*4.2. Impact of Silica Nanostructure*

FS exhibited superior stabilization capacity and was able to stabilize a FEN saturation level of 200% Seq. It was evident that saturation levels greater than 400% Seq surpassed silica's stabilization capability and was reflected in DSC thermograms and XRP diffractograms. The relative intensities of XRPD patterns of MPS and FS formulations propose the presence of greater amounts of crystalline FEN in MPS compared to FS formulations, suggesting the nanostructured matrix of silica microparticles influences the physical state of the drug. XRPD is highly sensitive to molecular packing; therefore, any changes in the internal structure of the drug can be reflected in the patterns obtained [39]. This was evident in the XRPD pattern of 200% MPS PG8 formulation, suggesting the presence of FEN in a polymorphic form. However, the identification and confirmation of the presence and stability of FEN in any polymorphic form needs to be further investigated. Therefore, 200% Seq was considered optimal in combination with FS as a solid carrier to maximize drug load while maintaining FEN in a solubilized non-crystalline state. These findings were in agreement with previous studies where

super-SLH containing IBU and AbA displayed an increased amount of crystalline drug with increase in drug loading and optimum drug loading, which was 227% for IBU and 150% for AbA [28,31].

During in vitro dissolution studies (Figure 5), unsaturated (at 80% Seq) MPS and FS formulations had a comparable performance, regardless type of lipid used. In contrast, both the rate and extent of dissolution from super-SLH (200%, 400%, and 600% Seq) were significantly enhanced when using FS compared to corresponding MPS formulations. At higher saturation levels (400% and 600% Seq), it was evident that FS was able to maintain excellent FEN dissolution despite the presence of "crystalline" FEN and less amount of lipid in the formulations. However, the relatively larger amount of crystalline FEN present at 600% FS PG8 and 600% FS C300 resulted in slower release kinetics. In contrast, a reduction in dissolution performance was observed by MPS formulations with an increase in drug load. These findings were consistent with data obtained from FEN aqueous solubilization during in vitro lipolysis, despite the complexity of the data generated due to FEN precipitation.

FEN is a commonly used model PWSD in LBFs, and its tendency to precipitate from LBFs has been previously reported in multiple studies [11,41–43]. In these studies, authors proposed that, as digestion proceeds, formulations lose their solubilization capacity along with dilution of formulation triggered by the transition into intestinal digestion, where FEN precipitates with increasing FA concentration [41]. However, in these studies, FEN precipitation during in vitro lipolysis did not correlate with poorin vivo performance. Thomas et al. observed more extensive in vitro FEN precipitation from supersaturated self-nanoemulsifying drug delivery systems (super-SNNEDS) than from unsaturated SNEDDS [11]. Nevertheless, all SNEDDS formulations displayed enhanced in vivo bioavailability. FEN was shown to precipitate in a crystalline form in vitro but the extent of FEN precipitation in vitro was not known. These studies indicate that in vitro lipolysis models are not predictive of in vitro performance of FEN in a LBFs and should be used carefully to interpret such data [42].

In a previous study, in order to understand the underlying mechanism of FEN absorption and possible precipitation in vivo, orlistat as a lipase inhibitor was co-administered with SNEDDS containing FEN [43]. The study showed that lipase inhibition did not influence FEN bioavailability from SNEDDS, therefore digestion does not impact FEN absorption but can be beneficial when FEN is present in crystalline form in the formulation. However, it was still not known whether FEN precipitation occurs in vivo and to what extent. Recently, Tanaka et al. [44] found that FEN precipitated in the stomach of rats in an amorphous form with 20% microemulsion (ME) and crystalline form with 90% ME. In both cases, rapid redissolution was attained in the duodenum and 90% ME achieved significantly enhanced bioavailability. It is important to note that precipitation was more pronounced in the closed in vitro lipolysis system, which implies the importance of adding an absorption mimicking step [45]. The authors concluded that precipitation from the supersaturated state in the gastrointestinal environment was suppressed by the rapid absorption process for a highly permeable drug, like FEN, creating an absorption sink that acts as a driving force for absorption. In addition, the authors suggested the possibility of enhanced solubilization capacity of GI fluid caused by the formation of colloidal species upon digestion of the lipid. Therefore, formulations that cause higher supersaturation in the GI fluid [46] and the inclusion of medium chain triglycerides might result in enhanced bioavailability, despite their tendency to precipitate in vitro. However, such conclusions cannot be made for this study as further in vitro studies are required to validate the full potential of the fabricated SLH formulations.

In the present study, a higher extent of lipid digestion, as reflected in FA release, correlated with more pronounced FEN precipitation. This was evidenced by the 80% liquid PG8 formulation exerting the greatest FEN solubilization but reduced FA release. Liquid lipid formulations investigated in this study formed large emulsion droplets during in vitro lipolysis, providing less surface area for lipase action. However, when lipid was internalized within a porous silica matrix, a high surface area for lipase-mediated digestion was provided [40], which led to a more pronounced FA release and subsequent FEN precipitation. In the case of SLH prepared by FS, supersaturation resulted in enhanced FEN solubilization. In contrast, SLH prepared with MPS achieved the best performance when unsaturated. This was in agreement with solubilization data and previous studies where AbA-loaded super-SLH (prepared with MPS) compromised its solubilization performance during lipolysis when compared to unsaturated SLH [25]. However, this was not the case for IBU (dissolution studies), which implies that the benefits of super-SLH are drug dependent and require further investigation and optimization. *Pharmaceutics* **2020**, *12*, x FOR PEER REVIEW 18 of 22

The difference in performance achieved by FS and MPS formulations can be attributed to their unique structure and resulting dynamic microenvironment when re-dispersed in biorelevant media (Figure 8). It is hypothesized that SLH microparticles prepared with FS de-aggregate and dissociate to their precursor silica nanoparticles upon dispersion in aqueous media, providing high surface area for lipase mediated digestion, dynamic mixed micelle formation, and subsequent drug release. For FEN, this results in a more profound precipitation in the closed in vitro digestion model, which lacks an absorption step. In contrast, FEN-loaded lipid droplets need to diffuse out of the 6 nm [29] structured nanopores of the MPS particles. This slows the kinetics of lipid/drug release, hindering lipolysis, which in turn leads to less pronounced FEN precipitation. In addition, MPS formulations were aggregated at all FEN saturation levels, suggesting poor lipid encapsulation which further reduced the surface area available for lipase action. The difference in performance achieved by FS and MPS formulations can be attributed to their unique structure and resulting dynamic microenvironment when re-dispersed in biorelevant media (Figure 8). It is hypothesized that SLH microparticles prepared with FS de-aggregate and dissociate to their precursor silica nanoparticles upon dispersion in aqueous media, providing high surface area for lipase mediated digestion, dynamic mixed micelle formation, and subsequent drug release. For FEN, this results in a more profound precipitation in the closed in vitro digestion model, which lacks an absorption step. In contrast, FEN-loaded lipid droplets need to diffuse out of the 6 nm [29] structured nanopores of the MPS particles. This slows the kinetics of lipid/drug release, hindering lipolysis, which in turn leads to less pronounced FEN precipitation. In addition, MPS formulations were aggregated at all FEN saturation levels, suggesting poor lipid encapsulation which further reduced the surface area available for lipase action.

**Figure 8.** A schematic representation of the proposed mechanism of FEN release and precipitation from SLH made with FS and MPS, during in vitro lipolysis. Created with BioRender.com. **Figure 8.** A schematic representation of the proposed mechanism of FEN release and precipitation from SLH made with FS and MPS, during in vitro lipolysis. Created with BioRender.com.

### *4.3. Impact of Lipid Type 4.3. Impact of Lipid Type*

were observed.

FEN precipitation from SLH formulations occurred either during the gastric phase or during the intestinal phase depending on drug load and type of lipid used. Unlike dissolution, this made discrimination between formulations challenging. However, data generated display the benefits provided by super-SLH evident in the enhanced performance of all SLH formulations when FEN precipitation from SLH formulations occurred either during the gastric phase or during the intestinal phase depending on drug load and type of lipid used. Unlike dissolution, this made discrimination between formulations challenging. However, data generated display the benefits provided by super-SLH evident in the enhanced performance of all SLH formulations when compared to crystalline drug and APO-fenofibrate.

compared to crystalline drug and APO-fenofibrate. Lipids utilized in this study differ in their chemical composition and the physical interaction when imbibed into the silica nanopores. PG8 consists of propylene glycol (PG) mono (> 90%)- and di (< 10%)-esters of caprylic acid, while C300 is a medium-chain triglyceride manufactured by the esterification of glycerin and fatty acids, mainly caprylic (≈ 70%) and capric acid [47]. This impacts their physical encapsulation within the silica nanopores, lipolysis kinetics, digestibility, and the arrangement of digestion products [40]. Subsequently, influencing drug release and solubilization. Therefore, at the same lipid dose, C300 can generate greater FA release when digested compared to Lipids utilized in this study differ in their chemical composition and the physical interaction when imbibed into the silica nanopores. PG8 consists of propylene glycol (PG) mono (>90%) and di (<10%)-esters of caprylic acid, while C300 is a medium-chain triglyceride manufactured by the esterification of glycerin and fatty acids, mainly caprylic (≈70%) and capric acid [47]. This impacts their physical encapsulation within the silica nanopores, lipolysis kinetics, digestibility, and the arrangement of digestion products [40]. Subsequently, influencing drug release and solubilization. Therefore, at the same lipid dose, C300 can generate greater FA release when digested compared to

PG8. This was reflected in the FA release-time profiles. Conversely, the greater extent of digestion observed in C300 formulations did not correlate with enhanced FEN solubilization as no significant differences in solubilization capacity between SLH C300 and corresponding SLH PG8 formulations were observed.

Lipid distribution within the particles was different for each lipid. As evidenced in the confocal images (Figure 4), more PG8, compared to C300, was adsorbed on the surface of the silica microparticles than the pores with large droplets in between the particles, suggesting incomplete lipid loading that, in turn, can lead to aggregation, as observed in SEM images (Figure 3) and more evident with supersaturated FEN formulations. The distribution profile of PG8 within the SLH was similar to what was previously reported by Schultz et al. [31]. However, unlike findings in this study, the solubilization capacity of all SLH prepared with PG8 was much greater than those prepared with Capmul MCM, despite its even distribution within the silica matrix. In addition, it is important to highlight the synergistic effect obtained by the slow release kinetics of MPS along with higher amount of digestion products produced from the digestion of C300 in 80% and 200% MPS C300 formulations. As evidenced by the enhanced solubilization of FEN, these were the only SLH formulations to significantly enhance FEN solubilization between 15–60 min during the intestinal phase of in vitro GI lipolysis (Figure 6).

In the present work, we were able to highlight the importance of understanding the nanostructured matrix and composition of SLH microparticles in order to seize their full potential in enhancing the biopharmaceutical performance in delivering PWSDs, such as FEN. Each type of silica provides competitive advantages that can be tailored according to the outcome desired from the formulation. It has become evident that changes in the nanostructure and surface chemistry of SLH particles, even if small, can have significant impact on lipase activity which, in turn, influences drug release kinetics and absorption. Therefore, Future work will be directed toward understanding the influence of internal silica particles nanostructure, microporosity, and surface area on drug solubilization.
