It was one of the toughest cases Gholson Lyon had come across — and it was not even his patient. Last year, Lyon, who is a practicing physician but also a research scientist at the Center for Applied Genomics at Children's Hospital of Philadelphia, was heading up a study on Ogden Syndrome, a rare Mendelian disorder associated with a distinct combination of an aged appearance, craniofacial anomalies, global developmental delays, and premature death in infants. The study, published in the American Journal of Human Genetics in June, sequenced the exons from the X-chromosome of an affected boy and four family members. During the course of the study, one of the boy's aunts became pregnant and wished to know whether she was a carrier of the disease. But because Lyon was running a research study, and because the test was not performed in a manner that was US Clinical Laboratory Improvement Amendments certified, he could not share her results with her. The baby was born with the disease, and died three months later.
Lyon eventually developed a CLIA-certified test for the disorder, but it took time. "I was working with the reference lab ARUP and it took months to develop the test. But by that time, she had already given birth and for five months I couldn't tell her," he says.
At Cold Spring Harbor Laboratory's Personal Genomes meeting this fall, Lyon suggested that all exome and whole genome sequencing studies be done in a CLIA-certified manner, so researchers could avoid similar situations. In a July Discovery Medicine paper, Lyon describesd how, during the course of a study on ADHD, he discovered an unrelated genetic finding that explained why one of the research participants had a case of idiopathic hemolytic anemia. Again, Lyon found himself in the same difficult position.
If CLIA-certified tests could be implemented in studies where findings of clinical relevance might be identified, such return-of-results situations could be avoided. But developing CLIA-certified tests is not something that a research group can do on the fly. That is why Lyon is advocating for the implementation of changes that essentially build a return-of-all-results component into the design, regardless of what genetic information is found, regardless of what genetic information is found.
"It's kind of a major mind-shift that myself and others are advocating, because up until now, many researchers take someone's blood and the current consent model basically says, 'We won't give you any results back, we won't tell you anything,'" Lyon says. "Yet I estimate that 50 percent of people who participate don't understand that. They're not fully informed that a researcher may discover some deleterious mutations in their genome that may impact the research participant or future generations in their family."
However, those advocating for open or portable consent might need to be careful what they wish for, says Michelle Lewis, a physician, attorney, and research scholar at the Johns Hopkins Berman Institute of Bioethics. The current workforce of geneticists and genetic counselors is not equipped to adequately address all of these issues, she adds. "Primary care doctors as well are not trained to interpret whole-genome sequence information into clinical care," she says. "So we're going to need to develop resources for both medical personnel and the lay public."
Lyon says human sequence data should be determined in a CLIA-certified way, so that when the community does know more, participants can benefit from down-the-road discoveries. "Right now the government is funding for thousands of exome and genomes to be sequenced, but it's not clear that any of them are being sequenced in a CLIA-certified manner and none of them are being linked to the medical record," Lyon says. "All of that sequencing data is not being optimally utilized, given that there is no regulation or guidance on performing the sequencing in a clinical-grade manner, thus making it very unwise for anyone to return that research information to research participants. Clinicians first and foremost must do no harm, and returning less than clinical grade sequencing results to patients can potentially cause such harm."