A new study by several genetic counselors involved in the University of Michigan's Mi-OncoSeq clinical cancer sequencing project has offered a window into the effort's approach to integrating genetic counseling into a large-scale precision oncology and personalized medicine program, with potential lessons for future efforts as the field of clinical sequencing continues to grow.
The authors, who published the study this month in the Journal of Genetic Counseling, wanted to offer a comprehensive look at the role they have had and continue to play in the ongoing Mi-OncoSeq effort, which Michigan began in 2011 and has continued under an $8 million, four-year grant from the National Human Genome Research Institute as one of several sites participating in the agency's Clinical Sequencing Exploratory Research (CSER) consortium.
Jessica Everett, the study's first author and a genetic counselor and clinical instructor at Michigan, told Clinical Sequencing News this week that the paper emerged out of a presentation she and her colleagues gave at the annual meeting of the Society of Genetic Counselors in 2012.
"At that point we were just starting to generate germline data [with Mi-OncoSeq] but one thing we were running into a lot was we were getting calls from other genetic counselors saying 'we think something like this might be coming to our institution, or we might start a clinical sequencing program,' and we were all trying to figure out what role we would have and what we needed to be prepared for," she said.
"So we thought it would be helpful for those out there [who are] thinking about these same issues, because of our involvement with this research study might put us ahead of the curve, for us to just describe our experience."
Everett and her colleagues' paper on their experiences in Mi-OncoSeq demonstrates that genetic counselors have played and continue to play an important role not only in the clinical facilitation of the project, but also in its construction and architecture.
According to the group, genetic counselors have been involved not only in discussing options for return of sequencing results, documenting family histories, and communicating results to patients, but also in reviewing germline sequencing data and participating in the project's multi-disciplinary tumor board, where they provide clinical context for interpretation of germline results and aid in the decision-making process on which findings meet the threshold for disclosure.
"It's been really great to work with this team that wanted to incorporate us from the beginning," Everett said. "I think that's the other emphasis we wanted to make with this paper, that this is something where input from a lot of different disciplines is really very crucial."
Traditionally, genetic counselors have not played much of a role in molecular testing of tumors. But within a clinical next-generation sequencing effort in which sequencing of whole exomes or genomes raises the necessity of dealing with so-called incidental, or unsought findings, it becomes clear where GCs might play a necessary role.
This is especially true, Everett and her colleagues wrote, in the context of recent recommendations by groups like the American College of Medical Genetics and Genomics, outlining a minimum list of alterations that should always be disclosed from clinical sequencing results of germline DNA.
In the paper, Everett and co-authors describe their impact, for example, in developing Mi-OncoSeq's informed consent strategy.
The project, like many others within CSER, has struck a middle path with regards to disclosure of incidental results. Participants are neither forced to receive incidental findings, nor denied all findings outside of those relevant to their cancer.
According to Everett's description, she and her colleagues participated in developing the project's consent strategy, which is based around a default position to disclose all actionable incidental germline findings, which participants can then opt out of if they so desire.
As Mi-OncoSeq has progressed over the last several years, Everett and her colleagues wrote that genetic counselors have also been part of the recruitment team for every patient. After patients are referred to the program, she wrote, they meet with a group that includes a study clinician/oncologist, a coordinator, and a genetic counselor. GCs obtain a four-generation cancer-focused family pedigree as part of this initial meeting, which plays a later role in the interpretation of patents' sequencing results.
The counselor also is responsible for explaining the difference between tumor and germline sequencing, which, Everett wrote, has proved confusing to patients undergoing tumor testing in other studies, according to published research. This is a lead-in to discussion of the possibility of incidental germline findings, and the process of consenting participants to receive or not receive incidental findings based on their choice.
Everett and her colleagues reported in their paper that from the start of Mi-OncoSeq through mid-2013, 167 adult patients from 164 families were sequenced. Only five declined all germline findings, and one declined results with relevance to family members. During the same period, 34 families enrolled in the pediatric arm of the project, Peds-OncoSeq, with only four declining all germline findings.
Everett said an important aspect of the Michigan team's experience, which they hope can be informative to other genetic counselors involved in projects like CSER, was the question of how to toe the line between participating in a research study and their role more broadly as counselors.
Participants enroll in Mi-OncoSeq as subjects in a research effort, a choice that doesn’t necessarily mean they are seeking genetic counseling more broadly. So how a counselor should act when a patient's history indicates something suggestive of inherited risk is somewhat an open question.
"When we presented all this at the NSGC meeting a few years ago, this got the most questions and comments at the microphone because I don’t think there was agreement among everyone on how to handle this," she said.
Everett said she and her colleagues in Mi-OncoSeq decided not to broach the topic of clinical genetic testing during a study visit unless a patient brought it up themselves. Instead they opted to share any indication of potential inherited risk with patients' oncologists, who could then choose to refer the subject back to a counselor.
"We tried to come up with a sort of compromise that we are not using family history as a way to impose genetic counseling on someone who is not seeking it, but at the same time we aren't ignoring it," she explained.
According to the authors' description of Mi-OncoSeq, once patients have been sequenced, counselors play yet another role. The project's protocol limits routine annotation of germline variants to a list of 160 genes in recognized cancer pathways, including all those recently recommended for disclosure by the ACMG.
GCs were involved both in selection of this 160-gene list and in setting parameters for variant calls, Everett and her co-authors wrote.
Based on this protocol, all variants with direct impact on treatment or other clinical decisions related to a patients' cancer are disclosed, including germline findings associated with cancer pathways. And any germline findings not relevant to a patient's cancer are categorized, in order for the project's tumor board to evaluate whether or not they should be returned based on patients' consented preferences.
Then, as a final bookend, counselors are also part of the process of sharing these results in cases where they are disclosed to participants.
As the project continues over the next few years, Everett said the participating genetic counselors are hoping the data emerging may hold answers not only to questions about the efficacy of clinical sequencing and its impact on patients and their families but also to questions about inherited cancer risk.
"There is this number floating around in cancer genetics, that about five to ten percent of people with a cancer diagnosis have some underlying inherited risk factor, so we are really curious what the numbers will look like [from Mi-OncoSeq]," she said.
Dan Robinson, director of clinical sequencing for Mi-OncoSeq, told CSN that to date the project has sequenced about 320 adult, and another 60 pediatric patients.
Not much has changed, he said, in terms of sequencing strategy, although the center did adopt technological updates in the last year for more rapid Illumina sequencing runs on the three HiSeq 2500s and one HiSeq 2000 that are available to the project.
A full evaluation of the impact of the effort in directing personalized therapy will have to wait till the end of the CSER consortium project. But in the meantime, Robinson said that the data already collected suggests that Michigan's approach has had some clear clinical impact.
"In the overwhelming majority of cases with a successful biopsy we get informative results — something that informs as to the genetic basis of the tumor," he said.
"Whether that is actionable depends on your definition. For example, you could say a KRAS mutation is not actionable, because there is no therapeutic out there. If we leave out things like that, we've [still found that] the majority of cases have an actionable mutation."
Unfortunately, he explained, issues of eligibility and other logistical realities — including the advanced disease state of most of the patients who are sequenced through Mi-OncoSeq — mean that such actionability does not always translate to action.
"Many trials are still focused on tumor site rather than molecular aberration, so right now there is a still a little bit of a mismatch in the world between our molecular view of suitable trials and the traditional view by organ or tumor site," he said.
That said, Robinson added, the project has resulted in real discoveries and real treatment success stories.
"We are very optimistic that this currently has utility today and that that utility is only going to increase in the future as trials are reorganized according to molecular data and as more targeted therapeutics come into play," he said.