Purpose: The purpose of the present study was to investigate the feasibility of implementing a 3D high speed imaging technique (Eyecon™) for in-line real time determination of granule size and distribution (GSD) during comil operations as part of High Shear Wet Granulation (HSWG) process development.
A formulation with 50% Active Pharmaceutical Ingredient (API), microcrystalline cellulose and croscarmellose sodium was used. The HSWG process used water as the granulating liquid. Granules were dried in Fluid Bed Drying (FBD) and comiled before tablet compression. A Design of Experiment (DoE) was used to study the effect of API particle size, amount of water for granulation, water addition time and granule Loss on drying (LOD) after FBD on GSD after comilling the dryed granules. Liquid-to-solid (L/S) ratios were in the range of 0.5 to 1.2. A custom-made chute was designed to connect the Eyecon™ to the lower part at the exit of the comil to capture in-line imaging data at high speed. Batches were tested in-line and off-line with Eyecon™ and with two other off-line techniques (traditional sieve analysis and the Malvern imaging analyzer Morphologi G3™). The D10, D50, and D90 values from the different techniques were calculated and compared. Statistical analysis (JMP™ by SAS) was applied to determine factors within statistical significant effects on the GSD. Wet granules collected after wet granulation were also dried by the Tray Drying (TD) method and compared to FBD. The final quality of the tablets was tested with NIR chemical imaging.
statistical analysis was performed to determine the processing factors with significant effects on the GSD determined by Eyecon™ imaging analysis. Statistically significant effects were identified if the calculated p-value was ≤ 0.05. The main effects %water for granulation and water addition time were statistically significant. High %water (represented as high L/S ratios) showed increased in the D10, D50, and D90 values. Longer water addition times showed the opposite effect with a decrease in D10, D50, and D90 values. Other factors such as the LOD end-point during FBD and the API particle size (or combination of these factors) didn’t show any statistical significant effect on the GSD. There was a very good correlation between the off-line and in-line GSD values determined by Eyecon™, demonstrating that the experimental set-ups for analysis were optimal. Comparison of sieve analysis demonstrated that although the absolute D10, D50, and D90 values from both methods were different, the general trends within the DoE were similar. Eyecon™ analysis showed significant difference in GSD of tray dried vs. FBD batches. Tray dried samples showed a much narrow GSD, with less %fines compared to FBD samples (where higher attrition was observed). Morphologi G3 was used as an alternative off-line direct imaging technique that provided additional morphology data. This technique allowed for a better detection of smaller particle size ranges, recording fines (5-1000 μm), but this technology isn’t optimized for granules and can’t be used in-line. GSD analysis supported the selection of optimal process parameters. Chemical imaging from compressed tablets demonstrated good content uniformity and tablet quality.
This study demonstrated the feasibility of using in-line Eyecon™ imaging analyzer to provide real-time GSD data to support process development studies of HSWG and comil operations. %water for granulation and water addition time were found to be statistically significant factors in the GDS data. In-line and off-line GSD data determined by Eyecon were compared demonstrating consistency between the two methods.
Daniel McNamara– Principal Scientist, Bristol-Myers Squibb Company
Kyle Martin– Principal Scientist, Bristol-Myers Squibb Company
Gregory Lane– SR RES INVESTIGATOR I, Bristol-Myers Squibb Company
Douglas Both– Research Fellow, Bristol-Myers Squibb Company