(M1030-02-12) Decoding Choice of Surrogate Compounds for TSMG Empirical Modelling of Gabapentin - A Case Study

Purpose: Twin-screw melt granulation (TSMG) is a viable manufacturing process for continuous granulation. However, feasibility studies require large amounts of material and costs, especially for thermolabile drugs like gabapentin (GABA). Initial attempts to develop a material-sparing model for stability assessment of GABA using DSC showed promise, with good correlation between measured and predicted degradation¹. One of the inputs for this model was peak granule temperature, measured through GABA empirical runs. Peak granule temperature results from local heat generation and is influenced by granule growth and strength which depends on  particle size, flow and rheological properties of the blend¹. It was hypothesized that using appropriate surrogate excipients in TSMG could mimic the powder flow, resulting in similar granule temperatures without GABA. Therefore, the main aim of this study was to investigate the selection of surrogate excipients to make empirical model more material-sparing during TSMG.

Methods: Mannitol and Lactose were chosen as surrogate excipients for investigation due to their widespread use as diluents in oral-solid dosage forms and immiscibility with binder². A D90 of 100-150 µm was selected for mannitol/lactose to maintain consistency with gabapentin. Klucel™ EXF (Hydroxypropylcellulose) was chosen as thermal binder and binder level was fixed at 10 % based on the developed empirical model with gabapentin¹. To ensure consistent packing, the D90 of mannitol and lactose was chosen to match the range of gabapentin.  Powder flow in twin-screw processing was evaluated using a real-time feed flow study using a Brabender volumetric feeder at a fixed speed. Granulation trials were conducted on Leistritz ZSE-18 twin-screw corotating extruder with a 25 L/D barrel configuration, equipped with 30 mm kneading block (Staggering angle 60°). The feed rate was kept constant while barrel temperature (130-140 °C) and screw speed (50-200 rpm) was varied. Peak granule temperature was measured at discharge using a Thermodyne FLIR E54 camera. The onset of granulation and morphology changes were studied using screw pullout technique, collecting samples from the different zones in the extruder.

Results: Granule growth was governed by the rheological properties, powder packing and flow through the twin-screws despite feeding the blend gravimetrically at a constant feed rate. Material flow through the twin-screws was monitored using a real-time 30-second moving average of the feed over a 2-min period.  Gabapentin and lactose are crystalline, free-flowing materials, and their blends exhibited similar feed rates of around 18-19 g/min at constant screw speed (Figure 2). The bulk density for both blends was approximately 0.5-0.6 g/mL, indicating similar powder packing and flow inside the extruder, which could lead to a comparable granule growth pattern. Mannitol, on the other hand, is a cohesive, poor-flowing powder, as reflected by the blend feed of 9 g/min at the same feeder speed (Figure 2). The bulk density of mannitol blend was approximately 0.35 g/mL, significantly lower than that of gabapentin/lactose which may lead to different powder packing and granule growth patterns. This was confirmed by the formation of an incompressible plug during TSMG trials across similar processing conditions. Addition of only 0.2% Aerosil 200 (silica) significantly improved the flow increasing blend feed to 21 g/min at the same feeder speed (Figure 2). Bulk density of the new blend was similar to gabapentin at 0.56 g/mL. Similar torque readings across various processing conditions for gabapentin, lactose and mannitol (w/ Aerosil 200) suggested similar mechanical energy input for granule growth. Statistical analysis showed no significant difference in peak granule temperature between gabapentin, lactose and mannitol (w/ Aerosil 200) blends during TSMG trials under varied processing conditions (Figure 3). This demonstrated the potential of using lactose or mannitol (w/ Aerosil 200) as a surrogate compound to make the developed empirical model more material sparing.

Conclusion: The developed model relied on empirical TSMG runs with gabapentin for peak-granule temperature measurements. Peak-granule temperatures for TSMG runs with mannitol and lactose are like gabapentin demonstrating the potential of surrogate compounds to make the model material-sparing. The choice of surrogate compounds was influenced by the material flow, rheological properties and particle characteristics. The untapped potential of a real-time feed flow study using a twin-screw feeder was observed, as varied material flow inside the extruder affected powder packing despite a constant feed rate during TSMG. Further research is being conducted to elucidate the changes in rheological properties at the kneading zone with these surrogate compounds.

References: 1. Pradhan, A., Yang, F., Karan, K., Durig, T., Schwing, Q., Haight, B., ... & Zhang, F. (2025). Material-sparing degradation-kinetics model for thermolabile drug stability assessment during twin-screw melt granulation–Insights with gabapentin. International Journal of Pharmaceutics, 674, 125421.
2. Pradhan, A., Costello, M., Yang, F., Bi, V., Durig, T., & Zhang, F. (2022). Using twin-screw melt granulation to co-process mannitol and hydroxypropylcellulose. Journal of Drug Delivery Science and Technology, 77, 103880.
3. Kittikunakorn, N., Paul, S., Koleng III, J. J., Liu, T., Cook, R., Yang, F., ... & Zhang, F. (2021). How does the dissimilarity of screw geometry impact twin-screw melt granulation? European Journal of Pharmaceutical Sciences, 157, 105645.

Acknowledgements: The study was sponsored by Ashland Specialty Ingredients

Figure 1. Bulk density of Gabapentin and surrogate compounds tested for TSMG using cup method.

Figure 2. Real time feed flow study demonstrating how different materials will flow differently inside the extruder despite similar particle size and a constant feed rates.

Figure 3. Peak granule temperature and torque readings across three processing conditions from the empirical model confirming similar values for gabapentin, lactose and mannitol (w/ Aerosil 200) blends