Microwave-Assisted Freeze-Drying: Impact of Microwave Radiation on the Quality of High-Concentration Antibody Formulations

Abstract

Microwave-assisted freeze-drying (MFD) offers significant time savings compared to conventional freeze-drying (CFD). While a few studies have investigated the stability of biopharmaceuticals with low protein concentrations after MFD and storage, the impact of MFD on high-concentration monoclonal antibody (mAb) formulations remains unclear. In this study, we systematically examined the effect of protein concentration in MFD and assessed protein stability following MFD, CFD, and subsequent storage using seven protein formulations with various stabilizers and concentrations. We demonstrated that microwaves directly interact with the active pharmaceutical ingredient (API), leading to decreased physical stability, specifically aggregation, in high-concentration antibody formulations. Furthermore, typically used sugar:protein ratios from CFD were insufficient for stabilizing mAbs when applying microwaves. We identified the intermediate drying phase as the most critical for particle formation, and cooling the samples provided some protection for the mAb. Our findings suggest that MFD technology may not be universally applicable to formulations well tested in CFD and could be particularly beneficial for formulations with low API concentrations requiring substantial amounts of glass-forming excipients, such as vaccines and RNA-based products.

Keywords: aggregation; freeze-drying; lyophilization; microwave; monoclonal antibody; protein; stability.

Figure 1

The solid-state properties of the lyophilizates and storage stability of LMU1 when sugar was subsequently replaced with mAb. Samples were analyzed after MFD (t0) and storage at 4 °C, 25 °C, and 40 °C over 6 months. (A) Specific surface area (bars) and residual moisture (symbols). The relative monomer yields (bars) and percentages of soluble aggregates (HMWS, symbols) from SEC are shown in (B). (C) The relative number of acidic and (D) basic variants from IEX. All values are means (n = 3) ± standard deviation.

Figure 2

The effect of the drying mechanism on critical quality attributes of highly concentrated LMU1 formulations. Following MFD and CFD (t0), the lyophilizates were stored at 4 °C, 25 °C, and 40 °C for 6 months. (A) The specific surface area (bars) and residual moisture (symbols) of the cakes. (B) The relative number of acidic and basic variants for F4 (left) and F5 (right) from IEX. (C) The relative monomer yield and the relative number of high-molecular-weight species (HMWS) was determined using SEC. Subvisible particles (SvP) detected with flow imaging microscopy: (D) >25 μm, (E) >10 µm, and (F) >1 μm. All values are means (n = 3) ± standard deviation. SvP measurements were conducted in technical duplicates.

Figure 3

Physical stability of LMU2 (formulation F6) following MFD and CFD. Samples were analyzed after lyophilization (t0) and storage at 4 °C, 25 °C, and 40 °C (MFD samples) and 40 °C (CFD samples). (A) The relative monomer yield and the relative number of high-molecular-weight species (HMWS). (B) Subvisible protein aggregates. All values are means (n = 3) ± standard deviation. Subvisible particle measurements were conducted in technical duplicates.

Figure 4

The impact of the microwave run time on protein aggregation during MFD. (A) Graphical overview of the lyophilization process readouts for P3. Microwave radiation was started immediately after the desired vacuum for primary drying was established and ran for 5 h. T_s denotes the shelf temperature; the chamber pressure is monitored via a Pirani gauge (Pirani) and capacitance gauge (Capacitance); T_p is the reading from the fiberoptic temperature sensors. (B) Process readouts for P3 when microwave radiation was applied for 5 h toward the end of the process. (C) Comparison of subvisible particle formation in the F1, F5, and placebo formulations, as detected via flow imaging microscopy, when microwave radiation was applied during the initial 5 h of drying (init) and for 5 h later in the process (late), using process P3. (D) Product temperature profiles recorded for P3 and P4 with the different microwave module run times. The arrows represent the switch off of microwave radiation. All temperature sensors shown in the process graphs (A,B,D) were placed in formulation F5. (E) Subvisible particle formation in the F1, F5, and placebo formulations when subjected to increasing microwave run times. The reported numbers of subvisible particles are means (n = 3 and technical duplicates per vial) ± standard deviation. MW, microwave irradiation.

Figure 5

The impact of microwave radiation on protein aggregate formation in lyophilized formulation F5. Initial subvisible particle counts (t0) were determined immediately after conventional freeze-drying. (A) Samples were exposed to 360 W for different durations, without chilling during exposure to microwave radiation. A polymeric vial was used as a spacer to insulate the samples from the rotating glass plate in the microwave oven. This setup was used for the following experiments, with the data shown in (B–E). (B) The formation of subvisible particles with increased residual moisture. The residual moisture content of 15% was adjusted in all processed samples shown in (C–E). (C) Comparison of convective heat transfer and microwave heating, with the drying cabinet temperature set to 80 °C. To mimic freeze-drying conditions, the vial was placed on a precooled stainless steel cylinder inside the microwave oven (red bars, without pattern). (D,E) Lyophilizates were subjected to three different energy sources. (D) The temperature within the cakes and (E) the corresponding formation of protein aggregates. The subvisible particle data represent the mean values of technical duplicates per vial ± standard deviation. MW, microwave irradiation.