Biomedical Engineering Reference
In-Depth Information
There is a working group within FDA, which has been working for the past 6
months, and it will likely be a few more months. It is a slow process, with the
writing and other steps. What specifi cations/limits would you envision for
EDA? For example, would you set a range on the ratio?
[Dr. Terry Tougas—in reply to Dr. Prasad Peri]: This is a big dilemma—
what should be the basis for setting specifi cations? Dr. Peri's slides (given at
this symposium—see http://www.ipacrs.com/PDFs/CI %20Workshop/5-CI
%20Workshop %20- %20Peri.pdf visited September 6, 2012) include “typical
ranges” for stage groupings—we could use those as a basis and translate them
into the EDA specifi cations. Ideally, specifi cations should be tied to clinical
performance but in the absence of a quantitative IVIVC, or something else that
links in vitro and clinical performance, we have to go with what has been
required historically.
[Dr. Marjolein Weda]: Why not use development batches' data to set
specifi cations?
[Dr. Terry Tougas]: That would be setting specifi cations based on capability
rather than QbD. From the engineering perspective, that would be a poor way
to set specifi cations. Today's limits are based on process capability, but there is
limited experience/data at the time of registration. This results in a band that is
too tight for real commercial processes. We can look at the data and use it as a
rough guide regarding expected performance and standard deviation, but the
modern engineering thought around “capable process” recommends a different
approach (e.g., ±6 sigma, instead of ±3 sigma). Currently in the pharmaceutical
industry, if a sponsor does a good job developing a product with a tight distribu-
tion, that sponsor will be penalized with very tight specifi cations. This is coun-
ter to the spirit of continuous improvement.
So, even though there is still an incomplete understanding of all the issues
surrounding AIM and EDA, the overall outcome from the conference was one
of probing the potential for both concepts to become part of the mainstream of
OIP in vitro performance testing.
13.2
Future Directions: Some Further Ideas
An outline for comprehensive product lifecycle management strategy in terms of
in vitro characterization of APSD has been described in Chap. 6 that is based on
simpler yet more statistically powerful effi cient data analysis metrics. This approach
is easily combined with abbreviated impactor measurements. The EDA/AIM
approach could be adopted as the norm for inhaler development and quality control,
but its effective implementation will need to be undertaken on a product-by-product
basis. Although it can strongly be argued that full-resolution multistage CI testing is
less than ideal for QC purposes, such measurements have their place in the initial
product development process, as the fi rst resort in the event of an OOS investigation
and also in OIP change management when in commercial production.
Search WWH ::




Custom Search