For cultivations where samples are taken twice a day and cultivation times exceed a

week, the rather low working volume might not be ideal. Furthermore, when

planning experiments, care should be taken that appropriate controls run in parallel,

for instance mock cultivations where seed virus is not added. Especially, when

switching to bioreactor cultivations at larger scale, sterility controls of samples after

important process steps and a parallel run in small scale might help to identify

possible contamination sources. For cell passaging it is equally helpful to use se-

parate media bottles and sometimes even separate batches per person to have

adequate backup solutions in case of contaminations.

Overall, optimization of cell growth of the host cell will take a considerable part

of process optimization (see Table 5.7). It will be a continuous process as new

medium formulations, sensors, or cultivation vessels are constantly developed and

then a re-evaluation of process options might be needed. Taken together, relatively

high costs of equipment and media together with time-consuming cultivations and

assays make process development and optimization a demanding task.

5.7.2.1

Seed Virus Generation

A crucial part is the selection and production of a proper seed virus. The virus can

either be isolated from an infected person or be obtained from a biological reference

material bank such as the National Institute for Biological Standards and Control

(NIBSC). As some viruses only enter and replicate in a small subset of GMP-

relevant cell lines, the choice of the appropriate host cell line is of utmost im-

portance. A cell line screening should be performed, where different cell lines are

infected with the virus at different MOI (and other infection conditions) to identify

the cell line with the highest virus yield. Such scouting experiments can be con-

ducted in small scale in well plates, T-flasks, or shake flasks. From there, the virus

is passaged serially in a cell line, while the sample with the highest infectious titer is

used for the next passage to try to adapt the virus to the cell line. This might result

in a more efficient production due to an accelerated virus propagation and increased

maximum virus titer. This adaptation can also be used to adapt to lower MOI for

instance.

5.7.2.2

Process Optimization Options

To obtain maximum virus yields and a high product quality, various parameters

need to be considered and optimized in process development (see Table 5.7).

Depending on vaccine type, the definition of product quality is variable. For live-

attenuated vaccines the infectious titer is important and if possible the ratio of non-

infectious to infectious particles should be low. However, if the non-infectious

particles function as an adjuvant by increasing the immune response, this might not

be too important. For sub-unit or split vaccines, the amount of functional antigens is

of interest with possible low contaminations with host-cell proteins and DNA. Viral

vectors or chimeric viruses generate the correct immune response by providing the

antigen in the correct formation with the correct structure and if important glyco-

sylation. Product quality is thus not only one attribute, but is an overall result of

several virus/antigen properties. Therefore, optimization of the production process

will need to consider on the one hand, how to obtain maximum yield and on the other

Upstream processing for viral vaccines

113