Formulation of spray-dried phenytoin loaded poly(ε-caprolactone) microcarrier intended for brain delivery to treat epilepsy

Materials

Poly(ε-caprolactone) (PCL, Tone® Polymer P-757) was kindly provided by Union Carbide (Danbury, CT). This polymer was used without further purification and processing. The following chemicals were purchased from commercial suppliers: Phenytoin, span 40, sodium azide and phosphate-buffered saline (PBS, pH 7.4) from Sigma Chemical Co. (St. Louis, MO); dichloromethane (DCM), reagent grade from VWR International, Inc. (West Chester, PA); Slide-A-Lyzer Dialysis Cassettes, 7K MWCO, 3–12 mL Capacity from PIERCE Chemical Co. (Rockford, IL). All chemicals were reagent grade and used as purchased without further purification. Distilled water was prepared with a Mega-Pure™ system from Cole-Palmer (Vernon Hills, IL).

Preparation of PHT Containing Spray-Dried Microcarrier

The PCL microcarriers preparation was adapted from the work of Baras et al.20 Briefly, PCL and span 40 were dissolved in DCM; PHT was dissolved in water. The two solutions were mixed and then emulsified using an Ultraturrax model T25 (IKA Laboratory Technology, Staufen, Germany) at high speed (11000–16000 rpm for 5 min) and room temperature. Microcarriers were obtained by spraying this emulsion through the nozzle of a Buchi Mini Spray Dryer Model B-290 (Buchi Laboratoriums—Technik AG, Flawil, Switzerland). Typically, the process parameters were set as follows: inlet temperature (50°C); pump setting (25); aspirator (setting 80); spray flow (30 mmHg and 357 L/h). The solid microcarriers that had precipitated into the bottom collector (32–33°C outlet temperature) were collected and kept at room temperature.

The Orthogonal Experimental Design

In these studies, the effects of formulation variables on microcarriers characteristics such as particle size and PHT entrapment were investigated. These factors included: (a) the presence of an emulsion stabilizer (span 40) with variable concentration in the organic phase, (b) the concentration of the polymer solution, (c) the concentration of the PHT solution, and (d) the ratio of organic/aqueous phase volumes. In order to obtain the optimal synthesis conditions, experiments have been orthogonally designed for arranging the four factors above with three levels for each factor as shown in Table 1. Nine different sets of experiments have been conducted under conditions of different parameter combinations according to the standard L₉(3⁴) table as shown in Table 2. Each experiment was repeated twice.

Characterization of PHT Containing Spray-Dried Microcarriers

Determination of Amount of PHT Encapsulated in PCL-Microcarriers

The HPLC method was adapted from Potschka and Loscher.13 To determine the amount of PHT encapsulated in PCL microcarriers, a 3-mg sample of particle powder was dissolved in 1 mL DCM and 10 mL acetonitrile–water (50:50, v/v) was then added after the complete evaporation of DCM. The solution was filtered into a vial and 150 µL of the solution was extracted for the HPLC (Varian, Inc., ProStar 330, Palo Alto, CA) measurement using a C₁₈ column with UV detection at 220 nm by an autoinjector (Varian Model 410). The mobile phase consisting of a mixture of acetonitrile and water (40:60, v/v) was delivered through the HPLC column at a flow rate of 1.0 mL/min. Each experiment was repeated three times. The percentage of PHT loading within the PCL microcarrier was then calculated from the following equation:21

(1)

Differential Scanning Calorimetry (DSC)

DSC was used to study the blending state of the drug and polymer matrix by observing the melting peaks of the sample. DSC scans were performed using a Mettler–Toledo differential scanning calorimeter DSC823e with a Julabo FT100 intracooler and nitrogen purge gas. Samples ranged from 1.3 to 2.5 mg and were sealed in aluminum pans. DSC scans from −10°C to 450°C at heating rates of 40°C/min. The temperature of the DSC was calibrated with indium and lead, and the heat flow of the DSC was calibrated with indium. At the melting temperature of the sample, the physical structure of the sample will transform and the sample starts to melt; an endothermic melting peak will appear in the DSC thermogram illustrating that there is additional heat flowing to the sample to supply the heat of melting. The area of the melting peak is related to the heat of fusion and degree of crystallinity of the sample. By studying the location and width of the melting peak, the crystallinity of the sample can be examined. The melting points are taken to be the onset of melting as is the convention in DSC. The heat of fusion (ΔH) reported for PCL is the value measured normalized by the percent PCL in the sample. The percent crystallinity of the PCL phase is determined by

(2)

where ΔH_U is the enthalpy of fusion for a perfect crystal and is 142.9 J/g for PCL.22

Powder X-Ray Diffraction Analysis

XRD of the fine powders of the samples were carried out with a Panalytical (former Philips-Norelco) powder diffractometer with CuKα-radiation (λ = 0.154184 nm) generated at 40 kV and 20 µA and monochromotized with a graphite crystal. The fine powders were scanned from 4° 2Θ to 44° 2Θ which was the range showing the most important X-ray reflections for these samples.

In Vitro Release Study

The in vitro release behavior of PHT from PCL-microcarrier sample was investigated as follows: 60 mg sample of drug loaded spray-dried PCL microcarriers was suspended in 4 mL of 0.1 M PBS (pH 7.4) containing sodium azide 0.01%(w/v), and then was placed into the Slide-A-Lyzer Dialysis Cassettes that was immersed into 250 mL 0.1 M PBS (pH 7.4) containing sodium azide 0.01%(w/v). The system was shaken at 60 rpm and temperatures 37.4°C. At predetermined intervals, 300 µL PBS solution was taken out from the 250 mL immersion medium and replaced by the same volume of fresh medium. The PHT content of the release solution was measured by HPCL using a C₁₈ column with UV detection at 220 nm.13 All release experiments were carried out in triplicate, and averages and standard errors of the mean are presented in Figure 7.

Determination of the Optimal Parameters

The analysis of preparation variables (the carrier size) is shown from Tables 2–3, and for drug loading is shown from Table 4 and Figure 1. As it is noteworthy, the size of microcarriers was remarkably affected by the organic/aqueous phase ratio. It was also affected by the amount of stabilizer. However, the concentration of polymer solution and the concentration of the PHT had a much less significant influence on particle size, but their impacts on the PHT loading of microcarriers cannot be overlooked. For the intended efficient brain delivery of microcarriers, our goal was to obtain small size microcarriers with high drug loading. For parameter A (concentration of polymer solution), the size of A1 was the smallest and the amount of PHT loading was also higher than that of A2 and A3 (Fig. 1), therefore A1 was chosen as one of the best preparation conditions. For parameter B (conc. of the PHT), the particle size with B1 was the smallest, but the drug loading amount was low (about 0.5%), which is not suitable for the successful application for brain delivery. The particle size with B2 was smaller than that of B3, and the loading amount was not different. So B2 was chosen. For parameter C (ratio of organic/aqueous phase), the size with C1 was the smallest, but the drug loading amount is low (about 1.0%), which is not suitable for the intended application into the brain. Although the drug loading amount of C3 was higher than that of C2, the size of C2 was smaller than that of C3. So the best parameter was C2. The particle sizes with D2 and D3 were not different, and they are smaller than that of D1, and the loading amount of D1, D2, and D3 were not remarkably different, leading us to choose D2. To summarize, A1B2C2D2 (i.e., 0.1% PCL, 1.0% PHT, 100/5 phase ratio, and 0.5% stabilizer) was the best process parameters combination. In the confirmation experiments under conditions of the best parameters combination, the particle size of the sample was about 4.0 µm (as shown in Fig. 2), and the loading amount was about 17.5% w/w.

Table 2. Experimental Results and Analysis of the Orthogonal Design L₉(3⁴)

No.	A	B	C	D	Particle Size (µm)			PHT Loading (% w/w)
	1	2	3	4	I	II	(I + II)/10	Mean Value
1	1	1	1	1	4.01	4.21	0.822	0.61
2	1	2	2	2	4.69	4.51	0.920	18.8
3	1	3	3	3	4.62	4.80	0.942	21.7
4	2	1	2	3	4.41	4.61	0.902	0.48
5	2	2	3	1	4.85	5.11	0.996	14.1
6	2	3	1	2	3.90	4.12	0.802	3.11
7	3	1	3	2	4.90	4.80	0.970	0.81
8	3	2	1	3	3.99	4.11	0.810	0.40
9	3	3	2	1	5.01	5.21	1.022	8.55
K1	2.684	2.694	2.434	2.840
K2	2.700	2.726	2.844	2.692
K3	2.802	2.766	2.908	2.654
R	22.3451	22.3395	22.4692	22.3562
R/3m	3.7242	3.7233	3.7449	3.7260
Sj	0.0014	0.0004	0.0220	0.0032

Effects of Each Factor on the PCL Microcarrier Formulation

Concentration of the PCL Solution in the External Organic Phase

An increase in the initial polymer concentration (from 0.1% to 1.0%, w/v) resulted in a significant decrease in PHT loading (Fig. 1) and was associated with an increase of the particle size (Tab. 2). These results could be related to the increased viscosity and density of the emulsion or fluid, which increased significantly with the polymer concentration.23 These factors would also decrease the air/fluid mass ratio (Eqs. 3 and 4) or decrease of the energy of atomization resulting in coarser spray-droplets and leading overall to an increase in the mean size of solid microcarriers.23 Moreover, a considerable loss of materials occurred during spray-drying. This loss depended on the technical features of the apparatus: much of the powder adhering to the cyclone walls was lost during spray-drying and was related to the polymer concentration as well.

The coupling of liquid feeding rate (V_lf) with atomizing air velocity (V_aa) and their influence on spray droplet size is defined by the so-called air/fluid mass ratio (n_a/f), which represents the energy available for atomisation:24

(3)

The diameter of a solid produced by spray-drying, D_s, is dependent on the diameter of the original spray-droplet, D_w, and also the total solids' content of the spray solution, w₂.

(4)

where ρ₂ is the density of the solid.24

The Morphology and Size of PHT Containing PCL Microcarriers

Particles prepared by spray-drying normally have a good spherical structure and uniform size distribution. A STEM of the PHT containing PCL microcarriers is shown in Figure 3 for the confirmation experiment under conditions of the best parameters combination (i.e., 0.1% PCL, 1.0% PHT, 100/5 phase ratio, and 0.5% stabilizer) and shows that the PCL-microcarriers have good spherical structure. SEM of the PHT containing PCL microparticles also shows spherical morphology and mainly uniform size (Fig. 4). The particle sizes are shown in Figure 2 and indicate that 100% of the PHT containing PCL microcarriers were in the range of 3.5–4.4 µm, with very narrow size distribution. The mean diameter of these particles is 4.3 ± 0.1 µm (n = 3), 90% of these particles had a size d₉₀ = 3.5 ± 0.1 µm (n = 3), 50% of these particles had a size d₅₀ = 3.9 ± 0.2 µm (n = 3).

DSC Analysis

(5)

where T_m(r) is the melting point of a spherical particle with radius r, T_m(∞) is the bulk melting temperature, ΔH_f(∞) is the bulk latent heat of fusion, ρ_s is the solid phase density, and σ_sl is the solid–liquid interfacial energy. The 40°C depression in the control sample of PHT and PCL indicates the magnitude of the effect of PCL on the melting point. The much larger melting point for PHT in the microcarrier indicates that the size of PHT particles dispersed in the microcarrier is significantly smaller than that of native PHT powder although the dispersion is not at the level of a molecular dispersion.

PXRD Analysis

Powder X-ray patterns of PHT, PCL, their mixture, before the processing, and one of the processed samples (compositions are given in Tab. 6) are displayed in Table 7 and Figure 6. In terms of 2-theta angles, the X-ray diffractogram of PHT powder (Fig. 6), sharp peaks at diffraction angle 2θ, 2.78°, 3.29°, 5.58°, 7.57°, 10.99°, 11.24°, 11.92°, 13.02°, 14.34°. The highest peak and characteristic peak of PHT was the one at 2θ, 10.99°. These observations were similar to the previously reported diffractogram of PHT.29 These suggested that the drug was present as crystalline material. In the X-ray diffractogram of blank PCL powder, a combination of large diffraction peaks with sharp peaks at diffraction angle 2θ, 10.70°, 10.95° and 11.85° indicated that the polymer is semicrystalline material. The highest peak and characteristic peak of PCL was the one at 2θ, 10.74°. These observations were similar to the previously reported diffractogram of PCL.30 In the PHT containing microcarrier, the characteristic peak of PHT at 2θ, 10.99 slightly overlap with the PCL peak at 2θ, 10.70° suggesting that the drug was encapsulated in the crystalline state within the PCL matrix.

As angles vary with the type of radiation, it is also important to discuss these data in term of d-value. The strong reflections of PHT at 0.404 and 0.395 nm were useful for examining the effects of the mixing with PCL (sample ZH7) and after the processing the mixtures (ZH1, ZH2, ZH3, and ZH4). The polymer PCL blank has a PXRD pattern with only two broad but well-defined reflections at 0.414 and 0.375 nm (Tab. 7). The patterns in Figure 6 clearly shows that PCL has a crystallinity much lower than that of PHT. However the strong 0.404 nm reflections (marked with an asterisk in Tab. 7) can serve as a good indicator about the state of PHT after processing. The XRD pattern of the mixture (ZH7) before processing and that of the processed mixture (ZH3) are shown in Figure 6 and they are indistinguishable. This clearly shows that the PHT has not changed during the process. This is also true for all the other three XRD patterns for the processed mixtures ZH1, ZH2, and ZH4 as the data in Table 1 indicate. Thus, PHT seems to maintain its crystallinity during these processes. This XRD observation was consistent with the DSC data analysis.

The crystal structures of PHT (diphenylhydantoin) was reported as sodium salt and crystallized as the free acid from aqueous solution at pH 11. The crystals are orthorhombic and the defining characteristics of the respective unit cells are: a = 6.230, b = 13.531, c = 15.523 Angstrom; space group Pn21a with Z = 4 molecules per unit cell.31 Moreover, the stereochemical similarities and three-dimensional conformations of AEDs have been suggested to be having implication in their bioactivities.32 Therefore, the conservation of the crystalline structure of drug inside the PCL system suggested a preservation of the PHT bioactivity that remains to be demonstrated in biological systems in future studies.

In Vitro Release Profile of PHT Containing Spray-Dried PCL Microcarriers

PHT release from PCL-microcarriers was investigated in PBS (pH 7.4) at temperature 37.4°C. PHT release from the PCL-microcarrier samples generally showed multi-stage release behavior, as shown in Figure 7. During the first 4 days, release was linear with about 1.5 mg PHT released per day. The release occurred in the first stage may be due to the dissolution of the PHT adsorbed at the external surface of PCL-microcarriers. These results are consistent with Dong's report.33 During the following 8 days, the release rate decreased to about 0.4 mg/day and then it approached complete (80%) of the PHT release from the microcarriers. The release rate in this stage became much slower than that in the first stage, and we hypothesize that release at this point was a consequence of diffusion through pores generated in the dosage form, which was formed by the coalescence of some PHT crystals, since they might stick to each other when the PCL suspensions containing PHT was spray-dried. At the last stage, the PHT release almost not observable, which might have been PHT strongly isolated within the delivery matrix, and the release rate at this stage may be mainly controlled by the decomposition kinetics of PCL, leading to very slow and sustained drug release.