This is important, especially for the development of biopharmaceuticals, where
drug substance for clinical trials may be produced as early as the laboratory
development phase, or in early pilot plant phase under cGMP guidelines. Once
regulatory clearance is obtained, the process is then scaled up to industrial conditions. Two important points must be considered here: (1) the product quality
characteristics affecting the efficacy of the biopharmaceutical candidate must comply with the specifications defined during clinical trials and (2) the process must be
economically feasible whilst producing sufficient quantities to supply the market.
Unfortunately, the increase in size has many implications for the process conditions
inside the bioreactor. That is, if an 80 L bioreactor is scaled up to 10,000 L whilst
maintaining a constant power input per unit volume, the mixing time increases
3 times, the impeller tip speed doubles, and the shear forces increase almost
10 times [46]. Oldshue showed that a scale-up design to satisfy mass transfer
(constant K l a criterion) from a 75 L pilot scale process to a 95,000 L production
scale would increase the shear rate by 180%, whereas maintaining a constant shear
rate between the two scales could only produce 40% of the mass transfer requirements of the culture in the large scale [47]. The most common consequence is an
inevitable increase in mixing times of up to 200 s in larger-scale bioreactors (since
scale-up is mostly based on K l a, P/V, impeller tip speed) [48].
In addition to the increased mixing times, fed-batch processes are fed with
concentrated substrates at localized feeding points, which are mechanically fixed.
The longer mixing times and the localized addition of highly concentrated viscous
substrates lead to the formation of concentration gradients in the bioreactor
[49, 50]. Cells that are traversing these gradients respond in many ways, by the
varied distribution of metabolic fluxes due to the changed uptake rates in different
positions of the bioreactor, and by specific gene expression profiles which include
both specific responses and general stress adaptation. The specific reaction of an
individual cell depends not only on its metabolic state and the current phase in the
cell cycle, but also on its specific historical situation, i.e. what conditions it has
experienced previously in the dynamic course of time [51]. This is currently being
investigated using fluid dynamic models by simulating cell lifelines. The sum of all
of these affects the fermentation efficiency in terms of yields and overall process
robustness.
When the characteristic time of relevant cellular processes (translation, cell
division) is close to the mixing time in large-scale bioreactors, there is a measurable
influence of gradients on the growth and metabolic behavior of the culture
[46, 52]. The inefficient mixing in large-scale bioreactors leads to the creation of
spatial concentration pockets of relevant process parameters, such as substrate
(glucose), dissolved oxygen, acidity, and temperature. Furthermore, GMP
manufacturing processes suffer from the rigidity of the process due to the difficulty
in using validated equipment for such studies, especially when the characterization
study requires minor retrofitting of the bioreactor, such as installing extra sensors. In
cases where bioreactor characterization has been done, companies consider the data
as confidential; therefore, the information is not available to the scientific research
community.
8
P. Neubauer et al.
drug substance for clinical trials may be produced as early as the laboratory
development phase, or in early pilot plant phase under cGMP guidelines. Once
regulatory clearance is obtained, the process is then scaled up to industrial conditions. Two important points must be considered here: (1) the product quality
characteristics affecting the efficacy of the biopharmaceutical candidate must comply with the specifications defined during clinical trials and (2) the process must be
economically feasible whilst producing sufficient quantities to supply the market.
Unfortunately, the increase in size has many implications for the process conditions
inside the bioreactor. That is, if an 80 L bioreactor is scaled up to 10,000 L whilst
maintaining a constant power input per unit volume, the mixing time increases
3 times, the impeller tip speed doubles, and the shear forces increase almost
10 times [46]. Oldshue showed that a scale-up design to satisfy mass transfer
(constant K l a criterion) from a 75 L pilot scale process to a 95,000 L production
scale would increase the shear rate by 180%, whereas maintaining a constant shear
rate between the two scales could only produce 40% of the mass transfer requirements of the culture in the large scale [47]. The most common consequence is an
inevitable increase in mixing times of up to 200 s in larger-scale bioreactors (since
scale-up is mostly based on K l a, P/V, impeller tip speed) [48].
In addition to the increased mixing times, fed-batch processes are fed with
concentrated substrates at localized feeding points, which are mechanically fixed.
The longer mixing times and the localized addition of highly concentrated viscous
substrates lead to the formation of concentration gradients in the bioreactor
[49, 50]. Cells that are traversing these gradients respond in many ways, by the
varied distribution of metabolic fluxes due to the changed uptake rates in different
positions of the bioreactor, and by specific gene expression profiles which include
both specific responses and general stress adaptation. The specific reaction of an
individual cell depends not only on its metabolic state and the current phase in the
cell cycle, but also on its specific historical situation, i.e. what conditions it has
experienced previously in the dynamic course of time [51]. This is currently being
investigated using fluid dynamic models by simulating cell lifelines. The sum of all
of these affects the fermentation efficiency in terms of yields and overall process
robustness.
When the characteristic time of relevant cellular processes (translation, cell
division) is close to the mixing time in large-scale bioreactors, there is a measurable
influence of gradients on the growth and metabolic behavior of the culture
[46, 52]. The inefficient mixing in large-scale bioreactors leads to the creation of
spatial concentration pockets of relevant process parameters, such as substrate
(glucose), dissolved oxygen, acidity, and temperature. Furthermore, GMP
manufacturing processes suffer from the rigidity of the process due to the difficulty
in using validated equipment for such studies, especially when the characterization
study requires minor retrofitting of the bioreactor, such as installing extra sensors. In
cases where bioreactor characterization has been done, companies consider the data
as confidential; therefore, the information is not available to the scientific research
community.
8
P. Neubauer et al.