| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
BACKGROUND: Severe exacerbation of chronic neuropathic pain often requires morphine in patients already treated with drugs such as tricyclic antidepressants, carbamazepine and gabapentin. However, it is unclear if a combination of these drugs intensifies the effects of morphine on the respiratory system and, if so, whether these effects are due to pharmacokinetic or pharmacodynamic interaction. METHODS: We gave rabbits (n = 6 per group) the following drugs daily for 4 days: subcutaneous normal saline 1 mL (control); amitriptyline subcutaneously 7 mg/kg; carbamazepine orally 100 mg/kg; gabapentin subcutaneously 25 mg/kg; and all three drugs concurrently (combination). On the fifth day, morphine 5 mg/kg was given IV, and Paco2, Pao2 and pH were measured. Morphine, morphine 3-glucoronide and morphine 6-glucoronide concentrations were measured in the plasma over the 4 h period after morphine injection. RESULTS: Compared with controls, premorphine baseline Paco2 was significantly higher (P < 0.05) in the amitriptyline group. Postmorphine Paco2 was significantly higher in the amitriptyline and combination groups at all time points over the 240 min, and in the gabapentin group at 10 and 30 min after morphine injection (P < 0.05). Peak Paco2 was significantly higher in the amitriptyline group (58.4 ± 1.6 mm Hg; mean SD, P < 0.005) and in the combination group (57.4 ± 1.0 mm Hg, P < 0.02) than in the control group (50.2 ± 5.2 mm Hg). Similarly, the area under the curve of Paco2 from zero to 240 min was significantly higher in the amitriptyline and combination groups than in the control (P < 0.001). There were no significant differences among the groups in plasma concentrations of morphine and its metabolites. CONCLUSIONS: We conclude that pretreatment with amitriptyline increases morphine- induced hypercarbia through pharmacodynamic processes. The effects of carbamazepine or gabapentin were not obvious in this model.
Morphine is one of the most commonly used opioid analgesics for severe pain.1 In order to achieve better pain control in the management of chronic pain, such as neuropathic pain, morphine is used in combination with other drugs, including tricyclic antidepressants (TCA),2–5 carbamazepine,6 and gabapentin.7 Similarly, short-term use of parenteral morphine is not uncommon to control acute pain episodes in patients who are already taking these drugs. One of the most serious adverse effects of morphine is respiratory depression characterized by hypercarbia.8,9 Although a cautious approach is warranted in a combination therapy of morphine and other drugs with sedative properties, the potential pharmacokinetic/ pharmacodynamic (PK/PD) interaction between morphine and TCAs, carbamazepine and gabapentine has not been well characterized. TCAs may cause respiratory depression in patients with chronic obstructive pulmonary disease (COPD).10,11 Reduced CO2 sensitivity has also been reported in patients receiving TCAs.12 The combined effect of fentanyl and imipramine on the respiratory system has been studied in a rabbit model,13 which failed to show additive effects of imipramine. When gabapentin was used with morphine, no major increase in the frequency of side effects was shown in humans.7,11,14,15 However, these studies were based on subjective variables of side effects, and may have lacked sensitivity. One case report16 described an elderly patient with obstructive lung disease, who developed respiratory failure during gabapentin therapy. The effects of carbamazepine on morphine-induced respiratory depression are not known. In addition to the knowledge gap discussed above, the morphine-related death of a 10-yr-old girl in Toronto, who was taking the combination of amitriptyline, gabapentine, and carbamazepine,17 prompted us to examine the potential interactions. The objectives of the present study, therefore, were to determine whether the co-administration of amitriptyline, carbamazepine and gabapentin magnify morphine-induced respiratory depression in a rabbit model. Potential PK interactions were evaluated by examining plasma concentration-time profiles of morphine with and without co-administration of amitriptyline, carbamazepine, and/or gabapentin.
Animals Adult male New Zealand rabbits weighing 3 to 4 kg were kept in a controlled environment and received standard rabbit pellets and water.
Study Design Five groups of rabbits (n = 6 per group) received the different drugs twice a day for 5 days before morphine injection. The control group was given subcutaneous normal saline 1 mL twice a day. The other groups received the following drugs: amitriptyline subcutaneously 7 mg · kg–1 · dose–1; carbamazepine orally 100 mg · kg–1 · dose–1; or gabapentin subcutaneously 25 mg · kg–1 · dose–1. The combination group received all three drugs (amitriptyline, carbamazepine and gabapentin) concurrently. The doses were chosen to achieve plasma concentrations similar to human therapeutic concentrations, at least for the first 2–3 hrs of administration when morphine injection was planned. On the fifth day, morphine 5 mg/kg was given IV 2 hrs after the last dose of the corresponding drugs to minimize the effects of the administration procedures of these drugs.
Drugs
Drug Administration and Blood Sampling IV drugs were given into the marginal ear vein through a 23-g butterfly needle. Subcutaneous drugs were injected into the back. For the measurement of morphine and its metabolites 2 mL of blood was collected into a heparin-containing tube, and plasma was separated and frozen at –80°C until analysis. For gabapentin, amitriptyline and carbamazepine, 1 mL of blood was collected, and the plasma was separated and frozen at –80°C until analysis.
Main Outcome Measures
Drug Level Measurement
Laboratory Analysis For the measurement of morphine and its metabolites 2 mL of blood was collected into a heparin-containing tube, and plasma was separated and frozen at –80°C until analysis. Morphine and its metabolite were extracted as described by Gerostamoulos and Drummer18 and assayed using 0.15 mL aliquot by high performance liquid chromatography according to the method described by Meng et al.19 with some modifications. Briefly, the following chromatographic conditions were used: mobile phase: 0.05M sodium dihydrogen phosphate buffer (pH 2.1) containing 2.5 mM dodecyl sulfate and 25% acetonitrile; flow rate: 1 mL/min.; column: Prodigy ODS3 5 µm, 150 x 4.6 mm at ambient temperature. A combination of electrochemical and fluorescent detectors was used. Two standard curves 1 to 100 ng/mL and 100 to 5000 ng/mL respectively, were constructed based on the analysis of morphine, morphine-3 glucuronide, and morphine-6 glucuronide using 5 calibrators for each curve. Lower limits of quantitation for morphine, morphine-6 glucuronide and morphine 3-glucuronide were 1 ng/mL, 1 ng/mL and 5 ng/mL respectively.
Data Analysis
Amitriptyline, Carbamazepine, and Gabapentin Plasma Concentrations The plasma concentrations of amitriptyline, carbamazepine, and gabapentin on day 5 before and after morphine injection are shown in Table 1. There was no statistically significant difference among the plasma concentrations of any of the drugs when given as a single drug or a combination.
Respiratory Variables
After morphine injection, Paco2 was significantly higher in the amitriptyline and the combination groups than in the control throughout the 4 h after injection period, and in the gabapentin group at 10 and 30 min after morphine injection (Fig. 1: ANOVA with post hoc Tukey test; P < 0.05). The carbamazepine group showed higher mean Paco2 levels than the control, but statistical significance could not be reached. As shown in Figure 2A, Peak Paco2 was higher in the amitriptyline group (58.4 ± 1.6 mm Hg) and the combination group (57.4 ± 1.0 mm Hg) than in the control (50.2 ± 5.2 mm Hg) (ANOVA with post hoc Tukey test; P < 0.005 and P < 0.02 respectively). Similarly, the area under the arterial Paco2 versus time curves from zero to 120 min (AUC120Pao2: not shown) and AUC240Paco2 (Fig. 2B) were significantly higher in the amitriptyline and combination groups than the control (ANOVA with post hoc Tukey test; P < 0.001 for both groups). However, the AUC240Paco2 corrected for baseline values (Paco2 10 min before morphine injection) was not statistically different among the groups (Fig. 2C), suggesting that the observed effect of morphine on Paco2 is dependent at least partly on the baseline value.
Pao2 declined rapidly, reached the trough level of about 70% of the baseline levels at 10 minutes after morphine injection, and recovered gradually over the following 4 hrs in all groups with no statistically significant differences among the groups at any of the time points. Morphine injection did not change the pH significantly in any of the groups (data not shown).
Morphine, Morphine 3-Glucoronide and Morphine 6-Glucoronide Plasma Concentrations
We observed that rabbits pretreated with amitriptyline- containing regimens developed accentuated hypercarbia after morphine injection compared to saline pretreatment. Since the plasma disposition of morphine and its metabolites remained relatively similar among the groups, the effect of amitriptyline on morphine-induced hypercarbia was likely PD rather than PK in nature. As only 16% of morphine is bound to plasma proteins in rabbits20 (as compared to 35% in humans), it is unlikely that changes in free morphine plasma concentrations were the cause of the observed effects. In cancer patients the co-administration of oral morphine and TCA increases plasma levels of morphine by more than 5021 probably by increasing the bioavailability of morphine. In the current study morphine was given IV, which excludes the effects of bioavailability changes. In this study, post morphine AUC240Paco2 was significantly increased in amitriptyline-containing regimens. When corrected for the baseline Paco2, however, the AUC240Paco2 values were not statistically different among the treatment groups (Fig. 2C), suggesting that the observed effect was not fully independent of the baseline premorphine Paco2. In rats, chronic treatment with imipramine, another TCA, increases the density of cells expressing mu-opioid receptors in the brain.22 Although, it is not known if amitriptyline causes similar changes, this may explain a synergistic component of the effects of amitriptyline on morphine-induced hypercarbia. The exact mechanism of the mixed additive and synergistic nature of the amytriptiline-morphine interaction we observed requires further study. The effect of the combined treatment with amitriptyline, carbamazepine, and gabapentin on the baseline Paco2 and morphine-induced hypercarbia was similar to, and no more than, that of amitriptyline alone. This apparent ceiling phenomenon also awaits further verification. Adverse effects of TCAs on the respiratory system are not common but have been described in humans. In overdose, TCAs can cause pulmonary edema.23 In therapeutic doses tricyclic antidepressants induced depression of the respiratory drive in a patient with COPD and a psychiatric disorder10 and were associated with a significant reduction in CO2 sensitivity.12 In contrast to these studies, nortriptyline had no effect on physiological measures reflecting pulmonary insufficiency in patients with COPD.23 Bergman et al.13 using the rabbit tooth pulp model for pain found that rabbits treated with the combination of fentanyl and imipramine did not have higher Paco2. However, the interaction may have been obscured because Paco2 was measured only once and imipramine was given as a single IV dose in their study. When clomipramine was given on a chronic basis to rats, treated animals had a larger reduction in respiratory rate after morphine injection compared to the untreated rats.24 Taken together, these findings, including ours, suggest that chronic administration of amitriptyline may cause borderline hypercarbia, which may predispose patients to exaggerated morphine-induced respiratory depression. Our approach in this study was in vivo animal experiments, instead of a human study, due to uncertainty of the severity of the potential interactions. Our findings, therefore, must be interpreted in the context of animal experiments, which have their own advantages and disadvantages. We chose rabbits because they metabolize morphine to the major metabolites found in humans (morphine-3-glucoronide and morphine-6-glucoronide).25 Like humans, rabbits are capable of generating quaternary amonium glucuronides.26 The uridine di-phosphate-glucuronosyltransferases (the catalytic enzymes involved in glucoronation of endogenous and exogenous compounds including morphine and tertiary and quaternary amines like TCAs and carbamazepine) activity is higher in rabbits in comparison with other species27 and is inhibited by the same inhibitors as human uridine di-phosphate-glucuronosyltransferases.28 Paco2 was chosen as the primary end-point of the current study because it is more sensitive than changes in respiratory rate as a measure of respiratory depression.29 Although, it is fair to assume that PK/PD profiles of morphine are not substantially different between rabbits and humans, the validity of this notion remains unknown. In one study using adult volunteers (average weight: 74 kg), Cmax was 250 ng/mL after a single IV dose of morphine 8.8 mg, which induced 30% respiratory depression as measured by CO2 responsiveness.30 Although the morphine dose in our study was higher than the human dose and the plasma concentrations were also higher, it is important to note that the dose was chosen for sufficient sensitivity to characterize the effects of morphine-drug interaction on peak Paco2 without compromising vital function in rabbits. Amitriptyline plasma concentrations in this study were comparable to human therapeutic concentrations (100–250 ng/mL [0.36–0.90 µM]) which were shown to block the voltage-gated Na+ channel.31 The serum level-effect relationship of gabapentin in neuropathic pain has not been clearly defined, but anti-epileptic concentrations may range from approximately 12 to 120 µM.32 In our experimental model, carbamazepine did not enhance morphine-induced respiratory depression. However, unlike gabapentin and amitriptyline, carbamazepine plasma concentrations were relatively low compared to the human anti-epileptic therapeutic concentration (17–50 µM). The possibility that higher plasma carbamazepine concentrations might have an effect on Paco2 cannot be excluded. In the treatment of neuropathic pain, TCAs are the first drug of choice, followed by drugs such as gabapentine and carbamazepine.33 Although one cannot directly extrapolate the data obtained from animal models to humans, the current study suggests that patients treated with amitriptyline may be at increased risk for morphine-induced hypercarbia. We cautiously suggest that IV morphine doses be reduced with careful titration, if patients are receiving concomitant treatment with TCAs. If acute morphine injection is to be used in patients already treated with these drugs, morphine-induced respiratory depression and resultant hypercarbia may be intensified.
Drs. Kozer, Taguchi and Garcia-Bournissen received fellowship awards from the Research Institute, The Hospital for Sick Children, Toronto, Canada. Dr. Koren holds the Ivey Chair in Molecular Toxicology, University of Western Ontario. The authors thank Dr. Norman Smith and his team at St Josephs Health Centre, London, Ontario Canada for the gabapentin measurements. We also thank the staff at the Clinical Pharmacology laboratory and the animal facility at the Hospital for Sick Children for their assistance.
Accepted for publication April 15, 2008. Supported by Hospital for Sick Children.
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|