Updates to the ASCO Policy on Genetic Testing in Oncology

November 10, 2015


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

More medical technology available for a lower price changes the way we interact with and use that technology. Nowhere is this more evident than in medical genetics and its explosive growth over the past couple of decades. To deal with the rapid expansion of next generation sequencing, the American Society of Clinical Oncology published an update to its policy statement on testing for genetic cancer susceptibility.

Five broad topics are covered in the policy update:

  • How to deal with potential germline-related findings during somatic mutation profiling

  • How to approach multi-gene panel testing in oncology

  • Regulatory and quality issues in genetic testing

  • Medical education and continuing education in oncology

  • Payment for and access to genetic testing services.

Germline Results During Somatic Testing

As with any discussion on managing the entire genetic testing process, a key issue is handling secondary findings and variants of unknown significance. The ASCO statement recognizes the challenges faced by medical providers in light of the volume of data generated by panel and NGS tests, and the declining cost of these technologies leading to increased use. The challenges include a high probability of finding germline mutations through somatic testing. Somatic mutations driving tumor progression are often the target of genetic testing, as results from these tests can guide subsequent care. However, germline mutations may also be revealed in the process, and in the case of whole-genome sequencing, not all of these will be cancer-related. What to do with these, since the implications go beyond immediate decisions about a patient’s tumor? In short, ASCO supports passing along information about unexpected germline variants when they are medically actionable. At the same time, this guideline must be managed in the context of patient preferences, “allow[ing] patients to decline receipt of germline information” if desired.

ASCO also recommends additional research to better understand how patients interact with these secondary or incidental findings, as well as putting resources towards improving the communication of these findings to patients. As is often the case, the general message is to make do with what we have, while recognizing that we don’t know enough and should work towards a deeper understanding of this sensitive intersection between medicine and emotion.

Panel Testing in Oncology

Related to the above issue of unexpected germline findings is the breadth of results that often come from panel testing. While some of the genes tested for are either high-probability candidates or known to confer a significant predisposition to cancer, others may present an intermediate level risk, with uncertain clinical implications.

In general, the statement encourages practitioners to focus on medically actionable genes that are selected based on family or patient history. Beyond that, the organization wants “providers with particular expertise in cancer risk assessment” to help with the decision to use panels with genes that lie outside this relatively narrow window.

In a companion editorial in the Journal of Clinical Oncology, Peter Yu, Julie Vose and Daniel Hayes discuss this issue. One of the potential problems is with bundling candidate susceptibility genes in with panels. The editorial notes that, regardless of how these unconfirmed genes are tested for, the simple fact that their role in cancer remains unclear means that there is a significant risk to testing for variants at all. Maybe the best solution to this problem is to get more data about current variants of unknown significance. Eventually, their role (or lack thereof) in a disease process will hopefully be uncovered and become clinically actionable. That is a long-term process, though, so in the meantime thorough genetic counseling will be critical to educate patients and manage their expectations.
With these challenges as background, ASCO takes a careful approach to panel tests. The group recommends limited testing based on patient history, focusing on “genes of established clinical utility.”

Quality, Regulation, Access and Education

The remaining three sections are relatively straightforward. ASCO supports FDA regulation of high-risk tests, trying to walk the line between protecting patients without crimping innovation (although one could argue that putting those two ideas in opposition with each other creates a false dichotomy).

In the section about access, ASCO simply reaffirms its commitment “to ensuring access to high-quality cancer genetic services […]” and avoiding payment policies that might limit this access.

Finally, the article runs through medical education and notes that medical training for oncology does not by default provide enough with regards to medical genetics in cancer. Thus, ASCO affirms the importance of multidisciplinary education that incorporates genetics into continuing medical education for oncologists, beyond what is currently offered. The statement notes that existing training requirements include baseline “competency in genetic testing for high-risk individuals.” Additionally, changes have been made and, “germline risk assessment is [now] regularly integrated into standard oncology practice”). Physicians need not also be medical geneticists or genetic counselors, but they do need to be able to interact with and incorporate the field. Therefore, more training is required as the field evolves.

These policy recommendations from ASCO don’t break any major new ground. However, they do provide valuable discussion points for the oncology field as the use of genetic testing and NGS expands. Additionally, the reaffirmation of the need for things like evidence-based and innovative regulation, a focus on patient preference, and expanded CME will hopefully provide a productive starting point for any new initiatives.

November 10, 2015


Welcome to the new Concert website! See what else is new: