Biomedical Engineering Reference
In-Depth Information
the process is robust. The final level of knowledge, first principles understanding,
reflects knowledge of the underlying “why” interrelationships occur, and as a result it is
highly predictive. Insights at this level form a basis for process innovations and
generalized to other process steps.
If the view is taken that the “process is the product,” the result is that the process is
described in terms of descriptive, and perhaps correlative, knowledge. The opportunity
that may seem to be missed is that higher levels of knowledge and understanding are
frequently not developed unless prompted by an excursion or a quality issue. Generally,
most companies have gone far beyond this level, and there is far more data to support their
products and processes; the shortcoming lies in documenting and demonstrating their
understanding to the regulatory bodies, perhaps in order to avoid prompting questions or
discussions that are not necessarily value added in terms of the impact on quality, but
resulting in an understandable tendency on the part of the regulators to take a
compliance-based approach to inspections and a restrictive view to change. Manufac-
turers who can demonstrate the highest levels of understanding, on the contrary, may
enjoy the confidence of regulators and deploy process strategies with greater freedom to
innovate.
13.5.2 Product Design
The area of product design is rich in opportunities for QbD approaches, but less so for
PAT, since once the optimal molecule has been chosen that part of the process is fixed.
The same can be said for the selection of clones and expression systems—there may be
considerable experimentation to determine the best combinations of growth character-
istics and product expression, but once clone selection is finalized that part of the design
is locked down. PAT tools can be employed to investigate how different product
candidates respond in scaled-down processes, and thus provide insight on manufactur-
ability, an important selection criteria when other factors may be closely similar or equal.
High-throughput screening tools are of considerable value in this area.
13.5.3 Process Design and Development
Developing a process that integrates PAT tools and principles and demonstrates science-
and risk-based process understanding will give rise to an approach that deviates from the
traditional filing strategy, which was focused on the process and resulted in largely
descriptive submissions. The Standard Practice for Process Design using PAT was
released in 2006 [30], and it describes eight basic practices that should be followed to
ensure proper design, reducevariation, andmeet the requiredacceptance levels for quality
(Table 13.4). The approach is equally applicable to drug substance and drug product.
Given that the emphasis from the agency has been on a risk-based approach to
quality, the first practice is risk management through assessment and mitigation, which
is applicable throughout design. The objective is to reduce variability and ensure safety
of the process through formal risk evaluation procedures. Risk assessment and control is
the subject of a separate standard currently under development; clinical safety is
addressed in clinical trials.
