Journal of Chromatographic Science Advance Access published August 8, 2012

Journal of Chromatographic Science 2012;0:1– 7

doi:10.1093/chromsci/bms127 Article

HPLC and Chemometric Methods for the Simultaneous Determination of Miconazole

Nitrate and Nystatin
Hala M. Heneedak1, Ismail Salama2, Samia Mostafa2* and Mohamed El-Sadek3
1
Chemical Laboratory legitimate, Department of Forensic Medicine, Ismailia, Egypt, 2Department of Pharmaceutical Chemistry,
Faculty of Pharmacy, Suez Canal University, Ismailia, Egypt and, and 3Department of Medicinal Chemistry, Faculty of Pharmacy,
Zagazig University, Zagazig, Egypt

*Author to whom correspondence should be addressed. Email: [email protected]

Received 6 April 2012; revised 30 May 2012

High-performance liquid chromatography (HPLC) and chemometric determination of MIC and NYS in their synthetic mixtures and
methods were applied to the simultaneous determination of the two vaginal suppositories. The studied chemometric spectrophoto-
nonsteroidal antifungal drugs, miconazole (MIC) and nystatin (NYS). metric methods are multivariate methods including classical
The applied chemometric techniques are multivariate methods in- least squares (CLS), principal component regression (PCR) and
cluding classical least squares, principal component regression and partial least squares (PLS).
partial least squares methods. The ultraviolet (UV) absorption spectra The primary advantages of these techniques are the higher
of the standard solutions of the training and validation sets in metha- speed of processing data concerning the values of concentra-
nol are recorded in the range of 280–320 nm at 0.2-nm intervals. The tion and absorbance of compounds with strongly overlapping
HPLC method depends on reversed-phase separation using a C18 spectra. Additionally, the errors of calibration models are mini-
column. The mobile phase consists of a mixture of methanol–aceto- mized by measuring the absorbance values at many points in
nitrile–ammonium acetate buffer (pH 6; 50 mM) (60:30:10 v/v/v). the wavelength range of the zero order spectra.
The UV detector was set at 230 nm. The developed methods were CLS, sometimes known as K-matrix calibration, is so called
validated and successfully applied to the simultaneous determination because it originally involved the application of multiple
of MIC and NYS in their tablets. The assay results obtained using the linear regression (MLR) to the classical expression of the
chemometric methods were statistically compared to those of the Beer-Lambert law of spectroscopy. PCR combines the principal
HPLC method and good agreement was observed. component analysis (PCA) spectral decomposition with an
inverse least square regression (ILS) to create a quantitative
model for complex samples. The eigenvectors resulting from
data decomposition represent the spectral variations that are
common to all of the spectroscopic calibration data. PLS is
Introduction closely related to PCR, but in PLS the concentration data
Miconazole nitrate [l-(2,4-dichloro-b-((2,4-chlorobenzyl)oxy) matrix is also used in the spectral decomposition. Both PCR
phenethyl)imidazole] is a synthetic imidazole derivative, and PLC produce robust models. They remove noise from the
applied widely as the nitrate salt (MIC) with broad-spectrum absorbance and concentration data.
antifungal activity (1, 2). It has been established as a useful
drug for the treatment of various systemic mycoses. It is also
active against Gram-positive bacteria. Nystatin (NYS) is a polyene
antifungal antibiotic that is of particular interest because it exhi- Experimental
bits remarkable action against a wide range of pathogenic and Instrumentation
non-pathogenic yeast and fungi. It is also used for the prophy- A double-beam Shimadzu (Kyoto, Japan) ultraviolet visible
laxis and treatment of candidiasis of the skin and mucous mem- (UV –Vis) spectrophotometer, model UV-1601 PC equipped
branes. For the treatment of oral candidiasis, NYS is with 1-cm quartz cells and connected to an IBM-compatible
administrated in either a suspension or suppositories (100,000 computer. The bundled software was UVPC personal spectros-
IU ¼ 20.5 mg) (3, 4). MIC is formulated with NYS as vaginal sup- copy software version 3.7 (Shimadzu). The spectral band-
positories for the treatment of vaginal or vulvovaginal candidiasis width was 2 nm and the wavelength scanning speed was
(moniliasis), vaginitis or vulvovaginitis caused by other sensitive 2,800 nm/min. CLS, PCR and PLS analyses were conducted
fungi or Gram-positive bacteria and vaginal mycosis secondarily using the Chemometrics toolbox 2.1 software (15) for use with
infected by Gram-positive bacteria. Several analytical methods MATLAB 7.10.
have been reported in the literature describing the determin- Liquid chromatography was performed on a Knauer (HP)
ation of miconazole nitrate (5–9) or nystatin (10–14) alone or in liquid chromatographic system (Berlin, Germany) equipped
combination with other drugs. Neither high-performance liquid with an A54103 smart line pump 100, an HP variable UV de-
chromatography (HPLC) nor chemometric spectrophotometric tector 2500, and an A1357 manual injection valve with 20 mL
methods have yet been reported for the simultaneous determin- sample loop. Eurochrom for Windows Basic Edition V3.05 was
ation of MIC and NYS in their pharmaceutical formulations. employed for data collecting and processing. A Phenomenex
In this study, chemometric spectrophotometric and Luna (Germany) C18 column (250  4.6 mm i.d., 5 mm ps) was
HPLC methods have been developed for the simultaneous used for the separation. The detector was set at l ¼ 230 nm.

# The Author [2012]. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
Materials and reagents volume was completed to 100 mL with methanol. Further dilu-
MIC and NYS were supplied by Medical Union Pharmaceuticals tions of the filtrate were conducted with mobile phase (for
(MUP) (Ismailia City, Egypt) and certified to contain 99.8 and HPLC method) or methanol (for spectrophotometric methods)
99.9%, respectively. to reach the calibration range.
Commercial Monicure Plus vaginal suppositories (batch no.
7361056) (Pharaonia Pharmaceuticals, Alexandria City, Egypt)
were used. Each vaginal suppository was labeled to contain Results and Discussion
100 mg MIC and 100.000 IU ¼ 20.5 mg NYS.
Figure 1 shows the zero-order absorption spectra of MIC and
Acetonitrile and methanol were HPLC grade (Tedia, Fairfield,
NYS in methanol. It is clear that the spectra of the two drugs
OH). Ammonium acetate and glacial acetic acid were analytical
display considerable overlap throughout the wavelength range.
grade.
Due to the presence of this spectral interference, HPLC and
chemometric methods were necessary for the simultaneous
HPLC conditions determination of the two drugs in vaginal suppositories.
The mobile phase was prepared by mixing methanol, aceto-
nitrile and 50 mM ammonium acetate buffer (apparent pH was
adjusted to 6 using glacial acetic acid) in a ratio of 60:30:10 HPLC method
(v/v/v). The flow rate was 1 mL/min. The injection volume The primary target in developing this LC method is to achieve
was 20 mL. Quantitation based on peak area was achieved using the simultaneous determination of MIC and NYS. The mobile
UV detection at 230 nm. All determinations were performed at phase composition and pH of 50 mM ammonium acetate were
ambient temperature. studied and optimized. A successful separation was obtained
with a mobile phase consisting of methanol, acetonitrile and
50 mM ammonium acetate buffer ( pH 6) in a ratio of 60:30:10
Standard solutions and calibrations v/v at a flow rate of 1 mL/min.
Stock standard solutions (1 mg/mL) of each drug were prepared Increasing the acetonitrile concentration to more than 40%
separately in methanol for spectrophotometric methods and for led to inadequate separation of MIC and NYS. At lower aceto-
HPLC. Suitable dilutions were made using the specified solvent. nitrile concentration (,30%), separation occurred, but with
excessive tailing for the MIC peak. Variation of the apparent
CLS, PCR and PLS methods pH of the 50 mM ammonium acetate buffer resulted in a low
A training set of 15 synthetic binary mixture solutions were capacity factor (K) value for MIC at apparent pH 3.5, with loss
prepared by further dilution of the stock solutions with metha- of peak symmetry for NYS. At apparent pH 5, improved resolu-
nol in the range of 10 –90 mg/mL (MIC) and 1.8 –18 mg/mL tions were observed for the two drugs. However, at apparent
(NYS). The UV absorption spectra were recorded over the pH 6, optimum resolution with reasonable retention time was
range 280 –320 nm. The data points of the spectra were col- observed.
lected at every 0.2 nm. A validation set containing 12 synthetic The specificity of the HPLC method is illustrated in Figure 2,
binary mixtures in the ranges of 20 –80 and 3.8 –16.2 mg/mL1 in which complete separation of the two drugs was observed.
for MIC and NYS, respectively, was prepared using the preced- The average retention time + standard deviation (SD) for NYS
ing stock solutions. and MIC were found to be 3.35 + 0.04 and 6.67 + 0.05 min, re-
PLS and PCR models were applied to the UV absorption spectively, for 10 replicates.
spectra of these mixtures using three latent variables for PLS
and three principal components for PCR for the determination
of each compound.

HPLC method
The standard solutions were prepared by further dilutions of
the stock standard solutions with mobile phase to reach the
concentration ranges of 4 –20 mg/mL for MIC and 10 –100 mg/
mL for NYS. Triplicate 20-mL injections were made for each
concentration and chromatographed under the specified condi-
tions. The peak area values versus corresponding concentra-
tions were plotted. Linear relationships were obtained.

Sample preparation
Five Monicure Plus vaginal suppositories were accurately
weighed and finely powdered in a mortar. An amount of the
suppository mass equivalent to one suppository content
(100 mg of MIC and 20.5 mg of NYS) was dissolved in 60 mL of
methanol. After 30 min of warming and mechanical shaking, Figure 1. UV zero-order absorption spectra of 100 mg/mL miconazole (dashed line),
the solution was filtered in a 100-mL volumetric flask. The 20 mg/mL nystatin (dotted line) and a mixture of 100 mg/mL miconazole and
residue was washed twice, each with 10 mL of the solvent. The 20 mg/mL nystatin (straight line) in methanol.

2 Heneedak et al.
 Table I
Concentrations of Different Mixtures of MIC and NYS used in the Training Set

Mixture MIC (mg/mL) NYS (mg/mL)

1 10 1.8
2 10 2
3 10 2.2
4 30 5.8
5 30 6
6 30 6.2
7 50 9.8
8 50 10
9 50 10.2
10 70 13.8
11 70 14
12 70 14.2
13 90 17.8
14 90 18
15 90 18.2

accuracy of predictions, and was recalculated upon addition of

Figure 2. HPLC chromatogram of a 20-mL injection of a synthetic mixture containing
10 mg/mL nystatin ( peak 1) and 4 mg/mL miconazole ( peak 2) in methanol.
each new factor to the PLS and PCR models:

RMSECV ¼ ðPRESS=nÞ1=2 ð1Þ

CLS, PCR and PLS methods where PRESS is the predicted residual error sum of squares and
n is the number of calibration samples (19):
Construction of CLS, PCR and PLS models
Xn
The composition of the training set was orthogonally designed PRESS ¼ ðCiPred  Ci
Experimental
Þ2 ð2Þ
to obtain maximum information on each drug from the cali- i¼1

bration procedure. The number of calibration mixtures in the

training set was selected according to the rule of five. This where CPred
i and CExperimental
i are predicted and true concentra-
rule states that using five times the number of samples as tions in mg/mL, respectively. Visual inspection was used for
components provides enough samples to reasonably represent selecting the optimum number of factors. Three factors were
all possible combinations of different concentration values found to be suitable for both PCR and PLS methods, as shown
(16). A training set was prepared, as shown in Table I. The ab- in Figures 3 and 4.
sorbance data matrix for this training set was obtained by An independent set of validation synthetic mixtures contain-
recording the absorbances within the wavelength range of ing MIC and NYS in the different compositions given in
280 –320 nm at 0.2-nm intervals. A CLS model was constructed Table II was prepared and analyzed for validation. The mean
with non-zero intercept. To build this model, the computer percentage recoveries, SD and relative standard deviations
was fed with the absorbance and concentration matrices for (RSD) are indicated in Table II.
the training set. The calculations to obtain the K matrix were The standard error of prediction (SEP), mean squared error
conducted. For the PCR and PLS models, the training set ab- of prediction (MSEP), RMSECV, variance of prediction (S2) and
sorbance and concentration matrices, together with relative error of prediction (REP) are also used for validation
PLS-toolbox 2.1 software, were used for calculations. The (16, 20). The accuracy and precision of prediction are defined
cross validation method was employed to eliminate only one by MSEP and RMSECV (20):
sample at a time, and then the remaining standard spectra
were calibrated by PCR and PLS (17, 18). With the utilization SEP ¼ ½PRESS=n  11=2 ð3Þ
of this calibration, the concentration of the remaining sample
was predicted. This process was repeated until each standard MSEP ¼ PRESS=n ð4Þ
had been left once. Xn Experimental
Bias ¼ i¼1
ðCiPred  Ci Þ=n ð5Þ
Xn Experimental
Selection of the optimum number of factors S2 ¼ i¼1
ðCi  CiPred  biasÞ2 =n  1 ð6Þ
The optimum number of factors (latent variables) to be
included in the calibration model was determined by comput- ^ Experimental Þ
REPð%Þ ¼ RMSECV  ð100=C ð7Þ
ing the prediction error sum of squares (PRESS) for the cross-
validated models, using a high number of factors (half the where C Experimental
i is the true concentration, CPred
i is the pre-
number of total standard þ 1). The predicted concentrations of dicted concentration, ĈExperimental is the average concentration
the two drugs in each sample were compared with the actual in the validation set and n is the total number of validation
concentrations in these calibration samples and root-mean- samples. The numerical values of SEP, MSEP, RMSECV and S2
square error of cross-validation (RMSECV) was calculated for are listed in Table III. Small values of the results indicate the
each method. RMSECV indicates both of the precision and negligible error of prediction and high ability of prediction.

HPLC and Chemometric Methods for the Simultaneous Determination of Miconazole Nitrate and Nystatin 3
Figure 3. RMSECV plot of a calibration set prediction using cross validation (PCR model).

Figure 4. RMSECV plot of a calibration set prediction using cross validation (PLS model).

Table II Table III

Assay Results of M_C and NYS Combinations in Synthetic Mixtures (Validation Mixtures) by the Statistical Parameters of the Validation of Synthetic Mixtures using the Proposed Chemometric
Proposed Chemometric Methods Methods

Validation Recovery (%) Parameters MIC NYS

mixtures
CLS PCR PLS CLS PCR PLS
Added (mg/mL) CLS PCR PLS
Intercept –0.99 –1.08 – 1.08 – 0.09 –0.28 –0.28
MIC NYS MIC NYS MIC NYS MIC NYS Slope 1.04 1.03 1.03 1.02 1.03 1.03
r 0.998 0.998 0.998 0.999 0.999 0.999
20 3.8 96.72 96.31 99.45 102.85 99.45 102.85 SE of intercept 0.81 0.72 0.72 0.13 0.08 0.08
20 4 99.74 101.86 101.44 103.01 101.44 103.01 SE of slope 0.01 0.01 0.01 0.01 0.01 0.01
20 4.2 101.6 99.06 99.99 99.54 99.99 99.54 Lower CL of intercept* –2.79 –2.69 – 2.69 – 0.38 –0.46 –0.46
40 7.8 97.66 98.73 102.11 101.79 102.11 101.79 Upper CL of intercept* 0.82 0.53 0.53 0.18 –0.1 –0.1
40 8 100.38 101.54 103.36 102.33 103.36 102.33 Lower CL of slope* 1 1 1 0.99 1.01 1.01
40 8.2 104.29 103.95 97.41 101.10 97.41 101.10 Upper CL of slope* 1.07 1.06 1.06 1.04 1.05 1.05
60 11.8 95.95 99.49 99.83 99.37 99.83 99.37 LOD (mg/mL) 3.58 3.21 3.21 0.57 0.35 0.35
60 12 95.49 96.91 100.93 100.17 100.93 100.17 LOQ (mg/mL) 10.85 9.74 9.74 1.73 1.07 1.07
60 12.2 97.51 97.97 96.33 99.66 96.33 99.66 SEP 1.77 1.19 1.19 0.2 0.18 0.18
80 15.8 96.47 99.39 98.57 97.77 98.57 97.77 MSEP 2.87 1.31 1.31 0.04 0.03 0.03
80 16 96.97 98.95 98.57 98.89 98.57 98.89 RMSECV 1.69 1.15 1.15 0.19 0.17 0.17
80 16.2 98.88 98.45 99.22 98.12 99.22 98.12 S2 2 1.32 1.32 0.035 0.031 0.031
Mean 98.47 99.38 99.76 100.38 99.76 100.38 REP (%) 3.39 2.29 2.29 1.95 1.69 1.69
+ SD 2.61 2.14 1.98 1.80 1.98 1.80
RSD 2.65 2.16 1.98 1.79 1.98 1.79
*Confidence limit; calculated at 95% confidence limit.

Plotting of the predicted versus true concentrations is also Validation of the methods
used for validation. A straight line is expected to be obtained
(20). The regression analysis for these linear relationships was Linearity
conducted and the results are shown in Table III. The absence The linearity of the HPLC method for determination of MIC
of bias was proved by determining the confidence limits for and NYS was evaluated by analyzing a series of different con-
the intercept and the slope at the 95% significance level (21). centrations of each drug. In this study, six concentrations were
Limit of quantification (LOQ) and limit of detection (LOD) are chosen, ranging between 4 –20 mg/ml for MIC and from
also given in Table III. 10– 100 mg/mL for NYS. Each concentration was repeated

4 Heneedak et al.
three times, to provide information on the variation in peak satisfactory precision and accuracy were observed for these
area values between samples of the same concentration. The methods. Consequently, the excipients in pharmaceutical for-
linearity of the calibration graphs was validated by the high mulation do not interfere in the analysis of these compounds
value of the correlation coefficient and the intercept value, in the pharmaceutical formulation.
which was not statistically ( p ¼ 0.05) different from zero
(Table IV). Characteristic parameters for regression equations System suitability
of the HPLC method obtained by least-squares treatment of the Resolution (Rs) is a measure of the degree of separation
results are given in Table IV. between adjacent peaks. A value of 1.5 for resolution implies a
LOQ and LODThe LOD and LOQ are given in Table IV. The complete separation of the two compounds (23). Additionally,
SD of the response and the slope was used for calculating the British Pharmacopoeia specifies that the symmetry factor of a
detection and quantitation limits (22) as follows: LOD ¼ 3SD/ principal peak must be between 0.8 and 1.5 (23). Resolutions
Slope and LOQ ¼ 10SD/Slope. and other system suitability parameters were calculated for MIC
and NYS. Their values were found to be acceptable (Table VII).
Precision
Ruggedness and robustness tests
The intra-day and inter-day variations of the method were
determined using three replicate injections of three different As recommended in the ICH guidelines and the Dutch
concentrations, which were prepared and analyzed on the Pharmacists guidelines, a robustness assessment was performed
same day and on three different days over a period of two during the development of the analytical procedure (24). The
weeks, respectively (Table V). These data indicate a consider- ruggedness (25) of the method was assessed by comparison of
able degree of precision and reproducibility for the method the intra-day and inter-day assay results for MIC and NYS that
both during one analytical run and between different runs. were performed by two analysts. The RSD values for intra-day
and inter-day assays of MIC and NYS in the Monicure Plus
vaginal suppositories performed in the same laboratory by two
Accuracy analysts did not exceed 3.8%, indicating the ruggedness of the
The interference of excipients in the pharmaceutical formula- method. In addition, the robustness of the method was investi-
tions was studied in detail by the CLS, PCR, PLS and HPLC gated under a variety of conditions, including changes of the
methods. For this reason, the standard addition method was flow rate, PH and mobile phase composition (26).
applied to the pharmaceutical formulation containing these
compounds. In application of standard addition method to the
pharmaceutical formulation, the mean percentage recoveries Table VI
and their standard deviations for the proposed methods were Determination of MIC and NYS in Monicure Plus Vaginal Suppositories using the Proposed
calculated (Table VI). According to the obtained results, Chemometric and HPLC Methods*

CLS PCR PLS HPLC

MIC
Table IV Mean recovery 100.14 + 0.55 100.12 + 0.55 100.12 + 0.55 99.98 + 0.34
Calibration Curve Data for MIC and NYS using the HPLC Method (%) + SD
t 1.32 1.32 1.32 (2.31)†
Regression parameters MIC NYS f 2.88 2.88 2.88 (6.39)†
Regression coefficient (r)* 0.9997 0.9993 NYS
Calibration range (mg/mL) 4 –20 10 –100 Mean recovery 100.45 + 0.78 100.39 + 0.78 100.39 + 0.78 100.12 + 0.23
LOD (mg/mL) 0.36 3.4 (%) + SD
LOQ (mg/ml) 1.1 10 t 0.88 0.88 0.88 (2.31)†
Slope + SD 3.47 + 0.06 0.5 + 0.01 f 1.22 1.22 1.22 (6.39)†
CL of the slope† –0.45 –1.71 0.49– 0.53 Standard addition technique‡
Intercept + SD 0.62 + 0.95 0.96 + 1.01
CL of the intercept† 3.4 –3.55 –0.36– 2.11 Mean recovery 100.21 + 0.66 100.21 + 0.66 100.21 + 0.66 100.15 + 0.6
Number of points (n) 6 6 (%) + SD for MIC
Mean recovery 100.13 + 0.57 100.13 + 0.57 100.13 + 0.57 100.17 + 0.51
*Degrees of freedom: 5. (%) + SD for NYS
†
Calculated at 95% confidence limit.
*Note: Monicure Plus vaginal suppositories labeled to contain 100 mg MIC and 20.6 mg NYS
per suppository.
†
Theoretical values for t and F at P ¼ 0.05.
‡
Table V For standard addition of 50% of the nominal content.
Intra-Day and Inter-Day Precision of MIC and NYS Standard Solutions by the HPLC Method

Compound Theoretical Inter-day measured Inter-day measured

concentrations concentrations concentrations Table VII
(mg/mL) (mg/mL) (mg/mL) Parameters Required for System Suitability Testing of the Proposed HPLC Method

Mean RSD% Mean RSD% Parameters MIC NYS

4 3.98 0.06 4.51 0.33 Resolution (Rs) 1.7
MIC 12 12.34 0.22 11.89 0.21 Selectivity (a) 2.8
20 19.75 0.14 20.45 0.13 Symmetry factor (T) 1.01 1.04
20 19.88 0.27 19.98 0.26 Capacity factor (k’) 3.44 1.23
NYS 60 59.69 0.52 59.66 0.62 Number of theoretical plates (N) 2,743 1,557
100 100.33 0.23 99.81 0.23 HETP (cm/plate) 0.01 0.02

HPLC and Chemometric Methods for the Simultaneous Determination of Miconazole Nitrate and Nystatin 5
Selectivity Acknowledgments
Method selectivity was achieved by preparing eight laboratory- The authors would like to acknowledge financial assistance
prepared mixtures of the studied compounds at various con- from Faculty of Pharmacy, Suez Canal University, Ismailia,
centrations within the linearity range. The laboratory-prepared Egypt.
mixtures were analyzed according to the procedures des-
cribed under the proposed methods. Satisfactory results were
obtained (Table VI), indicating the high selectivity of the
proposed methods for simultaneous determination of MIC References
and NYS. 1. Reynolds, J.E.F. (ed). Martindale: The extra pharmacopoeia.
29th edition, Pharmaceutical Press, London, UK, (1989), pp. 430,
666.
2. Hardman, J.G., Limbird, L.E. (eds). Goodman & Gilman’s: The
Analysis of MIC and NYS in Monicure Plus vaginal pharmacological basis of therapeutics. 9th edition, McGraw-Hill,
suppositories New York, NY, (1996).
The four methods were applied to the determination of MIC 3. United States Pharmacopeia (USP DI). Drug information for the
and NYS in commercial Monicure Plus vaginal suppositories. health care professional. 24th edition, Rockville, MD, (2004).
Recoveries were calculated using external regression equations. 4. Sweetman, S.C. (ed). Martindale: The complete drug reference,
35th edition. Pharmaceutical Press, London, UK, (2002), p. 392.
No interfering peaks were observed from any of the excipients.
5. Kobylinska, M., Kobylinska, K., Sobik, B.; High performance liquid
The assay results revealed satisfactory accuracy and precision, chromatographic analysis for the determination of miconazole
as indicated by the recovery and SD values (Table VI). in human plasma using solid-phase extraction; Journal of
Recovery data resulting from the proposed CLS, PCR, PLS Chromatography B, (1996); 685: 191–195.
and HPLC methods were statistically compared with those of 6. Goger, N.G., Gokcen, L.; Quantitative determination of miconazole
the reported HPLC methods: the method of Akay et al. (9) for in creams by second order derivative spectrophotometry;
MIC and that of Groll et al. (12) for NYS, using one way Analytical Letters, (1999); 32: 2595–2602.
7. Wrobel, K., Wrobel, K., de la Garza Rodriguez, I.M., Lopezde-Alba,
ANOVA. It was found that the calculated F values did not P.L., Lopez Martinez, L.; Determination of miconazole in pharma-
exceed the critical value of the F-ratio at a ¼ 0.05, indicating ceutical creams using internal standard and second derivative spec-
no significant differences between the proposed and reported trophotometry; Journal of Pharmaceutical and Biomedical
methods (Table VIII) Analysis, (1999); 20: 99 –105.
8. Khashaba, P.Y., El-Shabouri, S.R., Emara, K.M., Mohamed, A.M.;
Analysis of some antifungal drugs by spectrophotometric and spec-
trofluorimetric methods in different pharmaceutical dosage forms;
Conclusion Journal of Pharmaceutical and Biomedical Analysis, (2003); 22:
Chemometric and HPLC methods have been developed for the 363–376.
simultaneous determination of miconazole and nystatin in their 9. Akay, C., Özkan, S.A., Şentürk, Z., Cevheroğlu, Ş.; Simultaneous de-
termination of metronidazole and miconazole in pharmaceutical
synthetic binary mixtures and vaginal suppositories. The results
dosage forms by RP-HPLC; Il Farmaco, (2003); 57: 953– 957.
obtained using chemometric methods (CLS, PCR and PLS) 10. Lupan, L., Bandula, R., Vasilescu, M., Bercu, C.; Spectroscopic
were compared to those of the proposed HPLC method and no study on nystatin conformational modifications generated by its
significant difference was observed between the methods. The interaction with the solvent; Fresenius Journal of Analytical
HPLC method is superior with regard to identification and Chemistry, (1996); 355: 409– 411.
specificity. However, the chemometric methods are less expen- 11. Botsoglou, N.A., Fletouris, D.J; Rapid spectrophotometric method
sive and do not require sophisticated instrumentation prior for the assay of nystatin in feeds; Journal of Agriculture and Food
Chemistry, (1996); 44: 1271– 1274.
separation steps. The presented methods are suggested to be 12. Groll, A.H., Mickiene, D., Werner, K., Pistelli, S.C., Walsh, T.J.;
used in the routine analysis of miconazole and nystatin in their Determination of nystatin by RP-HPLC.; Journal of Chromatography
synthetic binary mixtures and vaginal suppositories. B, (1999); 735: 51–57.
13. Yeo, S.K., Lee, H.K., Li, S.F.Y.; Separation of antibiotics by
high-performance capillary electrophoresis with photodiode-array
detection; Journal of Chromatography A, (1991); 585:133– 137.
Table VIII 14. Raith, K., Althoff, E., Banse, J., Neidhardt, H., Neubert, R.H.H.; Two
ANOVA Test for Statistical Comparison of the Recovery Data Results Obtained by the Proposed examples of rapid and simple drug analysis in pharmaceutical for-
Chemometric and HPLC Methods and Reported Methods* mulations using capillary electrophoresis: Naphazoline, dexametha-
Compound Source of Sum of Degree of Mean sum of F-ratio† P-value sone and benzalkonium in nose drops and nystatin in an oily
variation squares freedom squares suspension; Electrophoresis, (1998); 19: 2907–2911.
15. Wise, B.M., Gallagher, N.B.; PLS-Toolbox Version 2.1, Eigenvector
MIC Between 2.78 4 0.69 Research Inc., Manson, WA, (2000).
groups
Within 13.62 20 0.68 1.02 0.42 16. Kramer, R.; Chemometric techniques for quantitative analysis.
groups Marcel Dekker Inc., New York, NY, (1998), pp. 20, 51– 142, 168,
Total 16.4 24 169.
NYS Between 3.21 4 0.8 17. Haaland, D.M., Thomas, E.V.; Partial least-squares methods for
groups spectral analyses. 1. Relation to other quantitative calibration
Within 18.06 20 0.91 0.89 0.49
groups methods and the extraction of qualitative information; Analytical
Total 21.27 24 Chemistry, (1998); 60: 1193– 1202.
18. Ni, Y., Gong, X.; Simultaneous spectrophotometric determination of
*Note: Reported methods are from Akay et al. (9) for MIC and Groll et al. (12) for NYS. mixtures of food colorants; Analytica Chimica Acta, (1997); 354:
†
The critical value of the F-ratio is 2.87 at a ¼ 0.05. 163– 171.

6 Heneedak et al.
19. El-Gindy, A.; HPLC and chemometric assisted spectrophotometric 23. The British Pharmacopoeia, Her Majesty’s Stationery Office.
methods for simultaneous determination of diprophylline, pheno- London, UK, (2008), pp. 404–405, 485– 486, 1080– 1081,1382,
barbitone and papaverine hydrochloride; Il Farmaco, (2005) 60: 2519–2520, 2559– 2560, 2592– 2593, 2766–2767, 2861.
745–753. 24. Zeaiter, M., Roger, J.M., Maurel, V.B., Rutledge, D.N.; Robustness
20. Kenneth, R.B., Randy, J.P., Seasholtz, M.B.; Chemometrics: A prac- of models developed by multivariate calibration Part I: The assess-
tical guide. John Wiley and Sons, Inc., New York, NY, (1998), ment of robustness; Trends in Analytical Chemistry, (2004); 23:
pp. 201, 204, 243–245, 278– 290. 157– 170.
21. Miller, J.N., Miller, J.C.; Statistics and chemometrics for analytical 25. Mulholland, M.; Ruggedness testing in analytical chemistry; Trends
chemistry, 4th edition. Pearson Education Limited, England, (2000), in Analytical Chemistry, (1988) 7: 383–389.
pp. 127, 128. 26. Heyden, Y.V., Nijhuis, A., Verbeke, J.S., Vandeginste, B.G.M.,
22. The European Agency for the Evaluation of Medical Products; ICH Massaret, D.L.; Guidance for robustness/ruggedness tests in
Topic Q2B Note for Guidance on Validation of Analytical method validation; Journal of Pharmaceutical and Biomedical
Procedures: Methodology GPMP/ICH/281/95, (1996). Analysis, (2001); 24: 723–753.

HPLC and Chemometric Methods for the Simultaneous Determination of Miconazole Nitrate and Nystatin 7