pharmaceutics Article Repurposing Disulfiram as an Antifungal Agent: Development of a New Disulfiram Vaginal Mucoadhesive Gel Maria Lajarin-Reinares 1,2, * , Iria Naveira-Souto 2 , Mireia Mallandrich 3 , Joaquim Suñer-Carbó 3 , Montserrat Llagostera Casas 1 , Maria Angels Calvo 4 and Francisco Fernandez-Campos 5 1 Department of Genetics and Microbiology, Campus Microbiology Unit, Autonomous University of Barcelona, 08193 Bellaterra, Spain; montserrat.llagostera@uab.cat 2 R & D Development, Reig Jofre Laboratories, 08970 Sant Joan Despí, Spain; inaveira@reigjofre.com 3 Department of Pharmacy and Pharmaceutical Technology and Physical Chemistry, School of Pharmacy and Food Sciences, University of Barcelona, 08028 Barcelona, Spain; mireia.mallandrich@ub.edu (M.M.); jsuner@ub.edu (J.S.-C.) 4 Department of Animal Health and Anatomy, Faculty of Veterinary, Autonomous University of Barcelona, 08193 Bellaterra, Spain; mariangels.calvo@uab.cat 5 R & D Department, Labiana Pharmaceuticals, 08757 Corbera Llobregat, Spain; francisco.fernandez@labiana.com * Correspondence: mlajarin@reigjofre.com; Tel.: +34-935-507-718 Abstract: Alternative formulations need to be developed to improve the efficacy of treatments administered via the vaginal route. Mucoadhesive gels with disulfiram, a molecule that was originally approved as an antialcoholism drug, offer an attractive alternative to treat vaginal candidiasis. The aim of the current study was to develop and optimize a mucoadhesive drug delivery system for the local administration of disulfiram. Such formulations were composed of polyethylene glycol and carrageenan to improve the mucoadhesive and mechanical properties and to prolong the residence time in the vaginal cavity. Microdilution susceptibility testing showed that these gels had antifungal Citation: Lajarin-Reinares, M.; activity against Candida albicans, Candida parapsilosis, and Nakaseomyces glabratus. The physicochemical Naveira-Souto, I.; Mallandrich, M.; properties of the gels were characterized, and the in vitro release and permeation profiles were Suñer-Carbó, J.; Llagostera Casas, M.; investigated with vertical diffusion Franz cells. After quantification, it was determined that the Calvo, M.A.; Fernandez-Campos, F. amount of the drug retained in the pig vaginal epithelium was sufficient to treat candidiasis infection. Repurposing Disulfiram as an Together, our findings suggest that mucoadhesive disulfiram gels have the potential to be an effective Antifungal Agent: Development of a alternative treatment for vaginal candidiasis. New Disulfiram Vaginal Mucoadhesive Gel. Pharmaceutics Keywords: disulfiram; Candida spp.; mucoadhesion; ATP-binding cassette; resistant; vaginitis 2023, 15, 1436. https://doi.org/ 10.3390/pharmaceutics15051436 Academic Editor: Natasa Skalko-Basnet 1. Introduction Received: 21 March 2023 Vaginal candidiasis (VC), often referred to as vulvovaginal candidiasis, is mainly Revised: 2 May 2023 caused by the Candida species and is the second most prevalent infection of the vaginal Accepted: 3 May 2023 mucosa after bacterial vaginosis [1]. VC affects approximately 75% of women of reproduc- Published: 8 May 2023 tive age at least once during their lives, and it is estimated that 40–50% will experience recurrence and 8% will develop chronic VC [2–4]. Candida spp. are commensal fungi in the human microflora that are found in the lower genital tract in 10–20% of healthy women. Although part of the normal microflora, Candida Copyright: © 2023 by the authors. can become opportunistic agents with a tendency to overgrow [2]. The most common Licensee MDPI, Basel, Switzerland. Candida species involved in VC is C. albicans (85–90%), followed by Nakaseomyces glabratus This article is an open access article (C. glabrata), C. tropicalis, Pichia kudriavzevii (C. krusei), and C. parapsilosis [5]. Approximately distributed under the terms and 25 to 40% of women who are culture-positive for Candida are asymptomatic. The natural conditions of the Creative Commons evolution of asymptomatic colonization to symptomatic infection is not clear. There are Attribution (CC BY) license (https:// multiple mechanisms that contribute to the colonization of Candida and its symptomatic creativecommons.org/licenses/by/ infection, such as genetic host factors (polymorphisms in the blood group), hormones, 4.0/). Pharmaceutics 2023, 15, 1436. https://doi.org/10.3390/pharmaceutics15051436 https://www.mdpi.com/journal/pharmaceutics Pharmaceutics 2023, 15, 1436 2 of 21 the use of antibacterial agents, age, sexual activity, certain pathologies (diabetes mellitus and immunodepression), idiopathic causes, pregnancy, and an imbalance in the vaginal microbiota [4,6,7]. The most common clinical manifestations of VC are pruritus, hyperemia, vaginal discomfort, leucorrhea, burning, soreness, dyspareunia, and vaginal erythema [7]. Furthermore, colonization of the vaginal epithelium contributes to the deterioration of mucosal surfaces, facilitates the formation of yeast reservoirs, or becomes a source for future infections, leading to chronic infections and serious consequences (i.e., infertility and sterility) [8]. The treatment of noncomplicated VC involves the use of short-term local therapy (up to 3–7 days) or single-dose oral antimycotics, mainly those in the azole family, which is effective in 90% of cases. The most common locally applied azoles are clotrimazole, buto- conazole, and miconazole. Fluconazole is usually given orally between 1 and 3 days [9]. Topical azoles may provide more immediate relief; however, some patients experience hypersensitivity, local itching, or burning [10]. For vaginal drug delivery, several phar- maceutical formulations are available for consumers, including creams, ovules, tablets, and gels [8,11,12]. These treatments exhibit some drawbacks, such as low permeation in the underlying epithelium (to remove reservoirs) and low residence time owing to the self-cleaning action of the vaginal tract. (It is recommended to be administered at night before lying in bed.) These factors could lead to low drug exposure and the development of resistance. Many studies have documented the ability of Candida to develop high-level resistance to antifungal therapy [13–16]. Resistance mechanisms include overexpression of or mutations to the target enzyme of azoles, lanosterol 14a-demethylase, and overexpres- sion of ATP-binding cassette (ABC) efflux proteins [17]. Among ABC transporters, Cdr1p plays a key role in azole resistance in C. albicans [18]. Disulfiram (D) is a dithiocarbamate used to treat alcoholism that was approved by the US Food and Drug Administration (FDA) in 1951. D produces discomfort in patients who drink alcohol as a strategy for their dishabituation, inhibiting aldehyde dehydrogenase (ALDH), which results in the specific accumulation of acetaldehyde [19]. In addition, D is characterized by a low molecular weight (296 Da), logP value of 3.88, melting point under 200 ◦ C (71.5 ◦ C), and fewer than five hydrogen bond donors and acceptors with a strong ability to chelate metals [20]. In accordance with its physicochemical properties, a good permeation profile is expected. Different in vitro and in vivo studies have shown that D could be an active substance against scabies, lice, and bacteria [21,22]. Furthermore, the effectiveness of D against fungi and cancer was reported [23,24]. Different mechanisms of action have been proposed to explain the antimicrobial activity of D: metal ion chelation, which affects cell homeostasis, and the presence of reactive thiol groups, which can impact the functions of proteins by forming incorrect disulfide bonds [22]. D was found to be potentially useful by inhibiting the function of glycoprotein P (P-gp) and has also been shown to act as a modulator of other multidrug transporters, such as MRP1 and MRP4 [25]. Because of the functional similarities between Cdr1p and human P-gp, D is a potent modulator of Cdr1p [17]. Therefore, D is an attractive agent to develop for candidiasis treatment. The aim of this work is to develop and optimize a mucoadhesive drug delivery system for the local administration of D. The in vitro antifungal efficacy of D was studied with different Candida species, and the mucoadhesion of the formulation was enhanced using polymers with different adhesive mechanisms. The proposed formulation represents an alternative to the commercially available treatments for VC due to the selected novel drug and mucoadhesive polymers that would increase the residence time of the treatment and yeast drug exposure. 2. Materials and Methods 2.1. Materials Sodium chloride (0.35% w/v) (Quimivita, Barcelona, Spain), potassium hydroxide (0.14% w/v), calcium hydroxide (0.02% w/v), acetic acid (0.1% w/v) (Scharlab, S.L., Pharmaceutics 2023, 15, 1436 3 of 21 Sentmenat, Spain), bovine serum albumin 0.002% (w/v) (Merck Life Science, S.L.U., Madrid, Spain), lactic acid (0.2% w/v) (Escuder, S.L., Rubi, Spain), glycerol 0.02% (w/v) (Caila & Pares, S.A., Barcelona, Spain), urea (0.04% w/v) (Quality Chemicals, S.L., Esparraguera, Spain) and glucose (0.5% w/v) (Quimidroga, S.A., Barcelona, Spain) were used to prepare simulated vaginal fluid (SVF) [26]. Mucoadhesive gels were produced with polyethylene glycol (PEG-90M) (Colorcon Limited, Kent, UK) and carrageenan (DuPont, Wilmington, DE, USA); benzylic alcohol (Scharlab, S.L., Sentmenat, Spain) was chosen as a preservative; and hydroxypropyl-beta- cyclodextrin (HPCD) (Pracofar, Martorelles, Spain) was used as an active solubilizing agent. D (Farchemia S.r.l., Treviglio, Italy) was the active pharmaceutical ingredient, and the selected solvent was water purified in-house. Sodium fluorescein (Scharlau, Barcelona, Spain) was used in the mucoadhesive study as a marker. 2.2. Antifungal Susceptibility Testing The fungal strains used in this study were purchased from the Spanish Collection of Type Cultures (CECT). C. albicans CECT 1002 (which corresponds to ATCC 18804), C. parapsilosis CECT 10434 (MCYC 2031), and C. glabrata CECT 1448 (ATCC 2001) were used for the antifungal susceptibility test. The minimum inhibitory concentration (MIC) of D was determined by the broth microdilution method in 96-well microplates according to guidelines M27-A3 of the Clinical and Laboratory Standards Institute (CLSI) standard [27]. A 24 h culture of Candida adjusted to a 0.5 McFarland was used to prepare the initial inoculum. A working suspension was generated by making a 1:100 dilution followed by a 1:20 dilution of the stock suspension with RPMI 1640 broth medium buffered with MOPS (3-(N morpholino)propanesulfonic acid) (Thermo Fisher Scientific, Waltham, MA, USA) at pH 7.0. Stock solutions (10 mg/mL) of D were prepared in ethanol, and D was tested from 64 (the highest soluble concentration) to 0.125 µg/mL. The final ethanol concentration was below 0.5% to not alter yeast growth. The negative control was the respective medium, and a control with the same ethanol concentration as that in the samples with D was evaluated. Furthermore, each fungal strain in medium was used as a positive control, and amphotericin B was used as a positive control of inhibition. The plates were incubated at 35 ◦ C for 48 h, and then, the MIC values were determined. The M27 methodology gave reproducible results when testing Candida, and the establishment of interpretive breakpoints has begun. However, it is becoming clear that the M27 methodology may be inadequate for certain organisms and antifungal agents. For example, determining the susceptibility of Candida isolates to amphotericin B may be a concern [28]. D is not described in these guidelines, and the suitability of the M27 method is not known. To resolve the inherent possible variability of MIC determination, MICs were also determined an alternative medium (Sabouraud dextrose broth). 2.3. Preparation and Optimization of the Mucoadhesive Gel For the mucoadhesive placebo gel preparation, purified water (q.s., 100% w/w) was heated to 50 ◦ C with HPCD (20% w/w). PEG-90M was added, it was stirred at 750 rpm with a mixing paddle, and then stirring was reduced to 150 rpm until complete dissolution. Finally, carrageenan was added with stirring at 75 rpm until complete homogenization. Disulfiram gels were produced in the same way by adding D (0.5% or 1% w/w) to purified water with HPCD. Fluorescein (0.01% w/w) mucoadhesive gels were produced to study the in vitro mucoadhesive properties. The properties of formulations were influenced by different experimental variables, such as the amounts of the polymers. To obtain the most suitable mucoadhesive gel for vaginal application, the influence of the concentrations of PEG-90M and carrageenan on the dependent variables (mucoadhesion, fluorescein adhesion, and viscosity) was studied (Table 1) with placebo formulations. Central composite design was employed to analyze the main and combined effects of these variables on the responses. For this, 5 central points were selected, and a default α value of 1.41 was used to determine the axial points. A Pharmaceutics 2023, 15, x FOR PEER REVIEW 4 of 21 the dependent variables (mucoadhesion, fluorescein adhesion, and viscosity) was studied Pharmaceutics 2023, 15, 1436 4 of 21 (Table 1) with placebo formulations. Central composite design was employed to analyze the main and combined effects of these variables on the responses. For this, 5 central points were selected, and a default α value of 1.41 was used to determine the axial points. A significance (α) of significance level (α) of 0.05 0.05was wasestablished. established.The Thechoice choice ofof the the model model was was made made based based on on 2 , adjusted 2R2 , and AIC (Akaike information criterion). In addition, the thethe values values of of R2R, adjusted R , and AIC (Akaike information criterion). In addition, the production production order order ofof the the formulations formulations was was randomized randomized to to reduce reduce the the chances chances of of external external design factors affecting our results. design factors affecting our results. Table 1. Levels Table 1. Levels studied studied for for surface surface response responsedesign. design. Factor Factor LowerLevel Lower Level Higher Level Higher Level % PEG-90M (w/w) % PEG-90M (w/w) 0.5 0.5 1.5 1.5 %%Carrageenan (w/w) Carrageenan (w/w) 1.0 1.0 2.0 2.0 Statistical analysis Statistical analysisofof the the variables variables studied studied was was carried carried out out using using Minitab Minitab 1717 statistical statistical software (Minitab, Inc., 2010, State College, PA, USA) to obtain mathematical equations software (Minitab, Inc., 2010, State College, PA, USA) to obtain mathematical equations for for each each model. model. Then, Then, optimization optimization of each of each model model waswas performed performed (Section (Section 2.7).2.7). 2.4. 2.4. Adhesive Strength of the Gels The The adhesive adhesive strengths strengths of of the the placebo placebo gels gels were were evaluated evaluated byby modifying modifying thethe devices devices from from Mei Meietetal.al.[11] (Figure [11] 1). 1). (Figure To To simulate the vaginal simulate epithelium, the vaginal agar hydrogels epithelium, (4% w/w) agar hydrogels (4% (Scharlab, Barcelona, w/w) (Scharlab, Spain) Spain) Barcelona, were prepared with type were prepared II pig with typemucin mucinw/w) II pig(0.34% (0.34%(Sigma– w/w) Aldrich, St. Louis,St. (Sigma‒Aldrich, MI,Louis, USA)MI, [29].USA) This [29]. gel was Thisattached gel wastoattached two planks, as shown to two planks,inasFigure shown 1. The formulation (0.5 g) was placed on the bottom plank, and to evaluate in Figure 1. The formulation (0.5 g) was placed on the bottom plank, and to evaluate the the effect of vaginal effect offluid on the vaginal formulations, fluid SVF (0.25 SVF on the formulations, mL) was (0.25applied mL) was toapplied the formulation before the to the formulation upper plank was pressed onto the bottom plank and the formulation. before the upper plank was pressed onto the bottom plank and the formulation. Figure 1. Figure 1. Schematic Schematic illustration illustration of of the the device device for for evaluating evaluating adhesive adhesive strength. strength. One One milliliter milliliter of of water water was was added added every every 15 15 ss to to aa tared tared beaker beaker fixed fixed on on the the opposite opposite side side of the balance. Once the two planks were separated, the weight of the water(W of the balance. Once the two planks were separated, the weight of the water (W11)) was was used 2 according to Equation (1): used to to calculate calculate the the tensile tensile stress stress (σ; (σ; mN/cm mN/cm2)) according to Equation (1): σ =σ(W = (W 1 ×1 × g)/s g)/s (1) (1) where g is gravity (m/s2), and s is the agar surface (cm2). where g is gravity (m/s2 ), and s is the agar surface (cm2 ). Pharmaceutics 2023, 15, x FOR PEER REVIEW 5 of 21 Pharmaceutics 2023, 15, 1436 5 of 21 2.5. Viscosity of the Mucoadhesive Gels 2.5. Viscosity of the Mucoadhesive Gels As a screening step in the selection of the final formulation composition, the As a screening step in the selection of the final formulation composition, the viscosi- viscosities of the mucoadhesive gels were evaluated using a rotatory viscosimeter ties of the mucoadhesive gels were evaluated using a rotatory viscosimeter (Brookfield, (Brookfield, Manassas, VA, USA). The viscosities of the samples were measured 24 h after Manassas, VA, USA). The viscosities of the samples were measured 24 h after production production of each formulation. Moreover, viscosities of the same samples with SVF in a of each formulation. Moreover, viscosities of the same samples with SVF in a ratio of ratio of 1:6.67 (SVF:gel) were measured after mixing and resting for 48 h. Ten milliliters of 1:6.67 (SVF:gel) were measured after mixing and resting for 48 h. Ten milliliters of each each sample sample was was applied applied to the to the sample sample container container at at 25.0±±0.5 25.0 0.5 ◦°C. C. Viscosity Viscosity measurements measurements were were conducted for 30 min at 10 rpm (spindle SC4-29). The ratioof conducted for 30 min at 10 rpm (spindle SC4-29). The ratio ofSVF SVFtoto gel gel used used was was considered to allow approximately 0.5–0.75 mL of SVF to be present at any one considered to allow approximately 0.5–0.75 mL of SVF to be present at any one time [30]. time [30]. 2.6. 2.6.Mucoadhesive MucoadhesiveProperties Properties For For the mucoadhesive mucoadhesive studies,studies,a avagina vaginaandandits its physiological physiological conditions conditions werewere sim- simulated ulated using using a tube a tube withwith a portion a portion of of agar agar (8(8 ××22cm)cm)with with0.34% 0.34%w/ww/w mucin, mucin, on on which which 11 gg of of mucoadhesive mucoadhesive gel gel with with fluorescein (0.01% w/w) w/w)was wasplaced placedon ontop topofofthe the tube. tube. The The tubes tubes were were mounted at an an approximate approximateangle angleofof70 ◦ under 70° underphysiological physiologicalconditions. conditions.SVF SVFat at 1 rpm was passed through the tube using 1 rpm was passed through the tube using a peristaltic a peristaltic pump (Shenchen Pump YZ1515x, (Shenchen Pump YZ1515x, Hebei, Baoding, China) at 0.0133 China) at 0.0133 mL/s, mL/s,thethe lowest flow lowest of the flow peristaltic of the pump. peristaltic pump.Friedl Friedlet et al.al. used a used total of of a total 6 mL 6 mLofof SVFSVFover over2424h h(v(v= =0.00006944 0.00006944mL/s), mL/s),taking takingsamples samplesatat44h,h,88h,h,12 12h, h,and and 24 24 hh [31]. [31]. Extrapolating Extrapolatingthese thesesampling samplingtimestimestotothe theflow flow of of our our system, system, the the determined determined samplingtimes sampling times were were as as follows: follows: 1.15 1.15min, min, 2.30 2.30 min, min, 3.45 3.45 min, min, and and 7.30 7.30 min. min. TheThe whole whole experimentalsetup experimental setup isis illustrated illustrated inin Figure Figure 2. 2. Figure Figure2.2.Schematic Schematicof ofthe thedevice deviceused usedto tomeasure measurethe themucoadhesive mucoadhesiveproperties. properties. At Atthe the end end of of the the study, study, the the concentration concentration of of fluorescein fluorescein released released (F(F22)) at at each each time time point was point wasdetermined determinedby bymeasuring measuringthe thefluorescence fluorescenceat ataawavelength wavelengthof of485 485nm nmusing usingaa Victor Nivo Victor Nivo multimode multimode plate plate reader reader (Perkin (Perkin Elmer, Elmer, Waltham, Waltham, MA, MA, USA). USA). Before Before sample sample measurement, aafluorescein measurement, fluoresceincalibration calibration curve curvewas wasprepared preparedby bymeasuring measuringthe thefluorescence fluorescence offluorescein of fluorescein solutions in in water waterfrom from7.6 10−−44µg/mL 7.6×× 10 µg/mLtoto22××1010−5− 5 µg/mL 2(R2 = 0.999928). µg/mL (R = 0.999928). Once the samples Once samples werewerequantified, quantified,the thefluorescent fluorescent dose remaining dose remaining in the in theagaragar (F3 ) (F was3) determined was determinedby calculating by calculatingthe difference the difference fromfrom the fluorescein dosedose the fluorescein initially seeded initially seeded(F1 ) following (F the the 1) following equation equation F3 = FF23−=FF1 2 − F1 The FF33 value The valuewaswasused as an used asindirect parameter an indirect of the mucoadhesion parameter of the formulation. of the mucoadhesion of the formulation. 2.7. Optimization of the Mucoadhesive Gels Pharmaceutics 2023, 15, 1436 6 of 21 2.7. Optimization of the Mucoadhesive Gels Once the mathematical models for each variable were obtained, the two factors (PEG- 90M and carrageenan concentration) were examined to optimize the mucoadhesive proper- ties and viscosity of the formulation. The concentration of polymers giving the formulation the highest mucoadhesion and lowest viscosity was determined. The mucoadhesive properties and viscosity of the optimized final formulation placebo and the formulations with 0.5 and 1% D (w/w) were studied. Furthermore, from the mathe- matical model obtained during the statistical study, the bias was calculated (Equation (2)) to determine the predictive capacity of the model. | Re f − Z1 | %Bias = × 100 (2) Re f where Ref is the theoretical reference value of the model, and Z1 is the mean of the experi- mental value obtained. Additionally, the rheological properties (Section 2.8) and syringeability (Section 2.9) of the placebo-optimized formulation and the release (Section 2.10) and permeation (Section 2.11) profile of the final D formulation were characterized. 2.8. Final Rheological Properties Placebo-optimized formulation rheology, thixotropy, and viscoelasticity measurements were made by performing rotational and oscillatory tests using a Haake Rheostress® 1 rheometer (Thermo Fisher, Karlsruhe, Germany). Rotational measurements. Steady-state measurements were made with a parallel pate and plate geometry (P35Ti L: 60 mm diameter, 2◦ angle). The shear stress (τ) was measured . . as a function of the shear rate (γ). Viscosity curves (η = f(γ)) and flow curves (τ = f(γ)) were recorded at 37 ± 0.1 ◦ C. The shear rate ramp program included a 3 min ramp-up period from 0 to 100 s−1 , a 1 min constant shear rate period at 100 s−1 , and a 3 min ramp-down period from 100 to 0 s−1 . Representative mathematical models were fit to the flow curves to search for the best descriptive model (Table 2) [32]. Selection of the best-fitting model was based on the correlation coefficient (observed vs. predicted) and Chi-square value. The apparent thixotropy (Pa/s) was estimated as the area of the hysteresis loop. Steady-state viscosity (η, Pa s) was determined from the constant shear section at 100 s−1 . Table 2. Rheological equations used to evaluate the prepared formulations. Rheological Model Equation . Newton τ = η×γ  . Bingham τ = τ0 + η0 × γ .n Ostwald–de Waele τ = K×γ .n Herschel–Bulkley τr= τ0 + K × γ . n   Casson τ = n τ0n + η0 × γ . . . n  Cross  τ = γ × (η ∞ + (η 0 − η∞ / 1 + γ/γ 0 . τ: Shear stress (Pa), η: viscosity (cP or Pa-s), γ: shear rate (s−1 ), τ0 : minimum yield stress (Pa), η0 : zero shear viscosity . (cP or Pa-s), K: consistency factor (Pa sn ), η∞ : infinite shear viscosity (cP or Pa-s), γ0 : zero shear rate (s−1 ). Oscillatory test. The oscillatory test was performed with parallel plate–plate geometry (P35Ti L: 60 mm diameter, 2◦ angle). A strain sweep test was performed with a strain range of 0.01 to 500 Pa and an oscillation rate of 1 s−1 to determine the linear viscoelastic region (LVR) of the samples. During each sweep stress test, the changes in the storage and loss moduli and the phase angle (G0 , G”, and δ, respectively) were plotted as a function of shear stress. Afterward, a frequency sweep test was carried out between 0.1 and 10 s−1 at a constant shear rate within the LVR to determine the related variations in the G0 , G”, and complex viscosity (η*), which were used for sample characterization. Pharmaceutics 2023, 15, 1436 7 of 21 All obtained data were analyzed with HAAKE RheoWin® Data Manager v. 4.88. 2.9. Syringeability of the Formulations Syringeability was determined by calculating the time needed for 5 g of the product to fall from a cannula with the application of a constant force of 9.688 N (applied weight of 987.6 g) [33]. 2.10. In Vitro Release Tests of the Disulfiram Gel The in vitro release of D from the mucoadhesive gels was studied using vertical Franz cells (Vidrafoc, Barcelona, Spain) with a 12 mL receptor compartment and an effective diffusion area of 1.54 cm2 . A 15% w/w HPCD solution in PBS (pH 5.5) was used as the receptor medium (RM) at 37 ◦ C with stirring at 500 rpm to maintain sink conditions throughout the experiment due to the poor water solubility of D (4.09 mg/L) [20] and according to previous studies reported [21,22]. A total of 0.0545 g/cm2 gel was applied in the donor compartment, corresponding to 420 µg of D for the 0.5% formulation (n = 5) and 840 µg of D for the 1% formulation (n = 5). A membrane with a 0.45 µm pore diameter was used (Supor PES membrane, Thermo Fisher, Waltham, MA, USA). Aliquots of 300 µL were taken at certain times (0, 0.5, 1, 1.5, 3, 4, 5, and 6 h). The samples were analyzed using High-Performance Liquid Chromatography (HPLC) with a method previously reported for similar studies [21] to quantify the amount of D that had diffused through the membrane. Kinetic modeling of the release data was carried out with the DD-solver Excel Add- on [34]. The mean values of the release curve were adjusted to nonlinear models, i.e., first-order, Higuchi, Korsmeyer–Peppas, and Weibull (Table 3), to select the population behavior. Then, individual data were adjusted according to the selected population model. Model selection was based on the AIC, which reflected the lowest deviation of the model with respect to the empirical data [35]. The mean and standard deviation of the parameters were reported. Table 3. Different kinetic models and equations tested. Kinetic Model Equation Parameter(s)   First-Order F = Fmax 1 − e(−K1 t) Fmax , K1 Higuchi F = KH × t1/2 KH Korsmeyer–Peppas F = KKP × tn KKP , n Weibull (− t T )β Fmax , α, β F = Fmax × 1 − e d F: Fraction of drug released, t: time, Fmax : maximum fraction of drug released, K1 : first-order constant, K H : Higuchi constant, KKP : Korsmeyer–Peppas constant, n: diffusional exponent, Td : the time at which 63.2% of the drug had been released, β: Weibull shape parameter. In Table 3, if n is between 0.43 and 0.85, then the release mechanism follows an anomalous transport mechanism. In the case of β, for values lower than 0.75, the release follows Fickian diffusion, either in Euclidian (0.69 < β < 0.75) or fractal (β < 0.69) spaces. Values (β) in the range of 0.75–1.0 indicate a combined mechanism, which is frequently encountered in release studies [36]. 2.11. Pig Vagina Permeation Pig vaginas were obtained at the time of sacrifice from a local abattoir (Barcelona, Spain). The full thickness vaginal mucosa (approximately 6.2 mm) was carefully debrided, cleaned with sterile saline solution, and transported to the laboratory at 4 ◦ C in saline solution. For this study, tissues were frozen by placing them in containers with a PBS mixture containing 4% albumin and 10% DMSO (as cryoprotective agents) and stored (for a maximum of 1 month) at −80 ◦ C in a mechanical freezer. DMSO produces adverse effects at room temperature; therefore, the addition of DMSO prior to freezing was performed at 4 ◦ C, whilst thawing involved immersion in a water bath filled with PBS at 37 ± 1 ◦ C and gentle shaking for 30 min, until total elimination of DMSO was achieved [37]. On Pharmaceutics 2023, 15, 1436 8 of 21 the day of the experiment, the vaginal pieces were thawed and mounted on Franz cells (n = 4 for each dose) with an effective diffusion area of 0.64 cm2 and approximately 5 mL of receptor volume capacity (HPCD 15% in PBS, pH = 5.5 with stirring at 600 rpm). A total of 0.0545 g/cm2 gel (0.5% or 1%) was administered in infinite doses under nonoccluded conditions. Samples from the receptor compartment (300 µL) were taken at regular time intervals of 1.5, 2.5, 3.5, 4.5, and 6 h. Samples were analyzed and quantified using the HPLC method described in the previous section. After drug quantification, the following permeation parameters were calculated: trans- mucosal flux (J, µg/cm2 h) (Equation (3)), permeability coefficient (Kp, cm/h) (Equation (4)), lag time (Tlag , h) (obtained by linear extrapolation of the x-axis of the points at steady state), dif- fusion parameter (Dif, 1/h) (Equation (5)), and partition parameter (P, cm/h2 ) (Equation (6)). dQ J= , (3) dT × S where J is the transdermal flux, dQ is the difference in the amount permeated, dT is the time differential, and S is the membrane diffusion surface. J Kp = , (4) Cd where Cd (µg/mL) is the concentration of the drug in the donor compartment. 1 Di f = . (5) 6T lag Kp P= . (6) Di f 2.12. Determination of the Concentration Retained in the Pig Vagina After the permeation studies, the retained D was extracted from the mucosal sections. The mucosa that had come in contact with the formulation was cut into 4 equal pieces (approximately 50 mg each) per Franz cell for introduction into the MagNa Lyser instrument (Roche, Sant Cugat del Valles, Spain) with 1 mL of mobile phase. Homogenization of the tissue was carried out with 5 cycles of 90 s each at 6500 rpm. Finally, the samples were analyzed using the HPLC method [21]. 3. Results and Discussion 3.1. Antifungal Susceptibility Testing The susceptibility of different species of Candida to D was determined in vitro by the microdilution broth method after 48 h of treatment in two different media. The results are presented in Table 4. Table 4. Summary of the MICs (n = 3) for disulfiram with three different species of Candida using two different test media. MIC (µg/mL) Test Media Candida spp. RPMI 1640 Sabouraud Dextrose C. albicans 2 2 C. glabrata 4 2 C. parapsilosis 8 8 RPMI 1640 with MOPS is the medium proposed by CLSI, but the suitability of this medium for D is not well known. To determine if RPMI is adequate, we compared the results with this medium with the results with Sabouraud dextrose, a medium classically Pharmaceutics 2023, 15, 1436 9 of 21 used for yeast testing [38,39]. In general, good agreement was obtained between the MIC values with the two different media, confirming the suitability of RPMI 1640. Regarding the MIC values obtained with RPMI 1640, D was effective against the three species tested at concentrations ranging from 2 to 8 µg/mL. These results correlate well with published data, reporting MIC values ranging from 2 to 8 µg/mL for C. albicans (ATCC 90028 and 36082), 16 µg/mL for C. parapsilosis (ATCC 22019), and 4 µg/mL for N. glabrata (ATCC 90030) [40,41]. The differences in the results could be due to the different strains used from different isolation sources. Furthermore, conventional antifungal (amphotericin B) was used as the control for the yeast susceptibility method. CLSI guidelines suggest that comparing the MICs of different antifungals should not be based solely on the numerical value but rather on how far the MIC is from the breakpoint [27]. Regarding D, there are no defined clinical breakpoints because there are no published data on natural mutants resistant to it. Therefore, it was difficult to use the classical approach to compare D activity with other antifungals. The key aspects of the disulfiram mode of action that could explain its antifungal activity can be understood by two effects. First, disulfiram is a chelator agent that can sequester copper and consequently alter various metabolic pathways in cells; and second, D has an affinity for the thiol groups of cysteine residues in several cellular targets [21,22]. D may be considered a potential agent for the treatment of candidiasis. Moreover, it has been noted that D inhibits the activities of ABC drug transport proteins that are associated with antifungal resistance [17]. Therefore, its synergism with other antifungals could be an option, and inhibition of ATP pumps would probably result in an increase in the intracellular concentration of the second drug, thus making the second drug, such as an azole, more effective in candidiasis therapy. Additional studies should be carried out to examine this possible effect in clinically relevant strains. 3.2. Preparation and Statistical Design to Evaluate the Effect of Experimental Variables on the Mucoadhesive Gel Properties Once the interesting antifungal properties of D were examined, a novel vaginal placebo formulation was developed to increase mucoadhesion with the vaginal epithelium. Mu- coadhesion is defined as the interaction between a synthetic or natural polymer and a mucin surface. It can be affected by different factors, including molecular weight, hydrophilicity, crosslinking, swelling, pH, and polymer concentration [42]. Polymer concentration is an important factor, as there is an optimal concentration at which mucoadhesion reaches a maximum [43]. Additionally, polymers affect the mechanical and rheological properties of the formulation; therefore, controlling the concentration of the polymer combination is important to allow the formulation to adapt to its site of application and ensure appropriate efficacy [44]. PEG-90M was selected as the polymer because of its mucoadhesive properties, its bio- compatibility, and its approval by different pharmacopoeias. In addition, PEG-90M presents controlled release properties, especially when used in solid oral dosage formulations [45]. Carrageenan is an anionic polymer with sulfate functional groups that can crosslink with monovalent cations (i.e., sodium and potassium) [46,47]. Carrageenan was selected because it can interact with these ions in the vaginal fluid and increase the hydrogel strength and the residence time in situ, avoiding possible loss of the formulation. Both components have been widely used for vaginal administration of other antimicrobial agents [48–52], including antimicrobial properties for carrageenan [53,54]. This combination has not been described and is of great interest due to the different and complementary mucoadhesion mechanisms of both polymers. Furthermore, due to the hydrophobic nature of D, 20% w/w HPCD was added to the formulation. To study the effects of the polymers on the formulation properties (viscosity and mucoadhesion), the percentages of PEG-90M and carrageenan were modified between 0.50% w/w and 1.50% w/w and 1.00% w/w and 2.00% w/w, respectively. As shown in Table 5, thirteen formulation batches were produced, and the viscosity, mucoadhesion, Pharmaceutics 2023, 15, x FOR PEER REVIEW 10 of 21 Pharmaceutics 2023, 15, 1436 10 of 21 fluorescein is a hydrophilic compound that would be more easily removed from the formulation and amountwhen SVF passes of adhered over. The fluorescein viscosity were and mucoadhesion characterized were Fluorescein for each of them. measured with was and without SVF to evaluate the effect of carrageenan crosslinking. selected as the model drug, representing the opposite case compared with D because fluorescein is a hydrophilic compound that would be more easily removed from the Table 5. Influences formulation when ofSVF PEG-90M and passes carrageenan over. on theand The viscosity physicochemical mucoadhesion parameters. were measured with and without SVF to evaluate the effect of carrageenan Viscosity crosslinking. Mucoadhesion Adhered PEG-90M Carrageenan Viscosity Mucoadhesio Batch with SVF with SVF Fluorescein (% w/w) (% w/w) (cP) n (mN/cm2) Table 5. on the physicochemical(mN/cm Influences of PEG-90M and carrageenan(cP) 2) parameters. (%) LP-83 0.50 1.50 19,910 16,170 50.6 60.5 91.40 LP-84 2.00 1.50 61,660 Viscosity 39,350 65.3 Mucoadhesion 79.3 Adhered 99.64 PEG-90M Carrageenan Viscosity Mucoadhesion Batch with SVF with SVF Fluorescein LP-86 (% w/w) 1.25 (% w/w) 2.00 52,840 (cP) 31,360 (mN/cm2 ) 71.9 80.9 99.50 (cP) (mN/cm2 ) (%) LP-87 1.25 1.50 38,000 27,070 67.3 70.8 99.52 LP-83 0.50 1.50 19,910 16,170 50.6 60.5 91.40 LP-88 LP-84 1.25 2.00 1.50 1.50 33,170 61,660 22,210 39,350 57.4 65.3 58.1 79.3 99.60 99.64 LP-89 LP-86 1.25 1.25 1.50 2.00 27,890 52,840 21,730 31,360 59.4 71.9 61.8 80.9 99.59 99.50 LP-87 LP-90 1.25 1.25 1.50 1.50 38,000 33,370 27,070 25,070 65.5 67.3 68.2 70.8 99.52 99.43 LP-88 1.25 1.50 33,170 22,210 57.4 58.1 99.60 LP-91 LP-89 1.25 1.25 1.50 1.50 32,260 27,890 23,250 21,730 53.6 59.4 63.3 61.8 99.60 99.59 LP-90 LP-92 1.25 1.78 1.50 1.14 33,370 29,750 25,070 26,180 47.9 65.5 50.3 68.2 99.43 97.33 LP-91 1.25 1.50 32,260 23,250 LP-93 0.72 1.85 20,100 16,790 48.3 53.6 51.9 63.3 99.60 98.10 LP-92 1.78 1.14 29,750 26,180 47.9 50.3 97.33 LP-94 LP-93 0.72 0.72 1.14 1.85 4720 20,100 10,490 16,790 34.9 48.3 38.8 51.9 90.28 98.10 LP-94 LP-95 0.72 1.78 1.14 1.85 4720 87,280 10,490 40,590 64.8 34.9 74.9 38.8 90.28 99.54 LP-95 1.78 1.85 87,280 40,590 64.8 74.9 99.54 LP-96 LP-96 1.25 1.25 1.00 1.00 10,000 10,000 13,690 13,690 38.4 38.4 47.8 47.8 92.42 92.42 As shown in Table 5, the addition of SVF reduced the formulation viscosity in most As shown in Table 5, the addition of SVF reduced the formulation viscosity in most cases, probably due to a dilution effect, and the effect of the crosslinking effect was now cases, probably due to a dilution effect, and the effect of the crosslinking effect was now observed. In contrast, the addition of SVF increased the mucoadhesion of the formulation, observed. In contrast, the addition of SVF increased the mucoadhesion of the formulation, although the dilution effect (seen with viscosity), mucoadhesion, and viscosity under although the dilution effect (seen with viscosity), mucoadhesion, and viscosity under simulated physiological conditions (in the presence of vaginal fluid) were not correlated. simulated physiological conditions (in the presence of vaginal fluid) were not correlated. First, the relationships among the different variables were studied by means of a First, the relationships among the different variables were studied by means of a corre- correlation matrix (Figure 3). This study also tried to reduce the variables in the lation matrix (Figure 3). This study also tried to reduce the variables in the experimental experimental design and remove highly correlated variables to avoid statistical redundancy. design and remove highly correlated variables to avoid statistical redundancy. Matrixplots Figure3.3.Matrix Figure plotsexplaining explainingthe therelationships relationshipsbetween betweenthe thedifferent differentvariables. variables. AsFigure As Figure33 shows, shows, mucoadhesion mucoadhesionand andmucoadhesion mucoadhesionwith withSVF SVFpresented presentedaalinear linear relationship. Therefore, because the formulation was applied to the vaginal epithelium, relationship. Therefore, because the formulation was applied to the vaginal epithelium, whichcontains which containsvaginal vaginalfluid, fluid,the the mucoadhesion mucoadhesiondata datawithout withoutSVF SVF were were discarded. discarded. The The same effect was observed with viscosity, but in this case, viscosity without SVF was Pharmaceutics 2023, 15, x FOR PEER REVIEW 11 of 21 Pharmaceutics 2023, 15, 1436 11 of 21 same effect was observed with viscosity, but in this case, viscosity without SVF was chosen because viscosity is related to the administration device (syringe or canula), and as previously seen, there was no relationship between the viscosity and mucoadhesion of the chosen because formulation. viscosity Finally, is related fluorescein to the did adhesion administration not presentdevice a linear(syringe or canula), and as relationship. previously Surface seen, there was response no relationship methodology was between applied the viscosity using and mucoadhesion Minitab software (withof the a formulation. Finally, fluorescein adhesion did not present a linear relationship. significance level of α = 0.05 for the level terms) to evaluate the effects of the tested Surface variables experimental response methodology (mucoadhesion waswith applied using SVF, Minitaband viscosity, software (with aadhesion). fluorescein significance level of α = 0.05 for the level terms) to evaluate the effects of the The fitted response surface model equations are shown in Table 6. Model selection was tested experimental variables based on the(mucoadhesion adjusted R2, whichwith SVF, viscosity, takes and fluorescein into account the differentadhesion). parameters The of fitted the response nested surfaceThe model equations areofshown in Table 6. Model selection was based on the adjusted 2 models. high R2 value the final equation indicates that it described most of theR , which takes variability into account the different parameters of the nested models. The high R value 2 experimental [55]. of the final equation indicates that it described most of the experimental variability [55]. Table 6. Response surface model equations for monitoring the effect of polymer concentration on Table the 6. Response properties surface model of mucoadhesive equations for monitoring the effect of polymer concentration on the gels. properties of mucoadhesive gels. Quadratic Polynomial Model Equation R2 Adj R2 Y1 =Quadratic Polynomial −0.8 + 14.40 X1 + 29.88 Model X2 Equation R2 0.702 Adj R2 0.642 Y2 =Y24,704 1 = − − 0.8 48,649 + 14.40 X X11 − + 23,052 29.88 X2X2 + 56,200 X1 X2 0.945 0.702 0.927 0.642 Y = Y3 = 28.85 2 24,704 − 48,649 X + 32.60 X1 + 56.06 1 − 23,052 X2 + 56,200 X2 − 6.652 X1 X2X22 − 7.48 X1X2 X12 − 13.21 0.945 0.987 0.927 0.978 Y3 = 28.85 + 32.60 X1: Concentration X1 + of (% w/w) 56.06 X2 − 6.652 PEG-90M; X12 − 13.21 X(% X2: concentration 22 − 7.48ofXcarrageenan. w/w) 1 X2 0.987 0.978 Y1: mucoadhesion X : Concentration with SVF (mN/cm 1 (% 2 w/w) of PEG-90M; X : concentration (% w/w) of ), Y2: Viscosity (cPs), Y3: fluorescein adhered (%). 2 carrageenan. Y1 : mucoadhesion with SVF 2 (mN/cm ), Y2 : Viscosity (cPs), Y3 : fluorescein adhered (%). The amounts of PEG-90M and carrageenan were significant with respect to The amounts mucoadhesion withof PEG-90M SVF andand (p = 0.017 carrageenan p = 0.003, were significant respectively). Bywith respect the increasing to mucoadhe- amount sion with SVF (p = 0.017 and p = 0.003, respectively). By increasing the of polymer, the mucoadhesion increased (Figure 4), as expected. As more polymeramount of polymer, was included in the formulation, there were more interactions with mucin, probably dueinto the mucoadhesion increased (Figure 4), as expected. As more polymer was included the formulation, the formation ofthere were more hydrogen interactions bonds. with mucin, The coefficient probably X2 is higher than due toshown X1, as the formation in Tableof hydrogen bonds. The coefficient X2 is higher than X1 , as shown in Table 6, so the effect of 6, so the effect of carrageenan on mucoadhesion is more important than that of PEG, carrageenan on mucoadhesion is more important than that of PEG, probably due to the probably due to the crosslinking effect with monovalent ions in SVF. crosslinking effect with monovalent ions in SVF. Figure 4. Plots of the main effects of PEG-90M and carrageenan on the mucoadhesive properties of Figure 4. Plots of the main effects of PEG-90M and carrageenan on the mucoadhesive properties of thethe gels. gels. Regarding viscosity, a similar effect was observed. The levels of PEG-90M and Regarding viscosity, a similar effect was observed. The levels of PEG-90M and carrageenan and the interaction between the two variables were significant (p = 0.000, carrageenan and the interaction between the two variables were significant (p = 0.000, p = p = 0.000, and p = 0.006, respectively). By increasing the polymer concentrations, the viscos- 0.000, and p = 0.006, respectively). By increasing the polymer concentrations, the viscosity ity increased (Figure 5A). In this case, the concentration of PEG-90M had a more significant increased (Figure 5A). In this case, the concentration of PEG-90M had a more significant effect on the viscosity values. Both PEG-90M and carrageenan can form hydrogen bonds; effect on the viscosity values. Both PEG-90M and carrageenan can form hydrogen bonds; so, the interaction between them influences viscosity. As Figure 5B shows, even though the so,interaction the interaction between between them influences the polymers viscosity. is significant, As Figure the viscosity 5B shows, even is proportional to thethough range of polymer concentrations used. Pharmaceutics 2023, 15, Pharmaceutics x FOR 2023, 15, PEER x FORREVIEW PEER REVIEW 12 of 21 12 of 21 Pharmaceutics 2023, 15, 1436 the interaction between the interaction the polymers between is significant, the polymers the viscosity is significant, is proportional the viscosity to theto12the is proportional of 21 rangerange of polymer concentrations of polymer used. used. concentrations FigureFigure 5. (A) Plots Figure 5.5.(A)of (A) the of Plots Plotsmaintheeffects ofthe main of PEG-90M maineffects effects and carrageenan ofofPEG-90M PEG-90M and on theon andcarrageenan carrageenan viscosity onthe (mPa×s) theviscosity viscosity of the (mPa×s) (mPa ×s)ofofthe the gels. (B) Viscosity gels. gels.(B) interaction (B)Viscosity Viscosity point. point. interaction interaction point. In theIn case Inthe theofcase fluorescein case adhesion, ofoffluorescein fluorescein a quadratic adhesion, adhesion, relationship a aquadratic quadratic was obtained relationship relationship was (Figure wasobtained obtained 6A), 6A), (Figure (Figure 6A), and a and significant andaasignificantinteraction significant between interaction interaction the studied between between variables the studied was found variables variables was (Figure was found found 6B). There (Figure (Figure 6B). 6B).There Thereis isa maximum is a maximum a maximum levellevel level ofof adherence adherence (expressed of adherence(expressed asasthethe (expressed fluorescein remaining as fluorescein in theinformulation) remaining the fluorescein remaining thein the that formulation) decreases that decreases formulation) at high that decreases concentrations at high of polymer. concentrations at high concentrations There of polymer. is an optimal ThereThere of polymer. concentration an optimalat is an isoptimal which mucoadhesion concentration at which concentration at which ismucoadhesion maximized. mucoadhesion This is maximized.observed is maximized.This effect This may observed be due effect observed mayto diffusion effect be theory; duebetodue may to diffusionwhen diffusion thetheory; theory; concentration when when of polymer the concentration the isof concentrationhigh (especially polymer of is high polymer carrageenan is (especiallybecause it has a larger carrageenan high (especially carrageenan because spatial it hasconformation because ita has larger spatial a larger than PEG-90M), conformation spatial the interpenetration than conformation PEG-90M), than PEG-90M), ofthe theinterpenetration polymericof the interpenetration chains theof with the mucin polymeric polymeric is hindered, chains with mucin chains withestablishing mucin weaker is hindered, is hindered, mucoadhesion. establishing weaker establishing mucoadhesion. weaker mucoadhesion. FigureFigure 6. (A) 6. Figure Plots of 6.(A)(A) the Plots main of Plots the effects main of the of effects PEG-90M effects main and carrageenan of PEG-90M and and of PEG-90M on fluorescein carrageenan adhesion. on fluorescein carrageenan (B) adhesion. on fluorescein (B) adhesion. Fluorescein adhesion Fluorescein interaction adhesion plot. interaction plot. (B) Fluorescein adhesion interaction plot. 3.3.Optimization Optimization 3.3. Optimization 3.3. ofofthe theMucoadhesive Mucoadhesive of the Mucoadhesive Gels Gels Gels Considering the high 2 Considering Considering the high the Rhigh 2 R Rvalues values 2 values from the from fromthethe obtained obtained model, obtained model, statistical model, statistical optimization optimization statistical of optimization of of the the formulation the formulation formulation waswas was carried carried out carried toout out to to obtain obtainmaximum maximum obtain maximum mucoadhesion mucoadhesion mucoadhesion andfluorescein fluorescein and fluorescein and adhesionadhesion effect effect adhesion effect and andminimal minimal and minimal viscosity viscosity (tofacilitate (to facilitate viscosity (to facilitate administration). administration). Toobtain To obtain administration). To obtain maximum maximum maximum mucoadhesion mucoadhesion and and minimum minimum viscosity viscosity (to (to facilitate facilitate syringeability), syringeability), mucoadhesion and minimum viscosity (to facilitate syringeability), within the design within within the the design design space space space limits, limits,limits, the the independent independent the independentvariables variables were were optimized. variables optimized. As As shown were optimized. shown in Table As shown in Table 7, 7, to obtain in Table to obtain a 7, to obtain aa mucoadhesion value of >70 mN/cm 2 , viscosity of 33,824 cP, and fluorescein adherence of mucoadhesion mucoadhesion value value of >70ofmN/cm >70 mN/cm 2 , viscosity 2 of 33,824 , viscosity cP, and of 33,824 cP,fluorescein adherence and fluorescein of adherence of 98.42%, 98.42%, ititisisnecessary it is necessary 98.42%, necessary to useto touse use0.866% 0.866% 0.866% PEG-90M PEG-90M PEG-90M and 2% and 2%carrageenan. carrageenan. carrageenan. and 2% Pharmaceutics 2023, 15, x FOR PEER REVIEW 13 of 21 Pharmaceutics 2023, 15, 1436 13 of 21 Table 7. Prediction of polymer concentrations for a formulation with maximum mucoadhesion and fluorescein adhesion and minimal viscosity. Table 7. Prediction of polymer concentrations for a formulation with maximum mucoadhesion and Fluorescein fluorescein adhesion and minimal Mucoadhesion Viscosity viscosity. Composite PEG-90M Carrageenan Adhesion Fit with SVF Fit Fit Desirability Fluorescein Mucoadhesion Composite 0.866 PEG-90M 2Carrageenan Adhesion 98.42 Fit 71.46 with SVF Fit 33,824 Viscosity Fit 0.76 Desirability 0.866 2 98.42 71.46 33,824 0.76 The quality by design study was carried out with the placebo. To check if the inclusion of the drug affected the dependent variables, three additional batches (placebo, 0.5%, and 1% TheDquality by design gel) were study produced wasthe with carried out with optimized the placebo. formula To PEG-90M (0.866% check if theand inclusion 2% carrageenan). The results, as well as model predictability and bias from the theoretical 0.5%, of the drug affected the dependent variables, three additional batches (placebo, and 1% predictions, areDreported gel) wereinproduced withobtained Table 8. All the optimized results formula were in (0.866% the 95% PEG-90M confidenceand 2% intervalcarrageenan). The was (CI), and the bias results, <10%as in well all as model predictability and bias from the theoretical cases. predictions, are reported in Table 8. All obtained results were in the 95% confidence interval (CI), Table 8. and the Results afterbias was <10% inof characterization allthe cases. final formulations with bias with respect to the theoretical value. Table 8. Results after characterization of the final formulations with bias with respect to the theoretical value. Theoretical Value 0.5% D Property Placebo Gel 1% D Gel Bias (%) (95% CI) Gel Mucoadhesion Theoretical Value Property Placebo Gel 0.5% D Gel 1% D Gel 71.5 (60.85; (95% CI)82.07) 72.9 68.2 69.4 1.86 Bias (%) (mN/cm2) Mucoadhesion Viscosity 71.5 (60.85; 82.07) (mN/cm2 ) 33,824 (22,885; 44,763) 72.9 68.2 69.4 1.86 29,835 30,400 32,780 8.30 (cP)Viscosity 33,824 (22,885; 44,763) 29,835 30,400 32,780 8.30 (cP) Fluorescein Fluorescein 98.42 98.42 (97.23; 99.62) (97.23; 99.62) 98.67 98.67 98.58 98.58 98.8198.81 0.27 0.27 adhered (%) (%) adhered 3.4. Rheological Properties 3.4. Rheological Properties As there As are small there differences are small in viscosity differences between in viscosity the the between optimized placebo optimized formula placebo formula and and the formulations with D, the following characterization was carried out the formulations with D, the following characterization was carried out with thewith the placebo. placebo. The rheological behavior was studied with and without SVF (Figure The rheological behavior was studied with and without SVF (Figure 7). 7). Figure 7. (A) Shear Figure 7. (A)stress Shearvs. shear stress vs.rate (red) shear and rate viscosity (red) vs. shear and viscosity vs.rate (blue) shear ratecurves (blue) of the placebo curves of the placebo formulation. (B) Shear formulation. (B)stress Shearvs. shear stress vs.rate (red) shear and rate viscosity (red) vs. shear and viscosity vs.rate (blue) shear ratecurves (blue) of the of the curves placeboplacebo formulation with SVF. formulation with SVF. As Figure 7 shows, As Figure 7 the formulation shows, presented the formulation thixotropy presented (area of 29.56 thixotropy (areaPa/s), and thus of 29.56 Pa/s), and the system thus therequired systemmore timemore required to recover time to its original recover internalinternal its original structure. The same structure. The same characteristics were observed characteristics for the placebo were observed for thegel with SVF placebo but with gel with SVFabutlower thixotropic with area a lower thixotropic area (17.95 (17.95 Pa/s). Pa/s). Additionally, Additionally, the viscosity the viscosity values under thevalues under the experimental experimental conditions conditions were 2212 ± 10.22were mPa×s and±1586 2212 2.75 × 10.22± mPa s and with mPa×s ± 2.75 1586 and mPa×SVF, without s with and withoutThe respectively. SVF, respectively. reduction in The reduction in viscosity and apparent thixotropy were probably viscosity and apparent thixotropy were probably due to the dilution effect of SVF. due to the dilution effect of SVF. Figure 7 shows the possible pseudoplastic behavior of the tested formulation (with and without SVF) Figure due 7toshows the possible the apparent pseudoplastic viscosity decrease behavior of the tested with an increasing formulation shear rate. To (with and without SVF) due to the apparent viscosity decrease with an increasing shear rate. To Pharmaceutics 2023, 15, 1436 14 of 21 confirm the rheological behavior, the experimental data were fit to different equations, as shown in Table 9. A lower Chi2 and a higher correlation factor represent the best fit. The model that best fits the experimental data is the Cross equation, which is usually employed to describe the rheological behavior of pseudoplastic materials with yield stress values [56]. Table 9. Rheological models with Chi2 and r values for the placebo gel and placebo gel with SVF. Placebo Gel Placebo Gel with SVF Rheological Model Chi2 r Chi2 r Newton 1.773 × 105 0.326 1.004 × 105 −0.137 Bingham 1.147 × 104 0.970 6127 0.968 Ostwald–de Waele 8.997 1.000 13.18 0.999 Herschel–Bulkley 7.020 1.000 12.900 0.999 Casson 9.220 0.988 1791 0.991 Cross 1.164 1.000 1.359 1.000 The Cross equation is a versatile equation that can estimate different models that converge into simpler models based on different assumptions. Table 10 shows the Cross equation parameters, where for shear-thinning materials, the value of n is between zero and one. In this case, the n values of the gels with and without SVF were lower than one, which confirms the pseudoplastic profile [56]. The placebo gel without SVF showed a higher viscosity at zero shear than with SVF due to the dilution effect of SVF. In contrast, the η ∞ in both cases were very similar, indicating that the gels converged toward similar viscosities at high flow rates with extremely small values compared to the viscosities . described previously (2212 and 1586 mPa×s). According to γ value, the placebo gels without SVF were less shear-thinning than they were with SVF. The rheological flow properties of polymers depend on the structural parameters; in this case, SVF crosslinks with carrageenan, affecting the rheological behavior of the gel. Table 10. Rheological model fitting of the placebo gels with and without SVF. Cross Equation Parameter Placebo Gel Placebo Gel with SVF η0 318.7 141.9 η∞ 0.1517 0.1600 . γ 0.0422 0.1235 N 0.6488 0.6853 In addition, a viscoelasticity study was carried out to evaluate the internal structure of the formulation and the effect of SVF on it. Figure 8 and Table 11 show the viscoelasticity parameters (storage modulus G0 and loss modulus G”). The formulations behaved as solid-like products (G0 > G”), and this value was con- firmed by tan(δ) < 1. The loss tangent (tan δ) is a measure of the energy lost to stored energy during cyclic deformation (tan δ = G”/G0 ). A value of tan δ < 1 indicates prevalent elastic behavior [57]. In this case, the same value of tan δ was obtained for both formulations, and it was <1. The values of both the storage and complex moduli increased in the presence of SVF. When SVF was added to the gel, G” remained essentially the same as that without SVF, but G0 increased. Carrageenan crosslinking with the monovalent ions of SVF forms a more rigid structure, and therefore, the storage modulus is higher than that without SVF. Despite the increasing values with SVF, the proportion was maintained, and therefore, the value of δ was the same in both cases. ɣ̇ 0.0422 0.1235 N 0.6488 0.6853 In addition, a viscoelasticity study was carried out to evaluate the internal structure Pharmaceutics 2023, 15, 1436 15 of 21 of the formulation and the effect of SVF on it. Figure 8 and Table 11 show the viscoelasticity parameters (storage modulus G′ and loss modulus G″). Figure 8. Storage Figure 8. Storage modulus, modulus, loss loss modulus, modulus, andand phase phase angle angle during during the the sweep sweep stress stress test test on the on the mucoadhesive gel 24 h after preparation at a frequency of 1 s−1. − (A) 1 Placebo Gel. (B) Placebo Gel with mucoadhesive gel 24 h after preparation at a frequency of 1 s . (A) Placebo Gel. (B) Placebo Gel SVF. with SVF. Table 11. Viscoelasticity parameters of the Placebo Gel and Placebo Gel with SVF (mean ± standard deviation (SD)). Parameter Formulation Mean ± SD Placebo gel 129.92 ± 25.20 Pa G0 Placebo gel with SVF 149.66 ± 25.33 Pa Placebo gel 57.90 ± 4.60 Pa G” Placebo gel with SVF 63.67 ± 4.01 Pa Placebo gel 22,709 ± 3657 mPa η* Placebo gel with SVF 25,947 ± 3649 mPa Placebo gel 24.20 ± 4.46◦ δ Placebo gel with SVF 23.67 ± 5.07◦ Placebo gel 0.45 ± 0.11 tan(δ) Placebo gel with SVF 0.44 ± 0.08 Under an external force, molecular chain orientation is caused by internal friction. When the molecular weight is lower than a certain value, there is a crossover point between the G” and G” curves, which means that there is a balance between the states of internal friction and disorientation. With an increase in molecular weight, the crossover point moves to a lower frequency because of restricted disorientation. This lower frequency allows sufficient time for molecular orientation. Therefore, the higher molecular weight molecules need more time for molecular orientation. In summary, the crossover point data demonstrate that the molecular weight decreases with an increase in frequency and polydispersity increases with a decrease in frequency [58]. In this case, as Figure 8 shows, the crossing point is higher without SVF (approximately τ = 140 Pa) than with SVF (approx- imately 75 Pa). With SVF, sodium ions intercalated into the carrageenan chains, increasing their rigidity and producing a more restricted orientation, so the value of τ was lower. 3.5. Syringeability of the Formulations Syringeability (the force required to extrude a formulation though a syringe) is another important parameter for practical administration with a syringe [50]. In this case, the time needed to empty the syringe was determined with the placebo formulation, which was 3.04 ± 0.0058 s. The time and force needed to empty the syringe were acceptable and would allow for the easy application of the product to the vaginal area. Pharmaceutics 2023, 15, 1436 16 of 21 3.6. In Vitro Release of Disulfiram from the Gel The release of disulfiram from the 0.5% and 1% mucoadhesive gels is depicted in Pharmaceutics 2023, 15, x FOR PEER REVIEW 16 of 21 Figure 9. After 6 h, 80% of D was released from the 0.5% gel, while more than 90% was released from the 1% gel. 100 80 % D Released 60 40 1% Disulfiram 20 0.5% Disulfiram 0 0 0.5 1 1.5 2 3 4 5 6 7 Time (h) Figure Figure 9.9. Release Releaseofofdisulfiram disulfiram from from thethe 0.5% 0.5% gel from gel and and from the 1%the gel.1% gel. of Mean Mean of the quantities the quantities released released of D expressed as a percentage respect total amount seeded from the 0.5% gel and from the of D expressed as a percentage respect total amount seeded from the 0.5% gel and from the 1% gel 1% gel after 6 h in vitro. after 6 h in vitro. After After fitting fitting the the mean mean release release data data to to the the different different mathematical mathematical models models (Table (Table 12), 12), it it was observed that the first-order model provided the best fit (AIC 30.16, adj was observed that the first-order model provided the best fit (AIC 30.16, adj R 99.78%) ofR 2 99.78%) of 2 the the experimental experimental data in the data in the case case of of the the 0.5% 0.5% (w/w) (w/w) gel gel formulation, formulation, whereas whereas the the Weibull Weibull model best fit the data from the 1% (w/w) D gel (AIC 55.19, adj model best fit the data from the 1% (w/w) D gel (AIC 55.19, adj R 97.33%). R 2297.33%). Once the mean release kinetic behaviors were determined, the means and standard deviations of the individual Table 12. Individual models andrelease data forforboth parameters the gels were release calculated, of D asand from the 0.5 reported in Table 12. 1% gels. Table Formulation Model 12. Individual models and parameters Parameters for the release of D from the 0.5 and 1%Value gels. K1 (h−1 ) 0.95 ± 0.48 0.5% D gel Formulation First-Order Model Parameters Value Fmax (%) 85.79 ± 2.13 K1 (h−1) 0.95 ± 0.48 0.5% D gel First-Order α 0.88 ± 0.43 1% D gel Weibull Fmax (%)β 85.79 0.73±±2.13 0.23 αFmax (%) 0.88 ± 0.43 93.42 ± 3.21 1% D gel Weibull β 0.73 ± 0.23 F max (%) 93.42 ± 3.21 Once the mean release kinetic behaviors were determined, the means and standard deviations of the individual release data for both gels were calculated, as reported in TableThe 12. first-order model confirmed that the D release mechanism from the 0.5% (w/w) gel was Thediffusion, first-orderaccording to Fick’sthat model confirmed Lawthe[33,51]. However, D release the Weibull mechanism from the model was a 0.5% (w/w) nonmechanistic gel was diffusion, equation. Papadopoulou according to Fick’s Lawet al. [51] determined [33,51]. However, athe relationship between Weibull model wasthea shape parameterequation. nonmechanistic β and thePapadopoulou release mechanism. In the et al. [51] case of thea1% determined (w/w) D gel, relationship the value between the of β was 0.73, which corresponds to the Fick diffusion release mechanism (β shape parameter β and the release mechanism. In the case of the 1% (w/w) D gel, the value < 0.75). The release of β was from 0.73,both formulations, which corresponds despite to the being explained Fick diffusion by different release mechanismmodels, (β