The "FDA Guideline General Principles on Process Validation" is
now 19 years old, and the process validation concept described in it
"may now be ready for the trash bin" - or at least that is what
Laura Bush writes in her article on process validation in the online
edition of the US magazine Pharmaceutical Technology. Since many
companies endeavour to understand their processes better, traditional
process validation becomes less important.
Experts agree that, in the course of time, the issue of process
validation has gradually turned into extensive documentation instead of
focussing on its proper meaning: assuring quality. A whole branch of
industry has emerged around the topic of process validation, leading to an
inflated documentation. And still there are processes that do not
"work correctly".
The new keyword is process understanding. Up till now, many processes
were not thoroughly understood. However, validation activities based on
insufficient process understanding do not make sense, according to
Professor Muzzio from Rutgers University. As further weaknesses of the
current validation model, Muzzio sees the practice that the Quality
Control has the task to guarantee product quality by discarding bad
batches. In his opinion, that is inefficient.
Another expert, Dr Philippe Cini from Tunnel Consulting, states that medicinal products do reach a high standard
relative to the Six Sigma quality approach. However, this
high quality is not built into the product during production, but is
achieved through quality management systems discarding bad products.
A further weakness of the current validation concept lies with the
avoidance of changes at the cost of even preventing improvements, since
changes could always be relevant on a regulatory level. This can go as far
as to refrain from generating new data in order to avoid the impression
(which could be evoked by the new data) that the company does not
understand its own processes.
In case of new products, the firms often know relatively little about
the corresponding processes, since the development phases are kept as
short as possible in order to be able to market the products soon. Every
day passed without the new product being on the market is considered as a
financial loss. Consequently, it is not until commercial-scale production
begins that larger amounts of data on the process are generated.
With regard to this practice, we are witnessing a change at present.
The companies have begun to invest more time in developing and improving
their processes.
Application examples of PAT studies by Pfizer, AstraZeneca and
Genentech confirm this change. Interestingly, these companies have
different objectives: Pfizer puts the emphasis on reducing deviations,
AstraZeneca aims at "real-time release", and Genentech places
special focus on early stages of process development. Here, Genentech
chooses Design of Experiments (DoE) as a means to achieve process understanding.
Surprisingly, when this method was first used for the development of the
monoclonal antibody Avastin, time could be saved because the really
critical parameters (in a narrow range) were found using DoE. Further
tests then focus on just these parameters.
For Genentech, this approach was also the key to quickly obtaining the
marketing authorisation, as their way of proceeding enabled them to hand
in statistically valid data and validation could be limited to a small
number of parameters.
In the future, industry and authorities will concentrate more on
"good science". To Ron Branning from Genentech, it is less
important to practise exhaustive science than to focus on the issues at
hand. Prof Muzzio, too, sees that, as soon as a new quality standard
is achievable, it leads to higher expectations.
Click here for the complete Pharmaceutical Technology article:
http://www.pharmtech.com/pharmtech/article/articleDetail.jsp?id=173672
|
