Biomedical Engineering Reference
In-Depth Information
TABL E 8.1. Essential Differences Between Current and Design Space Process
Characterization
Technical
Approach
Process
Variability
Typical Learning
Current
Empirical
Not systematically
assessed
If the manufacturing process
conditions match the test conditions,
virus clearance is reasonably assured
Milestone
Statistical and
mechanistic
Main effectors
understood
The impact to changes of the key
independent variables is understood.
Manufacturing deviations and process
improvements may be justifiable based
on knowledge generated
Design space
(end state)
Statistical and
mechanistic
Full system
understood
As above, and with a full understanding
of all effectors and interactions
between independent variables
manufacturing process design space has been effectively established will have greater
confidence that the manufacturer is making decisions from a strong knowledge base.
This will lessen the requirement for regulatory oversight, allowing regulators to
reallocate resources. In parallel, manufacturers gain freedom to change manufacturing
processes within a multidimensional window without onerous regulatory consequences.
The ultimate benefits of this better process understanding must translate directly to the
patient through consistent viral safety assurance.
8.4 TECHNICAL LIMITATIONS RELATED TO ADOPTION
OF QbD/DESIGN SPACE CONCEPTS IN VIRUS CLEARANCE
Developing an in-depth knowledge base for a virus clearance technology requires tools
and technologies for virus clearance quantification. Several of these technologies have
limitations that directly impact the cost and feasibility of achieving a fully characterized
virus clearance design space.
8.4.1 Definition of Product CQA
A QbD approach requires one to set critical quality attributes (CQAs). In areas such as
potency, purity, stability, and so on, the end points are assignable in real number, nonzero
terms. The virus clearance level CQA is commonly stated as “zero virus in the product”
or a “sufficient level of virus safety.” Neither of these end points is actionable. The former
suffers from the statistical issues inherent in claiming “zero” and the latter lacks
quantification. This issue is faced by the industry today, and finds strong analogy in
(e.g., steam) sterilization processes, for which the industry has established standards for
quantifying process effectiveness.
