Formulating Inhalable Dry Powders Using Two-Fluid

0 downloads 0 Views 2MB Size Report
tered problem occurs when the solubility of the API and the excipient(s) differs ... ing this approach is that organic solvents are not required to dissolve the ...... D. Co-spray-dried mannitol-ciprofloxacin dry powder inhaler formulation for cystic.
Pharm Res (2018) 35:247 https://doi.org/10.1007/s11095-018-2509-z

RESEARCH PAPER

Formulating Inhalable Dry Po wders Using Two -Fluid and Three-Fluid Nozzle Spray Drying Donglei Leng 1 & Kaushik Thanki 1 & Camilla Foged 1 & Mingshi Yang 1,2

Received: 16 July 2018 / Accepted: 24 September 2018 # Springer Science+Business Media, LLC, part of Springer Nature 2018

ABSTRACT Purpose The spray drying process is widely applied for pharmaceutical particle engineering. The purpose of this study was to investigate advantages and disadvantages of two-fluid nozzle and three-fluid nozzle spray drying processes to formulate inhalable dry powders. Methods Budesonide nanocomposite microparticles (BNMs) were prepared by co-spray drying of budesonide nanocrystals suspended in an aqueous mannitol solution by using a twofluid nozzle spray drying process. Budesonide-mannitol microparticles (BMMs) were prepared by concomitant spray drying of a budesonide solution and an aqueous mannitol solution using a spray drier equipped with a three-fluid nozzle. The resulting dry powders were characterized by using X-ray powder diffraction (XRPD), differential scanning calorimetry (DSC), dynamic mechanical analysis (DMA) and Raman microscopy. A Next Generation Impactor was used to evaluate the aerodynamic performance of the dry powders. Results XRPD and DMA results showed that budesonide remained crystalline in the BNMs, whereas budesonide was amorphous in the BMMs. Spray drying of mannitol into microparticles resulted in a crystalline transformation of mannitol, evident from XRPD, DSC and Raman spectroscopy analyses. Both BMMs and BNMs displayed a faster dissolution rate than bulk budesonide. The yield of BNMs was higher Electronic supplementary material The online version of this article (https://doi.org/10.1007/s11095-018-2509-z) contains supplementary material, which is available to authorized users. * Mingshi Yang [email protected] 1

Department of Pharmacy, Faculty of Health and Medical Science University of Copenhagen, Universitetsparken 2 DK-2100 Copenhagen, Denmark

2

Wuya College of Innovation Shenyang Pharmaceutical University, Wenhua Road 103 110016 Shenyang, China

than that of BMMs. The mass ratio between budesonide and mannitol was preserved in the BNMs, whereas the mass ratio in the BMMs was higher than the theoretical ratio. Conclusions Spray drying is an enabling technique for preparation of budesonide amorphous solid dispersions and nanocrystal-embedded microparticles. Two-fluid nozzle spray drying is superior to three-fluid nozzle spray drying in terms of yield.

KEY WORDS budesonide . spray drying . three-fluid nozzle spray drying . two-fluid nozzle spray drying

INTRODUCTION Dry powder inhalers (DPIs) are superior to pressurized metered dose inhalers (pMDIs) and nebulizers with respect to the physicochemical stability of the dosage form and aerosol performance (1–3). The most commonly used technique to manufacture powders for DPIs is jet milling of the active pharmaceutical ingredient (API), followed by blending the micronized API with excipient(s) (4). However, the spray drying process has recently become an attractive manufacturing method as it offers the opportunity for engineering of particles with customized particle properties, e.g., size, density, shape, and surface chemistry (5,6). Even though excipients are not always needed during spray drying, addition of excipient(s) may render the dry powders with improved stability, dispersibility, and the required bulk volume. However, a frequently encountered problem occurs when the solubility of the API and the excipient(s) differs, i.e., a suitable co-solvent to dissolve both the API and the excipient(s) does not exist. To overcome this challenge, three- and four-fluid nozzle spray driers have been introduced, allowing for the dissolution of the API and the excipient(s) in different solvents prior to feeding into the drying chamber using separate liquid passages. In fact, spray driers equipped with three- and four-fluid nozzles have been

247

Page 2 of 11

employed to prepare protein-loaded poly(DL-lactic-co-glycolic acid) microparticles (7), amorphous solid dispersion microparticles (8), two-API composite microparticles, wherein one API is water-soluble and the other is water-insoluble (9), and polymeric nanocomposite microparticles (10). An alternative to multi-fluid nozzle spray drying is to suspend one solid substance (e.g., the API) into a solution of the other substance(s) [e.g., the excipient(s)]. In this way, concomitant spray drying of both API and excipient(s) can be performed. This could be a strategy for formulating a poorly water-soluble API and water-soluble excipient(s), where the poorly water-soluble API is suspended in an aqueous solution of the excipient(s). An advantage of using this approach is that organic solvents are not required to dissolve the poorly water-soluble API, eventually resulting in a more environmentally friendly manufacturing process. However, a disadvantage of this approach includes adding extra operation units, e.g., micronization of the API and dispersion of the micronized API in the aqueous solution of the excipient(s). In addition, it is challenging to ensure a homogeneous feed suspension. The aim of this study was to investigate further the pros and cons of the aforementioned two approaches for preparing inhalable dry powders consisting of API and excipient. To this end, we used budesonide and mannitol as a model API and excipient, respectively. Budesonide is a corticosteroid, which is used in the long-term management of lung diseases, e.g., asthma and chronic obstructive pulmonary disease. After preparation using the two different approaches, the resultant powder particles were characterized with respect to yield, moisture content, particle size, morphology, drug content, solid state properties, aerodynamic performance and dissolution properties. For the process of spray drying using the three-fluid nozzle, budesonide was dissolved in ethanol, and the solution was fed into the drying chamber via one passage, while an aqueous solution of mannitol was fed into the drying chamber via the other passage of the nozzle. For the spray drying with the two-fluid nozzle, a budesonide nanosuspension was first prepared by using wet ball milling and subsequently spray-dried into nanocomposite microparticles.

MATERIALS AND METHODS Materials Budesonide was a gift from Hubei Gedian Renfu Pharmaceuticals (Hubei, Wuhan, China). D-mannitol and Lutrol F68 were purchased from Sigma-Aldrich (St. Louis, MO, USA). Additional chemicals were obtained commercially at analytical grade.

Pharm Res

(2018) 35:247

Methods Preparation of Budesonide Nanosuspensions Budesonide nanosuspensions were prepared by using a wet ball milling method, as previously reported (11). In brief, 300 mg of budesonide, 100 mg of F68 and 10 g of glass beads (1.0 mm in diameter: 0.5 mm in diameter = 3:1, w/w) were weighed into 10 ml of MilliQ water in a 100 ml glass bottle. At room temperature, the suspension with the milling media was stirred with a magnetic bar on a magnetic stirrer (IKA® RCT basic, IKA®-Werke GmbH & Co. KG, Staufen, Germany) at dial 7 for 5 h. The dial from 1 to 10 equals a speed range from 50 to 1100 rpm. The milled budesonide particles were collected using a syringe with a 21 gauge needle to separate them from the glass beads. Ultracentrifugation was used to remove F68 using the following centrifugation procedure: 6000 g for 5 min, 12,000 g for 5 min, 21,000 g for 5 min, 34,000 g for 5 min and 48,000 g for 10 min at 25°C (Optima™ Max-XP Ultracentrifuge, Beckman Coulter, CA, USA). The supernatant containing F68 was discarded. The pellet containing the budesonide nanocrystals was collected and resuspended in MilliQ water to a final concentration of 30 mg/ml. Characterization of Budesonide Nanosuspensions The intensity-weighted mean hydrodynamic particle diameter (z-average) and the polydispersity index (PDI) of the nanosuspensions were measured by using the photon correlation spectroscopy (PCS) technique. The PDI was used as a parameter to reflect the broadness of the particle size distribution. The measurements were conducted at 25°C using a Zetasizer Nano ZS (Malvern Instruments, Worcestershire, UK) equipped with a 633 nm laser and 173° detection optics. A viscosity value of 0.8872 and a dispersant refractive index of 1.330 were used. Malvern DTS v.7.11 software was used for data analysis. For sample preparation, 25 μl of budesonide nanosuspension was diluted with MilliQ water to 1 ml. Samples of three independent batches were measured in duplicate. The morphology was examined by microscopy: the nanosuspension was dried thoroughly at room temperature before recording images using a FEI Quanta 3D FEG (FEI, Hillsboro, OR, USA). Samples were sputter-coated with gold (6 nm) using a Leica EM ACE 200 (Vienna, Austria), and images were acquired at 100,000× magnification. The colloidal stability of the nanosuspensions was evaluated after 1, 2, 5 and 6 days of storage (30 mg/ml in MilliQ water). The nanosuspensions were kept in closed clear glass vials and stored at 4°C. At pre-determined time intervals, samples were visually inspected for sedimentation, and 25 μl of the suspensions was withdrawn after mixing and subjected to particle size analysis, as described above. Two independent batches were analysed. For solid state analysis, the nanosuspensions were dried at

Pharm Res

(2018) 35:247

Page 3 of 11 247

room temperature and analysed using an X’Pert PRO X-ray diffractometer (PANalytical, Almelo, The Netherlands). The samples were measured by using a CuKα radiation source (k = 1.54060 Å) at diffraction angles (2Theta) from 5° to 35° using a step size of 0.026°. The operating current and voltage were 40 mA and 45 kV, respectively. Data were collected and analysed using the X’Pert Data Collector software (PANalytical). Preparation of Microparticles by Spray Drying The BMMs and BNMs were prepared using a Büchi B-290 spray dryer (Büchi Labortechnik AG, Postfach, Switzerland). For preparation of the BMMs, two feed solutions containing budesonide dissolved in ethanol and mannitol dissolved in MilliQ water, respectively, were spray-dried using a spray drier equipped with a three-fluid nozzle (Büchi Labortechnik AG, Flawil, Switzerland), an inert loop B-295 (Büchi Labortechnik AG) and a dehumidifier (Büchi Labortechnik AG). The BNMs were prepared by spray drying of the budesonide nanosuspension in an aqueous mannitol solution. The formulation parameters and spray drying process parameters are shown in Table I. Characterization of Microparticles The yield of the spray-dried powders was calculated as the difference in weight of the collected dry powders divided by the total dry mass used for spray drying. The moisture content in the dry powders was not considered in the calculation. Three independent batches were analyzed. The moisture content of the spray-dried powders was measured using a Discovery Thermogravimetric Analyzer (TA Instruments, New Castle, USA). Approximately 10 mg of samples was placed in a 100 μl platinum pan and heated from room temperature to 150°C at a heating rate of 10°C/min. The weight

Table I

loss in percent between room temperature and 140°C was calculated and defined as the moisture content. The measurements were performed on three independent batches. Scanning electron microscopy (SEM) was used to examine the morphology of bulk budesonide and the microparticles. The samples were mounted on an aluminum tab using a conductive carbon tape, and gold sputter-coated. A Hitachi TM3030 Tabletop Microscope (Hitachi High-Technologies Corporation, Tokyo, Japan) was used to image the samples at an accelerating voltage of 15 kV. The particle size distribution of the dry powders was measured by laser diffraction using a Mastersizer 2000 connected to a dry dispersion unit (Scirrocco 2000 powder feeder, Malvern Instruments). A dispersive air pressure of 3 bar and a refractive index of 1.520 were used to measure the particles in air. The particle size distribution is described as the mass median diameter (d0.5) and the span. The span value was calculated using the following Eq. (1): Span ¼

ðd 0:9 −d 0:1 Þ d 0:5

ð1Þ

where d0.1 is the value of particle diameter at 10% in the cumulative distribution and d0.9 is the value of particle diameter at 90% in the cumulative distribution. The measurements of the dry powders were performed using three independent batches. A volume of 1 ml of MilliQ water was added to 7.5 mg of the powder samples to dissolve mannitol. After mixing, the samples were centrifuged as described above. Mannitol, which was assumed to be mainly in the supernatant, was discarded. The pellets containing budesonide particles were collected and redispersed in 1 ml of MilliQ water. The size was determined as described in the characterization of budesonide nanosuspensions. The measurements were performed on two independent batches. The size measured after redispersion is expected to reflect the particle size of the budesonide

Formulation Parameters and Spray Drying Process Parameters Two-fluid nozzle

Formulation parameters Spray drying process parameters

Concentration of budesonide (mg/ml) Concentration of mannitol (mg/ml) Feed rate (ml/min) Inlet temperature (°C) Outlet temperature (°C) Drying air flow rate (m3/h) Atomization air flow rate (l/h)

7.5 67.5 3.0 100 45–50 37.5 473

Three-fluid nozzle Inner feed solution (Mannitol/MilliQ water)

Outer feed solution (Budesonide/ethanol)

– 135.0 1.5 100 60–65 37.5 473

15.0 – 1.5

247

Page 4 of 11

Pharm Res

(2018) 35:247

particles since mannitol is expected to fully dissolve under the applied experimental conditions. The total amount of budesonide contained in the microparticles was determined by using high performance liquid chromatography (HPLC, see below) after dissolving approximately 5 mg of powder sample in 25 ml of 80% (v/v) ethanol. The drug content was calculated using the following Eq. (2):

activated for 2.4 s into the NGI at an air flow rate of 95 l/ min. Before each run, the NGI collection cups were coated with ethanol containing 10% (v/v) tween 20 to minimize particle bouncing after deposition. The powder dispersed in the Inhaler, the empty capsule shell, the throat, the stage cups and the micro-orifice collector (MOC) were rinsed and recovered with an ethanol-MilliQ water mixture (40%, v/v). The budesonide concentration was quantified by using HPLC (see Drug content below). The lower cut-off aerodynamic diameters of the NGI Actual amount of budesonide in the weighed powder  100% stages 1–7 at 95 l/min were 6.29, 3.51, 2.24, 1.34, 0.74, 0.42, ¼ Theoretical amount of budesonide in the weighed powder and 0.25 μm, respectively, calculated by using the equation ð2Þ specified in Table 2.9.18.-9 of the European Pharmacopoeia (12). The cumulative mass of the powder deposited at each stage (expressed as the percentage of the total mass in the Solid State Analysis stages) was plotted as a function of the cut-off diameter. The mass median aerodynamic diameter (MMAD), defined as the Diffractograms of all samples were collected using an diameter at 50% cumulative percentage, was calculated by X’Pert PRO X-ray diffractometer, as described above. interpolation from the plot of the cumulative versus cutoff aeroDifferential scanning calorimetry (DSC) was performed dynamic diameter. The emitted dose was defined as the using a Discovery DSC (TA instruments, New Castle, amount of powder emitted from the inhaler. The emitted USA). An amount of 2–4 mg of the powder samples efficiency was calculated as the percentage of drug emitted was filled into an aluminum Tzero. The samples were from inhaler relative to the total amount of recovered drug. heated at a heating rate of 10°C/min with a constant The fine particle fraction of total recovered drug (FPF total) nitrogen flow rate of 50 ml/min. Dynamic mechanical was defined as the percentage of drug particles with an aeroanalysis (DMA) was utilized to determine the glass trandynamic diameter below 5 μm relative to the total recovered sition temperature (Tg) of budesonide in the micropardrug. The FPF of emitted drug (FPF emitted) was defined as ticles. A stainless-steel powder pocket filled with the the percentage of drug particles smaller than 5 μm in diameter spray-dried powders was loaded into a 35 mm dual canin the emitted aerosol. The measurements were performed on tilever clamp of a DMA Q800 (TA Instruments). The three independent batches. measurements were performed in a multifrequencystrain mode at a constant frequency of 1 Hz and an amSolubility and Dissolution Testing plitude of 20.00 μm. The samples were heated from room temperature to 120°C at a heating rate of 3°C/min. Data The equilibrium dynamic solubility of bulk budesonide, physanalysis was performed using a TA Universal Analysis ical mixtures of budesonide and mannitol, and BMMs and 2000 software (TA Instruments), and the peak maximum BNMs were measured at room temperature. An amount of of the tan δ plot was taken as the Tg value. Three inde10 mg of bulk budesonide and a weighed amount of physical pendent batches were analyzed. Raman spectroscopy was mixture and microparticles equivalent to 10 mg of budesonide conducted with a Kaiser RXN1 Microprobe from Kaiser were placed in a test tube with 10 ml of 0.01 M phosphate Optical Systems (Ann Arbor, MI, USA) equipped with a buffer (pH 7.4). The suspension was mixed by using a headPhaT-probe. Samples were manually compacted into an over-head device (Stuart rotator SB3, Bibby Sterilin, aluminum slide for analysis. The Raman-scattered light Staffordshire, UK) at 30 rpm for 48 h, and subsequently filwas collected with the PhaT-probe using a 10× objective. tered through a 0.22-μm syringe filter (Q-Max Nylon, The Raman shift from 150 to 1890 cm−1 was measured Frisenette APS, Knebal, Denmark) and analyzed by HPLC at a wavelength of 785 nm. as described below. The experiments including bulk budesoIn Vitro Aerosolization Assessment The aerodynamic performance of the spray-dried powders was assessed by using a Next Generation Impactor (NGI, Copley Scientific, Nottingham, UK). A RS01 Monodose dry powder inhaler device (Plastiape, Osnago, Italy) with approximately 10 mg powder sample in a hydroxypropyl methylcellulose capsule (Size 3; Capsugel, West Ryde, Australia) was

nide and physical mixtures were conducted in triplicates, and the experiments of the BMMs and the BNMs were performed using three independent batches. A USP type II apparatus (Erweka DT70 dissolution tester, Erweka GmbH, Heusenstamm, Germany) in a modified, custom-made set-up with a rotating minipaddle at 50 rpm was used to evaluate the dissolution of bulk budesonide, physical mixtures of budesonide and mannitol, BMMs, budesonide nanosuspensions (30 mg/ml in MilliQ water) and

(2018) 35:247

Pharm Res

Page 5 of 11 247

BNMs. An amount of 4 mg of bulk budesonide, or an equivalent amount of budesonide, was weighed and filled into a hydroxypropyl methylcellulose capsule (Size 3; Capsugel). HPMC capsules were used because it was difficult to disperse the dry powders homogeneously in the dissolution medium, which is most likely due to poor wetting properties of the dry powders. Most powders were floating on the surface of the dissolution medium resulting in large deviation between measurements. When HPMC capsules were used, the dry powders were wetted more evenly, which largely prevented the powders from floating on the surface of the dissolution medium during the test. The purpose of the dissolution experiments was to compare the dissolution rates of different formulations, hence they do not necessarily reflect the dissolution conditionsin the lungs. Thus, in the current study, capsules were used in the dissolution tests. The capsule was placed in a custommade metal sinker (to ensure that the capsule was submerged during testing), which was subsequently placed in a 250 ml vessel containing 200 ml of 10 mM phosphate buffer (pH 7.4) as the dissolution medium at 37°C. Samples of dissolution medium (3 ml) were withdrawn at specific time points (10, 20, 30, 45 and 60 min, respectively), filtered through 0.22-μm syringe filters (Q-Max Nylon, Frisenette APS) and analyzed by HPLC, as described below. An equal amount of fresh dissolution medium was used for replacement after withdrawal of the samples. The in vitro release study was conducted in at least duplicate, and the average of percentage of cumulative drug release as a function of time was determined.

Quantification of Budesonide by HPLC Budesonide was quantified by HPLC using an Agilent 1260 infinity HPLC system (Agilent, Santa Clara, USA) equipped with an Agilent 1290 Diode Array Detector. An Agilent 5 TCC18 column (250 × 4.6 mm, 5 μm, Agilent) was used at room temperature, and the mobile phase (at a flow rate of 1.0 ml/ min) was a mixture of 35% (v/v) acetonitrile and 65% (v/v) monobasic sodium phosphate at pH 3.2 ± 0.05. A volume of 20 μl of drug solution was injected into the HPLC system, and the signal was detected at 254 nm by using an ultraviolet detector. The standard curve was linear in the range of 0.5– Table II Drying

300 μg/ml (R2 > 0.999). The limit of detection and the limit of quantification were 0.05 and 0.2 μg/ml, respectively. Statistics All experimental data are presented as mean values ± standard deviation (SD). To identify statistically significant difference, an unpaired t-test analysis was performed by using the GraphPad Prism software version 6.05 for Windows (GraphPad Software, La Jolla, CA, USA). Statistical significance was taken at p < 0.05.

RESULTS Characterization of Budesonide Nanosuspensions The mean hydrodynamic diameter of budesonide nanosuspensions was approximately 300 nm and the PDI value was below 0.3, suggesting a relatively narrow size distribution (Table II). A loose, thin layer sediment was observed when the nanosuspension was stored for 1, 2, 5 and 6 days, respectively. However, the sediment disappeared upon gentle mixing. The particle sizes were not significantly changed (p > 0.05) after storage for 6 days at 4°C (Table II). Bulk budesonide showed an irregular angular shape with a broad size distribution (Fig. 1a). After milling into nanosuspensions, the size of the particles was reduced, but the particles still displayed irregular shapes (Fig. 1b). Budesonide displayed distinct X-ray powder diffraction (XRPD) patterns before and after milling (2θ: 6.2, 11.5. 14.6, 15.5 and 16.2°), indicating no significant loss of crystallinity (Fig. 2). The observed diffraction patterns were in accordance with patterns published in literature (11,13,14). Some groups have reported problems related to erosion of glass beads due to attrition upon milling, which could result in bead residues in the product, eventually compromising the purity of the final product and causing safety concerns (15,16) However, in this study, the presence of glass residues was not evaluated, and their potential effects on the product properties were considered negligible.

Particle Size of Budesonide Nanosuspensions at Day 0, 1, 2, 5 and 6 after Storage at 4 °C and Redispersed Budesonide Nanosuspensions after Spray

z-average (nm) PDI

Day 0

Day 1

Day 2

Day 5

Day 6

Redispersed particles

295.0 ± 2.1 0.15 ± 0.03

291.7 ± 3.1 0.15 ± 0.02

284.4 ± 8.2 0.15 ± 0.02

287.1 ± 17.8 0.17 ± 0.02

300.8 ± 5.8 0.14 ± 0.02

299.7 ± 18.4 0.20 ± 0.07

For day 0, data represent mean values ± SD of three independent batches performed in duplicates (N = 3, n = 2); for the other days, data represent mean values ± SD of two independent batches (N = 2)

247

Page 6 of 11

Pharm Res

(2018) 35:247

Fig. 1 Representative SEM images of bulk budesonide (a), budesonide nanosuspension (b), BMMs (c) and BNMs (d).

Characterization of Spray-Dried Powders The data in Table II show that the average diameter as well as the PDI values of the nanosuspensions before and after spray drying did not change significantly (p > 0.05), suggesting that the BNMs can be readily redispersed into nanosuspensions. This can be attributed to the fact that mannitol in this formulation serves as a bulking agent and prevents aggregation of budesonide nanocrystals. Because the mass of mannitol was nine Fig. 2 Left: Representative X-ray diffractograms of bulk budesonide (A), F68 (B), a physical mixture of budesonide and F68 (C), budesonide nanosuspension (D), mannitol (E), a physical mixture of budesonide and mannitol (F), BMMs (G), and BNMs (H). Right: References of mannitol polymorphs, modified from (17). The diffractograms have been displaced on the y-axis for clarity.

times greater than the mass of the nanoparticles in the formulation, it can be assumed that the nanoparticles are embedded in the mannitol matrix after spray drying (17). When re-dispersing the BNMs, mannitol dissolved rapidly, and the nanoparticles were subsequently released. From the SEM micrographs (Fig. 1), it is apparent that both the BMMs (Fig. 1c) and the BNMs (Fig. 1d) had a spherical shape with a broad particle size distribution. A large fraction of the BMMs were larger than 5 μm. Laser diffraction analysis showed that the

Pharm Res

(2018) 35:247

Page 7 of 11 247

average particle size of the BMMs was significantly larger than that of the BNMs (Table III). The BNMs prepared using the two-fluid nozzle displayed a drug content of approximately 100%, while the drug content of budesonide in the BMMs spray dried using the threefluid nozzle was higher than 160% (Table III).

Solid State Analysis of Microparticles As shown in Fig. 2, bulk mannitol displays characteristic diffraction peaks of β-mannitol at 2θ degree of 10.6 and 14.7 (18). The XRPD pattern of BMMs suggests that this sample is composed of a mixture of α-mannitol (2θ degree of 13.8 and 17.3) and δ-mannitol (2θ degree of 9.8), while the XRPD pattern of BNMs suggests that this sample is composed of a mixture of α-mannitol and crystalline budesonide (19,20). The absence of the characteristic crystalline diffractions of budesonide at 2θ degree of 6.2, 15.5 and 16.2 in the BMMs suggests amorphization of budesonide in the particle matrix. This finding was further supported by DMA measurements. A Tg of 98.96 ± 2.05°C (N = 3) indicated that the Tg of budesonide was observed in the BMMs (Supplementary data, Fig. S1). However, no Tg was observed in BNMs (Supplementary data, Fig. S1). In the physical mixture and BNMs, diffractions at 2θ degree of 6.2 and 15.5 were still apparent, indicating a crystalline state of budesonide. DSC was used to assess the thermal behavior of the bulk material and the microcomposite formulations (Fig. 3). The thermogram of bulk budesonide displayed an endothermic melting peak at 258°C, which is well in agreement with a previous study (21). Mannitol showed a sharp endothermic peak at 167°C, which corresponds to the melting point of mannitol (22). The thermogram for the binary mixture (budesonide: mannitol = 1: 9, w/w) displayed an endothermic melting peak (approximately 247°C), which can be assigned to the melting of budesonide in the physical mixture. This slightly reduced melting point of budesonide may be explained by dissolution of budesonide in the melted mannitol (above 167°C), and the interaction between budesonide and mannitol led to a broadening of the melting peak of budesonide and a shift towards lower temperature, as compared to bulk

Table III Characteristics of the Spray-Dried Powder

Yield (%)

BMMs BNMs

32.5 ± 5.2 74.9 ± 8.7**

budesonide. The BMMs and the BNMs also displayed the endothermic peak at 247°C as the physical mixture. Similar results have been reported for mannitol with budesonide and mannitol with other APIs (23–25). For the thermogram of mannitol, a melting point at 167°C indicates the presence of α-mannitol and/or β-mannitol, and a melting point at 156°C demonstrates the presence of δ-mannitol (22,26). The BMMs displayed two endothermic peaks (167°C and approximately 156°C), indicating that mannitol in the BMMs exists as a mixture of two or three forms. Raman spectroscopy was also used to identify the polymorphic form of mannitol (Fig. 4). As reported in the literature, β-mannitol is the thermodynamically stable form at 20°C, whereas α-mannitol and δ-mannitol represent metastable forms (27). The Raman spectra showed that budesonide did not affect the analysis of mannitol. Bulk mannitol displayed one main peak at 876 cm−1 and two main characteristic peaks at 1119 and 1134 cm−1 in the spectral range of 1100–1200 cm−1, indicating that bulk mannitol displays a β-form. This was also the case for mannitol in the physical mixture. For mannitol in the BMMs and BNMs, two main peaks at 876 cm−1 and 887 cm−1 were apparent, indicating the prensence of αmannitol or δ-mannitol (26,28). Hence, it can be concluded that the mannitol of BMMs is transformed from β-form to a mixture of α- and δ- form during the spray drying process, and that α-mannitol is the most abundant form in the BNMs.

Aerodynamic Performance and Release Profiles of Microparticles As shown in Fig. 5, the majority of the powder was retained in the throat after aerosolization. Furthermore, the MMAD was calculated to be 3.7 ± 0.4 μm for the BMMs and 3.9 ± 0.2 μm for the BNMs, the emitted dose was 8.6 ± 0.5 mg for the BMMs and 8.7 ± 0.3 mg for the BNMs, the emitted efficiency was 83.6 ± 1.0% for the BMMs and 90.9 ± 7.6% for the BNMs. The FPF total was 21.9 ± 2.2% for BNMs, which is significantly higher than the value of the BMMs (15.2 ± 1.1%) (p < 0.01). The FPF emitted of BNMs (24.2 ± 3.0%) was also significantly higher (p < 0.05) than that of BMMs (18.2 ± 1.5%).

Moisture content (%)

0.3 ± 0.1 0.3 ± 0.1

Particle size

Drug content (%, w/w)

d0.5 (μm)

Span

6.5 ± 0.5 4.7 ± 0.6*

2.2 ± 0.3 2.0 ± 0.2

164.4 ± 9.7 105.7 ± 2.5

For the drug content, data represent mean values ± SD of two independent batches in duplicate (N = 2, n = 2); the other values represent mean values ± SD (N = 3) Results (except for drug content) that are significantly different are indicated: *p < 0.05 and **p < 0.01

247

Page 8 of 11

Pharm Res

(2018) 35:247

Fig. 3 Representative DSC thermograms of bulk budesonide (A), mannitol (B), a physical mixture of budesonide and mannitol (C), BMMs (D) and BNMs (E), obtained upon heating at 10°C/min. The thermograms have been displaced on the y-axis for clarity.

The solubility of bulk budesonide, the physical mixture, the BMMs and the BNMs were determined to be 25.8 ± 3.7 μg/ml (n = 3), 23.2 ± 1.1 μg/ml (n = 3), 48.2 ± 9.1 μg/ml (N = 3) and 29.7 ± 0.5 μg/ml (N = 3), respectively. The dissolution rate of bulk budesonide was very low with less than 10% of the drug being dissolved in 60 min (Fig. 6). The dissolution of budesonide from the physical mixture was slightly faster (14% in 60 min) than that from bulk budesonide. The dissolution of budesonide from the spray-dried microparticles was faster than the dissolution of bulk budesonide.

DISCUSSION Inhalable BMMs and BNMs were manufactured by spray drying using two-fluid and three-fluid nozzles, respectively. Although the same spray drying conditions, i.e., feed rate, inlet temperature, drying air flow rate, and atomization air flow

Fig. 4 Representative Raman spectra of bulk budesonide (A), mannitol (B), a physical mixture of budesonide and mannitol (C), BMMs (D) and BNMs (E). The spectra have been displaced on the y-axis for clarity.

rate, have been used, the resulting outlet temperatures of the two spray drying processes were different (Table I). This can be attributed to the fact that the solvent compositions in the feed solutions were different. The difference in the outlet temperature also implies that the resulting particle properties of BMMs and BNMs are be different. Based on the results of this study, the advantages and disadvantages of the two approaches are summarized in Table IV. Generally, the yield of spray drying at laboratory scale is low, and the loss of product is mainly a result of wall deposition and a low separation efficiency of the cyclone (29). In this study, the yield of the BNMs was significantly higher than the yield of the BMMs (Table III). In general, with an increase in the outlet temperature, more dry product is obtained, and the extent of wall deposition can be expected to be low, eventually resulting in a higher yield (30). However, a lower yield was obtained for the BMMs spray-dried at a higher outlet temperature (60–65°C), as compared to the yield of the BNMs (outlet temperature: 45–50°C). This could be attributed to the

Pharm Res

(2018) 35:247

Fig. 5 Aerodynamic assessment of BMMs (white bars) and BNMs (black bars). The bars represent mean values ± SD, N = 3.

amorphous state of budesonide in the BMMs. Amorphous materials are usually cohesive and have higher surface energy than crystalline materials, which tend to adhere to the inner wall of the spray dryer, eventually resulting in a low yield (31). It is assumed that the presence of the amorphous form of budesonide in the BMMs is caused by instantaneous evaporation of solvent, hence budesonide molecules did not have sufficient time to arrange into crystal structures but form amorphous materials (32). The evaporation of solvent in the spray drying process is mainly caused by the heat carried by the

Fig. 6 Dissolution profiles (from bottom to top) of bulk budesonide (■, n = 2), a physical mixture of budesonide and mannitol (●, n = 2), BMMs (▲, n = 2), budesonide nanosuspension (▽, N = 2, n = 2) and BNMs (◆, N = 2, n = 2).

Page 9 of 11 247

drying gas from the inlet of the dryer. In addition, atomization can dramatically enlarge the surface area of the droplets, which further facilitates the evaporation of the solvent. In this study, an inlet temperature of 100°C was used, which is higher than the boiling point of ethanol. The drying gas was in full capacity of 37.5 m3/h (aspiration rate of 100%). An instantaneous evaporation of solvent can be expected from the process. The amorphous form displays an enhanced dissolution rate because of its higher molecular mobility, relative to the crystalline form (33). However, this also poses a stability challenge, amorphous materials may transform to crystalline form within the shelf-life of the product (34). Compared to other carbohydrates used for spray drying, e.g., lactose, trehalose and sucrose, mannitol is less hygroscopic (35–37). However, the ratio of mannitol in the spray drying formulations could affect the moisture content of the final product, and a decreased moisture content will be obtained upon an increase in the ratio of the mannitol (17,38). In this study, the starting ratios between budesonide and mannitol (w/w) in the BMMs and BNMs were identical. These moisture contents of these two dry powders were below 1% (Table III). A higher drug content (> 160%) was observed for the BMMs prepared by using the three-fluid nozzle, as compared to the drug content in the dry powders prepared by using the two-fluid nozzle. This could be related to the large mass ratio between budesonide and mannitol, or because of the difference of solvent properties used in the three-fluid nozzle spray drying process (water/inner nozzle, ethanol/outer nozzle). Another possible reason could be the separate feeding passages give rise to separate mannitol particles and budesonide particles after atomization. It is assumed that the drug content could be controlled within the acceptable range, if the ratio difference between the two compounds is decreased, or by using the same solvent or solvents with comparable properties. Many factors could affect the aerosolization efficiency of micron-sized powders, including moisture sorption, particle morphology, particle size distribution and content of amorphous materials (39). From the current results, it is assumed that the amorphous state of budesonide in the BMMs could explain the lower FPF as compared to BNMs because amorphous materials are relatively cohesive and possess large surface energy, as discussed above. Overall, the FPFs of BMMs and BNMs from the study were low. The FPF could be increased by optimizing the formulation composition and the spray drying process parameters. For example, leucine could be added to the spray drying feed formulation since the addition of leucine could decrease the cohesion between particles and improve the flowability of the powders (40), or modifying particle morphology into wrinkled or pitted surface, because the degree of corrugation increased, dry particle adhesion could be reduced, hence resulting in an increase FPF (41).

247

Page 10 of 11

Table IV Pros and Cos of TwoFluid and Three-Fluid Nozzle Spray Drying for Processing BudesonideMannitol Formulations in this Study

Pharm Res

(2018) 35:247

Nozzle type

Advantages

Disadvantages

Two-fluid nozzle

Easy control of the drug content

Three-fluid nozzle

Enables flexible solvent selection

Prerequisite for preparation of a suspension before subjecting it to spray drying The process involves organic solvents, which are not environmentally friendly; Undesired drug transition to amorphous state; Uncontrollable drug content

The increased solubility (p < 0.05, compared to bulk budesonide) and the rapid dissolution of the BMMs can be attributed to the amorphous state of budesonide in the microparticles. The BNMs exhibited an almost identical dissolution profile as that of the budesonide nanosuspension. This phenomenon can be explained by rapid dissolution of the hydrophilic mannitol matrix once exposed to the dissolution medium. Nanosized budesonide was subsequently released from the mannitol matrix and showed similar release performance as the budesonide nanosuspension. The enhanced dissolution rate of the budesonide nanosuspension, relative to that of bulk budesonide is mainly due to the reduction of the particle size and the increase of surface area of the budesonide nanosuspension as described by the Noyes-Whitney equation (42). Residual F68 in the nanosuspension was not investigated in this study. However, residual F68 may be adsorbed on the surface of budesonide nanoparticles due to the amphiphilic properties of F68, but may have been negligible in the present study, as suggested by the solubility tests, i.e. the solubility of BNMs was comparable to the solubility of bulk budesonide and the physical mixture.

81573380) for funding raw materials. We gratefully acknowledge Eric Ofosu Kissi and Junwei Wang for valuable scientific discussions and technical support, Chengyu Wu for technical assistance with the spray drier, and Yongquan Li and Magnus Edinger for technical support. We thank the Core Facility for Integrated Microscopy, Faculty of Health and Medical Sciences, University of Copenhagen, for the morphology study of the budesonide nanosuspension. We acknowledge the Danish Agency for Science, Technology and Innovation for funding the Zetasizer Nano ZS and Novo Nordisk for providing the NGI.

REFERENCES 1.

2. 3.

4.

CONCLUSIONS 5.

This study demonstrates the feasibility of formulating budesonide into amorphous solid dispersion dry powders or nanocrystal-embedded dry powders using a spray dryer equipped with a two-fluid nozzle or a three-fluid nozzle. The nanocrystal-embedded dry powders obtained by twofluid nozzle spray drying were superior to the amorphous solid dispersion dry powders obtained by three-fluid nozzle spray drying with respect to the yield of the spray drying process, the homogeneity of drug content in the dry powders and the aerosol performance. This highlights the importance of selecting optimal processing approaches for rational design of pharmaceutical formulation. ACKNOWLEDGMENTS AND DISCLOSURES. This work was funded by a Faculty PhD stipend, the Faculty of Health and Medical Sciences, University of Copenhagen, Denmark. We thank the National Natural Science Foundation of China (No.

6.

7.

8.

9.

10.

Smith IJ, Parry-Billings M. The inhalers of the future? A review of dry powder devices on the market today. Pulm Pharmacol Ther. 2003;16(2):79–95. Dolovich MB, Dhand R. Aerosol drug delivery: developments in device design and clinical use. Lancet. 2011;377(9770):1032–45. Zhou QT, Leung SS, Tang P, Parumasivam T, Loh ZH, Chan HK. Inhaled formulations and pulmonary drug delivery systems for respiratory infections. Adv Drug Deliv Rev. 2015;85:83–99. Chow AH, Tong HH, Chattopadhyay P, Shekunov BY. Particle engineering for pulmonary drug delivery. Pharm Res. 2007;24(3): 411–37. Vehring R. Pharmaceutical particle engineering via spray drying. Pharm Res. 2008;25(5):999–1022. Wan F, Maltesen MJ, Bjerregaard S, Foged C, Rantanen J, Yang M. Particle engineering technologies for improving the delivery of peptide and protein drugs. Journal of Drug Delivery Science and Technology. 2013;23(4):355–363. Wan F, Maltesen MJ, Andersen SK, Bjerregaard S, Baldursdottir SG, Foged C, et al. Modulating protein release profiles by incorporating hyaluronic acid into PLGA microparticles via a spray dryer equipped with a 3-fluid nozzle. Pharm Res. 2014;31(11): 2940–51. Mizoe T, Ozeki T, Okada H. Application of a four-fluid nozzle spray drier to prepare inhalable rifampicin-containing mannitol microparticles. AAPS PharmSciTech. 2008;9(3):755–61. Ozeki T, Beppu S, Mizoe T, Takashima Y, Yuasa H, Okada H. Preparation of two-drug composite microparticles to improve the dissolution of insoluble drug in water for use with a 4-fluid nozzle spray drier. J Control Release. 2005;107(3):387–94. Ozeki T, Beppu S, Mizoe T, Takashima Y, Yuasa H, Okada H. Preparation of polymeric submicron particle-containing microparticles using a 4-fluid nozzle spray drier. Pharm Res. 2006;23(1): 177–83.

Pharm Res 11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

(2018) 35:247

Liu T, Han M, Tian F, Cun D, Rantanen J, Yang M. Budesonide nanocrystal-loaded hyaluronic acid microparticles for inhalation: in vitro and in vivo evaluation. Carbohydr Polym. 2018;181: 1143–52. European Directorate for Quality in Medicines and Healthcare (EDQM): European Pharmacopeia 9.0, Monograph 2.9.18. Preparations for Inhalation: Aerodynamic Assessment of Fine Particles. EDQM, Strasburg, France, 2017. Bandi N, Wei W, Roberts CB, Kotra LP, Kompella UB. Preparation of budesonide- and indomethacin-hydroxypropylbeta-cyclodextrin (HPBCD) complexes using a single-step, organic-solvent-free supercritical fluid process. Eur J Pharm Sci. 2004;23(2):159–68. Boraey MA, Hoe S, Sharif H, Miller DP, Lechuga-Ballesteros D, Vehring R. Improvement of the dispersibility of spray-dried budesonide powders using leucine in an ethanol–water cosolvent system. Powder Technol. 2013;236:171–8. Peltonen L, Hirvonen J. Pharmaceutical nanocrystals by nanomilling: critical process parameters, particle fracturing and stabilization methods. J Pharm Pharmacol. 2010;62(11):1569–79. Shegokar R, Muller RH. Nanocrystals: industrially feasible multifunctional formulation technology for poorly soluble actives. Int J Pharm. 2010;399(1–2):129–39. Yamasaki K, Kwok PC, Fukushige K, Prud'homme RK, Chan HK. Enhanced dissolution of inhalable cyclosporine nano-matrix particles with mannitol as matrix former. Int J Pharm. 2011;420(1): 34–42. Lee YY, Wu JX, Yang M, Young PM, van den Berg F, Rantanen J. Particle size dependence of polymorphism in spray-dried mannitol. Eur J Pharm Sci. 2011;44(1–2):41–8. Hulse WL, Forbes RT, Bonner MC, Getrost M. The characterization and comparison of spray-dried mannitol samples. Drug Dev Ind Pharm. 2009;35(6):712–8. Sharma VK, Kalonia DS. Effect of vacuum drying on proteinmannitol interactions: the physical state of mannitol and protein structure in the dried state. AAPS PharmSciTech. 2004;5(1):E10. Mezzena M, Scalia S, Young PM, Traini D. Solid lipid budesonide microparticles for controlled release inhalation therapy. AAPS J. 2009;11(4):771–8. Kaialy W, Nokhodchi A. Dry powder inhalers: physicochemical and aerosolization properties of several size-fractions of a promising alterative carrier, freeze-dried mannitol. Eur J Pharm Sci. 2015;68: 56–67. Kumon M, Kwok PC, Adi H, Heng D, Chan HK. Can low-dose combination products for inhalation be formulated in single crystalline particles? Eur J Pharm Sci. 2010;40(1):16–24. Malamatari M, Somavarapu S, Kachrimanis K, Bloxham M, Taylor KMG, Buckton G. Preparation of theophylline inhalable microcomposite particles by wet milling and spray drying: the influence of mannitol as a co-milling agent. Int J Pharm. 2016;514(1): 200–11. Hu J, Dong Y, Ng WK, Pastorin G. Preparation of drug nanocrystals embedded in mannitol microcrystals via liquid antisolvent precipitation followed by immediate (on-line) spray drying. Adv Powder Technol. 2018;29(4):957–63. Hao HX, Su WY, Barrett M, Caron V, Healy AM, Glennon B. A calibration-free application of Raman spectroscopy to the monitoring of mannitol crystallization and its polymorphic transformation. Org Process Res Dev. 2010;14(5):1209–14.

Page 11 of 11 247 27.

Maas SG, Schaldach G, Littringer EM, Mescher A, Griesser UJ, Braun DE, et al. The impact of spray drying outlet temperature on the particle morphology of mannitol. Powder Technol. 2011;213(1–3):27–35. 28. Burger A, Henck JO, Hetz S, Rollinger JM, Weissnicht AA, Stottner H. Energy/temperature diagram and compression behavior of the polymorphs of D-mannitol. J Pharm Sci. 2000;89(4):457– 68. 29. Sosnik A, Seremeta KP. Advantages and challenges of the spraydrying technology for the production of pure drug particles and drug-loaded polymeric carriers. Adv Colloid Interf Sci. 2015;223: 40–54. 30. Maury M, Murphy K, Kumar S, Shi L, Lee G. Effects of process variables on the powder yield of spray-dried trehalose on a laboratory spray-dryer. Eur J Pharm Biopharm. 2005;59(3):565–73. 31. Parumasivam T, Chang RY, Abdelghany S, Ye TT, Britton WJ, Chan HK. Dry powder inhalable formulations for anti-tubercular therapy. Adv Drug Deliv Rev. 2016;102:83–101. 32. Islam MIU, Langrish TAG. The effect of different atomizing gases and drying media on the crystallization behavior of spray-dried powders. Dry Technol. 2010;28(9):1035–43. 33. Hancock BC, Zografi G. Characteristics and significance of the amorphous state in pharmaceutical systems. J Pharm Sci. 1997;86(1):1–12. 34. Shetty N, Park H, Zemlyanov D, Mangal S, Bhujbal S, Zhou QT. Influence of excipients on physical and aerosolization stability of spray dried high-dose powder formulations for inhalation. Int J Pharm. 2018;544(1):222–34. 35. Debord B, Lefebvre C, Guyot-Hermann AM, Hubert J, Bouché R, Cuyot JC. Study of different crystalline forms of mannitol: comparative behaviour under compression. Drug Dev Ind Pharm. 2008;13(9–11):1533–46. 36. Ohrem HL, Schornick E, Kalivoda A, Ognibene R. Why is mannitol becoming more and more popular as a pharmaceutical excipient in solid dosage forms? Pharm Dev Technol. 2014;19(3):257– 62. 37. Naini V, Byron PR, Phillips EM. Physicochemical stability of crystalline sugars and their spray-dried forms: dependence upon relative humidity and suitability for use in powder inhalers. Drug Dev Ind Pharm. 1998;24(10):895–909. 38. Jensen DM, Cun D, Maltesen MJ, Frokjaer S, Nielsen HM, Foged C. Spray drying of siRNA-containing PLGA nanoparticles intended for inhalation. J Control Release. 2010;142(1):138–45. 39. Adi H, Young PM, Chan HK, Agus H, Traini D. Co-spray-dried mannitol-ciprofloxacin dry powder inhaler formulation for cystic fibrosis and chronic obstructive pulmonary disease. Eur J Pharm Sci. 2010;40(3):239–47. 40. Eedara BB, Rangnekar B, Doyle C, Cavallaro A, Das SC. The influence of surface active l-leucine and 1,2-dipalmitoyl-sn-glycero-3-phosphatidylcholine (DPPC) in the improvement of aerosolization of pyrazinamide and moxifloxacin co-spray dried powders. Int J Pharm. 2018;542(1–2):72–81. 41. Adi S, Adi H, Tang P, Traini D, Chan HK, Young PM. Microparticle corrugation, adhesion and inhalation aerosol efficiency. Eur J Pharm Sci. 2008;35(1–2):12–8. 42. Noyes AA, Whitney WR. The rate of solution of solid substances in their own solutions. J Am Chem Soc. 1897;19:930–4.