NEW YORK – Incorporating RNA sequence data may help reduce variants of uncertain significance and boost the detection of authentic pathogenic variants in individuals from underrepresented populations who undergo germline genetic testing for hereditary cancer susceptibility variants, according to new research by Ambry Genetics investigators.
At the American College of Medical Genetics and Genomics annual meeting on Wednesday, Ambry Genetics senior clinical research specialist Carolyn Horton presented findings from a variant reinterpretation study that involved retrospectively analyzing matched DNA and RNA on data for more than 34,800 individuals who had hereditary cancer predisposition testing with multigene panel tests between the spring of 2019 and April of 2020.
Of those, 29,804 participants were female and 5,041 were male, Horton noted. The participants had self-reported ethnicity from African American, Asian, Hispanic, and non-Hispanic white groups, though non-Hispanic white ethnicity was reported most often, making up almost 27,800 of the cases.
Consistent with past findings, the team found that rates of pathogenic/likely pathogenic variants, benign variants, and VUSs found by panel testing differed depending on participants' self-reported ancestry, with benign variant detection being higher — and VUS rates lower — in those who self-reported non-Hispanic white ethnicity.
Without representatives from diverse ethnicities in genetic studies and reference sets used to interpret them, investigators are left with lower-than-usual levels of evidence for classifying variants, leading to diminished genetic testing utility in the clinic.
"When there's a lack of representation in clinical testing cohorts, published literature, and population reference databases, that limits the evidence base that's available to leverage towards variant classification," Horton explained, adding that this "limits the ability to accurately interpret variants, decreases yield, and increases ambiguity through VUS rate, which ultimately reduces the clinical utility of genetic testing in racial and ethnic minorities."
The team reasoned that additional functional data provided by paired RNA-seq might start filling such ethnicity-related genetic evidence gaps, building on research presented by investigators from Ambry in the past.
In their retrospective reinterpretation study, Horton and her colleagues considered any type of variant reclassification achieved with RNA-seq data. Those included everything from VUS downgrades and classification shifts that did not alter actionability — such as a change from likely pathogenic to pathogenic — to the identification of medically actionable variants.
Based on multigene panel data alone, for example, pathogenic variants had been reported in a smaller proportion of non-white participants, while Hispanic/non-white participants had higher VUS rates than non-Hispanic white individuals.
By bringing in RNA-seq data, the researchers reclassified variants in 4.3 percent of Asian individuals, 1.4 percent of African American individuals, and 1.2 percent of Hispanic individuals, compared to just over 1.1 percent of non-Hispanic white individuals. Those classification changes included VUS downgrades as well as medically significant reclassifications that were considered actionable and may alter the participants' care.
"Our results indicate that RNA sequencing may be of particular benefit to underrepresented populations," Horton and her coauthors noted in an abstract accompanying the ACMG conference presentation, adding that RNA evidence "had a greater impact on [pathogenic variant] and VUS rate in non-white and Hispanic relative to [non-Hispanic white] individuals in our cohort."
Overall variant detection also jumped with the inclusion of RNA data, particularly in participants who self-reported as African American or Asian, as the uncertainty around variant classification diminished.
Ambry currently offers RNA-seq in combination with DNA testing for cancer indications, and on a case-by-case basis for other clinical indications, Horton explained. "When performed concurrently with DNA sequencing on all patients, RNA sequencing generates an additional line of evidence to fill existing gaps."
Even so, the VUS rate remained higher in non-white individuals, suggesting that there are still improvements to be made in variant classification across ethnicities. That, in turn, points to the need for continued efforts to increase genetic test access, Horton said, particularly for ethnic groups that are underrepresented in genetic databases.
"This study only attempted to address gaps in the evidence base. We still need to bolster efforts outside of traditional medical settings to engage communities and improve access to genetic testing overall," she noted. "In order to improve equity in genetic testing, it has to be a multimodal approach."