SEATTLE (GenomeWeb) – A new analysis from Invitae suggests a significant proportion of informative, pathogenic cancer risk variants may be missed by focusing germline genetic tests in cancer patients on relatively few genes with potential ties to an individual's cancer type.
Invitae's Edward Esplin outlined results from the study during a session on cancer risk assessment and management at the American College of Medical Genetics and Genomics annual meeting here on Friday.
For their study, he and his team focused on more than 113,000 individuals with a personal history of breast or ovarian cancer, colorectal cancer, pancreatic cancer, or prostate cancer, comparing the number of pathogenic or likely pathogenic hereditary cancer gene variants picked up with Invitae's 83 gene test compared to more limited panel testing focused on a few genes per patient.
National Comprehensive Cancer Network's guidelines currently recommend hereditary cancer gene testing for individuals with the cancer types that were included in the study, Esplin explained. But he noted that guidance is less clear on the use of expanded, multi-gene panels, and clinicians often opt for genetic tests that focus on germline variants in relatively small sets of genes.
In an abstract accompanying the talk, he and his co-authors noted that "NCCN guidelines say little about the role of expanded multi-gene panel testing in these patients," though studies done so far "have begun to show that germline testing using expanded multi-gene panels identifies clinically actionable mutations in genes other than those recommended per the patient's cancer type, at a very modest incremental cost of uncovering these additional mutations."
"Genetic test panels that include just a few of the clinically important genes provide incomplete genetic information for patients and their clinicians, both in terms of informing treatment choices and identifying additional health risks," co-author Robert Nussbaum, Invitae's chief medical officer, said in a statement. He contended that guidelines that recommend focusing on a few genes "date to a time when testing was much more expensive, multi-gene panels were not widely available, cancer risks resulting from pathogenic changes in genes other than BRCA1 or BRCA2 were not well described, and cost was prohibitive."
To explore the number of risk variants that might be missed with cancer type-specific gene testing, the team secured institutional review board approval to assess de-identified sequence data across 83 cancer-related genes in 113,107 patients, including 81,861 individuals with breast cancer and 11,124 ovarian cancer cases, as well as others with colorectal, pancreatic, or prostate cancer.
The team's results demonstrated that a significant proportion of individuals with these cancer types carry cancer risk variants that would have been missed by focusing on genes with ties to their cancer type alone — even after adjusting for single copy variants in genes such as MUTYH that involve recessive risk.
Overall, the team identified BRCA1 or BRCA2 risk variants in around 4 percent of patients who had limited testing on those two genes. The BRCA1/2 variants turned up in 3.7 percent of the breast cancer patients, 4.2 percent of individuals with pancreatic cancer, just over 5 percent of prostate cancer cases, and 8.2 percent of those with ovarian cancer.
In the subset of nearly 10,000 patients with colorectal cancer, meanwhile, the researchers noted that testing focused on a handful of Lynch syndrome genes unearthed informative alterations in 9.7 percent of cases.
In contrast, when the team looked at data across 83 cancer-related genes in Invitae's Multi-Cancer next-generation sequencing test, it picked up clinically actionable variants in 9,737 additional patients. The expanded test revealed pathogenic or likely pathogenic variants in almost 12 percent of individuals with breast cancer, for example, and more than 18 percent of ovarian cancer patients.
By combining results of their analysis with insurance claim data on the number of "limited" multi-gene tests reimbursed in the US each year, the researchers estimated that there may be some 7,000 cancer patients with clinically relevant cancer susceptibility variants in their germlines that are missed every year, despite the fact that they receive genetic testing.
The ability to detect additional risk variants may be important for patients and their family members, Esplin noted. For example, the Invitae team argued that these variants may prove useful for selecting patients for targeted pan-cancer trials, including trials of PARP inhibitors.
Likewise, the team noted that patients with pathogenic germline mutations in Lynch syndrome genes may be eligible for checkpoint inhibitor immunotherapy treatments, while the presence of pathogenic and likely pathogenic mutations in the patients in general would inevitably inform future screening programs for the patients and their family members.
Consequently, the authors argued in their ACMG abstract that "genetic testing guidelines should be expanded to include clear recommendations supporting multi-gene panel testing in patients with cancer, to improve the care of patients and their family members."
In a paper published in the American Journal of Human Genetics last year, an independent team led by researchers at the University of Cambridge sequenced the genomes of 460 cancer patients with multiple primary tumors or a family history of multiple primary tumors. Results in that patient population also suggested that expanded analyses of cancer-related genes can pick up pathogenic germline variants that might be missed by more targeted testing, including genes with or without known ties to the cancer types involved.