Industrial Process of Manufacturing
Optimization efforts in process development throughout the years have
been devoted to decreasing process time, investment and development
costs per antibody (Gronemeyer, Ditz, & Strube, 2014). Furthermore,
reducing mAb variants and process impurities, and delivering
batch-to-batch consistency is a key deliverable of process development.
The manufacturing process can be summarized into five steps (Steinmeyer
& McCormick, 2008):
- Development of a stable and productive cell line expressing the gene
of interest.
- Cell culture conditions that allow cells to grow and produce the
correct form of the desired antibody in high titres.
- Drug substance purification.
- Formulation steps for purification of the antibody and appropriate
dosage forms (Fill & Finish).
- Analytical testing methods to monitor the processes and evaluate the
final product.
The first two stages correspond to upstream processing (USP) operations,
while the latter three are considered downstream processing (DSP)
activities. Owing to the structural similarity of mAbs, platform process
technologies can be developed. The platform process includes the host
system for cell culture, pre-engineered vectors for transformation,
pre-defined cell amplification scheme, bioreactor conditions, high
performance purification systems and well-validated analytical methods
(Steinmeyer & McCormick, 2008).
During mAb production, one of the most critical steps is the choice of a
cell line. The cells must propagate well, be highly stable in culture
and produce high mAb titres in its active form, meaning the protein is
properly folded, glycosylated and not aggregated. Thus, based on these
requirements, mammalian cells are most commonly used, as they are
adapted for the production, processing and secretion of highly complex
molecules (Carvalho et al., 2017). Historically, the host cell line
commonly used was NS0, a murine myeloma cell line. Nonetheless, as the
expression system for tissue plasminogen activator, one of the first
approved biopharmaceuticals in 1986, Chinese hamster ovary (CHO) cells
were used and have remained the preferred choice for production since
(Kunert & Reinhart, 2016). The preference for CHO cells is attributable
to their rapid propagation and high expression titres, along with the
absence of two glycan epitopes for humans, which are present in murine
cell lines e.g. NS0 and Sp2/0 (Figure 1) (Gomord et al., 2010;
Gronemeyer et al., 2014; Jefferis, 2009; Kelley, 2009; Kunert &
Reinhart, 2016). Other human cell lines such as human embryonic kidney
(HEK) or PER.C6 cells are also considered an adequate expression host.
However, reports have shown increased in vivo heterogeneity in
comparison to CHO cells (Kunert & Reinhart, 2016).
Cell line engineering and development ultimately aims at achieving
higher cell titres to improve productivity and product quality.
Optimisation efforts in cell medium, feed development, bioprocess
development and scale-up experiments are reviewed by Gronemeyer et al.
(2014). Engineering of CHO cells, for example, has resulted in cell
lines capable of secreting up to 100 pg/cell/day of humanized mAbs (Page
& Sydenham, 1991) or 80-110 pg/cell/day of chimerized mAb (Fouser et
al., 1992) in perfusion. Nevertheless, the feed method implemented has a
direct influence on mAb titres. The different feed methods are batch,
fed-batch or perfusion. In the batch method, all nutrients are added
into the initial medium, whereas in fed-batch nutrients are added as
they become depleted. Perfusion means the medium is being circulated
through the growing culture while keeping the cells in the bioreactor
via filtration, removing waste and supplying fresh nutrients to the
cells (Dorceus et al., 2017). Fed-batch processing leads to mAb titres
of 1-5 g/L, with some companies reporting up to 13 g/L using extended
culturing conditions (Kelley, 2009). The different processes have been
compared in literature (Carvalho et al., 2017; Fan, Ley, & Andersen,
2018; Kunert & Reinhart, 2016; Ritacco, Wu, & Khetan, 2018)
Over the years, USP operations have seen titre and stability increases
and the focus has shifted to the optimization of DSP operations,
focusing on improving yield, purity and productivity. After mAb
production, DSP is responsible for the delivery of the active
pharmaceutical ingredient (API), also referred to as bulk drug substance
or drug substance (DS), to formulation and filling (Fill & Finish). The
formulated DS is then referred to as the drug product (DP) and is ready
to be administered to patients. Novel technologies and the establishment
of platform technologies based on Quality-by-Design (QbD) approaches has
allowed developments in DSP operations. Focusing on chromatographic
separations, mainly using Protein A chromatography, advances in column
characteristics including higher flow rate, longer life cycles, reduced
run times and increasing binding capacity has brought advances into DSP
operations. However, non-chromatographic separations are of increasing
interest due to the high costs associated with chromatography.
Non-chromatographic separation methods include the use of membrane-based
procedures, aqueous two-phase separations, precipitation or
crystallization methods. These allow for high-volume feeds and rapid
liquid removal, aiming at cost and process time reduction, as well as
improved impurities and yield losses (Gronemeyer et al., 2014; Großhans,
Wang, Fischer, & Hubbuch, 2018; Thömmes, Twyman, & Gottschalk, 2017).
Other technologies moving away from chromatographic methods and
supporting higher throughput include continuously-fed multicolumn
methods using disposable columns such as simulated moving-bed
chromatography. Continuous manufacturing has combined the advantages of
improved process performances and flexible manufacturing (Chahar,
Ravindran, & Pisal, 2020; Großhans et al., 2018; Pollard, Brower, Abe,
Lopes, & Sinclair, 2016). Trends have led to the use of disposables,
with the benefits of having lower capital investment and operational
costs, increased flexibility, improved production scheduling and higher
process replication (Eibl & Eibl, 2019; Gronemeyer et al., 2014).
Single-use bioreactors and technologies are applicable to Good
Manufacturing Practice (GMP) regulations and are available in sizes up
to 2000 L (Gronemeyer et al., 2014; Jossen, Eibl, & Eibl, 2019; Langer
& Rader, 2019).