In Situ - Forming Microparticles for Controlled Release of Rivastigmine: In Vitro Optimization and In Vivo Evaluation

Abstract

In this work, sucrose acetate isobutyrate (SAIB) and polylactic co-glycolic acid (PLGA) were used alone or in combination as a matrix-former (MF) to prepare long-acting injectable rivastigmine (RV) in situ-forming microparticles (ISM). RV-ISM were prepared by the emulsification of an internal phase, containing the drug and the matrix former(s), into an external oily phase containing a stabilizer. The statistical design, Central Composite Design (CCD), was adopted as a quality by design (QbD) approach to optimize the formulation of RV-ISM systems. The fabricated RV-ISM systems was designed to minimize the initial burst drug release and maximize the sustainment of RV release from the ISM and ease of injection. The influence of critical formulation variables such as the matrix-former to drug (MF/D) ratio and SAIB to PLGA (S/P) ratio in the internal phase with respect to critical quality attributes (CQAs), such as the percentage drug release within the first day (Q₁), the time required for 50% drug release (T_50%) and the rate of injection, were studied using the CCD. The optimal RV-ISM system with the highest desirability value (0.74) was predicted to have an MF/D ratio of 11.7:1 (w/w) and an S/P ratio of 1.64:1 (w/w). The optimal RV-ISM system was assessed for its release profile, injectability, rheological properties, morphology, effect on cell viability, tolerance to γ-sterilization and in vivo performance in male albino rabbits. In vitro release studies revealed that the optimal RV-ISM system released 100% of its drug content throughout a release period of 30 days with only 15.5% drug release within the first day (Q₁) and T_50% of 13.09 days. Moreover, the optimal system showed a high injection rate of 1.012 mL/min, pseudoplastic flow, uniform spherical globules with homogenous particle size, minimal cytotoxicity and high tolerability to γ-sterilization. In vivo pharmacokinetic (PK) studies revealed that the rate of absorption of RV from the optimal RV-ISM system was controlled compared to a drug solution following either intramuscular (IM) or subcutaneous (SC) injection. Furthermore, the optimal RV-ISM was found to follow flip-flop PK with poor correlation between in vitro release and in vivo findings. These findings suggest that the optimal RV-ISM is a promising tool to achieve a sustained release therapy for RV; however, further investigation is still required to optimize the in vivo performance of RV-ISM.

Keywords: depot release; in situ-forming microparticles; optimization; rivastigmine; sucrose acetate isobutyrate.

Figure 1

In vitro release profiles of RV from the prepared RV-ISM systems as per the experimental design in phosphate-buffered saline (PBS, pH 7.4) at 37 °C (A) F1–F3 systems prepared using PLGA, (B) F4–F6 systems prepared using SAIB and (C) F7–F9 systems prepared using SAIB/PLGA 1:1, in comparison to RV solution in NMP. Data points are mean ± SD (n = 3 except for F8 n = 9).

Figure 2

Response surface plots for the effect of MF/D and S/P ratios on: (A) Q₁ (%), (B) T_50% (days), and (C) rate of injection (mL/min).

Figure 3

Optimization of RV-ISM systems showing the response surface plot for the effect of using MF/D in the ratio 11.71:1 and S/P in the ratio 1.64:1 on the desirability value.

Figure 4

Effect of γ-sterilization on the rheological properties of the optimal RV-ISM system: (A) Before sterilization and (B) After sterilization; TEM of optimal RV-ISM system: (C) Before sterilization; and (D) After sterilization; and (E) In vitro release profiles of the optimal RV-ISM system in PBS (pH 7.4) at 37 °C before and after sterilization. Data points are mean ± SD (n = 3).

Figure 5

In vitro cytotoxicity results using EBTr cells treated for 48 h with different concentrations of the optimal drug-free-ISM system. Data points are mean ± SD (n = 3).

Figure 6

Mean (±SD) plasma RV concentrations following IM and SC injections of the optimal RV-ISM system and RV solution to albino rabbits. Data points are mean ± SD (n = 3).