Inherited mutations in the genes BRCA1 and BRCA2 are a known risk factor for several types of cancer. Women who inherit mutations in these genes have a greater risk of developing breast and ovarian cancers. Men are at a higher risk of developing breast and prostate cancers. And both are at an increased risk of pancreatic cancer.
Testing for BRCA mutations is becoming increasingly common. Not only is this information useful for genetic counseling of family members who may share a mutation, it can also inform the choice of treatment. Tumors driven by these mutations sometimes respond well to drugs called PARP inhibitors.
A new study published on July 10 in Nature finds a complicating wrinkle in this otherwise tidy picture. BRCA mutations in a tumor are sometimes unrelated to its development.
“Not all BRCA mutations are created equal,” says Barry Taylor, Associate Director of the Marie-Josée and Henry R. Kravis Center for Molecular Oncology (CMO) at MSK and a corresponding author on the paper. “In some cases, these mutations are just along for the ride, rather than driving tumor development. That has important therapeutic implications for patients today.”
When a BRCA Mutation Makes a Difference
The BRCA1 and BRCA2 genes provide instructions for making proteins that repair damaged DNA. When these proteins are mutated, cells accumulate more DNA damage, which can lead to cancer. Like all genes, BRCA1 and BRCA2 come in pairs; cells have two copies of each gene.
Typically, a person with a BRCA mutation is eligible for treatment with a drug called a PARP inhibitor. This form of targeted therapy prevents cells from repairing another kind of DNA damage. In cancer cells that are already deficient in DNA repair due to a mutated BRCA gene, the added insult of a PARP inhibitor causes them to die.
But this treatment approach assumes that a BRCA mutation in a tumor always makes a difference. The new results suggest that’s not the case.
Philip Jonsson, a computational biologist at MSK and the paper’s first author, analyzed DNA sequences from 17,152 MSK patients who were diagnosed with one of 55 cancer types. He determined the frequency of germline and somatic mutations in BRCA genes as well as the types of tumors in which these mutations were most frequently found. (Germline mutations are present in every cell in the body, while somatic mutations are present only in cancer cells.) Among tumors with a mutated BRCA gene, he then looked for signs of a mutation in the second copy of the relevant BRCA gene and of DNA repair deficiency. When both copies of a BRCA gene are mutated in a tumor, it’s a sign that DNA repair deficiency is important to its biology.
As expected, both germline and somatic BRCA mutations were most common in those cancers that have historically been thought of as BRCA driven: breast, ovarian, prostate, and pancreatic tumors. These same tumor types were also the ones in which both copies of a BRCA gene were mutated and evidence of DNA repair deficiency was found.
In other highly prevalent cancers — lung cancer and colorectal cancer, for example — germline and somatic BRCA mutations happened but were less common. And when they did occur, they were almost never associated with a mutation in the other gene copy. Nor did these cancers show evidence of DNA repair deficiency. To the investigators, this implied that the BRCA mutations in these tumors were largely incidental.
PARP Inhibitors Not for Everybody
These results have important implications for treatment. They suggest that the presence of a BRCA mutation in a tumor doesn’t necessarily mean the tumor is dependent on it — even in those cancers, such as breast and ovarian, that are thought of as highly BRCA driven. And if that’s the case, then a PARP inhibitor won’t provide any benefit.
“Not everyone who inherits a BRCA1 or BRCA2 mutation from their parents is ultimately going to get a tumor related to that abnormality,” Dr. Taylor says. “Some will, some will not, and that distinction is incredibly important biologically and incredibly important therapeutically.”
Unfortunately, this distinction is not captured by standard genetic testing. “Our results suggest that you can’t rely solely on germline testing to select patients for treatment with drugs like PARP inhibitors that target DNA damage repair,” Dr. Jonsson says. “Doing so would likely result in the inclusion of a significant number of patients with non-BRCA driven tumors.”
Instead, he says, germline information must be integrated with other information to guide optimal treatment — for example, evidence of DNA repair deficiency in a person’s tumor.
Michael Berger, a second corresponding author on the paper and Associate Director of the CMO, thinks that the results won’t necessarily change who is a candidate for genetic testing. But it will influence thinking about who gets what treatment.
“The implications for family members remain the same. They will still need genetic testing to find out if they are at risk. But the same BRCA variant in one person may have a different biological effect in another. Understanding what’s driving each individual person’s cancer can inform what therapies they might respond to or not respond to,” he says.
Physician-scientist David Solit, Director of the CMO, is a third corresponding author on the Nature paper.