Genetic testing has emerged as a powerful tool in personalized medicine, offering insights into individual health risks, disease predispositions, and treatment options. However, its effectiveness relies on the accuracy of the results. We know that individuals from non-White populations receive less informative genetic testing results compared to Europeans, meaning individuals from non-White populations are more likely to receive inconclusive results called variants of uncertain significance (VUS).  A genetic variant is classified as a VUS if there is not enough evidence to confirm it is either pathogenic (harmful) or benign (harmless). Medical decisions should not be made based on identification of a VUS.

Variants of uncertain significance (VUS) are more frequent in underrepresented racial and ethnic groups across a variety of clinical indications. This can create ambiguity for providers and for patients with regards to disease risk estimations and management recommendations, which may lead to confusion and a loss of trust in results for patients. The difference in accuracy is often due to limited data representation and insufficient understanding of population-specific genetic variations. While there are many socio-political factors that contribute to the disparities in genetic testing overall, there are several ways genetic testing laboratories can play a pivotal role in reducing VUS rates and promoting more accurate results for all. 

Population databases and archives of human variation

Genetic databases, which are an important resource in variant interpretation, are often biased towards individuals of European descent. This is true for population-based databases meant to represent natural variation among generally healthy individuals, as well as archives of variants thought to be causative in individuals with suspected hereditary disease. This bias can lead to misinterpretation of both benign and pathogenic variants. 

For example, a variant identified in a South Asian individual tested due to neurodevelopmental delay may in fact be prevalent in the general South Asian population, but poor representation in the database referenced may cause it to be reported as a VUS rather than benign. This is because one factor considered when determining if a variant is likely to be disease-causing is whether it is rare or it is commonly seen in healthy individuals.  Due to lack of representation, a variant could be mistaken to be rare. 

Conversely, when clinical registries lack diversity, a pathogenic variant that is specific to a sub-population can be difficult to identify. This is because there are too few patients in the registry from this sub-population to capture those who have overlapping phenotype or symptoms and the same population-specific pathogenic variant. This can prevent accurate classification of a variant as pathogenic. 

These examples demonstrate how important it is for laboratories to use the largest and most up-to-date population databases that enable robust stratification by racial and ethnic group. It also provides yet another argument towards the necessity to contribute to variant registries in order to reduce the historical data gaps and provide a more representative archive for variant interpretation.

Population-specific research

Laboratories can also engage in targeted research efforts to identify and understand genetic variations that are unique to specific populations. Much of the existing literature describes disease phenotypes as they present in Europeans. By actively engaging in population-specific initiatives, laboratories can uncover genetic variants and disease associations that may be overlooked in more homogeneous datasets. 

For example, the HOXB13 gene originally emerged as a prostate cancer predisposition gene based on the enrichment of one founder variant among European men with prostate cancer. Years later, another potential pathogenic variant was implicated, this time in men of West African descent. The large size of laboratory cohorts allowed internal case-control comparisons to reproduce and confirm preliminary published findings in what might otherwise be a small, statistically underpowered population. 

Similarly, laboratories must account for differences in the prevalence and penetrance of phenotypes across racial and ethnic groups. The incidence of several cancer types, such as melanoma or gastric cancer, vary greatly by ethnicity and must therefore be weighted differently for the purposes of variant classification in affected probands. 

It is critical for laboratories to prioritize study in underrepresented groups to improve the interpretation of test results within those communities. This approach not only enhances the accuracy of genetic testing but also fosters a deeper understanding of the genetic contribution to health disparities.

Generation of novel functional evidence

Just as population databases and published clinical reports have been historically skewed towards Europeans, the majority of data from functional assays are also derived from variants observed in White individuals. However, there are methods that can help labs generate functional evidence to fill these gaps.

Traditionally, functional assays are built to incorporate variants detected in high-risk individuals ascertained through clinical or research testing. New techniques, called Multiplexed Assays of Variant Effect (MAVEs), utilize high-throughput evaluation of all feasible variants at each codon. In this way, all variants are tested simultaneously, agnostic of racial group or the frequency of the variant in any given population. These assays have been shown to lead to more reclassifications in non-White individuals compared to White individuals. Paired DNA and RNA sequencing has also been shown to preferentially improve variant interpretation in underrepresented populations. 

Both of these methods generate new functional data to help fill existing evidence gaps, driving VUS rates down and positive rates up among individuals absent from existing reports. Investing in technological advancements is another way that labs can better serve their diverse patients. 

These examples of how laboratories can help improve genetic testing performance in minoritized populations are not meant to be exhaustive, nor should we overlook the many other systemic and biologic factors that play a role in health disparities. Rather, our hope is to participate in open dialogue, seek out diverse partnerships, and integrate advancements that improve accuracy of testing in all individuals so that we can play a part in a shared goal of more equitable genetic testing.