A Novel MATLAB®-Algorithm-Based Video Analysis to Quantitatively Determine Solution Creeping in Intact Pharmaceutical Glass Vials.

During the filling process of a biopharmaceutical drug product (DP), a liquid DP film might creep up the inner vial wall which is barely discernible, appears as milky-white haze after lyophilisation and is known as fogging. Creeping and fogging are mainly dependent on the primary packaging material surface and its hydration, vial preparation process as well as DP composition. The occurrence of both can impede visual inspection and might lead to DP rejection. Hence, our studies focused on the early detection of liquid solution and glass vial surface interaction directly after filling. For a fast and highly sensitive evaluation a novel video-based analysis was used. To our knowledge, this is the first time a MATLAB®-algorithm-based video analysis was applied to quantitatively determine creeping time-resolved. Furthermore, creeping in dependence of vial processing sites, surfactant type and concentration, filling temperature, and vial format were investigated. The results were verified using orthogonal conventional methods such as surface tension, wetting behaviour, and contact angle measurements, as well as ToF-SIMS, ICP-MS, and SEM. Additionally, the methods applied were assessed regarding their cross-validation capability. The observations indicate that the vial preparation process can have a pronounced impact on alteration of the glass vial surface and related creeping behaviour of the filled solution.

Establishment of a CFD-based kL a model in microtiter plates to support CHO cell culture scale-up during clone selection.

Microtiter plates are a common tool for clone selection in biopharmaceutical development. A way of visualizing and evaluating these systems and key processes parameters is the application of Computational Fluid Dynamics (CFD). CFD is a powerful tool for the modelling of hydrodynamics and mass transfer parameters. In this work, CFD was used to determine the specific surface area, the volumetric power input and the oxygen mass transfer coefficient kL a for two different microtiter plates with different scales (100 µL - 5 mL). For this purpose, a new method of predicting the kL a is presented and calibrated with literature data. Scaling effects in shaken microtiter plates are evaluated by comparing two culture volume scales under various operating conditions. To test validity of these models, three different Boehringer Ingelheim Pharma proprietary CHO production cell lines with different growth characteristics were cultivated using the respective microtiter plates under different conditions until limitations in growth and viability were observable. The cell culture data then was compared to different parameters obtained by CFD. The calculated kL a values match the cell culture performance in the 96-deepwell by predicting lowered oxygen transfer with increasing culture volume and decreasing orbital velocity. The same cells behave differently in the 6-deepwell scale. Here, the overall larger shear stress might cause physical stress for the cells. The kL a model predicts overall higher shear rates for this system, supporting the experimental findings. © 2018 American Institute of Chemical Engineers Biotechnol. Prog., 2018.