Pharmacokinetics of methadone in HIV-positive patients receiving the non-nucleoside reverse transcriptase efavirenz.

Abstract

We find that the work titled ‘The pharmacokinetics of methadone in HIV-positive patients receiving the non-nucleoside reverse transcriptase efavirenz’, authored by Clarke et al. [1], adequately describes an interaction with possible clinical significance between methadone and efavirenz. Nevertheless, from a pharmacokinetic viewpoint the study lacks the exploration or discussion of some important issues. In the introduction the authors write, ‘we designed a study to assess the effect of efavirenz, EFV, on methadone pharmacokinetics, the timing of withdrawal symptoms, and the requirement for a dose escalation in methadone’. Then, based only on Cmax (obtained by inspection of the data) and the area under the curve of plasma levels (AUC) from 0–24 h (noncompartmental analysis) Clarke et al. make prescription recommendations for efavirenz to injection drug users receiving oral methadone concurrently. The authors do not discuss any kinetic parameter that could serve to predict dosification schedules such as volume of distribution after a single dose, clearance at steady state, or half-life (in order to adjust the dosage interval). Methadone is characterized by large interindividual variation in pharmacokinetics and by a rapid and extensive distribution phase (half-life of 2–3 h) and a slow elimination phase (λz half-life of 15–60 h) [2, 3]. This alone could have important clinical implications. In a recent study [4], of steady state pharmacokinetics and pharmacodynamics in methadone maintenance patients, the only difference between individuals with strong and weak withdrawal symptoms was the significantly more rapid hourly rate of decline in the plasma concentration within the period from Cmax until the next dose. Also, no differences in methadone AUC were observed between the two symptom extremes. Clarke et al. do not discuss this possibility which could explain, at least in part, their finding of small changes (increases) required in methadone dosage after efavirenz (22%) although the AUC was reduced by over 50% (since the latter appears not be an adequate correlate of symptom severity.) Also, the pharmacokinetics of methadone in maintenance treatment for opiate users show adaptive changes. In a recent population study, changes in CL and V with the time of methadone treatment had to be considered in order to explain and predict plasma drug concentrations during continuous administration [5]. These authors attributed the time-dependence in the kinetics of methadone to an increase in CL with time due to an autoinduction of CYP3A4. The changes in V with time can reflect both up-and down-regulation of α1-acid glycoprotein (AAG), the major binding site for methadone in plasma [6, 7]. One more recent study [8] has evaluated the quantitative importance of the metabolic pathway and concludes that it is not possible to adapt the daily doses to the hepatic metabolic activity of the patients, suggesting that intestinal CYP3A4 is involved in the metabolism (efavirenz and the other antiretroviral drugs are administered orally). An effect of P-glycoprotein in intestinal absorption of methadone has also been suggested [9]. The results of Clarke et al. (Figure 1 [1]) with a lower AUC do not lead to the conclusion that the interaction described is only due to induction in the hepatic metabolism of methadone by efavirenz. A change in AUC, without primary pharmacokinetic changes, can be due to changes at the intestinal level (CYP3A4 or glycoprotein). Also, the fact of not observing appreciable changes, in this figure, in the elimination slope seems to be more in accord with this last mechanism, which is not discussed. Given that methadone is administered as a racemic mixture (R-(+)-methadone and S-(–)-methadone of which only the former shows significant opioid receptor affinity) the possibility of stereoselective interaction [10] could also be explored. Another possibility which the authors do not contemplate is the interaction with AAG binding which would lead to lower plasma levels of methadone as a result of elevated V and CL (methadone is a low extraction ratio drug). The half-life could again remain unaltered if V and CL changed proportionally [11]. Although there are no specific data about the degree of binding of efavirenz, other non-nucleoside reverse transcriptase inhibitors, such as GW420867X, are highly bound to plasma proteins (range 89% to 94%) [12]. Finally, the ranges of Cmax and AUC are very broad, even in those individuals only under methadone, which implies possibly appreciable interindividual variability in the kinetics. In this situation it is not advisable to establish dosage regimens without performing a proper population analysis and establishing subject-specific covariate models for the parameters. With these parameter estimates the authors could then perform simulations to explore the appropriate dose intervals for coadministration of oral methadone with efavirenz.