Preparation, and ex vivo and in vivo Characterization of Favipiravir-Loaded Aspasomes and Niosomes for Nose-to-Brain Administration

Introduction

Neurotropic viruses including RNA viruses, such as poliovirus, Zika virus, influenza A and B, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), mumps virus, and measles virus, are able to access the brain and infect the central nervous system (CNS), resulting in meningitis, myelitis, encephalitis, or meningoencephalitis.^1–3 Neurotropic viruses can access CNS by crossing the blood–brain barrier (BBB) or blood-cerebrospinal fluid barrier (BCSFB), nerve terminals in the nasal olfactory epithelium, or the pseudounipolar sensory neurons of the peripheral nervous system (PNS).^3–5 This viral infection of the CNS leads to alteration or degeneration of neuronal cell function, which causes several neurodegenerative diseases, such as Alzheimer’s disease, Parkinson’s disease, and amyotrophic lateral sclerosis.^5–7 Therefore, it has become urgent to search for potential drugs for the treatment of viral CNS infections. Considering these facts, intranasal delivery can be a proper choice to deliver the drug directly to the brain via the systemic pathway, olfactory and trigeminal nerve pathways.^8–10 Several possible candidates exist for the intranasal delivery of virucidal drugs and agents, such as favipiravir (FAV).^11,12

FAV (FAV, 6-fluoro-3-hydroxypyrazine-2-carboxamide, Figure 1), as an antiviral prodrug, was approved in Japan in 2014 for the treatment of new-onset or recurrent pandemic influenza as well as for experimental drug for Ebola virus infections. FAV is first converted by hypoxanthine-guanine phosphoribosyl transferase (HGPRT) to ribose 5′-monophosphate (T-705-RMP) and then metabolized to the triphosphate form (T-705RTP) by cellular kinases.^13,14 The compound selectively inhibits viral RNA polymerase in vivo by its triphosphorylated metabolite, translating to broad-spectrum inhibition of RNA viruses.^15,16

Figure 1 Chemical structure of FAV and its metabolic activation.14

Figure 2 EE% and DL% results of FAV-ASP8, FAV-ASP10 and FAV-NIO9. Results are expressed as means ± SD (n = 3), ***p-value < 0.001, ****p-value < 0.0001.

Figure 3 BBB-PAMPA results for FAV-ASP8, FAV-ASP10 and FAV-NIO9 in comparison with pure FAV. (A): permeability results, (B): flux results. Results are expressed as means ± SD (n = 6), **p-value < 0.01, and ***p-value < 0.001, ****p-value <0.0001.

Figure 4 Rapid equilibrium dialysis (RED) of the optimal formulations in comparison with pure FAV. Results are expressed as means ± SD (n = 6).

Figure 5 Cumulative in vitro release profile of the optimal formulations compared to pure FAV. Results are expressed as means ± SD (n = 3).

Figure 6 Ex vivo permeability study of FAV-ASP8 and FAV-NIO9 in comparison with pure FAV. Results are expressed as means ± SD (n = 3).

Figure 7 Plasma concentration–time profile after nasal administration of FAV-ASP8, FAV-NIO9 and pure FAV in rats. Results are expressed as means ± SD (n = 4), (ns) means non-significant.

Figure 8 The concentration of FAV in cerebrospinal fluid (CSF) after 1 hour of nasal administration. Results are expressed as means ± SD (n = 4), (ns) means non-significant, *p-value < 0.05, ***p-value < 0.001.

Figure 9 IVIVC graphs for the comparison of AUC 0-t values between the in vitro release and the in vivo PK data. Where (A): IVIVC for pure FAV, (B): IVIVC for FAV-ASP8, and (C): IVIVC for FAV-NIO9.

Figure 10 IVIVC graphs for the comparison of AUC 0-t values between the in vitro permeation and the in vivo PK data. (A): IVIVC for pure FAV, (B): IVIVC for FAV-ASP8, and (C): IVIVC for FAV-NIO9.

FAV is available in an oral solid dosage form (Avigan® 200 mg),¹⁷ with a recommended dosage of 1600 mg twice daily on day 1, followed by 600 mg twice daily on days 2 to 5 for the treatment of influenza infection,¹⁸ which leads to patient incompliance, and results in toxicity and limited oral bioavailability (due to its low water solubility, FAV is a Biopharmaceutics Classification System (BCS) class II drug).¹⁹ FAV has a short half-life, leading to rapid renal clearance in its hydroxylated form (inactive metabolite T-705M1) after being metabolized by aldehyde oxidase/xanthine oxidase enzymes (present in the liver, lungs, small and large intestine, kidney, prostate, and adrenal glands).²⁰ Thus, a novel formulation and alternative route of administration are critical to improve bioavailability and therapeutic efficacy while minimizing the adverse effects of FAV (that include mild to moderate diarrhoea, increase of blood uric acid and transaminase level, and decrease in the neutrophil counts).^21,22

Intranasal delivery of FAV means further advantages, including improved systemic bioavailability due to increased drug solubility, absorption and nasal permeability through a highly vascularized and less acidic environment, avoiding hepatic first-pass metabolism in comparison to the gastrointestinal tract.^23,24 FAV shows low brain penetrance because it has a low log P value (0.72), the critical factor for CNS penetration, and low lipophilicity, which decreases brain uptake.²⁵ Therefore, intranasal delivery can be utilized to deliver FAV directly to the brain, where it could be activated by the HGPRT enzyme within the infected cells by neurotropic viruses, as HGPRT is also be found in the CNS. Moreover, the administration of lower doses may be therapeutically effective with fewer systemic side effects.

Lipid nanoparticles, including liposomes and niosomes, are considered a suitable antiviral drug carrier to enhance the BBB penetrating, protect the drug against enzymatic degradation, and improve its bioavailability, hence, prevent viral spread.^26–28 Niosomes (NIOs), a newer generation of liposomes, rise from the self-assembly of non-ionic surfactants of the alkyl or dialkyl polyglycerol ether class and cholesterol (CH) with subsequent hydration in aqueous media.^29,30 Recently, amphiphilic materials, such as ascorbyl palmitate (AP), have been used to prepare bilayer vesicles called aspasomes (ascorbyl palmitate-based nanocarriers; ASPs), a newer generation of NIOs, to enhance the stability and biological activity of NIOs.³¹ ASPs contain AP in combination with CH and an anionic or nonionic surfactant in various molar ratios. AP is an ester form of ascorbic acid, which is amphiphilic by nature and more stable in liquid form than ascorbic acid.^31–35

These colloidal vesicles have several advantages, like the development of an effective drug delivery system to achieve a maximum effective concentration and the formation of vesicles that are able to entrap both hydrophilic and hydrophobic drug molecules with higher stability compared to liposomes. Moreover, modification of the nanoparticulate composition or surface can adjust the affinity for the target site and/or the drug release rate.

According to our knowledge, several FAV-nanocarriers for intranasal administration have been developed such as favipiravir-pyridinecarboxamide cocrystal nasal powder formulation,³⁶ crystalline favipiravir Sodium Salt liquid formulation,³⁷ favipiravir loaded PLGA nanoparticles,³⁸ and favipiravir loaded mucoadhesive chitosan–alginate nanoparticles.³⁹ Moreover, ASPs have been used for transdermal delivery, such as Acyclovir loaded aspasomal gel,⁴⁰ Quercetin loaded aspasomal gel,⁴¹ Idebenone/Naproxen co-loaded aspasomes,⁴² and based on the literature, there are no available reported ASP formulations for intranasal administration. Therefore, FAV-loaded aspasome formulations (FAV-ASPs) and FAV-loaded niosome formulations (FAV-NIOs) seem to be a novel approach.

In this study, our aim was to develop and compare novel FAV-ASPs and FAV-NIOs intended for intranasal administration in order to improve the poor solubility of FAV, as a potential treatment for viral infections of CNS. A further aim was the characterization of nasal applicability of optimized FAV-ASP and FAV-NIO formulations.

Materials and Methods

Chemicals and Solvents

FAV was provided by Egis Pharmaceuticals Plc. (Budapest, Hungary) with a purity of 99.6% w/w (according to supplier certificate of analysis). Methanol 99.99% v/v (HPLC grade), ortho-phosphoric acid 85% v/v (HPLC grade) and anhydrous disodium hydrogen phosphate were purchased from Molar Chemicals Kft. (Budapest, Hungary). Acetonitrile 99.8% v/v (HPLC grade) was purchased from PromoChem (Wesel, Germany). Ascorbyl acid-6-palmitate (AP), sorbitan monostearate (Span® 60), chloroform, ethanol 96% v/v, dimethyl sulfoxide (DMSO) and dodecane were purchased from Merck KGaA (Darmstadt, Germany). Porcine brain polar lipid extract, dicetyl phosphate (DCP) and cholesterol (CH) were purchased from Sigma Aldrich Co. Ltd. (Budapest, Hungary). For redispersion of formulations, pH 7.4 Dulbecco’s phosphate-buffered saline (DPBS) acquired from Capricorn Scientific GmbH (Ebsdorfergrund, Germany) was applied. As nasal dissolution medium, Simulated Nasal Electrolyte Solution (SNES) was freshly prepared, containing 8.77 g sodium chloride (NaCl), 2.98 g potassium chloride (KCl), and 0.59 g anhydrous calcium chloride (CaCl₂) in 1000 mL of deionized water at pH 5.6. These chemicals were acquired from Sigma-Aldrich Co. Ltd. (Budapest, Hungary). In all experiments, purified water was filtered using the Millipore Milli-Q® (Merck Ltd., Budapest, Hungary) Gradient Water Purification System.

Quantitative Analysis by RP-HPLC-DAD

FAV concentration in the experiments was analyzed using an Agilent 1260 hPLC (Agilent Technologies, San Diego, CA, USA) and Zorbax® SB-CN C18 column (5 μm, 250 mm × 4.6 mm, 100 Å). As mobile phase acetonitrile–disodium hydrogen phosphate anhydrous buffer (pH 3.1, 20 mm) in a ratio of 10:90 (v/v) was used, and FAV was eluted at an isocratic flow rate of 1.0 mL/min up to 10 min at 30°C. The injection volume was 10 µL. Chromatograms were detected at 323 nm using a UV-VIS diode-array detector.

Table 1 Composition of FAV-Loaded Aspasome Formulations (FAV-ASP)

Table 2 Composition of FAV-Loaded Niosome Formulations (FAV-NIO)

Table 3 Results of System Suitability (n = 6)

Preparation of FAV-ASP Formulations

A modified film hydration method was used for the preparation of FAV-ASPs^31,34 (Table 1). For anionic surfactant-based FAV-ASPs, various ratios of lipid mixture (AP:CH:DCP) (w/w) were dissolved in 10 mL of chloroform, whereas for nonionic surfactant-based FAV-ASPs, different ratio of AP/Span® 60 with a constant amount of CH (50 mg) were dissolved in 10 mL of chloroform. A fixed amount of FAV (30 mg) was dissolved in 5 mL of methanol. Then, the FAV solution was mixed with each lipid solution in a round bottom flask, and the organic solvent was evaporated at 60 °C and 633 mbar pressure using a Büchi R-210 rotary vacuum evaporator (Flawil, Switzerland), and the rotation was set at 100 rpm for 1 h, until the appearance of a thin film on the wall of the flask. Thereafter, the thin lipid film was hydrated with 15 mL of pH 7.4 phosphate-buffered saline (PBS) for 1 h.

Preparation of FAV-NIO Formulations

FAV-NIOs were formulated by a modified ethanol injection method.⁴³ Various amounts of Span® 60 and CH (as shown in Table 2) were dissolved in 2 mL ethanol using ultrasonication (for 10 min at 30°C). A fixed amount of FAV (30 mg) was dissolved in the ethanolic solution, followed by sonication until the formation of a clear solution. Then, it was injected rapidly in 15 mL of distilled water (previously heated to 60°C) under constant stirring, followed by probe sonication (0.5 cycles for 2 min at ultrasonic power input 80%). Then, it was stirred for 1 h at room temperature, followed by a complete evaporation of ethanol at 60°C and 633 mbar pressure using a rotary vacuum evaporator, and the rotation was set at 100 rpm for 1 h. To reach the final volume (15 mL) of niosomal dispersion, distilled water was added.

Lyophilization

For freeze-drying a Scanvac, CoolSafe 100–9 Pro type apparatus (LaboGeneApS, Lynge, Denmark) was applied. Formulations were divided in 1.5 mL portions, and 5% w/v of mannitol was added as a cryoprotectant. Freeze-drying was performed at −40°C for 12 h at reduced pressure (0.013 mbar) controlled by Scanlaf CTS16a02 software. After that, a secondary drying at 25°C was applied for 3 hours. Freeze-dried samples were stored in the refrigerator at 4°C until further analysis. The samples were redispersed in adequate medium before each test.

Vesicle Size Analysis, Polydispersity Index and Zeta Potential Determination

The mean hydrodynamic diameter of vesicles (Z-average), polydispersity index (PDI) and zeta potential (ZP) for FAV-ASP and FAV-NIO formulations were investigated using a Malvern NanoZS equipment (Malvern Instrument). Samples were redispersed in distilled water and suitably diluted (1:10) before assessments. All measurements were carried out in triplicate, and results are presented as means ± SD.

Determination of Drug Content, Drug Encapsulation Efficiency (EE%) and Drug Loading (DL%)

For drug content determination 1–1 mL of formulations were diluted with 4 mL of methanol using ultrasonication bath for 10 min. Then, the suspensions were filtered using 0.45 µm syringe filters and analyzed with HPLC to determine FAV concentration. Encapsulation efficiency (EE%), ie the percentage amount of FAV that has been loaded in the vesicles, was determined by using the dialysis technique in order to select the proper formulations with the highest EE% for further studies.⁴⁴ The formulations were redispersed in distilled water, and then 3 mL of each sample was transferred into a dialysis bag (12 kDa MWCO, Sigma-Aldrich, St. Louis, USA) immersed in 100 mL of purified water and constantly stirred for 30 min. The concentration of free FAV that diffused from the dialysis bag to the purified water was determined by using HPLC, and EE% and DL% were calculated as follows:⁴⁵

(1)

Residual Solvent Determination

To determine the residual organic solvents (ethanol, methanol and chloroform) in the freeze-dried formulations, gas chromatography-mass spectroscopy (GC-MS) measurements were performed using a Shimadzu GCMS-QP2010 SE (Shimadzu Europa GmbH, Duisburg, Germany) with a ZBWax-Plus column (length: 30 m, diameter: 0.25 mm) and helium carrier gas.

In vitro Permeability Measurements

A blood–brain barrier-specific parallel artificial membrane permeability assay (BBB-PAMPA) was used to investigate FAV permeability (cm/s) of initial API solution and in the selected formulations (according to the EE% results) in order to find the optimal one.⁴⁶ 10 mm FAV solution was prepared in dimethyl sulfoxide (DMSO), which was further diluted with DPBS solution (pH 7.4) to obtain reference solution with 100 µM concentration. The donor plate (Multiscreen™-IP, with pore size 0.45 μm; Millipore, Merck Ltd., Budapest, Hungary) was preliminary coated with 5 μL of lipid solution (24 mg porcine brain polar lipid extract dissolved in 840 μL hexane and 360 μL dodecane). Then, the donor plate was inserted to the acceptor plate (Multiscreen Acceptor Plate, Millipore, Merck Ltd., Budapest, Hungary), which contained 300 μL of DPBS solution (pH 7.4). 150–150 μL of reference FAV solution, as well as of the redispersed freeze-dried formulations with DPBS (with 2 mg/mL FAV nominal concentration) were transferred on the lipid membrane of the donor plate. The plates were incubated at 37°C for 4 h (Heidolph Titramax 1000, Heidolph Instruments, Schwabach, Germany). After that the PAMPA plates were separated and the FAV concentrations both in the donor and acceptor chambers were determined using HPLC. The effective permeability and membrane retention of drugs were calculated using the following equation (3):

(3)

where P_e means the effective permeability coefficient (cm/s), V_A is the volume of the acceptor well (0.3 cm³), A indicates the surface area of one well (0.24 cm²), t is the incubation time (s), τ_SS is the time to reach the steady state (s), C_A(t) is the concentration of the compound in the acceptor phase at time point t (mol/cm³), and S is the solubility of FAV in the donor phase. The flux of the samples was calculated as follows (4):

(4)

Six parallel measurements were performed, and data are presented as means ± SD.

Rapid Equilibrium Dialysis Measurement (RED)

RED method (Thermo Scientific™, Waltham, MA, USA) was applied for the determination of the time-dependent drug release profiles of FAV and the selected optimal formulations (according to PAMPA results) after nasal absorption. For the measurement, the reference was prepared by suspending 2 mg of FAV in 1 mL DPBS solution (pH 7.4) using an Eppendorf MixMate (Thermo Scientific™, Waltham, MA, USA) vortex mixer for 30 s. RED inserts (8K MWCO) were fitted into the reusable Teflon base plate, and then 150–150 μL of reference FAV solution, as well as of the lyophilized formulations redispersed in DPBS (with 2 mg/mL FAV nominal concentration) were pipetted into the donor chambers. Then, 300 μL of DPBS was pipetted to the acceptor chambers, and the device was covered with a sealing tape to avoid evaporation and incubated at 37°C for 4 h. Aliquots from the acceptor chambers were withdrawn at predetermined time points and replaced with fresh DPBS.⁴⁶ FAV concentrations were determined using HPLC. Six parallel measurements were conducted, and data were presented as means ± SD.

In vitro Drug Release Under Nasal Conditions

In vitro drug release study under nasal conditions is an important test during the development of intranasal formulations to determine the absorption through the nasal cavity. Using the modified paddle method (Hanson SR8 Plus, Teledyne Hanson Research, Chatsworth, CA, USA) drug release profile for reference FAV suspension, and the optimal freeze-dried formulations redispersed in SNES (with 2 mg/mL FAV nominal concentration⁴⁷) were determined at 32°C at 50 rpm paddle rotation speed in 100 mL SNES medium for 60 min. 1 mL of reference suspension and formulations were loaded in pre-treated dialysis bags (Spectra/Por® Dialysis Membrane with 12–14 kDa MWCO, Spectrum Laboratories Inc., Rancho Dominguez, CA, USA) sealed at both ends. Quantification of aliquots was performed by HPLC.

Evaluation of in vitro Drug Release Data

Drug release profiles were compared with model-independent approaches based on calculating the area under the curve (AUC), dissolution efficiency (DE), and mean dissolution time (MDT). In addition, the model-dependent kinetic analysis (zero order, first order, Higuchi, and Korsmeyer-Peppas kinetic profiles) was fit to evaluated. Microsoft Excel DDsolver® add-in software was used for the mathematical evaluation of the release kinetics, and the fitting of each model by comparing the correlation coefficient (R²), rate constant (K), Akaike Information Criterion (AIC), and Model Selection Criterion (MSC).^48–51

Storage Stability

The optimal formulations were stored at 4°C and the storage stability study was conducted for 4 weeks and analyzed in terms of Z-average, PDI, ZP to evaluate the physical stability and in terms of drug content to evaluate the chemical stability every week (as described previously).

Droplet Size Distribution Measurement

Laser diffraction method was used to evaluate the droplet size distribution of the optimal formulations.⁵² The evaluation was performed using a Malvern Spraytec® system (Malvern Instruments Ltd., Malvern, UK), equipped with a 300 mm lens capable of analysing droplet sizes in a range of 0.1–900 μm (Dv50: 0.5–600 μm). The tip of the nasal spray device was aligned and positioned at 45° from the horizontal plane. Measurements were conducted at room temperature. The formulations were redispersed in distilled water and placed in a nasal spray container. Each nasal spray was manually actuated three times and discharged into waste, allowing the device to function optimally. The data were analyzed using the Spraytec® software v4.00 (Malvern Panalytical Ltd., Malvern, UK), with volume diameter and described as 10% (Dv10), 50% (Dv50), and 90% (Dv90) of the cumulative volume distribution. The results were presented as means ± SD.

Ex vivo Nasal Permeability Study on Human Nasal Mucosa

The ex vivo transmucosal permeability of the optimized formulation (based on BBB-PAMPA, nasal dissolution, and RED results) and reference FAV suspension (with 2 mg/mL nominal FAV concentration) was studied in a modified Side-Bi-Side® type horizontal diffusion apparatus under artificial nasal conditions.⁵³ Human nasal mucosa were collected during routine nasal and sinus surgeries (septoplasty, Functional Endoscopic Sinus Surgery (FESS)) under general or local anesthesia. The surgical field was infiltrated with locally administered 1% lidocaine-adrenalin injection, and the mucosa was excised with a raspatorium or Cottle elevator. Excised nasal mucosa was stored in physiological saline until further investigation. All investigations were conducted freshly within 30 min after the removal of the tissue. The experiments have been performed in line with the principles of the Declaration of Helsinki under approval of University of Szeged’s institutional ethics committee (ETT-TUKEB: IV/3880-1/2021/EKU). The participants were briefed on the study procedures, and written informed consent was obtained from all subjects prior to conducting the procedure. Nasal mucosa was cut with a surgical scalpel into uniform segments with a diameter of 6 mm and inserted between donor and acceptor phases to provide an appropriate surface for permeability study.⁵⁴ To the donor Phase 8 mL of SNES was added, whereas to the acceptor Phase 9 mL of DPBS solution (pH 7.4) was pipetted. The temperature of both chambers was thermostated at 32 ± 0.5°C using a heating circulator (ThermoHaake C 10-P5, Sigma–Aldrich Co. Ltd., Budapest, Hungary). For the measurement, both FAV reference suspension and selected freeze-dried formulation were redispersed in 1 mL SNES (with 2 mg/mL FAV nominal concentration) and added to the donor compartment. Both compartments were continuously stirred at 300 rpm using magnetic stirrers. Aliquots were withdrawn (100 µL) from the acceptor phase at 5, 10, 15, 30, and 60 min and replaced with fresh DPBS. FAV concentration was determined using HPLC. The steady-state flux (J_ss), permeability coefficient (K_p), and enhancement ratio (ER) were calculated as follows:⁵⁵

(5)

where Jss is the steady-state flux (µg/cm²/h), mt is the quantity of FAV permeated through the nasal mucosa, A is the permeability surface of nasal mucosa (0.785 cm²), t is the duration of the investigation (h), K_p is the permeability coefficient (cm/h) and C_d is the drug concentration in the donor phase (µg/cm³).

The theoretical steady-state plasma concentration (C_ss) of the drug can be used for estimation of the concentration of drug that could be reached in the blood after nasal administration, was calculated as follows:⁵⁶

(8)

where J_ss is the steady-state flux, A is the surface area of nasal mucosa used for the permeation study (cm²), and Cl_p is the plasmatic clearance (human Cl_p value for FAV is 5.11 L/h).⁵⁷

In vivo Study

Animals and Samples Collection

The animals were kept at room temperature (approximately 23°C) and light was adjusted to alternate darkness and light for 12 hours. The rats were fed with normal rodent chow and tap water ad libitum. All animals received humane care, in compliance with the “Principles of Laboratory Animal Care” according to the National Society for Medical Research and the Guide for the Care and Use of Laboratory Animals, formulated by the National Academy of Sciences, and published by the National Institute of Health (NIH Publication No. 86–23, revised 1985). Healthy male Sprague-Dawley rats (339 ± 39 g, mean ± SD) were anaesthetized with an i.p. injection of ketamine (50 mg/kg) and xylazine (10 mg/kg). The femoral vein was cannulated for plasma collection.⁵⁸ A total of 50 µL FAV (2 µg/µL) or the corresponding formulation of FAV was administered nasally. Animals were divided into three groups (FAV, FAV-ASP8, and FAV-NIO9), and 0.5 mL of blood samples were taken at 0, 5, 10, 15, 30 and 60 min post-FAV administration, then cerebrospinal fluid samples were collected at the end of the experiments. After 45 min, the clot was removed by centrifuging at 1000 × g for 10 min. The resulting supernatant (non-hemolytic serum) was extracted with 2 volumes of acetonitrile. Accordingly, acetonitrile is suitable for precipitating and removing high-abundance proteins from the serum and eliminate the intra-molecular protein interactions. The precipitate was removed by centrifuging at 1000 × g for 10 min. The animal study protocol was approved by the Institutional Review Board (or Ethics Committee) of Institutional Animal Care and Use Committee of the University of Debrecen, Debrecen, Hungary. (Approval number: 2/2021/DEMÁB)

Sample Preparation

Real or spiked (standard) plasma or cerebrospinal fluid samples were precipitated with acetonitrile (1:2) and centrifuged at 1000 × g for 10 min. The supernatants were collected, and the samples were dried under a gentle stream of N₂ at 60°C using a Turbovap LV concentrator. The samples were re-dissolved in 50 µL pyridine, and then 50 µL N,O-bis(trimethylsilyl)tri- fluoroacetamide (BSTFA) with 1% of trimethylchlorosilane (TMCS) was added into the tubes. The solutions were transferred into 1.5 mL vials sealed, and derivatization was carried out at 80°C for 30 min at 500 rpm using an Eppendorf Thermomixer C device.⁵⁹ Finally, 1 µL of the reaction mixture was injected into GC-MS equipment.

Gas Chromatograph-Mass Spectrometric (GC-MS) Analysis

Derivatized FAV was measured by GC-MS using a Shimadzu GCMS-QP2010 (Shimadzu, Kyoto, Japan) equipment. The GC was equipped with an SLB-5 MS capillary column (Supelco, Bellefonte, PA, USA) (30 m × 0.25 mm i.d.; 0.25 µm film thickness). Operating conditions were as follows: carrier gas: He, flow rate 32 cm/s; column temperature program: 1 min at 60°C, 60–250°C at 40°C/min and finally 7 min at 250°C. The temperature of the injection port and the interface were 250°C and 270°C, respectively. 1 µL of the samples was injected into the GC-MS using an AOC-20i autosampler; the split ratio was 1:10. The MS was equipped with an electron ionization (EI) ion source, and the operating conditions were as follows: ionization energy 70 eV, ion source temperature 200°C, solvent cut time 4 min. The measurements were carried out in selected ion monitoring (SIM) mode. The registered ions were 301 m/z, 286 m/z and 270 m/z, which were selected based on the EI spectrum measured in SCAN mode.

Evaluation of Pharmacokinetic (PK) Parameters

The PK parameters of pure FAV, FAV-ASP8, and FAV-NIO9 were determined in the plasma and the CSF following a single intranasal dose to evaluate the brain targeting. PK Solver 2.0 software was used to determine pharmacokinetic parameters by non-compartmental analysis.⁶⁰ The maximum concentrations (C_max) and times taken to reach these concentrations (T_max) were determined from the mean concentration-time profiles. The elimination rate constant (k_e) was determined by plotting log-linear concentration versus time. The elimination half-life (t_1/2) was calculated using equation (9):⁶¹

(9)

The area under the plasma concentration versus time curve (AUC_0-t) was calculated using the linear trapezoidal method from time 0 to time t (the last time point to withdraw blood samples), then the AUC from 0 to infinity, AUC_0–∞ (µg hour/mL) was calculated using the following equation:^61,62

(10)

where C_t (µg/mL) is the last measured concentration at time t.

The clearance (C_L), the apparent volume of distribution (V_d), the mean residence time (MRT), and the relative bioavailability (F%) were calculated using the following equations:^46,63–66

(11)

where AUMC is the first moment of the concentration-time integral.

Evaluation of in vitro-in vivo Correlations (IVIVC)

A point-to-point IVIVC for pure FAV, FAV-ASP8, and FAV-NIO9 was mathematically examined based on the in vitro data (release and permeation) and the in vivo pharmacokinetics data (based on FDA guidance^67–69). The relationship between in vitro values of AUC_0-t (µg×min/mL) and in vivo values of AUC_0-t (µg×min/mL) were evaluated using linear regression. R² values have been calculated for each graph.

Statistical Analysis

All experimental results are expressed as the mean ± SD. For statistical analysis, GraphPad Prism version 10.12 software (GraphPad Software, San Diego, CA) was used. A one-way analysis of variance (ANOVA) and Tukey’s post-hoc test were performed to compare the groups. The significance level was set at p-value < 0.05, where (ns) means non-significant, *p-value < 0.05, **p-value < 0.01, and ***p-value < 0.001, ****p-value < 0.0001.

Results and Discussion

Quantitative Analysis by RP-HPLC-DAD

The optimized method showed a good retention time (4.95 ± 0.01 min), symmetrical peak shape (0.87 ± 0.02) and the required number of theoretical plates (12722.12 ± 89.21) as per United States Pharmacopoeia (USP) and International Council for Harmonization of Technical Requirements for Pharmaceuticals for Human Use (ICH) Q2 (R1) guidelines.^70,71 The correlation coefficient (R²) was 0.9996, which suggests an excellent correlation and good linearity for the method. System suitability is related to the reproducibility and capability of the method for routine analysis. The optimized method showed good system suitability (Table 3).

Preparation of FAV-ASP and FAV-NIO Formulations

Anionic surfactant-based FAV-ASP consisted of AP, CH and DCP. AP has a hydrophilic–lipophilic balance (HLB) value of about 8.4, which is suitable for vesicle formation.⁷² However, these AP vesicles are not stable. In contrast, AP may form stable ASPs when combined with CH and the negatively charged lipid DCP.^42,73,74 CH is necessary to improve physical stability as it provides rigidity to the bilayer membrane. DCP is used as a negative charge inducer to create vesicles with stronger interaction with the mucosal surface, achieving improved delivery. It also forms electrostatic repulsion between the vesicles, inhibiting aggregations. In addition, the double hydrocarbon chains in DCP results in tighter packing of the bilayer membrane, increasing the encapsulation efficiency in ASP formulations.^75,76

Nonionic surfactant-based FAV-ASPs were formed by combining AP, CH and nonionic surfactants (Span® 60). All Span types have the same head group but a different alkyl chain, and increasing the alkyl chain length leads to a higher encapsulation efficiency. In this study, Span® 60 (HLB = 4.7) was used as a vesicle stabilizer due to its long chain length (C18) and high transition temperature (53−55°C).³⁴ The temperature of the rotary evaporator was 60°C, above the gel-to-liquid phase transition temperature of Span® 60.

The NIO vesicles were formed by the combination of CH and Span® 60. The impact of CH and Span® 60 on the formation of the vesicles was described previously. According to the literature, Span® 60 can form vesicles-NIOs only in the presence of a suitable amount of cholesterol.^77,78

Table 4 The Critical Parameters for the Selection of the Optimal Formulations and the Justification

Table 5 Z-Average, PDI and ZP of Selected Formulations. Results are Expressed as Means ± SD (n = 3)

Table 6 The Summarized Results of the Optimal Formulations

Table 7 Comparison of Release Profiles Using Model-Independent Approaches. Results are Expressed as Means ± SD (n = 3)

Table 8 Droplet Size Distribution of the Optimal Formulations. Results are Presented as Means ± SD (n = 3)

Table 9 The Biopharmaceutical Parameters of the Formulations are Compared to the Initial FAV. Results are Expressed as Means ± SD (n = 3)

Table 10 Pharmacokinetic (PK) Parameters of Pure FAV, FAV-ASP8, and FAV-NIO9 Obtained After Non-Compartmental Analysis Using PK Solver Software. Results are Expressed as Means ± SD (n = 4)

The optimal formulations were selected based on the critical parameters of the drug carrier for the nose-to-brain delivery system, as summarized in Table 4. Firstly, we selected the formulations within the acceptable range of Z-average, PDI and ZP. Then, we selected those with the highest EE% and DL%. After that, depending on PAMPA results, we selected those with the highest permeability and flux values to evaluate the release rate, permeation through the human nasal mucosa and in vivo permeability. Finally, we selected the promising formulation based on the in vivo pharmacokinetics results.

Vesicle Size (Z-Average), Polydispersity Index (PDI) and Zeta Potential (ZP) Analysis

Table 5 shows the results of Z-average, PDI and ZP for the selected formulations (Z-average, PDI and ZP results of all formulations are presented in Supplementary Table 1). The results for FAV-ASPs showed that when the amount of AP was decreased, the vesicle size of the ASPs significantly increased (p-value < 0.0001). Furthermore, the vesicle size significantly increased with increasing CH amount (p-value < 0.0001). This result was similar to the findings of a previous study by Babar et al (2025), where they developed Acyclovir loaded aspasomal gel for transdermal delivery system. Their results revealed that the amount of AP and CH had a positive impact on Z-average.⁴⁰ CH intercalates in the bilayer with its polar head, and due to its hydrophobic properties, it occupies the interior portion of lipid bilayers and can fill the gaps. The nonionic surfactant-based FAV-ASP and FAV-NIO formulations showed that with the increase in Span® 60 amount, the Z-average significantly increased (p-value < 0.0001). This result was also in accord with a previous study by Mathure et al (2018). They developed Buspirone loaded niosomes for nose-to-brain delivery system. Their results revealed that Z-average increased by increasing Span® 60 amount from 20 to 40 mg.⁸⁷ Soni et al (2024) explained the effect of Span® 60 on Z-average by the high interfacial activity and a critical packing parameter between 0.5 and 1, where the lower value means tight packed vesicles, thus a smaller Z-average, while the further increase in Span® 60 amount led to an increase in Z-average due to increase CH amount to fill the gaps and counteract the effect of Span® 60 leading to higher rigidity of the Methotrexate loaded niosomes.⁸⁸

Anionic surfactant-based FAV-ASPs exhibited bigger vesicles, which could be attributed to the influence of DCP that had a negative charge, leading to an increase in the interbilayer distance and the surface area of the vesicles. Similar results obtained by Varshosaz et al (2003) for Insulin loaded niosomes.⁸⁹

Moreover, FAV-ASP formulations had a broad size distribution (PDI > 0.3). However, PDI values < 0.5 were reported to be acceptable.⁸¹ Whereas FAV-NIO formulations showed a significantly smaller particle size compared with FAV-ASP formulations (p-value < 0.0001), and a narrow size distribution (PDI < 0.3). The results of the prepared formulations showed an acceptably homogenous size distribution; this distribution may depend on the synthesis of mixed multilamellar and large unilamellar vesicles during the hydration stage, as reported in previous studies.^42,72

ZP is a predictive indicator of the intensity of repulsive and attractive forces between vesicles, which allows the evaluation of the vesicle stability. The formulations had negative values of ZP, which maintain electric repulsion. These negative values depended on the chemical properties of AP and CH, also because of the application of negatively charged lipids (anionic surfactant-based FAV-ASP formulations). These results agreed with data previously reported by Gopinath et al.³¹ According to these results, the produced FAV-ASPs and FAV-NIOs were physically stable, preventing vesicle aggregation, and they have a longer residence time inside the brain without being effluxed.⁹⁰

Z-average, PDI and ZP are critical parameters which could impact the release, transport through direct or indirect pathways, biodistribution and stability.^91,92 As reported in the literature, nanoparticles with a Z-average of 150–300 nm were successfully taken up into the brain.⁷⁹ Accordingly, we selected the formulations with a Z-average < 300 nm for further investigations.

Determination of Drug Content, Encapsulation Efficiency (EE%) and Drug Loading (DL%)

The evaluation of the selected formulations (with a Z-average < 300 nm) indicating that FAV was successfully encapsulated into the vesicles with drug content ranged from 1.60 ± 0.36 mg to 1.85 ± 0.60 mg, and EE% values ranging between 41.01 ± 1.07% and 55.33 ± 0.41%. Drug content of the selected formulations did not alter significantly with the compositions of the formulations (p-value > 0.05). The EE% of FAV-NIOs was varied, and it did not alter significantly with changing of CH amount (p-value > 0.05), and the same result was obtained by Gopinath et al,³¹ while increasing the ratio of Span® 60 to CH increased EE% (p-value < 0.01). This outcome might be explained by an increase in the rigidity of the NIO membrane as CH amount increased to fill the gaps and counteract the effect of Span® 60, leading to higher rigidity; thus, the membrane fluidity decreased.⁹³ El-Sayed et al (2017) studied the effect of CH amount on EE% of Flurbiprofen loaded niosomes, and the results revealed a significant increment in EE% upon increasing CH amount.⁹⁴ The results of FAV-ASPs demonstrated that nonionic surfactant-based FAV-ASPs had a significantly higher EE% compared to anionic surfactant-based FAV-ASPs (p-value < 0.0001), which could be related to the impact of using AP and Span® 60. Moreover, EE% increased significantly by increasing the amount of Span® 60 (p-value < 0.0001), or by increasing the amount of AP (p-value < 0.01), or by increasing both AP and Span® 60 amounts (p-value < 0.0001).

Most nanoparticle systems have relatively low drug loading (<10 w%). It is critical to increase drug loading in order to improve the therapeutic effect and minimize adverse effects.⁹⁵ The nano-carrier system can be considered efficient if EE% > 50% and DL% > 5%.⁸⁶ The results showed that using ASP and NIO formulations enhanced DL % of FAV (as shown in Figure 2). FAV-ASP8, FAV-ASP10 and FAV-NIO9 showed acceptable EE % (>50%), high DL % (>5%) and drug content. Accordingly, we selected these formulations for further investigation. The results of drug content, EE % and DL % of the selected formulations are presented in Supplementary Table 2.

Residual Solvent Determination

Residual solvents are classified into four classes based on the toxicity level and the degree to which they can be considered as an environmental hazard. Both chloroform and methanol belong to Class II solvents; thus, their residual concentration should be less than 60 ppm and 3000 ppm, respectively, while ethanol belongs to Class III solvents (the accepted residual limit is 5000 ppm) in the daily dose of final product as a requirement of the International Council of Harmonisation (ICH) Q3C (R5) guideline for residual solvents. The results of GC measurement showed that the residual organic solvent contents were under LOD (less than 1 ppm for chloroform, methanol and ethanol), which supports that they were totally evaporated from the ASP and NIO vesicles during the film hydration method, ethanol injection method and freeze-drying.

In vitro Permeability Measurements

PAMPA is a cost-effective, relative, and reproducible technique for rapidly predicting passive transcellular transport,^96,97 which is one of the absorption routes of lipophilic molecules through the nasal mucosa and BBB,^98,99 so it is unable to give a complete description of the permeability via the BBB. However, PAMPA would be useful for choosing the most promising formulation, hence reducing in vitro and in vivo testing.¹⁰⁰

The BBB-PAMPA results showed that the formulations (FAV-ASP8, FAV-ASP10 and FAV-NIO9) had significantly higher permeability than pure FAV (Figure 3A). The addition of the surfactant agent to the formulations enhanced the solubility of FAV and increased its membrane permeability through BBB.⁸⁷ The effect of Span® 60 could also be observed in the flux values in the PAMPA study (Figure 3B). Based on these results, we selected FAV-ASP8 and FAV-NIO9 as optimal formulations due to the high permeability and flux values through the porcine brain polar lipid extract, high EE% and DL%. The results of the optimal formulations were summarized in Table 6.

Rapid Equilibrium Dialysis Measurement (RED)

RED assay was used to investigate the in vitro drug release of FAV from ASPs and NIOs vesicles at blood circulation conditions (DPBS, pH = 7.4). The incubation time of 4 h was sufficient for test formulations to reach equilibrium. The result showed that the dissolution rate of the optimal formulations was increased by approximately 3.5-fold compared to pure FAV (p-value < 0.0001), which could be related to the nanosized particles and the effect of AP and Span® 60. As shown in Figure 4, the amount released of FAV increased rapidly during the first hour, and the equilibrium state of the formulations started after 2 h. This result could be related to the release of desorption FAV from the surface of the vesicles, followed by diffusion of FAV through the bilayers. Similar result obtained by Taymouri et al (2016).¹⁰¹

In vitro Drug Release Under Nasal Conditions

Figure 5 presents the cumulative in vitro release of FAV from the optimal formulations in SNES. The study indicated that all formulations had a significantly higher release rate than FAV suspension. FAV-ASP8 released 77.31 ± 1.96% of the loaded drug after 60 min (p-value < 0.001), whereas FAV-NIO9 released 57.01 ± 1.22% (p-value < 0.05), in comparison with pure FAV (46.31 ± 1.83%). AP is a more stable amphiphilic derivative of ascorbic acid. It reduces the surface tension, especially when integrated into a phospholipid monolayer.¹⁰² Therefore, the integration of AP with CH and Span® 60 could effect on the fluidity of the vesicle membrane resulting in higher release rate compared to NIOs. Moreover, AP is pH neutral, so the hydrolysis of AP at pH 5.6 is slow, which could impact FAV release by diffusion without affecting the vesicle membrane integrity. In addition, increasing Span® 60:CH ratio may improve the integrity of NIOs membrane and reduce the FAV leakage, resulting in less drug release from the vesicles compared to ASPs. The result indicated that the dialysis bag did not restrict the diffusion of FAV. Furthermore, the use of ASPs and NIOs as carriers enhanced the encapsulation of FAV and increased the release rate due to the small particle size of the carrier, which increased the surface area and its hydrophilic properties.

In vitro Drug Release Data Evaluation

The release kinetics of the optimal formulations were statically compared with pure FAV using DDsolver®. The results showed that FAV-ASP8 had the highest AUC and DE values (Table 7). The AUC and DE values of all the optimal formulations were significantly higher than the AUC value of pure FAV. Moreover, the MDT values of FAV-ASP8 were significantly lower than those of pure FAV, which means a faster release rate in comparison to pure FAV and FAV-NIO9, which could increase the absorption through the nasal cavity.^103,104

The model with the highest R², the smallest AIC, and the highest MSC values is the best kinetics model.^50,51,105 The release profiles of the optimal formulations were fitted Korsmeyer-Peppas kinetic model. The release exponent “n” values of FAV-ASP8 and FAV-NIO9 were 0.44 ± 0.01 and 0.38 ± 0.01, respectively, which refers to a quasi-Fickian diffusion (n < 0.45)^50,106 (Supplementary Table 3).

Storage Stability

The stability test is important for the evaluation of the ability of the vesicles to resist environmental conditions, such as temperature, and for the protection of the drug from degradation, thus determining the recommended storage conditions and the retest period for the drug. The stability results for the optimal formulations showed that there was a decrease in Z-average value after 4 weeks (p-value < 0.0001) in case of FAV-ASP8, which can be explained by the change of mean intensity distribution of the size (Z-average is calculated by the intensity of the particle),¹⁰⁷ and the effect of AP as antioxidant which prevents the oxidative degradation of lipid bilayer. However, FAV-NIO9 showed a significant increase in Z-average value (p-value < 0.0001), presumably because of the aggregation of NIOs over time due to attractive forces (Van der Waals forces) or electrostatic interactions (repulsive forces). Moreover, the evaluation of PDI for FAV-ASP8 and FAV-NIO9 showed a non-significant increase, and both formulations maintained acceptably homogenous size distribution. The evaluation of ZP over time showed that FAV-ASP8 had a non-significant change in ZP, while FAV-NIO9 exhibited a decrease in ZP (non-significant), indicating a higher possibility to aggregate over time.¹⁰⁸ Furthermore, the six hydroxyl groups of mannitol can form hydrogen bonds with FAV (using the hydroxyl and carboxamide groups), which could impact the crystallization of mannitol and form a partially amorphous state. The amorphous state is less chemical stability and more hygroscopic.^109,110 Therefore, hydrolysis or oxidative degradation could occur over time. FAV-NIO9 showed a higher reduction of FAV concentration compared to FAV-ASP8. Thus, we can conclude that FAV-ASP8 had a higher stability than FAV-NIO9 (Supplementary Figures 1–4).

Droplet Size Distribution Measurement

Targeting the CNS requires particle deposition in the posterior region of the nasal cavity (to reach the olfactory and trigeminal nerves).¹¹¹ The deposition depends on many factors, including the interaction of the spray device with the nose (the angle of insertion and the plume angle),¹¹² nasal cavity (the mucociliary clearance mechanism and the human variabilities),¹¹³ and the formulation characteristics (droplet size, viscosity, elasticity).¹¹⁴

US FDA and EMA recommended the droplet size distribution measurement as an indicator of in vitro bioavailability for the liquid nasal formulation.^113,114 The nasal spray is recommended to be placed 3 cm below the receiving lens,¹¹³ with an application angle of 45°. A study by Jüptner et al (2025) demonstrated a more posterior deposition profile for 45° compared to 60° due to increased deposition in the turbinate region, and the overall nasal deposition was excellent at ~98%, with minimal loss from the nose.¹¹¹

The recommended droplet size ranges between 20 to around 200 µm.¹¹² Droplets with Dv50 < 10 µm can be inhaled through the nasopharynx and reach the lungs,¹¹⁵ or deposited in the anterior region of the nasal cavity which enhances formulation clearance from the nasal cavity.¹¹³

The results demonstrated that FAV-ASP8 and FAV-NIO9 were suitable for nasal administration (Dv50 values were 174.80 ± 12.08 and 89.05 ± 7.33 µm, respectively), which indicates a more posterior deposition, thus, increasing the absorption through the nasal cavity.

FAV-ASP8 showed a significantly higher Dv50 values compared to FAV-NIO9 (p-value < 0.01), which could be attributed to the effect of AP and Span® 60. Span® 60, as a non-ionic surfactant, reduced the interfacial tension and facilitated the formation of smaller droplets (FAV-NIO9),¹¹⁴ while the incorporation of AP could counteract the effect of Span® 60 leading to larger droplets (FAV-ASP8). Moreover, FAV-ASP8 and FAV-NIO9 showed small Span values (no significant differences), indicating a narrow width of droplet size distribution¹¹⁶ (Table 8).

Ex vivo Nasal Diffusion Study on Human Nasal Mucosa

The transmucosal permeability measurement of the optimal formulations in comparison with pure FAV was performed using human nasal mucosa. The study showed that both FAV-ASP8 and FAV-NIO9 had a higher amount of FAV diffused through the human nasal mucosa to the acceptor chamber within 60 min (1311.74 ± 41.70 µg/cm² and 402.94 ± 22.32 µg/cm², respectively) compared to pure FAV (128.17 ± 13.64 µg/cm², p-value < 0.0001), which means an enhancement in the absorption and permeation through the nasal mucosa (Figure 6). These results could be attributed to the nanosized particles (particle sizes < 500 nm can pass through the aqueous, non-viscous holes of the mucin network, resulting in higher mucosal penetration and absorption into the mucosal cells), negative ZP value (avoid slow penetration by limiting adherence to the negatively charged membrane), in addition to the effect of vesicles components.¹¹⁷ Furthermore, FAV suspension has high fluidity and negligible mucosal adherence.³⁸ As a result of the shorter contact time with the nasal mucosa, FAV suspension has a lower permeability.

Moreover, FAV-ASP8 showed a significantly higher permeability compared to FAV-NIO9 (p-value < 0.0001), which could be attributed to the surfactant-like property of AP (like sorbitan derivatives), Its lipophilic properties which enhance the tissue penetration, and its deformability which allowed ASPs to bypass through the tight junctions of the nasal cavity. D’Avanzo et al (2022) evaluated the permeation profile of Idebenone/Naproxen co‐loaded aspasomes by human stratum corneum and viable epidermis (SCE) membrane, and the results demonstrated enhancement in the permeability due to the impact of AP.⁴²

Table 9 summarizes the biopharmaceutical parameters, including permeation and prediction parameters. FAV-ASP8 had higher steady-state flux (J_ss), permeability coefficient (K_p) and theoretical plasma concentration at the steady-state in humans compared to FAV-NIO9 and pure FAV.

In vivo Study

Pharmacokinetic Parameters

FAV has a low solubility in water (8.7 mg/mL) and a low CNS permeability, which could be attributed to the low passive permeability due to its three H-bonding donors.²⁵ The capacity of FAV-ASP8 and FAV-NIO9 to deliver FAV to the CNS via the intranasal route was evaluated using Sprague–Dawley rats.

The average plasma drug concentration–time curves after a single nasal dose of pure FAV, FAV-ASP8 and FAV-NIO9 are demonstrated in Figure 7. FAV was detected and reached a plasma concentration peak (C_max) of 26.24 ± 7.382, 27.03 ± 6.88, and 20.32 ± 11.08 µg/mL at 7.5 ± 2.88, 7.5 ± 2.88, and 10 ± 4.08 min (T_max) after the administration of pure FAV, FAV-ASP8 and FAV-NIO9, respectively. These results indicate that the nasal mucosa effectively and rapidly transports FAV into the systemic circulation. In addition, FAV-ASP8 showed the highest t_1/2 value (226.42 ± 337.06 min) compared to pure FAV and FAV-NIO9. This could be related to the use of AP to prepare the drug carrier, which may reduce the elimination rate and thus extend the duration in the systemic circulation (C_L value was 0.79 ± 0.56 mL/min). The relative bioavailability (F%) values of FAV-ASP8 and FAV-NIO9 were 124.53 ± 35.33% and 85.28 ± 33.24%, respectively (Table 10).

Due to restrictions in CNS tissue access, CSF drug concentrations are used as an alternative for CNS tissue delivery, with the hypothesis that there is a direct, though not absolute, link between CSF and CNS tissue penetration.¹¹⁸ Furthermore, the olfactory nerves in the nasal cavity provide additional link with the CSF in the subarachnoid space through the interstitial fluid surrounding the olfactory nerve bundle.^119,120 In our study, FAV was detected in the CSF samples (Figure 8). The evaluation of CSF samples showed a significant elevation of FAV concentration into CSF after 1 hour of nasal administration of FAV-ASP8 compared to FAV suspension (*p-value < 0.05) and FAV-NIO9 (***p-value < 0.001), while there was no significant increase in the case of FAV-NIO9 compared to FAV suspension. These results can be attributed to an increase in the extent of drug absorption through the nasal mucosa (olfactory nerves) due to the physical parameter, the lipophilicity nature of the vesicles and the presence of a permeation enhancer (Span® 60), which improves the membrane penetration. Furthermore, AP can overcome biological barriers, enter the brain¹²¹ and resist hydrolysis,¹²² thereby preventing the degradation of FAV. According to the chemical structure of FAV, amide hydrolysis and oxidation are the potential major degradation pathways.¹²³

Abou-Taleb et al (2018) developed Nefopam loaded niosomes for intranasal administration using CH:Span® 80 (1:1.85 molar ratio). Their study illustrated a 4.77-fold increase in bioavailability compared with oral solution of drug.⁶¹ Moreover, Ammar et al (2017) developed Diltiazem loaded niosomes using CH:Span® 60 (1:1 molar ratio), which showed the highest EE% and release rate. The in vivo study on male Wistar rats showed an increase in MRT, t_1/2 and AUC with a decrease in K_e.¹²⁴

Due to the lipophilic nature of AP, it can easy bypass phospholipid bilayers of membranes, especially neuronal ones and BBB.^121,125,126 AP has been used in nasal formulations as a cross-linker, as reported by J. Varshosaz et al (2004). They developed insulin chitosan microspheres by emulsification-cross linking process using ascorbic acid or ascorbyl palmitate as a cross-linker. The formulation consists of 400 mg of chitosan and 70 mg of AP showed a 67% reduction of blood glucose compared to IV route.¹²⁷

Evaluation of in vitro–in vivo Correlations (IVIVC)

IVIVC is a useful tool to predict the in vivo performance of FAV based on its in vitro data. Figure 9 demonstrates the IVIVC graphs for the comparison of AUC_0-t values between the in vitro release and the in vivo PK data. The correlation coefficient (R²) values were 0.971, 0.982, and 0.903 for pure FAV, FAV-ASP8, and FAV-NIO9, respectively. Furthermore, R² values obtained from IVIVC graphs for the comparison of AUC_0-t values between the in vitro permeation and the in vivo PK data (Figure 10) showed a good point-to-point correlation (0.977, 0.992, and 0.839 for pure FAV, FAV-ASP8, and FAV-NIO9, respectively). This result indicates that the use of in vitro release and permeation data to establish the IVIVC could be useful in predicting the in vivo properties of the formulations.

Conclusion

In the present study, ASP and NIO formulations have been developed as promising intranasal delivery systems for FAV to enhance its bioavailability. The optimized formulation FAV-ASP8 (consisting of FAV:AP:Span® 60:CH in 30:25:25:50 w/w ratio) can encapsulate FAV inside ASP vesicles with an encapsulation efficiency of 55.33 ± 0.41% drug loading of 12.79 ± 0.22%. The formulation showed a small particle size (292.06 ± 2.10 nm), a narrow PDI value (0.36 ± 0.03) and negative values of ZP (−74.73 ± 3.28 mV), indicating a high physical stability for FAV-ASP8. The other optimized formulation, FAV-NIO9 contained FAV, Span® 60 and CH in 30:30:40 w/w ratio with a Z-average value of 167.13 ± 1.60 nm, a narrow PDI value (0.07 ± 0.01), adequate ZP (−27.1 ± 1.24 mV), acceptable encapsulation efficiency (51.30 ± 0.69%) and advantageous drug loading (15.42 ± 0.38%).

FAV-ASP8 and FAV-NIO9 formulations enhanced the solubility of FAV, elevated its membrane permeability through BBB, and increased its release under both nasal (pH = 5.6) and blood circulation conditions (pH = 7.4), with Dv50 values ranged between 20 and 200 µm. The ex vivo study showed that FAV-ASP8 had a higher amount of FAV diffused through the human nasal mucosa than FAV-NIO9 and pure FAV. The in vivo study indicated that there was no significant difference in plasma concentrations for the optimal formulations in comparison to pure FAV. However, only FAV-ASP8 had higher CSF concentrations than FAV-NIO9 and pure FAV after a single intranasal dose. IVIVC showed a good point-to-point correlation between the in vitro and the in vivo PK data.

These encouraging results suggest that ASPs could be promising intranasal delivery systems for hydrophobic drugs to enhance solubility, avoid first-pass metabolism, and decrease systemic adverse effects.