Skip to main content

Ambry Genetics +RNA Insight Validation Data Suggests Notable Improvement in Diagnostic Yield

Premium

NEW YORK – Researchers from Ambry Genetics have published important validation data for the targeted RNA-seq addition the company made to its hereditary cancer test services last fall, concluding that obtaining RNA splicing data parallel to DNA enables the identification of pathogenic variants that would have been either missed or classified as inconclusive with DNA testing alone.

Launched in October, and more recently granted approval by the New York State Clinical Laboratory Evaluation Program (CLEP), +RNA Insight tests 18 genes — chosen based on their propensity to harbor the types of splice variants that RNA can either uncover or help better classify — in addition to the company's existing 34-gene CancerNext DNA panel or smaller custom panels if requested.

In the new study, published last week in NPJ Precision Oncology, investigators used the concurrent DNA and RNA sequencing technique on 1,000 samples submitted to the company's lab from patient with suspected hereditary cancer syndromes, concluding that the addition of RNA improved diagnostic yield by 9.1 percent compared to what would have been seen with the firm's previous DNA-only pipeline.

"The results were very promising and they were exciting for us because this is the first time we've ever tried this specific sort of targeted panel approach for RNA genetic testing," Ambry scientist Tyler Landrith, the study's first author, said this week.

"There are other workflows out there that are sort of reflex-based or whole-transcriptome-based. But this is the first time that we've ever tried this sort of parallel, scalable, targeted approach. And to see that it works as we as we hope is exciting," he added.

In previous retrospective research, Ambry has been able to show that RNA analysis performed post-hoc, in cases where a DNA splice variant couldn't be definitively classified as pathogenic or benign on its own, can reclassify up to 90 percent of these uncertain findings. Authors projected based on this data that upfront parallel RNA analysis could impact 1 in 43 individuals.

In the new study, investigators described their initial development of the 18-gene capture RNA-sequencing panel that underlies +RNA insight, which involved analyzing splicing profiles from 345 healthy donor samples to create a reference control dataset against which RNA splicing profiles from future clinical samples could be compared.

As an initial test, the group analyzed RNA data from 25 individuals previously identified as having germline variants already known to affect splicing of genes associated with hereditary breast and ovarian cancer, colorectal cancer predisposition, or hereditary diffuse gastric cancer.

The group calculated a percent splicing index (PSI) for all tested variants, and saw that this was greater than the mean PSI among controls. They then validated the results using a second orthogonal methodology, CloneSeq, which involves sequencing cloned transcripts produced from RT-PCR products, finding that results for the 25 positive controls were concordant between CloneSeq and the +RNA Insight capture method for all clinically significant splicing events.

For the main thrust of the validation, the team then applied their capture RNA-seq method in a clinical context, in parallel with DNA analysis in 1,000 consecutive patients referred to Ambry for clinical inherited cancer predisposition testing.

In total, 84 individuals received positive results (pathogenic or likely pathogenic variant classifications) after performing consecutive DNA sequencing and RNA-seq, and investigators projected that they would have only been able to return those results to 77 individuals if DNA sequencing was performed alone. The effect calculated out to a 9.1 percent relative increase in the diagnostic rate.

Although other hereditary cancer testing companies have instituted policies and pipelines that use RNA sequencing to resolve uncertain DNA splice variants after the fact, Ambry is the first company to institute this testing as an upfront parallel process.

As described last year, the addition of RNA is at no additional cost, and does not alter turnaround time from what is typical for Ambry's existing DNA platform.

The service also doesn't alter substantially the structure or nature of results physicians receive back. The company is still returning results defined as pathogenic, likely pathogenic, variants of unknown significance, likely benign, or benign, and reports them "on the DNA level," an Ambry spokesperson said in an email. It is simply able to classify more results that would have previously been unknown into one or another more definitive category.

In the study, investigators from the firm highlighted two cases from the 1,000-patient pilot that they wrote "exemplify the utility" of +RNA Insight in identifying pathogenic variants in the hereditary breast and ovarian cancer predisposition gene BRCA1.

The first case was a female diagnosed twice with breast cancer and then again with high-grade papillary serous ovarian cancer, who had multiple hereditary cancer tests over the last 18 years — all inconclusive.

Ambry's RNA splicing profile analysis identified an out-of-frame partial retention of intron 2, which was not detected in controls. Analysis of the DNA identified a cytosine-to-guanine substitution nine nucleotides upstream from BRCA1 intron 2, for which in silico data predicted significant consequences.

In the second case, a female proband of Ashkenazi Jewish descent diagnosed with breast cancer at the age of 49 years, testing revealed a skipping of exon 17 that was not seen in controls. In the DNA data, the group was able to find a corresponding BRCA1 variant six nucleotides downstream of exon 17 with downstream effects predicted to result in a non-functional protein.

Other splice-profile positive cases in the cohort included two with variants in ATM that were reclassified from inconclusive to positive based on RNA evidence. The first was detected in a Hispanic female and the second a Caucasian female diagnosed with invasive ductal carcinoma of the breast at the age of 33 years.

Other interesting results included a partial intron 7 retention in the gene MUTYH, in a 31-year-old female with no personal history of cancer. They linked the splice profile to an adenine-to-guanine substitution five nucleotides upstream from exon 8 in the gene, which had been previously identified in the Ambry lab as co-occurring with the founder pathogenic MUTYH variant p.G396D in two patients with severe inherited colorectal polyposis.

According to the study authors, a 9.1 percent relative increase in the diagnostic yield represents a clinically significant improvement, which matches, if not exceeds, the improvement that was seen with the introduction of DNA copy number variation analysis to hereditary cancer genomics pipelines.

Having now more precisely defined that impact, Landrith said, should "help support healthcare providers trying to make a decision about whether or not to implement this testing. It can be some supporting data for them to help make that decision."

Hinting at how the observed diagnostic improvement could in turn affect patient outcomes, the study investigators also wrote that by their calculations, six of the seven cases with RNA-seq-related positive test results would imply a need for substantial changes to medical management based on current guidelines "not only for the probands but also for family members who test positive."

Pathogenic variants in BRCA1 and BRCA2, for example "have implications not only for early detection and risk reduction of several associated cancer types but also have the potential to inform eligibility for PARP inhibitor therapy," the group wrote. "Similarly, early detection and prevention options are recommended for individuals with PMS2 [pathogenic variants, and] for individuals with pathogenic ATM variants, increased breast cancer surveillance is recommended."

According to Ambry, the company did not track if any of these implied changes were actually implemented in the current study. That said, the firm is "actively studying the impact of multigene panel testing on patient management via a pre/post-test survey administered through ordering providers."

As the company continues to test individuals clinically, Ambry should be able to further refine whether there are specific genes or phenotypic sub-groups where its RNA panel seems to offer the greatest benefit in improving the identification of pathogenic splice aberrations that can explain hereditary patterns and inform clinical management for individuals and their extended families.

"In the study itself, most of the genes that we found were in hereditary breast and ovarian cancer related genes … so the results of the pilot study suggest that we'd see a great impact in the hereditary breast and ovarian cancer space," Landrith said.

In an email, an Ambry spokesperson called the +RNA strategy a big differentiator. "The benefits of obtaining both DNA and RNA information in a single test has been very well received by the medical community [not only because of] improved diagnostic yield, but also decrease in inconclusive results."