Journal of Chromatography B, 810 (2004) 259–267

Development and validation of a liquid chromatography–mass

spectrometry method for the quantitation of naltrexone
and 6␤-naltrexol in guinea pig plasma
Satyanarayana Valiveti, Buchi N. Nalluri, Dana C. Hammell,
Kalpana S. Paudel, Audra L. Stinchcomb∗
Department of Pharmaceutical Sciences, College of Pharmacy, University of Kentucky, Lexington, KY 40536-0082, USA

Received 4 February 2004; accepted 9 August 2004

Abstract

A quantitative liquid chromatographic–electrospray ionization mass spectrometry method for the determination of naltrexone and 6␤-
naltrexol in guinea pig plasma has been developed and validated using naloxone as an internal standard. A single step precipitation-extraction
technique was carried out to extract the plasma samples using acetonitrile:ethyl acetate (1:1, v/v). The chromatographic separation was
performed on a C18 column using a mobile phase consisting of 35:65 (v/v) acetonitrile:2 mM ammonium acetate with 0.01 mM ammonium
citrate at a flow rate of 0.25 mL/min. The analyte was detected after positive electrospray ionization using selected ion monitoring (SIM)
mode. The mean recoveries for naltrexone, naltrexol, and naloxone were 91.7, 89.3, and 99.0%, respectively. The lower limit of quantification
(LLOQ) for naltrexone and 6␤-naltrexol was 1.25 ng/mL, and the limit of detection (LOD) was 0.75 ng/mL. The method was applied to a
pharmacokinetic study in order to assess the drug disposition of naltrexone in guinea pigs.
© 2004 Elsevier B.V. All rights reserved.

Keywords: Naltrexone; 6␤-Naltrexol; Pharmacokinetics

1. Introduction where the volume of plasma is very low (<200 ␮L). A va-
riety of quantitative analytical methods, including thin layer
Naltrexone, an opioid antagonist, commonly used for the chromatography (TLC) [4], gas chromatography (GC) [5–7],
treatment of narcotic addiction [1], has recently been pre- high-pressure liquid chromatography (HPLC) with electro-
scribed as an adjunct in the treatment of alcohol dependence chemical detection (ECD) [8–11], and GC–MS (mass spec-
[2,3]. Naltrexone undergoes extensive hepatic metabolism trometry) [12] have been reported for the quantification of
primarily via reduction to its major metabolite in humans, naltrexone and 6␤-naltrexol in plasma. The method based on
6␤-naltrexol. 6␤-Naltrexol is believed to be a major con- TLC may not be selective and sensitive for routine analysis
tributor to the pharmacologic effect of naltrexone [1]. For of the drugs in plasma. HPLC with ECD detection hinders
this reason, it is worthwhile to characterize the disposition the reproducibility and robustness of the method, because
of both naltrexone and 6␤-naltrexol. A sensitive and sim- the cell can be easily contaminated, especially in the analysis
ple analytical method is necessary for the pharmacokinetic of plasma samples. Disadvantages to using the GC and the
analysis of naltrexone and its metabolite, 6␤-naltrexol, in GC–MS methods are attributed to the elaborate sample prepa-
plasma samples from small animal models like guinea pigs, ration and various derivatization techniques required for these
assays. Two methods have been reported on the simultane-
∗ Corresponding author. Tel.: +1 859 323 6192; fax: +1 859 257 2787. ous analysis of naltrexone and 6␤-naltrexol by GC–MS/MS
E-mail address: astin2@email.uky.edu (A.L. Stinchcomb). [13,14] in biological specimens with a sensitivity of at least

1570-0232/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.jchromb.2004.08.016
260 S. Valiveti et al. / J. Chromatogr. B 810 (2004) 259–267

1 ng/mL. However, both methods required a derivatization acetonitrile). Working calibration standards at concentrations
technique and larger volumes (1 mL) of sample. Mason et of 1.25–500 ng/mL in plasma were prepared fresh daily. Five
al. reported a LC–MS/MS method [15] for quantification of levels of QC samples, 5, 50, 100, 200 and 400 ng/mL, were
naltrexone and 6␤-naltrexol in human plasma with a sen- prepared in plasma for the determination of inter-day accu-
sitivity of 0.25 ng/mL. However, no information is avail- racy and precision. A stock solution of naloxone (1 mg/mL)
able on the sample preparation and analytical conditions. was prepared in acetonitrile, from which a 500 ng/mL internal
In this manuscript, a relatively simple, selective, and sen- standard (IS) working solution was prepared in acetonitrile
sitive LC–MS method for the determination of naltrexone as well.
and 6␤-naltrexol in guinea pig plasma using a single-step
precipitation-extraction method is described. 2.3. Extraction procedure

All samples, QCs, and standards with a sample volume

2. Experimental of 0.1 mL spiked with 20 ␮L of IS working solution were
2.1. Materials and chemicals extracted with 1 mL of acetonitrile:ethyl acetate (1:1, v/v).
The mixture was vortexed for 30 s and centrifuged at 10,000
Naltrexone was obtained from Mallinckrodt Inc. (St. × g for 20 min. The supernatant was pipetted into a 3 mL glass
Louis, MO) and 6␤-naltrexol was obtained from the National test tube and evaporated at 37 ◦ C under nitrogen. The residue
Institute on Drug Abuse (NIDA Drug Supply, Research Tri- was reconstituted with 100 ␮L of acetonitrile and sonicated
angle Park, NC). The internal standard, naloxone, was ob- for 15 min. The samples were transferred into autosampler
tained from Sigma (St. Louis, MO). Ammonium acetate, vials containing low volume inserts and 20 ␮L was injected
ethyl acetate, and acetonitrile (HPLC grade) were obtained onto the HPLC column.
from Fisher Scientific (Fairlawn, NJ). Ammonium citrate was
obtained from Alfa Aesar (Ward Hill, MA). Water was pu- 2.4. LC–MS conditions
rified by a Barnstead nanopure DiamondTM Ultrapure water
system (Barnstead International, Dubuque, Iowa). Chromatography was performed on a Waters Symmetry®
C18 (2.1 mm × 150 mm, 5 ␮m) column at 35 ◦ C with a mobile
2.2. Calibration standards and quality control samples phase consisting of acetonitrile:ammonium acetate (2 mM)
containing 0.01 mM of ammonium citrate (35:65, v/v) set at a
Standards and quality control samples (QCs) were made flow rate of 0.25 mL/min. A Waters Symmetry® C18 (2.1 mm
from stock solutions (1 mg/mL, naltrexone and naltrexol in × 10 mm, 3.5 ␮m) guard column was also used.

Fig. 1. Full scan mass spectrum of naltrexone (m/z 324).

 S. Valiveti et al. / J. Chromatogr. B 810 (2004) 259–267 261

Fig. 2. Full scan mass spectrum of 6␤-naltrexol (m/z 344).

The LC–MS system consisted of a Waters Alliance (ESI) for ion production. Selected ion monitoring (SIM) was
2690 HPLC pump (Waters, Milford, MA, USA), a Waters performed in positive mode for naltrexone, m/z 324 [342 
Alliance 2690 autosampler, and a Micromass ZQ detector 324] (dwell time 0.30 s) (Fig. 1), naltrexol, m/z 344 [M + H]+
(Waters, Milford, MA, USA) using electrospray ionization (Fig. 2), and naloxone, m/z 310 [364  310] (dwell time

Fig. 3. Full scan mass spectrum of naloxone (m/z 310).

262 S. Valiveti et al. / J. Chromatogr. B 810 (2004) 259–267

Fig. 4. Typical HPLC/MS ion chromatograms spiked with 25 ng/mL of naltrexone, 6␤-naltrexol and 100 ng/mL of naloxone in guinea pig plasma (a) naltrexone
(4.80 min); (b) 6␤-naltrexol (3.22 min); (c) naloxone (5.98 min).
 S. Valiveti et al. / J. Chromatogr. B 810 (2004) 259–267 263

0.30 s) (Fig. 3). Capillary voltage was 4.5 kV and cone volt- 2.7. Pharmacokinetic analysis
age was 30 V. The source block and desolvation temperatures
were 120 and 250 ◦ C, respectively. Nitrogen was used as a The pharmacokinetic analysis of naltrexone plasma con-
nebulization and drying gas at flow rates of 50 and 450 L/h, centration versus time profiles after intravenous bolus admin-
respectively. The retention times for naltrexone, naltrexol and istration was carried out by fitting the data to a three compart-
naloxone were 4.81 ± 0.15, 3.20 ± 0.11, and 5.84 ± 0.20 min ment model (WinNonlin Professional, version 4.0, Pharsight
(Fig. 4), respectively. Calibration graphs were constructed Corporation, Mountain View, California) with the following
using a linear regression of the ratio of the drug peak- exponential expression:
area to internal standard versus nominal drug concentra-
tions. C(t) = Ae−αt + Be−βt + Ce−γt (1)

where C(t) is the plasma concentration of drug at time ‘t’; A,

2.5. Validation B and C are preexponential constants; fast distribution rate
constant, α; slow distribution rate constant, β; terminal elim-
The method was validated for accuracy, precision, selec- ination rate constant, γ; and t is time. The pharmacokinetic
tivity, calibration curve range, and reproducibility over a con- parameters, such as terminal elimination half-life, t1/2(␥) ; dis-
centration range of 1.25–500 ng/mL using five calibration tribution half lives, t1/2(␣) and t1/2(␤) ; steady-state volume
standards, each containing the two analytes of interest, and of distribution, Vss ; area under the curve from 0 to infinity,
three replicates of QC samples at each concentration level in AUC0−∞ ; and total body clearance (Cltot ) were estimated
three separate runs. using the software. The peak plasma concentration (Cmax )
The matrix effect (co-eluting, undetected endogenous after the IV bolus dose of naltrexone was used to calculate
matrix compounds that may influence the analyte ionization) the initial volume of distribution by the following equation:
was investigated by extracting “blank” normal plasma and
dose
reconstituting with acetonitrile containing a known amount V = (2)
of the analytes, analyzing the reconstituted extracts, and Cmax
then comparing the peak areas of the analytes with that
of analytes in acetonitrile. The extraction recoveries of
3. Results and discussion
naltrexone, naltrexol, and naloxone were calculated by
comparing the peak areas of extracted plasma standards
The initial development step for the LC–MS method
to the peak areas of post-extraction plasma blanks spiked
consisted of a mobile phase of 2mM ammonium acetate:
at corresponding concentrations. The extraction recoveries
acetonitrile (35:65) at a flow rate of 0.25 mL/min, but tailing
of naltrexone and 6␤-naltrexol in QC samples were also
was observed with naltrexone and 6␤-naltrexol. In order
performed to prove consistency across the complete dynamic
to improve the peak shapes, a concentration of 0.01 mM
range.
ammonium citrate was added. Typical ion chromatograms
obtained with blank guinea pig plasma spiked with 25 ng/mL
2.6. Stability studies naltrexone, naltrexol and IS working solution are shown in
Fig. 4. The representative chromatograms of processed blank
The stabilities of naltrexone, 6␤-naltrexol, and the IS were plasma are shown in Fig. 5. The total run time for each sample
investigated in the stock solutions and in the final extracts. was about 15 min. Naltrexone, 6␤-naltrexol, and naloxone
The stabilities of the analytes and IS in the stock solution were peaks were well resolved and free of interference from
determined at room temperature and at 4 ◦ C. The concentra- endogenous compounds in the plasma. Only three additional
tion of IS in the QC samples was 100 ng/mL. Freshly prepared peaks were observed at 1.69–1.95 min, and these were well
QC samples were stored for 48 h at room temperature, and 1 separated from the drug peaks. Standard curves prepared
week at 4 ◦ C. For each of the storage conditions, three repli- for naltrexone and 6␤-naltrexol in plasma were linear over a
cates were analyzed at five concentration levels. The analyte range of 1.25–500 ng/mL. The mean (n = 3) calibration curves
and IS samples were processed immediately at each individ- for naltrexone and 6␤-naltrexol were y = 0.0262x − 0.0749,
ual time point and compared with that of freshly prepared R2 = 0.999 and y = 0.0177x + 0.0069, R2 = 0.999, respectively,
solutions. The post-preparative stabilities of the analytes and where y and x are the peak area ratio of analyte to internal
IS in the final extracts were studied at three concentrations at standard and concentration (ng/mL) of analyte, respectively.
autosampler temperature (12 ◦ C) for 48 h. The drug concen- The mean absolute recoveries of naltrexone, 6␤-naltrexol,
trations in the final extract QC samples were compared at 0 and naloxone (IS) determined in triplicate in the concentra-
and 48 h. The analytes and IS were considered to be stable in tion range of 1.25–500 ng/mL were 91.7% (%CV 4.6), 89.3%
the final extract (post-preparative) when 85–115% of the ini- (%CV 7.2), and 99.0% (%CV 5.4), respectively. The abso-
tial concentration was found. The stability limit in the stock lute recoveries of naltrexone and 6␤-naltrexol in the QC sam-
solutions was set at 95–105% of the initial concentrations ples are listed in Table 1. The absolute recovery values for
[16]. QC samples were in between 82.2 and 95.5% for naltrexone
264 S. Valiveti et al. / J. Chromatogr. B 810 (2004) 259–267

Fig. 5. The representative HPLC/MS ion chromatograms of processed blank guinea pig plasma (a) naltrexone; (b) 6␤-naltrexol; (c) naloxone.
 S. Valiveti et al. / J. Chromatogr. B 810 (2004) 259–267 265

Table 1
Recovery data for QC samples of naltrexone and 6␤-naltrexol (n = 3)
Naltrexone 6␤-Naltrexol

Concentration (ng/mL) Recovery (%) %CV Concentration (ng/mL) Recovery (%) %CV
5 82.2 6.5 5 89.3 5.4
50 87.2 5.3 50 91.3 3.6
100 88.2 3.2 100 92.4 3.1
200 94.3 4.6 200 98.6 4.9
400 95.5 3.0 400 98.3 4.3

Table 2
Intra-day and inter-day quality control results of naltrexone
Intra-day variation Inter-day variation

Concentration Mean concentration %CVb %Accuracy Concentration Mean concentration %CVb %Accuracy
(ng/mL)a found (ng/mL) (ng/mL) found (ng/mL)
5 5.0 5.8 100.2 5 4.98 4.9 99.6
50 49.95 3.9 99.9 50 48.6 1.9 97.2
100 98.1 4.8 98.1 100 101.2 4.0 101.2
200 195.9 2.5 98.0 200 199.7 6.0 99.8
400 400.5 3.3 100.1 400 401.1 4.7 100.3
a n = 3.
b %CV: coefficient of variation.

and between 89.3 and 98.6% for 6␤-naltrexol. No significant verification standard at the beginning and at the end of each
matrix effect was observed for the analytes in the plasma batch indicated that the system response remained stable.
samples. The peak areas of analytes in the reconstituted QC The described method was applied to a pharmacokinetic
samples had a coefficient of variation of 6%, indicating that study of an intravenous dose of naltrexone in guinea pigs. All
the extracts were “clean” with no co-eluting compounds in- animal studies were approved by the University of Kentucky
fluencing the ionization of the analytes. IACUC. Representative plasma profiles of observed and pre-
The LLOQ, defined as that concentration of naltrexone dicted concentrations of naltrexone, and observed concen-
and naltrexol which can still be determined with acceptable trations of 6␤-naltrexol after an intravenous bolus dose of
[16] precision (%CV < 10) and accuracy, was found to be naltrexone in guinea pigs (3 mg/kg) are shown in Fig. 6. It
1.25 ng/mL and the LOD for naltrexone and 6␤-naltrexol was can be seen from the plasma profiles of naltrexone and 6␤-
0.75 ng/mL. Results of the intra-day and inter-day validation naltrexol that drug could still be detected even after 20 h.
assays presented in Tables 2 and 3 indicated that the accuracy The plasma profile of naltrexone in the guinea pig followed
of the assay was >95% and the CV did not exceed 7%. Nal- a three compartmental model. The observed plasma concen-
trexone, 6␤-naltrexol and the IS were stable (Table 4) in the tration of naltrexone was in good agreement (correlation =
stock solution at room temperature and at 4 ◦ C for the time pe- 0.978) with the predicted plasma concentration, and the phar-
riods studied. The post-preparative stability studies (Table 5) macokinetic parameters of naltrexone are shown in Table 6.
indicated that the stabilities of naltrexone, 6␤-naltrexol and The maximum plasma concentration of naltrexone obtained
the IS were guaranteed for at least 48 h at 12 ◦ C. Due to the after intravenous administration of 3 mg/kg in the guinea pigs
high selectivity of MS detection; no interfering peaks were was 1039.5 ± 612.3 ng/mL, and it sharply declined to 9.2 ±
found when blank plasma extracts were analyzed. The ioniza- 3.5 ng/mL after 2 h. The maximum plasma concentration of
tion response monitored by injecting a system performance the naltrexol metabolite was 60.7 ± 18.2 ng/mL with a Tmax

Table 3
Intra-day and inter-day quality control results of naltrexol
Intra-day variation Inter-day variation

Concentration Mean concentration %CVb %Accuracy Concentration Mean concentration %CVb %Accuracy
(ng/mL)a found (ng/mL) (ng/mL) found (ng/mL)
5 4.9 6.2 97.8 5 5.2 2.3 103.2
50 50.2 2.6 100.5 50 49.1 2.1 98.2
100 102.4 5.3 102.4 100 98.7 4.2 98.7
200 196.3 2.0 98.2 200 198.5 3.7 99.3
400 399.5 5.6 99.9 400 402.7 4.7 100.7
a n = 3.
b %CV: coefficient of variation.
266 S. Valiveti et al. / J. Chromatogr. B 810 (2004) 259–267

Table 4
Stability of naltrexone, 6␤-naltrexol and IS in stock solutions (n = 3)
Drug Storage condition Concentration Mean Concentration Mean Concentration %Deviation %CV
(ng/mL) (ng/mL) recovered (ng/mL)
Naltrexone At 25 ◦ C for 48 h 5 5.21 5.35 2.69 3.2
50 51.25 50.50 −1.46 1.33
100 100.52 99.58 −0.94 1.80
200 200.21 198.36 −0.92 5.65
400 398.01 400.19 0.55 1.36
At 4 ◦ C for 1 week 5 5.15 4.96 −3.69 2.00
50 49.00 50.02 2.08 2.55
100 100.89 99.88 −1.00 5.66
200 200.14 199.63 −0.25 2.01
400 397.21 400.41 0.81 3.32
6␤-Naltrexol At 25 ◦ C for 48 h 5 4.85 5.15 6.19 2.5
50 50.00 49.05 −1.90 1.02
100 100.35 99.25 −1.10 4.09
200 199.25 200.25 0.50 4.65
400 400.65 398.33 −0.58 4.01
At 4 ◦ C for 1 week 5 4.98 4.93 −1.0 3.4
50 51.66 48.62 −5.88 4.22
100 99.02 98.55 −0.47 4.01
200 201.62 197.32 −2.13 1.5
400 401.25 398.14 −0.78 1.89
IS At 25 ◦ C for 48 h 100 100.25 97.1 −3.14 3.85
At 4 ◦ C for 1 week 100 101.63 100.2 −1.41 1.02

Table 5
Post-preparative stability of naltrexone, 6␤-naltrexol and IS at 12 ◦ C for 48 h (n = 3)
Drug Concentration Mean concentration Mean concentration %Deviation %CV
(ng/mL) (ng/mL) recovered (ng/mL)
Naltrexone 5 4.98 4.86 −2.41 0.99
50 51.21 50.90 −0.61 1.65
100 99.20 101.15 1.97 3.69
200 199.01 200.21 0.60 2.5
400 400.25 399.85 −0.10 1.69
6␤-Naltrexol 5 5.21 5.35 2.69 1.8
50 49.52 51.56 4.12 2.01
100 99.60 99.00 −0.60 2.85
200 198.65 197.54 −0.56 5.63
400 399.60 400.54 0.24 6.01
IS 100 100.1 97.10 −3.00 5.30

Table 6
Pharmacokinetic parameters of naltrexone after intravenous administration
(3 mg/kg) in guinea pigs (n = 3)
Parameter Mean ± S.D.
Cmax (ng/mL) 1039.5 ± 612.3
AUC (ng/mL)h 430.7 ± 105.8
AUMC (ng/mL)h2 1095.2 ± 132.3
Cl (L/h) 7.14 ± 1.68
Vss (L/kg) 15.78 ± 3.44
α (1/h) 3.75 ± 0.77
β (1/h) 0.45 ± 0.03
γ (1/h) 0.07 ± 0.02
t1/2(␣) (h) 0.19 ± 0.04
t1/2(␤) (h) 1.54 ± 0.09
t1/2(␥) (h) 9.81 ± 2.43
Fig. 6. Mean (±S.D.) plasma profiles of naltrexone and 6␤-naltrexol after Initial Vd (L/kg) 3.49 ± 1.53
intravenous administration of naltrexone (3 mg/kg) in guinea pigs (n = 3). MRT (h) 8.53 ± 0.71
 S. Valiveti et al. / J. Chromatogr. B 810 (2004) 259–267 267

of 15 min (Fig. 6). The mean terminal elimination half-life [2] J.R. Volpicelli, A.I. Alterman, M. Hayashida, Arch. Gen. Psychiatry
(t1/2(␥) ), steady-state apparent volume of distribution (Vss ), 49 (1992) 876.
and total clearance (Cl) of naltrexone were 9.81 h, 15.78 L/kg [3] S.S. O’Malley, A. Jaffe, G. Chang, R.S. Schottenfeld, R. Meyer, B.
Rounsaville, Arch. Gen. Psychiatry 49 (1992) 881.
and 7.14 L/h, respectively. [4] M.E. Wall, D.R. Brine, M. Perez-Reyes, Drug Metab. Dispos. 9
(1981) 369.
[5] K. Verebey, M.J. Kogan, A. DePace, S.J. Mulle, J. Chromatogr. 118
4. Conclusion (1976) 331.
[6] K. Verebey, A. DePace, D. Jukofsky, J.V. Volavka, S.J. Mule, J.
Anal. Toxicol. 4 (1980) 33.
A LC–MS method for the estimation of naltrexone and [7] R.H. Reuning, S.B. Ashcraft, B.E. Morrison, NIDA Res. Monogr.
6␤-naltrexol, its metabolite, in guinea pig plasma was suc- 28 (1981) 25.
cessfully developed and validated. The method is sensitive [8] M.C. Meyer, A.B. Straughn, M.W. Lo, W.L. Schary, C.C. Whitney,
and simple with an LLOQ of 1.25 ng/mL for naltrexone J. Clin. Psychiatry 45 (1984) 15.
and 6␤-naltrexol using a 0.1 mL aliquot of sample. It has [9] H. Derendorf, A. El-Koussi, E.R. Garrette, J. Pharm. Sci. 73 (1984)
621.
been shown in a pharmacokinetic study with guinea pigs [10] E.F. O’Connor, S.W.T. Cheng, W.G. North, J. Chromotogr. Biomed.
that naltrexone and 6␤-naltrexol could be quantitated after a Appl. 491 (1989) 240.
3 mg/kg dose of naltrexone. Thus, the method is appropriate [11] P. Zuccaro, I. Altieri, P. Bette, R. Pacifici, G. Ricciarello, L.A. Pini,
for monitoring naltrexone and its metabolite, 6␤-naltrexol, E. Sternieri, S. Pichini, J. Chromotogr. Biomed. Appl. 567 (1991)
in pharmacokinetic studies. 485.
[12] K.M. Monti, R.L. Foltz, D.M. Chinn, J. Anal. Toxicol. 15 (1991)
136.
[13] C.C. Nelson, M.D. Fraser, J.K. Wilfahrt, R.L. Foltz, Ther. Drug
Acknowledgments Monit. 15 (1993) 557.
[14] C.F. Chan, G.M. Chiswell, R. Bencini, L.P. Hackett, L.J. Dusci, K.F.
This work was supported by the National Institutes Ilett, J. Chromatogr. B 761 (2001) 85.
[15] B.J. Mason, A.M. Goodman, R.M. Dixon, M.H. Hameed, T. Hulot,
of Health R01DA13425. The authors acknowledge the K. Wesnes, J.A. Hunter, M.G. Boyeson, Neuropsychopharmacology
National Institute on Drug Abuse (NIDA) for providing the 27 (2002) 596.
6␤-naltrexol. [16] Guidance for Industry Bioanalytical Method Validation, US Depart-
ment of Health and Human Services, Food and Drug Administration,
Center for Drug Evaluation and Research (CDER), Center for Vet-
erinary Medicine (CVM), BP, May 2001, http://www.fda.gov/cvm.
References

[1] K. Verebey, J. Volavka, S.J. Mule, R.B. Resnick, Clin. Pharmacol.

Ther. 20 (1976) 315.