NEW YORK – The push toward universal genetic risk testing for all breast cancer patients before there is evidence to support such an expansion may result in more harm that good, some experts fear.
The American College of Medical Genetics and Genomics released a statement today that all breast cancer patients should be evaluated to determine if they should receive germline genetic testing for assessing their inherited risk for the disease, according to the National Comprehensive Cancer Network's guidelines. However, the ACMG experts did not advocate for testing all breast cancer patients, as the American Society of Breast Surgeons (ASBrS) did earlier this year.
"Currently, there is insufficient evidence to recommend genetic testing for BRCA1/2 alone or in combination with multi-gene panels for all breast cancer patients," ACMG experts led by Tuya Pal from the Vanderbilt University Medical Center wrote in Genetics in Medicine, the official journal of the ACMG.
Doreen Agnese, a surgical oncologist and clinical geneticist at Ohio State University Comprehensive Cancer Center, who is also an author of the ACMG statement, noted that with greater use of multi-gene panels there is potential for mistreatment when providers who don't fully understand genetics are making decisions. "Sometimes, things are misinterpreted," she said.
Additionally, there is concern in the oncology community that the push for universal genetic testing could overwhelm the system with insurance coverage issues, since policies tend to track with NCCN guidelines. At OSUCCC, Agnese and her colleagues discuss the availability of genetic testing with all surgical patients and evaluate them to assess if they meet guidelines for testing, but they also tell patients that the test may not be covered if they don't meet guidelines. "We don't test everybody right off the bat," she said.
However, in their "Points to Consider" statement, the ACMG also recognized that the field is currently doing a poor job of ensuring that all high-risk breast cancer patients who currently meet NCCN guidelines have access to genetic testing and urged expert bodies to advance harmonized guidelines and lift existing barriers to care. "One thing all the authors of this paper agreed on is that we need to spend more time trying to get the people who do meet the criteria to get tested," said Mary Daly, a clinical geneticist who directs the risk assessment program at the Fox Chase Cancer Center and a co-author on the ACMG statement.
According to one estimate, only one-fifth of breast and ovarian cancer patients who meet guidelines actually receive testing, and the access gap is further widened in certain racial and ethnic groups. "If we had to spend our energies somewhere, wouldn't that be the place to do it first?" said Daly, who is also the head of the expert panel that authored the recently updated NCCN guidelines on genetic testing for assessing breast, ovarian and pancreatic cancer risk.
In the latest iteration, the NCCN continues to recommend clinicians take a risk-based approach when deciding which patients to genetically test. For breast cancer patients, personal and family history, and factors like age of diagnosis, should be considered. For certain subsets of breast cancer patients, the NCCN has found sufficient evidence to recommend testing all, such as those with triple negative breast cancer who are 60 years old or younger, or individuals with Ashkenazi Jewish ancestry.
The NCCN also recognizes that the availability of multi-gene panel testing has changed testing patterns in the field and that these tests can identify patients with mutations in genes with varying levels of evidence on cancer risk. As such, the latest guidelines discuss not only how to manage patients with mutations in high-risk or high-penetrance breast cancer genes, such as BRCA1/2, PALB2, TP53, and PTEN, but also state whether there is evidence supporting follow-up actions when patients have mutations in other genes that increase cancer risk more modestly (moderate penetrance), such as CHEK2 and ATM. The NCCN guidelines specifically note when the association between genetic variants and breast or other cancers isn't yet established.
The ACMG's "Points to Consider" document similarly emphasizes a risk-stratified approach and is "completely in line with NCCN," said Pal, who is also a member of the NCCN's expert panel and worked on the updated NCCN guidelines with Daly.
The evidence supporting testing for different genes may emerge at different times, but clinical geneticists involved in promulgating the NCCN guidelines are regularly evaluating this data, Daly said. "The best contribution the NCCN can make to the field is to review the data carefully and critically so we have the best evidence when we do make a recommendation," she added.
However, the ACMG's position is notably in contrast to the ASBrS, and others in the field who would like to see less restrictive testing guidelines in the era of multi-gene panels. The breast surgeons' group justified expanding testing, noting that in current medical practice, many patients with mutations in moderate penetrance genes, but also those with mutations in BRCA1/2, are not being identified. While the ASBrS acknowledged that greater access to testing will lead to more patients getting tested, the group reasoned that "breast surgeons are well positioned to be a resource for patients who may benefit from testing."
The push to broaden testing access is being spurred by dropping costs in the genetic testing industry. A decade ago it cost several thousands of dollars to test for mutations in a single gene. Now, it is possible to sequence hundreds of genes for the same price. For a few hundred dollars, people can go online and learn not only if they harbor mutations in BRCA1/2 genes, but dozens of other hereditary cancer risk genes, without having to step into a doctor's office.
Amid the enthusiasm for more testing, many of the experts involved in writing the ACMG statement and the NCCN guidelines would like to proceed more carefully, recognizing that testing all breast cancer patients for genes lacking clinical utility evidence have the potential to harm patients and increase healthcare costs. In their statement, the ACMG experts take issue with one particular study that the ASBrS cited in recommending that all breast cancer patients receive genetic testing for BRCA1/2 and PALB2, and if deemed appropriate based on clinical factors and family history, for other genes as well.
That study published in the Journal of Clinical Oncology involved breast cancer patients tested on an 80-gene panel, and showed a similar rate of pathogenic and likely pathogenic variants among patients who met the NCCN's testing guidelines and those who didn't. The study authors, co-led by Peter Beitsch from the Dallas Surgical Group and Robert Nussbaum, chief medical officer of the genetic testing firm Invitae, concluded that a testing strategy guided by NCCN criteria would miss nearly half of the patients with breast cancer who have a potentially actionable pathogenic or likely pathogenic variant. "We recommend that all patients with a diagnosis of breast cancer undergo expanded panel testing," they concluded in the paper, entitled "Underdiagnosis of Hereditary Breast Cancer: Are Genetic Testing Guidelines a Tool or an Obstacle?"
"If it hadn't been for the surgeons coming out and making their recommendation, this would have just been one more paper to consider and review as to how the faults in the paper could be improved upon with additional study," said Daly. "Most of us who do this work for a living, who take care of these families and make decisions about genetic testing, were sensitive to the flaws of the study and didn't feel that it warranted such a big change in policy."
Mainly, the ACMG experts challenge the study's conclusions that NCCN's risk-based criteria would miss around half of breast cancer patients. First, they note that the study excluded previously tested patients, who would have been more likely to meet NCCN guidelines for assessment of high-risk genes. Additionally, the 80-gene panel included genes with unclear association to breast cancer or genes for which there isn't enough evidence to recommend management actions.
"You can do more and find more stuff, but we don't know if that other stuff is going to lead to patients living healthier and longer," said Susan Domchek, director of the Basser Center for BRCA at the University of Pennsylvania, who wrote an editorial accompanying the ACMG statement in Genetics in Medicine.
The ACMG experts pointed out, however, that if the Beitsch et al. analysis had been restricted to likely pathogenic and pathogenic variants in highly penetrant breast cancer-linked genes, such as BRCA1/2, PALB2, TP53, and PTEN, then less than 1 percent of patients who harbored pathogenic mutations in these genes would be missed by current guidelines. The guidelines start missing more people, as more moderate risk genes, such as ATM and CHEK2 are considered, though not all the genes included in that analysis were clearly associated with breast cancer risk.
Pal thought it particularly concerning that Beitsch et al. included individuals who were heterozygous carriers of the MUTYH gene in their evaluation to see if NCCN criteria could identify breast cancer patients with pathogenic variants. However, there is no established association between breast cancer and heterozygous MUTYH mutations, which in and of themselves have a relatively high carrier frequency in the general population of between 1 percent and 2 percent.
One of the real-world challenges that many in the field are observing with the rapid dissemination of multi-gene panels is that some clinicians are managing patients with mutations in other genes as if they have mutations in high-risk BRCA1/2 genes associated with breast, ovarian and pancreatic cancers. "MUTYH mutation carriers are not BRCA mutation carriers," Pal said.
The growing popularity of consumer-initiated genetic testing is another complicating factor. The consumer genomics industry has been able to increase people's interest in testing by making it affordable, easy, and "fun" for people to order testing online, without consulting their own physicians.
On the one hand, this has enabled individuals who otherwise may not have been tested or couldn't previously afford it to learn their genetic risk for cancer and other serious diseases. On the other hand, physicians and genetic counselors who have interacted with these consumers, have recounted that they are coming away from this testing with a limited understanding of genetic risk and the test limitations.
In Domchek's view, the affordability of next-generation sequencing panel tests, which has made it possible to test not just for mutations in BRCA1/2 but for many genes at one time, has confounded this whole issue. "If we were still talking just about BRCA1/2 testing and it is as cheap as it is now, I don't think there would be as much concern about this whole concept," she said. "But now it is almost never someone who is just testing for BRCA1/2. There are these large panel tests and they have made this really complicated because … there are these rising mastectomy rates in the US and we have to be super careful that genetic risk isn't misunderstood."
One study showed that genetic testing rates and risk-reducing mastectomies increased after actress Angelina Jolie penned an editorial in 2013, discussing the preventive measures she took after finding out she was a BRCA1 mutation carrier. With the growth in multi-gene panels, clinicians are continuing to see patients who have taken or have wanted to take inappropriate surgical actions based on findings in a range of other genes. GenomeWeb sister publication Precision Oncology News is publishing an anonymized case series with My Gene Counsel that highlights some of these challenges.
Domchek recounted how she had seen several women recently who had no family history of breast cancer but had a CHEK2 mutation. According to the NCCN, certain types of CHEK2 mutations may confer an increased risk for breast cancer, but with other types of mutations in this gene, the risk may be lower. The NCCN states that clinical management decisions should be based on the best risk estimates for specific CHEK2 variants. Moreover, the expert group notes that while increased screening may be warranted for certain mutations, there isn't enough evidence to support risk-reducing mastectomies. And yet, several patients were planning on preventative bilateral mastectomy, Domchek said.
Pal and other researchers conducted a study in which they tracked cancer risk management practices in 238 women who had likely pathogenic or pathogenic mutations in BRCA1/2, PALB2, CHEK2, and ATM, and found that risk-reducing mastectomy and oophorectomy were appropriately high among BRCA1/2 mutation carriers. However, in this study (which has been presented but not yet published), mastectomies and oophorectomies were also relatively frequent in patients with pathogenic or likely pathogenic mutations in CHEK2 and ATM, as well as oophorectomies for individuals with PALB2 mutations, even though NCCN guidelines don't recommend risk-reducing surgeries in such cases, and the association of these genes to heightened ovarian cancer risk isn't fully established. This data signaled to Pal and her research colleagues that patients with mutations in non-BRCA genes are receiving surgeries even in the absence of evidence-based guidelines.
Testing more genes also detects more variants of unknown significance. One study documented how some patients with VUS in BRCA1/2 received bilateral mastectomy, despite guidelines recommending against making clinical decisions based on VUS.
At OSUCCC, Agnese recently provided a second opinion to a woman who had a VUS in the ATM gene, which moderately increases breast cancer risk. This patient had been told previously that she had a 60 percent chance of getting breast cancer, despite the fact that the vast majority of VUS are reclassified as benign. She had also been previously recommended for a double mastectomy, even though such a course isn't recommended for VUS results, or even for individuals who have a pathogenic mutations in this gene.
"Most surgical and medical oncologists have a pretty decent understanding of BRCA1/2, because [knowledge of those genes] have been around for a while and people are pretty comfortable with them," Agnese said. "Where it falls off is people don't have that depth of knowledge for some of the other genes that are in these high-risk panels now, so there's a potential for misinterpretation there." Although some doctors do their due diligence and look up the guidelines on specific genes, there are others who think of every gene as being similar to BRCA1/2 and make recommendations without doing their due diligence, she said.
Susan Klugman, VP of clinical genetics at ACMG, has also seen women who have chosen to have prophylactic surgeries after learning they have a VUS. She recounted the case of a breast cancer patient in her 50's who had a VUS. Her sister, who didn't have breast cancer, also received testing and upon learning she had inherited the same VUS, went to a surgeon and had a double mastectomy.
Because these types of cases are happening, Domchek said it's important to remind the field, as the ACMG is doing, that not all cases are equivalent in terms of genetic risk and that risk stratification still matters. "We don't want to miss the people who are most likely to have a mutation," she said, noting as an example that a 40-year-old woman with triple-negative breast cancer and an 80-year-old with estrogen receptor-positive breast cancer and no family history don't have the same risk of having a mutation in a high-penetrance gene. "Putting the concept of risk stratification front and center and keeping that in [the clinician's] mind is super important."
Ed Esplin, a clinical geneticist from Invitae, an author of the Beitsch et al. paper and an author of the ACMG statement, described the ACMG statement as a step forward in that it asks clinicians to consider evaluating all breast cancer patients to determine their need for testing, and it calls for research to continue to expand the evidence base for testing. Invitae has been aggressively driving down prices in genetic testing and is an industry proponent of broadening test access.
"We know that even patients who meet current criteria are under-tested," Esplin said. "Our hope is that this statement offers guidance to practicing clinicians today even as the community continues to build the evidence base to guide where testing is headed."
Although the ACMG statement doesn't include any specific proposals for how the access gap may be narrowed, institutions serving ethnically diverse and economically disadvantaged communities have implemented strategies to improve the testing rate among those who meet guidelines. For example, when patients come in to discuss a high-risk mutation they have with a genetic counselor, some institutions are providing a brochure they can share with relatives or a letter that they can send to family members they're not in regular touch with.
"It's very difficult to get some family members to come in for testing," said Klugman. In minority populations, many don't have insurance or are underinsured, and may not speak English. There may be distrust and fear in such communities about genetic testing and potentially losing medical insurance if they have it, Klugman said. Additionally, she noted that some hospitals are also conducting outreach to educate community providers about the latest guidelines on who should be genetically tested for inherited cancer risk, and that genetic testing is much more broadly covered than it was just a few years ago.
Domchek advised greater use of electronic health records, digital health platforms, telegenetics, and video-based point-of-care strategies. "Community oncologists need to be able to do this in a more efficient manner. We're long gone from the days when genetic testing was only happening in tertiary care facilities," she said. "Instead of debating how low we should go on the testing threshold, let's look for ways to test every single patient who is at high risk."