New Approaches to Cancer Variant Classification: Challenges and Opportunities

Each year, over 1.7 million Americans are diagnosed with cancer and more than half a million will die from their disease. When someone receives a cancer diagnosis, knowing the etiology can be helpful. 

A genetic predisposition to cancer, as compared with a sporadic onset of cancer, can matter to both the cancer patient and to his or her family. If the cancer is hereditary in nature, this can inform questions like: What medications might be appropriate? Should the patient be screened for other cancers? Should family members consider genetic testing and potentially take prophylactic steps to prevent cancer? 

Enter genetic testing and variant classification: the characterization of genes as disease-causing, or “pathogenic” helps to inform clinicians and the patients they care for in navigating cancer care, future risks, and familial testing opportunities.

Current challenges in classifying variants

Genetic testing is one step toward attempting to understand cancer etiology, but classifying the variants identified through testing comes with its own set of challenges. 

Every person inherently has many variants but not all are well-understood and are, therefore, labeled as variants of uncertain significance (VUS). There are often clear guidelines for patients with variants classified as pathogenic, but a VUS finding is less clear. Clinical guidance for patients with a VUS finding is to treat them as if no variants were found at all—just based on their personal and family clinical history. 

Furthermore, most VUS findings will turn out to be benign, but some will ultimately turn out to be pathogenic variants. Thus, a VUS finding may always confer a bit of a “what if” for clinicians. 

Benefit of evidence beyond phenotype

With some genes, such as those associated with Lynch Syndrome, a clear connection between genotype and phenotype exists: these patients get colon or endometrial cancer that has a characteristic features on laboratory testing. 

But not all cancers have this clear genotype-phenotype relationship. For example, breast cancer, which is very common, does not have any signatures that distinguishes a hereditary from a sporadic form of breast cancer. As such, a variant in a breast cancer gene in a patient with breast cancer is inconsequential: it is not evidence that the variant is pathogenic or benign. Therefore, a need exists for some other kind of evidence—besides phenotype—to see if a variant in a gene like BRCA1 or BRCA2 is pathogenic or not. 

In this scenario, functional data is especially important in helping us overcome the challenge of many variants having insufficient data to accurately interpret their clinical impact. Large-scale functional studies are emerging that offer promise to improve variant interpretation and two such large-scale functional studies for BRCA2 were published side-by-side in the January 2025 issue of the journal Nature from researchers at the Mayo Clinic and independently, at the National Cancer Institute. 

Each study documented the characterization of close to 7,000 BRAC2 variants theoretically integrating them into variant interpretation models and demonstrating how these data can help resolve variants of uncertain significance (VUS) and guide better clinical management.  

New variant interpretation methods may benefit marginalized groups

Functional data have another under-appreciated benefit: a variant can be assessed even if that variant has not previously been seen in a patient. 

With the functional data made available by these studies, clinicians do not have to depend on the observation of that same variant in additional people from clinical or general population databases to gather additional variant interpretation data. A functional study can contribute significant information towards variant interpretation, even for ultrarare variants observed only once as “private” mutations. 

Furthermore, functional data can help to reduce the disproportionate number of VUS detected in historically marginalized groups. 

This occurs because for many years, non-white patients did not get genetic testing as often as white patients, and few genomic studies focused on or sought to include these populations. As a result, the genetics of persons with non-white ancestry remain underrepresented in variant databases and, subsequently, small-scale functional studies. As persons from these groups receive genetic testing, large-scale functional data are immediately available for interpreting their variants reducing these disparities. 

Improved classifications provide clearer test results

Expanding our variant interpretation toolbox to improve classification will provide clearer test results and disambiguate management recommendations. While the potential to discover new variants (and be able to classify them) will persist, the use of large-scale functional studies mitigates the rate of VUS among recurrent and novel variants identified across all ethnicities, pointing the way toward ever more personalized approaches to cancer prevention and treatment.