(PDF) JSCS-3908

Electronic offprint for author’s personal use JSCSEN 74(10)1021–1153(2009) UDC 54:66 ISSN 0352–5139 ve ersion lectronic Available on line at VOLUME 74 No 10 BELGRADE 2009 www.shd.org.rs/JSCS/ The full search of JSCS is available through www.doaj.org Author's Personal Copy J. Serb. Chem. Soc. 74 (10) 1143–1153 (2009) UDC *digoxin:661.12.002.8:615:62–776 JSCS–3908 Original scientific paper Validation of an HPLC–UV method for the determination of digoxin residues on the surface of manufacturing equipment 1 1# 1# ZORAN B. TODOROVIĆ *#, MIODRAG L. LAZIĆ , VLADA B. VELJKOVIĆ 2# and DRAGAN M. MILENOVIĆ 1 Faculty of Technology, Bulevar oslobođenja 124, 16000 Leskovac, University of Niš and 2 “Zdravlje-Actavis“ Company, Analytical Department, Vlajkova 199, 16000 Leskovac, Serbia (Received 25 February, revised 15 May 2009) Abstract: In the pharmaceutical industry, an important step consists in the removal of possible drug residues from the involved equipments and areas. The cleaning procedures must be validated and methods to determine trace amounts of drugs have, therefore, to be considered with special attention. An HPLC–UV method for the determination of digoxin residues on stainless steel surfaces was developed and validated in order to control a cleaning procedure. Cotton swabs, moistened with methanol were used to remove any residues of drugs from stainless steel surfaces, and give recoveries of 85.9, 85.2 and 78.7 % for three concentration levels. The precision of the results, reported as the relative standard deviation (RSD), were below 6.3 %. The method was validated over a concentration range of 0.05–12.5 µg mL-1. Low quantities of drug residues were determined by HPLC–UV using a Symmetry C18 column (150×4.6) mm, 5 µm) at 20 °C with an acetonitrile–water (28:72, v/v) mobile phase at a flow rate of 1.1 mL min-1, an injection volume of 100 µL and were detected at 220 nm. A simple, selective and sensitive HPLC–UV assay for the determination of digoxin residues on stainless steel was developed, validated and applied. Keywords: cleaning validation; digoxin; swab analysis; residues. INTRODUCTION An important step in the manufacture of pharmaceutical products is the clean- ing of equipment and surfaces. The cleaning procedures for the equipment must be validated according to good manufacture practice (GMP) rules and guidelines. The main objective of cleaning validation is to avoid contamination between dif- ferent productions or cross-contamination. This cleaning is verified by deter- mining the amount of residues on surfaces involved in the manufacture process. * Corresponding author. E-mail:

# Serbian Chemical Society member. doi: 10.2298/JSC0910143T 1143 Available online at www.shd.org.rs/JSCS/ ___________________________________________________________________________________________________________________________ 2009 Copyright (CC) SCS Author's Personal Copy 1144 TODOROVIĆ et al. Cleaning validation consists of two separate activities: the first is the deve- lopment and validation of the cleaning procedure used to remove drug residues from manufacturing surfaces and the second involves the development and vali- dation of methods for quantifying residuals from the surfaces of manufacturing equipment. Furthermore, many sampling points of the manufacturing facility and the manufacturing equipment have to be tested to verify the occurrence of conta- mination. For these reasons, an analytical method for residue monitoring should also be rapid and simple.1 The acceptable limit for residue in equipment is not established in the cur- rent regulations. According to the FDA, the limit should be based on logical cri- teria, involving the risks associated with residues of a determined product. The calculation of an acceptable residual limit, the maximum allowable carryover of active products in production equipment should be based on therapeutic doses, the toxicological index and a general limit (10 ppm). Several mathematical for- mulas were proposed to establish the acceptable residual limit.1–7 Digoxin (Fig. 1) is a cardioactive glycoside isolated from the leaves of Digi- talis lanata and is the most frequently used drug in the treatment of congestive heart failure.8 Numerous methods have been published for the quantitative deter- mination of digoxin in raw materials,9–13 tablets,14–18 other solid dosage forms,19 human plasma, urine, serum20–23 and Digitalis leaves.24,25 Fig. 1. Chemical structure of di- goxin. A literature survey revealed that no validation of cleaning methods for di- goxin is to be found. Due to their high sensitivity and selectivity, analytical me- thods such as liquid chromatography were previously reported to be used for the determination of residues to control cleaning procedures.5–7,26–32 Taking the above-mentioned consideration into account, the aim of this stu- dy was to develop and validate a simple analytical method that allows the deter- mination of trace levels of digoxin residues in production area equipment and to confirm the efficiency of the cleaning procedure. The analytical method reported was validated considering selectivity, linearity, accuracy, precision and limits of detection (LOD) and quantification (LOQ). The stability of digoxin samples was also studied.33 Available online at www.shd.org.rs/JSCS/ ___________________________________________________________________________________________________________________________ 2009 Copyright (CC) SCS Author's Personal Copy HPLC DETERMINATION OF DIGOXIN RESIDUES 1145 EXPERIMENTAL Chemicals and reagents The certified digoxin, working standard was purchased from Roche Diagnostics GmbH (Mannheim, Germany). Acetonitrile and methanol (HPLC gradient grade) were purchased from J. T. Baker (Deventer, Holland). Purified water was obtained from Arium Laboratory Equipment (RO, UV) Sartorius AG (Gottingen, Germany). The extraction–recovery sampling was realized with Alpha® Swab polyester on a polypropylene handle – TX714A (ITW Tex- wipe®, Mahwah, NJ, USA). The mobile phase was filtered through a 0.45 µm Sartorius membrane filter (Gottingen, Germany). Equipment The HPLC system consisted of a degasser G1379B, a bin pump G1312A, an automatic injector G1329A, a thermostated column oven G1316A and a multiwavelength detector G1365B, all 1200 Series from Agilent Technologies, which were controlled by an HP Chem- station software (Waldbroon, Germany). The employed ultrasonic bath was from Elma, Trans- sonic 470/H (Singen, Germany) and the analytical balance was from Sartorius AG, CP224S – – OCE (Gottingen, Germany); accuracy of the balance: ± 0.0001 g. Chromatographic conditions All chromatographic experiments were performed in the isocratic mode. The mobile phase was acetonitrile-water (28:72, v/v) at a flow rate of 1.1 mL min-1. The separation was performed at 20 °C on a Symmetry C18 column ((150×4.6) mm, 5 µm) from Waters (Milford, MA, USA). UV detection was realized at 220 nm. The injection volume was fixed at 100 µL. Standard solutions preparation The stock solution of standard was prepared by accurately weighing digoxin standard (approximately 12.5 mg) and transferring it into a 100 mL volumetric flask. Approximately, 20 mL of methanol was added and the content of the flask was sonified for 15 min. The so- lution in the flask was diluted to volume with 20 mL acetonitrile and water; the final concen- tration being 125 µg mL-1 digoxin. Dilutions were later prepared with water to obtain solu- tions for calibration (0.05–12.5 µg mL-1) and standard solution for the positive swab control at three concentration levels (0.625, 3.125, and 15.625 µg per swab level). These solutions were filtered through a 0.45 µm regenerated cellulose filter before analysis and injected in trip- licate. Sample preparation The selected surfaces (5 cm×5 cm) of stainless steel, previously cleaned and dried, were sprayed with 500 µL of a standard solution, for the positive swab control at all concentration levels, and the solvent was allowed to evaporate (approximate time was 2 h). The surfaces were wiped with the first cotton swab soaked with methanol, passing it in various directions, to remove the residues from the stainless steel. The other dry cotton swab was used to wipe the wet surfaces. The swabs were placed into a 25 mL screw cap test tube, and 5.0 mL of the purified water was pipetted into adequate sample tubes. The background control sample was prepared from the extraction media. The negative swab control was prepared in the same way as the samples, using swabs, which had not been in contact with the test surface. In addition, test and excipient solutions were prepared according to the content of tablets to assure that they did not interfere with digoxin. Subsequently, the tubes were placed in an ultrasonic bath for 15 min and the solutions were analyzed by HPLC–UV. Available online at www.shd.org.rs/JSCS/ ___________________________________________________________________________________________________________________________ 2009 Copyright (CC) SCS Author's Personal Copy 1146 TODOROVIĆ et al. RESULTS AND DISCUSSION Acceptance limit calculation The maximum allowable carryover (MACO) is the acceptable transferred amount from the previous to the following product. The MACO is determined based on the therapeutic dose, toxicity and generally 10 ppm criterion. Once the maximum allowable residue limit in the subsequent product was determined, the next step was the determination of the residue limit in terms of the contamination level of active ingredient per surface area of equipment. The total surface area of the equipment in direct contact with the product was accounted for in the calcul- ations. The limit per surface area was calculated from the equipment surface area and the most stringent maximum allowable carryover (the most stringent crite- rion being based on the therapeutic dose in this case). The 0.1 % dose limit cri- terion is justified by the principle that an active pharmaceutical ingredient (API) at a concentration of 1/1000 of its lowest therapeutic dose will not produce any adverse effects on human health. The calculated limit per surface area (LSA) in the case of digoxin was 1.20 µg swab–1 pro 25 cm2. A stainless steel surface area of 5 cm×5 cm was chosen for practical reasons. Optimization of the chromatographic conditions To obtain the best chromatographic conditions, the wavelength for detection, the column and the mobile phase composition were adequately selected. The main objective was to develop an HPLC–UV method that, running in the isocratic mode, allowed the determination of digoxin residues collected by swabs, without interference of impurities originating from the swabs, plates and extraction media. For the analysis, 220 nm was selected for detection because the drug has a sufficient absorption at this wavelength and low quantities of digoxin may be de- tected correctly. Furthermore, the calibration curve obtained at 220 nm showed good linearity. Regarding the chromatographic procedure, different C18 columns were eva- luated: LiChrosorb RP-18, Nucleosil C-18, µBondapak C-18, but Symmetry C-18 ((150×4.6) mm, 5 µm) was preferred to improve the peak symmetry and to obtain an appropriate retention time. A mixture of acetonitrile–water in different proportions is frequently used as the mobile phase. In order to improve the selectivity and sensitivity, the amount of acetonitrile was varied between 20 and 40 vol. %. The best separation was achieved with the proposed mobile phase acetonitrile–water (28:72, v/v) at a flow rate of 1.1 mL min–1. Furthermore, with decreasing flow rate, the column back- pressure is lower and hence the column life is prolonged. The injection volume was varied between 10 and 100 µL and was finally set at 100 µL in order to increase the sensitivity of the method without sacrificing the chromatographic peak shape. Taking into account the results obtained with Available online at www.shd.org.rs/JSCS/ ___________________________________________________________________________________________________________________________ 2009 Copyright (CC) SCS Author's Personal Copy HPLC DETERMINATION OF DIGOXIN RESIDUES 1147 different C18 columns and the tested mobile phases, the already defined chroma- tographic conditions were finally chosen. The sensitivity achieved by this method was much better than the initial method given by Song et al.,9 which was mo- dified in order to accommodate the requirements of trace analysis specific for a cleaning validation study. Optimization of the sample treatment Cotton swabs were spiked with different quantities of digoxin and placed into glass tubes. After the addition of different solvents (water, methanol and the mobile phase), the tubes were sonified for different times (5, 15 and 30 min) and the solutions analyzed by HPLC. The optimum conditions were achieved with water as the extracting solvent and a sonification time of 15 min. In all cases, the best results were obtained using two cotton swabs (the first wetted with methanol and the second dry); hence, this technique was applied in the subsequent work. Validation of the method Once the chromatographic conditions had been selected, the method was va- lidated, whereby attention was paid to the selectivity, linearity, limit of detection, limit of quantification, precision, accuracy and sample stability. Suitability test. System suitability testing is essential for the assurance of the quality performance of a chromatographic system. During performing the system suitability tests, in all cases, RSD of the peak areas was < 2 %, the average num- ber of theoretical plates per column was > 5500 and the USP tailing factor was about 1.08. Selectivity. The selectivity of the method was checked by injecting the digo- xin standard, the background control sample, the negative swab control, the un- spiked stainless steel 5 cm×5 cm plate swabbed as described and the excipient mixture (according to the content of the tablets). The results are shown in Fig. 2, from which it can be observed that there were no mutual interferences. Linearity. Linearity data were obtained by plotting the area of the digoxin peak, expressed in area units, against the concentration of digoxin, expressed as µg mL–1. A linear regression least square analysis was performed in order to de- termine the slope, intercept and coefficient of determination. The standard curve was linear from 0.05–12.5 µg mL–1. The values of the slope, intercept and coef- ficient of determination of the calibration curve for digoxin are given in Table I. The high value of the coefficient of determination indicated good linearity. Limit of detection and quantification. LOD and LOQ were determined based on the standard deviation of the response (y-intercept) and the slope of the calib- ration curve at low concentration levels according to ICH guidelines. The LOD and LOQ for digoxin were found to be 0.010 and 0.030 µg mL–1, respectively. Available online at www.shd.org.rs/JSCS/ ___________________________________________________________________________________________________________________________ 2009 Copyright (CC) SCS Author's Personal Copy 1148 TODOROVIĆ et al. Fig. 2. Chromatograms obtained from: a) non-spiked stainless steel surface, b) excipient mixture and c) standard solution of digoxin (1.25 µg mL-1). TABLE I. Linear regression data in the analysis of digoxin Statistical parameters Values Concentration range, µg mL-1 0.050–12.50 Regression equation y = 64.342x + 4.7658 Coefficient of determination r2 = 0.9999 S(a)-error in intercept 2.684 S(b)-error in slope 0.428 S(y/x)-error for y-est 0.195 Precision and accuracy. The precision and accuracy of the procedure, repor- ted as relative standard deviation (RSD) and the recovery (%), respectively, were assessed by comparing the amount of analyte determined versus the known amount spiked at three different concentration levels with 6 replicates (n = 6) for each in- vestigated concentration level. The method precision and accuracy were determi- ned for spiked and dried swabs and plates. The recovery, 95 % confidence inter- val, and the RSD values obtained on spiked and dried swabs and plates (Table II) for each level illustrated the good precision and accuracy of the method. These precision and recovery results are acceptable for the purpose of residue moni- toring. Available online at www.shd.org.rs/JSCS/ ___________________________________________________________________________________________________________________________ 2009 Copyright (CC) SCS Author's Personal Copy HPLC DETERMINATION OF DIGOXIN RESIDUES 1149 TABLE II. Precision and accuracy of the results obtained from swabs and plates spiked with digoxin Amount added Amount found 95 % Confidence Recovery RSD / % Samples µg mL-1 µg mL-1 interval, % % n=6 Swabs 0.125 0.116 92.3–93.2 92.7 3.1 0.625 0.580 89.8–95.2 92.5 3.6 3.125 2.836 89.9–91.6 90.7 0.6 Plates 0.125 0.107 83.5–86.3 85.9 3.0 0.625 0.530 80.9–89.5 85.2 6.3 3.125 2.461 77.1–80.4 78.7 1.3 The intermediate precision of the method was investigated by performing five consecutive injections of standard solutions on two different days by different analysts. On both days, the RSD was calculated for peak area responses obtained for the digoxin peak. The data obtained suggested that the method exhibited an acceptable intermediate precision with less than 1.0 % RSD for the digoxin stan- dard solution when analyzed on two different days by two different analysts. Sample stability. The stability of the digoxin in the swab matrix was tested. The spiked samples at all concentration levels were stored after analyses in the injector vials in the auto-sampler tray at ambient temperature for 7 days. All the samples were injected into the appropriate HPLC system after 24 h, 48 h and 7 days against fresh standard solutions. The stability of the standard digoxin solu- tion (1.25 µg mL–1) was also inspected after storage for 7 days at ambient tempe- rature and refrigerated. No changes in the chromatography of the stored samples were found and no additional peaks were registered when compared with the chromatograms of the freshly prepared samples (Table III). TABLE III. Stability results obtained from digoxin swab extract samples a Sample (n = 3) Mean recovery±IC , % µg per swab 0h 24 h 48 h 7 days 0.625 92.2±1.7 93.5±1.8 92.6±1.0 93.2±0.6 3.125 90.2±1.3 90.6±1.5 90.7±0.8 90.1±0.8 15.62 88.9±1.2 89.3±0.8 89.3±1.1 88.9±1.3 a Interval of confidence Assay of swab samples collected from different locations within the equipment train Swab samples from different locations within the manufacturing equipment train were submitted to the laboratory for analysis of residual digoxin. These samples were prepared and analyzed by the proposed method. Some of the re- sults obtained for these samples are presented in Table IV. An example of a chro- matogram which originated from Glatt WSG (Bowl bottom mesh point), which is below the limit per surface area (LSA) and above the LOQ is shown in Fig. 3. Available online at www.shd.org.rs/JSCS/ ___________________________________________________________________________________________________________________________ 2009 Copyright (CC) SCS Author's Personal Copy 1150 TODOROVIĆ et al. TABLE IV. Results obtained for the determination of digoxin in actual swab samples col- 2 lected from 25 cm swabbed area from different locations within the equipment train Equipment swabbed Location swabbed Digoxin detected, µg per swab Material dispensing scoops Internal surface <LOQ Back round plate <LOQ External surface <LOQ High shear mixer – Diosna Bottom of gran. bowl <LOQ Chopper shaft <LOQ Impeller blade <LOQ Turbo sieve – Bohle Stainless steel inlet ring 0.24(<LSAa) Product inlet, side wall <LOQ Sieve unit <LOQ Fluid bed dryer – Glatt WSG Inside wall <LOQ Viewing window <LOQ Bowl bottom mesh 0.87(<LSA) Pillar hoist – FBD bowl inverter Internal surface-top <LOQ Internal surface-bottom <LOQ Collar <LOQ Washer – extractor Miele Drum back plate <LOQ Drum perforated surface <LOQ Door-middle <LOQ Metal detector – Lock Met 30+ In-feed chute 0.223(<LSA) Reject device, corner <LOQ Reject flap <LOQ Deduster – Kramer Inlet plate 0.277(<LSA) Dedusting helix <LOQ Outlet <LOQ Tablet press – Kilian The table 0.357(<LSA) Tablet chute cover <LOQ Main gate 0.355(<LSA) aLimit per surface area CONCLUSIONS In conclusion, a simple to use HPLC–UV method to quantify residues of the active pharmaceutical ingredient digoxin on swabs, in support of cleaning vali- dation of pharmaceutical manufacturing equipment, was developed. Validation studies showed that the HPLC–UV method is selective, linear, precise and ac- curate. To extract the digoxin residues from the surfaces, a wipe test procedure using two cotton swabs is recommended. The recoveries obtained from the stain- less steel surfaces were close to 80 % or higher and there was no interference from the cotton swab. Stability studies show that the digoxin swab samples are, at least, stable over the investigated 7 days. The overall procedure can be used as part of a cleaning validation program in the pharmaceutical manufacture of di- goxin. Available online at www.shd.org.rs/JSCS/ ___________________________________________________________________________________________________________________________ 2009 Copyright (CC) SCS Author's Personal Copy HPLC DETERMINATION OF DIGOXIN RESIDUES 1151 Fig. 3. Example of a chromatogram from a swab sample (Glatt WSG-Bowl bottom mesh point). Ц HPLC–UV Ђ Ш 1 1 1 2 . , . , . . 1 Tehnolo{ki fakultet, Bulevar oslobo|ewa 124, 16000 Leskovac, Univerzitet u Ni{u 2 “Zdravqe-Actavis“, Odeqewe analiti~kog razvoja, Vlajkova 199, 16000 Leskovac , - . ђ , ђ . ђ HPLC–UV ђ . , , - , 85,92, 85,22 78,74 %, . , (RSD), 6,3 %. ђ 0,05–12,5 µg mL-1. ђ HPLC Symmetry C18 ((150×4,6) mm, 5 µm), 20 °C, - – (28:72, v/v) 1,1 mL min-1, - 100 µL 220 nm. - , , ђ , HPLC–UV ђ . ( 25. , 15. 2009) Available online at www.shd.org.rs/JSCS/ ___________________________________________________________________________________________________________________________ 2009 Copyright (CC) SCS Author's Personal Copy 1152 TODOROVIĆ et al. REFERENCES 1. Guide to Inspections Validation of Cleaning Processes, U.S. Food and Drug Admini- stration, (FDA), Office of Regulatory Affairs, Washington, DC, pp. 1–6, http://www.fda.gov/ /ora/inspect_ref/igs/valid.html (accessed 20 April, 2009) 2. Cleaning Validation Guidance, Guidance on aspects of cleaning validation in active pharmaceutical ingredient plants, http://apic.cefic.org/pub/pub-cleaning-validation.pdf (accessed 20 April, 2009) 3. Guide to Cleaning Validation in API plants, Cleaning Validation in Active pharmaceu- tical ingredient manufacturing plants, available at: http://apic.cefic.org/pub/ /4CleaningVal9909.pdf (accessed 20 April, 2009) 4. World Health Organization, Supplementary guidelines on good manufacturing practices (GMP): Validation, Working document QAS/03.055/Rev.2, pp. 24–33, http://www.who.int/ /medicines/services/expertcommittees/pharmprep/Validation_QAS_055_Rev2combined.pdf (accessed 20 April, 2009) 5. R. Klinkenberg, B. Streel, A. Ceccato, J. Pharm. Biomed. Anal. 32 (2003) 345 6. T. T. Fazio, A. K. Singh, E. R. M. Kedor-Hackmann, M. I. R. M. Santoro, J. Pharm. Biomed. Anal. 43 (2007) 1495 7. T. Mirza, M. J. Lunn, F. J. Keeley, R. C. George, J. R. Bodenmiller, J. Pharm. Biomed. Anal. 19 (1999) 747 8. D. Belachew, K. M. McErlane, J. Pharm. Sci. 71 (1982) 777 9. M. Song, M. Li, X. Hu, CA 143:139350 [2005:435677] 10. B. Pekić, B. Slavica, Memoir of Matica Srpska for natural Sciences 84 (1993) 73 11. D. M. Popova, T. N. Konyakhina, V. A. Semenova, A. P. Arzamastsev, Farm. Zhurnal (Kiev) 3 (1983) 44 (in Russian) 12. D. Belachew, E. Kwong, K. M. McErlane, J. Chromatogr. 240 (1982) 137 13. M. C. Castle, J. Chromatogr. 115 (1975) 437 14. Q. Zhang, CA 144:94565 [2005:1317285] 15. X. Ouyang, Y. He, CA 142:397882 [2005:236821] 16. D. Ivanović, M. Medenica, D. Radulović, Z. Djugumović, Drug Dev. Ind. Pharm. 21 (1995) 1789 17. T. Hagiwara, K. Nakayama, K. Akiyama, M. Doguchi, CA 111:140376 [1989:540376] 18. T. Hagiwara, K. Akiyama, CA 106:144053 [1987:144053] 19. T. Sugiyama, R. Matsuyama, S. Usui, Y. Katagiri, K. Hirano, Biol. Pharm. Bull. 23 (2000) 274 20. V. S. M. Varma, N. Kapoor, M. Sarkar, R. Panchagnula, J. Chromatogr. B 813 (2004) 347 21. E. Reh, H. Jork, Fres. Zeitsch. Anal. Chem. 318 (1984) 264 22. A. J. Morais, A. R. Zlotecki, E. Sakmar, L. P. Stetson, G. J. Wagner, Res. Comm. Chem. Pathol. Pharmacol. 31 (1981) 285 23. S. Shi, Z. Li, H. Chen, F. Zeng, J. Chromatogr. B 875 (2008) 405 24. C. B. Fernao, K. Wolfgang, A. R. Rubens, B. O. Alaide, Quimica Nova 20 (1997) 481 (in Portuguese) 25. Y. Ikeda, Y. Fujii, M. Yamazaki, J. Nat. Prod. 55 (1992) 748 26. M. J. Nozal, J. L. Bernal, J. J. Jimenez, M. T. Martin, F. J. Diez, J. Chromatogr. A 1024 (2004) 115 27. M. J. Nozal, J. L. Bernal, L. Toribio, M. T. Martin, F. J. Diez, J. Pharm. Biomed. Anal. 30 (2002) 285 Available online at www.shd.org.rs/JSCS/ ___________________________________________________________________________________________________________________________ 2009 Copyright (CC) SCS Author's Personal Copy HPLC DETERMINATION OF DIGOXIN RESIDUES 1153 28. M. J. Nozal, J. L. Bernal, L. Toribio, M. T. Martin, F. J. Diez, J. Chromatogr. A 919 (2001) 87 29. J. Lambropoulos, G. A. Spanos, N. V. Lazaridis, J. Pharm. Biomed. Anal. 23 (2000) 421 30. N. Fakutsu, T. Konse, T. Kawasaki, K. Saito, H. Nakazawa, J. Pharm. Biomed. Anal. 41 (2006) 599 31. M. B. Boca, Z. Apostolides, E. Pretorius, J. Pharm. Biomed. Anal. 37 (2005) 461 32. D. M. Milenović, M. L. Lazić, V. B. Veljković, Z. B. Todorović, Acta Chromatogr. 20 (2008) 183 33. International Conference on Harmonization Q2 (R1), Validation of Analytical Proce- dures, Text and Methodology, http://www.ich.org/LOB/media/MEDIA417.pdf (accessed 20 April, 2009). Available online at www.shd.org.rs/JSCS/ ___________________________________________________________________________________________________________________________ 2009 Copyright (CC) SCS