Hbp

Ka

b

a

ARRAA

KBNGM

1

aapt[iaadas

nb

aa

sU

1h

Journal of Chromatography B, 939 (2013) 23– 31

Contents lists available at ScienceDirect

Journal of Chromatography B

j ourna l h om epage: www.elsev ier .com/ locate /chromb

igh-sensitivity analysis of buprenorphine, norbuprenorphine,uprenorphine glucuronide, and norbuprenorphine glucuronide inlasma and urine by liquid chromatography–mass spectrometry�

aren J. Reginaa, Evan D. Kharascha,b,∗

Department of Anesthesiology, Division of Clinical and Translational Research, Washington University in St. Louis, St. Louis, MO, United StatesDepartment of Biochemistry and Molecular Biophysics, Washington University in St. Louis, St. Louis, MO, United States

r t i c l e i n f o

rticle history:eceived 18 July 2013eceived in revised form 30 August 2013ccepted 1 September 2013vailable online 8 September 2013

a b s t r a c t

A new method using ultra-fast liquid chromatography and tandem mass spectrometry (UFLC–MS/MS)was developed for the simultaneous determination of buprenorphine and the metabolites norbuprenor-phine, buprenorphine-3�-glucuronide, and norbuprenorphine-3�-glucuronide in plasma and urine.Sample handling, sample preparation and solid-phase extraction procedures were optimized for maxi-mum analyte recovery. All four analytes of interest were quantified by positive ion electrospray ionization

eywords:uprenorphineorbuprenorphinelucuronideass spectrometry

tandem mass spectrometry after solid-phase microextraction. The lower limits of quantification in plasmawere 1 pg/mL for buprenorphine and buprenorphine glucuronide, and 10 pg/mL for norbuprenorphineand norbuprenorphine glucuronide. The lower limits of quantitation in urine were 10 pg/mL for buprenor-phine, norbuprenorphine and their glucuronides. Overall extraction recoveries ranged from 68–100% inboth matrices. Interassay precision and accuracy was within 10% for all four analytes in plasma and within15% in urine. The method was applicable to pharmacokinetic studies of low-dose buprenorphine.

. Introduction

Buprenorphine is an opioid used for decades in treating acutend chronic pain [1–3]. A transdermal formulation was recentlypproved for treating moderate-severe chronic pain. Buprenor-hine sublingual tablets or films are also used for opioid addictionherapy [4], and for reducing addiction-related infectious diseases5]. Clinical advantages of buprenorphine in addiction treatmentnclude approval for use by private (non-specialized) practitioners,nd a narrow dose range enabling rapid dose titration [5]. Reporteddvantages in pain therapy include a ceiling effect for respiratoryepression but not analgesia at clinically relevant doses, fewerdverse events than other opioids, and the rarity of withdrawalymptoms [6–10].

Buprenorphine is extensively metabolized, primarily toorbuprenorphine, and both also undergo glucuronidation, touprenorphine-3�-glucuronide and the secondary metabolite,

� National Institutes of Health grants R01-DA025931 and K24-DA00417 (to EDK)nd UL1-TR000448 (to the Washington University in St. Louis Institute of Clinicalnd Translational Sciences).∗ Corresponding author at: Department of Anesthesiology, Washington Univer-

ity in St. Louis, 660 S Euclid Ave, Campus Box 8054, St. Louis, MO 63110-1093,nited States. Tel.: +1 314 362 8796; fax: +1 314 362 8571.

E-mail address: kharasch@wustl.edu (E.D. Kharasch).

570-0232/$ – see front matter © 2013 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.jchromb.2013.09.004

© 2013 Elsevier B.V. All rights reserved.

norbuprenorphine-3�-glucuronide (Fig. 1) [11,12]. Plasmametabolite concentrations can approximate or exceed those of theparent drug, and relative norbuprenorphine, buprenorphine-3�-glucuronide, and norbuprenorphine-3�-glucuronide exposures,based on molar area under the plasma concentration vs timecurves, are 200%, 100%, and 600% those of buprenorphine [13].These buprenorphine metabolites are pharmacologically active inanimals [14–17]. Metabolism therefore may potentially constitutea bioactivation pathway, although the role of these metabolitesin mediating buprenorphine effects in humans is unknown.Additional pathways of buprenorphine metabolism have alsobeen recognized [18]. Clinical disposition and metabolism ofbuprenorphine are subject to well-described drug interactions[19–21].

Methods for quantification of buprenorphine and one or moremajor metabolites in plasma and/or urine include immunoassay[22,23], gas chromatography with various detectors [24–28], HPLCwith single quadrupole or tandem quadrupole mass spectrometry[13,29–35], and most recently UPLC methods with mass spectrom-etry [36–40]. Various methods of sample preparation have alsobeen described, including protein precipitation [40], liquid/liquidextraction [39], and solid-phase extraction [37,41,42]. The most

sensitive of the above assays generally have a lower limit ofquantification of approximately 0.1 ng/mL for buprenorphine and0.25 ng/mL for norbuprenorphine, and the glucuronide metabolitesare often not analyzed.

24 K.J. Regina, E.D. Kharasch / J. Chromatogr. B 939 (2013) 23– 31

ine, b

rsdtooitwii2dimpdstmaf

apgapl

2

2

naffoi

Fig. 1. Structures of buprenorphine and the metabolites norbuprenorph

Increasing interest in the clinical use of buprenorphine,ecognition of metabolites formation and their potential clinicalignificance, use of lower buprenorphine doses for pain, and theesire to study buprenorphine drug interactions in healthy volun-eers (at much lower doses than used in patients for prevention ofpioid withdrawal) have accentuated interest in analytical meth-ds with greater sensitivity for buprenorphine and metabolitesn both plasma and urine. Buprenorphine is commonly adminis-ered sublingually at 4–32 mg/d for addiction maintenance therapy,ith steady-state plasma buprenorphine concentrations averag-

ng 5–10 ng/mL, and 2–16 mg intravenous buprenorphine resultsn peak buprenorphine and norbuprenorphine concentrations of0–130 and 0.5–3.5 ng/mL respectively [43]. In contrast, loweroses (0.15–0.2 mg intravenously, 2 mg sublingually) must be used

n non-dependent healthy volunteers [44]. Assuming linear phar-acokinetics, 0.2 mg/kg intravenously would have an expected

eak plasma concentration (Cmax) of approximately 2 ng/mL,eclining to approximately 0.005 ng/mL after 72 h, and a 2 mgublingual dose gives an expected Cmax of 1.6 ng/mL [45]. Withransdermal buprenorphine patches delivering 5, 10 and 20 �g/h

aximum plasma buprenorphine concentrations were 176, 191nd 476 pg/mL, respectively [10]. These studies exemplify the needor assay sensitivity.

Therefore, the purpose of this investigation was to developnd validate an analytical method for the analysis of buprenor-hine and the metabolites norbuprenorphine, buprenorphine-3�-lucuronide, and norbuprenorphine-3�-glucuronide, in plasmand urine, with greater sensitivity, particularly over the expectedlasma concentration range of low pg/mL to low ng/mL. A target

imit of quantification of 1 pg/mL was established.

. Experimental

.1. Reagents

Buprenorphine, buprenorphine-d4, norbuprenorphine,orbuprenorphine-d3, buprenorphine-3�-glucuronide (B3G)nd norbuprenorphine-3�-glucuronide (N3G) were all purchased

rom Cerilliant Corporation (Round Rock, TX). Phosphoric acid,ormic acid, glacial acetic acid, methanol and acetonitrile were allbtained from Sigma-Aldrich (St. Louis, MO). Ammonium hydrox-de was from JT Baker (Center Valley, PA). Water was filtered by a

uprenorphine-3�-glucuronide and norbuprenorphine-3�-glucuronide.

Milli-Q water filtration system, Millipore (Billerica, MA). Outdatedhuman plasma and human urine were obtained from the localuniversity hospital. Oasis MCS 96 well microextraction plates formethod development purposes were kindly donated by Waters(Milford, MA).

2.2. Instrumentation

LC–MS/MS analysis was performed on an ultra-fast liquidchromatography system from Shimadzu Scientific Instruments(Columbia, MD) consisting of a CMB-20A system controller, twoLC-20ADXR pumps, a DGU-20A3 degasser, a SIL-20AC autosampler,and a CTO-20A column oven. An external Valco switching valve wasinstalled between the chromatography system and the mass spec-trometer. The chromatography system was coupled to an API 4000QTrap LC–MS/MS linear ion trap triple quadrupole tandem massspectrometer from Applied Biosystems/MDS Sciex (Foster City, CA).

2.3. Solutions

Methanol stock solutions of buprenorphine, norbuprenorphine,B3G and N3G (100 �g/mL) from Cerilliant Corporation werestored at −20 ◦C. Composite 1 �g/mL standard solutions containingbuprenorphine, norbuprenorphine, B3G, and N3G were preparedin duplicate in methanol from the 100 �g/mL stock solutions (CS1and CS2). CS1 was utilized to make the working standard solu-tions for the standard curve and CS2 was used to make the workingstandard solutions for the quality control samples. Composite stan-dards (10 �g/mL) were prepared and used to make urine calibratorsand quality control samples (UCS1 and UCS2) which were preparedin a similar manner to the plasma standards. Internal standardsolutions were made by spiking 100 ug/mL methanol stocks ofbuprenorphine-d4 and norbuprenorphine-d3 into 5% phosphoricacid to a final concentration of 0.5 ng/mL of each compound. Thissolution was found to be stable for several months stored at 4 ◦C.The solid-phase extraction (SPE) elution solution was acetonitrile,methanol and 30% ammonium hydroxide (12:8:1).

2.4. Standard and quality control preparation

Working plasma standards were prepared in methanol at thefollowing concentrations of buprenorphine, norbuprenorphine,

. Chro

Bt2as2ccoTwtt0

ctea00mHab0

2

oepam

Mcott2ufpswp

2

paar(a(grtrpp

K.J. Regina, E.D. Kharasch / J

3G and N3G by spiking the appropriate amount of CS1 or dilu-ions of CS1 into methanol: 0, 0.02, 0.06, 0.2, 0.6, 2, 6, 20, 60 and00 ng/mL. Working urine standards were prepared in methanolt the following concentrations of the four analytes of interest bypiking methanol with UCS1 or dilutions of UCS1: 0.2, 0.6, 1, 6,0, 60, 200, 600, 2000 and 6000 ng/mL. Working plasma qualityontrol solutions were prepared in methanol at the following con-entrations of all four analytes by spiking the appropriate amountf CS2 or dilutions of CS2 into methanol: 0.12, 1.2 and 12 ng/mL.hese working standard solutions were verified to be stable for foureeks when stored at −20 ◦C. Working urine quality control solu-

ions were also prepared in methanol for all for analytes by spikinghe appropriate amount of UCS2 or dilutions of UCS2 into methanol:.12, 1.2, 12, 120 and 1200 ng/mL.

Standards and quality control samples were prepared in pooledontrol plasma or urine from 5 anonymous donors. To a glass cul-ure tube containing 475 �l of control plasma or urine, 25 �l ofach working standard or working quality control solution wasdded. The final concentrations of standards were 0.001, 0.003,.01, 0.03, 0.1, 0.3, 1, 3 and 10 ng/mL for plasma and 0.01, 0.03,.1, 0.3, 1, 3, 10, 30, 100 and 300 ng/mL for urine. The final low,id and high quality control (QC) concentrations (LQC, MQC andQC respectively) were 0.001, 0.1, and 10 ng/mL for buprenorphinend buprenorphine glucuronide and 0.01, 0.3 and 10 ng/mL for nor-uprenorphine and norbuprenorphine glucuronide in plasma, and.06, 0.6, and 60 ng/mL for all analytes in urine.

.5. Sample preparation procedure

Plasma or urine (420 �l) from calibration standards, QC samples,r subject samples was pipetted to a clean glass culture tube. Anqual volume of internal standard mix, 0.5 ng/mL of each buprenor-hine d4 and norbuprenorphine d3 in 5% phosphoric acid, wasdded to each tube. Samples were capped and mixed on a vortexixer for approximately 30 s.Solid-phase microextraction was then performed on an Oasis

CX 96 well �extraction plate. The plate was prepared by firstonditioning the plate with 200 �l of methanol followed by 200 �lf MilliQ water. Next, a 750 �l aliquot of the plasma sample washen passed through the plate under vacuum as recommended byhe manufacturer (10–15 mm Hg). The plate was then washed with00 �l of 0.2% acetic acid, followed by 200 �l of methanol, againnder vacuum. The plates were thoroughly dried under vacuumor 2 min. The analytes were eluted into a clean 96 well collectionlate by vacuum with two separate additions of 25 �l of SPE elutionolution (acetonitrile, methanol, ammonium hydroxide). Eluentsere then transferred to autosampler vials with glass inserts andlaced in the autosampler tray.

.6. Analytical method

Chromatographic separation of buprenorphine, norbuprenor-hine, B3G, N3G and the internal standards buprenorphine-d4,nd norbuprenorphine-d3 was achieved using a Kinetix core shellnalytical column (100 × 2.1 mm, 2.6 �m) coupled with a Secu-ity Guard ULTRA cartridge UHPLC C18 (2.1 mm) guard cartridgePhenomenex, Torrance, CA). The flow rate was 0.7 mL/min with

mobile phase of 0.1% formic acid in water (A) and acetonitrileB). The column was equilibrated with 10% B for 0.5 min, a linearradient to 85% B was applied over 6.5 min. The column was thenamped linearly to 100% B over 0.1 min, held at 100% B for 0.5 min,

hen reverted back to 10% B over the next 0.3 min and allowed toe-equilibrate at 10% B for 2.2 min. The total run time was 10 miner sample. The column oven was kept at 55 ◦C and the autosam-ler cooler was maintained at 15 ◦C. The injection volume was 15 �l.

matogr. B 939 (2013) 23– 31 25

The same chromatographic conditions were utilized for plasma andurine samples.

The mass spectrometer was equipped with an electro-spray ion source which was operated in positive ion multiplereaction monitoring (MRM) mode for the detection of all ofthe analytes and internal standards. The [M+H]+ transitionswere optimized for each analyte and were selected as fol-lows: m/z 468 → 363 and m/z 468 → 101 for buprenorphine,m/z 414 → 115 and m/z 414 → 101 for norbuprenorphine, m/z644 → 396 and m/z 644 → 468 for buprenorphine-3-glucuronide,m/z 590 → 414 and m/z 590 → 101 for norbuprenorphine-3-glucuronide, m/z 472 → 101 for buprenorphine-d4 and m/z417 → 101 for norbuprenorphine-d3. The first mass pair forbuprenorphine, norbuprenorphine, B3G and N3G was used forquantitation and the second mass pair was used for qualitativeidentification. The mass spectrometer settings for the declusteringpotential (DP), the collision energy (CE) and the collision cell exitpotential (CXP) were optimized for each of these transitions. Theentrance potential (EP) was optimized at 10 V for all of the transi-tions. Buprenorphine was used to optimize the source temperatureand gas settings through flow injection. The optimal source tem-perature was 600 ◦C with the selected mobile phase and flow rate.Nitrogen was used for all gasses and the curtain gas was set at35 psig, CAD gas set to high, Gas 1 40 psig and Gas 2 60 psig.

Raw data were acquired and processed using Analyst softwareversion 1.5.2 (Applied Biosystems/MDS Sciex, Foster City, CA). Cal-ibration curves of peak area ratio vs analyte concentration were fitusing linear least-squares analysis and 1/x2 weighting.

2.7. Method validation

The validation was performed as per the United States Foodand Drug Administration (US-FDA) Bioanalytical Method Vali-dation Guidance [46]. The following assessments were made todemonstrate the rigor of this assay: recovery, accuracy and pre-cision, selectivity, stability and matrix effect. The method was thendeemed rigorous enough to analyze sub-clinical doses of buprenor-phine that had been dosed in healthy volunteers.

2.7.1. RecoveryRecovery of buprenorphine, norbuprenorphine, B3G and N3G

from plasma or urine was determined by comparing a sample ofpure compound with internal standard in solvent at the nominalhigh, medium or low concentrations, to the same concentration ofsample in plasma or urine extracted as described above.

2.7.2. Accuracy and precisionThe precision and accuracy for each analyte was determined at

a low, mid and high concentration levels in human control plasmaor urine over three separate batch runs. Precision was determinedas the coefficient of variation (%CV), calculated by dividing thestandard deviation by the mean. Accuracy was determined as thepercent relative error, calculated by subtracting the theoretical con-centration from the observed mean, and dividing by the theoreticalconcentration.

2.7.3. SelectivityChromatography of buprenorphine and metabolites in pooled

blank plasma and urine was optimized to achieve separation frompotential interferences. Signal to noise ratios were greater than 5:1

at the limit of quantification for each analyte, as recommended bythe FDA guidance [46]. No carryover was observed when a blankplasma sample or blank urine sample was injected directly follow-ing the highest calibration standard.

26 K.J. Regina, E.D. Kharasch / J. Chromatogr. B 939 (2013) 23– 31

Max. 1546.7 cps.XIC of +MRM (17 pairs): 468.327/101.100 Da fro...

6.05.55.04.54.03.53.02.52.0Time, min

0

200

400

600

800

1000

1200

1400

1547

Inte

nsity

, cps

BuprenorphineLOQ 0.001 ng/mL

Max. 280.0 cps.XIC of +MRM (17 pairs): 644.328/468.400 Da from...

6.05.55.04.54.03.53.02.52.0Time, min

0

100

200

300

400

500

600

Inte

nsity

, cps

Buprenorphine glucuronideLOQ 0.001 ng/mL

Max. 1946.7 cps.XIC of +MRM (17 pairs): 590.309/414.300 Da fro...

6.05.55.04.54.03.53.02.52.0Time, min

0

500

1000

1500

2000

2333

Inte

nsity

, cps

Norbuprenorphine glucuronideLOQ 0.01 ng/mL

Max. 1.7e4 cps.XIC of +MRM (17 pairs): 414.267/115.200 Da from...

6.05.55.04.54.03.53.02.52.0Time, min

0.0

2000.0

4000.0

6000.0

8000.0

1.0e4

1.2e4

1.4e4

1.6e4

1.7e4

Inte

nsity

, cps

Norbuprenorphine LOQ 0.01 ng/mL

Fig. 2. Representative MRM chromatogram of a human plasma standard containing each analyte of interest at the assay limit of quantification (0.001 ng/ml buprenorphine,0 0 ng/ma

2

mtfc

2

pieg

3

3

tR6t

.01 ng/ml norbuprenorphine, 0.001 ng/ml buprenorphine-3�-glucuronide, and 0.01nd the red trace is the analyte at the limit of quantification.

.7.4. StabilityStability of the spiked plasma and urine samples was deter-

ined at the low and high quality control concentrations afterhe samples were held at room temperature for 24 hr, after onereeze/thaw cycle (−20 ◦C), and after three −20 ◦C freeze/thawycles.

.7.5. Matrix effectMatrix effect was studied by comparing pure compound sam-

les prepared in mobile phase, to extracted quality control samplesn plasma. The chromatographic separation was optimized tonsure that any potentially interfering peaks would be chromato-raphically separated from the peak of interest.

. Results and discussion

.1. Method development

Various mobile phases and HPLC columns were evaluated for

heir usefulness. Initially investigated was a Zorbax Extend C18apid Resolution C18 HT 2.1 × 100 mm, 1.8 �m column with a00 bar limit (Agilent, Santa Clara, CA) and a mobile phase of ace-onitrile and 10 mM ammonium bicarbonate in water (pH 9.0). The

l norbuprenorphine-3�-glucuronide). In each panel, the blue trace is blank plasma,

small particle size was chosen to give sharper and taller peaks andtherefore improved sensitivity. However, interferences with nor-buprenorphine prevented a limit of quantitation below 50 ng/mLfor this analyte. Also, upon method validation, extended use of theammonium bicarbonate mobile phase with the extracted plasmasamples required frequent column washing to keep the operatingpressure within the recommended operating range. The mobilephase was switched to an ammonia/methanol mobile phase toassess if that would provide a lower LOQ for norbuprenorphineand a more robust column life and operation. The pressure againwas very high after each individual run requiring extensive columnwashing to regain adequate operating pressures. The Kinetix coreshell column ultimately used afforded the requisite selectivity, andhad a slightly larger particle size than the Zorbax extend column(2.6 �m vs 1.8 �m), necessitating less extensive column washingthan was necessary with the smaller particle size Zorbax column.

The mass spectrometer settings were optimized for each of theanalytes dissolved in methanol. Flow injection allowed optimiza-tion of the system gas settings and temperature with a mobile phase

system of 0.1% formic acid in water and either methanol or acetoni-trile, both with and without added formic acid at various flow rates.Buprenorphine and norbuprenorphine were used to optimize thegas and temperature settings.

K.J. Regina, E.D. Kharasch / J. Chromatogr. B 939 (2013) 23– 31 27

Max. 266.7 cps.XIC of +MRM (17 pairs): 468.327/101.100 Da from...

6.05.55.04.54.03.53.02.52.0Time, min

0

100

200

300

400

500

600

700

773

Inte

nsity

, cps

BuprenorphineLOQ 0.01 ng/mL

Max. 613.3 cps.XIC of +MRM (17 pairs): 644.328/468.400 Da from...

6.05.55.04.54.03.53.02.52.0Time, min

0

1000

2000

3000

4000

5000

5893

Inte

nsity

, cps

Buprenorphine glucuronideLOQ 0.01 ng/mL

Max. 2.1e4 cps.XIC of +MRM (17 pairs): 414.267/115.200 Da from...

6.05.55.04.54.03.53.02.52.0Time, min

0.0

2000.0

4000.0

6000.0

8000.0

1.0e4

1.2e4

1.4e4

1.6e4

1.8e4

2.0e42.1e4

Inte

nsity

, cps

NorbuprenorphineLOQ 0.01 ng/mL

Max. 5386.7 cps.XIC of +MRM (17 pairs): 590.309/414.300 Da fro...

5.04.54.03.53.02.52.01.51.0Time, min

0

1000

2000

3000

4000

5000

5880

Inte

nsity

, cps

Norbuprenorphine glucuronideLOQ 0.01 ng/mL

F each

p quant

tXpqnpti

rsw0wttAnt

buh

ig. 3. Representative MRM chromatogram of a human urine standard containinganel, the blue trace is blank plasma, and the red trace is the analyte at the limit of

Extraction procedures which were initially assessed all utilizedraditional solid-phase extraction techniques with either a StrataC plate (Phenomenex, Torrance, CA) or a Varian Bond Elute Plexalate (Agilent, Santa Clara, CA). These systems did not give ade-uate sample concentration to achieve the target limit of detectionecessary due to many processing steps, drying down, and transferrocedures which increased the loss of compound at each consecu-ive transfer step. This was most notable for buprenorphine, whichs known to stick to plasticware.

The final microextraction procedure was optimized for highecovery of the analytes from plasma and its applicability to urineamples was then confirmed. The microextraction procedure stepith the greatest effect on percent recovery was the wash step with

.2% acetic acid. Various acids and strengths of acid were tested,ith a more dilute acid giving a greater recovery than more concen-

rated acids. Optimization of recovery was of great importance ashe concentrations in subject samples was expected to be very low.lso, minimization of steps and elimination of the dry down stepsecessary with traditional solid-phase extraction plates increasedhe recovery of all analytes.

Sample concentration achieved by using microextraction, com-ined with a 2.6 �m HPLC column that could be run at pressuresp to 600 bar (thereby decreasing peak width and increasing peakeight), along with careful optimization of the mass spectrometer,

analyte of interest at the assay limit of quantification (0.01 ng/ml for all). In eachification.

resulted in a limit of quantitation in plasma for buprenorphineand buprenorphine glucuronide of 1 pg/mL and 10 pg/mL fornorbuprenorphine and norbuprenorphine glucuronide.

Fig. 2 shows representative chromatograms for buprenor-phine and metabolites in plasma at the limit of quantification,as well as chromatograms of blank plasma. Fig. 3 shows rep-resentative chromatograms for buprenorphine and metabolitesin urine at the limit of quantification, and chromatograms ofblank plasma. Calibration curves were linear over the concen-tration range 0.001–10, 0.01–10, 0.001–10, and 0.01–10 ng/mLin plasma for buprenorphine, norbuprenorphine, buprenorphineglucuronide and norbuprenorphine glucuronide, with typicalcorrelation coefficients (r2) of 0.999, 0.998, 0.999, and 0.997,respectively. Calibration curves in urine were linear over therange 0.01–100 ng/mL for all four analytes, with typical correla-tion coefficients of 0.997, 0.998, 0.998 and 0.997 for buprenorphine,norbuprenorphine, buprenorphine glucuronide and norbuprenor-phine glucuronide, respectively.

3.2. Method validation

Recovery data for the four analytes in plasma and urine are pro-vided in Tables 1 and 2, respectively. Plasma recovery ranged from92 to 104% for buprenorphine and buprenorphine glucuronide,

28 K.J. Regina, E.D. Kharasch / J. Chromatogr. B 939 (2013) 23– 31

Table 1Analyte recovery from plasma.

Analyte Theoretical concentration (ng/mL) Recovered concentration (ng/mL)* Recovery (%) CV (%)

Buprenorphine 0.001 0.00096 ± 0.00011 96.4 11.80.1 0.0966 ± 0.0058 96.6 6.0

10 10.4 ± 0.77 104 7.3

Norbuprenorphine 0.01 0.00815 ± 0.00064 81.5 7.90.3 0.247 ± 0.032 82.4 12.8

10 10.9 ± 1.7 109 16.0

Buprenorphine Glucuronide 0.001 0.00092 ± 0.00012 92.5 12.40.1 0.098 ± 0.010 98.2 10.5

10 9.45 ± 0.66 94.5 6.9

Norbuprenorphine Glucuronide 0.01 0.0068 ± 0.0006 68.4 8.70.3 0.252 ± 0.063 84.2 24.9

10 8.04 ± 2.00 80.4 25.0

* Mean ± SD (N = 3).

Table 2Analyte recovery from urine.

Analyte Theoretical concentration (ng/mL) Recovered concentration (ng/mL)* Recovery (%) CV (%)

Buprenorphine 0.06 0.0647 ± 0.0052 108 8.16 6.2 ± 0.41 103 6.5

60 53.5 ± 2.0 89.1 3.8

Norbuprenorphine 0.06 0.0636 ± 0.0029 106 4.56 6.54 ± 0.42 109 6.4

60 57.7 ± 3.5 96.2 6.1

Buprenorphine glucuronide 0.06 0.0594 ± 0.0082 99.0 13.86 5.63 ± 0.74 93.9 13.1

60 60.3 ± 5.9 100 9.8

Norbuprenorphine glucuronide 0.06 0.0627 ± 0.0050 105 8.06 6.08 ± 0.56 101 9.2

60 61.6 ± 5.3 103 8.6

* Mean ± SD (N = 3).

Table 3Accuracy and precision in plasma.

Buprenorphine Norbuprenorphine Buprenorphine glucuronide Norbuprenorphine glucuronide

Intra-assay precision and accuracyLow 99.9 ± 3.3 99.9 ± 7.4 99.0 ± 6.6 100.4 ± 13.0Medium 104.7 ± 7.7 102.8 ± 8.0 96.0 ± 12.1 98.6 ± 7.6High 88.9 ± 2.6 102.4 ± 5.5 101.3 ± 6.5 103.4 ± 5.2

Inter-assay precision and accuracyLow 99.3 ± 3.3 100.6 ± 8.9 100.7 ± 5.6 103.1 ± 12.8Medium 107.3 ± 7.0 97.2 ± 9.3 96.7 ± 13.2 98.9 ± 9.5High 89.3 ± 2.5 101.1 ± 5.3 103.6 ± 5.7 103.6 ± 5.9

Results are % of target value (mean ± standard deviation; N = 5 for intra-assay and N = 9 for inter-assay). Low, medium and high quality control concentrations were 0.001,0.1, and 10 ng/mL for buprenorphine and buprenorphine glucuronide and 0.01, 0.3 and 10 ng/mL for norbuprenorphine and norbuprenorphine glucuronide.

Table 4Accuracy and precision in urine.

Buprenorphine Norbuprenorphine Buprenorphine glucuronide Norbuprenorphine glucuronide

Intra-assay precision and accuracyLow 106.0 ± 6.2 101.4 ± 10.6 107.0 ± 9.9 111.0 ± 2.6Medium 99.1 ± 7.2 113.3 ± 2.1 85.8 ± 0.5 97.8 ± 7.2High 99.9 ± 9.6 100.0 ± 4.5 101.4 ± 5.2 99.8 ± 8.1

Inter-assay precision and accuracyLow 102.9 ± 9.7 96.2 ± 10.4 97.8 ± 12.2 105.7 ± 7.5Medium 99.3 ± 8.1 102.3 ± 9.0 94.9 ± 12.2 99.9 ± 5.9High 88.3 ± 5.5 93.8 ± 4.8 93.5 ± 10.3 89.3 ± 6.410 fold dilution 98.3 ± 9.8 100.5 ± 4.2 101.4 ± 5.2 99.3 ± 7.5

RL

esults are the % of target value (mean ± standard deviation; N = 3 for intra-assay and N =ow, medium and high quality control concentrations were 0.06, 0.6, and 60 ng/mL for al

9 for inter-assay).l analytes.

K.J. Regina, E.D. Kharasch / J. Chromatogr. B 939 (2013) 23– 31 29

Table 5Stability in plasma.

% of original value

Room temp. Freeze/thaw x1 Freeze/thaw x3

Low High Low High Low High

Buprenorphine 101.3 ± 14.1 102.1 ± 9.1 113.0 ± 10.3 104.4 ± 10.6 106.0 ± 6.6 101.5 ± 12.6Norbuprenorphine 99.1 ± 9.9 100.6 ± 1.4 109.3 ± 14.4 102.7 ± 5.2 101.4 ± 15.0 101.4 ± 6.3Buprenorphine glucuronide 108.4 ± 8.2 108.9 ± 7.4 116.9 ± 10.9 111.5 ± 6.6 114.8 ± 10.3 112.8 ± 4.9Norbuprenorphine glucuronide 114.2 ± 6.7 102.3 ± 14.2 105.1 ± 13.5 102.3 ± 8.5 116.7 ± 9.4 103.0 ± 5.6

Results are the mean ± standard deviation (N = 3). Low concentration was 0.01 ng/mL and high concentration was 10 ng/mL.

Table 6Stability in urine.

% of original value

Room temp Freeze/thaw x1 Freeze/thaw x3

Low High Low High Low High

Buprenorphine 87.7 ± 1.9 101.1 ± 5.9 92.9 ± 4.3 95.5 ± 3.4 94.1 ± 13.7 97.7 ± 9.8Norbuprenorphine 109.8 ± 1.1 104.1 ± 7.0 82.5 ± 7.3 89.4 ± 8.4 101.8 ± 5.6 92.1 ± 10.8Buprenorphine glucuronide 108.7 ± 3.5 82.4 ± 8.5 101.3 ± 11.6 100.4 ± 11.6 97.3 ± 7.1 96.6 ± 8.8

R L and

aaahapcpclaucdbrcbfs

3

pit(ia

3

mpnilsab

range needed was low pg/mL to low ng/mL with a target limitof quantification of 1 pg/mL. The degree of sensitivity afforded bythis assay was sufficient to determine buprenorphine concentra-tions in the clinical study, and achieved the target of a lower limit

Norbuprenorphine glucuronide 101.9 ± 10.1 86.2 ± 11.5

esults are the mean ± standard deviation (N = 3). Low concentration was 0.06 ng/m

nd from 89% to 107% from urine. Recovery of norbuprenorphinend norbuprenorphine glucuronide from plasma was lower, gener-lly approximately 80%. Analyte recovery from urine was generallyigher for all analytes, with all in the range of 90–100%. Inter-daynd intraday accuracy and precision for analyte determination inlasma and urine are shown in Tables 3 and 4, respectively. Pre-ision and accuracy of buprenorphine was high in both urine andlasma with a coefficient of variation of less than 10% in both matri-es. The other three analytes all had coefficients of variation ofess than 15% in both matrices. Optimized chromatographic sep-ration of analytes from matrix interferences in both plasma andrine minimized matrix interferences, thereby enabling the lowoefficients of variation. Precision and accuracy of 10 fold urineilution samples were also tested and passed acceptance criteria ofeing within 10% of the original measurement (Table 4). Stabilityesults in plasma and urine are shown in Tables 5 and 6. All fourompounds were stable in both plasma and urine for 24 h on theenchtop at room temperature, and after undergoing up to threereeze thaw cycles. It was also verified that prepared samples weretable up to one week when stored at 4 ◦C (data not shown).

.3. Application to a pharmacokinetic study

The method described here was successfully utilized to quantifylasma and urine concentrations of buprenorphine and metabolites

n a healthy volunteer subject enrolled in a clinical study to evaluatehe disposition of low dose intravenous (0.2 mg) and sublingual2 mg) buprenorphine. Plasma and urine concentrations are shownn Figs. 4 and 5, respectively. Analytes were quantifiable for 96 hfter dosing.

.4. Assay comparison

The investigation aimed to develop and validate an analyticalethod with improved sensitivity for analysis of buprenor-

hine, norbuprenorphine, buprenorphine-3�-glucuronide, andorbuprenorphine-3�-glucuronide, in plasma and urine. The

ntended application was analysis of pharmacokinetic studies of

ow dose intravenous and sublingual buprenorphine, with mea-urement of plasma and urine concentrations for up to 96 hfter dosing (the entire measurement period), to capture threeuprenorphine half-lives. The anticipated plasma concentration

105.6 ± 4.8 93.5 ± 8.6 101.5 ± 8.3 104.1 ± 10.6

high concentration was 60 ng/mL.

Fig. 4. Plasma concentrations of buprenorphine and metabolites in a researchsubject administered (A) intravenous buprenorphine (0.2 mg) or (B) sublingualbuprenorphine (2 mg).

30 K.J. Regina, E.D. Kharasch / J. Chro

Fig. 5. Urine concentrations of buprenorphine and metabolites in a research subjectap

ogbbtwpa

aolapfc2Apm[a2i(aim

[[[

[

[[

[

[

[[

[[

[

[

[

[

[[

[

[[

dministered (A) intravenous buprenorphine (0.2 mg) or (B) sublingual buprenor-hine (2 mg). Urine was collected in 24 hr intervals (days 1–4) after drug dosing.

f quantitation of 1 pg/mL for buprenorphine and buprenorphinelucuronide in plasma. Limits of quantification were higher for nor-uprenorphine and norbuprenorphine glucuronide. Nonetheless,ecause plasma concentrations of these metabolites were higherhan those of buprenorphine glucuronide, plasma concentrationsere quantifiable for up to 96 h (the experimental measurementeriod). Buprenorphine and all metabolites were readily quantifi-ble in urine.

No analytical methods with sufficient sensitivity, for all fournalytes, had been reported at the time this assay was devel-ped. Two recently reported higher sensitivity assays, one usingiquid–liquid extraction for buprenorphine and norbuprenorphine,nd another using unspecified solid-phase extraction for buprenor-hine and norbuprenorphine glucuronide [47]. Assay range for theour analytes was 20–10,000 pg/mL, and the lower limit of quantifi-ation was 20 pg/mL for buprenorphine and norbuprenorphine, and5 pg/mL for buprenorphine and norbuprenorphine glucuronide.

high-sensitivity assay for buprenorphine and norbuprenor-hine, but not their glucuronides, in plasma, using unspecifiedixed-mode cation extraction plates, was also recently reported

48]. Recovery of buprenorphine and norbuprenorphine was 76%nd 82%, respectively, and the lower limit of quantification was5 pg/mL for both. The present assay represents a significant

mprovement in sensitivity, and applicability to all four analytes

buprenorphine, norbuprenorphine, buprenorphine glucuronide,nd norbuprenorphine glucuronide), and in both matrices ofnterest (plasma and urine), compared with even these recent

ethods.

[[

[

matogr. B 939 (2013) 23– 31

4. Conclusion

A highly selective and sensitive UFLC–MS/MS assay forbuprenorphine and the metabolites norbuprenorphine,buprenorphine-3�-glucuronide and norbuprenorphine-3�-glucuronide in human plasma and urine was developed andvalidated. The method follows the guidance for biolanalyticalassays outlined by the US FDA. The method uses two deuteratedinternal standards for quantification, minimizes the number ofsteps to increase ruggedness of the assay. The use of solid-phasemicroextraction plates combined with UFLC instrumentationincreased both the speed of sample processing and the sensitivityof the assay. The results presented support the assay validation,including recovery, precision, accuracy, specificity, selectivityand stability. This is the first assay to present a robust methodthat can be used for routine sample analysis of buprenorphine,norbuprenorphine, buprenorphine glucuronide and norbuprenor-phine glucuronide in plasma in the pg/mL concentration rangeutilizing solid phase microextraction and core shell column tech-nology combined with a UFLC instrument. This method is fast,efficient, robust and has successfully been utilized to analyzeclinical samples.

References

[1] R.C. Heel, R.N. Brogden, T.M. Speight, G.S. Avery, Drugs 17 (1979) 81.[2] R.E. Johnson, P.J. Fudala, R. Payne, J. Pain Symptom Manage. 29 (2005) 297.[3] H.G. Kress, Eur. J. Pain 13 (2009) 219.[4] P. Lobmaier, M. Gossop, H. Waal, J. Bramness, Eur. J. Clin. Pharmacol. 66 (2010)

537.[5] D.S. Metzger, Y. Zhang, Curr. HIV/AIDS Rep. 7 (2010) 220.[6] S.L. Walsh, K.L. Preston, M.L. Stitzer, E.J. Cone, G.E. Bigelow, Clin. Pharmacol.

Ther. 55 (1994) 569.[7] A. Dahan, A. Yassen, R. Romberg, E. Sarton, L. Teppema, E. Olofsen, M. Danhof,

Br. J. Anaesth. 96 (2006) 627.[8] R.F. Wolff, D. Aune, C. Truyers, A.V. Hernandez, K. Misso, R. Riemsma, J. Kleijnen,

Curr. Med. Res. Opin. 28 (2012) 833.[9] G. Hans, D. Robert, J. Pain Res. 2 (2009) 117.10] G.L. Plosker, Drugs 71 (2011) 2491.11] J.P. Zacny, K. Conley, J. Galinkin, J. Pharmacol. Exp. Ther. 282 (1997) 1187.12] D.E. Moody, Y. Chang, W. Huang, E.F. McCance-Katz, Basic Clin. Pharmacol.

Toxicol. 105 (2009) 211.13] D.E. Moody, M.H. Slawson, E.C. Strain, J.D. Laycock, A.C. Spanbauer, R.L. Foltz,

Anal. Biochem. 306 (2002) 31.14] M. Ohtani, H. Kotaki, Y. Sawada, T. Iga, J. Pharm. Exp. Ther. 272 (1995) 505.15] M. Ohtani, H. Kotaki, K. Nishitateno, S. Yasufumi, T. Iga, J. Pharmacol. Exp. Ther.

281 (1997) 428.16] A. Yassen, J. Kan, E. Olofsen, E. Suidgeest, A. Dahan, M. Danhof, J. Pharmacol.

Exp. Ther. 321 (2007) 598.17] S.M. Brown, M. Holtzman, T. Kim, E.D. Kharasch, Anesthesiology 115 (2011)

1251.18] Y. Chang, D.E. Moody, E.F. McCance-Katz, Drug Metab. Dispos. 34 (2006) 440.19] R.D. Bruce, E. McCance-Katz, E.D. Kharasch, D.E. Moody, G.D. Morse, Clin. Infect.

Dis. 43 (Suppl 4) (2006) S216.20] V.A. Gruber, E.F. McCance-Katz, Curr. HIV/AIDS Rep. 7 (2010) 152.21] A.S. Saber-Tehrani, R.D. Bruce, F.L. Altice, Am. J. Drug Alcohol Abuse 37 (2011)

1.22] C.W. Hand, D. Baldwin, R.A. Moore, M.C. Allen, H.J. McQuay, Ann. Clin. Biochem.

23 (1986) 47.23] N. De Giovanni, N. Fucci, S. Scarlata, G. Donzelli, Clin. Chem. Lab. Med. 43 (2005)

1377.24] M. Ohtani, H. Kotaki, K. Uchino, Y. Sawada, T. Iga, Drug Metab. Dispos. 22 (1994)

2.25] J.J. Kuhlman Jr., S. Lalani, J. Magluilo Jr., B. Levine, W.D. Darwin, J. Anal. Toxicol.

20 (1996) 369.26] A.M. Lisi, R. Kazlauskas, G.J. Trout, J. Chromatog. B 692 (1997) 67.27] R.P. Nath, R.A. Upton, E.T. Everhart, P. Cheung, P. Shwonek, R.T. Jones, J.E.

Mendelson, J. Clin. Pharmacol. 39 (1999) 619.28] Y.S. Wang, D.L. Lin, S.C. Yang, M.Y. Wu, R.H. Liu, L.W. Su, P.S. Cheng, C. Liu, M.R.

Fuh, J. Chromatogr. A 1217 (2010) 1688.29] A. Tracqui, P. Kintz, P. Mangin, J. Forensic Sci. 42 (1997) 111.30] H. Hoja, P. Marquet, B. Verneuil, H. Lotfi, J.-L. Dupuy, G. Lachâtre, J. Anal. Toxicol.

21 (1997) 160.

31] R. Kronstrand, T.G. Selden, M. Josefsson, J. Anal. Toxicol. 27 (2003) 464.32] H. Yue, M.R. Borenstein, S.A. Jansen, R.B. Raffa, J. Pharmacol. Toxicol. Methods

52 (2005) 314.33] S. Hegstad, H.Z. Khiabani, E.L. Øiestad, T. Berg, A.S. Christophersen, J. Anal.

Toxicol. 31 (2007) 214.

. Chro

[

[[

[[

[[

[[

[

[

[

[

K.J. Regina, E.D. Kharasch / J

34] M.E. Rodriguez-Rosas, M.R. Lofwall, E.C. Strain, D. Siluk, I.W. Wainer, J. Chro-matogr. B 850 (2007) 538.

35] M. Concheiro, D.M. Shakleya, M.A. Huestis, Forensic Sci. Int. 188 (2009) 144.36] E.L. Oiestad, U. Johansen, A.M. Oiestad, A.S. Christophersen, J. Anal. Toxicol. 35

(2011) 280.37] R. Verplaetse, J. Tytgat, Forensic Sci. Int. 215 (2012) 136.38] G.A. McMillin, R. Davis, H. Carlisle, C. Clark, S.J. Marin, D.E. Moody, J. Anal.

Toxicol. 36 (2012) 81.

39] T. Berg, B. Jorgenrud, D.H. Strand, J. Anal. Toxicol. 37 (2013) 159.40] G. Luthi, V. Blangy, C.B. Eap, N. Ansermot, J. Pharm. Biomed. Anal. 77 C (2013)

1.41] A.I. Al-Asmari, R.A. Anderson, J. Anal. Toxicol. 31 (2007) 394.42] A.C. Liu, T.Y. Lin, L.W. Su, M.R. Fuh, Talanta 75 (2008) 198.

[

[

matogr. B 939 (2013) 23– 31 31

43] M.A. Huestis, E.J. Cone, S.O. Pirnay, A. Umbricht, K.L. Preston, Drug AlcoholDepend. 131 (2013) 258.

44] M. Escher, Y. Daali, J. Chabert, G. Hopfgartner, P. Dayer, J. Desmeules, Clin. Ther.29 (2007) 1620.

45] S.D. McAleer, R.J. Mills, T. Polack, T. Hussain, P.E. Rolan, A.D. Gibbs, F.G. Mullins,Z. Hussein, Drug Alcohol Depend. 72 (2003) 75.

46] US Food and Drug Administration, Center for Drug Evaluation and Research,2001, http://www.fda.gov/downloads/Drugs/Guidances/ucm070107.pdf,

(accessed 25.08.2013).

47] R.P. Kapil, A. Cipriano, G.H. Michels, P. Perrino, S.A. O’Keefe, M.S. Shet, S.V.Colucci, R.J. Noveck, S.C. Harris, Clin. Drug Investig. 32 (2012) 583.

48] N. Al-Tawil, I. Odar-Cederlof, A.C. Berggren, H.E. Johnson, J. Persson, Eur. J. Clin.Pharmacol. 69 (2013) 143.