Editor's note: Boston Children's Hospital and St. Jude Children's Research Hospital presented their successes, challenges and startup knowledge as they use pharmacogenomics to examine the role genes play in drug response at the 2017 Annual Leadership Conference.
Pharmacogenomics examines the role genes play in a patient's drug response, helping clinicians prescribe a more effective dose. This could lead to fewer adverse events and lower the cost of care.
By Megan McDonnell Busenbark
A little boy who was otherwise healthy died after surgery in Canada when he was just 2 years old. The New England Journal of Medicine reported this boy's case back in 2009: He received a codeine dose in the recommended range, but an autopsy revealed that morphine (a product of codeine metabolism) built up to toxic levels in his blood and likely depressed his breathing.
Genetic profiling revealed he was an "ultra-rapid codeine metabolizer," meaning his liver converted codeine to morphine in higher than normal amounts. This was due to a genetic variation in an enzyme called CYP2D6—an enzyme that plays a role in the metabolism of numerous drugs. This raises the question: What if this patient's health care team knew about his genetic variation before they prescribed the drug?
Enter pharmacogenomics. An emerging area of genetic testing that some experts call the latest in personalized medicine, pharmacogenomics is the study of how all genes (the genome) influence drug response. "Much like a director in a play or movie, genes give instructions to other parts of the body," says Shannon Manzi, Pharm.D., director of Clinical Pharmacogenomics Services at Boston Children's Hospital.
One of the things genes direct is the production of enzymes that are required to break down or metabolize the drugs a person takes. Enzymes influence how effective a drug might be for someone and how likely he or she is to experience side effects.
And that is one of the benefits pharmacogenomics brings to pediatrics—reducing adverse drug events and improving patient safety. The CDC says one in 250 Americans, including children, seeks emergency care each year due to an adverse drug event, and 27 percent of those visits require hospitalization. Adverse drug events cause 26,500 deaths in children each year in the U.S.
And experts say pharmacogenomics doesn't end with reducing adverse events. "This testing can impact outcomes—for length of stay, the cost of care and getting the patient to an effective dose sooner," says Laura B. Ramsey, Ph.D., assistant professor, Department of Pediatrics, University of Cincinnati College of Medicine, and co-director of Genetics Pharmacology Service at Cincinnati Children's Hospital Medical Center.
The work in action
Manzi and the team at Boston Children's are doubling down on their work—the hospital has two separate arms of Clinical Pharmacogenomics Services. On the research arm, Manzi leads InforMEDKids, a study at Boston Children's in partnership with the Medical College of Wisconsin (MCW) and the Children's Hospital of Wisconsin Research Institute.
The study enrolled 650 patients from the hospital's epilepsy, end-stage renal and inflammatory bowel disease programs and the anticoagulation clinic within cardiology. The goal: Use genetic information to understand children's reactions to medications, ultimately helping clinicians select a more individualized therapy for patients. In the end, this makes medications safer and more effective.
Since 2012, Boston Children's has also offered targeted pharmacogenomic clinical testing at the hospital and is putting results into the electronic medical record. The hospital started an outpatient service in 2014—one of only two in the U.S. at the time.
Today, the outpatient clinic sees children and adults who have experienced adverse drug reactions or have a history of non-response, says Manzi, who saw first-hand the difference this testing can make for patients.
"We have a lot of stories of patients who did well once they were placed on a drug that was more appropriate for them," she says. "Patients and families thank us for explaining what was happening because multiple providers had trouble explaining why they weren't responding to a medication—and in fact, it was in their genes."
About 1,300 miles south of the team in Boston, a clinical trial is taking place at St. Jude Children's Research Hospital in Memphis. More than 3,700 St. Jude patients are enrolled in the trial, called PG4KDS, with each being tested for hundreds of gene variations that might be important for drug use.
The study selectively moves test results for a few genes into the electronic medical record if the evidence strongly shows the result can help in better prescribing of drugs for patients.
The main purpose of the trial: to make pharmacogenetic test results easily available to St. Jude clinicians to inform prescribing—all with the goal of making this a part of standard care, according to Mary V. Relling, Pharm.D., chair of the Pharmaceutical Sciences Department at St. Jude's, and a member of the Steering Committee for the Clinical Pharmacogenetics Implementation Consortium (CPIC).
CPIC is an international organization that creates guidelines designed to help clinicians understand how available genetic test results should be used to optimize drug therapy, rather than whether tests should be ordered.
"At St. Jude's, we're looking at how frequently clinicians use test results to alter prescribing," says Relling. "And how common it is for all our patients to have genetic testing available in their electronic health record before they're prescribed a high-risk pharmacogenetic drug. The idea is to focus on genes and drugs with clear evidence genetic testing should be used to improve prescribing."
Still, it's not all about the genes. "We have to consider age, renal function, liver function and drug interactions," Relling says. "There are a lot of things that account for why medications work well for some patients and not others, and genetics is just one of those things."
So Relling and her team try not to oversell or undersell pharmacogenetics. "It's not going to eliminate all adverse drug effects—and it's not going to make every drug work perfectly for every patient," she says.
But, she adds, it's important to use this information for certain patients, especially as approaches to conducting genetic testing takes a stronger hold in health care.
"As more people have their genomes tested for important genomic variations, the question for clinicians is not whether to order a genetic test—the question is what genetic variations are so important that I should use them in prescribing medications or making decisions in clinic," Relling says.
At Cincinnati Children's, the Genetic Pharmacology Service dates back to 2004. That's when providers started genotyping every patient admitted to the psychiatry unit, and each genotype result came back with a recommendation on the best drug selection and dosing. Since then, about 22,000 patients have been through the testing process.
Today, the team is taking the program to the next level by making the data more accessible to inform drug choice and dosing decisions for patients. "This means getting all this information even more accurately reported into the electronic health record; then make links where providers can—with one click—see immediately what the result and the actionable recommendation are," says Alexander (Sander) A. Vinks, Pharm.D., Ph.D., FCP, co-director of Cincinnati Children's Genetic Pharmacology Service.
In one example of how this works, Vinks and his team at Cincinnati Children's identified voriconazole (an anti-fungal used in critically ill patients after transplants) was the right drug for a patient based on genetics and reduced the number of days to the effective dose from 29 to 6.5.
Currently, the cost of pharmacogenomic testing varies. Generally, a test can run about $1,000. But Manzi says her team runs a 225-gene panel for hundreds of dollars. However, certain tests can run into the thousands. The question is who shoulders the cost? Will insurance cover it?
In some cases, yes. At Cincinnati Children's, for example, the team uses pharmacogenomic testing as a standard of care in psychiatry, and when the hospital bills insurance, the organization has a good track record of getting it covered, Ramsey says.
In other cases, insurance does not cover the cost of testing. While Manzi has succeeded in convincing some insurance companies to authorize the testing, there are instances where they said, "No," and she has sent 20-pound binders full of reasons why the insurers should change their minds. Some insurance companies say pharmacogenomics is experimental and investigational. But medical experts disagree.
"We know it's not investigational, but there are insurance companies that are still willing to go through trial and error and spend two years trying to find the right medication, which can be detrimental to the patient," Manzi says. "If we had known the genetics up front, we may have started a more appropriate dose or drug. Especially if what we're trying to treat can be life threatening or have high morbidity." The costs for testing are coming down to the point where some families are willing to pay out of their own pocket if they have to, she says.
The cost of re-testing
Manzi says now, when providers run a laboratory test, the majority of insurance companies will pay to run it once and interpret it once. But the question is, will they pay for it a second time?
Here's the issue: the tests today are based on the pharmacogenes for which there are current CPIC guidelines. That is a limited number of genes. However, there are "hundreds and hundreds" of other pharmacogenes that clinicians will learn more about over time.
"Once we know more about a patient over time, we may need to re-test him or her," Manzi says. "Sometimes, we can take the data and rerun it through the pipeline, but other times, we have to rerun the test itself. And then, is the DNA available still, or do we have to get a new blood draw? If we do, is that a new charge? Will the insurance company pay for that a second time even though we've got a better mechanism now?" This is a challenge, particularly given the nature of pharmacogenomics, she adds.
"Payment models in the United States—particularly around lifelong results that require interpretation over time—are lacking," she says. "So, we have to come to grips with how we handle genetic results over the lifetime of patients. And it isn't just about pharmacogenes, it's about every gene, including disease states; all those things that may be informed by genetics."
Right mix, right time
Relling says childhood is just the right time to start with pharmacogenomic testing. And there are programs for babies providing genotypes and sequences at birth to see how that affects outcomes. "We could test a child and have genetic information that could be helpful for guiding all prescribing as he or she moves through adolescence, young adulthood, middle adulthood and elderly age—the lifetime of that patient," Relling says.
"People will have baseline genetic testing hopefully as newborns that follows them throughout life," says Manzi. "But the question is, how do you get that from one electronic health record to another? Genetic results are not small. They are unwieldy sometimes, and how do you store that data?"
Other considerations: How do you protect the data if you're going to put it in the cloud? What are the mechanisms for security and making sure that all HIPAA-compliant statutes are followed? There's a lot of work to be done on how providers share those results, Manzi says.
This work begins with identifying one system that can manage the data throughout the patient's life. "We don't have a single electronic health record per patient in the United States, or in any country," she says. "And that's the fundamental problem. We need to provide a uniform, comprehensive lifetime health care record to really make this effort work."
Privacy and the future
Some patients and families have expressed concern about the privacy—and impact—of the test results. While there are some exceptions, currently, patients who undergo genetic testing are protected by the Genetic Information Nondiscrimination Act of 2008 (GINA), which prohibits genetic information discrimination in employment or health insurance based on genetic testing. However, there is no provision for life- or long-term care insurance at this point, Manzi says.
And some pharmacogenomic markers are also disease state indicators, which can lead to a question of ethics and insurability. "Everyone wants to know if they can take a medication or not, but not everyone wants to know, 'I'm 23 years old right now, and I have an increased risk of developing Alzheimer's by the time I'm 50,'" Manzi says. "That is a different conversation, and one that experts in this field debate often."
Groups are split on the ethical debate: What should be mandatory for providers to have to return? What should be decided by the family or the patient?
Better care, cure and costs
Regardless of the challenges, pharmacogenomic testing presents an opportunity to reduce adverse drug events and improve patient care, as well as reduce costs and length of stay, according to Vinks. Not to mention, it creates a more personalized level of care. "There are multiple studies that show doing this type of testing up front will help providers make better choices and individualize care," says Vinks.
A study in the February issue of Clinical Therapeutics examined the potential for medication cost savings for payers and patients with mental illness. If a patient failed to respond to a single psychiatric drug, providers used a pharmacogenomic test to switch or add a new medication. The results showed a cost savings of about $4,000 per member per year.
Still, there is a tremendous need for increased awareness around pharmacogenomic testing: how it works, when it's right to test (or not) and the benefits of testing. "Education is very important," says Vinks. "Pharmacogenomics is complicated, and the world is divided into believers and non-believers. So one of the things we do frequently is give talks to our genetic counselors and our clinical teams as to where pharmacogenomics can make a difference for them."
"It will truly be useful in everyday selection of medications for patients so we can stop some of the trial and error and start at a more appropriate dose or on a more appropriate drug out of the box," Manzi says. "And I hope that decreases the time we spend trying to find the right answer, as well as decreasing severe side effects. That's the whole goal."
Megan McDonnell Busenbark is a writer and founder and principal of Encore Communications LLC, in New Fairfield, Connecticut. Send questions or comments to email@example.com.