Cost-effectiveness of Pharmacogenomic Testing: How to Measure the Value of Having the Right Dose of the Right Drug for the Right Patient

Posted on by Lu Shi, Zhuo Chen, W. David Dotson, Katherine Kolor, Scott Grosse, Muin J. Khoury {Office of Genomics and Precision Public Health, National Canter on Birth Defects and Developmental Disabilities, Centers for Disease Control and Prevention, Atlanta, Georgia}

a stethoscope with money, pills with DNA inside them, a crowd and a few people in the inner circle being target, and a doctor talking to a patient in a hospital bedA recent systematic review that assessed the cost-effectiveness of pharmacogenetic testing for drugs with existing guidelines concluded that most studies favored pharmacogenomic testing. The significance of this conclusion must be interpreted with caution and in the context of study factors, such as funding sources, geography, cohort, and the cost-effectiveness comparisons being made.

Pharmacogenomics (PGx) combines the sciences of pharmacology and genomics to assess how to use information about a person’s genetic makeup, or genome, to choose the drugs and drug doses that are likely to work best for that particular person. PGx applications have been characterized for a wide variety of drugs, such as analgesics, antidepressants, anticoagulants, proton pump inhibitors, and cardiovascular medications, that are prescribed to tens of millions of people in the United States. For example, variations in the liver enzyme gene CYP2D6 affect the human body’s processing of a quarter of all prescription drugs.

Realizing population health benefits from emerging PGx applications requires sifting through the available research studies to identify the gene-drug pairs with convincing scientific evidence for improving health outcomes. The Clinical Pharmacogenetics Implementation Consortium (CPIC®) is an international consortium that aims to address barriers to the clinical implementation of tests by creating, curating, and posting guidelines and evidence levels for gene-drug pairs. CPIC guidelines do not address whether PGx testing should be done, instead focusing on how to best use genetic information once it is available. Of 486 gene-drug pairs evaluated to date, CPIC has designated 95 gene-drug pairs as level A final, meaning at least one prescribing action is recommended based on genetic information, if available. CPIC guidelines also do not address the cost or cost-effectiveness of PGx testing. A call for evidence-based cost-benefit analyses for PGx interventions made more than ten years ago is still valid today.

Cost-Effectiveness of PGx Testing

In a recent systematic review conducted to evaluate the evidence on the cost-effectiveness of PGx applications, Morris and her colleagues evaluated 108 articles that addressed CPIC level A drug-gene pairs. PGx interventions that result in sufficient benefit or value compared with costs (i.e., reduction in cost or an incremental cost-effectiveness ratio below the willingness-to-pay threshold adopted for the study setting) were categorized as cost-effective. Morris and colleagues concluded that most of the PGx-guided treatments that were evaluated were cost-effective compared with standard treatment as the cost of PGx testing continued to decline. Of the 108 studies covering 39 drugs, 77 (71%) reported PGx testing as cost-effective; 21 (20%) reported as not cost-effective, and 10 (9%) were uncertain. However, that varied across clinical applications. Cost-effectiveness was reported in 22 of 23 cost-effectiveness analyses of PGx for clopidogrel and 9 of 11 studies of antidepressants but just 7 of 16 warfarin studies and 15 of 26 studies evaluating human leukocyte antigen (HLA) testing.

Study Characteristics: Funding, Geography, and Cohort

Morris and colleagues found that study characteristics may influence cost-effectiveness findings, but no statistical significance could be inferred for the associations because of the small number of studies. Not surprisingly, 13 out of 14 (93%) studies directly funded by pharmaceutical companies supported the cost-effectiveness of PGx testing. In contrast, 64 out of the remaining 94 (68%) studies concluded that PGx testing was cost-effective.  For example, 9 studies of antidepressants directly supported by industry concluded PGx was cost-effective whereas 2 other studies did not reach the same conclusion. For HLA testing, 3 industry studies reported cost-effectiveness compared with 12 of 23 other studies. Among the 23 “other” studies, 1 was funded by a grant from a pharmaceutical company, 10 reported government or institutional support, and 11 did not report funding sources.

Country or continent of origin may also be relevant to the study findings, with studies conducted in Asia less likely to find PGx testing cost-effective. Morris et al. also suggested that studies with hypothetical cohorts were likely to find PGx testing not cost-effective. Out of a total of 75 (69%) of the studies that used a hypothetical population with secondary data for cost-effectiveness modeling, 19  (25%) found PGx testing not cost-effective while only 2 (6%) out of the 33 based on a real-world cohort demonstrated PGx testing not cost-effective

Preemptive and Reactive PGx Testing

The evidence on the impact on the cost-effectiveness of the timing of PGx testing relative to drug initiation was inconclusive. Preemptive PGx testing refers to testing performed before administering a specific drug to predict a patient’s responses to future drug use, while reactive PGx testing is performed in reaction to the prescription. Preemptive PGx testing makes actionable phenotypes available to providers when prescribing. However, a qualitative study reported barriers for payers to develop cohesive reimbursement policies for PGx testing.

Future Direction and Public Health Implications: Are We Ready?

Future cost-effectiveness studies of PGx testing may need to expand their study timeframe to address lifetime costs and different age groups of patients. Studies with cohorts of younger persons may be more likely to demonstrate cost-effectiveness of PGx testing because germline information obtained from PGx testing can be useful lifelong. That presumes that the PGx information is stored in electronic medical records that follow the patient and are routinely accessed by all providers seen by the patient. The high proportion of studies based on hypothetical cohorts may reflect the lack of existing evidence of clinical efficacy and other parameters and the need to collect additional data on clinical parameters and information on cost.

Effective implementation of evidence-based PGx tests is critical to the success of precision medicine. However, while the cost barrier has largely diminished, other barriers persist, including fragmented use of electronic medical records and inadequate reimbursement policies for genetic testing. The uneven adoption and the lack of portability of electronic medical records across the United States hamper the transfer and use of genomic information from PGx testing.

Cost-effectiveness analyses are only as good as the data used to inform modeling assumptions. That includes the evidence base for the relative clinical utility of each drug-PGx pair relative to its comparators (alternatives), including both use of the drug without PGx and use of different medications. It also includes information on how often PGx tests are ordered by providers and the results used to guide prescriptions.

The Centers for Medicare & Medicaid Services has issued local coverage determinations designating PGx testing as reasonable and necessary in limited circumstances. Evidence of future cost-effectiveness analyses, particularly those from the payer’s perspective, may influence other payers to develop reimbursement policies for PGx testing.

Posted on by Lu Shi, Zhuo Chen, W. David Dotson, Katherine Kolor, Scott Grosse, Muin J. Khoury {Office of Genomics and Precision Public Health, National Canter on Birth Defects and Developmental Disabilities, Centers for Disease Control and Prevention, Atlanta, Georgia}Tags ,
Page last reviewed: May 8, 2023
Page last updated: May 8, 2023