Determination of Piroxicam and Degradation Products in Drugs by TLC

2009, 69, 351–356

2 _ Małgorzata Starek1,&, Jan Krzek1, Monika Tarsa2, Marek Zylewski 1

2

Department of Inorganic and Analytical Chemistry, Collegium Medicum, Jagiellonian University, 9 Medyczna Str, 30-688 Krako´w, Poland; E-Mail: [email protected] Department of Organic Chemistry, Collegium Medicum, Jagiellonian University, 9 Medyczna Str, 30-688 Krako´w, Poland

Received: 3 June 2008 / Revised: 4 September 2008 / Accepted: 25 September 2008 Online publication: 8 November 2008

Abstract A thin-layer chromatography method with densitometric detection for the determination of piroxicam and its degradation products is described. Separation was on silica gel TLC plates with ethyl acetate/toluene/butylamine (2+2+1, v/v/v) as the mobile phase. Densitometric detection was carried out at 360 nm. The method has high sensitivity, and satisfactory recovery. Based on 1H NMR and LC–MS–MS, it was found that piroxicam decomposes to produce pyridine-2-amine and 2-methyl-2,3-dihydro-4H-1k6,2-benzotiazin-1,1,4-trione.

Keywords Thin-layer chromatography Densitometry Piroxicam Drug analysis

Introduction Piroxicam (4-hydroxy-2-methyl-1,1-dioxoN-(pyridin-2-yl)-2H-1k6,2-benzothiazine3-carboxamide) belongs to a group of anti-inflammatory, analgesic and antipyretic drugs. Its characteristic feature is reversible action to COX and a long half-life (about 50 h), thus enabling it to be administered once a day. The biotransformation of piroxicam applies primarily to the pyridyl ring and the

Full Short Communication DOI: 10.1365/s10337-008-0883-0 0009-5893/09/02

resulting hydroxyl derivatives are coupled with glucuronic acid and are subject to 60% elimination in urine [1]. Piroxicam in pharmaceutical preparation is present in pure form as an ester along with cinnamic acid (Cinnoxicam) or in complexes with b-cyclodextrin (Cycladol) that increase its water solubility and absorption rate, while having no other effect on its clinical efficacy. To determine the concentration of piroxicam in pharmaceutical preparations and

body fluids, spectrophotometric [2–7] and LC methods are mainly used [5, 8–14]. An HPTLC method for quantification of piroxicam in the presence of 2-aminopyridine has been described [8, 15]. TLC-matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS) for the direct analysis of TLC plates have been reported for the determination of piroxicam [16]. Capillary electrophoresis [8, 17], potentiometric methods [6] and flow analysis methods [18] have also been employed stability indicating LC, HPTLC and CE methods for the determination of piroxicam have been described [8, 10, 11, 15]. Simultaneous spectrofluorimetric analysis of piroxicam and pyridine-2-amine has also been reported [4]. To continue our previous studies that proved the suitability of TLC/densitometry for determining impurities in drugs [19–21], the conditions for identification and quantitative analysis of piroxicam and its degradation products in pharmaceutical preparations were established. In addition, the effects of pH, temperature and incubation time on piroxicam stability were investigated and a decomposition pathway is indicated based on the identification of piroxicam degradation products. The presented method is simple, rapid and sensitive for simultaneous determination of piroxi-

Chromatographia 2009, 69, February (No. 3/4) Ó 2008 Vieweg+Teubner | GWV Fachverlage GmbH

351

Fig. 1. An example of densitograms, obtained after piroxicam decomposition in acidic (a) and basic (b) environment (1 = piroxicam; 2 = ZA; 3 = ZB)

cam and its degradation products in pharmaceutical preparations.

Experimental Equipment Densitometer TLC Scanner 3 with Cats 4 software (Camag, Switzerland); Sample applicator Linomat V (Camag, Switzerland); Silica gel aluminium TLC F254 plates No. 1.05554 (E. Merck, Darmstadt, Germany); TLC chamber of 18 9 9 9 18 cm in size (Sigma-Aldrich); Incubator ECOCELL55—BMT (Brno, Czech Republic); NMR Spectrometer Mercury VX 300 MHz (Varian, USA); Mass spectrometer API 2000 (Applied Biosystems MDX SCIEX, Concorde, Ontario, Canada); HPLC system with

352

Xterra column (Waters) and DAD detector (Agilent Technologies, Waldbronn, Germany).

Reagents and Chemicals Standard substances: 4-hydroxy-2-methyl-1,1-dioxo-N-(pyridin-2-yl)-2H-1k6, 2-benzothiazine-3-carboxamide (piroxicam; F.I.S., Italy); pyridine-2-amine (Roanal, Budapest, Hungary). Reagents: acetone, ethyl acetate, toluene, butylamine (POCH Gliwice, Poland).

s.G1F00 (APOTEX, Ontario, Canada); Feldene—injections cont. 20 mg of piroxicam s.612041202 (Pfizer, Orsay, France); Brexin—tablets cont. 191.2 mg of complex piroxicam—b-cyclodextrin (corr. 20 mg of piroxicam) s.G00452 (Pierre Fabre Med. Prod., Boulogne, Italy); Cycladol—tablets cont. 191.2 mg of complex piroxicam—b-cyclodextrin (corr. 20 mg of piroxicam) s.033689 (Chiesi, Courbevoie, France).

Solutions Preparations Used for Analysis The following drugs, chosen at random, were analyzed: Apo-Piroxicam— capsules cont. 10 mg of piroxicam

Standard solutions: at a concentration of 0.001% (w/v) for piroxicam and pyridine-2-amine were prepared by dissolving an appropriate amount of the substances in acetone.

Chromatographia 2009, 69, February (No. 3/4)

Full Short Communication

For samples from Apo-Piroxicam and Feldene: solutions were prepared by grinding the powder in the capsules weighed to 0.1 mg, corresponding approximately to 10 mg of piroxicam, and dissolving in acetone, shaken for 15 min, filtered and filled up to the volume of 100.0 mL. For Brexin and Cycladol, weighed amounts of the powdered drugs, each containing about 10 mg of piroxicam were dissolved in 50.0 mL 0.5 mol L-1 HCl and heated in a water bath at 60 °C for 1 h, then filled up with acetone to 100.0 mL. For analysis, the solutions were diluted with acetone at the ratio of 1:10.

Table 1. Kinetic and thermodynamic parameters describing degradation process of piroxicam

Results and Discussion

k Stability constant (h-1), t0.1 time, concentration concentration will decrease about 10% (h), t0.5 time, concentration will decrease about 50% (h), Ea the energy of activation (J mol-1 K)

Chromatographic Analysis The conditions for the analysis of piroxicam were established by experimental selection of the appropriate stationary and mobile phases. Experiments were performed on F254 TLC plates with ethyl acetate + toluene + butylamine (2+2+1, v/v/v) as the mobile phase. From 1 to 100 lL of solutions were applied with an applicator on the plates in the form of 1 cm bands. Plates were developed to different distances in a tank previously saturated with mobile phase vapor for 10 min at ca. 20 °C. Good separation and well developed peaks were obtained in approximately 20 min, by developing chromatograms over a distance of 10 cm. After drying at room temperature, peak area and absorption spectra were recorded directly from the plates, by UV densitometric detection in the wavelength range from 200 to 400 nm. On chromatograms of the solutions recorded after extracting piroxicam with acetone the presence of the main spot originated from piroxicam extracted with acetone was observed at RF & 0.53, while for drugs extracted with 0.5 mol L-1 hydrochloric acid by heating the solution in a water bath at 60 °C for 1 h, an additional spot of RF & 0.62 (ZA) was also present. Prolonged heating of the solutions led to a decrease in the piroxicam spot and an

Full Short Communication

Temperature

HCl or NaOH concentration/kinetic parameters

60 °C

1 mol L-1 HCl k = 8.32 9 10-4 t0.1 = 126.56 t0.5 = 832.93 3 mol L-1 HCl k = 2.91 9 10-3 t0.1 = 36.19 t0.5 = 238.14 0.5 mol L-1 HCl k = 3.35 9 10-1 t0.1 = 0.31 t0.5 = 2.07 1 mol L-1 HCl k = 2.26 9 10-1 t0.1 = 0.47 t0.5 = 3.07

120 °C

Ea = 1.02 9 105 (HCl) Ea = 9.58 9 104 (NaOH)

increasing ZA spot and occurrence of a second additional spot of RF & 0.82 (ZB). The spots differed also in absorption spectra recorded directly from the plates. The recorded spectra show characteristic absorbance maxima for each component at 293 and 360 nm for piroxicam, 233 and 297 nm for ZA, and 220 and 375 nm for ZB. These results lead to the conclusion that, in an acid solution and at increased temperature, piroxicam decomposes into two products. Such findings were used for further investigations to determine the stability of piroxicam in acidic and basic solutions depending on pH, temperature and incubation time.

Examination of Piroxicam Stability in Solution The effects of pH, temperature and incubation time on stability of piroxicam in solutions were investigated. For this purpose, weighed amounts of preparations containing approximately 10 mg of piroxicam were dissolved in 5 mL of hydrochloric acid or sodium hydroxide solution. The solutions were incubated for specified lengths of time at tempera-

1 mol L-1 NaOH k = 4.49 9 10-4 t0.1 = 234.52 t0.5 = 1,543.43 3 mol L-1 NaOH k = 1.25 9 10-3 t0.1 = 84.24 t0.5 = 554.40 0.5 mol L-1 NaOH k = 3.40 9 10-2 t0.1 = 3.10 t0.5 = 20.38 1 mol L-1 NaOH k = 8.79 9 10-2 t0.1 = 1.20 t0.5 = 7.88

tures of 60 and 120 °C. The samples were diluted with acetone (1+1, v/v) for quantitative analysis. In addition to the piroxicam peak, some additional peaks were observed in chromatograms depending on test conditions. It seems from the RF values obtained and recorded absorption spectra that two impurities (RF & 0.62, RF & 0.82), ZA and ZB, can be present in an acidic environment. In a basic environment only one impurity occurred (RF & 0.62), probably ZA (Fig. 1). The results obtained were used for kinetic and thermodynamic evaluation of the piroxicam decomposition process, by determining the reaction rate constants k, half-life t0,5 and the time t0,1 in which the concentration of piroxicam is reduced by 10%, as well as activation energy, according to the kinetics of a first order reaction [22]. The results obtained are presented in Table 1.

Analysis of Degradation Products To identify the piroxicam degradation products, 1H NMR and LC–MS–MS analyses were carried out. The piroxicam solutions after hydrolysis in 0.5 mol L-1 HCl were separated on a Chromatotron

Chromatographia 2009, 69, February (No. 3/4)

353

Fig. 2. A 1H NMR and LC–MS–MS spectras for substance ZA (a) and ZB (b) in tested sample

plate. After separation, the solvent was evaporated and the products were identified by their 1H NMR spectra.

354

The piroxicam spectrum shows signals coming from protons of the benzene ring (r 7.9 ppm, r 8.1 ppm), pyridyl ring

(r 8.4 ppm, r 8.2 ppm, r 7.25 ppm) and a hydroxyl group proton of the enol hydroxyl group (r 13.6 ppm), NH amide group proton (r 9.5 ppm) and a methyl group singlet (r 3.05 ppm). In the spectrum of ZA the presence of characteristic signal groups were detected for: CH group protons— (r 6.58–6.60 ppm, r 6.60–6.61 ppm, r 7.40–7.42 ppm, r 8.05 ppm), and NH2 group protons (r 4.51 ppm). The spectra of the samples tested show the presence of signals with chemical shifts corresponding to pyridine-2-amine (Fig. 2). The product ZB spectrum shows no OH and NH group protons. The benzene ring (r 8.1–7.8 ppm) and methyl group protons (r 2.9 ppm) are present. There is also an additional signal coming from two protons with r 4.45 ppm. The disappearance of signals resulting from the pyridine ring indicates hydrolysis of the amide bond and pyridine-2-amine removal. The absence of OH and COOH group signals and appearance of the new signal can be explained by products of further hydrolysis such as decarboxylation of the acid being formed and then tautomeric transformation of an OH enol group into a ketone one, thus causing a new CH2 group to be formed. Thus, another hydrolysis product may be the following compound: 2-methyl-2,3dihydro-4H-1k6,2-benzotiazin-1,1,4-trione (Fig. 2). To analyze the degradation products by LC–MS–MS, a triple quadrupole mass analyzer connected with to an LC system was used. The analysis was carried out by using acetonitrile + water (50 + 50, v/v) with addition of 10 lL L-1 formic acid, at a flow rate of 600 lL min-1, by using positive ionization and electrospray ion sources. The presence of the molecular ion at m/z = 95 coming from pyridine-2amine, m/z 94 (82%) M+ [C5H6N2]+; m/z 78 (100%) [C5H4N]+; m/z 51 (2%) [C4H3]+ and the m/z = 121 peak originated from an intermediate piroxicam decomposition product were recorded. The spectrum also shows an m/z = 211 peak coming from a product identified as ZB (2-methyl-2,3-dihydro-4H-1k6, 2-benzotiazin-1,1,4-trione) (Fig. 2).

Chromatographia 2009, 69, February (No. 3/4)

Full Short Communication

Table 2. The results of piroxicam determination in pharmaceutical preparations with statistical analysis Preparation

Declared concentration

Determined concentration

Apo-Piroxicam (APOTEX)

10 mg caps.-1

9.84 9.90 10.24 19.05 19.41 20.63 20.04 20.70 19.30 19.62 20.81 19.62

Feldene (Pfizer)

Brexin (P. Fabre Med.)

Cycladol (Chiesi)

-1

20 mg ini.

-1

20 mg tab.

20 mg tab.-1

Statistical analysis (n = 6)

9.74 9.97 9.74 19.92 20.03 19.68 19.24 20.08 20.02 20.82 20.11 20.62

xm = 9.91 S = 0.19 l = xm ± 0.21 xm = 19.79 S = 0.54 l = xm ± 0.57 xm = 19.90 S = 0.55 l = xm ± 0.57 xm = 20.27 S = 0.56 l = xm ± 0.59

Sxm ¼ 0:08 RSD = 1.92 Sxm ¼ 0:22 RSD = 2.73 Sxm ¼ 0:22 RSD = 2.76 Sxm ¼ 0:23 RSD = 2.76

xm Arithmetic mean, S standard deviation, Sxm standard deviation for arithmetic mean, l confidence interval at 95% probability, RSD relative standard deviation (%)

Quantitative Analysis for Piroxicam in Drugs Ten microliters of standard and preparation solutions were applied with an applicator onto the plates 9 9 10 cm cut from 20 9 20 cm before use, at the distance of 1 cm from the edge and 1 cm from the plate bottom, in the form of bands of 1 cm in width to determine piroxicam. One hundred microliters each was used for determining the impurities. Chromatograms were developed by using the mobile phase: ethyl acetate + toluene + butylamine (2+2+1, v/v/v). After developing, plates were dried at room temperature. Then, they were scanned and peak areas were recorded at 360 and 297 nm. The piroxicam concentration in preparations under investigation was computed by comparing the peak areas for standard and sample solutions. For each determination three measurements were made and the mean was considered to be the final result. The method was validated in accordance with ICH guidelines [23]. The specificity of the method was ascertained by analysis of standards and drug samples. The spots were identified by comparison of the RF values and spectra of separated spots with those obtained from the standards. For extracted solutions used for determination purposes, no additional spots were founds. The linearity was expressed as a relationship between peak area and

Full Short Communication

analyte concentration within a specified measuring range. The regression plot, its regression equation and the correlation coefficients are indicative of linearity. To determine linearity a series of six solutions (three bands per concentration) at concentration ranging from 2 to 110 lg mL-1 for piroxicam and from 10 to 240 lg mL-1 for pyridin-2-amine were prepared. Limits of detection (LOD) and quantitation (LOQ) were determined on the basic of the standard deviation and slope of the straight lines obtained from the equations: LOD = 3.3 SD/a and LOQ = 10 SD/a, where SD is the standard deviation of the response and a is the slope of the curve. The LOD and LOQ were found to be 0.07 and 0.20 lg per spot for piroxicam and 0.10 and 0.29 lg per spot for pyridine-2-amine, respectively. The accuracy of the method was defined as % recovery of analyte added from 80 to 120% of the substance in relation to the content of the pharmaceutical preparations. The recovery values obtained for piroxicam and pyridine-2-amine were 99.33 and 101.95%, respectively. The repeatability of sample application and peak area measurement was determined by analysis of six replicates of the same sample. The precision was expressed as %RSD = 0.95 for piroxicam and 2.07 for pyridine-2-amine. The degree of consistency of the results obtained for the same analyte

sample was checked by an analyst who made the analysis over a 1-month interval. In assessment all results within two standard deviations (S) from the mean (xm) were taken into account. The percentage concentration of piroxicam was determined: analysis I: xm = 19.84%; S = 0.27; t95% = xm ± 0.43; RSD = 1.36%; analysis II: xm = 19.93%; S = 0.25; t95% = xm ± 0.39; RSD = 1.25%. As a result a procedure for determining piroxicam and pyridin-2-amine were developed and tested for selected drugs. The results obtained for piroxicam are listed in Table 2. No results for impurities are presented as no impurities were found in the examined preparations.

Conclusions A new method for identification and quantitative determination of piroxicam and pyridin-2-amine in drugs and stability of piroxicam in acidic and basic environments has been developed. By using TLC F254 silica gel coated plates as a stationary phase and ethyl acetate + toluene + butylamine (2+2+1, v/v/v) as the mobile phase, good separation of the constituents under investigation was obtained. The results of kinetic and thermodynamic investigations indicate that the decomposition of piroxicam is a first order reaction and the high values of

Chromatographia 2009, 69, February (No. 3/4)

355

activation energy, Ea, suggest that piroxicam is a relatively stable product. Piroxicam demonstrates higher susceptibility to degradation in acidic than basic environment. The products of decomposition are pyridin-2-amine and 2-methyl-2,3-dihydro-4H-1k6,2-benzotiazin-1,1,4-trione. The method developed meets the acceptance criteria related to validation with specificity for piroxicam, a good linear relationship (r & 0.998), low LOD and LOQ limits, high recovery and good precision for both pyridin-2-amine and piroxicam. The results of determination for piroxicam concentration in preparations do not differ from those declared by the manufacturers.

References 1. Zejc A, Gorczyca M (2002) Chemia leko´w. PZWL, Warszawa

356

2. Abdollahi H, Sororaddin MH, Naseri A (2006) Anal Sci 22(2):263–267 3. Amin AS (2002) J Pharm Biomed Anal 29(4):729–736 4. Barary MH, Abdel-Hay MH, Sabry SM, Belal TS (2004) J Pharm Biomed Anal 34:221–226 5. Basan H, Goger NG, Ertas N, Orbey MT (2001) J Pharm Biomed Anal 26(2):171– 178 6. El-Ries MA, Mohamed G, Khalil S, El-Shall M (2003) Chem Pharm Bull 51(1):6–10 7. Escandar GM, Bystol AJ, Campiglia AD (2002) Anal Chim Acta 466:275–283 8. Bartsch H, Eiper A, Kopelent-Franf H (1999) J Pharm Biomed Anal 20:531–541 9. Council of Europe (2004) European Pharmacopoeia, 5th edn. Council of Europe, Strasbourg, France 10. Boneschans B, Wessels A, Van Staden J, Zovko M, Zorc B, Bergh J (2003) Drug Dev Ind Pharm 29(2):155–160 11. Gaudiano MC, Valvo L, Bertocchi P, Manna L (2003) J Pharm Biomed Anal 32:151–158 12. Ji HZ, Lee H, Kim ZH, Jeong DW, Lee HS (2005) J Chrom B Analyt Technol Biomed Life Sci 826(1–2):214–219

13. Rozou S, Voulgari A, Antoniadou-Vyza E (2004) Eur J Pharm Sci 21(5):661–669 14. Sultan M, Stecher G, Stoggl WM, Bakry R, Zaborski P, Huck CW, El Kousy Nagla M, Bonn GK (2005) Curr Med Chem 12(5):573–588 15. Puthli SP, Vavia PR (2000) J Pharm Biomed Anal 22:673–677 16. Crecelius A, Clench MR, Richards DS, Parr V (2004) J Pharm Biomed Anal 35:31–39 17. Chen ZL, Wu SM (2005) Anal Bioanal Chem 381(4):907–912 18. Al-Kindy SMZ, Al-Wishahi A, Suliman FEO (2004) Talanta 64(5):1343–1350 19. Krzek J, Starek M (2002) J Pharm Biomed Anal 28(2):227–243 20. Krzek J, Starek M (2004) Biomed Chromatogr 18:589–599 21. Krzek J, Starek M (2001) J AOAC Int 84(6):1703–1707 22. Pawełczyk E, Hermann T (1982) Podstawy trwałos´ ci leko´w. PZWL, Warszawa 23. ICH Steering Committee (2005) Validation of analytical procedures (Q2)

Chromatographia 2009, 69, February (No. 3/4)

Full Short Communication