- Email: [email protected]
Development of solid dispersions of artemisinin for transdermal delivery
International Journal of Pharmaceutics 457 (2013) 197–205
Contents lists available at ScienceDirect
International Journal of Pharmaceutics journal homepage: www.elsevier.com/locate/ijpharm
Development of solid dispersions of artemisinin for transdermal delivery Yasser Shahzad a,∗ , Sadia Sohail b , Muhammad Sohail Arshad b , Talib Hussain a , Syed Nisar Hussain Shah b a b
Pharmacy and Pharmaceutical Science, University of Huddersfield, Queensgate, Huddersfield HD1 3DH, UK Faculty of Pharmacy, Bahauddin Zakariya University, Multan 68000, Pakistan
a r t i c l e
i n f o
Article history: Received 14 July 2013 Received in revised form 20 September 2013 Accepted 22 September 2013 Available online 29 September 2013 Keywords: Artemisinin Solid dispersions Solubility parameter Miscibility Molecular dynamics Permeability coefficient
a b s t r a c t Solid dispersions of the poorly soluble drug artemisinin were developed using polymer blends of polyvinylpyrrolidone (PVP) and polyethylene glycol (PEG) with the aim of enhancing solubility and in vitro permeation of artemisinin through skin. Formulations were characterised using a combination of molecular dynamics (MD) simulations, differential scanning calorimetry (DSC), X-ray diffraction (XRD) and Fourier transform infrared spectroscopy (FT-IR). Solubility of artemisinin was determined in two solvents: de-ionised water and phosphate buffered saline (PBS; pH 7.4), while in vitro drug permeation studies were carried out using rabbit skin as a model membrane. MD simulations revealed miscibility between the drug and polymers. DSC confirmed the molecular dispersion of the drug in the polymer blend. Decrease in crystallinity of artemisinin with respect to polymer content and the absence of specific drug–polymer interactions were confirmed using XRD and FT-IR, respectively. The solubility of artemisinin was dramatically enhanced for the solid dispersions, as was the permeation of artemisinin from saturated solid-dispersion vehicles relative to that from saturated solutions of the pure drug. The study suggests that high energy solid forms of artemisinin could possibly enable transdermal delivery of artemisinin. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Malaria is the major and the most prevalent public health problem in many parts of the world (Chamchod and Beier, 2013). According to WHO, about half of the world’s population is exposed to the risk of malaria (World Health Organization, 2010a) and it is the foremost cause of morbidity and mortality in the world with a million of deaths every year (World Health Organization, 2010a; Eastman and Fidock, 2009). Resistance to conventional antimalarial drugs has led to changes in malaria control policies globally in favour of artemisinin. Currently, the treatment of choice for uncomplicated and severe malaria is based on administration of artemisinin and its derivatives (Shahzad et al., 2011), either used alone or in combination with other drugs (Eastman and Fidock, 2009; World Health Organization, 2010b). Average plasma concentration of artemisinin after single oral dose of 500 mg tablet formulation range between 400 and 700 ng/mL (Gordi et al., 2000). Artemisinin (Fig. 1) is a parent compound of a novel family of antimalarials extracted from the Chinese traditional plant, Artemisia annua, L. Asteraceae. It is effective against malaria
∗ Corresponding author. Tel.: +44 7910427497. E-mail addresses: [email protected], [email protected] (Y. Shahzad). 0378-5173/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ijpharm.2013.09.027
parasites, including the multidrug-resistant falciparum species due to its high potency and rapid onset of action (Klayman, 1985; Svensson et al., 1999; Thaithong and Beale, 1985). Structurally, it is a sesquiterpene lactone with an inner peroxide bridge which is responsible for its antimalarial activity. It is characterised as a poorly water-soluble drug with an octanol–water partition coefficient greater than 2 and a short half-life of 2–3 h, and is extensively metabolised by the liver. Thus, oral bioavailability can be as low as 32% (Sahoo et al., 2011; Svensson et al., 1999; Wong and Yuen, 2001). Although artemisinin has shown excellent permeability across the intestinal mucosa, it has low bioavailability because of its poor aqueous solubility, which can adversely affect its efficacy (Titulaer et al., 1991). This makes artemisinin a suitable candidate for transdermal drug delivery, which circumvents the first pass metabolism by the liver. Solid dispersion of an insoluble drug in hydrophilic polymers is an attractive technique for enhancing drug solubility and the dissolution rate, which for rapidly absorbing molecules such as artemisinin could enhance their drug bioavailability (Lima et al., 2011); (Craig, 2002). Most solid dispersions are prepared using highly water-soluble polymers as the carrier, where the polymer can be amorphous (for example polyvinylpyrrolidone; PVP) or partially crystalline (for example polyethylene glycol; PEG) (Zhu et al., 2012). Solid dispersions are generally prepared by either a solvent
198
Y. Shahzad et al. / International Journal of Pharmaceutics 457 (2013) 197–205
Leicestershire, UK); vacuum grease (Dow Corning, Midland, Michigan, USA). De-ionised water was used throughout to make up all solutions. 2.2. Quantitative analysis of artemisinin
Fig. 1. Structure of artemisinin.
method, whereby the drug and carrier are dissolved in a mutual solvent followed by the removal of solvent by evaporation, or by a melting method, whereby drug–carrier mixtures are prepared by co-melting/cooling (Bley et al., 2010). Upon dissolution of the solid dispersion in an aqueous medium the carrier will dissolve rapidly, releasing very fine colloidal drug particles of very high surface area resulting in dissolution rate enhancement (Serajuddln, 1999). The primary issue with solid dispersions is re-crystallisation of the drug on storage and also immediately on dissolution, which can be addressed by judicious choice of stabilising polymer(s) taking into account the drug–polymer intermolecular interactions, the viscosity of the polymer, and the glass transition temperature of the solid dispersion (Andrews et al., 2010). In the present study we investigate the development of solid dispersions of artemisinin using polyvinylpyrrolidone (PVP) grade PVP-K30 and polyethylene glycol (PEG) 4000 aimed at enhancing in vitro permeation of artemisinin across skin. Both PVP and PEG are polymers of choice for solid dispersions. PVP is an amorphous water-soluble polymer with a glass transition temperature between 110 and 180 ◦ C, while PEG is semicrystalline with a low melting point and is also water-soluble. We show that the permeability of artemisinin from the solid dispersions in enhanced in excess of 10 fold relative to that of a saturated solution of the original drug. The physical state of the drug in the solid dispersions was characterised through X-ray diffraction, Fourier-transform infra-red spectroscopy, and differential scanning calorimetry. The stability of the solid dispersions is linked with the drug–polymer and polymer–polymer (for mixed polymer systems) miscibility, which can be characterised by the solubility parameter. The solubility parameters are typically estimated by a group contribution method (Van Krevelen and Hoftyzer, 1976), but such an approach is challenged by directional hydrogen bonding and long-range electrostatic interactions (Langer et al., 2003). Here we employ the more rigorous approach of molecular dynamics (MD) simulation, which accounts for the individual atomic interactions, to estimate the solubility parameters, following the work of (Gupta et al., 2011). The results reveal that artemisinin and the polymers PVP and PEG are relatively miscible with each other encouraging the formation of a molecular dispersion of the drug in the polymer matrix and its stability, which indeed rationalises the enhanced solubility and permeation observed from the dispersions. 2. Materials and methods 2.1. Materials The following chemicals were used as purchased: artemisinin 99.9% purity (Alchem, New Delhi, India); PVP-K30 and PEG-4000 (Beijing chemical company, Beijing, China); methanol HPLC grade 99%+ (Merck, Darmstadt, Germany); potassium dihydrogen phosphate (VWR, Lutterworth, Leicestershire UK); sodium chloride and potassium chloride (Sigma–Aldrich, Poole, Dorset, UK); di-sodium hydrogen phosphate (Fischer Scientific Chemicals, Loughborough,
Artemisinin was quantified using HPLC method with UVdetection as described previously (Sahoo et al., 2009; Zhao and Zeng, 1985), in which artemisinin is transformed into a UVabsorbing compound through an alkali reaction, i.e. by heating the solution with 0.2% NaOH solution. The alkali reaction was carried out by adding 5 mL of 0.2% NaOH solution to each sample, then heated at 50 ± 1 ◦ C for 30 min and allowed to cool down in a refrigerator. Finally, 1 mL glacial acetic acid was added to each sample before injecting into the HPLC system. Briefly, the HPLC system comprised of a Severn Analytical solvent delivery system (SA 6410B), a Waters 2487 UV-detector (Waters, Hertfordshire, UK), a Rheodyne injector (Perkin Elmer 7725) fitted with a 20 L sample loop. The column used was Novapack C18 (4.6 cm × 15 cm, 4 m) (Waters, Hertfordshire, UK). The composition of mobile phase was 0.01 M di-sodium hydrogen phosphate (75%) and acetonitrile (25%) and the final pH was adjusted to 6.5 using glacial acetic acid. The mobile phase was pumped through the system with a flow rate of 0.8 mL/min. The UV-detector was set at a wavelength of 290 nm. 2.3. Formulation of solid dispersions Solid dispersions of artemisinin in two hydrophilic polymers, namely PVP and PEG were prepared by conventional solvent evaporation and lyophilisation methods as described previously (Shahzad et al., 2012). Solid dispersion were prepared with different drug to polymer ratios, i.e. 6:4, 5:5, 3:7, 2:8, and 1:9. It should be noted that the quantities of PVP and PEG were kept equal (i.e. 50:50 wt%) with respect to each other in all the formulations. For conventional dispersions (CD), accurately weighed quantities of PVP and PEG were dissolved in 100 mL of methanol in a pre-cleaned vessel. This was followed by the addition of accurately weighed quantities of artemisinin and allowed to dissolve completely by continuous stirring at ambient temperature. Methanol was evaporated on a rotary evaporator at reduced pressure. The resulting residue was dried under vacuum in a desiccator for 24 h at ambient temperature and the dried mass obtained was finally pulverised in a pre-cleaned pestle and mortar for about 5 min until a homogeneous mixture was obtained and then refrigerated in a closed vial at 5 ◦ C until further investigations. For lyophilised dispersions (LD), each of the drug–polymer mixture of respective ratio was dissolved in methanol until a clear solution was obtained. This solution was quickly solidified by immersing the flask in liquid nitrogen. Upon cooling, the flask was attached to the vacuum adapter of the lyophiliser for sublimation. After the solvent was completely removed, a porous and fluffy powder residue appeared that was kept in a refrigerator at 5 ◦ C until further investigations. The prepared formulations were subjected to drug content analysis. Three random samples of 10 mg drug equivalent from each formulation were dissolved in methanol and appropriately diluted and the drug content was determined by HPLC analysis after alkali reaction. The drug content in the solid dispersions prepared from both methods was found to be in the range of 88–91%. 2.4. Characterisation of dispersions 2.4.1. Drug–polymer miscibility determination A key consideration for developing a physically stable dispersion is the miscibility of drug in the polymer matrix. The cohesive
Y. Shahzad et al. / International Journal of Pharmaceutics 457 (2013) 197–205
199
Fig. 2. Energy minimised amorphous simulation cell of artemisinin.
energy density (CED) and Hansen’s solubility parameter (ı) are important parameters in this regard, which were estimated using molecular dynamics (MD) simulation. The MD simulations were performed using Materials Studio 6.0 (Accelrys, San Diego, CA) software. The artemisinin structure was imported from the Cambridge Structural Database (RSC, UK) while the homopolymer structure of PVP (chain of 10 monomers) and PEG (chain of 30 monomers) were constructed using the Build module within Materials Studio. Energy minimised (steepest decent/conjugate gradient) amorphous simulation cells for artemisinin (50 molecules), PVP (6 chains), PEG (3 chains) and polymer blend (2 PVP/4 PEG chains of 15 monomers each) were constructed as exemplified in Fig. 2. The ab initio COMPASS force field was assigned for the MD simulations (Rigby et al., 1997). The amorphous simulation cells were equilibrated under isothermal–isobaric (NPT) conditions using the Anderson thermostat (298 K) and barostat (1 bar) at a time step of 1 fs for 0.5 ns. The equilibrated cells were further subjected to dynamics for 200 ps simulation time and trajectories were captured every 10 ps for subsequent analysis. The non-bonded van der Waals and Coulomb interactions were truncated using a group based cut-off distance ˚ The captured trajectories were used for CED and ı calcuof 12.5 A. lation using Forcite molecular dynamics module within Materials Studio.
were obtained using a scan speed of 4◦ /min with 2 between 10◦ and 50◦ .
2.4.2. Differential scanning calorimetry (DSC) The DSC curves of pure artemisinin and dispersion system were recorded on a Mettler Toledo® DSC 1 Star thermal analyser. Thermal behaviour was studied by heating samples (5–6 mg) in a sealed aluminium crucible, using an empty crucible as reference, over the temperature range of 25–200 ◦ C at a rate of 10 ◦ C/min under constant nitrogen flow at a rate of 50 mL/min. Percentage crystallinity of the samples was calculated using the following equation (Rawlinson et al., 2007):
This study was conducted under the conditions that had been regulated and approved by the Animal Ethics Committee of Bahauddin Zakariya University, Pakistan. Full thickness skin was excised from the abdomen of male White New Zealand rabbits (3–4 kg) that had their hairs clipped short without damaging the skin to preserve its integrity. Skin samples were visually examined for any damage or abrasion that can compromise skin integrity. After excision, it was necessary to carefully remove adhering subcutaneous fat. Then, the skin was cut into samples that were just larger than the surface area of the diffusion cells. Prior to mounting in the cells, the skin samples were briefly immersed in normal saline (Meidan and Pritchard, 2010; Shah et al., 2012). Permeation measurements were made using Franz cells manufactured ‘in house’, exhibiting a diffusion area of 0.85 cm2 and a receptor cell volume of 5 mL. Each receptor compartment, containing a small Teflon coated magnetic stirrer, was initially filled with ultrasonically degassed PBS (pH 7.4). Subsequently, the rabbit skin was inserted as a barrier between the donor and receiver cells. Silicone grease was applied to create a good seal between the barrier and the two Franz cell compartments. To start each permeation experiment, 1 mL volume of the drug saturated solution (∼2 mg/mL) prepared from each ratio of conventional dispersion and lyophilised dispersion was deposited in the donor cell, which
Crystallinity (%) =
H HART × W
× 100
where H is the enthalpy of fusion (J/g) of solid dispersion, HART is the enthalpy of fusion (J/g) of pure artemisinin (assumed to be 100% crystalline), W is the weight fraction of artemisinin in the polymer blend (for example, W = 6/4 = 1.5). 2.4.3. Wide angle X-ray diffraction X-ray patterns for each of the dispersion formulations were obtained using a Bruker D8 Discover (Germany) apparatus. Measurement conditions included target (Cu K␣), voltage (35 kV), and current (35 mA). A system of diverging, receiving, and anti-scattering slits of 1◦ , 1◦ , 0.15◦ , respectively, was used. Patterns
2.4.4. Fourier transformed infrared spectroscopy (FT-IR) FT-IR spectra were measured on a Schimadzu, 2400 s Spectrometer. The samples were scanned over the range of 400–4000 cm−1 at a resolution of 2 cm−1 . 2.5. Solubility studies Solubility of pure artemisinin as well as from the various formulated dispersions was determined in two solvents, namely distilled water and phosphate buffered saline (PBS; pH 7.4). Essentially, excess amount of artemisinin, conventional dispersion and lyophilised dispersion was added to each of the solvent in separate glass vials with a Teflon coated magnetic flea. The glass vials were capped and placed on a stirring plate that was immersed in a water bath maintained at 37 (±1 ◦ C) for 48 h. The suspensions were then centrifuged at 4000 rpm for 30 min and a supernatant aliquot was taken out by a micro-pipette and analysed using HPLC after alkali reaction to determine the concentration. 2.6. Skin permeation studies
200
Y. Shahzad et al. / International Journal of Pharmaceutics 457 (2013) 197–205
was then covered with parafilm. Uniform receptor phase mixing was initiated by placing the diffusion cells on a magnetic stirring bed immersed in a thermostatic water bath maintained at 35 (±2 ◦ C). After 1 h, the receptor solution was removed and completely replaced with fresh, pre-warmed PBS. At scheduled times, an aliquot of 0.5 mL receiver fluid was withdrawn and the receiver phase was replenished with 0.5 mL fresh PBS maintained at 35 ± 2 ◦ C. Withdrawn aliquots were assayed by HPLC after alkali reaction with the derived artemisinin concentration values being mathematically corrected for the progressive dilutions caused by repeated sampling. Sink conditions existed throughout the experiment. Since skin exhibits large sample-to-sample permeability differences (Meidan and Pritchard, 2010), therefore, each experiment consisted of 5 replicate runs (n = 5). 2.7. Data analysis The steady state flux (J) was calculated from the slope of the steepest part (linear portion) of the plot of the cumulative amount permeated per unit area as a function of time. The flux is expressed as: J=
C0 KD = C0 KP L
From this relation the permeability coefficient was calculated using following Equation: KP =
J C0
The effectiveness of penetration enhancers (enhancement ratio, ER) was calculated from the ratio of artemisinin flux in the presence and absence of enhancers. All values are expressed as mean ± SD of 5 replicates. Statistical analysis was executed on Minitab software version 16. One-way ANOVA at 95% confidence level (p ≤ 0.05) was performed to check the significance in the values obtained and post hoc comparison was carried out using pairwise Turkey’s test, where the difference in means was significant. 3. Results and discussion 3.1. Characterisation of dispersions 3.1.1. Miscibility determination using MD simulations MD simulations were carried out to evaluate the solubility parameter ı of the drug, the individual polymers and the polymer blend (50:50 wt%) to provide an insight into the compatibility and miscibility between the drug and the polymers. The solubility parameter is linked to the cohesive energy density CED (Hildebrand and Scott, 1950): ı = (CED)0.5 The cohesive energy density CED is the energy required to vaporise a mole of liquid normalised per unit volume (Hildebrand and Scott, 1950): CED =
Hv − RT Vm
where Hv is the molar enthalpy for vaporisation, R is the gas constant, T the absolute temperature in Kelvins, and Vm is the molar volume. The Hv is estimated from molecular dynamics simulation, making the cohesive energy density and the solubility parameter accessible. Thus amorphous structures were generated for the drug and the two polymers, equilibrated at 298 K using molecular dynamics to constant density, and then samples of configurations were used to calculate an average enthalpy of
Fig. 3. DSC thermogram of (a) artemisinin, (b) CD 5:5, (c) LD 5:5, (d) CD 1:9 and (e) LD 1:9.
vaporisation. Molecular dynamics simulation enables the system to sample different molecular configurations consistent with the thermal energy at the set temperature, without the system being locked in a local minimum. The potential energy function employed the molecular mechanics approximation and included bond stretching, valence angle bending, dihedral, Coulombic interactions, and van der Waals interactions. The estimated CED and solubility parameters of artemisinin, polymers and polymer blend computed from the MD simulations at 298 K are presented in Table 1. It has been established that compounds with similar ı are thermodynamically miscible (Greenhalgh et al., 1999). This is owing to a balance between non-bond interaction energies during mixing. Generally, drug–polymer blends are considered miscible if the difference in solubility parameter ı is less than 7 MPa0.5 , or immiscible if ı greater than 10 MPa0.5 with respect to each other (Maniruzzaman et al., 2013). The MD simulated solubility parameter difference ı was 0.08 for artemisinin and PVP, 0.57 for artemisinin and PEG, and 0.15 for artemisinin and blend of polymers (PVP:PEG; 50:50 wt%). This implies that the polymers were miscible with the artemisinin to form stable solid dispersions suggesting a molecular dispersion of the drug in the polymer matrix. The simulated ı values of PVP and PEG reported here are in good agreement with the published values (Xiang and Anderson, 2013). 3.1.2. Thermal analysis Thermal behaviour of solid dispersions was evaluated using DSC. Fig. 3 shows the DSC thermograms of pure artemisinin, and selected solid dispersion formulations, both conventional dispersions (CDs) and lyophilised dispersions (LDs). A sharp melting endothermic peak of pure artemisinin was observed at 152.6 ◦ C (H = −80.7 J/g). Melting endotherm depression and shifting to a lower temperature was observed as the ratio of polymer was increased in the dispersion system (Fig. 3b–e), which indicates a significant degree of mixing at this melting temperature suggesting the drug is molecularly dispersed in the polymer matrix. These findings validate the MD simulated miscibility profile of artemisinin and polymers. Percentage crystallinity of artemisinin in the solid dispersions was determined from the enthalpy of fusion of pure artemisinin and its dispersions. Percentage crystallinity of artemisinin decreased with increasing content of the polymer in the dispersion system as shown in Fig. 4. Results also demonstrated that the percent crystallinity of artemisinin in lyophilised dispersions (i.e. 2:8 and 1:9) was slightly higher than that of the conventional solid dispersions.
Y. Shahzad et al. / International Journal of Pharmaceutics 457 (2013) 197–205
201
Table 1 Cohesive energy density and solubility parameter values estimated from molecular dynamics simulation at 298 K. Chemical
Molecules (repeating units)
Number of chains per unit cell
ART PVP PEG PVP/PEG blend
50 10 30 15 monomers each
– 6 3 2/4
Simulated density (g cm−3 ) 1.2 1.3 1.1 1.2
± ± ± ±
0.4 0.1 0.1 0.1
CED (×106 J m−3 ) 367 370 389 373
± ± ± ±
8.8 11.2 10.5 10.1
ı (MPa0.5 ) 19.15 19.23 19.72 19.30
± ± ± ±
0.23 0.29 0.38 0.26
ı (MD) – 0.08 0.57 0.15
3.2. Solubility studies
Fig. 4. Degree of crystallinity (%) of artemisinin estimated using the DSC in the various conventional (CD) and lyophilised dispersions (LD).
3.1.3. X-ray diffraction Wide angle X-ray scattering patterns were obtained for artemisinin and the solid dispersions in order to study the crystalline status of artemisinin in the formulations and are shown in Fig. 5. Artemisinin is a highly crystalline powder that has characteristic sharp peaks at diffraction angles of 11.96◦ , 14.54◦ , 20.24◦ , 21.92◦ , and 28.27◦ at the 2 scale (Shahzad et al., 2012). The solid dispersions show a significant peak reduction with increasing polymer concentration indicating the transformation of crystalline artemisinin to its amorphous state. These results are in accordance with the previous studies where artemisinin was transformed into an amorphous form in the solid dispersion system with PVP (Shahzad et al., 2012). In case of lyophilised dispersions, although the peak intensity was reduced as the polymer concentration was increased, peaks were still observed for the LDs containing higher concentrations of polymers (2:8 and 1:9). However, this fraction was slightly higher (not more than 2%) than those observed for the CD. These results are consistent with the DSC results where a slightly higher percentage crystallinity of artemisinin was observed for lyophilised dispersions. 3.1.4. FT-IR analysis The interaction between the drug and carrier often leads to peculiar changes in the FT-IR profile. Therefore, FT-IR spectra (Fig. 6) were recorded for pure artemisinin as well as its conventional and lyophilised dispersions. The FT-IR spectra of conventional solid dispersions and lyophilised dispersions of artemisinin (recall the polymer matrix comprises PVP and PEG) showed a significant broadening of the O H stretching vibration peak characteristic of PEG (large band between 3481 cm−1 and 3118 cm−1 ) and a C N stretching vibration peak characteristic of PVP (1239 cm−1 ). A band at 876 cm−1 (O O C linkage), which is characteristic of artemisinin, was observed without shifting. All the characteristic FT-IR peaks of artemisinin were visible without shifting in all formulation systems, suggesting an absence of significant specific interaction between artemisinin and the binary mixture of PVP and PEG. These findings were in good agreement to the previously published report where no interactions were observed between artemisinin and PVP within the formulations (Shahzad et al., 2012).
The effects of PVP and PEG on the solubility of artemisinin were investigated in two solvents, namely water and the phosphate buffered saline PBS (pH 7.4). The aqueous solubility of pure artemisinin was found to be 12.9 ± 0.2 g/mL, which increased to 15.7 ± 0.5 g/mL in PBS. The solid dispersions of artemisinin with various drug to polymer ratios (6:4, 5:5, 3:7, 2:8, and 1:9) exhibited a substantial enhancement (up to about 50 fold) in artemisinin solubility in the two solvents compared to that of the pure drug (p < 0.05). The solubility and corresponding solubility enhancement ratios (ERSol ) are presented in Table 2. Although, the presence of hydrophilic polymers increased the solubility of artemisinin in water and phosphate buffered saline, yet no clear relationship exists between polymer concentration and solubility enhancement. In fact, the solubility of artemisinin started to increase to a maximum value for conventional (3:7) and lyophilised dispersion (3:7) and then started to decrease with further increase of the polymers. Solubility enhancement can be explained on the basis of several mechanisms. We believe that primarily this comes from the drug being present mostly in the amorphous form in the polymer matrices. The amorphous form can have a substantially higher solubility than the crystalline forms. Any crystalline component of the drug in the polymer matrix is likely to be released in a fine colloidal form, which too will yield a higher solubility as described by the Ostwald–Freundlich and the Kelvin equations (Grant and Brittian, 1995). Enhanced solubility may also result from solubilisation action of the hydrophilic polymers.
3.3. Permeation studies For permeation studies, the solid dispersions were dissolved up to their saturated limits in two solvents, namely water and PBS, and the permeation of the artemisinin was investigated from these vehicles through rabbit skin using Franz type diffusion cells. The cumulative amount of drug permeated (g/cm2 ) as a function of time is shown in Fig. 7a–d. The steady-state flux (J) was calculated from the slope linear of the amount permeated vs. time graph. Since the permeation profiles appeared to be non-linear, therefore, the steepest part of the curve (linear portion) was used for the steady state flux determination. The results show a significant difference of artemisinin permeation from the saturated solution of solid dispersions as compared to the saturated solution of the drug alone. Difference in permeation patterns were also observed for saturated solutions of CDs prepared in water and PBS. This anomalous behaviour could be a result of existence of the non-sink conditions when CD saturated solutions in PBS were evaluated for artemisinin permeation. However, if there was any other hidden phenomena existed, this would also effect the permeation pattern of LDs saturated solutions. No significant difference in permeation pattern was observed for LDs saturated solutions prepared in water and PBS. The steady state flux for the pure artemisinin was found to be 0.16 ± 0.02 (g cm−2 min−1 ) while the permeability coefficient was 0.8 × 10−4 cm min−1 . In all cases, the total amount of artemisinin
202
Y. Shahzad et al. / International Journal of Pharmaceutics 457 (2013) 197–205
Fig. 5. XRD patterns of conventional dispersions (CD) and lyophilised dispersions (LD).
Fig. 6. FT-IR spectra of artemisinin (ART) and its various solid dispersion formulations.
Table 2 Solubility (g/mL) of artemisinin from CDs and LDs in two solvents. Drug:polymer ratio
Conventional dispersions (CD) Water
6:4 5:5 3:7 2:8 1:9
474.0 495.3 515.2 477.5 469.1
± ± ± ± ±
5.7 6.4 3.2 9.2 7.9
Lyophilised dispersions (LD)
ERSol
PBS
36.7 40.0 38.4 37.0 36.4
560.8 570.8 621.7 616.5 607.4
The results are presented as mean ± standard error; n = 5.
± ± ± ± ±
12.1 8.1 12.4 5.8 11.7
ERSol
Water
35.7 36.4 41.5 39.3 38.7
588.3 676.7 755.4 665.2 628.3
± ± ± ± ±
9.9 17.9 14.8 11.1 13.3
ERSol
PBS
45.6 52.5 58.6 51.7 48.7
669.2 873.4 889.2 797.1 781.3
ERSol ± ± ± ± ±
13.8 19.0 17.9 5.9 13.2
42.6 55.6 56.6 50.8 49.7
Y. Shahzad et al. / International Journal of Pharmaceutics 457 (2013) 197–205
203
Fig. 7. Cumulative amount of artemisinin permeated from (a) saturated solutions of CDs prepared in water, (b) saturated solutions of CDs prepared in PBS, (c) saturated solutions of LDs prepared in water and (d) saturated solutions of LDs prepared in PBS.
Table 3 Permeation parameters of artemisinin form conventional dispersion in two solvents. Drug:polymer dispersion
Saturated solutions in water J (g cm−2 min−1 )
CD (6:4) CD (5:5) CD (3:7) CD (2:8) CD (1:9)
1.12 1.32 1.41 1.43 1.52
± ± ± ± ±
0.14 0.10 0.06 0.09 0.22
Saturated solutions in PBS
KP × 10−4 (cm min−1 ) 5.6 6.6 7.1 7.2 7.6
± ± ± ± ±
0.1 0.2 0.4 0.9 0.6
permeated over 8 h of study from solid dispersion vehicles was significantly increased (5–11 fold increase) (p < 0.05) as compared to the pure drug permeation profile. The permeation parameters of artemisinin from its solid dispersions are presented in Tables 3 and 4. The flux enhancement ratio ER was calculated from the flux rate for the solid dispersion vehicles relative to that from the pure drug solution. The ER for the conventional dispersions in water varied from about 5–10 fold in both water and PBS. For the lyophilised dispersion the enhancement was about 5–11 fold in both water and PBS, which is marginally higher than that from the conventional dispersions. This enhancement in the flux was dependent on the polymer content in the formulation, resulting in an increase in ER with increasing polymer content.
ER (J)
J (g cm−2 min−1 )
7.0 8.3 8.8 8.9 9.5
0.87 0.88 1.02 1.16 1.33
± ± ± ± ±
0.11 0.14 0.20 0.21 0.10
KP × 10−4 (cm min−1 ) 4.4 4.4 5.1 5.8 6.7
± ± ± ± ±
0.3 0.5 0.5 0.2 0.3
ER (J) 5.4 5.5 6.4 7.3 8.3
One-way ANOVA confirmed a significant difference (p < 0.05) in the flux values of artemisinin from the saturated solutions of the various solid dispersions in the two solvents. The pairwise Turkey’s test revealed that the flux values obtained from saturated solutions of the conventional dispersions in water and PBS were significantly different (p < 0.05) from each other as well as from that of the lyophilised dispersions in water and PBS. However, there was no significant difference (p > 0.05) observed in the fluxes of artemisinin from the various ratios of lyophilised dispersions in water and PBS apart from the one with highest polymer content (1:9). Overall, both formulation systems (CDs and LDs) produced flux values of artemisinin that were significantly higher than that from the artemisinin alone. Furthermore, the flux and the associated permeability coefficient increase with increase in the polymer
Table 4 Permeation parameters of artemisinin form lyophilised dispersions in two solvents. Drug:polymer dispersion
Saturated solutions in water −2
J (g cm LD (6:4) LD (5:5) LD (3:7) LD (2:8) LD (1:9)
1.04 1.14 1.16 1.42 1.43
± ± ± ± ±
min
0.20 0.10 0.11 0.20 0.20
−1
)
Saturated solutions in PBS −4
KP × 10 5.2 5.7 5.8 7.1 7.2
± ± ± ± ±
0.2 0.5 0.4 0.7 0.5
−1
(cm min
)
ER (J)
J (g cm−2 min−1 )
6.5 7.1 7.3 8.9 8.9
1.03 1.06 1.33 1.41 1.86
± ± ± ± ±
0.30 0.10 0.11 0.20 0.26
KP × 10−4 (cm min−1 ) 5.2 5.3 6.7 7.1 9.3
± ± ± ± ±
0.3 0.5 0.2 0.1 0.3
ER (J) 6.4 6.6 8.3 8.8 11.6
204
Y. Shahzad et al. / International Journal of Pharmaceutics 457 (2013) 197–205
content in the formulation system. Increase in the KP values with increasing polymer content was a result of perturbation of the skin lipid bilayers, consequently enhancing the partitioning of permeant in that region. The fact is pyrrolidones have tendency to penetrate into the hydrophobic regions of the SC and reduce the barrier function in these areas which allows influx of permeant via polar routes of the stratum corneum (Southwell and Barry, 1983). This was consistent with the previously published reports where permeation rate of poorly water soluble drugs like celecoxib and meloxicam was increased as a function of polymer concentration (Saleem et al., 2010; Soliman et al., 2011). The fact that we are getting a substantial increase in the flux of artemisinin from the solid dispersion vehicles suggests that the cause is the massive increase in supersaturation (i.e. increase in the thermodynamic activity of the solution) that is attained when the dispersions are dissolved in the vehicle. Had the polymers merely increased the solubility by a solubilisation effect, the solution activity would be equivalent to that of the pure artemisinin saturated solution, which would have yielded a similar flux. There is the possibility that the polymers are acting as penetration enhancers (see Williams and Barry, 2012) but this effect is not easy to decouple (Williams and Barry, 2012). The important question is, is the enhanced permeability adequate for the transdermal delivery of artemisinin? The maximum achievable flux is about 1.86 g cm−2 min−1 . Here we used the predictive approach to estimate the steady state plasma concentration (Css ) that would penetrate after topical application using following equation (Flynn and Stewart, 1988): Css =
Jmax × A Clp
where Jmax is the maximum achievable flux, A is the hypothetical area of application and Clp is the plasmatic clearance. The Clp for artemisinin after a 500 mg oral dose is 130 mL/min (Ashton et al., 1998). Taking into account a 100 cm2 hypothetical area for a transdermal patch, the predicted plasma concentration turn out to be 1.43 g/mL which is higher than the therapeutic range of 0.23–0.58 g/mL (Chan et al., 1997) confirming the potential transdermal application of artemisinin. 4. Conclusion Artemisinin is a poorly soluble drug prone to first-pass metabolism, for which the bioavailability can be low and erratic. We have investigated the use of solid dispersions for enhancing its solubility and associated supersaturation with a view to increasing its permeability through skin. Solid dispersions were prepared using a 50:50 blend of PVP-K30 and PEG 4000 with the ratio of drug to polymer being varied. The dispersions were characterised using physical methods to ascertain whether the dispersed drug was crystalline or amorphous. Molecular dynamics simulations were also employed to estimate the miscibility of the drug and the polymer mix. The simulations revealed that the drug and the polymers were miscible, which rationalised the finding the observation that the dispersed drug was largely molecular dispersed or amorphous. The dispersions should substantial increase in solubility and hence associated supersaturation, and resulting in enhanced permeation through rabbit skin. Conflict of interest The authors declare no competing financial interest. Acknowledgments The authors thank the Bahauddin Zakariya University Multan, Pakistan for providing funding to conduct this research. The authors
would like to thank Prof. Jamshed Anwar for his help on MD simulations. We are also grateful to Malcolm Cutler, Managing Director FSC Development Services Ltd, UK for his support in this work. References Andrews, G.P., Abudiak, O.A., Jones, D.S., 2010. Physicochemical characterization of hot melt extruded bicalutamide–polyvinylpyrrolidone solid dispersions. J. Pharm. Sci. 99, 1322–1335. Ashton, M., Gordi, T., Hai, T.N., Van Huong, N., Sy, N.D., Nieu, N.T., Huong, D.X., Johansson, M., Công, L.D., 1998. Artemisinin pharmacokinetics in healthy adults after 250, 500 and 1000 mg single oral doses. Biopharm. Drug Dispos. 19, 245–250. Bley, H., Fussnegger, B., Bodmeier, R., 2010. Characterization and stability of solid dispersions based on PEG/polymer blends. Int. J. Pharm. 390, 165–173. Chamchod, F., Beier, J.C., 2013. Modeling Plasmodium vivax: relapses, treatment, seasonality, and G6PD deficiency. J. Theor. Biol. 316, 25–34. Chan, K.L., Yuen, K.H., Jinadasa, S., Peh, K.K., Toh, W.T., 1997. A high-performance liquid chromatography analysis of plasma artemisinin using a glassy carbon electrode for reductive electrochemical detection. Planta Med. 63, 66–69. Craig, D.Q.M., 2002. The mechanisms of drug release from solid dispersions in watersoluble polymers. Int. J. Pharm. 231, 131–144. Eastman, R.T., Fidock, D.A., 2009. Artemisinin-based combination therapies: a vital tool in efforts to eliminate malaria. Nat. Rev. Microbiol. 7, 864–874. Flynn, G.L., Stewart, B., 1988. Percutaneous drug penetration: choosing candidates for transdermal development. Drug Dev. Res. 13, 169–185. Gordi, T., Nielsen, E., Yu, Z., Westerlund, D., Ashton, M., 2000. Direct analysis of artemisinin in plasma and saliva using coupled-column high-performance liquid chromatography with a restricted-access material pre-column. J. Chromatogr. B 742, 155–162. Grant, D.J.W., Brittian, H.G., 1995. In: Brittian, H.G.E. (Ed.), Physical Characterisation of Pharmaceutical Solids. Marcel Dekker, New York. Greenhalgh, D.J., Williams, A.C., Timmins, P., York, P., 1999. Solubility parameters as predictors of miscibility in solid dispersions. J. Pharm. Sci. 88, 1182–1190. Gupta, J., Nunes, C., Vyas, S., Jonnalagadda, S., 2011. Prediction of solubility parameters and miscibility of pharmaceutical compounds by molecular dynamics simulations. J. Phys. Chem. B 115, 2014–2023. Hildebrand, J., Scott, R., 1950. The Solubility of Nonelectrolytes. Reinhold Publ. Corp, New York. Klayman, D.L., 1985. Qinghaosu (Artemisinin): an antimalarial drug from China. Science 228, 1049–1055. Langer, M., Höltje, M., Urbanetz, N.A., Brandt, B., Höltje, H.D., Lippold, B.C., 2003. Investigations on the predictability of the formation of glassy solid solutions of drugs in sugar alcohols. Int. J. Pharm. 252, 167–179. Lima, Á.A.N., Soares-Sobrinho, J.L., Silva, J.L., Corrêa-Júnior, R.A.C., Lyra, M.A.M., Santos, F.L.A., Oliveira, B.G., Hernandes, M.Z., Rolim, L.A., Rolim-Neto, P.J., 2011. The use of solid dispersion systems in hydrophilic carriers to increase benznidazole solubility. J. Pharm. Sci. 100, 2443–2451. Maniruzzaman, M., Morgan, D.J., Mendham, A.P., Pang, J., Snowden, M.J., Douroumis, D., 2013. Drug–polymer intermolecular interactions in hot-melt extruded solid dispersions. Int. J. Pharm. 443, 199–208. Meidan, V.M., Pritchard, D., 2010. A two-layer diffusive model for describing the variability of transdermal drug permeation. Eur. J. Pharm. Biopharm. 74, 513–517. Rawlinson, C.F., Williams, A.C., Timmins, P., Grimsey, I., 2007. Polymer-mediated disruption of drug crystallinity. Int. J. Pharm. 336, 42–48. Rigby, D., Sun, H., Eichinger, B.E., 1997. Computer simulations of poly(ethylene oxide): force field, PVT diagram and cyclization behaviour. Polym. Int. 44, 311–330. Sahoo, N.G., Abbas, A., Judeh, Z., Li, C.M., Yuen, K.H., 2009. Solubility enhancement of a poorly water-soluble anti-malarial drug: experimental design and use of a modified multifluid nozzle pilot spray drier. J. Pharm. Sci. 98, 281–296. Sahoo, N.G., Kakran, M., Li, L., Judeh, Z., Müller, R.H., 2011. Dissolution enhancement of a poorly water-soluble antimalarial drug by means of a modified multi-fluid nozzle pilot spray drier. Mater. Sci. Eng. C 31, 391–399. Saleem, M.A., Bala, S., Liyakat Aeajaz, A., 2010. Effect of different carriers on in vitro permeation of meloxicam through rat skin. Indian J. Pharm. Sci. 72, 710–718. Serajuddln, A.T.M., 1999. Solid dispersion of poorly water-soluble drugs: early promises, subsequent problems, and recent breakthroughs. J. Pharm. Sci. 88, 1058–1066. Shah, S.N.H., Mehboob, E.R., Shahzad, Y., Badshah, A., Meidan, V.M., Murtaza, G., 2012. Developing an efficacious diclofenac diethylamine transdermal formulation. J. Food Drug Anal. 20, 464–470. Shahzad, Y., Shah, S.N.H., Ansari, M.T., Riaz, R., Safdar, A., Hussain, T., Malik, M., 2012. Effects of drug–polymer dispersions on solubility and in vitro diffusion of artemisinin across a polydimethylsiloxane membrane. Chin. Sci. Bull. 57, 1685–1692. Shahzad, Y., Shah, S.N.H., Atique, S., Ansari, M.T., Bashir, F., Hussain, T., 2011. The evaluation of coated granules to mask the bitter taste of dihydroartemisinin. Braz. J. Pharm. Sci. 47, 323–330. Soliman, S.M., Malak, N.S.A., El Gazayerly, O.N., Rehim, A.A.A., 2011. Preparation of celecoxib solid dispersions for dermal application: in vitro characterization and skin irritation test. J. Drug Deliv. Sci. Technol. 21, 509–516. Southwell, D., Barry, B.W., 1983. Penetration enhancers for human skin: mode of action of 2-pyrrolidone and dimethylformamide on partition and diffusion
Y. Shahzad et al. / International Journal of Pharmaceutics 457 (2013) 197–205 of model compounds water, n-alcohols, and caffeine. J. Invest. Dermatol. 80, 507–514. Svensson, U.S.H., Sandström, R., Carlborg, O., Lennernäs, H., Ashton, M., 1999. High in situ rat intestinal permeability of artemisinin unaffected by multiple dosing and with no evidence of P-glycoprotein involvement. Drug Metab. Dispos. 27, 227–232. Thaithong, S., Beale, G.H., 1985. Susceptibility of Thai isolates of Plasmodium falciparum to artemisinine (qinghaosu) and artemether. Bull. World Health Organ. 63, 617–619. Titulaer, H.A.C., Zuidema, J., Lugt, C.B., 1991. Formulation and pharmacokinetics of artemisinin and its derivatives. Int. J. Pharm. 69, 83–92. Van Krevelen, D.W., Hoftyzer, P., 1976. Properties of Polymers: Their Estimation and Correlation with Chemical Structure. Elsevier, New York. Williams, A.C., Barry, B.W., 2012. Penetration enhancers. Adv. Drug Deliv. Rev. 64, 128–137.
205
Wong, J.W., Yuen, K.H., 2001. Improved oral bioavailability of artemisinin through inclusion complexation with - and ␥-cyclodextrins. Int. J. Pharm. 227, 177– 185. World Health Organization, 2010a. Guidelines for the treatment of malaria. In: Diagnosis of Malaria, second ed, 10–11. World Health Organization, 2010b. World Malaria Report. World Health Organization, Geneva. Xiang, T.-X., Anderson, B.D., 2013. Molecular dynamics simulation of amorphous indomethacin–poly(vinylpyrrolidone) glasses: solubility and hydrogen bonding interactions. J. Pharm. Sci. 102, 876–891. Zhao, S.S., Zeng, M.Y., 1985. Spectrometric high-pressure liquid chromatographic studies on analytical chemistry of Qinghaosu. Planta Med. 3, 233–237. Zhu, Q., Harris, M.T., Taylor, L.S., 2012. Modification of crystallization behavior in drug/polyethylene glycol solid dispersions. Mol. Pharm. 9, 546–553.
Contents lists available at ScienceDirect
International Journal of Pharmaceutics journal homepage: www.elsevier.com/locate/ijpharm
Development of solid dispersions of artemisinin for transdermal delivery Yasser Shahzad a,∗ , Sadia Sohail b , Muhammad Sohail Arshad b , Talib Hussain a , Syed Nisar Hussain Shah b a b
Pharmacy and Pharmaceutical Science, University of Huddersfield, Queensgate, Huddersfield HD1 3DH, UK Faculty of Pharmacy, Bahauddin Zakariya University, Multan 68000, Pakistan
a r t i c l e
i n f o
Article history: Received 14 July 2013 Received in revised form 20 September 2013 Accepted 22 September 2013 Available online 29 September 2013 Keywords: Artemisinin Solid dispersions Solubility parameter Miscibility Molecular dynamics Permeability coefficient
a b s t r a c t Solid dispersions of the poorly soluble drug artemisinin were developed using polymer blends of polyvinylpyrrolidone (PVP) and polyethylene glycol (PEG) with the aim of enhancing solubility and in vitro permeation of artemisinin through skin. Formulations were characterised using a combination of molecular dynamics (MD) simulations, differential scanning calorimetry (DSC), X-ray diffraction (XRD) and Fourier transform infrared spectroscopy (FT-IR). Solubility of artemisinin was determined in two solvents: de-ionised water and phosphate buffered saline (PBS; pH 7.4), while in vitro drug permeation studies were carried out using rabbit skin as a model membrane. MD simulations revealed miscibility between the drug and polymers. DSC confirmed the molecular dispersion of the drug in the polymer blend. Decrease in crystallinity of artemisinin with respect to polymer content and the absence of specific drug–polymer interactions were confirmed using XRD and FT-IR, respectively. The solubility of artemisinin was dramatically enhanced for the solid dispersions, as was the permeation of artemisinin from saturated solid-dispersion vehicles relative to that from saturated solutions of the pure drug. The study suggests that high energy solid forms of artemisinin could possibly enable transdermal delivery of artemisinin. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Malaria is the major and the most prevalent public health problem in many parts of the world (Chamchod and Beier, 2013). According to WHO, about half of the world’s population is exposed to the risk of malaria (World Health Organization, 2010a) and it is the foremost cause of morbidity and mortality in the world with a million of deaths every year (World Health Organization, 2010a; Eastman and Fidock, 2009). Resistance to conventional antimalarial drugs has led to changes in malaria control policies globally in favour of artemisinin. Currently, the treatment of choice for uncomplicated and severe malaria is based on administration of artemisinin and its derivatives (Shahzad et al., 2011), either used alone or in combination with other drugs (Eastman and Fidock, 2009; World Health Organization, 2010b). Average plasma concentration of artemisinin after single oral dose of 500 mg tablet formulation range between 400 and 700 ng/mL (Gordi et al., 2000). Artemisinin (Fig. 1) is a parent compound of a novel family of antimalarials extracted from the Chinese traditional plant, Artemisia annua, L. Asteraceae. It is effective against malaria
∗ Corresponding author. Tel.: +44 7910427497. E-mail addresses: [email protected], [email protected] (Y. Shahzad). 0378-5173/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ijpharm.2013.09.027
parasites, including the multidrug-resistant falciparum species due to its high potency and rapid onset of action (Klayman, 1985; Svensson et al., 1999; Thaithong and Beale, 1985). Structurally, it is a sesquiterpene lactone with an inner peroxide bridge which is responsible for its antimalarial activity. It is characterised as a poorly water-soluble drug with an octanol–water partition coefficient greater than 2 and a short half-life of 2–3 h, and is extensively metabolised by the liver. Thus, oral bioavailability can be as low as 32% (Sahoo et al., 2011; Svensson et al., 1999; Wong and Yuen, 2001). Although artemisinin has shown excellent permeability across the intestinal mucosa, it has low bioavailability because of its poor aqueous solubility, which can adversely affect its efficacy (Titulaer et al., 1991). This makes artemisinin a suitable candidate for transdermal drug delivery, which circumvents the first pass metabolism by the liver. Solid dispersion of an insoluble drug in hydrophilic polymers is an attractive technique for enhancing drug solubility and the dissolution rate, which for rapidly absorbing molecules such as artemisinin could enhance their drug bioavailability (Lima et al., 2011); (Craig, 2002). Most solid dispersions are prepared using highly water-soluble polymers as the carrier, where the polymer can be amorphous (for example polyvinylpyrrolidone; PVP) or partially crystalline (for example polyethylene glycol; PEG) (Zhu et al., 2012). Solid dispersions are generally prepared by either a solvent
198
Y. Shahzad et al. / International Journal of Pharmaceutics 457 (2013) 197–205
Leicestershire, UK); vacuum grease (Dow Corning, Midland, Michigan, USA). De-ionised water was used throughout to make up all solutions. 2.2. Quantitative analysis of artemisinin
Fig. 1. Structure of artemisinin.
method, whereby the drug and carrier are dissolved in a mutual solvent followed by the removal of solvent by evaporation, or by a melting method, whereby drug–carrier mixtures are prepared by co-melting/cooling (Bley et al., 2010). Upon dissolution of the solid dispersion in an aqueous medium the carrier will dissolve rapidly, releasing very fine colloidal drug particles of very high surface area resulting in dissolution rate enhancement (Serajuddln, 1999). The primary issue with solid dispersions is re-crystallisation of the drug on storage and also immediately on dissolution, which can be addressed by judicious choice of stabilising polymer(s) taking into account the drug–polymer intermolecular interactions, the viscosity of the polymer, and the glass transition temperature of the solid dispersion (Andrews et al., 2010). In the present study we investigate the development of solid dispersions of artemisinin using polyvinylpyrrolidone (PVP) grade PVP-K30 and polyethylene glycol (PEG) 4000 aimed at enhancing in vitro permeation of artemisinin across skin. Both PVP and PEG are polymers of choice for solid dispersions. PVP is an amorphous water-soluble polymer with a glass transition temperature between 110 and 180 ◦ C, while PEG is semicrystalline with a low melting point and is also water-soluble. We show that the permeability of artemisinin from the solid dispersions in enhanced in excess of 10 fold relative to that of a saturated solution of the original drug. The physical state of the drug in the solid dispersions was characterised through X-ray diffraction, Fourier-transform infra-red spectroscopy, and differential scanning calorimetry. The stability of the solid dispersions is linked with the drug–polymer and polymer–polymer (for mixed polymer systems) miscibility, which can be characterised by the solubility parameter. The solubility parameters are typically estimated by a group contribution method (Van Krevelen and Hoftyzer, 1976), but such an approach is challenged by directional hydrogen bonding and long-range electrostatic interactions (Langer et al., 2003). Here we employ the more rigorous approach of molecular dynamics (MD) simulation, which accounts for the individual atomic interactions, to estimate the solubility parameters, following the work of (Gupta et al., 2011). The results reveal that artemisinin and the polymers PVP and PEG are relatively miscible with each other encouraging the formation of a molecular dispersion of the drug in the polymer matrix and its stability, which indeed rationalises the enhanced solubility and permeation observed from the dispersions. 2. Materials and methods 2.1. Materials The following chemicals were used as purchased: artemisinin 99.9% purity (Alchem, New Delhi, India); PVP-K30 and PEG-4000 (Beijing chemical company, Beijing, China); methanol HPLC grade 99%+ (Merck, Darmstadt, Germany); potassium dihydrogen phosphate (VWR, Lutterworth, Leicestershire UK); sodium chloride and potassium chloride (Sigma–Aldrich, Poole, Dorset, UK); di-sodium hydrogen phosphate (Fischer Scientific Chemicals, Loughborough,
Artemisinin was quantified using HPLC method with UVdetection as described previously (Sahoo et al., 2009; Zhao and Zeng, 1985), in which artemisinin is transformed into a UVabsorbing compound through an alkali reaction, i.e. by heating the solution with 0.2% NaOH solution. The alkali reaction was carried out by adding 5 mL of 0.2% NaOH solution to each sample, then heated at 50 ± 1 ◦ C for 30 min and allowed to cool down in a refrigerator. Finally, 1 mL glacial acetic acid was added to each sample before injecting into the HPLC system. Briefly, the HPLC system comprised of a Severn Analytical solvent delivery system (SA 6410B), a Waters 2487 UV-detector (Waters, Hertfordshire, UK), a Rheodyne injector (Perkin Elmer 7725) fitted with a 20 L sample loop. The column used was Novapack C18 (4.6 cm × 15 cm, 4 m) (Waters, Hertfordshire, UK). The composition of mobile phase was 0.01 M di-sodium hydrogen phosphate (75%) and acetonitrile (25%) and the final pH was adjusted to 6.5 using glacial acetic acid. The mobile phase was pumped through the system with a flow rate of 0.8 mL/min. The UV-detector was set at a wavelength of 290 nm. 2.3. Formulation of solid dispersions Solid dispersions of artemisinin in two hydrophilic polymers, namely PVP and PEG were prepared by conventional solvent evaporation and lyophilisation methods as described previously (Shahzad et al., 2012). Solid dispersion were prepared with different drug to polymer ratios, i.e. 6:4, 5:5, 3:7, 2:8, and 1:9. It should be noted that the quantities of PVP and PEG were kept equal (i.e. 50:50 wt%) with respect to each other in all the formulations. For conventional dispersions (CD), accurately weighed quantities of PVP and PEG were dissolved in 100 mL of methanol in a pre-cleaned vessel. This was followed by the addition of accurately weighed quantities of artemisinin and allowed to dissolve completely by continuous stirring at ambient temperature. Methanol was evaporated on a rotary evaporator at reduced pressure. The resulting residue was dried under vacuum in a desiccator for 24 h at ambient temperature and the dried mass obtained was finally pulverised in a pre-cleaned pestle and mortar for about 5 min until a homogeneous mixture was obtained and then refrigerated in a closed vial at 5 ◦ C until further investigations. For lyophilised dispersions (LD), each of the drug–polymer mixture of respective ratio was dissolved in methanol until a clear solution was obtained. This solution was quickly solidified by immersing the flask in liquid nitrogen. Upon cooling, the flask was attached to the vacuum adapter of the lyophiliser for sublimation. After the solvent was completely removed, a porous and fluffy powder residue appeared that was kept in a refrigerator at 5 ◦ C until further investigations. The prepared formulations were subjected to drug content analysis. Three random samples of 10 mg drug equivalent from each formulation were dissolved in methanol and appropriately diluted and the drug content was determined by HPLC analysis after alkali reaction. The drug content in the solid dispersions prepared from both methods was found to be in the range of 88–91%. 2.4. Characterisation of dispersions 2.4.1. Drug–polymer miscibility determination A key consideration for developing a physically stable dispersion is the miscibility of drug in the polymer matrix. The cohesive
Y. Shahzad et al. / International Journal of Pharmaceutics 457 (2013) 197–205
199
Fig. 2. Energy minimised amorphous simulation cell of artemisinin.
energy density (CED) and Hansen’s solubility parameter (ı) are important parameters in this regard, which were estimated using molecular dynamics (MD) simulation. The MD simulations were performed using Materials Studio 6.0 (Accelrys, San Diego, CA) software. The artemisinin structure was imported from the Cambridge Structural Database (RSC, UK) while the homopolymer structure of PVP (chain of 10 monomers) and PEG (chain of 30 monomers) were constructed using the Build module within Materials Studio. Energy minimised (steepest decent/conjugate gradient) amorphous simulation cells for artemisinin (50 molecules), PVP (6 chains), PEG (3 chains) and polymer blend (2 PVP/4 PEG chains of 15 monomers each) were constructed as exemplified in Fig. 2. The ab initio COMPASS force field was assigned for the MD simulations (Rigby et al., 1997). The amorphous simulation cells were equilibrated under isothermal–isobaric (NPT) conditions using the Anderson thermostat (298 K) and barostat (1 bar) at a time step of 1 fs for 0.5 ns. The equilibrated cells were further subjected to dynamics for 200 ps simulation time and trajectories were captured every 10 ps for subsequent analysis. The non-bonded van der Waals and Coulomb interactions were truncated using a group based cut-off distance ˚ The captured trajectories were used for CED and ı calcuof 12.5 A. lation using Forcite molecular dynamics module within Materials Studio.
were obtained using a scan speed of 4◦ /min with 2 between 10◦ and 50◦ .
2.4.2. Differential scanning calorimetry (DSC) The DSC curves of pure artemisinin and dispersion system were recorded on a Mettler Toledo® DSC 1 Star thermal analyser. Thermal behaviour was studied by heating samples (5–6 mg) in a sealed aluminium crucible, using an empty crucible as reference, over the temperature range of 25–200 ◦ C at a rate of 10 ◦ C/min under constant nitrogen flow at a rate of 50 mL/min. Percentage crystallinity of the samples was calculated using the following equation (Rawlinson et al., 2007):
This study was conducted under the conditions that had been regulated and approved by the Animal Ethics Committee of Bahauddin Zakariya University, Pakistan. Full thickness skin was excised from the abdomen of male White New Zealand rabbits (3–4 kg) that had their hairs clipped short without damaging the skin to preserve its integrity. Skin samples were visually examined for any damage or abrasion that can compromise skin integrity. After excision, it was necessary to carefully remove adhering subcutaneous fat. Then, the skin was cut into samples that were just larger than the surface area of the diffusion cells. Prior to mounting in the cells, the skin samples were briefly immersed in normal saline (Meidan and Pritchard, 2010; Shah et al., 2012). Permeation measurements were made using Franz cells manufactured ‘in house’, exhibiting a diffusion area of 0.85 cm2 and a receptor cell volume of 5 mL. Each receptor compartment, containing a small Teflon coated magnetic stirrer, was initially filled with ultrasonically degassed PBS (pH 7.4). Subsequently, the rabbit skin was inserted as a barrier between the donor and receiver cells. Silicone grease was applied to create a good seal between the barrier and the two Franz cell compartments. To start each permeation experiment, 1 mL volume of the drug saturated solution (∼2 mg/mL) prepared from each ratio of conventional dispersion and lyophilised dispersion was deposited in the donor cell, which
Crystallinity (%) =
H HART × W
× 100
where H is the enthalpy of fusion (J/g) of solid dispersion, HART is the enthalpy of fusion (J/g) of pure artemisinin (assumed to be 100% crystalline), W is the weight fraction of artemisinin in the polymer blend (for example, W = 6/4 = 1.5). 2.4.3. Wide angle X-ray diffraction X-ray patterns for each of the dispersion formulations were obtained using a Bruker D8 Discover (Germany) apparatus. Measurement conditions included target (Cu K␣), voltage (35 kV), and current (35 mA). A system of diverging, receiving, and anti-scattering slits of 1◦ , 1◦ , 0.15◦ , respectively, was used. Patterns
2.4.4. Fourier transformed infrared spectroscopy (FT-IR) FT-IR spectra were measured on a Schimadzu, 2400 s Spectrometer. The samples were scanned over the range of 400–4000 cm−1 at a resolution of 2 cm−1 . 2.5. Solubility studies Solubility of pure artemisinin as well as from the various formulated dispersions was determined in two solvents, namely distilled water and phosphate buffered saline (PBS; pH 7.4). Essentially, excess amount of artemisinin, conventional dispersion and lyophilised dispersion was added to each of the solvent in separate glass vials with a Teflon coated magnetic flea. The glass vials were capped and placed on a stirring plate that was immersed in a water bath maintained at 37 (±1 ◦ C) for 48 h. The suspensions were then centrifuged at 4000 rpm for 30 min and a supernatant aliquot was taken out by a micro-pipette and analysed using HPLC after alkali reaction to determine the concentration. 2.6. Skin permeation studies
200
Y. Shahzad et al. / International Journal of Pharmaceutics 457 (2013) 197–205
was then covered with parafilm. Uniform receptor phase mixing was initiated by placing the diffusion cells on a magnetic stirring bed immersed in a thermostatic water bath maintained at 35 (±2 ◦ C). After 1 h, the receptor solution was removed and completely replaced with fresh, pre-warmed PBS. At scheduled times, an aliquot of 0.5 mL receiver fluid was withdrawn and the receiver phase was replenished with 0.5 mL fresh PBS maintained at 35 ± 2 ◦ C. Withdrawn aliquots were assayed by HPLC after alkali reaction with the derived artemisinin concentration values being mathematically corrected for the progressive dilutions caused by repeated sampling. Sink conditions existed throughout the experiment. Since skin exhibits large sample-to-sample permeability differences (Meidan and Pritchard, 2010), therefore, each experiment consisted of 5 replicate runs (n = 5). 2.7. Data analysis The steady state flux (J) was calculated from the slope of the steepest part (linear portion) of the plot of the cumulative amount permeated per unit area as a function of time. The flux is expressed as: J=
C0 KD = C0 KP L
From this relation the permeability coefficient was calculated using following Equation: KP =
J C0
The effectiveness of penetration enhancers (enhancement ratio, ER) was calculated from the ratio of artemisinin flux in the presence and absence of enhancers. All values are expressed as mean ± SD of 5 replicates. Statistical analysis was executed on Minitab software version 16. One-way ANOVA at 95% confidence level (p ≤ 0.05) was performed to check the significance in the values obtained and post hoc comparison was carried out using pairwise Turkey’s test, where the difference in means was significant. 3. Results and discussion 3.1. Characterisation of dispersions 3.1.1. Miscibility determination using MD simulations MD simulations were carried out to evaluate the solubility parameter ı of the drug, the individual polymers and the polymer blend (50:50 wt%) to provide an insight into the compatibility and miscibility between the drug and the polymers. The solubility parameter is linked to the cohesive energy density CED (Hildebrand and Scott, 1950): ı = (CED)0.5 The cohesive energy density CED is the energy required to vaporise a mole of liquid normalised per unit volume (Hildebrand and Scott, 1950): CED =
Hv − RT Vm
where Hv is the molar enthalpy for vaporisation, R is the gas constant, T the absolute temperature in Kelvins, and Vm is the molar volume. The Hv is estimated from molecular dynamics simulation, making the cohesive energy density and the solubility parameter accessible. Thus amorphous structures were generated for the drug and the two polymers, equilibrated at 298 K using molecular dynamics to constant density, and then samples of configurations were used to calculate an average enthalpy of
Fig. 3. DSC thermogram of (a) artemisinin, (b) CD 5:5, (c) LD 5:5, (d) CD 1:9 and (e) LD 1:9.
vaporisation. Molecular dynamics simulation enables the system to sample different molecular configurations consistent with the thermal energy at the set temperature, without the system being locked in a local minimum. The potential energy function employed the molecular mechanics approximation and included bond stretching, valence angle bending, dihedral, Coulombic interactions, and van der Waals interactions. The estimated CED and solubility parameters of artemisinin, polymers and polymer blend computed from the MD simulations at 298 K are presented in Table 1. It has been established that compounds with similar ı are thermodynamically miscible (Greenhalgh et al., 1999). This is owing to a balance between non-bond interaction energies during mixing. Generally, drug–polymer blends are considered miscible if the difference in solubility parameter ı is less than 7 MPa0.5 , or immiscible if ı greater than 10 MPa0.5 with respect to each other (Maniruzzaman et al., 2013). The MD simulated solubility parameter difference ı was 0.08 for artemisinin and PVP, 0.57 for artemisinin and PEG, and 0.15 for artemisinin and blend of polymers (PVP:PEG; 50:50 wt%). This implies that the polymers were miscible with the artemisinin to form stable solid dispersions suggesting a molecular dispersion of the drug in the polymer matrix. The simulated ı values of PVP and PEG reported here are in good agreement with the published values (Xiang and Anderson, 2013). 3.1.2. Thermal analysis Thermal behaviour of solid dispersions was evaluated using DSC. Fig. 3 shows the DSC thermograms of pure artemisinin, and selected solid dispersion formulations, both conventional dispersions (CDs) and lyophilised dispersions (LDs). A sharp melting endothermic peak of pure artemisinin was observed at 152.6 ◦ C (H = −80.7 J/g). Melting endotherm depression and shifting to a lower temperature was observed as the ratio of polymer was increased in the dispersion system (Fig. 3b–e), which indicates a significant degree of mixing at this melting temperature suggesting the drug is molecularly dispersed in the polymer matrix. These findings validate the MD simulated miscibility profile of artemisinin and polymers. Percentage crystallinity of artemisinin in the solid dispersions was determined from the enthalpy of fusion of pure artemisinin and its dispersions. Percentage crystallinity of artemisinin decreased with increasing content of the polymer in the dispersion system as shown in Fig. 4. Results also demonstrated that the percent crystallinity of artemisinin in lyophilised dispersions (i.e. 2:8 and 1:9) was slightly higher than that of the conventional solid dispersions.
Y. Shahzad et al. / International Journal of Pharmaceutics 457 (2013) 197–205
201
Table 1 Cohesive energy density and solubility parameter values estimated from molecular dynamics simulation at 298 K. Chemical
Molecules (repeating units)
Number of chains per unit cell
ART PVP PEG PVP/PEG blend
50 10 30 15 monomers each
– 6 3 2/4
Simulated density (g cm−3 ) 1.2 1.3 1.1 1.2
± ± ± ±
0.4 0.1 0.1 0.1
CED (×106 J m−3 ) 367 370 389 373
± ± ± ±
8.8 11.2 10.5 10.1
ı (MPa0.5 ) 19.15 19.23 19.72 19.30
± ± ± ±
0.23 0.29 0.38 0.26
ı (MD) – 0.08 0.57 0.15
3.2. Solubility studies
Fig. 4. Degree of crystallinity (%) of artemisinin estimated using the DSC in the various conventional (CD) and lyophilised dispersions (LD).
3.1.3. X-ray diffraction Wide angle X-ray scattering patterns were obtained for artemisinin and the solid dispersions in order to study the crystalline status of artemisinin in the formulations and are shown in Fig. 5. Artemisinin is a highly crystalline powder that has characteristic sharp peaks at diffraction angles of 11.96◦ , 14.54◦ , 20.24◦ , 21.92◦ , and 28.27◦ at the 2 scale (Shahzad et al., 2012). The solid dispersions show a significant peak reduction with increasing polymer concentration indicating the transformation of crystalline artemisinin to its amorphous state. These results are in accordance with the previous studies where artemisinin was transformed into an amorphous form in the solid dispersion system with PVP (Shahzad et al., 2012). In case of lyophilised dispersions, although the peak intensity was reduced as the polymer concentration was increased, peaks were still observed for the LDs containing higher concentrations of polymers (2:8 and 1:9). However, this fraction was slightly higher (not more than 2%) than those observed for the CD. These results are consistent with the DSC results where a slightly higher percentage crystallinity of artemisinin was observed for lyophilised dispersions. 3.1.4. FT-IR analysis The interaction between the drug and carrier often leads to peculiar changes in the FT-IR profile. Therefore, FT-IR spectra (Fig. 6) were recorded for pure artemisinin as well as its conventional and lyophilised dispersions. The FT-IR spectra of conventional solid dispersions and lyophilised dispersions of artemisinin (recall the polymer matrix comprises PVP and PEG) showed a significant broadening of the O H stretching vibration peak characteristic of PEG (large band between 3481 cm−1 and 3118 cm−1 ) and a C N stretching vibration peak characteristic of PVP (1239 cm−1 ). A band at 876 cm−1 (O O C linkage), which is characteristic of artemisinin, was observed without shifting. All the characteristic FT-IR peaks of artemisinin were visible without shifting in all formulation systems, suggesting an absence of significant specific interaction between artemisinin and the binary mixture of PVP and PEG. These findings were in good agreement to the previously published report where no interactions were observed between artemisinin and PVP within the formulations (Shahzad et al., 2012).
The effects of PVP and PEG on the solubility of artemisinin were investigated in two solvents, namely water and the phosphate buffered saline PBS (pH 7.4). The aqueous solubility of pure artemisinin was found to be 12.9 ± 0.2 g/mL, which increased to 15.7 ± 0.5 g/mL in PBS. The solid dispersions of artemisinin with various drug to polymer ratios (6:4, 5:5, 3:7, 2:8, and 1:9) exhibited a substantial enhancement (up to about 50 fold) in artemisinin solubility in the two solvents compared to that of the pure drug (p < 0.05). The solubility and corresponding solubility enhancement ratios (ERSol ) are presented in Table 2. Although, the presence of hydrophilic polymers increased the solubility of artemisinin in water and phosphate buffered saline, yet no clear relationship exists between polymer concentration and solubility enhancement. In fact, the solubility of artemisinin started to increase to a maximum value for conventional (3:7) and lyophilised dispersion (3:7) and then started to decrease with further increase of the polymers. Solubility enhancement can be explained on the basis of several mechanisms. We believe that primarily this comes from the drug being present mostly in the amorphous form in the polymer matrices. The amorphous form can have a substantially higher solubility than the crystalline forms. Any crystalline component of the drug in the polymer matrix is likely to be released in a fine colloidal form, which too will yield a higher solubility as described by the Ostwald–Freundlich and the Kelvin equations (Grant and Brittian, 1995). Enhanced solubility may also result from solubilisation action of the hydrophilic polymers.
3.3. Permeation studies For permeation studies, the solid dispersions were dissolved up to their saturated limits in two solvents, namely water and PBS, and the permeation of the artemisinin was investigated from these vehicles through rabbit skin using Franz type diffusion cells. The cumulative amount of drug permeated (g/cm2 ) as a function of time is shown in Fig. 7a–d. The steady-state flux (J) was calculated from the slope linear of the amount permeated vs. time graph. Since the permeation profiles appeared to be non-linear, therefore, the steepest part of the curve (linear portion) was used for the steady state flux determination. The results show a significant difference of artemisinin permeation from the saturated solution of solid dispersions as compared to the saturated solution of the drug alone. Difference in permeation patterns were also observed for saturated solutions of CDs prepared in water and PBS. This anomalous behaviour could be a result of existence of the non-sink conditions when CD saturated solutions in PBS were evaluated for artemisinin permeation. However, if there was any other hidden phenomena existed, this would also effect the permeation pattern of LDs saturated solutions. No significant difference in permeation pattern was observed for LDs saturated solutions prepared in water and PBS. The steady state flux for the pure artemisinin was found to be 0.16 ± 0.02 (g cm−2 min−1 ) while the permeability coefficient was 0.8 × 10−4 cm min−1 . In all cases, the total amount of artemisinin
202
Y. Shahzad et al. / International Journal of Pharmaceutics 457 (2013) 197–205
Fig. 5. XRD patterns of conventional dispersions (CD) and lyophilised dispersions (LD).
Fig. 6. FT-IR spectra of artemisinin (ART) and its various solid dispersion formulations.
Table 2 Solubility (g/mL) of artemisinin from CDs and LDs in two solvents. Drug:polymer ratio
Conventional dispersions (CD) Water
6:4 5:5 3:7 2:8 1:9
474.0 495.3 515.2 477.5 469.1
± ± ± ± ±
5.7 6.4 3.2 9.2 7.9
Lyophilised dispersions (LD)
ERSol
PBS
36.7 40.0 38.4 37.0 36.4
560.8 570.8 621.7 616.5 607.4
The results are presented as mean ± standard error; n = 5.
± ± ± ± ±
12.1 8.1 12.4 5.8 11.7
ERSol
Water
35.7 36.4 41.5 39.3 38.7
588.3 676.7 755.4 665.2 628.3
± ± ± ± ±
9.9 17.9 14.8 11.1 13.3
ERSol
PBS
45.6 52.5 58.6 51.7 48.7
669.2 873.4 889.2 797.1 781.3
ERSol ± ± ± ± ±
13.8 19.0 17.9 5.9 13.2
42.6 55.6 56.6 50.8 49.7
Y. Shahzad et al. / International Journal of Pharmaceutics 457 (2013) 197–205
203
Fig. 7. Cumulative amount of artemisinin permeated from (a) saturated solutions of CDs prepared in water, (b) saturated solutions of CDs prepared in PBS, (c) saturated solutions of LDs prepared in water and (d) saturated solutions of LDs prepared in PBS.
Table 3 Permeation parameters of artemisinin form conventional dispersion in two solvents. Drug:polymer dispersion
Saturated solutions in water J (g cm−2 min−1 )
CD (6:4) CD (5:5) CD (3:7) CD (2:8) CD (1:9)
1.12 1.32 1.41 1.43 1.52
± ± ± ± ±
0.14 0.10 0.06 0.09 0.22
Saturated solutions in PBS
KP × 10−4 (cm min−1 ) 5.6 6.6 7.1 7.2 7.6
± ± ± ± ±
0.1 0.2 0.4 0.9 0.6
permeated over 8 h of study from solid dispersion vehicles was significantly increased (5–11 fold increase) (p < 0.05) as compared to the pure drug permeation profile. The permeation parameters of artemisinin from its solid dispersions are presented in Tables 3 and 4. The flux enhancement ratio ER was calculated from the flux rate for the solid dispersion vehicles relative to that from the pure drug solution. The ER for the conventional dispersions in water varied from about 5–10 fold in both water and PBS. For the lyophilised dispersion the enhancement was about 5–11 fold in both water and PBS, which is marginally higher than that from the conventional dispersions. This enhancement in the flux was dependent on the polymer content in the formulation, resulting in an increase in ER with increasing polymer content.
ER (J)
J (g cm−2 min−1 )
7.0 8.3 8.8 8.9 9.5
0.87 0.88 1.02 1.16 1.33
± ± ± ± ±
0.11 0.14 0.20 0.21 0.10
KP × 10−4 (cm min−1 ) 4.4 4.4 5.1 5.8 6.7
± ± ± ± ±
0.3 0.5 0.5 0.2 0.3
ER (J) 5.4 5.5 6.4 7.3 8.3
One-way ANOVA confirmed a significant difference (p < 0.05) in the flux values of artemisinin from the saturated solutions of the various solid dispersions in the two solvents. The pairwise Turkey’s test revealed that the flux values obtained from saturated solutions of the conventional dispersions in water and PBS were significantly different (p < 0.05) from each other as well as from that of the lyophilised dispersions in water and PBS. However, there was no significant difference (p > 0.05) observed in the fluxes of artemisinin from the various ratios of lyophilised dispersions in water and PBS apart from the one with highest polymer content (1:9). Overall, both formulation systems (CDs and LDs) produced flux values of artemisinin that were significantly higher than that from the artemisinin alone. Furthermore, the flux and the associated permeability coefficient increase with increase in the polymer
Table 4 Permeation parameters of artemisinin form lyophilised dispersions in two solvents. Drug:polymer dispersion
Saturated solutions in water −2
J (g cm LD (6:4) LD (5:5) LD (3:7) LD (2:8) LD (1:9)
1.04 1.14 1.16 1.42 1.43
± ± ± ± ±
min
0.20 0.10 0.11 0.20 0.20
−1
)
Saturated solutions in PBS −4
KP × 10 5.2 5.7 5.8 7.1 7.2
± ± ± ± ±
0.2 0.5 0.4 0.7 0.5
−1
(cm min
)
ER (J)
J (g cm−2 min−1 )
6.5 7.1 7.3 8.9 8.9
1.03 1.06 1.33 1.41 1.86
± ± ± ± ±
0.30 0.10 0.11 0.20 0.26
KP × 10−4 (cm min−1 ) 5.2 5.3 6.7 7.1 9.3
± ± ± ± ±
0.3 0.5 0.2 0.1 0.3
ER (J) 6.4 6.6 8.3 8.8 11.6
204
Y. Shahzad et al. / International Journal of Pharmaceutics 457 (2013) 197–205
content in the formulation system. Increase in the KP values with increasing polymer content was a result of perturbation of the skin lipid bilayers, consequently enhancing the partitioning of permeant in that region. The fact is pyrrolidones have tendency to penetrate into the hydrophobic regions of the SC and reduce the barrier function in these areas which allows influx of permeant via polar routes of the stratum corneum (Southwell and Barry, 1983). This was consistent with the previously published reports where permeation rate of poorly water soluble drugs like celecoxib and meloxicam was increased as a function of polymer concentration (Saleem et al., 2010; Soliman et al., 2011). The fact that we are getting a substantial increase in the flux of artemisinin from the solid dispersion vehicles suggests that the cause is the massive increase in supersaturation (i.e. increase in the thermodynamic activity of the solution) that is attained when the dispersions are dissolved in the vehicle. Had the polymers merely increased the solubility by a solubilisation effect, the solution activity would be equivalent to that of the pure artemisinin saturated solution, which would have yielded a similar flux. There is the possibility that the polymers are acting as penetration enhancers (see Williams and Barry, 2012) but this effect is not easy to decouple (Williams and Barry, 2012). The important question is, is the enhanced permeability adequate for the transdermal delivery of artemisinin? The maximum achievable flux is about 1.86 g cm−2 min−1 . Here we used the predictive approach to estimate the steady state plasma concentration (Css ) that would penetrate after topical application using following equation (Flynn and Stewart, 1988): Css =
Jmax × A Clp
where Jmax is the maximum achievable flux, A is the hypothetical area of application and Clp is the plasmatic clearance. The Clp for artemisinin after a 500 mg oral dose is 130 mL/min (Ashton et al., 1998). Taking into account a 100 cm2 hypothetical area for a transdermal patch, the predicted plasma concentration turn out to be 1.43 g/mL which is higher than the therapeutic range of 0.23–0.58 g/mL (Chan et al., 1997) confirming the potential transdermal application of artemisinin. 4. Conclusion Artemisinin is a poorly soluble drug prone to first-pass metabolism, for which the bioavailability can be low and erratic. We have investigated the use of solid dispersions for enhancing its solubility and associated supersaturation with a view to increasing its permeability through skin. Solid dispersions were prepared using a 50:50 blend of PVP-K30 and PEG 4000 with the ratio of drug to polymer being varied. The dispersions were characterised using physical methods to ascertain whether the dispersed drug was crystalline or amorphous. Molecular dynamics simulations were also employed to estimate the miscibility of the drug and the polymer mix. The simulations revealed that the drug and the polymers were miscible, which rationalised the finding the observation that the dispersed drug was largely molecular dispersed or amorphous. The dispersions should substantial increase in solubility and hence associated supersaturation, and resulting in enhanced permeation through rabbit skin. Conflict of interest The authors declare no competing financial interest. Acknowledgments The authors thank the Bahauddin Zakariya University Multan, Pakistan for providing funding to conduct this research. The authors
would like to thank Prof. Jamshed Anwar for his help on MD simulations. We are also grateful to Malcolm Cutler, Managing Director FSC Development Services Ltd, UK for his support in this work. References Andrews, G.P., Abudiak, O.A., Jones, D.S., 2010. Physicochemical characterization of hot melt extruded bicalutamide–polyvinylpyrrolidone solid dispersions. J. Pharm. Sci. 99, 1322–1335. Ashton, M., Gordi, T., Hai, T.N., Van Huong, N., Sy, N.D., Nieu, N.T., Huong, D.X., Johansson, M., Công, L.D., 1998. Artemisinin pharmacokinetics in healthy adults after 250, 500 and 1000 mg single oral doses. Biopharm. Drug Dispos. 19, 245–250. Bley, H., Fussnegger, B., Bodmeier, R., 2010. Characterization and stability of solid dispersions based on PEG/polymer blends. Int. J. Pharm. 390, 165–173. Chamchod, F., Beier, J.C., 2013. Modeling Plasmodium vivax: relapses, treatment, seasonality, and G6PD deficiency. J. Theor. Biol. 316, 25–34. Chan, K.L., Yuen, K.H., Jinadasa, S., Peh, K.K., Toh, W.T., 1997. A high-performance liquid chromatography analysis of plasma artemisinin using a glassy carbon electrode for reductive electrochemical detection. Planta Med. 63, 66–69. Craig, D.Q.M., 2002. The mechanisms of drug release from solid dispersions in watersoluble polymers. Int. J. Pharm. 231, 131–144. Eastman, R.T., Fidock, D.A., 2009. Artemisinin-based combination therapies: a vital tool in efforts to eliminate malaria. Nat. Rev. Microbiol. 7, 864–874. Flynn, G.L., Stewart, B., 1988. Percutaneous drug penetration: choosing candidates for transdermal development. Drug Dev. Res. 13, 169–185. Gordi, T., Nielsen, E., Yu, Z., Westerlund, D., Ashton, M., 2000. Direct analysis of artemisinin in plasma and saliva using coupled-column high-performance liquid chromatography with a restricted-access material pre-column. J. Chromatogr. B 742, 155–162. Grant, D.J.W., Brittian, H.G., 1995. In: Brittian, H.G.E. (Ed.), Physical Characterisation of Pharmaceutical Solids. Marcel Dekker, New York. Greenhalgh, D.J., Williams, A.C., Timmins, P., York, P., 1999. Solubility parameters as predictors of miscibility in solid dispersions. J. Pharm. Sci. 88, 1182–1190. Gupta, J., Nunes, C., Vyas, S., Jonnalagadda, S., 2011. Prediction of solubility parameters and miscibility of pharmaceutical compounds by molecular dynamics simulations. J. Phys. Chem. B 115, 2014–2023. Hildebrand, J., Scott, R., 1950. The Solubility of Nonelectrolytes. Reinhold Publ. Corp, New York. Klayman, D.L., 1985. Qinghaosu (Artemisinin): an antimalarial drug from China. Science 228, 1049–1055. Langer, M., Höltje, M., Urbanetz, N.A., Brandt, B., Höltje, H.D., Lippold, B.C., 2003. Investigations on the predictability of the formation of glassy solid solutions of drugs in sugar alcohols. Int. J. Pharm. 252, 167–179. Lima, Á.A.N., Soares-Sobrinho, J.L., Silva, J.L., Corrêa-Júnior, R.A.C., Lyra, M.A.M., Santos, F.L.A., Oliveira, B.G., Hernandes, M.Z., Rolim, L.A., Rolim-Neto, P.J., 2011. The use of solid dispersion systems in hydrophilic carriers to increase benznidazole solubility. J. Pharm. Sci. 100, 2443–2451. Maniruzzaman, M., Morgan, D.J., Mendham, A.P., Pang, J., Snowden, M.J., Douroumis, D., 2013. Drug–polymer intermolecular interactions in hot-melt extruded solid dispersions. Int. J. Pharm. 443, 199–208. Meidan, V.M., Pritchard, D., 2010. A two-layer diffusive model for describing the variability of transdermal drug permeation. Eur. J. Pharm. Biopharm. 74, 513–517. Rawlinson, C.F., Williams, A.C., Timmins, P., Grimsey, I., 2007. Polymer-mediated disruption of drug crystallinity. Int. J. Pharm. 336, 42–48. Rigby, D., Sun, H., Eichinger, B.E., 1997. Computer simulations of poly(ethylene oxide): force field, PVT diagram and cyclization behaviour. Polym. Int. 44, 311–330. Sahoo, N.G., Abbas, A., Judeh, Z., Li, C.M., Yuen, K.H., 2009. Solubility enhancement of a poorly water-soluble anti-malarial drug: experimental design and use of a modified multifluid nozzle pilot spray drier. J. Pharm. Sci. 98, 281–296. Sahoo, N.G., Kakran, M., Li, L., Judeh, Z., Müller, R.H., 2011. Dissolution enhancement of a poorly water-soluble antimalarial drug by means of a modified multi-fluid nozzle pilot spray drier. Mater. Sci. Eng. C 31, 391–399. Saleem, M.A., Bala, S., Liyakat Aeajaz, A., 2010. Effect of different carriers on in vitro permeation of meloxicam through rat skin. Indian J. Pharm. Sci. 72, 710–718. Serajuddln, A.T.M., 1999. Solid dispersion of poorly water-soluble drugs: early promises, subsequent problems, and recent breakthroughs. J. Pharm. Sci. 88, 1058–1066. Shah, S.N.H., Mehboob, E.R., Shahzad, Y., Badshah, A., Meidan, V.M., Murtaza, G., 2012. Developing an efficacious diclofenac diethylamine transdermal formulation. J. Food Drug Anal. 20, 464–470. Shahzad, Y., Shah, S.N.H., Ansari, M.T., Riaz, R., Safdar, A., Hussain, T., Malik, M., 2012. Effects of drug–polymer dispersions on solubility and in vitro diffusion of artemisinin across a polydimethylsiloxane membrane. Chin. Sci. Bull. 57, 1685–1692. Shahzad, Y., Shah, S.N.H., Atique, S., Ansari, M.T., Bashir, F., Hussain, T., 2011. The evaluation of coated granules to mask the bitter taste of dihydroartemisinin. Braz. J. Pharm. Sci. 47, 323–330. Soliman, S.M., Malak, N.S.A., El Gazayerly, O.N., Rehim, A.A.A., 2011. Preparation of celecoxib solid dispersions for dermal application: in vitro characterization and skin irritation test. J. Drug Deliv. Sci. Technol. 21, 509–516. Southwell, D., Barry, B.W., 1983. Penetration enhancers for human skin: mode of action of 2-pyrrolidone and dimethylformamide on partition and diffusion
Y. Shahzad et al. / International Journal of Pharmaceutics 457 (2013) 197–205 of model compounds water, n-alcohols, and caffeine. J. Invest. Dermatol. 80, 507–514. Svensson, U.S.H., Sandström, R., Carlborg, O., Lennernäs, H., Ashton, M., 1999. High in situ rat intestinal permeability of artemisinin unaffected by multiple dosing and with no evidence of P-glycoprotein involvement. Drug Metab. Dispos. 27, 227–232. Thaithong, S., Beale, G.H., 1985. Susceptibility of Thai isolates of Plasmodium falciparum to artemisinine (qinghaosu) and artemether. Bull. World Health Organ. 63, 617–619. Titulaer, H.A.C., Zuidema, J., Lugt, C.B., 1991. Formulation and pharmacokinetics of artemisinin and its derivatives. Int. J. Pharm. 69, 83–92. Van Krevelen, D.W., Hoftyzer, P., 1976. Properties of Polymers: Their Estimation and Correlation with Chemical Structure. Elsevier, New York. Williams, A.C., Barry, B.W., 2012. Penetration enhancers. Adv. Drug Deliv. Rev. 64, 128–137.
205
Wong, J.W., Yuen, K.H., 2001. Improved oral bioavailability of artemisinin through inclusion complexation with - and ␥-cyclodextrins. Int. J. Pharm. 227, 177– 185. World Health Organization, 2010a. Guidelines for the treatment of malaria. In: Diagnosis of Malaria, second ed, 10–11. World Health Organization, 2010b. World Malaria Report. World Health Organization, Geneva. Xiang, T.-X., Anderson, B.D., 2013. Molecular dynamics simulation of amorphous indomethacin–poly(vinylpyrrolidone) glasses: solubility and hydrogen bonding interactions. J. Pharm. Sci. 102, 876–891. Zhao, S.S., Zeng, M.Y., 1985. Spectrometric high-pressure liquid chromatographic studies on analytical chemistry of Qinghaosu. Planta Med. 3, 233–237. Zhu, Q., Harris, M.T., Taylor, L.S., 2012. Modification of crystallization behavior in drug/polyethylene glycol solid dispersions. Mol. Pharm. 9, 546–553.