Chapter 4

PATTERNS OF CRAVING AND PHARMACOKINETICS IN LONG-TERM OPIATE ADDICTS IN METHADONE MAINTENANCE TREATMENT

J.W. de Vos¹, H. van Wilgenburg¹, W. van den Brink², C.D. Kaplan³ and M.W. de Vries³

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

² Amsterdam Institute for Addiction Research, Jacob Obrechtstraat 92, 1017 KR Amsterdam, The Netherlands.

³ International Institute for Psycho-social and Socio-ecological Research, University of Maastricht, Parallelweg 46, 6221 BD Maastricht, The Netherlands.

Printed in:

Addiction Research (1996) 3: 285-295.

Methadone maintenance treatment (MMT) is the most widely used (pharmaco)therapy for opiate addicts in the world today (Ling et al., 1994). The effectiveness of MMT has been extensively shown in the last two decades. In large-scale studies both client and program characteristics have been shown to influence treatment course and outcome (e.g. DARP: Simpson and Sells, 1982; TOPS: Hubbard et al., 1989; Ball and Ross, 1991). An important and much debated program characteristic is dosage schedule. From the start of MMT, high dosing of methadone (80-120 mg/day) has been propagated to achieve the therapeutic goals, i.e. suppression of withdrawal symptoms, reduction of opiate use and treatment retention (Dole and Nyswander, 1965; Dole, 1994). A number of studies support the claim that high dosages (> 50 mg/day) are more effective than low dosages (< 30 mg/day) in terms of illegal heroin use (Caplehorn et al., 1993; Strain et al., 1993a; Maremmani et al., 1994), treatment retention (Strain et al., 1993b; Caplehorn et al., 1994), and mortality (Caplehorn et al., 1994). On the other hand, there have been studies that did not find a clear dose-response relationship (Handal and Lander 1976; Maddux et al., 1991). In an extensive literature study comparing several dosage schedules, Hargreaves concluded that in 10% of the clients an oral dose of 50-100 mg methadone per day had a clear advantage over a lower dose in terms of treatment retention. The author, however, added that for a large group of clients a dose of 30 mg per day is also effective (Hargreaves, 1983). In summary it seems that dose-response studies do not allow the use of a universal dose-schedule or a single threshold value: most clients perform best on a daily methadone dose of 80, plus or minus 20 milligrams (Parrino, 1993), but some clients show a positive response with a relatively low methadone dose, whereas some other clients may show negative responses despite an adequate dose.

In order to control for individual pharmacokinetic differences in dose-response studies, a number of studies have been performed that use plasma methadone levels to predict treatment course and outcome (Horns et al., 1975; Holmstrand et al., 1978; Tennant et al., 1984; Bell et al., 1990; Loimer and Schmid, 1992). Similar to the dose-response studies, the results are not consistent: minimal plasma methadone levels vary from 100 ng/ml (Bell et al., 1988), through 150 ng/ml (Loimer and Schmid, 1992) to 200 ng/ml (Holmstrand et al., 1978), and many studies report unexplainable outliers who show high plasma methadone levels but poor performance (Horns et al., 1975; Tennant et al., 1984; Loimer and Schmid, 1992). For example, Loimer and Schmid (1992) report observing clients with clear withdrawal/abstinence symptoms and plasma methadone levels higher than 600 ng/ml.

The inconsistencies in the reported dose-response and plasma-response studies call for pharmacodynamic studies in order to explain poor performance in MMT clients in terms of client dissatisfaction and/or craving. Continued craving despite adequate methadone treatment can easily lead to client dissatisfaction, client versus therapist conflicts and finally to (increased) illegal drug use, treatment drop-out and negative treatment outcomes. As a consequence, the reduction of craving and withdrawal symptoms has high priority in the clinical management of MMT clients and, therefore, unexplained cravers are at the center of attention in methadone research.

The present study aims at a detailed description of the relationship between dosage and pharmacokinetic parameters of methadone on the one hand and a recently developed, dynamic measure of craving using the Experience Sampling Method (ESM) on the other in a group of long-term opiate addicts.

Methods

Sample

A consecutive series of 20 long-term opiate addicts was recruited from the closed ward of the Crisis Observation and Detoxification Department of the Jellinek Clinic in Amsterdam after having given written informed consent. All clients who were asked to participate entered the study with the exception of one client who refused due to the reluctance to intravenous blood sampling. Table 1 presents a description of the sample. The sample consisted of 11 males and 9 females with a mean age of 30 years (SD 4.5 years). Time in MMT ranged from 4 months to 13 years with a mean of 6.9 years. Methadone dose ranged from 10 to 225 mg/day with a mean of 60 mg/day. The daily methadone dose was tapered according to a reduction program in six clients. Urinalysis showed positive results for illegal opiates, cocaine and benzodiazepines in 16 (80%), 12 (60%) and 10 (50%) cases respectively.

Table 1 Client characteristics of the sample

sub.	sex	age	time in MMT	methadone use in last month	current dose‡	duration current dose	previous dose	tapering	drugs in urine*						medication°	medical status
no.	m/f		years	days	mg/d		mg	mg/d	Op	Qa	Bb	Bz	Am	Co
1	m	29	7	30	70	6 months	50		+	-	-	+	-	+	a	impotency
2	f	31	5	9	40	3 days	25		+	-	-	-	-	+	b	infection of apex dentis
3	m	32	12	30	55	1 day	90	5	+	-	-	+	-	-	-	subcutaneous abcess
4	m	24	9	20	40a	1 day	50	10	+	-	-	+	-	+	-	subcutaneous abcess
5	m	33	4.25	30	60	9 months	0		+	-	-	+	-	+	c	HIV+, tbc, HSV II
6	f	26	3	30	30	6 months	20		+	-	-	-	-	-	-	bronchitis
7	f	32	6.5	21	50	2 years	15		+	-	-	+	-	-	d	-
8	m	28	7.5	5	30	5 days	0		+	-	-	-	-	-	-	-
9	f	39	1	30	50	1 year	0		+	-	-	-	-	+	-	-
10	m	39	10.5	30	65b	1 day	75		-	-	-	+	-	-	-	-
11	m	24	7	6	70	6 days	0		+	-	-	-	-	+	-	-
12	f	21	0.33	30	30	1 day	35	5	-	-	-	-	-	-	-	subcutaneous abcess
13	f	31	6	30	60c	6 days	75		+	-	-	-	-	+	-	pregnancy
14	f	28	6	25	20	1 year	?		+	-	-	+	-	+	-	-
15	m	30	6	20	10	1 day	25	15	+	-	-	-	-	+	-	-
16	m	34	9	27	60	1.5 year	0		+	-	-	-	-	+	-	-
17	f	31	10	30	225	1 day	250	25	-	-	-	+	-	-	d	-
18	f	28	13	30	70	1 year	60		+	-	-	+	-	+	e	-
19	m	27	9	30	70	1 day	75	5	-	-	-	+	-	-	d	-
20	m	28	6	30	90	6 months	60		+	-	-	-	-	+	-	arthritis

‡Current dose: a - 20 mg at 11:23 and 20 mg at 12:30; b - schedule last 4 days: 80, 100, 75, 65 mg/day; c - dubious compliance (no methadone in urine). *Drugs in urine: Op - opiates; Qa - methaqualone; Bb - barbiturates; Bz - benzodiazepines; Am - amphetamines; Co - cocaine. °Medication: a - mesterolon/chlordiazepoxide; b - floctafenine; c - isoniazid/rifampicin/azidothymidine; d - chlordiazepoxide; e - doxepine.

Table 1 also provides a summary of the medical status of the sample. Because liver diseases and co-medication can influence terminal half-life of methadone (Kreek et al., 1980; Novick et al., 1981), liver enzyme profiles were established (aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, γ-glutamyltransferase) and co-medication was recorded. Liver enzymes were elevated in four cases, but the mean elimination half-life of methadone in these patients (28.6 hours) was not significantly different from the total group (31.2 hours). Co-medication was used by seven clients. Urine pH was measured to establish its influence on the total body methadone clearance (Baselt and Casarett, 1972; Bellward et al., 1977, Nilsson et al., 1982; Kell, 1994). The findings are reported in the results section.

Assessments

During the study period of four days, craving was assessed 10 times per day and a period of 24 hours was used for plasma sampling. Various daily amounts of d,l-methadone-HCL linctus (Brocades, Leiderdorp, The Netherlands) were prescribed by an independent physician.

Pharmacokinetics

The daily oral methadone dispensing time averages 0920 hours (range 0725 - 1230 hours). In one case (no. 4), the total daily methadone dose was given in two portions; the first at 1130 hours, the second at 1230 hours. Including two samples just before and after oral ingestion, 8 to 9 blood samples (10 ml) were taken by venipuncture into heparinized tubes during the next 24 hours. The blood samples were immediately centrifuged during 10 min at 1500 g. The supernatant plasma was frozen at -25^oC and stored until required for analysis. Prior to analysis all samples were preventative HIV-deactivated by incubation at 56^oC for 30 min. The HPLC blood samples analysis and the kinetic calculations have been published elsewhere (de Vos et al., 1995). Plasma methadone trough level is defined as the methadone concentration just before the next methadone dose is taken, which indicates the lowest plasma concentration in a 24 hour dispensing schedule.

Craving

For measurement of the individual daily craving levels the Experience Sampling Method (ESM) instrument was used. ESM has been developed to measure mental disorders in their natural setting (Csikszentmihalyi and Larson, 1987; de Vries, 1987). At random moments (n = 10) during the day (from 0800 to 2200 hours) a signal from a SEIKO^© wrist watch terminal prompted a self-report. A "craving module" in the ESM instrument consisting of six seven-point Likert scale items was scored by clients. Factor analysis with varimax rotation produced a single factor (Kaplan, 1992). The loadings of each item in a factor were used to construct a scale by multiplying the raw item score by the respective item loading. The scores in the sample range from a low craving (score = -0.25) to high craving (score = 20.57). Craving scores for each client were calculated by averaging the scores across a maximum of 40 beeps. The fluctuation of craving over the course of the day was plotted by averaging the 10 daily measurements from all 4 days on a diurnal time axis.

Statistical Analysis

Statistical analysis was performed using SPSS-PC (Norusis, 1988) and Excel for Windows. Descriptive statistics were calculated at both the subject and the group level. Pearson correlations (r) were used to express the relationship between methadone dose and pharmacokinetic parameters. Spearman rank correlations (r_s) were used to describe the relationship between the various pharmacokinetic parameters and craving.

Results

Pharmacokinetics

Figure 1 Client 10 methadone concentration and craving level over the course of the day

Fig. 1 shows the plasma methadone concentration for a single client (no. 10). This curve was representative for the individual curves found for all clients. During the first 20 minutes after oral methadone administration, a flat line can be observed indicating the passage through stomach and intestines. This is followed by a rapidly ascending line indicating methadone absorption in the blood. Maximum plasma levels are reached approximately 2.3 hours (range 1-4 hours) after methadone administration. The decay of the methadone concentration appears in two phases: a first rapid (distribution) phase, and a second slow (elimination) phase. This indicates the existence of a second compartment (tissue) where methadone is stored. The pharmacokinetic parameters show a wide individual variation. The elimination half-life varied between 13 and 53 hours (mean 31 hours). The volume of distribution during the postdistributive phase ranged from 1.87 to 7.95 l·kg^-1 (mean 4.03 l·kg^-1). The total body clearance, expressed as milliliter of blood which are cleared of methadone per minute and per kilogram body weight, varied from 0.76 to 4.26 ml·min^-1·kg^-1 (mean 1.64 ml·min^-1·kg^-1). No significant influence of urine pH on total body clearance was found. As a result of the large interindividual differences in methadone pharmacokinetics in this group, the weight-corrected methadone dose was only weakly correlated with plasma methadone trough levels (r = 0.50, P£ .025).

Craving

The total number of responses to the 10 daily signals of the wrist watch was 492, i.e. 24.6 (SD 5.5) responses per client or 6.2 (SD 2.4) responses per client per day. The mean craving for all clients amounted to 3.8 (SD 2.5; range 0.2 - 8.6). High average craving levels (ESM-score > 6.0) were observed in five and low craving levels (ESM-score < 6.0) in the remaining 15 clients (Table 2).

Table 2 Client craving, methadone dose and trough concentrations

sub.	craving	methadone dose	Ctrough
no.	ESM	mg/day	ng·ml^-1
11	8.58	70	254
19	7.52	70	278
12	7.46	30	191
20	6.56	90	276
5	6.48	60	115
8	5.89	30	65
10	4.49	65	282
18	4.46	70	426
16	4.42	60	224
7	3.82	50	279
3	2.73	55	137
1	2.60	70	487
6	2.36	30	287
9	1.96	50	451
4	1.53	40	305
13	1.43	60	114
15	1.40	10	69
14	1.38	20	88
2	0.85	40	305
17	0.15	225	630

Several daily craving patterns could be distinguished. In 15 cases (75%), there was a daily high peak craving level between 0800 and 1000 hours (Figure 1). In 13 of them the peak appeared just before the methadone dispensing time. The client who received methadone in two portions (1130 and 1230 hours) also showed the 0900 hours craving peak. In addition to the morning craving peak, 13 cases (65%) showed a - somewhat lower - craving peak around noon (Figure 1). Finally, 11 cases (55%) showed a significant increase in craving between 1400 and 2200 hours (2.8 to 4.8; P < .05).

Methadone and Craving

Figure 2 shows the correlations between craving and oral methadone dose for those clients with and without high dose regimes. The relationship between daily methadone dose and average craving showed a negative trend; craving seemed to decrease with increasing methadone dosage (r_s = -.33; P = .16). However, when only "normal" therapeutic dosages for MMT were considered (10-90 mg/day), a significant positive relationship emerged; higher craving levels were associated with higher methadone dosages. (r_s = .55; P = .015).

A clear relationship between craving levels and plasma methadone concentrations over the day could be observed in only five of the 20 cases (Figure 3). Apart from the early morning and noon peaks, most of the other craving curves did not show much fluctuation. As a consequence, no significant correlation was found between craving levels and plasma methadone concentrations (r_s = -.22; P = .36).

Figure 2 Correlation between craving and oral methadone dose for with and without high dose methadone

Figure 3 Client 16 methadone concentration and craving level over the course of the day

Discussion

In an attempt to discover some potential reasons for negative treatment outcomes among clients in MMT, the present study investigated the relationship between methadone dose and plasma levels and heroin craving among 20 long-term opiate addicts. In the underlying model, heroin craving (and abstinence symptoms) is considered to be an intermediate step in clinical outcome; craving can lead to client discomfort and client-therapist conflicts which in turn can cause (increased) heroin use and treatment drop-out. Assuming the existence of a negative relationship between methadone plasma levels and the level of craving, one would expect a lowering of the clients discomfort with increasing plasma methadone levels. In the present study, however, plasma methadone trough levels (ranging from 65 to 630 ng·ml^-1) were not significantly correlated with levels of craving. This finding is consistent with studies reporting MMT-clients with high levels of discomfort due to abstinence symptoms or illegal heroin use despite adequate levels of plasma methadone (Horns et al., 1975; Bell et al., 1990; Loimer and Schmid, 1992). In an attempt to explain this phenomenon, several authors have pointed to the co-occurrence of psychiatric disorders in "poor performers" (Treece and Nicholson, 1980; Roszell and Calsyn, 1986; Maremmani et al., 1993). In our study, however, no significant correlations were observed between heroin craving and general psychopathology or alcohol abuse/dependence (de Vos et al., 1997). It is, therefore, unlikely that a differential distribution of psychiatric disorders between high and low cravers is responsible for the absence of a significant correlation between plasma methadone levels and craving.

A significant positive correlation (r_s = .55; P = .015) was demonstrated between oral methadone dose and craving if the client with the extreme high dose of 225 mg/day was excluded from the analysis, indicating that higher average craving levels are associated with higher doses. This combination of continued high craving and high methadone dosages has also been reported by others (Whitehead, 1974; Horns et al., 1975; Goldstein et al., 1975; Bell et al., 1990; Loimer and Schmid, 1992). It seems that clients with high levels of craving stimulate or even force their therapists to provide them with higher dosages of methadone with no clear effect on the initial levels of craving.

The average craving level was considered to be low in 15 out of the 20 cases. However, high early morning (0800 - 1000 hours) and noon craving peaks were observed in 13 clients. In all but one of these cases, the early morning peak immediately preceded the methadone dispensing time, suggesting an anticipatory reflex as its causal mechanism (Childress et al., 1986; Powell et al., 1992; Robinson and Berridge, 1993). Similarly, the noon peak can be interpreted as a conditioned anticipatory reflex related to the ambulatory methadone dispensing which usually takes place around noon. An alternative explanation could be a conditioned anticipatory reflex related to an old drug taking schedule in which the client typically takes his first portion of heroin around noon. In one case methadone was given in two portions around noon (1130 and 1230 hours). An early morning craving peak was also present in this client; a finding that can be interpreted as an anticipatory craving reflex related to the regular methadone dispensing time in the clinic.

The average craving level was considered high in five of the 20 clients. With the exception of one client (no. 12) with a daily methadone dose of 30 mg, they all received maintenance doses of 60 to 90 mg/day. One of the high craving clients (no. 5) who was HIV-positive and suffered from an active form of tuberculosis took a combination of isoniazid, rifampicin and azidothymidine in addition to his daily dose of 60 mg of methadone. The specific co-medication might explain the relatively low plasma trough concentration (115 ng·ml^-1) and the relatively high craving level (ESM-score = 6.5) in this particular client. In the remaining three clients, the high average craving levels cannot be explained by low methadone plasma trough levels (254 to 276 ng·ml^-1), short plasma methadone half-life (21.7 to 25.7 hours) or other pharmacokinetic factors. It should be noted, however, that the client with the highest craving level (no. 11) took methadone only six days in the last month prior to admission to the clinic. His urine was positive for both heroin and cocaine. His infrequent participation in the MMT and the additional use of illegal drugs might explain his high level of craving despite his adequate plasma methadone trough level (254 ng·ml^-1). The second highest craver (no. 19) was a well stabilized MMT-client who was tapering his methadone dose during the study period. The reduction of his dosage by 5 mg/day might be (partly) responsible for his high level of craving. This leaves one client without any compromising circumstance (no. 20): a man of 28 years old who is in MMT for the last six years and took his maintenance doses of 90 mg/day all 30 days in the last month and had a methadone plasma trough level of 276 ng·ml^-1. As in most clients his urine was positive for both heroin and cocaine. These findings are in general agreement with those of Loimer and Schmid (1992) who reported some patients with plasma concentration greater than 600 ng·ml^-1 with high levels of craving.

The study presented here has both strengths and weaknesses. The most important weaknesses are the small and heterogeneous sample recruited from a rather specific treatment setting and the lack of true treatment outcome parameters. The strength of the study is the application of a dynamic measure of craving using, the Experience Sampling Method, that enables the construction of daily craving curves that can be related to plasma methadone concentration curves. Through this novel method we were able to identify the early morning and noon craving peaks in the majority of this heterogeneous sample of long-term opiate addicts.

The present study does not support the simple use of plasma methadone trough levels in clinical dosage adjustment in MMT clients with high levels of discomfort and/or craving. In individual clients MMT-status (e.g. stability, dose reduction), additional drug use, co-medication and medical status should also be taken into account. However, more than anything else, craving seems to be related to anticipatory conditional responses in reaction to environmental cues such as regular dispensing time. Together with plasma levels, these cues should be investigated and discussed with the client in order to obtain optimal dosage schedules and maximum treatment effectiveness.