Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Oct 1.
Published in final edited form as: Clin Pharmacol Ther. 2024 May 26;116(4):1005–1012. doi: 10.1002/cpt.3314

Use of CYP2D6 Inhibitors with CYP2D6 Opioids: Association with Emergency Department Visits for Pain

Noor Ahmed Nahid a, Caitrin W McDonough a, Yu-Jung Jenny Wei b, Emily J Cicali a, Yan Gong a, Roger B Fillingim c, Julie A Johnson a,d,*
PMCID: PMC11452273  NIHMSID: NIHMS1994968  PMID: 38797987

Abstract

Hydrocodone, tramadol, codeine, and oxycodone are commonly prescribed opioids that rely on activation by cytochrome P450 2D6 (CYP2D6). CYP2D6 inhibitors can significantly decrease CYP2D6 activity, leading to reduced generation of active metabolites, and impairing pain control. To understand this impact, we assessed emergency department (ED) visits in patients initiating these CYP2D6-dependent opioids while on CYP2D6 inhibitor antidepressants vs. antidepressants that do not inhibit CYP2D6. This retrospective cohort study included adult patients prescribed CYP2D6-dependent opioids utilizing electronic health records data from University of Florida Health (2015– 2021). The association between ED visits and inhibitor exposure was tested using multivariable logistic regression. The primary analysis had 12,118 patients (72% female; mean [SD] age, 55 [13.4]) in the hydrocodone/tramadol/codeine cohort and 5,547 patients (64% female; mean [SD] age, 53.6 [14.2]) in the oxycodone cohort. Hydrocodone/tramadol/codeine treated patients exposed to CYP2D6-inhibitor antidepressants(n=7,043) had a higher crude rate of pain-related ED visits than those taking other antidepressants (n= 5,075) (3.28% vs 1.87%), with an adjusted odds ratio (aOR) of 1.75 (95%CI: 1.36 to 2.24). Similarly, in the oxycodone cohort, CYP2D6-inhibitor antidepressant-exposed individuals (n=3,206) had a higher crude rate of ED visits than individuals exposed to other antidepressants (n=2,341) (5.02% vs 3.37%), with aOR of 1.70 (95%CI: 1.27 to 2.27). Similar findings were observed in secondary and sensitivity analyses. Our findings suggest patients with concomitant use of hydrocodone/ tramadol/ codeine or oxycodone and CYP2D6 inhibitors have more frequent ED visits for pain, which may be due to inadequate pain control.

Keywords: CYP2D6 inhibitors, opioids, pain, emergency department visits, drug-drug interaction, precision medicine

Introduction

In the U.S., approximately 100 million people suffer from pain, making it a major health concern1,2, and opioids are widely used to treat moderate to severe pain3. Meta-analyses on chronic pain reported that compared to placebo, the magnitude of pain relief by opioids is small and not clinically important4,5, yet opioids remain widely prescribed. Specifically, hydrocodone, tramadol, oxycodone, and codeine were the 16th, 35th, 54th, and 187th most commonly prescribed medications in the U.S. in 2020, accounting for 30.1M, 17.5M, 12.3M, 2.8M prescriptions, respectively6.

The drug-metabolizing enzyme cytochrome P450 2D6 (CYP2D6) is encoded by CYP2D6, which is highly polymorphic, with more than 100 alleles identified7. Certain variant alleles lead to little to no enzymatic activity, resulting in an intermediate (IM) and poor metabolizer phenotype (PM). Gene duplication can also occur, leading to an ultra-rapid metabolizer phenotype (UM)8. Hydrocodone, tramadol, and codeine as parent compounds have little to no pharmacological activity and are bioactivated by CYP2D6 into active metabolites8–10. Those who carry the IM or PM phenotype therefore have little to no ability to generate the active metabolite required for pain relief.11,12 CYP2D6 also metabolizes oxycodone to active metabolite, though oxycodone is also active, and data are mixed on the role of CYP2D6 on the analgesic effect of oxycodone13–15.

CYP2D6 phenotype is also influenced through drug interactions via enzyme inhibition by several commonly used drugs (e.g., fluoxetine, bupropion, duloxetine). The FDA defines specific drugs as strong, moderate or weak CYP2D6 inhibitors16, and strong and moderate inhibitors (Table S1) cause phenoconversion to PM or IM phenotypes14,17, while weak inhibitors are generally considered clinically insignificant16,18,19.

Several pharmacokinetic studies reported that CYP2D6 inhibitors significantly reduce the production of active metabolites of hydrocodone, tramadol, and codeine20–22, leading to decreased effectiveness of the opioids15. Data also suggest that 20–30% of opioid-treated patients are co-prescribed CYP2D6 inhibitors23,24. However, the impact of CYP2D6 inhibitors on pain-related outcomes with opioids is not well understood. A recent study utilized a machine learning approach and reported that patients taking CYP2D6-dependent opioids and SSRIs (CYP2D6 inhibitor and non-inhibitor SSRIs) had worse pain control than patients taking SSRIs and other opioids25; however, limitations of the study included a failure to account properly for concomitant exposure to the two drugs and inclusion of non-inhibitor SSRIs.

In the current study, we aimed to evaluate pain-related emergency department (ED) visits (as a measure of impaired analgesia) and compare the outcome between users of CYP2D6 opioids (i.e., hydrocodone, tramadol, or codeine) concomitantly with a CYP2D6 vs other antidepressant (primary analysis) and between users of these CYP2D6 opioids concomitantly with vs without any CYP2D6-inhibitors (secondary analysis). We hypothesized that patients treated with opioids and a concomitant CYP2D6-inhibitor antidepressant would present to the ED more often than those concomitantly taking a non-CYP2D6-inhibitor antidepressant. We focused on patients taking CYP2D6 opioids and antidepressants because the most used CYP2D6 inhibitors are all antidepressants. Similar primary and secondary analyses were also conducted for oxycodone users, and our hypothesis was that exposure to CYP2D6 inhibitor antidepressants drugs would not be associated with increased ED visits for pain in patients treated with oxycodone due to less reliance of oxycodone on CYP2D6 than the other drugs for its efficacy. To our knowledge, this study is the first population-based, real-world study assessing the association between this drug-drug interaction and pain-related clinical outcomes.

Methods

Data source and settings:

We conducted a retrospective cohort study utilizing electronic health record (EHR) de-identified data from the University of Florida Health (UFHealth) Gainesville and Jacksonville in the integrated data repository (IDR). The study was approved by UF Institutional Review Board (Approval #201903494). Routinely collected medical data from January 01, 2015, to December 31, 2021 were utilized.

Study participants and design:

We included adult patients (18 years or older) who were prescribed at least one CYP2D6-mediated opioid (hydrocodone, tramadol, codeine, and oxycodone). Figure 1 depicts the procedures used to define study cohort. To avoid including patients with short-term acute pain (e.g., from an accident or minor procedure), and to increase the likelihood of actual consumption of the tablets, we excluded those who were prescribed opioids for a duration of seven days or less. We also sought to exclude cancer pain, thus excluded patients with cancer-related International Classification of Diseases (ICD) 9 and 10 codes in the six months prior to the index date. Based on our hypotheses, we created two cohorts, one with patients on hydrocodone, tramadol, and codeine and another with patients on oxycodone. We conducted primary and secondary analyses on both cohorts. The primary analysis was conducted on those exposed to CYP2D6-dependent opioids (hereafter referred to as opioids) plus CYP2D6-inhibitor antidepressants versus those exposed to opioids and non-CYP2D6-inhibitor antidepressants (hereafter referred as other antidepressants) (Figure 1 and Table S1). In Figure 1 we show that the first differentiator within the two groups was exposed to a CYP2D6 inhibitor and not exposed to a CYP2D6 inhibitor. This group represents our secondary analysis cohorts. Because the most used CYP2D6 inhibitors are antidepressants, and to minimize indication bias, then our primary analysis was a subset of the initial group and focused on patients on opioid plus and antidepressant – either a CYP2D6 inhibitor antidepressant, or a non-CYP2D6 inhibitor antidepressant. For the latter group, this means they had no record of CYP2D6 moderate or strong inhibitor prescription history in their EHR data. Similarly, the exposed to CYP2D6-inhibitor antidepressant group is a subset of the exposed to CYP2D6 inhibitor group (exposed to any of the 11 CYP2D6 inhibitors). That means they have prescriptions of CYP2D6 inhibitor antidepressants at the time of opioid prescriptions. The study design is depicted in Figure 2, where patients were prescribed both opioid and CYP2D6-inhibitor/other antidepressant drugs (Table S1), with the antidepressant prescription starting at the same time or prior to the opioid prescription start date, and with concomitant use of the two prescriptions for at least three days before the day defined as the index date, which is defined as day zero and occurs following 3-day concomitant opioid-antidepressant use. Patients were then followed up for 60 days from the index date for any events (ED visit). Baseline covariates were assessed for patients 180 days prior to the index date. Secondary analysis was performed on those exposed to CYP2D6 opioid with or without any CYP2D6 inhibitor (Figure S1).

Figure 1. Flowchart of study cohort identification.

Figure 1.

Open in a new tab

Figure 2. Graphical representation of the study design.

Figure 2.

Open in a new tab

The inhibitor/ non-inhibitor antidepressant prescription was initiated before or at the start date of the opioid prescription, and then there was an overlap of at least three days between the two prescriptions. The index date was set three days after the start date of the opioid prescription start date. Follow-up started from the index date for any events(ED visit) and stopped at the end of 60 days or the end date of the opioid and antidepressant prescriptions overlap, whichever occurred first. Baseline covariates were assessed for patients 180 days prior to the index date.

Exposures:

We assessed the drug exposure based on dates of prescriptions. Most opioid prescriptions (>90%) were ‘as-needed’, and we calculated the end dates in a conservative approach, using the directions of the prescriptions, quantity, and refill numbers such that a prescription of 30 tablets to be taken 1–2 tablets every 4–6 hours as needed for pain was quantified as 2.5 days of therapy. We used the end dates in EHR data for the prescriptions with a set dosing schedule for opioid and antidepressant prescriptions. To accurately determine the exposure to concomitant opioid and antidepressant prescriptions, we implemented a stitching methodology. For opioids, we combined prescriptions of the same medication with a gap of fewer than 14 days between them and considered the end date of the last prescription as the end date of that opioid exposure. For antidepressant medications, we combined prescriptions of the same medication only if there was a gap of less than or equal to 3 days between them, and we considered the end date of the last prescription as the end date of that antidepressant medication prescription. Since opioid prescriptions were mostly ‘as needed’ (PRN), and we calculated the end date in a conservative way, we opted for a liberal approach while stitching two opioid prescriptions together. On the other hand, as antidepressant prescriptions are scheduled, we stitched those prescriptions together in a more conservative manner. Concomitant exposure was then determined based on the start date and calculated end date with a minimum of 3-day overlap (Figure 2).

Outcome:

Our outcome variable for this study was an emergency department (ED) visit for pain. We identified pain-related ED visits based on ICD 9 and 10 codes (Table S2) related to pain conditions from the ED visit. If the patients had an ED visit coded for with any of these pain diagnostic codes, we considered them as having had an ED visit for pain related problems.

Follow-up periods:

Follow-up started from the index date and stopped at the end of 60 days or the end date of the opioid and antidepressant prescriptions overlap, whichever occurred first (Figure 2).

Covariates:

We assessed 26 variables related to patient demographics, comorbidities, comedications, and other factors that might influence ED visits for pain. All the covariates are listed in Table S3.

Data analysis:

Demographics and baseline characteristics were compared between opioid users with CYP2D6 inhibitors vs other antidepressants as numbers and percentages (categorical variables) or mean ± S.D. (continuous variables). Differences were assessed using chi-square tests or two-sample t-test as appropriate.

For the primary analysis, multivariable logistic regression was used to test the association between exposure to CYP2D6 inhibitor vs other antidepressant medications with the presence of ED visit (yes/no) as a measure of pain control. Stepwise selection was used to identify which variables should be included in a multivariable logistic regression model. To be eligible for entry in the model, each covariate had a minimum significance level of 0.2. Covariates with p value < 0.05 were retained in the final model (Table S3). As age and sex are two critical demographic factors, we forced them into the model if they were removed in stepwise selection. We developed separate models for the hydrocodone/tramadol/codeine cohort and oxycodone cohort for both primary and secondary analyses.

We conducted a sensitivity analysis where we included only patients who had no ED visits within the 180 days prior to the index date, as patients with recent ED visits could bias the event occurrence. As some patients had multiple events of ED visits for pain-related conditions, we also examined the incidence rate ratio (IRR) of ED visits, counting the total number of ED visits for each patient during follow-up. Then, we utilized multivariable Poisson regression for the ED visit count data. In addition to that we conducted subgroup analyses on patients who were on tramadol and on patients who were on hydrocodone. To examine whether the association was consistent in different pain problems, we also conducted subgroup analyses such as injury vs. non-injury patients (i.e., no injury/accident in 180 days prior to the index date), patients with back pain, pain in hand, leg, joint, and arthritis patients. We reported crude odds ratios (OR) and adjusted odds ratio (aOR) and crude incidence rate ratios (IRR) and adjusted incidence rate ratio (aIRR) as appropriate. The 95% confidence intervals (CIs) or p-value < 0.05 were used to establish statistical significance. All analyses were performed using SAS version 9.4 (SAS, Cary, NC), and some graphical presentations were prepared in GraphPad Prism.

Results

Study population:

The initial data pull yielded 352,468 adults prescribed at least one of the four opioids- hydrocodone, tramadol, codeine, or oxycodone. Among these, 70,168 had active opioid prescriptions with a >7-day opioid supply and represented the analysis cohort (Figure 1).

The dataset for the hydrocodone/tramadol/codeine cohort included 48,178 patients, of which 15.9% were exposed to inhibitors and 84.1% were not. Among the 11 moderate and strong CYP2D6 inhibitor drugs, four are antidepressants (Table S1), though in our dataset, about 92% of those exposed to inhibitor were on CYP2D6 inhibitor antidepressants. The primary analysis had 12,118 patients (72% female; mean [S.D.] age, 55 [13.4]), with 7,043 and 5,075 individuals concomitantly using a CYP2D6-inhibitor antidepressant vs other antidepressant, respectively. The dataset for the oxycodone cohort included 30,746 patients, from which we included 5,547 patients for the primary analysis comprising 3,206 and 2,341 individuals concomitantly using a CYP2D6-inhibitor antidepressant vs other antidepressant, respectively. The demographic and clinical characteristics of the primary and secondary analyses patients for both data cohorts are provided in Table 1 and Table S4, respectively. Differences in patient characteristics between the two comparison groups were observed, stemming in part from incorporation of data from diverse pain conditions (e.g., acute pain, back pain, etc.). Focusing on the specific pain with no other pain problems reduced the differences in their clinical characteristics (Tables S5 and S6). Covariates included in various models following stepwise selection are detailed in Table S7.

Table 1.

Characteristics of the hydrocodone/tramadol/codeine and oxycodone cohort patients

Characteristics Hydrocodone/tramadol/codeine Oxycodone^
Exposed to CYP2D6 inhibitor antidepressants# Exposed to other antidepressants## P value Exposed to CYP2D6 inhibitor antidepressants# Exposed to other antidepressants## P value
n= 7,043 (%) n= 5,075 (%) n= 3,206 (%) n= 2,341 (%)
Age, years 54.08 ± 13.53 56.65 ± 13.36 <0.0001 52.90 ± 13.47 54.59 ± 15.57 <0.0001
Sex (Females) 5236 (74.34) 3521 (69.38) <0.0001 2172 (67.75) 1401 (59.85) <0.0001
Race
White 4765 (67.66) 3426 (67.60) 0.6288 2264 (70.62) 1672 (71.42) 0.5041
Black 1856 (26.35) 1355 (26.70) 768 (23.96) 558 (23.84)
Other 422 (5.99) 284 (5.60) 174 (5.43) 111 (4.74)
Pain diagnoses*
Back pain 3359 (47.69) 2001 (39.43) <0.0001 1680 (52.40) 971 (41.48) <0.0001
Pain in hand leg joint 4231 (60.07) 2843 (56.02) <0.0001 2018 (62.94) 1330 (56.81) <0.0001
Rheumatoid arthritis 221 (3.14) 119 (2.34) 0.0091 131 (4.09) 59 (2.52) 0.0015
Headache(including migraine) 1373 (19.49) 741 (14.6) <0.0001 773 (24.11) 373 (15.93) <0.0001
Neuropathic pain 286 (4.06) 96 (1.69%) <0.0001 188 (5.86) 67 (2.86) <0.0001
Fibromyalgia 874 (12.41) 260 (5.12) <0.0001 387 (12.07) 136 (5.81) <0.0001
Injury 3722(52.85) 2335 (46.01) <0.0001 2142 (68.81) 1430 (61.09) <0.0001
Comorbidities**
Depression 3471 (49.28) 2138 (42.13) <0.0001 1787 (55.74) 1119 (47.80) <0.0001
Psychoses 474 (6.73) 271 (5.34) 0.0011 289 (9.01) 174 (7.43) 0.0354
Anxiety 3022 (42.91) 1976 (38.94) <0.0001 1556 (48.53) 1031 (44.04) 0.0009
Opioid use disorder 113 (1.60) 69 (1.36) 0.2743 189 (5.90) 104 (4.44) 0.0169
Diabetes 754 (10.71) 637 (12.55) 0.0017 382 (11.92) 281 (12.00) 0.9203
Renal failure 592 (8.41) 595 (11.72) <0.0001 122 (3.81) 331 (14.14) <0.0001
Liver disease 368 (5.23) 300 (5.91) 0.1024 273 (8.52) 246 (10.51) 0.0118
Medication history**
Opioid or non-opioid pain med 4520(64.18) 3384 (66.68) 0.0043 1934 (60.32) 1786 (76.29) <0.0001
Benzodiazepines and other sedative hypnotics 2338( 33.20) 1681 (33.12) 0.9333 954 (29.76) 808 (34.52) 0.0002
Antipsychotics 541 (7.68) 360 (7.09) 0.2237 257 (8.02) 240 (10.25) 0.004
CNS meds/stimulants 658 (9.34) 313 (6.17) <0.0001 318 (9.92) 160 (6.83) <0.0001
Skeletal muscle relaxants 2290 (32.51) 1370 (27.0) <0.0001 906 (28.26) 690 (29.47) 0.3235
Opioid related medication use
MME/DAY 23.30 ± 12.39 23.21 ± 12.29 0.728 54.32 ± 41.48 52.55 ± 34.47 0.0834
Health Care Utilization**
ED Visits (last six mo) 0.44 ± 1.21 0.49 ± 1.17 0.0077 0.72 ± 1.70 0.94 ± 1.82 <0.0001
Health care system location (Gainesville) 2084 (29.59) 1670 (32.91) <0.0001 1239(38.65) 1060 (45.28) <0.0001
Open in a new tab

Data are presented as mean ±SD or N (percent)

*

Patients could have more than one pain diagnosis

**

Within six months prior to the index date

^

Some patients may have both codeine/hydrocodone/tramadol, and oxycodone prescriptions in different time

#

Exposed to CYP2D6 inhibitor antidepressants means exposed to antidepressants that are inhibitors of CYP2D6, i.e., bupropion, paroxetine, fluoxetine, duloxetine

##

Exposed to other antidepressants means exposed to antidepressants that are not inhibitors of CYP2D6; i.e., sertraline, vilazodone, citalopram, escitalopram, venlafaxine, desvenlafaxine, trazodone

Pain-related emergency department (ED) visit:

Hydrocodone/tramadol/codeine cohort:

Hydrocodone/tramadol/codeine treated patients exposed to CYP2D6-inhibitor antidepressants had an almost double crude rate of pain-related ED visits than those taking other antidepressants (3.28% vs 1.87%), with an adjusted odds ratio (aOR) of 1.75 (95% CI: 1.36– 2.24) after adjusting for several factors (Figure 3). Similar findings were observed when looking at hydrocodone alone (aOR: 2.14 ; 95% CI:1.44– 3.17) and tramadol alone (aOR: 1.54 ; 95% CI: 1.09– 2.16). As shown in Figure 3, sensitivity and other subgroup analyses also yielded consistent findings.

Figure 3. Association of ED visits with exposure to inhibitors in different analyses.

Figure 3.

Open in a new tab

In Poisson regression analysis, we observed a statistically significant 1.56-fold increase (95% CI, 1.26– 1.95) in the incidence rate of ED visit among CYP2D6 opioid users exposed to CYP2D6 inhibitor antidepressants compared to those exposed to other antidepressants (Table S8).

Similar findings emerged in the secondary analysis, where patients exposed to any CYP2D6-inhibitors went to the ED for pain-related problems at a higher rate than those taking only opioids with no inhibitors (3.25% vs. 1.76%; aOR of 1.43; 95% CI: 1.21 – 1.69) (Figure 4).

Figure 4. Association of ED visits with exposure to inhibitors in secondary analyses.

Figure 4.

Open in a new tab

Oxycodone cohort:

Oxycodone users who concomitantly used CYP2D6-inhibitor antidepressants had a higher crude rate of ED visits than those who used other antidepressants (5.02% vs. 3.37%), with an adjusted odds ratio (aOR) of 1.70 (95% CI: 1.27– 2.27) (Figure 3). Sensitivity, subgroup and secondary analyses resulted in similar findings (Figures 3 and 4). Poisson regression analysis revealed a 1.75-fold increase (95% CI: 1.36 −2.44) in the ED visit incidence rate for patients exposed to CYP2D6 inhibitor antidepressants compared to those exposed to other antidepressants. (Table S8).

Discussion

Our primary analysis of 12,118 individuals on hydrocodone/tramadol/codeine and 5,547 individuals on oxycodone showed that initiating those opioids in patients who are also taking a CYP2D6-inhibitor antidepressant was associated with significantly more pain-related ED visits compared to those taking a concomitant antidepressant that does not inhibit CYP2D6. Secondary analyses with 48,178 individuals on hydrocodone/tramadol/codeine and 30,746 individuals on oxycodone yielded similar findings.

Previous studies indicate the analgesic effects of hydrocodone, tramadol, and codeine are primarily associated with their CYP2D6-bioactivated metabolites rather than the parent compounds8–10. Thus, our findings are consistent with CYP2D6 inhibition impairing the analgesic efficacy of these drugs.

The role of CYP2D6 metabolism in oxycodone’s analgesic efficacy is less clear in the literature, though the data from this study suggest CYP2D6 may be clinically relevant to oxycodone’s efficacy. Our findings are consistent with CYP2D6 inhibitors impairing liver metabolism to the active metabolites for all four of these opioids, potentially impairing pain control and resulting in worse patient outcomes, as was observed in this study in those concomitantly taking a CYP2D6 inhibitor. These data suggest that co-prescription of a CYP2D6 inhibitor with hydrocodone, oxycodone, tramadol or codeine should be avoided.

Defining clinical phenotypes using real-world data poses challenges, particularly in measuring pain outcomes and establishing clinically significant pain reduction 26,27. The assumption underlying the selection of ED visits for pain as the response phenotype is that a patient who presents to the ED for pain while on opioid treatment is in a state of poorly controlled pain. The ED is seen by patients as the only or best option when the pain becomes unmanageable28.

We opted for logistic regression as the approach for adjustment in our analysis. This is suitable when there are at least eight events per confounder (EPV), and our study exceeded this threshold29. We selected this over a propensity score approach because empirical coverage probability decreases after eight or more EPV with propensity scoring29. Moreover, comparison of these approaches in another study revealed odds ratios generated from different propensity score methods and regression analysis were broadly similar30.

Our study did not consider the weak inhibitors and other substrates of CYP2D6. CPIC assigned the activity score 0 in the presence of CYP2D6 strong inhibitors14. So, the presence of another CYP2D6 substrate cannot further decrease the activity of CYP2D6 in the presence of a strong inhibitor. Moreover, the presence of another substrate and moderate inhibitor should not affect the activity of CYP2D6 significantly. If a substrate causes consistent and clinically relevant enzyme inhibition to change the AUC of another substrate to clinically significant folds16, then it would also be listed by the FDA as an inhibitor of the enzyme. On the other hand, some of the weak inhibitors (sertraline, escitalopram) are in our other group exposed to non-CYP2D6 inhibitor antidepressants. If they have any effect on the outcome, then it would bias our data towards the null. The other weak inhibitor drugs should be distributed similarly in both of our groups.

Our analysis assumed that antidepressant drugs were being used for a mental health individuation, but for two antidepressant drugs, duloxetine and trazodone, as they are also commonly used for neuropathic pain and sleep (although at lower doses), respectively. To confirm that inclusion of these drugs did not influence our findings, we did two analyses excluding the patients who were on duloxetine or trazodone. These analyses did not change our findings and exclusion of either of these drugs still led to findings of a significant association between the exposure to antidepressant inhibitor drugs and increased ED visit for pain related problems (Table S9).

It is noted that there were differences in the baseline patient characteristics between the two comparison groups. As different pain types show different comorbidities, these differences appear to be attributed in part to combining data from patients with various pain conditions (e.g., acute pain, back pain, etc.)27,31. Focusing on the specific pain with no other pain problems decreased the differences in their clinical characteristics (Tables S5 and S6). Furthermore, in tables S5 and S6, regarding the mental health related comorbidities, there are 18 different tests where only 3 came as significantly different. But none of them were statistically significant after correcting for multiple comparison correction ( 0.05/18= 0.0028), suggesting no significant difference for the mental health related comorbidities between the comparison groups, when we look at each pain condition separately.

We chose to exclude cancer patients from our analysis as the origin and treatment strategies of cancer pain are quite different than other pain types. This approach is consistent with previous acute and chronic pain studies, including meta-analyses, 4,5 that excluded cancer patients/cancer pain from their analysis.

To our knowledge, our study is the first population-based, real-world evidence assessing the clinical consequences of concomitant treatment with opioids and CYP2D6 inhibitors on pain control as assessed by pain-related ED visits. Previous studies were limited by assessing only pharmacokinetic parameters without adequately examining actual pain control, limited generalizability, poorly defined clinical outcomes of pain control, or co-exposure to CYP2D6 inhibitor drugs17,25. In this study we sought to minimize these and other limitations by employing several approaches. Perhaps most notable was narrowing the population for the primary analysis to only those being treated with an antidepressant, that decreased the possible confounding by indication or prescription. We also included sensitivity, subgroup and secondary analyses to provide additional insights into our findings, which enhance the generalizability of the findings.

Our findings are consistent with Frost et al.’s17 study, which focused on tramadol consumption in the presence of CYP2D6 inhibitors and had a limited number of patients (n=152)17. They reported significantly higher daily morphine equivalents in patients with CYP2D6 inhibitors compared to patients without inhibitors17. Another study aimed to examine the effect of the combination of SSRIs and prodrug opioids on the difference between preoperative and postoperative pain scores. Although it has several limitations, they concluded that for postoperative pain control prodrug opioids are less effective than non-prodrug opioids in patients on SSRIs15.

Oxycodone is metabolized to oxymorphone by CYP2D6 and noroxycodone by CYP3A4, with further metabolism to noroxymorphone by the activity of CYP2D6 and CYP3A4, respectively11. As oxycodone, oxymorphone, and noroxymorphone are active compounds, the impact of CYP2D6 activity on oxycodone’s analgesic effect has been purported to be less pronounced than on other CYP2D6-metabolized opioids 11,12. However, very few studies examined the effect of CYP2D6 inhibitors on analgesia by oxycodone. Some studies suggest central effects of oxycodone on both analgesia and respiratory depression are attributed more to parent oxycodone than to its metabolites32,33. However our findings suggest a crucial role for CYP2D6 in the analgesic efficacy of oxycodone (measured by ED visits), despite the parent drug also having analgesic effects. This suggests oxycodone metabolites play a critical role in the analgesic efficacy of oxycodone. Our results are aligned with several previous studies. One study found concentrations of oxymorphone observed following oral administration of oxycodone were of the same magnitude as when oxymorphone was directly administered, and oxymorphone is efficacious for pain management in humans. This suggests oxymorphone mediates most of the analgesic efficacy of oxycodone34. Several studies also found that CYP2D6 PMs lack sufficient analgesia with oxycodone, which could be due to reduced formation of its active metabolite10,35,36.

Limitations:

This study has several limitations. In EHR data, missingness and misclassification are very common. We could not confirm the medicines’ actual consumption, dose, or period of exposure in this retrospective study with EHR data. We also did not have information on illicit drug use. The genetics of CYP2D6 also has the potential to modulate the formation of the metabolites from CYP2D6-mediated opioids but CYP2D6 genotype information of the individuals were not available in the EHR data19. Similar studies in different health care systems can increase confidence about generalizability of the findings. Another limitation is notable differences in the baseline patient characteristics between the two comparison groups. These differences appear to be attributed in part to combining data from patients with various pain conditions (e.g., acute pain, back pain, etc.) as different pain types show different comorbidities 27,31. Focusing on the specific pain with no other pain problems reduced the differences in their clinical characteristics (Tables S5 and S6). We included patients with different pain conditions who had concomitant inhibitor prescriptions with opioids to increase the generalizability of the findings.

Clinical Implications and Recommendations

Our data suggest that patients who are on CYP2D6 inhibitor drugs while also taking hydrocodone, oxycodone, tramadol or codeine may have more pain-related ED visits, with the implication being that the inhibitors are impairing the generation of active metabolites of the opioids and thus reducing their analgesic effects. This study’s findings provide data to guide clinicians in decision-making. When prescribing hydrocodone, tramadol, codeine, and oxycodone, concomitant medications need to be considered routinely since the presence or absence of inhibitor drug prescriptions can change over time 14,19. CYP2D6 can regain its regular activity within approximately 10 days of discontinuing CYP2D6 inhibitors37. Consideration of concomitant CYP2D6 inhibitor prescriptions and CYP2D6 genotype (if available) while prescribing hydrocodone, tramadol, oxycodone, and codeine may provide the best available personalized therapy for patients13,14,19.

Supplementary Material

Supinfo

Study Highlights.

What is the current knowledge on the topic?

Several pharmacokinetic studies reported that CYP2D6 inhibitors significantly reduce the production of active metabolites of hydrocodone, tramadol, and codeine, leading to decreased effectiveness of the opioids. However, the impact of CYP2D6 inhibitors on pain-related outcomes with opioids is not well understood.

What question did this study address?

Is concomitant use of CYP2D6-dependent opioids and CYP2D6 inhibitors associated with increased emergency department visits for pain?

What does this study add to our knowledge?

Among nearly 18,000 patients treated with hydrocodone, oxycodone, tramadol or codeine, those concomitantly taking a CYP2D6 inhibitor antidepressant were nearly twice as likely to have a visit to an emergency department for pain as those taking an antidepressant that does not inhibit CYP2D6.

How might this change clinical pharmacology or translational science?

Our findings suggest CYP2D6 inhibitors taken concomitantly with hydrocodone, oxycodone, tramadol, or codeine may have impaired bioactivation of the opioids’ active metabolites, leading to increased emergency department visits for pain. This study’s findings provide data to guide clinicians in decision-making. When prescribing those opioids, concomitant medications need to be considered routinely to provide the best available personalized therapy for patients.

Funding:

The paper has been supported by a grant from the American College of Clinical Pharmacy Foundation Futures Grant to NAN and a voucher from University of Florida Clinical and Translational Science Institute, which is supported in part by the NIH National Center for Advancing Translational Sciences under award number UL1TR001427 to NAN.

JAJ has served as an expert advisor (minor) for United Health Group in the last two years. NAN has nothing to disclose other than the grants mentioned above.

Footnotes

SUPPORTING INFORMATION

Supplementary information accompanies this paper on the Clinical Pharmacology & Therapeutics website (www.cpt-journal.com).

Conflict of Interest: All other authors declared no competing interests for this work.

Reference:

  • 1.Institute of Medicine (US) Committee on Advancing Pain Research Cr, and, Education. Relieving Pain in America: A Blueprint for Transforming Prevention, Care, Education, and Research. 2011. [PubMed] [Google Scholar]
  • 2.Nahin RL. Estimates of pain prevalence and severity in adults: United States, 2012. J Pain. Aug 2015;16(8):769–80. doi: 10.1016/j.jpain.2015.05.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Persons APoPPiO. The management of persistent pain in older persons. J Am Geriatr Soc. Jun 2002;50(6 Suppl):S205–24. doi: 10.1046/j.1532-5415.50.6s.1.x [DOI] [PubMed] [Google Scholar]
  • 4.Busse JW, Wang L, Kamaleldin M, et al. Opioids for Chronic Noncancer Pain: A Systematic Review and Meta-analysis. JAMA. Dec 18 2018;320(23):2448–2460. doi: 10.1001/jama.2018.18472 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Abdel Shaheed C, Maher CG, Williams KA, Day R, McLachlan AJ. Efficacy, Tolerability, and Dose-Dependent Effects of Opioid Analgesics for Low Back Pain: A Systematic Review and Meta-analysis. JAMA Intern Med. Jul 01 2016;176(7):958–68. doi: 10.1001/jamainternmed.2016.1251 [DOI] [PubMed] [Google Scholar]
  • 6.clincalc.com The Top 300 of 2020 https://clincalc.com/DrugStats/Top300Drugs.aspx Accessed 06/09/2023.
  • 7.PharmVar.org. Pharmacogene Variation Consortium (PharmVar) at www.PharmVar.org (Gaedigk et al. 2018, CPT 103:399; Gaedigk et al. 2019, CPT 105:29) Last accessed on 08/06/2022. [Google Scholar]
  • 8.Crews KR, Gaedigk A, Dunnenberger HM, et al. Clinical Pharmacogenetics Implementation Consortium guidelines for cytochrome P450 2D6 genotype and codeine therapy: 2014 update. Clin Pharmacol Ther. Apr 2014;95(4):376–82. doi: 10.1038/clpt.2013.254 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Volpe DA, McMahon Tobin GA, Mellon RD, et al. Uniform assessment and ranking of opioid μ receptor binding constants for selected opioid drugs. Regul Toxicol Pharmacol. Apr 2011;59(3):385–90. doi: 10.1016/j.yrtph.2010.12.007 [DOI] [PubMed] [Google Scholar]
  • 10.Foster A, Mobley E, Wang Z. Complicated pain management in a CYP450 2D6 poor metabolizer. Pain Pract. Dec 2007;7(4):352–6. doi: 10.1111/j.1533-2500.2007.00153.x [DOI] [PubMed] [Google Scholar]
  • 11.Stamer UM, Musshoff F, Kobilay M, Madea B, Hoeft A, Stuber F. Concentrations of tramadol and O-desmethyltramadol enantiomers in different CYP2D6 genotypes. Clin Pharmacol Ther. Jul 2007;82(1):41–7. doi: 10.1038/sj.clpt.6100152 [DOI] [PubMed] [Google Scholar]
  • 12.Lötsch J, Skarke C, Schmidt H, et al. Evidence for morphine-independent central nervous opioid effects after administration of codeine: contribution of other codeine metabolites. Clin Pharmacol Ther. Jan 2006;79(1):35–48. doi: 10.1016/j.clpt.2005.09.005 [DOI] [PubMed] [Google Scholar]
  • 13.Smith DM, Weitzel KW, Elsey AR, et al. CYP2D6-guided opioid therapy improves pain control in CYP2D6 intermediate and poor metabolizers: a pragmatic clinical trial. Genet Med. 08 2019;21(8):1842–1850. doi: 10.1038/s41436-018-0431-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Crews KR, Monte AA, Huddart R, et al. Clinical Pharmacogenetics Implementation Consortium Guideline for CYP2D6, OPRM1, and COMT Genotypes and Select Opioid Therapy. Clin Pharmacol Ther. 10 2021;110(4):888–896. doi: 10.1002/cpt.2149 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Monte AA, Heard KJ, Campbell J, Hamamura D, Weinshilboum RM, Vasiliou V. The effect of CYP2D6 drug-drug interactions on hydrocodone effectiveness. Acad Emerg Med. Aug 2014;21(8):879–85. doi: 10.1111/acem.12431 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.fda.gov. U.S. Food and Drug Administration. Drug Development and Drug Interactions: Table of Substrates, Inhibitors and Inducers. https://www.fda.gov/drugs/developmentapprovalprocess/developmentresources/druginteractionslabeling/ucm093664.htm Accessed on 08/06/2022.
  • 17.Frost DA, Soric MM, Kaiser R, Neugebauer RE. Efficacy of Tramadol for Pain Management in Patients Receiving Strong Cytochrome P450 2D6 Inhibitors. Pharmacotherapy. 06 2019;39(6):724–729. doi: 10.1002/phar.2269 [DOI] [PubMed] [Google Scholar]
  • 18.Cicali EJ, Elchynski AL, Cook KJ, et al. How to Integrate CYP2D6 Phenoconversion Into Clinical Pharmacogenetics: A Tutorial. Clin Pharmacol Ther. 09 2021;110(3):677–687. doi: 10.1002/cpt.2354 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Nahid NA, Johnson JA. CYP2D6 pharmacogenetics and phenoconversion in personalized medicine. Expert Opin Drug Metab Toxicol. Nov 2022;18(11):769–785. doi: 10.1080/17425255.2022.2160317 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Long T, Cristofoletti R, Cicali B, et al. Physiologically Based Pharmacokinetic Modeling to Assess the Impact of CYP2D6-Mediated Drug-Drug Interactions on Tramadol and O-Desmethyltramadol Exposures via Allosteric and Competitive Inhibition. J Clin Pharmacol. Jan 2022;62(1):76–86. doi: 10.1002/jcph.1951 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Kathiramalainathan K, Kaplan HL, Romach MK, et al. Inhibition of cytochrome P450 2D6 modifies codeine abuse liability. J Clin Psychopharmacol. Aug 2000;20(4):435–44. doi: 10.1097/00004714-200008000-00008 [DOI] [PubMed] [Google Scholar]
  • 22.Overholser BR, Foster DR. Opioid pharmacokinetic drug-drug interactions. Am J Manag Care. Sep 2011;17 Suppl 11:S276–87. [PubMed] [Google Scholar]
  • 23.Radford H, Simpson KH, Rogerson S, Johnson MI. A Single Site Population Study to Investigate CYP2D6 Phenotype of Patients with Persistent Non-Malignant Pain. Medicina (Kaunas). May 28 2019;55(6)doi: 10.3390/medicina55060220 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Tirkkonen T, Laine K. Drug interactions with the potential to prevent prodrug activation as a common source of irrational prescribing in hospital inpatients. Clin Pharmacol Ther. Dec 2004;76(6):639–47. doi: 10.1016/j.clpt.2004.08.017 [DOI] [PubMed] [Google Scholar]
  • 25.Parthipan A, Banerjee I, Humphreys K, et al. Predicting inadequate postoperative pain management in depressed patients: A machine learning approach. PLoS One. 2019;14(2):e0210575. doi: 10.1371/journal.pone.0210575 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Younger J, McCue R, Mackey S. Pain outcomes: a brief review of instruments and techniques. Curr Pain Headache Rep. Feb 2009;13(1):39–43. doi: 10.1007/s11916-009-0009-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Dowell D, Ragan KR, Jones CM, Baldwin GT, Chou R. CDC Clinical Practice Guideline for Prescribing Opioids for Pain - United States, 2022. MMWR Recomm Rep. Nov 04 2022;71(3):1–95. doi: 10.15585/mmwr.rr7103a1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Poulin PA, Nelli J, Tremblay S, et al. Chronic Pain in the Emergency Department: A Pilot Mixed-Methods Cross-Sectional Study Examining Patient Characteristics and Reasons for Presentations. Pain Res Manag. 2016;2016:3092391. doi: 10.1155/2016/3092391 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Cepeda MS, Boston R, Farrar JT, Strom BL. Comparison of logistic regression versus propensity score when the number of events is low and there are multiple confounders. Am J Epidemiol. Aug 01 2003;158(3):280–7. doi: 10.1093/aje/kwg115 [DOI] [PubMed] [Google Scholar]
  • 30.Amoah J, Stuart EA, Cosgrove SE, et al. Comparing Propensity Score Methods Versus Traditional Regression Analysis for the Evaluation of Observational Data: A Case Study Evaluating the Treatment of Gram-Negative Bloodstream Infections. Clin Infect Dis. Dec 03 2020;71(9):e497–e505. doi: 10.1093/cid/ciaa169 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Angeletti C, Angeletti PM, Paesani M, et al. Assessment of Pain and Associated Comorbidities: A Survey of Real Life Experiences Among Nurses in Italy. J Pain Res. 2021;14:107–115. doi: 10.2147/JPR.S245792 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Lalovic B, Kharasch E, Hoffer C, Risler L, Liu-Chen LY, Shen DD. Pharmacokinetics and pharmacodynamics of oral oxycodone in healthy human subjects: role of circulating active metabolites. Clin Pharmacol Ther. May 2006;79(5):461–79. doi: 10.1016/j.clpt.2006.01.009 [DOI] [PubMed] [Google Scholar]
  • 33.Kaiko RF, Benziger DP, Fitzmartin RD, Burke BE, Reder RF, Goldenheim PD. Pharmacokinetic-pharmacodynamic relationships of controlled-release oxycodone. Clin Pharmacol Ther. Jan 1996;59(1):52–61. doi: 10.1016/S0009-9236(96)90024-7 [DOI] [PubMed] [Google Scholar]
  • 34.Deodhar M, Turgeon J, Michaud V. Contribution of CYP2D6 Functional Activity to Oxycodone Efficacy in Pain Management: Genetic Polymorphisms, Phenoconversion, and Tissue-Selective Metabolism. Pharmaceutics. Sep 14 2021;13(9)doi: 10.3390/pharmaceutics13091466 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Andreassen TN, Klepstad P, Davies A, et al. Is oxycodone efficacy reflected in serum concentrations? A multicenter, cross-sectional study in 456 adult cancer patients. J Pain Symptom Manage. Apr 2012;43(4):694–705. doi: 10.1016/j.jpainsymman.2011.05.008 [DOI] [PubMed] [Google Scholar]
  • 36.Maddocks I, Somogyi A, Abbott F, Hayball P, Parker D. Attenuation of morphine-induced delirium in palliative care by substitution with infusion of oxycodone. J Pain Symptom Manage. Sep 1996;12(3):182–9. doi: 10.1016/0885-3924(96)00050-4 [DOI] [PubMed] [Google Scholar]
  • 37.O’Mathúna B, Farré M, Rostami-Hodjegan A, et al. The consequences of 3,4-methylenedioxymethamphetamine induced CYP2D6 inhibition in humans. J Clin Psychopharmacol. Oct 2008;28(5):523–9. doi: 10.1097/JCP.0b013e318184ff6e [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supinfo
NIHMS1994968-supplement-Supinfo.docx (195.1KB, docx)

RESOURCES