Sustained release from hot-melt extruded matrices based on ethylene vinyl acetate and polyethylene oxide

European Journal of Pharmaceutics and Biopharmaceutics 82 (2012) 526–533

Contents lists available at SciVerse ScienceDirect

European Journal of Pharmaceutics and Biopharmaceutics journal homepage: www.elsevier.com/locate/ejpb

Research paper

Sustained release from hot-melt extruded matrices based on ethylene vinyl acetate and polyethylene oxide A. Almeida a, L. Brabant b, F. Siepmann c, T. De Beer d, W. Bouquet a, L. Van Hoorebeke b, J. Siepmann c, J.P. Remon a, C. Vervaet a,⇑ a

Laboratory of Pharmaceutical Technology, Ghent University, Harelbekestraat 72 B-9000, Gent, Belgium Centre for X-ray Tomography (UGCT), Department of Physics and Astronomy, Ghent University, Ghent, Belgium College of Pharmacy, University of Lille, Lille, France d Laboratory of Process Analytical Technology, Ghent University, Ghent, Belgium b c

a r t i c l e

i n f o

Article history: Received 21 April 2012 Accepted in revised form 10 August 2012 Available online 8 September 2012 Keywords: Hot-melt extrusion Ethylene vinyl acetate Polyethylene oxide Sustained release

a b s t r a c t The aim of the present study was to evaluate the importance of matrix flexibility of hot-melt extruded (HME) ethylene vinyl acetate (EVA) matrices (with vinyl acetate (VA) contents of 9%, 15%, 28% and 40%), through the addition of hydrophilic polymers with distinct swelling capacity. Polyethylene oxide (PEO 100 K, 1 M and 7 M) was used as swelling agent and metoprolol tartrate (MPT) as model drug. The processability via HME and drug release profiles of EVA/MPT/PEO formulations were assessed. Solid state characteristics, porosity and polymer miscibility of EVA/PEO matrices were evaluated by means of DSC, X-ray tomography and Raman spectroscopy. The processability via HME varied according to the VA content: EVA 40 and 28 were extruded at 90 °C, whereas higher viscosity EVA grades (EVA 15 and 9) required a minimum extrusion temperature of 110 °C to obtain high-quality extrudates. Drug release from EVA matrices depended on the VA content, PEO molecular weight and PEO content, matrix porosity as well as pore size distribution. Interestingly, the interplay of PEO leaching, matrix swelling, water influx and changes in matrix porosity influenced drug release: EVA 40- and 28-based matrices extruded with PEO of higher MW accelerated drug release, whereas for EVA 15- and 9-based matrices, drug release slowed down. These differences were related to the distinct polymer flexibility imposed by the VA content (lower VA content presents higher crystallinity and less free movement of the amorphous segments resulting in a higher rigidity). In all cases, diffusional mass transport seems to play a major role, as demonstrated by mathematical modeling using an analytical solution of Fick’s second law. The bioavailability of EVA 40 and 28 matrices in dogs was not significantly different, independent of PEO 7 M concentration. Ó 2012 Elsevier B.V. All rights reserved.

1. Introduction Hot-melt extrusion (HME) is a valuable and versatile processing technology for drug delivery systems with an exciting future within the pharmaceutical industry. Independent of the final dosage form (granules, pellets, mini-matrices, beads, tablets, implants, vaginal rings), one of the main applications of HME as drug delivery technology is to control drug release over an extended period of time, using a variety of polymers: ethylcellulose [1,2], polyvinyl acetate [3], poly(L-lactic acid) [4], poly(lactic-co-glycolic acid) [5], polycaprolactone [6], silicone [7], ammonium metacrylate copolymers [8], lipid matrices [9] and ethylene vinyl acetate (EVA) [10]. EVA has been identified not only as a suitable polymer for HME to produce implants, films, transdermal patches and vaginal rings, but also as a matrix former in hot-melt extruded matrices for oral ⇑ Corresponding author. Tel.: +32 (0)9 264 8069; fax: +32 (0)9 222 82 36. E-mail address: [email protected] (C. Vervaet). 0939-6411/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ejpb.2012.08.008

sustained-release formulations [11]. The low glass transition temperature of EVA polymers (around 25 °C, independent of the vinyl acetate content) allows easy processing via HME, and its hydrophobic chains ensure sustained-release capacity. One of the main features of EVA matrices with higher VA content is its flexibility, which was responsible for a partial elastic rearrangement of the matrix once part of the drug had leached from the tablet. While the EVA 40 matrix can be initially (i.e., after HME) structurally supported by drug crystals, the structure might partially collapse after drug release, reducing the pathways available for diffusioncontrolled release of the remaining drug fraction [11]. In order to further understand the flexible structural behavior of hot-melt extruded EVA formulations and its influence on drug release, PEO was added in the present study to such matrices as a swelling agent. PEO has been successfully used to increase drug release from hydrophobic matrices as the PEO chains swell in contact with aqueous media, thus opening the matrix structure and allowing easier access of the dissolution liquid in the EVA matrix [12]. The

A. Almeida et al. / European Journal of Pharmaceutics and Biopharmaceutics 82 (2012) 526–533

physicochemical properties (drug release, solid state properties via DSC and Raman spectroscopy, and porosity via X-ray tomography) and in vivo behavior of hot-melt extruded matrices containing PEO and EVA were characterized during this study. 2. Materials and methods 2.1. Materials Different ethylene vinyl acetate (EVA) grades (ElvaxÒ 40w, 260, 550 and 750 with a vinyl acetate (VA) content of 40, 28, 15 and 9 wt.%, respectively) were kindly donated by DuPont (Geneva, Switzerland) and used as hydrophobic carriers. Metoprolol tartrate (MPT) (10 lm) (EQ Esteve, Barcelona, Spain) was selected as model drug. To modulate the drug release profile, polyethylene oxides (PEO, Sentry™ Polyox™, Dow Chemical Company, Midland, USA) with different molecular weights (MW) were used: WSR N10, N12K and 303 with a MW of 100,000 (100 K), 1,000,000 (1 M) and 7,000,000 (7 M), respectively. LutrolÒ F 68 (recently named Kolliphor P 188; polyoxypropylene–polyoxyethylene block copolymer, BASF, Burgbernheim, Germany) was also added to the formulations to increase drug release. 2.2. Hot-melt extrusion: production of the mini-matrices Physical mixtures of metoprolol tartrate, PEO, Lutrol and EVA (homogenized using mortar and pestle) were fed into a co-rotating twin-screw mini-extruder (Haake MiniLab II Micro Compounder, Thermo Electron, Karlsruhe, Germany) and extruded at a screw speed of 60 rpm and a processing temperature of 90–110 °C (Table 1). The extruder was equipped with a pneumatic feeder, two conical intermeshing screws and a cylindrical die of 2 mm. The extrudates were cooled down to room temperature under ambient conditions and afterward manually cut, using surgical blades, into mini-matrices of 2 mm length. 2.3. Extrudate characterization 2.3.1. Raman spectroscopy A RamanRxn 1 Microprobe (Kaiser Optical systems, Ann Arbor, USA) equipped with an air-cooled CCD detector (back-illuminated deep depletion design) was used to evaluate the PEO crystallinity in the hot-melt extruded matrices. Spectra were collected on a vertical cross-section of the mini-matrices using a 10 long working distance objective lens (laser spot size: 50 lm). To evaluate PEO distribution in the tablet (horizontal cross-section), five areas

Table 1 Composition of the formulations, extrusion temperature and surface structure of the hot-melt extruded formulations. Batch

EVA type

EVA/drug ratio

PEO (7 M) (%)

Extrusion temperature (°C)

Surface structure

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

EVA EVA EVA EVA EVA EVA EVA EVA EVA EVA EVA EVA EVA EVA EVA

50:50 50:50 50:50 50:50 60:40 60:40 60:40 60:40 50:50 60:40 60:40 60:40 60:40 70:30 70:30

5 5 5 5 5 5 5 5 10 10 10 10 10 10 10

90 90 110 110 90 90 110 110 90 90 90 110 110 110 110

Smooth Shark skin Shark skin Shark skin Smooth Smooth Smooth Smooth Shark skin Smooth Smooth Shark skin Shark skin Smooth Smooth

40 28 15 9 40 28 15 9 40 40 28 15 9 15 9

527

(2200  1200 lm, four at the edges and one in the middle of the mini-matrices) were scanned in a point-by-point mapping mode with a step size of 50 lm in both the x and y directions. The laser wavelength during the experiments was the 785 nm line from a 785 nm Invictus NIR diode laser. All spectra were recorded at a resolution of 4 cm1 using a laser power of 400 mW and a laser light exposure time of 20 s per spectrum. Before data analysis, spectra were baseline-corrected (Pearson’s method). Data collection and analysis were done using the HoloGRAMS™ data collection software package, the HoloMAP™ data analysis software package and the MatlabÒ software package (version 6.5). 2.3.2. Thermal analysis Glass transition temperature (Tg), crystallization temperature (Tc), melting point (Tm) and heat of fusion (DH) of pure components (EVA grades, MPT and PEO), physical mixtures and extruded samples were analyzed by differential scanning calorimetry (DSC) and modulated differential scanning calorimetry (MDSC). The DSC instrument (Q2000, TA Instruments, Leatherhead, UK) was run in standard mode and equipped with a refrigerated cooling system (RCS). Samples (±5 mg) were run in T0 aluminum plans over a temperature range from 60 to 180 °C with a heating rate of 10 °C/ min. (M)DSC measurements were carried out using dry nitrogen (flow rate of 50 ml/min) to purge the DSC cell. During MDSC tests, the temperature amplitude was 0.3 °C, the period 50 s and the underlying heating rate 2 °C/min. The samples were evaluated over a temperature range from 60 to 180 °C. All results were analyzed using the TA Instruments Universal Analysis 2000 software. 2.3.3. In vitro drug release Drug release from EVA-based matrices was determined using USP apparatus 1 (baskets), in a VK 7010 dissolution system combined with a VK 8000 automatic sampling station (VanKel Industries, New Jersey, USA). The mini-matrices (n = 8, length: 2 mm) were placed in demineralized water (900 ml, 37 ± 0.5 °C), while the rotational speed of the baskets was 100 rpm. Samples of 5 ml were withdrawn at 0.5, 1, 2, 4, 6, 8, 12, 16, 20 and 24 h (without medium replacement) and spectrophotometrically analyzed for MPT at 222 nm by means of a Perkin–Elmer Lambda 12 UV–VIS double beam spectrophotometer (Zaventem, Belgium). Each batch was evaluated in triplicate. 2.3.4. X-ray tomography The internal 3D-structure of the mini-matrices (total porosity and equivalent diameter) was evaluated by means of X-ray tomography. Total porosity is expressed as the percentage of a material’s pore volume to its total volume. Equivalent diameter is the diameter of a sphere with the same volume as the pore. EVA samples extruded in combination with MPT and 5% PEO were scanned immediately after HME, and after 24 h dissolution testing (to avoid collapse of the samples after dissolution testing, these samples were freeze-dried prior to analysis). X-ray tomography was performed using the high resolution micro-CT scanner of the Ghent University Centre for X-ray Tomography [13]. The system is composed of a FeinFocus FXE160.51 X-ray tube with Tungsten transmission target and diamond exit window (FeinFocus, Garbsen, Germany), a Micos UPR160-F Air high precision airbearing rotation stage (Micos, Eschbach, Germany) and a Varian Paxscan 2520V a-Si flat panel detector (Varian Medical Systems, Palo Alto, CA, USA). The sample was rotated over 360° in 0.36° steps, with radiographic images recorded at every step. The 1000 shadow images were processed and reconstructed to 1500 cross-sectional images of 1100  1100 pixels with Octopus software (inCT, Ghent, Belgium). Reconstructions were performed using the Modified Bronnikov Algorithm (MBA) [14–16].

528

A. Almeida et al. / European Journal of Pharmaceutics and Biopharmaceutics 82 (2012) 526–533

The in-house developed software package Morpho+ [17] was used for 3D image analysis. As the voxel pitch was about 2 lm, pores with a smaller diameter were not visible on the CT scans. However, they do affect the reconstructed image by reducing the density inside that voxel. 2.3.5. Extrudate morphology The extrudates were visually inspected for surface defects (e.g., shark skinning), while the deformation (due to cutting) of the mini-matrices was assessed using a KH-7700 digital microscope (Hirox, Japan), equipped with a high resolution zoom lens (MXG10C model, using co-axial vertical lighting for high-level optical observation) and an OL-70 II objective lens with a magnification capacity of 70–700. Swelling of the matrices was evaluated by measuring diameter and height during dissolution testing (at 0, 0.5, 1, 2, 4, 6, 8, 12 and 24 h). 2.3.6. Viscosity The kinematic viscosity of PEO 7 M and PEO 7 M/Lutrol (9/1, w/w) aqueous solutions (n = 3) (PEO concentration: 2.5 mg/ml) was measured using a capillary viscosimeter (micro-Ubbelohde viscosimeter, type 53720/II) (Schott-Geräte, Mainz, Germany) at 25 °C. 2.3.7. Mathematical modeling Drug release from the cylindrical extrudates was quantitatively described using an analytical solution of Fick’s second law of diffusion, considering axial as well as radial mass transfer, perfect sink conditions, an initially homogeneous drug distribution within the mini-matrices (before exposure to the release medium, t = 0) as well as negligible mass transfer resistance due to liquid unstirred boundary layers at the systems’ surfaces [18]. Under these conditions, the following equation can be derived [19]:

 2  X 1 1 Mt 32 X 1 q 1 ¼1 2   exp  n2  D  t  2 M1 p n¼1 qn R ð2  p þ 1Þ2 p¼0 ! ð2  p þ 1Þ2  p2  exp   D  t H2

ð1Þ

where Mt and M1 represent the absolute cumulative amounts of drug released at time t and infinite time, respectively; qn is the roots of the Bessel function of the first kind of zero order [J0(qn) = 0], R and H denote the radius and height of the cylinder, and D represents the apparent diffusion coefficient of the drug within the matrix. For the implementation of the mathematical model, the programming language C++ was used. 2.4. In vivo study All procedures were performed in accordance with the guidelines and approval of the local Ethics Committee on Animal Experimentation. To investigate the influence of VA content and PEO in the hotmelt extruded mini-matrices on the bioavailability of metoprolol

tartrate, different formulations (Table 2) were administered to male mixed-breed dogs (n = 6, 22.0–41.5 kg). The mini-matrices were filled in hard-gelatin capsules no. 000, each capsule containing 200 mg metoprolol tartrate. SlowLopressorÒ 200 DivitabsÒ (1 tablet) was used as sustained-release reference formulation. The formulations were administered in randomized order with a washout period of 1 week. On the experimental days, the dogs were fasted for 12 h prior to the study period, although water was available. Before administration of the formulations, a blank blood sample was obtained. The formulations were orally administered with 10 ml water, and blood samples (3 ml at each sampling) were collected in dry heparinized tubes 0.5, 1, 2, 3, 4, 5, 6, 8 and 12 h after intake of the formulations. No food was administered to the dogs during the entire test period, but water could be taken freely. Within 1 h after collection, blood was centrifuged for 10 min at 1500g and kept frozen at 20 °C until analyzed. 2.4.1. Metoprolol tartrate assay The metoprolol tartrate plasma concentrations were determined based on the HPLC-fluorescence method used by Quinten et al. [20] using bisoprolol as internal standard and acetonitrile/sodium dihydrogen orthophosphate buffer (2 M)/water (4/0.5/95.5, v/v/v) as mobile phase. The peak plasma concentration (Cmax), the time to reach Cmax (Tmax) and the extent of absorption (AUC0– 12h) were determined. The relative bioavailability (Frel) was calculated as the ratio of AUC0–12h values between a test formulation and the sustained-release reference formulation (Slow-LopressorÒ 200 DivitabsÒ). The sustained-release capacity of the formulations with 5% and 15% PEO 7 M was evaluated by the time span during which the plasma concentrations were at least 50% of the Cmax value (HVDt50%Cmax, corresponding to the width of the plasma concentration profile at 50% of Cmax) [21]. The effect of formulation composition on bioavailability was statistically evaluated by repeated-measures ANOVA (univariate analysis) using SPSS 17 (SPSS, Chicago, USA). To further compare the effects of the different treatments on the pharmacokinetic parameters, a multiple comparison among pairs of means was performed using a Bonferroni post hoc test with P < 0.05 as significance level. 3. Results and discussion 3.1. Addition of swelling agents to EVA matrices PEO polymers are non-ionic, hydrophilic, linear, semi-crystalline, homo-polymers manufactured by the heterogeneous catalytic polymerization of ethylene oxide and are available in a broad range of molecular weights (MW). When in contact with water, attraction forces between polymer and water weaken the interactions between polymer segments and the polymer chains start to swell. Due to this gel formation property, PEO polymers were added to EVA matrices (processed via hot-melt extrusion) in order to evaluate their effect on drug release from EVA matrices with different VA content.

Table 2 Mean pharmacokinetic parameters (±SD) after oral administration of 200 mg metoprolol tartrate to dogs (n = 6). Composition (%) EVA

MPT

EVA 40 47.5 47.5 EVA 40 42.5 42.5 EVA 28 55 40 EVA 40 60 40 Slow-LopressorÒ 200 DivitabsÒ a,b

Cmax (lg/ml)

Tmax (h)

AUC0–12h (lg h/ml)

Frel (%)

1.4 ± 0.1a 1.8 ± 0.3a 1.5 ± 0.2a 0.7 ± 0.3b 1.7 ± 0.4a

3.0 ± 0.0a 2.2 ± 0.4a 2.7 ± 0.5a 2.2 ± 0.4a 2.2 ± 0.4a

6.6 ± 1.6a 5.1 ± 0.9a 6.0 ± 2.0a 2.4 ± 0.7b 6.2 ± 2.4a

107.2 ± 45.7 82.7 ± 18.0 97.5 ± 26.1 38.4 ± 24.6 –

PEO 5 15 5 0

Means in the same column with different superscript are different at the 0.05 level of significance.

529

A. Almeida et al. / European Journal of Pharmaceutics and Biopharmaceutics 82 (2012) 526–533

100 80 60 40 20

Drug release from EVA matrices was also affected by the incorporation of PEO as swelling agent. Both, the PEO MW as well as the PEO content influenced the release rate. Although no differences were observed during the first 8 h in the case of EVA 40 matrices (EVA/MPT, 50/50), subsequent drug released differed: for instance, 66%, 81% and 85% MPT were released after 24 h when 5% PEO 100 K, 1 M and 7 M were added (compared to 58% in case of PEO-free extrudates) (Fig. 1). Higher PEO 7 M concentrations in such matrices accelerated MPT release: almost complete release

Cumulative drug release (%)

100 80 60 40 20 0 0

4

8

12

16

20

24

Time (h) Fig. 3. MPT release from EVA 40/MPT matrices (50/50, w/w) (), containing 10% PEO 7 M (j), PEO 1 M (N) and PEO 100 K ().

A

20

Matrix volume (mm3)

Cumulative drug release (%)

The addition of PEO of variable molecular weight (100 K, 1 M and 7 M) to MPT/EVA blends affected the HME processability. Table 1 lists the surface structure and extrusion temperature required to process the different blends containing PEO 7 M (formulations containing PEO 100 K and 1 M showed similar results, data not shown). While EVA/MPT mixtures (ratio: 50/50, w/w) could be processed into smooth-surfaced extrudates over a temperature range from 60 to 110 °C [11], the incorporation of 5% PEO 7 M required a higher minimal process temperature (i.e., 90–110 °C) to yield smooth extrudates. Hot-melt processing of higher viscosity EVA grades (EVA 9 and 15) in combination with MPT and PEO only resulted in high-quality extrudates at an extrusion temperature of 110 °C. The maximum achievable drug load in EVA/PEO matrices depended on the EVA grade: a maximum of 50% MPT in EVA 40 formulations (containing 5% PEO), while drug load was limited to 40% in case of EVA 28 and EVA 15. On the other hand, in the case of EVA 9 (the most viscous polymer), it was not possible to obtain good quality extrudates when PEO was added to the formulation. Although at 40% drug content smooth extrudates based on EVA 9 could be produced (batch 8), the latter changed their color after HME processing. A similar observation was made at lower drug content and higher PEO content (batch 15), due to the extended residence time and high shearing forces of the material in the extrusion chamber when processing this highly viscous formulation (torque increased to about 0.9 Nm, but remained constant during extrusion). At the die exit, swelling of all formulations was observed, for example, the diameter of EVA 15 matrices (containing 5% PEO) increased from 2 mm (i.e., diameter of extrusion die) to 2.9 mm.

16

12

8

4

0 0

4

8

12

16

20

24

0 0

Time (h)

4

8

12

16

20

24

Time (h) Fig. 1. MPT release from EVA 40/MPT matrices (50/50, w/w) (), containing 5% PEO 7 M (j), PEO 1 M (N) and PEO 100 K ().

B 200 175

(%) Liquid uptake

Cumulative drug release (%)

100 80 60 40

150 125 100 75 50 25

20

0 0

0 0

4

8

12

16

20

24

4

8

12

16

20

24

Time (h)

Time (h) Fig. 2. MPT release from EVA 40/MPT matrices (50/50, w/w), with 0% (), 5% (j), 10% (N) and 15% PEO 7 M (d).

Fig. 4. Changes in matrix volume (A) and liquid uptake (B) of EVA 40/MPT matrices (50/50, w/w), containing 0% (j) and 5% PEO 7 M () upon exposure to the release medium.

530

A. Almeida et al. / European Journal of Pharmaceutics and Biopharmaceutics 82 (2012) 526–533

was already observed after 12 and 8 h in the case of matrices containing 10% and 15% PEO 7 M, respectively (Fig. 2). In addition to the increased hydrophilicity of these matrices, the leaching of PEO into the bulk fluid increased MPT mobility within the matrices. The interplay between the flexibility of the EVA matrix [11] and viscosity of the PEO polymers was complex and the effect of PEO MW on drug release was not straightforward. For example, Fig. 3 shows drug release from EVA 40/MPT (50/50, w/w) matrices containing 10% PEO 100 K, 1 M and 7 M. In contrast to the respective systems containing only 5% PEO, drug release was faster for PEO 1 M. The same was true for extrudates containing 15% PEO (data not shown). In all these cases, drug release from PEO containing extrudates was slowest for PEO 100 K (still being faster than in the case of PEO-free devices). One of the effects of the presence of PEO in the extrudates was an increase in device swelling (Fig. 4A) and fluid uptake (Fig. 4B). Interestingly, despite the effect of PEO on the structure of the matrix, none of the EVA/PEO formulations completely released their drug content over the 24 h test period, indicating that a small fraction of the drug remained entrapped in the hydrophobic EVA matrix. Although PEO 7 M (Tmelt 76 °C) completely melted during extrusion due to thermal (Textr 90 °C) and shear stress (screw speed 60 rpm), at a concentration of 5% PEO the hydrophilic polymer recrystallized after cooling down to room temperature, as shown by the thermogram of the extrudate (Fig. 5). These clusters of PEO crystals were identified via Raman spectroscopy and X-ray tomography and are distributed throughout the entire EVA matrix. Mapping of the PEO signal (via Raman spectroscopy at 521– 549 cm1) on a horizontal cross-section detected PEO organized in crystalline agglomerates between 100 and 250 lm. These find-

ings were confirmed by X-ray tomography (Fig. 6), as the lighter gray zones (characteristic of denser material such as semi-crystalline polymers) embedded in the tablet structure corresponded to PEO clusters. X-ray tomography also allowed visualization of the phases in the formulation: the less dense EVA polymer appears as the darker gray zones (representing the bulk of the material), MPT crystals as small white pixels (uniformly dispersed throughout the matrix) and pores as black zones (corresponding to zero density). Quantification of the pore network via X-ray tomography not only confirmed the preferential location of larger and smaller pores in the matrix (i.e., larger pores mainly in the core and smaller ones toward the surface due to compression of the melt during passage through the die), it also showed that the addition of PEO increased matrix porosity (from 7.2% [11] to 14.4%) and pore size (Fig. 7A), and broadened the pore size distribution. Matrices with 5% PEO had pores with an average equivalent diameter of 52 lm, whereas without PEO it was only 38.4 lm. As the addition of semi-crystalline PEO increased matrix porosity and swelling of hydrated PEO chains stretched the EVA 40 matrix, these changes in pore structure contributed to the increase in drug release rate (Figs. 1 and 2). The total porosity of the samples after dissolution testing increased up to 23.5%, and the pore size distribution became even broader (the largest equivalent pore diameter in the matrix after dissolution was 274 lm, vs. 111 lm prior to dissolution) (Figs. 6B and 7A). The reduction in bulk volume after dissolution as observed during X-ray tomography renderings (Fig. 6B) was caused by shrinking of the sample (when exposed to room temperature) due to drying prior to scan. Fitting Eq. (1) – an analytical solution of Fick’s second law of diffusion – to the experimentally deter-

Fig. 5. DSC thermograms of pure MPT (1), EVA 40 (2), PEO 7 M (3), physical mixture (4) and extrudate (5) of EVA 40/MPT (50/50, w/w) with 5% PEO. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

531

A. Almeida et al. / European Journal of Pharmaceutics and Biopharmaceutics 82 (2012) 526–533

A

B Fig. 6. X-ray tomography renderings of radial cross-sections of extrudates based on EVA 40/MPT (50/50, w/w) containing 5% PEO 7 M, (A) before dissolution testing and (B) after 24 h dissolution testing.

2,1E+08 1

A

Volume (µm3)

1,8E+08 1,5E+08 1,2E+08 9,0E+07

2

6,0E+07 3,0E+07 3

0,0E+00 0

50

100

150

200

250

300

Cumulative drug release (%)

A

mined drug release kinetics indicated that diffusional mass transport is likely to play a major role in the control of MPT release from the investigated matrices. Fig. 9A shows some examples of these fittings. As it can be seen, good agreement between theory (curves) and experiment (symbols) was observed in all cases. This was true, irrespective of the EVA type and absence/presence of PEO (data not shown). Based on these calculations, the apparent diffusion coefficients of MPT in the investigated EVA extrudates could be determined (Fig. 9B). The effect of the addition of PEO on MPT release from EVA extrudates also depended on the VA content. While drug release after 24 h from EVA 40 matrices (EVA/MPT, 60/40, w/w) increased from 36% to 85% upon the incorporation of 10% PEO 7 M (Fig. 8A), MPT release only increased from 57% and 77% in the case of EVA 28 after the same exposure time to the bulk fluid. Also, the relative increase in the apparent diffusion coefficient of MPT upon addition of 5% PEO 7 M was much more pronounced in the case of EVA 40 compared to EVA 28 (Fig. 9B). In contrast, the addition of 5% PEO 7 M to EVA 9 (data not shown) and EVA 15 (Fig. 8B) matrices slowed down drug release (also indicated by lower apparent drug diffusivities in 5% PEO 7 M containing EVA 9 and EVA 15 based matrices compared to the respective PEO-free extrudates, Fig. 9B). This decrease in release was caused by the creation of a highly viscous PEO solution/ gel phase [22] (instantly formed within the pore network of the EVA matrix upon contact with the dissolution medium). Such a highly viscous phase can act as a diffusion barrier for the drug. Due to lower flexibility of EVA 9 and 15, swelling of both formulations was limited (radial increase of 0.02 and 0.01 mm after 24 h exposure of EVA 15 and 9 matrices to the bulk fluid, respectively). The higher flexibility of EVA 28 and EVA 40 allowed a higher MPT release rate upon PEO addition: being more flexible, the matrices

Equivalent diameter (µm)

100 80 60 40 20 0 0

8

12

16

20

24

Time (h) 1

1,8E+08

Volume (µm3)

4

2,1E+08

B

1,5E+08 1,2E+08 3

9,0E+07

4

2

6,0E+07 3,0E+07 0,0E+00 0

50

100

150

200

250

300

Equivalent diameter (µm)

Cumulative drug release (%)

B

100

80

60

40

20

0 Fig. 7. Volume distribution of the equivalent pore diameter obtained by X-ray tomography; (A) EVA 40/MPT (50/50, w/w) matrices before dissolution experiments with 0% (1) and 5% PEO 7 M (2) and after 24 h dissolution experiments with 5% PEO 7 M (3); (B) EVA/MPT (50/50, w/w) matrices formulated with EVA 40 (1), EVA 28 (2), EVA 15 (3) and EVA 9 (4). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

0

4

8

12

16

20

24

Time (h) Fig. 8. MPT release from EVA 40 (A) and EVA 15 (B) matrices (EVA/MPT, w/w, 60/ 40), containing: 0% PEO 7 M (j), 5% PEO 7 M (N), 10% PEO 7 M (), 5% PEO 7 M/ Lutrol (9/1, w/w) (), or 5% PEO 100 K (d).

532

A. Almeida et al. / European Journal of Pharmaceutics and Biopharmaceutics 82 (2012) 526–533

Therefore, the swelling capacity of the hydrophilic PEO chains (higher MW corresponded to higher viscosity) improved drug release from EVA 40 and 28-based hydrophobic matrices due to polymer flexibility. The stretching observed in the matrices, the hydrophilicity imposed by PEO chains and the increased porosity, facilitated drug percolation inside the matrix and a faster dissolution into the medium. However, this viscous PEO gel was not able to enhance drug release from less flexible EVA matrices (EVA 15 and 9) entrapping the drug inside the matrix.

100

Cumulative drug release (%)

A

75

50

25

3.2. In vivo performance 0 0

4

8

12

16

20

24

Time (h)

B

4

D, 10-8 cm2/s

3

2

1

As the EVA grade (i.e., VA content), swelling agent and porosity had a pronounced impact on MPT release, the effect of these parameters on the in vivo performance of the EVA-based matrices was evaluated. Fig. 10 shows the mean plasma concentration–time profiles after oral administration of the formulations, while the pharmacokinetic parameters are reported in Table 2. The slow drug release from EVA 40/MPT matrices (60/40, w/w) was correlated with a low bioavailability. However, the addition of PEO 7 M to the EVA 40 matrix enhanced drug release from the matrix and improved bioavailability. While the PEO concentration had a significant effect of in vitro drug release from the EVA 40 matrices (50% drug release after about 10 and 3 h from tablets formulated with 5% and 15% PEO 7 M, respectively), the AUC0–12h values were not significantly different (P > 0.05), both showing a sustainedrelease capacity in vivo. However, the 5% PEO formulation tended to sustain plasma levels over a longer period (HVDt50%Cmax = 4.1 and 2.3 for 5% and 15% PEO, respectively), and tmax was increased.

0 EVA 40

EVA 28

EVA 15

EVA 9 2.5

0% PEO 7M 5% PEO 7M

Fig. 9. (A) Theory (curves, Eq. (1)) and experiments (symbols): MPT release from EVA 28-based matrices containing 0% PEO 7 M (j), 5% PEO 7 M (N) or 5% PEO 7 M/ Lutrol (9/1, w/w) (). (B) Apparent diffusion coefficients of MPT in EVA-based matrices, containing 0% PEO 7 M, 5% PEO 7 M, or 5% PEO 7 M/Lutrol (9/1, w/w). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

2

MPT conc. (µg/ml)

5% PEO 7M + Lutrol

1.5

1

0.5

0 0

2

4

6

8

10

16

20

12

Time (h) 100

Cumulative drug release (%)

allow for more water influx and, thus, limit a potential increase in the viscosity of the liquid filling the matrix pores. Moreover, in the case of EVA 15 and 9 larger pores in combination with a reduced number of small pores were detected (Fig. 7B), lowering the interconnectivity between the pores. The interplay between matrix swelling, porosity and fluid viscosity within the liquid filled pores was also reflected in the case of EVA 15 matrices containing lower MW PEO. The addition of 5% PEO 100 K (which forms lower viscous solution) to EVA 15 resulted in faster drug release compared to PEO-free extrudates (Fig. 8B). This is explained by a faster leaching of PEO 100 K out of the systems compared to PEO 7 M, limiting the potential effects of increased liquid viscosity within the matrix pores discussed above. Furthermore, the addition of Lutrol in combination with a high molecular weight PEO grade (i.e., PEO 7 M) increased significantly drug release from EVA 9 (data not shown) and EVA 15 (Fig. 8B) matrices (reflected by increased apparent MPT diffusivities, Fig. 9B). This surfactant lowers the surface tension, and it decreases the viscosity of a PEO solution, that is, the kinematic viscosity of a 2.5 mg/ml aqueous PEO solution decreases from 10.1 to 4.5 cStokes upon addition of 10% Lutrol. Therefore, PEO with higher MW increased drug release from EVA 40 and 28 matrices, whereas for EVA 15 and 9, only lower MW PEO (with reduced viscosity) was able to enhance drug release.

80 60 40 20 0 0

4

8

12

24

Time (h) Fig. 10. (A) Mean plasma concentration–time profiles (±SD, n = 6) after oral administration of 200 mg metoprolol tartrate to dogs and (B) in vitro drug release profiles of the same formulations: EVA 40/MPT (50/50, W/W) containing 5% PEO 7 M (j) or 15% PEO 7 M (N); EVA 28/MPT (55/40, w/w) containing 5% PEO 7 M (); EVA 40/MPT (60/40, w/w), free of PEO (d) and 1 tablet of Slow-LopressorÒ 200 DivitabsÒ (reference formulation) (- - -).

A. Almeida et al. / European Journal of Pharmaceutics and Biopharmaceutics 82 (2012) 526–533

Substitution of EVA 40 by EVA 28 in the formulation had no significant effect on the plasma drug concentration profile. The bioavailability profile of the hot-melt extruded EVA/PEO combinations was similar with Slow-LopressorÒ 200 DivitabsÒ used as reference formulation. 4. Conclusion The impact of the PEO grade and concentration in EVA-based matrices varied depending on the VA content: the interplay of PEO leaching, matrix swelling, water influx and changes in matrix porosity accelerated drug release from EVA matrices with higher VA content (due to increased mobility of the polymer chains), whereas at lower VA content, the addition of PEO with higher MW reduced drug release. In all cases, diffusional mass transport played a major role. The bioavailability profile of the hot-melt extruded EVA/PEO combinations was similar with a reference formulation containing metoprolol tartrate. Acknowledgements The authors wish to thank Mr. Daniël Tensy for his valuable contribution to the in vivo study. The Special Research Fund of the Ghent University (BOF) is acknowledged for the doctoral grant to Loes Brabant. References [1] E. Verhoeven, T.R.M. De Beer, G. Van den Mooter, J.P. Remon, C. Vervaet, Influence of formulation and process parameters on the release characteristics of ethylcellulose sustained-release mini-matrices produced by hot-melt extrusion, Eur. J. Pharm. Biopharm. 69 (2008) 312–319. [2] T. Quinten, Y. Gonnissen, E. Adriaens, T. De Beer, V. Cnudde, B. Masschaele, L. Van Hoorebeke, J. Siepmann, J.P. Remon, C. Vervaet, Development of injection moulded matrix tablets based on mixtures of ethylcellulose and lowsubstituted hydroxypropylcellulose, Eur. J. Pharm. Sci. 37 (2009) 207–216. [3] G.A.G. Novoa, J. Heinamaki, S. Mirza, O. Antikainen, A.I. Colarte, A.S. Paz, J. Yliruusi, Physical solid-state properties and dissolution of sustained-release matrices of polyvinylacetate, Eur. J. Pharm. Biopharm. 59 (2005) 343–350. [4] C.G. Pitt, A.R. Jeffcoat, R.A. Zweidinger, A. Schindler, Sustained drug delivery systems. 1. Permeability of poly(epsilon-caprolactone), poly(DL-lactic acid), and their copolymers, J. Biomed. Mater. Res. 13 (1979) 497–507. [5] J.K. Li, N. Wang, X.S. Wu, A novel biodegradable system based on gelatin nanoparticles and poly(lactic-co-glycolic acid) microspheres for protein and peptide drug delivery, J. Pharm. Sci. – US 86 (1997) 891–895.

533

[6] N. Follonier, E. Doelker, E.T. Cole, Evaluation of hot-melt extrusion as a new technique for the production of polymer-based pellets for sustained-release capsules containing high loadings of freely soluble drugs, Drug Dev. Ind. Pharm. 20 (1994) 1323–1339. [7] R.K. Malcolm, A.D. Woolfson, C.F. Toner, R.J. Morrow, S.D. McCullagh, Longterm, controlled release of the HIV microbicide TMC120 from silicone elastomer vaginal rings, J. Antimicrob. Chemother. 56 (2005) 954–956. [8] N. Follonier, E. Doelker, E.T. Cole, Various ways of modulating the release of diltiazem hydrochloride from hot-melt extruded sustained-release pellets prepared using polymeric materials, J. Control. Release 36 (1995) 243–250. [9] C. Reitz, P. Kleinebudde, Solid lipid extrusion of sustained release dosage forms, Eur. J. Pharm. Biopharm. 67 (2007) 440–448. [10] J.A.H. van Laarhoven, M.A.B. Kruft, H. Vromans, Effect of supersaturation and crystallization phenomena on the release properties of a controlled release device based on EVA copolymer, J. Control. Release 82 (2002) 309–317. [11] A. Almeida, S. Possemiers, M.N. Boone, T. De Beer, T. Quinten, L. Van Hoorebeke, J.P. Remon, C. Vervaet, Ethylene vinyl acetate as matrix for oral sustained release dosage forms produced via hot-melt extrusion, Eur. J. Pharm. Biopharm. 77 (2011) 297–305. [12] M. Varma, A.K. Singla, S. Dhawan, Release of diltiazem hydrochloride from hydrophilic matrices of polyethylene oxide and carbopol, Drug Dev. Ind. Pharm. 30 (2004) 545–553. [13] B.C. Masschaele, V. Cnudde, M. Dierick, P. Jacobs, L. Van Hoorebeke, J. Vlassenbroeck, UGCT: new X-ray radiography and tomography facility, Nucl. Instrum. Methods A 580 (2007) 266–269. [14] A. Groso, R. Abela, M. Stampanoni, Implementation of a fast method for high resolution phase contrast tomography, Opt. Exp.. 14 (2006) 8103–8110. [15] M. Boone, Y. De Witte, M. Dierick, J. Van den Bulcke, J. Vlassenbroeck, L. Van Hoorebeke, Practical use of the modified Bronnikov algorithm in micro-CT, Nucl. Instrum. Methods B 267 (2009) 1182–1186. [16] M.N. Boone, Y. De Witte, M. Dierick, A. Almeida, L. Van Hoorebeke, Improved signal-to-noise ratio in laboratory-based phase contrast tomography, Microsc. Microanal.: Offic. J. Microsc. Soc. Am.; Microbeam Anal. Soc.; Microsc. Soc. Can. (2012) 1–7. [17] L. Brabant, J. Vlassenbroeck, Y. De Witte, V. Cnudde, M.N. Boone, J. Dewanckele, L. Van Hoorebeke, Three-dimensional analysis of high-resolution X-ray computed tomography data with Morpho+, Microsc. Microanal. 17 (2011) 252–263. [18] J. Siepmann, F. Siepmann, Mathematical modeling of drug delivery, Int. J. Pharm. 364 (2008) 328–343. [19] J. Siepmann, F. Siepmann, Modeling of diffusion controlled drug delivery, J. Control. Release (2011). [20] T. Quinten, T. De Beer, F.O. Onofre, G. Mendez-Montealvo, Y.J. Wang, J.P. Remon, C. Vervaet, Sustained-release and swelling characteristics of xanthan gum/ethylcellulose-based injection moulded matrix tablets: in vitro and in vivo evaluation, J. Pharm. Sci. – US 100 (2011) 2858–2870. [21] J. Meier, E. Nuesch, R. Schmidt, Pharmacokinetic criteria for the evaluation of retard formulations, Eur. J. Clin. Pharmacol. 7 (1974) 429–432. [22] L. Maggi, L. Segale, M.L. Torre, E.O. Machiste, U. Conte, Dissolution behaviour of hydrophilic matrix tablets containing two different polyethylene oxides (PEOs) for the controlled release of a water-soluble drug. Dimensionality study, Biomaterials 23 (2002) 1113–1119.