NEW YORK (GenomeWeb) – Multi-gene germline testing of breast cancer patients doubles the detection of disease-causing variants that could result in a change of care, compared to BRCA1- and BRCA2-only testing, though it does not increase prophylactic breast removal, according to a large new retrospective population-based study.
However, the study, published today in JAMA Oncology by researchers at Stanford University, the University of Michigan, and elsewhere, found that multi-gene sequencing resulted in a higher rate of variants of unknown significance, particularly in racial minorities, and was often delayed until after surgery, thus reducing the potential impact of the results. Â
The researchers, led by corresponding author Allison Kurian of Stanford University School of Medicine, set out to study the impact of comprehensive breast cancer germline genetic testing on patients. That type of testing became available after the US Supreme Court struck down Myriad's BRCA patents.
They focused on participants in the iCanCare study, which includes women recently diagnosed with stage-0 to stage-2 breast cancer who are part of the Surveillance, Epidemiology, and End Results (SEER) registries in Georgia or Los Angeles County. Between July 2013 and August 2015, they sent surveys to 7,810 patients, of whom 70 percent responded.
Complete data was available for 5,026 patients — a racially diverse group that included 51 percent white, 18 percent black, 19 percent Hispanic, 9 percent Asian, and 2.5 percent unknown-ethnicity patients.
About a quarter of these, or 1,316, had a hereditary cancer genetic test result, 45 percent from a multi-gene sequencing test and 55 percent from a BRCA1/2-only test. Testing was performed by four laboratories — Ambry Genetics, GeneDx, Invitae, and Myriad Genetics — and the results were merged with the SEER clinical data and survey responses.
Over the course of the study, testing switched from BRCA1/2-only to multi-gene panels: At the beginning, multi-gene sequencing accounted for 25.6 percent, and BRCA1/2-testing for 74.4 percent of tests, which switched to 66.5 percent for multi-gene panels and 33.5 percent for BRCA1/2-testing at the end.
Surgeons were equally likely to order either type of test, whereas medical oncologists favored BRCA1/2-tests and genetic counselors were more likely to order a multi-gene sequencing test. Test results were most often discussed with patients by genetic counselors, in particular if they were from a multi-gene test. An earlier analysis of the survey results, published last year, found that many patients undergoing genetic testing did not see a genetic counselor.
Most patients received testing after their diagnosis but before surgery. But those who received a multi-gene test were more likely to have their test after surgery than those who got a BRCA1/2-test. This, the researchers wrote, might limit the impact of multi-gene testing because results obtained after surgery cannot be used to make surgical choices. The reason for the delay of a multi-gene test "may reflect clinicians' recognition of its greater complexity and need for genetics expertise," they wrote, which clinicians do not or cannot receive quickly enough.
Pathogenic variants for which guidelines advise a change in care were detected twice as often in patients who had multi-gene sequencing than in those who only had BRCA1 and BRCA2 analyzed. Those variants occurred in a variety of genes, including CHEK2, ATM, PALB2, APC, BRIP1, PMS2, and RAD51C.Â
However, VUS were also ten times more frequent in multi-gene-testers — about 30 percent received one — than in BRCA1/2-only testers, who only had a 3 percent VUS rate. This was especially true for certain groups: 51 percent of Asian patients tested with a multi-gene panel, for example, had a VUS, and 45 percent of black patients. "The cause of this racial VUS gap is more social than biological," the researchers wrote, as more testing has been performed in white patients in the past. However, the VUS gap has already been narrowed for BRCA1/2 over the last two decades, they added, and is likely to be narrowed for the other genes, too, as more multi-gene testing is performed in other racial and ethnic groups.
Patients with a VUS, or with a pathogenic variant in a gene other than BRCA1/2, did not opt for prophylactic mastectomy more often than those with a negative test result: While 79 percent of patients with a pathogenic BRCA1/2 variant had their breasts removed, about a third each of the other groups decided to do so. "[O]ur observation that the rate of prophylactic mastectomy was no higher for those with VUS vs negative results is reassuring," the researchers wrote, especially since previous studies, including their own, had found that clinicians may misinterpret VUS as an indication for prophylactic mastectomy. Â
"Our findings demonstrate that sequencing more genes offers the potential advantage of more clinically valuable information and also pinpoints gaps that must be bridged to realize that potential," the authors concluded.