*3.5. Design of Experiment (DoE) for the Optimization of Rivastigmine-Loaded NLC Formulation*

A DoE was used to evaluate the effects of critical parameters related to the CMAs (i.e., formulation variables) and CPPs (i.e., instrumental parameters) on CQAs, namely for particle size, PDI, ZP, and EE.

Figure 1 shows the Ishikawa diagram used as a visualization tool for the two parts of the optimization process.

3.5.1. Part 1: Optimization of Formulation Variables by Central Composite Design (CCD)

In the first part of the optimization of the rivastigmine-loaded NLC formulation, we tested different solid lipid and liquid lipid (SL/LL) ratios, which were selected using the results of lipid-drug solubility tests and different ratios of surfactants.

*Pharmaceutics* **2020**, *12*, 599 7 of 27

**Figure 1.** Ishikawa diagram showing the effects of critical material attributes (CMAs) and critical process parameters (CPPs) on the critical quality attributes (CQAs) of rivastigmine-loaded NLC formulation. **Figure 1.** Ishikawa diagram showing the effects of critical material attributes (CMAs) and critical process parameters (CPPs) on the critical quality attributes (CQAs) of rivastigmine-loaded NLC formulation.

3.5.1. Part 1: Optimization of Formulation Variables by Central Composite Design (CCD) In the first part of the optimization of the rivastigmine-loaded NLC formulation, we tested different solid lipid and liquid lipid (SL/LL) ratios, which were selected using the results of lipiddrug solubility tests and different ratios of surfactants. The influence of the different ratios of SL/LL and surfactants on CQAs or dependent responses, namely particle size, PDI, ZP, and EE, were studied using a CCD with α rotatability of 1.4142, two-factors, and 3 levels. Table 1 shows the DoE used to test the rivastigmine-loaded NLC formulation variables.

namely particle size, PDI, ZP, and EE, were studied using a CCD with α rotatability of 1.4142, twofactors, and 3 levels. Table 1 shows the DoE used to test the rivastigmine-loaded NLC formulation variables. **Table 1.** Design of experiment (DoE) using six central composite design (CCD) for rivastigmine-loaded nanostructured lipid carriers (NLC) formulations with different critical material attributes (CMAs).

The influence of the different ratios of SL/LL and surfactants on CQAs or dependent responses,


dependent responses were studied at low (−1), medium (0), and high (+1) levels. Six DoE with 10 experimental runs were generated by the statistic software with two factors or independent variables corresponding to the CMAs, namely solid lipids (Precirol® ATO 5, SL), vitamin E (LL), and surfactants (Tween® 80 and Phospholipon® 90G, Tw/Ph). Their effects on CQAs or dependent responses were studied at low (−1), medium (0), and high (+1) levels.

E (LL), and surfactants (Tween® 80 and Phospholipon® 90G, Tw/Ph). Their effects on CQAs or

For the experiments, rivastigmine-loaded NLC formulations were produced by employing the ultrasound technique previous described by Silva et al. [72], involving high-speed homogenization at 13400 rpm and a sonication amplitude of 75%. The formulation with the most suitable values for CQAs, namely having a lower particle size, PDI of around 0.2–0.3, a ZP value close to 30mv, and an EE value > 90%, was selected for the next part of the optimization, which is related to the instrumental parameters [31,33,34,42]. It is important to note that ZP values close to |30| mv are desired to ensure more stable NLC formulations, since the electrostatic repulsion between the nanoparticles prevents aggregation. However, negative ZP values reduce the residence time of the formulation in the nasal mucosa, since interactions between NLC and mucin, a negatively charged glycoprotein of the nasal cavity, may not occur [73,74]. To overcome this limitation, mucoadhesive polymers were added to the optimized rivastigmine-loaded NLC formulations to form in situ gels without negative charge. This strategy was used to develop NLC formulations for nose-to-brain delivery. For example, Rajput and Butani developed a NLC-based in situ gel for intranasal administration of resveratrol [75], while Abouhussein et al. developed a NLC-based in situ gel to improve the brain target of rivastigmine after intranasal administration [9].
