Systematic study on the chemical stability of the prodrug antitumor agent carzelesin (U-80,244)

Article abstract of DOI:10.1021/js960005n

The chemical stability of the novel anticancer agent carzelesin in aqueous buffer/acetonitrile (1:1, v/v) mixtures has been investigated utilizing a stability-indicating reversed-phase high-performance liquid chromatographic assay. The degradation kinetics of carzelesin has been studied as a function of pH, buffer composition, ionic strength, and temperature. Degradation of carzelesin follows (pseudo-) first-order kinetics. A pH-rate profile, using rate constants extrapolated to zero buffer concentration, was constructed demonstrating that carzelesin is most stable in the pH region 1-4. The degradation rate of carzelesin was not significantly affected by buffer components and by the ionic strength. In addition to the formation of the degradation products U-76,073, U-76,074, and aniline in alkaline medium and in acetate buffer solution, another degradation product was formed in acetate buffer solution. In perchloric acid buffer solution (pH* < 3), U-76,073 and U-76,074 could not be detected as degradation products.

Full text of DOI:10.1021/js960005n

+
+
Systematic Study on the Chemical Stability of the Prodrug Antitumor Agent
Carzelesin (U-80,244)
X
†
†
J. D. JONKMAN
-DE
V
RIES* , W. G. DOPPENBERG†, R. E. C. HENRAR‡, A. BULT
,
AND J. H. BEIJNEN
*
Received January 8, 1996, from the *Department of Pharmacy, Slotervaart Hospital/Netherlands Cancer Institute, Louwesweg 6, 1066 EC
Amsterdam, The Netherlands, †Department of Pharmaceutical Analysis, Faculty of Pharmacy, Utrecht University, Sorbonnelaan 16, 3584
CA Utrecht, The Netherlands, and ‡European Organization for the Research and Treatment of Cancer
s
New Drug Development Office,
Free University Hospital, Gebouw Zuid, Amstelveenseweg 601, 1081 BT Amsterdam, The Netherlands.
Final revised manuscript
received July 9, 1996 .
Accepted for publication July 10, 1996X.
of the drug in this formulation was dependent upon the
interbatch variability of the excipients used in the vehicle.
Accelerated stability studies (at 40 °C) of carzelesin in the
PET formulation showed that the variable stability charac-
teristics were caused mainly by the pH differences between
batches of PEG 400.7 This stimulated us to investigate
systematically the chemical stability of carzelesin in solutions,
which can aid to further optimize the pharmaceutical formu-
lation, and to get more insight into the degradation mecha-
nism.
Abstract
in aqueous buffer/acetonitrile (1:1, v/v) mixtures has been investigated
utilizing stability-indicating reversed-phase high-performance liquid
chromatographic assay. The degradation kinetics of carzelesin has been
studied as a function of pH, buffer composition, ionic strength, and
temperature. Degradation of carzelesin follows (pseudo-) first-order
0 The chemical stability of the novel anticancer agent carzelesin
a
kinetics. A pH
buffer concentration, was constructed demonstrating that carzelesin is
most stable in the pH region 1 4. The degradation rate of carzelesin
−rate profile, using rate constants extrapolated to zero
−
was not significantly affected by buffer components and by the ionic
strength. In addition to the formation of the degradation products U-76,-
073, U-76,074, and aniline in alkaline medium and in acetate buffer
solution, another degradation product was formed in acetate buffer
solution. In perchloric acid buffer solution (pH* < 3), U-76,073 and U-76,-
074 could not be detected as degradation products.
Experimental Section
Ma ter ia lssChemicalssCarzelesin, U-76,073, and U-76,074 were
synthesized by The Upjohn Company (Kalamazoo, MI)8,9 and provided
through the New Drug Development Office (NDDO) of the EORTC
(Amsterdam, The Netherlands). Aniline was obtained from Aldrich
Chemical Co. Ltd. (Axel, The Netherlands). All other chemicals used
were of analytical grade, and deionized water was used through-
out.
Buffer SolutionssDue to the poor aqueous solubility of carzelesin,
the kinetic studies were performed in water/acetonitrile mixtures (1:
1, v/v). The aqueous components were 0.02 or 0.03 M buffer solutions,
except in the experiments where the effects of buffer concentration
were studied. The following buffers were used: 0 e pH < 3, perchloric
acid; 3 e pH < 6, acetate; 6 e pH < 9, phosphate; and pH g 9,
carbonate. The pH values of the aqueous buffer solutions were
measured using an Ingold Semimicro combination pH electrode and
a Model 654 pH meter (Metrohm AG, Herisau, Switzerland), which
had been calibrated on fully aqueous standard solutions. The pH
values of the buffer solutions were adjusted with either 0.1 M
perchloric acid solution or 0.1 M sodium hydroxide solution to the
appropriate pH values. Then, acetonitrile was added to the buffer
solutions and the apparent pH values (called pH*) of these mixtures
(except for pH* 0, which was calculated), at the study temperature,
were measured using the same electrode system.
Introduction
Carzelesin (U-80,244; NSC 619029; Figure 1) is a syntheti-
cally derived cyclopropylpyrroloindole (CPI) analog of the
highly potent, alkylating, antitumor agent antibiotic CC-
1065.1 The CPI analogs are DNA minor-groove binders
containing a cyclopropyl group, which mediates N3-adenine
covalent adduct formation in a sequence-selective fashion
with no intercalation.1 It has been speculated that these
CPI derivatives selectively inhibit the binding of regulatory
proteins to DNA, thereby modifying the transcription of
specific genes that are important for neoplastic cell growth.2
Carzelesin showed activity against a broad panel of human
tumor xenografts in preclinical in vivo studies in mice1
and in vitro in a number of gynecologic cancer cell lines.3
Carzelesin was designed to be an inactive prodrug, requir-
ing chemical or enzymatic activation to the DNA-reactive
form.1 Activation of carzelesin requires two steps, i.e., (a)
hydrolysis of the carbanilate substituent to form U-76,073
followed by (b) ring closure of the chloromethyl group to
form the cyclopropyl-containing DNA-reactive U-76,074 (Fig-
ure 1).
A constant ionic strength (µ) of 0.3 was maintained for each solution
by addition of the appropriate amount of sodium chloride, except in
experiments where the effect of the ionic strength on the degradation
of carzelesin was investigated and where the OH- and H+ concentra-
tions were higher than 0.3 M.
Degr a d a tion Kin eticssKinetic MeasurementssThe kinetic stud-
ies were carried out over at least 3 half-lives in the dark (in order
to prevent possible photolytic degradation), and the temperature
of the study was 25 °C in the pH* range 6-10, 60 °C in the pH*
range 1.5-5.5, and 70 °C at pH* 0. The reaction was initiated by
addition of 100 µL of a solution of carzelesin in absolute ethanol (1
mg/mL) to 5.0 mL of preheated buffer/acetonitrile solution (1:1, v/v)
to give an initial concentration of approximately 20 µg/mL (2.7 × 10-5
M). The reaction solutions were kept in screw-capped brown glass
vials in a thermostatically controlled water bath. At appropriate time
intervals 200-µL samples were withdrawn and directly analyzed for
undegraded carzelesin and the degradation products U-76,073 and
U-76,074, by a stability-indicating high-performance liquid chroma-
tography (HPLC) assay. All kinetic studies were performed in
duplicate.
Recently, a study4 has been conducted dealing with the
characterization and pharmaceutical formulation of carzelesin
according to the “Guidelines for the formulation of investiga-
tional cytotoxic drugs”, which were drawn up by the J oint
Formulation Working Party of the EORTC/CRC/NCI.5
A
stable parenteral formulation of carzelesin in a polyethylene
glycol 400 (PEG 400)/absolute ethanol/Tween 80 (6:3:1, v/v/v;
PET formulation) solution was developed,4 which is now being
tested in phase I clinical trials.6 During the manufacturing
process of clinical batches of carzelesin in the PET formu-
lation, observations were made indicating that the stability
Influence of TemperaturesThe effect of temperature on the
degradation of carzelesin was studied at pH* 1.5 (0.1 M perchloric
X Abstract published in Advance ACS Abstracts, October 1, 1996.
© 1996, American Chemical Society and
American Pharmaceutical Association
S0022-3549(96)00005-6 CCC: $12.00
Journal of Pharmaceutical Sciences / 1227
Vol. 85, No. 11, November 1996

+
+
Figure 1sChemical structures of carzelesin and the major degradation products U-76,073 and U-76,074.
acid solution; µ ) 0.3) in the temperature range 60-80 °C and
at pH* 7.2 (0.02 M phosphate buffer; µ ) 0.3) between 25 and
45 °C.
Calibration curves of carzelesin and the degradation products in
the mobile phase were linear (r > 0.990) in the concentration range
of interest.
Ultraviolet/ Visible (UV/ Vis) Spectrophotometry-UV/vis absorption
spectra of carzelesin and degradation mixtures were recorded with a
Model GBC 918 UV/vis spectrophotometer (GBC Scientific Equipment
Ltd., Victoria, Australia). The 1-cm quartz cells were kept in a
thermostatically controlled cell compartment at 25 or 40 °C. Degra-
dation reactions were initiated by addition of 40 µL of a solution of
carzelesin in absolute ethanol (1 mg/mL) to 3.0 mL of preheated
buffer/acetonitrile solution (1:1, v/v) to give an initial concentration
of approximately 13 µg/mL (1.8 × 10-5 M).
An a lytica l P r oced u r essHPLCsThe liquid chromatographic sys-
tem consisted of a Model 510 HPLC pump, a Model 441 absorbance
detector (both from Millipore Waters Chromatography, Milford, MA),
a Model ISS-100 autosampler (Perkin-Elmer, Uberlingen, Germany),
and a Model SP 4270 integrator (Thermo Separation Products (TSP),
San J ose, CA). Ultraviolet detection at 254 nm was used. A µ
Bondapak C-18 analytical column (300 × 3.9 mm i.d.; particle size
10 µm; Millipore Waters Chromatography) was used. The mobile
phase consisted of 2 mM phosphate buffer pH 6.5/acetonitrile (25:65,
v/v). The flow rate was 1.0 mL/min, and the injection volume was 20
µL. Chromatography was performed at room temperature. Quan-
titation of carzelesin, U-76,073, and U-76,074 was based on peak area
measurements using a Model SP 4270 integrator (TSP). Chromato-
graphic data processing was performed with the WINner/286 data
system (TSP).
In addition, UV/vis absorbance spectra of reference standards
of the degradation products were recorded for identification pur-
poses.
Identification of Degradation ProductssFor the confirmation of the
identity of the degradation products, U-76,073 and U-76,074, and for
recording spectra of other degradation products, a Model Waters 996
1228 / Journal of Pharmaceutical Sciences
Vol. 85, No. 11, November 1996

+
+
Figure 2sHPLC chromatogram of partly degraded (t ) t1 2) carzelesin in acidic
/
medium (0.1 M perchloric acid buffer solution; pH* 1.5; µ ) 0.3). Peak 1:
carzelesin.
Figure 4
s
First-order plots of carzelesin at pH* 0.0 (hourglass symbol), 4.6 (0×),
5.2 ( ), 7.7 (0), and 9.8 (9), respectively.
2
Figure 3sHPLC chromatogram of partly degraded (t ) t1 2) carzelesin in alkaline
/
medium (0.03 M phosphate buffer; pH* 7.2; µ ) 0.3). Peak 1: carzelesin. Peak
2: U-76,073. Peak 3: U-76,074. Peak 4: aniline.
Figure 5
s
Typical degradation curve of carzelesin (
2) and appearance curves of
photodiode array (PDA) detector (Millipore Waters Chromatography)
was used on-line with the HPLC system.
U-76,073 (
9
) and U-76,074 ( ) (concentrations have been calculated as percent
0
of the initial carzelesin peak area at t
µ ) 0.3).
) 0) at pH* 7.2 (0.03 M phosphate buffer;
Results and Discussion
lesin concentration versus time (eq 1),
ln[carzelesin]t ) ln[carzelesin]0 - kobs
An a lytica l P r oced u r essCarzelesin is poorly soluble in
water (1-2 µg/mL). Furthermore, carzelesin, dissolved in
aqueous solution, tends to adsorb to container materials. This
problem could be solved by performing the degradation
studies in aqueous buffer/acetonitrile mixtures. The optimal
buffer/acetonitrile ratio, in terms of carzelesin solubility,
adsorption prevention, and buffer capacity requirements, was
1:1 (v/v).
UV/vis spectrophotometry measurements showed similar
spectra for carzelesin and its degradation products, thus
indicating that this is not an adequate technique to study the
degradation reactions. A stability-indicating HPLC method
was, therefore, used to resolve the degradation products from
the parent carzelesin. Typical HPLC chromatograms of the
degradation mixtures of carzelesin in acidic and alkaline
media are depicted in Figures 2 and 3, respectively.
Degr a d a tion Kin eticssOrder of ReactionssThe degrada-
tion of carzelesin in acidic as well as alkaline media displays
(pseudo-) first-order kinetics over several half-lives. The
observed (pseudo-) first-order rate constant (kobs) for the
overall degradation rate was calculated by linear regression
analysis of a plot of the natural logarithm of residual carze-
t
(1)
where [carzelesin]t and [carzelesin]0 are the concentration of
carzelesin at time t and the initial concentration, respectively.
Figure 4 shows first-order plots of carzelesin at several pH*
values. Figure 5 shows a typical degradation curve of carze-
lesin and the subsequent appearance curves of the degrada-
tion products U-76,073 and U-76,074. After hydrolysis of the
carbanilate ring moiety, carzelesin was converted into U-76,-
073, which was then, after ring closure, transformed into
U-76,074. The formation of U-76,074 was characterized by a
lag time, which is in accordance with the proposed mechanism
where U-76,073 is formed first before subsequent transforma-
tion into U-76,074 can take place. The mass balance of
carzelesin and its degradation products adds up to ap-
proximately 90%. Since the amounts of U-76,073 and U-76,-
074 formed were calculated relative to the amount of carze-
lesin initially present at t ) 0, the missing 10% can be
explained by a slight decrease in molar absorptivities at the
detection wavelength (254 nm). These observations have also
been confirmed by UV/vis spectrophotometric experiments.
Journal of Pharmaceutical Sciences / 1229
Vol. 85, No. 11, November 1996

+
+
Figure 7
s
UV/vis spectra of carzelesin in 0.01 M buffer solutions at pH* 0, 4.7,
7.9, and 10, recorded at t
) 0.
Figure 6spH*−rate profile of carzelesin at 25 °C.
Table 1 Influence of Acetate, Phosphate, and Carbonate Buffer
−
Concentration and pH on kobs of the Degradation Reactions of Carzelesin
at 25 °C
Standard Deviation in kobssThe standard deviation (SD)
of the overall rate constant, kobs, was determined for the HPLC
assay. These statistical experiments were performed at
pH* 6.4 (0.03 M acetate buffer solution; µ ) 0.3) and 60 °C
and at pH* 8.2 (0.03 M phosphate buffer solution; µ ) 0.3)
and 25 °C, respectively. The value of kobs ((SD) at pH* 6.4
was 1.5 × 10-6 ( 7.3 × 10-8 s-1 (n ) 6), and the relative SD
is equal to 4.8%. The value of kobs ((SD) at pH* 8.2 was 7.3
× 10-5 ( 2.2 × 10-6 s-1 (n ) 4), and the relative SD is equal
to 3.0%.
Influence of pHsFrom previous stability experiments of
carzelesin in the pharmaceutical formulation, it was concluded
that the degradation of carzelesin is strongly influenced by
pH.7 Buffering is especially important for carzelesin because
the ring closure step produces hydrochloric acid in addition
to U-76,074. The overall rate constant (kobs) for the degrada-
tion of carzelesin in buffer solutions is defined as
[Buffer],
M
[Buffer],
M
1
1
pHa
pH*b
kobs, s-
pHa
5.4
pH*b
6.2
kobs, s-
8
9
9
9
9
8
8
8
8
8
8
8
8
8
8
8
8
7
7
7
7
7
7
7
7
6
6
6
6
6
7
7
6
6
6
6
6
6
6
6
6
6
5
5
5
5
5
5
5
4
4
4
4
3
3
3
0.0
1.3
1.7
2.1
2.5
3.2
0.0
1.5
1.8
2.2
2.7
3.2
1.0
0.1
5.0
3.5
3.0
5.8
2.5
2.3
2.2
1.9
1.8
5.1
4.4
3.8
3.7
9.9
8.2
7.2
6.7
2.4
2.1
2.0
1.9
5.6
5.8
5.6
5.5
2.6
2.2
1.6
1.6
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
0.01
0.03
0.05
0.01
0.03
0.05
0.0025
0.005
0.01
0.02
0.03
0.04
0.05
0.01
0.03
0.05
0.005
0.01
0.03
0.05
0.01
0.03
0.05
0.005
0.01
0.03
0.05
1.0
8.1
6.9
2.7
2.4
2.3
9.0
8.1
7.8
7.8
7.6
7.3
7.4
2.5
2.8
2.6
6.1
7.8
8.3
8.3
1.3
1.9
2.2
3.7
1.1
2.7
3.1
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
0.03
0.01
0.003
0.005
0.01
0.03
0.05
0.005
0.01
0.03
0.05
0.005
0.01
0.03
0.05
0.005
0.01
0.03
0.05
0.005
0.01
0.03
0.05
0.005
0.01
0.03
0.05
5.9
6.4
6.7
7.2
3.5
3.9
4.2
4.7
5.2
4.6
5.0
5.4
5.9
6.4
7.0
7.5
7.7
8.2
kobs ) k0 + kH[H+] + kOH[OH-] + kbuffer[buffer] (2)
where k0 is the (pseudo-) first-order rate constant for the
degradation in solvent only and kH and kOH are the second-
order rate constants for proton- and hydroxyl-catalyzed
degradation, respectively. The term kbuffer[buffer] represents
the sum of the second-order rate constants for the degradation
catalyzed by each of the buffer components multiplied by its
concentration. For each pH the (pseudo-) first-order rate
constant for [buffer] ) 0, kobs′ ) k0 + kH[H+] + kOH[OH-], was
calculated as the intercept from the linear part of the plot of
kobs vs [buffer] at a fixed pH.
8.0
8.5
8.7
9.8
a pH measured in the aqueous buffer solution. b pH* is the pH in the buffer/
acetonitrile (1:1, v/v) mixture.
As the degradation of carzelesin at low pH and 25 °C was
relatively slow, the kinetic experiments at low pH were
performed at 60 °C in the pH* range 1.5-5.5 and at 70 °C at
pH* 0. The effect of temperature on the degradation of
carzelesin was studied at pH* 1.5 in the temperature range
60-80 °C and at pH* 7.2 in the temperature range 25-45
°C, respectively. The Arrhenius relationship, ln kobs ) ln A
- Ea/RT, between the natural logarithm of the kobs and the
reciprocal of the absolute temperature holds. At pH* 1.5, ln
higher pH values. Overall the degradation of carzelesin is
catalyzed mainly by hydroxyl ions, which constitutes the
major part of the profile. At pH* > 5, the pH*-rate profile
has a slope of +0.9; this is close to unity and indicates that
specific hydroxyl ion catalysis occurs.
With the use of eq 3,
A is equal to 26.49 (0.098) s-1 and Ea to 111.8 (6.75) kJ mol-1
and at pH* 7.2, ln A equals 40.97 (0.037) s-1 and Ea is equal
to 130.7 (2.08) kJ mol-1
Since the activation energy was
,
kobs′ ) k0 + kH[H+] + kOH[OH-]
(3)
.
determined to be constant over the temperature range, these
parameters were used to extrapolate kobs′ values obtained at
60 °C to kobs′ values at 25 °C for the construction of the pH*-
rate profile at 25 °C (Figure 6). The profile shows distinct
regions with a pH independent region between approximately
1 and 4, with a predominantly hydroxyl catalyzed part at
the specific rate constants for the degradation of carzelesin
were calculated. The values of the specific rate constants kH,
k0, and kOH were 2.41 × 10-8 M-1 s-1, 7.62 × 10-7 s-1, and
3.50 × 101 M-1 s-1, respectively. The pH at which the
minimum occurs in the pH-rate profile was determined
1230 / Journal of Pharmaceutical Sciences
Vol. 85, No. 11, November 1996

+
+
Figure 8sProposed overall degradation scheme of carzelesin in alkaline medium.
graphically to be pH 2-3. The pH*-rate profile does not show
inflections that could be attributed to acid-base equilibria of
the compound. UV/vis spectra of carzelesin recorded in four
buffer/acetonitrile solutions at pH* 0, 4.7, 7.9, and 10.0,
however, showed that the spectrum at pH* 0 was significantly
different from the other three spectra (Figure 7). Carzelesin
Journal of Pharmaceutical Sciences / 1231
Vol. 85, No. 11, November 1996

+
+
probably becomes protonated at this extremely low pH value.
This might also explain the relatively slow degradation rate
in this acidic pH region at pH* < 2. Protonated carzelesin
will be positively charged and, therefore, less susceptible to
proton-catalyzed degradation.
It is clear that the stability of carzelesin is extremely pH
dependent with maximal stability around pH 3 (Figure 6).
Degradation mainly involves hydroxyl-catalyzed reactions
where kOH >> kH. In an earlier study it was found that the
pH values of PEG 400 obtained from different suppliers were
variable and had a major impact on the stability of the
pharmaceutical formulation. It is advisable to adjust the pH
of the PET formulation to around 3 in order to guarantee
maximal drug stability. Shortly before intravenous infusion
the PET formulation is diluted with 5% dextrose infusion
solution, which increases the pH within physiologically ac-
cepted limits.4
Influence of BufferssExperiments were performed where
kobs was measured at varying buffer concentrations, while pH,
ionic strength (µ ) 0.3) and temperature (25 or 60 °C) were
kept constant. Table 1 shows the influence of buffer ions on
the degradation of carzelesin. All four different types of
buffers appear to have a small, however, not significant effect
on the degradation rate of carzelesin.
Influence of Ionic StrengthsThe influence of ionic strength
on the degradation of carzelesin was investigated by adding
various amounts of sodium chloride to buffer solutions of fixed
pH, pH* 1.5 (0.1 M perchloric acid buffer solution) and pH*
7.2 (0.02 M phosphate buffer solution). The influence of ionic
strength on the degradation of carzelesin at pH* 7.2 is
negligible. However, at pH* 1.5 the ionic strength seems to
have a stabilizing effect on the degradation rate of carzelesin.
The higher the ionic strength, the lower the degradation rate
of carzelesin.
Degradation MechanismsFigures 2 and 3 show HPLC
chromatograms of degradation products of carzelesin in acidic
(perchloric acid buffer solution) and alkaline media, respec-
tively. It appears that different degradation products are
formed in the two regions of the pH*-rate profile. Figure 8
shows the proposed degradation mechanism of carzelesin in
alkaline medium. Carzelesin degrades, in the presence of
OH- ions, mainly into U-76,073 and a carbanilate ion. U-76,-
073 is subsequently converted into U-76,074 and hydrochloric
acid. Carbanilate is unstable and decomposes into aniline and
CO2. The peak with a retention time of 3.6 min in Figure 3
is aniline. All three degradation products, U-76,073, U-76,-
074, and aniline, were confirmed by photodiode array detec-
tion. The retention times in the HPLC chromatograms and
UV/vis spectra were compared with those of reference stan-
dard solutions and were found to be identical.
In acidic medium (perchloric acid buffer solution) many
peaks are formed. The intensity, however, is limited, and
none of the peaks coeluted with U-76,073 and U-76,074. The
stability of U-76,073 in a perchloric buffer solution (pH* 2.2;
60 °C) was then investigated, and it was found that the
degradation half-life of U-76,073 was approximately 4 h, while
the half-life of carzelesin was 130 h at this pH (60 °C). In
perchloric acid buffer solutions, carzelesin may be converted
into U-76,073; however, due to the short degradation half-
life of U-76,073, this will not be detected. Several carzelesin
degradation products formed at pH* 2.2 had retention times
identical to that of the U-76,073 degradation products gener-
ated under these conditions.
Figure 9
074 (
50% degradation of carzelesin (
s
Quantification of the degradation products U-76,073 (
9
) and U-76,-
0
) (calculated as percent of the initial carzelesin peak area at t
), versus the pH*.
) 0), at
2
U-76,074. In addition, a peak with a retention time of 4.4
min is seen in the HPLC chromatogram, with a peak area of
up to 40% of the amount of carzelesin present at t ) 0. The
on-line recorded UV/vis spectrum of this product was identical
to that of U-76,073. Experiments to isolate and to elucidate
the chemical structure of this intermediate degradation
product are ongoing.
In order to determine the influence of pH on the amount of
formed degradation products, at each pH* the amount of
formed degradation products was quantitated as a function
of 50% degradation of carzelesin. The results are displayed
in Figure 9. The higher the pH*, the higher the amount of
identified degradation products formed (calculated as the
percentage of carzelesin present at t ) 0). In the pH* range
6-10, the maximum amounts of U-76,073 and U-76,074
formed were constant and equal to 10 and 30%, respectively.
Conclusions
Degradation of carzelesin follows (pseudo-) first-order kinet-
ics. The pH*-rate profile of carzelesin is determined mainly
by specific hydroxyl catalysis. The degradation rate and
mechanism of carzelesin degradation are strongly pH-depend-
ent. Carzelesin is most stable in the pH* region 1-4.
The degradation rate of carzelesin is not significantly
affected by buffer components and by the ionic strength at
pH* 7.2. At pH* 1.5, however, the ionic strength seems to
have a stabilizing effect on the degradation of carzelesin.
In addition to the formation of the degradation products
U-76,073, U-76,074, and aniline in alkaline medium and in
acetate buffer solution, another degradation product (tR 4.4
min) was formed in acetate buffer solution. In perchloric acid
buffer solution (pH* < 3), U-76,073 and U-76,074 were not
detected. The degradation mechanism is not completely
understood yet.
References and Notes
1. Li, L. H.; DeKoning, T. F.; Kelly, R. C.; Krueger, W. C.;
McGovern, J . P.; Padbury, G. E.; Petzold, G. L.; Wallace, T. L.;
Ouding, R. J .; Prairie, M. D.; Gebhard, I. Cancer Res. 1992, 52,
4904-4913.
2. Broggini, M.; Ponti, M.; Ottolenghi, S.; D’Incalci, M.; Mongelli,
N; Mantovani, R. Nucleic Acids Res. 1989, 17, 1051-1059.
3. Hightower, R. D.; Sevin, B. U.; Perras, J .; Nguyen, H.; Angioli,
R.; Untch, M.; Averette, H. Cancer Invest. 1993, 11, 276-282.
The degradation profile of carzelesin in acetate buffer
solutions (3 e pH < 6) is different from that in perchloric acid
solution at identical pH. In acetate buffer solutions, carzelesin
degrades into U-76,073, which is subsequently converted into
1232 / Journal of Pharmaceutical Sciences
Vol. 85, No. 11, November 1996

+
+
4. J onkman-de Vries, J . D.; De Graaff-Teulen, M. J . A.; Henrar,
R. E. C.; Kettenes-van den Bosch, J . J .; Bult, A.; Beijnen, J . H.
Invest. New Drugs 1994, 12, 303-314.
5. Davignon, J . P.; Slack, J . A.; Beijnen, J . H.; Vezin, W. R;
Schoemaker, T. J . Eur. J . Cancer Clin. Oncol. 1988, 24, 1535-
1538.
9. Hurley, L. H.; Warpehoski, M. A.; Lee, C. S.; McGovern, J . P.;
Scahill, T. A.; Kelly, R. C.; Mitchell, M. A.; Wicnienski, N. A.;
Gebhard, I.; J ohnson, P. D.; Bradford, V. S. J . Am. Chem. Soc.
1990, 112, 4633-4649.
6. Awada, A.; Punt, C. J . A.; Gil, Th.; Kwakkelstein, M. O.; Van
Tellingen, O.; Wagener, D. J . Th.; Piccart, M. J . Proc. Am. Assoc.
Cancer Res. 1995, 237.
Acknowledgments
7. J onkman-de Vries, J . D.; Rosing, H.; Henrar, R. E. C.; Bult, A.;
Beijnen, J . H. J . Pharm. Sci. Technol. 1995, 49, 283-288.
8. Kelly, R. C.; Gebhard, I.; Wicnienski, N. A.; Aristoff, P. A.;
J ohnson, P. D.; Martin, D. G. J . Am. Chem. Soc. 1987, 109,
6837-6838.
The authors would like to thank Dr. C. V. Pesheck (Upjohn
Laboratories, The Upjohn Company, Kalamazoo, MI) for reviewing
this manuscript.
J S960005N
Journal of Pharmaceutical Sciences / 1233
Vol. 85, No. 11, November 1996