(T1530-02-12) Optimization of Spray-Drying Process Parameters for the Preparation of PVP Microparticles Suitable for Nose-to-Brain Drug Delivery

Purpose: Spray-drying is a well-established manufacturing technique to prepare engineered microparticles having attributes suitable for nose-to-brain drug delivery. Particle size is a critical quality attribute (CQA) that determines the deposition pattern of solid microparticles in the nasal olfactory region following intranasal administration. The purpose of this work was to prepare a solid microparticle platform using a mucoadhesive polymer, polyvinyl pyrrolidone (PVP; Kollidon 12 pf), while optimizing the key spray-drying process variables. A design of experiments (DoE) approach was used to optimize the process to produce microparticles suitable for nose-to-brain drug delivery (median diameter ̴ 20 µm).

Methods: PVP microparticles were prepared using a Yamato ADL 311S mini spray-dryer, equipped with a 0.7 mm 2-fluid spray nozzle. Feed solutions were prepared by dissolving the polymer in deionized water at three different solids contents: 30%, 44% and 58% w/v. Dissolution of the polymer was aided by sonication of feed solutions for up to 45 min. Atomization air pressure (P_a) was evaluated at three levels: 0.05, 0.1 and 0.15 MPa. The remaining spray-drying process parameters (inlet temperature = 180°C, flow rate = 8.2 mL/min and blower speed = 0.41 m³/min) were kept constant for all experiments. The diameter and size distributions for microparticles from each experimental run were characterized using a BX-51 optical microscope. The Feret’s diameter for approximately 5050 particles was determined from images obtained using a 10X objective and processed using ImageJ software.

Results: A two-factor, three-level (3²) full factorial design was developed using JMP® Pro version 15.0.0 (SAS Institute Inc., Cary NC) statistical software. Preliminary studies demonstrated that the feed solution solids content (x₁) and P<sub>a (x₂) were critical process parameters that could be used to control the spray-dried microparticle size distribution. The DoE results indicated that d₅₀ increased with increasing solids content of the feed solution and decreased with increasing P<sub>a (Figure 1 and 2). Notably, the experiment that combined a feed solution solids content of 58% w/v and P<sub>a = 0.05MPa resulted in microparticles having d₅₀ ̴21 µm, consistent with values needed to minimize inertial impaction when administered by a commercial insufflator (1). Statistical analysis of the data was performed using two-way ANOVA. The coefficient value of the intercept was 15.53, solids content = 0.153, P<sub>a = -66.7 and interactions = -1.61 with p ˂0.05 for all the parameters. These results demonstrate that solids content, P<sub>a and their interactions influenced the d₅₀. It was also noted that P<sub>a had the greatest impact on d₅₀ followed by the interaction term and the solids content. The model exhibited a good fit to the measured particle size data (Figure 3) with an R² = 0.98931, RMSE = 0.4632 and p< 0.0001.

Conclusion: Effects of solids content, atomization pressure and their interaction were evaluated to screen the spray-drying process parameters for significance. All of the parameters studied in this work had a significant influence on the d₅₀ of PVP microparticles. In particular, the P<sub>a had the greatest impact on d₅₀. Optimization of the process parameters used to spray-drying PVP resulted in solid microparticles having a controllable d₅₀ of approximately 21 µm and distribution ranging from 10-48 µm. These attributes are both consistent with target values suitable for particulate platforms used for nose-to-brain drug delivery.

References: 1. Rigaut C, Deruyver L, Goole J, Haut B, Lambert P. Instillation of a dry powder in nasal casts: parameters influencing the olfactory deposition with uni-and bi-directional devices. Frontiers in Medical Technology. 2022;4.

Figure 1: Graphical representation of the coefficient values for the intercept, solids content, atomization air pressure and interaction of the input variables.

Figure 2: (A) Contour plot showing the unshaded region with a combination of upper level of solids content and lower level of atomization air pressure can result in microparticles with d50 above 20 µm and (B) surface plot of the response (median particle size) using full factorial design.

Figure 3: Actual vs predicted plot showing relationship between solids content, atomization pressure and their interaction with the median particle size for PVP