Skip to main content
Premium Trial:

Request an Annual Quote

ACMG Plans Update of Secondary Findings Gene List by Year End, Considers PGx Genes for Next Round


NEW YORK (GenomeWeb) – The American College of Medical Genetics and Genomics is working on a first update to its gene list for reporting secondary findings from exome or genome sequence data with the aim of finalizing the revisions by the end of the year, GenomeWeb has learned.

In addition, an ACMG committee tasked with the update has been looking into how pharmacogenomics genes can be added to the list without unduly burdening clinical labs reporting the results, but PGx genes will only be implemented in future updates.

According to Christa Martin, director of the Autism and Developmental Medicine Institute at Geisinger Health System, and David Miller, medical geneticist at Boston Children's Hospital, who co-chair the ACMG Secondary Findings Working Group, the first update will likely involve the addition of five genes to the list and the removal of one gene. The proposed changes, based on suggestions from ACMG members, are still subject to approval by ACMG's board of directors, which will consider them later this month.

ACMG first put out a policy statement with recommendations for reporting incidental findings in clinical exome and genome sequencing data in 2013. The statement included a list of 57 genes — later revised to 56 — that are involved in hereditary cancer syndromes, cardiomyopathies, and other inherited disorders. The following year, ACMG revised those recommendations to allow patients to opt out of secondary findings and to incorporate feedback from a member survey, but the gene list remained the same. 

The current working group, which started in 2014, was charged with developing a process for reviewing and updating the gene list, based on suggestions from ACMG members. Under a nomination process devised by the group, members can nominate a gene for addition or removal from the list and submit their evidence via email.

Genes need to be "medically actionable" — meaning effective interventions must exist that alter the course of the disease — and deleterious mutations in the gene need to be associated with a clear phenotype with serious medical implications.

The group also developed "a more rigorous process" that uses semi-quantitative metrics to evaluate the actionability of genes, Miller said, "to be a little more objective about deciding on the merits of having that gene either added to or removed from the list."

For example, some interventions have not clearly demonstrated that they modify the outcome of the disease, "and that's one of the things that we want to take very seriously for the genes that are on this list — we don't want people to know about disease risk and there is not really a clear path forward on what they can do about it," he said.

So far, the group has received six nominations for genes involved in inherited cancer and in metabolic disorders. Five of them were requests for adding a gene, and one for removing a gene from the list that does not meet the standards for actionability, he said.

The group is currently finalizing a manuscript that will describe its work in more detail and discuss its plans going forward.

According to Martin, the goal is to open the nominations process, currently restricted to ACMG members, to the broader medical community, for example, to professional societies of other specialties that want to nominate genes in their disease area.

Another aim is to incorporate pharmacogenomics genes in the list. "Everybody agreed that it's important, we just have to figure out the right way to do it," Martin said. Stanford University's Teri Klein, director of the Pharmacogenomics Knowledgebase (PharmGKB) and a co-principal investigator of the Clinical Pharmacogenetics Implementation Consortium (CPIC), has joined the working group specifically to provide expertise in this area, she said.

Pharmacogenomics genes "don't fit into the same model of disease which we're used to for the other genes," Miller explained, which is why the group has been developing a modified process for evaluating nominations for PGx genes. "We want to do it in a way that is helping people to learn about actionable information without having to add too much of an increased burden on the laboratory providers and the clinical providers," he said.

PGx genes are fundamentally different from the genes that are currently on the list because they do not predispose to disease or an adverse event on their own, but only in combination with exposure to a drug, Miller said. Secondly, the information gleaned from PGx genes may not be immediately actionable but might only be relevant in the future, when a drug is prescribed.

In addition, the frequency of reportable pharmacogenetic variants — depending on which ones are included — could be much higher than the frequency of pathogenic variants in the current gene list. So far, only about 5 percent or less of patients have had secondary findings in the 56 genes, Miller said, but a majority of patients may harbor some type of pharmacogenetic result. That, in turn, would increase the workload for clinical labs having to report these findings.  

Finally, pharmacogenomics results often don't just involve isolated variants but haplotypes that need to be reconstructed, which represents a technical hurdle, and many are found in non-coding regions of the genome that are not included in exome data.

For PGx gene nominations, the Secondary Findings Working Group will thus prioritize genes associated with medications that patients are more likely going to be exposed to. "We are going to have to draw a line somewhere — there is no way that we can put all pharmacogenetic variants in, so we are going to have to figure out the criteria," Miller said. 

In any case, PGx genes won't become part of the list until at least the next round of revisions. The working group currently plans to make updates every six months at most. "Every time you update that list, laboratories have to change their pipelines for analysis, and we do not want to overburden labs trying to keep up with that," Martin said.