Mechanistic Modeling of Continuous Lyophilization for Biopharmaceutical Manufacturing

TL;DR

The paper presents the first mechanistic model for continuous lyophilization of suspended vials, covering freezing, primary drying, and secondary drying within a fully continuous framework. It combines a preconditioning/ VISF-enabled freezing scheme with a hybrid solidification approach, a 1D heat/mass transfer description for primary drying, and 1D desorption-dominated secondary drying; all components are integrated for rapid, cycle-to-cycle predictions. Validation against Capozzi et al. suspended-vial data shows close agreement for temperature profiles, ice/water fractions, and bound-water trajectories, and the model runs in under 1 s on a laptop, enabling design, optimization, and control applications including digital-twin development. The authors provide ContLyo as an open-source MATLAB and Julia tool to facilitate adoption and future extension toward model-based control and uncertainty quantification in continuous biopharmaceutical lyophilization.

Abstract

Lyophilization (aka freeze drying) is a typical process in pharmaceutical manufacturing used for improving the stability of various drug products, including its recent applications to mRNA vaccines. While extensive efforts have been dedicated to shifting the pharmaceutical industry toward continuous manufacturing, the majority of industrial-scale lyophilization is still being operated in a batch mode. This article presents the first mechanistic model for a complete continuous lyophilization process, which comprehensively incorporates and describes key transport phenomena in all three steps of lyophilization, namely freezing, primary drying, and secondary drying. The proposed model considers the state-of-the-art lyophilization technology, in which vials are suspended and move continuously through the process. The validated model can accurately predict the evolution of critical process parameters, including the product temperature, ice/water fraction, sublimation front position, and concentration of bound water, for the entire process. Several applications related to model-based process design and optimization of continuous lyophilization are also demonstrated. The final model is made available in MATLAB and Julia as an open-source software package called ContLyo, which can ultimately be leveraged for guiding the design and development of future continuous lyophilization processes.

Figures (14)

• Figure 1: (A) Conventional batch lyophilization of unit doses. A number of vials are placed on the cooling/heating shelf. (B) Continuous lyophilization of suspended vials. A number of vials are suspended and move continuously through the lyophilizer.
• Figure 2: Modeling strategies for the freezing, primary drying, and secondary drying steps in lyophilization. The strategies used in this article are highlighted.
• Figure 3: Schematic diagram showing the mechanistic modeling of continuous lyophilization via suspended vials for (A) freezing, (B) primary drying, and (C) secondary drying.
• Figure 4: Schematic diagram showing the mechanistic modeling of the solidification step. For simplification, it is assumed that the liquid part retains a cylindrical shape with the same aspect ratio as the initial solution before nucleation starts.
• Figure 5: Model validation for the lyophilization of suspended vials using the experimental data from Capozzi2019ContLyo_SuspendedVials. Panel A shows the model prediction and experimental data for the product temperature during the freezing step. Panel B shows the model prediction and experimental data for the product temperature (assumed to be measured at the bottom surface) during the primary drying step. The maximum shelf temperatures for Cases 2a and 2b are 263 and 313 K, respectively. Panel C shows the model prediction and experimental data for the average concentration of bound water during the secondary drying step. The initial concentrations for Cases 3a and 3b are 0.088 and 0.075 kg water/kg solid, respectively.