Sustained Release Micro Spheres Of Metoclopramide Using

J. MICROENCAPSULATION ,

2000,

VOL.

17,

NO.

4, 425± 435

Sustained release microspheres of metoclopramide using poly(D,L-lactide-co-glycolide) copolymers S. A. ELKHESHEN{* and M. A. RADWAN{ { Department of Pharmaceutics, { Department of Clinical Pharmacy, College of Pharmacy, King Saud University, P.O. Box 22452, Riyadh 11495, Saudi Arabia (Received 23 February 1999; revised 6 October 1999; accepted 20 October 1999) Metoclopramide was encapsulated with poly(D,L-lactide co glycolide) copolymers of diå erent molecular weights using the emulsi® cation/solvent evaporation technique. These polymers included poly(D,L-lactide-co-glycolide) 50:50 with inherent viscosity (i.v.) 0.2, and average molecular weight 8000, poly(D,Llactide-co-glycolide) 50:50 with i.v. 0.8 and average molecular weight 98 000 and poly(D,L-lactide-co-glycolide) 85:15 with i.v. 1.4 and average molecular weight 220 000. The eå ect of the polymers’ molecular weights as well as the polymer-to-drug ratios on the yield, the particle size distribution, and the drug content of the microspheres was investigated. The release rate of the drug was studied for 96 h in a phosphate buå er of pH 7.4. The study also investigated the eå ect of the new poly(lactide-co-glycolide)-H series on the characteristics of the prepared microspheres. Data revealed that a higher yield was obtained with polymers of lower molecular weights. A lower yield was also obtained with increasing the drug-to-polymer ratios for all the investigated polymers. The drug content of the microspheres was lower than expected, ranging from 49± 85%, which suggested a chemical interaction between the drug and the polymers, as proved by diå erential scanning calorimetry (DSC) and infra red (IR) studies. A higher interaction was obtained with the H-series of the copolymers. The release of the drug mainly followed zero order kinetics on increasing either the polymers’ molecular weights or the polymer-to-drug ratios. Diå usion kinetics was observed only with those batches prepared with low polymer-to-drug ratios. The release rate was a function of both the polymers’ molecular weights and the drug-to-polymer ratios. Keywords: Metoclopramide, microspheres, sustained release, poly(lactide-coglycolide).

Introduction Metoclopramide is widely used for the treatment of some forms of nausea and vomiting, such as those associated with migraine, cancer therapy, or those following surgery (Shaughnessy 1985). Because of its relatively short half life (4 h), it needs to be administered three or four times a day in a dose of 10± 15 mg (Khidr et al. 1995). It is in the ® eld of treating nausea and vomiting associated with cancer therapy that the preparation of injectable controlled release microspheres for intramuscular administration becomes advantageous. These microspheres can

* To whom correspondence should be addressed. e-mail: [email protected] Journal of Microencapsulation ISSN 0265± 2048 print/ISSN 1464± 5246 online # 2000 Taylor & Francis Ltd http://www.tandf.co.uk/journals

426

S. A. Elkheshen and M. A. Radwan

be designed to release the drug continuously and eå ectively over a period of a week or more, which is considered more convenient for patients. Polymers of lactic and glycolic acids and their copolymers are biocompatible and safe for parenteral (subcutaneous and intramuscular) administration, since the biodegradable polymers do not need to be removed from the body (Conti et al. 1992). They have been widely used for the encapsulation of various drugs for diå erent purposes. Recently, these polymers were used for the encapsulation of peptides and proteins (Sah et al. 1995, Uchida et al. 1996, Jeyanthi et al. 1997, McGee et al. 1997), stabilization of hormones (Johnson et al. 1997), and vaccines (Chang and Gupta 1996). The most important application of these polymers is controlling drug release (Labhasetwar et al. 1994, El Corre et al. 1997, Yang and Cleland 1997). Depending on the composition and/or the molecular weight of these polymers, they slowly hydrate in the body and release the drug (Deasy et al. 1989). Increasing the molecular weight or changing the composition of the polymers also aå ects their rate of biodegradation. The L-lactic acid polymers are faster in biodegradation than the D,L series. Poly-lactic acid polymers are faster than poly-lactide and the lactic acid series are faster than the glycolic acids. The co-polymers of lactic and glycolic acids are of intermediate degradation rate between the two pure polymers (Deasy et al. 1989). The present work investigates the eå ect of the molecular weight of poly(lactide-co-glycolide) copolymers and the drug-to-polymer ratio on the microsphere characteristics and the drug release kinetics and rate. The study also investigates the eå ect of the new poly(lactide-coglycolide)-H series, which was developed with more free carboxylic end groups to increase the water uptake and the degradation of the polymer, on the characteristics of the prepared microcapsules. Experimental Materials Metoclopramide was kindly supplied by the Saudi Pharmaceutical Industrial and Medical Application (SPIMACO, Rhiyadh, Saudi Arabia). Poly(D,L-lactideco-glycolide) 50:50, i.v. of 0.2, average molecular weight of 8000 (Resomer RG 502 H), poly(D,L-lactide-co-glycolide) 50:50, i.v. of 0.8, average molecular weight of 98 000 (Resomer RG 506), and poly(D,L-lactide-co-glycolide) 85:15, i.v. of 1.4, average molecular weight of 220 000 (Resomer RG 858), were a generous gift from Boehringer Ingelheim (Ingelheim, Germany). Light liquid paraæ n and n-hexane, from BDH (Poole, UK), Acetonitrile, from Fisher Chemical (Fair Lawn, New Jersey, USA), Span 80, from Fluka Chemie (Buchs, Switzerland) and Arlacel 186; from Atlas Chemicals (USA), were used as received. Preparation of microspheres The emulsion solvent evaporation method was applied for the preparation of metoclopramide (Meto) microspheres using diå erent poly(D,L-lactide-co-glycolide) copolymers of diå erent i.v. including, 0.2 (PLGA-I), 0.8 (PLGA-II) and 1.4 (PLGA-III). The dispersed phase was prepared by dissolving 100 mg of the drug in the polymer solution. The solution of the polymer was prepared by dissolving diå erent amounts of the polymer, based on drug-to-polymer ratios of 1:2, 1:4, 1:6, 1:8 and 1:10, in acetonitrile heated to 558C. The concentration of the polymer in

Sustained release microspheres

427

acetonitrile was kept constant (10%) to avoid the eå ect of changing the viscosity of the polymer solution on the droplet size of the dispersed phase and consequently on the particle size of the formed microspheres. The only eå ect will be due to increasing the drug-to-polymer ratio. The dispersed phase was added dropwise to liquid paraæ n (200 ml) containing 1% Span 80 heated to 558C and stirred at a constant stirring rate of 500 rpm, using a mechanical stirrer (Fisher Scienti® c, USA, connected to Fisher Dyna-Mix for speed adjustment). Stirring was continued at 558C for 45 min to aid the evaporation of the solvent under ambient pressure. Microspheres were then collected by decantation and/or centrifugation and washed twice with n-hexane containing Arlacel to remove the mineral oil, and twice with n-hexane to remove traces of surfactant. Microspheres were then ® ltered, air dried and kept refrigerated in a dessiccator until the time of evaluation. Evaluation of microspheres Microspheres’ yield. Microspheres were sieved to remove polymeric sheets and microsphere aggregates. The yield was calculated as a percentage of the original amount of polymer and drug. Data are the average of three batches of each formulation. Drug content. An amount of the microspheres, equivalent to 5 mg of the drug, was dissolved in 10 ml aceonitrile in a volumetric ¯ ask. The solution was measured spectrophotometrically at 272 nm after suitable dilution. The amount of the drug was determined using a standard curve, developed in acetonitrile (a ˆ ¡0:0035, b ˆ 0:04706, R2 ˆ 0:9998). Polymers did not interfere with the absorbance of the drug at the speci® ed wavelength. The drug content was calculated as a percentage of the originally encapsulated amount of Meto and expressed as the mean of six experiments (two replicates for each of the three batches). Particle size. Particle size determination was done by taking microscopic photographs of three diå erent mounts of each formulation together with a scale. A hundred particles from each batch were examined for their particle size. The particle size range as well as the average particle size were calculated from a frequency distribution curve. Morphological examination. Scanning electron microscopy (Jeol Scanning Electron Microscope, T300, Tokyo, Japan) was used to study the morphology and surface characteristics of the microspheres. Drug release. The release of the drug from the microspheres was studied in 500 ml of Sorensen phosphate buå er of pH 7.4 using USP apparatus, paddle system (Erweka Apparatebau, Heusenstamm, Germany). An amount of the microspheres equivalent to 15 mg drug content (free drug) was added to the medium at time zero. The stirring rate was maintained at 100 rpm and the temperature at 378C. Samples of 3 ml were removed through a membrane ® lter at diå erent time intervals up to 96 h, measured spectrophotometrically at 272 nm (PU8800 UV/VIS Spectrophotometer, Pye Unicam, Cambridge, UK) and returned to the medium. The dissolution apparatus cups were covered ® rmly to

428

S. A. Elkheshen and M. A. Radwan

guard against the evaporation of the dissolution medium. The released amount of the drug was determined using a standard curve constructed in the same medium (a ˆ 0:025, b ˆ 0:0408, R2 ˆ 0:999). The release data were ® tted to the following equation developed by Peppas (1985) to determine the release kinetics. Mt ² Ktn M1

…1†

where the Mt =M1 is the fractional drug released, K is the kinetic constant, t is the release time and n is the release exponent characteristic for the drug transport mechanism. A value of n close to 0.5 has been shown to correspond to Fickian release, a value very close to 1 indicates a zero order release pro® le, while a value of n higher than 0.5 and lower than 1 indicates a non-Fickian release behaviour (Peppas 1985, Vigoreaux and Ghaly 1994). A value of n less than 0.5 indicates that the release of the drug is highly dependent on the amount of the drug remaining in the microspheres (Rizk et al. 1994). The release constant was calculated from the equation that best ® t the release data and, consequently, the time for 50% of the drug to be released was calculated (t50%). Results and discussion The morphological studies (® gures 1 and 2) revealed that, with the exception of those batches prepared with a drug-to-polymer ratio of 1:2, microspheres were spherical, with a smooth surface and no adsorbed drug crystals. The eå ect of the polymer molecular weights and the drug-to-polymer ratio on the microsphere yield, the drug content, the microsphere particle size, and kinetics of drug release were investigated. Microsphere yield A higher yield (table 1) was observed with microspheres prepared with PLGA-I followed by PLGA-II then PLGA-III. The lower yield observed with polymers of higher i.v. is due to their higher tackiness and, consequently, liability for aggregation, forming polymeric sheets and/or sticking to the wall of the container or the surface of the stirrer shaft and blade (Elkheshen 1996). A lower yield was obtained with increasing the drug-to-polymer ratio for all the investigated polymers, although the high drug content was expected to act as a detacki® er Table 1.

The yield of microspheres (SD) prepared with diå erent polymers and diå erent drug-to-polymer ratios. Polymers

Drugs-topolymer ratio

PLGA-I

PLGA-II

1:2 1:4 1:6 1:8 1:10

80.3 83.0 85.2 88.3 97.5

71.0 84.1 85.4 86.7 92.4

(1.6) (3.7) (2.4) (1.9) (2.1)

(1.4) (0.8) (0.5) (0.8) (0.3)

PLGA-III 63.4 78.3 81.7 82.5 87.1

(2.3) (1.3) (1.0) (0.7) (0.7)

Sustained release microspheres

429

Figure 1. Electron scanning micrograph of microspheres prepared with PLGA-I and drug-to-polymer ratio 1:4, showing smooth surface with no adsorbed drug crystals.

Figure 2. Electron scanning micrograph of microspheres prepared with PLGA-I and drug-to-polymer ratio 1:10, showing much smaller particle size.

430

S. A. Elkheshen and M. A. Radwan

Table 2. The particle size distribution of microspheres prepared with diå erent polymers and diå erent drug-to-polymer ratios. Polymers PLGA-I

PLGA-II

PLGA-III

Drugs-to- Particle size Peak of the Particle size Peak of the Particle size Peak of the polymer range distribution range distribution range distribution ratio (mm) (mm) (mm) (mm) (mm) (mm) 1:2 1:4 1:6 1:8 1:10

Table 3.

300± 100± 60± 50± 30±

400 200 125 125 100

400 150 100 60 50

100± 100± 75± 75± 75±

250 225 200 200 200

125 125 125 125 100

100± 100± 100± 100± 60±

400 300 300 225 200

250 150 175 150 125

Per cent drug content (SD) of microspheres prepared with diå erent polymers and diå erent drug-to-polymer ratios. Polymers

Drugs-topolymer ratio

PLGA-I

PLGA-II

1:2 1:4 1:6 1:8 1:10

63.6 61.0 58.5 58.9 48.9

64.5 63.4 60.2 60.4 55.6

(1.4) (1.9) (2.2) (2.8) (0.4)

(2.8) (2.0) (0.6) (1.4) (2.8)

PLGA-III 85.2 72.5 65.8 61.3 58.6

(6.2) (2.4) (2.9) (2.5) (3.2)

and prevent aggregation of microsphere to larger or deformed ones. As the concentration of the polymer in the dispersed phase was kept constant, the volume of the dispersed phase was variable, decreasing with increasing the drug-topolymer ratio. The high concentration of the drug in the dispersed phase at the lower polymer ratios aå ected the size of the dispersed droplets and their shapes. With a drug-to-polymer ratio of 1:2, the volume of the dispersed phase was not suæ cient to completely dissolve the whole amount of the drug, and microcapsules were abnormally large in size (table 2), irregular in shape, with drug crystals adsorbed on the surface. On the contrary, at lower drug-to-polymer ratios, the volume of the dispersed phase was suæ cient to dissolve the drug and the formed microcapsules were regular and small in size (® gures 1 and 2). The drug content The drug content was unexpectedly low (table 3) compared to a previous ® nding by Elkheshen (1996) with the encapsulation of nicotinic acid as a watersoluble drug in poly(lactide-co-glycolide) using the same system. Under these conditions, the loss of the water-soluble drugs to the dispersion medium (liquid paraæ n) is unexpected. The diå erence between the two conditions was the basicity of Meto and its solubility in acetonitrile, while nicotinic acid was insoluble. The solubility of both Meto and the polymers in the dispersion medium during the microsphere preparation allowed a chance for interaction between the drug and

Sustained release microspheres

431

the polymers. The presence of an amino group on the drug molecule and the presence of free carboxylic groups on the polymers suggests the possibility of hydrogen bonding and/or salt formation if there is no steric hindrance to suppress this interaction. Possible interaction products may have the following structures: (a) Hydrogen bonding;

(b) Salt formation:

Further investigation of the possibility of interaction was done by conducting DSC (DSC-4, Perkin Elmer, Norwaki, Connecticut, USA) and IR Spectroscopy (783-IR, Perkin Elmer, Oak Brook, Illinois, USA) on the drug, the polymer, their physical mixture and the encapsulated drug. The DSC results revealed that the drug showed an endothermic peak at 182.168C. The same peak was obtained with the physical mixture but it disappeared with the encapsulated drug. Assessment of the IR charts proved the possibility of salt formation between the basic amino group on the drug and the acidic carboxylic group on the polymer. This was obvious from the disappearance of the stretching of the amino group at 3400 cm ¡1 from the spectrum of the encapsulated drug but not from that of the drug± polymer physical mixture. This ® nding indicated that the interaction took place during the encapsulation process, while both the drug and the polymer were soluble in acetonitrile (the dispersed phase). Obtaining a drug content of 49± 85%of the originally encapsulated amount after dissolving the microspheres in acetonitrile proved that not all the interaction products are ® rm and some of them can dissociate, releasing the free drug. This was also obvious from the release study. The lower drug content observed with the PLGA-I (one of the H-series) was due to the fact that the H series of PLGA polymers were developed with hydrophilic adjusted Homo and copolymer with more free carboxylic end groups to fasten the degradation and to increase the water uptake than the standard polymer series. The presence of more free carboxylic end groups, which are not stirrically hindered, allowed for more chemical interaction and lower free drug content of the microspheres. Moreover, increasing the polymer-to-drug ratio decreased the drug content of the microspheres. This ® nding may be explained on the basis that the drug, the polymer and the drug± polymer complex are present in equilibrium in the dispersed phase (acetonitrile). This equilibrium was shifted towards more complex formation by increasing the polymer-to-drug ratio, leading to a decrease in the drug content of the microspheres.

432

S. A. Elkheshen and M. A. Radwan

Figure 3.

The release of metoclopramide from microspheres prepared with diå erent drugto-polymer ratios using PLGA-II.

Table 4.

The time for 50%drug relase (t50%) from microcapsules prepared with diå erent polymers and drug-to-polymer ratios.

Drug-topolymer ratio

t50% (h) PLGA-I

1:2 1:4 1:6 1:8 1:10 c

00.6 21.7 28.9 26.9 23.7

PLGA-II

PLGA-IIIc

6.4 13.5 25.9a 30.7a 41.0b

34.4 37.5 41.9 83.6 92.7

a t50% was calculated after 5 h lag period; b t50% was calculated after 7 h lag period; and t50 % was calculated after 24 h lag period.

The drug release kinetics The release of the drug from microspheres prepared with diå erent drug-topolymer ratios using PLGA-I is presented in ® gure 3. Only the formulation prepared with a drug-to-polymer ratio of 1:2 (formulation 1) showed a burst eå ect due to the adsorbed drug crystals. All the other drug-to-polymer ratios (1:4± 1:10) showed a steady release with no burst eå ect. Analysing the data using equation (1) revealed that, with the exception of formulation 1, the release of the drug followed nearly zero order kinetics with an n equal to 1 § 0:0928. The release rate of the drug was decreasing, as observed from the t50% (table 4), as the polymer-to-drug ratio was increasing up to 1:6. Formulations prepared with drug-to-polymer ratios of 1:8 and 1:10 showed higher release rates, which may be attributed to the decrease in the particle size of the microcapsules (table 2) and, consequently, the increase in the surface area in contact with the dissolution medium. This is beside

Sustained release microspheres

433

Figure 4.

The release of metoclopramide from microspheres prepared with diå erent drugto-polymer ratios using PLGA-II.

Figure 5.

The release of metoclopramide from microspheres prepared with diå erent drugto-polymer ratios using PLGA-III.

the fact that PLGA-I, being one of the H-series, has a high ability for hydration and water uptake. The increase in the polymer-to-drug ratio did not signi® cantly decrease the release rate of the drug after complete encapsulation. The release of the drug from microcapsules prepared with PLGA-II is shown in ® gure 4. The release of the drug followed diå usion kinetics from the ® rst two formulations (1:2 and 1:4 drug-to-polymer). The release of the drug (for up to 30 h) from the other three formulations with higher polymer content followed zero

434

S. A. Elkheshen and M. A. Radwan

order kinetics after a lag period of 5± 7 h, with an n equal to 1:0 § 0:125. Increasing the ratio of the polymer to the drug gradually decreased the rate of release, as re¯ ected by the increase in t50% (table 4). The release from microcapsules prepared with PLGA-III is shown in ® gure 5. The release from the microcapsules prepared with this polymer at all ratios showed a lag period of almost 24 h before the drug was steadily released. The extent of release did not exceed 40%after 96 h from formulations 4 and 5 (1:8 and 1:10 drugto-polymer ratios) and exceeded 90% for lower polymer-to-drug ratios. The kinetic analysis of the release data proved that the release of the drug followed zero order kinetics after ignoring the lag period from all the batches. References CHANG , A., and GUPTA, R. K., 1996, Stabilization of tetanus toxoid in poly(DL-lacticco-glycolic acid) microspheres for the controlled release of antigen. Journal of Pharmaceutical Science, 85, 129± 132. CONTI, B., PAVANETTO, F., and GENTA, I., 1992, Use of polylactic acid for the preparation of microencapsulate drug delivery systems. Journal of Microencapsulation, 9, 153± 166. DEASY, P. B., FINAN, M. P., and MEEGAN, M. J., 1989, Preparation and characterization of lactic/glycolic acid polymers and copolymers. Journal of Microencapsulation, 6, 369± 378. EL CORRE, P., RYTTING, J. H., GAJAN, V., CHEVANNE , F., and EL VERGE, R., 1997, In vitro controlled release kinetics of local anesthetics from poly(D,L-lactide and poly(lactideco-glycolide) microspheres. Journal of Microencapsulation, 14, 243± 255. ELKHESHEN, S., 1996, Simplex lattice design for the optimization of the microencapsulation of a water soluble drug using poly(lactic acid) and poly(lactide-co-glycolide)copolymer. Journal of Microencapsulation, 13, 447± 462. JEYANTHI, R., MEHTA, R. C., THANOO, B. C., and DELUCA, P. P., 1997, Eå ect of processing parameters on the properties of peptide-containing PLGA microspheres. Journal of Microencapsulation, 14, 163± 174. JOHNSON, O. L., JAWOROWICZ, W., CLELAND , J. L., BAILEY, L., CHARNIS, M., DUENAS, E., WU, C., SHEPARD, D., MAGIL, S., LAST, T., JONES, A. J. S., and PUTNEY, S. D., 1997, The stabilization and encapsulation of human growth hormone into biodegradable microspheres. Pharmaceutical Research, 17, 730± 735. KHIDR, S. H., NIAZY, E. M., and EL-SAYED, Y. M., 1995, Preparation and in-vitro evaluation of sustained-release metoclopramide hydrochloride microspheres. Journal of Microencapsulation, 12, 193± 202. LABHASETWAR , V., UNDERWOOD, T., GALLAGHER , M., MURRHY , G., LANGBERG, J., and LEVY, R. J., 1994, Sotalol controlled-release systems for arrhythmias: in-vitro characterization, in vivo drug disposition, and electrophysiologic eå ects. Journal of Pharmaceutical Science, 83, 156± 164. MCGEE, J. P., SINGH, M., LI, X.-M., QIU, H., and O’ HAGAN , D. T., 1997, The encapsulation of model protein in poly(D,L lactide-co-glycolide) microparticles of various sizes: an evaluation of process reproducibility. Journal of Microencapsulation, 14, 197± 210. PEPPAS, N. A., 1985, Analysis of Fickian and non-Fickian drug release from polymers. Pharmaceutica Acta Helvetiae, 60, 110± 111. RIZK, S., DURU, C., GAUDY , D., JACOB , M., COLOMBO, P., and MASSIMO, G., 1994, Natural polymer hydrophilic matrix in¯ uencing drug release factors. Drug Developments and Industrial Pharmacy, 20, 2563± 2594. SAH, H. K., TODDYWALA , R., and CHIEN, Y. W., 1995, Biodegradable microcapsules prepared by a w/o/w technique: eå ect of shear force to make a primary w/o emulsion on their morphology and protein release. Journal of Microencapsulation, 12, 59± 69. SHAUGHNESSY , A. F., 1985, Potential uses for metoclopramide. Drug Intelligence in Clinical Pharmacy, 19, 723± 728.

Sustained release microspheres

435

UCHIDA, T., YAGI, Y., ODA, Y., and GOTO, S., 1996, Microencapsulation of ovalbumin in poly(lactide-co-glycolide) by an oil-in-oil (o/o) solvent evaporation method. Journal of Microencapsulation, 13, 509± 518. VIGOREAUX, V., and GHALY , E. S., 1994, Fickian and relaxational contribution quanti® cation of drug release in a swellable hydrophilic polymer matrix. Drug Developments and Industrial Pharmacy, 20, 2519± 2526. YANG, J., and CLELAND, J. L., 1997, Factors aå ecting the in vitro release of recombinant human interferon-(rhIFN-) from PLGA microspheres. Journal of Pharmaceutical Sciences, 86, 908± 914.