(W0930-09-58) Continuous Wet Granulation and Drying: Moving from Technical Feasibility to Commercial Scale

Purpose: Many industrial processes use continuous manufacturing due to its operational efficiency and reduced facility requirements. Pharmaceutical manufacturing has traditionally used batch processing, though there is a growing trend towards continuous manufacturing backed by operationally validated commercial processes. Among pharmaceutical processes, wet granulation has been more challenging to adapt to continuous processing. The adaptation of twin-screw extruders used for polymer processing and recently improved designs for continuous dryers has allowed for simplified transition from batch to continuous manufacturing. The goal of this study was technical evaluation of new equipment for pilot scale (1 – 5 kg/h) continuous granulation and drying, comparison to batch processing, and scale-up to commercial scale (10 – 25 kg/h) continuous wet granulation and drying.

Methods: Pilot scale testing was with placebo formulations of microcrystalline cellulose and lactose as well as active granule formulations containing two proprietary APIs, MK-A & MK-B. Granules were manufactured using a pilot scale QbCon®1 (L.B.Bohle) integrated systems of a loss-in-weight feeder, continuous twin-screw wet granulator with liquid injection, and continuous vibratory fluid dryer. These granules were compared with conventional batch wet granulation and twin-screw wet granulation with a 18 mm twin-screw extruder (Leistritz) followed by tray drying. Commercial scale testing was performed with placebo and an active granule formulation of one of the proprietary APIs, MK-B. Granules were made with the QbCon®25 (L.B.Bohle) granulation and drying equipment and a 27 mm twin-screw extruder (Leistritz) followed by tray drying. For each granulation equipment, a range of solid feed rates, screw speeds, and granulation fluid levels (GFLs), controlled by the ratio of the liquid and solid addition rates, were studied. Drying airflow, air temperature, and bed acceleration were adjusted for the QbCon® systems to produce granules with < 2% loss in drying (LOD, measured with HR 73, Mettler Toledo). Dried granules were milled, then tested for bulk and tap density (Autotap, Anton Paar), particle size distribution (Qicpic with Rodos dry disperser, Sympatec), and imaged using scanning electron microscopy (Quanta FEG 250, FEI). Granules at selected process conditions were tumble blended (Turbula, WAB AG) with extragranular excipients and tableted across typical tableting forces using compaction simulation (Roland Research Devices, Inc. & HB100, Huxley-Bertram).

Results: The QbCon® systems enabled rapid evaluation of process parameters as the conditions were altered and representative granules were collected in a few minutes. The vibratory fluid-bed dryer (VFBD) dried the granules to < 2% LOD within a few minutes, at up to the 30% GFL studied. The QbCon®25 was also able to dry to this target across a wide range of process conditions. Dried, milled granule size distributions, densities, and tableting performance were similar for the QbCon®1 system and reference batch processes. Figure 1 shows the mean granule size across the process space studied, with larger granules at higher GFL and specific throughput (ratio of solid feed rate to screw speed). Figure 2 shows the tablet tensile strength versus compaction pressure. Tablets of granules made with the QbCon®1 and 18 mm Leistritz were much stronger than tablets made with the reference HSWG process. Unmilled granules from the QbCon®25 and Leistritz 27 mm were larger than from the pilot scale processes; after milling, the granules were similar in size and quality parameters.

Conclusion: The results across the studied formulations demonstrate that translation to continuous processing is straight-forward and can result in similar or improved granule quality and product robustness. The process can also readily be scaled up from pilot scale. These positive technical results increase confidence that the operational advantages of continuous wet granulation and drying over batch processing can be realized without compromising product quality. Beyond technical readiness, comparing these trials to typical batch processing trials and scale-up shows the potential for accelerated development from early clinical batches through commercial supply.

Acknowledgements: Thanks to Mike Fazio, Imre Homolya, Martin Hack from L.B. Bohle, LLC for their support in evaluations using the QbCon®1 and QbCon®25. Thanks to Augie Machado and Steve Post from Leistritz for their support in process trials with the 27 mm twin-screw extruder.

Figure 1: Contour plot of mean granule size by L/S and specific throughput, placebo and MK-B formulations from QbCon®1

Figure 2: Comparison of tabletability for MK-B formulations made at pilot scale with different equipment