Purpose The development and implementation of a pharmacist-managed clinical pharmacogenetics service are described.
Summary A pharmacist-managed clinical pharmacogenetics service was designed and implemented at an academic specialty hospital to provide clinical pharmacogenetic testing for gene products important to the pharmacodynamics of medications used in the hospital’s patients. A series of accredited educational seminars were conducted for our pharmacists to establish competencies in providing pharmacogenetic consults for the genes to be tested by the clinical pharmacogenetics service. The service was modeled after and integrated with an already-established clinical pharmacokinetics service. A steering committee was formed to evaluate the use of available tests, new evidence for implementation of additional tests, and other service quality metrics. All clinical pharmacogenetic test results are first reported to one of the pharmacists, who reviews the result and provides a written consultation. The consultation includes an interpretation of the result and recommendations for any indicated changes to therapy. In 2009, 136 clinical pharmacogenetic tests were performed. The service has been met with positive clinician feedback. The successful implementation of this service highlights the leadership role that pharmacists can take in moving pharmacogenetics from research to patient care.
Conclusion The development of and experience with a pharmacist-managed clinical pharmacogenetics service are described. The program’s success has depended on collaboration between the clinical laboratory and pharmacists, and pharmacists’ pharmacogenetic recommendations have been well accepted by prescribers.
Therapeutic drug monitoring (TDM) involves the measurement of drug concentrations in biological fluid and the interpretation of those concentrations in order to individualize therapy. TDM applies the principles of both pharmacokinetics and pharmacodynamics. Traditional TDM requires close collaboration between the clinical laboratory and the pharmacist to ensure appropriate sample collection and timely interpretation of results.1 The benefits of pharmacist-directed TDM through a clinical pharmacokinetics service have been well documented and include decreased adverse effects, which translates into decreased length of treatment, decreased length of hospital stay, and reduced costs.2–4
Pharmacist-managed clinical pharmacokinetics services are now common in most hospitals, and the American Society of Health-System Pharmacists (ASHP) has stated that clinical pharmacokinetic monitoring is a “fundamental responsibility of all pharmacists.”5 Results of a 2008 ASHP survey showed that pharmacists at 83.7% of hospitals surveyed provided clinical pharmacokinetics consultations, and 97.2% of respondents agreed that pharmacokinetic dosage adjustment is an essential pharmacy service.6 Reports in the literature have demonstrated clear clinical and economic benefits of these pharmacist-managed services.2–4
Pharmacogenetics is the study of genetic factors that influence the variability in drug response among patients. Pharmacogenetics integrates knowledge of pharmacokinetics and pharmacodynamics with modern advances in genetic testing. Of particular relevance are polymorphisms in genes encoding drug-metabolizing enzymes, drug transporters, and drug targets and the effects of these polymorphisms on drug efficacy and toxicity in individual patients.7 Over the past decade, pharmacogenetics has been widely incorporated into pharmacologic research and drug- development initiatives.8 The potential role for pharmacogenetics in reducing adverse drug reactions (ADRs) was highlighted in a systematic review, which found that the disposition of the majority of 27 drugs frequently involved in ADRs is influenced by genetic factors, suggesting that many ADRs may be prevented by individualizing drug therapy based on patients’ genetic profiles.9
In 2001, Ensom et al.10 predicted that TDM of the future would include pharmacogenetics-oriented TDM and that pharmacists would play a major role in interpreting the results of these new laboratory-based tests. In recent years, the Food and Drug Administration (FDA) approved the incorporation of genetic information to guide dosing in product labeling for specific drugs, such as mercaptopurine, irinotecan, cetuximab, trastuzumab, abacavir, clopidogrel, and warfarin.11,12 The incorporation of pharmacogenetics into product labeling, combined with the increasing availability of genotyping tests from reference laboratories certified by the Clinical Laboratory Improvement Amendments of 1988, has facilitated the transition of pharmacogenetics from a research endeavor to a clinical practice.13,14
Pharmacogenetics-based testing can help determine which individuals may benefit from a specific drug and the most appropriate drug dosage.15 Unlike traditional TDM, which is not performed until after a drug is administered, pharmacogenetics-oriented TDM can be conducted even before treatment begins, decreasing the trial-and-error period or eliminating certain drugs from consideration.10 Optimally, clinical pharmacogenetics and TDM can play complementary roles in managing patients’ drug therapy to achieve positive outcomes. The following factors may be considered in assessing the importance of clinical pharmacogenetic testing, as suggested by Phillips et al.9: (1) widely used drugs have a high rate of associated toxicity (i.e., narrow therapeutic range), (2) prevalence of variant alleles of relevant gene is high enough to warrant use of genetic information, (3) severe toxicity is associated with use, (4) current methods for monitoring response or evaluating toxicity are inefficient or inadequate, (5) sufficient evidence exists for a relationship between genetic factors and response or patient outcomes, (6) an assay is available that can rapidly, reliably, and inexpensively detect the variant alleles, and (7) clinicians are able to interpret the results and use the information.
Several studies have suggested clinical pharmacogenetic testing can result in cost reduction.15 Screening of patients treated with thiopurines for thiopurine methyltransferase (TPMT) polymorphisms has been shown to be cost-effective in the treatment of several diseases.16–18 Other research has shown genotyping for the gene encoding for cytochrome P-450 (CYP) isoenzyme 2C19 (CYP2C19) to be cost-effective when incorporated into treatment decisions for Helicobacter pylori infection,19 HLA-B*5701 genotyping to be cost-effective in preventing abacavir hypersensitivity in patients with human immunodeficiency virus (HIV),20 and HER2 testing to be cost-effective when incorporated into treatment decisions for metastatic breast cancer.21
Research studies conducted over the past two decades have shown that TPMT-deficient patients treated with conventional doses of thiopurines are predisposed to drug-induced complications due to accumulation of excessive intracellular concentrations of thioguanine nucleotide metabolites.22 Clinical interest in TPMT pharmacogenetics is based on studies showing that the TPMT genotype or phenotype can identify patients at high risk of hematopoietic toxicity after thiopurine therapy.23,24 Moreover, studies have shown that all patients with at least one TPMT-variant allele may respond better to thiopurine therapy compared with patients who have two wild-type TPMT alleles, with no difference in long-term treatment efficacy.25–27
With these findings firmly established in the research setting, we sought to move the use of genotype testing for relevant genes, such as TPMT, into the clinical care of our patients. This report describes the development and implementation of a pharmacist-managed clinical pharmacogenetics service at an academic pediatrics hospital. The review of patient charts was approved by the hospital’s institutional review board.
St. Jude Children’s Research Hospital is a 78-bed inpatient facility that also has nearly 60,000 outpatient visits per year. Since the early 1980s, the hospital’s pharmaceutical department has operated both the clinical and laboratory aspects of a clinical pharmacokinetics service, which provides TDM for a variety of drugs administered to patients with catastrophic illnesses (e.g., cancer, HIV) in a clinical research setting. The pharmaceutical department employs 10 advanced-practitioner pharmacists with postgraduate residency training and substantial professional experience who spend the majority of their time focused on clinical activities. These pharmacists practice as part of multidisciplinary inpatient and outpatient care teams and also are involved in clinical research.
TDM samples are usually ordered by pharmacists through collaborative practice agreements, and samples are subsequently assayed by one of five medical technologists in the clinical pharmacokinetics laboratory. All TDM results are first reported to one of our pharmacists, who reviews and evaluates the results to provide a consultation documented within the computerized medical record regarding any indicated changes to drug therapy. A pharmacist provides a consultation for every concentration or series of concentrations for a course of therapy. In 2009, our clinical pharmacokinetics laboratory processed 6405 clinical TDM specimens, and our pharmacists conducted 4205 clinical pharmacokinetics consultations involving a wide variety of drugs.
The clinical pharmacogenetics service was designed to integrate with and mirror the functions of our clinical pharmacokinetics service. Any member of the hospital’s clinical staff may order these tests when clinically indicated. As with TDM samples, pharmacists may order pharmacogenetic tests under a collaborative practice agreement. The genotyping tests offered through the clinical pharmacogenetics service are clinical tests and do not require special consent to be obtained. All clinical pharmacogenetic test results are first reported to a pharmacist who reviews the result and provides a written consultation, which includes an interpretation of the test result and recommendations for any changes to therapy.
Program development and implementation
In 2005, the initial steps in implementing a clinical pharmacogenetics service were taken. We began by communicating the usefulness of clinical pharmacogenetic testing to pharmacists and other clinicians involved in direct patient care. The possibility of offering this service was discussed at weekly multidisciplinary clinical research program meetings and clinical pharmacy conferences. Enthusiastic support for the service was provided by clinicians throughout the institution.
Using the hospital’s formulary, we identified drugs that are metabolized by polymorphic enzymes with commercially available genotyping tests and drugs meeting the criteria previously described by Phillips et al.9 that are used in our patient population. Based on these criteria, initial efforts were focused on two clinical pharmacogenetic tests: those for TPMT and the uridine glucuronosyltransferase 1A1 gene (UGT1A1). TPMT catalyzes the S-methylation of azathioprine, mercaptopurine, and thioguanine,28 all of which are inactive prodrugs that require metabolic conversion to thioguanine nucleotides or inactivation via TPMT. UGT1A1 is a polymorphic gene involved in the inactivation pathway of irinotecan, a camptothecin analogue used to treat colorectal cancer. This polymorphism is also associated with Gilbert syndrome, a mild form of indirect hyperbilirubinemia. In 2007, we expanded the service to include a third test for the CYP2D6 gene (CYP2D6). CYP2D6 is a polymorphic gene involved in the transformation of up to 25% of clinically useful drugs, including codeine, tamoxifen, antiar-rhythmics, neuroleptics, and tricyclic antidepressants (Table 1).
Over three months in 2005, our faculty conducted a series of educational seminars, accredited by the Accreditation Council for Pharmacy Education, for our pharmacists to establish competencies in providing pharmacogenetic consultation for the genes to be tested by the clinical pharmacogenetics service. Understanding of the material was assessed via examination. The pharmacists most likely to use pharmacogenetics in their practice took advantage of additional self-study and collaboration with other clinicians. Laboratory staff members were trained on the proper handling of pharmacogenetic samples and results reporting. Test names and written consultation templates were created in our institution’s electronic medical record system for each of the genes to be measured. The templates facilitate patient-specific written consultations by pharmacists for each pharmacogenetic test result. Each consultation contains relevant test data, patient-specific information, and the pharmacist’s clinical assessment. As part of the electronic medical record, the completed consultation is available electronically to all clinicians.
A cost analysis was performed to determine the feasibility of conducting the selected genotyping tests inhouse. Based on our institution’s anticipated usage (less than 100 tests per year for each test) and associated costs, we chose to send our samples for genotyping to an outside laboratory that offers pharmacogenetic testing services. Insurance reimbursement was established according to the American Medical Association’s Current Procedural Terminology (CPT) codes for each test; however, no insurance reimbursement was established for the pharmacist’s consultations.
The mission statement of the clinical pharmacokinetics laboratory was broadened to include pharmacogenetic testing: “Provide state of the art therapeutic drug monitoring and pharmacogenetic testing that will be interpreted by pharmacists to assure optimal drug dosing.” A new departmental policy was created to provide direction to pharmacists and laboratory staff involved in the ordering and reporting of clinical pharmacogenetic test results. The policy lists the genes available for testing and the relevant drugs metabolized by each polymorphic enzyme and mandates that a pharmacy consultation accompany all clinical pharmacogenetic test results in the electronic medical record. Each test requires 2–5 mL of whole blood, which is sent overnight to the testing laboratory where DNA is extracted for the genotyping tests and the genotyping test is performed using a polymerase chain-reaction method. Results are then transmitted to the clinical pharmacokinetics laboratory.
The availability of this new service was communicated to our medical staff. Because many of our patients are treated on clinical trials incorporating pharmacogenetics research objectives, our medical staff is familiar with the concept of pharmacogenetics and overall has been accepting of the services provided by the clinical pharmacogenetics service. Information about the available tests and pharmacist services was placed on the clinical pharmacokinetics laboratory’s intranet site, which is accessible by all hospital employees.
To assess the use and relevance of the pharmacogenetic tests and services offered, a steering committee was formed, consisting of department leaders and the clinical pharmacokinetics laboratory director. The committee meets to evaluate the use of available tests, new evidence for implementation of additional tests, and other service quality metrics (e.g., test turnaround times).
Experience with the program
Implementation of the clinical pharmacogenetics service occurred in fall 2005, offering tests for TPMT and UGT1A1. In 2007, a genotyping test for CYP2D6 was added. Table 2 summarizes the number of each type of pharmacogenetic test ordered by year. In 2009, 136 clinical pharmacogenetic tests were performed, consisting of 66 TPMT tests, 65 CYP2D6 tests, and 5 UGT1A1 tests. Each clinical pharmacogenetics pharmacy consultation requires between 20 minutes and two hours of a pharmacist’s time, depending on the complexity of the result and the consultation. The median turnaround time for these tests has been 11 days (range, 4–25 days). The relatively long turnaround time associated with sending tests to a reference laboratory could mean that the results might not be available swiftly enough to be useful to avoid toxicity with early doses of the monitored drug and to maximize efficacy. For this reason, as much as possible, we have implemented preemptive pharmacogenetic testing of TPMT and CYP2D6 genotyping, beginning with patients on our frontline treatment protocol for acute lymphoblastic leukemia, a population who will receive thiopurines and is at risk for pain control issues. With preemptive testing, a blood sample is obtained from a patient who is likely to need a medication. The genotyping results can be generated up-front (i.e., soon after a patient begins treatment and before the patient is expected to require the pharmacogenetically monitored drug) and could then be available to the clinician when he or she decides whether the drug should be prescribed and what dosage should be given. The proactive use of pharmacogenetic tests is one of the ultimate goals for integration of pharmacogenetics into clinical care.
One challenge in implementing a clinical pharmacogenetics service is the speed with which the science and information advance. The clinical pharmacogenetics service steering committee has proved valuable in reviewing the tests offered on a regular basis to be certain the tests remain relevant and useful. Requests from clinicians to implement new pharmacogenetic tests are evaluated by this committee. In one such use, the steering committee evaluated the evidence for UGT1A1 genotyping. In the intervening time since UGT1A1 testing was made available by our service, clinical findings emerged to indicate that the UGT1A1 genotype does not correlate closely with irinotecan toxicity in pediatric patients with cancer receiving the schedule of irinotecan used at our institution.29 After evaluating the clinical data, the steering committee elected to continue offering UGT1A1 testing but not to advocate its use for all patients receiving irinotecan.
Another potential obstacle to incorporating pharmacogenetics into clinical practice is the issue that many practicing health care professionals have not had formal training in the field of pharmacogenetics.9 For pharmacogenetics to be a useful clinical tool, clinicians must be able to interpret the results and appropriately use the information to make decisions about drug therapy. We addressed this concern by ensuring that our pharmacists were trained and educated in the pharmacogenetic considerations surrounding the drugs covered by our service. The written policy of our service guarantees that a pharmacogenetic test result will always appear in the electronic medical record along with a pharmacy consultation, which provides interpretation and suggestions for changes to drug therapy, when indicated. The clinical laboratory staff alerts the pharmacist on the service of each new pharmacogenetics test that is ordered so that the order can be reviewed for appropriateness and discussed with the patient’s care team if necessary.
Of the 66 patients whom we genotyped for TPMT through the clinical pharmacogenetics service in 2009, 5 (8%) patients were determined to have one variant allele, which results in intermediately low TPMT enzyme activity, placing these patients at increased risk for toxicity to thiopurines.24 For each patient determined to be heterozygous for a variant TPMT allele, the pharmacists’ recommendation was to prescribe a decreased dose of mercaptopurine, at a maximum of 80% of a normal daily dose for patients with acute lymphoblastic leukemia (e.g., maximum dose of 60 mg/m2/day for heterozygous patients instead of a full dose of 75 mg/m2/day). By genotyping for TPMT preemptively, the information of each patient’s heterozygous status was placed in the chart before the first dose of mercaptopurine was prescribed.
We conducted CYP2D6 genotyping for 65 patients in 2009. Of those, 49 (75%) patients were determined to be extensive metabolizers (i.e., carrying one or two [normal] functional alleles), 9 (14%) were intermediate metabolizers (i.e., carrying either two reduced-function alleles or one nonfunctional and one reduced-function alleles), 4 (6%) were poor metabolizers (i.e., carrying two nonfunctional alleles), and 3 (5%) were determined or presumed to be ultrarapid metabolizers (i.e., carrying three or more [normal] functional alleles). For patients determined to be poor metabolizers, our service alerts the clinical team that these patients are at high risk for a lack of response to codeine, as the ability to form the active metabolite from codeine requires a functional CYP2D6 gene. Likewise, for patients determined to have a duplication of the CYP2D6 gene, we alert the clinical team that this patient may be an ultrarapid metabolizer of drugs metabolized by CYP2D6, such as codeine, and that these patients are at high risk for toxicity to normal doses of codeine. A choice of another analgesic is recommended for pain control in these patients. We have had a 100% acceptance rate by physicians of the therapeutic recommendations made by pharmacists in the pharmacogenetics consultations.
While new clinical findings have brought into question the usefulness of the UGT1A1 genotype in choosing an irinotecan dose in our patient population,29 our experience is that the UGT1A1 genotype test remains a valuable clinical tool. This test is beneficial for diagnosing Gilbert syndrome, which aids in ruling out drug-related toxicity in patients with persistent hyperbilirubinemia. We determined the UGT1A1 genotype in five patients being treated for cancer who had hyperbilirubinemia in 2009; the results revealed that all five patients had two variant UGT1A1*28 alleles. This clinical information has been added to each patient’s permanent medical record, allowing the patient care team to make informed decisions about whether to hold further doses of chemotherapy and to individualize future therapy for drugs that are metabolized by UGT1A1.
To make certain that genetic information is used for drug therapy decision-making throughout a patient’s care at our hospital, automated clinical-decision-support rules have been created to facilitate the consistent application of pharmacogenetic information in specific instances. For example, in the case of CYP2D6 and codeine, when the pharmacist interprets the genetic test result and determines a patient to be a poor or ultrarapid metabolizer, this information is recorded in our hospital’s electronic medical record in two ways: through a clinical pharmacy consultation and as a codified problem list entry. When any codeine-containing medication is ordered electronically for a patient with a problem list entry of “CYP2D6 ultrarapid metabolizer,” a warning message reminds the clinician that this patient may be expected to have a higher risk of adverse effects than normal and that alternative drug therapy should be considered (Figure 1).
Our goal was to establish a clinical pharmacogenetics service to improve the dosing of drugs that are metabolized by polymorphic enzymes and to help catalyze the appreciation of pharmacogenetics as a key clinical tool by other health care providers in our system. The involvement of the clinical laboratory was essential to the establishment of the clinical pharmacogenetics service, as the laboratory staff set up the tests in the electronic medical record and established a method for posting results to the medical record. In our situation, laboratory aspects of the service were expedited, since the clinical laboratory for these services is part of the pharmaceutical department, but other pharmacy departments can achieve similar results through close collaboration with the clinical laboratory at their hospital. Because not all clinicians are familiar with the interpretation of pharmacogenetic test results, we trained pharmacists to provide test result interpretation.
To date, the clinical pharmacogenetics service at our institution focuses on pharmacogenetics in patients with cancer, but clinical pharmacogenetics also plays a role in an increasing number of diseases and drug therapies. A recent study reviewed 1200 FDA-approved drug labels and found that 121 labels contained pharmacogenomic information.11 The authors then queried the medication use of 36.1 million patients who filled prescriptions through a major pharmacy benefit management company and found that approximately one fourth of patients received a prescription for a medication that included pharmacogenomic information on the label.
In the new era of personalized drug therapy, clinical pharmacogenetics services will assist in identifying the safest and most effective drug and dosage from the outset of therapy, and pharmacists are well positioned to lead these services.
Our experiences demonstrate the feasibility of the design and function of a pharmacist-managed clinical pharmacogenetics service at an academic specialty hospital. This clinical pharmacogenetics service, one of the first of its kind, can serve as a model for other health systems. Pharmacogenetic testing has been made readily available to our patients receiving relevant drugs through collaboration between the clinical laboratory and pharmacists.
The development and implementation of this service illustrate the maturation of pharmacogenetics as a discipline and the growing awareness of the value of incorporating pharmacogenetic testing into clinical best practices and quality initiatives. As the science of pharmacogenetics advances, some clinical application of pharmacogenetic testing will likely exist in every patient population and health system. Existing pharmacokinetics services can serve as a platform for the introduction of clinical pharmacogenetic testing. As drug therapy experts, pharmacists should lead the effort to incorporate pharmacogenetic information into patient care.
The development of and experience with a pharmacist-managed clinical pharmacogenetics service are described. The program’s success has depended on collaboration between the clinical laboratory and pharmacists, and pharmacists’ pharmacogenetic recommendations have been well accepted by prescribers.
An audio interview with, which supplements the information in this article, is available on AJHP’s website at www.ajhp.org/site/misc/podcasts.xhtml.
Supported in part by grants GM-92666 and CA 21765 from the National Institutes of Health and from the American Lebanese Syrian Associated Charities. The authors acknowledge the pharmacists who operated the clinical pharmacogenetics service discussed herein, as well as all the other clinicians, patients, and families involved in this effort.
The authors have declared no potential conflicts of interest.