- Email: [email protected]
Effect of carrier size on the dispersion of salmeterol xinafoate from interactive mixtures
Effect of Carrier Size on the Dispersion of Salmeterol Xinafoate from Interactive Mixtures NAZRUL ISLAM,1 PETER STEWART,1 IAN LARSON,1 PATRICK HARTLEY2 1
Department of Pharmaceutics, Victorian College of Pharmacy, Monash University, 381 Royal Parade, Parkville Vic 3052, Australia 2
Division of Molecular Science, CSIRO, Clayton Vic 3168, Australia
Received 18 March 2003; revised 30 June 2003; accepted 27 August 2003
ABSTRACT: The objective of this study was to determine the influence of lactose carrier size on drug dispersion of salmeterol xinafoate (SX) from interactive mixtures. SX dispersion was measured by using the fine particle fractions determined by a twin stage impinger attached to a Rotahaler1. The particle size of the lactose carrier in the SX interactive mixtures was varied using a range of commercial inhalation-grade lactoses. In addition, differing size fractions of individual lactose samples were achieved by dry sieving. The dispersion of SX appeared to increase as the particle size of the lactose carrier decreased for the mixtures prepared from different particle size commercial samples of lactose and from different sieve fractions of the same lactose. Fine particles of lactose (<5 mm) associated with the lactose carrier were removed from the carrier surface by a wet decantation process to produce lactose samples with low but similar concentrations of fine lactose particles. The fine particle fractions of SX in mixtures prepared with the decanted lactose decreased significantly (analysis of variance, p < 0.001) and the degree of dispersion became independent of the volume mean diameter of the carriers (analysis of variance, p < 0.05). The dispersion behavior is therefore associated with the presence of fine adhered particles associated with the carriers and the inherent size of the carrier itself has little influence on dispersion. ß 2004 Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci 93:1030–1038, 2004
Keywords: drug dispersion; lactose carrier size; salmeterol xinafoate; decantation; dry powder inhalation; interactive mixtures
INTRODUCTION Dry powders for inhalation are formulated either as loose agglomerates of micronized drug particles or as carrier-based interactive mixtures with micronized drug particles adhered onto the surface of large lactose carriers. For the carrier-based interactive mixtures, three major factors are of prime concern for desirable therapeutic effects: 1.
Correspondence to: Peter Stewart þ61399039517; Fax: 61399039583; E-mail: [email protected])
(Telephone:
Journal of Pharmaceutical Sciences, Vol. 93, 1030–1038 (2004) ß 2004 Wiley-Liss, Inc. and the American Pharmacists Association
1030
the detachment of drug particles from the large carriers; 2. dispersion of the detached drug particles during aerosolization by an inhalation device; and 3. subsequent deposition in the lower airways of lung. Therefore, any factors influencing these processes will affect the therapeutic activity. For example, the geometry of the drug particle and the respiratory tract will influence dispersion and particles with aerodynamic particle sizes of <5 mm are necessary for effective deep lung deposition.1 However, given that micronized particles are cohesive and are likely to agglomerate, and are adhesive and are likely to adhere to lactose carriers, the extents of detachment and subsequent dispersion of micronized drug
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 4, APRIL 2004
DISPERSION OF SALMETEROL XINAFOATE
particles are key elements in determining the efficiency of the formulation for respiratory delivery. The mechanism by which drugs are dispersed is not fully understood. The magnitude of particle interactions between drugs and between drugs and excipients will be important to the dispersion process. One theory that has been advanced to explain dispersion involves the interaction of fine lactose present in the lactose excipient with the high-energy, active sites on the carrier’s surface causing saturation of these active sites, leaving low-energy, passive sites available for drug adhesion.2,3 The reduced adhesion between drug and carrier particles increased drug detachment. Recently, a hypothesis based on competitive adhesion and redistribution of the drug between the carrier surface and fine excipient to produce mixed agglomerates of drug and fine excipient was proposed to explain improved drug dispersion.4 Greater efficiencies were likely to result through facilitated drug detachment from the carrier surface as agglomerates and better dispersion of mixed agglomerates by the dry powder inhaler. Several innovations to improve the delivery of drugs have been reported. These studies have focused on changing particle and in particular surface morphology, using alternate carriers and utilizing fine lactose or other particles to enhance deposition. Some of these studies are highlighted as follows. Smoothing the carrier surface improved the respirable fraction of albuterol sulfate.5,6 Fine particles of ternary materials added to the interactive mixtures increased the fine particle fraction (FPF) of several drugs. Inclusion of magnesium stearate, L-leucine, and micronized glucose in formulations produced higher in vitro deposition of salbutamol sulfate4,7 and the addition of fine particles of lactose to dry powder formulations improved dispersion and deposition of salbutamol sulfate2,3,8 and beclomethasone dipropionate.9 Using the Twin Stage Impinger (TSI), the FPF of salbutamol sulfate increased with increasing concentration of fine particles of lactose (<10 mm).10 The FPF of salbutamol sulfate increased with increasing the concentration of fine carrier of synthesized sugars.11 Addition of intermediate-sized lactose and micronized glucose also increased the FPF of salbutamol sulfate.3,12 A ternary mixture of salbutamol sulfate, fine and coarse lactose prepared by first blending the coarse and fine lactose before mixing with drug, produced higher FPFs of salbutamol sulfate.13 An increased FPF of spray dried bovine serum albumin occurred with
1031
increasing concentrations of fine lactose (5.4 mm) up to 5%, but no further increase of FPF was observed at higher concentration (7.5–10%).14 Recently, a higher FPF and dispersibility of salbutamol sulfate was found using more elongated crystals of lactose carriers.6 Carrier particle size is important in the design of dry powder inhalers and the effect of carrier size on drug dispersion has been reported.15–19 Greatest dispersion of cromolyn sodium from an interactive mixture with lactose particles (70–100 mm) at a flow rate of 60 L min1 was observed.15 Using binary mixtures of salbutamol sulfate and synthesized sugar, the FPF decreased with increasing carrier particle size.10 A reduction in carrier size improved respirable fraction of albuterol sulfate16,18 and budesonide.19 However, a higher respirable fraction of terbutaline sulfate was obtained from coarser lactose (53–105 mm) than from a finer lactose (<53 mm).20 From the studies above, the particle size distribution of the carrier influenced the respiratory in vitro deposition of drugs. It is not clear from these studies whether the intrinsic carrier size or the proportion of fine lactose particles associated with the carrier influenced dispersion because the proportion of small lactose particles adhered to the carrier surface are likely to be greater as the particle size distribution is decreased. In this study, we have investigated the effect of only carrier size on drug dispersion by removing the fine lactose from the large lactose by sieving and wet decantation. In this way, we hope to determine specific particle size effects on drug dispersion; such determinations are important in understanding the mechanism of dispersion. The study was conducted using salmeterol xinafoate (SX) as the model drug and a range of commercial samples of lactose.
EXPERIMENTAL Materials SX, micronized (inhalation grade; <4 mm), was obtained from GlaxoSmithKline, Australia. Aeroflo 95, Aeroflo 65, and Aeroflo 20, inhalation grades of a-lactose monohydrate, were donated by Foremost Farms, Rothschild, WI; other lactose grades of Inhalac 120, Inhalac 230, and Sorbolac 400 were donated by Meggle GmbH, Germany. Ammonium acetate (Analar, BDH, Victoria), methanol [high-performance liquid JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 4, APRIL 2004
1032
ISLAM ET AL.
chromatography (HPLC) grade, Biolab, Victoria], and absolute alcohol (HPLC grade, CSR, Victoria) were used as supplied. Dry Sieving and Fractionation Fractions (45–63, 63–90, 90–106, and >106 mm) of Aeroflo 95 and 65 were produced by dry sieving for 30 min using a sieve shaker (Fritsch, Germany) and test sieves (Endecotts, UK), and were stored in a desiccator over silica gel. Decantation and Removal of Fine Lactose Fine lactose particles were removed from the unsieved and sieved fractions of lactose carrier (Aeroflo 65 and 95) by a decantation method.21 Approximately 20 g of lactose was dispersed in absolute alcohol, presaturated with lactose for 5 min to make a homogeneous suspension and allowed to settle for a predetermined time. The supernatant liquid was decanted and replaced by fresh saturated ethanol, vigorously mixed, and the decantation process repeated (15 times) until the supernatant liquid was clear. Although 15 repeats produced a clear supernatant,22 Heywood23 estimated that approximately 50 repeats would be necessary to eliminate 99% of particles near the size of separation. Particle Size of Lactose The particle size of the lactose carriers was measured by laser diffraction (Malvern Mastersizer S; Malvern Instruments Ltd., UK) using the 300 RF lens and the small-volume sample presentation unit (capacity 150 mL). Approximately 500 mg of lactose powder was dispersed in 5 mL of butan-1-ol with the aid of a sonication in a waterbath for 3 min. Sonication was necessary because unsonicated samples showed significant agglomeration. The sonicated sample was added drop-wise into the sample cell containing 150 mL of ethanol until an obscuration between 10–30% was obtained. Size measurement of each sample was performed using 2000 sweeps and analyzed with the reference refractive index of lactose (1.533) and ethanol (1.36) and calculated imaginary refractive index of lactose is 0.1. The average particle size distribution was measured from five replicates of each sample. The size distributions were characterized by the 10, 50, and 90 percentiles (d10%, d50%, and d90%, respectively). The volume mean diameter (VMD), which is defined JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 4, APRIL 2004
as Snd4/Snd3, where n is the number of particles and d is the equivalent spherical diameter, was determined from the output of the laser diffraction particle sizing and was used as the size parameter to characterize the lactose carrier distributions. The percentage of particles <5 mm was determined using the cumulative frequency distribution (undersized) curve. The residual value was always <1%. Preparation of Interactive Mixtures and Filling of Capsules The interactive mixtures were prepared by a hand mixing method.24,25 Binary interactive mixtures of SX (2.5%) and different grades of lactoses were prepared in 5-g batches. The drug powders were placed between two layers of carrier powders in a glass test tube with three ceramic beads of approximately 10-mm diameter. After stoppering, the test tube was then vigorously shaken by hand for 5 min, where the ceramic beads provided a ball-milling effect for breaking up the drug agglomerates formed during mixing. The powder formulations were filled (20 mg) into hard gelatin capsules (size 3; Fawns and McAllan Pty Ltd., Australia) manually. Homogeneity Test of Interactive Mixtures The homogeneity of each powder mixture prepared for this experiment was assessed by the mean drug content and variability between samples. Twenty samples, each of approximately 20 mg, were dissolved in 40% methanol (HPLC grade) and the amount of SX was determined by a validated UV assay. An acceptable degree of homogeneity was achieved with a mean drug content within 100 5% (mean SD) of the theoretical value and a coefficient of variation <5% for all samples.26 Drug Dispersion by TSI Using a Rotahaler1 (Glaxo Wellcome), the in vitro aerosol deposition of the powder formulations was determined using a TSI (Apparatus, A; British Pharmacopoea, 2000) (Copley, UK). Seven and thirty milliliters of 60% methanol (HPLC grade) were placed into stage one and stage two of the TSI, respectively. The air flow was drawn through the TSI using a vacuum pump (model OD5/2; Dynavac Engineering, Australia) and the air flow rate was adjusted to 60 L/min at the mouthpiece,
DISPERSION OF SALMETEROL XINAFOATE
before each measurement (MODEL 10A3567SAX; Fisher and Porter, UK). The FPF was defined as the percent of drug particles deposited in the lower stage of the TSI of the recovered dose (RD). The RD was the total amount of drug collected from inhaler, stage one (S1), and stage two (S2). The emitted dose was the percent of drug delivered from the inhaler of RD. HPLC Analysis of SX SX was analyzed by HPLC using a C18 column (mBondapakTM, 3.9 300 mm; Waters) and an UV detector (Waters Tunable Absorbance Detector) at a wavelength of 252 nm. A mixture of methanol and 0.2% (w/v) ammonium acetate solution (55:45, pH 6.9) was used as a mobile phase running at a flow rate of 1.0 mL/min by an HPLC pump (Waters 510). The peak area was recorded by integration (Shimadzu CR6A Chromatopack, Japan). The retention time of SX was 4.2–4.9 min (the variation from batch to batch ranged from 4.2–4.9 min whereas a single batch showed a constant retention time). The calibration plot of standard SX solution was linear over the range of 0.4–10 mg/mL with r2 ¼ 1.00. Five replicates of standards samples of 0.4, 1.0, 4.0, and 10.0 mg/mL solutions were performed for assay validation and the mean accuracy was 91.4 2.4%, 100.4 5.3%, 101.3 1.1%, and 99.8 0.6%, respectively (mean SD). The precision was tested before each experiment by analyzing 4.0 mg/mL standard solution of SX using five replicates and the coefficient of variation was <2%. Statistical Analysis Linear regression analysis (Sigmastat; Jandel Scientific) was used in the calibration of HPLC and UV analyses and testing relationships between FPF and VMD. Comparison between different groups of FPF was performed by using
1033
one way analysis of variance (ANOVA) (Sigmastat; Jandel Scientific). Covariance Analysis by Minitab (Minitab Corporation) was undertaken to compare the slopes of regression lines.
RESULTS AND DISCUSSION Influence of Carriers on Dispersion SX (2.5%) was mixed with six different grades of lactose having different inherent size distributions summarized in Table 1. Interactive mixtures were formed with the drug particles adhering to the surface of the lactose; these mixtures showed excellent homogeneity with means within 100 2.5% of the target and coefficients of variation <3%. The FPF of SX was determined for each of the mixes and increased as the VMD of the lactose carrier decreased (Fig. 1). There is a significant difference between the FPF of SX at the differing VMD of the lactose carriers (ANOVA, p < 0.05). Such findings have been supported by other studies in our laboratory using 5% binary mixtures of salbutamol sulfate and a series of different lactose samples from three different manufacturers.11 The percentage of small particles (e.g., <5 mm) present in each carrier was seen to increase as the particle size distribution of the carrier (characterized by the VMD) decreased (Table 1). The presence of small lactose particles associated with the carriers has been shown in a number of studies to increase dispersion of micronized drug particles from mixtures of drug and lactose.2,3,8–14 Therefore, the results shown in Figure 1 could be associated either with a specific particle size effect of the lactose carriers where decreasing carrier particle size inherently improves SX dispersion or with the effect of increasing concentrations of fine lactose particles present on the lactose carrier surface as the particle size of the lactose carrier decreases. To identify the real cause of this
Table 1. Particle Size Distributions of Different Lactose Carriers Sample Sorbolac 400 Aeroflo 20 Aeroflo 65 Aeroflo 95 Inhalc 230 Inhalc 120
d10% 1.5 3.2 9.9 19.6 69.1 114.2
(0.1) (0.1) (0.2) (0.3) (0.3) (1.7)
d50%
d90%
8.5 (0.1) 22.3 (0.3) 72.8 (1.3) 101.0 (0.8) 110.3 (0.2) 156.5 (1.2)
20.3 (0.2) 48.7 (2.0) 167.9 (2.8) 222.4 (1.6) 156.4 (0.9) 209.4 (2.9)
<5 mm (%) 30.3 14.3 6.2 3.9 1.6 1.2
(0.8) (0.2) (0.0) (0.1) (0.0) (0.1)
VMD (mm) 9.99 (0.9) 24.9 (0.6) 82.5 (1.3) 112.7 (0.9) 110.6 (0.3) 157.6 (2.1)
Data are mean SD, n ¼ 5. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 4, APRIL 2004
1034
ISLAM ET AL.
Figure 1. Relationship between FPF of SX and VMD of six different lactose carriers, determined using the TSI from 2.5% SX interactive mixtures using the Rotahaler1.
behavior, fractions of Aeroflo 65 and 95 were selected to determine the effect of the removal of lactose fines from the surface of these carriers on the dispersion of SX. Classification of Lactose Carrier Fractions Fractions of 45–63, 63–90, 90–106, and >106 mm of Aeroflo 65 and 95 were obtained by sieve classification. The particle size distributions of these carriers were determined by laser diffraction (Table 2). The sieve fractions of Aeroflo 65 and 95 were found to be significantly different from each other (ANOVA, p < 0.001) with the VMD of the sieved fractions of Aeroflow 65 being significantly lower than those of the corresponding fractions of Aeroflo 95 (Tukey test, p < 0.05). This might be due to the presence of greater amounts of the fine lactose in the Aeroflo 65 that adhered to the large lactose carrier surface even after sieving. The sieved fractions of Aeroflo 65 and 95 were subjected to wet decantation to remove the fine lactose associated with these lactose carriers.
Table 2 shows that the VMD increased after decantation because of the removal of fine particles of lactose and that wet decantation decreased the percentage of fine lactose in the sieved fractions of Aeroflo 65 and 95 with the decanted fractions showing about 60–70% removal of fine lactose after 15 repeat wet decantation cycles. After decantation, the Aeroflo 65 and 95 fractions contained similar amounts of fine lactose. It was more difficult to remove adhered fines from the smallest fractions of Aeroflo 65 and 95. The 45–63 mm fraction of Aeroflo 65 and 95 contained 2.8 and 3.0% of particles <5 mm, respectively, whereas all other fractions showed more consistent concentrations of fine particles <5 mm at about 1–2%. Effect of Carrier Size (VMD) on Drug Dispersion The FPFs of SX (2.5%) mixtures with the dry sieved and decanted fractions of Aeroflo 65 and 95 were determined. The FPF of SX increased with decreased VMD of different dry sieved fractions of Aeroflo 95 as expected (Fig. 2). FPFs of SX were significantly different for the sieved fractions (45–63, 63–90, 90–106, and >106 mm) of Aeroflo 95. Pair-wise comparisons found significant differences between 45–63 mm fraction versus the other three fractions, and 63–90 mm fraction versus 90–106 mm fraction and >106 mm fraction, but no difference between 90–106 mm fraction and >106 mm fraction, with p < 0.05 level of significance. The relationship between FPF of SX and VMD of different sieved fractions of Aeroflo 95 showed good linearity and a significant corelationship (r2 ¼ 0.922, p < 0.04). Interestingly, after decantation, the FPFs of SX of all fractions were significantly reduced in comparison with those of the corresponding fractions before decantation (p < 0.001), reinforcing the influence of the fine adhered lactose in the dispersion process. Except for the 45–63 mm fraction, which
Table 2. Particle Size Distributions of Dry Sieved and Decanted Fractions of Aeroflo 95 and Aeroflo 65 VMD (mm) Sieved
% FL (<5 mm) Decanted
Size range
Aeroflo 95
Aeroflo 65
Aeroflo 95
Aeroflo 65
45–63 63–90 90–106 >106
45.9 (0.1) 82.1 (1.2) 120.7 (0.5) 184.4 (0.7)
38.3 (0.7) 70.2 (0.9) 106.3 (0.9) 166.6 (1.6)
62.7 (0.12) 100.1 (0.3) 131.3 (0.1) 189.6 (0.9)
62.2 (0.6) 90.5 (0.8) 115.0 (0.5) 169.2 (0.4)
Data are mean SD, n ¼ 5. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 4, APRIL 2004
Sieved Aeroflo 95 6.9 4.1 2.8 1.6
(0.1) (0.1) (0.0) (0.0)
Decanted
Aeroflo 65 7.7 5.0 3.5 1.5
(1.0) (0.7) (0.7) (0.2)
Aeroflo 95
Aeroflo 65
3.0 (0.1) 1.9 (0.0) 1.5 (0.0) 1.2 (0.1)
2.7 (0.0) 1.9 (0.0) 1.6 (0.0) 1.2 (0.0)
DISPERSION OF SALMETEROL XINAFOATE
1035
cant difference in FPF of SX (p < 0.05), except fraction 45–63 mm, which contained a higher amount of fine lactose (2.7% compared with about 1.5% for the other fractions). When considering decanted fractions with similar concentrations of fine lactose, the FPF of SX was independent of carrier size. Effect of Decantation on Dispersion
contained a higher amount of fine lactose (3% in comparison to about 1.5% for the other fractions), there was no significant difference between FPF of the 63–90, 90–106, and >106 mm fractions (p < 0.05). The outcome of the dispersion testing of the sieve fractions of Aeroflo 65 was similar to that of Aeroflo 95 (Fig. 3). The FPFs of SX of dry sieved and decanted fractions was found to be significantly different (p < 0.001). Good linearity and a strong co-relationship existed between FPF of SX and VMD of dry sieved fractions (r2 ¼ 0.98, p ¼ 0.010). Decanted fractions showed no signifi-
From the data presented in Figures 2 and 3, it was clear that the decanting process had a significant effect on drug dispersion and that the fine lactose particles influenced dispersion in some manner. The data also demonstrated that SX mixtures, using different decanted lactose fractions that contained similar concentrations of fine lactose, produced similar FPFs of SX. This information helped in interpreting the role of the fine lactose in the dispersion process. The theoretical surface area of the lactose fractions increased as the particle size decreased (Fig. 4). Increased surface area of the carrier would effectively increase the number of adhesion sites on the lactose carrier; such a proposition is similar to the strategies used to improve the adsorptive capacity of charcoal. If the assumption is true that the distribution of the strength of interaction of the adhesion sites on the lactose surface was likely to be similar for the different fractions of both carriers because the lactose is from the same source and has undergone similar solvent treatment, then the number of high-energy adhesion
Figure 3. Relationship between FPF of SX and VMD of different size fractions (45–63, 63–90, 90–106, and >106 mm) of dry sieved and decanted fractions of Aeroflo65 determined using the TSI from 2.5% SX interactive mixtures using the Rotahaler1. Regression lines are fitted to the data (mean, n ¼ 5).
Figure 4. Relationship of FPF of SX and calculated specific surface area (SSA) with VMD of decanted fractions of Aeroflo 95. SSA has been calculated from the particle size distribution assuming the lactose particles are spherical in shape and the surface is smooth.
Figure 2. Relationship between FPF of SX and VMD of different size fractions (45–63, 63–90, 90–106, and >106 mm) of dry sieved and decanted fractions of Aeroflo95 determined using the TSI from 2.5% SX interactive mixtures attached to the Rotahaler1. Regression lines were fitted to the data (mean, n ¼ 5).
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 4, APRIL 2004
1036
ISLAM ET AL.
Figure 5. Scanning electron micrographs of lactose particles before and after decantation. (A) Aeroflo 95 before decantation, (B) Aeroflo 95 after decantation, (C) Aeroflo 65 before decantation, and (D) Aeroflo 65 after decantation.
sites on the lactose carrier will increase as the surface area increases. If the role of the fine adhered lactose was to occupy high-energy sites, allowing SX to be less strongly bound to the lowenergy sites, then it would be expected that the FPF of SX on the decanted lactose fractions would decrease as the particle size decreased. The proportion of high-energy sites occupied by the residual fine lactose would be less, leaving a greater concentration of higher energy sites on the lactose surface available for interaction with SX. SX particles would therefore adhere more strongly on the lactose surface minimizing disJOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 4, APRIL 2004
persion. Data in Figures 2 and 3 clearly show that this does not happen. The results of this study again have demonstrated the importance of adhered fine lactose on the carrier surface. The mechanism by which the small lactose particles improve the dispersion of drugs was still not clear; however, the data were not incompatible with the proposed hypothesis that particle redistribution occurred on the carrier surface producing mixed agglomerates or multilayers of fine lactose and drug and that drug removal from the coarse lactose surface occurred via the agglomerates or multilayers.4
DISPERSION OF SALMETEROL XINAFOATE
Some studies have shown that treatment of lactose with 95% ethanol induced small asperities and cavities into the surface causing fine particle dose of albuterol sulfate to decrease.27 Examination of the scanning electron micrographs of the decanted lactose used in this current study demonstrated that little evidence of asperities and cavities resulted from the treatment of the lactose with absolute alcohol presaturated with lactose, but did result in the removal of fine adhered lactose particles from the surface (Fig. 5). However, microscopic indentations and etchings to the surface may not have been visible at the magnification used. In addition, when fine lactose (<5 mm) was added to the decanted unsieved fractions of Aeroflo 95 and 65, the FPF of SX from the resulting mixture was restored to values not significantly different from the original samples of Aeroflo 95 and 65 (p < 0.05). For example, for Aeroflo 95, when fine lactose (<5 mm) was added to the decanted lactose to restore the original content to 3.9%, the FPF of SX from the mixture increased from 5.9 1.2% to 12.3 1.6%, which was not significantly different from 13.8 2.7% for the original sample (ANOVA, p ¼ 0.277). Also, for Aeroflo 65, when fine lactose (<5 mm) was added to the decanted lactose to restore the original content to 6.2%, the FPF of SX from the mixture increased from 4.6 1.0% to 18.9 1.5%, which was not significantly different from 21.2 2.0% for the original sample (ANOVA, p ¼ 0.084). These results demonstrate that the decantation process did not produce any major changes to the surface that influenced dispersion, other than removal of fine adhered lactose.
CONCLUSIONS A decantation method was applied to remove fine lactose from the coarse lactose carrier surface to investigate the effect of carrier size on the dispersion of SX (2.5%) in interactive mixtures with lactose. The major outcome of this research was that inherent particle size of the lactose carrier did not influence the degree of dispersion of SX. This was significant because this finding had not been seen in other studies. The decantation methodology was an innovative approach to remove fine lactose from the surface of the lactose carriers and to control the concentration associated with each of the lactose carriers used in this study. In contrast, the presence of fine lactose adhered on the surface of the lactose carrier
1037
contributed to increased dispersion of SX. The relationship between FPF and VMD seen with the dry sieved fractions was due to the presence of fine lactose on the carrier surface, because smaller sized fractions possessed a greater concentration of fine adhered lactose. This study therefore has identified a formulation variable that does not influence dispersion, at least over the particle range seen in this study. Such an outcome is important in the design of carrier systems. The results of this study that dispersion was not intrinsically influenced by carrier VMD is not incompatible with the hypothesis that adhered lactose on the carrier surface is involved in interaction with the drug through redistribution processes and probably forms multilayers or agglomerates with the drug before dispersion. One limitation of the study was that the conclusions were based on data from laboratory scale powder mixing processes. Because the state of inter-particulate interactions in the powder formulation will depend to some extent on mixing energies, extending the interpretations of this work to other mixing processes should be undertaken cautiously.
ACKNOWLEDGMENTS N.I. was supported by International Postgraduate Research and Monash Graduate Scholarships. Lactose samples were donated by Foremost Farms, USA and Meggle GmbH, Germany. We extend our thanks to Dr. Graham Heys, CSIRO, Clayton, Australia, for providing decanting apparatus and thoughtful suggestion in the decantation process and Dr. Aidan Sudbury, School of Mathematical Science, Monash University, for advice on the statistical analysis.
REFERENCES 1. Newman SP, Clarke SW. 1983. Therapeutic aerosols. 1. Physical and practical considerations. Thorax 38:881–886. 2. Zeng XM, Tee SK, Martin GP, Marriot C. 1996. Effect of mixing procedure and particle size distribution of carrier particles on the deposition of salbutamol sulphate from dry powder inhaler formulations. Proc Drug Deliv Lung (Lond) 7:40–43. 3. Zeng XM, Martin GP, Tee SK, Marriot C. 1998. The role of fine particle lactose on the dispersion and deaggregation of salbutamol sulphate in the air stream in vitro. Int J Pharm 176:99–110. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 4, APRIL 2004
1038
ISLAM ET AL.
4. Louey MD, Stewart PJ. 2002. Particle interactions involved in aerosol dispersion of ternary interactive mixtures. Pharm Res 19(10):1524–1531. 5. Ganderton D, Kassem NM. 1992. Dry powder inhalers. In: Ganderton D, Jones T, editors. Advances in pharmaceutical sciences, Vol 6. London: Academic Press, pp 165–191. 6. Zeng XM, Martin GP, Marriott C, Pritchard J. 2000. The influence of carrier morphology on drug delivery by dry powder inhalers. Int J Pharm 200: 93–106. 7. Staniforth JN. 1996. Improvement in the dry powder inhaler performance: Surface passivation effects. In: Proceedings drug delivery to the lung. VII. London: Aerosol Society, pp 86–89. 8. Zeng XM, Tee SK, Martin GP, Marriot C. 1996. Improving the delivery efficiency of dry powder inhaler DPIs by adding the fine carrier particles to powder formulations. Thorax 51(3):195. 9. Zeng XM, Pandhal KH, Martin GP. 2000. The influence of lactose carrier on the content homogeneity and dispersibility of beclomethasone dipropionate from dry powder aerosols. Int J Pharm 197:41–52. 10. Srichana T, Martin GP, Marriot C. 1998. On the relationship between drug and carrier deposition from dry powder inhalers in vitro. Int J Pharm 167:13–23. 11. Iida K, Leuenberger H, Fueng LM, Rudi MW, Okamoto H, Danzo K. 2000. Effect of mixing of fine carrier particles on dry powder inhalation property of salbutamol sulfate (SS). Yakugaku Zasshi 120(1): 113–119. 12. Louey MD, Razia S, Stewart PJ. 2003. Influence of physico-chemical carrier properties on the in vitro aerosol deposition from interactive mixtures. Int J Pharm 252:87–98. 13. Zeng XM, Martin GP, Tee SK, Ghoush AA, Marriot C. 1999. Effects of particle size and adding sequence of fine lactose on the deposition of salbutamol sulphate from the dry powder formulation. Int J Pharm 182:133–144. 14. Lucas P, Anderson K, Staniforth JN. 1998. Protein deposition from dry powder inhalers: Fine particle multiplets as performance modifiers. Pharm Res 15(4):562–569.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 4, APRIL 2004
15. Bell JH, Hartley PS, Cox JSG. 1971. Dry powder aerosols. I. A new powder inhalation device. J Pharm Sci 60(10):1559–1564. 16. Ganderton D. 1992. The generation of respirable clouds from coarse powder aggregate. J Biopharm Sci 3(1/2):101–105. 17. French DL, Edward DA, Niven RW. 1996. The influence of formulation on emission, deaggregation, and deposition of dry powders for inhalation. J Aerosol Sci 27(5):769–783. 18. Kassem NM, Ho KKL, Ganderton D. 1989. The effect of air flow and carrier size on the characteristics of an inspirable cloud. J Pharm Pharmacol 41:14P. 19. Steckel H, Muller BW. 1997. In vitro evaluation of dry powder inhalers. II. Influence of carrier particle size and concentration on in vitro deposition. Int J Pharm 154:31–37. 20. Byron PR, Jashnani R, Germain S. 1990. Efficiency of aerosolization from dry powder blends of terbutaline sulphate and lactose NF with different particle size distribution. Pharm Res 7:S81. 21. Denis FK, Clifford JR, Philip SBS. 1974. Technical note on an improved cyclosizing technique. Proc Aust Inst Min Metall 251:9–10. 22. Allen T. 1997. Powder sampling and particle size measurement. In: Allen T, editor. Particle size and measurement. 5th ed. London: Chapman and Hall, pp 216–219. 23. Heywood H. 1952. Fundamental principles of subsieve particle size measurement. Symposium on recent developments in mineral dressing. Inst Min Met Lond p 31. 24. Alway B, Sangchantra R, Stewart PJ. 1996. Modelling the dissolution of diazepam in lactose interactive mixtures. Int J Pharm 130:213–224. 25. Liu J, Stewart PJ. 1998. Deaggregation during the dissolution of benzodiazepines in interactive mixtures. J Pharm Sci 87(12):1632–1638. 26. Crooks MJ, Ho R. 1976. Ordered mixing in direct compression of tablets. Powder Technol 14:161– 167. 27. Zeng XM, Martin GP, Marriott C, Pritchard J. 2001. Lactose as a carrier in dry powder formulations: The influence of surface characteristics on drug delivery. J Pharm Sci 90(9):1424–1434.
Department of Pharmaceutics, Victorian College of Pharmacy, Monash University, 381 Royal Parade, Parkville Vic 3052, Australia 2
Division of Molecular Science, CSIRO, Clayton Vic 3168, Australia
Received 18 March 2003; revised 30 June 2003; accepted 27 August 2003
ABSTRACT: The objective of this study was to determine the influence of lactose carrier size on drug dispersion of salmeterol xinafoate (SX) from interactive mixtures. SX dispersion was measured by using the fine particle fractions determined by a twin stage impinger attached to a Rotahaler1. The particle size of the lactose carrier in the SX interactive mixtures was varied using a range of commercial inhalation-grade lactoses. In addition, differing size fractions of individual lactose samples were achieved by dry sieving. The dispersion of SX appeared to increase as the particle size of the lactose carrier decreased for the mixtures prepared from different particle size commercial samples of lactose and from different sieve fractions of the same lactose. Fine particles of lactose (<5 mm) associated with the lactose carrier were removed from the carrier surface by a wet decantation process to produce lactose samples with low but similar concentrations of fine lactose particles. The fine particle fractions of SX in mixtures prepared with the decanted lactose decreased significantly (analysis of variance, p < 0.001) and the degree of dispersion became independent of the volume mean diameter of the carriers (analysis of variance, p < 0.05). The dispersion behavior is therefore associated with the presence of fine adhered particles associated with the carriers and the inherent size of the carrier itself has little influence on dispersion. ß 2004 Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci 93:1030–1038, 2004
Keywords: drug dispersion; lactose carrier size; salmeterol xinafoate; decantation; dry powder inhalation; interactive mixtures
INTRODUCTION Dry powders for inhalation are formulated either as loose agglomerates of micronized drug particles or as carrier-based interactive mixtures with micronized drug particles adhered onto the surface of large lactose carriers. For the carrier-based interactive mixtures, three major factors are of prime concern for desirable therapeutic effects: 1.
Correspondence to: Peter Stewart þ61399039517; Fax: 61399039583; E-mail: [email protected])
(Telephone:
Journal of Pharmaceutical Sciences, Vol. 93, 1030–1038 (2004) ß 2004 Wiley-Liss, Inc. and the American Pharmacists Association
1030
the detachment of drug particles from the large carriers; 2. dispersion of the detached drug particles during aerosolization by an inhalation device; and 3. subsequent deposition in the lower airways of lung. Therefore, any factors influencing these processes will affect the therapeutic activity. For example, the geometry of the drug particle and the respiratory tract will influence dispersion and particles with aerodynamic particle sizes of <5 mm are necessary for effective deep lung deposition.1 However, given that micronized particles are cohesive and are likely to agglomerate, and are adhesive and are likely to adhere to lactose carriers, the extents of detachment and subsequent dispersion of micronized drug
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 4, APRIL 2004
DISPERSION OF SALMETEROL XINAFOATE
particles are key elements in determining the efficiency of the formulation for respiratory delivery. The mechanism by which drugs are dispersed is not fully understood. The magnitude of particle interactions between drugs and between drugs and excipients will be important to the dispersion process. One theory that has been advanced to explain dispersion involves the interaction of fine lactose present in the lactose excipient with the high-energy, active sites on the carrier’s surface causing saturation of these active sites, leaving low-energy, passive sites available for drug adhesion.2,3 The reduced adhesion between drug and carrier particles increased drug detachment. Recently, a hypothesis based on competitive adhesion and redistribution of the drug between the carrier surface and fine excipient to produce mixed agglomerates of drug and fine excipient was proposed to explain improved drug dispersion.4 Greater efficiencies were likely to result through facilitated drug detachment from the carrier surface as agglomerates and better dispersion of mixed agglomerates by the dry powder inhaler. Several innovations to improve the delivery of drugs have been reported. These studies have focused on changing particle and in particular surface morphology, using alternate carriers and utilizing fine lactose or other particles to enhance deposition. Some of these studies are highlighted as follows. Smoothing the carrier surface improved the respirable fraction of albuterol sulfate.5,6 Fine particles of ternary materials added to the interactive mixtures increased the fine particle fraction (FPF) of several drugs. Inclusion of magnesium stearate, L-leucine, and micronized glucose in formulations produced higher in vitro deposition of salbutamol sulfate4,7 and the addition of fine particles of lactose to dry powder formulations improved dispersion and deposition of salbutamol sulfate2,3,8 and beclomethasone dipropionate.9 Using the Twin Stage Impinger (TSI), the FPF of salbutamol sulfate increased with increasing concentration of fine particles of lactose (<10 mm).10 The FPF of salbutamol sulfate increased with increasing the concentration of fine carrier of synthesized sugars.11 Addition of intermediate-sized lactose and micronized glucose also increased the FPF of salbutamol sulfate.3,12 A ternary mixture of salbutamol sulfate, fine and coarse lactose prepared by first blending the coarse and fine lactose before mixing with drug, produced higher FPFs of salbutamol sulfate.13 An increased FPF of spray dried bovine serum albumin occurred with
1031
increasing concentrations of fine lactose (5.4 mm) up to 5%, but no further increase of FPF was observed at higher concentration (7.5–10%).14 Recently, a higher FPF and dispersibility of salbutamol sulfate was found using more elongated crystals of lactose carriers.6 Carrier particle size is important in the design of dry powder inhalers and the effect of carrier size on drug dispersion has been reported.15–19 Greatest dispersion of cromolyn sodium from an interactive mixture with lactose particles (70–100 mm) at a flow rate of 60 L min1 was observed.15 Using binary mixtures of salbutamol sulfate and synthesized sugar, the FPF decreased with increasing carrier particle size.10 A reduction in carrier size improved respirable fraction of albuterol sulfate16,18 and budesonide.19 However, a higher respirable fraction of terbutaline sulfate was obtained from coarser lactose (53–105 mm) than from a finer lactose (<53 mm).20 From the studies above, the particle size distribution of the carrier influenced the respiratory in vitro deposition of drugs. It is not clear from these studies whether the intrinsic carrier size or the proportion of fine lactose particles associated with the carrier influenced dispersion because the proportion of small lactose particles adhered to the carrier surface are likely to be greater as the particle size distribution is decreased. In this study, we have investigated the effect of only carrier size on drug dispersion by removing the fine lactose from the large lactose by sieving and wet decantation. In this way, we hope to determine specific particle size effects on drug dispersion; such determinations are important in understanding the mechanism of dispersion. The study was conducted using salmeterol xinafoate (SX) as the model drug and a range of commercial samples of lactose.
EXPERIMENTAL Materials SX, micronized (inhalation grade; <4 mm), was obtained from GlaxoSmithKline, Australia. Aeroflo 95, Aeroflo 65, and Aeroflo 20, inhalation grades of a-lactose monohydrate, were donated by Foremost Farms, Rothschild, WI; other lactose grades of Inhalac 120, Inhalac 230, and Sorbolac 400 were donated by Meggle GmbH, Germany. Ammonium acetate (Analar, BDH, Victoria), methanol [high-performance liquid JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 4, APRIL 2004
1032
ISLAM ET AL.
chromatography (HPLC) grade, Biolab, Victoria], and absolute alcohol (HPLC grade, CSR, Victoria) were used as supplied. Dry Sieving and Fractionation Fractions (45–63, 63–90, 90–106, and >106 mm) of Aeroflo 95 and 65 were produced by dry sieving for 30 min using a sieve shaker (Fritsch, Germany) and test sieves (Endecotts, UK), and were stored in a desiccator over silica gel. Decantation and Removal of Fine Lactose Fine lactose particles were removed from the unsieved and sieved fractions of lactose carrier (Aeroflo 65 and 95) by a decantation method.21 Approximately 20 g of lactose was dispersed in absolute alcohol, presaturated with lactose for 5 min to make a homogeneous suspension and allowed to settle for a predetermined time. The supernatant liquid was decanted and replaced by fresh saturated ethanol, vigorously mixed, and the decantation process repeated (15 times) until the supernatant liquid was clear. Although 15 repeats produced a clear supernatant,22 Heywood23 estimated that approximately 50 repeats would be necessary to eliminate 99% of particles near the size of separation. Particle Size of Lactose The particle size of the lactose carriers was measured by laser diffraction (Malvern Mastersizer S; Malvern Instruments Ltd., UK) using the 300 RF lens and the small-volume sample presentation unit (capacity 150 mL). Approximately 500 mg of lactose powder was dispersed in 5 mL of butan-1-ol with the aid of a sonication in a waterbath for 3 min. Sonication was necessary because unsonicated samples showed significant agglomeration. The sonicated sample was added drop-wise into the sample cell containing 150 mL of ethanol until an obscuration between 10–30% was obtained. Size measurement of each sample was performed using 2000 sweeps and analyzed with the reference refractive index of lactose (1.533) and ethanol (1.36) and calculated imaginary refractive index of lactose is 0.1. The average particle size distribution was measured from five replicates of each sample. The size distributions were characterized by the 10, 50, and 90 percentiles (d10%, d50%, and d90%, respectively). The volume mean diameter (VMD), which is defined JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 4, APRIL 2004
as Snd4/Snd3, where n is the number of particles and d is the equivalent spherical diameter, was determined from the output of the laser diffraction particle sizing and was used as the size parameter to characterize the lactose carrier distributions. The percentage of particles <5 mm was determined using the cumulative frequency distribution (undersized) curve. The residual value was always <1%. Preparation of Interactive Mixtures and Filling of Capsules The interactive mixtures were prepared by a hand mixing method.24,25 Binary interactive mixtures of SX (2.5%) and different grades of lactoses were prepared in 5-g batches. The drug powders were placed between two layers of carrier powders in a glass test tube with three ceramic beads of approximately 10-mm diameter. After stoppering, the test tube was then vigorously shaken by hand for 5 min, where the ceramic beads provided a ball-milling effect for breaking up the drug agglomerates formed during mixing. The powder formulations were filled (20 mg) into hard gelatin capsules (size 3; Fawns and McAllan Pty Ltd., Australia) manually. Homogeneity Test of Interactive Mixtures The homogeneity of each powder mixture prepared for this experiment was assessed by the mean drug content and variability between samples. Twenty samples, each of approximately 20 mg, were dissolved in 40% methanol (HPLC grade) and the amount of SX was determined by a validated UV assay. An acceptable degree of homogeneity was achieved with a mean drug content within 100 5% (mean SD) of the theoretical value and a coefficient of variation <5% for all samples.26 Drug Dispersion by TSI Using a Rotahaler1 (Glaxo Wellcome), the in vitro aerosol deposition of the powder formulations was determined using a TSI (Apparatus, A; British Pharmacopoea, 2000) (Copley, UK). Seven and thirty milliliters of 60% methanol (HPLC grade) were placed into stage one and stage two of the TSI, respectively. The air flow was drawn through the TSI using a vacuum pump (model OD5/2; Dynavac Engineering, Australia) and the air flow rate was adjusted to 60 L/min at the mouthpiece,
DISPERSION OF SALMETEROL XINAFOATE
before each measurement (MODEL 10A3567SAX; Fisher and Porter, UK). The FPF was defined as the percent of drug particles deposited in the lower stage of the TSI of the recovered dose (RD). The RD was the total amount of drug collected from inhaler, stage one (S1), and stage two (S2). The emitted dose was the percent of drug delivered from the inhaler of RD. HPLC Analysis of SX SX was analyzed by HPLC using a C18 column (mBondapakTM, 3.9 300 mm; Waters) and an UV detector (Waters Tunable Absorbance Detector) at a wavelength of 252 nm. A mixture of methanol and 0.2% (w/v) ammonium acetate solution (55:45, pH 6.9) was used as a mobile phase running at a flow rate of 1.0 mL/min by an HPLC pump (Waters 510). The peak area was recorded by integration (Shimadzu CR6A Chromatopack, Japan). The retention time of SX was 4.2–4.9 min (the variation from batch to batch ranged from 4.2–4.9 min whereas a single batch showed a constant retention time). The calibration plot of standard SX solution was linear over the range of 0.4–10 mg/mL with r2 ¼ 1.00. Five replicates of standards samples of 0.4, 1.0, 4.0, and 10.0 mg/mL solutions were performed for assay validation and the mean accuracy was 91.4 2.4%, 100.4 5.3%, 101.3 1.1%, and 99.8 0.6%, respectively (mean SD). The precision was tested before each experiment by analyzing 4.0 mg/mL standard solution of SX using five replicates and the coefficient of variation was <2%. Statistical Analysis Linear regression analysis (Sigmastat; Jandel Scientific) was used in the calibration of HPLC and UV analyses and testing relationships between FPF and VMD. Comparison between different groups of FPF was performed by using
1033
one way analysis of variance (ANOVA) (Sigmastat; Jandel Scientific). Covariance Analysis by Minitab (Minitab Corporation) was undertaken to compare the slopes of regression lines.
RESULTS AND DISCUSSION Influence of Carriers on Dispersion SX (2.5%) was mixed with six different grades of lactose having different inherent size distributions summarized in Table 1. Interactive mixtures were formed with the drug particles adhering to the surface of the lactose; these mixtures showed excellent homogeneity with means within 100 2.5% of the target and coefficients of variation <3%. The FPF of SX was determined for each of the mixes and increased as the VMD of the lactose carrier decreased (Fig. 1). There is a significant difference between the FPF of SX at the differing VMD of the lactose carriers (ANOVA, p < 0.05). Such findings have been supported by other studies in our laboratory using 5% binary mixtures of salbutamol sulfate and a series of different lactose samples from three different manufacturers.11 The percentage of small particles (e.g., <5 mm) present in each carrier was seen to increase as the particle size distribution of the carrier (characterized by the VMD) decreased (Table 1). The presence of small lactose particles associated with the carriers has been shown in a number of studies to increase dispersion of micronized drug particles from mixtures of drug and lactose.2,3,8–14 Therefore, the results shown in Figure 1 could be associated either with a specific particle size effect of the lactose carriers where decreasing carrier particle size inherently improves SX dispersion or with the effect of increasing concentrations of fine lactose particles present on the lactose carrier surface as the particle size of the lactose carrier decreases. To identify the real cause of this
Table 1. Particle Size Distributions of Different Lactose Carriers Sample Sorbolac 400 Aeroflo 20 Aeroflo 65 Aeroflo 95 Inhalc 230 Inhalc 120
d10% 1.5 3.2 9.9 19.6 69.1 114.2
(0.1) (0.1) (0.2) (0.3) (0.3) (1.7)
d50%
d90%
8.5 (0.1) 22.3 (0.3) 72.8 (1.3) 101.0 (0.8) 110.3 (0.2) 156.5 (1.2)
20.3 (0.2) 48.7 (2.0) 167.9 (2.8) 222.4 (1.6) 156.4 (0.9) 209.4 (2.9)
<5 mm (%) 30.3 14.3 6.2 3.9 1.6 1.2
(0.8) (0.2) (0.0) (0.1) (0.0) (0.1)
VMD (mm) 9.99 (0.9) 24.9 (0.6) 82.5 (1.3) 112.7 (0.9) 110.6 (0.3) 157.6 (2.1)
Data are mean SD, n ¼ 5. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 4, APRIL 2004
1034
ISLAM ET AL.
Figure 1. Relationship between FPF of SX and VMD of six different lactose carriers, determined using the TSI from 2.5% SX interactive mixtures using the Rotahaler1.
behavior, fractions of Aeroflo 65 and 95 were selected to determine the effect of the removal of lactose fines from the surface of these carriers on the dispersion of SX. Classification of Lactose Carrier Fractions Fractions of 45–63, 63–90, 90–106, and >106 mm of Aeroflo 65 and 95 were obtained by sieve classification. The particle size distributions of these carriers were determined by laser diffraction (Table 2). The sieve fractions of Aeroflo 65 and 95 were found to be significantly different from each other (ANOVA, p < 0.001) with the VMD of the sieved fractions of Aeroflow 65 being significantly lower than those of the corresponding fractions of Aeroflo 95 (Tukey test, p < 0.05). This might be due to the presence of greater amounts of the fine lactose in the Aeroflo 65 that adhered to the large lactose carrier surface even after sieving. The sieved fractions of Aeroflo 65 and 95 were subjected to wet decantation to remove the fine lactose associated with these lactose carriers.
Table 2 shows that the VMD increased after decantation because of the removal of fine particles of lactose and that wet decantation decreased the percentage of fine lactose in the sieved fractions of Aeroflo 65 and 95 with the decanted fractions showing about 60–70% removal of fine lactose after 15 repeat wet decantation cycles. After decantation, the Aeroflo 65 and 95 fractions contained similar amounts of fine lactose. It was more difficult to remove adhered fines from the smallest fractions of Aeroflo 65 and 95. The 45–63 mm fraction of Aeroflo 65 and 95 contained 2.8 and 3.0% of particles <5 mm, respectively, whereas all other fractions showed more consistent concentrations of fine particles <5 mm at about 1–2%. Effect of Carrier Size (VMD) on Drug Dispersion The FPFs of SX (2.5%) mixtures with the dry sieved and decanted fractions of Aeroflo 65 and 95 were determined. The FPF of SX increased with decreased VMD of different dry sieved fractions of Aeroflo 95 as expected (Fig. 2). FPFs of SX were significantly different for the sieved fractions (45–63, 63–90, 90–106, and >106 mm) of Aeroflo 95. Pair-wise comparisons found significant differences between 45–63 mm fraction versus the other three fractions, and 63–90 mm fraction versus 90–106 mm fraction and >106 mm fraction, but no difference between 90–106 mm fraction and >106 mm fraction, with p < 0.05 level of significance. The relationship between FPF of SX and VMD of different sieved fractions of Aeroflo 95 showed good linearity and a significant corelationship (r2 ¼ 0.922, p < 0.04). Interestingly, after decantation, the FPFs of SX of all fractions were significantly reduced in comparison with those of the corresponding fractions before decantation (p < 0.001), reinforcing the influence of the fine adhered lactose in the dispersion process. Except for the 45–63 mm fraction, which
Table 2. Particle Size Distributions of Dry Sieved and Decanted Fractions of Aeroflo 95 and Aeroflo 65 VMD (mm) Sieved
% FL (<5 mm) Decanted
Size range
Aeroflo 95
Aeroflo 65
Aeroflo 95
Aeroflo 65
45–63 63–90 90–106 >106
45.9 (0.1) 82.1 (1.2) 120.7 (0.5) 184.4 (0.7)
38.3 (0.7) 70.2 (0.9) 106.3 (0.9) 166.6 (1.6)
62.7 (0.12) 100.1 (0.3) 131.3 (0.1) 189.6 (0.9)
62.2 (0.6) 90.5 (0.8) 115.0 (0.5) 169.2 (0.4)
Data are mean SD, n ¼ 5. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 4, APRIL 2004
Sieved Aeroflo 95 6.9 4.1 2.8 1.6
(0.1) (0.1) (0.0) (0.0)
Decanted
Aeroflo 65 7.7 5.0 3.5 1.5
(1.0) (0.7) (0.7) (0.2)
Aeroflo 95
Aeroflo 65
3.0 (0.1) 1.9 (0.0) 1.5 (0.0) 1.2 (0.1)
2.7 (0.0) 1.9 (0.0) 1.6 (0.0) 1.2 (0.0)
DISPERSION OF SALMETEROL XINAFOATE
1035
cant difference in FPF of SX (p < 0.05), except fraction 45–63 mm, which contained a higher amount of fine lactose (2.7% compared with about 1.5% for the other fractions). When considering decanted fractions with similar concentrations of fine lactose, the FPF of SX was independent of carrier size. Effect of Decantation on Dispersion
contained a higher amount of fine lactose (3% in comparison to about 1.5% for the other fractions), there was no significant difference between FPF of the 63–90, 90–106, and >106 mm fractions (p < 0.05). The outcome of the dispersion testing of the sieve fractions of Aeroflo 65 was similar to that of Aeroflo 95 (Fig. 3). The FPFs of SX of dry sieved and decanted fractions was found to be significantly different (p < 0.001). Good linearity and a strong co-relationship existed between FPF of SX and VMD of dry sieved fractions (r2 ¼ 0.98, p ¼ 0.010). Decanted fractions showed no signifi-
From the data presented in Figures 2 and 3, it was clear that the decanting process had a significant effect on drug dispersion and that the fine lactose particles influenced dispersion in some manner. The data also demonstrated that SX mixtures, using different decanted lactose fractions that contained similar concentrations of fine lactose, produced similar FPFs of SX. This information helped in interpreting the role of the fine lactose in the dispersion process. The theoretical surface area of the lactose fractions increased as the particle size decreased (Fig. 4). Increased surface area of the carrier would effectively increase the number of adhesion sites on the lactose carrier; such a proposition is similar to the strategies used to improve the adsorptive capacity of charcoal. If the assumption is true that the distribution of the strength of interaction of the adhesion sites on the lactose surface was likely to be similar for the different fractions of both carriers because the lactose is from the same source and has undergone similar solvent treatment, then the number of high-energy adhesion
Figure 3. Relationship between FPF of SX and VMD of different size fractions (45–63, 63–90, 90–106, and >106 mm) of dry sieved and decanted fractions of Aeroflo65 determined using the TSI from 2.5% SX interactive mixtures using the Rotahaler1. Regression lines are fitted to the data (mean, n ¼ 5).
Figure 4. Relationship of FPF of SX and calculated specific surface area (SSA) with VMD of decanted fractions of Aeroflo 95. SSA has been calculated from the particle size distribution assuming the lactose particles are spherical in shape and the surface is smooth.
Figure 2. Relationship between FPF of SX and VMD of different size fractions (45–63, 63–90, 90–106, and >106 mm) of dry sieved and decanted fractions of Aeroflo95 determined using the TSI from 2.5% SX interactive mixtures attached to the Rotahaler1. Regression lines were fitted to the data (mean, n ¼ 5).
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 4, APRIL 2004
1036
ISLAM ET AL.
Figure 5. Scanning electron micrographs of lactose particles before and after decantation. (A) Aeroflo 95 before decantation, (B) Aeroflo 95 after decantation, (C) Aeroflo 65 before decantation, and (D) Aeroflo 65 after decantation.
sites on the lactose carrier will increase as the surface area increases. If the role of the fine adhered lactose was to occupy high-energy sites, allowing SX to be less strongly bound to the lowenergy sites, then it would be expected that the FPF of SX on the decanted lactose fractions would decrease as the particle size decreased. The proportion of high-energy sites occupied by the residual fine lactose would be less, leaving a greater concentration of higher energy sites on the lactose surface available for interaction with SX. SX particles would therefore adhere more strongly on the lactose surface minimizing disJOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 4, APRIL 2004
persion. Data in Figures 2 and 3 clearly show that this does not happen. The results of this study again have demonstrated the importance of adhered fine lactose on the carrier surface. The mechanism by which the small lactose particles improve the dispersion of drugs was still not clear; however, the data were not incompatible with the proposed hypothesis that particle redistribution occurred on the carrier surface producing mixed agglomerates or multilayers of fine lactose and drug and that drug removal from the coarse lactose surface occurred via the agglomerates or multilayers.4
DISPERSION OF SALMETEROL XINAFOATE
Some studies have shown that treatment of lactose with 95% ethanol induced small asperities and cavities into the surface causing fine particle dose of albuterol sulfate to decrease.27 Examination of the scanning electron micrographs of the decanted lactose used in this current study demonstrated that little evidence of asperities and cavities resulted from the treatment of the lactose with absolute alcohol presaturated with lactose, but did result in the removal of fine adhered lactose particles from the surface (Fig. 5). However, microscopic indentations and etchings to the surface may not have been visible at the magnification used. In addition, when fine lactose (<5 mm) was added to the decanted unsieved fractions of Aeroflo 95 and 65, the FPF of SX from the resulting mixture was restored to values not significantly different from the original samples of Aeroflo 95 and 65 (p < 0.05). For example, for Aeroflo 95, when fine lactose (<5 mm) was added to the decanted lactose to restore the original content to 3.9%, the FPF of SX from the mixture increased from 5.9 1.2% to 12.3 1.6%, which was not significantly different from 13.8 2.7% for the original sample (ANOVA, p ¼ 0.277). Also, for Aeroflo 65, when fine lactose (<5 mm) was added to the decanted lactose to restore the original content to 6.2%, the FPF of SX from the mixture increased from 4.6 1.0% to 18.9 1.5%, which was not significantly different from 21.2 2.0% for the original sample (ANOVA, p ¼ 0.084). These results demonstrate that the decantation process did not produce any major changes to the surface that influenced dispersion, other than removal of fine adhered lactose.
CONCLUSIONS A decantation method was applied to remove fine lactose from the coarse lactose carrier surface to investigate the effect of carrier size on the dispersion of SX (2.5%) in interactive mixtures with lactose. The major outcome of this research was that inherent particle size of the lactose carrier did not influence the degree of dispersion of SX. This was significant because this finding had not been seen in other studies. The decantation methodology was an innovative approach to remove fine lactose from the surface of the lactose carriers and to control the concentration associated with each of the lactose carriers used in this study. In contrast, the presence of fine lactose adhered on the surface of the lactose carrier
1037
contributed to increased dispersion of SX. The relationship between FPF and VMD seen with the dry sieved fractions was due to the presence of fine lactose on the carrier surface, because smaller sized fractions possessed a greater concentration of fine adhered lactose. This study therefore has identified a formulation variable that does not influence dispersion, at least over the particle range seen in this study. Such an outcome is important in the design of carrier systems. The results of this study that dispersion was not intrinsically influenced by carrier VMD is not incompatible with the hypothesis that adhered lactose on the carrier surface is involved in interaction with the drug through redistribution processes and probably forms multilayers or agglomerates with the drug before dispersion. One limitation of the study was that the conclusions were based on data from laboratory scale powder mixing processes. Because the state of inter-particulate interactions in the powder formulation will depend to some extent on mixing energies, extending the interpretations of this work to other mixing processes should be undertaken cautiously.
ACKNOWLEDGMENTS N.I. was supported by International Postgraduate Research and Monash Graduate Scholarships. Lactose samples were donated by Foremost Farms, USA and Meggle GmbH, Germany. We extend our thanks to Dr. Graham Heys, CSIRO, Clayton, Australia, for providing decanting apparatus and thoughtful suggestion in the decantation process and Dr. Aidan Sudbury, School of Mathematical Science, Monash University, for advice on the statistical analysis.
REFERENCES 1. Newman SP, Clarke SW. 1983. Therapeutic aerosols. 1. Physical and practical considerations. Thorax 38:881–886. 2. Zeng XM, Tee SK, Martin GP, Marriot C. 1996. Effect of mixing procedure and particle size distribution of carrier particles on the deposition of salbutamol sulphate from dry powder inhaler formulations. Proc Drug Deliv Lung (Lond) 7:40–43. 3. Zeng XM, Martin GP, Tee SK, Marriot C. 1998. The role of fine particle lactose on the dispersion and deaggregation of salbutamol sulphate in the air stream in vitro. Int J Pharm 176:99–110. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 4, APRIL 2004
1038
ISLAM ET AL.
4. Louey MD, Stewart PJ. 2002. Particle interactions involved in aerosol dispersion of ternary interactive mixtures. Pharm Res 19(10):1524–1531. 5. Ganderton D, Kassem NM. 1992. Dry powder inhalers. In: Ganderton D, Jones T, editors. Advances in pharmaceutical sciences, Vol 6. London: Academic Press, pp 165–191. 6. Zeng XM, Martin GP, Marriott C, Pritchard J. 2000. The influence of carrier morphology on drug delivery by dry powder inhalers. Int J Pharm 200: 93–106. 7. Staniforth JN. 1996. Improvement in the dry powder inhaler performance: Surface passivation effects. In: Proceedings drug delivery to the lung. VII. London: Aerosol Society, pp 86–89. 8. Zeng XM, Tee SK, Martin GP, Marriot C. 1996. Improving the delivery efficiency of dry powder inhaler DPIs by adding the fine carrier particles to powder formulations. Thorax 51(3):195. 9. Zeng XM, Pandhal KH, Martin GP. 2000. The influence of lactose carrier on the content homogeneity and dispersibility of beclomethasone dipropionate from dry powder aerosols. Int J Pharm 197:41–52. 10. Srichana T, Martin GP, Marriot C. 1998. On the relationship between drug and carrier deposition from dry powder inhalers in vitro. Int J Pharm 167:13–23. 11. Iida K, Leuenberger H, Fueng LM, Rudi MW, Okamoto H, Danzo K. 2000. Effect of mixing of fine carrier particles on dry powder inhalation property of salbutamol sulfate (SS). Yakugaku Zasshi 120(1): 113–119. 12. Louey MD, Razia S, Stewart PJ. 2003. Influence of physico-chemical carrier properties on the in vitro aerosol deposition from interactive mixtures. Int J Pharm 252:87–98. 13. Zeng XM, Martin GP, Tee SK, Ghoush AA, Marriot C. 1999. Effects of particle size and adding sequence of fine lactose on the deposition of salbutamol sulphate from the dry powder formulation. Int J Pharm 182:133–144. 14. Lucas P, Anderson K, Staniforth JN. 1998. Protein deposition from dry powder inhalers: Fine particle multiplets as performance modifiers. Pharm Res 15(4):562–569.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 4, APRIL 2004
15. Bell JH, Hartley PS, Cox JSG. 1971. Dry powder aerosols. I. A new powder inhalation device. J Pharm Sci 60(10):1559–1564. 16. Ganderton D. 1992. The generation of respirable clouds from coarse powder aggregate. J Biopharm Sci 3(1/2):101–105. 17. French DL, Edward DA, Niven RW. 1996. The influence of formulation on emission, deaggregation, and deposition of dry powders for inhalation. J Aerosol Sci 27(5):769–783. 18. Kassem NM, Ho KKL, Ganderton D. 1989. The effect of air flow and carrier size on the characteristics of an inspirable cloud. J Pharm Pharmacol 41:14P. 19. Steckel H, Muller BW. 1997. In vitro evaluation of dry powder inhalers. II. Influence of carrier particle size and concentration on in vitro deposition. Int J Pharm 154:31–37. 20. Byron PR, Jashnani R, Germain S. 1990. Efficiency of aerosolization from dry powder blends of terbutaline sulphate and lactose NF with different particle size distribution. Pharm Res 7:S81. 21. Denis FK, Clifford JR, Philip SBS. 1974. Technical note on an improved cyclosizing technique. Proc Aust Inst Min Metall 251:9–10. 22. Allen T. 1997. Powder sampling and particle size measurement. In: Allen T, editor. Particle size and measurement. 5th ed. London: Chapman and Hall, pp 216–219. 23. Heywood H. 1952. Fundamental principles of subsieve particle size measurement. Symposium on recent developments in mineral dressing. Inst Min Met Lond p 31. 24. Alway B, Sangchantra R, Stewart PJ. 1996. Modelling the dissolution of diazepam in lactose interactive mixtures. Int J Pharm 130:213–224. 25. Liu J, Stewart PJ. 1998. Deaggregation during the dissolution of benzodiazepines in interactive mixtures. J Pharm Sci 87(12):1632–1638. 26. Crooks MJ, Ho R. 1976. Ordered mixing in direct compression of tablets. Powder Technol 14:161– 167. 27. Zeng XM, Martin GP, Marriott C, Pritchard J. 2001. Lactose as a carrier in dry powder formulations: The influence of surface characteristics on drug delivery. J Pharm Sci 90(9):1424–1434.