A novel gastro-floating multiparticulate system for dipyridamole (DIP) based on a porous and low-density matrix core: In vitro and in vivo evaluation

International Journal of Pharmaceutics 461 (2014) 540–548

Contents lists available at ScienceDirect

International Journal of Pharmaceutics journal homepage: www.elsevier.com/locate/ijpharm

A novel gastro-floating multiparticulate system for dipyridamole (DIP) based on a porous and low-density matrix core: In vitro and in vivo evaluation Zhao Li, Heming Xu, Shujuan Li, Qijun Li, Wenji Zhang, Tiantian Ye, Xinggang Yang, Weisan Pan ∗ Department of Pharmaceutics, Shenyang Pharmaceutical University, 103 Wenhua Road, 110016 Shenyang, China

a r t i c l e

i n f o

Article history: Received 9 August 2013 Received in revised form 28 November 2013 Accepted 15 December 2013 Available online 22 December 2013 Keywords: Dipyridamole Gastro-floating Sustained release pellets Porous matrix cores Bioavailability

a b s t r a c t The study was aimed to develop a novel gastro-floating multiparticulate system based on a porous and low-density matrix core with excellent floatability. The gastro-floating pellets (GFP) were composed of a porous matrix core, a drug loaded layer (DIP and HPMC), a sub-coating layer (HPMC) and a retarding layer (Eudragit® NE 30D). The porous matrix cores were evaluated in specific. EC was chosen as the matrix membrane for its rigidity and minimal expansion to large extent. The porous matrix core was achieved by the complete release of the bulk water soluble excipient from the EC coated beads, and mannitol was selected as the optimal water soluble excipient. SEM photomicrographs confirmed the structure of porous matrix cores. The compositions of GFP were investigated and optimized by orthogonal array design. The optimized formulation could sustain the drug release for 12 h and float on the dissolution medium for at least 12 h without lag time to float. The pharmacokinetic study was conducted in beagle dogs, and the relative bioavailability of the test preparation was 193.11 ± 3.43%. In conclusion, the novel gastro-floating pellets can be developed as a promising approach for the gastroretentive drug delivery systems. © 2013 Elsevier B.V. All rights reserved.

1. Introduction The oral controlled drug delivery system is the most preferable and most reliable route for drug administration, benefiting from excellent patient compliance, low cost of therapy and flexibility in formulation. But for drugs that act locally in the proximal gastrointestinal tract, exhibit good solubility at an acidic pH but poor solubility at an alkaline pH, or have a narrow absorption in the stomach or in the upper small intestine, the conventional oral controlled drug delivery system cannot provide satisfactory therapeutic efficiency (Davis, 2005; Streubel et al., 2006). In these cases, the gastro-retentive systems show super advantages in enhanced bioavailability by releasing drug in a controlled and prolonged manner. The main approaches to achieve gastro-retention include: (a) High density systems, which can settle in the lower part of the antrum for a prolonged period of time by employing a heavy inert material such as barium sulphate, titanium dioxide, zinc

∗ Corresponding author. Tel.: +86 24 23986313. E-mail address: [email protected] (W. Pan). 0378-5173/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ijpharm.2013.12.024

oxide to increase the total density (Clarke et al., 1993; Guan et al., 2010). (b) Swelling and expanding systems, which after swallowing, imbibe lots of water before swell or unfold to prevent their passage through the pyloric sphincter (Johnson et al., 1997; Klausner et al., 2003). (c) Muco-adhesive systems, which are usually hydration, bonding or receptor mediated by adhering to the gastric epithelial cell, thereby prolonging the gastric retention time (Chun et al., 2005; Park and Robinson, 1984). (d) Floating systems, which can be effervescent or none effervescent in nature to reduce the bulk density to float on the surface of the gastric fluid (Gröning et al., 2007; Sungthongjeen et al., 2008). Among all the gastro-retentive systems, floating drug delivery systems (FDDS) are considered preferable and promising, since they do not adversely affect the motility of the gastrointestinal tract (GIT) (Kotreka and Adeyeye, 2011; Reddy and Murthy, 2002; Strusi et al., 2008). The fact can also confirm the superiority and reliability of the FDDS that many floating dosage forms have been commercialized and marketed (Bardonnet et al., 2006; Arora et al., 2005; Udaya and Kotreka, 2011). The wide range of dosage forms

Z. Li et al. / International Journal of Pharmaceutics 461 (2014) 540–548

Fig. 1. The schematic diagram of the novel gastro-floating pellets (GFP).

developed as FDDS include single unit FDDS and multiparticulate FDDS, and the multiparticulate FDDS are more advantageous than the single unit FDDS for the absence of “all or none” emptying pattern and minimized variation in distribution in the GIT. Dipyridamole (DIP) inhibits thrombus formation when given chronically and causes vasodilation when given at high doses over a short time. DIP is a weakly basic drug with a pKa value of 6.4 (Zhou et al., 2005), and thereby exhibits a pH-dependent solubility with good solubility at a low pH value and poor solubility at a high pH value (Zhang et al., 2009). Since DIP is classified as class II drug (high permeability and low solubility) according to the Biopharmaceutics Classification System (BSC) (Butler and Dressman, 2010), it shows a narrow absorption in the stomach and duodenum (Zhang et al., 2012). In addition, DIP has a short biological half-life of 2–3 h. Research data indicated that the individual difference in drug absorption and bioavailability of DIP is significant (Mahony et al., 1982; Terhaag et al., 1986). Recently, there are some searches on the DIP gastro-retentive dosage forms (Senthil et al., 2011; Birajdar Shivprasad and Darveshwar Jagdeep, 2005). All characterizations mentioned above make DIP an ideal drug to develop into a gastro-retentive dosage form. In this study, multi-layered gastro-floating pellets (GFP) of DIP were designed to provide sustained drug release in the stomach. Fig. 1 showed a schematic diagram of the novel gastro-floating pellets. Water soluble excipient, possessing 80% in weight, were completely released from the cellulose (EC) coated core beads. After dried at 40 ◦ C, the matrix cores were achieved and then sequentially coated with three different layers: a drug loaded layer (DIP and HPMC), a sub-coating layer (HPMC) and a retarding layer (Eudragit® NE 30D) by using the fluid bed coater. The air entrapped in the matrix cores confers buoyancy to retain in the stomach for a prolonged period of time. The composition of the porous matrix cores and the three layers were investigated. The porous matrix cores were evaluated for the final weight, mechanical strength and morphology, and the gastro-floating pellets were tested for in vitro floatability, in vitro drug release and morphology. Orthogonal array design was employed to clarify the significance of the influencing factors: the particle size of the matrix cores, the coating levels of the sub-coating layer (HPMC E5) and the retarding release layer (Eudragit® NE 30D), and then to optimize the formulation. The pharmacokinetic study was conducted in beagle dogs. 2. Material and method 2.1. Materials Dipyridamole (DIP, the purity of which was over 99.9%) was presented by Shenyang No. 1 Pharmaceutical Factory, Shenyang,

541

China; microcrystalline cellulose (MCC, Avicel PH-101) was a gift from FMC, USA; lactose, mannitol, sucrose, sodium chloride (NaCl) and sodium acetate anhydrous were purchased from Tianjin Bodi Chemical Industry, Tianjin, China; ethyl cellulose (EC, 10cp) and hydroxylpropyl methylcellulose (HPMC, E5) were gifts from Colorcon, shanghai, China; polyvinyl pyrrolidone (PVP k30) was a gift from BASF, Germany; Eudragit® NE 30D (B111112096) was a gift from Evonik Röhm Pharma, Darmstadt, Germany; sodium taurocholate and pepsin (1:3000) were purchased from Beijing Biotopped Science & Technology Co. Ltd., Beijing, China; Lipoid S 75 was a gift from Lipoid GmbH, Ludwigshafen, Germany; methanol, glacial acetic acid and ethyl acetate were purchased from Jiangsu Hanbon Sci. & Tech. Co. Ltd., Jiangsu, China. All the others were of either analytical grade or HPLC grade. Conventional sustained release pellets (CP) with the similar release profile were prepared as the reference preparation in the pharmacokinetic study. 2.2. Preparation of GFP 2.2.1. Preparation of the porous matrix cores The core beads were prepared by using extrusion/ spheronization technique (E-50/S250, Chongqing Enger Granulating & Coating Technology Co., Ltd., China). Water soluble excipient (NaCl, mannitol, lactose or sucrose) and MCC (4:1, w/w) were uniformly mixed by passing through an 80 mesh screen. Moderate amount of distilled water was added to make damp mass, and then it was extruded through a 1.0 mm screen at 35 rpm. The extrudates were spheronized at 700 rpm for 20 min. The collected beads were dried overnight at 40 ◦ C in an oven, and then they were sieved for next study. The beads were coated with a coating solution which prepared by dissolving EC and PVP k30 in alcohol–water (80:20, v/v), and the mass ratios of EC to PVP k30 were 10:1, 7:1, 5:1 and 4:1. The PVP k30 was added as pore forming agent. The coating solution was sprayed onto the beads in a fluid bed coater (FD-MP-01, Powrex, Japan). The levels of the coating weight gain were 10%, 20% and 30%, separately. The coating conditions were as follows: temperature: 30 ± 2 ◦ C, spray rate: 1.0 ml/min, atomization pressure: 0.2 bar, air flow frequency: 35 Hz. After coating, the pellets were dried at 40 ◦ C for 12 h. The matrix cores were achieved by using the USA TypeI (basket) dissolution test apparatus (RCZ-6B, Shanghai Huanghai Drug Inspection Instrument Co., Shanghai, China) in 1000 ml distilled water at 50 rpm and room temperature (25 ◦ C). After 6 h, the distilled water was replaced by fresh distilled water. Pellets were picked out at 12 h and dried at 40 ◦ C to constant weight. They were weighed accurately before and after dissolution. 2.2.2. Coating of the three successful layers: drug layer, sub-coating layer (HPMC) and retarding layer (Eudragit® NE 30D) Firstly, a drug-binder suspension was sprayed onto the matrix core using the fluid bed coater. DIP passing through a 200 mesh sieve was dispersed in the HPMC solution. The PEG6000 (10%, w/w of HPMC) was added as plasticizer. The bottom spray technique was optimized for different parameters, such as atomization pressure, spray rate and air flow frequency. During the optimization process, each parameter was varied while the others were kept constant. The process parameters were as follows: temperature: 35 ± 2 ◦ C, atomization pressure: 0.2 bar, spray rate: 0.7 ml/min, air flow frequency: 25 Hz. After coating, the pellets were fluidized in the fluid bed coater for 15 min to remove the residual moisture. Then HPMC solution plasticized by PEG6000 was sprayed onto the drug-layered matrix cores as a sub-coating in the fluid bed coater. The coating parameters were as follows: temperature:

542

Z. Li et al. / International Journal of Pharmaceutics 461 (2014) 540–548

Table 1 Levels and factors affecting the drug release property. Level

1 2 3

Factors A (mesh)

B (%)

C (%)

18–24 24–30 30–40

0 5 10

5.5 6 6.5

35 ± 2 ◦ C, atomization pressure: 0.2 bar, spray rate: 0.5 ml/min, air flow frequency: 30 Hz. Finally, DIP layered matrix cores were coated with Eudragit® NE 30D in the fluid bed coater. The aqueous dispersion was diluted by distilled water to offer a 10% concentration of acrylic copolymer. The coating parameters were as follows: temperature: 25 ± 2 ◦ C, spray rate: 0.5 ml/min, atomization pressure: 0.2 bar, air flow frequency: 30 Hz. After coating, the pellets was mixed with 0.5% (w/w) fine talc and cured at 40 ◦ C for 24 h to form a uniform film. 2.3. Orthogonal array design (OAD) Orthogonal array design was used to design the experiment and to optimize the compositions of the gastro-floating pellets. Based on the result of the preliminary investigation, in a specific range the amount of drug layered on the porous matrix cores had little effect on the drug release, and in this study it was fixed at the level of 30% (w/w). Thus only three factors were investigated: (A) the particle size of the matrix cores, (B) the coating level of sub-coating layer (HPMC) and (C) the coating level of retarding layer (Eudragit® NE 30D). Three levels were set for each factor based on the results of single factor experiment. Factors and levels tested were reported in Table 1. The orthogonal table OA9 (34 ) was designed, in which a blank column was designed for the error evaluation. The critical response (R) was calculated using the following formula: R = (|w2h − 20%| + |w6h − 50%| + |w10h − 80%|) × 100%

(1)

where w2h , w6h , w10h is the cumulative percentage drug release at 2 h, 6 h and 10 h. The significant levels of different influencing factors were clarified by Range analysis and analysis of variance (ANOVA). 2.4. Evaluation of the porous matrix cores 2.4.1. Final weight study Since the buoyancy originated in the porous matrix cores, the more that the water soluble excipient was released, more pores was formed in the matrix cores and the better floatability we achieved. From formulation and technological point of view, the EC coated core beads should release the water soluble excipient completely in 12 h. In theory, the final weight (m1 ) of the porous matrix cores after dried could be calculated as follows: m1 = m0 · (1 − w1 ) + m0 · w2 · w3

(2)

where m0 is the weight of core pellets, w1 is the fraction of mannitol in the core beads, w2 is the level of EC coating weight gain and w3 is the fraction of EC in the EC-PVP membrane. When the value of the final weight (m1 ) by weighing fell in the range of (100 ± 2) % · m1 , it was believed that the water soluble excipient was released completely. 2.4.2. Mechanical strength study Considering the process of drug layered afterwards, the matrix cores should be strong enough to withstand the powerful mechanical collision in the fluid bed coater and maintain integrity. 5 g of the matrix cores weighed accurately were placed in the fluid bed coater to be fluidized for 20 min under extreme condition. Afterward the wrecks were removed, and all of the remained matrix cores were weighed accurately. The instrumental parameters were as follows: temperature: 35 ± 2◦ C, atomization pressure: 0.2 bar, air flow frequency: 45 Hz. Here, a mechanical strength index (wm ) was introduced in the study. The mechanical strength increased as the wm increased. The fraction of the remained matrix cores was used to calculate the mechanical strength index (wm ) according to the following formula: m3 wm = × 100% (3) m2 where m3 is the weight of the matrix cores remaining integrity and m2 is the initial weight of the total matrix cores. All the tests were conducted in triplicate. 2.4.3. Scanning electron microscopy (SEM) The morphologies of the surface and cross-section of the matrix cores were examined under a SEM (SSX-550, Shimadzu, Japan). The sphere of a single pellet was taken as a whole or was sectioned into two halves by a sharp blade. The pellets were mounted on a glass stub with double side adhesive tape. Afterward, the pellets were coated under an argon atmosphere with gold palladium. 2.5. Evaluation of GFP 2.5.1. In vitro buoyancy study To evaluate the buoyancy of GFP, the counting method described by Zhang et al. (2012) was adopted. This test was carried out using USA type II (paddle) dissolution apparatus (RCZ-6B, Shanghai Huanghai Drug Inspection Instrument Co., Shanghai, China) in 500 ml 0.1 N HCl, fasted state simulated gastric fluid (FaSSGF) or fed state simulated gastric fluid (FeSSGF) at 50 rpm and 37 ◦ C, and the sample compositions of the FaSSGF and FeSSGF were listed in Table 2. A precise number of pellets (between 100 and 150) were placed in the medium, and the number of sank pellets was observed visually. Experiments were performed in triplicate. The fraction of the floating pellets (wf ) was calculated by the following formula: wf =

nt − ns nt

(4)

Table 2 Sample compositions for fasted state simulated gastric conditions (FaSSGF) and fed state simulated gastric fluid (FeSSGF). FaSSGF pH 1.6 Sodium taurocholate Lecithin Pepsin NaCl HCl conc Deionized water Osmolality (mOsmol/kg) Surface tension (mN/m)

FeSSGF pH5.0

qs ad ad

80 ␮M 20 ␮M 0.1 mg/ml 34.2 mM pH 1.6 ad 1 l 120.7 ± 2.5 42.6

NaCl Acetic acid Sodium acetate Milk/acetate buffer HCl conc. Osmolality (mOsmol/kg) Buffer capacity (mEq/pH/L)

qs ad

237.02 mM 17.12 mM 29.75 mM 1:1 pH 5.0 400 25

Z. Li et al. / International Journal of Pharmaceutics 461 (2014) 540–548

where nt is the number of total pellets and ns is the number of the sank pellets. 2.5.2. In vitro dissolution test Pellets equivalent to 100 mg DIP were studied for dissolution test by USA apparatus II (paddle) (RCZ-6B, Shanghai Huanghai Drug Inspection Instrument Co., Shanghai, China) at 37 ◦ C in 500 ml 0.1 N HCl, FaSSGF or FeSSGF. 5 ml samples were withdrawn and filtered at predetermined time, replacing by fresh dissolution medium with the same volume. The samples withdrawn from 0.1 N HCl medium were analyzed in an ultraviolet spectrophotometer (Beijing Beifenruili Analytic Instrument Co., Beijing, China) at 282 nm, and the samples withdrawn from FaSSGF and FeSSGF media were determined by HPLC method built in Section 2.7.2. All tests were carried out in triplicate. The similarity factor (f2 ) suggested by FDA was used to evaluate the similarity of the dissolution profiles (Pillay and Fassihi, 1998; Shah et al., 1998). f2 = 50 log

⎧ ⎨ ⎩

1+

n 1

n

−(1/2) (Rt − Tt )2

⎫ ⎬

× 100

t=1

⎭

(5)

wherein Rt and Tt stand for the dissolution value at time t of the reference preparation and the test preparation, respectively; n is the number of time points. When the similar factor (f2 ) was not less than 50, the two drug release profiles were considered similar.

543

used in the study were performed in accordance with the guidelines approved by the Ethics Committee of Shenyang Pharmaceutical University. 2.7.2. Sample preparation and analysis of plasma concentration by HPLC 50 ␮l internal standard solution (10 ␮g/ml diazepam methanol solution) was added into 500 ␮l plasma in a 5 ml tube, and mixed by vortex for 30 s. Then 2 ml ethyl acetate was added, and mixed by vortex for 10 min to extract the drug in the plasma. Next it was centrifuged at 4000 rpm for 10 min to separate the ethyl acetate from the plasma. Finally the supernatant was transferred to another 5 ml tube, and dried under nitrogen at 50 ◦ C. The residue was reconstituted by 100 ␮l methanol. 20 ␮l methanol solution was directly injected to HPLC. The HPLC conditions were as follows: the HPLC apparatus was equipped by Shimadzu LC-20AT pump and Shimadzu SPD-M20A detector; separation was achieved by Agela C18 (25 mm × 4.6 mm, 5 ␮m) reverse phase column; the mobile phase was that methanol and pH 5.4 ammonium acetate (7:3, v/v) pumped at a flow rate of 1.0 ml/min; the detection wavelength was 282 nm.

2.5.3. Scanning electron microscopy (SEM) The morphologies of the surface and cross-section of the novel light pellets were examined following the procedure described in Section 2.3.3.

2.7.3. Statistical analysis The analysis of the pharmacokinetic data was conducted by DAS 2.0 software (Mathematical Pharmacology Professional Communities of China, Shanghai, China). The Cmax , Tmax , t1/2 , AUC0–∞ and AUC0–t were calculated. The relative bioavailability (F) was calculated as: F = AUC0–∞,T /AUC0–∞,R . The one-way analysis of variance (ANOVA) was performed to clarify the significance of the difference between the relevant pharmacokinetic parameters of the two preparations.

2.6. Drug release kinetics

3. Results and discussion

The in vitro drug release data of the optimized formulation was evaluated in various mathematical models, namely zero order, first order, Higuchi, and Korsmeyer–Peppas models (Banker and Rhodes, 2002; Hadjiioannou et al., 1993; Higuchi, 1963; Korsmeyer et al., 1983).

3.1. Design of the porous matrix cores

Zero-order Model: F = K0 t, where F represents the fraction of drug released in time t, and K0 is the apparent release rate constant or zero-order release constant. First-order Model: ln(1 − F) = −K1 t, where F represents the fraction of drug released in time t, and K1 is the first-order release constant. Higuchi Model: F = KH tH is the Higuchi dissolution constant. Korsmeyer–Peppas Model: F = KP tn , where F represents the fraction of drug released in time t, KP is the rate constant and n is the release exponent, which indicates the drug release mechanism.

3.1.1. Screening of the water soluble excipient In this study, the water soluble excipient was employed to form the matrix cores of porous structure by their complete release. The air trapped in the porous matrix cores conferred buoyancy for the gastro-floating pellets. The water soluble excipient candidates were the common substances, such as sodium chloride (NaCl), mannitol, lactose and sucrose. The desirable candidate should be able to be released slowly but completely in the distilled water without rupturing the membrane. In the early stage of release, there was a saturated solution in the membrane. So the osmotic pressure of saturated solution should be moderate. By Van’t Hoff’s law, osmotic pressure formula could be calculated as follow:  =i·R·T ·c =

2.7. In vivo pharmacokinetic study in beagle dogs 2.7.1. Protocol A two-period crossover pharmacokinetic study was conducted in beagle dogs. The conventional sustained pellets (CP), which was prepared by exchanging a porous matrix core of GFP with a MCC core, were chosen as the reference preparation. Six healthy beagle dogs, weighing 15–20 kg, were randomly divided into two groups, and they were fasted overnight with free access to water before oral administration of GFP or CP. 2 ml blood was collected at predetermined time points before and after administration GFP or CP, respectively. Then the blood was centrifuged at 4000 rpm for 10 min, and the plasma was transferred into a 5 ml tube. Finally, the plasma was stored at −20 ◦ C for analysis. Between the two periods, there was a wash-out period of one weak at least. All procedures

i · R · T · c i·R·T ·m = MV M

(6)

where  is the osmotic pressure, i is the Van’t Hoff factor, c is the molar or molal concentration, c is the saturated solubility, m is the weight of the dissolved water soluble excipient, M is the relative molecular mass, V is the volume of the solution. For these water soluble excipients, R and T were constant. To compare the osmotic pressure under the saturated condition was to compare the value of i · c /M, which were listed in Table 3 (Brito and Giulietti, 2007; Shimoyama et al., 2009). NaCl with a lower relative molecular had a higher saturated solubility. So when the pellets were composed of NaCl and MCC, the osmotic pressure in the membrane was so large that the membrane was ruptured easily. The others had relatively moderate osmotic pressure under the saturated condition. But the pellets containing lactose or sucrose were easy to stick to each other when dried after dissolution. For

544

Z. Li et al. / International Journal of Pharmaceutics 461 (2014) 540–548

Table 3 The relevant parameters of the soluble excipients. Water soluble excipients

Van’t Hoff factor (i)

The saturated solubility (c , g/l, 25 ◦ C)

The relative molecular mass (M)

Value of ic /M

NaCl Mannitol Lactose Sucrose

2 1 1 1

36.2 49 18.8 55.5

58.44 182.17 342.3 342.3

1.24 0.27 0.05 0.16

these reasons, mannitol was chosen as the desirable water soluble excipient.

3.1.2. Composition and characterization of the matrix membrane The membrane coated on the core beads should be rigid under high osmotic pressure and permeable enough to release mannitol slowly but completely in distilled water; on the other hand, the membrane should be strong enough to withstand the powerful mechanical collision in the fluid bed coater and maintain integrity. EC with high levels of pore forming agent (PVP k30) and coating weight gain was the optimal coating material (Krögel and Bodmeier, 1999; Rouge et al., 1996). Taking the final weight and the mechanical strength index as evaluation indexes, the ratio of EC to PVP k30 and the levels of coating weight gain were investigated, separately. It could be seen from Fig. 2 that the cumulative release amount of mannitol at 12 h raised significantly with the increase of the amount of PVP k30, and decreased with the increase of coating weight gain. This was attributed to that when the level of coating weight gain was kept constant, with the increase of PVP k30, more pores were formed during the dissolution, resulting in that more mannitol was released; and when the amount of PVP k30 was kept constant, with the increase of coating weight gain, the permeability of the membrane was decreased, causing less mannitol was released. Fig. 3 depicted that the yield of matrix cores after fluidized in the fluid bed coater decreased with the increase of the amount of PVP k30, and increased with the increase of the coating weight gain. That was because suitable amount of PVP k30 was necessary for the mannitol release, but too much PVP k30 made the membrane so fragile that high level of coating weight gain was obligatory. Under the condition that the water soluble excipient was completely released in 12 h and not less than 95% of the matrix cores remained integrity, the lower level of the coating weight gain was desired. Thus the optimal composition of matrix cores was that the ration of EC to PVP k30 was 4:1, with a coating weight gain of 20% (w/w).

Fig. 3. Effect of the ratios of EC to PVP k30 and the coating weight gain on the mechanical strength of the porous matrix cores.

3.2. The results of the OAD The range analysis was aimed to confirm the significant level of different factors on the drug release, and those most significant factors could be disclosed basing on the result of range analysis. The layout of the design and the range analysis result of the effect of different factors on the drug release property were showed in Table 4. And the drug release profiles of the OAD formulations were given in Fig. 4. The K value for each level of certain parameter was the average of the three values shown in Table 4, and the range value (R) for each factor was the difference between the maximal and minimal value of the three levels. Based on the results of range analysis, all the investigated influencing factors in the order of significance were: B (the coating level of sub-coating layer), C (the coating level of sustained release layer) and A (the particle size of the matrix cores).

Table 4 Data of range analysis for the drug release property. Trial No.

Factors A

Fig. 2. Effect of the ratios of EC to PVP k30 and the coating weight gain on the final weight of the matrix cores.

R B

C

D

1 2 3 4 5 6 7 8 9

18–24 18–24 18–24 24–30 24–30 24–30 30–40 30–40 30–40

K1 K2 K3

85.12 90.47 55.58

48.13 102.47 80.57

105.18 60.03 65.95

77.41 79.35 74.41

R

34.89

54.34

45.15

4.94

0 5 10 5 10 0 10 0 5

5.5 6 6.5 6.5 5.5 6 6 6.5 5.5

1 2 3 1 2 3 1 2 3

83.92 94.17 109.47 98.29 146.31 34.45 61.43 23.63 85.05

Z. Li et al. / International Journal of Pharmaceutics 461 (2014) 540–548

545

Table 5 Analysis of variance (ANOVA) for the drug release property.

Fig. 4. Release profiles of the formulations of the orthogonal array design.

Source

SS

DF

MS

F

p

A B C Error

235.32 498.39 401.39 4.13

2 2 2 2

117.6599 249.1948 200.6945 2.0657

56.9599 120.6367 97.1574

0.017 0.008 0.010

mean squared deviation (MS) were determined and summarized in Table 6. Through comparing the obtained p value with the theoretical one for specific level, the significant level could be determined for each factor. As shown in Table 5, the coating level of sub-coating layer had shown significant influence on the drug release property at a level of p < 0.01, and all three studied factors showed significant influence on the drug release property at a level of p < 0.05. Considering the results of the range analysis and the analysis of variance, it could be concluded that the most important factor contributing to the drug release was factor B: the coating level of sub-coating layer, 5.0%; followed by factor C: the coating level of sustained release layer, 5.5%; and lastly, factor A: the particle size of the matrix core, 24–30 mesh. 3.3. In vitro drug release study

Fig. 5. The release profiles of GP in 0.1 N HCl, FaSSGF and FeSSGF.

In this experiment, a blank column was set in the orthogonal table to estimate the error; so it was unnecessary to repeat the experiment for each protocol. In analysis of variance (ANOVA), the sum of squares of deviation (SS), degree of freedom (DF) and

To predict the in vivo behavior of the optimized formulation, biorelevant dissolution media such as FaSGF and FeSGF were chosen for GFP, and the dissolution profiles were shown in Fig. 5. Due to the strong pH dependent solubility of DIP, the dissolution rate in FaSGF (pH 1.6) was slower than that in 0.1 N HCl, and no drug was released in FeSGF (pH 5.0). So the fasted stomach was the most favorable state for DIP to dissolve. To analyze the drug release kinetics and mechanism, the in vitro dissolution data of GFP was applied to various mathematical models, such as zero order, first order, Higuchi, and Korsmeyer–Peppas models. The results of the curve fitting into these mathematical models above were given in Table 6. It was found that the in vitro dissolution data was well fitted to the first order mathematical model (R = 0.9978), indicating the drug release was concentration dependent (Banker and Rhodes, 2002). Thus, it could be deduced that the whole process of the drug release was as follows: the dissolution medium flowed in through the semi-permeable membrane, and then the drug was dissolved and released. The initial drug concentration exceeded the drug solubility, so released molecules were replaced by the dissolution of drug crystals, resulting in constant saturated solution inside the membrane. Considering the sink condition was provided throughout the release process, the concentration difference between the inside and outside of the membrane was constant for the first few hours, and then decreased at last. The drug release profiles of the optimal formulation of GFP and CP were illustrated in Fig. 6. The f2 factor was calculated as 55, which meant that in vitro dissolution behaviors of the test and the reference preparations were similar. 3.4. In vitro buoyancy study The floating test was conducted to investigate the floatability of the novel gastro-floating pellets. Fig. 7 showed the floating Table 6 Results of curve fitting of drug release data of GFP in vitro from optimized formulation.

Fig. 6. The release profiles of GFP and CP for the pharmacokinetic study.

Models

Zero order

First order

Higuchi

Korsmeyer–Peppas

R

0.9769

0.9978

0.9961

0.9896

546

Z. Li et al. / International Journal of Pharmaceutics 461 (2014) 540–548

condition of the optimal formulation at different time in 0.1 N HCl, FaSSGF and FeSSGF. All the formulations showed immediate floatability without lag time to float, indicating that the novel pellets owed a density lower than that of the gastric content (1.004 g/ml). The huge volume of air imbibed in the porous matrix cores provided a small enough bulk density to confer an excellent floating property. And during drug release process, few pellets sank. At 12 h, the percentage of floating pellets in the above three media was 96 ± 2.3%, 97 ± 2.6% and 97 ± 1.4%, separately, which indicated that GFP can float on the gastric fluid under either fed or fasted condition. Afterwards, the pellets sank gradually. This might be the consequence of the entrapped air out and the fluid in. Thus GFP complied with the following criteria that the floating device should be easily removed from the stomach, once its purpose had been served (Streubel et al., 2003; Pandey et al., 2012). 3.5. Morphology in SEM The morphology of the surface and the cross-section of the pellets were examined and photographed by the SEM. It was widely known that structure determines function, and the SEM photographs revealed the unique structure behind the excellent floating property. In Fig. 8, A and B showed the morphology of the surface of the matrix core. It could be clearly seen that the matrix core was spherical and there were many pores in the EC membrane, forming by the dissolved PVP k30 in the EC membrane. The pores were the route from which mannitol was released, and the large numbers of pores guaranteed that mannitol was released completely. In Fig. 8,C showed the morphology of the cross-section of the matrix core, and D further showed that the internal structure of the matrix cores which was formed by the left MCC after mannitol was released was porous. Much air entrapped in the porous structure conferred the buoyancy. E and F depicted the surface morphology of GFP. It was spherical and still a little rough. Due to that the surface of the drug layered matrix cores was too rough to perform the coating of the retarding layer; a sub-coating

Table 7 Pharmacokinetic parameters of GFP and CP. Parameters

Unit

GFP

AUC0–t AUC0–∞ t1/2 Tmax Cmax

␮g/ml h ␮g/ml h h h ␮g/ml

18.11 23.56 2.78 8.33 2.09

CP ± ± ± ± ±

3.82 10.87 0.62 0.52 0.37

10.77 12.20 1.49 4.33 1.55

± ± ± ± ±

1.59* 3.17* 0.60* 0.82* 0.25*

All the data are presented in the from of mean ± SD. * p < 0.05.

layer was necessarily added. A sub-coating enhanced the surface smoothness of the GFP. Though it was still a little bumpy, the drug release could be controlled effectively. Also the cross-section of GFP was shown in Fig. 8G and H. In G, It was obvious that there was a porous core inside GFP. Though the hierarchy configuration of the whole coating layer might not be so clear, it still could be seen the loose EC layer, the drug layer and the retarding layer. H further revealed the drug layer, where the drug crystals were apparent. 3.6. Pharmacokinetic study in beagle dogs When both in vitro release and in vitro floating condition were taken into account, the fasted beagle dogs could meet the need of pharmacokinetic study. The plasma concentration–time profiles of GFP and CP were given in Fig. 9, and the relevant pharmacokinetic parameters were listed in Table 7. There was significant difference between the two preparations in Tmax , AUC0–∞ , AUC0–t , t1/2 and Cmax at a level of p < 0.05. It could be clearly seen that in the first 4 h the plasma concentration–time profiles of the two preparations were similar, because that in the first few hours, both GFP and CP retained in the stomach and they had similar in vitro dissolution profiles in the acidic environment. The AUC0–t value of GFP and CP were 18.11 ± 3.82 and 10.77 ± 1.59, respectively. Since DIP was classified as class II drug (high

Fig. 7. Floating sequence in 0.1 N HCl, FaSSGF and FeSSGF of the optimal formulation: (A) 0 min, (B) and (C) 12 h in 0.1 N HCl; (D) 0 min, (E) and (F) 12 h in FaSSGF; (G) 0 min, (H) and (I) 12 h in FeSSGF.

Z. Li et al. / International Journal of Pharmaceutics 461 (2014) 540–548

547

Fig. 8. SEM photomicrographs of (A) and (B) the matrix core, (C) and (D) the cross-section of the matrix core, (E) and (F) GFP, (G) the cross section of GFP, and (H) the drug layer.

permeability and low solubility) according to the BSC (Butler and Dressman, 2010), the high AUC0–t value of GFP might be caused by the sustained drug release and continuous absorption in the stomach. Compared with CP, the Tmax of GFP was delayed from 4.33 ± 0.82 h to 8.33 ± 0.52 h, and the t1/2 was prolonged from 1.49 ± 0.60 h to 2.78 ± 0.62 h. On the other hand, the Cmax of GFP was increased from 1.55 ± 0.25 ␮g/ml to 2.09 ± 0.37 ␮g/ml, in comparison to CP. The changes of the Tmax and Cmax again not only

demonstrated the prolonged gastric retention time (GRT) and sustained drug release, but also the better control over fluctuations in peak plasma concentrations. In addition, the relative bioavailability of GFP calculated by the ratio of AUC0–∞ was 193.11 ± 3.43% to CP. All the results clarified the successful achievement of the prolonged GRT and sustained drug release, indicating that the GFP was an effective approach to prolong the GRT and improve the bioavailability.

548

Z. Li et al. / International Journal of Pharmaceutics 461 (2014) 540–548

Fig. 9. The plasma concentration–time profiles of GFP and CP.

4. Conclusion In conclusion, GFP with excellent gastro-retentive property was formulated and investigated. Mannitol was chosen as the optimal water soluble excipient, and the core beads was composed of MCC and mannitol (1:5, w/w), which was coated with EC (poring by PVP k30, 5:1, w/w). The porous matrix core was achieved by left MCC after the complete release of mannitol in the distilled water. Then the matrix core (24–30 mesh) was coated with a drug loaded layer (DIP and HPMC), a sub-coating layer (HPMC) and a retarding layer (Eudragit® NE 30D). The optimized formulation could sustain the drug release for 12 h and float on dissolution medium for at least 12 h without lag time to float. The dissolution data well fitted to the first order drug release model, indicating the drug release was concentration dependent. It could be clearly seen from the SEM photomicrographs that the matrix core was porous with lots of air entrapped to confer buoyancy. According to the result of the pharmacokinetic study in the fasted beagle dogs, high AUC0–t , prolonged Tmax and increased Cmax together demonstrated the prolonged gastric retention and sustained drug release, resulting in enhanced bioavailability. To sum up, GFP was a promising strategy of developing drugs that acted locally in the proximal GIT, or only had a high solubility at low pH value, or unstable in alkaline environment to get retained in stomach. Further study will be performed to explore the ideal matrix core film material, which can meet the demand well with lower level of coating weight gain, and comparison of this system with other floating dosage forms still needs investigation. Acknowledgment This project is supported by special construction projects fund which belongs to “Taishan Scholar—Pharmacy Specially Recruited Experts”. References Arora, S., Ali, J., Ahuja, A., Khar, R.K., Baboota, S., 2005. Floating drug delivery systems: a review. AAPS PharmSciTech 6, E372–E390. Banker, G.S., Rhodes, C.T., 2002. Modern Pharmaceutics, Fourth Edition: Basic Principles and Systems. Marcel Dekker Incorporated. Bardonnet, P.L., Faivre, V., Pugh, W.J., Piffaretti, J.C., Falson, F., 2006. Gastroretentive dosage forms: overview and special case of Helicobacter pylori. J. Control. Release 111, 1–18. Birajdar Shivprasad, M., Darveshwar Jagdeep, D., 2005. Development and evaluation of floating-mucoadhesive dipyridamole tablet. Asian J. Pharm. Res. Health Care 4, 78–89.

Brito, A.B.N., Giulietti, M., 2007. Study of lactose crystallization in water–acetone solutions. Cryst. Res. Technol. 42, 583–588. Butler, J.M., Dressman, J.B., 2010. The developability classification system: application of biopharmaceutics concepts to formulation development. J. Pharm. Sci. 99, 4940–4954. Chun, M.-K., Sah, H., Choi, H.-K., 2005. Preparation of mucoadhesive microspheres containing antimicrobial agents for eradication of H. pylori. Int. J. Pharm. 297, 172–179. Clarke, G.M., Newton, J.M., Short, M.D., 1993. Gastrointestinal transit of pellets of differing size and density. Int. J. Pharm. 100, 81–92. Davis, S.S., 2005. Formulation strategies for absorption windows. Drug Discovery Today 10, 249–257. Gröning, R., Cloer, C., Georgarakis, M., Müller, R.S., 2007. Compressed collagen sponges as gastroretentive dosage forms: in vitro and in vivo studies. Eur. J. Pharm. Sci. 30, 1–6. Guan, J., Zhou, L., Nie, S., Yan, T., Tang, X., Pan, W., 2010. A novel gastric-resident osmotic pump tablet: in vitro and in vivo evaluation. Int. J. Pharm. 383, 30–36. Hadjiioannou, T.P., Christian, G.D., Koupparis, M.A., Macheras, P.E., 1993. Quantitative Calculations in Pharmaceutical Practice and Research. Wiley. Higuchi, T., 1963. Mechanism of sustained-action medication. Theoretical analysis of rate of release of solid drugs dispersed in solid matrices. J. Pharm. Sci. 52, 1145–1149. Johnson, F.A., Craig, D.Q.M., Mercer, A.D., Chauhan, S., 1997. The effects of alginate molecular structure and formulation variables on the physical characteristics of alginate raft systems. Int. J. Pharm. 159, 35–42. Klausner, E.A., Lavy, E., Friedman, M., Hoffman, A., 2003. Expandable gastroretentive dosage forms. J. Control. Release 90, 143–162. Korsmeyer, R.W., Gurny, R., Doelker, E., Buri, P., Peppas, N.A., 1983. Mechanisms of solute release from porous hydrophilic polymers. Int. J. Pharm. 15, 25–35. Kotreka, U.K., Adeyeye, M.C., 2011. Gastroretentive floating drug-delivery systems: a critical review. Crit. Rev. Ther. Drug Carrier Syst. 28, 47–99. Krögel, I., Bodmeier, R., 1999. Floating or pulsatile drug delivery systems based on coated effervescent cores. Int. J. Pharm. 187, 175–184. Mahony, C., Wolfram, K.M., Cocchetto, D.M., Bjornsson, T.D., 1982. Dipyridamole kinetics. Clin. Pharmacol. Ther. 31, 330–338. Pandey, A., Kumar, G., Kothiyal, P., Barshiliya, Y., 2012. A review on current approaches in gastro retentive drug delivery system. Asian J. Pharm. Med. Sci. 2, 61–77. Park, K., Robinson, J.R., 1984. Bioadhesive polymers as platforms for oral-controlled drug delivery: method to study bioadhesion. Int. J. Pharm. 19, 107–127. Pillay, V., Fassihi, R., 1998. Evaluation and comparison of dissolution data derived from different modified release dosage forms: an alternative method. J. Control. Release 55, 45–55. Reddy, L.H., Murthy, R.S., 2002. Floating dosage systems in drug delivery. Crit. Rev. Ther. Drug Carrier Syst. 19, 36. Rouge, N., Buri, P., Doelker, E., 1996. Drug absorption sites in the gastrointestinal tract and dosage forms for site-specific delivery. Int. J. Pharm. 136, 117–139. Senthil, A., Hardik, T., Ravikumar, Narayana Swamy, V.B., 2011. Design and evaluation of gastroretentive floating tablets of dipyridamole influence of formulation variables. Asian J. Biochem. Pharm. Res. 1, 182–190. Shah, V., Tsong, Y., Sathe, P., Liu, J.-P., 1998. In vitro dissolution profile comparison—statistics and analysis of the similarity factor, f2 . Pharm. Res. 15, 889–896. Shimoyama, Y., Higashi, H., Tsuzaki, S., Okazaki, F., Kitani, Y., Iwai, Y., Arai, Y., 2009. Effect of cation species on solubilities of metal chlorides in water vapor at high temperatures and pressures. J. Supercrit. Fluids 50, 1–5. Streubel, A., Siepmann, S., Bodmeier, R., 2003. Multiple unit gastroretentive drug delivery systems a new preparation method for low density microparticles. J. Microencapsulat. 20, 329–347. Streubel, A., Siepmann, J., Bodmeier, R., 2006. Drug delivery to the upper small intestine window using gastroretentive technologies. Curr. Opin. Pharmacol. 6, 501–508. Strusi, O.L., Sonvico, F., Bettini, R., Santi, P., Colombo, G., Barata, P., Oliveira, A., Santos, D., Colombo, P., 2008. Module assemblage technology for floating systems: in vitro flotation and in vivo gastro-retention. J. Control. Release 129, 88–92. Sungthongjeen, S., Sriamornsak, P., Puttipipatkhachorn, S., 2008. Design and evaluation of floating multi-layer coated tablets based on gas formation. Eur. J. Pharm. Biopharm. 69, 255–263. Terhaag, B., Donath, F., Le Petit, G., Feller, K., 1986. The absolute and relative bioavailability of dipyridamole from different preparations and the in vitro-in vivo comparison. Int. J. Clin. Pharmacol. Ther. Toxicol. 24, 298–302. Udaya, K., Kotreka, M.C.A., 2011. Gastroretentive floating drug-delivery systems a critical review. Crit. Rev. Ther. Drug Carrier Syst. 28, 47–99. Zhang, Z.-h., Nie, S.-f., Wang, Q.-m., Peng, B., Liu, H.-f., Pan, W., 2009. Absorption kinetics of dipyridamole in rat gastrointestinal. Chin. Pharm. J. 44, 1229– 1233. Zhang, C., Xu, M., Tao, X., Tang, J., Liu, Z., Zhang, Y., Lin, X., He, H., Tang, X., 2012. A floating multiparticulate system for ofloxacin based on a multilayer structure: in vitro and in vivo evaluation. Int. J. Pharm. 430, 141–150. Zhou, R., Moench, P., Heran, C., Lu, X., Mathias, N., Faria, T.N., Wall, D.A., Hussain, M.A., Smith, R.L., Sun, D., 2005. pH-dependent dissolution in vitro and absorption in vivo of weakly basic drugs: development of a canine model. Pharm. Res. 22, 188–192.