NEW YORK — Multigene panel germline testing of cancer patients can pick up unexpected pathogenic variants, supporting its wider use in this group, according to a new study presented at the virtual annual conference of the National Society of Genetic Counselors last week.
Genetic counselors have increasingly turned to germline panel testing for cancer patients in the last five years, and some studies have indicated that up to 10 percent of pathogenic variants are missed if multigene panels are not ordered, said Chloe Phillips, a genetic counselor at Allegheny Health Network, during a conference presentation.
Last year, the American Society of Breast Surgeons called for broad germline genetic testing of all newly diagnosed breast cancer patients, but the American College of Medical Genetics and Genomics released a statement urging all patients to undergo assessment for testing. At last year's NSGC meeting, a panel was divided on whether newly diagnosed breast cancer patients should undergo multigene panel testing. Additionally, in October, researchers from the Interrogating Cancer Etiology Using Proactive Genetic Testing (INTERCEPT) study found that more than one in eight cancer patients harbored germline pathogenic variants associated with increased cancer risk.
Philips and her colleagues sought to determine how often panel testing at their clinic revealed unexpected results among cancer patients suspected of having Hereditary Breast and Ovarian Cancer (HBOC) syndrome or Lynch syndrome. For nearly 20 percent of patients, their pathogenic variants did not align with their clinical indicators, they found.
"We are picking up pathogenic variants we definitely do not expect to find in some of our patients," she said during her talk.
For their study, the researchers sought to examine all cancer genetic counseling patients at their site who met National Comprehensive Cancer Network criteria for either HBOC or Lynch syndrome testing. To be included, the patients had to elect to undergo germline multigene panel testing and have at least one pathogenic variant. The panel test had to include high-risk breast cancer genes like ATM, BRCA1 and BRCA2, CHEK2, and TP53, and Lynch syndrome genes like MLH1, MSH2, and MSH6, among others.
Between the beginning of January 2018 and the end of December 2019, 2,633 cancer genetic counseling patients underwent testing and 199 met the study criteria. Of those, 178 were tested due to suspicion of HBOC and 21 for Lynch syndrome.
Of those tested for HBOC, 88 percent had pathogenic variants in HBOC-related genes, most often in CHEK2, BRCA2, ATM, and BRCA1, but 12 percent had pathogenic variants in non-HBOC-related genes such as RAD50, CFTR, and HOXB13.
Additionally, some patients tested for HBOC had two pathogenic variants. For instance, one patient with a history of childhood leukemia and a family history of breast and ovarian cancer had pathogenic variants in CHEK2 and the Lynch syndrome gene MSH6, despite not meeting Lynch syndrome criteria.
Meanwhile, 71 percent of pathogenic variants found among patients tested for Lynch syndrome were in non-Lynch syndrome-related genes like CHEK2, BRCA2, and MUTYH. Twenty-nine percent, though, were in Lynch syndrome genes, most commonly MSH6 and MLH1. Again, the researchers uncovered patients with two pathogenic variants, but none were in Lynch syndrome-related genes.
In all, 18.6 percent of patients had pathogenic variants in genes not associated with the condition for which they met clinical testing criteria.
"This is higher than we might expect based on the data published in previous studies," Phillips said. However, she noted that the finding that 12.3 percent of patients tested for HBOC had non-HBOC pathogenic variants was closer in line with previous findings.
She added that the high proportion of patients tested for Lynch syndrome who had variants in non-Lynch syndrome-related genes could be in part due to the low number of patients who met Lynch syndrome criteria in the study and may not be representative.
But the finding that 47.6 percent of patients tested for Lynch syndrome had HBOC gene variants is not uncommon, she noted, as both syndromes are associated with ovarian cancer.
These findings, Phillips said, support the adoption of broad germline multigene panel testing for patients with suspected inherited disease. She noted that results from such tests will find unexpected pathogenic variants that affect treatment planning as well as preventative screening and surgical recommendations and could provide important information for patients' families.
"There is really not a lot of downside to offering a broad multigene panel testing nowadays," Phillips said. Panel testing does not significantly add to cost or turnaround times, she added, and patients generally do not appear to suffer stress from learning about variants of uncertain significance, which had been a common reason cited for not offering panels previously.
She and her colleagues plan to evaluate next how patients' outcomes may change after learning about these unexpected genetic alterations.