Physical Causes of APSD Changes in Aerosols from OIPs and Their Impact on CI Measurements - Good Cascade Impactor Practices, AIM and EDA for Orally Inhaled Products - page 74

an insufficient number of nozzles for a statistically meaningful calculation of limits
for stages 1 and 2, but it is reasonable to expect their tolerances to be at least as good
as for lower stages based on the fact that the nozzles are larger and therefore easier to
measure precisely. Roberts [ 85 ] also showed that D eff and stage d 50 are related through
the expression:

13 /

13 /

4
9

r
phSt

C

Q
n

=

cp

/
()

23

D d

(3.17)

eff 50

50

where Q is the volumetric flow rate, n is the number of nozzles for the stage in ques-
tion, St 50 is the particle Stokes number at which the stage collection efficiency is
50%, and the other terms have been previously defined. Importantly, in the present
context, Roberts reported that the limiting precision associated with stage d 50 can be
kept close to 1.5% of the nominal value by keeping the uncertainty in flow rate to
3% of nominal operating value and at the same time as keeping the uncertainty in
the nozzle diameters to 1% [ 85 ]. The ability to control flow rate within the limit
indicated by Roberts is well within the capability of current flow measuring equip-
ment. In an experimental estimation of the precision of commercially available opti-
cal image analysis systems used to stage mensurate CIs, Chambers et al. [ 86 ]
confirmed that their overall capability was within the current pharmacopoeial stage
specifications for two Andersen 8-stage “nonviable” cascade impactor “reference”
stages that were representative of jet sizes for this instrument type (stages 2;
d eff = 0.914 ± 0.0127 mm and 7; d eff = 0.254 ± 0.0127 mm). These findings confirm
that the 1% uncertainty in this d eff advocated by Roberts [ 85 ] is also a feasible
proposition, in association with a regular program of stage mensuration for a
given CI.
On the basis of these assessments, the capability of the CI method to resolve
APSD shifts is dominated by limitation (1) above.
Changes to OIP aerosol APSDs may also occur in terms of an increase or
decrease in the absolute magnitude of the mass of API that is collected within the
size-fractionating portion of the CI system. Here, the capability of the method is
controlled by the sensitivity of the recovery and assay method for the API [ 38 ]. It
follows that for a given analytical sensitivity, the method will become less capable
of resolving small differences as more stages are incorporated into the system. This
trade-off is especially true when adding stages whose d 50 values are located furthest
from the MMAD of the aerosol being detected, as by definition they will capture the
lowest mass of API per determination. Increasing the number of actuations per
determination is one way to offset such loss of sensitivity, but this approach has
been discouraged by regulators on the basis that the clinical dose of the OIP may be
as small as one or two actuations [ 87 ]. If this option is not available, then an AIM-
based approach, in which the number of intermediate stages is minimized, becomes
an attractive proposition for detecting changes in APSD amplitude. The underlying
rationale for introducing the AIM approach to the assessment of OIP APSDs is