ORLANDO – A large swath of people whose genetic variants put them at increased risk of developing cancer are missed under current genetic testing criteria, according to a study presented Tuesday during the American Association for Cancer Research (AACR) annual meeting.
The findings lend further fuel to a passionate contingent of oncologists who say genetic testing should be universally offered, or at the very least, that the criteria become less stringent.
"With decreasing whole-exome sequencing costs of below $500 and the ability to complete whole-exome sequencing on saliva samples of thousands of patients, we can now approach comprehensive, population-based whole-exome genomic sequencing to identify genetic carriers of disease," Jewel Samadder, director of the high-risk cancer clinic at Mayo Clinic, said in a presentation to members of the media.
Through the ongoing TAPESTRY study, researchers have so far performed whole-exome sequencing on more than 44,300 participants across three Mayo Clinic sites in Arizona, Florida, and Minnesota. The study's enrollment criteria are deliberately broad; most anyone treated at one of these three sites for any condition could consent to testing regardless of whether they have a cancer diagnosis or family cancer history.
Emily Gay, a graduate genetic counseling student at the University of Arizona, presented an analysis of the results during AACR. Of 550 people found to have either hereditary breast or ovarian cancer syndrome (HBOC) or Lynch syndrome — 1.24 percent of the participants screened so far — 39.2 percent would not have qualified for cancer risk testing under current National Comprehensive Cancer Network (NCCN) guidelines, Gay said.
Among 387 patients with HBOC, who had pathogenic or likely pathogenic variants in the BRCA1 or BRCA2 genes, 32 percent wouldn't have qualified under current NCCN screening guidelines, and for 387 people with Lynch syndrome, with pathogenic or likely pathogenic variants in MLH1, MSH2, MSH6, PMS2, or EPCAM, 56.2 percent wouldn't have qualified for genetic testing under the guidelines.
Having one of these germline genetic variants significantly increases cancer risk and can guide preventative measures such as prophylactic mastectomies or more frequent and aggressive imaging. For instance, people with HBOC have an 80 percent lifetime risk of developing breast cancer and significantly higher risk than the general population of developing ovarian cancer, prostate cancer, pancreatic cancer, or melanoma.
People with Lynch syndrome, meanwhile, have an 80 percent lifetime risk of developing colorectal cancer, a 60 percent lifetime risk of developing endometrial cancer, and an increased risk of developing upper gastrointestinal, urinary tract, skin, or other cancers.
Receiving a diagnosis of HBOC or Lynch syndrome can help people catch these cancers earlier or take steps to mitigate them. The benefit of knowing about these variants, in fact, is great enough that the US Centers for Disease Control and Prevention categorizes them as Tier 1 genetic conditions (CDCT1), signifying that testing can improve public health outcomes. But too few people are getting screened for these variants, according to the Mayo Clinic researchers.
"Our results really emphasize the need for increased access to genomic screening for CDC Tier 1 genetic conditions, and potentially the use of exome sequencing in large populations," Samadder said.
As Gay recapped during her presentation, the NCCN criteria for genetic testing for cancer risk variants, which providers and payors often use to determine who to screen and which tests to reimburse, consider factors such as personal and family cancer history, age of cancer diagnosis, and tumor features. However, according to the Mayo Clinic findings, these factors seem to be missing well over a third of patients with HBOC and Lynch syndrome.
For the TAPESTRY study, the researchers collaborated with Helix to perform whole-exome sequencing on participants' saliva samples, which covers less than 2 percent of the genome but more than 85 percent of all disease-related variants, Gay said. They considered electronic health records to determine who would have been eligible for screening under the NCCN criteria.
Broken down by genes, the researchers determined that 63.8 percent of people with variants in MSH6 would not have met NCCN screening criteria. The same was true for 83.7 percent of those with variants in PMS2.
The top reasons for falling outside the testing guidelines included a lack of personal cancer history, an insufficient number of relatives with cancer history, or a personal or family history of a cancer type not related to the genetic syndrome, Gay said.
Even among people who did meet NCCN genetic testing criteria, 34 percent were not aware of their HBOC or Lynch syndrome diagnoses before the Mayo Clinic study because they had not had prior testing.
"This suggests that the NCCN guidelines are underutilized in clinical practice, potentially due to the busy schedule of clinicians, or because of the complexity of using these criteria," Gay said.
Subanalysis finds troubling disparities
In a subanalysis of their findings, Gay, Samadder, and colleagues found that 49 percent of participants from underrepresented minorities whose whole-exome sequencing tests identified Lynch syndrome or HBOC wouldn't have met the NCCN criteria for testing, whereas the same was true for just 32 percent of white individuals.
Looking ahead, Gay and Samadder hope to garner additional data on these underrepresented minority participants as the TAPESTRY study moves toward its 100,000-person target enrollment. In the meantime, Samadder suggested the disparity could point to underlying systemic biases in how doctors screen for cancer risk variants and how the NCCN criteria were written.
"It is an interesting and concerning finding that the NCCN criteria seem to work less well in those underrepresented minority populations," Samadder said. "The why of this is likely because those criteria were created with European populations that were studied in terms of their personal and family cancer history … that may lead to a systematic bias in the way these guidelines propose genetic evaluation and testing in other populations."
In light of the study's findings, especially the apparent cancer testing disparities among minority patient populations, Samadder and Gay were insistent that the criteria need to be broadened. In an ideal world, they said, they'd like to see everyone receive whole-exome sequencing — a suggestion that most payors reject given the current cost.
But once the cost for whole-exome sequencing decreases further, and more cost-benefit studies show that population screening costs offset the cost of prevented cancer treatments the health system and its payors could come around, Samadder suggested.
"If exome sequencing costs fall within a few hundred dollars, it really can be implemented across the board in the general population," he said. "We already do it with every newborn … for metabolic genetic disorders, so in theory, similar testing could be done at a certain point in your life."
Acknowledging that genomic sequencing testing costs won't drop immediately and cost-benefit studies won't emerge overnight, Samadder suggested it would be valuable to at least broaden NCCN criteria for the time being.
"I think the intermediate step [is], broad genomic testing and genetic testing for germline disorders can be offered for people coming in with a cancer diagnosis," he suggested.
Gay echoed this, saying that if she had the reins right now, she would suggest anyone over the age of 30 who wanted to get this testing could do so. In the meantime, testing criteria could be broadened by loosening family member requirements. As of now, under the NCCN criteria, patients need to have multiple family members with certain cancer types to qualify for testing. "Especially as we see families get smaller and smaller, especially in the United States, that family data isn't necessarily as available or as useful," Gay said. In addition, it could be valuable to broaden the types of cancer diagnoses that would warrant genetic testing.
Other oncologists at AACR pointed out potential challenges to broader germline testing. Chief among them is the fact that there aren't enough genetic counselors.
"Hiring is very difficult. We need genetic counselors. We need them badly," said Andrew Godwin, director of the division of genomics diagnostics at the University of Kansas Medical Center, and the lack of patient counseling prevents expansion of genetic testing. "Even at our cancer center, we should be sequencing every high-grade serous ovarian cancer patient and we don't … we're still underutilizing testing in those individuals."