(T1230-09-52) Exploring Particle Behaviour during Dissolution Testing of Long-Acting Injectable Suspensions - A Potential Role for Image Analysis

Purpose: Dissolution/release testing is an in vitro test in specific pharmacopeial apparatuses to measure drug release over time. This test is not always predictive of in vivo release, but vital in formulation development and quality control testing. Long acting injectable (LAI) suspensions provide drug to a patient’s systemic circulation over time through dissolution of the formulation in the injection depot in the muscular/subcutaneous tissue. Where there is no pharmacopeial testing method for a product, the FDA’s ‘Dissolution Methods Database’ provides recommended dissolution test methods.¹ Shadowgraph imaging (SGI) is a technique that can be used to characterise particle size and dispersal during dissolution in the flow-through apparatus (FTA). This technique records the shadow formed by particles against a lighter background enabling estimation of particle properties during dissolution testing such as such as particle size.^2,3 During analysis, a threshold (Th) value must be set for particle versus background image classification. Depo-Medrone, an aqueous suspension containing methylprednisolone acetate (MPdA)- a poorly soluble steroid drug- is the medication explored in the study. The ‘Dissolution Methods Database’ recommends a 120-minute in vitro dissolution test with 0.55% SLS for MPdA, in the FTA.¹ LAI dissolution in vivo would be much slower than this, especially with limited space in the depot for particle motion, therefore it is of interest to develop a biopredictive test. The aims of this study were to explore the impact of flow rate and surfactant concentrations to identify simple conditions resulting in dissolution over a longer timeframe, and an initial exploration of the potential for SGI to characterise LAI particle behaviour during these tests.

Methods: Dissolution tests were performed in triplicate in a SOTAX CE7 smart FTA (USP IV) as an open-loop system. An 80mg/2ml Depo-Medrone injection was tested in a 22.6 mm diameter cell. Filters were placed at cell outlet- GF/D and GF/F (pore size 2.7 µm and 0.7µm, respectively) glass microfiber filter (GE Healthcare Life Sciences Whatman^TM). Medium containing 0.55% or 0.05% sodium lauryl sulphate (SLS) was used in the dissolution tests with flow rates: 4, 8 and 16 ml/min. Dissolution tests were 70-360 minutes with 4ml samples taken from outlet reservoirs every 10-minutes for 0.55% SLS dissolution tests and longer time intervals for 0.05% SLS tests. Samples were analysed using UV-Vis spectrophotometry (PharmaSpec 1700, Shimadzu) at a wavelength of 247 nm. The SGI set-up (n=1) included a lens and camera (Grasshopper 3 GSB-U3-23S6C (Teledyne FLIR, formerly Point Grey, Ludwigsburg, Germany)) with a Sony IMX174 imaging sensor facing the FTA cell with a sheet to diffuse light and a light emitting diode (LED) opposite the camera lens. The camera was positioned and focussed on the centre of the cell, above the glass beads with a 7 X 11 mm² field of view. Fifteen second videos were taken at the same time as dissolution sampling. Videos taken were analysed with an in-house image processing code in MatLab (MathWorks®, Natick, MA, US) and processed using Th values of 0.1 and 0.05.

Results: The manipulation of simple in vitro dissolution conditions such as decreasing flow rate or surfactant concentration decreases dissolution rate, as seen in Figure 1, but a greater effect is seen from surfactant than flow rate. Although still faster than in vivo dissolution rates, preliminary results illustrate the potential for simpler changes in medium and agitation conditions to reduce the dissolution rate. In all conditions, most of the powder settled on the bed of beads and/or at the top of the cell, (Figure 2), acting as one large agglomerate. The SGI results suggest challenges in identifying particles in 0.55% SLS medium, possibly due to bubbles and rapid dissolution. Although 0.05% SLS conditions proved more promising, image characteristics varied with different flow rates. Preliminary results suggest more and larger particles detected at 16 ml/min.  At 4 ml/min, visual examination of images showed many small particles present, which are therefore difficult to detect and remained undetected in most frames (example of detected and undetected particles Figure 3c-d respectively). Particle detection can be easier when particles are larger and in a more dilute system, where there are fewer particles per frame, (Figure 3a, 16 ml/min). At these flow rates, thresholding did not have an effect on detection numbers. At 8ml/min, thresholding resulted in many false positives being identified through Th 0.1 settings (Figure 3b). Imaging can be useful for insight into particle behaviour, but awareness of the impact of hydrodynamics and thresholding used in accepting and interpreting outputs must be taken into account.

Conclusion: Reducing surfactant concentration and flow rate can decrease the dissolution rate, potentially enabling more biorelevant timeframes and creating the sedimented agglomerates that are more likely to reflect the in vivo environment for LAIs. With further optimisation, SGI is a promising method to characterise the particulate suspension which is not part of the agglomerate.

References: 1. FDA Dissolution Methods Database, https://www.fda.gov/drugs/drug-approvals-and-databases/dissolution-methods-database
2. D’Arcy, DM., Persoons, T., Mechanistic Modelling and Mechanistic Monitoring: Simulation and Shadowgraph Imaging of Particulate Dissolution in the Flow-Through Apparatus, 2011. Journal of Pharmaceutical Sciences, 100(3), 1102-1115.
3. Taseva, A.R., D’Arcy, DM., Persoons, T., Application of an AI image analysis and classification approach to characterise dissolution and precipitation events in the flow through apparatus, Eur. J. Pharm. Biopharm., 189, 36-47

Acknowledgements: This work was supported by Science Foundation Ireland 18/EPSRC-CDT/3587 and the Engineering and Physical Sciences Research Council EP/S023054/1.

Figure 1: Dissolution in USP IV FTA with 0.55% SLS under three flow rates: 16ml/min (dark blue), 8ml/min (medium blue), 4ml/min (light blue); and with 0.05% SLS under three flow rates: 16ml/min (dark red), 8ml/min (medium red), 4ml/min (light red).

Figure 2: Photograph of the FTA cell during 8ml/min 0.05% SLS in vitro dissolution at 14 minutes.

Figure 3: A screenshot from the 60-minute time point of 0.05% SLS dissolutions at: a) 16ml/min, (a typical frame with larger, easily detected particles); b) 8ml/min, Th 0.1, false positives evident; and c) 4ml/min frame at 13 seconds showing particle detections d) 4ml/min frame at 11 seconds (a typical frame at this flow rate with many small undetected particles).