Engineering gas-foamed large porous particles for efficient local delivery of macromolecules to the lung

European Journal of Pharmaceutical Sciences 41 (2010) 60–70

Contents lists available at ScienceDirect

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

Engineering gas-foamed large porous particles for efficient local delivery of macromolecules to the lung Francesca Ungaro a , Concetta Giovino a , Ciro Coletta b , Raffaella Sorrentino b , Agnese Miro a , Fabiana Quaglia a,∗ a b

Department of Pharmaceutical and Toxicological Chemistry, University of Naples Federico II, Via D. Montesano 49, 80131 Naples, Italy Department of Experimental Pharmacology, University of Naples Federico II, Via D. Montesano 49, 80131 Naples, Italy

a r t i c l e

i n f o

Article history: Received 5 March 2010 Received in revised form 11 May 2010 Accepted 19 May 2010 Available online 25 May 2010 Keywords: Pulmonary delivery Large porous particles PLGA Phospholipids Macromolecules

a b s t r a c t Gas-foamed large porous particles (gfLPP) based on poly(lactic-co-glycolic) acid (PLGA) have been recently suggested as potential carriers for pulmonary drug delivery. In this work, we attempt to engineer gfLPP for efficient local delivery of macromolecules in the lungs. Particles were fabricated by the double emulsion-solvent evaporation technique using ammonium bicarbonate as porogen. To improve particle technological properties, two lipid aid excipients, namely dipalmitoylphosphatidylcholine (DPPC) and 1,2-dioleoyl-3-trimethylammonium-propane (DOTAP), were tested. Preliminary technological studies performed on unloaded gfLPP showed that the addition of an appropriate amount of NH4 (HCO3 ), which spontaneously produces CO2 and NH3 during solvent evaporation, is essential to achieve a homogeneous population of highly porous particles with optimal aerodynamic properties. Then, the effect of the presence of DPPC or DOTAP upon the properties of gfLPP containing a model hydrophilic macromolecule, rhodamine B isothiocyanate–dextran (Rhod-dex), was assessed. We found that in the case of hydrophilic macromolecules unable to interact with PLGA end-groups, such as Rhod-dex, excipient addition is essential to increase the amount of drug entrapped within gfLPP, being as high as 80% only for DPPCor DOTAP-engineered gfLPP. Also Rhod-dex release profile from gfLPP was strongly affected by excipient addition in the initial formulation, with lipid-engineered gfLPP allowing for a more prolonged release of Rhod-dex as compared to excipient-free gfLPP. A further modulation of Rhod-dex initial release rate could be achieved when DOTAP was used, likely due to the electrostatic interactions occurring between macromolecule and cationic phospholipid. Conceiving the developed gfLPP for drug inhalation, DPPCand DOTAP-engineered gfLPP displayed optimal MMADexp values falling within the range 6.1–7.6 ␮m and very low geometric standard deviations (GSD) varying between 1.2 and 1.3. In vivo deposition studies performed after intra-tracheal administration of gfLPP in rats confirmed the ability of the developed dry powders to deposit along bronchia and bronchioles. In perspective, lipid-engineered gfLPP represent a viable alternative to LPP developed so far to achieve local and prolonged release of hydrophilic macromolecules, such as nucleic acids, in the lungs. © 2010 Elsevier B.V. All rights reserved.

1. Introduction In recent years, the inhalation of macromoleculecular drugs has generated tremendous interest as an alternative and more convenient way for both systemic non-invasive delivery of protein therapeutics (Agu et al., 2001; Laube, 2005; Kumar et al., 2006; Patton and Byron, 2007) and local treatment of chronic lung diseases (e.g., COPD, cystic fibrosis) via nucleic acids (Densmore, 2006; Moschos et al., 2008; Séguin and Ferrari, 2009). To this end, the use

∗ Corresponding author. Tel.: +39 81678707; fax: +39 81678707. E-mail address: [email protected] (F. Quaglia). 0928-0987/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.ejps.2010.05.011

of carrier-based strategies has been regarded as a useful mean to improve drug therapeutic index by increasing not only the amount of macromolecule reaching the target, but also its stability (Cryan, 2005; Mohamed and van der Walle, 2008). In particular, biodegradable large porous particles (LPP) made of poly(lactic-co-glycolic) acid (PLGA) hold great promise for the sustained delivery of macromolecules in the lungs (Edwards et al., 1998; Ungaro et al., 2006, 2009). Typically, LPP display an aerodynamic diameter much lower than geometric one, respectively facilitating their deep lung deposition and reducing macrophage-mediated escape (Edwards et al., 1998). So far, PLGA-based LPP have been adequately engineered into aerosols meeting several important criteria for systemic delivery of

F. Ungaro et al. / European Journal of Pharmaceutical Sciences 41 (2010) 60–70

macromolecules to the lungs, that are: (i) dry powder formulations, which are currently regarded as the choice for macromolecule delivery to the lungs (Van Campen and Venthoye, 2002; Cryan, 2005; Patton and Byron, 2007); (ii) mass mean aerodynamic diameters suitable for deposition in deep lung (Ungaro et al., 2006, 2009); (iii) controlled release of the macromolecule in its bioactive form (Edwards et al., 1998; Ungaro et al., 2009). In so doing, engineered PLGA-based LPP have been shown to sustain deslorelin delivery via the deep lungs (Koushik et al., 2004), allow efficient pulmonary adsorption of low-molecular weight heparin (Rawat et al., 2008) and increase insulin systemic availability (Edwards et al., 1998; Ungaro et al., 2009). Nonetheless, much less is known on the potential of PLGA-based LPP for local delivery of macromolecules, such as nucleic acids, in the lungs, with particular regard to those technological factors allowing optimization of LPP aerodynamic and release properties to exert a local effect. Actually, local inhalation therapies can benefit from LPP deposition over upper lung (i.e., along bronchia and bronchioles) rather than reach absorption sites (i.e., alveoli) (Patton and Byron, 2007; Kleinstreuer et al., 2008). On the other hand, slow release can prolong drug residence time in the airways, sustaining its pharmacological effect meanwhile diminishing the rate of drug appearance in the bloodstream (i.e., unspecific distribution to non-target tissues). From a technological standpoint, PLGA-based LPP for inhalation can be typically obtained by the double emulsion-solvent evaporation technique generating a difference in the osmotic pressure between the internal and the external aqueous phases of the emulsion (Ungaro et al., 2006, 2009; Know et al., 2007; Rawat et al., 2008). This can be achieved adding osmogens in the internal aqueous phase, thus causing water influx from the external to the internal aqueous phase during solvent evaporation (i.e., particle hardening) (De Rosa et al., 2008). In this sense, to control PLGA particle porosity and, consequently, flow and aerosolization properties of the developed dry powders, hydroxypropyl-␤-cyclodextrin (HP␤CD) can be used as osmogen (Ungaro et al., 2006, 2009). However, a severe limitation of the osmogen-based approach is represented by the poor control of drug encapsulation efficiency, ascribable to mass exchanges and consequent drug loss occurring between the two phases during particle hardening. This phenomenon, as well as rapid drug release due to the macroporous structure of the system, can be particularly dramatic in the case of highly hydrophilic macromolecules, such as nucleic acids (De Rosa et al., 2003). An alternative formulation strategy to achieve PLGA-based LPP relies on the use of an effervescent agent, namely ammonium bicarbonate, which decomposes into ammonia and carbon dioxide during emulsification, forming a porous matrix as the gas products escape (i.e., gas-foamed open porous microparticles) (Kim et al., 2006; Yang et al., 2009). Since pore formation depends on effervescence rather than on diffusional mass exchanges between aqueous phases, this technique has been demonstrated to allow efficient encapsulation of a model macromolecule, namely lysozyme, in highly porous PLGA particles (Yang et al., 2009). Nonetheless, poor control over release properties was anyway achieved without further modification of formulation conditions. The aim of this work was to develop PLGA-based gas-foamed LPP (gfLPP) for local and prolonged release of hydrophilic macromolecules in the lungs. To this end, gfLPP made of PLGA were fabricated by the double emulsion-solvent evaporation technique using ammonium bicarbonate as porogen. To control gfLPP release properties, two lipid aid excipients were tested. Our first choice was dipalmitoyl phosphatidylcholine (DPPC), the major component of human lung surfactant, gaining increasing research interest in the development of respirable dry powders for a number of reasons, including its biocompatibility (Evora et al., 1998; Tsapis et al.,

61

2002; Newhouse et al., 2003; Minne et al., 2008; Kaye et al., 2009). As an alternative lipid excipient, we investigated the potential of 1,2-dioleoyl-3-trimethylammonium-propane (DOTAP), a cationic lipid extensively studied as transfection agent for nucleic acids (Simberg et al., 2004). Even if first attempts have been made to deliver drugs to the lung via DOTAP-based formulations (Peek et al., 2008; Garbuzenko et al., 2009), the potential of the cationic additive in the development of PLGA-based LPP for the delivery of anionic macromolecules, such as nucleic acids, in the lungs has not been investigated yet. In the light of these observations, a deep technological study was carried out to assess formulation parameters controlling the properties of lipid-engineered gfLPP, with particular regard to their aerodynamic behavior and release features. For encapsulation/release studies, we chose rhodamine B isothiocyanate–dextran (Rhod-dex) as a model macromolecule, due to its molecular weight (10 kDa) and potential interactions with the positively charged surfactant DOTAP. Optimized formulations were tested in vivo for their potential to deposit over bronchia/bronchioles to exert a local effect. 2. Materials and methods 2.1. Chemicals Poly(d,l-lactide-co-glycolide) (50:50) (PLGA) (Resomer RG 504 H; Mw 41.9 kDa; inherent viscosity 0.5 dL/g) was purchased from Boehringer Ingelheim (Germany). Rhodamine B isothiocyanate–dextran (Mw 10 kDa, Rhod-dex), polysorbate 80, polyvinylalcohol (PVA, Mowiol® 40-88), sodium azide, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC), and ammonium hydrogen carbonate were obtained from Sigma–Aldrich (Italy). 1-Oleoyl-2-[6-[(7-nitro-2-1,3-benzoxadiazol-4-yl) amino] hexanoyl]-3-trimethylammonium propane (chloride salt) (DOTAP) was purchased from Lipoid (Germany). Analytical grade sodium chloride, potassium chloride, sodium phosphate dibasic anhydrous, sodium bicarbonate, methylene chloride, were supplied by Carlo Erba (Italy). 2.2. Preparation of gas-foamed large porous particles (gfLPP) PLGA-based gfLPP were prepared by a modified double emulsion-solvent evaporation technique. Briefly, 0.25 mL of water containing ammonium bicarbonate as porogen (5, 10, and 20%, w/v) were poured into 2.5 mL of methylene chloride containing different amounts of PLGA (15 or 20%, w/v) in the presence of DPPC or DOTAP (0.1%, w/v). The primary emulsion (w1 /o) was generated by a high-speed homogenizer (model Ystral equipped with a tool 6G, Heidolph, Germany) operating at 15 000 rpm for 3 min. Afterwards, the emulsion was added to 25 mL of 1% (w/v) aqueous PVA (Mowiol® 40-88) solution and homogenized at 11 000 rpm (tool 10F) for 2 min to produce the multiple emulsion (w1 /o/w2 ). Solvent evaporation and subsequent particle hardening was achieved under magnetic stirring (MR 3001K, Heidolph, Germany) at room temperature. After 3 h, particles were collected, washed three times with distilled water by centrifugation (Hettich Zentrifugen, Universal 16R) and freezed in liquid nitrogen. Samples were then dried for 36 h by a Modulyo freeze-drier (Edwards, UK) operating at 0.01 atm and −60 ◦ C. gfLPP achieved using ammonium bicarbonate in w1 without phospholipids in o as well as particles containing DPPC or DOTAP without ammonium bicarbonate in w1 were produced for comparison. Composition of unloaded gfLPP is reported in Table 1. gfLPP encapsulating Rhod-dex at the theoretical loading of 0.3% (0.3 mg of Rhod-dex per 100 mg of gfLPP) were also achieved by adding the model macromolecule to ammonium bicarbonate-

62

F. Ungaro et al. / European Journal of Pharmaceutical Sciences 41 (2010) 60–70

Table 1 Composition of unloaded and Rhod-dex loaded gfLPP. Formulationa

DPPC20 DOTAP20 DPPC20SC e DPPC20AB5 DPPC20AB10 DPPC20AB20 DPPC15AB10 DOTAP15AB10 DOTAP20AB10 Rhod-15 Rhod-DPPC15 Rhod-DPPC20 Rhod-DOTAP15 Rhod-DOTAP20

PLGA

Ammonium bicarbonateb

DPPCc

DOTAPc

Rhod-dex theoretical loading

(%, w/v)

(%, w/v)

(%, w/v)

(%, w/v)

(%)d

20 20 20 20 20 20 15 15 20 15 15 20 15 20

– – – 5 10 20 10 10 10 10 10 10 10 10

0.1 – 0.1 0.1 0.1 0.1 0.1 – – – 0.1 0.1 – –

– 0.1 – – – – 0.1 0.1 – – – – 0.1 0.1

– – – – – – – – – 0.3 0.3 0.3 0.3 0.3

a The acronyms used are composed by character strings indicating: presence of Rhod-dex (Rhod), presence and type of lipidic excipient added in o (DPPC or DOTAP), PLGA concentration in o (15 or 20%, w/v), presence and concentration of ammonium bicarbonate (AB) in w1 (5–20%, w/v). The last formulation variable was inferred for loaded gfLPP, which were all prepared at 10% AB in w1 . b NH4 (HCO3 ) was added in the internal water phase of the double emulsion. c DPPC or DOTAP were added in the organic phase of the double emulsion. d Milligrams of Rhod-dex per 100 mg of gfLPP. e Control particles prepared using sodium chloride as porogen agent in the internal water phase (9%, w/v).

Table 2 Overall properties of unloaded gfLPP. Formulation

Volume mean diameter (d)a (␮m ± SDb )

DPPC15AB10 DPPC20AB10 DOTAP15AB10 DOTAP20AB10

31.2 23.8 32.5 36.9

a b c d

± ± ± ±

MMADt c (␮m ± SDb )

Tapped density () (g/mL ± SDb )

1.5 2.2 6.0 2.0

0.044 0.038 0.052 0.064

± ± ± ±

0.0012 0.0048 0.015 0.0033

6.8 4.6 7.3 9.5

± ± ± ±

0.7 0.1 0.3 0.3

Carr’s indexd 6.2 5.5 5.2 13

± ± ± ±

1.3 1.9 1.7 0.6

Mean geometric diameter as determined by laser diffraction. Standard deviation of values calculated on three different batches. Mass mean aerodynamic diameter estimated on the basis of Eq. (2). Powder flowability estimated on the basis of Eq. (1).

containing w1 (10%, w/v) in the presence or not of DPPC/DOTAP in o (0.1%, w/v) (Table 2). 2.3. Characterization of gfLPP Particle shape and morphology were analyzed by scanning electron microscopy (SEM) (Leica S440, Germany). The samples were stuck on a metal stub and coated with gold under vacuum for 90–120 s. The mean geometric diameter and size distribution of the particles were determined by laser light scattering (Coulter LS 100Q, USA) on a dispersion of freeze-dried particles in 0.2% (w/v) aqueous PVA. Particle size is expressed as volume mean diameter ± SD of values collected from three different batches. Powder density was estimated by tapped density measurements according to Ph. Eur. VI Ed. A known weight of particles (100 mg) was transferred to a 10 (±0.05) mL graduated cylinder and the initial volume recorded. The cylinder was then mechanically tapped 1250 times up to volume plateau, by mean of a tapped density tester (Mod. IG/4, Giuliani, Italy). Tapped density of particles () was expressed as the ratio between sample weight (g) and the volume occupied after 1250 tappings (mL). To achieve information about gfLPP flow properties, the compressibility index or Carr’s index was estimated through the relative percent difference between bulk and tapped density as stated by US Pharmacopoeia: Carr’s index =

1 −   i



× 100

(1)

where  and i are tapped and bulk density of the powder, respectively. On the basis of Carr’s index value, powder flowability is defined as: 5–12%, excellent; 12–18%, good; 18–21%, fair; 21–25%, poor, fluid; 25–32%, poor, cohesive; 32–38%, very poor; >40%, extremely poor. The theoretical mass mean aerodynamic diameter (MMADt ) of the particles was also estimated on the basis of the definition: MMADt = d

  1/2 0 X

(2)

where d is the geometric mean diameter, 0 is a reference density of 1 g/mL and X is the dynamic shape factor, which is 1 for a sphere. In the case of porous particles of approximately spherical shape:  ≈ s (1 − ε)

(3)

where s is the skeletal mass density of the particle as measured by pycnometry, ε is the particle porosity. An approximate bulk measure of  as defined by Eq. (3) is provided by tapped density (Vanbever et al., 1999). Loaded gfLPP were further characterized for Rhod-dex distribution inside microspheres and actual loading. Rhod-dex distribution inside microspheres was investigated by CLSM analysis carried out on a LSM 510 Zeiss confocal inverted microscope equipped with a Zeiss 63x/1.25 oil objective lens (Carl Zeiss, Germany). An argon laser (excitation = 541 nm; emission = 572 nm) was used. The amount of Rhod-dex actually encapsulated within gfLPP was determined by a solvent extraction method. Briefly, 5 mg of microspheres were dissolved into 1 mL of methylene chloride and

F. Ungaro et al. / European Journal of Pharmaceutical Sciences 41 (2010) 60–70

Rhod-dex was extracted into 1 mL of water. The suspension was centrifuged (5000 rpm, room temperature, 15 min) and the supernatant analyzed for Rhod-dex content. Rhod-dex in solution was quantified by spectrofluorimetric analysis using a Shimadzu R1501 spectrometer (Shimadzu, Japan) fitted out with a 0.1-cm quartz cell (Hellma® Italia, Italy). The emission of the fluorescent label, after excitation at 556 nm, was measured at 575 nm. The linearity of the response was verified over the concentration range 0.1–2 ␮g/mL (r2 > 0.99). The limit of detection (LOD) (estimated as 3 times the background noise) was 4.16 ␮g/L. The limit of quantitation (QOD) (estimated as 10 times the background noise) was 13.9 ␮g/L. Rhoddex actual loading was calculated as mg of Rhod-dex encapsulated per 100 mg of microspheres. Results are expressed as encapsulation efficiency (ratio of actual and theoretical Rhod-dex loading × 100). 2.4. In vitro release studies In vitro release of Rhod-dex from gfLPP was monitored by membrane dialysis in phosphate buffer (120 mM NaCl, 2.7 mM KCl, 10 mM phosphate salts) (PBS) at 37 ◦ C containing 0.05% (w/v) sodium azide as preserving agent. Prior to use, pH was adjusted to 7.2 with 0.01 M HCl. A known amount of Rhod-dex loaded gfLPP (5 mg) was suspended in 0.35 mL of PBS and placed in a dialysis membrane bag (MWCO: 50 000 Da, Spectra/Por® ). The sample was dropped into 5 mL of PBS (sink condition) and kept at 37 ◦ C. At scheduled time intervals, 1 mL of external medium was withdrawn and replaced by the same amount of fresh PBS. The withdrawn medium was analyzed for Rhod-dex content by spectrofluorimetric analysis as described above. Results are expressed as percent of Rhod-dex released versus time. 2.5. Aerosolization properties of gfLPP The aerosolization properties of Rhod-dex loaded gfLPP were tested in vitro after delivery from Turbospin® , a breath-activated, reusable DPI working with single unit capsule containing the dry powder (PH&T Pharma, Milano, Italy). For each test, a hard gelatin capsule (size 2, Capsugel) was filled with about 20 mg of the powder and placed in the Turbospin® . The pulmonary deposition of Rhod-dex loaded gfLPP was investigated in vitro using a Astra Draco Multi-Stage Liquid Impinger (MSLI), Type ALI 1000 (Erweka, Italy). To determine the amount of Rhod-dex deposited into the impactor, the same extraction method of Rhod-dex from microparticles was used. In particular, 10 mL of water and 10 mL of methylene chloride were poured into each of the four stages of the impinger to wet the collection surfaces. The capsule was then pierced and the liberated powder drawn through the impactor operating at 60 L/min for 4 s using an electronic digital flow meter (DFM mode). This allowed the aspiration of 4 L of air through the apparatus as recommended by Ph.Eur. VI Ed. The powder deposited on the four MSLI stages was recovered by agitating the apparatus, removing the initial solution and rinsing with additional fractions of water/methylene chloride mixture (1:1, v/v). The powder deposited in the induction port and on the final collection site (i.e., filter) was also recovered by washing with an aqueous solution. In each case, the suspension achieved was centrifuged (5000 rpm, 15 min) and the supernatant analyzed for Rhod-dex content by spectrofluorimetric analysis as previously described. Control experiments were run to verify the ability of the method in recovering all the powder collected from the stage. A known amount of Rhod-dex loaded microparticles (5 mg) was suspended in 20 mL of water/methylene chloride mixture (1:1, v/v), transferred into the stage 1 of MSLI and treated as described above. The experiment was run in triplicate and the recovery was found to be 67.6 ± 9.9%. Thus, all data were corrected for recovery.

63

The emitted dose (ED) was calculated by accurately weighing the capsule before and after Turbospin® actuation. Results are expressed as percentage of powder actually delivered ± SD of values collected from three different batches. The fine particle fraction (FPF), the experimental mass median aerodynamic diameter (MMADexp) and the geometric standard deviation (GSD) were calculated according to Ph.Eur. VI Ed. deriving a plot of cumulative mass of powder detained in each stage (expressed as percent of total mass recovered in the impactor) versus cut-off diameter of the respective stage. The cut-off diameter of each individual stage (D) was determined as D = D60

 60 1/2 Q

(4)

where D60 is the cut-off diameter at a flow rate of 60 L/min (i.e., 13.0, 6.8, 3.1 and 1.7 ␮m for stages 1–4, respectively) and Q is the flow rate employed in the test. The FPF was calculated by interpolation from the plot as the percentage of powder emitted from the inhaler with an aerodynamic diameter less than 5 ␮m. The MMAD of the particles was determined from the same graph as the particle size at which the line crosses the 50% mark and the GSD was defined as GSD =

 Size X 1/2 Size Y

(5)

where size X is the particle size for which the line crosses the 84% mark and size Y the 16% mark. 2.6. Data analysis Experimental data are reported as mean ± standard deviation (SD) of values collected from three different batches (n = 3–6). Statistical analysis was performed using KaleidaGraph (Synergy Software, Perkiomen Avenue, USA) by one way analysis of variance (ANOVA) followed by Bonferroni’s post test. P-values less than 0.05 were considered as significant. 2.7. Animals Male Wistar rats (Charles River, Italy) were used for ex vivo fluorescence studies. Animals were housed in an environment with controlled temperature (21–24 ◦ C) and lighting (12:12 h light–darkness cycle). Standard chow and drinking water were provided ad libitum. A period of 7 days was allowed for acclimatisation of rats before any experimental manipulation was undertaken. Animals use was in accordance with the guidelines of Italian (N. 116/1992) and European Council Law (N. 86/609/CEE) for animal care. The experimental procedures were approved by the Animal Ethics Committee of the University of Naples Federico II (Italy). 2.8. In vivo deposition studies In vivo deposition studies of Rhod-dex loaded gfLPP were performed on male Wistar rats. Rats were anesthetized by an intraperitoneal injection of urethane (1 mg/kg). The trachea was exposed, and an endotracheal tube was inserted through the mouth and about 3 mg of gfLPP were intra-tracheally delivered using the low-scale DPI Model DP-4 from Penn-Century (USA). Immediately after delivery, the rat abdominal cavity was incised and a catheter, connected to an infusion pump (Harvard pump type 22, WatsonMarlow), was inserted in the posterior vena cava. Both carotids arteries were severed and lung vasculature was then treated by the infusion of two different solutions trough the vena cava. The first infusion (10 mL/min for 5 min) was performed by phosphate buffer at pH 7.4 (138 mM NaCl, 8.1 mM Na2 HPO4 , 1.1 mM KH2 PO4 , 2.7 mM KCl), whereas the second infusion (5 mL/min for 5 min) was

64

F. Ungaro et al. / European Journal of Pharmaceutical Sciences 41 (2010) 60–70

Fig. 1. SEM micrographs of DPPC-containing unloaded gfLPP. Control particles prepared using sodium chloride as porogen are reported for comparison (DPPC20SC ). For key legend see Table 1.

carried out by formaldehyde (4%, w/v) to fix tissues. Afterward respiratory apparatus was removed and cross-sectioned (section thickness: 30 ␮m) (Accu-Cut® SRMTM 200 Rotary Microtome, Sakura Finetek, USA). Tissue slices were analyzed by a fluorescent microscope (Leica DMRB) and images acquired by Leica Application Suite Program V2.8.1. Control experiments were run to assess the autofluorescence of the pulmonary tissue after intra-tracheal delivery of unloaded gfLPP and than tissues were fixed as described above. 3. Results Gas-foamed large porous particles (gfLPP) based on poly(lactide-co-glycolic) acid (PLGA) were prepared by the double emulsion-solvent evaporation technique and efficiently engineered for local and prolonged delivery of an hydrophilic macromolecule in the lungs. The effect of ammonium bicarbonate used as porogen, as well as the addition of phospholipids within the formulation, on bulk and aerodynamic properties of PLGA-based gfLPP was elucidated. Then, the influence of the presence/nature of lipid excipients upon in vitro/in vivo aerosolization properties of gfLPP containing a model hydrophilic macromolecule, rhodamine B isothiocyanate–dextran (Rhod-dex), as well as their potential to control gfLPP release properties was assessed.

3.1. Effect of formulation conditions on technological/aerodynamic properties of unloaded gfLPP Preliminary formulation studies were carried out on unloaded particles prepared in different fabrication conditions (Table 1). In particular, three formulation variables were considered, that is presence and concentration of ammonium bicarbonate in the internal aqueous phase (w1 ), presence of DPPC or DOTAP in the organic phase (o), and PLGA concentration in o. As expected, the technological/aerodynamic properties of unloaded gfLPP were strongly affected by the type and amount of porogen added in w1 (Table 2). When powders were prepared without salts, regularly shaped non-porous particles unsuitable for lung delivery were achieved (DPPC20 in Fig. 1). Sodium chloride-engineered LPP displayed large, isolated and scattered pores (DPPC20SC in Fig. 1). An optimal particle morphology could be achieved only when 10% (w/v) of NH4 (HCO3 ) was used irrespective of PLGA concentration (DPPC20AB10 and DPPC15AB10 in Fig. 1). When NH4 (HCO3 ) was added in w1 at 5% or 20% (w/v), less porous or collapsed particles were achieved (DPPC20AB5 and DPPC20AB20 in Fig. 1, respectively). On the contrary, spherical and highly porous particles were obtained also with DOTAP when NH4 (HCO3 ) was used at 10% (w/v), independently upon PLGA concentration in o (Fig. 2).

Fig. 2. SEM micrographs of DOTAP-containing unloaded gfLPP. For key legend see Table 1.

F. Ungaro et al. / European Journal of Pharmaceutical Sciences 41 (2010) 60–70

65

Table 3 Overall properties of Rhod-dex loaded gfLPP. Formulation

Volume mean diameter (d)a (␮m ± SDb )

Rhod-15AB10 Rhod-DPPC15AB10 Rhod-DPPC20AB10 Rhod-DOTAP15AB10 Rhod-DOTAP20AB10

33.9 34.4 34.9 28.4 25.8

a b c d e

± ± ± ± ±

8.20 11.8 4.8 0.2 4.0

Tapped density () (g/ml ± SDb ) 0.082 0.066 0.078 0.050 0.047

± ± ± ± ±

0.059 0.015 0.018 0.014 0.013

MMADt c (␮m ± SDb ) 9.8 8.6 9.6 6.3 5.5

± ± ± ± ±

4.0 2.0 0.2 0.8 0.1

Carr’s indexd 6.5 9.4 7.9 13 13

± ± ± ± ±

3.6 2.4 1.7 6.8 4.6

Encapsulation efficiencye (% ± SDb ) 17.1 75.1 76.9 33.8 82.0

± ± ± ± ±

0.57 6.0 5.0 6.1 3.3

Mean geometric diameter as determined by laser diffraction. Standard deviation of values calculated on three different batches. Mass mean aerodynamic diameter estimated on the basis of Eq. (2). Powder flowability estimated on the basis of Eq. (1). Rhod-dex theoretical loading was 0.3% (0.3 mg per 100 mg of gfLPP).

As can be seen in Table 2, gfLPP formulations prepared at 10% (w/v) of NH4 (HCO3 ) displayed a very low tapped density of about 0.05 g/mL and a high geometric diameter (≈30 ␮m). Notably, very low values of Carr’s index ranging between 5.5% and 13% were calculated. Furthermore, MMADt values, preliminarily estimated by normalizing the geometric diameter with the powder tapped density according to Eq. (1), felt within the range 5–10 ␮m (Table 2).

3.2. Effect of formulation conditions upon the technological/aerodynamic properties of gfLPP loaded with an hydrophilic macromolecule Rhod-dex was loaded in gfLPP optimized formulations prepared at 10% NH4 (HCO3 ). As can be seen in Table 1, two formulation variables were taken into account, that is PLGA amount and lipid excipient (i.e., DPPC and DOTAP) added in

Fig. 3. SEM micrographs of Rhod-dex loaded gfLPP. For key legend see Table 1.

66

F. Ungaro et al. / European Journal of Pharmaceutical Sciences 41 (2010) 60–70

the organic phase. In each case, highly porous Rhod-dex loaded gfLPP were obtained with good yields, ranging between 70% and 77% (Fig. 3 and Table 3). Nonetheless, DPPC-containing gfLPP were characterized by very small and well-distributed surface pores (Rhod-DPPC15 and Rhod-DPPC20 in Fig. 3), whereas DOTAP-containing gfLPP displayed a much more porous structure (Rhod-DOTAP15 and Rhod-DOTAP20 in Fig. 3). Large and unhomogenous pores were also observed on excipient-free gfLPP surface (Rhod-15 in Fig. 3). In Fig. 4, CLSM images of a single Rhod-dex loaded gfLPP for all the batches are reported. Notably, in the case of gfLPP containing DPPC or DOTAP, fluorescent spots distributed evenly in the polymer matrix were observed irrespective of PLGA concentration employed. When lipids were not added in the organic phase, Rhod-dex was localized mainly on particle surface (Rhod-15 in Fig. 4). Results confirmed also good flow/aerodynamic properties of Rhod-dex loaded gfLPP, displaying a very low tapped density and MMADt ranging between 5.5 and 9.6 ␮m (Table 3). Again, very low values of Carr’s index ranging between 7.8% and 13% were calculated. While Rhod-dex was entrapped within excipient-free gfLPP with an efficiency lower than 20%, the addition of DPPC or DOTAP within the formulation resulted in entrapment efficiencies as high as 80%. Only in the case of Rhod-DOTAP15 the encapsulation efficiency resulted to be around 34%.

Fig. 5. In vitro release profiles of Rhod-dex from gfLPP. Panel A: gfLPP prepared at 15% (w/v) of PLGA in methylene chloride. Panel B: gfLPP prepared at 20% (w/v) of PLGA in methylene chloride. Data are reported as mean ± standard deviation (SD).

3.3. Effect of formulation conditions upon release profile of an hydrophilic macromolecule from gfLPP

Fig. 4. CLSM images of Rhod-dex loaded gfLPP. For key legend see Table 1.

Results of in vitro release studies performed by dialysis at physiological pH and temperature (pH 7.2 and 37 ◦ C) are reported in Fig. 5 as percentage of Rhod-dex released over time. In each case, gfLPP displayed a two-stage release profile characterized by an initial rapid release phase (i.e., burst), followed by a modulated and progressive release of Rhod-dex (Fig. 5). In particular, while Rhod-dex release from Rhod-15 was almost complete after 20 days, both Rhod-DPPC15 and RhodDOTAP15 gfLPP released Rhod-Dex for longer time frames (up to 2 months) (Fig. 5, panel A). Furthermore, DOTAP-engineered gfLPP displayed lower burst as compared to DPPC-containing particles (11.7% versus 17.1% for Rhod-DOTAP15 and Rhod-DPPC15, respectively), this effect being more evident in the case of gfLPP prepared at 20% (w/v) of PLGA in the organic phase (Fig. 5, panel B).

F. Ungaro et al. / European Journal of Pharmaceutical Sciences 41 (2010) 60–70 Table 4 In vitro aerosolization properties of Rhod-dex loaded gfLPP. Formulation Rhod-15 Rhod-DPPC15 Rhod-DPPC20 Rhod-DOTAP15 Rhod-DOTAP20

ED (% ± SDa ) 27.4 80.1 97.1 90.4 81.1

± ± ± ± ±

6.7 1.0 3.9 11 0.4

FPF (% ± SDa ) b

ND 38.7 ± 2.9 44.9 ± 1.6 44.7 ± 1.5 46.3 ± 0.4

MMADexp (% ± GSD) NDb 7.6 ± 1.3 6.8 ± 1.2 6.0 ± 1.2 6.4 ± 1.2

ED = emitted dose; FPF = fine particle fraction; MMADexp = experimental mass mean aerodynamic diameter; GSD = geometrical standard deviation. a Standard deviation of values calculated on three different batches. b Not determined due to the poor flow properties of excipient-free powders.

3.4. Effect of formulation conditions upon in vitro/in vivo aerosolization properties of loaded gfLPP The aerosolization properties of Rhod-dex loaded gfLPP were analyzed in depth according to Pharmacopoeial test. The emitted dose (ED), fine particle dose (FPD), fine particle fraction (FPF) and mass mean aerodynamic diameter (MMADexp ) are listed in Table 4. ED reached a value as high as 80% for all the formulations except for Rhod-15, which showed a significantly lower ED of about 27.4%. Worthily, no significant difference in terms of FPF and MMADexp was observed between formulations. The FPF appeared to be less than 50% and the experimental MMADexp was lower than 8 ␮m regardless of the type of formulation studied. In particular MMADexp values felt within the range 6.1–7.6 ␮m with low geometric standard deviations (GSD) varying between 1.2 and 1.3. The deposition pattern of the developed gfLPP was confirmed in vivo after administration of a model formulation in rats by a lowscale dry powder inhaler. In Fig. 6, fluorescence microscopy images of rat airway slices after intra-tracheal delivery of Rhod-DPPC15 gfLPP are reported. The autofluorescence of lung tissue in the red channel was found to be negligible at the microscopy settings used in this study (data not shown). Thus, red fluorescence was unequivocally attributed to Rhod-dex loaded gfLPP. Notably, after administration, several red spots were evident through the trachea, the bronchia and the bronchioles but not on alveolar surface. 4. Discussion Gas-foamed large porous particles (gfLPP) based on poly(lactide-co-glycolic) acid (PLGA) have been recently suggested as potential carriers for pulmonary drug delivery (Yang et al., 2009). In this work, we attempt to engineer gfLPP for efficient local delivery of macromolecules to the lung. To this end, three crucial technological aspects of gfLPP were taken into account, that is: (i) primary bulk properties (e.g., morphology, size and density), determining powder handling and dispersibility as well as gfLPP potential to escape local macrophage uptake; (ii) encapsulation efficiency and release properties, which should be tuned to prolong in situ delivery of the encapsulated macromolecule to achieve a long-term therapeutic effect; (iii) aerosolization properties, affecting gfLPP deposition along the respiratory apparatus and, thus, their potential to spread over upper airways to exert a local effect Preliminary formulation studies, carried out on unloaded gfLPP prepared in different fabrication conditions, highlighted that ammonium bicarbonate plays a crucial role in determining the technological/aerodynamic properties of gfLPP (Table 2). Sodium chloride, a typical osmogen which could be in theory used for the production of porous PLGA particles by double emulsion-solvent evaporation technique (Pistel and Kissel, 2000), demonstrated unsuitable for the production of LPP (DPPC20SC in Fig. 1), likely due to its detrimental effect on primary emulsion stability. On the other hand, the addition of an appropriate amount of ammo-

67

nium bicarbonate, which spontaneously produces CO2 and NH3 during solvent evaporation, resulted in a homogeneous population of highly porous particles (DPPC20AB10 in Fig. 1), ascribable to gas migration toward the organic phase of the double emulsion (Yang et al., 2009). Actually, it has been previously observed that the evolution of small gas bubbles from the surface of internal water droplets may sterically prevent them from coalescence (Kim et al., 2006). Further stabilization of gas bubbles, as well as internal water droplets, was originated from lipid surfactant addition, likely resulting in numerous, small and well-distributed pores within the generated particles, as confirmed by CLSM studies (Fig. 4). In agreement with previous literature findings (Kim et al., 2006), less porous particles were achieved in the presence of a lower amount of effervescent salt. On the other hand, particle collapse, observed when higher amounts of ammonium bicarbonate were employed, is likely ascribable to the generation of excessive effervescence during gfLPP production. The fact that a homogeneous population of highly porous particles was achieved also when DOTAP was used independently upon PLGA concentration in the organic phase, suggested that surfactant type is not critical for gfLPP morphology. Conceiving engineered gfLPP for pulmonary delivery, an important aspect to consider is their flow properties. On this matter, the fact that gfLPP formulations prepared at 10% (w/v) of ammonium bicarbonate displayed very low tapped densities and high geometric diameters, suggested a great potential as dry powders for inhalation. Nonetheless, from a technological standpoint, it should be considered that a powder with a low tapped density may be more prone to consolidation (i.e., densification or packing of the particles), which is undesirable during powder filling or product shipping (Van Campen and Venthoye, 2002). Furthermore, the presence of DPPC or DOTAP in the PLGA-containing organic phase may strongly affect the physico-chemical properties of the polymer matrix (e.g., surface energy, net charge, hygroscopic nature) and, in so doing, gfLPP flow properties. If DPPC disposition on gfLPP surface may reduce particle surface energy and hence powder cohesion (Evora et al., 1998; Vanbever et al., 1999), DOTAP has been especially used to induce agglomeration of PLGA particles by adjusting their surface charge (Peek et al., 2008). In the light of these observations, the extent of inter-particulate interactions and powder consolidation was estimated through the relative percent difference between bulk and tapped density (expressed as Carr’s index) as stated by US Pharmacopoeia. The very low values of Carr’s index calculated for both DPPC- and DOTAP-engineered gfLPP indicated powders with excellent flowability despite gfLPP very low tapped density and addition of lipid surfactants. The preliminary estimation of MMADt confirmed the good flow properties of the developed gfLPP and an aerodynamic behavior suitable to reach bronchia/bronchioles (Table 2). In the light of the preliminary technological results, a model hydrophilic macromolecule, namely Rhod-dex, was loaded in gfLPP optimized formulations, which varied for PLGA amount and lipid excipient present in the organic phase. In this case, SEM analysis underlined a different behavior between DPPC and DOTAP, with DOTAP-engineered gfLPP displaying large and unhomogenous pores, very similar to those achieved for excipient-free gfLPP (Fig. 3). As a rule, because of their amphiphillic nature (i.e., hydrophilic “head” plus hydrophobic “tail”), surfactant aid excipients are expected to preferentially locate at both surfaces and interfaces of the dispersed system. In the case of DOTAP-engineered gfLPP, its role at the interfaces/surfaces is much less predictable as compared to DPPC, likely due to its potential involvement also in the formation of ion pairings with both PLGA carboxylic ends and negatively charged Rhod-dex. This may result in a minor effect of DOTAP on the stabilization of the evolving gas bubbles before particle hardening. Results confirmed also good flow/aerodynamic

68

F. Ungaro et al. / European Journal of Pharmaceutical Sciences 41 (2010) 60–70

Fig. 6. Localization of Rhod-dex loaded gfLPP in rat respiratory system. Photomicrographs show localization of fluorescent gfLPP (red) in 12 ␮m sections of trachea (A), left primary bronchus (B), and lobar bronchus (C). No particles were found in alveolar ducts (D). Bars are 100 ␮m. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)

properties of the developed gfLPP, displaying a very low tapped density, a MMADt suitable for deposition along the lung irrespective of PLGA concentration and excipient employed (Table 4). Again, very low values of Carr’s index ranging between 7.8% and 13% were calculated, indicating powder excellent flowability. When dealing with porous drug delivery systems, such as gfLPP, a critical feature to ensure batch to batch reproducibility is the achievement of an efficient drug distribution within the polymer matrix, high encapsulation efficiency and control over drug release rate. On this matter, CLSM results (Fig. 4) indicated that the addition of DPPC or DOTAP in the formulation modulates not only particle porosity but also drug distribution within the particles, which is critical for drug release rate. Concerning encapsulation efficiency, double emulsion technique assisted by gas-foaming has been recently demonstrated as a viable alternative to more widely employed osmogens to increase lysozyme encapsulation efficiency within PLGA-based LPP (Yang et al., 2009). Nonetheless, in the case of hydrophilic macromolecules unable to interact with PLGA end-groups, such as Rhod-dex, we found that excipient addition is essential to increase the amount of drug entrapped within gfLPP, being as high as 80% when DPPC or DOTAP were used in appropriate formulation conditions (Table 3). Lipid excipients played a crucial role also in determining release profile of the encapsulated hydrophilic macromolecule, with particular regard to initial drug release phase (i.e., burst) and overall release duration (Fig. 5). The main factors controlling drug release

from highly porous particles have not been fully elucidated yet. As a rule, release profile of a macromolecule from PLGA particles result from the complex interplay of drug diffusion through the polymeric matrix, polymer erosion, and particle porous microenvironment (Batycky et al., 1997; Lemaire et al., 2003). Diffusion of the macromolecule through the particle pore network formed during fabrication (i.e., macropores) and subsequent enhanced diffusion via erosion-induced pore evolution have been regarded as the predominant release mechanisms of PLGA particles. In particular, the initial burst can be defined as the amount of drug that escapes from microparticle prior to the onset of erosion-mediated release phase (i.e., after initial hydration phase), involving those molecules with easy access to particle surface (Batycky et al., 1997; Lemaire et al., 2003; Allison, 2008). On the basis of the pre-existing highly porous structure of gfLPP, likely facilitating drug dissolution and subsequent diffusion out of the device, high burst were expected. Once the drug is not chemically bound to the polymer but only physically entrapped into the matrix, a homogeneous distribution of drug-containing internal cavities, as that achieved in the presence of DPPC and DOTAP (Fig. 4), can be supposed to slow down release rate in the diffusion phase. Nonetheless, release data suggest that also other factors play a role in modulating Rhod-dex release from gfLPP and, in particular, the initial release phase. This is the case of DOTAP-containing gfLPP, where the contribution of the electrostatic interactions occurring between conjugated dextran and the cationic phospholipid on release rate cannot be neglected. Actually,

F. Ungaro et al. / European Journal of Pharmaceutical Sciences 41 (2010) 60–70

isothiocianate–dextran modified rhodamine is negatively charged and likely able to interact with this lipid excipient at the interfaces. Thus, these results highlight the importance of using phospholipids to modulate release of hydrophilic macromolecules from gfLPP and, in particular, to employ a cationic phospholipid when gfLPP are conceived for lung delivery of anionic drugs, such as nucleic acids. A crucial factor to consider when developing inhalable dry powders relies on their aerosolization properties. Thus, the deposition pattern of Rhod-dex loaded gfLPP was assessed in vitro by a MSLI after delivery from a breath-actuated DPI (Turbospin® ) characterized by an intermediate device resistance (0.08 cm (H2 O)1/2 min L−1 ) which represents a good balance between the patient inspiratory effort and the airflow effect on particle deaggregation (i.e., low inter-patient variability) (Meakin et al., 1996; Newman and Busse, 2002; Frijlink and De Boer, 2004). The superior flow properties of DPPC-containing gfLPP as compared to gfLPP prepared without additives can be reasonably attributed to the lipid surfactant disposition at the oil/water interface formed during gfLPP fabrication by the double emulsion method, likely resulting in its adsorption on gfLPP surface, decreasing particle surface energy and hence powder cohesion (Evora et al., 1998; Vanbever et al., 1999). A similar behavior has been recently reported for DOTAP, which has been demonstrated to line the surface of PLGA particles produced by emulsion methods (Díez et al., 2009). As already suggested by their optimal flow properties (Table 3), electrostatic repulsions between the resulting positively charged particles likely prevents gfLPP aggregation and enhances their aerosolization properties. This result, together with optimal MMADexp achieved independently upon formulation conditions, suggests a great potential for the developed gfLPP to deposit along bronchia and bronchiole level and effectively escape from local macrophages. It is well known that the main disadvantage of in vitro test developed so far to this purpose, comprising inertial impactors (i.e., ACI and NGI), is that none of them actually resembles in vivo airway complexity. Furthermore, significant different in vitro data may be achieved testing a dry powder in the same conditions (i.e., same flow rate and DPI) by different impactors (Taki et al., 2010). Thus, even if liquid impingment is a Pharmacopeial standard method for in vitro assessment of aerodynamic deposition of inhaled formulations, we decided to confirm in vivo the deposition pattern of the developed gfLPP. The fact that after administration several red spots, unequivocally attributed to loaded gfLPP, were evident through trachea, bronchia and bronchioles but not on alveolar surface of rat lung, confirmed that the developed gfLPP exhibit aerosolization properties suitable to distribute homogeneously along the respiratory tree. The value of the developed gfLPP as carriers for local delivery of macromolecules to the lung is further strengthened by their in vivo incapacity to reach alveoli, meaning that gfLPP have a reduced potential for drug systemic absorption (i.e., unspecific accumulation in non-target tissues). 5. Conclusions Gas-foamed large porous particles (gfLPP) prepared by ammonium bicarbonate-assisted double emulsion technique have been efficiently engineered for local delivery of hydrophilic macromolecules to the lung. Results demonstrated that the addition of appropriate amounts of porogen within initial formulation is essential to produce gfLPP with optimal morphology and flow properties. When a hydrophilic macromolecule was loaded within gfLPP, the addition of DPPC or DOTAP was crucial to control encapsulation efficiency and release properties of the developed particles. As compared to excipient-free gfLPP, lipid-engineered gfLPP displayed very good in vitro aerosolization properties, suggesting

69

a great potential for inhalation. Results were confirmed in vivo by deposition studies performed after intra-tracheal administration of gfLPP to rats, showing a homogeneous distribution of the developed dry powders along the respiratory tree. In perspective, lipid-engineered gfLPP represent a viable alternative to LPP developed so far to achieve local and prolonged release of hydrophilic macromolecules, such as nucleic acids, in the lungs. Acknowledgements The financial support of Fondazione per la Ricerca sulla Fibrosi Cistica-Onlus (FFC#5/2007) is gratefully acknowledged. We warmly thank Dr. Giovanna Benvenuto and Dr. Franco Iamunno for their technical assistance in CLSM and SEM analysis. References Agu, R.U., Ugwoke, M.I., Armand, M., Kinget, R., Verbeke, N., 2001. The lung as a route for systemic delivery of therapeutic proteins and peptides. Respir. Res. 2, 198–209. Allison, S.D., 2008. Analysis of initial burst in PLGA microparticles. Expert Opin. Drug Deliv. 5, 615–628. Batycky, R.P., Hanes, J., Langer, R., Edwards, D.A., 1997. A theoretical model of erosion and macromolecular drug release from biodegrading microspheres. J. Pharm. Sci. 86, 1464–1474. Cryan, S.A., 2005. Carrier-based strategies for targeting protein and peptide drugs to the lungs. AAPS J. 7, E20–E41. Densmore, C.L., 2006. Advances in noninvasive pulmonary gene therapy. Curr. Drug Deliv. 3, 55–63. De Rosa, G., Quaglia, F., Bochot, A., Ungaro, F., Fattal, E., 2003. Long term release and improved intracellular penetration of oligonucleotide–polyethylenimine complexes entrapped in biodegradable microspheres. Biomacromolecules 4, 529–536. De Rosa, G., La Rotonda, M.I., Quaglia, F., Ungaro, F., 2008. Use of additives in the design of poly(lactide-co-glycolide) microspheres for drug delivery. In: Ravi Kumar, M.N.V. (Ed.), Handbook of Particulate Drug Delivery. American Scientific Publisher, CA, USA, pp. 61–91. Díez, S., Navarro, G., de Ilarduya, C.T., 2009. In vivo targeted gene delivery by cationic nanoparticles for treatment of hepatocellular carcinoma. J. Gene Med. 11, 38–45. Edwards, D.A., Ben Jebria, A., Langer, R., 1998. Recent advances in pulmonary drug delivery using large, porous inhaled particles. J. Appl. Physiol. 85, 379–385. Evora, C., Soriano, I., Rogers, R.A., Shakesheff, K.N., Hanes, J., Langer, R., 1998. Relating the phagocytosis of microparticles by alveolar macrophages to surface chemistry: the effect of 1,2-dipalmitoylphosphatidylcholine. J. Control. Release 51, 143–152. Frijlink, H.W., De Boer, A.H., 2004. Dry powder inhalers for pulmonary drug delivery. Expert Opin. Drug Deliv. 1, 67–86. Garbuzenko, O.B., Saad, M., Betigeri, S., Zhang, M., Vetcher, A.A., Soldatenkov, V.A., Reimer, D.C., Pozharov, V.P., Minko, T., 2009. Intratracheal versus intravenous liposomal delivery of siRNA, antisense oligonucleotides and anticancer drug. Pharm. Res. 26, 382–394. Kaye, R.S., Purewal, T.S., Alpar, H.O., 2009. Simultaneously manufactured nanoin-micro (SIMANIM) particles for dry-powder modified-release delivery of antibodies. J. Pharm. Sci. 98, 4055–4068. Kim, T.K., Yoon, J.J., Lee, D.S., Park, T.G., 2006. Gas foamed open porous biodegradable polymeric microspheres. Biomaterials 27, 152–159. Kleinstreuer, C., Zhang, Z., Donohue, J.F., 2008. Targeted drug-aerosol delivery in the human respiratory system. Annu. Rev. Biomed. Eng. 10, 195–220. Know, M.J., Bae, J.H., Kim, J.J., Na, K., Lee, E.S., 2007. Long acting porous microparticles for pulmonary protein delivery. Int. J. Pharm. 333, 5–9. Koushik, K., Dhanda, D.S., Cheruvu, N.P., Kompella, U.B., 2004. Pulmonary delivery of deslorelin: large-porous PLGA particles and HPbetaCD complexes. Pharm. Res. 21, 1119–1126. Kumar, T.R., Soppimath, K., Nachaegari, S.K., 2006. Novel delivery technologies for protein and peptide therapeutics. Curr. Pharm. Biotechnol. 7, 261–276. Laube, B.L., 2005. The expanding role of aerosols in systemic drug delivery, gene therapy, and vaccination. Respir. Care 50, 1161–1176. Lemaire, V., Bélair, J., Hildgen, P., 2003. Structural modeling of drug release from biodegradable porous matrices based on a combined diffusion/erosion process. Int. J. Pharm. 258, 95–107. Meakin, B.J., Woodcock, P.M., Segu, A., Niccolai, F., Vecchio, C., 1996. Evaluation of the resistance to airflow through Turbospin in comparison with dry powder inhalers available on the market. Boll. Chim. Farm.-Anno 135 (125), 107–109. Minne, A., Boireau, H., Horta, M.J., Vanbever, R., 2008. Optimization of the aerosolization properties of an inhalation dry powder based on selection of excipients. Eur. J. Pharm. Biopharm. 70, 839–844. Mohamed, F., van der Walle, C.F., 2008. Engineering biodegradable polyester particles with specific drug targeting and drug release properties. J. Pharm. Sci. 97, 71–87.

70

F. Ungaro et al. / European Journal of Pharmaceutical Sciences 41 (2010) 60–70

Moschos, S.A., Spinks, K., Williams, A.E., Lindsay, M.A., 2008. Targeting the lung using siRNA and antisense based oligonucleotides. Curr. Pharm. Des. 14, 3620–3627. Newhouse, M.T., Hirst, P.H., Duddu, S.P., Walter, Y.H., Tarara, T.E., Clark, A.R., Weers, J.G., 2003. Inhalation of a dry powder tobramycin PulmoSphere formulation in healthy volunteers. Chest 124, 360–366. Newman, S.P., Busse, W.W., 2002. Evolution of dry powder inhaler design, formulation, and performance. Respir. Med. 96, 293–304. Patton, J.S., Byron, P.R., 2007. Inhaling medicines: delivering drugs to the body through the lungs. Nat. Rev. Drug Discov. 6, 67–74. Peek, L.J., Roberts, L., Berkland, C., 2008. Poly(d,l-lactide-co-glycolide) nanoparticle agglomerates as carriers in dry powder aerosol formulation of proteins. Langmuir 24, 9775–9783. Pistel, K.F., Kissel, T., 2000. Effects of salt addition on the microencapsulation of proteins using W/O/W double emulsion technique. J. Microencapsul. 17, 467–483. Rawat, A., Majumder, Q.H., Ahsan, F., 2008. Inhalable large porous microspheres of low molecular weight heparin: in vitro and in vivo evaluation. J. Control. Release 128, 224–232. Séguin, R.M., Ferrari, N., 2009. Emerging oligonucleotide therapies for asthma and chronic obstructive pulmonary disease. Expert Opin. Investig. Drugs 18, 1505–1517. Simberg, D., Weisman, S., Talmon, Y., Barenholz, Y., 2004. DOTAP (and other cationic lipids): chemistry, biophysics, and transfection. Crit. Rev. Ther. Drug Carrier Syst. 21, 257–317.

Taki, M., Marriott, C., Zeng, X.M., Martin, G.P., 2010. Aerodynamic deposition of combination dry powder inhaler formulations in vitro: a comparison of three impactors. Int. J. Pharm. 388, 40–51. Tsapis, N., Bennett, D., Jackson, B., Weitz, D.A., Edwards, D.A., 2002. Trojan particles: large porous carriers of nanoparticles for drug delivery. Proc. Natl. Acad. Sci. U.S.A. 99, 12001–12005. Ungaro, F., De Rosa, G., Miro, A., Quaglia, F., La Rotonda, M.I., 2006. Cyclodextrins in the production of large porous particles: development of dry powders for the sustained release of insulin to the lungs. Eur. J. Pharm. Sci. 28, 423–432. Ungaro, F., d’Emmanuele di Villa Bianca, R., Giovino, C., Miro, A., Sorrentino, R., Quaglia, F., La Rotonda, M.I., 2009. Insulin-loaded PLGA/cyclodextrin large porous particles with improved aerosolization properties: in vivo deposition and hypoglycaemic activity after delivery to rat lungs. J. Control. Release 135, 25–34. Vanbever, R., Mintzes, J.D., Wang, J., Nice, J., Chen, D., Batycky, R., Langer, R., Edwards, D.A., 1999. Formulation and physical characterization of large porous particles for inhalation. Pharm. Res. 16, 1735–1742. Van Campen, L., Venthoye, G., 2002. Inhalation, dry powder. In: Swarbrick, J., Boylan, J.C. (Eds.), Encyclopedia of Pharmaceutical Technology. Marcel Dekker Inc., New York, pp. 1529–1544. Yang, Y., Bajai, N., Xu, P., Ohn, K., Tsifansky, D., Yeo, Y., 2009. Development of highly porous large PLGA microparticles for pulmonary drug delivery. Biomaterials 30, 1–7.