By Monica Heger
While genome sequencing is increasingly demonstrating value in the clinical setting, researchers and clinicians are facing difficult questions when it comes to when and how to return sequencing results to patients — and what kinds of results to return.
Because sequencing technology has progressed well ahead of the regulatory space, there are currently no clear guidelines about how to go about returning results. Ethical questions remain regarding whether and how to provide information on incidental findings in clinical sequencing studies, for example, as well as whether the results from research studies should be made available to participants if the sequencing uncovers deleterious mutations.
As such, researchers in the field have been charting their own courses, sometimes successfully, other times not.
At the Personal Genomes meeting at Cold Spring Harbor Laboratory earlier this month, a number of researchers presented specific cases where they had to grapple with such questions. In some cases, returning results worked to better patients' health, while in others there were complex issues involved in conveying results to a patient's physician. In other cases, researchers had to determine whether to share medically relevant genomic data with patients' next of kin.
Increasingly, "lines between clinicians and investigators, clinical care and research, and individual research participants and biobank studies" are being "blurred," said Ellen Wright Clayton, director of the Center for Biomedical Ethics and Society at Vanderbilt University, during a panel discussion at the meeting.
Sequencing under CLIA
In a presentation at the meeting, Gholson Lyon, a psychiatrist at the University of Utah, presented an example that definitely blurred these lines.
He and his team were investigating a previously undescribed rare, Mendelian disorder that causes rapid aging and eventual death in infants. To identify the genetic cause of the disease, the team sequenced the exons of the X chromosome of a sample from an affected boy, as well as 11 other family members, including the boy's mother. The study was published earlier this year in the American Journal of Human Genetics.
Throughout the course of the study, one of the women that had been sequenced became pregnant with a boy. She wanted to know whether she was a carrier of the disease.
Lyon said his conflicting roles made the situation ethically complicated: he was not the woman's physician; his goal was to perform a research study, but the findings had clinical impact; and he was personally involved with the family.
"I'm not a genetic counselor, but I delivered about ten to 12 hours of counseling to this family," he said.
Ultimately, the "IRB protocol dictated that the results could not be returned" because Lyon and his team had sequenced the woman for research purposes and not in a CLIA-certified setting. Additionally, at the time, no clinically validated test existed for that disease, "so I couldn't tell her that she was a carrier," Lyon said.
The baby was born in March 2011. He was affected with the disease and died in June, Lyon said.
Lyon and his team have since developed a CLIA-certified test for the disease, but he said the case illustrates the challenges of next-gen sequencing for human studies.
As a result, Lyon is calling for all human sequencing to be done in CLIA-certified laboratories, so medically relevant results do not have to be withheld from patients.
Research vs. Clinical Aims
Typically, patients are sequenced not by their physician, but by a researcher who then conveys the results to the physician, who then determines how to act on those results.
This can add yet another layer of complexity, particularly when the researchers doing the sequencing and the clinicians have different goals, Elaine Mardis, co-director of the Genome Institute at Washington University, said at the meeting.
She presented the case of a 60-year old male patient whose esophageal cancer had metastasized to his lung and lymph nodes. He had been referred to the Genome Institute for sequencing to determine his next course of therapy and consented to having results returned. The Wash U team sequenced tissue extracted from the metastatic tumor in the lymph node.
Analysis identified a mutation to the DDR2 gene, which is in close proximity to known cancer hot spots FLT3 and KIT, suggesting it was actionable.
Mutations to FLT3 are typically associated with acute myeloid leukemia, while KIT mutations are common in gastrointestinal tumors, and have also been seen in other cancers like AML and testicular seminoma.
Mardis said the team then reached out to an expert, who said that the best candidate drug for treatment would be Pfizer's Gleevec (imatinib), but "it was still a guess because it had never been tested" on DDR2 mutations. However, certain mutations in the KIT gene are susceptible to the drug.
The Wash U team delivered a case report to the patient's oncologist, detailing its findings. The oncologist switched the patient to imatinib therapy, and his metastatic lesions diminished and his condition improved.
However, Mardis said that it is unclear whether the imatinib therapy was responsible for the patient's improvement or whether the drugs he was previously on finally kicked in.
There wasn't a proper "wash out" period, to give the patient time to flush out the existing drugs from his system, before the second treatment was started. Additionally, the oncologist did not do a CT or PET scan prior to the imatinib treatment to see if the metastatic tumors were starting to show signs of diminishing from the previous therapy.
The case, Mardis said, highlights the sometimes conflicting motivations between clinicians and researchers. While both ultimately have the patient's health as a priority, from a research perspective, Mardis said it would have been helpful to know conclusively whether mutations to DDR2 were predictive of a good Gleevec response.
When results are returned is just as important a consideration as how they are returned, said Jennifer Ivanovich, a genetic counselor at Washington University, during the panel discussion at the conference.
Often, participants agree to research studies, and may even consent to having results returned, but the lag between the study's start and the discovery of medically relevant information can be great.
For example, in cancer or other disease sequencing projects, a volunteer may consent to having his or her genome sequenced and the results returned, but may then pass away before the study has completed.
Communicating genomic test results about a person who has passed away poses unique challenges. For instance, the patient's "reason for participating in the study is gone," she said.
Legally, results should then be shared with the next of kin, which for a married couple would be the spouse. The spouse may be grieving, and delivering test results may bring back that pain, Ivanovich said. While the legal responsibility dictates that the information be communicated to the next of kin, it's important to consider the ethical responsibility as well, she added in a follow-up e-mail, which may be "to the children, siblings, or other family members of the affected proband, whose medical care and family may be impacted."
For example, a Wash U team published results from a sequencing study in the Journal of American Medical Association earlier this year, where they found de novo germline mutations in a woman with aggressive breast cancer that had developed chemotherapy-induced AML (CSN 4/26/2011).
She passed away before the study was complete, but the results have significance for her surviving family members.
Under the university's "moveable firewall" protocol, the team was able to return results to the woman's oncologist.
However, Ivanovich suggests that such approaches need to be more tailored to specific families. "Just as genomic results are of an intensely individualized nature, an individualized approach is needed for each family," she added.
Have topics you'd like to see covered by Clinical Sequencing News? Contact the editor at mheger [at] genomeweb [.] com.