05_2011.pdf

This art icle was downloaded by: [ Leonardo Meneghini] On: 07 Novem ber 2011, At : 08: 50 Publisher: Taylor & Francis I nform a Lt d Regist ered in England and Wales Regist ered Num ber: 1072954 Regist ered office: Mort im er House, 37- 41 Mort im er St reet , London W1T 3JH, UK Journal of Liquid Chromatography & Related Technologies Publicat ion det ails, including inst ruct ions f or aut hors and subscript ion inf ormat ion: ht t p: / / www. t andf online. com/ loi/ lj lc20 CHEMOMETRIC EVALUATION OF DARIFENACIN HYDROBROMIDE USING A STABILITY-INDICATING REVERSED-PHASE LC METHOD L. Z. Meneghini C. F. Codevilla a a , C. Junqueira , P. E. Fröehlich a a , A. S. Andrade b & A. M. Bergold , F. R. Salazar a , a a Programa de Pós-Graduação em Ciências Farmacêut icas, Faculdade de Farmácia, Universidade Federal do Rio Grande do Sul-UFRGS, Port o Alegre, Brasil b Laborat ório de Pest icidas e Resíduos Vet erinários-Minist ério da Agricult ura do Brazil, Port o Alegre, Brazil Available online: 03 Nov 2011 To cite this article: L. Z. Meneghini, C. Junqueira, A. S. Andrade, F. R. Salazar, C. F. Codevilla, P. E. Fröehlich & A. M. Bergold (2011): CHEMOMETRIC EVALUATION OF DARIFENACIN HYDROBROMIDE USING A STABILITY-INDICATING REVERSED-PHASE LC METHOD, Journal of Liquid Chromat ography & Relat ed Technologies, 34: 18, 2169-2184 To link to this article: ht t p: / / dx. doi. org/ 10. 1080/ 10826076. 2011. 585486 PLEASE SCROLL DOWN FOR ARTI CLE Full t erm s and condit ions of use: ht t p: / / www.t andfonline.com / page/ t erm s- and- condit ions This art icle m ay be used for research, t eaching, and privat e st udy purposes. Any subst ant ial or syst em at ic reproduct ion, redist ribut ion, reselling, loan, sub- licensing, syst em at ic supply, or dist ribut ion in any form t o anyone is expressly forbidden. The publisher does not give any warrant y express or im plied or m ake any represent at ion t hat t he cont ent s will be com plet e or accurat e or up t o dat e. The accuracy of any inst ruct ions, form ulae, and drug doses should be independent ly verified wit h prim ary sources. The publisher shall not be liable for any loss, act ions, claim s, proceedings, dem and, or cost s or dam ages what soever or howsoever caused arising direct ly or indirect ly in connect ion wit h or arising out of t he use of t his m at erial. Journal of Liquid Chromatography & Related Technologies, 34:2169–2184, 2011 Copyright # Taylor & Francis Group, LLC ISSN: 1082-6076 print/1520-572X online DOI: 10.1080/10826076.2011.585486 Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011 CHEMOMETRIC EVALUATION OF DARIFENACIN HYDROBROMIDE USING A STABILITY-INDICATING REVERSED-PHASE LC METHOD L. Z. Meneghini,1 C. Junqueira,1 A. S. Andrade,2 F. R. Salazar,1 C. F. Codevilla,1 P. E. Fröehlich,1 and A. M. Bergold1 1 Programa de Pós-Graduação em Ciências Farmacêuticas, Faculdade de Farmácia, Universidade Federal do Rio Grande do Sul-UFRGS, Porto Alegre, Brasil 2 Laboratório de Pesticidas e Resı´duos Veterinários-Ministe´rio da Agricultura do Brazil, Porto Alegre, Brazil & A set of chemometric tools including in silico simulation and design of experiments (DOE) was used to develop and validate an RP-LC=UV method for darifenacin hydrobromide. Retention factor (k0 ) from darifenacin and resolution (R) between darifenacin and degradation product were assessed by an exploratory study using organic modifier type, organic modifier content, mobile phase pH, flow rate, and column oven temperature as factors in a full factorial 25 design. The significant factors (organic modifier content, mobile phase pH, and column oven temperature) were selected and applied in a central composite design (CCD) with axial points positioned in a ¼ 1.681. The optimum condition was obtained with Derringer’s desirability. The validation was performed with a C8 (150 mm
4.6 mm, 5 lm) column, wavelength 205 nm, 1.1 mL  min1 flow rate, mobile phase with acetonitrile (MeCN) and orthophosphoric acid (OPA) 0.01% (v=v) pH 3.0 in a ratio 28:72 and column maintained at room temperature (25  1 C). Forced decomposition was performed in agreement with the guidelines and collected data were confirmed with LC-MS=MS. The obtained k0 and R were 3.22 and 2.62, respectively. Additionally a Plackett-Burman design was done to study robustness, and the results agreed with the guideline requirements for a stability-indicating method. Keywords darifenacin hydrobromide, Derringer’s desirability, design of experiments, high performance liquid chromatography, stability-indicating INTRODUCTION Development and optimization may be excessively time-consuming in analytical separations, when the technique uses a broad range of factors for this purpose, and more than one response must be optimized concurrently.[1] Reversed-phase liquid chromatography (RP-LC) is a traditional Address correspondence to Ana Maria Bergold, Programa de Pós-Graduação em Ciências Farmacêuticas, Faculdade de Farmácia, Universidade Federal do Rio Grande do Sul-UFRGS. Av. Ipiranga, 2752, CEP 906610-000, Porto Alegre, RS, Brazil. E-mail: ana.bergold@ufrgs.br Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011 2170 L. Z. Meneghini et al. technique for pharmaceutical dosage. It utilizes many experimental conditions (concentration of organic modifier, pH, flow rate, and others) associated with more than one response (retention time, capacity factor, resolution, among others). Since a one-factor-at-a-time optimization could be a complicated process, a chemometric approach involving optimization and validation steps is recommended for a quick, economic, precise, and accurate LC method.[2–4] This paper aimed at developing and validating a simple stabilityindicating LC=UV method, employing the design of experiments (DOE) methodology to determine darifenacin hydrobromide (2,3-dihydrobenzofuran-5-yl)ethyl)-3-pyrrolidinylg-2,2-diphenylacetamide hydrobromide) in an extended-release tablet formulation with 15 mg of the drug. Darifenacin hydrobromide (DF) (Figure 1) is an anticholinergic drug used in patients with an overactive bladder who do not achieve symptom relief and quality of life improvement with conservative management.[5] In the literature, LC=APCI-MS=MS was used for the determination of DF enantiomers in human plasma.[6] In addition, other chiral chromatography methods have been used for the determination of DF in bulk drug and extended release tablets.[7,8] However, they employed techniques that require a special column and complex mobile phase unusual in routine analysis. Furthermore, no degradation products resulting from forced-degradation in the validation study have been reported. DOE provides a variety of designs and chemometric tools for the LC application. In the development stage, two-level factorial design is useful for screening, but it does not provide a detailed model of the system investigated.[1,9–11] The central composite design (CCD) is often used to find the optimum condition and to provide a mathematical model to predict how one or more responses relate to the values of various factors.[9,12,13] When the situation involves simultaneous consideration of multiple responses, a multi-criteria decision making (MCDM) chemometric tool is necessary. There are many deals for MCDM, including the desirability functions (Derringer’s desirability). After an appropriate building response surface model for each response, this technique converts the latter into an individual desirability FIGURE 1 Darifenacin hydrobromide (molecular weight 507.5). Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011 Chemometric Evaluation of Darifenacin Hydrobromide 2171 function (di) to achieve a global optimization.[13–15] Once the method has been developed it must be validated. In official compendia it is encouraged that some requirements, like robustness, be performed with DOE. Many types of factorial designs using small variations in the method parameters can be performed to observe method reliability in routine analysis.[16–18] In the present work, initial chromatographic conditions in the RP-LC were investigated by in silico simulation. An exploratory two-level factorial design was used to assess some factors that could affect the DF retention factor and the resolution. Next, a CCD design was applied for a response surface methodology (RSM) study. Derringer’s desirability has been used to define the global optimum conditions in RSM models and to predict the best chromatographic system. In addition, a Plackett-Burman (PB) design was performed to test robustness in the validation stage. EXPERIMENTAL Materials and Reagents Darifenacin hydrobromide (99.70%) was purchased from Waterstone Technology (USA). Methanol, acetonitrile, and tetrahydrofuran LC grade are from J.T Baker (USA). Orthophosphoric acid, formic acid, and acetic acid were obtained from Merck (Germany). The 0.45-mm nylon filters were purchased from Millipore (USA). Analysis was performed on a Shimadzu chromatographic system with LC20A binary pump, SPD-AVvp UV=VIS photodiode array detector, SIL 20-AC autosampler, and CTO20A oven. Alternatively, for the robustness study, a Shimadzu system with LC20A quaternary pump, SPD-AV UV=VIS photodiode array detector, and Rheodyne 7725i manual injector was used. Samples from forced degradation study were analyzed by LC Agilent 1100 series (quaternary pump) coupled MS=MS API 5000 (LINAC) in electrospray positive mode. The chromatographic separations were achieved with Shimadzu C8 column (150 mm
4.6 mm, 5 mm) for the experimental designs and validation; Merck C8(250 mm
4.6 mm
5 mm), Merck C8 (125 mm
4.6 mm
5 mm) columns in the robustness study; Macherey-Nagel C18 (150 mm
4.6 mm
5 mm), Macherey-Nagel C18 (250 mm
4.6 mm
5 mm), and Merck C8 (250 mm
4.6 mm
5 mm) columns for the in silico simulations. Softwares Osiris 4.2.0.0 version (DATALYS Inc., France) was used for the in silico chromatographic simulations. The experimental designs were analyzed by Design-Expert 7.0.0 version (Stat–Ease Inc., USA) and the other statistical tests (linearity, ANOVA) by STATISTICA 5.0 version (Stat-Soft Inc., USA). 2172 L. Z. Meneghini et al. Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011 Chemometric Methods Initial chromatographic conditions were assessed by simulation using Osiris. This software provides many interfaces (tabs) and it was necessary to perform some runs under specific conditions (inputs) to obtain predictions (outputs) from modifications in the chromatographic parameters. The mobile phase (MP) – pH tab was used with different stationary phases (RP-C18, RP-C8); eluent composition (acetonitrile-water, methanol-water) and pH range (2.5-3.5) were the inputs and the DF retention time was the output. After this trial, two chromatographic systems were performed with samples from the stress testing. In order to improve the chromatographic parameters retention factor (k0 ) and resolution (R), an exploratory 25 general factorial design (Table 1) was applied. In factorial design 2k, k is the number of factors and 2 is the number of levels for each factor. It was therefore possible to select significant factors and to apply a rotatable CCD design with eight factorial points, six axial points (positioned in a ¼ 1.681), and six central points. Once the design has been performed, the first step in RSM is to find a suitable approximation for the true functional relationship between response and the set of independent variables.[9] For an experimental design with three factors, a low-order polynomial model is usually employed: y ¼ b0 þ b1 x1 þ b2 x2 þ b3 x3 þ b12 x1 x2 þ b13 x1 x3 þ b23 x2 x3 ð1:0Þ In this first order response surface model, y represents the estimated response (e.g., resolution); b0, b1, b2, bm is a set of unknown parameters where b0 is the average experimental response, b1 to b3 are the estimated effects (main effects) on the factors (x1,x2,x3), and b12 to b23 are estimated effects with interaction. After model assessment a higher order (quadratic or cubic) polynomial model and=or mathematical transformation in the response is sometimes necessary.[19] TABLE 1 General Full-Factorial Design Levels Factors Units (1) (þ1) A: Organic Modifier Type Category acetonitrile tetrahydrofuran B: Organic Modifier Content % (v=v) 20 40 C: Mobile Phase pH – 2.5 5.0 0.90 1.20 D: Flow Rate mL  min1  E: Column Oven Temperature C 25.0 50.0  (1) and (þ1) are ‘‘low’’ and ‘‘high’’ levels. Chemometric Evaluation of Darifenacin Hydrobromide 2173 Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011 Stress Studies Forced decomposition was performed in agreement with the guidelines and specific literature[16,20] in methanolic solutions containing drug substance or drug product. Intentional degradation was carried out by exposing 5 mL of reference=test stock solution to 5 mL of hydrochloric acid or sodium hydroxide at different concentrations. This procedure was repeated using additional 50 C heat in a water bath. The solutions were transferred to a 10-mL volumetric flask, allowed to attain room temperature, and then neutralized with acid or base (when necessary). Oxidative degradation was performed by mixing equal volumes of standard stock solution with 13% hydrogen peroxide solution in a 10-mL volumetric flask. The samples were allowed to reach ambient temperature and diluted with methanol. The solutions were kept in a dry oven at 50 C and 80 C for different lengths of time to perform a thermal stress study. Photolytic studies were carried out with a 2 mL sample in a quartz container and exposed to light (UV 352 nm) in a photostability chamber (controlled temperature). Blank solutions were prepared by the aforementioned procedure, wherein instead stock solutions methanol was used. The analytical data of the method were collected by PDA detector for the peak purity evaluation. When secondary peak was detected, an ESI LC=MS=MS analysis was performed for confirmation. Validation Validation was performed according to the ICH guidelines Q2(R1)[16] and US Pharmacopeia.[21] System Suitability and Solution Stability The stability of DF in solution was assessed at room temperature, under ultrasound effects, and refrigeration condition at different periods. Prior to beginning the method validation stage, the accuracy and precision of LC data collected were checked. For this, ten injections were performed and the chromatographic parameters were evaluated by RSD. Linearity and Range Solutions of DF were prepared at seven concentrations (0.005, 0.010, 0.015, 0.020, 0.025, 0.030, 0.035 mg  mL1) spanning the range of 25– 175% of the target DF assay concentration (0.020 mg  mL1) for the linearity experiments. Ordinary least-squares method was applied to calibration curve construction y ¼ ax  b, where y is the response (peak area), x is the concentration, and a and b are the slope and intercept, respectively. The 2174 L. Z. Meneghini et al. model was evaluated by determination coefficient regression significance, lack-of-fit, and residual analysis. Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011 Precision Fifty tablets were weighed and finely powdered. A portion of powder was accurately weighed, placed for 15 min in an ultrasound bath and diluted to 0.020 mg  mL1 in mobile phase. Then, the sample solutions were filtered through a 0.45-mm membrane and transferred to vials for LC quantification. Nine experiments were carried out in triplicate on three different days with different analysts and instruments. The RSD was assessed for repeatability and intermediate precision. Accuracy Method accuracy was calculated as a percentage of recovery by assaying the known added amount of analyte in the samples. The concentrations were 75%, 100%, and 125% from nominal concentration. The recoveries RSD were evaluated. Limit of Detection (LOD) and Limit of Quantification (LOQ) LOQ was determined by two approaches: based on signal-to-noise ratio (RS=N), with 10:1 ratio and based on standard deviation of response and slope (SDLOQ). This limit was subsequently validated by the assay of a set of samples. Likewise, LOD was assessed on RS=N (generally 3:1) and SDLOD; additionally the LOD was justified with an experimental test. The equations for LOQ and LOD involving standard deviation of the slope and intercept have been estimated by the following equations: SDLOQ ¼ 10:0
s=S TABLE 2 SDLOD ¼ 3:3
s=S ð2:0Þ Factors and Levels Investigated in the Placket-Burman Design Levels Factors A: MeCN Content(%) in MP B: pH Aquose Component in MP C: Flow Rate D: Wavelength E: Autosampler Injection Volume F: Column Oven Temperature G: Acid Type H: Column Type I–L: Dummies 1.2. 3  Units (1) (þ1) (0) % (v=v) – mL=min nm mL  C Categoric Categoric 27 2.90 1.0 206 10 23 Acetic Merck, 120 mm – 29 3.10 1.2 210 30 27 Formic Merck, 250 mm – 28 3.00 1.1 208 20 25 OPA Shimadzu, 150 mm – – (1), (þ1), (0) are ‘‘low’’, ‘‘high’’ and nominal levels, respectively. Chemometric Evaluation of Darifenacin Hydrobromide 2175 where s is the standard deviation of the response (blank) and S is the slope of calibration curve. Robustness Eleven factors were selected to construct the centered Plackett-Burman design (Table 2), but three factors, named dummy factors represent unreal physical changes in the chromatographic system. Nominal levels are the conditions used for method validation. Pareto chart (with t statistics) and half-normal probability plot were used to indicate relevant effects. Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011 RESULTS AND DISCUSSION Forced Decomposition Study The results are shown in Table 3. A stress condition with hydrogen peroxide 13% for 72 hr resulted in significant degradation of DF drug substance and drug product. The data were collected from a PDA detector and confirmed with LC-MS=MS, (Figure 2). Mass balance between non-degraded DF solution and degraded samples agreed (information not shown). DOE In silico chromatographic simulation demonstrated that C8 (250 mm
4.6 mm
5 mm, Merck) was the best choice for the column. An isocratic elution employing a mixture of acetonitrile (MeCN), methanol (MeOH) and orthophosphoric acid (OPA) (pH 2.5, 0.01%v=v) at a ratio 20:40:40 resulted in adequate parameters for the drug product assay. This HLPC TABLE 3 Conditions Applied to the Forced Degradation Study Assay Stress Agent HCl 0.1 M HCl 1 M NaOH 1 M NaOH 1 M Heat 50 C Heat 80 C Heat 50 C þ HCl 0.1 M Heat 50 C þ HCl 0.1 M H2O2 1.3% H2O2 13% UV 352 nm UV 352 nm Time (hours) Drug Substance Drug Product 24 72 24 72 24 72 14 14 72 72 24 72 99.98 99.75 99.34 100.26 99.70 100.24 99.57 100.12 99.62 92.54 100.38 100.24 99.64 99.11 99.13 99.34 99.10 99.32 99.83 98.83 99.39 94.23 99.32 100.22 Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011 2176 L. Z. Meneghini et al. FIGURE 2 Mass spectrum comparison between DF (a) and peak from stress degradation (b); PDA scanning from DF peak (l) and product from stress degradation (2). (Color figure available online.) system was performed with samples from stress studies. Peak impurity was detected in the 13% hydrogen peroxide solution of the samples. New MeCN: OPA 0.01% ratios were attempted but not enough peak purity was achieved. As selectivity is a function of type and concentration of organic modifier,[11,22] the full-factorial 25 was used to assess the differences between MeCN and tetrahydrofuran (THF) and other conditions. Significant effects with the k0 response were found for all factors analyzed (Figure 3a). Flow rate did not present a significant mean effect or interaction effect for R (Figure 3b). No significant R difference was found between THF and MeCN; than the last one was chosen to be investigated in the CCD (less expensive and damaging for the apparatus). The results of CCD performed in twenty runs are shown in Table 4. Statistical analysis demonstrated that a second order polynomial was most appropriate for the relationship between factors and responses, as the response surface graphs show in the Figures 4 and 5. Based on the k0 results (Figure 4) it can be seen that the k0 value increases in the region with lower pH, higher MeCN content and higher column oven temperature. For R (Figure 5), the lowest values of pH, MeCN content and column oven temperature results in the highest resolution value. The polynomials 2177 Chemometric Evaluation of Darifenacin Hydrobromide Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011 FIGURE 3 Perturbation plots from full-factorial 25 showing the significant effect of the examined factors on the responses (a) Retention factor, (b) Resolution. Where the organic modifier type is (1) for the MeCN and (þ1) for the THF, B is an organic modifier content, C is mobile phase pH, D is a flow rate and E is a column oven temperature. were obtained after ‘‘backward elimination’’ of the nonsignificant coefficients. Derringer’s desirability was used to optimize the k0 and R. The first step in this procedure was to convert each response into an individual desirability function (di) which varies over the range 0  di  1, where di ¼ 1 is the best TABLE 4 Experimental Design for DCC Factors and Levelsa Runs 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 a MeCNb 22.0 28.0 22.0 28.0 22.0 28.0 22.0 28.0 20.0 30.0 25.0 25.0 25.0 25.0 25.0 25.0 25.0 25.0 25.0 25.0 (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (-1.681) (þ1.681) (0) (0) (0) (0) (0) (0) (0) (0) (0) (0) pHc 3.0 3.0 5.0 5.0 3.0 3.0 5.0 5.0 4.0 4.0 2.3 5.7 4.0 4.0 4.0 4.0 4.0 4.0 4.0 4.0 (1) (1) (þ1) (þ1) (1) (1) (þ1) (þ1) (0) (0) (1.681) (þ1.681) (0) (0) (0) (0) (0) (0) (0) (0) Responses Oven Temperatured 26.0 26.0 26.0 26.0 45.0 45.0 45.0 45.0 35.5 35.5 35.5 35.5 19.5 51.5 35.5 35.5 35.5 35.5 35.5 35.5 (1) (1) (1) (1) (þ1) (þ1) (þ1) (þ1) (0) (0) (0) (0) (1.681) (þ1.681) (0) (0) (0) (0) (0) (0) (1), (þ1), (0) are ‘‘low’’, ‘‘high’’ and nominal levels, respectively. Acetonitrile content (%) in mobile phase. c pH of the mobile phase. d Column oven temperature in degree celsius. b k0 R 11.40 4.41 19.73 6.01 7.34 3.12 13.93 4.32 16.83 3.23 5.30 13.70 8.39 4.31 6.45 6.08 6.46 6.92 7.21 6.87 3.68 2.51 1.90 2.53 3.25 2.15 0.73 1.68 2.08 1.88 3.07 1.57 3.03 2.17 2.60 2.37 2.42 2.62 2.60 2.55 Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011 2178 L. Z. Meneghini et al. FIGURE 4 RSM graphs showing the effect of factors on retention factor (k0 ): (a) MeCN content (A) and pH (B) in mobile phase; (b) MeCN content (A) and column oven temperature (C); (c) mobile phase pH (B) and column oven temperature (C). (Color figure available online.) FIGURE 5 RSM graphs showing the factor effects on resolution (R): (a) MeCN content (A) and mobile phase pH (B); (b) mobile phase pH (B) and column oven temperature (C). (Color figure available online.) Chemometric Evaluation of Darifenacin Hydrobromide 2179 alternative and di ¼ 0 when the response is outside the optimal region. Then the design variables are chosen to maximize global desirability (D), D ¼ ðd1
d2
. . . dm Þ1=m ð3:0Þ where there are m responses and d1,d2 . . . dm are the individual desirabilities. If the objective or target T for the response y is a maximum value: Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011 d¼ 8 < 0;  yL T L : 1; r y<L LyT y>T ð3:1Þ when weight r ¼ 1, the desirability function is linear. Choosing r > 1 places more emphasis on being close to the target value, and choosing 0 < r < 1 makes this less important. If the target for the response is a minimum value: d¼ 8 < 1;  U y U T : 0; r y<T T yU y>U ð3:2Þ the two-sided desirability function (3.3) assumes that the target is located between the lower (L) and the upper (U) limits, and is defined as: 8 0; >  r 1 > > > < yL T L d¼  r 2 > U y > > > : U T 0; y<L LyT ð3:3Þ LyU y>U FIGURE 6 Desirability’s Derringer RSM graph showing optimum operability region of the chromatographic method where (A) is MeCN content and (B) is pH in mobile phase. (Color figure available online.) 2180 L. Z. Meneghini et al. TABLE 5 Parameters Used in Derringer’s Desirability Criteria Response= Parameter Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011 k0 R Column oven temperature Lower Limit Upper Limit Goal Importance 2 2 23 5 4 28 Target ¼ 3.5 Range Target ¼ 25 C 2 2 1 The results of Derringer’s function are demonstrated in Figure 6. The conditions proposed for criteria selection have been established in the appropriate parameters for routine analysis (Table 5). Room temperature was chosen to be used for method simplification. Ten solutions have been obtained from the response surface producing a desirability value (D) of TABLE 6 Requirements Assessed for the RP-HPLC=UV Method Validation System Suitability Test (n ¼ 10) Area RSDa Parameter R2 b ¼ 743 a ¼ 70109 Rt b 1397478 8.45 0.03 0.47 Linearity CLg Seh – 12877 13152 6283 588 281 k0 Tc Nd 3.22 0.12 1.25 0.35 p-value – 0.91 <0.01 RS=N 79.69 Re 4712 2.62 0.22 0.46 LQ analysisf SDLOD Experimental 11.78 400.00 LD analysisf 3.89 20.00 29.30 Model Assessment Parameter Statistical Test Regression Lack-of-Fit Autocorrelation in Residuals Homoscedasticity ANOVA; F-value ANOVA; p-value Durbin-Watson; D-value Levene; p-value Barllell; p-value Ryan-Joiner; p-value Normality Test Kolmogorov–Smirnoff; p-value Anderson-Darling; p-value Dixon; s-value Outlier Assumption a Relative standard deviation. Retention time. c Tail. d Theorical plates. e Resolution. f Concentration in ng mL1. g Confidence limit. h Standard error. b Critical Value Calculated Value 3.87 0.05 dL ¼ 1.22; dv ¼ 1.42 0.05 0.05 0.05 62240.80 0.99 2.72 0.72 0.15 0.12 0.15 0.05 0.05 0.976 0.30 0.680 2181 Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011 Chemometric Evaluation of Darifenacin Hydrobromide 0.726, where the MeCN content, mobile phase pH, and column oven temperature considering the global optimum condition were 28%, 3.0, and 25.0 C, respectively; the predicted k0 and R were 4.46 and 2.45, respectively. The mobile phase pH and organic modifier content play an important role for elution and selectivity in RP=LC mechanisms and could have a dramatic effect on separation of the ionizable compounds.[11,22] The 8.80 pKa and basic behavior in aqueous solution were predicted for DF by ALOGPS2.1 simulator and the molecule until 7.8 would be in the ionized form. The changes in flow rate did not demonstrate any effect on the resolution parameter, probably because the DF has a poor basic behavior, without a free primary amine which could interact with the free silanols in the stationary phase.[11,23] Method Validation The samples showed stability in 45 d under refrigeration (methanolic solution) and 2 d at room temperature (mobile phase solution). The parameters utilized for the present validation were 205 nm (wavelength), 1.1 mL  min1 flow rate, mobile phase with MeCN and OPA (pH 3.0, 0.01% v=v) in at a 28:72 ratio and column maintained at room temperature (25  1 C). The results are presented in Table 6, Table 7; and the chromatogram is shown in Figure 7. Resolution between peaks of darifenacin and degradation product was adequate, resulting in peak purity (UV=PDA) for the darifenacin peak. As seen in Table 6, LOQ and LOD results, using different approaches, are not similar. The greatest difference occurred between TABLE 7 Analysis for Accuracy and Precision for the Method Validation Accuracy Range (%) Mean (%) RSDa (%) 97.45–99.18 98.33 0.88 Precision Conditionb Repeatability N 1 N 2 N 3 Intermediary a Mean assay (%) RSD (%) 99.43 101.12 98.33 0.52 0.78 0.60 99.63 1.32 Standard error. N  1: Day 1, Analyst 1, Instrument 1; N  2: Day 2, Analyst 2, Instrument 2; N  3: Day 3, Analyst 1, Instrument 2. b Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011 2182 L. Z. Meneghini et al. FIGURE 7 Chromatogram showing hydrogen peroxide (1), DF (2) and secondary peak from oxidation stress (3) performed on validation chromatographic parameters. SDLOQ and the experimental value, because the point in LOQ is not only the lowest quantification level but a compromise with the precision[18] and this was only possible with 400.00 ng  mL1 level (RSD < 1.0%). For the precision study (Table 7), although were used variations involving analysts, days, and instruments; the method showed adequate values for the RSD. The results for the robustness study are presented in Figure 8. The method can be considered robust because the significant factors pointedout in a half-normality curve (not shown) and Paretto chart (F factor for percentage assay and A factor for tail) have been statistically different from nominal condition, but they do not exceed the limits from the model requirements (2% RSD for the repeatability and 2.0 as the maximum value for tail). When one or more dummy factors are excluded from the model, this situation is maintained (not shown). Also, the other chromatographic parameters (theoretical plates, retention factor, and resolution) presented values within appropriate specifications. FIGURE 8 Pareto charts (a, b) for the effects analysis showing the effects on assay % response (a) and (b) showing the effects on tail, where A is acetonitrile content (%, v=v) and F is column over temperature. Chemometric Evaluation of Darifenacin Hydrobromide 2183 Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011 CONCLUSIONS The study successfully developed and validated a stability-indicating RP-LC=UV method. The chemometric methodologies employed were useful to improve information about the influence of the factors on responses R and k0 without taking a lot of time and being cost-consuming (no buffers, low organic modifier content). It was only possible to optimize the chromatographic method using Derringer’s desirability, because resolution and retention factor presented distinct behaviors when the factors were changed. A robustness study provided more complete evaluation about deliberate variations in the method. Finally, it can be said that the method developed is fully applicable to the routine assay of the drug product. ACKNOWLEDGMENTS The authors thank CAPES and CNPq of Brazil, LANAGRO (Brazil) and Labex Inc. (USA). REFERENCES 1. Harvey, D. Modern Analytical Chemistry, McGraw-Hill: New York, NY, 2000. 2. Jancic-Stojanovic, I. J.; Malenovic, A.; Markovic, S.; Ivanovic, D.; Medenica, M. Optimization and Validation of an RP-HPLC Method for Analysis of Hydrocortisone Acetate and Lidocaine in Suppositorie. J. AOAC Int. 2010, 93 (1), 102–107. 3. Barmpalexis, P.; Kanaze, F. I.; Georgarakis, E. Developing and Optimizing a Validated Isocratic Reversed-Phase High-Performance Liquid Chromatography Separation of Nimodipine and Impurities in Tablets Using Experimental Design Methodology. J. Pharm. Biomed. Anal. 2009, 49, 1192–1202. 4. Vignaduzzo, S. E.; Castellano, P. M.; Kaufman, T. S. Experimentally Designed, Validated HPLC Simultaneous Determination of Pridinol and Diclofenac in Their Combined Pharmaceutical Formulations, Which Allows Limiting Diclofenac Related Compound A. J. Liq. Chromatogr. R. T. 2010, 33 (19), 1720–1732. 5. Novara, G.; Galfano, A.; Secco, S.; D’Elia, C.; Cavalleri, S.; Ficarra, V.; Artibani, W. A Systematic Review and Meta-Analysis of Randomized Controlled Trials with Antimuscarinic Drugs for Overactive Bladder. Eur. Urol. 2008, 54, 740–764. 6. Kaye, B.; Herron, W. J.; Macrae, P. V.; Robinson, S.; Stopher, D. A.; Venn, R. F.; Wild, W. Rapid, Solid Phase Extraction Technique for the High-Throughput Assay of Darifenacin in Human Plasma. Anal. Chem. 1996, 68, 1658–1660. 7. Venn, R. F.; Goody, R. J. Synthesis and Properties of Molecular Imprints of Darifenacin: The Potential of Molecular Imprinting for Bioanalysis. Chromatographia 1999, 50, 407–414. 8. Radhakrishnanand, P.; Rao, D. V. S.; Himabindu, V. H. A Validated LC Method for Determination of the Enantiomeric Purity of Darifenacin in Bulk Drug and Extended Release Tablets. Chromatographia 2008, 68, 1059–1062. 9. Montgomery, D. C. Design and Analysis of Experiments, Wiley: New York, NY, 2001. 10. Brereton, R. G. Chemometrics: Data Analysis for the Laboratory and Chemical Plant, Wiley: New York, 2003. 11. Kazakevich, Y.; Lobrutto, R. LC for Pharmaceutical Scientists, Wiley: New Jersey, 2007. 12. Myers, R. H. Response Surface Methodology: Process and Product Optimization Using Designed Experiments, Wiley: New Jersey, 2009. Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011 2184 L. Z. Meneghini et al. 13. Jancic-Stojanovica, B.; Malenovica, A.; Ivanovica, D.; Rakica, T.; Medenica, M. Chemometrical Evaluation of Ropinirole and Its Impurity’s Chromatographic Behavior. J. Chromatogr. A 2009, 1216, 1263–1269. 14. Lewis, G. A.; Roger, D. M.; Phan, T. L. Pharmaceutical Experimental Design, Marcel Dekker: New York, 1999. 15. Sivakumar, T.; Manavalan, R.; Muralidharan, C.; Valliappan, K. Multi-Criteria Decision Making Approach and Experimental Design as Chemometric Tools to Optimize HPLC Separation of Domperidone and Pantoprazole. J. Pharm. Biomed. 2007, 43, 1842–1848. 16. ICH. International Conference on Harmonisation-Harmonised Tripartite Guideline: Validation of Analytical Procedures: Text and Methodology, Q2(R1), 2005. 17. Dejaegher, B.; Heyden, Y. V. Ruggedness and Robustness Testing. J. Chromatogr. A 2007, 1158, 138–157. 18. Shabir, G. A. Validation of High-Performance Liquid Chromatography Methods for Pharmaceutical Analysis Understanding the Differences and Similarities Between Validation Requirements of the US Food and Drug Administration, the US Pharmacopeia and the International Conference on Harmonization. J. Chromatogr. A 2003, 987, 57–66. 19. Mason, R. L.; Gunst, R. F.; Hess, J. L. Statistical Design and Analysis of Experiments, Wiley: New Jersey, 2003. 20. Bakshi, M.; Singh, S. Development of validated stability-indicating assay methods—critical review. J. Pharm. Biomed. Anal. 2002, 28, 1011–1040. 21. USP. 31th edition. The United States Pharmacopeia, Rockville, United States, 2008. 22. Glajch, J. L.; Kirkland, J. J.; Squire, K. M. Optimization of solvent strength and selectivity for reversed-phase liquid chromatography using an interactive mixture-design statistical technique. J. Chromatogr. 1980, 199, 57–59. 23. McCalley, D. V. Effect of Temperature and Flow-Rate on Analysis of Basic Compounds in HighPerformance Liquid Chromatography using a Reversed-Phase Column. J. Chromatogr. A 2000, 902, 311–321.