Quality control and stability study using HPTLC: applications to cyclophosphamide in various pharmaceutical products.

Abstract

Cyclophosphamide is an alkylating agent widely used from cancer chemotherapy to immunotherapy purposes. In paediatrics oncology, oral cyclophosphamide prescribed at low dosages for a long time treatment is currently investigated. This treatment is a putative well tolerated regimen for children treated for a wide variety of recurrent solid tumours. For these purposes, new oral formulations more convenient for children than cyclophosphamide 50mg tablets are needed. Thus, we present a rapid method for the assay of cyclophosphamide in various pharmaceutical preparations using high-performance thin-layer chromatography (HPTLC) and derivatization with phosphomolybdic acid. This method is accurate and precise and allows quantitation of cyclophosphamide in aqueous solutions from 400 to 1200 microg/mL. It is suitable for quantitation and stability studies of cyclophosphamide in pharmaceutical products, i.e. capsules and infusion bags prepared in a hospital pharmacy. According to pharmaceutical guidelines, we demonstrated that low dose cyclophosphamide capsules, extemporaneously prepared for children, are stable at least for 70 days.

Full text

Page 1

Journal of Pharmaceutical and Biomedical Analysis 38 (2005) 180–185
Short communication
Quality control and stability study using HPTLC: applications to
cyclophosphamide in various pharmaceutical products
J´ erˆ ome Bouliganda,b, Thomas Stormea, Isabelle Lavillea, Lionel Merciera,
Odile Oberlinc, Gilles Vassalb,c, Philippe Bourgeta, Angelo Pacia,b,∗
aDepartment of Clinical Pharmacy, Institut Gustave Roussy, 39 rue Camille Desmoulins, 94800 Villejuif, France
bUPRES EA3535 Pharmacology and New Cancer Treatments, Institut Gustave Roussy and Paris XI University,
39 rue Camille Desmoulins, 94800 Villejuif, France
cDepartment of Paediatric Oncology, Institut Gustave Roussy, 39 rue Camille Desmoulins, 94800 Villejuif, France
Received 9 July 2004; received in revised form 30 November 2004; accepted 1 December 2004
Abstract
Cyclophosphamide is an alkylating agent widely used from cancer chemotherapy to immunotherapy purposes. In paediatrics oncology, oral
cyclophosphamide prescribed at low dosages for a long time treatment is currently investigated. This treatment is a putative well tolerated
regimen for children treated for a wide variety of recurrent solid tumours. For these purposes, new oral formulations more convenient for
children than cyclophosphamide 50mg tablets are needed. Thus, we present a rapid method for the assay of cyclophosphamide in various
pharmaceutical preparations using high-performance thin-layer chromatography (HPTLC) and derivatization with phosphomolybdic acid.
This method is accurate and precise and allows quantitation of cyclophosphamide in aqueous solutions from 400 to 1200?g/mL. It is suitable
for quantitation and stability studies of cyclophosphamide in pharmaceutical products, i.e. capsules and infusion bags prepared in a hospital
pharmacy. According to pharmaceutical guidelines, we demonstrated that low dose cyclophosphamide capsules, extemporaneously prepared
for children, are stable at least for 70 days.
© 2004 Elsevier B.V. All rights reserved.
Keywords: Cyclophosphamide; HPTLC; Stability study; Quality control
1. Introduction
Cyclophosphamide (C7H15Cl2N2O2P·H20, CAS 6055-
19-2, CPM) is used in a wide variety of therapeutic indi-
cations from cancer to inflammatory and immune diseases.
This drug is registered to both European [1] and US phar-
macopoeia [2]. CPM is an inactive drug and its activity is
provided through the metabolism by cytochromes P450 [3]
leadingtotheformationofaphosphoramidemustard,theac-
tive compound. CPM is administered at doses ranging from
20mg/m2to 1.2g/m2. CPM is mostly administered by intra-
venous route because this formulation allows avoiding vari-
∗Corresponding author. Tel.: +33 1 42 11 47 30; fax: +33 1 42 11 52 77.
E-mail address: apaci@igr.fr (A. Paci).
ability due to oral route. Oral use is usually preferred when
protracteddailyadministrationisconsidered.Inhematology,
CPM is a component of high dose chemotherapy regimens
to prepare allogeneic hematopoietic stem cell transplanta-
tion [4]. At conventional doses, CPM is a major drug for
the treatment of leukemias, especially in children, and non-
Hodgkin lymphoma. This drug is used at several dosages
for the treatment of a variety of solid malignancies such as
breast cancer [5], in adults as well as in children. Mesna
(UROMITEXAN®) is a drug available to reduce the vesical
toxicity due to the acrolein metabolite of CPM given at high
doses.
An oral formulation of CPM has been developed a long
timeago;whichiscurrentlyindicatedforalimitednumberof
therapeutics applications, such as autoimmunes diseases [6].
0731-7085/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.jpba.2004.12.001

Page 2

J. Bouligand et al. / Journal of Pharmaceutical and Biomedical Analysis 38 (2005) 180–185
181
Oncology is one of the few areas of medicine where most
patients are treated intravenously rather than receiving oral
drugs. Oral chemotherapy is attractive because of its conve-
nience and ease of administration at home [7]. In paediatrics
oncology, oral CPM is currently investigated for its putative
efficiency as a well tolerated regimen against a wide variety
of recurrent solid tumours [8]. The availability of an oral for-
mulation could make possible to use it during a long lasting
low dose maintenance treatment (25mg/m2daily). For this
purpose, a new paediatric formulation more convenient than
CPM tablets set at 50mg is needed for the oral route. When
considering the development of a new formulation it is nec-
essary to perform a validated analytical method. Thus, we
developed a rapid and suitable analytical method for identi-
fying and quantifying CPM in pharmaceutical preparations,
i.e. infusions bags or capsules. Analytical methods to assay
CPM in pharmaceutical products already exist [9,10] and
most of these methods are HPLC ones [11,12] such as the
standard USP method [2]; moreover most available methods
were performed for pharmacokinetics and not for pharma-
ceutical purpose [13]. CPM and metabolites pharmacokinet-
ics were determined in children [14] using a sensitive high-
performance thin-layer chromatography (HPTLC) including
derivatization with 4-nitrobenzylpyridine; since it is a rapid
method,wedevelopedanotherHPTLCmethodtoassayCPM
notinbiologicalsamplesbutinvariouspharmaceuticalprod-
ucts. This new method uses HPTLC before a derivatization
procedure using phosphomolybdic acid. The range of con-
centration,setfrom400to1200?g/mL,isadequateforphar-
maceuticalapplicationswithhighconcentrationsofCPMina
simple medium composed of pharmaceutical adjuvants. This
method was helpful to perform a stability study of CPM cap-
sules at two low dosages, 10 and 25mg. These capsules were
developed for children receiving oral CPM at the daily dose
of 25mg/m2during 28 days. This work is part of the qual-
ity control program for pharmaceutical products prepared in
the Department of Clinical Pharmacy at the Institut Gustave
Roussy [15].
2. Materials and methods
2.1. Materials and chemicals
2.1.1. Chemicals and solvents
Cyclophosphamide
mesna UROMITEXAN®were pharmaceutical products
from Baxter Oncology (Maurepas, France). Phosphomolyb-
dic acid was purchased from Sigma-Aldrich (Lyon, France).
Acetic acid and ethanol were analysis grade from Carlo
Erba (Rodano, Italy). Methanol and dichloromethane were
HPLC grade from Carlo Erba (Rodano, Italy). Carmin, lac-
tose and capsules n◦4 were purchased from Cooper (Melun,
France). Injection grade water, isotonic saline solution and
5% dextrose were purchased from Macopharma (Tourcoing,
France).
monohydrate ENDOXAN®
and
2.1.2. HPTLC CAMAG®analytical station
HPTLC CAMAG®is an high-performance thin-layer
chromatography platform provided by Camag (Muttenz,
Switzerland). The informations relative on the HPTLC an-
alytical platform were published previously [13,15].
2.1.3. Stationary phase
Thestationaryphases(Lichrospher®Si60F254nm),man-
ufacturedbyMerck,weremadeofuniform0.2mmthinsilica
layers that were placed on a glass surface of 20×10cm. The
granulometry phase (2–10?m) is guaranteed by the manu-
facturer.
2.2. Methods
2.2.1. Qualitative and quantitative HPTLC
densitometric assay
2.2.1.1. Samples,standards
(QCs) preparation. Cyclophosphamide stock solutions
(10mg/mL) were prepared using commercially available
cyclophosphamide 500mg ENDOXAN®vials dissolved in
50mL injection water, as specified by the manufacturer. This
stock solution was diluted 1/5 in injectable water to obtain
a 2mg/mL working solution. Cyclophosphamide standards
were made by diluting working solution in methanol/water
(1:1). Standard concentrations were set at 400, 600, 800,
1000, 1200?g/mL. To validate the method, three QCs were
prepared at 500, 700 and 1100?g/mL from another stock
solution, according to the same procedure. For routine
purpose, only the lower and upper QCs are used (QCLand
QCH). All these solutions were transposed in glass snap-ring
clipped vials which were arranged on the autosampler’s
rack.
andqualitycontrols
2.2.1.2. Sample application. Automated TLC sampler III®
devices take into account defined parameters such as the vol-
ume,sizeofthesprayedbandandaccuratepositioningonthe
chromatography plate. These parameters were computerized
by ATS III®software. The system is washed with an iso-
propanol/methanol/water mixture (33:33:33, v/v/v) between
eachdeposition.Twomicrolitersofsolutionaresprayedonto
the plate to form 2mm width and 4mm long bands, 6mm
apart. Bands were sprayed at 15mm from the left edge and
the delivery speed was 200nL/s. During the same run, it is
possible to assay five standards, two QCs and 24 samples
placed within the QCs, i.e. low and high QCs.
2.2.1.3. Mobile phase and migration. Chromatographic de-
velopment was standardized with a 5min horizontal sand-
wich migration method, using a dichloromethane/methanol/
acetic acid mixture (97:3:2, v/v/v).
2.2.1.4. Derivatization procedure. After chromatographic
development, the HPTLC plate was dried with a hair dryer.
Derivatization was performed by automated soaking of the
plate in a phosphomolybdic acid ethanolic solution (1.25%,

Page 3

182
J. Bouligand et al. / Journal of Pharmaceutical and Biomedical Analysis 38 (2005) 180–185
w/v). After drying, the HPTLC plate was heated at a tem-
perature of 190◦C for 10min on a TLC plate heater. Previ-
ously published HPTLC method used 4-nitrobenzylpyridine
for derivatization of busulfan and a triethylamine/ethanol so-
lution for revelation [16]. Cyclophosphamide was used as an
internal standard because of the unstable nature of the chro-
mophore as shown previously by Tasso et al. [14]. A new
method for cyclophosphamide derivatization was developed
using a one step procedure without the need of an internal
standard.
2.2.1.5. Densitometric analysis. The plate was analysed on
a TLC scanner 3®densitometer driven by the CATS 4®
software (4.05 version) as previously reported [13,15,16].
The densitometric measurement was performed at the max-
imum of absorbance of the derivatized product wavelength,
i.e. 700nm, as shown in Fig. 1.
2.2.1.6. Identification and quantitation procedure. In TLC,
each compound can be characterized with its retention fac-
tor (Rf) which is the ratio between substance migration
and solvent migration distances. With this parameter, cy-
clophosphamide can be qualitatively assayed in pharmaceu-
tical preparations. The cyclophosphamide peak surface was
automatically measured for each sample, QCs or standards.
QCsandsampleswerecalculatedaccordingtotheirresponse
and the calibration curve equation.
2.2.2. Validation of the method
2.2.2.1. Selectivity. Cyclophosphamide was co-eluted with
five products which were included in its pharmaceuticals
forms, i.e. carmin, lactose, gelatin, dextrose and mesna.
2.2.2.2. Calibration. The calibration function, i.e. relation-
ship between cyclophosphamide peak areas and its applied
amount was determined by linear regression over a defined
range from 400 to 1200?g/mL. For routine use, each cal-
ibration curve is validated using two QCs prepared at 500
and 1100?g/mL. Samples are placed on the rack within
the QCs. Six calibration curves (tested with three QCs, i.e.
500, 700 and 1100?g/mL) were done to ensure that the lin-
ear regression model was the most accurate for quantitation
purposes.
2.2.2.3. Accuracy. Accuracyprovidesinformationaboutthe
recovery of the analyte from the sample through the analy-
sis of in-system calibration of sample solutions of known
substance content. The solutions were spiked with three dif-
ferent known concentrations of cyclophosphamide, i.e. 500,
700 and 1100?g/mL, respectively. These low, medium and
high QCs were analysed individually six times. Means and
relative standard deviations (R.S.D.) were calculated.
2.2.2.4. Precision. InaccordancewithInternationalConfer-
ence of Harmonisation guidelines (ICH Q2A and Q2B), pre-
cisionincludesthreecomponents:repeatability,intermediate
precision and reproducibility. Here, reproducibility was not
studied.Wethusanalysedrepeatabilityandintermediatepre-
cision as we previously reported for other HPTLC analytical
methods [13,16].
3. Results and discussion
3.1. Validation of the method
3.1.1. Selectivity, specificity
Cyclophosphamide is separated from the other chemical
compounds detected by the process as shown in Fig. 2. The
CPM retention factors was 0.55. For the infusion bag assay,
Fig. 1. Visible spectrum of derivatized cyclophosphamide. This figure shows a spectrum of cyclophosphamide after derivatization with phosphomolybdic
acid. Absorbances (in % of the maximum value) are plotted on Y-axis while wavelength in nanometer from 400 to 800nm are plotted on X-axis. Absorbance
wavelength was set at 700nm for quantitative purpose.

Page 4

J. Bouligand et al. / Journal of Pharmaceutical and Biomedical Analysis 38 (2005) 180–185
183
Fig. 2. Specificity – HPTLC analysis of CPM with various chemical adjuvants. CPM was mixed with lactose, carmin, gelatin, sodium chloride, dextrose
and mesna. These mixtures were thus analysed by HPTLC. This figure shows the superposition of 26 chromatograms with adjuvants alone or mixed with
cyclophosphamide after chromatographic development, derivatization procedure and densitometric analysis (700nm). Intensity (arbitrary unit) are plotted on
Y-axis and Rfare plotted on X-axis.
weverifiedthattherewasnoanalyticalinterferencewithdex-
trose, sodium chloride and mesna. For the capsules, the lack
ofinterferencewithlactose,gelatinandcarminwaschecked.
All adjuvants tested presented Rfvery close to 0 and thus
there were no analytical interference with CPM (Fig. 2).
3.1.2. Calibration
The calibration function was determined by linear re-
gression over 400–1200?g/mL, with a r2=0.9921 (six
determinations). The slope of the mean equation was
β=0.1547±0.0080 (mean±S.D.) and the Y-intercept was
α=29.62±6.76(mean±S.D.).Thus,themeanequationac-
cording to linear regression was Y=0.155X+29.62; Y is the
response (arbitrary unit) and X the concentration (?g/mL).
3.1.3. Accuracy
TheresultsaresummarizedinTable1.Meansandrelative
standard deviation (R.S.D.) values were calculated from the
six determinations of each QC (500, 700 and 1100?g/mL).
The method was considered accurate according to the values
which were below 2.2%.
3.1.4. Precision
The CV values for repeatability (CVr) and for intermedi-
ate precision (CVi) are summarized in Table 2. For quality
control of pharmaceutical products, the method can be con-
sidered precise enough since these values were below 6.3%.
Table 1
Results of the accuracy study
Mean R.S.D. Bias (%)
QCL500?g/mL
QCM700?g/mL
QCH1100?g/mL
Values calculated were based on six different measurements.
510.6
698.9
1101.0
28.9
16.7
45.5
2.13
−0.15
0.1
3.2. Application
3.2.1. Development of cyclophosphamide capsules
In our institution, a clinical trial with CPM 25mg/m2/day
during 28 days is ongoing. Since previously reported sta-
bility studies of CPM solutions [11,12] had shown that the
drug in water solution was stable at 4◦C no more than 14
days, we decided to develop dry forms of CPM. For each
children,oralcapsulessetat10or25mgwillbepreparedex-
temporaneously. These oral formulations are prepared with
cyclophosphamide monohydrate (ENDOXAN®500mg),
lactose, carmin and gelatin capsules. To prepare capsules at
the correct dosage given in mg of cyclophosphamide base,
we have to take into account a conversion factor, i.e. 500mg
of cyclophosphamide base is corresponding to 534.5mg of
monohydrate cyclophosphamide (weighted powder).
3.2.1.1. Quality control – cyclophosphamide capsules.
Mass uniformity of 20 randomized capsules in all batches
was verified as specified by the European Pharmacopoeia
(4th edition) [17]. To validate the manufacturing procedure,
wecheckedtheamountofCPMincapsulespreparedasspec-
ified by the European Pharmacopoeia (4th edition) [18]. The
resultsofsuchatestoncapsulessetat10mgarepresentedin
Table 3. The calculated amounts were easily included in the
acceptance criteria defined by the European Pharmacopoeia,
i.e. no more than one capsule outside the 85–115% limit but
within the 75–125% limit of the target value. To assay CPM
Table 2
Validation of the repeatability and intermediate precision study
CVrrepeatability (%)CViintermediate
precision (%)
QCL500?g/mL
QCM700?g/mL
QCH1100?g/mL
Values calculated were based on six different measurements.
5.7
2.4
4.1
5.6
4.4
6.3

Page 5

184
J. Bouligand et al. / Journal of Pharmaceutical and Biomedical Analysis 38 (2005) 180–185
Table 3
Dose in 10 CPM capsules with 10mg as the theoretical value
Capsules CPM (mg)Amount ?g/mL in waterDilution factorResults ?g/mL Bias %
1
2
3
4
5
6
7
8
9
10
10
10
10
10
10
10
10
10
10
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.651
0.670
0.683
0.706
0.715
0.684
0.689
0.692
0.712
0.681
−7
−4
−2
1
2
−2
−2
−1
2
10
The powder content in each capsule is poured into 10mL of a methanol/water mixture (1:1, v/v). This CPM solution is thus diluted in a hydroalcoholic solution.
−3
content, capsules were opened and the capsule and its pow-
dered content, i.e. monohydrate cyclophosphamide, lactose
and carmin, were poured into 10mL of methanol/water (1:1)
mixture. This hydroalcoholic solution was correctly diluted
to obtain a concentration between 400 and 1200?g/mL. The
sample was then assayed by HPTLC.
3.2.1.2. Stability study – cyclophosphamide capsules. Be-
cause the formulation was performed for patients who will
receivethetreatmentduring28days,wedecidedtoperforma
stabilitystudybyassayingtheexactdoseofCPMincapsules
set at 10 and 25mg during a 70 days period. Two batches of
50 capsules were manufactured for each dose. The dose was
checked in six capsules of each batch at different times point
alongthestudy(day0andonceaweekupto70days).Results
are presented in Fig. 3. During the study period, there was
no evidence of degradation of CPM in the preparation. All
capsules tested were fulfilling acceptance criteria defined by
European Pharmacopoeia (4th edition) [17,18]. Thus, we are
able to warrant the stability of these extemporaneous prepa-
rations during the treatment duration (28 days).
Fig. 3. Stability of CPM capsules set at 10 and 25mg. Two batches of 50
capsules set at 10mg (EDX 10-1 and EDX 10-2) and two batches of 50
capsules set at 25mg (EDX 25-1 and EDX 25-2) were prepared. The exact
amounts of CPM were measured at different time-points along a 70 days
period. On this figure, calculated amounts of CPM are plotted on the Y-axis
and days are plotted on the X-axis. Dash lines represent limit of acceptance
for 10 and 25mg capsules (±15%). All tested capsules were conforms and
there was no trend of CPM degradation.
3.2.2. Quality control – infusion bags
Wedemonstratedthattherewasnoanalyticalinterference
withsodiumchloride,dextroseandmesna.Mesnaisaddedto
some infusion bags according to the manufacturer. Infusion
bagscontaincyclophosphamidefrom1to15g/Landthusour
method is adequate to check the dose of cyclophosphamide
in infusion bags after an appropriate diluting step. Since July
2001, the HPTLC platform has strongly contributed to oper-
ators training and involvement in the product process. It has
resulted in reduction in the non-conformity rate. This quality
managementtoolisapartofISO9001certificationprocedure
[15].
Acknowledgements
We are grateful to Baxter Oncology for supporting this
work.
References
[1] European Pharmacopoeia (4th edition Fr). Cyclophosphamide
01/2002:0711, 2002, pp. 1080.
[2] USP27-NF22, Cyclophosphamide (2003) 527–528.
[3] A.V. Boddy, S.M. Yule, Clin. Pharmacokinet. 38 (2000) 291–304.
[4] G.W. Santos, Semin. Oncol. 20 (1993) 12–16.
[5] C. Sweeney, C.B. Ambrosone, L. Joseph, A. Stone, L.F. Hutchins,
F.F. Kadlubar, B.F. Coles, Int. J. Cancer 103 (2003) 810–
814.
[6] H.L. Weiner, J.A. Cohen, Mult. Scler. 8 (2002) 142–154.
[7] V.J. O’Neill, C.J. Twelves, Br. J. Cancer 87 (2002) 933–
937.
[8] D.C. Bowers, V.M. Aquino, P.J. Leavey, R.O. Bash, J.M. Journey-
cake, G. Tomlinson, A.F. Mulne, H.J. Haynes, N.J. Winick, Pediatr.
Blood Cancer 42 (2004) 93–98.
[9] Z.H. Mohamed, S.M. Amer, A.M. el Kousasy, J. Pharm. Biomed.
Anal. 12 (1994) 1131–1136.
[10] A. Salgueiro, M.A. Egea, R. Valls, M. Espina, M.L. Garcia, J. Pharm.
Biomed. Anal. 21 (1999) 611–618.
[11] A. Mittner, Z. Vincze, K. Jemnitz, Pharmazie 54 (1999) 224–225.
[12] C. Menard, C. Bourguignon, J. Schlatter, N. Vermerie, Ann. Phar-
macother. 37 (2003) 1789–1792.
[13] A. Paci, L. Mercier, P. Bourget, J. Pharm. Biomed. Anal. 30 (2003)
1603–1610.

Page 6

J. Bouligand et al. / Journal of Pharmaceutical and Biomedical Analysis 38 (2005) 180–185
185
[14] M.J. Tasso, A.V. Boddy, L. Price, R.A. Wyllie, A.D. Pear-
son, J.R. Idle, Cancer Chemother. Pharmacol. 30 (1992) 207–
211.
[15] P. Bourget, A. Paci, J.B. Rey, L. Mercier, S. Demirdjian, Eur. J.
Pharm. Biopharm. 56 (2003) 445–451.
[16] J. Bouligand, A. Paci, L. Mercier, G. Vassal, P. Bourget, J. Pharm.
Biomed. Anal. 34 (2004) 525–530.
[17] European Pharmacopoeia (4th edition Fr) (2.9.5.), 2002, pp. 215.
[18] European Pharmacopoeia (4th edition Fr). (2.9.6.), 2002, pp. 215-
216.