Albendazole (ALB) was granted by Alexandria Co. (Cairo, Egypt). L-α phosphatidylcholine (PC), Pluronic L121 and Pluronic P188 were acquired from Sigma-Aldrich (St. Loius, MO, United States). Ceramide III and ceramide IIIB were purchased from Evonik Co. (Essen, Germany). Stearyl amine (SA) was acquired from Fluka Chemical Co. (Buchs, Switzerland). Chloroform and Methanol were procured from El-Nasr Pharmaceutical Chemicals Co. (Cairo, Egypt).

Elastic cerosomes loaded Albendazole (EC-ALB) were developed through the use of thin film hydration technique (Ahmed et al., 2024) Briefly, ALB (2 mg), PC (100 mg) together with variable amounts of Pluronics (L121 or P188) and ceramides (III or IIIB) were weighed and dissolved in 10 mL chloroform in round-bottom flask (250 mL). The organic solvent was evaporated under vacuum using a rotatory evaporator (Rotavapor, Heidolph VV 2000, Burladingen, Germany) that rotated for 30 min at 90 rpm and 60 °C. After that, 10 mL of ultra-pure distilled water was used to hydrate the resulting clear film for 45 min at 60 °C under normal pressure. Finally, to avoid aggregation, the formed elastic cerosomesʼ dispersions were subjected to varying durations of sonication (10, 20 or 30 min) at 25 °C using ultra-sound bath (Model SH 150-41; United States) (Joseph et al., 2020).

To evaluate the individual and combined effects of the formulation variables on the characteristics of the nanosystems, EC-ALB formulations were prepared using a D-optimal design generated by Design-Expert® software (version 7, Stat-Ease, Inc., Minneapolis, MN, USA). A total of 19 formulations were prepared according to the experimental design. The independent variables were SAA amount (X1), sonication time (X2), ceramide type (X3), and SAA type (X4), while the dependent variables included entrapment efficiency (EE%; Y1), particle size (PS; Y2), and polydispersity index (PDI; Y3). The best-fitting mathematical model was selected to obtain the highest predictive R² value (Eldeeb et al., 2019). The ranges of the studied factors and the constraints applied to the responses are summarized in Table 1.

Predicting the optimal conditions for preparing the best formula which may not be included in the actual design is one of the most important advantages of D-optimal design (Eldeeb et al., 2019). Maximizing EE% and minimizing PS and PDI were the selected criteria for preparing the optimal EC-ALB formulation (Table 1). By applying desirability function, the optimal formulation was suggested, fabricated, in-vitro characterized, and the results were compared with software predicted responses. The predicted optimal EC-ALB was also used for more investigations.

SA-EC-ALB formulation was prepared utilizing the same ingredients as the optimal EC- ALB along with varying amounts of SA (5, 10, 15 mg). The formula was fabricated utilizing the same thin film hydration technique applied to fabricate EC- ALB, and SA was also dissolved in the chloroform. The prepared SA-EC-ALB formulations were evaluated for EE%, PS, and ZP. The results were analyzed statistically using one-way ANOVA in SPSS software (version 17.0, SPSS Inc., Chicago, IL), followed by Tukey's HSD post-hoc test. A p-value of less than 0.05 was considered statistically significant. The optimal SA concentration was selected for the preparation of SA-EC-ALB and subsequently subjected to further in-vitro and in-vivo studies.

The total amount of drug released after 6 h (Q6h) was calculated utilizing the United States Pharmacopeia (USP) dissolution apparatus (Pharma Test, Hainburg, Germany) for 6 h at 37 °C. An amount of 2 mL samples from SA-EC-ALB, EC-ALB and ALB aqueous suspension (each containing 200 µg ALB) were enclosed in cylindrical plastic tube (3.14 cm² permeation area). One end of these tubes with one end securely sealed with a cellulose membrane (Spectra/Por^® 12,000–14,000 molecular weight cutoff, Spectrum Laboratories Inc., USA) while the other one connected to the USP dissolution apparatus shafts instead of the baskets. The formulations were submerged in 50 mL of buffer solution with a pH of 7.4 (to maintain sink condition). At 1, 2, 3, 4, 5, and 6 h, aliquots were withdrawn (3 mL), and analyzed for the presence of ALB using UV spectrophotometer at λ_max 295 nm. In order to preserve sink conditions, an equivalent volume of fresh release medium was added immediately. The experiment was repeated three times. A schematic illustration of the release setup is provided in the Supplementary Material (Supplementary Figure S1) for better visualization. The following equation was used to determine the amount of ALB released. %   R e l e a s e =   A m o u n t   o f   A L B   r e l e a s e d   a t   a t i m e   t T o t a l   a m o u n t   o f   A L B × 100

The SA-EC-ALB was morphologically examined by TEM (Joel JEM 1230, Tokyo, Japan). On a carbon-coated copper grid, one drop of the SA-EC-ALB dispersion was deposited as a thin layer following staining with 2% w/v phosphotungstic acid. After drying, The sample was analyzed using a TEM at 80 KV (Ahmed et al., 2025).

ALB and the optimal SA- EC- ALB were subjected to thermal investigation utilizing differential scanning calorimeter (Shimadzu DSC-60, Shimadzu Corp., Kyoto, Japan) that was calibrated with 99.9% purified indium. Each sample (about 5 mg) was placed in a conventional aluminium pan and heated between 10 °C and 300 °C at a scanning rate of 10 °C/min and a continuous nitrogen purging rate of 25 mL/min (Elmahboub et al., 2024).

The antitumor activity of SA-EC-ALB was assessed in a solid Ehrlich tumor model using adult male Swiss albino mice (n = 6) with an average weight of 22 ± 2 g. The group size of six mice per treatment group was determined based on ethical considerations aligned with the 3Rs (Replacement, Reduction, Refinement) principle, aiming to minimize animal use while ensuring scientific validity. Mice were kept in polypropylene cages in the animal house of the Faculty of Pharmacy, Beni-Suef University, in a light period of 12 h/day and at a temperature of 25 °C ± 2 °C and humidity of 50% ± 5%. The mice were randomly distributed into four groups. The study protocols followed the recommendations of the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH Publication No. 8023, revised 1978). No consent from animal owners was required, as the study involved laboratory mice (Council, 2011). The group (1) was a normal control group and injected intraperitoneally (I.P) with vehicle solution. Groups (2-4) were injected with Ehrlich carcinoma cell line (0.2 mL/2–2.5 × 10⁶ cells/mouse) subcutaneously into the right thigh of each mouse to establish tumor in mice (Salem et al., 2018). The solid Ehrlich tumor in each mouse was developed after 7 days. Group 2 served as the positive control group while, groups 3 and 4 received ALB aqueous suspension and SA- EC- ALB (100 µg/mL) (Kang et al., 2015), respectively, in a dose of 2 mg/kg (440 µL), I.P/3 days/week for 3 weeks. The tumor volume (TV) was determined using digital caliber at the end of each week and after the last day of treatment administration, all mice were anesthetized with thiopental (30 mg/kg) and then sacrificed with cervical dislocation, and the right thighs were isolated for histopathological study. Tumor masses were isolated and fixed in 10% formol saline. Tumor tissue sections (5 μm thick) from the four groups were stained with hematoxylin and eosin (H&E) and examined histopathologically using a light microscope (Leica Microsystems GmbH, Wetzlar, Germany) (Bancroft and Gamble, 2008). GraphPad Prism software (Version 9.2.0.332 – San Diego, United States) was used for data analysis. Statistical significance is calculated using one-way ANOVA. Additionally, differences between the two groups are made using Tukey’s multiple comparisons test.

EE% of elastic cerosomes loaded Albendazole (EC-ALB) fluctuated from 66.52% ± 2.37% to 94.00% ± 3.58% (Table 3 and Figure 1). The resulting model equation in terms of coded factor was as follows: EE % = + 75.72 + 0.82 * X 1 + 1.04 * X 2 + 0.017 * X 3 − 5.13 * X 4 + 0.98 * X 1 * X 2 − 2.77 * X 1 * X 3 − 2.80 * X 1 * X 4 − 0.49 * X 2 * X 3 + 3.44 * X 2 * X 4 + 2.90 * X 3 * X 4 + 12.41 * X 1 2 − 8.81 *   X 2 2 .

Clear interaction had been observed between SAA amount (X₁) and SAA type (X₄) (P < 0.0001).

PS values of all EC-ALB were in nano range (273.9 ± 7.90 to 740 ± 16.66 nm) as shown in Table 3 and Figure 2. The resulting model equation in terms of coded factor was as follows: PS  nm = + 454.86 + 12.82 *   X 1 − 9.87 *   X 2 + 9.06 *   X 3 − 162.47 *   X 4 − 14.75 *   X 1   *   X 2 − 18.35 *   X 1   *   X 3 + 11.00   *   X 1 * X 4 + 25.89 * X 2 * X 3 − 7.23 * X 2 * X 4 − 6.50 * X 3 * X 4 + 51.11 * X 1 2 + 7.32 * X 2 2 .

SAA amount (X₁), ceramide type (X₃) and SAA type (X₄) significantly affected the PS of EC-ALB (P = 0.0143 for SAA amount, 0.0358 for ceramide type and ˂0.0001 for SAA type).

PDI of EC-ALB ranged from 0.40 ± 0.12 to 0.81 ± 0.12 (Table 3). Some of EC-ALB were highly polydisperse, e.g., F3 (0.77 ± 0.06), F6 (0.81 ± 0.12) and F8 (0.76 ± 0.02) (Figure 3). The resulting model equation in terms of coded factor was as follows: PDI = + 0.45 − 0.041 * X 1 − 0.070 * X 2 − 0.035 * X 3 − 0.053 * X 4 − 0.051 * X 1 * X 2 + 3.688 * 10 − 3 * X 1 * X 3 + 0.096 * X 1 * X 4 + 0.031 * X 2 * X 3 − 0.18 * X 2 * X 4 − 0.016 * X 3 * X 4 − 0.055 * X 1 2 + 0.3 * X 2 2 + 0.18 * X 1 * X 2 * X 3 − 0.16 * X 1 * X 3 * X 4 + 0.035 * X 2 * X 3 * X 4 .

All four variables significantly affected the PDI of EC-ALB (P = 0.010 for SAA amount, 0.0003 for sonication time, 0.0016 for ceramide type and 0.0005 for SAA type).

Considering ZP, All the prepared formulation are negatively charged and ranged from −11.60 ± 0.01 to −26.30 ± 0.70 mV.

The total desirability of the optimal EC-ALB was 0.862 and it was suggested to be prepared using ceramide III (15 mg) and 50 mg Pluronic L121 with sonication time of 19.41 min. The formula was prepared and evaluated. Table 4 showed the small residual between the observed and predicted responses of the optimal EC-ALB which confirmed the rationality of the optimization process. Thus, optimal EC-ALB formulation was guaranteed to be used as a nucleus for preparing (SA-EC-ALB).

SA–EC–ALB formulations were developed with respect to EE% PS and ZP and examined to determine how positive charge inducers impact the physicochemical attributes of EC–ALB. As shown in Table 5, SA-EC-ALB prepared using 5 mg SA showed significantly higher EE% compared to those prepared using higher amounts of SA. Considering PS, SA-EC-ALB prepared using 10 and 15 mg were significantly larger than 5 mg SA-EC-ALB. Cerosomes prepared using 5 mg SAA showed ZP values of 38.2 ± 1.65 mV.

The in-vitro release profile of ALB from SA-EC-ALB and EC-ALB compared to ALB aqueous suspension is shown in Figure 4A. The release rate of ALB from SA–EC–ALB and EC–ALB formulations was significantly higher compared to the drug suspension. Furthermore, the extent of ALB released from SA-EC-ALB and EC-ALB after 6 h was significantly higher than ALB aqueous suspension.

As shown in Figure 4B, SA-EC-ALB displayed elongated fiber like vesicles. Spherical vesicles occasionally appeared together with tubular ones.

Figure 4C represents the thermograms of pure ALB and optimal ALB elastic cerosomes. The DSC scan of pure ALB showed sharp endothermic peak at 204 °C corresponding to its melting point (Trandafirescu et al., 2019). This sharp peak disappeared in the thermogram of the optimal ALB elastic cerosomes.

ZP of mucin suspension was found to be −8.47 ± 0.13 mV and exhibited changes upon mixing with mucoadhesive SA-EC-ALB. Mucoadhesive SA-EC-ALB showed a shift from negative mucin values −8.47 ± 0.13 mV to positive 15.00 ± 1.00 mV of mucin ZP value.

D-optimal design was selected in this study as its criterion depends on maximizing the determinant value of the information matrix to minimize the overall variance associated with estimating the model coefficients (Eldeeb et al., 2019). Furthermore; in D-optimal design, the software selected factorial points, axial check points, center of edges and overall center points (Shamma et al., 2012). These points’ diversity helps in providing a complete knowledge of responses with small number of experimental runs. Hence, a 19 run, four-factor, two-level D-optimal design was used for characterization and optimization of EC-ALB using Design-Expert^® software. The measured responses of the prepared 19 EC-ALB are shown in Table 3.

Selection of the model that best fits the data is the first step in statistical analysis. Therefore, sequential model comparison was performed for each response and showed that quadratic model is significant for responses Y₁ and Y₂ which indicates the fitting of the model to the measured response. For Y₃, cubic model was aliased since more experiments are needed to be performed to estimate the model term accurately (Al Hagbani et al., 2018). The predicted R² and adjusted R² values are preferred to be in a reasonable agreement to confirm the fitting of the selected model to the data (Aziz et al., 2022a). Referring to design analysis results (Supplementary Table S1), it is clearly noticed that predicted R² values are in good harmony with adjusted R² values in all responses except PDI. Therefore, the variability in PDI might be attributed to noise (Al Hagbani et al., 2018) Furthermore, signal to noise ratio is measured by calculating adequate precision to make sure that the selected model can be used for design space evaluation (Aziz et al., 2022a). Adequate precision with a ratio greater than 4 is preferable and it was observed in all responses as shown in Supplementary Table S1.

Elastic cerosomes loaded Albendazole (EC-ALB) prepared using lower concentration of Pluronic L121 (SAA amount, X₁) showed higher EE% compared to those prepared using higher SAA concentration. On the other hand, Pluronic P188 based vesicles showed increased EE% by increasing Pluronic P188 amount. This could be related to the higher hydrophilicity of Pluronic P188 (HLB = 29) compared to L121 (HLB = 1). Kadam Y. et al., reported that micelles prepared using hydrophobic polymers (higher % of PPO units) showed decrease in hydrodynamic size at higher concentrations with resultant decrease in the amount of solubilized drug (Kadam et al., 2009). In contrast, the hydrophilic Pluronics (e.g., P188) oppositely showed increase in micellar size with solubilized drug at higher concentration.

With respect to sonication time (X₂), ANOVA results showed that EE% of EC-ALB prepared using sonication time 10 and 30 min was significantly lower than those prepared using intermediate sonication periods (15 and 20 min). In another words, it is noticed that EE% gradually increased by increasing the sonication time from 10 to 20 min. This might be attributed to that by increasing the sonication time in a suitable range, the number of vesicles per unit volume increased with consequent augmentation of the hydrophobic ambiance of the lipid bilayer to incorporate more hydrophobic drug (ALB) with resultant increase in EE%. These findings came in agreement with that presented by He Y. et al., who showed that the highest EE% of Ibuprofen liposomes achieved at the intermediate sonication time (20 min) (He et al., 2019). On the other hand, EE% significantly decreased by further increase in sonication time (X₄) (30 min) as the too long sonication time might cause phospholipid degradation and disruption of the EC-ALB structures with resultant drug leakage and low EE% (Gala et al., 2015).

With respect to SAA amount (X₁), it was clearly noticed that the mean diameter of the prepared EC-ALB significantly decreased by initial increasing in SAA amount from 50 to 62.5 mg. This could be explained by the ability of SAA at higher levels to decrease the interfacial tension of the prepared nanosystem and increase the vesicular curvature with resultant decrease in PS (Albash et al., 2021). Furthermore; SAA, when used in sufficient amount, can increase the system stability by forming a steric barrier on its surface that prevents particles’ aggregation (Yousry et al., 2016) In contrast, PS significantly increased by further increasing in SAA amount (68.75 and 75 mg). This can be attributed to depletion-flocculation mechanism of SAA. In addition, SAAs beyond certain extent, form micelles in continuous phase rather than orientation on particles’ surface with resultant diminution of continuous phase between particles which consequently caused their aggregation and PS increase (Hasani et al., 2015).

With respect to ceramide type (X₃); ceramide III-derived cerosomes were significantly larger in diameter than those prepared using ceramide IIIB. The major differences between both ceramides is that ceramide IIIB has one unsaturated bond in its fatty acid chain and with lower molecular weight (581.95) compared to ceramide III (583.98). Unsaturated ceramide molecules with lower molecular weight have more mobility to migrate between membrane leaflets compared to saturated larger ones which accumulated in the leaflets with consequent alteration in membrane curvature and increasing PS (Hasani et al., 2015; Albash et al., 2021).

Considering SAA type (X₄), ANOVA results also showed that EC-ALB prepared using Pluronic P188 were significantly larger than those prepared using Pluronic L121. This could be explained based on the higher hydrophilicity of Pluronic P188 compared to L121 which consequently increases water uptake and causes PS enlargement (Aziz et al., 2022a).

PDI values of the prepared nano-formulation reflect the width of their size distribution which fluctuated from 0 (homogenous dispersion) to 1 (highly polydisperse particles with wide PS range) (Younes et al., 2018a). The results indicated that all EC-ALB formulations are homogenous with adequate PS distribution (Elsayed et al., 2020). The high PDI values of F3, F6 and F8 might be attributed to the irregularity of EC-ALB vesicular shape (Albash et al., 2021). Therefore, these systems were excluded in the optimization step.

With respect to SAA amount (X₁), EC-ALB prepared using the highest amount of SAA (75 mg) were homogenously dispersed compared to those prepared using lower concentrations of SAA. This was attributed to the ability of Pluronics in lipid phase to induce steric stabilization which hinders particles’ aggregation (Younes et al., 2018a).

Considering sonication time (X₂), ANOVA results showed that EC-ALB prepared using intermediate sonication periods (15 and 20 min) showed relatively more uniform particle distribution than those prepared using extreme sonication periods (10 and 30 min). As previously mentioned under EE% section, utilizing intermediate sonication periods increased the internalization of ALB within the formed EC-ALB and therefore, less free ALB would be available to form aggregates (Younes et al., 2018b).

For ceramide type (X₃), EC-ALB prepared using ceramide III were relatively more homogenously dispersed than ceramide III-B -based EC-ALB. This could be attributed to the relatively smaller mean diameter of ceramide III-B-based vesicles. As a result, more vesicles per unit volume will be formed with consequent reduction in the external phase water volume level. Hence, less nucleation sites would be available for drug solubilization with resultant increased drug precipitation and increased PDI values (Nour et al., 2016).

Considering SAA type (X₄), PDI values of EC-ALB prepared using Pluronic L121 were smaller than Pluronic P188-based EC-ALB. This could be explained based on the difference in their Mwt. Pluronic P188 showed higher Mwt (8,400 Da) compared to Pluronic L121 (4,400 Da) which caused less kinetically restricted entrapment of the drug with consequent increase in PDI (Younes et al., 2018b).

The higher EE% of SA-EC-ALB prepared using 5 mg SA might be attributed to that, increasing positive charge within lipid bilayer results in changing the lateral packing of the vesicular bilayer and consequently decreasing EE% (Villasmil-Sánchez et al., 2010). Furthermore, it displayed smaller PS compared to other formulation. This can be interpreted by that increasing the amount of charge inducer increased the repulsive force between the vesicular adjacent bilayers with resultant increase in the spacing between them and formation of relatively large vesicles (Naguib and Makhlouf, 2021). With respect to vesicular charge (ZP), it was shown that ZP values ˃ 20 mV indicate an acceptable colloidal stability due to the presence of sufficient electrical charge on the nano-vesicular surface that could induce electrostatic repulsion between particles and prevent their aggregation. Furthermore, SA-EC-ALB prepared using 5 mg SA showed acceptable ZP values. Therefore, it was selected for further characterization.

The higher rate and extent of ALB release from SA-EC-ALB compared to ALB aqueous suspension might be related to the solubilization of ALB by the phospholipid based-vesicular carrier (van Hoogevest et al., 2011). The phospholipid’s surface-active properties are widely recognized for enhancing the solubility of the poorly water-soluble drug enclosed within (Hills, 2002) On the other hand, the presence of SAA (Pluronic), might aid in the micelles formation within the bilayer which consequently may increase the membrane permeability and augments drug release (Albash et al., 2019).

TEM imaging is valuable for describing the shape of the prepared system as well as for confirming the results of Malvern zetasizer (Abd-Elsalam et al., 2018b). Elongated fibers appeared due to ceramide partitioning into PC bilayer which consequently caused flattening of the vesicular bilayer curvature upon preparation. The occasional appearance of spherical vesicles together with tubular ones could be related to the un-even distribution of ceramide III in the vesicular bilayers which consequently caused the appearance of flattened ceramide rich domains together with spherical ceramide poor ones (Abdelgawad et al., 2017; Albash et al., 2021).

DSC is a valuable tool for investigating the potential phase changes and the drug-excipient interaction upon ALB encapsulation. The disappearance of the sharp endothermic peak of ALB in the thermogram of the optimal ALB- elastic cerosomes confirmed ALB transformation from crystalline to amorphous form due to its complete dispersion in the prepared nanosystem (Aziz et al., 2022a).

The assemblies arising from mucin and SA-EC-ALB yielded a positive value, this suggests that the positive charge of SA–EC–ALB bound to the mucin surface neutralized its negative charge. The results further indicated that an adhesive component with mucoadhesive characteristics can alter the surface properties of mucin (Teaima et al., 2022).