Panel Testing Part 2: Actionable Data…again

By: David Shifrin, PhD
Science Writer, Filament Life Science Communications

Continuing on from the previous post broadly describing the issues around genetic panel testing, this post will go into a bit more detail regarding panel tests and the ever-present issue of actionable data and variants of unknown significance.

Researchers and clinicians are pressing forward in determining the best ways to incorporate panel testing into diagnostic and therapeutic workflows. Among the products being used for this work are the TruSight Sequencing Panels from Illumina, which cover a range of genes and disease groupings. Nazneen Rahman, PhD, who works at the Institute of Cancer Research and Royal Marsden Hospital in London, United Kingdom, helped design the panel. Rahman is taking her work ever further than just developing the panels, however. She also runs the Mainstreaming Cancer Genetics Programme in the UK. According to their website, “The goal of this programme is to make genetic testing part of routine cancer patient care.”

With that routine must come actionable data. In the case of Rahman’s work, it appears that the plan is to test for known cancer predisposition genes. Thus, most data obtained from an individual should be useable to help guide lifestyle changes and risk mitigation before diagnosis, improve diagnosis, and potentially guide therapeutic strategies. “In addition, as stated in this article from Medscape discussing Rahman’s work, ‘cancer predisposition genes offer us opportunities to implement risk reduction and cancer prevention in relatives […]’” Thus, this project in the UK by all accounts is focused on using the TruSight panels to look exclusively for known risk factors, and they are bringing together counselors, physicians and patients in order to ensure the data is actionable. Rahman is also clear that, well, her patients are clear on the implications: “’Every single patient we offered this test to wanted it,’ Dr Rahman reported. All patients said they were satisfied to have been tested during an oncology appointment, and 98% said they understood that the results could have implications for themselves and their families.”

Actionability is not always present, of course. Even in looking through material on the MCGP, there is language that seems to hedge a bit, seemingly acknowledging the fact that the goal of fully actionable data is a ways off. Last fall, the New York Times published an article titled “Finding Risks, Not Answers, In Gene Tests.” In it, the authors tell the story of a woman who underwent genetic testing for breast cancer predisposition and ended up accepting testing “for 20 other genes linked to various cancers.” Her results revealed that she carried a mutation linked to stomach cancer, “a troubling result that her doctors have no idea how to interpret.” It was, needless to say, a stressful time for the individual involved. Another interviewee, also with several non-BRCA mutations, describes her waffling on whether or not to receive the results (she ultimately did in hopes it might help her son, even if the impact on her was unclear).

From a clinical standpoint, it does seem that the trend is shifting in favor of increased panel testing. A year ago, investigators at Stanford used a custom risk assessment panel to look at the clinical effectiveness of these tests. Samples came from women referred for BRCA testing. About 10% of the individuals without BRCA mutations had other potentially pathogenic variants among the 42 cancer-related genes screened. These results “warrant[ed] discussion of more intensive screening or prevention. This is a significant yield of potentially actionable results […]”

Additionally, on April 8 of this year, a large, multi-center research group led from the UK published a study covering over 65,000 women in the Journal of the National Cancer Institute. The researchers took a set of 77 breast cancer-associated SNPs and created a “polygenic risk score” for each individual. By using so many SNPs, and interactions between them, the group was able to accurately “stratify breast cancer risk in women both with and without a family history.” Therefore, the study concludes, this type of multi-gene testing is a potentially valuable method for informing women of their cancer risk and guiding prevention and treatment options.

With this apparent trend in mind, the genetics community is working to codify best practices for panel testing. This will be the subject of the next post.