Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Oct 1.
Published in final edited form as: Drug Alcohol Depend. 2021 Sep 1;227:109025. doi: 10.1016/j.drugalcdep.2021.109025

Effect of HIV, antiretrovirals, and genetics on methadone pharmacokinetics: Results from the methadone antiretroviral pharmacokinetics study

Gavin Bart a,*, Le Minh Giang b, Hoang Yen b, James S Hodges c, Richard C Brundage d
PMCID: PMC8767566  NIHMSID: NIHMS1754398  PMID: 34482033

Abstract

Background:

Methadone treatment of opioid use disorder in HIV-infected individuals is complicated by drug-drug interactions. Genetic and other cofactors further contribute to interindividual variability in methadone pharmacokinetics. We used population pharmacokinetics to estimate the effect of drug-drug interactions, genetics, and other cofactors on methadone pharmacokinetics in a methadone maintained population in Vietnam.

Methods:

Plasma R- and S-methadone levels were measured in 309 methadone maintained individuals just before and 2–5 h following methadone dosing. A linear one-compartment population pharmacokinetic model with first- order conditional estimation with interaction was used to evaluate methadone clearance (CL/F) and volume of distribution (V/F). The influence of covariates on parameter estimates was evaluated using stepwise covariate modeling. Covariates included HIV status, antiretroviral use (efavirenz or nevirapine), weight, BMI, age, methadone dose, and 8 single nucleotide polymorphisms in across the CYP2B6, ABCB1, and NR1I3 genes.

Results:

Taking either efavirenz or nevirapine increased R-methadone CL/F 220%. Nevirapine and efavirenz increased S-methadone CL/F by 404% and 273%, respectively. Variants in NR1I3 increased R- and S-methadone CL/F by approximately 20% only in patients taking efavirenz. Different alleles in ABCB1 rs2032582 either increased or decreased R-methadone CL/F by 10%. The CYP 2B6*4 variant decreased S-methadone CL/F by 18%. HIV-infection increased R- and S-methadone CL/F and V/F by 24%–39%.

Conclusions:

The HIV antiretrovirals nevirapine and efavirenz significantly increase methadone clearance. Variants inNR1I3 increased the effect of efavirenz on methadone clearance. Other variants affecting methadone CL/F were also confirmed. To our knowledge, this is the first report of HIV itself affecting methadone pharmacokinetics.

Keywords: Methadone, HIV, Pharmacokinetics, Drug-drug interactions, Genetics

1. Introduction

Injection drug use contributes to nearly one-third of HIV infections outside sub-Saharan Africa (Wolfe et al., 2010). Injection opioid use is the main drug use associated with HIV infections; harm reduction strategies and medications for opioid use disorder (MOUD), such as methadone and buprenorphine, can reduce HIV transmission (MacArthur et al., 2012; Miller et al., 2018). MOUD also reduce all-cause, drug-related, and AIDS-related mortality by more than half (Nosyk et al., 2015; Sordo et al., 2017).

By dramatically reducing cycles of opioid withdrawal and drug use, MOUD improve engagement in the HIV continuum of care, including receipt of and adherence to antiretroviral therapy (ART) and suppression of HIV viral load (Low et al., 2016). Reducing illicit drug use and retaining patients on MOUD, therefore, are important for treating OUD and managing HIV. Reduced opioid use and retention in treatment are dose-responsive, with higher doses of MOUD also associated with better HIV outcomes (Lappalainen et al., 2015). There are, however, significant drug-drug interactions (DDI) between methadone and several antiretroviral drugs (Bruce et al., 2013; Gruber and McCance-Katz, 2010), which generally induce methadone metabolism, lowering plasma concentrations and shortening methadone half-life. These DDI can impact methadone’s effectiveness by precipitating opioid withdrawal, return to illicit opioid use, non-adherence to MOUD, and, consequently, non-adherence to ART (Tossonian et al., 2007). To overcome these DDI, methadone dose often is increased or split from once-daily to twice-daily. Yet large interindividual variability remains in these DDI, with methadone dose adjustments between 15%–250% required to overcome clinical instability (Clarke et al., 2001a, b). Little is known about whether genetic variants explain variability in methadone DDI with ART. Better understanding of DDI between methadone and ART may lead to strategies that limit their negative clinical consequences.

Methadone is the most commonly used MOUD throughout the world. It is a full agonist at mu-opioid receptors, a racemic mixture with the R- enantiomer acting as an opioid agonist and both R- and S-enantiomers acting as N-methyl-d-aspartate (NMDA) receptor antagonists. In treating opioid use disorder, methadone is generally dosed once daily and has a half-life of approximately 27h. P-glycoprotein efflux channels influence methadone passage across the intestinal lumen and blood-brain barrier and cytochrome P450 enzymes (CYP) metabolize methadone into inactive constituents in the liver and intestine (Eap et al., 2002). While CYP 2B6 is the principal metabolizing enzyme, it has preference for S-methadone (Totah et al., 2007). CYP 3A4 and, to a lesser extent, CYP 2C19, CYP 1A2 and CYP 2D6 are also involved in methadone metabolism. We and others have identified variants in genes encoding P-glycoprotein (ABCB1) and various CYPs (CYP2B6, CYP3A4, CYP2C19, CYP1A2, and CYP2D6) that influence methadone clearance (Ahmad et al., 2018; Bart et al., 2014; Dennis et al., 2014).

Many ART interact with methadone, with the non-nucleoside reverse transcriptase inhibitors nevirapine and efavirenz among the most profound interactions. While nevirapine mostly has been phased out, efavirenz remains an alternative first-line therapeutic and is among the most widely used ART in the world (World Health, 2019). Efavirenz and nevirapine induce expression and activity of CYP 2B6 and CYP 3A4 (Hariparsad et al., 2004); some genetic variants that influence methadone pharmacokinetics also affect efavirenz. Besides being a substrate at CYP 2B6, efavirenz binds to the constitutive androstane receptor (CAR). A variant in the gene expressing CAR (NR1I3) also increases CYP 2B6 expression and activity (Svard et al., 2010). Therefore, there may be genetic influences on methadone-ART DDI that are unique to efavirenz.

We previously used population pharmacokinetics (POPPK) to identify effects of ethnicity and genetic variants in ABCB1 and CYP2B6 on methadone pharmacokinetics in a population not taking medications known to interact with methadone (Bart et al., 2014). These variants reduce clearance of R- and S-methadone and S-methadone, respectively. Population pharmacokinetics is a US Food and Drug Administration recommended approach that uses sparse data sampling to study sources and correlates of variability in drug concentrations and pharmacokinetics within and between individuals and populations (Food and Drug Adminstration, 1999; Sheiner and Ludden, 1992). Through sparse sampling, POPPK allows larger samples of persons than traditional, densely sampled, pharmacokinetics, which increases the ability to evaluate the influence on pharmacokinetic parameters of multiple covariates such as genetic variants and clinical characteristics (Sheiner et al., 1977). The objective of this research was to use POPPK to evaluate methadone pharmacokinetics and identify variables that influence methadone DDI with efavirenz and nevirapine. This study’s aims were to confirm our previous findings of the effect of ABCB1 and CYP2B6 variants on methadone exposure, to characterize DDI between methadone and nevirapine and efavirenz, and to test the novel hypothesis that variants in NR1I3 accentuate the DDI between methadone and efavirenz.

2. Materials and methods

2.1. Participants

Participants in the Methadone Antiretroviral Pharmacokinetics Study (MAPS) were recruited from community methadone treatment programs in Hanoi or Ho Chi Minh City, Vietnam. Clinical and embedded research staff approached clinic patients to participate. Vietnam was selected because of its high prevalence of HIV among methadone patients (10%–30%), relative genetic homogeneity, and centralized approach to methadone and HIV treatment, which limited variability in dosing strategies and available ART combinations. MAPS was not considered a clinical trial and thus not required to be registered in clinicaltrials.gov. It was approved by the Human Subjects Research Committee of Hennepin Healthcare Research Institute and the Socialist Republic of Vietnam’s Ministry of Health Ethics Review Committee and conducted in accordance with the Helsinki Declaration of 1975 (2013 revision).

Participants were at least 18 years old (minimum age allowed for methadone treatment in Vietnam) and met International Classification of Diseases-10 criteria for opioid dependence when enrolled in methadone treatment. HIV+ and HIV− methadone patients were recruited. We planned to enroll HIV+ patients about to initiate ART to evaluate change in methadone pharmacokinetics pre- to post-ART initiation and through time, but most HIV+ methadone patients willing to be on ART were already on ART when the study began, so few HIV+ participants not on ART were enrolled. The study included a cross-sectional (single day of study) and a prospective arm (five study visits over 6 months). For population pharmacokinetic modeling, data from both arms were combined. Participants were excluded if they were in the 2nd or 3rd trimester of pregnancy, had end-stage liver disease, were HIV+ and receiving ART that did not include nevirapine or efavirenz, or could not provide written informed consent due to mental health or cognitive condition. All research staff were Vietnamese and all consents, research forms, and interviews were in Vietnamese.

2.2. Procedures

Following receipt of written signed consent, participants provided demographic and basic medical information, including current medications, comorbid illnesses, and HIV status. HIV status was confirmed by chart review of test results, CD4 count, and viral load, or for those with no or older than 90-day test results, with direct testing for antibodies to HIV-1/2 with reflex CD4 and viral load testing if positive. Methadone clinic records were abstracted for date of methadone initiation, previous 5-day methadone dosing record (all methadone dosing in Vietnam is directly observed), and hepatitis B and C serology.

Approximately 24 h after their previous methadone dose, 20 mL of venous blood was collected into lithium heparin blood collection tubes for methadone plasma analysis. The Clinical Opioid Withdrawal Scale (COWS) was assessed just before the participant took their daily methadone dose (Wesson and Ling, 2003). The COWS (range 0–48) combines objective and subjective opioid withdrawal symptoms with scores ≥5 indicating withdrawal. Another blood sample for methadone levels was obtained 2−5 hours after their daily methadone dose. With separate consent, blood was collected into PAXgene Blood DNA Tubes (Qiagen, Germantown, MD) for DNA extraction and genetic analysis. Urine was collected for point-of-care drug testing. Between blood draws, participant weight and height were collected for body mass index (BMI; weight in kilograms/height in meters2). Participants completed the Center for Epidemiologic Studies Depression Scale (CES-D), a screening tool for major depressive disorder that has been validated among HIV+ individuals in Vietnam (Radloff, 1977; Thai et al., 2016). The CES-D is scored 0–60; score ≥16 indicates significant depressive symptoms. Participants taking ART completed the ACTG Adherence Questionnaire, which measures past 4 day ART adherence (Chesney et al., 2000).

2.3. Assays

Blood specimens for HIV testing, CD4 count, and HIV viral load were processed within 6 h of blood draw by the National Hospital for Tropical Diseases in Hanoi using standard clinical assays. Urine drug testing used point-of-care immunoassay (Abon Biopharm, Hangzhou, China) and detected opiates, methamphetamine, marijuana, and methylenedioxymethamphetamine (MDMA).

Plasma levels of R- and S-methadone were determined by the University of Minnesota Center for Forecasting Drug Response using a previously-described protocol (Bart et al., 2014; Liang et al., 2004). Briefly, detection and quantification of total R- and S-methadone in human plasma used a high-performance liquid chromatograph (Agilent 1100 Series, Santa Clara, CA) coupled with a API 4000 triple quadrupole instrument (MDS-SCIEX, Concord, Ontario, Canada). Chromatographic separation used a Chiral-AGP column (5 um, 2.0 mm × 50 mm). The assay was linear in the range of 1.25–1250 ng/mL, using 1/X weighting. Method validation accuracy was 101.2% and total variability was 6.1% (within-day 5.1% and between-day 2.9%).

2.4. Genetic analysis

Whole blood for DNA extraction was shipped to RUCDR Infinite Biologics (Piscataway, NJ). Purified DNA (mean concentration 10 ng/μL) was shipped to the University of Minnesota Genome Center where the Agena Bioscience iPLEX Gold assay and MassARRAY system (San Diego, CA) genotyped 8 single nucleotide variant sites across the ABCB1 (rs2032582, rs1045642, and rs1128503), CYP2B6 (rs3745274, rs2279343, and rs36079186), and NR1I3 (rs2307424 and rs3003596) genes (see supplemental materials for primer specifications). The selected variants have been associated with altered methadone or efavirenz pharmacokinetics (Bart et al., 2014; Svard et al., 2010).

2.5. Population pharmacokinetic analysis

2.5.1. Fixed effects model

Based on previous work, a one-compartment disposition model was assumed (Bart et al., 2014). Since concentration data were collected shortly before a dose (about 24 h after the previous dose) and about two hours after dosing, these data could not provide information on either absorption or rapid distribution processes. Hence, absorption was modeled as first order with the first-order rate constant fixed to 1.5 h−1 (Wolff et al., 1993). The PK model was parameterized using apparent clearance (CL/F) and apparent volume of distribution (V/F). Total concentrations (bound and unbound methadone) were analyzed independently for R-methadone and S-methadone. Covariate fixed effects in each analysis were body size (body weight or BMI); HIV status (+/−); concomitant therapy with either efavirenz (EFV +/−, or nevirapine +/−); and these single nucleotide polymorphism variants: rs1045642, rs1128503, rs2032582, rs2279343, rs2307424, rs3003596, rs3745274.

2.5.2. Random effects model

Between-subject variability (BSV) for CL/F and V/F was assumed to be log-normally distributed, which expresses BSV as a coefficient of variation (CV%). Correlation between CL/F and V/F was also estimated. Concentration data were available for up to five visits per participant, permitting estimation of between-occasion variability (BOV) within participants as an additional random variable. Residual unexplained variability (RUV) was modeled using a proportional-error model which, estimates RUV as a CV%.

2.5.3. Model building

Defining dummy variables for HIV+/−, EFV+/−, and NVP+/−, a model was constructed to estimate methadone CL/F for these sub-groups separately: HIV-/ART-; HIV+/ART-; HIV+/EFV+; and HIV+/NVP +. Subsequently, groups could be combined by setting a regression parameter to its null value to determine if there was a statistical difference between the two models. For example, EFV+ and NVP+ could be combined (i.e., ART+) to test if the effects on CL/F differed between EFV and NVP. Such a combined model is nested within the larger model so a significance test uses the likelihood ratio test and the Chi-square distribution. With two subpopulations combined, the objective function value (OFV; a goodness-of-fit measure similar to a sum-of-squares) needs to increase by 6.6 units (Chi-square, df = 1; p < 0.01) to claim the model with two separate subpopulations (EFV+ and NVP+) fits significantly better than the model combining the two subpopulations (ART+). Using this approach, we tested the significance of HIV status and antiretroviral status on methadone CL/F.

After those subpopulations were combined when appropriate, the effects of genotype on CL/F of the individuals was tested in a forward inclusion manner. When adding an effect of a variant, the OFV needed to decrease by 6.6 units (Chi-square, df = 1; p < 0.01) to claim adding the variant significantly improved the model.

2.5.4. Software

This nonlinear mixed-effects model was implemented using NONMEM (v. 7.4.3, ICON Development Solutions, Ellicott City, MD). Supporting software included Pirana v2.9.7, R v4.0.1, R-studio v1.3.959, and Pearl-speaks-NONMEM (PsN) v4.8.1. NONMEM specifications included ADVAN2/TRANS2 subroutines, and the FOCE-I estimation method.

2.5.5. Model qualification

Model adequacy was evaluated using diagnostic plots and a prediction-corrected visual predictive check. Parameter estimates were evaluated for statistical significance guided by change in OFV, biological relevance, clinical importance, and statistical precision of the estimate, i.e., standard errors and 95% confidence intervals. Also, a non-parametric bootstrap approach was used to estimate the 2.5th, 50th, and 97.5th percentile from 1000 bootstrap data sets sampled with replacement from the original data set.

2.5.6. Population subgroup confounding

In this study, all HIV- participants were ART−. Similarly, all ART+ participants were HIV+. Thus, attempts to compare ART+ to ART− or HIV+ to HIV- may be confounded, so a secondary analysis was conducted in which the subset including just ART- participants was used to test the effect of HIV status, and another subset including only HIV + participants was used to test ART status.

3. Results

From May 2016 through September 2018, 325 people consented to participate in MAPS. Two participants withdrew before enrollment, 13 withdrew before their first study visit, and blood samples could not be obtained from 2 participants (see CONSORT diagram in supplemental materials). Thus 309 participants and 2009 R-methadone and 2006 S-methadone plasma concentrations were included in the final analyses. All participants were receiving methadone maintenance therapy, 175 were taking no ART, 65 were taking nevirapine (NVP), and 69 were taking efavirenz (EFV). Table 1 presents baseline characteristics. Consistent with the general methadone population in Vietnam’s major cities, participants were overwhelmingly male and self-identified as Kinh ethnicity. The mean methadone dose in ART- participants was 67 mg (SD 33) and for those on NVP or EFV, 133 mg (SD 76) and 154 mg (SD 63), respectively (p < 0.05). Compared to those not taking ART, significantly more participants on NVP or EFV endorsed opioid withdrawal symptoms, were at risk for depression, and had antibodies consistent with viral hepatitis infection.

Table 1.

Baseline characteristics.

No ART (n = 175) NVP (n = 65) EFV (n = 69)
Age years (SD) 39 (8.4) 40 (6.7) 38 (6.1)
Sex
 Male 166 (95 %) 61 (94 %) 66 (96 %)
Race
 Kinh 174 (99 %) 65 (100 %) 68 (99 %)
 Other 1 (1 %) 0 1 (1 %)
HIV Status
 Positive 20 (11.5 %) 65 (100 %) 69 (100 %)
*HIV viral load ≤ 200 copies/ mLa
 Yes 1 (5 %) 59 (91 %) 62 (90 %)
 No 119 (95 %) 16 (9 %) 17 (10 %)
*CD4 count cells/mm3 (SD) 333 (195) 440 (227) 433 (251)
Time on methadone, years (SD)a 1.44 (1.78) 3.60 (2.81) 2.00 (1.75)
Methadone dose, mg (SD)a 67 (33) 133 (76) 154 (63)
COWS ≥ 5 (%)a 3 (1.7) 19 (29.2) 14 (20.3)
BMI kg/m2 (SD) 21 (2.9) 19 (2.6) 20 (2.9)
Viral hepatitisa
 None 87 (50 %) 20 (31 %) 10 (15 %)
 HCV 74 (42 %) 42 (65 %) 54 (78 %)
 HBV 9 (5 %) 1 (1.5 %) 0
 HCV HBV coinfection 5 (3 %) 2 (3 %) 5 (7 %)
Depression (CES-D)a
 At risk 27 (15 %) 6 (9 %) 18 (26 %)
Open in a new tab
a

denotes group difference, p < 0.05;

*

For no ART group: only calculated for the HIV + records (n = 20).

Genetic analysis found deviations from Hardy-Weinberg equilibrium for rs2032582, rs3003596, and rs3745274. Because all participants had the same TT genotype for rs36079186, this was not analyzed. Combining rs3745274 and rs2279343 to evaluate the CYP 2B6*6 haplotype did not alter the objective function value in the pharmacokinetic model so each variant was evaluated separately. Body size (weight or BMI) was not related to any pharmacokinetic parameters and is not included in the results below.

3.1. R-Methadone

Parameter estimates in the HIV- group without genetic variants for R-methadone CL/F and V/F were 6.11 L/h and 316 L, respectively, with a half-life of 35.8 h (Table 2). No difference between EFV and NVP was found in CL/F and the groups could be combined into a single group taking ART. Being HIV+, taking ART, presence of the rare ABCB1 rs2032582 T allele, and combined presence of the NR1I3 rs2307424 GG and rs3003596 GG genotypes if taking EFV each significantly increased R-methadone CL/F. The ABCB1 rs2032582 AA genotype significantly decreased R-methadone CL/F. Being HIV + significantly increased R-methadone V/F.

Table 2.

R-Methadone parameters.

PHARMACOKINETIC PARAMETERS
Parameter Estimate RSE % 95 % CI Bootstrap 50th %ile Bootstrap 2.5–97.5%ile CI
CL (L/hr) HIV- 6.11 2.7 % 5.79–6.43 6.11 5.76–6.43
CL (L/hr) HIV + ART- 7.86 6.8 % 6.82–8.90 7.82 6.90–8.92
CL (L/hr) HIV + ART+ 13.4 3.3 % 12.5–14.3 13.44 12.46–14.54
CL (L/hr) ABCB12677AA 5.54 3.7 % 5.14–5.94 5.54 5.13–5.95
CL (L/hr) ABCB12677xT 6.66 4.1 % 6.11–7.21 1.08 6.08–7.21
CL (L/hr) CAR/EFV 7.15 5.4 % 6.42–7.88 7.15 6.42–9.94
VC (L) HIV- 316 2.9 % 298–334 316 298–334
V (L) HIV+ 392 4.0 % 360–423 392 363– 423
BETWEEN SUBJECT VARIABILITY (BSV)
Estimate RSE % Shrinkage (%) Bootstrap 50th %ile Bootstrap 2.5–97.5%ile CI
BSV CL 28.9 % 5.7 % 8 % 29.1 % 25.9–32.8%
BSV V 28.8 % 7.2 % 19 % 29.1 % 24.6–33.5%
CL,V correlation (r) 0.572 9.4 % - 0.572 0.483–0.625
BOV CL 15.5 % 5.2 % 37 % 15.5 % 13.9–17.2%
RESIDUAL UNEXPLAINED VARIABILITY (RUV)
Estimate RSE% Shrinkage (%) Bootstrap 50th %ile Bootstrap 95 %-ile CI
RUV 12.6 % 5.0 % 29 % 12.6 % 11.4–14.0%
Open in a new tab

RSE%: % relative standard error; BOV: between occasion variability; CI: confidence interval.

3.2. Model evaluation

Median parameter estimates for R-methadone from 1000 bootstrap simulations were comparable to the population pharmacokinetic model estimates, differing less than 10% from those in the final model (Table 2). A visual predictive check indicated no anomaly (see supplemental material).

3.3. S-Methadone

Parameter estimates in the HIV- group without genetic variants for S-methadone CL/F and V/F were 6.91 L/h and 205 L, respectively, with a half-life of 20.6 h (Table 3). Being HIV+/ART−, HIV+/EFV+, or HIV+/NVP+ all increased S-methadone CL/F. Notably, increases in CL/F among ART−, EFV+, and NVP+ all differed significantly from each other. Estimated CL/F was also increased in the presence of the NR1I3 rs3003596 GG genotype if taking EFV. The CYP2B6 variant rs2279343 G allele was associated with decreased S-methadone CL/F. Being HIV+ significantly increased S-methadone V/F.

Table 3.

S-Methadone parameters.

PHARMACOKINETIC PARAMETERS
Parameter Estimate RSE % 95 % CI Bootstrap 50th %ile Bootstrap 2.5–97.5%ile CI
CL (L/hr) HIV- 6.91 3.4 % 6.46–7.37 6.90 6.46–7.36
CL (L/hr) HIV+ ART- 9.41 9.2 % 7.71–11.1 9.40 7.88–11.40
CL (L/hr) HIV+ EFV+ 18.9 5.1 % 17.0–20.8 19.00 16.93–21.14
CL (L/hr) HIV+ NVP+ 27.9 5.2 % 25.1–30.7 27.84 24.95–30.10
CL (L/hr) NR1I3 8784 with EFV 8.29 7.0 % 7.19–9.40 8.22 7.05–9.60
CL (L/hr) CYP2B6 785G 5.67 3.6 % 5.27–6.06 5.69 5.29–6.08
VC (L) HIV- 205 3.0 % 193–217 205 195–219
V (L) HIV+ 285 4.1 % 262–308 287 262–310
Random Effects Estimate RSE % Shrinkage (%) Bootstrap 50th %ile Bootstrap 2.5–97.5%ile CI
BETWEEN SUBJECT VARIABILITY (BSV)
BSV CL 37.9 % 5.1 % 6.5 % 39.0 % 35.0–43.8 %
BSV V 30.8 % 6.4 % 15 % 31.1 % 27.1–35.7 %
CL,V correlation (r) 0.727 6.8 % - 0.731 0.691–0.758
BOV CL 16.3 % 6.3 % 37 % 16.3 % 14.3–18.4 %
RESIDUAL UNEXPLAINED VARIABILITY (RUV)
RUV 17.8 % 4.8 % 27.5 % 17.7 % 16.1–19.4 %
Open in a new tab

RSE%: % relative standard error; BOV: between occasion variability; CI: confidence interval.

3.4. Model evaluation

Median parameter estimates for S-methadone from 1000 bootstrap simulations were comparable to the population pharmacokinetic model estimates, differing less than 10% from those in the final model (see Table 3). A visual predictive check indicated no anomaly (see supplemental material).

3.5. Secondary analysis

Considering HIV+ participants only, R-methadone CL/F was 8.6 L/hr for ART- participants, 14.3 L/hr for those on EFV, and 13.0 L/hr for those on NVP. Similarly, for S-methadone, CL/F for ART- participants was 9.22 L/hr, 18.2 L/hr for EFV+ participants, and 24.2 L/hr for NVP+ participants. The differences due to ART status were significant and these findings are consistent with the primary analysis.

Considering ART- participants only, R-methadone CL/F was 6.09 L/hr for HIV- participants and 6.20 L/hr for HIV+ participants. Similar values for S-methadone were found (6.11 L/hr and 6.51 L/hr, respectively). HIV status had no significant impact on methadone CL/F in the ART- subset of participants.

4. Discussion

This population pharmacokinetics study of methadone patients in Vietnam found significant clinical and genetic influences on methadone pharmacokinetic parameters. Our hypotheses were partially upheld. The antiretroviral therapeutics efavirenz and nevirapine significantly increased methadone clearance. We confirmed that genetic variants in the ABCB1 and CYP2B6 genes decreased methadone clearance. In a novel finding, NR1I3 gene variants significantly increased methadone clearance only in the presence of efavirenz.

Interactions between methadone and efavirenz and nevirapine are well known. Most prior reports of these interactions have been clinical descriptions of induced opioid withdrawal or small clinical pharmacokinetic studies in healthy HIV- volunteers (Altice et al., 1999; Kharasch et al., 2012; Marzolini et al., 2000). The few small studies in HIV+ methadone patients found methadone exposure declined about 50% in patients starting either EFV or NVP (Clarke et al., 2001a, b; Pelet et al., 2011). These studies measured total methadone, not the separate enantiomers. Because EFV and NVP increase expression and function of CYP 2B6, which has preference for S-methadone, total methadone may misrepresent clinically relevant effects because reduction in total methadone disproportionately decreases S-methadone, which lacks opioid effect. Kharasch measured both enantiomers in a non-clinical population and found that methadone clearance following EFV was increased by 2- and 2.5−3- fold for R- and S-methadone, respectively (Kharasch et al., 2012). Our results for S-methadone are consistent with Kharasch’s for EFV. We have no comparison for NVP’s effect on S-methadone clearance although our observed four-fold increase may be consistent with a nearly threefold difference between R- and S-methadone concentrations observed in patients on NVP (Esteban et al., 2008). Because EFV and NVP are anticipated to have less effect on R-methadone than on S-methadone, it is not surprising the best fit model combined them for R-methadone. When participants on either EFV or NVP were combined, our results were consistent with a twofold increase in R-methadone clearance and our baseline difference in methadone dose requirements between participants on EFV or NVP versus those not on ART.

We found that the CYP 2B6*4 (rs2279343) variant independently reduced S-methadone clearance. This variant confers a gain of function effect on CYP 2B6 and, therefore, would be expected to increase S-methadone clearance rather than reduce it, as we observed (Gadel et al., 2015). We cannot explain this unexpected finding, although when the rs2279343 variant is combined with rs3475274 and rs3211371 variants it forms the loss of function CYP2B6*7 variant, which we did not assess. In a secondary analysis combining rs2279343 and rs3475274 for the loss of function CYP2B6*6 variant, we found no significant effect on S-methadone clearance. We could not confirm our previous finding of decreased S-methadone clearance for the CYP 2B6*9 (rs3475274) variant. This may be explained in part by more genetic heterogeneity in our previous study compared to this uniformly Vietnamese population.

We also found an independent effect of rs2032582 variants in the ABCB1 gene on R-methadone clearance. We previously found this variant genotype was associated with approximately 20% lower clearance of both R- and S-methadone and cannot explain the current lack of finding for S-methadone (Bart et al., 2014). Here we observed a 10% decrease for R-methadone only. Other studies have not found an effect of this allele by itself on methadone dose or plasma levels, although when included in various haplotype analyses it is associated with higher methadone trough levels, consistent with our findings (Crettol et al., 2006; Levran et al., 2008). We also observed a 10% increase in R-methadone clearance for the rare T allele, something not found in prior methadone studies likely because its frequency is <1% in these studies’ European and African-American populations.

The interactive effect of variants in the NR1I3 gene and efavirenz, increasing methadone clearance, is a novel finding. A variant in NR1I3 is associated with increased EFV levels (Swart et al., 2012). Because EFV can increase CYP 2B6 function and impact methadone pharmacokinetics independently of its transcriptional effect on the CYP 2B6 gene (Kharasch et al., 2012), we hypothesized that increased EFV levels in carriers of the NR1I3 variant would cause an increased methadone-EFV drug--drug interaction compared to those without the variant. Our observed 20% increase in methadone clearance is of a magnitude generally considered clinically relevant and may necessitate further study of methadone dose requirements and withdrawal symptomology in patients with the variants taking EFV and, if confirmed, could provide a rationale for using ART regimens without EFV in these patients. Further study is needed to understand why the effect on R-methadone clearance required both NR1I3 variants (rs2307424 GG and rs3003596 GG) and that on S-methadone required the rs3003596 GG genotype only.

Our finding that HIV independently increases R- and S-methadone clearance and volume of distribution is novel. Further, this effect is independent of HIV markers such as viral load and CD4 count (data not shown). While BMI differed significantly between HIV+ and HIV− participants, our pharmacokinetic model did not identify BMI or fat-free mass as influencing methadone pharmacokinetics. The magnitude of increase in clearance (CL/F) and volume of distribution (V/F) may coincidentally be similar, or such parallel increases could represent an effect on bioavailability (F) because decreasing F would increase each proportionately. As the gut is a major organ for HIV infection, alterations in F seem feasible given that lack of gastrointestinal immune reconstitution has been documented despite chronic ART treatment with suppressed viral load and restored CD4 count (Guadalupe et al., 2006). While there is no information specific to NVP’s effect on methadone bioavailability, Kharasch found that EFV decreased R- and S-methadone bioavailability in healthy HIV- subjects (Kharasch et al., 2012).

4.1. Limitations

Aside from the confounding between HIV and ART status, this study has other limitations. There may be further confounding by HCV status as participants with HIV had a significantly higher rate of HCV antibody positivity and an independent effect of HCV on methadone pharmacokinetics requires further elucidation (Ključcević et al., 2018; Talal et al., 2020; Wu et al., 2013). Also, we targeted specific genetic variants rather than conducting larger microarray studies of multiple variants, so our ability to identify haplotypes or novel variants associated with methadone pharmacokinetics is limited. Clinical correlates to our findings on treatment outcome remain unexplored, although we note that the EFV and NVP groups were more likely to have at least mild opioid withdrawal compared to ART- participants, which aligns with our observed increases in methadone clearance in these groups. Our approach was further limited by our inability to conduct studies of methadone-ART DDI using a pre-post ART initiation design.

We cannot rule out a confounding effect of ART on F given that most participants with HIV were also taking ART. In our secondary subpopulation analysis comparing HIV− to HIV+ participants not taking ART, we found no significant effect of HIV on methadone pharmacokinetics. Therefore, confirming the independent effect of HIV on methadone pharmacokinetics (and possibly bioavailability) seen in our primary analysis will require a pharmacokinetic study of oral and intravenous methadone in an HIV+ population not taking ART.

5. Conclusions

Our use of population pharmacokinetics is an important step towards precision medicine in addiction treatment. We identified drug interactions and genetic and clinical variables that impact methadone pharmacokinetics. Most prior literature on methadone clinical response or genetic influences has used either methadone dose or plasma concentration as the independent variable (Levran et al., 2008; Talal et al., 2020; Victorri-Vigneau et al., 2019). This is problematic because pharmacodynamic variables influence dose requirements and plasma concentrations at any given time depend on dose. A pharmacokinetic parameter such as clearance is independent of pharmacodynamics and dose and is, therefore, more easily linked to clinical outcomes over time and across dose ranges. Prospective population pharmacokinetic studies in patient populations over the course of methadone treatment may help identify factors that predict treatment outcome and rationale drug dosing regimens based on these variables.

Supplementary Material

MAPS Supp

Acknowledgements

The authors would like to acknowledge Dr. Jag Khalsa for early support of the grant, Drs. Tam Nguyen Thi Minh and Thanh Van Ta for supporting implementation of this study in Vietnam, Mr. James Fisher for establishing and conducting all methadone assays, and the research assistant support of Ms. Linnae Baird, Ms. Carnalya Johnson, and the HMU CREATA team.

Role of the funding source

Research reported in this publication was supported by the National Institute on Drug Abuse of the National Institutes of Health under Award Number R01DA040510. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Footnotes

Declaration of Competing Interest

The authors report no declarations of interest.

Appendix A. Supplementary data

Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.drugalcdep.2021.109025.

References

  1. Ahmad T, Valentovic MA, Rankin GO, 2018. Effects of cytochrome P450 single nucleotide polymorphisms on methadone metabolism and pharmacodynamics. Biochem. Pharmacol. 153, 196–204. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Altice FL, Friedland GH, Cooney EL, 1999. Nevirapine induced opiate withdrawal among injection drug users with HIV infection receiving methadone. AIDS 13 (8), 957–962. [DOI] [PubMed] [Google Scholar]
  3. Bart G, Lenz S, Straka RJ, Brundage RC, 2014. Ethnic and genetic factors in methadone pharmacokinetics: a population pharmacokinetic study. Drug Alcohol Depend. 145, 185–193. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bruce RD, Moody DE, Altice FL, Gourevitch MN, Friedland GH, 2013. A review of pharmacological interactions between HIV or hepatitis C virus medications andopioid agonist therapy: implications and management for clinical practice. Expert Rev. Clin. Pharmacol. 6 (3), 249–269. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Chesney MA, Ickovics JR, Chambers DB, Gifford AL, Neidig J, Zwickl B, Wu AW, 2000. Self-reported adherence to antiretroviral medications among participantsin HIV clinical trials: the AACTG adherence instruments. Patient Care Committee &Adherence Working Group of the Outcomes Committee of the Adult AIDS Clinical Trials Group (AACTG). AIDS Care 12 (3), 255–266. [DOI] [PubMed] [Google Scholar]
  6. Clarke SM, Mulcahy FM, Tjia J, Reynolds HE, Gibbons SE, Barry MG, Back DJ, 2001a. Pharmacokinetic interactions of nevirapine and methadone and guidelinesfor use of nevirapine to treat injection drug users. Clin. Infect. Dis. 33 (9), 1595–1597. [DOI] [PubMed] [Google Scholar]
  7. Clarke SM, Mulcahy FM, Tjia J, Reynolds HE, Gibbons SE, Barry MG, Back DJ, 2001b. The pharmacokinetics of methadone in HIV-positive patients receiving the non-nucleoside reverse transcriptase inhibitor efavirenz. Br. J. Clin. Pharmacol. 51 (3), 213–217. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Crettol S, Deglon JJ, Besson J, Croquette-Krokar M, Hammig R, Gothuey I, Monnat M, Eap CB, 2006. ABCB1 and cytochrome P450 genotypes and phenotypes: influence on methadone plasma levels and response to treatment. Clin. Pharmacol. Ther. 80 (6), 668–681. [DOI] [PubMed] [Google Scholar]
  9. Dennis BB, Bawor M, Thabane L, Sohani Z, Samaan Z, 2014. Impact of ABCB1 and CYP2B6 genetic polymorphisms on methadone metabolism, dose and treatment response in patients with opioid addiction: a systematic review and meta-analysis. PLoS One 9 (1), e86114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Eap CB, Buclin T, Baumann P, 2002. Interindividual variability of the clinical pharmacokinetics of methadone: implications for the treatment of opioid dependence. Clin. Pharmacokinet. 41 (14), 1153–1193. [DOI] [PubMed] [Google Scholar]
  11. Esteban J, Pellin ML, Gimeno C, Barril J, Gimenez J, Mora E, Garcia-Perez AG, 2008. Increase of R-/S-methadone enantiomer concentration ratio in serum of patients treated with either nevirapine or efavirenz. Drug Metab. Lett. 2 (4), 269–279. [DOI] [PubMed] [Google Scholar]
  12. Food and Drug Adminstration, 1999. Guidance for industry: population pharmacokinetics. In: Center for Drug, E., Research, Center for Biologicals, E., Research (Eds.), Food and Drug Administration. US Department of Health and Human Services, Rockville, MD. [Google Scholar]
  13. Gadel S, Friedel C, Kharasch ED, 2015. Differences in methadone metabolism by CYP2B6 variants. Drug Metab. Dispos. 43 (7), 994–1001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Gruber VA, McCance-Katz EF, 2010. Methadone, buprenorphine, and street drug interactions with antiretroviral medications. Curr. HIV/AIDS Rep. 7 (3), 152–160. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Guadalupe M, Sankaran S, George MD, Reay E, Verhoeven D, Shacklett BL, Flamm J, Wegelin J, Prindiville T, Dandekar S, 2006. Viral suppression and immune restoration in the gastrointestinal mucosa of human immunodeficiency virus type 1-infected patients initiating therapy during primary or chronic infection. J. Virol. 80 (16), 8236–8247. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Hariparsad N, Nallani SC, Sane RS, Buckley DJ, Buckley AR, Desai PB, 2004. Induction of CYP3A4 by efavirenz in primary human hepatocytes: comparison with rifampin and phenobarbital. J. Clin. Pharmacol. 44 (11), 1273–1281. [DOI] [PubMed] [Google Scholar]
  17. Kharasch ED, Whittington D, Ensign D, Hoffer C, Bedynek PS, Campbell S, Stubbert K, Crafford A, London A, Kim T, 2012. Mechanism of efavirenz influence on methadone pharmacokinetics and pharmacodynamics. Clin. Pharmacol.Ther. 91 (4), 673–684. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Ključević Ž, Benzon B, Ključević N, Veršić Bratinčević M, Sutlović D, 2018. Liver damage indices as a tool for modifying methadone maintenance treatment: a cross-sectional study. Croat. Med. J. 59 (6), 298–306. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Lappalainen L, Nolan S, Dobrer S, Puscas C, Montaner J, Ahamad K, Dong H, Kerr T, Wood E, Milloy MJ, 2015. Dose-response relationship between methadone dose and adherence to antiretroviral therapy among HIV-positive people who use illicit opioids. Addiction 110 (8), 1330–1339. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Levran O, O’Hara K, Peles E, Li D, Barral S, Ray B, Borg L, Ott J, Adelson M, Kreek MJ, 2008. ABCB1 (MDR1) genetic variants are associated with methadone doses required for effective treatment of heroin dependence. Hum. Mol. Genet. 17 (14), 2219–2227. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Liang HR, Foltz RL, Meng M, Bennett P, 2004. Method development and validation for quantitative determination of methadone enantiomers in human plasma by liquid chromatography/tandem mass spectrometry. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 806 (2), 191–198. [DOI] [PubMed] [Google Scholar]
  22. Low AJ, Mburu G, Welton NJ, May MT, Davies CF, French C, Turner KM, Looker KJ, Christensen H, McLean S, Rhodes T, Platt L, Hickman M, Guise A, Vickerman P, 2016. Impact of opioid substitution therapy on antiretroviral therapy outcomes: a systematic review and meta-analysis. Clin. Infect. Dis. 63 (8), 1094–1104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. MacArthur GJ, Minozzi S, Martin N, Vickerman P, Deren S, Bruneau J, Degenhardt L, Hickman M, 2012. Opiate substitution treatment and HIV transmission in people who inject drugs: systematic review and meta-analysis. BMJ 345. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Marzolini C, Troillet N, Telenti A, Baumann P, Decosterd LA, Eap CB, 2000. Efavirenz decreases methadone blood concentrations. AIDS 14 (9). [DOI] [PubMed] [Google Scholar]
  25. Miller WC, Hoffman IF, Hanscom BS, Ha TV, Dumchev K, Djoerban Z, Rose SM, Latkin CA, Metzger DS, Lancaster KE, Go VF, Dvoriak S, Mollan KR, Reifeis SA, Piwowar-Manning EM, Richardson P, Hudgens MG, Hamilton EL, Sugarman J, Eshleman SH, Susami H, Chu VA, Djauzi S, Kiriazova T,Bui DD, Strathdee SA, Burns DN, 2018. A scalable, integrated intervention to engage people who inject drugs in HIV care and medication-assisted treatment (HPTN 074): a randomised, controlled phase 3 feasibility and efficacy study. The Lancet 392 (10149), 747–759. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Nosyk B, Min JE, Evans E, Li L, Liu L, Lima VD, Wood E, Montaner JSG,2015. The effects of opioid substitution treatment and highly active antiretroviral therapy on the cause-specific risk of mortality among HIV-positive people who inject drugs. Clin. Infect. Dis. 61 (7), 1157–1165. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Pelet A, Favrat B, Cavassini M, Eap CB, Besson J, Monnat M, 2011. Usefulness of methadone plasma concentration measurement in patients receiving nevirapine or efavirenz. Am. J. Drug Alcohol Abuse 37 (4), 264–268. [DOI] [PubMed] [Google Scholar]
  28. Radloff LS, 1977. The CES-D scale:a self-report depression scale for research in the general population. Appl. Psychol. Meas. 1 (3), 385–401. [Google Scholar]
  29. Sheiner LB, Ludden TM, 1992. Population pharmacokinetics/dynamics. Annu. Rev. Pharmacol. Toxicol. 32 (185–209), 185–209. [DOI] [PubMed] [Google Scholar]
  30. Sheiner LB, Rosenberg B, Marathe VV, 1977. Estimation of population characteristics of pharmacokinetic parameters from routine clinical data. J. Pharmacokinet. Biopharm. 5 (5), 445–479. [DOI] [PubMed] [Google Scholar]
  31. Sordo L, Barrio G, Bravo MJ, Indave BI, Degenhardt L, Wiessing L, Ferri M, Pastor-Barriuso R, 2017. Mortality risk during and after opioid substitution treatment: systematic review and meta-analysis of cohort studies. BMJ 357, j1550. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Svard J, Spiers JP, Mulcahy F, Hennessy M, 2010. Nuclear receptor-mediated induction of CYP450 by antiretrovirals: functional consequences of NR1I2 (PXR) polymorphisms and differential prevalence in whites and sub-Saharan Africans. J. Acquir. Immune. Defic. Syndr. 55 (5), 536–549. [DOI] [PubMed] [Google Scholar]
  33. Swart M, Whitehorn H, Ren Y, Smith P, Ramesar RS, Dandara C, 2012. PXR and CAR single nucleotide polymorphisms influence plasma efavirenz levels in South African HIV/AIDS patients. BMC Med. Genet. 13, 112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Talal AH, Ding Y, Venuto CS, Chakan LM, McLeod A, Dharia A, Morse GD, Brown LS, Markatou M, Kharasch ED, 2020. Toward precision prescribing for methadone: determinants of methadone deposition. PLoS One 15 (4), e0231467. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Thai TT, Jones MK, Harris LM, Heard RC, 2016. Screening value of the Center for epidemiologic studies –depression scale among people living with HIV/AIDS in Ho Chi Minh City, Vietnam: a validation study. BMC Psychiatry 16 (1), 145. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Tossonian HK, Raffa JD, Grebely J, Trotter B, Viljoen M, Mead A, Khara M, McLean M, Duncan F, Fraser C, DeVlaming S, Conway B, 2007. Methadone dosing strategies in HIV-infected injection drug users enrolled in a directly observed therapy program. J. Acquir. Immune. Defic. Syndr. 45 (3), 324–327. [DOI] [PubMed] [Google Scholar]
  37. Totah RA, Allen KE, Sheffels P, Whittington D, Kharasch ED, 2007. Enantiomeric metabolic interactions and stereoselective human methadone metabolism. J. Pharmacol. Exp. Ther. 321 (1), 389–399. [DOI] [PubMed] [Google Scholar]
  38. Victorri-Vigneau C, Verstuyft C, Bouquie R, Laforgue EJ, Hardouin JB, Leboucher J, Le Geay B, Dano C, Challet-Bouju G, Grall-Bronnec M, 2019. Relevance of CYP2B6 and CYP2D6 genotypes to methadone pharmacokinetics and response in the OPAL study. Br. J. Clin. Pharmacol. 85 (7), 1538–1543. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Wesson DR, Ling W, 2003. The Clinical Opiate Withdrawal Scale (COWS).J. Psychoactive Drugs 35 (2), 253–259. [DOI] [PubMed] [Google Scholar]
  40. Wolfe D, Carrieri MP, Shepard D, 2010. Treatment and care for injecting drug users with HIV infection: a review of barriers and ways forward. Lancet 376 (9738), 355–366. [DOI] [PubMed] [Google Scholar]
  41. Wolff K, Hay AWM, Raistrick D, Calvert R, 1993. Steady-state pharmacokinetics of methadone in opioid addicts. Eur. J. Clin. Pharmacol. 44, 189–194. [DOI] [PubMed] [Google Scholar]
  42. World Health, O, 2019. Update of Recommendations on First- and Second-Line Antiretroviral Regimens (Accessed May 13, 2020). https://apps.who.int/iris/bitstream/handle/10665/325892/WHO-CDS-HIV-19.15-eng.pdf?ua=1.
  43. Wu SL, Wang SC, Tsou HH, Kuo HW, Ho IK, Liu SW, Hsu YT, Chang YS, Liu YL, 2013. Hepatitis C virus infection influences the S-methadone metabolite plasma concentration. PLoS One 8 (7), e69310. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

MAPS Supp
NIHMS1754398-supplement-MAPS_Supp.pdf (347.5KB, pdf)

RESOURCES