Powerful new analyses could improve breast cancer risk assessment on a massive scale microbiologystudy

Imagine that you have a history of breast cancer in your family, and you want a better idea of what your personal risk is. You consult your physician, and they recommend that you test to see if you have a genetic variant that would increase your chances of developing breast cancer. After weeks of nervously waiting, the results come in and you learn that your variants of BRCA 1 and 2, the genes associated with breast cancer risk, are… of “unknown clinical significance.”

Melissa Cline, a researcher for the UC Santa Cruz Genomics Institute and program manager for the BRCA Exchange platform that consolidates data on BRCA 1 and 2 genetic variants, hears this kind of story all the time. There are roughly 34,000 known variants of the BRCA genes that have been reported in clinical testing according to the NIH ClinVar repository, and over a third of those variants have unclear clinical significance, which means that we do not know what their impact on a person’s risk of breast cancer might be.

“I get emails from women who have found out that they have a variant of unknown significance and they are worried and want to know what I can tell them about their variant,” Cline said.

Before, Cline wasn’t able to tell them much, but now she and her team are rolling out a powerful new feature on the BRCA Exchange that will make risk much easier to assess. In alignment with the Evidence-based Network for the Interpretation of Germline Mutant Alleles (ENIGMA) consortium, they are providing “provisional evidence code assignments,” computationally generated analyses that will help scientists determine whether or not a given variant is pathogenic for thousands of previously unclassified variants.

Although these analyses will be “provisional” until they are able to be reviewed by the ENIGMA ClinGen Variant Curation Expert Panels (VCEPs), the official expert body for BRCA variant curation, the evidence code assignments they provide have been carefully vetted by both VCEP coordinators and BRCA Exchange developers and are unlikely to change after review by biocurators. That means that clinicians will already be able to use many of these designations as guidelines when helping their patients to make decisions about preventative care.

Cline herself is a “previvor,” a term for a person who has an increased genetic risk of breast cancer but has not yet developed any symptoms of the disease. Knowing her risk has allowed her to take precautions like increasing the frequency with which she gets mammograms, and given her more of a sense of control over her health. She is very excited that the BRCA Exchange is now able to share provisional data that will allow more people with previously unknown risk levels to have more clarity about the risk they carry.

New evidence codes follow stringent guidelines

BRCA Exchange is able to roll out the provisional data because the ClinGen Variant Curation Expert Panel (VCEP) for the ENIGMA Consortium just published its variant curation standards in August. The work is published in The American Journal of Human Genetics.

These objective standards are based on more general guidelines from the American College of Medical Genetics (ACMG) and the Association for Molecular Pathology (AMP) that have been used to describe how variants of various genes can be curated as Benign, Likely Benign, Pathogenic, or Likely Pathogenic.

Cline is a member of the ENIGMA VCEP, and says that part of the challenge in analyzing BRCA variants of unknown significance is that the guidelines are very complex to account for a great depth of knowledge on the nuances of the BRCA genes, and therefore difficult to apply.

“We were first prompted to take on this project because the rules for variant assessment are not easy to apply manually,” Cline said. “Even the biocurators were having trouble with them.”

The detailed nature of the standards has made variant assessment a prime candidate for computational analysis to systematically apply them for each of the variants in their database. Cline and her team are in the process of revealing evidence codes that they have computed according to the new rules. Evidence codes are broken down into those that support either a benign (labeled as B) or pathogenic (labeled as P) classification, as well as a ranking of how strong that evidence is, from “very strong” to “supporting.”

The codes fall into several categories, each of which provides a different piece of evidence that points to the likelihood that a variant is pathogenic. Recently, the team has released the evidence codes for the “population frequency” evidence category. The more frequently a variant appears in the human population, the more likely it is to be benign, and Cline estimates that population evidence alone will be strong enough to rule out the possibility of increased breast cancer risk for close to 3,000 BRCA variants.

Later this calendar year, the BRCA team plans to combine the population frequency evidence codes with provisional evidence codes for null variation and computational prediction. These codes analyze genomic evidence to see if a variant leads to a loss of function, which would make the variant likely pathogenic; or conversely, if the variant falls within a class of variants that do not disrupt function and are rarely pathogenic, which would suggest that the variant is benign. The combination of the population, null, and computational prediction evidence codes are expected to provide clarity on tens of thousands of variants.

Although people of all backgrounds will benefit from the clarity that these assessments bring, statistically the impact of reducing the number of variants of unknown significance will be most strongly felt among populations of non-European descent.

This is because historical inequalities in participation in research and in access to testing and treatment have made variants that are most common in non-white populations less likely to be studied, leading to higher percentages of “variants of unknown significance” among populations of non-European ancestry. This disparity in turn means that many individuals of non-European ancestry have fewer options for understanding and managing their inherited cancer risk.

“One exciting possibility is that this work will catalyze variant curation by labs in places in the world that don’t have a white majority population” Cline said. “To this day, people with African, Asian, and Hispanic ancestry have an elevated rate of variants of uncertain significance due to historical disparities in research, and I’m very hopeful that our work here could help lower those disparities.”

To further broaden the reach of the variant assessments, BRCA data will also now be available on a new BRCA Exchange track hub hosted by the UCSC Genome Browser. This hub provides all of the variants and provisional code assignments in a genome context, but with a user interface that many scientists are already accustomed to using. Cline is hoping that this hub will help speed both further research of BRCA variants, and use of the evidence codes by clinicians and genetic counselors.

“We’re anticipating that the data from BRCA Exchange will be used by clinical reference labs to help curate the new BRCA variants that they see in their patients,” Cline said. “This is going to have a huge impact and bring a lot of people peace of mind.”