Methadone adverse reaction presenting with large increase in plasma methadone binding: a case series

The use of methadone as an analgesic is on the increase [1], but it is widely recognized that the goal of predictable and reproducible dosing is confounded by considerable variability in methadone pharmacokinetics [2, 3]. The unpredictability of methadone's effects results in a high incidence of inappropriate overdosing and underdosing, which can lead to severe adverse events including sedation, respiratory depression and cardiac arrhythmias [4‐6]. Many such events, including death, occur beyond the reach of medical care or observation. As a result a complete understanding of the mechanisms underlying individual adverse events has rarely been possible. We present here a scenario in which we were able to carefully investigate the cause of an accidental methadone overexposure. To the best of our knowledge this is the first report of an association between accidental methadone overexposure and increased plasma protein binding.

We describe here the cases of two patients who experienced markedly different responses to the same dose of methadone during co-administration of letrozole. Both patients were post-menopausal Caucasian women who were among healthy volunteers participating in a clinical trial (depicted in Table 1). The protocol for this study was approved by the Indiana University School of Medicine Institutional review Board (IRB), both patients signed informed consent before participation in the trial, and all procedures were conducted in accordance with the guidelines of the Declaration of Helsinki. The trial was designed to test the hypothesis that the aromatase inhibitor, letrozole, would alter the pharmacokinetics of methadone, based on pre-clinical data indicating that methadone is metabolized by aromatase in vitro [7]. Under the trial protocol both patients received 6 mg of intravenous methadone before, and then after taking letrozole for seven days.

No interaction between methadone and letrozole had been previously described. As a result of this adverse event, the investigators immediately informed the IRB, and put the trial on hold so that the cause of this adverse event could be carefully investigated before any adjustments were made to the trial design.

Unanticipated adverse events and fatalities caused by methadone are a significant public health problem [4‐6]. In this case report, we observed a marked increase in plasma methadone concentrations and symptomatic overexposure during co-administration of letrozole to a single patient. Pharmacokinetic sampling was limited to the first 24 hours after methadone dosing during the adverse event, but large pharmacokinetic changes were obvious. Since such large intra-individual changes may have occurred in other patients, we estimated pharmacokinetic parameters in order to explore the multiple possible causes of this adverse reaction.

First, although the data appear similar to those that might be obtained as the result of a dosing error, the amount of methadone remaining (single vial) before and after this event were measured, and showed that the correct dose was used. The measured decrease in plasma f _u also makes an overdose seem impossible.

Second, the increase in methadone exposure was not due to decreased metabolism, since the parent to metabolite ratio in both plasma and urine decreased in our patient.

Third, although the observed renal clearance of methadone significantly decreased (Table 3), reduced urinary clearance was not the cause of the pharmacokinetic changes observed in our patient since the total urinary drug excreted over 12 hours was higher rather than lower (Table 3).

Fourth, a small decrease in plasma volume could contribute to the total increase in drug concentrations, but serum electrolyte concentrations were similar after both methadone doses, suggesting our patient was not substantially hypovolemic. A small decrease in plasma volume cannot explain such large pharmacokinetic changes.

Fifth, a change in transporter activity could in theory increase plasma concentrations. The transport system involved would have to counteract the rapid tissue diffusion of methadone, a lipophilic drug, by stimulating active transport of drug back into the plasma. While this is theoretically possible, methadone has not been shown to be vulnerable to transporter-mediated interactions, and no effect of letrozole on drug transporters has ever been described previously.

Sixth, we did observe a greater than fourfold decrease in the estimated V_d (Table 4), which could result in increased plasma and urine concentrations. V_d can be described using the following equation:

where ECF is the volume of the extra-cellular fluid, Φ is the tissue-binding factor, f _u is the plasma fraction unbound, and ICF is the volume of the intra-cellular fluid [8]. Relative to the terms in the second half of the equation, ECF is a tiny contributor to the total V_d, especially for a drug with a large V_d such as methadone [9]. A decrease in tissue binding would not be expected to reduce the f _u in plasma, and cannot explain the increase in plasma binding actually observed. In contrast, we observed a decrease in f _u due to increased plasma binding of methadone and its metabolites. A 3.7-fold change in f _u is of similar scale to that in V_d, and therefore can account for most of the change. This change was also accompanied by an increase in free methadone concentration from 1.36 to 3.18 ng/mL in our affected patient. The concentration of bound methadone increased from 8.72 to 78.0 ng/mL. While it is clear that the increased concentration of unbound methadone contributed to this symptomatic overexposure, the large change in bound methadone may also have contributed by making available a large reservoir of free drug for transport or diffusion across the blood brain barrier. In addition, we observed an increase in plasma protein concentrations, and this is consistent with the increase in plasma methadone binding. Of note, increased plasma protein concentration has been previously reported to result in decreased methadone f _u [10].

Methadone has been shown to bind to a number of different plasma proteins [11, 12], including α₁-acid glycoprotein (AAG), β-globulin [11] and lipoprotein fractions [12]. When we examined changes in the plasma proteome after dose D relative to dose C, we noted small increases in the concentrations of albumin and AAG, but prominent approximately twofold increases in proteins in the coagulation pathway: in concentrations of thrombin and its precursors, fibrinogen and its precursors and complement factor 1. Which of these might be responsible for the change in methadone binding observed in our patient is at present unclear, but the suggestion that changes in the coagulation pathway may result in altered drug binding is an observation that may have significance in some clinical scenarios.

The observed change in binding could not have been due to the ingestion of other drugs, since neither of the women were taking any medicine known to alter methadone binding. It is possible that letrozole caused these changes in plasma protein concentrations via its effect on estrogen concentrations [13, 14]. While few data are available on such effects of estrogen depletion, it is clear that estrogen supplementation with hormone replacement therapy can decrease the concentrations of AAG [15]. It follows that AAG may increase during therapy with an aromatase inhibitor, as was observed in this case. It is also possible that letrozole brings about these changes via an 'off-target' effect, or that the change we observed in methadone plasma binding is not due to letrozole, but due to some uncontrolled factor.

This study illustrates a novel mechanism underlying intra-individual changes in methadone pharmacokinetics and pharmacodynamics: increases in plasma binding that could result in increased effects. This mechanism might help explain the unpredictability of methadone effects. A similar mechanism may be responsible for interactions with other drugs that alter the concentrations of plasma binding proteins. The consequent changes in V_d may represent important causes of clinically meaningful drug interactions.

Written informed consent was obtained from both patients for publication of this case report and any accompanying images. Copies of the written consents are available for review by the Editor-in-Chief of this journal.