JOURNAL OF

CHROMATOGRAPHY B:
BIOMEDICALAPPLICATIONS
ELSEVIER Journal of ChromatographyB, 675 (1996) 287-294

High-performance liquid chromatographic assay for the

measurement of azathioprine in human serum samples
• a
Torsten Bmscheck ' , H a r t m u t M e y e r b, H a n s H e r b e r t W e l l h r n e r a
'~lnstitute of Toxicology, Medical School of Hannover, Konstanty-Gutschow-Str. 8, D-30625 Hannover, Germany
blnstitute of Organic Chemistry, University of Hannover, Schneiderberg IB, D-30167 Hannover. Germany
Received 27 February 1995; re+ised 31 July 1995: accepted 30 August 1995

Abstract

A quantitative, highly sensitive HPLC-based method for the direct measurement of azathioprine is described,
introducing a newly synthesized 9-methyl derivative of this immunosuppressant as internal standard in combination with
isocratic HPLC and UV-absorbance measurement at 285 nm. Analysis was performed on a RP18 select B column with
acetonitrile-0.01 M potassium phosphate buffer (12:88, v/v) at pH 2.3 as mobile phase. Results of precision analysis from
serum samples spiked with 3.125, 12.5, and 25 ng azathioprine, respectively, were (mean+_S.D.): 3.148+_0.259 ng (8.22%),
12.594_+0.571 ng (4.53%), and 25.016+-0.658 ng (2.63%) with C.V. values in parentheses for n = 5 . The accuracy of the
assay ranged from - 7 . 6 to 0.7% (expressed as % bias) tested on five consecutive days. The limit of quantification was at
2.5+-0.256 ng (C.V. 10.25%), thus allowing drug monitoring in long-term patients. The method can also be used to evaluate
individual pharmacokinetic parameters of a single patient, as well as for drug monitoring of a cohort of patients who suffer
from azathioprine-induced symptoms of toxicity. An example of the pharmacokinetic behaviour in an individual is given in
this paper.

Keywords : Azathioprine

1. I n t r o d u c t i o n mercaptopurine (6-MP), which acts as an an-

timetabolite to DNA and RNA synthesis (Fig. 1).
Azathioprine [6-( 1-methyl-4-nitroimidazol-5- This results in a mild antiproliferative effect of
ylthio)purine] is an immunosuppressant agent which azathioprine aimed at the proliferation of lympho-
has been clinically used for the past thirty years in cytes. Therefore, azathioprine is considered to be a
the prevention of organ rejection [1-5] and in the prodrug of 6-mercaptopurine. In vitro studies provide
management of several autoimmunological disorders evidence that azathioprine also acts directly on DNA
(e.g. rheumatoid arthritis, colitis ulcerosa, autoim- synthesis. So far, it is unknown to what extent
munological hepatitis) [6-9]. Its metabolism has azathioprine or its major metabolite 6-mercap-
been studied extensively [10-15]. In humans, it is topurine contribute to the immunosuppressant effect
mainly a non-enzymatic nucleophilic attack that and toxicity [ 1 1,16]. Azathioprine is orally applied in
converts azathioprine to the active metabolite 6 - doses from 0.5 to 3 mg/kg body weight. Intravenous
injection is avoided in patients because of unwanted
*Corresponding author. side effects and is of no practical use for long-term

0378-4347/96/$15.00 © 1996 Elsevier Science B.V. All rights reserved

SSDI 0378-4347(95)00365-7
288 T. Binscheck et al. / J. Chromatogr. B 675 (1996) 287-294

NO2 NO2 fine and detected by its UV absorbance [13,19-22].

Methods aiming at the detection of azathioprine itself
are not sensitive enough as to record the low plasma
levels ( < 10 ng/ml) expected during oral steady-state
treatment [21,23]. To achieve satisfactory precision
N
in quantification in this low concentration range, the
o) b) internal standard (I.S.) method was employed. Lin et
N al. [23] used benzazimide as I.S., which was unavail-
\
CH~ able to us. We therefore synthesized a new com-
'SH pound serving as an I.S. To qualify for use as an I.S.,
a compound would have to be more lipophilic than
N
azathioprine --resulting in an extended retention
time-- while the extraction behaviour would have to
c) N resemble that of azathioprine to yield constant ratios
of peak areas over a broad concentration range. A
Fig. 1. Structural formulas of azathioprine, 6-mercaptopurine
and 9-methylazathioprine. (a) Azathioprine [6-(1-methyl-4-
compound meeting these requirements was a 9-
nitroimidazol-5-ylthio)purine] is rapidly converted to (c) 6- methyl derivative of azathioprine (9-M-Aza), which
mercaptopurine by non-enzymatic, nucleophilic attack ($) was obtained by methylation of the parent substance
and successive cleavage of its internal thioether-bond. (b) A with methyliodide in absolute dimethylformamide in
9-methyl derivative of azathioprine [9-methyl-6-(1-methyl-4- the presence of potassium carbonate. The N-methyla-
nitroimidazol-5-ylthio)purine] was synthesized as an internal
standard for quantitative HPLC-analysis of serum samples. tion of azathioprine in the 9-position to 9-M-Aza
(Fig. lb) was demonstrated by spectroscopic and
molecular data. The melting point of 174---5°C
treatment. Some patients experience unpredictable distinguishes it from the previously synthesized 7-
acute myelosuppression, consecutive leucopenia and methyl derivative with a melting point of 254-5°C
hepatotoxicity [17,18], often necessitating dose re- [26].
duction or even cessation of therapy. It is not yet
clear whether the toxic effect of azathioprine is a
consequence of the pharmacokinetics of the drug
2. Experimental
(e.g. enhanced bioavailability, which might result in
significantly higher plasma levels in affected in-
2.1. Reagents
dividuals). Therefore we have developed a simple,
highly sensitive HPLC-based method for the direct
Azathioprine and Imurek were provided by Wel-
measurement of azathioprine levels in plasma sam-
lcome (Burgwedel, Germany). All solvents were
pies, allowing studies on steady state concentrations
HPLC-grade from Merck (Darmstadt, Germany).
in oral, long-term azathioprine therapy and evalua-
Water for eluent preparation was from a Millipore
tion of the drug's pharmacokinetics in individuals.
purification system. Methyliodide and absolute di-
Azathioprine (Mr 277.3, pK a 8.2, Fig. la) is a pale
methylformamide were obtained from Fluka (Neu-
yellow powder, insoluble in water and poorly soluble
Ulm, Germany).
in ethanol and chloroform, and moderately soluble in
acetonitrile and in dilute solutions of alkali hy-
droxides, although the latter is associated with some 2.2. Instrumentation and chromatographic
decomposition. Its melting point is about 238°C at conditions
which temperature it decomposes, thus evading
analysis by gas chromatography. Previously pub- The HPLC system used consisted of a Model 480
lished methods employing HPLC are based on the high-precision pump, a degassing device, an elec-
detection of its 6-MP metabolite (Fig. lc), which can tronically controlled column oven, and a photodiode-
be more easily extracted from plasma than azathiop- array UV detector (UVD 320) equipped with a 12-/xl
 T. Binscheck et al. / J. Chromatogr. B 675 (1996) 287-294 289

flow cell, all from Gynkotek (Munich, Germany). ( 4 × 4 . 6 mm, particle size 5 /xm). The mobile phase
Both Aza and 9-M-Aza were measured at a wave- consisted of 12% acetonitrile and 88% 0.01 M
length of 285 nm (Fig. 2). Data were sampled and potassium phosphate buffer (v/v), pH 2.3, and was
analyzed on a Gynkotek C-R5A integrator. For passed through a disposable filter (0.2-/~m pore
acquisition of spectral data and peak-purity trials the width) prior to use. After each run with a plasma
UVD 320 detector was used in combination with a sample the column was flushed for 2 min with
microcomputer-based analysis program (Gynkosoft acetonitrile-buffer (1:1) to wash out contaminating
V5.21). Chromatographic separation was performed substances with long retention times. The flow-rate
at 22°C on a Merck LiChrospher 60 RP-select B was 1000 /xl/min with t o = l . 9 min. The injection
column (250×4.6 mm, particle size 5/xm), protected volume was 0.2 ml. Under these conditions the
by a Merck LiChrospher 100 RP 18 precolumn number of theoretical plates was approximately
12 000 for calibration standards and approximately
8000 for serum samples. The precolumn was
a) b) changed every 100 serum sample runs and the
column was replaced when the plate number had
330 300 270 240 200 0.10 1,00 2. ~ 2
i e e i i 1 5
i
. 0
~
~
~
decreased below 6000.

,a n..i ¢

2.3. Synthesis of 9-methyl azathioprine (9-M-Aza)

A-400 mg (2.90 mmol) quantity of dried potas-

sium carbonate and 0.2 ml (3.2 mmol) of methyl-
iodide were added to a solution of 220 mg (0.79
mmol) of 6-(1-methyl-4-nitroimidazol-5-ylthio)-
purine in 7 ml absolute dimethylformamide at a
temperature of 0-5°C. After stirring in a nitrogen
atmosphere for 24 h at room temperature, the cloudy
reaction mixture was diluted with 14 ml of de-
mineralized water. The product was precipitated by
neutralization with 1 M hydrochloric acid and 1 M
sodium hydrogencarbonate solution, collected by
suction, washed with water and dried in a vacuum
yielding 99.4 mg (44%) of 9-methyl-6-(1-methyi-4-
nitroimidazol-5-ylthio)purine (9-M-Aza; yield not
optimized) with a melting point at 171_+2°C. Thin-
a -,n __ _1 mn b
layer chromatography (TLC) on silica gel 60 F 254
Fig. 2. (a) Contour plot and (b) chromatogram of serum (Merck) in acetone; Aza: RF=0.24; 9-M-Aza: RF=
sample containing 25 ng/ml of Aza and 9-M-Aza. (a) Con- 0.35. Analytically pure 9-M-Aza with the melting
tour plot of absorbance scans from 360 to 200 nm of a serum point of 174-175°C was obtained by repeating the
sample spiked with 25 ng/ml of azathioprine (Aza) and precipitation from dimethylformamide solution with
9-methylazathioprine (9-M-Aza); ** indicates the detection
wavelength of 285 nm; ($) shows an unknown serum com- water.
ponent with a peak absorbance at ca. 240 nm which would IR (KBr): 3400, 3120, 3103, 3070, 1587, 1566,
substantially interfere with the detection of 9-methyiazathiop- 1533, 1500, 1377, 1333, 1301, 1242, 1220, 933, 833,
rine at 220 nm, since both substances are not well resolved 641 cm l
under these chromatographic conditions. (b) Absorbance IH-NMR (DMSO-2H6)=3.45 (s, H20); 3.74 (s,
chromatogram of the same run at 280 nm. Besides Aza and
C H 3 ) ; 3.86 (s, C H 3 ) ; 8.29 (s, CH); 8.55 (s, CH); 8.64
9-M-Aza, an unknown serum component could be detected
at this wavelength (*). Superior chromatographic resolution (s, CH) ppm.
allows it to be clearly separated from both peaks of interest. 13C-NMR (DMSO-2H6)=29.91 (CH3); 33.08
290 T. Binscheck et al. / J. Chromatogr. B 675 (1996) 287-294

(CH3); 117.20; 130.44; 139.51 (CH); 146.82 (CH); eluent by vortex-mixing for 10 s, followed by 10 min
149.94; 150.22; 151.59 (CH); 155.14 ppm. sonification prior to injection.
MS (150°C) m/e=291 (0.03; M+), 275 (5), 261
(13), 246 (17), 245 (100), 230 (8), 204 (6), 177 2.6. Calibration and recovery of azathioprine from
(11), 160 (15), 150 (13), 133 (64), 106 (15), 79 serum samples
(18).
Anal. cal. for CIoH9NvO2.H20: C 38.84; H 3.59; An aliquot of 1 ml serum was spiked with
N 31.71; found: C 38.92; H 3.64; N 31.65. different concentrations of Aza (3.125 to 50 ng/ml,
All concentrations for 9-M-Aza given in this paper five samples per concentration) and 50 ng/ml of
are expressed as its monohydrate. 9-M-Aza. Peak areas for Aza and 9-M-Aza were
related to the amount/response ratios in chromato-
2.4. Preparations of standards for amount/ grams obtained from calibration standards. For
response calibration evaluation of accuracy and precision of the assay
three 5-ml serum aliquots were spiked with 3.125 ng,
Stock solution was 5% dimethylsulfoxide 12.5 ng, and 25 ng of azathioprine, respectively, and
(DMSO) in acetonitrile containing caffeine (1 mg/ analyzed on five consecutive days.
ml), azathioprine (1 mg/ml) and 9-methylazathiop-
rine (1 mg/ml). Caffeine was used as a hydrophilic 2. 7. Stability of azathioprine and 9-
component of the calibration standard mixture to methylazathioprine in blood
monitor elution of substances with short retention
times. Calibration standards were prepared by dilut- The thioether bond within the azathioprine mole-
ing the stock solution (625 ng/ml of each substance) cule is cleaved by non-enzymatic nucleophilic attack
with an appropriate amount of the eluent. Stored at (e.g. by glutathione) in the liver and during contact
room temperature, they were stable for four weeks with erythrocytes. No data exist for 9-
before they were prepared afresh. methylazathioprine, but its structural similarity to
azathioprine, especially in the region of the sulphur
2.5. Serum sample preparation bridge, suggests a similar breakdown for this deriva-
tive. To examine whether azathioprine and 9-
Serum was spiked by addition of appropriate methylazathioprine concentrations decrease during
amounts of stock solution to the desired serum prolonged contact with human blood, a 50-ml fresh
concentrations of azathioprine and 9-methylazathiop- blood sample was spiked with 25 ng/ml of each
rine, respectively, in screw-capped polyethylene substance and gently agitated at 37°C. Test samples
tubes. Prior to further processing, samples remained of 5 ml were drawn after 5, 15, 30, 60, 90, and 120
at room temperature for 30 min to allow equilibra- min and processed as described.
tion between free and serum-bound portions of the
compounds before further processing was initiated.
Blood samples were centrifuged immediately after 3. Results
collection. Following coagulation of the plasma
fraction the serum was removed and centrifuged 3.1. Chromatographic data for azathioprine and 9-
once to remove coagulated protein clots. methylazathioprine
Two 0.5-ml volumes of serum were transferred
into two polyethylene tubes and 4.5 ml of ethylace- Both substances were eluted from the column with
tate were added to each. After vigorous shaking sufficient resolution. Aza, which is more hydrophilic
(vortex) for 1 min, followed by centrifugation (! than 9-M-Aza had a shorter retention time than
min), the solvent phase were combined in a glass 9-M-Aza. The k' values were 6.20--_0.12 for Aza and
tube. This extraction was repeated once. The com- 11.53_+0.20 for 9-M-Aza as obtained from twelve
bined solvent phases were vacuum-dried at 35°C and consecutive runs of spiked serum samples. The
the residues were reconstituted with 0.25 ml of delayed elution of 9-methylazathioprine resulted in a
 T. Binscheck et al. / J. Chromatogr. B 675 (1996) 287-294 291

substantial broadening of the peak, but this did not 3.4. Calibration and quantitative analysis o f spiked
interfere with its correct integration. The main serum samples
metabolite of Aza is 6-mercaptopurine (6-MP). This
substance was not detectable at the used wavelength, Drug-free serum aliquots were spiked with the
because it has an absorbance maximum at 330 nm. indicated concentrations of Aza and 50 ng/ml of
Due to its short retention time (4.5 min) 6-MP was 9-M-Aza. Representative chromatograms of pro-
concealed by high-absorbance background induced cessed serum samples are depicted in Fig. 3b. Peak
by serum components which could not sufficiently be areas (PA) for both substances were determined and
separated from the injection sample without decreas- peak ratios (Aza/M-Aza=Q) were derived and the
ing the recovery of Aza. resulting calibration curve was analyzed for linearity
yielding:
Q
3.2. Spectral data of azathioprine and 9- Aza ( n g / m l -
methylazathioprine (1.815"10 =_+1.785"10 4)

Serum samples were spiked with 25 ng/ml of Aza +( 1.628" 10-2_+5.904 • 10 -3 )

and 9-M-Aza. Spectra covering the range from 220
to 400 nm were continuously scanned from 15 to 40 Means+_standard errors were derived from regres-
rain. Both substances had one peak absorbance sion analysis of serum spiked with 3.125, 6.25, 12.5,
below 220 nm and another peak absorbance at 285 18.75, 25, and 50 ng azathioprine/ml with five
nm, the latter appearing the most suitable for detec- independent samples identically processed for each
tion (Fig. 2). At that wavelength no disturbance of concentration Table 2. The accuracy and precision
the absorbance signal by contaminating serum con- was tested by analyzing serum samples containing
stituents could be observed. Peak-purity trials per- 3.125 ng, 12.5 ng, and 25 ng of azathioprine with 50
formed at 285 nm gave results >95% for both Aza ng of 9-M-Aza added to each sample on five
and 9-M-Aza. consecutive days. Results were (mean2S.D.):
3.14820.259 ng (8.22%), 12.59420.571 ng
(4.53%), and 25.016+0.658 ng (2.63%) with C.V.
3.3. Amount~response calibration values in parentheses. The limit of quantification
(LOQ) was determined as a concentration of Aza, at
Suitable volumes of calibration standards were which the C.V. values exceeded 10% and computed
chosen so as to allow Aza and 9-M-Aza injections in at 2.5 ng Aza/ml.
the range from 3.125 ng to 50 ng onto the column.
The peak areas (PA) of responses for both substances 3.5. Stabilit 3'
were determined, and a linear regression was per-
formed (for results see Table 1). Best results were Measured concentrations of Aza and 9-M-Aza
obtained by weighted fitting of the peak areas to were fitted by non-linear regression to a first-order
l/(amount) 2. The coefficient of variation (C.V.) for exponential function (Fig. 4 and Table 3). The
absolute peak areas was below 4.0% throughout all obtained k values were converted to half-life times
concentrations of both, Aza and 9-M-Aza. tl/2 for both substances. In serum samples stored at

Table 1
Parameters of linear regression from amount/response chromatograms
Compound Slope (mean+_S.D.) (mAU s ng ~) Intercept(mean+S.D.) (mAU s) r:
Azathioprine 4.338+_0.06988 0.8839+0.4352 0.99968
9-Methyl-azathioprinc 4.053+_0.09799 1.252+0.6245 0.99989
292 T. Binscheck et al. / J. Chromatogr. B 675 (1996) 287-294

a) 6,25 ng 2$ng 3O

CAza CAza 25
E
-m 20
8e~
•-~ 15
c
0

,o

~o ~, ~o m~ ~'o d0 do .,.

b) 0 / I I I I I
20 40 60 80 1O0 120
time [min]

Fig. 4. Time course of concentration decrease of (O) Aza and

(O) 9-M-Aza during prolonged contact with blood. For
experimental procedures and curve-fitting results please refer
to text. Values are means of three independent analysis
procedures; error bars indicate S.D.

3.6. Individual pharmacokinetic parameters

control
l 1~) 2o :~o min 10 20 30 min
A 30-year old, healthy, male volunteer (body
weight 82 kg) ingested 150 mg of azathioprine (three
Fig. 3. Chromatograms of (a) calibration runs and (b) serum
samples. (a) Suitable amounts of calibration standards were tablets of Imurek) at t=0. Serum samples were
used to achieve concentrations of caffeine (C), azathioprine drawn at times indicated in Fig. 5 and processed as
(Aza) and 9-methylazathioprine (9-M-Aza), respectively, as described. Serum concentrations increased from 15
indicated. For chromatographic conditions refer to text. Note to 45 rain reaching a maximum of ca. 33 ng/ml. This
the different calibration of the ordinate; increased sensitivity invasion was followed by a decline of azathioprine
of detection (figure on the left) leads to enhancement of
baseline signal noise. (b) Plain (control) and spiked serum concentration to ca. 13 ng/ml after 120 rain. The
samples were processed as described in the text and analyzed volume of distribution Vo, the fraction of resorption f
by HPLC. In the resulting control chromatogram, * and ** and the time constants for invasion k a and evasion k e
indicate retention times of Aza and 9-M-Aza. from the blood compartment were determined by
fitting the time course of serum concentrations to the
-20°C no significant decrease in the concentrations following model:
of Aza and 9-M-Aza could be detected for up to four
weeks.
Table 2
Recovery of azathioprine from serum samples
with t in min. Parameters estimated by nonlinear
Concentration Azathioprine 9-Methylazathioprine
(ng/ml) regression were: D = 1 5 0 mg, f = 1 2 . 8 % , VD=I.8
Mean C.V. Mean C.V. 1/kg, ka=0.029 and k e - - 0 . 0 3 . These data agree
(%) (%) (%) (%) well with pharmacokinetic data obtained from
6.25 39.54 9.88 n.d. several other sources [13,22,24]. Comparative
12.5 41.88 10.38 n.d studies using the i.v. administration route were
25 55.70 4.10 n.d not performed at this point of time due to the
50 63.57 9.9 75.25 11.26 substantial unwanted effects accompanied by i.v.
n = 5 for each concentration. application of the drug.
 T. Binscheck et al. / J. Chromatogr. B 675 (1996) 2 8 7 - 2 9 4 293

Table 3
Parameters of exponential decrease of azathioprine and 9-methylazathioprine during prolonged contact with blood
Compound C(, (mean-+S.D.) (ng/ml) k (mean_+S.D.) ( m i n - ' ) t~/2 (min)
Azathioprine 29.03-+1.638 10.16-10 3-+0.83.10 3 68
9-Methylazathioprine 25.70-+2.787 15.18.10-3-+2.34.10 3 45
C ( t ) _ C o . e k,
n 3 individually processed samples.

4. Discussion coextracted serum components. To set the detec-

tion wavelength for both substances to 285 nm
The HPLC method presented here is suffi- therefore represents a compromise with respect
ciently sensitive to quantitate azathioprine con- to sensitivity and specificity. The extraction pro-
centrations with an LOQ of 3 ng/ml, allowing cedure itself is rapid and avoids complex steps or
drug monitoring in long-term oral treatment with toxic reagents in sample preparation. It is desir-
this substance. The newly synthesized 9-methyl able to improve the method for simultaneous
derivative of azathioprine appeared to be a monitoring of the serum concentration of 6-MP,
compound suitable as internal standard, since it because it is the main metabolite of Aza. The
has a linear absorption response and an extrac- short retention time of 6-MP in this assay makes
tion behaviour similar to that of azathioprine. it difficult to clearly separate it from serum
Both compounds have an absorbance maximum constituents having similar UV-absorbances at
at 285 nm which allows the described technique 330 nm. Performing less sensitive preparation
also to be used with single wavelength detection. methods to clear contaminating substances from
Although there is an even higher absorbance at the sample has led to substantial loss of Aza and
220 nm we were concerned about high back- a decrease of recovery in preliminary experi-
ground absorbance at this wavelength caused by ments.
Studies on the stability of Aza and 9-M-Aza
indicate that the serum has to be separated from
40
blood samples as early as possible to prevent
35 degradation of azathioprine. This fact has also
~" 30 been emphasized by several other groups of
researchers [10,11,25]. Similarly, 9-
25
methylazathioprine undergoes a putative non-
20 enzymatic cleavage which appears even to be
accelerated. This may be due to the facilitated
entry of this more lipophilic substance into
erythrocytes. If processed, however, serum sam-
ples could be stored for at least four weeks with
0~ i
no significant decline of either Aza or 9-M-Aza
concentrations.
-5 I I I I

0 30 60 90 120 Unexpectedly the absorbed fraction of orally

applied azathioprine was as low as 12.8%. In-
time [min]
creased absorption would surely result in sig-
Fig. 5. Serum concentration of azathioprine after oral applica- nificantly higher serum concentrations of
tion. Experimental details are given in the text. Symbols azathioprine, which may be accompanied by
represent the mean of three individually processed serum
samples at indicated times (abscissa), error bars denote S.D.
toxic effects of the drug. At least two indepen-
The bars show the residuals after fitting the raw data to the dent processes contribute to the decline of
model described in the text. azathioprine blood levels. The first one is the
294 T. Binscheck et al. / J, Chromatogr. B 675 (1996) 287-294

passage of the substance from the intravascular [3[ D.F. du Toit and J.J. Heydenrych, S. Afr. Med. J., 70
space into the tissues and cells. The second (1986) 687.
[4] G.L. Chan, D.M. Canafax and C.A. Johnson, Phar-
mechanism that is involved in the decrease of
macotherapy, 7 (1987) 165.
azathioprine blood levels is the fast degradation [5] G.B. Elion and G.H. Hitchings, Handbook of Ex-
of the substance by reductive cleavage of the perimental Pharmacology, Vol. XXXVIII/2, Springer
thioether-bond. This was demonstrated by in- Verlag, Berlin, 1975.
cubating spiked blood samples and measuring [6] D.A. Fox and W.J. McCune, Concepts Immunopathol., 7
declining azathioprine concentrations. In future (1989) 20.
studies serum samples drawn from azathioprine- [7] T.J. Counihan and C. Feighery, Ir. J. Med. Sci., 160
(1991) 199.
treated patients should be analyzed and unto-
[8] V.C. Ho and D.M. Zloty, Dermatol. Clin., 11 (1993) 73.
ward effects monitored to detect a possible [9] D.J. Nashel, Med. Clin. North Am., 69 (1985) 817.
correlation between abnormal serum concentra- [10] K.G. Van Scoik, C.A. Johnson and W.R. Porter, Drug
tions of azathioprine and signs of toxicity. Metab. Rev., 16 (1985) 157.
Azathioprine is sometimes used in combination [11] L. Lennard, Eur. J. Clin. Pharmacol., 43 (1992) 329.
with other immunosuppressant drugs (e.g. [12] G.L.C. Chan, G.R. Erdmann, S.A. Gruber, P. Stock, S.
glucocorticoids or cyclosporin A). Analysis of Chen, N.L. Ascher and D.M. Canafax, Eur. J. Clin.
Pharmacol., 36 (1989) 265.
samples drawn from patients treated with combi-
[13] B. Odlind, P. Hartvig, B. Lindstr6m, G. L6nnerholm, G.
nation therapies gave no evidence for any inter- Tufveson and N. Grefberg, Int. J. Immunopharmacol.,
ference of these coadministered substances with 8(1) (1986) 1.
the presented assay. The study of pharmaco- [14] K . G . v . Scoik, C.A. Johnson and W.R. Porter, Drug
kinetics in a sufficient number of azathioprine- Metab. Rev., 16 (1985) 157.
treated individuals may allow insight into the [15] K.G.v. Scoik and C.A. Johnson, Drug Metab. Rev., 16
(1 and 2) (1985) 157.
underlying parameters and reveal how changes in
[16] C.E. Voogd, Mutat. Res., 221 (1989) 133.
these could account for alteration in the drug's [17] S.J. Rossi, T.J. Schroeder, S. Hariharan and M.R. First,
properties and effects in man. Drug Safety, 9 (1993) 104.
[18] D.I. Min and A.P. Monaco, Pharmacotherapy, 11 (1991)
119S.
Acknowledgments [19] V. Kurowski and H. Iven, Cancer Chemother. Phar-
macol., 28 (1991) 7.
[20] H.J. Wypior, A. Schilling, L. v. Meyer and W. Land, Res.
We thank Ch. Hotopp-Hergesell for her excel- Exp. Med. (Ber.), 181 (1882) 21.
lent technical assistance. Furthermore we are [21] J.L. Maddocks, Br. J. Clin. Pharmacol., 8 (1979) 273.
indebted to Dr. G. Erdmann and Dr. W. Martz [22] T.L. Ding and L.Z. Benet, Drug Metab. Dispos., 7(6)
for critical reading of the paper and helpful (1979) 373.
discussions. [23] S. Lin, K. Jessup, M. Floyd, T.-EF. Wang, T. v. Buren,
R.M. Caprioli and B.D. Kahan, Transplatation, 29
(1980) 290.
[24] R. Venkataramanan, K. Habucky, G.J. Burckart and R.J.
References Ptachcinski, Clin. Pharmacokin., 16 (1989) 134.
[25] L. Lennard and J.L. Maddocks, J. Pharm. Pharmacol., 35
[1] S.M. Lim and D.J. White, Ann. Acad. Med. Singapore, (1983) 15.
20 (1991) 144. [26] U.S.S.R. patent 384822 by P.M. Kochergin, V.S. Korsun-
[2] A.I. Lazarovits, Clin. Invest. Med., 12 (1989) 311. skii and V.S. Shlikhunova, 1973.