Abstract

To date, assessments of suspension creaming profile have been performed by different methods based on visual inspection[1], dose uniformity for long term stability studies[2], and light backscattering [3-5]. However, as the technological complexity increases so does its cost. Cosuspensions of drug and spray dried porous lipid particles have been shown to generate robust aerosol delivery performance in pressurized metered dose inhalers (pMDI) formulations [6,7]. The micronized drug crystals and the porous particles form stable ensembles, which create slow-creaming suspensions in propellant [6]. Analysis of creaming profile is critical to suspension stability is a requirement in this formulations to assure dose uniformity, however it is difficult and requires expensive instrumentation. To develop the shadow motion analysis (SMA) technique, as a low cost alternative to investigate the effect of the concentration of a drug on the creaming process of pMDI co-suspensions. The suspension contained in the glass vial cast a shadow of the particles suspended in the liquid propellant when the LED light is on. The shadow is projected toward the monochromatic camera, which captures the data as intensity pixel values. Then, the shadow motion is assessed by image analysis. Figure 2 summarized the average value of the three measurements, showing the change of the area by time. Representative images of the frames collected and analyzed at comparable time points are displayed in the figure to ease the analysis. A general equation for describing the creaming rate has been developed (figure 3). The data was fitted to a nonlinear, four parameter logistic model to determine the decay rate of the creaming process. For this equation (Areat ) is the area of the shadow at different time points, (Areaf ) is the area measured in the last frame captured in the study, K is the decay rate (creaming rate) and Ip is the inflection point in the curve. Additionally, figure 3 shows a plot of observed vs fitted curve data for equation 1, using the data from which the equation was derived. Reproducibility of the results was time dependent: The SMA system demonstrated being highly reproducible when analyzing the beginning of the creaming process, however, the results obtained at the end of the process were variable depending on the drug loads (as shown by the SD in figure 2). We hypothesize that an increased particle-particle interactions leads to more flocculation and more variability in the high dose compared to low dose formulation. Creaming rate are similar even though the creaming process starts later: The examination of the data obtained by image analysis allowed to identify a sigmoidal behavior that governs the creaming process (Figure 3). Data obtained from the image analysis was fitted to the sigmoidal model with a high correlation between the fit model and the observed areas at all tested conditions The SMA method is potentially a novel & low cost tool to aid development of MDIs. The creaming profiles obtained by SMA allowed to compare and identify the starting point and the creaming rate in an easy way .Identifying the reason for the time-dependency variability in the creaming profiles for the formulation at high drug load is out of the scope of this study. However, we hypothesize that the increase in micronized drug particles on the surface of the carriers affects the interaction between the lipid porous particles and the final particle density. This translates in a delay of the flocculation and the creaming formation, which causes a less consistent process. Further studies will focus on the limitations of the 2D approach in the SMA method, by evaluating the effect of the diameter and the material of the canister on the creaming profile.