Mechanistic Evaluation of Early Phases of Drug Release From Polymer Microparticles Prepared by Atomization Techniques

Abstract

Poly(lactic-co-glycolic acid) (PLGA) and poly(lactic acid) (PLA) injectable microparticles have been an excellent platform for the controlled-release of small and large molecules, and their manufacturing has been dominated by two encapsulation techniques, solvent evaporation and coacervation. The work herein focuses on two less used atomization techniques with potential advantages—spray drying and electrospraying—and mechanistic evaluation of their early and rapid release periods. Leuprolide acetate, a model peptide drug, was encapsulated in PLGA microparticles by spray-drying. Conventional sample-and-separate and a continuous monitoring system was used to study the initial burst release of 16 compositionally similar formulations and 1-month Lupron Depot® (LD). The continuous monitoring revealed a rapid primary (1°) phase followed by a constant-rate secondary (2°) release phase. The primary phase duration was divided by the characteristic diffusion time to give a dimensionless diffusion time that was highly correlated to the primary phase release. Spray-dried microspheres showed a strong in vitro-in vivo correlation (IVIVC) between the 24-h pharmacokinetic parameters, continuous monitoring in vitro release, and in vitro dimensionless times, whereas the sample-and-separate release was poorly correlated with in vivo absorption of leuprolide. The correlations found demonstrate that diffusion through the polymer matrix as a major release mechanism for highly porous spray-dried microspheres. The poor correlation of LD and a fixed lower limit on the burst release implies that there exists one or more competing release mechanisms. While initial formulations of core shell microsparticles (csMPs) made through coaxial electrospraying were able to achieve high drug loading of the contraceptive hormone—etonogestrel (ENG), they had elevated release rates over the first month and suffered from low yields, System modifications increased particles yield from 10-15% to 25-40%. Initial release rates were associated with the amorphous drug fraction with decreased steadily and steady decreases over the first month. Various methods of exposure with ethanol for the newly formed csMPs were found to increase crystallinity and reduce residual solvents, which led to a slight decrease in the rapid release. 40% ethanol in the collector solution reduced rapid phase release from 20-40% in other treatments to 5%. SEM imaging showed that increasing ethanol concentrations created particles with increasing sphericity. These results were replicated using the strong surfactant, Triton X-100, in place of ethanol, which implicated the importance of collector solution surface tension in defining final csMP morphology. Additionally, some tunability in the rapid release fraction and long-term rates was demonstrated by using blends of PLA or alternative polymers such as high-molecular weight polyethylene glycol (PEG) in shell. Eight disc-shaped csMP formulations were tested in vivo over 3 months. An IVIVC was developed using power law-scaling of in vitro and in vivo times. Mixed effects modeling of the linearized power law time scaling was utilized to analyze individual formulations. The IVIVC model based on in vitro release explained 81% of absorption variability. The predictive power was determined through a leave-one-out analysis. The predicted absorption curves displayed high f2 similarity, and log ratio accuracies of plasma concentrations were greater than 75% for a majority of formulations. In summary, the present work demonstrates the utility of mechanistic evaluation of release and the potential of these analysis for the guided development of coaxial electrosprayed and spray-dried PLGA and PLA microparticles.