Lack of Utility Evidence Still Hampers Adoption of Pharmacogenomic Testing: Benefits Seen in the Field of Psychiatry

For years it had been hoped that pharmacogenetic (PGx) testing would be a leading application for personalized medicine, but adoption of PGx testing has not lived up to its potential and remains in an early stage of real-world use due to a lack of clinical evidence of utility and low provider awareness.

It is recognized that genotype can impact the efficacy and/or toxicity of medications. The U.S. Food and Drug Administration (FDA) includes PGx considerations on the boxes of nearly 200 medications and several large studies have found that actionable PGx variants are common in the general patient population. Additionally, the FDA has approved PGx tests and many more have been developed as laboratory developed tests.

Yet, despite availability of PGx tests and some evidence for their utility, it remains largely unclear under what circumstances PGx tests should be ordered and how best to address the logistic challenges of incorporating PGx test results into medical records, workflow (preemptive testing versus at the time of drug ordering), and clinical decision making.

A Scan of Commercially Available PGx Testing
In what is described as the first published horizon scan of commercially available PGx tests, researchers from Duke University assessed the range and types of PGx tests offered by laboratories nationally. The results, published in the May issue of Health Affairs, show that available PGx tests vary in type and makeup, which presents a challenge for providers, health system, and insurers, alike.

In order to identify the numbers and types of PGx tests offered by clinical laboratories in the United States from September 2017 to January 2018 the researchers used data from the National Institutes of Health (NIH) Genetic Testing Registry, the McKesson Diagnostics Exchange, the Association for Molecular Pathology’s Test Directory, published literature, and internet searches. They excluded laboratories that offered testing for research purposes only or those offering only companion diagnostics for targeted therapies.

Particular interest was paid to tests that include the 13 genes (CYP2C9, CYP2C19, CYP2D6, CYP3A5, CYP4F2, DPYD, G6PD, HLA-B, IFNL3, SLCO1B1, TPMT, UGT1A1, and VKORC1) included in the Clinical Pharmacogenetics Implementation Consortium’s (CPIC’s) guidelines.

Acknowledging that they likely underestimated the number of laboratories offering PGx testing, the researchers identified 111 clinical labs offering PGx testing, but were only able to confirm the PGx offerings of 76. Of the confirmed laboratories, 31 offered only tests for single genes; 30 offered only tests for multiple genes; and 15 offered both.

Among the 46 laboratories that offered single-gene PGx testing, a total of 219 tests were available for 13 genes ranked as having high evidence (grade A) by CPIC. However, five genes (CYP2C9, CYP2C19, CYP2D6, CYP3A5, and VKORC1) accounted for nearly three-quarters of these tests. None of the laboratories offered individual PGx tests for all thirteen genes, but they offered a mean of five tests each.

In total, 45 laboratories offered 114 multigene panel tests covering 295 genes. Specialty labs, which offered only PGx testing, were the most common providers of multigene PGx panels (nearly 30 percent), followed by genetic testing labs that offered a range of genetic tests (nearly 25 percent). Laboratories offered a mean of three multigene PGx panels (range one to nine panel tests). Panel sizes ranged from 2 to 231 genes, with a mean of 14 genes covered. Most of the panel tests included at least some of the thirteen genes ranked “A” by CPIC. Again, the most commonly included gene was CYP2D6.

Panel testing can theoretically reduce the testing burden and cost on physicians, laboratories, patients, and payers, and may hold value in the realm of preemptive testing, multigene panel tests may present challenges for test comparison, test selection, insurance coverage, and patient counseling.

“With the exception of the gene pairs CYP2C9/VKORC1 for warfarin and CYP2C9/HLA-B for phenytoin (brand name Dilantin), all of the CPIC guidelines focus on a single gene-drug or gene–drug class interaction,” writes coauthor Susanne Haga, from Duke University in Durham, N.C. “The combined effect of variations in multiple genes relevant to a given medication remains largely unknown, though some test developers have suggested that combinatorial PGx testing … may provide a more comprehensive prediction of drug response by using proprietary algorithms to predict drug safety and may reduce medication costs.”

A Case Study for PGx Policy Recommendations
With uncertainty about the clinical utility of PGx testing, particularly for multigene PGx panels, the experiences of health systems that have been early PGx adopters can inform other provider groups.

As an integrated system with a common electronic health record that includes both

laboratory test results and pharmacy records, the Veterans Health Administration (VHA) is uniquely positioned to incorporate and benefit from PGx testing in routine patient care. The VHA’s Clinical Pharmacogenetics Subcommittee is charged with making recommendations for standardizing PGx testing across VHA care facilities. The subcommittee recently published a case study in the June issue of Genetics in Medicine elucidating how the process for reviewing the scientific evidence for and making policy recommendations about routine PGx use.

The subcommittee used an “ACCE model” that evaluates each PGx test by its

• Analytic validity – ability to accurately identify the genotype of interest
• Clinical validity – the likelihood that the test result (genotype) provides information about therapeutic efficacy or toxicity
• Clinical utility – the likelihood that PGx information will lead to a change in clinical management that improves health outcomes
• Ethical, legal, and social implications

After developing these consensus definitions, the subcommittee identified 30 relevant drug–gene pairs (from the NIH’s PharmGKB and CPIC recommendations) that would be potentially applicable in the VHA setting. Companion diagnostics were excluded. Reviewers used databases of clinical and research PGx groups from the government and private sector in order to complete a review template that included the indications for drug use, mechanism of action, pharmacokinetics, and PGx of the medication for each drug-gene pair. The subcommittee discussed each drug-gene pair and classified each test as strongly recommended (13 percent), recommended (40 percent), or not routinely recommended (47 percent) before drug initiation.

The subcommittee acknowledged that consideration of other factors aside from the recommendations, such as feasibility, cost, and patient and provider acceptance, will inform ultimate national VHA policy for PGx testing.

“The subcommittee’s recommendations do not contradict the work of PharmGKB or CPIC but, rather, highlight the need for demonstrated improvements in patient outcomes before large health care systems might broadly implement PGx testing outside of a research context,” writes lead author Jason Vassy, M.D., from the VA Boston Healthcare System in Massachusetts, on behalf of the VHA Clinical Pharmacogenetics Subcommittee. “No one approach to evidence review and policymaking will apply to all health care contexts. … Continual evidence review and rigorous outcomes research will help promote the translation of PGx discovery to health care.”

Application of PGx to Prescribing for Depression
While PGx initially focused heavily on warfarin testing to prevent bleeding for those initiating anticoagulation therapy, there is increasing interest on the application of PGx to aid dosing of antidepressants and other psychiatric agents. FDA’s box list of warnings includes 27 medications commonly used in psychiatric practice to treat mood and anxiety disorders. Two laboratories offering commercially available PGx panels for psychiatric dosing recently presented the results of large clinical trials.

Patients with depression whose medication selection was guided by the GeneSight Psychotropic PGx test saw significant improvements in remission, response and symptoms when treated by both primary care physicians and psychiatrists, according to Myriad Genetics (Salt Lake City). The company recently presented these results from its Individualized Medicine: Pharmacogenetics Assessment and Clinical Treatment (IMPACT) study at the American Society of Clinical Psychopharmacology’s annual meeting (May 29-June 1; Miami, Fla.).

The IMPACT study evaluated the clinical utility of the GeneSight test in selecting medications for 2,025 patients with moderate to severe major depressive disorder.  All patients were assessed using the Beck Depression Inventory at baseline (Day 0) and at follow-up (Week 8-12).

The researchers found that when clinicians used the GeneSight test results to guide medication selection, patients saw a significant 28 percent mean reduction in symptoms.  Additionally, 26 percent of patients responded to treatment and 17 percent achieved remission. Perhaps most importantly, there were significantly greater improvements in outcomes for patients treated by primary care providers versus psychiatrists. (This difference may be explained by the fact that psychiatrists may care for complex cases.)

“Most patients with depression are treated by primary care providers who select antidepressant medications using a trial and error approach,” said Bryan Dechairo, Ph.D., executive vice president of clinical development at Myriad Genetics, in a statement. “The IMPACT study demonstrated the clinical value of the GeneSight test to guide medication selection in the primary care setting.”

The benefits of PGx in the psychiatric care stetting may also include economic savings resulting from decreased utilization of health care, according to a study (funded by Genomind) published May 7 in the journal Depression and Anxiety.

The researchers used Aetna claims data to identify patients with non-newly diagnosed mood disorders who had two or more failed treatments. The intervention group (n=817; assay-guided treatment group) had received the commercially available Genecept assay (Genomind Inc.; King of Prussia, Penn.) from January 2012 through December 2015. The usual treatment group (n=2,745 controls) were propensity matched based on diagnosis, duration of illness, comorbidities, number of prior treatment failures, age, gender, and socioeconomic status.

The researchers found that those who had assay-guided treatment experienced 40 percent fewer emergency room visits and 58 percent fewer inpatient hospitalizations than individuals in the control group during the six-month follow-up period. The two groups did not differ significantly in number of psychotropic medications prescribed or mood disorder–related hospitalizations.

The decrease in utilization of health care services translated to a significant reduction in overall health care costs. Based on claims data, the researchers estimated that the subsequent six-month cost of care was $1,948 lower per individual in the assay-guided treatment group versus the control group, although these costs did not include cost of the PGx test, estimated to be $750 for contracted health plans.

Takeaway: PGx testing remains in an early stage of clinical adoption. Additional outcomes research is necessary to promote the translation of PGx discovery into patient care.