buprenorphine in human plasma by electron- capture capillary gas chromatography: application to pharmacokinetics of sublingual buprenorphine. E.Thomas ...
Clinical Chemistry 43:12 2292–2302 (1997)
Drug Monitoring and Toxicology
Subnanogram-concentration measurement of buprenorphine in human plasma by electroncapture capillary gas chromatography: application to pharmacokinetics of sublingual buprenorphine E.Thomas Everhart,* Polly Cheung, Peter Shwonek, Karen Zabel, Eileen C. Tisdale, Peyton Jacob III, John Mendelson, and Reese T. Jones
We describe a sensitive and specific method for the measurement of buprenorphine in human plasma. The method involves a structural analog as an internal calibrator, careful control of pH during sample extraction to maximize drug recovery, and back-extraction into acid followed by reextraction to eliminate endogenous interferences. After evaporation, sample residues are derivatized with heptafluorobutyric anhydride and analyzed by separation on a fused-silica polymethylsiloxane capillary column and electron-capture detection. Calibration curves were linear in the ranges 0.1–2.0 mg/L and 2.0 –20 mg/L, with within-run CVs of 9.7% at 0.1 mg/L to 5.0% at 20 mg/L, and total CVs of 15.9% at 0.1 mg/L to 6.5% at 10 mg/L. The limit of quantification was 0.1 mg/L. The method was utilized in studies to determine the absolute bioavailability of sublingual doses of 2 mg of buprenorphine in 1 mL of 300 mL/L ethanol and the bioequivalence of sublingual 8-mg tablet and 300 mL/L ethanol solution formulations. Buprenorphine, 21-cyclopropyl-7-a-[(S)-1-hydroxy-1,2,2trimethylpropyl]-6,14-endo-ethano-6,7,8,14-tetrahydrooripavine, is an analgesic with a long duration of action [1, 2], 25– 40 times more potent than morphine [2– 4], which elicits partial agonist effects at the m receptor and antagonist effects at the k receptor [5–10]. Buprenorphine has been used for the control of both acute postoperative pain and the chronic pain associated with terminal cancer [4, 6, 11–21]. Because of its relatively low abuse potential
Drug Dependence Research Center, Department of Psychiatry, University of California, San Francisco, San Francisco, CA 94143. *Address correspondence to this author at: Langley Porter Psychiatric Institute, Box CPR– 0984, University of California, San Francisco, 401 Parnassus Ave., San Francisco, CA 94143-0984. Fax 415-476-7690. Received January 27, 1997; revision accepted August 22, 1997.
[22, 23] and because it produces less respiratory depression than methadone [14, 16, 24], buprenorphine is currently in clinical trials as a pharmacotherapeutic agent for the treatment of opiate dependence [14, 17, 23, 25–31]. Because the analgesic dosage of buprenorphine in humans is low [13–15, 17–21, 32], 0.3–1 mg parenterally and 4 – 8 mg sublingually, and because the parent drug is metabolized by hepatic enzymes to the dealkylated product, norbuprenorphine , and conjugated to b-glucuronic acid , concentrations of buprenorphine in plasma often fail to exceed 1 mg/L [32, 35]. Analysis of buprenorphine concentrations in plasma is, therefore, challenging . The literature of the analytical chemistry of buprenorphine is extensive, with methods based on HPLC with ultraviolet [37–39], fluorescence [37, 40], electrochemical [41– 46], and ion spray–mass spectrometry (HPLC/ISPMS)  detection; gas chromatography (GC) interfaced with nitrogen–phosphorus , electron-capture (ECD) [33, 48, 49], and MS detectors [electron impact (EI) [34, 46, 50], chemical ionization , and tandem negative chemical ionization ]; RIA [46, 53]; radioreceptor assay ; and enzyme-linked immunoassay  applied to the analysis of buprenorphine in plasma and urine.1 HPLC, with the exception of the recently published HPLC/ ISP-MS method , appears not to possess the routine sensitivity required for the reliable measurement of buprenorphine in plasma, whereas RIA, potentially the most sensitive analytical technique, has cross-reactions with norbuprenorphine and b-glucuronide conjugates, which,
1 Nonstandard abbreviations: ISP, ion spray; MS, mass spectrometry; GC, gas chromatography; ECD, electron-capture detection; EI, electron impact; NIDA, National Institute on Drug Abuse; QC, quality control; AUC, area under the curve; and SIM, selected-ion monitoring.
Clinical Chemistry 43, No. 12, 1997
because of metabolite buildup in the plasma, presents particular problems for the measurement of the drug in chronically dosed subjects . Of the published methods, GC-MS  initially appeared to be the most suitable for our needs, but we were unable to achieve adequate sensitivity. The tandem negative chemical ionization method of Kuhlman et al.  had not been published at the time of the present work. Cone et al.  reported a GC-ECD method with a detection limit of 10 mg/L for the measurement of buprenorphine, norbuprenorphine, and demethoxybuprenorphine, the acid-catalyzed degradation product of buprenorphine , in urine and fecal extracts. The method utilizes etorphine as the internal calibrator, conversion of the analytes to the pentafluoropropionyl esters, and chromatography on a packed column. By refining the extraction and derivatization procedure of Cone et al., using a capillary column and a structural analog of buprenorphine, N-n-propylnorbuprenorphine, as an internal calibrator, we developed a method for the measurement of buprenorphine in plasma with a limit of quantification of 0.1 mg/L. This method was used to obtain pharmacokinetic data from human volunteers given intravenous and sublingual tablet and solution formulations of buprenorphine. This data is being used in support of a new drug application filed by the National Institute on Drug Abuse (NIDA) and Reckitt and Colman Products with the Food and Drug Administration for approval of the use of buprenorphine as a pharmacotherapeutic agent for the treatment of opiate addiction.
Materials and Methods Materials. Buprenorphine hydrochloride and norbuprenorphine free base were obtained from the Research Technology Branch, Division of Basic Research Resources of the NlDA. The following analytical reagent-grade materials were used: acetone, t-amyl alcohol, sodium bicarbonate, 10 mol/L sodium hydroxide, and concentrated sulfuric acid (Fisher Scientific). Toluene (Fisher Scientific) and n-butyl acetate (Aldrich) were of HPLC grade. Ethyl acetate and heptane were of ultraresianalyzed grade (J.T. Baker). Heptafluorobutyric anhydride was obtained in 1-mL prescored ampules (Pierce). All extraction tubes and GC microvial inserts were deactivated with water-soluble siliconizing fluid (Aquasil, Pierce). Synthesis of the internal calibrator, N-n-propylnorbuprenorphine. A solution of norbuprenorphine (28.4 mg, 63.1 mmol) in 11 mL of methanol:acetic acid (10:1 by vol) was treated with 0.500 mL (403 mg, 6.93 mmol) of freshly distilled propionaldehyde and the mixture allowed to stir at room temperature for 30 min. After cooling the mixture to 278 °C in a dry ice–acetone bath and slow addition of sodium borohydride (923 mg, 24.4 mmol) followed by the addition of another 10 mL of methanol, the mixture was allowed to stir for 1 h and then warm to room tempera-
ture. Evaporation of the solvent under reduced pressure was followed by cooling to 0 °C and slow addition of 20 mL of water. The pH of the aqueous phase was adjusted to 9.1 with 85% phosphoric acid and the crude product extracted with 3 3 20 mL of ethyl acetate. The combined extracts were washed with 50 mL of saturated sodium chloride, dried over anhydrous magnesium sulfate, filtered, and evaporated under reduced pressure to yield a viscous oil, which was dissolved in 5 mL of 2-propanol, cooled to 0 °C, and treated with 125 mL of 1 mol/L HCl. Addition of 50 mL of ether caused the precipitation of the hydrochloride salt, which was separated from solution, washed with fresh ether, and air-dried to yield 20.3 mg (41.2 mmol, 60%) of white, crystalline material that was used without further purification.
preparation of plasma calibrator, control, and internal calibrator solutions Buprenorphine. One stock solution of 100 mg/L buprenorphine was prepared in 0.01 mol/L sulfuric acid and diluted to a concentration of 10 mg/L. The dilution was used in the preparation of the plasma calibrators. A second 100 mg/L solution, also diluted to 10 mg/L, was used for the preparation of the plasma control solutions. Individual calibrators and controls were prepared by diluting the appropriate volume of 10 mg/L stock solution with enough blank plasma to attain a total volume of 100 mL. The solutions were then mixed well and aliquoted into single-run amounts in 1-dram, screw-capped glass vials, which were stored at 220 °C until needed. Calibrators were prepared at concentrations of 0.1, 0.2, 0.3, 0.4, 0.5, 1, 2, 5, 10, and 20 mg/L. Plasma controls were prepared at 0.1, 0.2, 0.5, 2, and 10 mg/L. Internal calibrator. A stock solution of 100 mg/L N-npropylnorbuprenorphine was prepared in 0.01 mol/L sulfuric acid, and 500 mL was diluted with 99.5 mL of 0.01 mol/L sulfuric acid to produce a solution at 500 mg/L. This solution was aliquoted into single-run amounts (about 7.0 mL per vial) in 2-dram, Teflon-lined, screwcapped vials, which were stored at 220 °C until needed. Calibration graphs. Quantification was achieved by integration of detector responses and constructing calibration curves of the response (peak-height) ratios of analyte/ internal calibrator vs amount (concentration) ratios of analyte/internal calibrator by linear regression. For this assay, the linear range was from at least 0.1 to 20 mg/L. Sample preparation. To 1 mL of plasma in a 16 3 100 mm silanized, screw-top culture tube was added 100 mL of 500 mg/L internal calibrator (N-n-propylnorbuprenorphine), 4 mL of ethyl acetate:heptane (4:1 by vol), and 0.5 mL of pH 9.13 sodium bicarbonate buffer. The tube was then capped, vortex-mixed for ;10 min, centrifuged for 20 min at 1500g, and subsequently immersed in a dry ice–acetone bath to freeze the bottom aqueous layer. The top organic
Everhart et al.: Buprenorpine measurement in plasma
layer was decanted into a clean 16 3 100 mm silanized culture tube containing 0.5 mL of 0.25 mol/L sulfuric acid, and the tube was recapped, vortex-mixed for ;10 min, centrifuged for 10 min at 1500g, and immersed in a dry ice–acetone bath to freeze the aqueous layer. The organic layer was discarded. Two milliliters of toluene:t-amyl alcohol (9:1 by vol) was added, and after the aqueous layer had thawed, the tube was recapped, vortex-mixed for ;5 min, centrifuged for 10 min at 1500g, and immersed in a dry ice–acetone bath to freeze the aqueous layer. The organic layer was discarded. Four milliliters of ethyl acetate:heptane (4:1 by vol) and enough pH 9.45 sodium bicarbonate buffer (usually 0.5– 0.9 mL) to adjust the aqueous phase to pH 9.13 were added to the tube, which was recapped, vortex-mixed for ;10 min, centrifuged for ;10 min at 1500g, and immersed in a dry ice–acetone bath to freeze the aqueous layer. The organic layer was decanted into a clean 16 3 100 silanized culture tube and evaporated to dryness under a stream of dry nitrogen. Two milliliters of dry heptane was added and evaporated to dryness. To the residue was added 100 mL of dry toluene and 50 mL of heptafluorobutyric anhydride. The tube was recapped and the mixture was vortex-mixed and allowed to stand at room temperature for at least 1 h and no longer than overnight. After uncapping, the excess reagent was evaporated to dryness at room temperature under a stream of dry nitrogen. The residue was dissolved in 20 mL of n-butyl acetate and transferred to an autosampler vial in a 200-mL silanized insert. The vial was crimp-capped and loaded onto the autosampler tray, which was water-thermostated to the ambient temperature of the laboratory, and the sample analyzed by ECD GC. Chromatography. GC analyses were performed with a Hewlett-Packard (HP) 5890 GC equipped with an HP 7673A autosampler, an HP G1223A ECD, and an HP 5895A data collection and processing system. An HP Ultra-1 fused-silica capillary column (25 m x 0.2 mm i.d.) was used, with a stationary phase of cross-linked methylsilicone gum (0.33-mm film thickness). A Merlin Microseal (Merlin Instruments) was used instead of a standard silicone–rubber septum. Helium was used as the carrier gas, with a head pressure of 200 kPa, which resulted in a flow rate of approximately 3 mL/min (37 cm/s) at 150 °C. The sample (2 mL) was injected in the splitless mode via the autosampler, with a septum purge on-time of 1 min and an injector-port temperature of 285 °C. The column-oven temperature was programmed from 150 °C (after a hold of 1 min) to 325 °C at a rate of 10 °C/min and held for another 6 min. The detector (ECD) temperature was 325 °C. Typical retention times for the heptafluorobutyryl derivatives were as follows: 19.2 min (internal calibrator), 20.4 min (buprenorphine). Before commencement of an analytical run, active sites within the chro-
matographic system were plugged with 8 –10 replicate injections of a derivatized extract of a plasma sample supplemented with 50 ng of both the internal calibrator and buprenorphine. Repeatability, precision, and accuracy. The repeatability of the method was estimated by comparing the linear regression slopes, intercepts, and correlation coefficients of the calibrator plots from individual runs. Precision and accuracy were determined from the analysis of supplemented plasma controls at five concentrations ranging from 0.1–10 mg/L. The precision of the method was expressed as the within-run and total (n 5 40)  CVs, and the accuracy as the ratio of the average calculated concentrations to their supplemented values. Long-term and freeze–thaw stability. The long-term stabilities of both plasma controls and clinical samples were determined. Replicate analyses (n 5 7) of aliquots of four different plasma control concentrations spanning the range 0.2–20 mg/L were performed and the mean calculated concentrations, CVs, and accuracies determined. After storage at 220 °C for 14 months, replicate analyses (n 5 4) of aliquots of the same four batches of plasma controls were performed and the results compared. After storage at 220 °C for 14 months, 20 previously measured and randomly selected clinical samples were reanalyzed. The original and repeat measurements were compared, the differences calculated, and the average change determined. Freeze–thaw stability was determined for five consecutive cycles of freezing and thawing. Supplemented plasma quality-control (QC) samples were prepared at 0.5 and 10 mg/L, aliquoted, and frozen at 220 °C immediately after preparation. For each freeze–thaw cycle, the samples at each concentration were thawed to room temperature, allowed to stand for 1 h, and then refrozen. After duplicate samples at each concentration had been through 1–5 cycles of freezing, all samples were thawed and analyzed for buprenorphine. Recovery. The recovery of buprenorphine from extracted plasma samples was determined at four concentrations spanning the range from 0.5 to 50 mg/L. A stock solution of buprenorphine was prepared at a concentration of 100 mg/L in methanol, and serial dilutions of an aliquot of this stock solution produced buprenorphine solutions at 10, 1, 0.1, and 0.01 mg/L. An aliquot of the 10 mg/L solution was diluted with blank plasma to yield a concentration of 50 mg/L. Serial dilutions of the 50 mg/L plasma with the appropriate volumes of blank plasma resulted in buprenorphine plasma concentrations of 10, 2, and 0.5 mg/L. Four 1-mL aliquots at each plasma concentration were extracted by the standard procedure, with the final extracts decanted into tubes previously supplemented with
Clinical Chemistry 43, No. 12, 1997
50 ng of the internal calibrator. The extracts were evaporated to dryness and derivatized by the standard method. The 0.1 and 0.01 mg/L stock solutions were used to supplement four 16 3 100 mm silanized screw-top culture tubes that had been previously supplemented with 50 ng of internal calibrator, with each of the following amounts of buprenorphine: 50, 10, 2, and 0.5 ng. The solutions were evaporated to dryness and the residues derivatized with heptafluorobutyric anhydride. GC analysis of the samples was followed by integration and calculation of the peak-height ratios of the buprenorphine/internal calibrator peaks in each chromatogram. The mean peak-height ratio for each concentration was calculated for both the extracted and unextracted samples. The ratio of averages at each concentration was considered to represent the recovery of buprenorphine at each concentration. Derivative stability. Stability of the derivatized extracts during the chromatographic run was established by using two identical and concurrently prepared sets of calibrator and QC samples. One set was chromatographed immediately; the second set was injected after sitting on the temperature-thermostated autosampler tray for 36 h. The first set was then reinjected.
pharmacokinetic studies This method has been used in human pharmacokinetic studies of the bioavailability of a sublingual 2-mg, 300 mL/L ethanol solution , and the bioequivalence of an 8-mg tablet compared with an ethanolic solution formulation. Bioavailability. Six healthy subjects, five men and one woman between the ages of 21 and 36 years, opiate experienced but non-opiate dependent by DSM-IV criteria , were tested. Written consent was obtained. In a three-session balanced design at weekly intervals, subjects were tested under the following experimental conditions: (a) buprenorphine 2 mg in 1 mL of a 300 mL/L ethanol solution held sublingually for 3 min; (b) buprenorphine 2 mg in 1 mL of a 300 mL/L ethanol solution held sublingually for 5 min; (c) buprenorphine 1 mg by intravenous infusion over 30 min. Sublingual treatments were random in sequence; the intravenous treatment was always given between the sublingual treatments. Sublingual doses were administered with a 1-mL tuberculin syringe. The buprenorphine solution was placed in the right posterior sublingual area at the base of the tongue. Dose exposure was terminated by swallowing, and subjects were instructed not to swallow until told to do so by an observer. After 3 or 5 min elapsed, subjects immediately swallowed once and thereafter were allowed to swallow ad libitum. The intravenous dose of buprenorphine was infused into a forearm vein at a rate of 1 mL/min under syringe pump control over a 30-min period. Blood samples were
obtained through an intravenous catheter placed in the opposite forearm. Blood samples (10 mL) were collected into iced, heparinized Vacutainer Tubes (Becton Dickinson) before buprenorphine administration and at 5, 20, 30, 40, 60, 90, 120, 180, 240, 300, 360, 480, 600, and 720 min postdose. Samples were centrifuged and the plasma separated and frozen at 220 °C within 30 min of blood collection. Statistical analyses were performed on areas under the curves (AUCs), peak concentration, and peak time by using the SAS Statistical Analysis System program . Bioequivalence. Six healthy male subjects between the ages of 23 and 42 years who met the same eligibility requirements as the participants in the bioavailability study gave informed written consent and were tested in a twosession, balanced crossover design approximately 7 days apart, under the following experimental conditions: (a) buprenorphine 8 g/L in 1 mL of a 300 mL/L ethanol solution held sublingually for 5 min; (b) a prototype buprenorphine 8-mg tablet held sublingually for 5 min. The liquid formulations were administered in the same manner as for the bioavailability study. The tablet was placed in the midportion of the lateral sublingual space. At 5 min after tablet dosing, the sublingual area was inspected briefly for any residual tablet. If tablet fragments were visible, the subject was instructed to continue sublingual holding and not to swallow until the tablet had subjectively completely dissolved or for an additional 5 min, whichever occurred first. If full dissolution was confirmed by visual examination, the time was recorded, the subject swallowed once, and then swallowed ad libitum. Blood samples (10 mL) were obtained through an intravenous catheter placed in the forearm of the nondominant hand. Samples were collected into heparinized Vacutainer Tubes immediately before dosing and at 0.25, 0.50, 1, 1.5, 2, 3, 4, 6, 8, 10, 12, 24, 36, and 48 h after administration. Samples were centrifuged and the plasma separated and frozen at 220 °C within 30 min of blood collection. Statistical analyses were performed by the same methodology as that used for the bioavailability study.
Results Extraction procedure. The recovery of buprenorphine and the internal calibrator was found to be a very sensitive function of pH, with the best and most consistent results obtained with an extraction pH of 9.13, corresponding to the midpoint of the tertiary-amino and phenolic pKas of 8.24 and 10.0, respectively, of buprenorphine . Therefore, the initial extraction from plasma was performed by using a saturated sodium bicarbonate buffer adjusted to pH 9.13. Before final extraction from the 0.5 mL, 0.25 mol/L sulfuric acid back extract, a pH 9.45 saturated NaHCO3 buffer was used to titrate a measured volume of 0.25 mol/L H2SO4 to pH 9.13. The appropriate volume of
Everhart et al.: Buprenorpine measurement in plasma
this buffer was then added to each tube before reextraction. Derivatization. Derivatization of buprenorphine with heptafluorobutyric anhydride results in the formation of the heptafluorobutyryl esters of the parent compound as well as its acid-catalyzed degradation products, 21cyclopropylmethyl-7-[2-(3,3-dimethyl-1-butenyl)]-6,14-endoethanotetrahydrooripavine (buprenorphine olefin) and 21cyclopropylmethyl-6,14-endo-ethano-29,39,49,59,7,8hexahydro-49,49,59,59-tetramethylfurano- [29,39:6,7]normorphide (demethoxybuprenorphine) (Fig. 1). These degradation products  result from simple elimination of water and elimination of the elements of methanol accompanied by cyclization, respectively, from the 7-a side chain of the tetrahydrooripavine nucleus. The internal calibrator, N-n-propylnorbuprenorphine, undergoes analogous elimination and cyclization. When the derivatization reaction is carried out in silanized culture tubes, the losses of analytes via this pathway are minimized, typically averaging ,10%. However, when the acylation is performed in unsilanized tubes, losses may approach 25% or more (Fig. 2). The side reaction may also be suppressed by pretreatment of the culture tubes for several hours with dilute aqueous disodium EDTA. Silanization of microvial inserts was necessary to prevent irreversible adsorption of the derivatives to the glass walls of the inserts. Chromatography. A typical chromatogram of a derivatized extract of a 1-mL blank plasma sample supplemented with 50 ng of internal calibrator and 20 ng of buprenorphine is shown in Fig. 3. Expanded chromatograms of derivatized extracts of a 1-mL blank plasma supplemented with 50 ng of the internal calibrator and 0.5 ng of
Fig. 2. Expanded chromatogram of the reaction product of heptafluorobutyric anhydride with a mixture of buprenorphine and the internal calibrator (N-n-propylnorbuprenorphine) in a silanized culture tube (A) and in an unsilanized culture tube (B). 1, 1a, and 1b are the heptafluorobutyryl esters of the internal calibrator (N-n-propylnorbuprenorphine) and its demethoxy and olefin analogs, respectively; 2, 2a, and 2b are the heptafluorobutyryl esters of buprenorphine and its demethoxy and olefin analogs, respectively.
Fig. 1. Reaction of buprenorphine with heptafluorobutyric anhydride. 1, buprenorphine; 2, 3, and 4, heptafluorobutyryl esters of buprenorphine, buprenorphine olefin, and demethoxybuprenorphine, respectively.
buprenorphine, and a 1-mL blank plasma sample are displayed in Figs. 4 and 5, respectively. No extraneous peaks that would interfere with the analysis at concentrations above the quantification limit were found in either blank plasma or in clinical samples. Additionally, a study was performed to examine the possibility of interference with the assay from some commonly abused narcotic
Clinical Chemistry 43, No. 12, 1997
Fig. 3. Full chromatogram of derivatized extract of a 1-mL blank plasma sample supplemented with 50 ng of the internal calibrator (1) and 20 ng of buprenorphine (2).
analgesics, opiate antagonists, amphetamines, anticonvulsants, antidepressants, antipyretics, benzodiazepines, hallucinogens, b-blockers, trimethylxanthines, and the major metabolite of buprenorphine, norbuprenorphine. Although most of the drugs examined yielded derivatives upon reaction with heptafluorobutyric anhydride, none presented any interference with the assay. The results are summarized in Table 1.
Fig. 5. Expanded chromatogram of a derivatized extract of a 1-mL blank plasma sample. The retention time of buprenorphine (2) is marked with an arrow.
Calibration curves. Calibration plots were broken into low (0.1–2.0 mg/L) and high (2.0 –20 mg/L) concentration ranges. The curves were linear in each range, with average coefficients of determination (r2) of 0.998 and 0.999, respectively, for the low and high ranges. Sensitivity. The limit of detection for accurate quantification was 0.1 mg/L, defined by a total CV of ,20%; however, concentrations of 0.05 mg/L could still be detected. Precision and accuracy. The within-run and total means, accuracies, SDs, and CVs of the method were determined with blank plasma samples supplemented with known amounts of buprenorphine at five different concentrations spanning the analytical range of the method, and included in single runs performed on 20 different days. For the study, two replicate samples were analyzed at each concentration. CVs were ,10% at every concentration with the exception of the 0.1 and 0.2 mg/L, where the total CVs were 15.9% and 12.5%, respectively (Table 2).
Fig. 4. Expanded chromatogram of a derivatized extract of a 1-mL blank plasma sample supplemented with 50 ng of the internal calibrator (1) and 0.5 ng of buprenorphine (2).
Long-term and freeze–thaw stability. Aliquots (n 5 7) of supplemented QC samples at four different concentrations were measured immediately after preparation. The remaining sample at each concentration was aliquoted into glass vials in single-run amounts and frozen at 220 °C. Fourteen months later the samples were thawed and remeasured (n 5 4) with no evidence of decomposition (Table 3). Twenty randomly chosen clinical samples were examined for stability in the same manner. However, statistical
Everhart et al.: Buprenorpine measurement in plasma
Table 1. GC retention times of heptafluorobutyryl derivatives of drugs processed by the buprenorphine assay procedure. Compound
Acetaminophen Amitriptyline Amphetamine Bufotenine Buprenorphine Caffeine Cocaine Codeine Desipramine Diazepam Diphenylhydantoin Doxepin Ephedrine Fentanyl Fluoxetine Haloperidol Heroin Imipramine MDA MDMA Meperidine Methadone Methamphetamine Morphine Nalorphine N-n-propylnorbuprenorphine Norbuprenorphine Nortriptyline Oxazepam Pentazocine Phenobarbital Propanolol Propoxyphene Tetrahydrocannabinol a b
Retention time, min a
No derivative b
9.87 20.75 No derivative No derivative 13.59 13.69, 14.33, 15.37 14.41 a
No derivative 4.70 No derivative 5.90, 10.85 No derivative No derivative 13.53, 13.87 6.46, 7.88 7.62, 9.00 No derivative No derivative b
13.51 14.57 19.52 18.81 14.07 No derivative 11.56 a
11.74, 13.09 No derivative a
Table 2. Precision and accuracy for the determination of buprenorphine in supplemented plasma QC samples. Known concentration, mg/L
Concentration found (mean 6 SD), mg/L
Within-run (n 5 40) 0.100 0.200 0.500 2.00 10.0 Total (n 5 40) 0.100 0.200 0.500 2.00 10.0
0.101 6 0.010 0.193 6 0.014 0.495 6 0.036 2.07 6 0.085 10.4 6 0.517
9.65 7.51 7.33 4.09 4.99
101 96.5 99 104 104
0.101 6 0.016 0.193 6 0.024 0.495 6 0.042 2.07 6 0.176 10.4 6 0.671
15.9 12.5 8.54 8.51 6.47
101 96.5 99 104 104
Stability of derivatives. Decomposition of the derivatives was found to be accelerated by the heat from the hot injector port transferred to the samples on the autosampler tray. To retard this decomposition, cold water was circulated through the sample trays throughout the course of the run. Stability was established with two identical and concurrently prepared sets of calibration and QC samples. The first set was chromatographed immediately. Twenty-four hours after completion of the first run, corresponding to 36 h after initiation of the first run, the second set of samples was chromatographed. Finally, the first set was rechromatographed and the data analyzed. No evidence of decomposition was found. Pharmacokinetic studies. Fig. 6 shows the mean concentration–time curves for buprenorphine obtained from six patients after receiving, in three different sessions approximately 7 days apart: 1 mg of buprenorphine, 30-min intravenous infusion; 2 mg of buprenorphine in 1 mL of 300 mL/L ethanol, held sublingually for 3 min before
Drug is eliminated during back-extraction into acid. Drug is lost during final evaporation step.
Table 3. Stability of frozen (220 °C) supplemented plasma QC samples.
analysis of the data indicated some decomposition, averaging 10%, over the same time frame. Supplemented plasma QC samples were prepared at 0.5 and 10.0 mg/L, aliquoted, and frozen at 220 °C immediately after preparation. The samples were put through repetitive cycles of freezing and thawing. For each freeze–thaw cycle, the samples were thawed to room temperature, allowed to stand for 1 h, and then refrozen. When duplicate samples at each concentration had been through 1–5 cycles of freezing, all samples were thawed and analyzed for buprenorphine, with no evidence of decomposition. Recovery. The recovery of buprenorphine averaged 77% over two orders of magnitude of concentration, and ranged from 71% at 10 mg/L to 85% at 0.5 mg/L (Table 4).
Concentration found (mean 6 SD), mg/L Known concentration, mg/L
0.200 0.500 2.00 10.0
1 week (n 5 7)
14 months (n 5 4)
0.210 6 0.018 0.521 6 0.011 1.96 6 0.060 10.5 6 0.276
0.199 6 0.009 0.506 6 0.011 2.12 6 0.066 11.0 6 0.260
Table 4. Absolute recovery of buprenorphine from supplemented plasma samples (n 5 4). Concentration, mg/L
0.500 2.00 10.0 50.0
Recovery (mean 6 SD), %
85.4 6 3.5 72.3 6 3.4 69.3 6 3.7 80.8 6 4.0
4.1 4.7 5.3 4.9
Clinical Chemistry 43, No. 12, 1997
Fig. 6. Plasma concentration–time profile of buprenorphine (mean 6 SD) in six human volunteers receiving 1 mg of buprenorphine intravenously for 30 min (f), 2 mg of buprenorphine in 1 mL of 300 mL/L ethanol sublingually for 3 min (Œ), and 2 mg of buprenorphine in 1 mL of 300 mL/L ethanol sublingually for 5 min (‚).
swallowing was allowed; and 2 mg of buprenorphine in 300 mL/L ethanol, held sublingually for 5 min. Table 5 lists the corresponding pharmacokinetic parameters: AUC, peak concentration, and peak time. Bioavailability from the 3-min sublingual administration, based on extrapolated AUC, was 36.5 6 13.4% (6SD). The corresponding parameter for the 5-min sublingual administration was 33.1 6 13.4% (6SD) with no significant difference in the extent of bioavailability between the two sublingual holding times. Figure 7 shows the mean plasma concentration–time curves for buprenorphine obtained from six patients after receiving, in two different sessions approximately 7 days apart: buprenorphine 8 g/L in 1 mL of 300 mL/L ethanol, held sublingually for 5 min; and a prototype buprenorphine 8-mg tablet, also held sublingually for 5 min. Table 5 displays the corresponding pharmacokinetic parameters derived from these curves. On the basis of these param-
Fig. 7. Plasma concentration–time profile of buprenorphine (mean 6 SD) in six human volunteers receiving 8 mg of buprenorphine in 1 mL of 300 mL/L ethanol sublingually for 5 min (E) and an 8-mg buprenorphine tablet sublingually for 5 min (F).
eters, we concluded that the two products differed in both rate and extent of bioavailability of buprenorphine. The relative bioavailability of buprenorphine from the prototype 8-mg tablet, when using the solution as a reference, was 58.4% 6 31.8% for extrapolated AUC. These results are discussed in detail in a separate publication .
Discussion HPLC with electrochemical detection [41– 46] and HPLC-MS  have been used for the determination of subnanogram concentrations of buprenorphine in plasma. Electrochemical, or coulometric, detection can be a difficult technique with which to detect subnanogram concentrations on a consistent basis . Recently, Traqui et al.  reported an HPLC assay for the simultaneous measurement of buprenorphine and norbuprenorphine in plasma. The method involves an atmospheric pressure, positive chemical ionization technique called HPLC/ISP-
Table 5. Estimates of bioavailability parameters. Parameter
3-min sublingual dose; 2 mg (n 5 6) (mean 6 SD) Dose A
5-min sublingual dose; 2 mg (n 5 5) (mean 6 SD) Dose B
Intravenous infusion; 1 mg (n 5 6) (mean 6 SD) Dose C
AUC (extrapolated), h z mg/L Peak concentration, mg/L Peak time, h
14.3 6 8.7 1.60 6 0.66 1.25 6 0.42
13.2 6 8.8 1.72 6 0.87 1.62 6 0.55
18.4 6 6.5 14.3 6 3.0 0.44 6 0.09
Bioavailability of an 8-mg tablet relative to that from an 8-mg solution
AUC (extrapolated) h z mg/L Peak concentration, mg/L Peak time, h
Liquid (n 5 6) (mean 6 SD) Dose A
Tablet (n 5 6) (mean 6 SD) Dose B
35.7 6 13.9 7.14 6 2.80 0.85 6 0.27
17.8 6 7.5 3.12 6 0.52 1.20 6 0.27
Everhart et al.: Buprenorpine measurement in plasma
MS, a modification of electrospray MS. The reported lower limits of detection were 0.10 and 0.05 mg/L for buprenorphine and norbuprenorphine, respectively. The first reported GC method for the analysis of buprenorphine in plasma was the GC-MS method of Lloyd-Jones et al. , which involved a complex liquid– liquid extraction of horse plasma, followed by derivatization with heptafluorobutyric anhydride for GC separation and analysis. The reported limit of quantification was 20 mg/L. Ohtani et al.  found that selected-ion monitoring (SIM) in the EI mode, even with extensive sample cleanup before derivatization with pentafluoropropionic anhydride, resulted in chromatograms with high background noise, precluding analysis of buprenorphine at subnanogram concentrations in plasma. SIM of the same derivatized extracts in the positive chemical ionization mode resulted in a method with a limit of detection of 0.2 mg/L. A GC method involving solid-phase extraction and ECD of the heptafluorobutyryl ester of buprenorphine, and claiming a limit of detection of 0.5 mg/L, was reported by Martinez et al. . This method, however, requires 2– 4 mL of plasma and still does not achieve the desired limit of quantification. Debrabandere et al. , using an extensive liquid–liquid extraction procedure, silylation of the final extract, and SIM in the EI mode, claimed limit of detection of 0.2 mg/L for buprenorphine in plasma. We initially attempted to use the GC-MS method of Blom et al. , with a reported limit of detection of 0.15 mg/L of buprenorphine in human plasma, but abandoned this method when we could not consistently reproduce the reported acid-catalyzed degradation of the 7-a side chain. Subsequently, it became apparent that this degradation was neither necessary nor advantageous for the chromatography, and that the amount of elimination and cyclization that occurs upon reaction with heptafluorobutyric anhydride could be controlled by appropriate deactivation of the culture tubes utilized for the derivatization reaction (Fig. 2). To obtain maximum sensitivity we decided to use ECD. The method of Cone et al. , developed for the analysis of buprenorphine in urine and feces, and with a reported limit of detection of 10 mg/L in urine, appeared to be a good starting point. Preliminary experiments determined that the extraction procedure described in this report provided cleaner extracts than those obtained with solid-phase extraction. The method was optimized by using N-n-propylnorbuprenorphine, a closer structural analog of buprenorphine than the etorphine used by Cone et al. The pH was maintained at the isoelectric point during the extraction procedure to maximize recovery, and silanizing the culture tubes used for derivatization minimized decomposition. Heptafluorobutyryl esters were prepared instead of the pentafluoropropionyl esters because the heptafluorobutyryl esters are more stable to hydrolysis. Microvial inserts were deactivated to prevent adsorption of the
heptfluorobutyryl derivatives of buprenorphine and the internal calibrator, which increased dramatically with time in the absence of silanization. The autosampler tray was temperature-thermostated to retard thermal degradation of the derivatives. Finally, modern capillary GC was used instead of packed-column chromatography. Implementation of these modifications resulted in a method with the precision, accuracy, specificity, and sensitivity (0.1 mg/L) required for the measurement of buprenorphine in plasma. After the completion of this work, a GC tandem (GC/MS-MS) mass-spectrometric method  was reported for the simultaneous measurement of buprenorphine and norbuprenorphine in plasma. The method involves a triple-stage quadrupole instrument in the negative chemical ionization mode. The reported limits of quantification are 0.2 mg/L for buprenorphine and 0.03 mg/L for norbuprenorphine. A separate publication  reports the use of this method to study the pharmacokinetics of intravenous, sublingual, and buccal buprenorphine. The method described here has the necessary specificity, sensitivity, precision, and accuracy for pharmacokinetic studies of buprenorphine. The extensive sample cleanup procedure is more labor-intensive than the simple extraction procedures utilized in either the HPLC/ ISP-MS  or GC/MS-MS  methods described above. The advantage of our method is that it can be performed by any laboratory that owns a capillary GC equipped with a data system, autosampler, and an ECD. The instruments required for the HPLC/ISP-MS and GC/MS-MS methods, although state of the art, require the capital investment of several hundred thousand dollars, and are thus beyond the means of most analytical laboratories. Measurement of the metabolite norbuprenorphine, which has a much lower analgesic potency than buprenorphine , was not required for our studies. However, norbuprenorphine is extracted and derivatized by the method (Table 1). Consequently, the method could be appropriately modified to measure norbuprenorphine as well.
This work was supported in part by the NIDA, contract no. N01DA-4 – 8306 and grant DA01696. We thank Kaye Welch for administrative assistance and for preparation of the manuscript.
References 1. Hambrook JM, Rance MJ. The interaction of buprenorphine with the opiate receptor: lipophilicity as a determining factor in drugreceptor kinetics. In: Kosterlitz HW, ed. Opiates and endogenous peptides. Amsterdam: North Holland, 1976:295–301. 2. Heel RC, Brogden RN, Speight TM, Avery GS. Buprenorphine: a review of its properties and therapeutic efficacy. Drugs 1979;17: 81–110. 3. Orwin JM. In: Harcas AW, Smith RB, Whittle BA, eds. Pain—new
Clinical Chemistry 43, No. 12, 1997
13. 14. 15.
16. 17. 18.
perspectives in measurements and management. Edinburgh: Churchill Livingstone, 1977:141–58. Kay B. A double-blind comparison of morphine and buprenorphine in the prevention of pain after operation. Br J Anaesth 1978;50: 605–9. Lewis JW. Ring C-bridged derivatives of thebaine and oripavine. In: Braude MC, Harris LS, May EL, eds. Narcotic antagonists: advances in biochemical psychopharmacology. New York: Raven, 1974:8:123– 6. Martin WR, Eades CG, Thompson JA, Huppler RE, Gilbert PE. The effects of morphine- and nalorphine-like drugs in the nondependent and morphine-dependent chronic spinal dog. J Pharmacol Exp Ther 1976;197:517–32. Cowan A, Lewis JW, Macfarlane IR. Agonist and antagonist properties of buprenorphine, a new antinociceptive agent. Br J Pharmacol 1977;60:537– 45. Rossano C, De Luca LF, Firetto V, Fossi F, Castiglioni CL, Carminati P. Activity and tolerability of buprenorphine after parenteral and sublingual administration. Clin Ther 1982;5:61– 8. Sade´e W, Rosenbaum JS, Herz H. Buprenorphine: differential interaction with opiate receptor subtypes in vivo. J Pharmacol Exp Ther 1982;223:157– 62. Takemori AE, Ikeda M, Portoghese PS. The mu, kappa, and delta properties of various opioid agonists. Eur J Pharmacol 1986;123: 357– 61. Adriaensen H, Van De Walle J. Clinical use of buprenorphine in chronic administration. Acta Anaesthiol Belg 1976;27:187–91. De Castro J, Parmentier P. Buprenorphine in analgesic anesthesia. Report on 200 cases. Sixth World Congress of Anaesthesiology, Mexico City, Mexico. Societe d’ Anesthesie de Charleroi, Belgium. 1976:103–55. Robbie DS. A trial of sublingual buprenorphine in cancer pain. Br J Clin Pharmacol 1979;7:315S–7S. Mello NK, Mendelson JH. Buprenorphine suppresses heroin use by heroin addicts. Science 1980;207:657–9. Bullingham RES, McQuay HJ, Moore RA. Clinical pharmacokinetics of narcotic agonist–antagonist drugs. Clin Pharmacokinet 1983; 8:332–3. Budd K. Buprenorphine. In: Bullingham RES, ed. Opiate analgesia. London: WB Saunders,1984:147–52. Reisinger M. Buprenorphine as a new treatment for heroin dependence. Drug Alcohol Depend 1985;16:257– 62. Pasqualucci V, Tantucci C, Paoletti F, Dottorini ML, Bifarini G, Belfiori R, et al. Buprenorphine vs morphine via the epidural route: a controlled comparative clinical study of respiratory effects and analgesic activity. Pain 1987;29:273– 86. Carl P, Crawford, ME, Madsen NB, Ravlo O, Bach B, Larsen AI. Pain relief after major abdominal surgery: a double-blind controlled comparison of sublingual buprenorphine, intramuscular buprenorphine, and intramuscular meperidine. Anesth Analg 1987;66: 142– 6. Lipp M, Daublander M, Lanz E. 0.15 mg Intrathecal buprenorphine applied for postoperative analgesia: a clinical double-blind study. Anaesthesist 1987:36:233– 8. Payne KA, Murray WB, Barrett HS. Intramuscular buprenorphine compared with morphine for postoperative analgesia. Afr Med J 1987;71:359 – 61. Cowan A, Lewis JW, Macfarlane IR, Whittle BA. Analgesic and dependence studies with oripavine partial agonists. Br J Pharmacol 1971;43:461–2. Jasinski DR, Pevnick JS, Griffith JD. Human pharmacology and abuse potential of the analgesic buprenorphine: a potential agent for treating narcotic addiction. Arch Gen Psychiatry 1978;35:501– 16.
24. Banks CD. Overdosage of buprenorphine: case report. N Z Med J 1979;89:255–7. 25. Mello NK, Mendelson JH, Kuehnle JC. Buprenorphine effects on human heroin self-administration: an operant analysis. J Pharmacol Exp Ther 1982;223:30 –9. 26. Blom Y, Bondesson U, Gunne LM. Effects of buprenorphine in heroin addicts. Drug Alcohol Depend 1987;29:1–7. 27. Bickel WK, Stitzer ML, Bigelow GE, Liebson IA, Jasinski DR, Johnson RE. A clinical trial of buprenorphine: comparison with methadone in the detoxification of heroin addicts. Clin Pharmacol Ther 1988;43:2–78. 28. Kosten TR, Kleber HD. Buprenorphine detoxification from opioid dependence: a pilot study. Life Sci 1988;42:635– 41. 29. Kosten TR, Kleber HD, Morgan C. Treatment of cocaine abuse with buprenorphine. Biol Psychiatry 1989;26:637–9. 30. Kosten TR, Schottenfeld R, Zedonis D, Falconi J. Buprenorphine versus methadone maintenance for opioid dependence. J Nerv Ment Dis 1993;181:358 – 64. 31. Strain EC, Stitzer ML, Liebson A, Bigelow GE. Buprenorphine versus methadone in the treatment of opioid-dependent cocaine users. Psychopharmacology 1994;116:401– 6. 32. Bullingham RES, McQuay MS, Porter EJB, Allen MC, Moore RA. Sublingual buprenorphine used postoperatively: ten hour plasma drug concentration analysis. Br J Clin Pharmacol 1982;13:665– 73. 33. Cone EJ, Gorodetzky CW, Yousefnejad D, Buchwald WF, Johnson RE. The metabolism and excretion of buprenorphine in humans. Drug Metab Dispos 1984;12:577– 81. 34. Blom Y, Bondersson U, A¨nggård E. Analysis of buprenorphine and its N-dealkylated metabolite in plasma and urine by selected-ion monitoring. J Chromatogr 1985;338:89 –98. 35. Bullingham RES, McQuay HJ, Moore RA, Bennett MRD. Buprenorphine kinetics. Clin Pharmacol Ther 1980;28:667–72. 36. Moore RA. Analysis. In: Cowan A, Lewis JL, eds. Buprenorphine: combatting drug abuse with a unique opioid. New York: Wiley-Liss, 1995:105–12. 37. Garrett ER, Chandran VR. Pharmacokinetics of morphine and its surrogates. VI: bioanalysis, kinetics, solvolysis kinetics, solubility, pKa values, and protein binding of buprenorphine. J Pharm Sci 1985;74:515–24. 38. Tebbett IR. Analysis of buprenorphine by high-performance liquid chromatography. J Chromatogr 1985;347:411–3. 39. Hackett IP, Dusci LJ, Ilett KF. Sensitive screening method for buprenorphine in urine. J Chromatogr 1986;374:400 – 4. 40. Ho ST, Wang JJ, Ho W, Hu OYP. Determination of buprenorphine by high-performance liquid chromatography with fluorescence detection: application to human and rabbit pharmacokinetic studies. J Chromatogr 1991;570:339 –50. 41. Noda J, Kojima T. Quantitative determination of buprenorphine in human plasma by high-performance liquid chromatography. Hiroshima J Med Sci 1985;34:261–5. 42. Mullersman G, Toufflin S, Derendorf H. HPLC analysis of buprenorphine in plasma and urine using coulometric detection. Pharm Biomed Sci 1987;5:303– 8. 43. Debrabandere L, Van Boven M, Daenens P. Analysis of buprenorphine in urine specimens. J Forensic Sci 1991;37:82–9. 44. Debrabandere L, Van Boven M, Daenens P. High-performance liquid chromatography with electrochemical detection of buprenorphine and its major metabolite in urine. J Chromatogr 1991;564: 557– 66. 45. Schleyer E, Lohmann R, Rolf C, Gralow A, Kaufmann CC, Unterhalt M, Hudderman W. Column-switching solid-phase trace-enrichment high-performance liquid-chromatographic method for measurement of buprenorphine and norbuprenorphine in human plasma
Everhart et al.: Buprenorpine measurement in plasma
and urine by electrochemical detection. J Chromatogr 1993;614: 275– 83. Debrabandere L, Van Boven M, Laruella L, Daenens P. Routine detection of buprenorphine in horse urine: possibilities and limitations of the combined use of radioimmunoassay, liquid chromatography and gas chromatography–mass spectrometry. Anal Chim Acta 1993;275:295–303. Tracqui A, Kintz P, Mangin P. HPLC/MS determination of buprenorphine and norbuprenorphine in biological fluids and hair samples. J Forensic Sci 1997;42:111– 4. Martı´nez D, Jurado MC, Repetto M. Analysis of buprenorphine in plasma and urine by gas chromatography. J Chromatogr 1990; 528:459 – 63. Cone EJ, Gorodetzky CW, Yousefnejad D, Darwin WD. 63Ni electron-capture gas-chromatographic assay for buprenorphine and metabolites in human urine and feces. J Chromatogr 1985;337: 291–300. Lloyd-Jones JG, Robinson P, Henson R, Biggs SR, Taylor T. Plasma concentration and disposition of buprenorphine after intravenous and intramuscular doses to baboons. Eur J Drug Metab Pharmacokinet 1980;5:233–9. Ohtani M, Shibuya F, Kotani H, Uchino K, Saitoh Y, Nakagawa F, Nishitateno K. Quantitative determination of buprenorphine and its active metabolite, norbuprenorphine, in human plasma by gas-chromatography chemical-ionization mass spectrometry. J Chromatogr 1989;487:469 –75. Kuhlman JJ Jr, Magliulo J Jr, Cone E, Levine B. Simultaneous assay of buprenorphine and norbuprenorphine by negative chemical ionization tandem mass spectrometry. J Anal Toxicol 1996; 20:229 –35. Hand CW, Baldwin D, Moore RA, Allen MC, McQuay HJ. Radioim-
munoassay of buprenorphine with iodine label: analysis of buprenorphine and metabolites in human plasma. Ann Clin Biochem 1986;23:47–53. Olley JE, Tiong GKL. Plasma levels of opioid material in man following sublingual and intravenous administration of buprenorphine: exogenous/endogenous opioid interaction? Pharm Pharmacol 1988;40:666 –7. Cone EJ, Gorodetzky CW, Darwin WD, Buchwald WF. Stability of the 6,14-endo-ethanotetrahydrooripavine analgesics: acid-catalyzed rearrangement of buprenorphine. J Pharm Sci 1984;73: 243– 6. National Committee for Clinical Laboratory Standards. The evaluation of precision performance of clinical chemistry devices, 2d ed. Tentative guidline. NCCLS document EP5–T2. Villanova, PA: 1992. Mendelson J, Upton RA, Everhart ET, Jacob P III, Jones RT. Bioavailability of sublingual buprenorphine. J Clin Pharmacol 1997;37:31–7. American Psychiatric Association. Diagnostic and statistical manual of mental disorders, DSM-IV; 1st M. B. ed. Washington, DC: American Psychiatric Association, 1994. SAS software. Statistical Analysis Systems (version 6.10). Cary, NC: SAS Institute, 1991. Kuhlman JJ Jr, Lalani S, Magliulo J Jr, Levine B, Darwin WD, Johnson RE, Cone EJ. Human pharmacokinetics of intravenous, sublingual, and buccal buprenorphine. J Anal Toxicol 1996;20: 369 –78. Ohtani M, Kotaki H, Sawada Y, Iga T. Comparative analysis of buprenorphine- and norbuprenorphine-induced analgesic effects based on pharmacokinetic–pharmacodynamic modeling. J Pharmacol Exp Ther 1995;272:505–10.