(W0930-11-72) Investigating the Role of Critical Formulation and Process Variables Using Hybrid-Design Approach on the Redispersibility of Spray Dried Co-precipitated Nanoamorphous Micellar Dispersion

Purpose: In the previous work, the anti-solvent precipitation process was optimized to develop a nanoamorphous micellar dispersion of ritonavir (RTV-NAMD) using different machine learning tools such as multiple linear regression (MLR) and verified the computational predictability of the optimized critical parameters with the multivariate analytical (MVA) tool: Partial least square (PLS) model (Figure 1, Left section). In continuation to our previous work, the objective of the current study was to screen and optimize the critical formulation and process variables for converting liquid-state RTV-NAMDs into stable and redispersible spray-dried (rSD) powder form of co-precipitated NAMD (RTV-rSD-NAMD) using polymethacrylate polymer (Eudragit EPO).

Methods: An RTV-NAMD was previously optimized with the aim of enhancing the dissolution of a weakly basic and poorly water-soluble drug RTV (Figure 1, Left section). For the present work, a systematic hybridized-design approach of screening and optimization of variables was performed for stable RTV-rSD-NAMD. Initially, a fractional factorial screening design [FFD, resolution IV, 7 variables (formulation variables: concentration of the drug, wetting agent (Poloxamer 407), and polymethacrylate polymer and process variables: feed and aspiration rate, inlet temperature and spray gas flow), 18 runs, 2 center points, 1/8 fraction] was employed. Later, a Box-Behnken optimization design (BBD, resolution V, 3 variables, 15 runs, 3 center points) was utilized to study the effect of screened significant formulation parameters (drug concentration and polymethacrylate polymer concentration) and process parameter [Feed rate (%)] on particle size (PS), polydispersibility index (PDI), zeta potential (ZP), redispersibility index [(RDI) = Avg. PS upon SD redispersion in DI water/ Initial Avg. PS of RTV-NAMD], percent yield (PY), average d50, and outlet temperature (OT). Poloxamer 407 and prescreened disaccharide(s) were used as wetting agents and matrix former, respectively. The data were modeled with an MLR equation to decode the statistically significant impact of design variables, their two-way interactions, and quadratic effects. The optimized formulation predicted by the developed model was further validated experimentally. The optimized RTV-rSD-NAMD were characterized by the X-ray powder diffraction (XRPD) technique, Fourier transform infrared spectroscopy, and evaluated for short-term (8 weeks) storage stability at 25 °C/60% RH and 40 °C/75% RH. The optimized RTV-rSD-NAMD were evaluated under a two-stage pH shift biorelevant dissolution method and compared against No_EPO_RTV-rSD-NAMD (without polymethacrylate polymer), and Norvir® Oral Powder (marketed formulation) in non-sink dissolution conditions.

Results: The observed values from optimized verification batches for PS, PDI, RDI, and PY were 292.0 nm, 0.267, 1.05, and 65.94% which is in good agreement with the predicted values by each of the MLR models of BBD. Note: ZP was excluded from the model as it reached Zeta_max (due to polymethacrylate)¹ with > +30 mV for all formulations rendering the model insignificant. The d50 (28.4 µ) and OT (43.0 °C) showed < 20% prediction error, which has no influential implications on the critical quality attributes of optimized RTV-rSD-NAMDs. PXRD diffractogram confirmed the amorphicity for RTV-rSD-NAMD at Day 1. RTV-rSD-NAMD displayed the avg. PS of 306.1 ± 5.9 nm after redispersion at Day 1. Finally, RTV-rSD-NAMD maintained its amorphicity (Figure 1, Right section-center picture) and redispersibility index (1.00; PS: 292.9 nm and PDI: 0.289) over a period of 8 weeks at 25 °C/60% RH. However, RTV-rSD-NAMD could not maintain its critical quality attributes of amorphicity and redispersibility at 40 °C/75% RH showing recrystallization (disaccharide) and RDI >10, respectively (Figure 2). Dissolution of RTV-rSD-NAMD has displayed RTV release for a greater extent maintaining the supersaturation for at least 60 min in intestinal fasted media before declining. While the maximum amount of RTV released from RTV-rSD-NAMD, NO_EPO_RTV-rSD-NAMD, and Norvir® oral powder at pH 6.5 was 62.1 ± 16.7%, 15.9 ± 3.7%, and 25.7 ± 1.2 %, respectively (Figure 1, Right section-bottom picture).

Conclusion: The prediction error (%) for PS, PDI, ZP, and PY using DoE-based MLR models was found to be less than 10% which conforms to the maximum computational predictability and compliments the verification of the optimized RTV-rSD-NAMD. Clearly, a systematic hybridized-design approach of screening and optimization of variables provides a better solution to develop a stable RTV-rSD-NAMD with maintaining the critical attributes of amorphicity and nanoredispersibility. The dissolution behavior of RTV-rSD-NAMD may have resulted due to the complex interplay of nanosization, amorphicity, and in-situ micellarization.

References: 1. Barbosa, J. A., Abdelsadig, M. S., Conway, B. R., & Merchant, H. A. (2019). Using zeta potential to study the ionisation behaviour of polymers employed in modified-release dosage forms and estimating their pKa. International journal of pharmaceutics: X, 1, 100024.

Acknowledgements: No conflict of interest was stated in this poster presentation. The study was conducted at Long Island University.

Figure 1: Graphical abstract representing our previous work to keep it anonymous (Left-section) and present work (right-section) depicting a cartoon of RTV-rSD-NAMD (Right section: top picture) and In-vitro dissolution studies (right-section: bottom picture) of RTV-rSD-NAMD, No_EPO_RTV-rSD-NAMD, and Norvir® Oral Powder.

Figure 2: Stability of RTV-rSD-NAMDs at 25 °C/60% RH and 40 °C/75% RH for 8 weeks. (Note: Orange circles are used to indicate the recrystallization of the disaccharide component in the system. Poloxamer 407 recrystallization at Tzero does not affect the critical quality attributes of RTV-rSD-NAMDs).