the AJMG SEQUENCE
The tricky matter of secondary genomic findings
ACMG plans to issue recommendations
Article first published online: 18 JUN 2012
Copyright © 2012 Wiley Periodicals, Inc.
American Journal of Medical Genetics Part A
Volume 158A, Issue 7, pages ix–x, July 2012
How to Cite
(2012), The tricky matter of secondary genomic findings. Am. J. Med. Genet., 158A: ix–x. doi: 10.1002/ajmg.a.35521
- Issue published online: 18 JUN 2012
- Article first published online: 18 JUN 2012
Clinical geneticists may soon have guidance on revealing to patients who get genomic testing specific disease-causing variants that are unrelated to phenotypic reasons for sequencing.
Currently, no guidelines exist for return of such secondary results, but the American College of Medical Genetics and Genomics (ACMG) expects to finalize such recommendations this summer.
The ACMG will likely recommend reporting germ line conditions and associated genes that are both well enough understood to determine their penetrance (whether having the mutation means developing a disease) and have treatments that can prevent or ameliorate symptoms, says Michael Watson, PhD, ACMG's Executive Director.
An ACMG workgroup has been formulating the guidelines since the beginning of the year. As a precursor to the guidelines, in March, the workgroup issued a position paper on genomic testing that defines terms to be used in the guidelines and describes indications for genomic testing, issues to consider before testing, and when genomic sequencing is appropriate for screening purposes. Robert C. Green, MD, a researcher at Boston's Brigham and Women's Hospital, and Leslie Biesecker, MD, Chief and Senior Investigator, Genetic Diseases Research Branch, National Human Genome Research Institute, chaired the work.
That position paper says that a clear distinction should be made between clinical and research testing, and encourages clinical laboratories to share genotypic data from genomic testing in public databases to more quickly generate information that will lead to improved care.
Dr. Watson describes the forthcoming guidelines as “first phase” because he and the workgroup anticipate the list of reportable genes and conditions growing longer as research reveals more about them. He hopes that the list will spur inclusion of more clinical and screening data into centralized databases, so the medical genetics field can learn from a population whose mutations aren't associated with phenotypes. Such data will provide “a much better general population perspective of the true conditions,” Dr. Watson says.
The ACMG will likely recommend reporting germ line conditions and associated genes that are both well enough understood to determine their penetrance and have treatments that can prevent or ameliorate symptoms.
Commentary Supports Scenario
The forthcoming ACMG guidelines are in line with a strategy suggested in an April Genetics in Medicine commentary by the journal's Editor-in-Chief, James P. Evans, MD, and Barbara Rothschild, MD. Dr. Evans is Bryson Distinguished Professor of Genetics and Medicine at the University of North Carolina at Chapel Hill (UNC-CH), Editor-in-Chief of Genetics in Medicine, and a member of the now disbanded Secretary's Advisory Committee on Genetics, Health, and Society. Dr. Rothschild is Assistant Professor of Social Medicine at the UNC-CH School of Medicine.
There's an obligation to return secondary findings of variants firmly associated with high risk of diseases with specific strategies that ameliorate expected morbidity and mortality, the pair write, noting that currently, variations in fewer than 100 genes meet these criteria [Evans and Rothschild, 2012]. Drs. Evans and Rothschild call for devising a list of genes and variants that meet these specifications, and a centralized process for evaluating and adding new ones.
In an interview, Dr. Evans says he sees the forthcoming ACMG guidance as a starting point. Having the guidance “doesn't mean we have to move in lockstep,” he adds. “People can implement a given list in an intelligent way. That list can be revised according to new data,” he says.
Agreeing on a List
A paper in the same issue of Genetics in Medicine concludes that getting agreement on a list of what secondary results to reveal could be difficult, but is feasible.
The paper details responses from 16 specialists in clinical genetics and molecular medicine. The specialists were tasked with selecting disease-causing variants in 99 common conditions that should be returned to the ordering physician if incidentally discovered through genomic sequencing. Dr. Green is the paper's first author [Green et al., 2012].
The 16 respondents, who include Dr. Evans, agreed that pathogenic variants for 21 diseases in adults and 4 such variants in children, when discovered incidentally, should be revealed. In general, the respondents disagreed about the relative value of different criteria when making decisions about disclosure. Some thought whether the patient was an adult or child was the most important reason for disclosing or not, while others thought the fact of established pathogenicity was the central issue.
Dr. Evans sees these results as a reason for optimism about the ACMG guideline effort. “No one [in the study] set up criteria for incidental results return or tried to achieve consensus,” he says, adding the paper shows that setting criteria and agreeing on specific mutations and diseases is a realistic goal.
The View from the Lab
Like medical geneticists, labs are also grappling with how to handle challenges presented by genomic sequencing. In Salt Lake City, Utah, the large reference lab ARUP has offered whole genome and exome sequencing for research purposes since February and began clinical exome sequencing in May. At that time, ARUP did not return secondary results for clinical testing, according to Elaine Lyon, PhD, ARUP's Director of Molecular Genetics and Associate Professor of Pathology at University of Utah
That's because “analysis is guided by disease and family history,” she says. “Looking for incidental findings may need a different analysis. We don't want it to be expected, so we won't look outside of the symptoms at issue.”
When ACMG guidelines recommend a list of pathogenic variants for reporting as secondary results, ARUP may change its policy, Dr. Lyon says. It is possible ARUP may allow patients to choose to receive such results. But that may be another service with an additional fee, she adds, emphasizing that a laboratory's policy of what will or won't be returned should be clearly stated in the test description and the reports.
While Dr. Lyon welcomes a policy on results return, she sees problems in the details. In particular, she wonders how the authors of the ACMG guideline will define what diseases are treatable and how much penetrance a gene needs to be recommended as reportable.
Dr. Lyon also questioned whether the guideline would address the common practice of testing parents as part of diagnosing a child. Would labs be required to reveal parents' secondary results? She gave as an example BRCA mutations that confer high risk of breast cancer, wondering whether the ACMG guidelines would direct labs and clinicians to report such a finding in a mother, or to report risk for this adult-onset disease in a child.
Another paper in the April Genetics in Medicine points out other potential problems with a policy of returning incidental and secondary genomic results to research subjects. Bioethicists argue that while rare instances may warrant it, such results return as widespread practice or obligation could result in more legal liability [Clayton and McGuire, 2012]. The bioethicists argue that if a physician does research involving patients, he or she may have a clinical obligation to follow up regarding incidental results. But the authors add that monitoring all aspects of health on a continuous basis is not a clinical geneticist's responsibility. Any obligation to return secondary genomic findings “does not necessarily extend to investigators or scientists who have no clinical relationship with the participant,” they add.
“Looking for incidental findings may need a different analysis. We don't want it to be expected, so we won't look outside of the symptoms at issue.” —Elaine Lyon, PhD
Dr.Lyon calls for more communication between labs and clinicians as genomic sequencing becomes more common. “I don't think either side can figure out the exome by itself. I envision more back and forth between the doctor and lab to achieve the best interpretation,” she says.
Those involved in genomic sequencing should develop principles for returning secondary and other results so payors can make good decisions about payment and follow-up and avoid establishing multiple standards of care, Dr. Watson says. To harness the power of the genome to improve care, he says geneticists and researchers must also “find ways to capture as much genomic data as possible in an organized way and learn from it. There's no doubt we have to get return of results right.”
- 2012. Return of results: not that complicated. Gen Med 14:358–360. , .
- 2012. Exploring concordance and discordance for return of incidental findings from clinical sequencing. Genet Med. 14:405–410. , , , , , , , , , , , , , , , , , .
- 2012. The legal risks of returning results of genomics research. Genet Med 14:473–477. , .