Tailoring controlled-release oral dosage forms by combining inkjet ...

European Journal of Pharmaceutical Sciences 47 (2012) 615–623

Contents lists available at SciVerse ScienceDirect

European Journal of Pharmaceutical Sciences journal homepage: www.elsevier.com/locate/ejps

Tailoring controlled-release oral dosage forms by combining inkjet and flexographic printing techniques Natalja Genina a,⇑, Daniela Fors b, Hossein Vakili a, Petri Ihalainen b, Leena Pohjala a, Henrik Ehlers a, Ivan Kassamakov c, Edward Haeggström c, Pia Vuorela a, Jouko Peltonen b, Niklas Sandler a a b c

Pharmaceutical Sciences Laboratory, Department of Biosciences, Abo Akademi University, Tykistökatu 6A, FI-20520 Turku, Finland Center of Excellence for Functional Materials, Laboratory of Physical Chemistry, Abo Akademi University, Porthaninkatu 3-5, FI-20500 Turku, Finland Electronics Research Laboratory, Department of Physics, Division of Materials Physics, University of Helsinki, P.O. Box 64, FI-00014 Helsinki, Finland

a r t i c l e

i n f o

Article history: Received 11 May 2012 Received in revised form 18 July 2012 Accepted 28 July 2012 Available online 9 August 2012 Keywords: Inkjet printing Flexographic printing Controlled-release Coating Tailored release Cytotoxicity

a b s t r a c t We combined conventional inkjet printing technology with flexographic printing to fabricate drug delivery systems with accurate doses and tailored drug release. Riboflavin sodium phosphate (RSP) and propranolol hydrochloride (PH) were used as water-soluble model drugs. Three different paper substrates: A (uncoated woodfree paper), B (triple-coated inkjet paper) and C (double-coated sheet fed offset paper) were used as porous model carriers for drug delivery. Active pharmaceutical ingredient (API) containing solutions were printed onto 1 cm  1 cm substrate areas using an inkjet printer. The printed APIs were coated with water insoluble polymeric films of different thickness using flexographic printing. All substrates were characterized with respect to wettability, surface roughness, air permeability, and cell toxicity. In addition, content uniformity and release profiles of the produced solid dosage forms before and after coating were studied. The substrates were nontoxic for the human cell line assayed. Substrate B was smoothest and least porous. The properties of substrates B and C were similar, whereas those of substrate A differed significantly from those of B, C. The release kinetics of both printed APIs was slowest from substrate B before and after coating with the water insoluble polymer film, following by substrate C, whereas substrate A showed the fastest release. The release rate decreased with increasing polymer coating film thickness. The printed solid dosage forms showed excellent content uniformity. So, combining the two printing technologies allowed fabricating controlled-release oral dosage forms that are challenging to produce using a single technique. The approach opens up new perspectives in the manufacture of flexible doses and tailored drug-delivery systems. Ó 2012 Elsevier B.V. All rights reserved.

1. Introduction Delivering an accurate dose of an API as well as controlling its release are key factors regarding API efficacy in pharmacotherapy. For today’s manufacturing of oral solid dosage forms, such as tablets and capsules, the possibilities to produce personalized medicines and flexible doses are limited. Commercial drugs are usually present in merely two or possibly three different strengths, which may insufficiently cover the needs of patients of different age. For example, for pediatric patients, the therapeutic dose is calculated based on the weight of the child, meaning that the strength of the active component should be selectable. Moreover, elderly people often consume several drugs at the same time. Preferably, these therapies could be fabricated as multidrug systems with individually adjusted doses and release profiles. With current dosage forms, tunable dosing and tailoring multidrug product proper⇑ Corresponding author. Tel.: +358 2 215 4018; fax: +358 2 241 0014. E-mail address: natalja.genina@abo.fi (N. Genina). 0928-0987/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ejps.2012.07.020

ties is difficult. In this context, flexibility in the production of more individualized medicines would be beneficial. Printing technologies such as inkjet and flexographic printing, offer possibilities to deposit a variety of functional materials onto different types of carrier surfaces or substrates (Ihalainen et al., 2011; Määttänen et al., 2011, 2010). The main advantage of inkjet printing includes the ability to dispense uniform droplets in the picoliter range with high degree of accuracy, where droplet formation and 3D placement can be controlled by a personal computer (Le, 1998). However, the ink formulation has to be designed with respect to its viscosity and surface tension to guarantee continuous printing and high reproducibility of the forming droplets (Di Risio and Yan, 2007). The use of this technology for printing functional materials has grown over the past years. A variety of biomaterials have been transferred onto solid support structures in single or multiple layers in either 2D or 3D arrangements at ambient temperature with the bioactivity being retained for extended periods (Abe et al., 2008; De Gans et al., 2004; Di Risio and Yan, 2007; Gonzalez-Macia et al., 2010; Hasenbank et al., 2008; Ilkhanizadeh

616

N. Genina et al. / European Journal of Pharmaceutical Sciences 47 (2012) 615–623

et al., 2007; Jordan et al., 1996). The food industry has used inkjet printers to decorate baked goods and other foodstuffs (Stewart and Gratiot, 2003). Flexographic printing offers ease of use and great versatility regarding ink formulation. Flexography is a direct rotary printing method, where different types of inks can be printed onto nearly any substrate (De Micheli, 2000; Kipphan, 2001). Unfortunately, flexographic printing lacks dosing precision, which can be a problem when preparing pharmaceutical dosage forms. In addition, every ink has to be printed separately, which makes it impossible to print multicomponent systems simultaneously. By combining both techniques one could potentially have both high dosing precision and the flexibility. Recently, simultaneous use of these two printing techniques was applied to prepare functional paper for point-of-care diagnostics tests and (bio)sensors in food packaging (Airo and Erho, 2010; Määttänen et al., 2011). Inkjet printing technology has been used in pharmaceutics (Goodall et al., 2002; Kastra et al., 2000; Melèndez et al., 2008; Rowe et al., 2000; Voura et al., 2011). Inkjet printing to directly deposit drug solutions or suspensions containing API and excipients onto carrier materials such as porous substrates and biodegradable films, offers new ways for fabrication of oral solid dosage forms with controlled API crystallinity (Buanz et al., 2011; Sandler et al., 2011). Recently, drug release profiles have been altered by inkjet printing of API(s) and polymer(s) mixtures at different molar ratios (Scoutaris et al., 2012, 2011). Inkjet printing of multicomponent systems containing several APIs and time-release layers of different polymers was proposed by Voura et al. (2011). These ideas open new perspectives when designing individual dosage forms. The goal of the present work was to determine whether one could combine inkjet and flexographic printing to fabricate pharmaceutical solid dosage forms with controlled release properties of drug substances (Fig. 1). Propranolol hydrochloride (PH) and riboflavin sodium phosphate (RSP) were deposited by inkjet printing on different porous paper substrates, and the printed drugs were then coated with ethylcellulose (EC) polymer using flexographic printing. The main aim was to study and gain understanding of the effects of different substrate properties as well as the impact of the printed polymer coating layers on the release of the drugs. To best of our knowledge this is the first report where inkjet printing is combined with flexographic printing to produce drug delivery systems with accurate doses and tailored drug release. 2. Materials and methods

sodium salt, Ph. Eur., Fluka Analytical, Sigma–Aldrich, France). Propylene glycol (PG) (P99.5%, Sigma–Aldrich, Germany), glycerol (P99.5%, J.T. Baker, Deventer, Holland), ethanol (P99.7%, Etax Aa, Altia OYj, Finland) and purified water (Milli-Q) were used in the preparation of ink formulations. Water insoluble ethylcellulose (EC) (E8003, Sigma–Aldrich, USA; ethoxy content, 48–49.5%; viscosity, 45 cP) and ethanol (P96.1%, Etax A, Altia OYj, Finland) were used to prepare a coating solution. CellTiter Glo Luminescent Cell viability Assay reagent (Promega, Madison, USA) and resazurin (Sigma–Aldrich, St. Louis, MO, USA) were used in the cell viability assays. 2.2. Ink formulations and coating solution Two different ink formulations were used to print the drug substances onto the substrates. The propranolol containing ink was prepared by dissolving 50 mg/ml propranolol powder in PG:water mixture (30:70, vol%). The RSP containing ink was made by mixing 31.5 mg/ml of RSP powder with glycerol:ethanol:water (10:10:80, vol%). The ink solutions were filtered with 0.45 lm and 0.2 lm polypropylene membrane filters (Whatman™, GE Healthcare, Piscataway, NJ, USA) before printing. The polymeric coating solution was prepared by dissolving 5% (wt%) EC in ethanol under continuous stirring. 2.3. Viscosity and surface tension Viscosity measurements were conducted with a Malvern Kinexus rheometer (Malvern, UK) at 23 °C (only for EC ink) and 30 °C. Surface tension values of the ink solutions were measured using the pendant drop method at 22 °C (CAM 200; KSV Instruments Ltd., Helsinki, Finland). The viscosity and surface tension measurements were performed with the ink solution without APIs. The tests were done in triplicate. 2.4. Printing substrates Three model paper substrates provided by Sappi Fine Paper Europe: A (alkyl ketene dimer-sized uncoated woodfree paper), B (triple-coated inkjet paper) and C (double-coated sheet fed offset (SFO) paper) were studied. A water impermeable polyethylene terephthalate (PET) film (MylarÒ A, Dupont Teijin Films Europe, Luxembourg) was used as reference substrate. All substrates are commercially available.

2.1. Materials 2.5. Inkjet printing of drug formulations Two different water soluble model drugs were used: propranolol hydrochloride (PH) (P98%, Sigma–Aldrich, China) and riboflavin sodium phosphate (RSP) (riboflavin 50 -monophosphate

Inkjet printing was performed with a Dimatix DMP-2800 inkjet printer (Fujifilm Dimatix Inc., Santa Clara, California). In the

Fig. 1. Schematic representation of the basic concept of combining inkjet and flexographic printing techniques.

N. Genina et al. / European Journal of Pharmaceutical Sciences 47 (2012) 615–623

printer, a MEMS-based cartridge-styled printhead with 16 nozzles linearly spaced at 254 lm that produce a typical drop size of 10 pl was used. Each cartridge could hold 1.5 ml of ink. Printing was performed in ambient conditions (45.5 ± 5% relative humidity (RH) and 21 ± 1 °C) with a single nozzle (20 lm orifice diameter) at firing voltages of 30 V and 35 V for PH and RSP, respectively. The cartridge temperature was 30 °C, and the drops were deposited at a drop spacing of 10 lm. The drugs were printed in squares (n = 10–20 for each substrate and ink formulation) of 1 cm  1 cm, making according to the Dimatix software the total number of drops in one square equal to 1,002,001, which theoretically corresponded to approximately 10.02 ll cm2, when assuming a drop volume of 10 pl.

2.6. Flexographic printing of coating solution Flexographic printing of the polymeric coating layers was carried out with a laboratory scale printability tester (IGT Global Standard Tester 2, IGT Testing system, The Netherlands) (Fig. 2). The coating ink is applied to a small area of the anilox roll behind the doctor blade. During printing the ink is transferred onto the anilox roll engraved with thousands of small cells. The volume of the cells on the anilox roll defines the amount of ink that is transferred onto the printing plate containing a relief pattern. The excess ink from the anilox roll is removed by the doctor blade. When the printing starts, the anilox roll accelerates to its printing speed and rotates to distribute the ink, simultaneously as the printing plate and impression roll rotate through one revolution to transfer the ink onto the substrate. In this study, printing was carried out at 23 °C and 50% RH. The amount of EC ink was controlled by an anilox roll having a cell angle of 45°, a cell volume of 20 ml/m2 and a line count of 40 l/cm. An unpattern photopolymer plate (Ohkaflex, Espoon Painolaatta, Finland) was utilized to transfer the ink to the printing substrate. The pressure between the anilox roll and the printing plate was 100 N, while the pressure between the printing plate and the substrate was 50 N. The printing speed was 0.5 m/s. To study the effect of coating on drug release 5–30 layers of EC were printed on both sides of the substrate. All printing was per-

617

formed so that, after printing five ink layers, more ink was added with a syringe onto the ink reservoir area on the anilox roll. Typically, the anilox roll was washed after printing 10–15 ink layers. 2.7. Storage of the printed and polymer-coated samples The samples were stored in a desiccator over silica granules (0– 10% relative humidity). The desiccator was covered with an aluminum foil to minimize the destructive effect of light, and was stored at 4–8 °C. 2.8. Analysis of content uniformity Printed drug areas on the A substrate (n = 11 for PH and n = 4 for RSP) were immersed into 50 ml of 0.1 N HCl. The volumetric flasks were shaken vigorously and were let to stay for 1 h. Consequently, the absorbance of the obtained solutions was measured at 220 nm for propranolol and 267 nm for riboflavin with a UV–VIS spectrophotometer (PerkinElmer, Lambda 25, USA). 2.9. Dissolution experiments The dissolution rate of the APIs from the paper substrates was determined using the USP paddle method. The dissolution testing was carried out using the Sotax AT7 Smart dissolution tester (SOTAX, Switzerland) and a UV/Vis spectrophotometer (PerkinElmer, Lambda 25, USA). 500 ml of 0.1 N HCl was used as a dissolution medium (pH = 1) to mimic conditions in the stomach. The paper samples were put, printed drug side facing outwards, in spiral capsule sinkers to prevent floating. The dissolution experiments were carried out at 267 nm for RSP and 220 nm for PH. The rotation speed was 50 rpm, the temperature was 37 ± 0.5 °C. 2.10. Optical microscopy and digital imaging Optical microscopy (OM, Evos XL, AMG, USA) in connection with a digital camera (DC) was used to capture the top view as well as the cross-section of the printed samples at magnifications of 10 and 40. The printed samples were photographed using another DC (Canon, SX230 HS, Japan). 2.11. Contact angle Contact angle measurements were conducted by applying a 2 ll drop of purified water onto the printing substrates and by then using a CAM 200 contact angle goniometer (KSV instruments Ltd., Helsinki, Finland). Measurements were conducted in ambient conditions (20% RH and 21 °C). 2.12. Scanning white light interferometry (SWLI)

Fig. 2. A picture of the flexography unit.

The surface topography of the samples were measured using a customized scanning white light interferometer with a 10 Mirau objective (Nikon CF IC Epi Plan DI; Japan), piezoelectric scanner (Physik Instrumente P-723.10; Karlsruhe Germany) and a high-resolution CCD-camera (Hamamatsu C11440 Orka Flash2.8, Hamamatsu City, Japan). The total magnification was 6.3. The dataanalysis and image construction was performed using Mountainsmap Imaging Topography 6.0-software (Digitalsurf, Avencon, France). Root mean square height (Sq), i.e. the standard deviation of the height distribution of the scanned area (0.8 mm  1.0 mm), was calculated according to ISO 25178 standard using Mountainsmap Imaging Topography 6.0-software.

618

N. Genina et al. / European Journal of Pharmaceutical Sciences 47 (2012) 615–623

2.13. Air permeability

3. Results and discussion

Bendtsen air permeation (ml/min) of the substrates was measured according to ISO standards: ISO 5636-3 Paper and board  Determination of air permeation (medium range)  Part 3: Bendtsen method. The measurements were automatically done in triplicate and only the arithmetic mean was provided by operating software.

3.1. Properties of the substrates

2.14. Thickness and specific volume Thickness (lm) and specific volume (cm3/g) of the substrates were measured according to ISO standards: ISO 534 Determination of thickness and apparent bulk density or apparent sheet density (EN 20534). The measurements were automatically done in triplicate and only the arithmetic mean was provided by operating software. The EC coating thickness was measured using a Digimatic caliper (Absolute™ Digimatic, Mitutoya Corporation, Kawasaki, Japan). Ten sheets of PET substrates, firstly, before and then, after coating with 30 layers of EC were stacked together and the thickness in mm was determined at 10 different spots. The thickness of a single layer was calculated. 2.15. Cell viability assays Human epithelial HL cells from respiratory tract (Cavallaro and Monto, 1972) were grown in RPMI1640 medium supplemented with 7.5% fetal bovine serum (FBS), 2 mM L-glutamine and 20 lg/ ml gentamycin. For the cell viability assays, the cells were seeded into 96-well plates at a density of 60,000 cells/well and incubated at 37 °C and 5% CO2 overnight before starting the exposure. The 1 cm2 paper samples were placed into 3 ml aliquots of the growth medium and incubated overnight at room temperature. The medium from the 96-wellplates containing cells was removed and aliquots of 200 ll/well of the samples were added in six replicates per plate. Usnic acid (200 lM) was added as a positive control in both assays (n = 4). The plates were incubated for 24 h after which the samples were removed and the wells were washed with 100 ll phosphate-buffered saline (PBS). To the plate used for ATP level determination, 50 ll PBS was added and the plate was equilibrated at room temperature. Then, 50 ll of CellTiter Glo Luminescent Cell viability Assay reagent was added and the plate was shaken for 10 min prior to measuring the luminescent signal in each well with a Varioskan Flash plate reader (Thermo Fischer Scientific Inc., Waltham, MA, USA). To the plate used for the resazurin assay, 20 lM resazurin solution in PBS was added into each well (V = 200 ll) and the plate was incubated at 37 °C for 2 h. The fluorescent signal (excitation/emission 560/590 nm) resulting from the reduced form of resazurin was read with the Varioskan Flash. The results were expressed as viability percentages normalized using a blank sample as a negative control. The obtained data was statistically treated, using Student’s unpaired t-test.

Three commercial paper grades (A, B and C) having different characteristics were chosen as model substrates for this study. Description of the papers and some of their physical properties are presented in Table 1. Depending on the intended end-use, paper is typically sized or coated to modify its physical characteristics and/or improve printability. Substrate A was an uncoated, sized paper, whose two sides differed in terms of surface roughness. Substrates B and C were double- or triple-coated papers, which had considerably smaller air permeability, specific volume and surface roughness than paper A. Sizing and coating of the paper had an influence on the contact angle and surface energetic properties of the substrate. This is illustrated in Fig. 3a, which represents the contact angle data of the abovementioned papers. Substrate A had, as expected, the higher contact angle than B and C substrates, since sizing typically provides hydrophobicity to the paper. On the other hand, substrates B and C were more hydrophilic owing to their coating composition. The wettability data revealed that the glossy side of the substrate A was more hydrophobic than the matt side. The contact angle of a water droplet on the matt face was 115° ± 4°, whereas for the glossy side this value was 124° ± 5° at 3 s from the beginning of the measurements. The root mean square height (Sq) values for the matt and glossy sides were 4.45 and 2.09 lm, respectively (Table 1). Clear differences in surface topography are seen in the SWLI images of the substrates (Fig. 4). Paper A was thinnest, followed by substrate C and substrate B, being the thickest (Table 1). The air permeability values for substrate A were similar for both surfaces. Substrate C permeated air two times faster than substrate B and ten times slower than substrate A. These differences were related both to distinctions in the paper manufacturing process as well as composition of the samples. In addition, substrate B possessed the smoothest surface (Sq = 180 nm), whereas for C substrate this value was 440 nm. The beginning of wetting of substrate C with water was faster than that of substrate B; however, the contact angles were similar to each other (Fig. 3a). The pure unprinted paper substrates were tested for cytotoxic effects on human cell viability, using two different methods: determination of intracellular ATP levels and resazurin reduction assay. Since the possible harmful effect of the samples is not known, two endpoint methods will deliver more reliable data than one as discussed by Pohjala et al. (2007a, 2007b). The cell viability was not below 90% after the exposure (Fig. 5). Consequently, none of the leachable components of the paper samples affected cell viability significantly (p > 0.05) in these conditions. The insoluble part of the papers, similar to pharmaceutically approved cellulose and its insoluble derivatives (Dahl, 2009; Guy, 2009; Podczeck, 2009), were considered to be of little toxic potential due to its excretion without systematic absorption by per oral administration.

Table 1 Properties of the substrates.

Substrate A, matt side Substrate A, glossy side Substrate B Substrate C a b

Grade type

Thickness (lm)a

Air permeability (ml/min)a

Specific volume (cm3/g)a

Sq (lm)b

Uncoated, sized paper Uncoated, sized paper Triple-coated inkjet paper, contains SB latex and PVA Double-coated SFO paper, contains SB latex and PVA

86.8 86.8 100.7 93.3

387 390 1.5 3

1.458 1.458 0.876 0.829

4.45 2.09 0.18 0.44

Measurements were done n = 3. Measurements were done n = 1 over scan range of 800 lm  1000 lm.

N. Genina et al. / European Journal of Pharmaceutical Sciences 47 (2012) 615–623

619

Fig. 3. Wettability properties: (a) pure substrates; (b) after printing of propranolol and (c) after printing of propranolol and coating with ethyl cellulose polymer film (30 layers). Each point represents the mean of three to six measurements.

Fig. 4. Visualization of the substrates measured with scanning white light interferometry: (a) substrate A, matt side; (b) substrate A, glossy side; (c) substrate B and (d) substrate C.

3.2. Properties of the ink formulations The properties of liquid ink are important in controlled and continuous printing. The main criteria for successful printing are optimal viscosity and surface tension of the fluid applied. The desired ink characteristics with respect to surface tension and dynamic viscosity are 25–50 mN/m and 1–30 mPa s for inkjet printing (Dimatix, 2010; Lee, 2003). Rheological properties and the surface tension of pure solvents such as water and ethanol are not suitable for inkjet printing. Therefore, surface tension and viscosity modifying agents are added to the formulation to reach optimal characteristics of the jetting liquids. The ink should also contain a moisturizing substance to prevent clogging of the firing nozzle in case of highly concentrated ink formulation or rapidly volatile solvents. Another important aspect is the solubility of the components in the printing fluid. Clear solutions are preferred in inkjet printing. Nanosuspensions can be jetted; however, the particle size range should be less than 200 nm to avoid blocking of the nozzle (Dimatix, 2010). In our study, we used previously optimized aqueous formulation, containing propylene glycol (PG) as a viscosity modifier

and moisturizer (Sandler et al., 2011) for water soluble API PH. PG was the most suitable additive if compared to other pharmaceutically approved liquid viscosity modifiers, such as glycerol, polyethylene glycol, and sodium carboxymethyl cellulose, to ensure continuous inkjet printing. However, this printing liquid was unsuitable for riboflavin. Despite its high solubility in water, RSP precipitated immediately in the presence of PG. Therefore, glycerol was used as a viscosity modifier and moisturizer for the RSP ink. Both ink formulations possessed viscosity and surface tension values in the desired range (Table 2). To obtain sustained release of the drug from the porous paper substrates, another ink formulation containing hydrophobic polymer ethylcellulose (EC) in alcohol was formulated and used to coat the printed APIs with water-insoluble films. However, ink-jetting the polymeric solution in ethanol was problematic: even 0.25% EC in 30% PG with ethanol as a solvent clogged the nozzles during printing, whereas printing 1% EC was impossible from the beginning. In addition, such a low concentration of EC with the presence of the water soluble PG would not provide a sufficient time-release barrier for the printed APIs. Hence, we employed contacting

620

N. Genina et al. / European Journal of Pharmaceutical Sciences 47 (2012) 615–623

layers of substrates B and C formed a densely packed structure (little light from the microscope passed through those substrates). The components of substrate A were loosely packed, where the substrate forming cellulose fibers could be observed. Cross-section images of the papers featuring printed API were taken to analyze the penetration depth of the ink into the substrate (Fig. 6, top). The API containing liquid penetrated deep into the structure in substrate A (both sides), whereas in substrates B and C, the API was concentrated near the surface. The paper coating and the dense structure of substrates B and C prevented ink penetration into the substrates. None of the substrates passed the ink formulation through as the back side of the printed areas remained always free of APIs. Printing water soluble APIs improved more than twice the wettability of all substrates (Fig. 3 b). This further confirmed at least partial presence of printed APIs on the paper surfaces. 3.4. API content Fig. 5. Cell viability measured through cellular ATP levels and metabolic activity of the cell culture (resazurin reduction test). Data are presented as mean ± standard deviation (n = 4–6). p