Future Directions for the AIM and EDA Concepts - Good Cascade Impactor Practices, AIM and EDA for Orally Inhaled Products - page 393

by interactive discussions with the audience around technical and statistical
aspects of EDA and AIM and the lifecycle approach to incorporating EDA and
AIM in product development, registration, and manufacturing. The presentations
and a summary report are posted on the IPAC-RS public website [ 1 ]. At the outset,
it is important to note that the conference mandate was on the application to OIP
quality control, so issues concerning the potential application of AIM to the acquisi-
tion of more clinically appropriate data, discussed in Chap. 12 , were not discussed.
Responses to the fi rst group of questions came from the panelists covering the
topic Technical Aspects of EDA and AIM: Incorporating EDA and AIM into the
Development Cycle . Editorial comments are inserted between parentheses “[]”
where needed for clarity, and some responses have added interpretation, based on
information located elsewhere in this topic.

1. Would we need a separate impactor for each product?
No, we need only a couple of different cut points for the boundary between
large and small particles, because the EDA method is robust with respect to
shifts of the boundary (in the studied example presented by Tougas [ 2 ], it dem-
onstrated stable performance when the LPM / SPM ratio was between 0.3 and
3.0). Moreover, all respiratory drugs need to get into the lung, so the sizes of
interest will always be around 2-3 m aerodynamic diameter. Setting the cut

point at a different size (much below or much above 2-3

µ

m) could be consid-
ered if there is an interest in controlling a specifi c part of the distribution.
2. For solution MDIs, would you require CI data for quality control, or would
delivered dose + bulk particle size testing be suffi cient?
Solutions are simpler than suspensions. Laser diffraction could be used for
sizing droplets, but you still need traceability to the API to satisfy regulatory
requirements. For nebulizers, it is possible to build up the traceability chain.
EDA can be extended to apply to nebulizers too.
3. Can EDA detect if something is wrong with the product, or if it is a wrong
product?
The short answer is yes. EDA is an enhancement of the current CI tests.
However, it should be applied only after you have defi ned the aerodynamic
particle size distribution (APSD) of the “right” product. EDA is more sensitive
to changes in APSD than current methods, so yes, you will get a signal that
something is “wrong” if APSD changes, but you would need to do an investiga-
tion, potentially using a full-resolution impactor and/or other tests, to determine
the cause. OOS investigations are done in the current system as well.
4. Analysts and managers need to know how to go about using EDA—not just the
science and statistics but the practical how-to's and what's, so people can get on
with it. We are waiting for IPAC-RS to come up with all the answers.
We share this goal. Part of the process of getting there is obtaining feedback
from users in settings like this one. We are also developing a topic that will set
down “ground rules” for EDA and AIM. Members of the European
Pharmaceutical Aerosol Group (EPAG) are collaborating with IPAC-RS on the
topic writing. In addition, a series of articles has already been published [ 3 ],
and more articles are in the works and will be published in due course.

µ