(M0930-07-42) Beyond Particle Size: Impact of Milling Techniques on Agglomeration Behavior and In Vitro Performance of Long-Acting Injectable Suspensions

Purpose: Long-acting injectable (LAI) suspensions offer several advantages including patient compliance, reduced dose frequency, and maintenance of stable drug levels. Despite these therapeutic advantages, development remains limited due to insufficient understanding of formulation-performance relationships. The complex nature of these formulations hinders product understanding, particularly the relationship between formulation parameters (particle size, aggregation behavior etc.) and in vitro–in vivo performance. Variations in processing methods may alter particle interactions, impacting suspension stability and drug release profiles. Additionally, standardized particle size analysis is critical for ensuring consistency across studies, as current methods lack harmonization, complicating data comparison. This study investigates how particle size reduction techniques affect particle agglomeration and in vitro performance of medroxyprogesterone acetate (MPA) suspensions. By clarifying these factors, this research aims to optimize LAI formulations, addressing gaps in product knowledge and guiding future development. Improved characterization of particle dynamics could lead to more predictable in vivo outcomes.

Methods: Three different techniques (antisolvent recrystallization, bead milling & jet milling) were used to prepare MPA crystals with comparable mean particle sizes (small & large). MPA obtained from Spectrum Chemicals was used for processing. Jet milled API was obtained from Medisca. Thermal characterization was conducted to detect polymorphic changes. Powder X-ray diffraction (PXRD) was used to assess crystallinity. Morphology was studied using polarized light microscopy (PLM) and scanning electron microscopy (SEM). Q1/Q2 equivalent formulations of Depo-Provera^® 150, named as S-BM, S-SR, S-JM, L-BM and L-SR, were prepared and characterized. Formulations S-BM, S-SR, and S-JM contained small MPA crystals prepared via bead milling, solvent recrystallization, and jet milling, respectively. Formulations L-BM and L-SR were prepared using large MPA crystals from bead milling and solvent recrystallization, respectively. Particle size distribution of the Q1/Q2 formulations was performed via PLM and Malvern Mastersizer. Surface area was analyzed using BET surface area analyzer. To investigate sedimentation behavior, formulations were stored in vials and kept undisturbed for a week. Agglomeration behavior was investigated under different stress conditions (e.g., low-energy sonication and prolonged stirring). In vitro release testing used USP apparatus IV along with in-house adapters. Release testing was performed using 1% SDS media at 37°C and MPA was analyzed via Ultra-Performance Liquid Chromatography (UPLC).

Results: MPA crystals with similar smaller and larger size particles were obtained using the different preparation techniques. Thermal analysis (DSC and TGA) and PXRD confirmed the crystalline nature of the particles without any change in the MPA polymorphic form (fig 1B, 1C). The S-SR and S-BM formulations had D50 values of ~3 μm, while the S-JM formulation had a larger D50 of ~14 μm (likely due to particle agglomeration). Formulations L-SR and L-BM had D50 values of ~ 20 μm. Despite its larger size, S-JM demonstrated the fastest in vitro release (fig 2A), attributed to the presence of agglomerates containing smaller primary particles and loose flocculates as revealed by SEM and PLM. These observations were supported by the sedimentation value of formulation S-JM, which indicated the presence of loose flocculates (fig 3C). Unexpectedly, L-SR (D50 ~ 20 μm) also showed fast in vitro release (fig 2B), potentially due to its agglomerated structure of small primary particles. These results were supported by surface area measurements which revealed higher surface area for formulations S-JM and L-SR (fig 2D). Low-energy sonication significantly reduced the particle size of S-JM (70–75%), indicating loose agglomerates (fig 3A). While prolonged stirring significantly decreased the particle size of L-SR (65–70%), suggesting initial relatively strong agglomerates (fig 3B). These findings highlight that particle agglomeration and surface area impact the expected release kinetics, and that processing methods can significantly influence suspension behavior regardless of particle size.

Conclusion: This study highlights that beyond the mean particle size, the nature and extent of particle agglomeration are critical determinants of the in vitro performance of long-acting injectable (LAI) MPA suspensions. The agglomeration behavior was influenced by the processing techniques. Although similar D50 values were achieved using different size reduction methods, significant variation in the drug release profiles and sedimentation behavior were observed. These differences were primarily linked to the formation of loose flocculates and the presence of smaller primary particles within agglomerated structures. For example, formulations such as S-JM and L-SR showed unexpectedly fast drug release, highlighting that surface area and particle agglomeration can have a significant impact on release rates. These findings highlight that particle size alone may not be a sufficient predictor of in vitro and potentially in vivo performance. Understanding agglomeration behavior and its impact on surface area is equally important. This study also demonstrates that techniques such as low-energy sonication and prolonged stirring can be instrumental in understanding the agglomeration tendencies of LAI suspensions. The results also highlight the need for harmonized and robust particle characterization protocols, especially when dealing with agglomeration-prone systems. By addressing key formulation-performance relationships, this research provides valuable insight into the design of predictable and effective LAI products.

Acknowledgements: Funding for this project was made possible by a U.S. Food and Drug Administration grant (1U01FD008304-01). Disclaimer: The views expressed here do not reflect official policies of the U.S. FDA or the Department of Health and Human Services, nor does any mention of trade names imply endorsement by the U.S. Government.


Figure1: (A) SEM Images of MPA crystals obtained using different milling techniques revealing different morphology; (B) DSC thermograms of MPA as received and processed using antisolvent precipitation, bead milling and jet milling; (C) PXRD analysis of MPA obtained using different techniques depicting the crystalline nature with no change in the polymorphic form of the API.


Figure 2: In vitro release profiles of formulations S-BM, S-SR, S-JM (A) and formulations L-BM, L-SR; (B) conducted at 37C using 1% SDS release media; (C) Particle size distribution of Q1Q2 equivalents of DepoProvera® 150; and (D) Surface area measurements of MPA particles from different formulations, determined by BET surface area analysis.


Figure 3: Influence of: low-energy sonication during particle size determination (A) and stirring (B) on the particle size distribution of formulations S-JM and L-SR, highlighting the aggregation/de-aggregation behavior. (C) Pictorial representation of sedimentation of suspended particles when kept unstirred for 24 hours. (D) Polarized light microscopy images of formulations S-JM and L-SR at different time points under continuous stirring.