(W1030-03-13) Inhalation Dry Powders of Monoclonal Antibodies Made by Thin-Film Freeze-Drying

Purpose: Thin-film freeze-drying (TFFD) is a two-step, ultra-rapid freezing followed by lyophilization technology. It is a bottom-up particle engineering method that can be used to prepare highly porous, low density, brittle matrix powders for inhalation into the lungs. The present study is designed to assess the feasibility of producing stable, aerosolizable dry powders of monoclonal antibodies (mAbs) for pulmonary delivery by TFFD.

Methods: Dry powders of mAbs of various compositions were prepared by TFFD using anti-programmed cell death protein (anti-PD-1 mAb), anti-TNF-a mAb, or an anti-SARS-CoV-2 spike protein (AUG-3387). A series of experiments on mAb powder characterization were performed, including aerosol performance evaluation (using next generation impactor (NGI)), modulated differential scanning colorimetry (mDSC), X-ray powder diffraction (XRPD), scanning electron microscopy (SEM) and moisture content measurement. Size exclusion chromatography (SEC), sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and micro-flow imaging (MFI) were performed to evaluate the chemical and physical stability of the mAbs after being subjected to TFFD. Enzyme-linked immunosorbent assay (ELISA) and virus neutralization assay were performed to measure the activity of mAbs.

Results: We first formulated an IgG antibody with lactose/leucine (60:40 w/w) or trehalose/leucine (75:25) and tested their aerosol performance. The IgG 1% (w/w) formulated with lactose/leucine (60:40 w/w) in phosphate buffered saline (PBS) (IgG-1-LL-PBS) and processed by TFFD was found to produce a powder with desirable aerosol properties. We then replaced the IgG with a specific antibody, anti-PD-1 mAb, to prepare a dry powder (anti-PD1-1-LL-PBS), which performed similarly to the IgG-1-LL-PBS powder. The aerosol properties of the anti-PD1-1-LL-PBS dry powder were significantly better when TFFD was used to prepare the powder than when conventional shelf freeze-drying (shelf FD) was used, at the composition and lyophilization cycle tested (figure 1). The dry powder had a porous structure with nanoaggregates and had a T_g value between 39-50ºC. When stored at room temperature, the anti-PD-1 mAb in the TFFD powder was more stable than that of the same formulation stored as a liquid. The addition of polyvinylpyrrolidone in the formulation raised the T_g to 152ºC, which is expected to further increase the storage stability of the mAbs. The PD-1 binding activity of the anti-PD-1 mAbs after TFFD was not different from before TFFD. While protein loss, likely due to binding to vials and the TFF apparatus, was identified, we were able to minimize the loss by increasing mAb concentration (i.e., from 1% to 13.2%) (figure 2). MFI revealed that the excipients and PBS affected subvisible aggregate formation. More subvisible mAb aggregates were generated when PBS was used (figure 3), but the mAb content in the dry powders did not significantly affect the total subvisible aggregate count. Dry powders prepared with mannitol as the excipient resulted in the least amount of subvisible mAb aggregates. Finally, we showed that anti-TNF-α mAb and AUG3387 can also be thin-film freeze-dried to dry powders with a similar composition, and for the AUG3387, its SARS-CoV-2 neutralization activity after being subjected to TFFD and reconstitution was not different from before TFFD.

Conclusion: We conclude that it is feasible to use TFFD to produce stable, aerosolizable dry powders of mAbs for pulmonary delivery, but the formulations must be optimized to maximize aerosol performance and minimize protein aggregation.

References: Emig, C. J., Mena, M. A., Henry, S. J., Vitug, A., Ventura, C. J., Fox, D., ... & Christensen, D. J. (2021). AUG-3387, a human-derived monoclonal antibody neutralizes SARS-CoV-2 variants and reduces viral load from therapeutic treatment of hamsters in vivo. bioRxiv.

Acknowledgments: This work was in part supported by TFF Pharmaceuticals, Inc. (to ROW and ZC). The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: [ROW and ZC report financial support provided by TFF Pharmaceuti- cals, Inc. ZC reports a relationship with TFF Pharmaceuticals, Inc. that includes: equity or stocks and funding grants. ROW reports a relation- ship with TFF Pharmaceuticals, Inc. that includes: consulting or advi- sory, equity or stocks, and funding grants. HX, SS, CM report a relationship with TFF Pharmaceuticals, Inc. that includes: consulting or advisory. ZC, ROW, SH, XH, & LQMC have a patent pending to UT Austin. LQMC has extensive clinical research consulting in the arena of immunotherapy and reports: minor de minimus personal advisory board consulting fees and prior institutional research grant funding from Merck, Pfizer, Dynavax and Novartis, Astra-Zeneca in the last three years and de minimus personal advisory board consulting fees and current institutional research grant funding from Alkermes. She reports current institutional grant funding only from Oncorus. She reports minor de minimus consulting/advisory board fees from Cullinan and Elicio, Gilead, Regeneron, Sanofi-Genzyme and Daiichi Sankyo, Ipsen, Nano- biotix, Beigene, Jazz Pharmaceuticals. Research grant funding only provided to her prior institution from Bristol Myers Squibb, Genentech, Seattle Genetics, Lilly/Imclone].

Figure 1. Aerosol performance properties of anti-PD1-1-LL-PBS TFFD powder (TFFD) compared to anti-PD1-1-LL-PBS shelf FD powder (Shelf FD)



Figure 2. Antigen (PD-1) binding capacity of anti-PD-1 mAbs in liquid before TFFD and after being subjected to TFFD and reconstituted in the initial liquid volume. (A) ELISA data normalized to protein content and (B) quantification of protein loss based on SDS-PAGE (* indicates difference from respective liquid samples (before TFFD). Data are mean ± SD (n = 3), and n.s. indicates no significance.



Figure 3. Subvisible aggregates in anti-PD-1 mAb TFFD powders after reconstitution. Shown are MFI counting data comparing subvisible aggregates in anti-PD-1 mAb dry powders of different compositions (lactose/leucine (LL), 60:40; trehalose/leucine (TL), 75:25; sucrose (S) alone with or without PBS). For powders prepared with anti-PD-1 mAbs at 1% (w/w) with or without PBS, the total particle counts in samples prepared with PBS were significantly higher than in respective samples prepared without PBS (*, p < 0.05, t-test).