Biomedical Engineering Reference
In-Depth Information
scale-down models should be designed such that, to the extent possible, key and critical
parameters are maintained across scales. Moreover, the performance of the laborator-
y-scale models used to develop the design space needs to be representative of the
manufacturing-scale process with regard to process performance and product compara-
bility [10-12]. A comparison of the manufacturing-scale process data to the predictive
models developed during process characterization will provide evidence for the validity
of the scale-down models. When manufacturing data do not exist, this comparison will
need to be done retrospectively after the process has been performed at full scale. If
sufficient manufacturing-scale data exist, predetermined acceptance criteria can be
developed prior to the scale-down runs or it may be demonstrated that the small-scale
data are within the full-scale historical range [11]. Alternatively, for a true side-by-side
comparison, the validity of models can be demonstrated with a scaled-down “satellite”
process that is run in parallel with and starting with cell culture media taken from the
full-scale manufacturing process.
7.3 APPLICATIONS OF DESIGN SPACE
Several potential uses of design space are illustrated in the context of a typical
manufacturing process shown in Fig. 7.1 for a monoclonal antibody (mAb) or an
antibody-like (fc) fusion protein (FP) produced in mammalian cell culture. The process
consists of cell removal by filtration or continuous flow centrifugation, followed by a
recombinant protein A affinity “capture” column, then a low pH hold for virus
inactivation, and subsequently further purification by one or two additional chromatog-
raphy steps. The latter two steps commonly comprise some combination of anion
exchange, cation exchange, hydrophobic interaction, or ceramic hydroxyapatite chro-
matography. Following the chromatography steps, the product is passed through a
virus-removal filter and subsequently concentrated and transferred (by diafiltration) into
the formulation buffer by using cross-flow ultrafiltration (UF). The purification process
provides the removal of process and host-cell related impurities such as cell culture
additives, host-cell protein, and DNA. In addition, the purification process may control
the level of product-related impurities and isoforms including clipped species, charge
variants, aggregates, and glycoforms. Lastly, for proteins produced in mammalian cell
culture, the downstream process must provide assurance that retroviruses and other
adventitious agents potentially present during cell culture are sufficiently removed by the
purification process [13, 14]. In the remainder of this chapter, several examples of how
design space could be applied to downstream processing are described. Each example is
taken from an actual manufacturing process for an antibody or fusion protein similar to
that outlined in Fig. 7.1.
7.4 CELL HARVEST AND PRODUCT CAPTURE STEPS
Similar to the process outlined in Fig. 7.1, the production process for a humanized
monoclonal antibody (mAb1) uses continuous-flow centrifugation for the cell removal
