Accepted Manuscript

CYP2D6 drug-gene and drug-drug-gene interactions among

patients prescribed pharmacogenetically actionable opioids

Mitchell R. Knisely, Janet S. Carpenter, Claire Burke Draucker,

Todd Skaar, Marion E. Broome, Ann M. Holmes, Diane Von Ah

PII: S0897-1897(17)30193-3
DOI: doi:10.1016/j.apnr.2017.10.001
Reference: YAPNR 50965
To appear in: Applied Nursing Research
Received date: 21 April 2017
Revised date: 5 October 2017
Accepted date: 6 October 2017

Please cite this article as: Mitchell R. Knisely, Janet S. Carpenter, Claire Burke Draucker,
Todd Skaar, Marion E. Broome, Ann M. Holmes, Diane Von Ah , CYP2D6 drug-gene and
drug-drug-gene interactions among patients prescribed pharmacogenetically actionable
opioids. The address for the corresponding author was captured as affiliation for all
authors. Please check if appropriate. Yapnr(2017), doi:10.1016/j.apnr.2017.10.001

This is a PDF file of an unedited manuscript that has been accepted for publication. As
a service to our customers we are providing this early version of the manuscript. The
manuscript will undergo copyediting, typesetting, and review of the resulting proof before
it is published in its final form. Please note that during the production process errors may
be discovered which could affect the content, and all legal disclaimers that apply to the
journal pertain.
 ACCEPTED MANUSCRIPT
CYP2D6 Drug-Gene and Drug-Drug-Gene Interactions Among Patients Prescribed
Pharmacogenetically Actionable Opioids

Authors:
Mitchell R. Knisely, PhD, RN, Department of Health Promotion & Development, School of Nursing,
University of Pittsburgh, 3500 Victoria Street, 360D, Pittsburgh, PA 15261

Janet S. Carpenter, PhD, RN, FAAN, Distinguished Professor & Associate Dean for Research,
Science of Nursing Care Department, School of Nursing, Indiana University, 600 Barnhill Drive, NU
340G, Indianapolis, IN 46202

Claire Burke Draucker, PhD, RN, FAAN, Professor & Angela Barron McBride Endowed Professorship

T
in Mental Health Nursing, Community and Health Systems Department, School of Nursing, Indiana

IP
University, 600 Barnhill Drive, NU W409, Indianapolis, IN 46202

Todd Skaar, PhD, Associate Professor of Medicine, Division of Clinical Pharmacology, Department of

CR
Medicine, School of Medicine, Indiana University, 950 W. Walnut St., Research II Room E402,
Indianapolis, IN 46202

US
Marion E. Broome, PhD, RN, FAAN, Dean & Ruby Wilson Professor of Nursing, School of Nursing,
Duke University, 307 Trent Dr., 4142 Pearson Bldg., Durham, NC 27710

Ann M. Holmes, PhD, Associate Professor, Health Policy & Management, School of Public Health,
AN
Indiana University, 1050 Wishard Blvd., RG 5138, Indianapolis, IN 46202

Diane Von Ah, PhD, RN, FAAN, Chair & Associate Professor, Community & Health Systems
M

Department, School of Nursing, Indiana University, 600 Barnhill Drive, NU 407, Indianapolis, IN 46202

Acknowledgements: The authors would like to thank Dr. Joseph Ipe, the Regenstrief Institute, the
ED

Indiana Biobank, contributors who collected samples used in this study, and subjects whose
participation made this work possible.
Funding Acknowledgements: This work was supported by the Indiana University School of Nursing
PT

100th Anniversary Scholars Fellowship; William & Doris Rodie Dissertation Award; National Institute of
Nursing Research [Award Numbers T32NR007066 & T32NR009759]; National Human Genome
Research Institute [Award Number 5U01HG007762]; and the Indiana Clinical and Translational
CE

Sciences Institute which is funded in part from the National Institutes of Health, National Center for
Advancing Translational Sciences, Clinical and Translational Sciences Award [Award Number
UL1TR001108] and the National Center for Research Resources, Construction Grant [Award Number
RR020128] and the Lilly Endowment. The content is solely the responsibility of the authors and does
AC

not necessarily represent the official views of the mentioned funding agencies.
Declaration of Conflicting Interests: The authors declare no potential conflicts of interest with respect
to the research, authorship, and/or publication of this article.
Corresponding author: Mitchell R. Knisely, University of Pittsburgh School of Nursing, 3500 Victoria
Street, 360D, Pittsburgh, PA 15261, phone 260-243-1486, email [email protected].
 ACCEPTED MANUSCRIPT
Abstract
Purpose: When codeine and tramadol are used for pain management, it is imperative that nurses are
able to assess for potential drug-gene and drug-drug-gene interactions that could adversely impact
drug metabolism and ultimately pain relief. Both drugs are metabolized through the CYP2D6 metabolic
pathway which can be affected by medications as well the patient’s own pharmacogenotype. The
purpose of this brief report is to identify drug-gene and drug-drug-gene interactions in 30 adult patients
prescribed codeine or tramadol for pain.
Methods: We used three data sources: (1) six months of electronic health record data on the number

T
and types of medications prescribed to each patient; (2) each patient’s CYP2D6 pharmacogenotype,

IP
and (3) published data on known CYP2D6 gene-drug and drug-drug-gene interactions.
Results: Ten patients (33%) had possible drug-gene or drug-drug-gene interactions. Five patients had

CR
CYP2D6 drug-gene interactions indicating they were not good candidates for codeine or tramadol. In
addition, five patients had potential CYP2D6 drug-drug-gene interactions with either codeine or

US
tramadol.
Conclusion: Our findings from this exploratory study underscores the importance of assessing and
AN
accounting for drug-gene and drug-drug-gene interactions in patients prescribed codeine or tramadol.
M

Keywords: Drug-gene interaction, drug-drug-gene interaction, pharmacogenomics, pharmacogenetic

testing, opioid, pain management, polypharmacy
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT
Drug-gene and drug-drug-gene interactions can interfere with patients’ abilities to metabolize
codeine and tramadol for pain relief. Drug-gene interactions (DGIs) are defined as the influence of a
patient’s genetic makeup on the ability to metabolize or clear a drug.1 Drug-drug-gene interactions
(DDGIs) are defined as the effect of a patient’s genetic makeup and another medication in their
regimen (e.g., cytochrome P450 inhibitor or inducer) on that patient’s ability to metabolize or clear a
certain medication.1 A medication that is a cytochrome P450 inhibitor prevents the abilities of a drug
metabolizing enzyme (e.g., CYP2D6) to metabolize a medication, whereas a medication that is an
inducer increases the abilities of an enzyme to metabolize a medication.2 Therefore, DDGIs can alter a

T
person’s drug metabolizing status (e.g., changing a person from a normal to a poor metabolizer) and

IP
increase the risk of adverse events such as treatment non-response or toxicity.3 Understanding both
drug-gene and drug-drug-gene interactions are beneficial in predicting medication response. Current

CR
efforts to implement widespread pharmacogenetic testing in clinical practice require that nurses be
knowledgeable about this field.4

US
The goal of implementing pharmacogenetic testing in clinical practice is to help address wide
interindividual variations in response to medication therapies. Ideally, this testing would lead to
AN
improved clinical and economic outcomes through avoidance of adverse drug effects and medication
non-response.5-7 However, current clinical practice guidelines for pharmacogenetic testing account for
drug-gene interactions but do not account for the complexity of patients’ medication regimens.8
M

Polypharmacy, or the concurrent taking of multiple medications, may increase the potential for adverse
drug effects or medication non-response as a result of known or unknown drug-drug-gene interactions.9
ED

For example, a person with a normal pharmacogenotype may not adequately metabolize a given
medication if they are taking a concurrent medication that inhibits or induces the metabolic pathway for
PT

that given medication. While there is a paucity of research exploring drug-gene and drug-drug-gene
interactions in the context of pain management, clinicians and researchers must consider the potential
CE

impact of both when evaluating the effects of codeine and tramadol for pain relief.
The purpose of this brief report was to identify drug-gene and drug-drug-gene interactions in 30
AC

adult patients prescribed codeine or tramadol for pain. Two pharmacogenetically actionable opioids
(i.e., codeine and tramadol) were chosen as exemplar medications because they have evidence-based
guidelines to guide drug or dosage changes based on pharmacogenetic test results.10
Methods
Design & Setting
This study was a retrospective analysis of 6-months of de-identified data from electronic health
records (EHRs) and banked DNA specimens from a random selection of patients who were newly
prescribed a pharmacogenetically actionable opioid in a primary care clinic within a large public health
system that provides care for an underserved and low income population. Due to the de-identified
 ACCEPTED MANUSCRIPT
nature of this study, it was deemed non-human subjects research by the Office of Human Subjects
Research.
Sample
The sample included 30 patients randomly selected from the population of patients who met the
following inclusion criteria: 1) part of a managed care program for individuals falling at or below 200
percent of the federal poverty level, 2) first prescription of record for codeine or tramadol in at least one
of the primary care clinics within the health system, 3) had a banked blood sample, 4) age 21 and older,
and 5) had no documentation of substance abuse in the electronic health record. A ‘first prescription of
record’ was defined as either codeine or tramadol being prescribed between January 1, 2010 and

T
December 31, 2014 and no information in the patient’s record that indicated that either medication had

IP
been previously prescribed.

CR
Data & Procedures
Six months of EHR data for each patient were obtained from comprehensive data repositories.

US
The 6-month time period started with the date of the new prescription for either codeine or tramadol and
ended six months later. Data were accessed, extracted, and de-identified by a trained data analyst. The
main repository where data originated was the Indiana Network for Patient Care, which is a large, multi-
AN
institutional information exchange that captures EHR data from more than 25,000 individual providers,
106 hospitals, and 110 clinics and surgery centers.11 The repository allows for capturing information
M

about patients’ healthcare encounters across multiple clinics and health systems and information about
medication orders and/or dispensing data from the safety-net health system, other health systems, and
ED

outpatient pharmacies within the region where this study was conducted. Other sources included
clinical and administrative data repositories associated with the managed care program and public
PT

health system. The EHR data were linked to existing DNA specimens obtained from the biobank.
Patient characteristics and medication regimens. Demographic characteristics at the time of
CE

the new prescription for the pharmacogenetically actionable opioid were extracted from the EHR,
including age, gender, and race. The medication regimen prescribed for each patient was identified
through data elements including medication name, dose, dose frequency, route, administration
AC

instructions, and fill or refill dates. The mean number and range of medications prescribed per patient
was determined over the 6-month time period. We also classified type of medication based on its
indication (e.g., pain medications, cardiac medications, psychiatric medications, etc.) and calculated
frequencies, mean number, and range of medications for that type.
CYP2D6 drug-gene interaction. A DGI occurs when the metabolization or clearance of a
medication is influenced by a gene or genes.1 In this study, we identified the effects of a patient’s
cytochrome P450 2D6 (CYP2D6) genotype on their ability to metabolize or clear codeine or tramadol.
Both codeine and tramadol are considered prodrugs, meaning they are inactive and will not produce
analgesic effects until metabolized into the active form of the drug. Therefore, we completed CYP2D6
 ACCEPTED MANUSCRIPT
genotype analyses on the banked DNA specimens for all patients in this study. The genotype results
were then used to calculate the CYP2D6 enzyme activity score.10 The activity score is used to identify
the drug metabolizing status (i.e., phenotype) classification and is determined by the sum of values
assigned to each CYP2D6 star allele making up the diplotype.10 The drug metabolizing status
classification includes ultra-rapid (activity score: >2), normal (activity score: 1-2), intermediate (activity
score: 0.5), and poor (activity score: 0). The genotyping was performed for the CYP2D6 star alleles *2,
*3, *4, *5, *6, *9, *10, *17, *29, and *41 using the QuantStudio (Thermo Fisher Scientific, Inc., Grand
Island, NY) and following the manufacturer’s instructions of the Taqman Genotyping Assays (Applied

T
Biosystems, Inc., Foster City, CA). These star alleles were chosen because they have common

IP
variants that influence both codeine and tramadol drug disposition and response, as well as their
frequencies in Caucasian, African and their descendant populations.12 We completed quality controls

CR
for each sample batch and repeated assays for 20% of samples. We had a genotype call rate, defined
as the proportion of genotypes per marker with non-missing data13, of 99.4%. This underscores the

US
quality of our samples and genotyping procedures.
CYP2D6 drug-drug-gene interactions. A DDGI occurs when there is an interaction between
an individual’s CYP2D6 genotype and a medication that affects a person’s ability to appropriately
AN
metabolize a different medication (e.g., codeine, tramadol).1 In this study, we identified a potential
DDGI was present if two medications had known CYP2D6 interactions and that the supply amount
M

dispensed for each medication overlapped. For example, if a 30-day supply of tramadol was dispensed
10 days following a 60-day supply of fluoxetine being dispensed, we assumed that these two
ED

medications would be taken concurrently and coded these data as a DDGI present. The Flockhart
TableTM for cytochrome P450 drug interactions served as the reference for identifying DDGIs in this
PT

sample (http://medicine.iupui.edu/clinpharm/ddis/main-table).2
To determine the effects of the DDGI on the CYP2D6 drug metabolizing status, we then
CE

adjusted the CYP2D6 enzyme activity score (identified from the CYP2D6 genotyping processes
described above) for each patient based on identified DDGIs that would alter the CYP2D6 enzyme
AC

activity.2 For drug interactions with strong inhibitors of the opioid, the final activity score of 0 was
assigned. For drug interactions with moderate inhibitors of the opioid, the CYP2D6 enzyme activity
score was multiplied by 0.5 to identify the final activity score.14 Consequently, the drug metabolizing
status for each patient was then determined from the final activity score.10
Results
The sample consisted of 30 adults (16 female), of whom 60% were White and 40% were Black
or Biracial. The mean age was 50.9 years (SD = 10.4; range = 23 – 65 years). The majority of patients
were prescribed tramadol (n = 24), with fewer prescribed codeine (n = 6). Patients received an average
of 10 different medication prescriptions over the 6-month time period (range = 2-24). A mean of 3
different pain medications were prescribed per patient, of which one was either codeine or tramadol per
 ACCEPTED MANUSCRIPT
inclusion criteria. Apart from the pain medications, the most common class of medications prescribed
were for cardiovascular conditions (n = 23, 77% of sample), with 2.9 medications per patient for
indications such as hypertension, dyslipidemia, and anticoagulation. In addition, 60% (n = 18) of the
sample were prescribed medications for psychiatric problems such as depression or anxiety.
There were a total of 10 patients (33%) adversely affected by a DGI or DDGI. Adverse DGIs
were present in 17% of the sample (n=5) (Table 1). Two patients were poor metabolizers, one patient
an intermediate metabolizer, and two patients were considered ultra-rapid metabolizers. These patients
would not be good candidates for codeine or tramadol due to the increased likelihood they would

T
experience adverse drug effects or treatment non-response. The remaining patients were normal

IP
metabolizers (n=25) of the codeine or tramadol.
DDGIs were present in 17% of the sample (n=5) (Table 2). Based solely on the

CR
pharmacogenetic genotyping results, and not considering the effects of the DDGI on the patient’s
CYP2D6 drug metabolizing status, all of these patients were initially considered normal metabolizers.

US
However, the identified DDGIs involved either a strong (n = 4) or moderate inhibitor (n = 1) of the
prescribed opioid, which altered patients’ CYP2D6 metabolizer status. As a result, four additional
AN
patients were ultimately classified as CYP2D6 poor metabolizers and one was classified as a CYP2D6
intermediate metabolizer.
Discussion
M

Findings from this exploratory study suggest that both drug-gene and drug-drug-gene
interactions can be an issue in both clinical practice and research. Of note, 33 percent (n = 10) of our
ED

sample had altered CYP2D6 metabolizing status as a result of such interactions. Five different patients
were identified with CYP2D6 drug-gene interactions and five additional patients were identified with
PT

potential CYP2D6 drug-drug-gene interactions with the pharmacogenetically actionable opioids. All
interactions would result in altering their metabolizing status and result in no or decreased analgesia or
CE

enhanced side effects from tramadol or codeine.10 These findings are consistent with larger studies
evaluating similar interactions.1, 9
Our findings highlight a number of clinical and research considerations. The fact that numerous
AC

patients in this study had altered CYP2D6 metabolizing status underscores the potential benefits of
implementing pharmacogenetic testing in practice. Without determining these patients CYP2D6
pharmacogenotype, these drug-gene and drug-drug-gene interactions would go undetected. In
addition, this study provides important illustrations of why pharmacogenetic test results and complete
medication regimens should be considered in tandem by nurses. Consequently, nurses need to
consider drug-gene and drug-drug-gene interactions when assessing and educating patients about
their medication regimens. Furthermore, future research is necessary to confirm and extend our
findings. Studies that identify CYP2D6 drug-gene and drug-drug-gene interactions in a larger sample
 ACCEPTED MANUSCRIPT
and the impact of these interactions on patient outcomes (e.g., pain relief, adverse events) are
warranted.
Conclusion
The findings from this exploratory study underscore the importance of assessing and accounting
for drug-gene and drug-drug-gene interactions in patients prescribed codeine or tramadol. We found
that pharmacogenetic test results that identified drug-gene interactions were only part of the story and
that the complexity of complete medication regimens added significantly to understanding patients’
abilities to appropriately metabolize these opioid medications. Therefore, future research and clinical

T
guidelines must also account for potential for drug-drug-gene interactions.

IP
Table 1. Patients with Adverse Cytochrome P450 2D6 (CYP2D6) Drug-Gene Interactions
CYP2D6 Drug Patient CYP2D6 CYP2D6 Prescribed Clinical Interpretation
Metabolizing ID Diplotype Activity PGxA Opioid

CR
Status Score
12 *4/*5 0 Tramadol Patients are carrying no
Poor functional CYP2D6 alleles.
Metabolizer 22 *4/*4 0 Codeine Increased risk for lack of opioid

US
efficacy.
Patient carrying a reduced
function CYP2D6 allele and a
Intermediate
21 *5/*41 0.5 Tramadol non-functional CYP2D6 allele.
AN
Metabolizer
Increased risk for reduced
opioid efficacy.
Patients are carrying more
5 *1/*17 >2 Tramadol
M

Ultra-rapid than 2 copies of functional

Metabolizer CYP2D6 alleles. Increased risk
17 *2/*2xN >2 Tramadol
for potential opioid toxicity.
ED

Note: This table only includes patients with adverse CYP2D6 drug-gene interactions and does not include
patients (n = 25) who were considered to be normal CYP2D6 drug metabolizers.
PGxA: pharmacogenetically actionable
PT

Table 2. Patients with Cytochrome P450 2D6 (CYP2D6) Drug-Drug-Gene Interactions & Clinical
Interpretations
Patient CYP2D6 CYP2D6 CYP2D6 DIs Final Final Clinical Interpretation
CE

ID Diplotype Genotype CYP2D6 CYP2D6

Activity Activity Drug
Score Score Metabolizing
AC

Status
1
07 *1/*41 1.5 Diphenhydramine – 0 Poor Tramadol: Avoid
tramadol Metabolizer medication due to lack
of efficacy.
1
11 *2/*5 1 Diphenhydramine – 0 Poor Tramadol: Avoid
tramadol Metabolizer medication due to lack
of efficacy.
1
16 *1/*5 1 Fluoxetine – tramadol 0 Poor Tramadol: Avoid
Metabolizer medication due to lack
of efficacy.
1
19 *1/*4 1 Bupropion – codeine 0 Poor Codeine: Avoid
Metabolizer medication due to lack
of efficacy.
2
23 *1/*4 1 Duloxetine – tramadol 0.5 Intermediate Tramadol: Avoid
Metabolizer medication due to
reduced efficacy.
 ACCEPTED MANUSCRIPT
1 2
DI: Drug interaction; CYP2D6 strong inhibitor: Final CYP2D6 Activity Score = 0; CYP2D6 moderate
inhibitor: multiply CYP2D6 Genotype Activity Score by 0.5 to determine Final CYP2D6 Activity Score.

References
1. Verbeurgt, P., Mamiya, T., & Oesterheld, J. (2014). How common are drug and gene interactions?
Prevalence in a sample of 1143 patients with CYP2C9, CYP2C19 and CYP2D6 genotyping.
Pharmacogenomics, 15(5), 655-665.
2. Love, R. R., Desta, Z., Flockhart, D., Skaar, T., Ogburn, E. T., Ramamoorthy, A., . . . Jarjoura, D. (2013).
CYP2D6 genotypes, endoxifen levels, and disease recurrence in 224 Filipino and Vietnamese women
receiving adjuvant tamoxifen for operable breast cancer. Springerplus, 2(1), 52.

T
3. Brixner, D., Biltaji, E., Bress, A., Unni, S., Ye, X., Mamiya, T., . . . Biskupiak, J. (2016). The effect of
pharmacogenetic profiling with a clinical decision support tool on healthcare resource utilization and

IP
estimated costs in the elderly exposed to polypharmacy. J Med Econ, 19(3), 213-228.
4. Levy, K., Decker, B., Carpenter, J., Flockhart, D., Dexter, P., Desta, Z., & Skaar, T. (2014). Prerequisites to

CR
implementing a pharmacogenomics program in a large healthcare system. Clinical Pharmacology &
Therapeutics, 93(3), 307-309.
5. Arnaout, R., Buck, T. P., Roulette, P., & Sukhatme, V. P. (2013). Predicting the cost and pace of
pharmacogenomic advances: an evidence-based study. Clin Chem, 59(4), 649-657.

US
6. Ross, S., Anand, S. S., Joseph, P., & Pare, G. (2012). Promises and challenges of pharmacogenetics: an
overview of study design, methodological and statistical issues. JRSM Cardiovasc Dis, 1(2), 1-13.
7. Zdanowicz, M. (2010). Concepts in Pharmacogenomics. Bethesda, MD: American Society of Health-
AN
System Pharmacists.
8. Bahar, M. A., Setiawan, D., Hak, E., & Wilffert, B. (2017). Pharmacogenetics of drug-drug interaction and
drug-drug-gene interaction: a systematic review on CYP2C9, CYP2C19 and CYP2D6. Pharmacogenomics,
M

18(7), 701-739.
9. Hocum, B. T., White, J. R., Jr., Heck, J. W., Thirumaran, R. K., Moyer, N., Newman, R., & Ashcraft, K.
(2016). Cytochrome P-450 gene and drug interaction analysis in patients referred for pharmacogenetic
ED

testing. Am J Health Syst Pharm, 73(2), 61-67.

10. Crews, K. R., Gaedigk, A., Dunnenberger, H. M., Leeder, J. S., Klein, T. E., Caudle, K. E., . . . Skaar, T. C.
(2014). Clinical Pharmacogenetics Implementation Consortium guidelines for cytochrome P450 2D6
PT

genotype and codeine therapy: 2014 update. Clin Pharmacol Ther, 95(4), 376-382.
11. Indiana Health Information Exchange. (n.d.). Indiana Network for Patient Care. Retrieved January 11,
2016, from http://www.ihie.org/indiana-network-for-patient-care
CE

12. Hicks, J. K., Bishop, J. R., Sangkuhl, K., Muller, D. J., Ji, Y., Leckband, S. G., . . . Gaedigk, A. (2015). Clinical
Pharmacogenetics Implementation Consortium (CPIC) Guideline for CYP2D6 and CYP2C19 Genotypes
and Dosing of Selective Serotonin Reuptake Inhibitors. Clin Pharmacol Ther, 98(2), 127-134.
AC

13. Anderson, C. A., Pettersson, F. H., Clarke, G. M., Cardon, L. R., Morris, A. P., & Zondervan, K. T. (2010).
Data quality control in genetic case-control association studies. Nature protocols, 5(9), 1564-1573.
14. Borges, S., Desta, Z., Jin, Y., Faouzi, A., Robarge, J. D., Philips, S., . . . Li, L. (2010). Composite functional
genetic and comedication CYP2D6 activity score in predicting tamoxifen drug exposure among breast
cancer patients. J Clin Pharmacol, 50(4), 450-458.
 ACCEPTED MANUSCRIPT
Highlights

 CYP2D6 drug-gene and drug-drug-gene interactions can interfere with patients’ abilities to
metabolize codeine and tramadol for pain relief.
 33% of the sample were affected by a CYP2D6 drug-gene or drug-drug-gene interaction.
 Findings suggest that both drug-gene and drug-drug-gene interactions can be an issue in both
clinical practice and research.

T
IP
CR
US
AN
M
ED
PT
CE
AC