pleasing overall correlation with the values measured experimentally and demonstrate the applicability of the unstructured, segregated growth model. By using
determined growth parameters from cultivation studies, the cell growth, glucose
consumption, lactate production, and ammonium production could be proficiently
approximated. The greatest deviations in cell density were in the range of 3–20% for
the cells in suspension and 4–24% for the cells on the MCs. The glucose, lactate, and
ammonium timelines also correspond to this pattern, even though the specific
substrate consumption and metabolite production rates were prone to errors. However, the models provide reliable predictions for the MC-based hMSC growth in the
two spinner flask types.
5 Conclusions and Outlook
In this review, the current state of the art of the in vitro expansion of hMSC and the
use of numerical tools to support the development of MC-based hMSCs expansions
as well as the establishment of “Digital Twins” have been presented. It has been
emphasized that different CFD model approaches are described in the scientific
literature which can be successfully applied for the characterization of SU bioreactors, especially for the process development of hMSC expansion processes.
The CFD case study presented clearly demonstrates that numerical models are
valuable tools for the biochemical engineering characterization of small-scale spinner flasks, especially for the determination of parameters that are difficult to determine experimentally. A good correlation was always found between the parameters
predicted by the CFD and those measured experimentally. This observation was also
in agreement with the literature data. The Euler-Euler and Euler-Lagrange models
gave adequate predictions of the MC distributions within the spinner flask systems
and were correlated qualitatively with experimental observations. The EulerLagrange approach allowed the calculation of particle histories due to its discrete
particle formulation, which can be combined with experimental cultivation studies.
Thus, Euler-Lagrange modelling should be favored in the future in order to derive
hydrodynamic stresses over time instead of volume-weighted data. The scientific
literature summarized also shows that different model approaches for the simulation
of the hMSC growth are available, even though only a few are applicable for the
MC-based growth simulation in a stirred bioreactor. The unstructured and segregated growth model presented gives a good description of the MC-based hMSC
expansion process in the two spinner flask systems. Thus, MC-based hMSC cell
growth can be predicted. However, the further development of descriptive, or even
predictive, models for hMSCs will be important in the future for exact scheduling of
the preparation of the cell material and the subsequent autologous therapy.
Numerical Methods for the Design and Description of In Vitro Expansion. . .
221
determined growth parameters from cultivation studies, the cell growth, glucose
consumption, lactate production, and ammonium production could be proficiently
approximated. The greatest deviations in cell density were in the range of 3–20% for
the cells in suspension and 4–24% for the cells on the MCs. The glucose, lactate, and
ammonium timelines also correspond to this pattern, even though the specific
substrate consumption and metabolite production rates were prone to errors. However, the models provide reliable predictions for the MC-based hMSC growth in the
two spinner flask types.
5 Conclusions and Outlook
In this review, the current state of the art of the in vitro expansion of hMSC and the
use of numerical tools to support the development of MC-based hMSCs expansions
as well as the establishment of “Digital Twins” have been presented. It has been
emphasized that different CFD model approaches are described in the scientific
literature which can be successfully applied for the characterization of SU bioreactors, especially for the process development of hMSC expansion processes.
The CFD case study presented clearly demonstrates that numerical models are
valuable tools for the biochemical engineering characterization of small-scale spinner flasks, especially for the determination of parameters that are difficult to determine experimentally. A good correlation was always found between the parameters
predicted by the CFD and those measured experimentally. This observation was also
in agreement with the literature data. The Euler-Euler and Euler-Lagrange models
gave adequate predictions of the MC distributions within the spinner flask systems
and were correlated qualitatively with experimental observations. The EulerLagrange approach allowed the calculation of particle histories due to its discrete
particle formulation, which can be combined with experimental cultivation studies.
Thus, Euler-Lagrange modelling should be favored in the future in order to derive
hydrodynamic stresses over time instead of volume-weighted data. The scientific
literature summarized also shows that different model approaches for the simulation
of the hMSC growth are available, even though only a few are applicable for the
MC-based growth simulation in a stirred bioreactor. The unstructured and segregated growth model presented gives a good description of the MC-based hMSC
expansion process in the two spinner flask systems. Thus, MC-based hMSC cell
growth can be predicted. However, the further development of descriptive, or even
predictive, models for hMSCs will be important in the future for exact scheduling of
the preparation of the cell material and the subsequent autologous therapy.
Numerical Methods for the Design and Description of In Vitro Expansion. . .
221