
  8. 8. vittles 9:34 am 03/22/2013

    Yes, and I don’t believe anyone is suggesting that those data not be shared with patients/care providers. The recommendation is to withhold the information that no one knows how to interpret (hence “total” reporting).

    Link to this
  9. 9. DavidJerk 10:32 am 03/22/2013

    It doesn’t matter if someone knows how to interpret it or not, withholding information is unethical.

    Link to this
  10. 10. vittles 12:19 pm 03/22/2013

    Withholding actual information may be unethical, and there has been a recommendation to share the discovery of all established (or confirmed) markers for diseases, particularly those that may be averted by preventative measures (See: This includes markers uncovered while looking for other variants or conducting other tests.

    Perhaps the issue here is not so much divulging information as it is divulging data for which there is not a verified connection to a particular disorder or disease. Information implies knowledge. What we are really discussing is premature release of data for which there is insufficient understanding of the implications of those data. It could be argued that sharing these data is unethical in that it holds the potential for misinterpretation of the data, and may lead to an undue course of action.

    Consider this scenario: there is a preliminary study that links a specific DNA sequence variant to ovarian cancer. Women may be advised to undergo a hysterectomy based their genotype for that sequence. Or they may be told they have an increased risk of developing ovarian cancer, causing them to undergo frequent screenings (and stress). If subsequent studies find that the link does not actually exist, then people have been unnecessarily harmed in the interim. One could argue that it is ‘more’ unethical to share those data than it would be to withhold them.

    Link to this
  11. 11. T. E. Hieatt 1:39 am 03/27/2013

    Christine G. wrote: “T.E.H wrote: ‘simply explain the information better.’ That’s the whole point. The genetic information can’t be better explained. No one knows what it means.

    The new guidelines assume there will be updates in the future, as new information becomes available. But this is the best advice from a range of genetic experts for the situation as it exists right now.”

    I think myself and others are quite capable of handling the ambiguity of the information. Withholding information is unethical and deceitful. It is not up to a lab geneticist or a bureaucratic panel to decide what information I should receive about my body.

    Link to this

Add a Comment
You must sign in or register as a member to submit a comment.

More from Scientific American

Email this Article