I.v. fentanyl decreases the clearance of midazolam

British Journal of Anaesthesia 1997; 79: 740–743

I. v. fentanyl decreases the clearance of midazolam

I. HASE, Y. ODA, K. TANAKA, K. MIZUTANI, T. NAKAMOTO AND A. ASADA

Summary We have examined the effect of fentanyl on the pharmacokinetics of midazolam in patients undergoing orthopaedic surgery. Thirty patients were allocated randomly to receive fentanyl 200 ␮g and midazolam 0.2 mg kg91 (fentanyl group, n:15) or placebo and midazolam 0.2 mg kg91 (placebo group, n:15) in a double-blind manner for induction of anaesthesia. Anaesthesia was maintained with nitrous oxide and isoflurane. Systemic clearance of midazolam was decreased by 30% (P:0.002) and elimination half-time was prolonged by 50% (P:0.04) in the fentanyl group compared with the placebo group. There were no differences in the distribution half-time or volume of distribution at steady state between the two groups. These findings indicate that elimination of midazolam was inhibited by fentanyl during general anaesthesia. (Br. J. Anaesth. 1997; 79: 740–743). Key words Hypnotics benzodiazepine, midazolam. Analgesics opioid, fentanyl. Enzymes, cytochrome P450. Pharmacokinetics, midazolam. Pharmacokinetics, fentanyl.

Midazolam is a widely used anaesthetic with a short duration of action.1 2 It is metabolized by the hepatic microsomal cytochrome P4503A (CYP3A),3 4 which is inhibited by various drugs, including antibiotics, antimycotics and calcium channel blocking agents, resulting in prolonged drowsiness and inhibition of psychomotor performance.5–7 Midazolam is used frequently in combination with opioids for anaesthesia and sedation. CYP3A plays a role in the metabolism of fentanyl8 and it may occur that one drug inhibits the metabolism of another drug when they are metabolized by the same P450 isozymes.9 Therefore, the metabolism of midazolam may also be inhibited by simultaneously administered fentanyl, with prolongation of the duration of action of midazolam. Although the pharmacokinetics of midazolam have been studied extensively,1 2 5–7 the effect of fentanyl on the pharmacokinetics of midazolam has not been elucidated. As midazolam frequently induces hypoxaemia and apnoea when used with fentanyl,10 it is important to examine the effect of fentanyl on the pharmacokinetics of midazolam. Our objective was to investigate the

effect of i.v. fentanyl on the pharmacokinetics of i.v. midazolam.

Patients and methods The reference substance of midazolam was a gift from Hoffmann-La Roche Ltd (Nutley, USA). Diazepam and fentanyl were purchased from Takeda Pharmaceuticals Co. Ltd (Tokyo, Japan) and from Sankyo Pharmaceuticals Co. Ltd (Tokyo, Japan), respectively. Other reagents and organic solvents were obtained from Wako Pure Chemical Industries (Tokyo, Japan). A C18-column (TSKgel ODS-120T) was obtained from the Tosoh Corp. (Tokyo, Japan). STUDY DESIGN

We used a double-blind, randomized, placebo-controlled study design with parallel groups. Based on our preliminary study, we determined that 15 patients would be needed in each group to demonstrate a 50% increase in the elimination half-time of midazolam at a level of significance of P:0.05 and power of 80%. Prolongation of the elimination halftime of midazolam inhibits psychomotor performance of patients and delays tracheal extubation after operation.5 6 After obtaining approval from the Institutional Ethics Committee and written informed consent, we studied 30 patients, ASA I, aged 17–40 yr, undergoing orthopaedic surgery. Patients were allocated randomly to receive fentanyl or placebo. All staff in the operating room and in the recovery room were unaware of the randomization code. Exclusion criteria were hepatic dysfunction, heart disease, treatment with any known inhibitor or inducer of CYP3A enzymes, and morbid obesity. Patients undergoing laparotomy were also excluded. All patients were premedicated with secobarbitone 100 mg and atropine 0.5 mg i.m., given 60 min and 30 min before induction of anaesthesia, respectively. Anaesthesia was induced with fentanyl 200 ␮g i.v.,

ICHIRO HASE, MD, YUTAKA ODA, MD, KAZUO TANAKA, MD, KOH MIZUTANI, MD, TATSUO NAKAMOTO, MD, AKIRA ASADA, MD, Department of Anaesthesiology and Intensive Care Medicine, Osaka City University Medical School, 1–5–7 Asahimachi, Abeno-ku, Osaka 545, Japan. Accepted for publication: July 22, 1997. Correspondence to Y. O.

Interaction between midazolam and fentanyl followed by midazolam 0.2 mg kg91 i.v. 1 min later (fentanyl group). Fentanyl was diluted with saline to a total volume of 10 ml and injected over 1 min. In the placebo group, the same volume of saline was administered instead of fentanyl, followed by midazolam 0.2 mg kg91. Isoflurane and 66% nitrous oxide in oxygen were used to maintain anaesthesia. Tracheal intubation was facilitated with vecuronium 0.1 mg kg91 i.v. After tracheal intubation, the lungs were ventilated to maintain end-tidal carbon dioxide concentration at 4.6–5.3 kPa. End-tidal carbon dioxide tensions and concentrations of isoflurane were measured with a gas analyser (Capnomac Ultima, Datex Instrumentarium Corp., Helsinki, Finland). The minimum alveolar anaesthetic concentration hours (MAC-h) exposure was calculated from the percent anaesthetic concentration and duration of exposure. MAC values of 1.15% were used for isoflurane.11 The trachea was extubated after operation and patients were transferred to the recovery room. BLOOD SAMPLING AND MEASUREMENT OF PLASMA CONCENTRATIONS OF MIDAZOLAM

To measure plasma concentrations of midazolam, blood samples were obtained from an indwelling arterial cannula before and 5, 10, 15, 20, 30, 45, 60, 90 and 120 min after administration of midazolam, and every 60 min thereafter until 360 min. Additional blood samples were obtained to measure plasma concentrations of fentanyl at 5, 15, 60 and 180 min after administration of midazolam. Plasma was separated within 30 min and stored at 940 ⬚C until analysis. The amount of blood collected for the study was less than 100 ml, and insignificant relative to surgical loss. Plasma concentrations of midazolam were measured by high-pressure liquid chromatography (HPLC) using a method described previously with modifications.12 Briefly, HPLC was carried out with a mobile phase consisting of 50% acetonitrile and 0.1% acetate (v/v) in distilled water at a flow rate of 1.5 ml min91 at 45⬚C, with midazolam and an internal standard, diazepam, detected at a wavelength of 250 nm. The limit of quantitation was 1 ng ml91. The coefficients of day-to-day variation were 8.6% at 30 ng ml91 (n:18) and 5.1% at 500 ng ml91 (n:17). Fentanyl was quantified by a radioimmunoassay technique as reported previously.13 14

741 was calculated by the trapezoidal rule, and other pharmacokinetic variables were calculated using standard formulae.1 All data are expressed as mean (SD). Patient characteristics and pharmacokinetic variables of the groups were compared using the Student’s t test. Data were analysed with the Statview J version 4.5 (Abacus Concepts, Inc., Berkeley, USA) and differences were considered significant if P:0.05.

Results There were no significant differences between the two groups in patient age, weight, height or duration of anaesthesia (table 1). There were no significant differences in isoflurane MAC-h between the two groups. Arterial pressure and heart rates were almost unchanged and there were no differences between the groups at any blood sampling time during the study. All patients awoke immediately after Table 1 Patient data and details of anaesthesia for the placebo and fentanyl groups (mean (SD or range) or number). Values were compared between groups using the Student’s t test. No significant differences between groups Placebo group (n:15) Sex (M/F) 10/5 Age (yr) 28 (17–40) Weight (kg) 58 (11) Height (cm) 167 (6.9) Duration of anaesthesia (min) 268 (92) Dose of isoflurane (MAC-h) 3.58 (1.25)

Fentanyl group (n:15) 9/6 28 (18–40) 60 (10) 166 (9.4) 219 (123) 3.38 (2.11)

Table 2 Pharmacokinetic variables for midazolam: area under the plasma concentration–time curve (AUC), systemic clearance (Cl), steady-state volume of distribution (Vss), distribution halftime (T1/2␣) and elimination half-time (T1/2␤). Values were compared between groups using the Student’s t test Placebo group Fentanyl group AUC (0-) (ng ml91 Cl (ml min91 kg91) Vss (litre kg91) T1/2␣ (min) T1/2␤ (min)

h) 418.5 (96.9) 8.4 (2.1) 1.2 (0.5) 17.0 (9.5) 146 (62)

645.7 (226.8) 5.9 (1.9) 1.3 (0.5) 23.3 (8.2) 217 (114)

PHARMACOKINETIC CALCULATIONS AND STATISTICS

The pharmacokinetics of midazolam were analysed with a two-compartment open model using SAAMII for the Macintosh computer (SAAM Institute, Seattle, WA, USA) as reported previously.2 The pharmacokinetics of midazolam were characterized by systemic clearance (Cl), distribution and elimination half-times (T1/2␣ and T1/2␤, respectively) and steady-state volume of distribution (Vss). Systemic clearance was defined as dose/areas under the plasma concentration–time curves (AUC). AUC

Figure 1 Plasma concentrations of midazolam after i.v. administration of midazolam (mean, SD). Anaesthesia was induced with fentanyl 200 ␮g and midazolam 0.2 mg kg91 (fentanyl group) or with saline and midazolam 0.2 mg kg91 (placebo group).

P 0.001 0.002 ns ns 0.04

742 discontinuation of inhalation anaesthesia, and the trachea was extubated. No patient developed hypoxia or hypercapnia after operation. Duration of anaesthesia was less than 360 min for 13 patients in the placebo group and for 13 patients in the fentanyl group, and blood sampling was performed in the recovery room after the end of operation. Systemic clearance, Vss, T1/2␣ and T1/2␤ for midazolam in the placebo group were similar to those reported previously.5 6 There were no differences in Vss or T1/2␣ between the placebo and fentanyl groups (table 2). In the fentanyl group, systemic clearance of midazolam was decreased by 30% (P:0.002) and T1/2␤ was prolonged by 49% (P:0.04) compared with the placebo group, suggesting that elimination of midazolam was prolonged by a single dose of fentanyl administered 1 min before midazolam (fig. 1, table 2). Plasma concentrations of fentanyl were 3.57 (1.29), 1.70 (0.54), 0.78 (0.23) and 0.48 (0.12) ng ml91 at 5, 15, 60 and 180 min after administration of midazolam (n:6), and comparable with those reported previously.15

Discussion There are many reports describing the pharmacokinetics of midazolam with or without opioid analgesics,1 2 5–7 and combined use of midazolam with such analgesics has been shown to have a supraadditive hypnotic effect and to induce respiratory depression.10 16 However, there have been no studies of the effect of fentanyl on the pharmacokinetics of midazolam. In this study, we found that fentanyl reduced significantly the clearance of midazolam in a anaesthetic procedure commonly used in clinical practice. Plasma concentrations of midazolam were less than the amnesic level at the end of operation, and no patient exhibited delayed emergence from anaesthesia. However, the prolonged effects of midazolam were reported to be a result of inhibition of clearance of midazolam,6 7 suggesting that persistent amnesia and respiratory depression may occur when midazolam is used with fentanyl. The pharmacokinetics of midazolam are influenced by agents which are metabolized by CYP3A or influence CYP3A activities, including antibiotics, antimycotics and calcium channel blocking agents.5–7 As fentanyl is also metabolized selectively by CYP3A,8 the decrease in systemic clearance of midazolam observed in our study was probably induced by competitive inhibition of CYP3A by fentanyl. Isoflurane is also metabolized by cytochrome P450, whereas CYP2E1 is involved exclusively in the metabolism of isoflurane, and isoflurane has no effect on CYP3A activity.17 18 We used secobarbitone for premedication in both groups. Although some barbiturates induce CYP3A isoforms, it takes several days to increase the level of CYP3A in the liver and to influence drug metabolism in vivo.19 20 In this study, plasma concentration of midazolam was measured within 7 h after administration of secobarbitone, suggesting that secobarbitone had no effect on the CYP3A level and pharmacokinetics of midazolam.

British Journal of Anaesthesia Metabolism of midazolam may also be influenced by hepatic blood flow, as the hepatic extraction ratio of midazolam is intermediate.1 Hepatic blood flow is decreased by hypocapnia, positive pressure ventilation and laparotomy.21 22 In our study, normocapnia was maintained and no patient underwent laparotomy. Although isoflurane may also affect hepatic blood flow,23 the mean concentration of isoflurane was comparable in the two groups. Moreover, the pharmacokinetic variables in the placebo group were comparable with those reported for healthy volunteers,5 6 suggesting that isoflurane and positive pressure ventilation had negligible effects on the pharmacokinetics of midazolam in our study. To our knowledge, there are no reports on the effect of fentanyl on hepatic blood flow in humans in vivo. In a study using animals, fentanyl had no effect on hepatic blood flow.24 Plasma concentrations of fentanyl in the fentanyl group were comparable with those after i.v. administration to volunteers without concomitant drugs.15 As wide inter-individual variations are reported for fentanyl pharmacokinetics and pharmacokinetic variables were not obtained as a result of the small number of sampling points in this study,15 the reason for this finding remains unclear. Recently, a contribution of CYP3A in the intestine to the metabolism of midazolam has been reported.25 Because midazolam undergoes first-pass elimination in the intestine, intestinal CYP3A activities should be considered when the pharmacokinetics of orally administered midazolam are analysed. In summary, we found that the clearance of midazolam was decreased by a single dose of fentanyl, probably as a result of competitive inhibition of CYP3A activity.

Acknowledgement Supported in part by a Grant-in-Aid for Research from the Ministry of Education, Science and Culture of Japan, No. 09771183.

References 1. Allonen H, Ziegler G, Klotz U. Midazolam kinetics. Clinical Pharmacology and Therapeutics 1981; 30: 653–661. 2. Persson P, Nilsson A, Hartvig P, Tamsen A. Pharmacokinetics of midazolam in total I.V. anaesthesia. British Journal of Anaesthesia 1987; 59: 548–556. 3. Kronbach T, Mathys D, Umeno M, Gonzalez FJ, Meyer UA. Oxidation of midazolam and triazolam by human liver cytochrome P450IIIA4. Molecular Pharmacology 1989; 36: 89–96. 4. Wandel C, Böcker R, Böhrer H, Browne A, Rügheimer E, Martin E. Midazolam is metabolized by at least three different cytochrome P450 enzymes. British Journal of Anaesthesia 1994; 73: 658–661. 5. Olkkola KT, Aranko K, Luurila H, Hiller A, Saarnivaara L, Himberg J-J, Neuvonen PJ. A potentially hazardous interaction between erythromycin and midazolam. Clinical Pharmacology and Therapeutics 1993; 53: 298–305. 6. Olkkola KT, Ahonen J, Neuvonen PJ. The effect of the systemic antimycotics, itraconazole and fluconazole, on the pharmacokinetics and pharmacodynamics of intravenous and oral midazolam. Anesthesia and Analgesia 1996; 82: 511–516.

Interaction between midazolam and fentanyl 7. Ahonen J, Olkkola KT, Salmenperä M, Hynynen M, Neuvonen PJ. Effect of diltiazem on midazolam and alfentanil disposition in patients undergoing coronary artery bypass grafting. Anesthesiology 1996; 85: 1246–1252. 8. Tateishi T, Krivoruk Y, Ueng Y-F, Wood AJJ, Guengerich FP, Wood M. Identification of human liver cytochrome P-450 3A4 as the enzyme responsible for fentanyl and sufentanil N-dealkylation. Anesthesia and Analgesia 1996; 82: 167–172. 9. Serlin MJ, Sibeon RG, Mossman S, Breckenridge AM, Williams JRB, Atwood JL, Willoughby JMT. Cimetidine: Interaction with oral anticoagulants in man. Lancet 1979; 2: 317–319. 10. Bailey PL, Pace NL, Ashburn MA, Moll JWB, East KA, Stanley TH. Frequent hypoxemia and apnea after sedation with midazolam and fentanyl. Anesthesiology 1990; 73: 826–830. 11. Stevens WC, Dolan WM, Gibbons RT, White A, Eger EI II, Miller RD, de Jong RH, Elashoff RM. Minimum alveolar concentrations (MAC) of isoflurane with and without nitrous oxide in patients of various ages. Anesthesiology 1975; 42: 197–200. 12. Puglisi CV, Pao J, Ferrara FJ, de Silva JAF. Determination of midazolam (Versed) and its metabolites in plasma by high-performance liquid chromatography. Journal of Chromatography 1985; 344: 199–209. 13. Michiels M, Hendriks R, Heykants J. A sensitive radioimmunoassay for fentanyl. European Journal of Clinical Pharmacology 1977; 12: 153–158. 14. Schüttler J, White PF. Optimization of the radioimmunoassays for measuring fentanyl and alfentanil in human serum. Anesthesiology 1984; 61: 315–320. 15. McClain DA, Hug CC jr. Intravenous fentanyl kinetics. Clinical Pharmacology and Therapeutics 1980; 28: 106–114. 16. Short TG, Plummer JL, Chui PT. Hypnotic and anaesthetic interactions between midazolam, propofol and alfentanil. British Journal of Anaesthesia 1992; 69: 162–167.

743 17. Kharasch ED, Thummel KE. Identification of cytochrome P450 2E1 as the predominant enzyme catalyzing human liver microsomal defluorination of sevoflurane, isoflurane, and methoxyflurane. Anesthesiology 1993; 79: 795–807. 18. Graham SG, Tateishi T, Guengerich FP, Wood AJJ, Wood M. Inhalational anesthetics inhibit CYP2E1 activity specifically. Anesthesiology 1995; 83: A342. 19. Murayama N, Shimada M, Yamazoe Y, Sogawa K, Nakayama K, Fujii-Kuriyama Y, Kato R. Distinct effects of phenobarbital and its N-methylated derivative on liver cytochrome P450 induction. Archives of Biochemistry and Biophysics 1996; 328: 184–192. 20. Nakamoto T, Oda Y, Imaoka S, Funae Y, Fujimori M. Effect of phenobarbital on the pharmacokinetics of lidocaine, monoethylglycinexylidide and 3-hydroxylidocaine in the rat: Correlation with P450 isoform levels. Drug Metabolism and Disposition 1997; 25: 296–300. 21. Hughes RL, Mathie RT, Fitch W, Campbell D. Liver blood flow and oxygen consumption during hypocapnia and IPPV in the greyhound. Journal of Applied Physiology 1979; 47: 290–295. 22. Gelman SI. Disturbances in hepatic blood flow during anesthesia and surgery. Archives of Surgery 1976; 111: 881–883. 23. Goldfarb G, Debaene B, Ang ET, Roulot D, Jolis P, Lebrec D. Hepatic blood flow in humans during isoflurane–N2O and halothane–N2O anesthesia. Anesthesia and Analgesia 1990; 71: 349–353. 24. Gelman S, Dillard E, Bradley EL jr. Hepatic circulation during surgical stress and anesthesia with halothane, isoflurane, or fentanyl. Anesthesia and Analgesia 1987; 66: 936–943. 25. Thummel KE, O’Shea D, Paine MF, Shen DD, Kunze KL, Perkins JD, Wilkinson GR. Oral first-pass elimination of midazolam involves both gastrointestinal and hepatic CYP3A-mediated metabolism. Clinical Pharmacology and Therapeutics 1996; 59: 491–502.