Pharmacogenetics: From Discovery to Patient Care

Jaekyu Shin; Steven R. Kayser; Taimour Y. Langaee


Am J Health Syst Pharm. 2009;66(7):625-637. 

In This Article

Factors Influencing the Application of Pharmacogenetic Discoveries to Patient Care

Many factors can influence the application of pharmacogenetic discoveries to patient care. These include mechanisms to introduce a pharmacogenetic test into clinical practice, turnaround time, cost, reimbursability, and interpretation of a test.

Test Regulation

There are two mechanisms by which a pharmacogenetic test can be introduced to clinical practice. In the first mechanism, FDA regulates in vitro diagnostic devices (IVDs) or test kits, which manufacturers produce, package, and sell with all ingredients and instructions needed to perform the test. Table 6 lists pharmacogenetic tests that have been approved by the FDA as IVDs for clinical use.[94,95]

In the second mechanism, an individual clinical laboratory develops and offers a test.[95] These so-called "home-brew" tests account for the vast majority of the more than 1300 genetic tests available for clinical use.[96] These tests do not require FDA approval. Instead, the quality of testing in the clinical laboratories is regulated under the Clinical Laboratory Improvement Amendment of 1988 (CLIA). Both the Centers for Medicare and Medicaid Services and the Centers for Disease Control and Prevention are responsible for ensuring the quality of the clinical laboratories.[95] Under CLIA, clinical laboratories performing tests that are classified as moderate to highly complex are required to be enrolled in a proficiency test program in order to maintain a high quality of testing. Although genetic testing, including pharmacogenetic testing, is classified as moderately or highly complex, laboratories performing genetic testing are not currently required to be enrolled in a proficiency test program.[97] Thus, whenever possible, it is important to have the pharmacogenetic tests performed by a reliable and experienced laboratory.

Test Availability, Cost, and Reimbursement

Despite technological improvement in pharmacogenetic tests, which can genotype multiple loci in a short time, test availability limits application of pharmacogenetic discoveries to patient care. A recent survey found that only 8% of U.S. laboratories offer pharmacogenetic testing. Table 7 lists some of the clinical laboratories that offer pharmacogenetic tests for clinical use.[98,99,100,101,102,103,104,105,106,107] Limited test availability also influences its turnaround time for test results. The turnaround time for the results of a pharmacogenetic test performed in an inhouse laboratory may be within a day because an assay itself usually takes only about two to six hours to perform. If, however, pharmacogenetic testing must be conducted by an outside laboratory, turnaround may take several days.[99,100,101,102,103,104,105,106,107] The significance of the turnaround time depends on the purpose for testing. If a test is performed for a drug that should be administered immediately, such as warfarin, the turnaround time is crucial for clinical decision-making. In contrast, if the purpose of testing is to obtain genotype information for future use, a fast turnaround time is not as important.

The price of testing ranges from $250 to $500.[108,109,110] The cost of pharmacogenetic testing required by FDA is generally reimbursed by most insurance plans. The cost of testing not required by FDA may be covered by an insurance plan if the test is considered medically necessary. This usually requires high-quality evidence for the clinical utility of the testing.[95] Currently, few pharmacogenetic tests have evidence to support their clinical utility because many of them have been recently introduced. Thus, most insurance plans consider a vast majority of pharmacogenetic tests "experimental."[95] This lack of high-quality study results and limited reimbursability may delay widespread adoption of pharmacogenetic testing to clinical practice. Interestingly, Medicare's "Coverage with Evidence Development" policy may cover a pharmacogenetic test if a patient has "appropriate" indications for an "experimental" test or if the patient participates in a registry to help develop evidence to support testing.[95]

Test Interpretation

Interpretation of a pharmacogenetic test result is particularly important for a test that influences a dosage of a drug in clinical practice. In its draft guidelines, National Academy of Clinical Biochemistry (NACB) recommends that clinical laboratories should not indicate a specific dosage of a drug in the laboratory report.[111] The package insert of a drug with pharmacogenetic information on the label does not generally provide a specific dosage of the drug for patients with a particular genotype. However, in the case of atomoxetine, FDA recommends that the starting dosage should be based on the patient's phenotype. For example, the recommended starting dosage of atomoxetine hydrochloride is 0.5 mg/kg daily in poor metabolizers of CYP2D6 who weigh 70 kg or less.[112] Some experts have proposed clinical guidelines for the use of CYP2C19/ CYP2D6 polymorphism testing, which provide dosing recommendations for antidepressants and antipsychotics according to CYP2C19/ CYP2D6 genotype.[30]

Given the complex interplay among the many factors that influence drug dosage, determination of an appropriate dosage of a particular drug for a given patient will eventually require knowledge about genetic and nongenetic factors that affect drug disposition and pharmacodynamics. One way to determine a drug dosage with genotype information is to use a dosing algorithm that accounts for genetic and nongenetic factors that cause dose variability of the drug. Although algorithms are useful, clinicians should be aware of advantages of and limitations in using an algorithm, which has been well illustrated for warfarin dosing algorithms.

The warfarin dosing algorithms are essentially a linear regression model that predicts an individualized warfarin dosage based on genetic and nongenetic variables obtained from an individual patient.[113,114,115] While all warfarin dosing algorithms require genotype information from at least three loci (CYP2C9*2, CYP2C9*3, and VKORC1-1639G/A [or its equivalent]), the required nongenetic variables (e.g., age, race, interacting drugs, smoking status, target INR) for dosage calculation vary by algorithm.[48,113,114,115] Despite this, it appears that the predicted warfarin dosages do not statistically differ among the algorithms.[116] The R2 value of the algorithms ranges from 0.4 to 0.7, suggesting that 40-70% of the variability in warfarin dosage is explained by the regression models.[48,113,114,115] When compared with models using only nongenetic variables, the models including both nongenetic and genetic variables had 20-40% higher R2 values, indicating a substantial contribution of genetic variables to warfarin dosage variability.

Other factors should also be considered when a dosing algorithm is used. The dosing algorithms cannot predict who will be outliers from the regression line. In addition, most dosing algorithms may not be useful when adjusting the dosage after warfarin is given. Thus, an individual patient's genotype data should be obtained before warfarin is prescribed. Finally, the algorithms do not predict when a therapeutic INR is reached. Thus, it is still important to closely monitor the INR and to adjust the dosage even when a dosing algorithm is used.

Given the many factors that influence dosage variability among individuals and some limitations in the algorithms, a dosing algorithm using pharmacogenetic discoveries should be viewed as a tool to decrease uncertainty about a patient's dosage in the early phase of the drug treatment; subsequent doses should be adjusted based on the patient's clinical response.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: