ACMG guidelines on IFs – responding to the response…

StethCollage1The ongoing debate about whether, what, when and how to feedback incidental findings (IFs) from whole genome sequencing continues to rage on both sides of the Atlantic following the American College of Medical Genetics and Genomics’ controversial recommendations on reporting IFs released last month. In an unexpected twist, the authors of the guidance have now written “a clarification” in response to the many criticisms that have been raised including here on GenomesUnzipped. The clarification covers five points – autonomy, children, labs, communication and interpretation.

(1) Patient autonomy – the recommendations state that patients cannot opt-out of having an extended search for known pathogenic variants in 57 genes when receiving a diagnostic genome sequence. According to the clarification, the clinician will ‘contextualize’ these results and can participate in a ‘shared decision-making process regarding the return of results’. I interpret this to mean that although the lab should not withhold results from the clinician, the clinician can withhold results from the patient. This potentially places clinicians in a rather sticky situation – deliberating over and deciding whether or not to report an ‘incidental finding’ (potentially at the behest of a well-informed, opinionated and obstinate patient), which could undermine honesty and trust in the clinical relationship. Why put clinicians and patients in this uncomfortable situation in the first place by deliberately doing non-voluntary opportunistic screening? The authors emphasise that the recommendations only relate to “unequivocally pathogenic mutations in genes where pathogenic variants lead to disease with very high probability”. Presumably the point here is that everyone will want to know about these variants because they are both definitely pathogenic and treatable? I would be interested to see a list of the specific variants to be reported, along with their population ascertained frequencies – as I’ve said before, the association between a genotype and disease phenotype may be weaker than we think, particularly for rare Mendelian variants where almost all our knowledge about the impact and penetrance comes from affected families.

(2) Incidental findings in children – perhaps the most controversial of all the ACMG’s recommendations is that they apply equally to adults and children. In particular, children who are being tested (to diagnose a developmental disorder, for example) should be screened for a number of adult onset conditions because of potential benefits to their parent who probably also carries the same variant. This flies in the face of numerous internationally agreed guidelines (including from the ACMG itself) specifically recommending against testing children for adult onset diseases. The clarification attempts to affirm both these positions by saying that opportunistic screening of children for adult onset conditions they probably don’t have is fine, but diagnostic testing for conditions they are at high risk for is not. That logic doesn’t make any sense – either children should be tested for adult onset conditions because of potential future benefits to them and their relatives through cascade testing, or they should not because it undermines their future autonomy and may unfairly bias their upbringing. Can a professional society really sit on both sides of this fence?

(3) Clinical laboratories – the clarification recognises that whole exome/genome sequencing may not have adequate coverage in all of the 57 genes to be screened, and that labs should make this clear in their reporting. This is an important point, and I think their position is fair enough – screening tests aren’t usually as accurate as diagnostic tests, and I’m sure that any finding reported as a result of opportunistic genome screening would be validated before being acted on. The aim of genome sequencing is still diagnosis, so the test should be optimised for that.

(4) Result communication – the clarification underlines the importance of returning genomic results to the patient via a knowledgeable clinician, who is able to deal with the patient, take a decent family history and consult with other specialists should IFs arise. Presumably this point was raised due to inevitable though inappropriate comparisons between the ACMG recommendations and direct-to-consumer genome testing companies that give genomic results relating to hundreds of conditions.

(5) Predicting disease likelihood – finally, the authors use the clarification to explain why they chose to draw this controversial line in the sand: “to develop an open and transparent means by which the community can offer input into the further curation of the list of recommended conditions/genes/variants”. They acknowledge that there is much still to be learned about predicting disease from genome sequences, and that a central genotype-phenotype repository needs to be developed to further the process. The importance of these recommendations is therefore not in the exact details of the genes or diseases to be screened, which we can all argue about and ultimately more data are needed to make an evidence-based decision, but in the crux of their argument – that opportunistic genome screening could and should be performed where there is evidence of clear benefit. I think most people can agree with that sentiment, but unfortunately the devil is in the detail. Until it is possible to determine that a ‘pathogenic variant’ is clinically useful in an individual with no family history, why not offer opportunistic screening as part of a research study rather than a clinical service?

  • Digg
  • StumbleUpon
  • del.icio.us
  • Facebook
  • Twitter
  • Google Bookmarks
  • FriendFeed
  • Reddit

2 Responses to “ACMG guidelines on IFs – responding to the response…”


  • “The clarification attempts to affirm both these positions by saying that opportunistic screening of children for adult onset conditions they probably don’t have is fine, but diagnostic testing for conditions they are at high risk for is not. That logic doesn’t make any sense”

    ACMG’s logic does make sense if you prioritize the prevention of disease for at-risk family members.

    Before next-gen sequencing, the way a clinician would know a healthy child was at risk for a medically actionable, adult-onset condition would be if someone else in the family was diagnosed with it. In this situation, there is already an opportunity for all the older at-risk relatives to be warned and the child’s testing can be deferred until s/he is an adult and/or clinical screening begins.

    After next-gen sequencing, a child’s incidental finding may inform family members of their risk for a serious disease amenable to screening/treatment before anyone in the family has become sick. It seems preferable to incidentally detect a hereditary disease in a family through a child’s exome sequencing versus after a family member develops an illness that could have been prevented or detected at a stage more responsive to treatment.

  • Bob Cook-Deegan

    NCBI has done a really nice list of the items on the ACMG “must report” list: http://www.ncbi.nlm.nih.gov/clinvar/docs/acmg/

    Re “must report,” it means the labs must report to the clinical intermediary, not necessarily the clinical person to the patients being sequenced (this is clearer in the “clarification” after the initial letter, and in Amy McGuire’s Policy Forum article in Science last week, although the information flow map is still a bit unclear from lab to person tested).

    I don’t think ACMG is trying to sit on both sides of the fence re notifying children. The conditions listed are mainly cancers and cardiac diseases that might affect a child, or for which childhood risk factors might influence outcomes. There are some exceptions (e.g., BRCA–is that “actionable” in childhood?). But most are conditions that might induce action during childhood.

    Finally, I have not heard any clinicians who run labs or counseling services who think the figure of 1 percent (fraction of those who would get notified of a deleterious mutation in one of the 50-some genes) is that low. Have heard figures 4 to 8 times higher than that.

Comments are currently closed.

Page optimized by WP Minify WordPress Plugin