Providers begin to use genomic testing in mapping patient care
As part of the pre-appointment activities before an annual physical, patients at NorthShore University HealthSystem complete an electronic risk-assessment questionnaire to determine if genomic testing for certain inherited cancers, cardiovascular diseases or other medical conditions is appropriate for them.
The four-hospital health system, located in an affluent suburban area north of Chicago, targets diseases where there are “clinical actions that could be done to improve care,” says Peter Hulick, MD, director of the center for personalized medicine at NorthShore. Examples include breast or colon cancers for which preventive screening is available or lipid disorders for which diet, exercise and statins are routinely prescribed.
Since NorthShore launched the program in March 2017, more than 60,000 patients have completed the risk-assessment questionnaire, and several hundred have had genomic testing, which primary-care providers order via the electronic medical record.
NorthShore is among a small but growing number of health systems that have precision medicine programs incorporating genomic sequencing at the point of care. Daryl Pritchard, senior vice president of science policy at the Personalized Medicine Coalition, an advocacy group, says providers with these programs include both academic medical centers and community health systems. Pritchard says the Personalized Medicine Coalition is conducting survey research to learn more about these precision medicine programs.
Many health systems or academic medical centers started their work with genomic medicine on the research side of the house. But now, “we’re in the phase where the results of that research are in,” Pritchard says. “They’re looking at the value that it has in clinical care and to their bottom line.”
While genetic counseling and testing have been a part of medical care for decades, particularly for neonatal and newborn care and, more recently, breast and other cancers, adding DNA sequencing and analysis to personalized medicine programs that are available broadly has occurred more recently as sequencing prices have dropped, scientific knowledge has increased, and consumer interest has mushroomed.
It took 13 years and $3 billion to sequence the first human genome—a project completed in 2003—compared with $1,000 today. Tests that look at targeted gene panels, such as for breast and ovarian cancer, cost less.
The science also has advanced rapidly. Researchers have uncovered more than 3 million genetic variations, many of which may be involved in either inherited risk for disease or biological reactions to medications, according to the National Institutes of Health.
Meanwhile, the availability of direct-to-consumer tests, such as 23andMe, has fueled patients’ curiosity about their genetic makeup and what it might mean about their inherited risk for diseases.
To respond, health systems have added new offerings, including consumer-driven full genome sequencing, targeted tests for patients with a family history of an inherited disease, and pharmacogenomics, which studies how specific genetic variants impact people’s biological response to medications.
Despite the enthusiasm for such programs, however, challenges to widespread adoption remain, including getting payer reimbursement for testing, scaling programs to reach large numbers of consumers, recruiting physicians and counselors with expertise in genetics, and integrating test results and associated decision-support tools into electronic health records.
Pharmacogenomics is one area in genetic and genomic medicine that is expanding rapidly.
The field looks at how people metabolize medications, move them in and out of the bloodstream and bind them to a target in the body. As Pritchard explains, pharmacogenomics “focuses on what is an effective medicine and what is an effective dose as well as a safe dose.”
So far, pharmacogenomics researchers have produced the most evidence about which genetic variants are associated with how fast or slow people metabolize different medications.
NorthShore launched its pharmacogenomics program in March 2015, according to Henry “Mark” Dunnenberger, the program’s director, who joined the health system in 2014 after completing a pharmacy residency in pharmacogenomics at St. Jude Children’s Research Hospital.
About 3,000 patients have taken NorthShore’s pharmacogenomics testing, which analyzes 20 genes and 120 places within the DNA, Dunnenberger says. Patients pay $225 directly for the testing and are responsible for following up with their insurance plans about possible reimbursement.
Nearly all patients—or 98 percent—have at least one variant in their DNA that corresponds to a specific medication; many have two, three or four variants, he says. The test covers nearly 92 medications—a number that grows as new evidence is discovered—including those for heart disease, depression, pain management, epilepsy, gastrointestinal disorders and other conditions.
Even if a variant is not associated with potentially life-threatening adverse drug events, the information in the pharmacogenomics testing still helps improve clinical outcomes, such as for choosing between a class of depression medications known as selective serotonin reuptake inhibitors, or SSRIs.
This is a promising clinical area because between 30 percent and 50 percent of patients do not respond to the first antidepressant they try, prolonging the amount of time it takes for them to feel better. “We can begin to pick and choose within that drug class,” Dunnenberger says. “That is what makes this really powerful.”
Physicians order the pharmacogenomics test for their patients via the EHR. Patients then watch a four-minute web-based educational video about the test and results. A test kit, involving a cheek swab, is sent to patients’ homes, which they complete and send back to NorthShore. The health system analyzes the sample at its in-house molecular lab, and Dunnenberger and other staff members also review the results. Patients and physicians both receive a PDF report explaining the results in terms of drugs to avoid and drugs to use with caution.
Working with ActX, a vendor of genomic testing and decision support tools, NorthShore integrated the test results into its EHR as discrete variables that are tied to clinical decision support.
For patients with pharmacogenomics test results in the EHR, providers will be able to click on a link to see a list of problematic medications.
Sometimes the issue could lead to a life-threatening drug reaction. In those cases, providers will see a popup alert if they try to order the medication. Blood thinners, such as Coumadin and Plavix, are medicines in this category. If patients metabolize the medications too slowly, they could be at risk for bleeding problems. But if they metabolize the medications too quickly, they could be at risk for blood clots.
Like NorthShore, Inova Health System focuses on blood thinners for one piece of its pharmacogenomics program.
Rapid DNA testing
Plavix, a common anti-platelet drug, isn’t appropriate for everyone. Indeed, it does not work effectively for about 30 percent of patients.
This was a problem for patients admitted to Inova’s cardiac catheterization laboratory, often after a heart attack, because they need to start taking an anti-clotting drug immediately. They do not have the luxury of waiting for results to come back from an outside genomic sequencing lab, which can take days or weeks.
That is why Inova purchased rapid testing technology from Spartan Bioscience, which focuses on DNA variants involved in metabolizing some common medications, including Plavix. The Spartan technology is located inside the cath lab, making it easy to obtain a cheek-swab sample from patients as soon as they arrive for their procedure.
“We put it in the machine, and by the time the cardiologist is done, we know if (patients) should get Plavix or one of the other medications,” says John Deeken, MD, a medical oncologist and COO of Inova’s Translational Medicine Institute, a genomic research lab that is part of Inova’s Center for Personalized Health.
Since Inova began the initiative in 2017 to screen patients before prescribing Plavix, it has tested hundreds of patients, Deeken says.
Inova applies pharmacogenomics in other areas of the health system, too. It offers a broad panel covering several dozen medications for free to all newborns born at one of its hospitals. Since it began the program two years ago, Inova has tested 15,000 newborns.
“We think this should be the standard of care,” Deeken says. “We are trying to change the field, change the mindset, change the way we treat patients by incorporating (genomic medicine) on Day 1,” he says.
The test results are shared with newborns’ parents and pediatrician of record and are added to Inova’s electronic medical record.
As far as Inova executives are aware, none of the babies have had a reason to use the information contained in the reports. But the reports could be useful not only in adulthood but also in childhood for a variety of reasons. One example is codeine, which is sometimes given to treat pain in children, particularly after an injury or surgical procedure. However, if they metabolize codeine too quickly, it could lead to some life-threatening consequences, such as respiratory distress. If, on the other hand, they metabolize codeine too slowly, they might not get the pain relief they need.
Inova also offers a comprehensive pharmacogenomics panel, called MediMap Plus, as well as specialized panels, such as MediMap Mind, MediMap ADHD and MediMap Heart.
Inova analyzes the genetic material at its in-house lab, and then contracts with Translational Software to turn test results into usable information for providers. Translational Software pushes the results into Inova’s EHR as a lab report and provides decision-support to alert physicians when they are treating a patient with a pharmacogenomics profile.
“Having the knowledge base to interpret those results and tell you what they mean and, therefore, what drugs are impacted by those results is sort of the secret sauce,” Deeken says. He also noted that the scientific evidence in pharmacogenomics is changing quickly, and Translational Software revises its product as necessary to keep Inova’s personalized medicine program up-to-date.
But adding precision medicine, including genomics, at the point of care is not easy. Inova, NorthShore and other health systems have had to overcome numerous hurdles before launching these programs.
One issue is how to scale the programs to reach more patients, given the fact that the demand for expertise in genetics and genomics exceeds the supply of practitioners.
In the United States, there are just under 3,000 board-certified physicians in genetics and genomics, but many of those providers are primarily involved in research and not direct patient care, according to Charis Eng, MD, chairwoman of the Genomic Medicine Institute and director of the Center for Personalized Genetic Healthcare at the Cleveland Clinic.
Similarly, there are not nearly enough genetic counselors—4,242 in 2017—to keep pace with rising demand, according to a study published in the February 2018 issue of the Journal of Genetic Counseling. The study’s authors projected that supply and demand for genetic counseling would reach equilibrium sometime between 2024 and 2030.
That is why the Cleveland Clinic is adopting an array of digital strategies to expand the reach of its genetics and genomics experts.
Providers at the Cleveland Clinic’s Center for Personalized Genetic Healthcare, the clinical arm of the Genomic Medicine Institute, log about 6,000 outpatient visits annually. Currently, five physicians, 12 counselors and one pharmacist work at the Center for Personalized Genetic Healthcare, which plans to add three physicians and three genetic counselors to the staff in 2018.
“There are not enough of us. We cannot train fast enough. The supply is tiny and the demand is huge and exponentially growing,” says Eng, who also is a professor and vice chairman of genetics department at the Case Western Reserve University School of Medicine. She was recruited by the Cleveland Clinic in 2005 to launch its Genomic Medicine Institute.
More recently in May, the Cleveland Clinic hired David B. Flannery, MD, as director of telegenetics and digital genetics. Flannery, who had been the medical director of the American College of Genetics and Genomics, is leading the Cleveland Clinic’s efforts to incorporate digital strategies into genetic and genomic medicine.
He is expanding the clinic’s use of telemedicine consults—a service the Center for Personalized Genetic Healthcare launched in 2016. For example, the Cleveland Clinic hired a medical geneticist in Miami who sees patients in Cleveland using “flip” telemedicine. In these cases, patients sitting in an exam room in Cleveland consult with the out-of-town physicians using telemedicine. And a second geneticist, located near Detroit, will begin offering “flip” telemedicine consultations later this year.
The Center for Personalized Genetic Healthcare also is investigating the use of chatbot technology to discuss results with patients who do not have positive findings in their genomic tests. “You don’t need a board certified genetic counselor to call up the negative cases. You can fire off this chatbot to the patient and it will go through the algorithm of post-test genetic counseling,” Eng says.
The Cleveland Clinic is working with a software vendor and hopes to launch the program this year.
Another digital offering in development is an eConsult. The idea is that physicians would request a consultation with a geneticist about a patient’s case rather than simply referring the patient to the Center for Personalized Genetic Healthcare. The clinic wants to create a structured process, which would ensure that a summary of the consult is incorporated into the patient’s electronic record, Eng says.
NorthShore University HealthSystem also has tapped into information technology to expand the reach of genetic and genomic medicine, such as its pharmacogenomics program.
When it launched the pharmacogenomics program initially, patients scheduled an appointment with Dunnenberger to get the testing. “What we found out is that I can see 10 to 15 patients a week and have a waiting list of eight to 12 weeks. That is not really meeting the needs of our patients and providers, so we had to come up with a new way to test patients quicker and still deliver high quality care,” Dunnenberger says.
By integrating the process with the Epic EHR, primary care providers now can order the test on behalf of their patients. Since implementing the revised program in 2017, NorthShore has increased the number of patients tested each month from 40 to 100.
Convincing payers to cover the cost of testing is another challenge.
“Payers are just looking for evidence at this point that this is going to save money; that these tests are valuable; that they are going to be clinically effective and improve patient outcomes,” Pritchard says.
Reimbursement is one reason why NorthShore developed its risk assessment tool. In cases where the tool indicates the potential of an inherited gene variant, payers typically will pay for a limited genomic panel targeted at a specific disease or group of diseases. “Most of our patients—more than 90 percent—are getting this covered,” Hulick says.
If patients want a comprehensive test, which NorthShore calls the Healthy Gene Panel, they pay $475 out-of-pocket. The panel tests for variants in the DNA of more than 130 different genes. Patients get results that cover genomic variants associated with increased risk for certain diseases as well as how they metabolize medications.
Inova also has had mixed results with reimbursement. It pays for its newborn pharmacogenomics testing program internally but has billed payers successfully for the rapid testing of DNA variants associated with metabolizing Plavix.
Patient education is another area ripe for improvement.
NorthShore’s Dunnenberger would like to see health systems provide test results to patients in an interactive fashion to enhance their understanding about how their medications can be customized based on their individual DNA. “I think we have to look at more creative and more innovative ways of delivering results outside of a PDF. A PDF is kind of boring.” The goal would be to improve patients’ knowledge and confidence about their medication regimens, and therefore, the likelihood that they will adhere to them, he says.
In addition to issues with reimbursement, scaling and patient education, health systems also worked though information technology challenges, particularly with data standardization and interoperability among systems.
The information may be reported differently by different practitioners or different institutions. “If the data isn’t inputted in a standard way, the system is going to have trouble integrating all of that,” Pritchard says.
NorthShore has been slogging through these issues as part of its genetic health and wellness assessment program.
NorthShore worked with an external lab to develop a standardized and secure way for NorthShore to transmit data sets, including patients’ demographic and insurance information, securely to the lab, and then for the lab to transmit test results back to NorthShore. The results currently are stored in Epic as a PDF linked to the original order for the test.
NorthShore now is working on refining the system. Hulick wants to build electronic best-practice pathways tied to specific genetic variants to help ensure patients receive the appropriate follow up care, such as more frequent mammograms, gynecological exams or colonoscopies.
Inova’s Deeken also would like to see clinical decision support in genetic and genomic medicine become more prescriptive. “It is one thing to say, ‘Hey, the patient had a test. Go look at the result.’ It is another thing to say, ‘The patient’s test result is (X), and therefore, you should do Y with that drug choice. Hopefully, that will be more robust in the future in terms of what the result is and what you should do with the result.”
And, as NorthShore’s Dunnenberger notes, genetic and genomic medicine is but one source of information to inform patient-care decisions, which may also include patients’ treatment preferences, behavior, socioeconomic factors or other variables. “When I take care of a patient, I don’t have a genetic answer and a clinical answer. I have the best answer to take care of my patient. I have to integrate all of that data together to make an informed decision.”