virus, between 2−4 days after TOI (Figure 5.6). Typically, the infectious virus titer

peaks earlier than the total virus titer (infectious plus non-infectious virus particles),

and the total virus titer remains more or less constant after reaching its maximum,

whereas the infectious titer often decreases with time depending on the virus stability.

This is important for generation of seed virus and when infectious virus material is the

product (e.g., live-attenuated vaccines, viral vectors, oncolytic viruses). For lytic

viruses, the cytopathic effect leads to the termination of the process due to cell death.

Depending on titers and the level of contaminating by-products (e.g., host cell DNA,

proteins), the virus harvest is collected, clarified (depth filtration, centrifugation) and

inactivated by (e.g., formaldehyde, β-propiolactone or binary ethyleneimine (BEI) for

manufacturing of inactivated vaccines). Subsequently, it is subjected to DSP and

formulation (Figure 5.1). After sterilization of the equipment or exchange of the

single-use equipment, a new batch cycle can be conducted. Monitoring of the pro-

duction process is carried out by measuring the concentrations of cells and metabo-

lites, pH value, total and infectious virus titer, DNA and (host cell) protein levels (see

also chapter 8). Most of these measurements still rely on manual sampling. For off-

line analytics, samples are stored at −80°C and should only be thawed once for ti-

trations as viruses are sensitive to freeze-thaw cycles. For other assays, heat or other

inactivation of samples should be considered with respect to biosafety and virus

contaminations of equipment.

5.7.2

PROCESS DEVELOPMENT AND OPTIMIZATION

Ideally, process development should be based on comprehensive data available for

cultivations that were performed at various scales and under different conditions

with at least three replicates using a design of experiments (DoE) approach. To

obtain maximum virus titers, experiments should include the adaptation of viruses

to the specific host cell line selected and variations in TOI and MOI. Moreover, the

replicates should be performed from different precultures and with different ali-

quots from the virus seed. As animal cell culture is time consuming and many of the

assays to determine virus titers require several days before read-outs are available,

the total time for a set of experiments including analytics can require several weeks

(typically 4−6 weeks). In addition, contamination risks and handling issues with

small bioreactors can be challenging.

Furthermore, regarding data evaluation and the selection of optimum process

conditions, the error of the respective assays has to be taken into account. This

concerns in particular, virus titrations based on dilution series which often display

errors in the ±0.3 log range. Together with relatively high batch-to-batch variations

of cultivations performed in triplicates, the final selection of optimum process

conditions is not trivial. Depending on resources (staff, equipment, analytics), it’s

generally advisable to first perform scouting experiments (possibly in triplicates) at

very small scales (well plates, T-flasks, shake flasks) and then plan for DoE ap-

proaches at the bioreactor scale for a low number of (pre-)selected parameters only.

Another option is the use of small-scale or micro-bioreactors such as Ambr^®

(Sartorius) or BioLector (m2plabs). However, working volumes and sampling have

to be selected carefully as for some analytics sample volumes of 1 mL are required.

112

Bioprocessing of Viral Vaccines