(T1430-12-83) Design of a Re-Dispersible High Drug Load Amorphous ABT-530 Formulation: Impact of Formulation Composition on Processing and Performance

Purpose: High drug load (60% w/w) re-dispersible amorphous nanoparticles (ANPs) formulation provide an alternative approach to conventional ASD technologies including hot melt extrusion (HME) and spray drying dispersions (SDD) which typically have low drug load ( < 20% w/w). Literature reports indicate that drug rich nanoparticles are formed in situ when congruent release of API and polymer occurs from the ASD. These nanoparticles are postulated to maximize flux through GI membrane, which combined with the higher apparent solubility ultimately enhances bioavailability. ANP formulation delivers directly engineered drug rich nanoparticles, thus eliminating excipients requirement for in situ nanoparticle generation, resulting in a much higher drug load. In this study, ANP formulation using a non-ionic surfactant Vitamin E TPGS (TPGS) was compared to ANP formulation containing sodium dodecyl sulfate (SDS) used in our previous work.¹The focus here is to understand the impact of formulation composition on processing and performance.

Methods: ANP formulations of model compound ABT-530 (BCS class IV) were prepared using solvent/anti solvent nanoprecipitation, organic solvent removal through thin film evaporation (TFE) or tangential flow filtration (TFF), followed by spray drying (SD) or freeze drying (FD) to achieve a solid intermediate form of drug loading ≥ 60% (w/w. Characterization of ANPs by dynamic light scattering (DLS), ζ-potential, filtrate potency assay (FPA), scanning electron microscopy (SEM), and differential scanning calorimetry (DSC) revealed differences in surface property and thermal property of nanoparticles afforded by surfactants. The performance of ANP formulations was assessed by re-dispersion of solid intermediates in pH 2 and pH 5-6 media.

Results: Nanoparticles of average diameter around 150 nm were obtained for both formulations containing ABT- ABT-530/PVPVA/TPGS-90/5/5 (S_TPGS^IJ) and 530/PVPVA/SDS-90/5/5 (S_SDS^IJ). ζ<-potential showed a close to neutral value of -12 mV for S_TPGS^IJ and around -50 mV for S_SDS^IJ, not surprisingly indicating significant impact of surfactant charge on surface charge. Organic solvent was removed from S_SDS^IJ using TFE (S_SDS^TFE), while it was necessary to use TFF for S_TPGS^IJ (S_TPGS^TFE), because particle size increase was observed with TFE. Wet glass transition (Tg) of nanoparticles in suspension after IJ and after organic solvent removal are shown in Table 1. Single Tg was detected in S_SDS^IJ and S_SDS^TFE. Additionally, there was significant increase in Tg after removal of organic solvent, from 66 °C to 94 °C. Single Tg was also detected in S_TPGS^IJ at around 59 °C, similar to S_SDS^IJ; however two Tg events were detected in S_TPGS^TFF, 55 °C and 90 °C. The difference in thermal property necessitated processes (including solvent removal and drying) to operate at a lower temperature to ensure re-dispersion after drying. The difference in surface charge between SDS and TPGS ANPs indicates that the stabilization mechanism is different. Surface charge provides electrostatic repulsion in addition to steric hindrance, thus SDS ANPs had stronger stabilization. In drying, the stronger stabilization translates to smaller/no additional quantities of excipitents to ensure redispersion. As a result, SDS ANP was able to achieve 90% DL in solid intermediate, whereas TPGS ANP required additional Copovidone (PVPVA) resulting in 60% DL. Re-dispersion of SDS and TPGS ANPs in pH 5-6 medium (DI water) back to primary nanoparticles within 10 minutes was successful [Figure 2 (a,d)]. However, redispersion of SDS ANP in pH 2 medium showed rapid precipitation following initial re-dispersion, leading to settled solids with a clear solution phase [Figure 2(b)]. On the other hand, redispersion of TPGS ANP was independent of pH, resulting in milky suspensions in both low pH and high pH media [Figure 2(a-b)]. Re-dispersion of ANPs in these pH ranges are relevant for oral administration where stomach fluid is typically at a lower pH than the intestinal fluid. It is hypothesized that dodecyl sulfate ion interacted with ABT-530 cation (ionized in pH 2 medium) on the surface of the nanoparticles, leading to precipitation.

Conclusion: In this work, a re-dispersible 60% (w/w) drug loaded amorphous nanoparticle formulation of ABT-530/PVPVA/TPGS was successfully prepared using the IJ – TFF – FD process. It was demonstrated that this formulation had excellent pH independent redispersibility. By comparing it with SDS ANP formulation from our previous work, it was clear that surfactant selection and formulation compositions of amorphous nanoparticle had an important role in determining nanoparticle behavior, process considerations and formulation performance. The impact of these findings on in vivo performance and stability are subject of future work. It is important to note that the anionic surfactant (SDS) allowed high temperature processing and higher drug loading but the formulation exhibited pH-dependent dispersibility which would necessitate enteric protection and be detrimental for compounds that exhibit regional absorption. Whereas most ASDs on the market use non-ionic surfactant and here we find that TPGS necessitated lower temperature processing and lower drug loading than SDS but a pH independent release.

References: 1. Oberoi, H. S.; Arce, F.; Purohit, H. S., et al. Design of a re-dispersible high drug load amorphous formulation. J Pharm Sci 2023, 112 (1), 250-263. DOI: 10.1016/j.xphs.2022.10.002.

Figure 1. (a) Solvent/antisolvent nanoprecipitation was used to generate nanoparticles for SDS and TPGS ANPs. (b) Size distributions of SDS and TPGS ANPs in suspension. (b) ζ-potentials of SDS and TPGS ANPs in suspension.

Figure 2. Redispersion of P_SDS in DI water (pH 5-6) and 0.01 N HCl (pH 2) by visual and DLS (a-c). Redispersion of P_TPGS in DI water (pH 5-6) and 0.01 N HCl (pH 2) by visual and DLS (d-f).

Table 1. Wet glass transition temperature (Tg) of samples S_SDS^IJ, S_SDS^TFE, S_TPGS^IJ, and S_TPGS^TFF. All Tg values were reported to be the midpoint of the glass transition event.