Enhanced powder dispersion of dual-excipient spray-dried powder formulations of a monoclonal antibody and its fragment for local treatment of severe asthma

International Journal of Pharmaceutics 644 (2023) 123272

Available online 25 July 2023 0378-5173/© 2023 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by- nc-nd/4.0/).

Enhanced powder dispersion of dual-excipient spray-dried powder formulations of a monoclonal antibody and its fragment for local treatment of severe asthma

Harry W. Pan a, Jinlin Guo b, Lingqiao Zhu b, Susan W.S. Leung a, Chenghai Zhang b,*, Jenny K.

W. Lam a,c,* a Department of Pharmacology and Pharmacy, Li Ka Shing Faculty of Medicine, The University of Hong Kong, 21 Sassoon Road, Pokfulam, Hong Kong SAR, China b R&D Department, Shanghai MabGeek Biotech Co. Ltd., Room 304, No. 1011 Halei Road, Zhangjiang Hi-tech Park, Shanghai, 201203, China c Department of Pharmaceutics, UCL School of Pharmacy, University College London, 29-39, Brunswick Square, London WC1N 1AX, United Kingdom

A R T I C L E I N F O Keywords:

Antibody fragment Asthma Cyclodextrin Inhalation Leucine

Pulmonary delivery Spray drying A B S T R A C T The advent of biologics has brought renewed hope for patients with severe asthma, a condition notorious for being hampered by poor response to conventional therapies and adverse drug reactions owing to corticosteroid dependence. However, biologics are administered as injections, thereby precluding the benefits inhalation therapy could offer such as increased bioavailability at the site of action, minimal systemic side effects, non- invasiveness, and self-administration. Here, 2-hydroxypropyl-beta-cyclodextrin and ʟ-leucine were co-spray- dried, as protein stabiliser and dispersion enhancer, respectively, at various weight ratios to produce a series of formulation platforms. Powder aerosolisation characteristics and particle morphology were assessed for suitability for pulmonary delivery. The selected platform with the best aerosol performance, a 1:1 ratio of the excipients, was then incorporated with a monoclonal antibody directed against IL-4 receptor alpha or its antigen- binding fragment. The dual-excipient antibody formulations exhibited emitted fraction of at least 80% and fine particle fraction exceeding 60% in cascade impactor study, while the residual moisture content was within a desirable range between 1% and 3%. The in vitro antigen-binding ability and inhibitory potency of the spray- dried antibody were satisfactorily preserved. The results from this study corroborate the viability of inhaled solid-state biomacromolecules as a promising treatment approach for asthma.

1. Introduction The number of inflammatory diseases embracing the use of biologics in their arsenal of treatment modalities is expanding, and asthma is no exception. A major chronic disease of the respiratory system, asthma affected more than 260 million individuals across the world in 2019 (Institute for Health Metrics and Evaluation, 2020). Severe asthma, which accounts for around 3% to 10% of people with asthma, typically depends on a maximal high-dose inhaled corticosteroid with a long- acting beta2-agonist (Chung et al., 2014; Global Initiative for Asthma,

2022). For patients who experience exacerbations or poor symptom control despite a high-dose inhaled therapy, or depend on maintenance oral corticosteroids, which are associated with severe long-term adverse effects (Lefebvre et al., 2015), biologic therapies are an option. Since severe asthma is often driven by type 2 inflammation that is characterised by airway eosinophilia and cytokines secreted by T-helper

2 cells such as interleukins (ILs) 4, 5, and 13 (Israel and Reddel, 2017), immunoglobulins (Ig) in clinical use and research strategically target these pathways.

There are a handful of monoclonal antibodies (mAbs) that have been approved by the U.S. FDA for the treatment of severe asthma, including omalizumab (IgE antagonist), mepolizumab and reslizumab (IL-5 an­ tagonists), dupilumab (IL-4 receptor antagonist), benralizumab (IL-5 receptor antagonist), and tezepelumab (thymic stromal lymphopoietin antagonist). All of them are administered parenterally, which increases the risk of off-target systemic side effects and may require larger doses compared to local therapy (Borghardt et al., 2018; Irvine et al., 2013). A noninvasive route of delivery is no doubt more popular among patients and healthcare personnel, particularly in the management of a chronic condition like asthma. Exemplified by the array of inhaled products for

* Corresponding authors.

E-mail addresses: chenghai.zhang@mabgeek.com (C. Zhang), jenny.lam@ucl.ac.uk (J.K.W. Lam).

Contents lists available at ScienceDirect International Journal of Pharmaceutics journal homepage: www.elsevier.com/locate/ijpharm https://doi.org/10.1016/j.ijpharm.2023.123272

Received 16 March 2023; Received in revised form 18 July 2023; Accepted 24 July 2023

International Journal of Pharmaceutics 644 (2023) 123272

2 asthma on the market, targeted delivery to the lungs through oral inhalation has advantages encompassing rapid onset of action, reduced systemic adverse reactions, high local bioavailability, and potential dose reduction (Liang et al., 2020). Although pulmonary delivery of anti­ bodies has been investigated using dry powder inhalers (DPIs) or neb­ ulisers (Hickey and Stewart, 2022; Matthews et al., 2020), the former group of inhalation devices are more compatible with labile bio­ macromolecules inasmuch as they do not generate heat (Shoyele and

Slowey, 2006) or create a vast air–liquid interface (Fr¨ohlich and Salar- Behzadi, 2021). In nebulisation, proteins are in the liquid state which necessitates the infamous cold chain that increases the manufacturing costs, negatively impacts the environment, and complicates the logis­ tical operations (Sharma et al., 2021). On the other hand, proteins delivered through DPIs are in the solid state which, with proper formulation, can confer prolonged shelf-life and stability (Chang and

Pikal, 2009), given that chemical and physical degradation reactions are driven hydrolytically (Lai and Topp, 1999). It has been proposed that a pressure drop of no less than 1 kPa, produced by the patient’s inspiratory effort, is required to effectively fluidise and disperse the powder from a

DPI (Clark et al., 2020). At the airflow resistances of many commercially available DPIs, the majority of patients with asthma, including those with severe asthma, are able to generate sufficient inspiratory flow rates necessary for adequate drug delivery to the lungs (Laube et al., 2011).

To produce biologics in powder form for inhalation, several particle- engineering technologies are available, including freeze drying-milling, spray freeze drying, and thin-film freeze drying (Chang et al., 2021).

Spray drying was used in this study in view of its scalability on an in­ dustrial level, ability to generate particles with uniform size distribution, and the lower costs involved (Chow et al., 2007; Sharma et al., 2021). It is also an established drying technique to manufacture the now-defunct

Exubera® (human insulin), one of the rare inhaled protein products that had received regulatory authorisation (White et al., 2005). A one-step process, spray drying involves the atomisation of a liquid formulation by means of a nozzle into a chamber of hot gas stream and the solvent in the atomised droplets is removed via evaporation (Ziaee et al., 2019).

The high temperature of the drying gas subjects thermolabile molecules to heat stress while atomisation introduces mechanical shear forces and induces protein adsorption at an extensive air–liquid interface that may destabilise biomacromolecules (Chow et al., 2007). For these reasons, stabilising excipients are incorporated into protein formulations for their protective effects against the various stresses encountered during the dehydration stage (Chang et al., 2021).

Among the well-known classes of solid-state protein stabilisers, which include sugars, polyols, surfactants, and salts (Depreter et al.,

2013), cyclodextrins, in particular, 2-hydroxypropyl-beta-cyclodextrin (2HPβCD) has gained attention as a promising excipient for both its protein-stabilising abilities (Serno et al., 2011) and capability to pro­ duce inhalable particles for pulmonary delivery (Ramezani et al., 2017).

The mechanisms through which 2HPβCD stabilises proteins have been postulated to include water replacement effect, vitrification, and surface activity (Serno et al., 2011). Like some other cyclodextrin derivatives,

2HPβCD is found in commercialised pharmaceutical products as an excipient for the intramuscular, intravenous, and oral routes (Jansook et al., 2018). Even though it is not yet licenced for the respiratory route, presumably due to insufficient safety data, 2HPβCD administered by intranasal spray to healthy human volunteers was well-tolerated (al- Nakib et al., 1989) and short-term exposure in mice via nebulisation did not lead to overt lung toxicity (Evrard et al., 2004).

Our previous work characterised spray-dried (SD) and spray-freeze- dried (SFD) powder formulations of an anti-IL-4 receptor alpha (IL-4Rα) mAb with 2HPβCD in terms of physicochemical properties, aerosol performance, protein stability, and biological activity (Pan et al., 2022).

Here, we sought to address the issues of high residual moisture content and suboptimal aerosolisation by the incorporation of leucine, which is a very hydrophobic amino acid. Leucine has been investigated as a dispersion enhancer in SD formulations of pharmaceuticals intended for inhalation, including biologics (Alhajj et al., 2021). We aimed to study the physical and aerodynamic characteristics of dual-excipient powder formulation platforms that comprised 2HPβCD and leucine in different weight ratios prepared by spray drying. The excipient ratio that ach­ ieved the most desirable outcomes was then applied to antibody- containing formulations. Besides the intact full-length mAb, the antigen-binding fragment (Fab) would be featured in the formulations as well. In addition, the stability and in vitro bioactivity of the SD antibody were evaluated.

2. Materials and methods 2.1. Materials Humanised anti-IL-4Rα mAb (~18 mg/mL) and the Fab (~16 mg/ mL) in phosphate-buffered saline (PBS) were developed and provided by

Shanghai MabGeek Biotech Co. Ltd. (Shanghai, China), and stored at

−80 ◦C. The antibody (IgG4) was generated from mouse hybridoma and expressed by stably-transfected CHO-K1 cells (ATCC® CCL-61™, Man­ assas, VA, USA). The Fab was also expressed by CHO-K1 cells, by tran­ siently transfecting with the Fab sequence of the mAb. 2HPβCD, ʟ- leucine, dithiothreitol (DTT), Coomassie Brilliant Blue R-250, trisodium phosphate (Na3PO4), bovine serum albumin (BSA), and Tween® 20 were purchased from Sigma-Aldrich (St. Louis, MO, USA). Recombinant human (rh) IL-4Rα, rhIL-4, rhIL-13, granulocyte–macrophage colony- stimulating factor (rhGM-CSF), and substrate reagent pack (consisting of colour reagents A and B, stabilised hydrogen peroxide and stabilised tetramethylbenzidine, respectively) were purchased from R&D Systems (Minneapolis, MN, USA). Horseradish peroxidase-conjugated polyclonal goat F(ab’)2 directed against human IgG (detection antibody), was ob­ tained from Abcam (Cambridge, UK). Prestained protein ladder (Pag­ eRuler™ Plus), RPMI-1640 medium powder, qualified foetal bovine serum (FBS), and Antibiotic-Antimycotic (penicillin–streptomycin- amphotericin B) were acquired from Thermo Fisher Scientific (Waltham,

MA, USA). Cell Counting Kit-8 (CCK-8) was procured from MedChe­ mExpress (Monmouth Junction, NJ, USA) and Dual-Glo® Luciferase

Assay System was obtained from Promega (Madison, WI, USA). Sulfuric acid (H2SO4) was bought from BDH Chemicals (Poole, England). Ul­ trapure water was retrieved from a laboratory water purification system with a 0.2-µm pore size rating (Barnstead NANOpure Diamond™, APS

Water Services, Van Nuys, CA, USA).

2.2. Formulations and spray drying Dual-excipient formulation platforms were prepared by weighing the appropriate amount of leucine and adding it to a concentrated solution

Table 1 Composition of the feed solutions for spray drying.

Formulation 2HPβCD content (% w/w) ʟ-leucine content (% w/w)

Antibody format Antibody content (% w/w) Excipient-only formulation platforms

E0 100 0 – – E1 95 5 – – E2 90 10 – – E3 80 20 – –

E4 50 50 – – Antibody-containing formulations F1 45

45 Fab 10 F2 37.5 37.5 Fab 25 F3 25 25 Fab 50 M1 45

45 mAb 10 M2 37.5 37.5 mAb 25 M3 25 25 mAb 50 2HPβCD: 2-hydroxypropyl-beta-cyclodextrin; Fab: antigen-binding fragment; mAb: full-length monoclonal antibody.

H.W. Pan et al.

International Journal of Pharmaceutics 644 (2023) 123272

3 of 2HPβCD in ultrapure water according to the proportions stated in

Table 1. Antibody concentration of the stock solutions was measured by

UV absorbance at 280 nm. To prepare the antibody-containing formu­ lations, the stock antibody solution was added to the excipient-only platform and the resultant mixture was swirled gently. The final solute concentration for all feed solutions was adjusted to 2% w/v with ul­ trapure water.

The feed solutions were pumped peristaltically at a 3% rate (~0.9 mL/min) into a mini spray dryer (B-290, BÜCHI Labortechnik AG, Fla­ wil, Switzerland) operated at 100% aspiration rate (corresponding to a gas flow rate of about 35 m3/hour) with nitrogen as the spray gas at a flow rate of 742 L/hour. These operating parameters were adopted from previous studies (Pan et al., 2022; Qiu et al., 2019). The inlet temper­ ature was preset at 150 ◦C for the formulation platforms and 100 ◦C for the antibody-containing formulations. An integrated two-fluid nozzle of

0.7 mm internal diameter was used to atomise the solutions into the spray cylinder. The SD powder was transferred from the product collection vessel into a transparent glass vial and stored in an auto dry box (Eureka Dry Tech, Taipei, Taiwan) at air-conditioned temperature (around 22 ◦C) and a controlled relative humidity of approximately

25%.

2.3. Scanning electron microscopy (SEM) Morphology and geometric size of the SD particles were studied using SEM. Powder samples were sprinkled onto aluminium specimen stubs using adhesive carbon tape. To enhance sample conductivity and prevent overheating, the surface of the mounted samples was coated with approximately 13 nm of gold–palladium for 120 seconds at 30 mA in an argon-rich environment (Q150T ES Plus, Quorum Technologies,

East Sussex, UK). Photomicrographs of the particles were taken by a field emission scanning electron microscope (Hitachi S-4800, Tokyo, Japan) at 5,000 × and 10,000 × magnifications, an accelerating voltage of 5 kV, and 4.6–6.6 mm working distance.

2.4. Differential scanning calorimetry (DSC) Thermal behaviour of the SD powders was investigated by DSC.

Approximately 2–3 mg of each powder formulation was weighed into a

5.4 × 2.0 mm aluminium hermetic pan (Jingyi Chemical Materials,

Shanghai, China) and covered with a perforated lid. The pans were sealed before loading onto an indium-calibrated differential scanning calorimeter (DSC 250, TA Instruments, New Castle, DE, USA). The DSC was programmed to hold isothermal for 10 min at 0 ◦C, followed by heating at a ramp rate of 10 ◦C/min to 300 ◦C. The thermograms were plotted using OriginPro® software (version 2022b, OriginLab®,

Northampton, MA, USA).

2.5. Thermogravimetric analysis (TGA) Moisture content of the SD antibody formulations was measured using a thermogravimetric analyser (TGA 550, TA Instruments, New

Castle, DE, USA). An open platinum sample pan was loaded with 2 ± 0.5 mg of each powder formulation and heated at a constant rate of 10 ◦C/ min from ambient temperature to 105 ◦C. The weight loss as monitored by a microbalance was deemed to account for the water that had evaporated from the solid sample during heating. Each measurement was done in triplicate.

2.6. Aerosol performance The aerodynamic characteristics of the SD formulations were assessed by a Next Generation Impactor (NGI) attached to a right-angled induction port (Copley Scientific, Nottingham, UK). A size 3 gelatin capsule (Capsugel®, Morristown, NJ, USA) was filled with 10 ± 0.5 mg of powder and placed into a high-resistance handheld RS01 inhaler (Plastiape, Osnago, Italy). The air flow rate was adjusted to approxi­ mately 54 L/min to reach a pressure drop of 4 kPa across the inhaler. The powder was dispersed into the apparatus for 4.4 seconds to allow 4 litres of air to be withdrawn. The assessment was conducted thrice for each formulation.

2HPβCD concentration was quantified by high- performance liquid chromatography (Agilent Technologies, Santa

Clara, CA, USA) and peak integration was performed on OpenLab CDS

ChemStation Edition software (version C.01.06, Agilent Technologies) as previously described (Pan et al., 2022).

The aerosol performance is described by the following metrics: recovered dose (RD), emitted dose (ED), emitted fraction (EF), fine particle dose (FPD), fine particle fraction (FPF), mass median aero­ dynamic diameter (MMAD), and geometric standard deviation (GSD).

With regard to 2HPβCD, the RD is the total mass assayed from the whole

NGI set-up; the ED is the mass discharged from the inhaler; and the FPD is the mass with aerodynamic diameter (dae) < 5 µm. The formulae for

EF and FPF are furnished below. MMAD and GSD were calculated ac­ cording to the methods stipulated in USP on Compounding (The United

States Pharmacopeial Convention, 2014). MMAD is the dae at which half of the aerosolised particle mass lie under, while GSD signifies the width of the dae distribution (Finlay and Darquenne, 2020).

EF = ED RD FPF = FPD RD 2.7. Sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS- PAGE)

Electrophoresis was utilised to verify the molecular weight of the antibody and the Fab post-spray drying. Powder formulations of the antibody were reconstituted with ultrapure water and the protein con­ centration was adjusted to 0.2 mg/mL in both non-reducing and reducing (containing 5 mM DTT) SDS buffers. Unprocessed antibodies and aliquots of the feed solutions were included as reference. The sample solutions were boiled at 95 ◦C for 5 min in a dry bath, except for the non- reduced full-length mAb to minimise aggregation. Each well of the 10%

Bis-Tris gels was loaded with 4 µg of antibody. The electrophoresis (Mini-PROTEAN® Tetra System, Bio-Rad), gel staining and destaining, and photography (G:BOX Chemi XR5 gel documentation system, Syn­ gene, Cambridge, UK) steps have been outlined elsewhere (Pan et al.,

2022).

2.8. Size-exclusion chromatography (SEC) SEC separates proteins on the basis of molecular size and was employed to monitor the monomer content after spray drying. The chromatography system consisted of a Yarra™ 3 µm SEC-3000 column (phenomenex®, Torrance, CA, USA) connected to a diode array detector (Agilent Technologies) set at a UV wavelength of 214 nm. The flow rate of the mobile phase (150 mM aqueous Na3PO4, pH 6.8) was 0.8 mL/min and the stoptime was 18 min. Unprocessed antibodies were included as controls and 50 µL of each sample, adjusted with the buffer to a protein concentration of 0.2 mg/mL, was injected at 25 ◦C in triplicate. Inte­ gration of the peaks was performed using OpenLab CDS ChemStation

Edition software (version C.01.03, Agilent Technologies) and the percent monomer content was calculated.

2.9. Enzyme-linked immunosorbent assay (ELISA) The binding capacity of the antibody of selected formulations was determined by labelled immunoassay. The ELISA protocol is detailed elsewhere (Pan et al., 2022). In brief, the capture antigen, rhIL-4Rα, was coated (50 ng/well) overnight and blocked with reagent diluent (2% w/ v BSA in PBS). The protein concentration of the samples was adjusted to

100 and 1 μg/mL with the reagent diluent and the samples were added in duplicate to the wells. The peroxidase reaction was arrested by the

H.W. Pan et al.

International Journal of Pharmaceutics 644 (2023) 123272

4 addition of 2 N H2SO4. Readings at 570 nm were subtracted from those at 450 nm to correct for optical imperfections in the plate. A total of three replicates were run for the ELISA experiment. The bar chart showing the optical density values was plotted using GraphPad Prism (version 8.2.1, San Diego, CA, USA).

2.10. Cellular antiproliferation assay The inhibitory potency of the SD antibody was evaluated by an antiproliferation assay using human erythroleukaemic TF-1 cells (ATCC® CRL-2003™) in which the antibody competed with IL-4 or IL- 13 for binding to IL-4R, a shared receptor target. Inhibition of consti­ tutive growth factors, such as GM-CSF, IL-4, and IL-13, abates the long- term proliferation of TF-1 cells (Drexler et al., 1997). The cells were cultured in complete growth medium (CGM) composed of RPMI-1640 supplemented with 2.0 g/L NaHCO3, 10% heat-inactivated FBS, 1%

Antibiotic-Antimycotic, and 2 ng/mL rhGM-CSF. For the assay, cells were centrifuged at 100 × g for 10 min and resuspended in the assay medium (CGM less rhGM-CSF) at a concentration of 6.25 × 105 cells/ mL. The harvested cells were added to a flat-bottom 96-well microplate (TPP®, Trasadingen, Switzerland) at a volume of 120 μL per well and starved for 24 hours at 37 ◦C and 5% CO2. The selected antibody for­ mulations were reconstituted in ultrapure water and the solutions, including the unprocessed antibody controls, were serially diluted 3-fold in the assay medium to produce 10 antibody concentrations. Each of these was supplemented with either 30 µL of rhIL-4 or rhIL-13 to salvage the cells, such that the final reaction mixture contained an interleukin at a fixed concentration of 8 ng/mL, with the antibody at a concentration that spanned from 5 ng/mL to 100 μg/mL. The test solutions were added in duplicate and the plate was incubated for a further 48 hours. Three hours prior to the end of this period, 10 μL of CCK-8 was added to each well. At the end of the entire incubation period, the plate was mixed gently on a microplate mixer and the absorbance was read at 450 nm using a microplate spectrophotometer (Thermo Scientific Multiskan

GO). The assays were run in triplicate and data fitted with nonlinear regression to a sigmoidal equation using GraphPad Prism. The optical density values were normalised to the best-fit top (100%) and bottom (0%) values of the unprocessed antibody control for each run and plotted as a function of the common logarithm of the antibody con­ centration to obtain the half-maximal inhibitory concentration (IC50).

2.11. Inhibition of STAT6 activation TF-1 cells transfected with a STAT6-driven luciferase (Luc) reporter construct (designated TF-1/STAT6-Luc) were cultured in CGM. Between

12 and 16 hours prior to the experiment, the medium was changed to

RPMI-1640 with 5% FBS and the cells were left to incubate overnight at

37 ◦C. On the day of the experiment, 12 different concentrations of antibody (Fab: 9.41 pg/mL–1.67 µg/mL; full-length antibody: 27.8 pg/ mL–4.93 µg/mL), diluted 3-fold, were added at 50 µL per well to a 96- well microplate in duplicate. The cells were harvested and resus­ pended in RPMI-1640 with 5% FBS, and inoculated at 50 µL into each well. The plate was left in the incubator for 30 min. Thereafter, 50 µL of rhIL-4 or rhIL-13 was added to each well to achieve a final interleukin concentration of 2 (rhIL-4) or 20 ng/mL (rhIL-13) and cell density of 0.5

× 105 cells/mL. This was followed by a five-hour incubation and 50 µL of luciferase substrate was added to each well. After 10 min, the plate was read at 560 nm. The ratio of firefly:Renilla/TK luminescence in relative light units (RLU) for each well was calculated. A four-parameter regression was performed by GraphPad Prism, and the RLU-antibody concentration curves were plotted, from which the IC50 could be ob­ tained. The experiment was performed three times.

2.12. Statistical analysis All data are reported as mean ± standard deviation where applicable. Differences in optical density and IC50 values were assessed using paired two-tailed Student’s t-test on GraphPad Prism. P values <

0.05 were considered statistically significant.

3. Results 3.1. Spray drying The inlet and outlet temperatures for all 11 formulations and their processing yield are shown in Table 2. The processing yield is defined as the percentage of powder weight collected over the total solute mass in the feed solution. The inlet temperature was lowered for the antibody formulations to minimise heat stress. The outlet temperature ranged from 98 to 100 ◦C for the excipient platforms, and from 62 to 63 ◦C for the antibody formulations. Among the excipient platforms, the addition of leucine, even at 5% w/w, increased the production from approxi­ mately 46% to above 60%. The yields for the antibody formulations were higher, between 73% and 84%.

3.2. Particle morphology Representative scanning electron photographs of the SD particles are shown in Fig. 1. The particle morphology exhibited striking variations as the leucine content is increased. At low leucine concentrations of 5% and 10% w/w (E1 and E2), the particles were similar to 100% SD

2HPβCD (E0). However, at higher leucine concentrations, E3 was highly creased with interparticle fibre-like features on a rough surface, whereas the globoid structure was largely absent in E4 and was instead replaced by hollow particles that were heterogeneously irregular in shape. At a high antibody concentration of 50% w/w (F3 and M3), the particles were more globular in shape, with multiple dimple-like features on a rough surface. In contrast, the formulations with lower antibody con­ centrations (F1,2 and M1,2) generally resembled E4; the microparticles manifested a collapsed hollow sphere morphology by enfolding upon themselves. The surface of these particles also appears more rugose.

Most of the visualised particles were smaller than 5 µm in diameter.

3.3. Thermal analyses The DSC thermograms of the leucine-containing powder formula­ tions (all except E0) are distinct from that of the 2HPβCD-only formu­ lation (E0) in respect of the additional endothermic peaks at temperatures between 230 and 270 ◦C, due to the presence of crystalline leucine (Fig. 2). These peaks shift rightwards, closer to the known melting point of leucine (~287 ◦C), and become more pronounced, as the leucine component begins to dominate in the formulations (Yal­ kowsky et al., 2010). The existence of residual water in the powders is evident in the other downward endothermic peaks below 100 ◦C that are much broader. Although this peak is present in all the formulations, it is slightly more prominent in F3 and M3, which contained a lower con­ centration of the hydrophobic leucine.

The residual water content of the antibody formulations was deter­ mined by TGA after spray drying (Table 2). The Fab formulations con­ tained marginally more moisture compared with the mAb formulations, and the water content was highest in the formulations that had the lowest concentration of leucine (i.e., F3 and M3). Nevertheless, the mean water content of all the powders tested was low and ranged be­ tween 1.4 ± 0.2% and 2.6 ± 0.3% w/w, demonstrating the drying ef­ ficiency of the formulation and process to keep moisture at bay.

3.4. Aerosol performance The EF and FPF obtained from the cascade impactor experiments, representing the dispersibility and respirable fraction of a powder formulation, respectively, are shown in Fig. 3. For the excipient plat­ forms, the EF rises from around 54% to 81% as the relative leucine

H.W. Pan et al.

International Journal of Pharmaceutics 644 (2023) 123272

5 content is gradually increased. Notably, formulation E4 had the highest

EF as well as FPF of nearly 70%, therefore a 1:1 leucine-to-2HPβCD ratio was chosen as the excipient foundation for the antibody formulations.

The mean EF of all six antibody formulations was above 80%. Apart from M3, which had an FPF of about 61%, the mean FPF of the rest of the antibody formulations was between 72% and 75%. The best aerosoli­ sation characteristics belonged to the formulations containing 25% antibody and 75% excipients, i.e., F2 (EF, 86.5 ± 1.7%; FPF, 75.3 ±

2.7%) and M2 (EF, 85.6 ± 2.4%; FPF, 75.2 ± 4.4%).

The MMAD and GSD values calculated from the aerosolisation per­ formance analysis are also displayed in Table 2. With the exception of formulation E3, the dae of the particles was within the critical size range of 1 to 3 µm that is optimal for pulmonary deposition (Chow et al., 2007;

Malcolmson and Embleton, 1998), while the GSD values (≥1.22)

Table 2 Spray-drying outcomes: outlet temperature, processing yield, mass median aerodynamic diameter (MMAD), geometric standard deviation (GSD), and residual moisture content (RMC).

Formulation Inlet temperature (◦C) Outlet temperature (◦C)

Processing yield (%) MMAD (µm) GSD RMC (% w/w) Excipient-only formulation platforms

E0 150 100 46.4 1.83 2.09 – E1 150 99 60.4 2.04 2.26

– E2 150 99 61.6 2.77 1.94 – E3 150 98 71.9 4.19 2.07

– E4 150 98 67.5 1.64 2.34 – Antibody-containing formulations

F1 100 62 73.0 1.46 2.33 2.1 ± 0.3 F2 100 62 78.0 1.68

2.24 1.9 ± 0.3 F3 100 62 83.3 1.95 2.16 2.6 ± 0.3 M1

100 63 73.7 1.50 2.30 1.4 ± 0.2 M2 100 63 76.4 1.59

2.24 1.8 ± 0.1 M3 100 63 84.1 2.38 2.17 2.0 ± 0.3 Fig. 1. Scanning electron microscopy (SEM) images of the (a) excipient-only formulations and (b) antibody-containing formulations. Scale bar = 5 µm.

H.W. Pan et al.

International Journal of Pharmaceutics 644 (2023) 123272

6 indicate a heterodisperse distribution, which is typical for therapeutic aerosols (Labiris and Dolovich, 2003; Laube et al., 2011). In the anti­ body formulations, raising the protein-to-excipient ratio increased the

MMAD, which appears to have a moderate yield-enhancing effect. This is probably related to the separation efficiency of the cyclone, as smaller particles are more difficult to be segregated from the gas stream (Prinn et al., 2002).

Fig. 2. Differential scanning calorimetry (DSC) thermograms of the (a) excipient platforms and (b) antibody formulations.

Fig. 3. Aerosolisation performance of the excipient platforms and antibody formulations assessed using a Next Generation Impactor (NGI), n = 3. EF: emitted fraction; FPF: fine particle fraction (cutoff aerodynamic diameter < 5 µm).

Fig. 4. SDS-PAGE gel images of the antibody formulations. DTT: dithiothreitol; Fab: antigen-binding fragment; mAb: full-length monoclonal antibody; PM: physical mixture (feed solution); SD: spray-dried; up: unprocessed.

H.W. Pan et al.

International Journal of Pharmaceutics 644 (2023) 123272

7 3.5. Antibody stability Electrophoretic protein separation was used to study the structural integrity of the antibody after spray drying, given that the high tem­ peratures involved in the process could potentially cause thermal degradation, leading to fragmentation (Vlasak and Ionescu, 2011). The gel in the left panel of Fig. 4 shows that the Fab migrated under non- reducing conditions with an apparent molecular weight of ~45 kDa, while a thick band between 115 and 185 kDa was observed for the intact mAb samples. The additional lighter bands at around 20 kDa in the Fab samples were probably due to an extraneous protein contaminant in the stock antibody solution. The disulfide bonds of the antibodies on the gel in the right panel was reduced, which gave rise to a band at approxi­ mately 25 kDa for the Fab formulations (F1–3), and two primary bands at about 50 and 25 kDa for the mAb formulations (M1–3). Importantly, the patterns of the bands of the SD samples are essentially the same as those of the untreated antibody, indicating that there is no evidence of heat-induced fragmentation as a result of the spray-drying process (Nowak et al., 2017).

Stresses during spray drying and storage can destabilise proteins and induce aggregation, which remains a huge challenge in the development of biotechnology products as it can lead to immunogenicity and other adverse effects, batch-to-batch variability, and impaired efficacy (Lowe et al., 2011; Wang et al., 2010). To monitor antibody aggregation, the monomer content of the SD formulations was measured using SEC at two time points, 1 week and 10 months post-spray drying (Fig. 5). Monomer content is a crucial parameter of physical stability as unfolded proteins, even partially, are inherently aggregation-prone (Wu et al., 2014) and this has implications in the safety and efficacy of antibody-based phar­ maceutical products (Hickey and Stewart, 2022). The unprocessed antibody was included at the first time point to discern aggregation attributable to the drying process. One week after spray drying, the average monomer content of the Fab formulations was close to 100% (99.8–99.9% vs. Fab-up 99.8%). For the full-length mAb formulations, the difference in monomer content between SD and unprocessed anti­ bodies was small (86.0–99.0% vs. 93.6%, respectively). At 10 months post-spray drying, the largest decrement in monomer content was detected in F3 (99.9% to 90.0%), whereas the other formulations registered reductions of no more than 2.8%. This is fairly remarkable considering the samples were not frozen or stored under refrigeration throughout the study.

3.6. Antigen-binding ability and inhibitory potency

Two representative formulations, one each from the Fab and intact mAb formulations, advanced to in vitro experiments for assessment of the biological activity of the SD antibodies relative to their respective unprocessed (up) counterparts. Among the six antibody formulations, F2 and M2 were chosen in view of their better aerosol performance, since the other characteristics, such as particle morphology, residual water content, and protein structural integrity, did not demonstrate superior­ ity of any particular formulation. The binding ability of the antibody to plate-coated rhIL-4Rα was assessed by ELISA (Fig. 6) at two empirically selected antibody concentrations, 1 and 100 µg/mL. There was no sta­ tistically significant difference between the Fab-up and F2 (1 µg/mL: p

= 0.4815; 100 µg/mL: p = 0.9904) or between mAb-up and M2 (1 µg/ mL: p = 0.5810; 100 µg/mL: p = 0.1549).

Given that the antibody interferes with the binding of IL-4 and IL-13 to IL-4R, the antiproliferation assay used TF-1 cells that have an absolute dependence on specific growth factors for survival (Drexler et al., 1997) to compare the inhibitory activity of the SD antibody as IC50 with that of the unprocessed (up) antibody. The TF-1 cells were separately exposed to rhIL-4 and rhIL-13, and the antibody concentration–response curves are shown respectively in the top and bottom panels of Fig. 7. In each of the four graphs, the two curves largely overlap each other, underscoring the almost identical inhibitory profiles of the tested antibody samples. In the presence of rhIL-4, the IC50 values were 0.534 ± 0.279 µg/mL (Fab- up), 0.535 ± 0.336 µg/mL (F2), 0.964 ± 0.458 µg/mL (mAb-up), and

1.262 ± 0.248 µg/mL (M2). There was no statistically significant dif­ ference in the IC50 between the SD antibody and the unprocessed anti­ body (F2 vs. Fab-up, p = 0.9861; M2 vs. mAb-up, p = 0.1465). With rhIL- 13, the IC50 values were 0.081 ± 0.006 µg/mL (Fab-up), 0.082 ± 0.024 µg/mL (F2), 0.123 ± 0.055 µg/mL (mAb-up), and 0.134 ± 0.030 µg/mL

Fig. 5. Antibody monomer content of antibody formulations at 1 week and 10 months post-spray drying quantified by size-exclusion chromatography (SEC), n = 3.

Fab: antigen-binding fragment; mAb: full-length monoclonal antibody; up: unprocessed.

Fig. 6. Antigen-binding ability of the antibody in the selected Fab (F2) and mAb (M2) formulations determined by ELISA (n = 3). 96-well plates were coated with rhIL-4Rα. Fab: antigen-binding fragment; mAb: full-length mono­ clonal antibody; up: unprocessed.

H.W. Pan et al.

International Journal of Pharmaceutics 644 (2023) 123272

8 (M2). Again, no statistically significant difference was found between the formulated antibody and the untreated antibody control (F2 vs. Fab- up, p = 0.9822; M2 vs. mAb-up, p = 0.5696).

The TF-1 cells used to study the inhibitory effect of the antibody on

STAT6 signalling pathway (Fig. 8) were stably transfected to express a luciferase reporter gene regulated by a STAT6-responsive promoter.

With rhIL-4-induced STAT6 activation, the IC50 values were 9.79 ±

0.34 ng/mL (Fab-up), 8.75 ± 0.41 ng/mL (F2), 15.93 ± 1.89 ng/mL (mAb-up), and 14.97 ± 0.62 ng/mL (M2). The difference in IC50 be­ tween the formulated and untreated antibodies was not statistically significant (F2 vs. Fab-up, p = 0.0589; M2 vs. mAb-up, p = 0.3229).

When induced by rhIL-13, the IC50 values were 9.79 ± 0.21 ng/mL (Fab- up), 9.60 ± 0.55 ng/mL (F2), 15.09 ± 3.28 ng/mL (mAb-up), and 12.32

± 1.82 ng/mL (M2). Similarly, there was no statistically significant difference in the IC50 between the SD antibody and the unprocessed antibody (F2 vs. Fab-up, p = 0.5707; M2 vs. mAb-up, p = 0.4350). The

ELISA results, cellular antiproliferation assay, and the inhibition of

STAT6 activity effectively illustrate that the antigen-binding capacity and inhibitory potency of both SD Fab and full-length mAb were retained relative to their initial state before spray drying.

Fig. 7. Concentration-response curves of the inhibitory effect of the antibody in the selected Fab (F2) and intact mAb (M2) formulations on the proliferation of TF-1 cells (n = 3) stimulated by (a) rhIL-4 or (b) rhIL-13. The optical density is normalised with respect to the best-fit top (100%) and bottom (0%) values of the un­ processed antibody curve in the individual runs. [Ab]: concentration of antibody; Fab: antigen-binding fragment; mAb: full-length monoclonal antibody; up: unprocessed.

Fig. 8. Concentration-response curves of the inhibitory effect of the antibody in the selected Fab (F2) and intact mAb (M2) formulations on the activation of the

STAT6 signalling pathway by (a) rhIL-4 or (b) rhIL-13 in TF-1/STAT6-Luc cells (n = 3). [Ab]: concentration of antibody; Fab: antigen-binding fragment; mAb: full- length monoclonal antibody; up: unprocessed.

H.W. Pan et al.

International Journal of Pharmaceutics 644 (2023) 123272

9 4. Discussion In our previous work, the anti-IL-4Rα mAb was spray-dried and spray-freeze-dried with 2HPβCD as the protein stabiliser (Pan et al.,

2022). Although the antibody was decently stabilised and retained its antigen-binding ability and in vitro biological activity, powder dis­ persibility was suboptimal, plausibly impaired by the high residual water content (7–9% w/w) detected. Spray drying at a higher temper­ ature might ameliorate this, since the relative humidity of the drying gas would be increased, leading to a dryer product. Particles with lower moisture content are less likely to be fused together by solid bridges to form larger agglomerates (Li et al., 2016). Nonetheless, excessively high temperatures should be avoided in preparations containing heat- sensitive biologics. Not only is moisture suspected to be a factor affecting the aerosolisation of powder formulations, water can also act as a plasticiser to potently lower the glass transition temperature of the excipient and increase local mobility of the embedded bio­ macromolecules, which destabilises the protein (Mensink et al., 2017).

To enhance drying efficiency and circumvent moisture adsorption, leucine, a hydrophobic amino acid, was co-formulated at different weight ratios with 2HPβCD to investigate whether its incorporation would improve the aerodynamic characteristics of the powders compared with 2HPβCD alone.

Numerous studies have examined amino acid-based multi-compo­ nent SD powder preparations for respiratory administration. They encompassed excipient-only platforms (Mangal et al., 2015; Sou et al.,

2013) and formulations containing small-molecule drugs (Mah et al.,

2019; Seville et al., 2007), small interfering RNA (Chow et al., 2017; Xu et al., 2022), antibodies (Faghihi et al., 2019; Shepard et al., 2021), and vaccines (Gomez et al., 2021; Lovalenti et al., 2016). However, in these studies, the co-excipients applied together with amino acids were pri­ marily trehalose and mannitol, and none included any cyclodextrin. As such, 2HPβCD plus leucine was a novel combination in this regard.

Mannitol crystallises easily (Eedara et al., 2021), which is detrimental to the biomacromolecule due to the combined effects of shear stresses and loss of protein–sugar interactions that are important for protein stabi­ lisation (Izutsu et al., 1994; Mensink et al., 2017). Unlike mannitol, cyclodextrins tend to form amorphous glasses (Serno et al., 2011). Be­ sides, it is usual for spray drying to generate amorphous glassy matrix (Chow et al., 2007), an essential aspect of the vitrification mechanism of protein stabilisation by sugars in the solid state (Mensink et al., 2017).

The first recognisable benefit of including leucine in the formulations was the increase in processing yield, a vital consideration in the scale-up of biopharmaceutical manufacturing (Hernandez, 2016). Although on an industrial level spray drying can attain close to 100% yield, some powder adhesion and accumulation on the interior of the cyclone was visibly noted here, which could explain for the loss of product (Bowey et al., 2013).

The value of leucine in the formulations was clearly in enhancing the aerosolisation of the powder formulations. The mechanisms through which leucine promotes dispersion are multifaceted, from keeping water out to modifying the surface composition and morphology of the SD particles (Alhajj et al., 2021). It has been suggested that powder flow­ ability or dispersibility is influenced by surface corrugation in two ways, altering the contact area between particles and mechanical interlocking or entrapment of adjacent particles (Chew and Chan, 2001; Walton and

Mumford, 1999b). From the SEM images, leucine plus 2HPβCD yielded surface-corrugated microparticles, and the extent of corrugation and collapse of the outer shells increases with the leucine content. These morphological changes were consistent with observations in other studies that employed leucine as a dispersion enhancer (Mangal et al.,

2015; Sou et al., 2013). The higher EF and FPF exhibited by E4 were perhaps ascribed to the high surface rugosity of the irregularly-shaped particles which reduced interparticle contact. Furthermore, the folded and hollow structure of the particles secondary to the inflation, rupture, and collapse of the crystalline crusts formed during the initial phase of drying lowered the density and thus the MMAD (Walton and Mumford,

1999a). However, overcorrugation can cause smaller particles to become embedded in the grooves of larger particles (Sou et al., 2013), and this effect may have ramifications for aerosolisation performance as the interparticulate cohesive forces intensify, inducing agglomeration, and the particles encounter more resistive airflow, which impacts powder fluidisation (Cui et al., 2018).

Another possible explanation for the improved aerosolisation by the addition of leucine into the formulation could be its ability to regulate water content in the powder by anti-hygroscopic effect. Even with the inclusion of a protein up to 50% w/w in the formulation, the residual moisture content of the assessed powders was no greater than 3% by weight. Such aridity is enough to inhibit crystallisation of the amor­ phous components and confer excellent powder flowability (Mah et al.,

2019). Bevacizumab was co-spray-dried with trehalose and leucine (at

40%, 40%, 20% w/w, respectively) for pulmonary delivery in a study and the measured moisture content was between 3% and 4%, while the

FPF (cutoff dae < 5 µm) was 82% (Shepard et al., 2021). Notwith­ standing the somewhat different formulation compositions and experi­ mental conditions, these results were comparable to those obtained here. There are two projected moisture-protective mechanisms, the crystallinity of leucine and its enrichment on the particle surface. With a sufficient concentration of leucine in the feed solution, e.g., greater than

10% w/w, leucine should reach supersaturation early in the spray- drying process and crystallise, considering that it is sparingly soluble in water (~22 mg/mL at room temperature) (Yalkowsky et al., 2010) and migrates to the droplet surface (Mangal et al., 2015; Vehring, 2008).

The moisture uptake by crystalline leucine on SD particle surface is limited (Li et al., 2016). The accumulation of leucine on the surface of

SD particles quantified by X-ray photoelectron spectroscopy has previ­ ously been reported (Mah et al., 2019; Mangal et al., 2015; Xu et al.,

2022). The chemical structure of a leucine molecule comprises distinct hydrophilic (amino and carboxylic acid groups) and lipophilic (aliphatic isobutyl side chain) moieties (Chow et al., 2017). This amphiphilicity of leucine predisposes its accretion on the particle surface (Weissbuch et al., 1990). Protein molecules too display the tendency to amass at air- solution interfaces due to their surface-active nature, however, their much larger molecular weight retards diffusion towards the core (Adler et al., 2000; Grasmeijer et al., 2016). Hence, it was hypothesised that leucine would compete with the antibody for accumulation on the sur­ face, decreasing the relative amount of proteins there. This not only enhances protein stability, since adsorption at interfaces promotes unfolding and aggregation of the protein (Wang et al., 2010), but also increases the level of powder dryness.

In the context of dehydrated biopharmaceuticals, a higher level of dryness is usually preferred, however, this is not always the case, as established in studies involving lyophilised excipient-free tissue-type plasminogen activator (Hsu et al., 1992) and plasma-derived IgG with sucrose (Duralliu et al., 2020). Beyond a certain threshold, denaturation resumes, therefore, it can be interpreted that an optimum residual moisture content may exist for each protein. Over-drying (e.g., < 1% water content) can transpire even in the presence of stabilising sugar excipients, especially when large rigid polysaccharides, for instance, some cyclodextrins, are used. This is perhaps due to inefficient hydrogen bonding, which makes local mobility more facile (Mensink et al., 2017).

Moreover, electrostatic charges on particle surfaces may build up in extreme dryness (Elajnaf et al., 2006), facilitating agglomeration and could seriously hinder powder fluidisation (Kaialy, 2016). Nevertheless, based on the physical stability observed in the gel electrophoretic and filtration analyses, the SD antibody formulations here did not appear to be overdried.

The binding ability of the antibody to IL-4Rα was investigated using

ELISA and found to be tantamount to that of the unformulated antibody.

Likewise, in the antiproliferation assays, the cytokine rescue of the starved TF-1 cells that were bereft of essential growth factors was inhibited in a concentration-dependent manner to similar extents by the

H.W. Pan et al.

International Journal of Pharmaceutics 644 (2023) 123272

10 formulated and unprocessed antibodies. These results were construed as adequate stabilisation of the SD antibody. The smaller IC50 values observed in the rhIL-13 assays compared to those in the rhIL-4 could be explained by the existence of an additional signalling pathway, inde­ pendent of STAT6 activation, for the IL-4/IL-4Rα complex to produce growth factors that stimulate TF-1 proliferation. Indeed, this complex recruits the common γc subunit to activate gene transcription through a downstream signalling adaptor molecule, insulin receptor substrate-2 (McCormick and Heller, 2015). At the single cytokine concentration (8 ng/mL) added to the cells, the stimulatory effect was saturated for IL- 4, but not for IL-13, which mediated roughly 60% maximal effect. This could presumably be rationalised by the different binding affinities and proliferative responses reported for IL-4 and IL-13. IL-4 binds to IL-4Rα with a much higher affinity than IL-13 to IL-13Rα1 (LaPorte et al.,

2008). The stimulation index of IL-4 is also higher than that of IL-13 (Drexler et al., 1997). In addition, the number of binding sites on TF-1 cells for IL-4 is double that for IL-13 (Lefort et al., 1995). Consequen­ tially, the EC50 is higher for IL-13 compared with IL-4.

The active ingredients presented in this study included not only a full-length mAb, but also the Fab. In local inhaled therapy, larger bio­ macromolecules may be preferable since unintended absorption from the lung (Patton, 1996) and degradation by lung proteases (Fr¨ohlich and

Salar-Behzadi, 2021) are both inversely related to molecular weight.

However, utilising just the Fab without the crystallisable fragment (Fc) region that is present in the full-length mAb is logical given that the antibody mechanism of action relies on the Fab engaging with IL-4Rα for therapeutic efficacy in asthma. Furthermore, the role of the Fc domain is less pertinent in inhaled therapy where systemic absorption and extended serum half-life is not sought (Chow et al., 2023). Future work might be pursued in two directions. Firstly, the in vivo biological activity of the SD antibody in animal models should be explored, in conjunction with pulmonary toxicity assessment of the excipients in the respiratory tract. Secondly, more comprehensive characterisation of the powder formulations, for example, in relation to the protein-excipient in­ teractions (Chen et al., 2021), antibody-antigen binding kinetics (Yang et al., 2017), and surface area-porosity analysis (Ji et al., 2016), could be undertaken. In spite of their medical utility and applications in phar­ maceutical research, certain relevant physicochemical properties of cyclodextrins, in particular, hygroscopicity and density, remain ambig­ uous (Day et al., 2020), thus a better understanding of these properties would also be useful.

5. Conclusions This study co-formulated 2HPβCD with leucine to produce a dual- excipient platform by spray drying for pulmonary delivery of solid- state biomacromolecules. The incorporation of leucine into the formu­ lation boosted the processing yield and aerosol performance, with a 1:1 weight ratio of both excipients exhibiting the most favourable outcomes among the tested powders. The improvement in powder aerosolisation was imputed to particle surface and morphology modifications as well as water repellent by leucine. Spray-dried powders of an anti-IL-4Rα mAb and its Fab founded upon this excipient ratio emulated the physico­ chemical and aerodynamic characteristics of their formulation platform predecessor. The moisture content of the powder antibody formulations was within an acceptable range of between 1% and 3% w/w, while only minor changes in antibody monomer content over a period of 10 months at ambient storage conditions were detected. The in vitro antigen- binding capability and cellular inhibitory potency of the dehydrated antibody were comparatively well-preserved. Collectively, the results here validate the feasibility of a successful inhaled biologic product for the local treatment of severe asthma.

CRediT authorship contribution statement Harry W. Pan: Conceptualization, Methodology, Validation, Formal analysis, Investigation, Writing – original draft, Writing – review & editing, Visualization. Jinlin Guo: Investigation. Lingqiao Zhu: Writing

– review & editing. Susan W.S. Leung: Writing – review & editing.

Chenghai Zhang: Supervision. Jenny K.W. Lam: Conceptualization,

Methodology, Resources, Writing – review & editing, Supervision,

Project administration, Funding acquisition.

Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability Data will be made available on request.

Acknowledgements This project was funded by the Seed Funding for Strategic Interdis­ ciplinary Research Scheme, The University of Hong Kong (HKU). H.W.P. is a recipient of the Hong Kong PhD Fellowship, Research Grants

Council, Hong Kong (PF18-13277). The authors thank Mr Ray Lee and

Ms Xinyue Zhang (Department of Pharmacology and Pharmacy, HKU) for their kind assistance with the TGA measurements.

References Adler, M., Unger, M., Lee, G., 2000. Surface Composition of Spray-Dried Particles of

Bovine Serum Albumin/Trehalose/Surfactant. Pharm Res 17, 863–870.

Alhajj, N., O’Reilly, N.J., Cathcart, H., 2021. Leucine as an excipient in spray dried powder for inhalation. Drug Discov Today 26, 2384–2396.

Al-Nakib, W., Higgins, P.G., Barrow, G.I., Tyrrell, D.A., Andries, K., Bussche, G.V.,

Taylor, N., Janssen, P.A., 1989. Suppression of colds in human volunteers challenged with rhinovirus by a new synthetic drug (R61837). Antimicrob Agents Chemother

33, 522–525.

Borghardt, J.M., Kloft, C., Sharma, A., 2018. Inhaled therapy in respiratory disease: The complex interplay of pulmonary kinetic processes. Can Respir J 2018, 2732017.

Bowey, K., Swift, B.E., Flynn, L.E., Neufeld, R.J., 2013. Characterization of biologically active insulin-loaded alginate microparticles prepared by spray drying. Drug Dev Ind

Pharm 39, 457–465.

Chang, R.Y.K., Chow, M.Y.T., Khanal, D., Chen, D., Chan, H.-K., 2021. Dry powder pharmaceutical biologics for inhalation therapy. Adv Drug Deliv Rev 172, 64–79.

Chang, L., Pikal, M.J., 2009. Mechanisms of protein stabilization in the solid state.

J Pharm Sci 98, 2886–2908.

Chen, Y., Ling, J., Li, M., Su, Y., Arte, K.S., Mutukuri, T.T., Taylor, L.S., Munson, E.J.,

Topp, E.M., Zhou, Q.T., 2021. Understanding the impact of protein-excipient interactions on physical stability of spray-dried protein solids. Mol Pharm 18,

2657–2668.

Chew, N.Y.K., Chan, H.-K., 2001. Use of solid corrugated particles to enhance powder aerosol performance. Pharm Res 18, 1570–1577.

Chow, M.Y.T., Qiu, Y., Lo, F.F.K., Lin, H.H.S., Chan, H.-K., Kwok, P.C.L., Lam, J.K.W.,

2017. Inhaled powder formulation of naked siRNA using spray drying technology with ʟ-leucine as dispersion enhancer. Int J Pharm 530, 40–52.

Chow, M.Y.T., Pan, H.W., Seow, H.C., Lam, J.K.W., 2023. Inhalable neutralizing antibodies – promising approach to combating respiratory viral infections. Trends

Pharmacol Sci 44, 85–97.

Chow, A.H.L., Tong, H.H.Y., Chattopadhyay, P., Shekunov, B.Y., 2007. Particle

Engineering for Pulmonary Drug Delivery. Pharm Res 24, 411–437.

Chung, K.F., Wenzel, S.E., Brozek, J.L., Bush, A., Castro, M., Sterk, P.J., Adcock, I.M.,

Bateman, E.D., Bel, E.H., Bleecker, E.R., Boulet, L.-P., Brightling, C., Chanez, P.,

Dahlen, S.-E., Djukanovic, R., Frey, U., Gaga, M., Gibson, P., Hamid, Q., Jarjour, N.

N., Mauad, T., Sorkness, R.L., Teague, W.G., 2014. International ERS/ATS guidelines on definition, evaluation and treatment of severe asthma. Eur Respir J 43, 343–373.

Clark, A.R., Weers, J.G., Dhand, R., 2020. The confusing world of dry powder inhalers: it is all about inspiratory pressures, not inspiratory flow rates. J Aerosol Med Pulm

Drug Deliv 33, 1–11.

Cui, Y., Zhang, X., Wang, W., Huang, Z., Zhao, Z., Wang, G., Cai, S., Jing, H., Huang, Y.,

Pan, X., Wu, C., 2018. Moisture-Resistant Co-Spray-Dried Netilmicin with ʟ-Leucine as Dry powder inhalation for the treatment of respiratory infections. Pharmaceutics

10, 252.

Day, C.P.F., Miloserdov, A., Wildish-Jones, K., Pearson, E., Carruthers, A.E., 2020.

Quantifying the hygroscopic properties of cyclodextrin containing aerosol for drug delivery to the lungs. Phys Chem Chem Phys 22, 11327–11336.

Depreter, F., Pilcer, G., Amighi, K., 2013. Inhaled proteins: Challenges and perspectives.

Int J Pharm 447, 251–280.

H.W. Pan et al.

International Journal of Pharmaceutics 644 (2023) 123272

11 Drexler, H.G., Zaborski, M., Quentmeier, H., 1997. Cytokine response profiles of human myeloid factor-dependent leukemia cell lines. Leukemia 11, 701–708.

Duralliu, A., Matejtschuk, P., Stickings, P., Hassall, L., Tierney, R., Williams, D.R., 2020.

The Influence of moisture content and temperature on the long-term storage stability of freeze-dried high concentration immunoglobulin G (IgG). Pharmaceutics 12, 303.

Eedara, B.B., Alabsi, W., Encinas-Basurto, D., Polt, R., Mansour, H.M., 2021. Spray-dried inhalable powder formulations of therapeutic proteins and peptides. AAPS

PharmSciTech 22, 185.

Elajnaf, A., Carter, P., Rowley, G., 2006. Electrostatic characterisation of inhaled powders: Effect of contact surface and relative humidity. Eur J Pharm Sci 29,

375–384.

Evrard, B., Bertholet, P., Gueders, M., Flament, M.P., Piel, G., Delattre, L., Gayot, A.,

Leterme, P., Foidart, J.M., Cataldo, D., 2004. Cyclodextrins as a potential carrier in drug nebulization. J Control Release 96, 403–410.

Faghihi, H., Najafabadi, A.R., Daman, Z., Ghasemian, E., Montazeri, H., Vatanara, A.,

2019. Respiratory administration of infliximab dry powder for local suppression of inflammation. AAPS PharmSciTech 20, 128.

Finlay, W.H., Darquenne, C., 2020. Particle Size Distributions. J Aerosol Med Pulm Drug

Deliv 33, 178–180.

Fr¨ohlich, E., Salar-Behzadi, S., 2021. Oral inhalation for delivery of proteins and peptides to the lungs. Eur J Pharm Biopharm 163, 198–211.

Global Initiative for Asthma, 2022. Global Strategy for Asthma Management and

Prevention. Global Initiative for Asthma (GINA), Fontana, WI.

Gomez, M., McCollum, J., Wang, H., Bachchhav, S., Tetreau, I., Gerhardt, A., Press, C.,

Kramer, R.M., Fox, C.B., Vehring, R., 2021. Evaluation of the stability of a spray- dried tuberculosis vaccine candidate designed for dry powder respiratory delivery.

Vaccine 39, 5025–5036.

Grasmeijer, N., Frijlink, H.W., Hinrichs, W.L.J., 2016. Model to predict inhomogeneous protein–sugar distribution in powders prepared by spray drying. J Aerosol Sci 101,

22–33.

Hernandez, R., 2016. Scale-Up of Complex Biologics. BioPharm Int 29, 26–30.

Hickey, A.J., Stewart, I.E., 2022. Inhaled antibodies: Quality and performance considerations. Hum Vaccin Immunother 18, 1940650.

Hsu, C.C., Ward, C.A., Pearlman, R., Nguyen, H.M., Yeung, D.A., Curley, J.G., 1992.

Determining the optimum residual moisture in lyophilized protein pharmaceuticals.

Dev Biol Stand 74, 255–270.

Institute for Health Metrics and Evaluation, 2020. GBD 2019 Cause and Risk Summary:

Asthma—Level 3 cause, 17 Oct 2020 ed. Institute for Health Metrics and Evaluation (IHME), Seattle, WA, pp. S108-109. https://www.healthdata.org/results/gbd_s ummaries/2019/asthma-level-3-cause.

Irvine, D.J., Su, X., Kwong, B., 2013. Routes of Delivery for Biological Drug Products. In:

Gad, S.C. (Ed.), Pharmaceutical Sciences Encyclopedia. John Wiley & Sons Inc,

Hoboken, NJ, pp. 1–48.

Israel, E., Reddel, H.K., 2017. Severe and difficult-to-treat asthma in adults. N Engl J Med

377, 965–976.

Izutsu, K.-I., Yoshioka, S., Terao, T., 1994. Effect of mannitol crystallinity on the stabilization of enzymes during freeze-drying. Chem Pharm Bull 42, 5–8.

Jansook, P., Ogawa, N., Loftsson, T., 2018. Cyclodextrins: structure, physicochemical properties and pharmaceutical applications. Int J Pharm 535, 272–284.

Ji, S., Thulstrup, P.W., Mu, H., Hansen, S.H., van de Weert, M., Rantanen, J., Yang, M.,

2016. Effect of ethanol as a co-solvent on the aerosol performance and stability of spray-dried lysozyme. Int J Pharm 513, 175–182.

Kaialy, W., 2016. A review of factors affecting electrostatic charging of pharmaceuticals and adhesive mixtures for inhalation. Int J Pharm 503, 262–276.

Labiris, N.R., Dolovich, M.B., 2003. Pulmonary drug delivery. Part I: Physiological factors affecting therapeutic effectiveness of aerosolized medications. Br J Clin

Pharmacol 56, 588–599.

Lai, M.C., Topp, E.M., 1999. Solid-state chemical stability of proteins and peptides.

J Pharm Sci 88, 489–500.

LaPorte, S.L., Juo, Z.S., Vaclavikova, J., Colf, L.A., Qi, X., Heller, N.M., Keegan, A.D.,

Garcia, K.C., 2008. Molecular and Structural Basis of cytokine receptor pleiotropy in the interleukin-4/13 system. Cell 132, 259–272.

Laube, B.L., Janssens, H.M., de Jongh, F.H.C., Devadason, S.G., Dhand, R., Diot, P.,

Everard, M.L., Horvath, I., Navalesi, P., Voshaar, T., Chrystyn, H., 2011. What the pulmonary specialist should know about the new inhalation therapies. Eur Respir J

37, 1308–1417.

Lefebvre, P., Duh, M.S., Lafeuille, M.-H., Gozalo, L., Desai, U., Robitaille, M.-N.,

Albers, F., Yancey, S., Ortega, H., Forshag, M., Lin, X., Dalal, A.A., 2015. Acute and chronic systemic corticosteroid-related complications in patients with severe asthma. J Allergy Clin Immunol 136, 1488–1495.

Lefort, S., Vita, N., Reeb, R., Caput, D., Ferrara, P., 1995. IL-13 and IL-4 share signal transduction elements as well as receptor components in TF-1 cells. FEBS Lett 366,

122–126.

Li, L., Sun, S., Parumasivam, T., Denman, J.A., Gengenbach, T., Tang, P., Mao, S.,

Chan, H.-K., 2016. ʟ-Leucine as an excipient against moisture on in vitro aerosolization performances of highly hygroscopic spray-dried powders. Eur J

Pharm Biopharm 102, 132–141.

Liang, W., Pan, H.W., Vllasaliu, D., Lam, J.K.W., 2020. Pulmonary delivery of biological drugs. Pharmaceutics 12, 1025.

Lovalenti, P.M., Anderl, J., Yee, L., Nguyen, V., Ghavami, B., Ohtake, S., Saxena, A.,

Voss, T., Truong-Le, V., 2016. Stabilization of live attenuated influenza vaccines by freeze drying, spray drying, and foam drying. Pharm Res 33, 1144–1160.

Lowe, D., Dudgeon, K., Rouet, R., Schofield, P., Jermutus, L., Christ, D., 2011.

Aggregation, stability, and formulation of human antibody therapeutics. In:

Donev, R. (Ed.), Advances in Protein Chemistry and Structural Biology. Academic

Press, Oxford, UK, pp. 41–61.

Mah, P.T., O’Connell, P., Focaroli, S., Lundy, R., O’Mahony, T.F., Hastedt, J.E., Gitlin, I.,

Oscarson, S., Fahy, J.V., Healy, A.M., 2019. The use of hydrophobic amino acids in protecting spray dried trehalose formulations against moisture-induced changes. Eur

J Pharm Biopharm 144, 139–153.

Malcolmson, R.J., Embleton, J.K., 1998. Dry powder formulations for pulmonary delivery. Pharm Sci Technol Today 1, 394–398.

Mangal, S., Meiser, F., Tan, G., Gengenbach, T., Denman, J., Rowles, M.R., Larson, I.,

Morton, D.A.V., 2015. Relationship between surface concentration of ʟ-leucine and bulk powder properties in spray dried formulations. Eur J Pharm Biopharm 94,

160–169.

Matthews, A.A., Ee, P.L.R., Ge, R., 2020. Developing inhaled protein therapeutics for lung diseases. Mol Biomed 1, 11.

McCormick, S.M., Heller, N.M., 2015. Commentary: IL-4 and IL-13 receptors and signaling. Cytokine 75, 38–50.

Mensink, M.A., Frijlink, H.W., van der Voort Maarschalk, K., Hinrichs, W.L.J., 2017. How sugars protect proteins in the solid state and during drying (review): Mechanisms of stabilization in relation to stress conditions. Eur J Pharm Biopharm 114, 288–295.

Nowak, C., K. Cheung, J., M. Dellatore, S., Katiyar, A., Bhat, R., Sun, J., Ponniah, G.,

Neill, A., Mason, B., Beck, A., Liu, H., 2017. Forced degradation of recombinant monoclonal antibodies: A practical guide. mAbs 9, 1217-1230. https://doi.org/10.1

080/19420862.2017.1368602.

Pan, H.W., Seow, H.C., Lo, J.C.K., Guo, J., Zhu, L., Leung, S.W.S., Zhang, C., Lam, J.K.W.,

2022. Spray-dried and spray-freeze-dried powder formulations of an anti- interleukin-4Rα antibody for pulmonary delivery. Pharm Res 39, 2291–2304.

Patton, J.S., 1996. Mechanisms of macromolecule absorption by the lungs. Adv Drug

Deliv Rev 19, 3–36.

Prinn, K.B., Costantino, H.R., Tracy, M., 2002. Statistical modeling of protein spray drying at the lab scale. AAPS PharmSciTech 3, 32–39.

Qiu, Y., Man, R.C.H., Liao, Q., Kung, K.L.K., Chow, M.Y.T., Lam, J.K.W., 2019. Effective mRNA pulmonary delivery by dry powder formulation of PEGylated synthetic KL4 peptide. J Control Release 314, 102–115.

Ramezani, V., Vatanara, A., Seyedabadi, M., Nabi Meibodi, M., Fanaei, H., 2017.

Application of cyclodextrins in antibody microparticles: potentials for antibody protection in spray drying. Drug Dev Ind Pharm 43, 1103–1111.

Serno, T., Geidobler, R., Winter, G., 2011. Protein stabilization by cyclodextrins in the liquid and dried state. Adv Drug Deliv Rev 63, 1086–1106.

Seville, P.C., Learoyd, T.P., Li, H.Y., Williamson, I.J., Birchall, J.C., 2007. Amino acid- modified spray-dried powders with enhanced aerosolisation properties for pulmonary drug delivery. Powder Technol 178, 40–50.

Sharma, A., Khamar, D., Cullen, S., Hayden, A., Hughes, H., 2021. Innovative drying technologies for biopharmaceuticals. Int J Pharm 609, 121115.

Shepard, K.B., Vodak, D.T., Kuehl, P.J., Revelli, D., Zhou, Y., Pluntze, A.M., Adam, M.S.,

Oddo, J.C., Switala, L., Cape, J.L., Baumann, J.M., Banks, M., 2021. Local treatment of non-small cell lung cancer with a spray-dried bevacizumab formulation. AAPS

PharmSciTech 22, 230.

Shoyele, S.A., Slowey, A., 2006. Prospects of formulating proteins/peptides as aerosols for pulmonary drug delivery. Int J Pharm 314, 1–8.

Sou, T., Kaminskas, L.M., Nguyen, T.-H., Carlberg, R., McIntosh, M.P., Morton, D.A.V.,

2013. The effect of amino acid excipients on morphology and solid-state properties of multi-component spray-dried formulations for pulmonary delivery of biomacromolecules. Eur J Pharm Biopharm 83, 234–243.

The United States Pharmacopeial Convention, 2014. Aerosols, Nasal Sprays, Metered- Dose Inhalers, and Dry Powder Inhalers, USP on Compounding – A Guide for the

Compounding Practitioner. United States Pharmacopeia (USP), Rockville, MD, pp.

152-178.

Vehring, R., 2008. Pharmaceutical particle engineering via spray drying. Pharm Res 25,

999–1022.

Vlasak, J., Ionescu, R., 2011. Fragmentation of monoclonal antibodies. MAbs 3,

253–263.

Walton, D.E., Mumford, C.J., 1999a. The morphology of spray-dried particles: The effect of process variables upon the morphology of spray-dried particles. Chem Eng Res Des

77, 442–460.

Walton, D.E., Mumford, C.J., 1999b. Spray dried products—characterization of particle morphology. Chem Eng Res Des 77, 21–38.

Wang, W., Nema, S., Teagarden, D., 2010. Protein aggregation—Pathways and influencing factors. Int J Pharm 390, 89–99.

Weissbuch, I., Frolow, F., Addadi, L., Lahav, M., Leiserowitz, L., 1990. Oriented crystallization as a tool for detecting ordered aggregates of water-soluble hydrophobic α-amino acids at the air-solution interface. J Am Chem Soc 112,

7718–7724.

White, S., Bennett, D.B., Cheu, S., Conley, P.W., Guzek, D.B., Gray, S., Howard, J.,

Malcolmson, R., Parker, J.M., Roberts, P., Sadrzadeh, N., Schumacher, J.D.,

Seshadri, S., Sluggett, G.W., Stevenson, C.L., Harper, N.J., 2005. EXUBERA®: pharmaceutical development of a novel product for pulmonary delivery of insulin.

Diabetes Technol Ther 7, 896–906.

Wu, H., Kroe-Barrett, R., Singh, S., Robinson, A.S., Roberts, C.J., 2014. Competing aggregation pathways for monoclonal antibodies. FEBS Lett 588, 936–941.

Xu, Y., Harinck, L., Lokras, A.G., Gerde, P., Selg, E., Sj¨oberg, C.-O., Franzyk, H.,

Thakur, A., Foged, C., 2022. Leucine improves the aerosol performance of dry powder inhaler formulations of siRNA-loaded nanoparticles. Int J Pharm 621,

121758.

H.W. Pan et al.

International Journal of Pharmaceutics 644 (2023) 123272

12 Yalkowsky, S.H., He, Y., Jain, P., 2010. Handbook of Aqueous Solubility Data, 2nd ed.

CRC Press, Boca Raton, FL.

Yang, D., Singh, A., Wu, H., Kroe-Barrett, R., 2017. Determination of High-affinity

Antibody-antigen Binding Kinetics Using Four Biosensor Platforms. J Vis Exp e55659.

Ziaee, A., Albadarin, A.B., Padrela, L., Femmer, T., O’Reilly, E., Walker, G., 2019. Spray drying of pharmaceuticals and biopharmaceuticals: Critical parameters and experimental process optimization approaches. Eur J Pharm Sci 127, 300–318.

H.W. Pan et al.