Stereoselective determination of the epimer mixtures of itraconazole in human blood plasma using HPLC and fluorescence detection

Article abstract of DOI:10.1002/chir.20932

Itraconazole is an antifungal drug widely used in a variety of fungal infections, which have become a significant public-health problem in recent decades. Itraconazole is a chiral drug consisting of two diastereoisomeric racemates, i.e., four stereoisomers. Data in the literature suggests that stereochemistry may play a significant role in the action and disposition of the drug and therefore stereoselective analytical methods for the determination of the drug in biological fluids are needed for the elucidation of that role. We report a stereoselective HPLC method that incorporates solvent extraction, the use of an internal standard, two chiral stationary phases in series, and fluorescence detection. The procedure is enantioselective and partially diastereoselective and provides the concentrations in blood plasma of the two epimer mixtures 2R,4S,2'R/2R,4S2'S and 2S,4R,2'R/2S,4R,2'S, respectively, each of which is a combination of the two epimers that differ in the configuration at the sec-butyl group. The analytical method has suitable sensitivity, recovery, precision, and accuracy. Analysis of the plasma of a human subject six hours after the oral administration of a single 200-mg dose of itraconazole showed a 3.4-fold difference between the concentrations of the epimer mixtures. The method has certain advantages over the published alternative procedure that uses LC-MS. Copyright

Full text of DOI:10.1002/chir.20932

CHIRALITY 23:495–503 (2011)
Stereoselective Determination of the Epimer Mixtures of Itraconazole
in Human Blood Plasma using HPLC and Fluorescence Detection
1
CHRISTINA PYRGAKI,1 STEVE J. BANNISTER,1 LAJOS GERA,2 JOHN G. GERBER,1 AND JOSEPH GAL
*
1Division of Clinical Pharmacology, University of Colorado School of Medicine, Aurora, CO, USA
2Department of Biochemistry and Molecular Genetics, University of Colorado School of Medicine, Aurora, CO, USA
ABSTRACT
Itraconazole is an antifungal drug widely used in a variety of fungal infections,
which have become a significant public-health problem in recent decades. Itraconazole is a chi-
ral drug consisting of two diastereoisomeric racemates, i.e., four stereoisomers. Data in the liter-
ature suggests that stereochemistry may play a significant role in the action and disposition of
the drug and therefore stereoselective analytical methods for the determination of the drug in
biological fluids are needed for the elucidation of that role. We report a stereoselective HPLC
method that incorporates solvent extraction, the use of an internal standard, two chiral station-
ary phases in series, and fluorescence detection. The procedure is enantioselective and partially
diastereoselective and provides the concentrations in blood plasma of the two epimer mixtures
2R,4S,2’R/2R,4S2’S and 2S,4R,2’R/2S,4R,2’S, respectively, each of which is a combination of the
two epimers that differ in the configuration at the sec-butyl group. The analytical method has
suitable sensitivity, recovery, precision, and accuracy. Analysis of the plasma of a human subject
six hours after the oral administration of a single 200-mg dose of itraconazole showed a 3.4-fold
difference between the concentrations of the epimer mixtures. The method has certain advan-
tages over the published alternative procedure that uses LC-MS. Chirality 23:495–503,
C
VV
2011.
2011 Wiley-Liss, Inc.
KEY WORDS: Itraconazole; HPLC; bioanalysis; enantioselectivity; diastereoselectivity;
fluorescence; antifungal; azole; stereoselectivity
INTRODUCTION
genic centers. The same is true for the relationship of B
and C.
There is considerable interest in the literature in the relation-
Fungal infections have become a major public health prob-
lem during the past ca. thirty years as a result of increased
incidence of immunodeficiency due to the advent of HIV/
AIDS, the widespread use of cancer chemotherapy, and
immunosuppression in organ transplantation.1,2 The ‘‘azoles’’
are important synthetic antifungal drugs and are so named
for the presence of an imidazole or triazole moiety in the mol-
ecule. Itraconazole (ITZ), a triazole, is an important and
widely used agent in the treatment of a variety of fungal
infections. Its chemical structure is chiral and includes three
stereogenic centers and therefore eight stereoisomers are
theoretically possible. However, the drug is defined and
used clinically as a mixture of the two cis racemates, i.e., two
pairs of enantiomers, four stereoisomers (Fig. 1a). Stereoiso-
mers A and B in Fig. 1a are enantiomerically related, as are
C and D. Diastereoisomers A and C are epimers since they
have the same (2R,4S) configuration at the stereogenic cen-
ters of the dioxolane ring and differ in the configuration at
only one (the sec-butyl center) of the three stereogenic cen-
ters in the molecules (it is worthwhile to recall the definition3
of epimers: ‘‘Among molecules containing asymmetric atoms,
epimers are diastereoisomers which differ in the configura-
tion of any one (and only one) of the several asymmetric
atoms’’).. Mixtures of A and C will henceforth be referred to
as epimer mixture (2R,4S). Similarly, B and D have the same
(2S,4R) configuration at the dioxolane ring and differ only in
the configuration at the sec-butyl center, and will henceforth
be referred to as epimer mixture (2S,4R). For the sake of
completeness, it should be added that stereoisomers A and
D are related as diastereoisomers but not as epimers, since
they differ in the configuration at two of the three stereo-
ship between the pharmacokinetics and pharmacodynamics of
ITZ, including the potential utility of therapeutic drug monitor-
ing (TDM), i.e., the use of the concentrations of the drug in the
blood of patients for guiding the physician in the optimization of
therapy.4,5,6,7,8 Despite such interest, however, little attention
has been paid to the stereochemical aspects of the antifungal ac-
tivity, toxicity, and disposition of ITZ. It has been reported in a
U.S. patent9 that epimer mixture (2S,4R) is 4-fold more potent
than epimer mixture (2R,4S) against Candida albicans in vitro,
but the antifungal activity of the four separate stereoisomers of
ITZ has not been described. Kunze et al. studied the stereo-
chemical aspects of the in vitro and in vivo metabolism of ITZ
and found that only A and C are metabolized by Cytochrome
P450 3A4 (CYP3A4).10 The authors also reported that all four
stereoisomers of ITZ induced a type II binding spectrum with
CYP3A4 and all four inhibited the CYP3A4-catalyzed hydroxyla-
tion of midazolam. In addition, they also found that the human
pharmacokinetics of ITZ was highly stereoselective, with epimer
Christina Pyrgaki is currently at Niswander Laboratory, Department of Pedia-
trics, University of Colorado Denver and HHMI, Mailstop 8133, Aurora, CO
80045, USA
Steve J. Bannister is currently at Hightower Pharmaceutical Services Corp,
5035 Barrowe Dr, Tampa, FL 33624, USA
John G. Gerber is retired.
*Correspondence to: Prof. Joseph Gal, University of Colorado Hospital, Clini-
cal Laboratories A-022, Aurora, CO 80045, Telephone: (720) 848-6032, Fax:
(720) 848-7073. E-mail: joe.gal@ucdenver.edu
Received for publication 9 January 2010; accepted in revised form 9 Septem-
ber 2010; Accepted 28 October 2010
DOI: 10.1002/chir.20932
Published online 10 June 2011 in Wiley Online Library
(wileyonlinelibrary.com).
C
VV
2011 Wiley-Liss, Inc.

496
PYRGAKI ET AL.
Fig. 1. The chemical structures of (a) the four stereoisomers of ITZ; A and B form one racemate and C and D form another; the third Cahn-Ingold-Prelog
descriptor in the configurational specification of each stereoisomer is that of the stereogenic center in the sec-butyl group; (b) OH-ITZ. The hydroxyl group intro-
duced during the biotransformation of ITZ to form OH-ITZ and the new stereogenic (chiral) center thus created are indicated; (c) IBDAK (the IS). The stereo-
chemical wedge bonds in (b) and (c) imply only relative configuration.
mixture (2S,4R) reaching higher concentrations than epimer
mixture (2R,4S).10
ing a large number of samples. In the same study SFC was
also examined and four peaks were obtained for ITZ, but two
were only partially resolved, and the total run time was
>60min.13 Finally, in the above-cited study10 on the stereose-
lective disposition of ITZ a LC-MS method was used. The
chromatographic conditions employed produced 3 peaks for
the drug and it was demonstrated that the epimers in the
(2R,4S) mixture were not resolved. Moreover, the total run
time was ca. 47 min and no internal standard was used in the
quantification of the stereoisomers. The pharmacokinetics of
the drug was characterized only for the two epimer mixtures,
as mentioned above.10
The limited published data suggests that stereochemistry
may play an important role in the pharmacodynamics and
pharmacokinetics of ITZ and thus stereoselective analytical
methods for the drug in biological fluids are of considerable
interest. In the present article a HPLC method for the stereo-
Limited work has been published on stereoselective ana-
lytical methodology needed for studies of the stereoselectiv-
ity of ITZ action and disposition. Breadmore and Thormann
described a CE-based method for ITZ which produced two
peaks, but the identity of the components giving rise to the
peaks was not determined.11 It is therefore not clear whether
the conditions used11 produced a separation of the racemates
or of some other mixtures of stereoisomers. Another CE-
based method12 separated the four stereoisomers of ITZ but
the technique was not applied to their determination in bio-
logical fluids. Using a custom-made CSP, one of the present
authors and coworkers developed13 a HPLC separation of
the four ITZ stereoisomers, but the retention times were
excessively long (total run time >100 min) and therefore the
separation was unsuitable for pharmacokinetic studies involv-
Chirality DOI 10.1002/chir

STEREOSELECTIVE HPLC BIOANALYSIS OF ITRACONAZOLE
497
50% B. Total run time was 25 min, and detection was at 235nm. The
eluted product was collected and concentrated at water aspirator pres-
sure (ca. 36 mm Hg) at 358C. The residue was dissolved in ethyl acetate
and the solution was washed with 3x15 ml sodium hydrogen carbonate
(3%) and then with brine (15 ml). The organic layer was dried over so-
dium sulfate, filtered, and evaporated at 358C using a water aspirator.
The residue was dissolved in dioxane (10 ml) and the solution was fil-
tered and lyophilized. The yield of IBDAK was 160.2 mg (68.8%).
HPLC Instrumentation for ITZ analysis
The HPLC system consisted of a Shimadzu SIL-10AD auto injector,
two Shimadzu LC-10AD pumps, and a Shimadzu RF-10AXL fluorescence
detector. The chromatographic separation was achieved on a two-column
system consisting of a Cyclobond I 2000 RSP (250 mm x 4.6 mm i.d.)
(Astec, Whippany, NJ, USA) followed in series by a Daicel Chiralcel OD
column of 4.6 mm id and 250 mm length (Chiral Technologies, Inc.,
Exton, PA). The columns were protected with a lBondapak Cyano guard
column (Waters, Milford, MA). The fluorescence of the eluate was moni-
tored at the excitation wavelength of 269 nm and emission at 374 nm.
Scheme 1. The synthesis of the internal standard. a. Ketoconazole, 85%
powdered potassium hydroxide, 1-butanol, reflux 24hr. b. Remove solvent in
vacuo, suspend residue in water, extract with chloroform, purify by HPLC. c.
Isobutyric acid, HATU, DMF, N,N-diisoproylethylamine (for details see
EXPERIMENTAL). The stereochemical bonds indicate only relative con-
figuration; only the racemate was synthesized.
selective analysis of the two epimer mixtures in human blood
plasma (‘‘plasma’’ henceforth) using fluorescence detection
is described.
Chromatographic conditions for ITZ analysis
The columns were maintained at 408 C using a model TCM column
heater (Waters) and the separation was carried out using a gradient sys-
tem. The mobile phase consisted of two components: solvent A was hex-
ane and solvent B was a mixture of 2-propanol, denatured ethanol (i.e.
ethanol containing 5% v/v 2-propanol), and acetonitrile, 42.5%, 42.5%, 15%
v/v, respectively. In the gradient program the initial conditions were 75%
A and 25% B. The linear gradient run was initiated at 0.01 min, and the
final composition of 38% A and 62% B was reached over 22 min. At the
end of the ramp the composition of the mobile phase was changed to 30%
A and 70% B and this composition was maintained for 8 min. Initial condi-
tions were then re-established. The flow rate was 1.5 ml/min throughout.
These conditions provided two peaks for ITZ, the less retained material
being the epimer mixture (2R,4S) while the longer-retained material was
the epimer mixture (2S,4R), determined by injecting separately the indi-
vidual epimer mixtures. The resolution RS of the two peaks was deter-
mined according to the standard definition [10]. ITZ drug substance gave
the two peaks in 1:1 ratio within experimental error (63%).
EXPERIMENTAL
Chemicals
ITZ and hydroxyitraconazole were purchased from Janssen Pharma-
ceutical (Titusville, NJ, USA); ketoconazole and boric acid were from
Sigma (St. Louis, MO, USA); HPLC-grade dioxane, isobutyric acid, N,N-
diisopropylethylamine (99.5%), and anhydrous sodium sulfate were
obtained from Aldrich (Milwaukee, WI, USA); trifluoroacetic acid (TFA,
HPLC grade) was from Chem-Impex International Inc. (Wood Dale, IL,
USA; O-(7-azabenzotriazol-1-yl)-N,N,N’,N’-tetramethyluronium hexafluor-
ophosphate (HATU) was obtained from Perseptive Biosystems (War-
rington, England); sodium hydrogen carbonate (A.C.S. grade), anhy-
drous sodium sulfate (A.C.S. grade), HPLC-grade acetonitrile, ethyl ace-
tate, hexane, methanol, and 2-propanol were purchased from Fisher
Scientific (Pittsburgh, PA, USA); dimethylformamide (DMF, sequencing
grade) was from Fisher; HPLC-grade methyl t-butyl ether (MTBE), etha-
nol, and denatured ethanol were purchased from Aldrich. Sodium hy-
droxide was obtained from JT Baker (Phillipsburg, NJ, USA).
Sample Preparation
The two epimer mixtures of ITZ were obtained via semipreparative
HPLC separation of ITZ carried out by Chiral Technologies, Inc (West
Chester, PA, USA). The semipreparative separation was based on our
earlier analytical separation.13 The separation used Chiralcel OD CSP
(10mm x 50 cm) and ethanol as the mobile phase at 258 C; and detection
was by UV at 270nm. The earlier-eluted peak was due to the dextrorota-
tory epimer mixture (2R,4S) and the later-eluting peak to the levorota-
tory epimer mixture (2S,4R) as determined with an optical-rotation detec-
tor13. The stereoisomeric purity of each mixture collected was >99%.
A 500-ll plasma sample was placed into a 10-ml glass tube with a PTF-
lined screw cap. A 100-ll aliquot of the internal standard solution (25
lg/ml in ethanol), 0.5 ml borate buffer (pH 10) and 5 ml of MTBE were
added. The mixture was then mixed on a Barnstead-Thermolyne Lab-
quake shaker (Fisher Scientific) for 20 min and centrifuged for five
minutes at 10,000 rpm. The organic layer was transferred into a silanized
tube and evaporated to dryness at 508 under a gentle stream of nitrogen.
The residue was reconstituted in 150 ll of ethanol and 75 ll of the solu-
tion was injected in the HPLC.
Synthesis of the Internal Standard (Scheme 1)
Extraction Recovery
(6)-N-Isobutyryldesacetylketoconazole (IBDAK, Fig. 1c), used as in-
ternal standard (IS), was synthesized by acylation of N-desacetylketoco-
nazole (DAK), in turn obtained by base-catalyzed deacetylation of keto-
conazole according to a published procedure.14 The acylation of DAK
was carried out as follows: Isobutyric acid (38ll, 0.41 mmole), HATU,
(155.9 mg, 0.41 mmole) were dissolved in 15 ml DMF in a 100-ml round-
bottom flask with magnetic stirring. N,N-Diisoproylethylamine (143ll,
0.82 mmole) was added and the mixture was allowed to stand at room
temperature for 3 min with magnetic stirring. (6)-DAK (200.7 mg, 0.41
mmole) was added and the solution was stirred overnight at room tem-
perature. The solvent was removed under vacuum (ca. 3 mm Hg) at 308
and the residue was purified by gradient-elution HPLC. The stationary
phase was ODS, contained in a 25 cm x 25 mm i.d. column. The gradient
elution used a mobile phase consisting of two components: solvent A
was 0.1% TFA in water and solvent B was 0.08% TFA in acetonitrile. The
initial composition was 70% A and 30% B, final composition 50% A and
Extraction recovery from plasma was determined at 1.25 lg/ml for
each stereoisomeric mixture, and the extraction procedure followed was
as described above but with the IS solution replaced with an equal vol-
ume of the vehicle. Before evaporation of the extraction solvent the IS
was added to the samples. Peak areas obtained upon HPLC analysis of
the extracts were compared to the peak areas obtained upon analysis of
unextracted samples containing equivalent concentrations. A similar pro-
cedure was used to determine the extraction recovery of the IS from
plasma, adding ITZ to the extract before the evaporation.
Calibration Curves
Calibration curves were constructed using plasma samples spiked
with ITZ at total ITZ concentrations of 50 ng/ml, 250 ng/ml, 1.00, 2.50,
5.00, 10.00 lg/ml (i.e. each stereoisomeric mixture having half the total
concentration).
Chirality DOI 10.1002/chir

498
PYRGAKI ET AL.
Fig. 2. HPLC chromatogram obtained upon analysis of a sample containing authentic standards of IS, ITZ, and OH-ITZ. The peaks are identified with IS for
the internal standard, dextro and levo for epimer mixture (2R,4S) and epimer mixture (2S,4R), respectively, for ITZ, and with OH-ITZ for the hydroxy metabolite.
Intraday and Interday Precision and Accuracy
Chromatography
Intraday precision and accuracy were determined at total ITZ concen-
trations of 50 ng/ml, 100 ng/ml, 250 ng/ml, and 10.0 lg/ml and interday
precision and accuracy at the concentrations of 50 ng/ml and 10.0 lg/ml.
A stereoselective HPLC separation of ITZ was developed
based on the Chiralcel OD chiral stationary phase (CSP).
The separation produced two peaks, i.e. each peak was due
to two stereoisomers contained in the drug (see DISCUS-
SION). The less-retained mixture was dextrorotatory while
the longer-retained material was levorotatory.13 Using vibra-
tional circular dichroism it was shown that the configuration
(at the dioxolane chiral centers) of the epimers of the first
peak was 2R,4S (A and C, Fig.1a) and the configuration of
the epimers of the second peak was 2S,4R (B and D,
Fig.1a).15 That is, in each epimeric mixture the two stereoiso-
mers differ only in the configuration of the sec-butyl group
(see DISCUSSION). However, the above separation proved
inadequate, because the hydroxy metabolite of ITZ (OH-ITZ)
(Fig. 1b) co-eluted with ITZ under these conditions. Chro-
matographic separation of the metabolite from the parent
drug proved to be a challenging problem that was solved
with the use of an additional column, Cyclobond I 2000 RSP
containing a derivatized cyclodextrin CSP, in combination
with gradient elution (Fig. 2). The peaks due to OH-ITZ elute
well after those of ITZ and no longer interfere. The resolu-
tion RS for the peaks of the two epimer mixtures was found
to have the value of 2.05. The IS was also well separated
from the ITZ and OH-ITZ peaks and eluted as a single peak,
i.e., its enantiomers were not resolved.
Specificity
Drug-free plasma was analyzed without the addition of internal stand-
ard to determine potential interference from endogenous components.
Five fluorescent drugs, ketoconazole, prazosin, propranolol, quinidine,
and quinine, were examined for potential interference.
Limit of Detection and Limit of Quantification
The limit of detection (LOD) was determined by injecting decreasing
amounts of ITZ into the HPLC system under the conditions of the analyt-
ical method. The lowest amount injected before reaching a signal-to-
noise ratio of 3 was considered the LOD in the assay. The lower limit of
quantification (LLOQ) was defined as the lowest plasma concentration of
each epimer mixture that gave an intraday CV ꢀ 15% and an intraday ac-
curacy (deviation from target) of ꢀ 10%.
Drug Administration
A healthy white 50-year-old male subject weighing 79.5 kg was admi-
nisted a single oral dose of 200 mg ITZ (Sporanox1 capsule). A plasma
sample was drawn 6 hours after the dose and analyzed using the method
described in the present article.
RESULTS
Extraction Recovery
The Internal Standard
The mean recovery for epimer mixture (2R,4S) was
90.060.1% (n55), for epimer mixture (2S,4R) 90.460.1%
(n55), and for the IS 90.560.0% (n53).
IBDAK, the internal standard, was synthesized using the
reaction sequence shown in Scheme. 1. The yield was 69%
and the matrix-assisted-laser-desorption/ionization-mass-
spectrum (MALDI MS) of the product was consistent with
the structure of IBDAK, giving an envelope of peaks corre-
sponding to the cluster of protonated molecular ions (M11)
arising from the chlorine isotopes (35, 37) for the two chlo-
rine atoms in the structure (C28H32Cl2N4O4, M11: 559, 561,
563). The purity of IBDAK thus obtained was found to be
>98% by HPLC.
Linearity
A linear relationship was found between the drug-to-inter-
nal standard peak-area ratios and the concentrations for both
stereoisomeric mixtures. Calibration curves were linear for
concentrations between 0.05 lg/ml and 10 lg/ml of total
ITZ. The goodness of fit (r2) for ITZ was consistently greater
Chirality DOI 10.1002/chir

STEREOSELECTIVE HPLC BIOANALYSIS OF ITRACONAZOLE
TABLE 1. Intraday precision and accuracy (n 5 6)
Precision % Accuracy %
499
In the present work a HPLC method was developed that is
suitable for the determination of the epimeric mixtures of
ITZ in plasma. In collaboration with a commercial supplier
we obtained the two epimer mixtures using a semiprepara-
tive separation. The identity of the diastereoisomeric compo-
nents in the separation was then established using vibra-
tional circular dichroism as (1)-2R,4S,2’R, (1)-2R,4S,2’S, (-)-
2S,4R,2’R, and (-)-2S,4R,2’S (where 2’ refers to the sec-butyl
stereogenic center, Fig. 1a).15
Concentration
(per ml)
(1)-ITZ
(-)-ITZ
(1)-ITZ
(-)-ITZ
25ng
50ng
14.12
11.16
11.22
7.94
10.92
6.96
10.17
7.76
24.00
25.60
4.80
20.04
25.96
4.80
125ng
5.00lg
0.66
0.05
In an earlier study we found that the commercially avail-
able Chiralcel OD CSP gives two peaks for ITZ, with a total
run time of ca. 25 min.13 Using optical-rotation monitoring it
was determined that the peak of shorter retention time was
produced by the two dextrorotatory stereoisomers, i.e., epi-
mer mixture (2R,4S) (A and C in Fig.1a) and the longer-
retained peak was due to the two levorotatory diastereoisom-
ers, i.e,, epimer mixture (2S,4R) (A and D, Fig. 1).15 It is
clear therefore that under the chromatographic conditions
used the Chiralcel OD CSP does not separate the epimers
differing in the configuration at the stereogenic center in the
sec-butyl group (Fig. 1a). The separation obtained can be
described as enantioselective since each set of enantiomeri-
cally related molecules is resolved (no enantiomerically
related substances co-elute) but only partially diastereoselec-
tive, inasmuch as each set of epimers at the sec-butyl group
stereogenic center remains unresolved. Obviously, some dia-
stereoselectivity is achieved since stereoisomer A (Fig. 1a) is
separated from D, and C is separated from B.
than 0.99. The mean for the slopes of five calibration curves
obtained on five separate days was 1.08060.076 for epimer
mixture (2R,4S) and 1.07360.073 for epimer mixture
(2S,4R).
Precision and Accuracy
The precision and accuracy data obtained are presented in
Tables 1 and 2.
LOD and LLOQ
The LOD was found to be 1.3 ng of each epimer mixture
injected and the LLOQ was 25 ng/ml for each epimer mix-
ture.
Typical chromatograms of plasma extracts are shown in
Fig. 3.
The Chiralcel OD-based chromatographic separation was
used in the development of a stereoselective analytical
method for the determination of the epimer mixtures of ITZ
in plasma. An immediate problem was encountered in that
OH-ITZ, the major metabolite of ITZ, coeluted with ITZ. This
was clearly not acceptable since the circulating concentra-
tions of the metabolite are significant16 and would interfere
with the determination of ITZ. The problem proved challeng-
ing inasmuch as the Chiralcel OD CSP did not separate ITZ
from OH-ITZ under several conditions examined. The
desired separation was finally achieved with the use of a
derivatized beta-cyclodextrin CSP, Cyclobond I 2000 RSP,
connected in front of and in series with the Chiralcel OD col-
umn. A gradient-elution-based separation was developed that
produced the best separation of the metabolite from the par-
ent drug without unduly extending the total run time (Fig.
2). The gradient also prevents the significant deterioration in
the signal-to-noise ratio which would be caused by band
broadening in isocratic elution from the 50-cm-long column
set. Under these conditions good peak shapes were
obtained; all components of interest (IS, ITZ, and OH-ITZ)
are well separated from each other; and baseline resolution
of the two epimer mixtures is achieved, with resolution RS 5
2.05 (Fig.2). We also found that the Cyclobond column alone
did not separate the two ITZ epimer mixtures, that is, the ste-
Assay Specificity
No interference was seen when drug-free plasma sam-
ples were analyzed without the addition of the IS (Fig.
3a). Potential interference by the fluorescent compounds
ketoconazole, propranolol, quinidine, and quinine was
examined. Ketoconazole did not interfere but propranolol,
quinidine, quinine, and prazosin were found to interfere
in the procedure.
DISCUSSION
Numerous non-stereoselective HPLC methods for the
determination of ITZ in plasma or serum have been pub-
lished16,17,18,19,20,21 but little has appeared on stereoselective
analytical procedures. Thus, only one article10 described an
appropriate procedure for a stereoselective analysis but this
method did not separate all four stereoisomers of ITZ and
provided pharmacokinetic data only for the two epimer mix-
tures. In fact, no method has appeared for the stereoselective
determination of all four stereoisomers in biological fluids, as
discussed under INTRODUCTION above. A practical separa-
tion of the four stereoisomers with sufficient sensitivity and a
reasonable run time for the processing of a large number of
samples has indeed remained a challenging task, and this
may be explained in part by the nature of the structure and
stereochemistry of ITZ. The two chiral moieties (the dioxo-
lane ring and the sec-butyl group, see Fig. 1a) are distant
from each other in a large and elongated molecule, and the
sec-butyl group is one of the smallest (non-hydrogen-isotope-
based) chiral groups. These factors make the development
of a practical chromatographic separation of the four stereo-
isomers challenging indeed. The published CE method12
that separates the four stereoisomers of ITZ has not been
applied to bioanalytical work and it is not clear whether it
would have the required sensitivity and specificity.
TABLE 2. Interday precision and accuracy (n 5 5)
Precision %
Accuracy %
Concentration
(per ml)
(1)-ITZ
(-)-ITZ
(1)-ITZ
(-)-ITZ
25ng
5.00lg
11.45
6.00
8.24
5.62
2.12
20.44
25.84
0.10
Chirality DOI 10.1002/chir

500
PYRGAKI ET AL.
Fig. 3. HPLC chromatograms obtained upon analysis of (a) drug-free plasma without addition of IS; (b) plasma sample spiked with ITZ at 2.5 lg/ml; (c)
plasma sample collected 6 hours after the administration of a 200-mg dose of ITZ to a male subject.
reoselectivity in our chromatographic system is provided by
the Chiralcel OD CSP.
As seen in Fig. 2, authentic OH-ITZ gave an envelope of
five incompletely resolved components. OH-ITZ contains
four stereogenic centers but because the trans configuration
at the dioxolane ring is excluded from the drug only eight
stereoisomers are possible for the metabolite. Thus, the sep-
aration of the eight stereoisomers of OH-ITZ under the chro-
Chirality DOI 10.1002/chir

STEREOSELECTIVE HPLC BIOANALYSIS OF ITRACONAZOLE
501
matographic condition used was incomplete. The chromato-
graphic mechanism responsible for the separation of ITZ
from OH-ITZ produced by the addition of the Cyclobond
CSP is not clear. It is possible that selective inclusion com-
plexation, a common mechanism for cyclodextrin-based
retention and selectivity, produces the separation, but an al-
ternative interaction of the hydroxyl group of OH-ITZ with
the unmodified external hydroxyl groups on the cyclodextrin
‘‘cone’’ cannot be ruled out as the mechanism for the
separation.
ITZ is a fluorescent compound and fluorescence has been
frequently used in HPLC-based non-stereoselective determi-
nations of the drug in biological fluids.16,17,19,20,21 In the pres-
ent procedure the fluorescence wavelengths were optimized
for the conditions used but were similar to those used in
other work.19,20,21
No interference was seen when drug-free plasma samples
were analyzed without the addition of the IS (Fig. 3a). Fig. 3b
shows the chromatogram obtained upon analysis of a plasma
sample spiked with ITZ. The procedure inherently incorporates
considerable specificity due to several features. Extraction at
pH 10 excludes acidic compounds ionized at that pH (e.g., the
over-the-counter medications aspirin, ibuprofen, naproxen,
etc.); the nonpolar extraction solvent excludes highly polar
compounds; the chromatographic separation eliminates inter-
ference from compounds that do not coelute with the com-
pounds of interest. In addition, fluorescence detection at spe-
cific excitation and emission wavelengths provides a great deal
of specificity, since compounds that are non-fluorescent (e.g.,
acetaminophen) or that are fluorescent at other wavelengths
remain undetected. Overall it is clear, therefore, that a large
number of commonly used drugs do not interfere. In addition,
we evaluated potential interference by several fluorescent com-
pounds that were expected on the basis of their structures to
follow ITZ during the sample preparation. Ketoconazole did not
interfere but propranolol, quinidine, quinine, and prazosin were
found to interfere in the procedure. Thus, the latter compounds
should not be co-administered with ITZ when this analytical
procedure is used for the determination of ITZ in biological flu-
ids. Also, in all applications of the method potential interference
from other agents that may be present should be evaluated.
Fig. 3c shows the chromatogram obtained after analysis of
a plasma sample collected 6 hours after the administration of
a single oral dose of 200 mg ITZ to a male subject. It is seen
that the two epimer mixtures differ considerably in their
plasma concentrations, in agreement with the data reported
by Kunze et al.10 Also noteworthy in Fig. 3c is the appear-
ance of the peaks due to OH-ITZ. When these peaks are
compared to the peaks produced by synthetic HO-ITZ (Fig.
2) it is seen that the metabolite formed in vivo consists of
fewer stereoisomers than the synthetic material. Here too,
our observations are in agreement with the earlier study.10
Based on literature reports of ITZ concentrations (sum of
all four stereoisomers) in the serum or plasma of patients
during multi-day dosing regimens or antifungal treatment,
the LLOQ (25 ng/ml for each epimer mixture) of our analyti-
cal method provides suitable sensitivity for such studies. For
example, treatment with a 200-mg daily oral ITZ dose during
a 3-month period produced plasma steady state concentra-
tions of total ITZ (sum of four stereoisomers) in the 800-1000
ng/ml range.8 In another study,22 200mg intravenous ITZ
(the approved formulation with hydroxypropyl-b-cylcodex-
trin) was administered in a 2-hour infusion to healthy sub-
jects. The dose was given every 12 hr during days 1 and 2
and every 24 hr during the remaining 5 days. On day 7 the
mean peak plasma concentration of total ITZ was 3349 ng/ml
and the trough 916 ng/ml. In the only published report10 on
the pharmacokinetics of the two epimer mixtures of ITZ, af-
ter a 7-day administration of an oral solution of 100 mg ITZ
to six healthy volunteers daily, the mean peak and trough
concentrations of the (2R,4S) epimer mixture were deter-
mined as 356.3 ng/ml and 33.2 ng/ml, respectively. The cor-
responding values for the (2S,4R) epimer mixture were 520.8
ng/ml and 128.4 ng/ml, respectively. Based on non-stereose-
lective analytical methods used in clinical studies, it has
been said that a trough value of at least 500 ng/ml (total)
ITZ is required for the successful prevention or treatment of
invasive fungal disease.7,22 It is clear that our LLOQ readily
meets this requirement and the stereochemical information
Identifying and obtaining a suitable internal standard was
also a challenging task, for several reasons. First, fluores-
cence was the detection method used, which obviously
requires a similarly fluorescent compound as IS; second, a
difficult separation of two peaks of the drug and several
peaks of the metabolite was involved which puts constraints
on the retention properties of the IS; third, the commercially
available racemic close analog of ITZ gave two peaks that
were not separable from those of ITZ under a variety of con-
ditions examined; fourth, no other suitable commercially
available compound was found. We were therefore faced
with the need to synthesize a suitable candidate for IS. Ca.
30 analogs of ITZ were synthesized, most of which were
found to be entirely unsuitable as IS for one reason or
another. However, one compound, IBDAK (Fig. 1c), a deriva-
tive of the related antifungal agent ketoconazole, did prove to
be highly suitable as IS. This compound was synthesized
(Scheme 1) via base-catalyzed N-deacetylation of ketocona-
zole and acylation of the resulting secondary amine with iso-
butyric acid, i.e, standard reactions of amides and secondary
amines, respectively, and a good yield of the final product
was obtained. Only the racemic mixture was synthesized but
no resolution of its enantiomers was obtained under the chro-
matographic conditions used. Thus, the IS eluted as a single
peak before the ITZ components (Fig. 2), with a suitable
retention time. The chemical structure of the IS is similar to
that of ITZ (Fig. 1), differing only in the nature of the moiety
on N4 of the piperazine ring and in the identity of the 5-mem-
bered nitrogen-containing substituent on the dioxolane ring
which is an imidazole group in the IS and a triazole in ITZ
(Fig. 1). Overall, the extraction, chromatographic, and fluo-
rescence properties of the IS are similar to those of ITZ and
the compound is suitable for the role of IS, as demonstrated
by the extraction-recovery, precision, and accuracy data
obtained. IBDAK is not available from commercial sources.
The sample preparation used is simple and includes addi-
tion of the IS, adjustment of the pH, and a single liquid-liquid
extraction, followed by evaporation of the solvent and recon-
stitution of the residue for injection into the HPLC system.
Extraction recovery was 90% for all three components, i.e.,
the two epimer mixtures and the IS. Linear calibration curves
were routinely obtained with r2>0.99 and the curves for the
epimer mixtures within each run were not significantly differ-
ent. The standard deviations of the means of the slopes of
the curves in five separate runs (different days) were <7.6%.
The method demonstrated acceptable intra- and interday pre-
cision and accuracy (Tables 1 and 2).
Chirality DOI 10.1002/chir

502
PYRGAKI ET AL.
our method can provide may be important in defining the
role of the various stereoisomers of ITZ in antifungal treat-
ment and prophylaxis.
agents such as ketoconazole, terconazole, saperconazole, etc.
Indeed, we recently applied a similar analytical procedure to
the enantiospecific determination in plasma of ketoconazole, a
drug administered as the racemic mixture, and the procedure
appeared to provide the basis of a useful method for the deter-
mination of this drug (unpublished data).
In single-dose pharmacokinetic studies the plasma concen-
trations achieved are of course significantly lower. For exam-
ple, Templeton et al reported23 that after an oral single dose
of 100 mg ITZ administered to six healthy volunteers the
mean peak plasma concentration of total ITZ was 255.4 ng/
ml. After six hours the concentration declined to ca. 56 ng/
ml and at 12 hr the concentration was ca. 35 ng/ml (esti-
mated from the concentration vs. time curves). In the
study10of the pharmacokinetics of the two epimer mixtures,
in the only subject reported on individually the peak concen-
trations on day 1 of the 7-day study (100-mg oral ITZ solu-
tion) were 213.1 ng/ml and 69.9 ng/ml for the the (2S,4R)
and the (2R,4S) mixtures, respectively, and the mean values
for six subjects were similar to those of the single individual.
However, the trough values after the first dose were below
the LLOQ of our procedure and in fact the use of LC-MS was
required in that study for their determination. The concentra-
tions in the last two studies (both of which used the lower
dose of 100 mg) should be contrasted to the results our study,
in which 6 hr after the administration of a single 200-mg oral
dose of ITZ the concentrations were 449 ng/ml and 134 ng/
ml for the (2S,4R) and (2R,4S) epimer mixtures, respectively.
It has been reported that ITZ displays dose-dependent phar-
macokinetics and marked intra- and interpatient variability.7
One or more of these factors, in combination with the higher,
200-mg, dose used in our study, may be the explanation for
the considerably higher plasma concentrations achieved in
our study compared to the concentrations found by Kunze
et al.10 who used 100-mg doses (it should be noted that the
200-mg dose is commonly used in pharmacokinetic studies
and in antifungal therapy, and even 300-mg doses have been
used22 in some studies; moreover, in antifungal treatment
daily oral doses of up to 400 mg can be used24). Concerning
the limitations of our LLOQ in single-low-dose pharmacoki-
netic studies, it would seem that applying LC-MS detection to
our method could provide the sensitivity needed for such
studies. Such a combination of our use of a highly suitable in-
ternal standard with LC-MS detection would likely further
enhance the utility and robustness of our analytical method.
An eventual full characterization of the stereoselectivity in
the disposition of ITZ will require a suitable analytical
method that resolves all four stereoisomers. Nevertheless,
the procedure described in the present communication
should prove a useful tool for studying the stereoselectivity
of the disposition of the drug insofar as the chirality of the
dioxolane ring is concerned. That the stereochemistry of the
dioxolane ring of ITZ is important for its actions is demon-
strated by the observation10 that only the (2R,4S) stereoiso-
mers (regardless of the configuration at the sec-butyl group)
are metabolized by CYP3A4. Our method has some advan-
tages over the other procedure10 in the literature, e.g., a sig-
nificantly shorter run time, the use of the considerably sim-
pler and less costly fluorescence detector instead of LC-MS,
and, importantly, the inclusion of an internal standard. How-
ever, LC-MS detection can provide higher sensitivity, which
may be required in some studies using single-low-dose drug
administration, as discussed above.
ACKNOWLEDGMENTS
Supported by grant 5-RO1 AI48000 from the National Insti-
tutes of Health, Bethesda, MD, USA.
REFERENCES
1. Groll AH, Shah PM, Mentzel C, Schneider M, Just-Nuebling G, Huebner
K. Trends in the postmortem epidemiology of invasive fungal infections
at a university hospital. J Infect 1996;33;23–32.
2. Singh N. Trends in the epidemiology of opportunistic fungal infections:
Predisposing factors and the impact of antimicrobial use practices. Clin
Infect Dis 2001;33:1692–1696.
3. Mislow K. Introduction to stereochemistry. New York: W. A. Benjamin;
1965. p. 91.
4. Baietto L, D’Avolio A, Ventimiglia G, De Rosa FG, Siccardi M, Simiele M,
Sciandra M, Di Perri G. Development, validation, and routine application
of a high-performance liquid chromatography method coupled with a sin-
gle mass detector for quantification of itraconazole, voriconazole, and
posaconazole in human plasma. Antimicrob Agents Chemother
2010;54:3408–3413.
5. Timmers GJ, Kessels LW, Wilhelm AJ, Veldkamp AI, Bosch TM, Beijnen
Jh, Huijgens PC. Effects of Cyclosporine A on single-dose pharmacoki-
netics of intravenous itraconazole in patients with hematologic malignan-
cies. Ther Drug Monit 2008;30:301–305.
6. Yao M, Srinivas N. Bioanalytical methods for the determination of itraco-
nazole and hydroxyitraconazole: overview from clinical pharmacology,
pharmacokinetic, and metabolism perspectives. Biomed Chromatogr
2009;23:677–691.
7. Poirier J-M, Cheymol G. Optimization of itraconazole therapy using tar-
get drug concentrations. Clin Pharmacokinet 1998;35:461–473.
8. Havu V, Brandt H, Heikkila¨ H, Hollmen A, Oksman R, Rantanen T, Saari
S, Stubb S, Turjanmaa K, Piepponen T. Continuous and intermittent
itraconazole dosing schedules for the treatment of onychomycosis: a
pharmacokinetic comparison. Brit J Dermatol 1999;140:96–101.
9. Koch P, Rossi RF Jr, Senanayake CH, and Wald SA, inventors, Sepracor,
Inc., assignee. 2R,4S-Hydroxyitraconazole isomers. U.S. Patent 6,147,077,
2000 Nov 14.
10. Kunze ,KL, Nelson WL, Kharasch ED, Thummel KE, and Isoherranen N.
Stereochemical aspects of itraconazole metabolism in vitro and in vivo.
Drug Metab Disp 2006;34:583–590.
11. Breadmore MC, Thormann W. Capillary electrophoresis evidence for the
stereoselective metabolism of itraconazole in man. Electrophoresis
2003;24:2588–2597.
12. Castro-Puyana M, Crego AL, and Marina ML. Separation and quantification
of the four stereoisomers of itraconazole in pharmaceutical formulations by
electrokinetic chromatography. Electrophoresis 2006;27:887–895.
13. Thienpont A, Gal J, Aeschlimann C, Felix G. Studies on stereoselective sep-
arations of the ‘‘azole’’ antifungal drugs ketoconazole and itraconazole using
HPLC and SFC on silica-based polysaccharides. Analusis 1999;27:713–718.
14. Chapman DR, Bauer L, Waller DP, and Zanewald LJD. Synthesis of diaster-
eomeric ketoconazole analogs .1. J Hererocyc Chem 1990;27:2063–2068.
15. Dunmire D, Freedman TB, Nafie LA, Aeschlimann C, Gerber JG, andGal.
Determination of the absolute configuration and solution conformation of
the antifungal agents ketoconazole, itraconazole, and miconazole with
vibrational circular dichroism. Chirality 2005;17:S101–S108.
16. M. Poirier JM, Lebot M, Descamps P, Levy M, and Cheymol G. Determi-
nation of itraconazole and its active metabolite in plasma by column liq-
uid-chromatography. Ther Drug Monit 1994;16:596–601.
17. Poirier JM and Cheymol G. A rapid and specific liquid chromatographic
assay for the determination of itraconazole and hydroxyitraconazole in
plasma. Ther Drug Monitor 1997;19;247–248.
Our procedure may also be applicable to the analysis of
drugs that have the same (or a very similar) dioxolane-based
chiral structural element, e.g., in several other azole antifungal
18. Koks CH, Sparidans RW, Lucassen G, Crommentuyn KM, and Beijnen JH.
Selective high-performance liquid chromatographic assay for itraconazole
Chirality DOI 10.1002/chir

STEREOSELECTIVE HPLC BIOANALYSIS OF ITRACONAZOLE
503
and hydroxyitraconazole in plasma from human immunodeficiency virus-
infected patients. J Chromatogr B 2002;767:103– 110.
hydroxyitraconazole in human plasma by high-performance liquid chro-
matography. J. Chromatog. A 2004;1031;307–313.
19. Compas D, Touw DJ, and deGoede PN. Rapid method for the analysis
of itraconazole and hydroxyitraconazole in serum by high-performance
liquid chromatography. J. Chromatogr B Biomed Appl 1996;687:453–
456.
22. Mouton JW, van Peer A, de Beule K, Van Vliet A, Donnelly JP, Soons
PA. Pharmacokinetics of itraconazole and hydroxyitraconazole in healthy
subjects after single and multiple doses of a novel formulation. Antimi-
crob Agent Chemother 2006;50:4096–4102.
20. Yao M, Chen L, and Srinivas NR. Quantitation of itraconazole in rat hep-
arinized plasma by liquid chromatography-mass spectrometry. J. Chroma-
togr B Biomed Appl 2001;752:9–16.
23. Templeton IE, Thummel KE, Kharasch ED, Kunze KL, Hoffer C, Nelson
WL, Isoherranen N. Clin Pharmacol Ther 2008;83:77–85.
24. Bennett JE. Antimicrobial agents. In: Hardman JG, Limbird LE editors:
Goodman and Gilman’s the pharmacological basis of therapeutics. 10th
Edition. New York: McGraw-Hill; 2001. p. 1304.
21. Srivatsan V, Dasgupta AK, Kale P, Datla RR, Soni D, Patel M, Patel R,
and Mavadhiya C. Simultaneous determination of itraconazole and
Chirality DOI 10.1002/chir