Characterization of oxalic acid derivatives as new metabolites of metamizol (dipyrone) in incubated hen's egg and human

Article abstract of DOI:10.1016/j.ejps.2005.11.010

Metamizol (dipyrone, 1), a widely used drug with effective analgesic and antispasmodic properties, shows severe side effects like agranulocytosis and anaphylactic shock reactions, the reasons of which are not known until today. After oral administration 1

Full text of DOI:10.1016/j.ejps.2005.11.010

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
european journal of pharmaceutical sciences 2 8 ( 2 0 0 6 ) 15–25
available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/ejps
Characterization of oxalic acid derivatives as new
metabolites of metamizol (dipyrone) in incubated
hen’s egg and human
Julia C. Wessela, Magdalena Matyjaa, Michael Neugebauerb, Heiko Kieferb,
Thomas Daldrupc, Fuad A. Tarbahc, Horst Webera,∗
a
¨
¨
¨
¨
Institut fur Pharmazeutische und Medizinische Chemie, Heinrich-Heine-Universitat Dusseldorf, Universitatsstrasse 1,
¨
40225 Dusseldorf, Germany
b
¨
Pharmazeutisches Institut, Rheinische Friedrich-Wilhelms-Universitat Bonn, Kreuzbergweg 26, 53115 Bonn, Germany
Institut fur Rechtsmedizin, Heinrich-Heine-Universitat Dusseldorf, Moorenstrasse 5, 40225 Dusseldorf, Germany
c
¨
¨
¨
¨
a r t i c l e i n f o
a b s t r a c t
Article history:
Metamizol (dipyrone, 1), a widely used drug with effective analgesic and antispasmodic
properties, shows severe side effects like agranulocytosis and anaphylactic shock reactions,
the reasons of which are not known until today. After oral administration 1 is completely
metabolized. All hitherto known metabolites have an intact pyrazolinone ring structure like
the parent compound and are completely extractable from urine with polar organic solvents.
However, only a fractional amount of the applied dosage can be recovered by this procedure.
To clarify the reason of this deficit of unknown metabolites we followed the hypothesis of
oxidative rupture of the heterocyclic ring during metabolism of 1. On the basis of former in
vitro results we now were able to identify in quality three oxalic acid derivatives and one
acetic acid phenylhydrazide as new metabolites of metamizol in the allantoic fluid (AF) of
incubated hen’s eggs as well as in human urine by means of GC–MS analysis and comparison
with unequivocally synthesized authentic reference compounds. Whereas the oxamazide 7,
the phenylhydrazide 8 and N-methyloxamic acid 9 are only present in trace concentrations
and therefore cannot account for the deficit in the balance of metabolites, the oxalic acid
monohydrazide 11 seems to be excreted in higher amount. But quantitative determination
of this new metabolite would be required to answer the open questions concerning the
biotransformation of metamizol and thereby to detect new facts about mode of action and
side effects of this drug.
Received 17 August 2005
Received in revised form 24
November 2005
Accepted 28 November 2005
Published on line 18 January 2006
Keywords:
Drug metabolism
Metamizol (dipyrone)
Pyrazolinone ring opening
Oxalic acid derivatives as drug
metabolites
© 2005 Elsevier B.V. All rights reserved.
1.
Introduction
used in several countries in Europe, Asia and South America.
In numerous developing countries it is the first line anal-
For 80 years, the sodium salt of metamizol (dipyrone 1) is a
widely used therapeutic drug. Due to its effective analgesic
and antispasmodic properties it is indicated for severe pain
and for colic of gall bladder and kidney and is commonly
gesic, because other analgetic drugs are not readily available
(Schonhofer et al., 2003; Bensenor, 2001). In contrast to this, 1
has been banned in other regions (USA, UK, Sweden) because
of its adverse drug reactions, notably agranulocytosis and ana-
∗
Corresponding author. Tel.: +49 211 8113846; fax: +49 211 204336.
E-mail address: Horst.Weber@uni-duesseldorf.de (H. Weber).
0928-0987/$ – see front matter © 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.ejps.2005.11.010

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
16
european journal of pharmaceutical sciences 2 8 ( 2 0 0 6 ) 15–25
phylactic shock reactions (Arznei-telegramm, 1974; Arellano
and Sacristan, 1990). In Germany and other countries the use
is restricted to special kinds of diseases.
While the International Study of Agranulocytosis and
Aplastic Anemia (IAAAS) estimated a low risk of agranulocyto-
sis after administration of 1 (1.1 cases per 1 million users, The
IAAAS Group, 1986), a Swedish investigation quoted a higher
rate of at least 1:1439 prescriptions (Hedenkalm and Spigset,
2002). The reasons for these severe side effects are not known
and the responsible metabolites could not yet been identified.
Even the mechanism of action of metamizol is still
unknown. It has been proposed that the antinocizeptive action
of metamizol is mediated by its inhibition of cyclooxygenase-
isoforms (Lanz et al., 1985; Weithmann and Alpermann, 1985;
Campos et al., 1999; Chandrasekharan et al., 2002). On the
other hand, a centrally mediated action has been suggested,
including opioid receptors (Taylor et al., 1998; Beirith et al.,
1998; Collares and Vinagre, 2003). In spite of the knowledge
that metamizol is a prodrug and spontaneously hydrolyzed
after oral intake in the gastric fluid to its effective metabolite 4-
methylaminoantipyrine (2), most of the in vitro investigations
has been carried out using 1 itself. Consequently, the results
of the studies using the prodrug are not suitable to clear this
problem. In addition, the metabolism of this classical drug is
to date not completely clarified. The metabolites, which gen-
erate the excellent analgetic action as well as the severe side
effects, are still unknown. Due to these unsolved problems the
therapeutic use of 1 is still discussed controversially.
Fig. 1 – Main metabolites of 1 in human, extractable from
urine with polar solvents.
short communication the authors affirmed the isolation of
in total “29 mostly polar peaks”. Using NMR-spectroscopy
and mass spectrometry they were able to propose chem-
ical structures for 26 of those peaks. Unfortunately, these
results have never been published elsewhere in detail but
it is noteworthy that one of the identified compounds was
assigned to the phenylhydrazide 8 (Fig. 2) and one other as
a derivative of 8 with unknown additional substituents. The
total amount of elucidated metabolites in urine of man was
added by these authors up to 86% of the given dose. So it still
remains a deficit of unknown metabolites and pyrazolinone
ring opening during metabolism of 1 seems to be possible.
Previous own experiments showed that the pyrazolinone ring
of 2 is easily opened by various oxidants under physiological
conditions (Duchstein et al., 1988; Weber and Wollenberg,
1988; Weber and Bresser, 1996). The resulting oxamazide 7
is gently hydrolyzed at pH 7.4 and 37 ◦C to give 1-methyl-2-
phenylacetohydrazide (8) and N-methyloxamic acid (9). The
hydrazide 8, however, is very unstable in the presence of
oxidants, forming a variety of following products (Bresser and
Weber, 1996). Therefore, the in vivo determination of 8 seems
not to be an appropriate method for proving the ring opening
of 2 as a real biotransformation reaction of 1.
For that reason, we followed another strategy considering
9 as a piece of evidence for the metabolic ring opening of 1.
This acid is a fragment resulting from the oxamazide 7 and
is – as far as known – no subject of further biotransformation
reactions.
In order to elucidate the metabolic stability of 9 we
designed a study with a small number of rats treated with an
orally administrated aqueous solution of this compound. For
this purpose, it was necessary to develop a reliable analytical
method for the qualitative and quantitative determination of
this oxamic acid in urine of rats, which was also suitable for
other biological samples. Later on we applied this method to
investigate the biotransformation of 1 and 7 and the recov-
ery of 9 by using incubated hen’s eggs. This is a widely used
In order to clarify these open questions it would be neces-
sary to have reliable facts with regard to the metabolic prop-
erties of this well known drug.
So far it is known that after oral administration of [14C]-
metamizol to man the absorption is rapid and uniform. After
48 h 96% and after 72 h 99% of the administrated radioactiv-
ity can be found in the human urine as quantity of radiation
without regarding special metabolites (Christ et al., 1973). Four
main metabolites 2, 3, 4 and 5 have been identified account-
ing only for 65–70% of a single dose of 480 mg of metami-
zol (Fig. 1). This apparent deficit of unidentified metabolites
could be attributable to inadequate analytical methods and/or
decomposition of metabolites (Volz and Kellner, 1980). Later
on, pharmacokinetic of metamizol was studied in healthy sub-
jects after various single doses of metamizole. The cumulative
amount of main metabolites 2–5 excreted in urine after 72 h
was determined by reversed-phase HPLC after extraction with
chloroform. But only less than 50% of the corresponding dose
could be recovered as defined metabolites (Vlahov et al., 1990).
These results have been confirmed by a series of studies pre-
sented in an extensive review of dipyrone pharmacokinetics
showing that the rate of metabolism and the composition of
metabolites obviously depend on individual differences of pro-
bationers like rapid or slow acetylators (Levy et al., 1995).
Further metabolites like 4-hydroxyantipyrin (6) as sulfate
or glucuronide conjugates and a number of other pyrazoli-
nones derived from metamizol by chemical modification
of the pyrazolinone substituents only account for minor
quantities. In the literature, non-heterocyclic metabolites
are mentioned only in a meeting abstract describing the
biotransformation after oral and inravenous administration
of 14C- and 13C-labeled metamizol (Volz et al., 1993). In this

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
european journal of pharmaceutical sciences 2 8 ( 2 0 0 6 ) 15–25
17
Fig. 2 – Synthesis of oxalic acid derivatives.
model of toxicological and metabolic research of xenobiotics
in biology and medicine (Rosenbruch, 1994; Lu¨ pke, 1987; Kiep
and Bekemeier, 1986; Neugebauer, 1997). This method offers
the possibility to apply a relatively high dosage of test com-
pounds. Our procedure based on previous experiments (Kiep
et al., 2002), who compared the biotransformation of 1 in incu-
bated hen’s eggs with the metabolism in human. The authors
concluded that the incubated hen’s egg is an appropriate ex
vivo model for simulating the biotransformation of 1 in vivo,
because they were able to recover all the well-known metabo-
lites of 1 with an intact pyrazolinone ring. On the basis of
these findings we decided to use this model to investigate the
production of oxamic acids and oxalylhydrazides during bio-
transformation of 1. In this connection, we also elaborated an
analytical method for the detection of 7 and 8 to control the
production of these compounds in the incubated hen’s eggs
after application of 1.
Finally, we explored the formation of oxalic acid derivatives
from 1 in human by analyzing the urine of two probation-
ers after a single oral dose of each 1 g of metamizol. By this
way, we intended to obtain the definitive evidence for the for-
mation of oxalic acid derivatives in the biotransformation of
1 in human. To this end, it was necessary to synthesize the
phenylhydrazides 7 and 8, the oxalic acid derivatives 9 and
11 and their volatile esters 10 and 12 as authentic reference
substances for GC–MS analysis (Fig. 2).
ter (Bruker Analytik GmbH, Rheinstetten, Germany) and peak
positions are given in parts per million (ı) downfield from
tetramethylsilane as internal standard, whereas coupling con-
stants (J) are given in Hertz. The HMBC-spectrum was recorded
on a Bruker DHX (500 MHz). Infrared spectra were measured on
a Perkin-Elmer 1600 FT-IR-Spectrometer. Mass spectra of refer-
ence compounds were obtained by a Finnigan MAT 4000 (Finni-
gan GmbH, Puchheim/Munchen, Germany) and by GC–MS
analysis (see Section 2.3.5.3). The elemental analyses were
determined by a Perkin-Elmer PE 2400 CHN Elemental Ana-
¨
lyzer (Perkin-Elmer Instruments GmbH, Rodgau-Jugesheim,
Germany). Column chromatography was performed using sil-
ica gel 60, 230–400 mesh ASTM (Merck Eurolab GmbH, Darm-
stadt, Germany). Solutions in organic solvents were dried over
anhydrous sodium sulfate. Melting points were obtained in
open capillary tubes and are uncorrected. Strongly acid cation
exchanger (Amberlite® IRA-120) and strongly basic anion
exchanger (Amberlite® IRA-402) were obtained from Merck.
¨
2.2.
Synthesis of compounds
2.2.1. Methylaminooxoacetic
acid-2-acetyl-2-methyl-1-phenylhydrazide (7)
Synthesis and spectral data according to literature (Weber and
Wollenberg, 1988).
2.2.2. 1-Methyl-2-phenylacetohydrazide (8)
Synthesis and spectral data according to literature (Weber and
Bresser, 1996).
2.
Materials and methods1
2.1.
Materials
2.2.3. N-Methyloxamic acid (9)
The synthesis based on a procedure described in literature
(Wallach, 1877).
Diethyl oxalate 14.6 g (0.1 mol) was dissolved in 50 ml
ethanol. Dimethylamine (0.1 mol, 40% aqueous solution) was
added and stirred for 4 h at room temperature. This mixture
Nuclear magnetic resonance (1H NMR and 13C NMR) spectra
were determined by a Bruker AC-200 (200 MHz) spectrome-
1
More details are given in the dissertation thesis Wessel (2004).

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
18
european journal of pharmaceutical sciences 2 8 ( 2 0 0 6 ) 15–25
2.2.6. Methyl-2-(Nꢀ-acetyl-Nꢀ-methyl-N-
phenyl)-hydrazino-2-oxo-acetate (12)
was alkalized by adding 0.1 mol sodium hydroxide in 50 ml
water and heated to reflux for 1 h. After cooling the mixture
was diluted with 200 ml water and extracted with chloroform.
The extracted aqueous phase was acidified by adding 10%
hydrochloric acid and evaporated to dryness. The dry residue
was triturated with dry magnesium sulfate and extracted
several times with diethyl ether. After filtration the organic
layer was evaporated in vacuum giving 4.12 g of a solid (yield
40%). The product was purified by crystallization from ace-
tone/petroleum ether, mp = 150–151 ◦C. 1H NMR (DMSO-d6):
ı = 2.65 (d, J = 4.8 Hz, 3H, NHCH3), 8.8 (broad s, 1H, NHCH3),
13.6 (broad s, 1H, COOH); 13C NMR (DMSO-d6): ı = 25.8 (NHCH3),
158.8 (CONH, correlation on the basis of HMBC), 162.0 (COOH,
A solution of 0.25 g (1.1 mmol) 11 in 5 ml methanol was
mixed with a diazomethane solution in diethyl ether (De
Boer and Backer, 1963) until the mixture retained yellow.
After stirring for 20 min methanol was removed by vac-
uum. The residual viscous oil (yield 100%) was used with-
out further purification. 1H NMR (DMSO-d6): ı = 2.06/2.14/2.19
(s,
(s,
(s,
3
4
4
rotameric isoforms, H, NCOCH3), 3.04/3.11/3.26/3.30
rotameric isoforms, 3H, NCH3), 3.56/3.64/3.83/3.88
rotameric isoforms, 3H, COOCH3), 7.3–7.6 (m, 5H,
ar). IR (KBr): ꢀ = 1751, 1704, 1677 cm−1 (C O). MS: m/z 59
(70%), 77 (86%), 92 (24%), 121 (100%), 208 (17%), 250 (M+,
0.3%). C, H, N: C12H14N2O4 (250.25), C calcd. 57.59, found
correlation on the basis of HMBC). IR (KBr): ꢀ = 1758, 1677 cm−1
.
MS: m/z 44 (100%), 58 (84%), 59 (43%), 103 (M•+, 3%). C, H, N:
C3H5NO3 (103.08) C calcd. 34.96, found 34.67, H calcd. 4.89,
found 4.88, N calcd. 13.59, found 13.32. Calculation of pka = 2.37
by Advanced Chemistry Development ACD, Software Solaris V
4.67, Chemical Abstracts.
57.23,
10.95.
H calcd. 5.64, found 5.69, N calcd. 11.19, found
2.3.
Methods
2.3.1. Collection of urine from rats after application
of 9
2.2.4. N-Methyloxamic acid methyl ester (10)
A solution of diazomethane was prepared by a method
described elsewhere (De Boer and Backer, 1963). To a solu-
tion of 9 (0.31 g, 3 mmol) in 5 ml methanol the solution of
diazomethane in diethyl ether was added until the mix-
ture retained yellow. After stirring for 20 min methanol was
removed by vacuum. The residual solid weights 0.35 g (yield
100%). The product was purified by crystallization from diethyl
ether/petroleum ether, mp = 78 ◦C. 1H NMR (DMSO-d6): ı = 2.66
(d, 3H, NHCH3), 3.76 (s, 3H, OCH3), 8.9 (broad, s, 1H, NHCH3).
IR (KBr): ꢀ = 1688 cm−1 (C O). MS: m/z 58 (100%), 89 (19%),
117 (M•+, 3%); GC–MS: m/z 58 (100%), 60 (45%), 117 (2%),
118 ([M + 1]+, 10%). C, H, N: C4H7NO3 (117.10), C calcd. 41.03,
found 40.89, H calcd. 6.03, found 5.95, N calcd. 11.96, found
11.67.
Wistar rats (female, 200 g) were divided into a verum group
(n = 3) and a control group (n = 3). The animals were kept under
controlled light/dark cycle and room temperature and fed
on standard pellet diet. One milligrams of 9 in 1 ml water
was administered by gavage to rats of the verum group. The
animals of the control group received 1 ml of pure water
under the same conditions. Rats were then kept in individ-
ual metabolic cages for collection of urine at 12, 24, 36 and
48 h after application. The collected samples were stored at
−20 ◦ C. These experiments were approved by the local ethical
committee.
2.3.2. Collection of allantoic fluid (AF) from incubated
hen’s egg after inoculation of 1, 7 and 9
For this purpose, the appropriate xenobiotic was inoculated
into the fertilized hen’s egg before or after the beginning
of incubation. After completing the incubation period the
metabolites can be identified in the AF, the excretion medium
of the embryo. This method depends on the description in
literature (Neugebauer, 1995). Fertilized eggs (White Leghorn)
were obtained from Johann Figgemeier, D-33397 Rietberg, Ger-
many. Before starting the incubation 0.1 ml of an aqueous
solution containing 5 or 10 mg of the corresponding substrate
was inoculated into the albumen of 11 eggs. In a control group
the same volume of pure water was inoculated. The eggs
were placed horizontally in a thermostatic oven and incu-
bated at 38.0 0.5 ◦C and 64 5% of relative humidity. They
were rotated every 3 h. The experiments were stopped on
the 11th or 15th day of incubation by storing the eggs for
30 min at −20 ◦C. The aspirated AF was pooled and stored
at −20 ◦C.
2.2.5. Oxalic
acid-(Nꢀ-acetyl-Nꢀ-methyl-N-phenyl)-hydrazide (11)
A solution of 1.05 g (3 mmol) 1 in 75 ml buffer (pH 7.4) and
30 mg horseradish peroxidase in 75 ml of the same buffer were
mixed. A dilution of 3.6 ml 30% hydrogen peroxide (36 mmol) in
300 ml water was added very slowly drop by drop. After stirring
for 24 h at room temperature the mixture was extracted with
dichloromethane and the aqueous solution acidified to pH 5
by adding 5% potassium hydrogen sulfate solution. A second
extraction with dichloromethane followed. Both organic layers
were dispensed. The remaining aqueous phase was brought
to dryness by evaporation and the dry residue was triturated
with excess potassium hydrogen sulfate and then extracted
with chloroform:isopropanol 9:1. The combined extracts were
dried and the solvents evaporated in vacuum yielding 0.25 g of
a colorless and viscous oil (yield 35%), crystallized as ammo-
nium salt of 11, mp = 183–184 ◦C. 1H NMR (DMSO-d6): ı = 2.01
(s, 3H, NCOCH3), 3.04 (s, 3H, NCH3), 7.2 (m, 1H, p-ar), 7.4
(m, 4H, o-, m-ar). IR (KBr): 1681, 1624 cm−1 (C O). C, H, N:
C11H15N3O4 (253.26), C calcd. 52.17, found 51.82, H calcd. 5.97,
found 5.76, N calcd. 16.59, found 16.31. The acid 11 is sta-
ble towards hydrolysis in refluxing acidic medium (pH 3, 1 h)
but is readily hydrolyzed in alkaline solution to 8 and oxalic
acid.
2.3.3. Collection of urine from human after a single
dose of 1
Metamizol (1) (two tablets of each 500 mg novaminsulfon
ratiopharm®) was administrated to two volunteers (male, 62a,
70 kg, and female, 27a, 55 kg) as a single oral dose. The 48 h
urine was collected in two fractions of each 24 h and stored in
plastic bottles at −20 ◦C.