Chapter 2

PHARMACOKINETICS OF METHADONE AND ITS PRIMARY METABOLITE IN TWENTY OPIATE ADDICTS

J.W. de Vos¹, P.J. Geerlings², W. van den Brink³, J.G.R. Ufkes¹, H. van Wilgenburg¹

¹ Department of Pharmacology, Academic Medical Centre, University of Amsterdam, Meibergdreef 15, 1105 AZ Amsterdam, The Netherlands.

² Jellinek Centrum, Jacob Obrechtstraat 92, 1017 KR Amsterdam, The Netherlands.

³ Amsterdam Institute for Addiction Research, Jacob Obrechtstraat 92, 1017 KR Amsterdam, The Netherlands.

printed in:

European Journal of Clinical Pharmacology (1995) 48: 361-366.

Since methadone maintenance treatment (MMT) was introduced (Dole and Nyswander, 1965), many studies have been conducted to establish the pharmacokinetics of methadone. Unfortunately they did not lead to a general consensus with regard to well-defined dosage schedules related to clinical and therapeutic efficacy. This might be explained in part by the very divergent results of most of the studies performed so far. For example, plasma half-lifes of methadone in tolerant subjects under MMT have varied from 19 h (Nilsson et al., 1983) to even 75 h (Ånggård et al., 1979). Both studies were performed in closed metabolic wards. The same variability has applied for bioavailability, which has ranged from 36% to 106 % (mean 87 %) among 12 tolerant opiate addicts (Nilsson et al., 1982), also in a closed metabolic ward. Another study revealed a mean bioavailability of 79 % (range 41-99 %)(Meresaar et al., 1981). Furthermore, there has been no consensus as to whether the disposition of methadone follows a one-(Olsen et al., 1981; Denson et al., 1990; Wolff et al., 1993) or a two-(Nilsson et al., 1982; Meresaar et al., 1981; Plummer et al., 1988) compartment model. However, most of these studies were performed under divergent, hardly comparable circumstances. The pharmacokinetics in non-tolerant ("naive") subjects might differ from those in tolerant subjects (Verebely et al., 1975; Ånggård et al., 1975). Also short-term methadone treatment might give different results from long-term (steady-state) treatment (Såwe et al., 1981; Horns et al., 1975). However, it seems that determinations of plasma methadone in outdoor addicts show wider scatters in elimination half-lifes than those performed in closed metabolic wards: 17.8-63.8 h (Wolff et al., 1993) and 18.9-43 h (Nilsson et al., 1983), respectively.

The metabolism of methadone has been studied intensively. Many metabolites have been traced and identified in human urine (Pohland et al., 1971; Beckett et al., 1968). Nevertheless essential knowledge about the formation and disposition of the major metabolite of methadone, EDDP (1,5-dimethyl-3,3-diphenyl-2-ethylidene-pyrrolidine) in plasma is still lacking.

The aim of our study was to establish the steady-state pharmacokinetics of methadone and its major metabolite in 20 opiate addicts under well-controlled conditions. For this reason we developed a high-performance liquid chromatography (HPLC) method to monitor methadone and EDDP concentrations in plasma simultaneously. The design of this study enabled us to supplement incomplete or conflicting data from earlier studies.

Methods

Subjects

Twenty long-term opiate addicts, each on MMT for at least 4 months, joined our study on a voluntary basis after their written informed consent. The study, performed in a closed metabolic ward (Jellinek Centrum, Amsterdam), which completely excluded illicit methadone or other drug supplementation, involved 4 days during which a 24-h period was used for plasma sampling. Standardized meals were supplied at fixed times. All participants were habitual smokers (20-30 cigarettes per day). The participants were subjected to physical examination, including clinical blood and urine analysis. Subject characteristics are shown in Table 1. Approval was obtained from the Medical Ethics Committee (Academic Medical Centre, University of Amsterdam) before the study was started.

Table 1 Subject characteristics

sub.	sex	age	weight	dose	blood analysis							urine	comedication
no.	m/f	y	kg	mg/day	(1)	(2)	(3)	(4)	(5)	(6)	(7)	pH	+/-
1	m	29	69	70	9.2	95	18	18	71	-	6	7.5	+^a
2	f	31	62	40	9.4	90	8	6	88	13	23	5	+^b
3	m	32	70	55	10.2	80	9	10	85	-	6	6	-
4	m	24	53	40	7.6	62	21	40	76	-	29	7	-
5	m	23	58	60	9.3	79	27	21	92	51	64	6	+^c
6	f	26	49	30	8.1	75	17	25	59	10	2	5	-
7	f	32	60	50	8.1	83	18	22	43	40	-	5.5	+^d
8	m	28	78	30	8.8	95	10	8	53	14	4	5	-
9	f	39	52	50	8.8	88	71	64	66	31	34	5.5	-
10	m	39	69	65	9.1	91	19	15	66	11	2	6.5	-
11	m	24	68	70	8.4	91	34	46	43	-	5	5	-
12	f	21	68	30	7.5	76	12	11	87	17	20	5	-
13	f	31	75	60	7.8	54	8	7	45	9	8	8	-
14	f	28	52	20	8.6	73	14	16	57	14	6	7	-
15	m	30	50	10	9.6	81	8	14	57	15	3	5	-
16	m	34	83	60	9.1	103	18	26	66	20	1	7	-
17	f	31	72	225	9.3	79	16	29	80	26	14	5.5	+^d
18	f	28	75	70	8.9	84	68	132	78	41	8	5	+^e
19	m	27	67	70	8.7	89	24	18	82	24	6	6	+^d
20	m	28	72	90	8.5	73	15	11	51	13	2	5	-

Blood analysis: (1) haemoglobin (mmol·l^-1); (2) creatinine (μmol·l^-1); (3) aspartate aminotransferase (U·l^-1); (4) alanine aminotransferase (U·l^-1); (5) alkaline phosphatase (U·l^-1); (6) γ-glutamyltransferase (U·l^-1); (7) erythrocyte sedimentation rate (mm·h^-1)

Comedication: ^aMesterolon/chlordiazepoxide; ^bFloctafenine; ^cIsoniazid/rifampicin/azidothymidine; ^dChlordiazepoxide; ^eDoxepine.

Drug Administration and Blood Sampling

The 20 subjects in the study were prescribed various daily amounts of d,l-methadone-HCl (Brocades, Leiderdorp, The Netherlands), prepared as a methadone linctus of 5 mg·ml^-1, depending on the previous amount of opiate use. They received the daily oral dose (average 60 mg, range 10-225 mg methadone-HCl, see Table 1) at about 0920 hours just after the standardized breakfast. at 0, 0.5, 1,2, 4, 6, 8, 12 and 24 h thereafter blood samples (10 ml) were taken by venipuncture and placed in heparinized tubes. The blood samples were immediately centrifuged for 10 min at 1500 g. The supernatant plasma was stored frozen at -25^oC until required for analysis. Prior to analysis all samples were preventatively HIV-deactivated by incubation at 56^oC for 30 min.

Sample Preparation

Subject plasma (0.5 ml) was added with 50 μl methanol, 100 μl trihexyphenidyl-HCl (2 μg·ml^-1) as internal standard (IS) to a stoppered glass tube and alkalized with 0.5 ml 2M K₂CO₃. This mixture was extracted into 3.5 ml n-hexane by gently agitating for 90 min at room temperature. After centrifugation (1500 g for 5 min) the tube contents were chilled to -25^oC, which enabled us to separate the unfrozen upper layer (n-hexane layer) from the frozen lower layer. The n-hexane layer was evaporated to dryness under a gentle stream of dry nitrogen. The dry residue was redissolved in 100 μl KH₂phosphate buffer (25 mM, pH 2.5). In order to prepare calibration curves the same procedure was performed using plasma from healthy, non-drug-using volunteers, which was spiked with methadone-HCl and EDDP-HClO₄ (Sigma, St.Louis, USA) over a concentration range of 5 to 800 ng·ml^-1.

Analytical Equipment

The HPLC system consisted of an HP 1050 Series Quarternary Pump and Variable Wavelength Detector (Hewlett Packard, USA), a Model 7125 Sample Injector (Rheodyne, USA) fitted with a 50-μl loop and a HP 3395 Integrator (Hewlett Packard, USA) in combination with a BD41 Kipp Recorder (Kipp & Zonen, Delft, The Netherlands). Separation was performed on a Supelcosil LC-ABZ column (50x4.6 mm ID) packed with 5-μm-diameter particles and protected by a 20-mm Supelguard column (Supelco, USA). The mobile phase consisted of KH₂phosphate buffer (25 mM, pH 2.5) mixed with acetonitrile (78.5:21.5, v/v). The flow rate was set at 1.5 ml·min^-1, whit UV detection at 206 nm. The analytical procedure was performed in an airconditioned room at about 20^oC.

Calculations

The methadone and EDDP levels in the plasma of our subjects were calculated by comparing the UV-absorption values of the extracted subject samples with those of the extracted spiked plasma samples (calibration curves) using the internal standard method. Recovery values were calculated by comparing the UV absorption values of the extracted spiked plasma samples with those of the unextracted standard solutions of methadone HCl and EDDP HClO4 in KH2phosphate buffer in the concentration range 5-800 ng·ml^-1. The area under the plasma concentration - time curve during the dosing interval, AUC_(0-24h), for each subject at steady-state was calculated by the trapezoid rule (Gibaldi and Perrier, 1982). Time to peak concentration (t_max) of methadone and EDDP for each subject was the period (h:min) between the methadone administration at time 0 (0920 hours) and the peak concentration detected of methadone and EDDP, respectively. The steady-state concentration (C_ss) was calculated as the mean of the plasma concentration just before methadone administration and the plasma concentration 24 h later, just before the next methadone administration. By means of the method of residuals (Gibaldi and Perrier, 1982), values of k_a, α, ß, A en B were calculated. Using these values in MW/Pharm, an integrated computerprogram with a nonlinear curve-fitting module for the assesment of pharmacokinetic parameters (Proost and Meijer, 1992), the data from plasma concentration measurements were fitted according to the two- or three-exponential equation to calculate the correlation coefficient (r) between observed and predicted values. Additionally the secondary rate constants, i.e. k₁₀, k₁₂, and k₂₁ as well as the volumes of the central compartment (V_C), the volumes of distribution during the post-distributive phase (V_β) and the body clearances (CL) were calculated by the computer using the appropriate formula (Gibaldi and Perrier, 1982). The steady-state level in each curve was thus simulated using an imaginary loading dose calculated as: D·F·B / B-C_ss , where D is the individual daily dose (in milligrams) and F is the bioavailability which is assumed to be 0.90 (Inturissi et al., 1987).

Statistics

Data were expressed as the mean (SD) or the 95 % confidence interval as appropriate. The paired t-test was used to compare t_max values for methadone with those for EDDP. Student's t-test was used to compare methadone elimination rate constants (β) of the male subjects with those of the female subjects. Correlation analysis was performed using Spearman's rank-order correlation method.

Results

HPLC-assay validation

The determination was found to be both sensitive and specific for both methadone and its metabolite EDDP. Retention times for EDDP, IS and methadone were 2.5, 3.9 and 5.5 min respectively. Detection limits in plasma (signal-to-noise ratio at least 3) appeared to be about 4 ng·ml^-1 for EDDP and 6 ng·ml^-1 for methadone. The calibration curves for both EDDP and methadone in plasma showed linearity (r³0.998) in the concentration ranges 5-400 ng·ml^-1 and 10-800 ng·ml^-1, respectively. The calculated recovery values were 60.1 % (9.4), n = 95 for EDDP, 77.5 % (7.5), n = 97 for methadone and 80.6 % (6.2), n = 88 for IS. The day-to-day assay coefficient of variation was 6.7 % for EDDP, 7.7 % for methadone and 3.1 % for IS (n = 104).

Pharmacokinetics

Figure 1 shows a representative plasma concentration - time curve for methadone and EDDP (subject 13) after the oral ingestion of 60 mg methadone-HCl at time 0 (0920 hours). As can be seen the shape of the EDDP-curve paralleled that of the methadone curve, but at a substantially lower level. Another phenomenon observed in 19 out of the 20 methadone curves, is the rapid decrease after the peak concentration followed by a considerably slower disposition, indicating pharmacokinetics according to a two-compartment model. This was confirmed using the curve-fitting computerprogram , which preferred a two-compartment model to a one-compartment model in all these cases. However, in one case (subject 1) the pharmacokinetics of methadone was best described using a one-compartment model.

Fig 1 Plasma concentration-time curve for methadone and EDDP in subject No. 13 after the oral ingestion of 60 mg methadone HCl at time 0

The 20 data sets, including steady-state concentration (C_ss), peak concentration (C_max), time to peak concentration (t_max) and AUC_(0-24h) for methadone and EDDP, are shown in Table 2. The mean values of t_max for methadone (0220 hours, range 0106-0408 hours) differed significantly (P = 0.010, paired t-test) from those of EDDP (0149 hours, range 0057-0404 hours). The calculated ratios between AUC_(0-24h) for methadone and AUC_(0-24h) for EDDP varied from 5.9 - 44.6.

The pharmacokinetic parameters for methadone calculated by means of the residual method are shown in Table 3. The computer calculated secondary rate constants (k₁₀, k₁₂ and k₂₁), the distribution volumes (V_C and V_β) and the body clearances (CL) are also given together with the correlation coefficient (r) between observed and predicted values. The mean elimination rate constant (β) was 0.026 h^-1 (0.011) and the mean plasma half-life t_½β as calculated from β, was 31.2 h (12.4) with a 95 % confidence interval from 25.6 to 37.0 h. Differences were found between the elimination rate constants (β) of the male subjects [0.030·h^-1 (0.012), n = 11] and those of the female subjects [0.021·h^-1 (0.057), n = 9], significance level: P = 0.067 (Student's t-test).

Table 2 Analytical data sets for the 20 subjects

		Methadone determinations				EDDP determinations				Ratios
sub.	dose	Css	Cmax	tmax	AUC(0-24h)	Css	Cmax	tmax	AUC(0-24h)	AUC(0-24h)methadone /
no.	mg·kg^-1	ng·ml^-1	ng·ml^-1	h:min	mg·h·l^-1	ng·ml^-1	ng·ml^-1	h:min	mg·h· l^-1	AUC(0-24h)EDDP
1	1.01	487	699	2:32	13.08	33	53	1:27	0.81	16.2
2	0.65	305	668	1:53	11.20	15	27	0:58	0.47	23.8
3	0.79	137	324	2:19	5.05	13	42	2:19	0.45	11.2
4	0.75	305	651	2:30	10.03	8	23	1:00	0.23	44.6
5	1.03	115	323	3:02	3.60	12	50	1:26	0.61	5.9
6	0.61	287	639	2:00	9.44	15	40	0:57	0.42	22.6
7	0.83	279	531	2:28	8.96	26	65	1:40	0.86	10.5
8	0.38	65	180	2:09	2.61	5	10	2:09	0.16	16.0
9	0.96	451	755	2:36	13.69	16	34	4:02	0.49	27.8
10	0.94	282	586	1:59	8.95	25	81	1:03	0.89	10.4
11	1.03	254	570	4:04	9.52	17	39	4:04	0.54	17.6
12	0.44	191	400	1:47	6.25	11	20	1:47	0.28	22.1
13	0.80	114	279	2:02	4.07	7	37	1:00	0.37	11.1
14	0.38	88	160	2:18	2.59	9	16	2:18	0.24	10.6
15	0.20	69	124	2:08	2.03	9	15	2:08	0.20	10.4
16	0.72	224	519	2:28	7.58	11	31	2:28	0.31	24.8
17	3.13	630	1255	1:06	18.45	55	301	1:06	2.55	7.2
18	0.93	426	1108	1:20	14.28	32	89	1:20	1.08	13.2
19	1.04	278	477	4:08	8.93	17	42	2:03	0.51	17.6
20	1.25	276	703	1:58	10.54	18	68	1:58	0.82	12.8
mean	0.89	263	548	2:20	8.54	17.7	54.15	1:51	0.61	16.82

Table 4 shows the calculated correlation coefficients between the daily dose (in milligrams per kilogram body weight), the steady-state concentration (C_ss), C_ss normalized to a dose of 1 mg·kg^-1 (C_ss-norm.), the AUC_(0-24h) and AUC_(0-24h) normalized to a dose of 1 mg·kg^-1 (AUC_{(0-24h)-norm.}), slow disposition rate constant (β), elimination rate constant (k₁₀) and the body clearance (CL).

Table 3 Pharmacokinetics including the graphically calculated parameters (A, B, C_ss, ka, α, β) and the computer-calculated parameters (k₁₀, k₁₂, k₂₁, VC, Vβ, CL) with an assumed bioavailibility of 0.90; r is the correlation coefficient between observed and predicted values

sub.