enhanced sensitivity method for measuring gemcitabine in human plasma
TRANSCRIPT
Enhanced Sensitivity Method for MeasuringGemcitabine in Human Plasma
Amit Khatri1, James E. Fisher1, Mark N. Kirstein1,2,&
1 Department of Experimental and Clinical Pharmacology, Clinical Pharmacology Analytical Services, College of Pharmacy,Minneapolis, MN, USA
2 Masonic Cancer Center, University of Minnesota, Minneapolis, MN, USA; E-Mail: [email protected]
Received: 17 June 2010 / Revised: 12 August 2010 / Accepted: 16 August 2010Online publication: 8 September 2010
Abstract
A simple and reliable method for the separation and sensitive quantitation of gemcitabine inhuman plasma was developed. Samples were concentrated and purified with solid phaseextraction. Extracts were then injected onto Spherisorb 4.6 mm 9 150 mm, 3 lm ODS2column, and isocratically eluted with mobile phase containing 11 mM 1-octanesulfonatesodium as ion pairing agent. The assay was validated and intermediate precision (% CVt)was �6.40%. The assay was linear over a concentration range from 0.01 to 50 lM, and thelower limit of quantitation (LLOQ) was 0.01 lM. Solid phase extraction has enabledachievement of sensitivity which is at least 20-fold less than that reported for other LC assays,and is comparable to MS–MS.
Keywords
Column Liquid ChromatographySolid phase extractionPharmacokinetic studyGemcitabine in human plasma
Introduction
Gemcitabine (dFdC, 20,20-difluorodeoxy-
cytidine) is a deoxycytidine analogue
active against a variety of solid tumors.
It is a prodrug, which undergoes intra-
cellular phosphorylation to form the
active di- (dFdCDP) and triphosphate
(dFdCTP) metabolites. Gemcitabine also
undergoes metabolism by plasma and li-
ver cytidine deaminase (CDA) to form
20,20-difluorodeoxyuridine (dFdU)[1].
Several analytical approaches, includ-
ing F-NMR, enzyme linked immunosor-
bance assay (ELISA), liquid chromatog-
raphy and tandem-mass spectrometry
(MS–MS) have been reported for gemcit-
abine concentration measurements [2–5].
To evaluate gemcitabine population
pharmacokinetics, we developed and
validated an LC method (LLOQ 2 lM)
[5]. Gemcitabine elimination clearance
ranges from 200 to 600 L h-1 [6, 7];
therefore, concentrations rapidly fall be-
low the LLOQ. To help preventing model
misspecification, a more sensitive assay
was needed.
Experimental
Chemicals and Reagents
Gemcitabine as analytical standard was
provided by Eli Lilly (Indianapolis, IN,
USA). Internal standard, 20-deoxycyti-
dine (20dC), and perchloric acid 70%
were from Sigma (St. Louis, MO, USA).
Tetrahydrouridine (THU) was from
Calbiochem (San Diego, CA, USA).
Ion-pairing reagent, 1-octanesulfonate
sodium, was from Regis technologies
(Morton Grove, IL, USA). LC grade
acetonitrile and methanol were obtained
from Fisher (Fairlawn, NJ, USA). MCX
Oasis solid phase extraction cartridges,
3 cc, were from Waters (Milford, MA,
USA). Heparinized human plasma was
from Biological Specialties (Colmar, PA,
USA). All other chemicals were of
analytical grade and obtained from
commercial sources.
2010, 72, 1005–1008
DOI: 10.1365/s10337-010-1751-20009-5893/10/11 � 2010 Vieweg+Teubner Verlag | Springer Fachmedien Wiesbaden GmbH
Limited Short Communication Chromatographia 2010, 72, November (No. 9/10) 1005
Preparation of CalibrationStandards
Gemcitabine (1 mg mL-1) and 20dC (50
lg mL-1) stock solutions were accu-
rately prepared in methanol and stored
at -20 �C. Standard calibration solu-
tions were then further diluted with
methanol. At the time of assay, 20 lL of
each solution was added to 500 lL of
heparinized human plasma to obtain the
standard calibration curve at the fol-
lowing concentrations: 0.01, 0.02, 0.1,
0.5, 2, 10, 40 and 50 lM.
Preparation of Quality ControlSamples
The quality control samples were pre-
pared prior to assay validation. To pre-
vent metabolism of gemcitabine by
plasma cytidine deaminase, tetrahydro-
uridine (THU) was added prior to the
addition of gemcitabine (40 lL of
10 mg mL-1 THU in water per 5.0 mL
of plasma). The samples were then
stored at -80 �C. The internal standard,20dC, was added to obtain a final con-
centration of 9 lM. Gemcitabine quality
control concentrations were 0.075, 0.6,
and 5 lM.
Specimen Preparation andSolid-Phase Extraction
Samples were concentrated with solid
phase extraction (SPE). The SPE
cartridges were washed with 2 mL of
methanol and conditioned with 2 mL of
50 mM NaH2PO4 (pH 2.9) buffer. For
the preparation of calibration standards
in blank plasma, 20 lL each of the stan-
dard calibration solutions and 20 lL of
the internal standard stock solution were
added to the 12 9 75 mm centrifuge tube
and evaporated under nitrogen at 37 �Cfor 5 min. 500 lL of blank plasma con-
taining THU (1:125, THU: plasma) was
added to the dried 12 9 75 mm centri-
fuge tubes, vortex mixed, and acidified
with 500 lL of 4% o-phosphoric acid.
Patient plasma samples were processed
similarly with 20 lL of the internal stan-
dard stock solution. Next, the acidified
plasma was transferred to the condi-
tioned SPE cartridges and centrifuged at
759g for 5 min. The cartridges were then
washed with 2 mL of 50 mM NaH2PO4
(pH 2.9) followed by 2 mL of methanol.
Each washing step was followed by cen-
trifugation at 759g for 3 min. Finally,
gemcitabine was eluted with 1 mL of
freshly prepared methanolic NH4OH
(2% v/v). The eluates were evaporated
under nitrogen at 37 �Cand reconstituted
with 50 lL of mobile phase. After vortex
mixing, 25 lL were injected onto the LC
column.
Instrumentation andChromatographic Conditions
The LC system and data capture soft-
ware has been described previously [5].
25 lL of the standard and sample ex-
tracts were injected onto a Waters
Spherisorb 4.6 mm 9 150 mm, 3 lmODS2 column (Waters Corporation,
Milford, MA, USA) with A-102X in-line
filter, 0.5 lM Frit (Chromtech, Apple
Valley, MN, USA). Analytes were iso-
cratically eluted with a mobile phase
of acetonitrile:methanol:50 mM sodium
phosphate buffer 5:5:90 (v/v) containing
11 mM 1-octanesulfonate sodium. The
aqueous buffer was adjusted to pH 2.9
with 85% o-phosphoric acid. The mobile
phase was filtered with 0.22 micron
Nylon filter (GE Water and Process
Technology, USA). The flow rate was
1.2 mL min-1 and the column tempera-
ture was 40 �C. The wavelength was
267 nm as described previously [5, 8].
Assay Validation
Peak height was used for analysis of gem-
citabine in plasma samples, as described
previously [5]. The linear regression of the
ratio of peakheight of gemcitabine topeak
height of the internal standard (20dC) with
nominal concentrations of gemcitabine in
the samples was weighted by 1/x. The lin-
earity of the calibration curve was evalu-
ated by the squared correlation coefficient.
The procedure was validated over 5 days
by analysis of quality control samples in
quintuplicate to determine the within-day
(repeatability) and total variability (inter-
mediate precision) along with assay accu-
racy and precision [9]. A single set of
calibrators were used for each validation
run.
The lower limit of quantitation
(LLOQ) was defined as the peak height
ratio of signal/noise �10. Gemcitabine
concentrations of 0.01 and 0.02 lMwere
tested in quintuplicate for 5 days during
the validation for determination of
LLOQ. The LLOQ was defined as pre-
cision and accuracy of �10%. Accuracy
parameters were determined with analy-
sis of variance (ANOVA). Percentage
recovery following solid phase extraction
was determined as follows:
Percent recovery
¼ Peak height ratio of extracted plasma standard
Peak height ratio of non - treated standard
� 100:
Results and Discussion
Sample Preparation
During method development, we found
that a protein precipitation step before
solid phase extraction does not provide
any additional benefit. We found that
washing with 2 mL of methanol is suffi-
cient for washing the SPE column, and
1 mL of methanolic NH4OH (2% v/v)
was sufficient for the elution step. This
solid phase extraction enabled us to
concentrate the samples and to achieve
an LLOQ (0.01 lM), which is at least 20
fold less than the previously reported
values from LC-UV methods and com-
parable to a 0.01–0.04 lM LLOQ range
in methods [2, 10, 11].
Separation
Gemcitabine was separated from endoge-
nous plasma components using a Waters
Spherisorb 4.6 mm 9 250 mm, 5 lmODS2 column with isocratic elution.
We used reversed-phase chromatography
with a mobile phase at pH 2.9 and
11.0 mM 1-octanesulfonate sodium as
an ion-pairing reagent for the best
results. We initially started with 3 mM
1006 Chromatographia 2010, 72, November (No. 9/10) Limited Short Communication
1-octanesulfonate sodium as described
previously [5]; however, further adjustment
to 11 mM was required to achieve better
separation from an interfering endogenous
substance in the plasma. Due to the higher
concentration of ion-pairing agent used
and the presence of a late eluting endoge-
nous substance in the plasma, the run
time was about 25 min on the 250 mm
column. Hence further method develop-
ment and validation was performed on a
4.6 mm 9 150 mm, 3 lm ODS2 column,
which resulted in shorteningof the run time
to 16 min. Acetonitrile 5% (test range
4–12%) and methanol 5% (test range
4–12%) in the mobile phase with a column
temperature of 40 �C produced the best
resolution from endogenous substances.
The retention times for 20dC and gemcita-
bine were 6.8 and 9.5 min, respectively, at
which no interfering endogenous peaks
were detected in human plasma.
Assay Validation
Figure 1 shows representative chro-
matogram overlays of increasing gemcit-
abine concentrations with internal
standard in human plasma. Accuracy and
precision data for gemcitabine calibra-
tion standards are shown in Table 1.
Accuracy of the calibration standards
ranged from 96 to 108% and precision,
from 1 to 7%. The mean (±SD) weighted
linear regression slope, intercept, and
coefficients of determination (r2 value)
were as follows: slope = 0.087994 ±
0.0048, intercept = -0.00014 ± 0.00025,
and r2 = 0.9998 ± 0.00017, respectively.
Assessment of within-day and total vari-
ability in assay performance, precision,
and accuracy was performed by an anal-
ysis of variance procedure (Table 1). The
lowest concentration of the calibration
curve, 0.01 lM, had an accuracy and
precision of 102 and 7.2%, respectively;
hence assessed to be the LLOQ of the
assay. The calibration curve was linear up
to 50 lM. The average recoveries for
gemcitabine and internal standard were
62.63 and 82.68%, respectively.
Patient Samples
Application of this assay enabled us to
measure additional patient plasma sam-
ples that were below a measurable limit
with the previously published assay [5].
The concentration–time data from a
representative patient were modeled in
ADAPT-II, and the clearance was esti-
mated as 420.7 L h-1 [12].
Conclusion
We developed a simple and reliable
method with enhanced sensitivity for the
separation and quantitation of gemcita-
bine in human plasma. The lower limit of
quantitation for gemcitabine in our assay
was 0.01 lM,which is at least 20-fold less
than that reported for other LC-UV
assays, and comparable to those reported
for MS–MS methods.
References
1. Veltkamp SA, Pluim D, van EijndhovenMA, Bolijn MJ, Ong FH, GovindarajanR, Unadkat JD, Beijnen JH, Schellens JH(2008) Mol Cancer Ther 7(8):2415–2425
2. Honeywell R, Laan AC, van GroeningenCJ, Strocchi E, Ruiter R, Giaccone G,Peters GJ (2007) J Chromatogr B AnalytTechnol Biomed Life Sci 847(2):142–152
3. Abbruzzese JL, Grunewald R, Weeks EA,Gravel D, Adams T, Nowak B, MineishiS, Tarassoff P, Satterlee W, Raber MNet al (1991) J Clin Oncol 9(3):491–498
4. Edzes HT, Peters GJ, Noordhuis P, Ver-morken JB (1993) Anal Biochem 214(1):25–30
0 2 4 6 8 10 12 14 16
150
130
110
90
70
50
Time (min)m
V
Internal standard Gemcitabine
Blank plasma
Fig. 1. Representative chromatogram overlays of human plasma as blank, and with internalstandard and 0.1, 0.5, 2 and 10 lM gemcitabine calibration standards. Inset shows magnifiedchromatogram overlays for the lowest concentration (0.01 lM) and human plasma as blank
Table 1. Accuracy and precision of gemcitabine calibration standards and quality controls inhuman plasma
Calibrator Concentration (lM) Accuracy (%)
0.01 96.00.02 108.00.1 97.00.5 100.92.0 96.210.0 100.240.0 101.050.0 101.7
Quality Control Concentration (lM) Accuracy (%) CVW (%) CVt (%)
0.075 95.6 3.80 4.500.60 98.9 4.10 6.005.00 103.6 4.70 6.40
CVw within-run component CV (repeatability), CVt total CV (intermediate precision)
Limited Short Communication Chromatographia 2010, 72, November (No. 9/10) 1007
5. Kirstein MN, Hassan I, Guire DE, WellerDR, Dagit JW, Fisher JE, Remmel RP(2006) J Chromatogr B Analyt TechnolBiomed Life Sci 835(1–2):136–142
6. ThamLS,WangLZ, SooRA, LeeHS, LeeSC, Goh BC, Holford NH (2008) CancerChemother Pharmacol 63(1):55–64
7. Venook AP, Egorin MJ, Rosner GL,Hollis D, Mani S, Hawkins M, Byrd J,
Hohl R, Budman D, Meropol NJ, RatainMJ (2000) J Clin Oncol 18(14):2780–2787
8. Keith B, Xu Y, Grem JL (2003) J Chro-matogr B Analyt Technol Biomed Life Sci785(1):65–72
9. Rodbard D (1974) Clin Chem 20(10):1255–1270
10. Khoury H, Deroussent A, Reddy LH,Couvreur P, Vassal G, Paci A (2007) J
Chromatogr B Analyt Technol BiomedLife Sci 858(1–2):71–78
11. Xu Y, Keith B, Grem JL (2004) J Chro-matogr B Analyt Technol Biomed Life Sci802(2):263–270
12. D’Argenio D (1997) ADAPT II user’sguide: pharmacokinetic/pharmacodynamicsystems analysis software. In: Biomedicalsimulations resource, Los Angeles
1008 Chromatographia 2010, 72, November (No. 9/10) Limited Short Communication