G Model
IJP 14898 1–8 International Journal of Pharmaceutics xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
International Journal of Pharmaceutics journal homepage: www.elsevier.com/locate/ijpharm
1 2
3 Q1 4 5 6 7 Q2
Formulating powder–device combinations for salmeterol xinafoate dry powder inhalers Mireille Hassoun a , Shirlene Ho a , Joanna Muddle a , Francesca Buttini a,c , Mark Parry b , Mark Hammond b , Ben Forbes a, * a b c
Department of Pharmacy, King’s College London, 150 Stamford Street, London SE1 9NH, UK Intertek-Melbourn Scientific Limited, Saxon Way, Melbourn SG8 6DN, UK Department of Pharmacy, University of Parma, Parco Area delle Scienze 27/a, Parma 43124, Italy
A R T I C L E I N F O
A B S T R A C T
Article history: Received 31 March 2015 Received in revised form 8 May 2015 Accepted 11 May 2015 Available online xxx
Using salmeterol xinafoate (SX) as an active pharmaceutical ingredient, the effects of carrier lactose particle type, total lactose fines content and device resistance on dry powder inhaler performance were investigated in vitro. To mimic drug levels in commercial preparations, interactive mixtures containing 0.58% w/w SX were prepared by low shear tumble mixing. Three types of milled inhalation grade lactose were used (Lactohale1 LH 200, Respitose1 ML006 and ML001) and the concentration of fine lactose (Lactohale1 300) added was varied. The in vitro deposition of each mixture was studied using a next generation impactor and inhaler devices exhibiting different resistances, Rotahaler1 < Aerolizer1 < Handihaler1. Aerosol performance was evaluated based on the emitted dose (ED), mass median aerodynamic diameter (MMAD) geometric standard deviation (GSD) and fine particle fraction (FPF). Increases of up to eight-fold in FPF were observed with increasing intrinsic fine lactose content. The addition of extra fine lactose increased the FPF further, although the effect diminished as more fines were added. The Aerolizer produced the best aerosol performance with any given powder blend, although suitable formulations were identified for each device as defined by the a priori success criteria: >80% ED and MMAD GSD between 1–5 mm. The results confirmed the factors under investigation to be important determinants of product performance, but demonstrated using realistic conditions how individual factor impact may be enhanced or mitigated by inter-dependency. ã 2015 Published by Elsevier B.V.
Keywords: Orally inhaled product (OIP) Salmeterol xinafoate (SX) Dry powder inhaler Lactose Rotahaler Aerolizer Handihaler
8
1. Introduction
9
Dry powder inhalers (DPI) are increasingly popular products for delivering drugs to the lungs due to their ease of operation, environmental sustainability and formulation stability (Telko and Hickey, 2005). Historically, DPI have delivered only 20–30% of the emitted dose to the lungs (De Koning, 2001), but more efficient products are emerging (Friebel et al., 2012). Although the scientific literature now provides well-established principles with which to design efficient and effective DPIs (Friebel et al., 2012; Hoppentocht et al., 2014), there is surprisingly little reported regarding the
10 11 12 13 14 Q3 15 16 17
Abbreviations: SX, salmeterol xinafoate; DPI, dry powder inhaler; PSD, particle size distribution; ED, emitted dose; MMAD, mass median eerodynamic diameter; GSD, geometric standard deviation; FPF, fine particle fraction; FPD, fine particle dose. * Corresponding author. Tel.: +44 207 8484823. E-mail address: [email protected] (B. Forbes).
practicalities of formulating powder blends for different inhaler Q4 devices. Drug particles within the respirable size range, 1–5 mm, are generally highly cohesive and tend to form agglomerates, resulting in poor flow properties, non-uniform dispersion and low dose uniformity. Therefore, drug particles are usually blended with coarse carrier particles, typically a-lactose, to help provide bulk and form an ordered mix (Pilcer and Amighi, 2010; Javadzadeh et al., 2012). Since the active drug and lactose powders exhibit different sized particles, the formulation and blending process is important to achieve uniform drug distribution and provide consistency of delivered dose (Pilcer and Amighi, 2010; Saleem and Smyth, 2008). Powders need to be sufficiently adhesive to possess good flow properties without compromising their ability to generate respirable aerosols efficiently and reproducibly (Begat et al., 2004). Powder dispersion can be manipulated by modifying the properties of microparticle drug surface or the carrier’s surface morphology and size (Buttini et al., 2008; Pomazi et al., 2013; Zeng et al., 2001; Islam et al., 2004; Zellnitz et al., 2011). A proportion of
http://dx.doi.org/10.1016/j.ijpharm.2015.05.028 0378-5173/ ã 2015 Published by Elsevier B.V.
Please cite this article in press as: Hassoun, M., et al., Formulating powder–device combinations for salmeterol xinafoate dry powder inhalers. Int J Pharmaceut (2015), http://dx.doi.org/10.1016/j.ijpharm.2015.05.028
18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
G Model
IJP 14898 1–8 2 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84
M. Hassoun et al. / International Journal of Pharmaceutics xxx (2015) xxx–xxx
fine lactose in the formulation, either as intrinsic or added fines, will further improve dispersion and the de-aggregation of the drug particles (Jones and Price, 2006; Beilmann et al., 2007; Kinnunen et al., 2014; Shur et al., 2008; Grasmeijer et al., 2014; Thalberg et al., 2012). Device properties also determine respirable aerosol generation from a carrier-based powder formulations (Martinelli et al., 2015; Shur et al., 2012). The airflow velocity must also overcome the adhesion forces between the drug and carrier particles to detach the micronized drug particles and disperse them in the aerosol. To aid dispersion, inhaler devices may be engineered with a more tortuous airflow that increases the resistance within the device (Coates et al., 2006; Selvam et al., 2010). Capsule-based inhalers with different resistances are available, which enables the effect of device resistance on dry powder aerosolisation to be investigated. Although aerosolisation is governed by the powder formulation and the inhaler device design inter-dependently (Frijlink and de Boer, 2004), devices are generally developed as platform technology with which to deliver different drugs to the lungs. Scientific studies aimed at understanding DPI performance often examine the influence of device or formulation as independent variables, or use model systems. For example, effect of adding fine excipient particles has been investigated by looking at the effect of fine lactose on the aerosolisation of salmeterol xinafoate (SX) 1–5% w/w interactive powder mixtures (Islam et al., 2004; Adi et al., 2008). The present study explored the inter-dependency of powder formulation and device factors in typical capsule-based DPI products by varying the carrier particle size distribution and the amount of added lactose fines. Three DPI devices having different intrinsic resistance were employed to aerosolize the SX blends. A 32 factorial design was constructed to understand the relative influence of these variables on fine particle fraction (FPF), emitted dose (ED), mass median aerodynamic diameter (MMAD) and the geometric standard deviation (GSD). 2. Materials and methods Micronised SX was supplied by Vamsi Labs Ltd., Maharashtra, India. Milled inhalation a-lactose monohydrate grades (Respitose1 ML006, Respitose1 ML001, Lactohale1 LH 200 and Lactohale1 300) were obtained as samples from DFE Pharma (Veghel, Netherlands). Size 3 hard gelatin capsules were acquired from Farillon Limited, (Romford, UK). Rotahaler1, Aerolizer1 and Handihaler1 devices were retrieved from commercial products. HPLC grade methanol and hexane and laboratory grade disodium hydrogen orthophosphate dodecahydrate were supplied by Fisher Chemicals (Loughborough, UK). Magnesium stearate and analytical grade citric acid monohydrate were supplied by Sigma–Aldrich Company Limited (Dorset, UK) while silicone oil was obtained from VWR International Limited (Lutterworth, UK).
2.1. Internal resistance of inhaler devices
85
Each DPI inhaler was loaded with an empty pierced capsule and attached to the dosage unit sampling apparatus (Copley Scientific Limited, Nottingham, UK) connected to a high-capacity vacuum pump (Model HCP5, Copley Scientific Limited, Nottingham, UK) via a critical flow controller (Model TPK, Copley Scientific Limited, Nottingham, UK). For each inhaler device, triplicate measurements of the pressure drop were recorded across the flow rate range of 30–100 L min1 at 10 L min1 intervals. A linear plot of the square root of the mean pressure drop against flow rate was obtained and the internal resistance of the DPI device was defined as the slope of the linear plot (Clark and Hollingworth, 1993).
86
2.2. Lactose particle size
97
To characterise the coarse lactose carriers, an amount of each a-lactose grade (60 mg) was dispersed in 15 mL of propan-2-ol with the aid of sonication for 1 min. The particle size distribution (PSD) of each lactose grade was characterised in terms of Dv10, Dv50 and Dv90 values (the volume diameter of 10%, 50% and 90% of aerosol droplets respectively); measured using a Spraytec1 (Malvern Instruments, Malvern, UK), equipped with a wet-cell dispersion unit. The span value provided an indication of the width of PSD and the percentage of particles with an equivalent volume diameter of less than 10 mm was also determined.
98
2.3. SX blends preparation
108
Based on a 32 factorial design, nine interactive mixtures, each containing 0.58% w/w SX and 0.08% w/w of magnesium stearate (to aid mixing), were prepared in 50 g batches with 10% overage, using different powder components (Table 1). The binary interactive mixtures consisted of SX particles and coarse carrier. Coarse lactose (2 g) was first used to ‘sandwich’ SX and the powders were mixed using a Spinmix Vortexer (Gallenkamp, Loughborough, UK). Coarse lactose was subsequently added in multiple stages (2.5 g, 5 g, 10 g, 10 g, 10 g and 10.21 g) and the powder mixture blended using a Turbula1 T2F shaker-mixer (Willy A. Bachofen AG, Basel, Switzerland) for 30 min at 67 rpm. The ternary interactive mixtures were three-component blends of SX particles, coarse carrier and added fines. The fine lactose and half of the coarse lactose were pre-blended for 30 min at 67 rpm, followed by the addition of the other half of the coarse lactose. This was blended for a further 30 min, before mixing with SX, similarly to the preparation of the binary mixtures, by gradual addition of the coarse-fine lactose preblend to the SX blends and mixing for 30 min at 67 rpm between each addition of the pre-blend.
109
Table 1 Composition of each interactive mixture prepared in accordance to the 32 factorial design, each containing 0.58% w/w salmeterol xinafoate and 0.08% magnesium stearate. Interactive mixture
Concentration of fine lactose added (%w/w)
Coarse lactose grade
M6F0 M6F5 M6F10 M1F0 M1F5 M1F10 LHF 0 LHF 5 LHF 10 LHF 20a
0 5 10 0 5 10 0 5 10 20
Respitose1 Respitose1 Respitose1 Respitose1 Respitose1 Respitose1 Lactohale1 Lactohale1 Lactohale1 Lactohale1
a
ML006 ML006 ML006 ML001 ML001 ML001 LH200 LH200 LH200 LH200
An additional mixture prepared to investigate poor blend homogeneity.
Please cite this article in press as: Hassoun, M., et al., Formulating powder–device combinations for salmeterol xinafoate dry powder inhalers. Int J Pharmaceut (2015), http://dx.doi.org/10.1016/j.ijpharm.2015.05.028
87 88 89 90 91 92 93 94 95 96
99 100 101 102 103 104 105 106 107
110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
G Model
IJP 14898 1–8 M. Hassoun et al. / International Journal of Pharmaceutics xxx (2015) xxx–xxx 128
2.4. Blend content uniformity
129
Blend homogeneity was evaluated by dissolving ten randomly selected samples (25 0.5 mg) of each interactive mixture in 65% v/v methanol in water mixture and quantifying SX content using high-performance liquid chromatography (HPLC). An interactive mixture was considered as homogeneous if no more than one sample was outside the limits of 85–115% of the mean salmeterol content and none was outside the limits of 75–125% of the mean salmeterol content (European Pharmacopeia, 2015).
130 131 132 133 134 135 136 137
2.5. In vitro deposition in the next generation impactor
138
The in vitro aerosolisation of each interactive mixture in different inhaler devices was evaluated using a next generation impactor (NGI). Ten size 3 hard gelatin capsules were loaded with the interactive mixture (12.5 0.5 mg) and discharged in each NGI test. The NGI collection stages were coated with 1% v/v silicone oil in hexane and the pre-separator filled with 15 mL deionised water. The NGI was connected to a high capacity vacuum pump via a critical flow controller. A leak test was carried out (to ensure the leak rate was less than 100 Pa/s) and the P3/P2 ratio at the test flow rate determined (flow rate assumed to be stable if P3/P2 0.5) (U.S. Pharmacopeia, 2015). Each dose was actuated at 100 L min1 for 2.4 s with the Rotahaler and Aerolizer, and at 60 L min1 for 4 s with the Handihaler. Impactor testing was performed at room temperature (22.0 3 C) and a relative hsumidity of 50 10%. Each formulation–device combination was tested four times. SX was recovered by rinsing each component of the NGI with 65% v/v methanol in water mixture, into appropriate volumetric flasks. The salmeterol content was quantified using HPLC.
139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156
2.6. HPLC analysis of salmeterol
157
SX was analysed by HPLC. A mixture of methanol and filtered buffer solution, consisting of 0.05 M citric acid and 0.1 M di-sodium hydrogen orthophosphate dodecahydrate solution (65:35, pH 4.0 0.1), was used as the mobile phase at a flow rate of 1.1 mL min1, with UV detection at 228 nm. The HPLC system consisted of a 15 cm (4.6 mm i.d.) column packed with 2.7 mm C-18 (Agilent Technologies, Cheshire, UK) with the column temperature maintained at 15 C. With reference to an external standard, quantification of SX content was carried out by integration of the peak corresponding to the retention time of 3.5 min. The method was validated over a concentration range of 1–50 mg/ml and the linearity was verified (R2 = 0.999) and the method was shown to be fit for purpose with a limit of quantification within this range, a precision of 2% and accuracy of 99%.
158 159 160 161 162 163 164 165 166 167 168 169 170 171 172
2.7. Data analysis
173
181
The emitted dose (ED), expressed as a percentage of the nominal dose (50 mg), was denoted as the total amount of drug recovered in the impactor per capsule. The mass median aerodynamic diameter (MMAD) and geometric standard deviation (GSD) were derived from a plot of cumulative fraction of salmeterol against NGI cut-off diameters (European Pharmacopoeia, 2015). The fine particle dose (FPD) was the mass of drug particles with aerodynamic diameter lower than 5 mm and the fine particle fraction (FPF) was calculated as the percentage ratio between FPD and ED.
182
2.8. Statistical analysis
183
Data was analysed for normality using an Anderson–Darling test. GraphPad Prism (v5.0) was used to perform statistical analysis
174 175 176 177 178 179 180
184
3
for all graphs. Two-way analysis of variance (ANOVA, SPSS v17.0) was used to determine the statistical significance of results.
185
3. Results
187
3.1. Internal resistance of inhaler devices
188
Internal resistance was calculated for each inhaler from the linear plots (R2 > 0.99) of pressure drop versus inspiratory flow rate (Fig. 1). Internal resistance was different for each type of inhaler p p device: 0.0122 kPa L min1 for the Rotahaler, 0.0182 kPa L min1 p 1 for the Aerolizer and 0.0419 kPa L min for the Handihaler and compared favourably with those reported previously by Frijlink and de Boer (2004). In vitro deposition studies were conducted using the Rotahaler, Aerolizer and Handihaler to represent devices with low and high resistance and at the range of 1.5–6.3 kPa pressure drops typically attained by the patient population, in order to investigate the effect of this variable on the performance (Table 2).
189
3.2. Blend homogeneity
201
All three coarse lactose grades displayed monomodal distributions (data not shown). They differed in terms of volume median diameter (Dv50), span and intrinsic fines content (Table 3). An inverse relationship between Dv50 and intrinsic fines content was evident. The dimensions of the fine particles were not measured in this study but have been reported previously; SX (Jaffari et al., 2014) and Lactohale LH 300 (Kinnunen et al., 2014). Only six of the nine interactive blends fulfilled the criteria for accuracy and uniformity outlined in the European Pharmacopoeia, exhibiting mean drug contents ranging from 95 to 105% of the theoretical salmeterol content and inside the acceptance limit of relative standard deviations (RSDs) of
IJP 14898 1–8 International Journal of Pharmaceutics xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
International Journal of Pharmaceutics journal homepage: www.elsevier.com/locate/ijpharm
1 2
3 Q1 4 5 6 7 Q2
Formulating powder–device combinations for salmeterol xinafoate dry powder inhalers Mireille Hassoun a , Shirlene Ho a , Joanna Muddle a , Francesca Buttini a,c , Mark Parry b , Mark Hammond b , Ben Forbes a, * a b c
Department of Pharmacy, King’s College London, 150 Stamford Street, London SE1 9NH, UK Intertek-Melbourn Scientific Limited, Saxon Way, Melbourn SG8 6DN, UK Department of Pharmacy, University of Parma, Parco Area delle Scienze 27/a, Parma 43124, Italy
A R T I C L E I N F O
A B S T R A C T
Article history: Received 31 March 2015 Received in revised form 8 May 2015 Accepted 11 May 2015 Available online xxx
Using salmeterol xinafoate (SX) as an active pharmaceutical ingredient, the effects of carrier lactose particle type, total lactose fines content and device resistance on dry powder inhaler performance were investigated in vitro. To mimic drug levels in commercial preparations, interactive mixtures containing 0.58% w/w SX were prepared by low shear tumble mixing. Three types of milled inhalation grade lactose were used (Lactohale1 LH 200, Respitose1 ML006 and ML001) and the concentration of fine lactose (Lactohale1 300) added was varied. The in vitro deposition of each mixture was studied using a next generation impactor and inhaler devices exhibiting different resistances, Rotahaler1 < Aerolizer1 < Handihaler1. Aerosol performance was evaluated based on the emitted dose (ED), mass median aerodynamic diameter (MMAD) geometric standard deviation (GSD) and fine particle fraction (FPF). Increases of up to eight-fold in FPF were observed with increasing intrinsic fine lactose content. The addition of extra fine lactose increased the FPF further, although the effect diminished as more fines were added. The Aerolizer produced the best aerosol performance with any given powder blend, although suitable formulations were identified for each device as defined by the a priori success criteria: >80% ED and MMAD GSD between 1–5 mm. The results confirmed the factors under investigation to be important determinants of product performance, but demonstrated using realistic conditions how individual factor impact may be enhanced or mitigated by inter-dependency. ã 2015 Published by Elsevier B.V.
Keywords: Orally inhaled product (OIP) Salmeterol xinafoate (SX) Dry powder inhaler Lactose Rotahaler Aerolizer Handihaler
8
1. Introduction
9
Dry powder inhalers (DPI) are increasingly popular products for delivering drugs to the lungs due to their ease of operation, environmental sustainability and formulation stability (Telko and Hickey, 2005). Historically, DPI have delivered only 20–30% of the emitted dose to the lungs (De Koning, 2001), but more efficient products are emerging (Friebel et al., 2012). Although the scientific literature now provides well-established principles with which to design efficient and effective DPIs (Friebel et al., 2012; Hoppentocht et al., 2014), there is surprisingly little reported regarding the
10 11 12 13 14 Q3 15 16 17
Abbreviations: SX, salmeterol xinafoate; DPI, dry powder inhaler; PSD, particle size distribution; ED, emitted dose; MMAD, mass median eerodynamic diameter; GSD, geometric standard deviation; FPF, fine particle fraction; FPD, fine particle dose. * Corresponding author. Tel.: +44 207 8484823. E-mail address: [email protected] (B. Forbes).
practicalities of formulating powder blends for different inhaler Q4 devices. Drug particles within the respirable size range, 1–5 mm, are generally highly cohesive and tend to form agglomerates, resulting in poor flow properties, non-uniform dispersion and low dose uniformity. Therefore, drug particles are usually blended with coarse carrier particles, typically a-lactose, to help provide bulk and form an ordered mix (Pilcer and Amighi, 2010; Javadzadeh et al., 2012). Since the active drug and lactose powders exhibit different sized particles, the formulation and blending process is important to achieve uniform drug distribution and provide consistency of delivered dose (Pilcer and Amighi, 2010; Saleem and Smyth, 2008). Powders need to be sufficiently adhesive to possess good flow properties without compromising their ability to generate respirable aerosols efficiently and reproducibly (Begat et al., 2004). Powder dispersion can be manipulated by modifying the properties of microparticle drug surface or the carrier’s surface morphology and size (Buttini et al., 2008; Pomazi et al., 2013; Zeng et al., 2001; Islam et al., 2004; Zellnitz et al., 2011). A proportion of
http://dx.doi.org/10.1016/j.ijpharm.2015.05.028 0378-5173/ ã 2015 Published by Elsevier B.V.
Please cite this article in press as: Hassoun, M., et al., Formulating powder–device combinations for salmeterol xinafoate dry powder inhalers. Int J Pharmaceut (2015), http://dx.doi.org/10.1016/j.ijpharm.2015.05.028
18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
G Model
IJP 14898 1–8 2 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84
M. Hassoun et al. / International Journal of Pharmaceutics xxx (2015) xxx–xxx
fine lactose in the formulation, either as intrinsic or added fines, will further improve dispersion and the de-aggregation of the drug particles (Jones and Price, 2006; Beilmann et al., 2007; Kinnunen et al., 2014; Shur et al., 2008; Grasmeijer et al., 2014; Thalberg et al., 2012). Device properties also determine respirable aerosol generation from a carrier-based powder formulations (Martinelli et al., 2015; Shur et al., 2012). The airflow velocity must also overcome the adhesion forces between the drug and carrier particles to detach the micronized drug particles and disperse them in the aerosol. To aid dispersion, inhaler devices may be engineered with a more tortuous airflow that increases the resistance within the device (Coates et al., 2006; Selvam et al., 2010). Capsule-based inhalers with different resistances are available, which enables the effect of device resistance on dry powder aerosolisation to be investigated. Although aerosolisation is governed by the powder formulation and the inhaler device design inter-dependently (Frijlink and de Boer, 2004), devices are generally developed as platform technology with which to deliver different drugs to the lungs. Scientific studies aimed at understanding DPI performance often examine the influence of device or formulation as independent variables, or use model systems. For example, effect of adding fine excipient particles has been investigated by looking at the effect of fine lactose on the aerosolisation of salmeterol xinafoate (SX) 1–5% w/w interactive powder mixtures (Islam et al., 2004; Adi et al., 2008). The present study explored the inter-dependency of powder formulation and device factors in typical capsule-based DPI products by varying the carrier particle size distribution and the amount of added lactose fines. Three DPI devices having different intrinsic resistance were employed to aerosolize the SX blends. A 32 factorial design was constructed to understand the relative influence of these variables on fine particle fraction (FPF), emitted dose (ED), mass median aerodynamic diameter (MMAD) and the geometric standard deviation (GSD). 2. Materials and methods Micronised SX was supplied by Vamsi Labs Ltd., Maharashtra, India. Milled inhalation a-lactose monohydrate grades (Respitose1 ML006, Respitose1 ML001, Lactohale1 LH 200 and Lactohale1 300) were obtained as samples from DFE Pharma (Veghel, Netherlands). Size 3 hard gelatin capsules were acquired from Farillon Limited, (Romford, UK). Rotahaler1, Aerolizer1 and Handihaler1 devices were retrieved from commercial products. HPLC grade methanol and hexane and laboratory grade disodium hydrogen orthophosphate dodecahydrate were supplied by Fisher Chemicals (Loughborough, UK). Magnesium stearate and analytical grade citric acid monohydrate were supplied by Sigma–Aldrich Company Limited (Dorset, UK) while silicone oil was obtained from VWR International Limited (Lutterworth, UK).
2.1. Internal resistance of inhaler devices
85
Each DPI inhaler was loaded with an empty pierced capsule and attached to the dosage unit sampling apparatus (Copley Scientific Limited, Nottingham, UK) connected to a high-capacity vacuum pump (Model HCP5, Copley Scientific Limited, Nottingham, UK) via a critical flow controller (Model TPK, Copley Scientific Limited, Nottingham, UK). For each inhaler device, triplicate measurements of the pressure drop were recorded across the flow rate range of 30–100 L min1 at 10 L min1 intervals. A linear plot of the square root of the mean pressure drop against flow rate was obtained and the internal resistance of the DPI device was defined as the slope of the linear plot (Clark and Hollingworth, 1993).
86
2.2. Lactose particle size
97
To characterise the coarse lactose carriers, an amount of each a-lactose grade (60 mg) was dispersed in 15 mL of propan-2-ol with the aid of sonication for 1 min. The particle size distribution (PSD) of each lactose grade was characterised in terms of Dv10, Dv50 and Dv90 values (the volume diameter of 10%, 50% and 90% of aerosol droplets respectively); measured using a Spraytec1 (Malvern Instruments, Malvern, UK), equipped with a wet-cell dispersion unit. The span value provided an indication of the width of PSD and the percentage of particles with an equivalent volume diameter of less than 10 mm was also determined.
98
2.3. SX blends preparation
108
Based on a 32 factorial design, nine interactive mixtures, each containing 0.58% w/w SX and 0.08% w/w of magnesium stearate (to aid mixing), were prepared in 50 g batches with 10% overage, using different powder components (Table 1). The binary interactive mixtures consisted of SX particles and coarse carrier. Coarse lactose (2 g) was first used to ‘sandwich’ SX and the powders were mixed using a Spinmix Vortexer (Gallenkamp, Loughborough, UK). Coarse lactose was subsequently added in multiple stages (2.5 g, 5 g, 10 g, 10 g, 10 g and 10.21 g) and the powder mixture blended using a Turbula1 T2F shaker-mixer (Willy A. Bachofen AG, Basel, Switzerland) for 30 min at 67 rpm. The ternary interactive mixtures were three-component blends of SX particles, coarse carrier and added fines. The fine lactose and half of the coarse lactose were pre-blended for 30 min at 67 rpm, followed by the addition of the other half of the coarse lactose. This was blended for a further 30 min, before mixing with SX, similarly to the preparation of the binary mixtures, by gradual addition of the coarse-fine lactose preblend to the SX blends and mixing for 30 min at 67 rpm between each addition of the pre-blend.
109
Table 1 Composition of each interactive mixture prepared in accordance to the 32 factorial design, each containing 0.58% w/w salmeterol xinafoate and 0.08% magnesium stearate. Interactive mixture
Concentration of fine lactose added (%w/w)
Coarse lactose grade
M6F0 M6F5 M6F10 M1F0 M1F5 M1F10 LHF 0 LHF 5 LHF 10 LHF 20a
0 5 10 0 5 10 0 5 10 20
Respitose1 Respitose1 Respitose1 Respitose1 Respitose1 Respitose1 Lactohale1 Lactohale1 Lactohale1 Lactohale1
a
ML006 ML006 ML006 ML001 ML001 ML001 LH200 LH200 LH200 LH200
An additional mixture prepared to investigate poor blend homogeneity.
Please cite this article in press as: Hassoun, M., et al., Formulating powder–device combinations for salmeterol xinafoate dry powder inhalers. Int J Pharmaceut (2015), http://dx.doi.org/10.1016/j.ijpharm.2015.05.028
87 88 89 90 91 92 93 94 95 96
99 100 101 102 103 104 105 106 107
110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
G Model
IJP 14898 1–8 M. Hassoun et al. / International Journal of Pharmaceutics xxx (2015) xxx–xxx 128
2.4. Blend content uniformity
129
Blend homogeneity was evaluated by dissolving ten randomly selected samples (25 0.5 mg) of each interactive mixture in 65% v/v methanol in water mixture and quantifying SX content using high-performance liquid chromatography (HPLC). An interactive mixture was considered as homogeneous if no more than one sample was outside the limits of 85–115% of the mean salmeterol content and none was outside the limits of 75–125% of the mean salmeterol content (European Pharmacopeia, 2015).
130 131 132 133 134 135 136 137
2.5. In vitro deposition in the next generation impactor
138
The in vitro aerosolisation of each interactive mixture in different inhaler devices was evaluated using a next generation impactor (NGI). Ten size 3 hard gelatin capsules were loaded with the interactive mixture (12.5 0.5 mg) and discharged in each NGI test. The NGI collection stages were coated with 1% v/v silicone oil in hexane and the pre-separator filled with 15 mL deionised water. The NGI was connected to a high capacity vacuum pump via a critical flow controller. A leak test was carried out (to ensure the leak rate was less than 100 Pa/s) and the P3/P2 ratio at the test flow rate determined (flow rate assumed to be stable if P3/P2 0.5) (U.S. Pharmacopeia, 2015). Each dose was actuated at 100 L min1 for 2.4 s with the Rotahaler and Aerolizer, and at 60 L min1 for 4 s with the Handihaler. Impactor testing was performed at room temperature (22.0 3 C) and a relative hsumidity of 50 10%. Each formulation–device combination was tested four times. SX was recovered by rinsing each component of the NGI with 65% v/v methanol in water mixture, into appropriate volumetric flasks. The salmeterol content was quantified using HPLC.
139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156
2.6. HPLC analysis of salmeterol
157
SX was analysed by HPLC. A mixture of methanol and filtered buffer solution, consisting of 0.05 M citric acid and 0.1 M di-sodium hydrogen orthophosphate dodecahydrate solution (65:35, pH 4.0 0.1), was used as the mobile phase at a flow rate of 1.1 mL min1, with UV detection at 228 nm. The HPLC system consisted of a 15 cm (4.6 mm i.d.) column packed with 2.7 mm C-18 (Agilent Technologies, Cheshire, UK) with the column temperature maintained at 15 C. With reference to an external standard, quantification of SX content was carried out by integration of the peak corresponding to the retention time of 3.5 min. The method was validated over a concentration range of 1–50 mg/ml and the linearity was verified (R2 = 0.999) and the method was shown to be fit for purpose with a limit of quantification within this range, a precision of 2% and accuracy of 99%.
158 159 160 161 162 163 164 165 166 167 168 169 170 171 172
2.7. Data analysis
173
181
The emitted dose (ED), expressed as a percentage of the nominal dose (50 mg), was denoted as the total amount of drug recovered in the impactor per capsule. The mass median aerodynamic diameter (MMAD) and geometric standard deviation (GSD) were derived from a plot of cumulative fraction of salmeterol against NGI cut-off diameters (European Pharmacopoeia, 2015). The fine particle dose (FPD) was the mass of drug particles with aerodynamic diameter lower than 5 mm and the fine particle fraction (FPF) was calculated as the percentage ratio between FPD and ED.
182
2.8. Statistical analysis
183
Data was analysed for normality using an Anderson–Darling test. GraphPad Prism (v5.0) was used to perform statistical analysis
174 175 176 177 178 179 180
184
3
for all graphs. Two-way analysis of variance (ANOVA, SPSS v17.0) was used to determine the statistical significance of results.
185
3. Results
187
3.1. Internal resistance of inhaler devices
188
Internal resistance was calculated for each inhaler from the linear plots (R2 > 0.99) of pressure drop versus inspiratory flow rate (Fig. 1). Internal resistance was different for each type of inhaler p p device: 0.0122 kPa L min1 for the Rotahaler, 0.0182 kPa L min1 p 1 for the Aerolizer and 0.0419 kPa L min for the Handihaler and compared favourably with those reported previously by Frijlink and de Boer (2004). In vitro deposition studies were conducted using the Rotahaler, Aerolizer and Handihaler to represent devices with low and high resistance and at the range of 1.5–6.3 kPa pressure drops typically attained by the patient population, in order to investigate the effect of this variable on the performance (Table 2).
189
3.2. Blend homogeneity
201
All three coarse lactose grades displayed monomodal distributions (data not shown). They differed in terms of volume median diameter (Dv50), span and intrinsic fines content (Table 3). An inverse relationship between Dv50 and intrinsic fines content was evident. The dimensions of the fine particles were not measured in this study but have been reported previously; SX (Jaffari et al., 2014) and Lactohale LH 300 (Kinnunen et al., 2014). Only six of the nine interactive blends fulfilled the criteria for accuracy and uniformity outlined in the European Pharmacopoeia, exhibiting mean drug contents ranging from 95 to 105% of the theoretical salmeterol content and inside the acceptance limit of relative standard deviations (RSDs) of
