(194c) Troubleshooting Micronization Performance for a Steroid Inhalation API – Conventional and Alternative Approaches

Direct delivery of a drug into the lung requires small API particle size (usually less than 2-3 microns) and a tight Particle Size Distribution (PSD). Both requirements impose significant limitations and challenges on the Drug Substance (DS) manufacturing process. Micronization via jet milling is usually the process of choice to define the DS PSD. However, certain classes of compounds, i.e. steroids, are difficult to micronize and methods to improve micronization performance or even bypass micronization are clearly desirable. In this paper we describe our efforts to troubleshoot the micronization performance of a steroid inhalation compound. Early studies with an Andersen Cascade Impactor (ACI) have shown that an acceptable Fine Particle Fraction (FPF) is obtained when the median (x50) of the DS PSD is smaller than 1.8 µm. Jet milling runs in the pilot plant with 4 and 8? micronizers exhibited very poor performance. In all cases the micronization chamber clogged in less than 1-5 minutes and had to be taken apart for cleaning. In addition, several times the desired PSD target was not met. We attempted to solve the above issues by exploring several alternatives. These included the optimization of the micronization conditions, designing crystallization processes which would either meet directly the PSD target or provide an input to the micronization process with better processing characteristics, exploring alternative micronization techniques or a suitable combination of the above approaches. Amongst the various techniques investigated, the lead candidates were High Pressure Homogenization (HPH), a continuous antisolvent crystallization from THF-heptane systems and co-micronization with lactose. HPH is a wet micronization technique that consistently delivered API material with a median particle size of 1.5-1.6 µm. It was shown to be tolerant of a wide range of input particle size and not susceptible to clogging. Two isolation protocols were tested with HPH, conventional filtration and drying vs. spray drying of the resultant slurry. We found that spray drying results in smaller particle sizes because it reduces aggregation. Another added benefit of spray drying is the elimination of an otherwise lengthy filtration step. Data mining indicated that the API BET SSA (Brunauer, Emmet and Teller specific surface area) prior to micronization could be an attribute controlling micronization performance, which seemed to improve with increasing API surface area. The latter was controlled at the crystallization step. The continuous crystallization process developed, produced API with considerably higher surface area (20-30 m2/g) compared to the previous crystallization process (1-5 m2/g). This API improved considerably micronization performance while still meeting the particle size specification. The time required to clog the micronization chamber was increased from 1-5 min to over 40-60 min, which represents at least an order of magnitude increase in throughput (from 1-2 kg/day to 40-50 kg/day) and alleviated any concerns for the feasibility of commercial production. Similar improvements in micronization throughput were obtained by co-micronization of the API with lactose.