Study Models
As discussed previously in this review, validation and the development
of viral filtration processes require scaled-down models studies.
However, these studies can be laborious, costly, and time-consuming.
Inadequate study design can translate into improperly sized filters.
Under extreme circumstances, reprocessing may be triggered due to
filtration failure or fouling at process scale. The advent of newer
strategies to improve model design would reduce some of these
hinderances to representative modeling at small scale. One such strategy
is the use of high throughput process development (HTPD) proposed for
resin screenings for viral filtration
(Brown, Johnson et al. 2017,
Brown, Burnham et al. 2018,
Brown, Burnham et al. 2018). Filter plate
technologies in combination with automated liquid and plate handling
systems allow for high throughput analytics in filter process
development. Screening of process conditions (e.g. load conditions) that
may impact critical process parameters (e.g. flux decay ) can be
performed rapidly. Initial work by Tang et al
(Tang, Ramos et al. 2020) provided a
framework for HTPD filter plate screening with recombinant proteins.
These tools can rapidly screen out failure mode conditions early in
process development.
Other new technologies that can aid in refining filter process
development include particle tracking technologies. These particle
tracking technologies use gold or fluorescent particles, as mentioned
previously, but can only track aggregate particle movement. These
technologies can be used to model flow and the impacts of flow
interruptions. One drawback is that these studies tend to have high
material costs. The advent of single-particle tracking technologies such
as the nanoparticles presented by Wu et al 2020 have been used in PVDF
membrane filters. n these technologies, nanoparticles are continuously
imaged, such as with an optical microscope, and tracking algorithms
implemented to localize particle positions and generate trajectories
(Wu and Schwartz 2020) which
theoretically could be adapted to viral filtration flow studies.