(343b) Development of An API Micronization Design Space Using a Design of Experiments Approach

The design space for a micronization process used to reduce the particle size of a low solubility active pharmaceutical ingredient (API) has been developed using a resource-efficient D-optimal design of experiments (DoE) study. Micronization trials were completed using an Isopak-2 SuperJet Microniser (APTM, Switzerland).  An experimental plan was designed to examine the impact of three micronization process parameters; feed rate, mill pressure, and Venturi pressure, on the D(v,0.5) and D(v,0.9) particle size averages.  A broad range of the parameter space was examined in the DoE study, but with an additional imposed constraint limiting the difference between the mill pressure and the Venturi pressure to guarantee proper micronizer performance.  Due to the irregular polygonal shaped parameter space caused by the imposed constraint, a 10-run D-optimal design was constructed to ensure both adequate coverage across the experimental region and precise modeling capability.

Statistical models generated from the measured D(v,0.5) and D(v,0.9) values demonstrated that these particle size averages have a strong dependence on the mill pressure and a slight dependence on the Venturi pressure.  The feed rate had no measurable effect on the particle size averages, implying latitude with that parameter.  Three subsequent micronization trials were run to independently verify the model predictions for the D(v,0.5) and D(v,0.9) particle size averages.  Target ranges for the particle size averages were established with consideration of bioavailability, content uniformity, and drug product manufacturability.  Due to its low aqueous solubility, the particle size of the API is limited to D(v,0.5) of no more than 8 microns and D(v,0.9) of no more than 17 microns to ensure bioavailability.  By comparing the model predictions with these target ranges, the design space for the micronization process was identified with the constraint of no less than 2.0 bar for the mill pressure.

Use of a D-Optimal DoE for the efficient development of a robust API micronization design space with parameter constraints has been demonstrated as an alternative to traditional micronization development studies.