Issue link: https://www.e-digitaleditions.com/i/1229406
20 April 2020 Inhalation size-fractionated mass for this product. In other words, virtually the entire APSD is contained within the FPD. A critical consequence of this is that large shifts in the size dimension of the APSD (significant coars- ening, for example) will elicit only small changes in the FPD. In the present example, the APSD of the newly manufactured pMDI batch (red data) is dramatically coarser than those of the historical batches, as reflected Why does fine particle dose fail? So why does FPD fail to pick out such a glaring outlier in this example? Why is the metric not sensitive to such a large shift in the APSD? ere are two key reasons. First, it needs to be understood that APSD data have two orthogonal dimensions: mass and aerodynamic size. An APSD can undergo changes in either dimen- sion separately, or in both dimensions simultaneously, as illustrated in the examples shown in Figure 2. For example, changes in the delivered dose (i.e., mass per actuation) might impact only the mass dimension of the APSD. On the other hand, a change in the formula- tion composition could impact only the size dimension of the APSD. And, of course, some factors (or combi- nations of factors) could simultaneously impact both the size and mass dimensions of a product's APSD. e bottom line is that to adequately control an APSD, QC testing must be sensitive to changes in either dimension. In more general terms, a system of metrics attempting to describe a distribution must account for both the x- and y-dimensions. is is fundamentally challenging with only a single metric. e chosen boundary of 5 µm is a further limitation of FPD in the QC environment. e cumulative distri- butions from the example (provided again in Figure 3) help illustrate this limitation. Looking at the typical product performance defined by the gray data, we can see that the FPD typically accounts for 95-99% of the Figure 2 An APSD has two orthogonal dimensions: mass and size. For quality control of the APSD of an OIP, the testing and metrics must be sensitive to changes in both dimensions. Mass Mass Mass Mass Aerodynamic Diameter (microns) Aerodynamic Diameter (microns) Aerodynamic Diameter (microns) Aerodynamic Diameter (microns) 0 5 10 0 5 10 0 5 10 0 5 10 Median Median Total Mass Total Mass Figure 3 Cumulative APSDs of a solution pMDI product. The APSD from a new batch (red line) is juxtaposed with APSDs from 200 typical batches (gray). Reference lines show the 5 µm FPD cut-off for each set of data. 100% 80% 60% 40% 20% 0% 0 1 2 3 4 5 6 7 8 9 Cumulative % Mass Aerodynamic Diameter (microns) 99% 95% 85%