(T1230-09-50) Downstream Processing of HPMCAS-Based Amorphous Solid Dispersion (ASD) of Fenofibrate: From Hot-Melt Extrudate to Tablet

Purpose: Downstream processing of amorphous solid dispersion (ASD) formulations into tablets remains a challenge due to the low tabletability of the milled ASD extrudates prepared by hot-melt extrusion (HME). The high viscoelasticity and reduced brittleness of the extrudates makes it difficult to micronize and compact these formulations into tablets with acceptable hardness and tensile strength (1). In addition, for drugs with high tensile strength, melting could occur upon milling causing disruption of the process. In this research we prepared ASD formulations of fenofibrate and HPMCAS-(AS-MMP) (using HME) and investigated the effect of different milling process parameters (milling speed, temperature, and feeding rate, with and without cryo-feeding) on the critical quality attributes (CQAs) of the prepared ASD tablet formulations.

Methods: Fenofibrate (FF) was utilized in this study as a poorly soluble model drug with high tensile strength to prepare the ASD formulations. A blend of FF: AS-MMP (1:4) was extruded using a Pharma 11 corotating twin screw extruder at 160 ^oC using 1 mm die. The extruded strands pre-milled with a hammermill (HW-6, Frewitt) and into the pin-mill (PM-160, Frewitt). Two full factorial DOEs (3²) were utilized for these studies. In DOE 1 the milling feedstock was deep frozen in liquid nitrogen (-196 ^oC) in the cryogenic feeder, then fed into the milling chamber with fixed feed rate (3.3 kg/h) using 3 milling speeds (10,000, 15,000 and, 21,000 rpm), and 3 milling temperatures (-80, -40, and 0 ^oC). In DOE 2 the milling feedstock was at ambient temperature and was fed into the milling chamber at 3 different feeding rates (3.3, 9.6, 16.5 kg/h) using 3 the same milling speeds used in DOE1. The resulting milled extrudates were characterized using laser diffraction (for particle size), differential scanning calorimetry and compacted into tablets using a single punch press. The CQAs of the prepared tablets (tablet tensile strength, in vitro dissolution behavior in buffer pH 6.8 with 0.025 M SLS were evaluated and statistical modelling of the independent and response variables was performed using JMP^© statistical software to study the effect of each process parameter on the response variables.

Results: In DOE 1 (Fixed feed rate, 3.3 kg/h, with cryo-feeding), increasing the milling speed was associated with significantly smaller particle size D50 (p < 0.05) and reduction of milling chamber temperature was associated with a modest decrease in particle size distribution. At the highest milling speed (21,000 rpm) and lowest milling chamber temperature (- 80 C), an average D50 of 20.9 µm was obtained. At the lowest milling speed (10,000 rpm) and milling chamber temperature of (0 ^oC), with no cryo-feeding, an average D50 of 109 µm was obtained. Increasing the milling speed and lowering the milling chamber temperature was associated with a narrower particle size distribution with a span value of < 2 at the highest milling speed (21,000 rpm) and lowest milling chamber temperature (- 80 ^oC). Compacted into tablets, the average tablet tensile strength ranged from 1.407 – 2.527 MPa with milling speed and temperature of 10,000 rpm, and -12 C, and 21,000 rpm, and -40 C, respectively. The in vitro dissolution experiments showed approximately 60 – 70% of fenofibrate released by 30 minutes and 90% released by 2 hours of beginning of the release study regardless of the milling speed and temperature. In DOE 2, (fixed milling temperature, ^o0 C, no cryo-feeding), increasing the milling speed was associated with a significant decrease in the particle size D50 (p < 0.05), while increasing the feed rate had no effect on the particle size. Compacted into tablets, the average tablet tensile strength ranged from 1.168 – 2.554 MPa with a fixed feed rate (16.5 kg/h) and milling speed of 10,000 rpm, and 21,000 rpm, respectively. In the vitro dissolution experiments, a maximum of approximately 60% of fenofibrate was released from the tablets regardless of the feeding rate and the milling speed. Upon increasing the milling chamber temperature to 25 °C, even a D50 of 168 μm was achieved (similar to hammer-milled extrudates at room temperature) without sticking of material to the pins. DSC thermograms of the milled extrudates showed no significant evidence of re-crystallization of fenofibrate.

Conclusion: We demonstrated that pre-cooling of the extrudates (using liquid nitrogen) before feeding into the pin-mill is necessary for successful milling of ASD extrudates of an API with low Tg and high tensile strength (elasticity) such as fenofibrate. Very fine powder of the milled extrudates can be obtained (similar to spray dried powders) with the use of the appropriate milling parameters. This may enable the transfer of ASD drug products prepared by spray drying to hot-melt extrusion, as particle size has a major impact on the CQA of the tablets. Finally, a wide temperature range (- 80 ^oC to room temperature) can be used for the downstream processing of ASD extrudates prepared by HME which simplifies the process.

References: Sauer A, Warashina S, Mishra SM, Lesser I, Kirchhöfer K. Downstream processing of spray-dried ASD with hypromellose acetate succinate – Roller compaction and subsequent compression into high ASD load tablets. Int J Pharm X. 2021 Dec 1;3.

Figure 1: A- Torque values of FF:ASMMP (20:80) during HME process, B- DSC scans of FF:ASMMP (20:80) milled extrudates at different process parameters, C and D include the SEM photomicrographs of the FF:ASMMP (20:80) milled extrudates.


Figure 2: A-FF:ASMMP (20:80) ASD tablets tensile strength at different process parameters according to the design of experiments 1 and 2.


Figure 3: A- In vitro dissolution studies of FF:ASMMP (20:80) milled extrudates at different process parameters according to the design of experiments 1 and 2.