**5. Conclusions**

A rivastigmine-loaded NLC formulation was successfully optimized using the QbD approach and other tools, namely an Ishikawa diagram, DoE, Pareto chart, and response surface plots. The developed HPLC method was found to be simple, linear, precise, selective, and robust for rivastigmine quantification. Regarding the optimization of the rivastigmine-loaded NLC formulation, the central composite design used to select the best ratios of lipids and surfactants and the Box–Behnken design used to obtain the best instrumental parameters were considered suitable and statistically significant for the CQAs, providing the selection of the most suitable formulations with a 95% confidence level.

The optimized rivastigmine-loaded NLC formulation had a solid lipid/liquid lipid ratio of 7.94:1.94 (%, *w*/*w*) and a surfactant/co-surfactant ratio of 4.5:0.5 (%, *w*/*w*). Regarding the production methods, the most adequate conditions were an emulsification rate of 13,400 rpm with 85% sonication amplitude for the ultrasound technique and an emulsification rate of 14,000 rpm with 18 cycles for the HPH method. The latter was considered the most suitable method to prepare the rivastigmine-loaded NLC formulation with the desirable CQAs, although the ultrasound technique also showed effectiveness.

The results showed that the optimized formulations produced by ultrasound technique and HPH method presented respective particle sizes of 114.0 ± 1.9 nm and 109.0 ± 0.9 nm, PDI values of 0.221 ± 0.003 and 0.196 ± 0.007, ZP values of −30.6 ± 0.3 mV and −30.5 ± 0.3 mV, and EE values of 97.0 ± 0.5% and 97.2 ± 0.3%. In addition, no significant changes in these CQAs were observed after 90 days of storage at different temperatures. In vitro studies showed the achievement of a biphasic release profile, resulting from the occurrence of an initial fast release followed by prolonged release of rivastigmine from the NLC formulations produced by both techniques over 48 h.

The results of our study suggest that the optimized rivastigmine-loaded NLC formulation produced by the HPH method is stable and can be used as an alternative delivery system for the nose-to-brain delivery of rivastigmine. However, this application must be confirmed with more in vitro and in vivo animal experiments before reaching clinical studies. In addition, QbD has proven to be a very useful approach for the optimization of NLC formulations with specific requisites.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/1999-4923/12/7/599/s1, Figure S1. Calibration plot of areas (mean) *versus* rivastigmine concentration (*n* = 3). Figure S2. Chromatogram of the supernatant of placebo-NLC formulation. Figure S3. Chromatogram of standard 1200 µg/mL rivastigmine solution. Figure S4. Pareto chart showing the effects of CMAs on CQAs, *viz*., size (Z-Ave, D50 and D90) (left: A-C), PDI, ZP and EE (right: D-F). Figure S5. Contour plot for CQAs, *viz*., size (Z-Ave, D50 and D90) (left: A-C); and PDI, ZP and EE (right: D-F). Figure S6. Filter paper showing the results of screening of drugs and lipids, where the absence of oil droplets resulting from the solubilisation of drug in the lipid mixture is observed. Table S1: System suitability parameters. Table S2. Results achieved for the intra-day precision and inter-day precision. Table S3: Results obtained for the instrumental precision. Table S4. Drug recovery for method accuracy. Table S5: Detection and quantification limits. Table S6. Results of the method robustness after variation the flow rate. Table S7: Results of the method robustness after variations in the mobile phase. Table S8: Results of encapsulation efficiency (EE) and loading capacity (LC) of rivastigmine-loaded NLC formulations. Table S9: ANOVA models and respective R

squared (R<sup>2</sup> ). Table S10: ANOVA models and respective *R* 2 for instrumental parameters: emulsification speed and number of HPH cycles. Table S11. ANOVA models and respective R<sup>2</sup> for instrumental parameter: sonication amplitude. Table S12: Critical quality attributes (CQAs) values of rivastigmine-loaded NLC formulations before and after the pH and osmolarity adjustment by addition of HCl and glycerin.

**Author Contributions:** Conceptualization, A.C.S. and S.C.; methodology, S.C. and C.P.C.; software, S.C., C.P.C. and J.A.; validation, J.A. and J.A.L.; formal analysis, J.A.L. and A.F.P.; investigation, S.C.; writing—original draft preparation, S.C.; writing—review and editing, A.C.S., B.F., J.M.S.L. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by Fundação para a Ciência e a Tecnologia (FCT), Portugal (SFRH/ 131074/2017), by the Applied Molecular Biosciences Unit-UCIBIO (UID/Multi/04378/2019), FP-ENAS (UID/Multi/04546/2019), and LEPABE (UIDB/00511/2020—through the FCT/MCTES (PIDDAC)). FCT supported J.A.L. under the Scientific Employment Stimulus-Institutional Call—[CEECINST/00049/2018].

**Conflicts of Interest:** The authors declare no conflict of interest. The funders/company had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
