Implementation of QbD strategies in the inoculum expansion of a mAb production process

Abstract

The quality by design approach was introduced to the biopharmaceutical industry over 15 years ago. This principle is widely implemented in the characterization of monoclonal antibody production processes. Anyway, the early process phase, namely the inoculum expansion, was not yet investigated and characterized for most processes. In order to increase the understanding of early process parameter interactions and their influence on the later production process, a risk assessment followed by a design of experiments approach was conducted. The DoE included the critical parameters methotrexate (MTX) concentration, initial passage viable cell density and passage duration. Multivariate data analysis led to mathematical regression models and the establishment of a designated design space for the studied parameters. It was found that the passage duration as well as the initial viable cell density for each passage during the inoculum expansion have severe effects on the growth rate and viability of the early process phase. Furthermore, the variations during the inoculum expansion directly influenced the production process responses. This carry-over of factor effects highlights the crucial impact of early process failures and the importance of process analysis and control during the first part of mAb production processes.

Keywords: CHO; PAT; QbD; inoculum expansion; mAb.

FIGURE 1

Schematic strategy for implementation of QbD and determination of a process design space in a biopharmaceutical production process. CPP, critical process parameter; CQA, critical quality attribute; DoE, design of experiments; QTPP, quality target product profile

FIGURE 2

Schematic representation of the process workflow of monoclonal antibody cell cultivation in larger scale

FIGURE 3

Ishikawa diagram for the inoculum expansion. Bold parameters were determined as critical and therefore examined during this study

FIGURE 4

Viable cell densities during the inoculum expansion for experiments with –1/0/1 levels for the passage duration. Cells were split to 0.2 × 10⁶ cells/mL for each passage

FIGURE 5

Factor (parameter) and its interaction effect plots for the inoculum expansion and production process responses. Not shown parameters as well as parameters whose confidential interval includes zero were determined to have no significant effect on the respective response (Pas, passage duration, MTX, TX concentration, VCD, initial passage viable cell density)

FIGURE 6

Observed versus predicted plots for the inoculum expansion and production process responses

FIGURE 7

Response contour plots showing results for inoculum expansion and production process responses

FIGURE 8

Determined design space for the inoculum expansion and production process viability/growth rate and the antibody titer with colour coded probability of failure