By Deborah Borfitz

May 11, 2022 | A pioneering program of Geisinger Health System has returned clinically actionable results from exome sequencing to a record-breaking 3,400 individuals at increased risk for potentially life-threatening conditions such as hereditary breast and colon cancers, familial hypercholesterolemia (FH), and heart disease. The MyCode Community Health Initiative, enabled by a 2014 agreement with Regeneron Pharmaceuticals, draws from participants in a sizable DNA biobank who agree to the sequencing and results reporting, according to Adam Buchanan, director of Geisinger’s Genomic Medicine Institute.

As newly named interim director of the Institute’s Genomic Screening and Counseling (GSC) program, Buchanan leads review of the exome sequences for findings where a medical intervention, preventative approach, or early detection is available for patients and their physician to act on, he says. Geisinger launched the broader MyCode biobank under the leadership of David Carey, Ph.D. (director of the Weis Center for Research) 15 years ago to learn about links between genes and disease, and to date it has enrolled more than 300,000 patients of the Danville, Pennsylvania-based health system.

Of those, 185,000 have had their DNA analyzed. The rest either haven’t yet submitted a blood sample or Regeneron has not yet completed exome sequencing on it, says Buchanan.

Relative to the overall Geisinger patient population, the MyCode cohort is a bit older and more racially and ethnically homogenous, as indicated in statistics shared in a 2020 article in Genetics in Medicine (DOI: 10.1038/s41436-020-0876-4). A forthcoming paper will explore the representativeness question in more detail, Buchanan says.

Geisinger hired Kim Drumgo as its first chief diversity, equity, and inclusion officer last fall and the Genomic Medicine Institute will be working with her to better align the MyCode group to the larger patient population that is already predominantly people of European ancestry, he adds. The MyCode cohort skews older partly because children weren’t being enrolled for the first several years and most participants sign up during clinic visits that tend to become more commonplace as people age. 

Approved List

Among healthcare systems, MyCode is one of the biggest biobanks in existence and likely the largest when it comes to the reporting of actionable gene-based findings, Buchanan says. The MyCode GSC program analyzes DNA samples for changes in genes known to increase the risk of developing more than 30 health conditions which, in total, are thought to affect between 2% and 3% of the general population.

These include hereditary breast and ovarian cancer (BRCA1 and BRCA2 genes); Lynch syndrome, which can cause early colon, uterine, and other cancers (MLH1, MSH2, MSH6, and PMS2 genes); and familial hypercholesterolemia (FH), which can cause early heart attacks and strokes (LDLR, APOB, and PCSK9 genes)—all recognized by the Centers for Disease Control and Prevention as having tier 1 evidence for interventions that reduce morbidity and mortality among at-risk individuals.

The GSC program also returns genomic risk results for cardiomyopathies, arrythmias, and other cancers and noncancerous conditions (e.g., malignant hyperthermia), Buchanan says.

The approved list of genetically-linked health conditions and diseases that MyCode participants would be told about if found in their genes is based on published guidance for reporting secondary findings put out by the American College of Medical Genetics and Genomics, he continues. As those recommendations get updated, so will the list used by Geisinger.

Spectrum Of Responses

Buchanan is focused on the genetic counseling portion of the initiative, which is triggered when a clinical diagnostic laboratory confirms a pathogenic variant in a gene of interest detected through exome sequencing. A genetic counselor phones patients and their families to disclose and discuss findings, which also get documented in the electronic health record (EHR), says Buchanan. Their primary care physician also gets notified. 

Although the informed consent process when enrolling in MyCode involves agreeing to be told about any clinically actionable information, he says, about 10% of participants will “passively decline,” mostly due to a move. “We send everyone a packet of information by mail… but sometimes it bounces back.”

With the other 90%, Buchanan continues, “there is a real spectrum of engagement with the information.” On the one end are patients who find the information medically useful and quickly follow through with recommended next steps, including getting other family members tested, undergoing various risk management procedures, and taking preventive steps indicated for their condition. On the other end are individuals who for one reason or another—for example, other health problems of more pressing concern—find the information less valuable or at least not as impactful to their current daily life.

Research Directions

As reported by Google Scholar, Geisinger MyCode has been tapped for more than 200 published articles, including several in the prestigious New England Journal of Medicine. The database tends to be used to better understand either the genetic basis of a disease or its clinical presentation, says Buchanan.

For example, Christopher M. Haggerty, Ph.D., who co-directs the cardiac imaging technology laboratory in Geisinger’s department of imaging science and innovation, is leading a study on pathogenic variants known to be associated with arrhythmogenic right ventricular cardiomyopathy, or ARVC, to see if affected individuals have the expected structural heart problems and arrhythmias.

Researchers from Geisinger are also using data from the MyCode Community Health Initiative to study the role of genetic variation in cancer, including DICER1 syndrome putting individuals at heightened risk for multiple tumor types, often at a relatively young age, Buchanan says. The intention is to expand the list of genes that can be used to guide genetic cancer screening.

In terms of gene discovery, participants in Geisinger’s MyCode Community Health Initiative were also part of a study published last year in Science (DOI: 10.1126/science.abf8683) by the Regeneron Genetics Center that discovered rare mutations in the GPR75 gene were associated with protection against obesity. Regeneron is now actively pursued multiple therapeutic based on the newly discovered genetic target.

A few thousand patients in the MyCode cohort had some pharmacogenomic testing done as part of a study by the Electronic Medical Records and Genomics (eMERGE) Network funded by the National Institutes of Health (NIH), says Buchanan. Separate from MyCode, Geisinger is also developing a clinical program whereby patients over the age of 65 on multiple medications get pharmacogenomic testing done.

At its current size, the biobank has the statistical power to start better understanding links between genes and disease—at least for more commonly occurring conditions and genetic variants, he says. It can also serve as either a validation cohort or be combined with another large biorepository, such as the UK Biobank or biosamples stored by the All of Research Program of the NIH, to detect less common disease-causing variants.

Pragmatic Studies

A study Buchanan leads is currently recruiting for a pragmatic trial that will look at the reaction of families, and their doctor, to adult-onset findings from exome sequencing, he adds. As with similar translational implementation science research, “it is very much a question of what it would look like in clinical practice—would it be acceptable to physicians, would they know what to do with the information... It is very much a ‘how’ series of questions.” That could potentially include a prompt in the EHR when a clinical diagnosis is documented to automate review of the exome data.

In another ongoing pragmatic study, Buchanan says, the focus is instead on improving the diagnosis of FH using both exome sequencing and algorithms that review the EHR looking for clinical features of concern. Part of the investigation will look at how best to communicate findings to relatives using innovative tools such as chatbots.

Another group of Geisinger researchers is working on developing a tool to diagnose genetic disorders in real time. The proposed High Impact Phenotype Identification System is expected to shorten the time between onset of symptoms and discovery of a genetic basis for 13 high-impact medical conditions, among them pediatric epilepsy, heart disease, type 2 diabetes, and congenital kidney disease.

In this case, researchers will create models to identify patients with documented clinical signs and symptoms of the targeted conditions, allowing healthcare providers to immediately screen for and diagnose a genetic basis. The clinical workflow will also be analyzed to determine the best points to present genetic information to providers.

“That study is focused on understanding what conditions would be necessary to have that happen as part of regular care… leveraging what we already have and using that as medical problems arise,” Buchanan explains. Marc Williams, M.D., professor at the Genomic Medicine Institute, is serving as principal investigator for the project.