Drug shortage of oral corticosteroids: Development of a prednisone 5 mg capsule formulation adapted to an automated preparation process

Introduction
Faced with a constant increase in drug shortages and following the publication of new Good Preparation Practices allowing the production of larger batches, the automation of preparations seems to be an essential alternative for our center, which has an automatic capsule filler (INCAP SE, Bonapace). In response to tensions in the supply of oral corticosteroids, our unit has begun a feasibility study into automated manufacturing. The aim of this work is to validate a formulation of prednisone 5 mg capsules and to qualify the automated process enabling this formulation to be used.

Material and method
Two excipients developed for automated capsule formulation, composed of microcrystalline cellulose (MC) (PROSOLV SMCC 90HD^®, JRS) with silica or without (VIVAPUR 302^®, JRS) were selected. For each production run, 300g of powder was prepared using a three-dimensional blender at speed 8 for 20 minutes (Inversina 2L, Bioengineering) to produce 600 prednisone size 2 capsules using the INCAP SE capsule filler. In order to validate the formulation, the parameters studied were the flow of powder in the hopper and the fouling of the capsule filler. To qualify the process, the mass of the capsules was used to calculate the process capability indices and produce control charts using R software (qcc library). An analysis of the active ingredient content using high-performance liquid chromatography coupled with a UV-visible spectrometer was carried out on the powder mixture before filling (surface, middle, bottom) and on the capsules (start, middle and end of production) (n=3). The aim of these assays was to investigate the homogeneity obtained during the mixing phase and the heterogeneity (unmixing) during capsule filling.

Results
As the formulation without silica did not meet the flow criteria, the final formulation with silica was tested in two production runs. The average masses of the capsules of the final formulation are 235.29±1.8 mg and 231.59±2.35 mg with a capacity of 2.17 and 1.92 indicating a process under control. The productions obtained less than 5% deviation from the target mass. The average prednisone content of the powders prepared with the mixer was 5.04 mg±0.11 and 5.08±0.1 mg, with no variation depending on the sampling point. The average contents of the filled capsules were 5.29±0.05 mg and 5.33±0.08 mg respectively (± 6% theoretical value). The control charts for this formulation do not show any capsules outside the limits.

Conclusion and discussion
The formulation, composed of microcrystalline cellulose with silica and prednisone 5mg, was validated. The data obtained ensures that the process is under control and that Prednisone 5mg capsules are highly productive.

See the poster