Enhanced brain targeting efficiency using 5-FU (fluorouracil) lipid–drug conjugated nanoparticles in brain cancer therapy

Abstract

The present investigation was aimed to synthesize, optimize, and characterize lipid/drug conjugate nanoparticles for delivering 5-fluorouracil (5-FU) to treat brain cancer. The Box–Behnken design was used to optimize the formulation, evaluate the particle size, entrapment efficiency, morphology, in vitro drug release study, and stability profiles. The in vitro performance was executed using cell line studies. The in vivo performance was carried out for pharmacokinetic studies, sterility test, biodistribution studies, and distribution lipid–drug conjugated (LDC) nanoparticles in the brain. Particle size, zeta potential, entrapment efficiency, and morphology of the optimized formulation demonstrated desirable results. In vitro release pattern showed initial fast release, followed by sustained release up to 48 h. Cytotoxic effects of blank stearic acid nanoparticles, LDC nanoparticles, and 5-FU solution on human glioma cell lines U373 MG cell showed more cytotoxicity by LDC-NPs compared to others. The values reported for LDC (AUC = 19.37 ± 0.09 µg/mL h and VD 2.4 ± 0.24 mL) and pure drug (AUC = 8.37 ± 0.04 µg/mL h and VD = 5.24 ± 0.29 mL) indicate higher concentrations of LDC in systemic circulation, while pure 5-FU was found to be largely available in tissue rather than blood circulation. The t_1/2 for LDC represents an approximate rise by ninefold, while MRT (12.10 ± 0.44 h) denotes 12-fold rise than pure 5-FU indicating the prolonged circulation of LDC. Free 5-FU concentration in the brain was maximum (5.24 ± 0.01 μg/g) after 3 h, while for the optimized formulation of LDC it was twofold greater estimated as 11.52 ± 0.32 μg/g. In conclusion, the efficiency of 5-FU to treat the brain is increased when it is formulated with LDC nanoparticles.

Introduction

5-FU is an analog of the naturally occurring pyrimidine uracil and is metabolized through the same metabolic pathways as uracil. 5-FU is an antimetabolite anticancer agent used in brain cancer treatment. In vivo phosphorylation of 5-FU enhances its binding to thymidylate synthetase and inhibits the synthesis of DNA and RNA. The cytotoxic action of 5-FU is significant in multiple types of cancers including brain cancer. The major limitation associated with the use of 5-FU in brain cancer is its higher water solubility and less bioavailability by oral routes (Longley et al. 2003). The oral route is considered to be the most common and convenient when it comes to patient compliance (Carneiro et al. 2012; Agrawal et al. 2017). Due to the high polarity of 5-FU, it may not cross the blood–brain barrier (BBB), hence brain cancer is not treated effectively through the oral route. However, very little bioavailability is reported by the oral route due to the hydrophilicity, shorter half-life, and first-pass metabolism of the drug candidate (Shelke et al. 2016a; Yang et al. 2012). Moreover, 5-FU has side effects like stomatitis and ulceration if it is released in the stomach (Ahn et al. 2012). The half-life is reported to be around f 12.3 ± 7.3 min (Lin et al. 2016). Faster absorption is favorable to maintain the therapeutic level, but earlier elimination (within 3 h) is the leading drawback encountered for 5-FU. In the treatment of brain cancer, 5-FU needs to be infused for a longer duration of time for effective therapeutics. The major drawback reported with oral route and infusion is the appearance of metallic taste, neutropenia, alopecia, stomatitis, GI ulceration, anemia, etc. (Kratz 2008). To evade the drug delivery by injection or continuous infusion, controlled parenteral delivery is the dosage form of choice, whereby the side effects can be minimized or nullified with reduced dosing frequency (Arya et al. 2011).

Nanotechnology and nanocarriers are a potential area of research due to their extensive applications in the treatment of multiple diseases with targeted therapy. There is a significant need to develop sustained release brain targeting dosage form for brain cancers. Various nanocarriers, viz., carbon nanotubes, dendrimers, micelles, quantum dots, and gold nanoparticles have been reported to treat brain cancer (Dinda et al. 2012). Colloidal drug delivery systems such as LDC nanoparticles, solid lipid nanoparticles (SLN), and nanostructured lipid carrier (NLC) are very good examples of lipid-based nanoparticles where drugs were successfully incorporated preventing physiological degradation (Meng et al. 2011; Shinde et al. 2020). Hydrophilic drugs can also be incorporated into liposome and can be used for cancer treatment, but due to its stability problems, its use as a carrier has limitations. Hence, LDC can be the choice to overcome the drawbacks of the liposome (Sun et al. 2006). LDC nanotechnology is an extensively growing field with numerous potential applicability in research and development. Fabrication of the LDC results in unique particle size characteristics that would be effectively applied in therapeutics. The ability to incorporate drugs into nanocarriers offers a new prototype in drug delivery that could be used for drug targeting (Yang et al. 2014). Hence, LDC holds great promise for reaching the goal of controlled and site-specific drug delivery. A major problem addressed in the formulation of SLN is the low capacity to load hydrophilic drugs due to partitioning effects during the production process, whereby only a potent dose of hydrophilic drugs may suitably incorporate in the solid lipid matrix (Wang et al. 2004). Hence, LDC nanoparticles are developed to overcome the limitation by improving the drug loading efficiency to 35%. LDC has the benefits of encapsulating both lipophilic and hydrophilic drugs, excellent biocompatibility, formulation into control and targeted drug release, no bio-toxicity of the carriers, enhance the bioavailability of encapsulated bioactive compounds, better control over release kinetics of encapsulated compounds, and improved stability of formulations (Akrami et al. 2016).

The hydrophilic drug is transformed to a more lipophilic, insoluble molecule by conjugation with a lipidic compound. Principally, conjugation can be carried out by covalent linkage or simply by the formation of a salt with a fatty acid type of lipid. The LDC nanoparticles demonstrate very poor solubility in water, exhibiting an approximate 50–100 °C melting range which can be transformed into nanoparticles using a high-pressure homogenization method (Balaji and Gothandam 2016). Considering the molecular weight of the two fractions in the conjugate molecule, i.e., of the drug itself and the lipid part, a drug loading of approximately 30–50% is achievable (about 33% formulated as diminazene aceturate–acid conjugate with palmitic acid/stearic acid) (Banerjee and Ravishankar 2017). LDC nanoparticles can be made from the conjugated drug only or solid lipids can be additionally admixed to form a mixed matrix of LDC and lipid. These lipid-based nanoparticles represent an alternative drug carrier system to traditional colloidal carriers such as liposomes, nanoemulsions, polymeric nanoparticles, and polymeric microparticles, and also possess advantages of controlled drug release, drug targeting, increase in intestinal permeability and increase in bioavailability. Moreover, they avoid the need for use of organic solvents (Gindy and Prud’homme 2009).

Brain cancer therapy with efficient dug delivery shows better results than the conventional system. The drug delivery of pure 5-FU is having the limitation of less systemic circulation and inadequate biodistribution which can be overcome by the formation LDC nanoparticles. Hence, the research work was aimed to formulate and optimize LDC nanoparticles of hydrophilic drug (5-fluorouracil) and study its brain targeting efficiency by in vitro and in vivo performances (Heath and Davis 2008).

Materials and methods

Materials

5-FU was obtained as a gift sample from the Biochem industry, Mumbai, India. Stearic acid and Polysorbate 80 were purchased from Hi-Media Lab Pvt. Ltd., Mumbai, India. Ethanol was purchased from Central Drug House, Delhi, India. All other reagents used in the study were of analytical grade.

Compatibility studies

The compatibility studies were carried out at room temperature using Fourier transform infrared (FTIR) spectroscopy (Bruker, ALPHA-II) to investigate any incompatibility between the drug and the excipients. The pure drug and physical mixture were separately mixed with IR-grade KBr and subjected to FTIR studies alone and in combination and scanned over a wavenumber range of 4000 to 400 cm⁻¹ (Inkielewicz-Stepniak et al. 2014).

Preparation of bulk material of LDC

5-FU: stearic acid was taken in a different molar ratio (1/1, 1/3, 1/5) and dissolved in 0.1 N KOH and solvent system (water/ethanol ratio 3:1). The solvent was evaporated by rota evaporator (EquitronRoteva, India) at controlled pressure (at 60 rpm and 200–400 psi) for 2–3 h. Solid LDC bulk material was collected after the evaporation of the organic solvent. The powder was collected and reduced to a fine powder using mortar and pestle (Sharma et al. 2012).

Preparation and optimization of LDC nanoparticles using experimental design

5-FU nanoparticles were prepared by a solid dispersion method. LDC powder was dispersed in different concentrations of Tween 80 aqueous solution (3, 5, 7% w/v). The mixture was then subjected to homogenization at 8000 rpm for 15 min using a high-pressure homogenizer (HPH) (GEA NiroSoavi, China) to form the predispersion. This predispersion was then subjected to HPH (GEA NiroSoavi,) (800 bar) for a consecutive number of cycles. The obtained suspension was subjected to freeze-drying using the lyophilizer (Allied Frost, New Delhi, India) to obtain the dry solid powder. The product obtained can be reconstituted using the sterile water for injection and stored at a suitable temperature (Liu et al. 2009).

Based on preliminary investigations, the Box–Behnken design was employed to study the effect of independent variables [lipid/drug ratio (X₁), % (w/v) surfactant concentration (X₂) and HPH cycles (X₃)] on dependent variables [particle size (Y₁) and entrapment efficiency (Y₂)]. Design Expert^® 9.0.2.0 software was used for determining the influence of factors on the selected responses and for optimization of formulations (performed in triplicate). The responses (Y) were measured for each experiment and then a simple linear equation, interactive equation, or quadratic model was generated by carrying out multiple regression analysis and F-statistics to identify statistically significant terms (Zheng et al. 2019; Zhang et al. 2020).

Fifteen batches of different combinations were prepared by taking values of selected variables X₁, X₂, and X₃ at different levels. Coded values and actual values of three independent variables, drug/lipid ratio (X₁), surfactant concentration (X₂), and HPH cycle (X₃) and response parameters are recorded and represented in Table 1 (Brigger et al. 2002).

Table 1.

Coded values for selection of independent variables and optimization of formulation variables using Box–Behnken design

Coded value (molar ratio)	Low (− 1)	Medium (0)	High (+ 1)
X1: lipid:drug ratio	1:1 (1)	3:1(3)	5:1 (5)
X2: % w/v surfactant concentration	3%	5%	7%
X3: HPH cycles	15	20	25

Composition and responses for Box–Behnken design for LDC NPs (mean ± SD, n = 3)
Batch	Variable levels in actual form			Response variables
	Lipid: drug molar ratio	% surfactant concentration	HPH cycles	Particle size (nm)	Entrapment efficiency (%)
F1	1	3	20	174.2 ± 2.32	61.2 ± 2.51
F2	5	3	20	150.5 ± 1.21	68.1 ± 3.34
F3	1	7	20	134.1 ± 1.02	80 ± 2.11
F4	5	7	20	168.4 ± 2.19	85.3 ± 1.16
F5	1	5	15	154.5 ± 3.25	65 ± 1.39
F6	5	5	15	155.8 ± 3.19	81.2 ± 2.51
F7	1	5	25	228.5 ± 1.17	65.4 ± 1.31
F8	5	5	25	201.6 ± 1.06	83.2 ± 2.82
F9	3	7	25	100.9 ± 3.24	85.7 ± 1.23
F10	3	7	15	185.2 ± 2.21	83.8 ± 2.23
F11	3	3	25	157.9 ± 2.07	67.1 ± 1.91
F12	3	7	25	162.5 ± 1.91	68.3 ± 2.32
F13	3	5	20	191.6 ± 2.32	64.2 ± 2.10
F14	3	5	20	125.2 ± 1.61	65.9 ± 1.09
F15	3	5	20	157.1 ± 1.58	80.2 ± 2.21

Evaluation of LDC nanoparticles

Particle size distribution, zeta potential, and morphological studies

Particle size distribution and zeta potential of LDC nanoparticles were analyzed using Malvern Zeta Sizer (Nano ZS, Malvern Instruments Ltd, UK) to determine particle size at room temperature. A suspension at 1 mL was diluted up to 100 mL using deionized water and the sample was placed in clear zeta cells and results were recorded. Before putting the fresh sample, the cuvette was washed with deionized water and rinsed using the sample to be measured before each experiment. The morphological characteristic of the optimized formulation was confirmed by scanning electron microscopy (SEM). The freeze-dried nanoparticles were scattered on to 12 mm diameter double-slide adhesive carbon pads which were attached to SEM specimen mounts and the specimen was examined with an electron microscope. (Philips XL 30 ESEM TMP + EDAX, Netherland) (Din et al. 2017).

Entrapment efficiency

The nanoparticles samples were centrifuged at 3000 rpm (Eppendorf) at 4 °C to pelletize the unencapsulated drug. The supernatant was centrifuged at 10,000 rpm to pelletize the drug-loaded LDC nanoparticles. The sample was centrifuged at 3000 rpm again to pelletize the drug alone. The supernatant was removed and the pellet was resuspended and the concentration of the encapsulated drug was measured as absorbance at 267 nm using a UV–visible spectrophotometer (Shimadzu 1800, Japan). The absorbance was converted into drug concentration using a standard curve (Gupta and Gupta 2005).

$$\text{Percent}\text{entrapment}\text{efficiency}=\frac{\text{total}\text{amount}\text{of}\text{drug}\text{loaded}\text{in}\text{LDC}-\text{free}\text{drug}\text{in}\text{supernatent}}{\text{total}\text{amount}\text{of}\text{drug}\text{used}\text{in}\text{the}\text{formulation}}\times 100.$$ 1

Analytical method development

An analytical method was developed as per the International Conference on Harmonization (ICH) guidelines for the estimation of 5-FUin drug-loaded LDC nanoparticles using a Shimadzu high-performance liquid chromatography (HPLC) (LC 10, Japan). The method was performed as per ICH guidelines for the following analytical parameters: specificity, precision, accuracy, linearity, solution stability, ruggedness, and robustness, using 5-FU nanoparticles. The developed method was used for estimation of the drug in formulation studies. The analysis was performed using the mobile phase [water/acetonitrile (90/10)] at a flow rate of 1.2 mL/min, injector with 20 μL fixed loop using the 2998 PDA UV–visible detector at a wavelength of 267 nm (run time 10 min). A standard solution was prepared for the pure drug in the concentration range of 5–60 μg/mL 5-FU to plot the standard curve (Hiremath et al. 2018).

In vitro drug release

Drug release from the suspension was determined using a modified diffusion cell (Dolphin pharmacy instruments, Pvt. Ltd., India) using a cellophane dialysis membrane. Before the start of the experiment, the dialysis membrane was soaked in 5mL phosphate buffer, pH 7.4. The suspension (5-FU) was taken in a dialysis bag (molecular weight: 12,000–40,000 kDa) and suspended in 500 mL phosphate buffer, pH 7.4 (dissolution medium). The system was maintained at a temperature of 37 °C using a thermostat with the standard stirring speed at 500 rpm. Aliquots, each of 1 mL, were withdrawn from the receptor compartment at appropriate time intervals of 0–50 h and were replaced with a fresh dissolution medium to maintain the sink condition. The samples were diluted suitably and analyzed for drug content using HPLC at a wavelength of 267 nm (Jitendar et al. 2013).

Stability studies

Out of all the 15 formulations, F9 was taken for stability studies. Stability studies were carried according to ICH guidelines Q1A [R2] (Stability Testing for new drug substance and Products), and formulation F9 was individually divided into two sample sets. The first sample set was stored at 25 °C ± 2 °C and 65% ± 5% RH (relative humidity), while another sample set was stored at 40 °C ± 2 °C and 75% ± 5% RH. At the interval of 7 days for 1 month (instead of 6 months as mentioned in Q1A [R2]), encapsulation efficiency of all the formulations and size were measured (Ramesh et al. 2006).

In vitro anticancer (cell line study)

The anticancer activity of the optimized formulation was done by using a cell line study with the U373 cell line. Human cancer cells were produced from ACTREC (Advanced Center for Treatment Research and Education on Cancer, Maharashtra, India). SRB assay method was selected to find the anticancer activity. The cell lines were grown in RPMI 1640 medium containing 10% fetal bovine serum and 2 mM l-glutamine, and 100 μg/mL streptomycin. For the screening experiment, cells were inoculated into 96-well microtiter plates in 90 μL at plating densities depending on the doubling time of individual cell lines. After cell inoculation, the microtiter plates were incubated at 37 °C, 5% CO₂, 95% air, and 100% relative humidity for 24 h. To execute the screening experiment, cell seeding was done into 96-well microtiter plates (10,000 cell/well) in 90 µL at plating densities depending on the doubling time of individual cell lines. After cell seeding, the microtiter plates were incubated at 37 °C, 5% CO₂, 95% air, and 100% relative humidity for 24 h before the addition of 5-FU LDC nanoparticles. After 24 h, one plate of each cell line was fixed in situ with trichloroacetic acid (TCA), to represent a measurement of the cell population for each cell line at the time of drug addition (Tz). Experimental drugs (pure 5-Fu and 5-FU LDC nanoparticles) were solubilized in an appropriate solvent at 400-fold the desired final maximum test concentration and stored frozen before use. At the time of drug addition, an aliquot of frozen concentrate was thawed and diluted to ten times the desired final maximum test concentration with a complete medium containing test article at a concentration of 10^–3. Additional tenfold serial dilutions were made to provide a total of four drug concentrations plus control. Aliquots, each of 10 µL, of these different drug dilutions were added to the appropriate microtiter wells already containing 90 µL of the medium, resulting in the required final drug concentrations (Nerkar et al. 2012; Chen et al. 2020).

Sterility test

The sterility test of the optimized formulation was carried out by direct inoculation. Nanoparticle suspensions were first pre-filtered through a membrane with a pore size of 0.45 mm. Thereafter, the suspensions were sterilized by filtering through a filter unit with a 0.22 mm Millipore membrane. Fluid Thioglycolate Medium (F.T.M) and Soyabean Casein Digest Medium (S.C.D.M) were prepared, dispersed in 10 mL quantities, and sterilized by autoclave at 121 °C, 15 psi for 20 min. For the test, 1 mL of sample was taken and added to F.T.M and S.C.D.M media, respectively, under aseptic laminar airflow condition. For −ve control, S.C.D.M and F.T.M media were kept as such for incubation. For +ve control, standard aerobic organism Bacillus subtilis and Staphylococcus aureus and standard fungal organism, Candida albicans, were inoculated into F.T.M and S.C.D.M media, respectively. The test for needle and syringe used in the sterility test was carried out by pipetting out and keeping back the F.T.M media into the test tubes. All the test samples and controls were incubated at 35 °C for F.T.M media and 28 °C for S.C.D.M media for 14 days. Data were observed and recorded (Zhuang et al. 2012).

In vivo bioavailability studies

Bioanalytical method development

A selective and sensitive analytical method was developed to quantitate 5-FU in plasma and brain tissue using a modified HPLC method. Brain tissue (80–400 µL) was homogenized using a microcentrifuge at 8000 rpm for 10 min. The supernatant was transferred to the conical tube where the drug was extracted with 10 mL of methanol. The separated organic phase was then evaporated to dryness at 40 °C under reduced pressure. The residue was reconstituted in 100 µL mobile phase [water/acetonitrile (90:10)], whereby 50 µL from the mixture was injected into the HPLC system to quantitate the drug present in the plasma. Chromatographic separation was achieved isocratically on a C18 column (Inertsil C18, 5 m, 150 mm × 4.6 mm) using a mobile phase [water/acetonitrile (90:10)] at the optimum flow rate (1.2 mL/min) with the detection wavelength of 267 nm. The residue was dissolved in 75 µL of methanol and 20 µL injected into an HPLC system consisting of an LC-20AT VP solvent system and SPD-20 A UV detector (Shimadzu, Japan) (Loira-Pastoriza et al. 2014).

In vivo (pharmacokinetic) studies

Pharmacokinetic studies were performed using either sex mice (100–200 g). The protocol was followed as per the proforma B for animal studies, submitted to Parul Institute of Pharmacy, Vadodara, India, with approval no. 94/PO/AC/05/CPCSEA. Albino mice of either sex were fasted overnight and divided into three groups each containing three mice. The group under treatment was segregated into three groups: group I: tumor control, group II: pure 5-FU and group III: LDC nanoparticles (Ma and Mumper 2013).

The group I received normal saline buffer solution through the tail vein of mice. Similarly, groups II and III also received 10 mg/kg dose of pure drug solution in the saline buffer and LDC nanoparticles, respectively, after 7 days of tumor implantation when the solid tumor was sufficiently grown to a specific volume. The blood samples were withdrawn at an interval of 1, 6, 12, 24 and 48 h. The pharmacokinetic profile and organ distribution of both pure drug and LDC nanoparticles were evaluated in mice bearing tumor. The distribution profile of 5-FU in the organs including plasma and tumor was measured by HPLC analysis (Mudshinge et al. 2011; Cai et al. 2020).

Biodistribution study

All the experiments conducted on animals were approved by the Institute of Animal Ethical Committee (registration no-921/PO/AC/05/CPCSEA). The study was conducted on six healthy New Zealand albino rabbits (6 weeks old) weighing between 1.5 and 2 kg, divided equally into two groups. The rabbits were housed individually in stainless steel cages and fed on the commercial rabbit diet and were allowed water ad libitum. Rabbits were anesthetized with an intraperitoneal high-dose barbiturate (40 mg/kg). 5-FU LDC nanoparticle formulation and 5-FU solution were given through intravenous injection to both the groups, respectively, whereas 0.5 mL (dose equivalent 5 mg/kg) of LDC nanoparticles and the marketed formulation were delivered through the femoral vein (Raut et al. 2010).

Blood samples (0.5 mL) were withdrawn through retro-orbital venous plexus puncture with the aid of glass capillary at 0, 0.5, 1, 1.5 and 2 h. All the samples were collected in heparinized tubes containing EDTA (3.2 mg/mL) and centrifuged (Remi Instruments Ltd., Mumbai, India) at 5000 rpm for 15 min at ambient temperature. Drug samples were extracted by a solid–liquid extraction technique. The rabbits were then killed and the brain was carefully excised, rinsed immediately with saline solution, and blotted with filter paper. The samples were weighed and homogenized with 10 mL methanol by tissue homogenizer (Eletrocraft Pvt. Ltd. India). The supernatant was separated and stored in a deep freezer (at a temperature of 20° C) until analyzed (Ahmadi and Adibhesami 2017).

Distribution of LDC particles in the brain

The procedure involved determining whether the 5-FU concentration in brain tissue was similar in a plasma pharmacokinetic study designed for each group where mice were killed at a determined interval of time. Aliquots of blood samples were collected at a specified interval and analyzed for drug content by HPLC. Thereafter the brain tissues were removed, weighed, and homogenized in PBS, pH 7.4. The homogenates were stored at − 80 °C until further analysis and drug content was analyzed by HPLC. The sample was analyzed using the Hiber C18 column with a Shimadzu LC-10AT solvent delivery gradient pump equipped with a Rheodyne sample injector and SPD-10A VP PDA detector (Shimadzu Analytical Pvt. Ltd., Chennai, India). The mobile phase consisted of acetonitrile/potassium dihydrogen orthophosphate (40/60% v/v), pH was adjusted to 4.5 with orthophosphoric acid and was used at a flow rate of 1 mL/min and the elution was monitored at 254 nm (Chen et al. 2015).

Result and discussion

Compatibility study

Drug excipient interaction was recorded using FTIR spectroscopy. The spectra observed for 5-FU, stearic acid, and LDC nanoparticles bulk material alone and in combination revealed no interaction suggesting the compatibility of the drug and the excipients. The characteristic peaks of pure 5-FU showed IR absorption at 1247 cm⁻¹ (C–N stretching), whereas the peak of stearic acid demonstrated the absorbance at 1706 cm⁻¹ (C=O stretching) and 2849 cm⁻¹ (O–H stretching). In LDC bulk material, 1659 cm⁻¹ (–NH₂) was absent and it confirmed the formation of bulk material (Fig. 1). The amino group present in 5-FU and carboxylic acid, which was present in stearic acid, was absent in LDC bulk material. Therefore, it was confirmed that the latter was formed successfully because carboxylic acid group in stearic acid had reacted with hydroxyl or amine group present in the drug and formed ester or amide linkage during lipid–drug conjugation in water and water-miscible solvent system (Koopaei et al. 2014).

Fig. 1.

FTIR spectra of a 5-FU, b stearic acid. c LDC nanoparticle bulk material

Preparation and optimization of LDC nanoparticles by using experimental design

LDC nanoparticles of 5-FU were successfully prepared using the solid dispersion method and optimized by applying Box–Behnken design approach. The effect of various formulation variables was analyzed to optimize the formulation using Design-Expert^® software. The concentration of lipid drug ratio, surfactant, and HPH cycles were the determinant factors as these demonstrated a significant effect on the response variables, i.e., particle size and percentage entrapment efficiency (Locatelli and Franchini 2012).

In the present investigation, the effect of concentration of drug/lipid ratio (X₁), surfactant (X₂), and HPH cycles (X₃) on particle size (Y₁) and percentage entrapment efficiency (Y₂) was studied using Box–Behnken design. Polynomial equation comprising the individual main effects and interaction factors were selected considering the statistical parameters, viz., multiple correlation coefficient (R²), adjusted multiple correlation coefficient (adjusted R²), and the predicted residual error sum of squares (PRESS) provided by the Design-Expert^® software (Stat-Ease Inc., Minneapolis, MN, USA) (Augustine and Rajarathinam 2012).

The smaller the PRESS statistic, the better the model fitted the data points. It was indicated by the measure of the fit of the model to the points in the design. The resultant equations for percentage particle size (Y₁) and encapsulation efficiency (Y₂) are presented as follows:

$$\begin{array}{cc}\hfill \text{Particle}\text{size}(Y_1)& =192.25+21.79X_1-18.79X_2-16.18X_3+40.33X_1{X}_2-41.25X_1{X}_3\hfill \\ \hfill & -55.60X_2{X}_3-6.64X_1^2-9.79X_2^2-8.79X_3^2.\hfill \\ \hfill \end{array}$$ 2

Reduced equation:

$$\begin{array}{cc}\hfill \text{Particle}\text{size}(Y_1)& =178.09+21.79X_1-18.79X_2-16.18X_3+40.33X_1{X}_2\hfill \\ \hfill & -41.25X_1{X}_3-55.60X_2{X}_3.\hfill \\ \hfill \end{array}$$ 3

The regression analysis for response Y₁ (particle size) is shown in Eq. (2). The particle size varied from 100.8 to 230 nm indicating the positive effect of X₁ (lipid: molar ratio) on the particle, while X₂ (% w/v surfactant concentration) factors and X₃ (HPH cycles) factor had a negative impact on particle size. Results of the regression indicate (according to p value in ANOVA table, i.e., Table 2A) that the effect of and X₂ (% w/v surfactant concentration) and X₃ (HPH cycles) is more significant than X₁ (drug/lipid ratio).

$$\begin{array}{cc}\hfill \text{Entrapment efficiency}(Y_2)& =80.50+1.76X_1-0.13X_2-0.80X_3-0.44X_1{X}_2+0.45X_1{X}_3\hfill \\ \hfill & -0.23X_2{X}_3-2.78X_1^2-0.73X_2^2+9.10X_3^2.\hfill \\ \hfill \end{array}$$ 4

Table 2.

Results of ANOVAfor measured response for A—particle size and B—percent entrapment efficiency

Source	Sum of squares	Degree of freedom	Mean square	F value	p value
A—ANOVA (particle size)
Model	34,878.51	9	3875.39	10.06	0.0201	Significant
A: lipid–drug molar ratio	3797.56	1	3797.56	9.85	0.0349
B: surfactant concentration	2823.76	1	2823.76	7.33	0.0537
C: HPH cycles	2093.05	1	2093.05	5.43	0.0802
AB	6504.42	1	6504.42	16.88	0.0148
AC	6806.25	1	6806.25	17.66	0.0137
BC	12,365.44	1	12,365.44	32.09	0.0048
A2	140.98	1	140.98	0.37	0.5779
B2	306.54	1	306.54	0.80	0.4229
C2	223.78	1	223.78	0.58	0.4885
Residual	1541.48	4	385.37
Corr. total	36,420.00	13
B—ANOVA (percent entrapment efficiency)
Model	355.32	9	39.48	103.24	0.0002	Significant
A: lipid–drug molar ratio	24.68	1	24.68	64.52	0.0013
B: surfactant concentration	4.31	1	4.31	11.26	0.0284
C: HPH cycles	15.74	1	15.74	41.15	0.0030
AB	0.78	1	0.78	2.05	0.2256
AC	0.79	1	0.79	2.07	0.2235
BC	8.12	1	8.12	21.24	0.0100
A2	15.23	1	15.23	39.81	0.0032
B2	5.63	1	5.63	14.72	0.0185
C2	231.27	1	231.27	604.75	< 0.0001
Residual	1.53	4	0.38
Corr. total	356.85	13

Reduced equation:

$$\begin{array}{cc}\hfill \text{Entrapment}\text{efficiency}(Y_2)& =80.50+1.76X_1-0.13X_2-0.80X_3+0.45X_1{X}_3-2.78X_1^2\hfill \\ \hfill & -0.73X_2^2+9.10X_3^2.\hfill \\ \hfill \end{array}$$ 5

The regression analysis for response Y₂ (% entrapment efficiency) is shown in Eq. (4). Percentage EE varied from 60 to 85. It is clear from the equation that the X₁ (drug/lipid molar ratio) and X₂ (% w/v surfactant concentration) factors had a positive effect on % EE. Results of the regression indicate (according to p value in ANOVA table, i.e., Table 2B) that the effect of X₂ (surfactant concentration) and X₁ (drug/lipid ratio) is more significant than that of X₃ (HPH cycles).

The particle size and percentage entrapment efficiency were observed in the range of 100.8–228.5 nm and 60–85%, respectively. The regression analysis for the response Y₁ (Eqs. 1, 2) indicates that the particle size is directly proportional to the concentration of drug/lipid ratio (Y₁) and inversely proportional to the surfactant concentration (X₂). The HPH cycles (X₃) were found to be more significant than the lipid drug ratio (X₁). Equation (3) elaborates the regression analysis for percentage entrapment efficiency where it demonstrates increased entrapment efficiency with an increase in the concentration of lipid drug molar ratio and surfactant. It has been revealed that an increase in the concentration of lipids subsequently increases entrapment efficiency. Lipid drug tends to solubilize into lipids with ease resulting in improved entrapment efficiency, which suggests that the drug/lipid ratio is the leading factor responsible for an increase in entrapment efficiency (Foldbjerg et al. 2011).

F value and p value recorded from the ANOVA for the measured response confirmed the significance of the model for particle size (0.02) and entrapment efficiency (0.0005). Response surface curve and contour plots were used to study the influence of X₁, X₂, and X₃ on the response obtained for Y₁ and Y₂, by keeping the third factor at a constant level (Figs. 2, 3).

Fig. 2.

3D response surface plots for particle size analysis, a response surface plot showing the effect of the amount of lipid:drug molar ratio (X₁) and surfactant concentration (X₂) on particle size (nm) (Y₁). b Response surface plot showing the effect of the amount of lipid:drug molar ratio (X₁) and HPH cycles (X₃) on particle size (nm) (Y₁). c Response surface plot showing the effect of the amount of surfactant concentration (X₂) and HPH cycles (X₃) on particle size (nm) (Y₁)

Fig. 3.

3D response surface plots for entrapment efficiency: a response surface plot showing the effect of the amount of lipid:drug molar ratio (X₁) and surfactant concentration (X₂) on % EE (Y₂). b Response surface plot showing the effect of the amount of lipid:drug molar ratio (X₁) and HPH cycles (X₃) on% EE (Y₂). c Response surface plot showing the effect of the amount of surfactant concentration (X₂) and HPH cycles (X₃) on% EE (Y₂). 3D plots were used to study the influence of two factors on the response at a time, when the third factor was kept at a constant level

Formulation F9 was an optimized formulation considering desirable properties (with the particle size of 100.9 ± 3.24 and entrapment efficiency of 85.7 ± 1.23) obtained from the design approach when compared with other formulations (Shrivastava et al. 2007). The checkpoint batch was prepared by lipid/drug ratio of 2.42:1, a surfactant concentration of 6.94% w/v, and 23 cycles of HPH. Table 3 elaborates the record obtained of checkpoint batch from the equations and the actual study performance on the formulation prepared as per the checkpoint method. Values of particle size and percentage entrapment efficiency were found to be 100.6 nm and 81.32%, respectively, which were found to be similar without any significant difference. The results confirmed the equation obtained from the data analysis of both the dependent variables. The analysis showed that there were no significant differences between the actual and predicted values on response Y₁ and Y₂.

Table 3.

Comparison of result for checkpoint batch with performed batch

Results of actual checkpoint batch	Results of performed checkpoint batch
Y1: particle size Y1 = 101.8 nm	Y1: particle size Y1 = 100.6
Y2: % entrapment efficiency Y2 = 83.32%	Y2: % entrapment efficiency Y2 = 81.32%

Evaluation of LDC 5-FU nanoparticles

Nanoparticles were evaluated for particle size, entrapment efficiency, and morphological studies. Particle size was a significant parameter in delivering the drug. Capillary distribution and permeation have been observed with the particle size of less than 1 µm to the targeted sites. Most solid tumors have elevated levels of vascular permeability. Particles less than 400 nm can cross vascular endothelia and accumulate at the tumor site through the EPR effects. The optimized nanoparticles were found to be 100.9 nm which will act as the effective carrier for the drug. The entrapment efficiency varied from 60 to 85%, where it indicated the direct correlation of lipid/drug molar ratio and percentage of surfactant concentration. It was found to be directly proportional to the lipid drug molar ratio and inversely proportional to the concentration of the surfactant. The entrapment efficiency of the optimized formulation was found to be 85.7%. Zeta potential analysis is shown in Fig. 4. The optimized batch (LDC nanoparticles) demonstrates the negative surface charge (− 20.8) indicating stabilization of the formulation. Scanning electron microscopy (SEM) of the optimized nanoparticles was carried out to find the morphological characteristics of the nanoparticles. The obtained image (Fig. 5) confirms that the optimized nanoparticles were nearly spherical (Chauhan and Jain 2013).

Fig. 4.

a Particle size distribution (F9) and b zeta potential of LDC nanoparticles (F9)

Fig. 5.

SEM image of LDC nanoparticles

In vitro drug release studies

In vitro drug release studies were carried out using a modified diffusion cell. Cumulative drug release from the optimized nanoparticle was found to be 48.41% after 48 h, while the drug release from the 5-FU solution was found to be 88.32% (Fig. 6). The release profile indicates that LDC nanoparticles release the drug for a maximum duration of time. The release data of gel formulation (Table 4) were kinetically analyzed by different mathematic models, viz., zero order, first order, Higuchi, Korsmeyer–Peppas, and Hixson–Crowell representing goodness of fit in terms of R² values as illustrated in Fig. 7 (Shelke et al. 2016b). The linearity found in Higuchi model (r² = 0.97) suggested dissolution rate-limited drug release from LDC nanoparticles. The value of release exponent (n) determines the release mechanism from the formulation. Non-Fickian diffusion is predicted if the values of n are between 0.5 and 1. The value of n = 0.5 indicates Fickian diffusion, whereas a value greater than 1 (n) indicates super case II transport. As presented in Table 4, Korsmeyere–Peppas equation (n = 0.5) demonstrates Fickian diffusion suggesting dissolution control release of the drug (Jain et al. 2008).

Fig. 6.

In vitro drug release study of LDC nanoparticles and 5-FU solution

Table 4.

Release kinetics of optimized batches of LDC nanoparticles

Release kinetics of optimized batches of LDC nanoparticles
Zero-order release		First-order release		Higuchi model		Hixson–Crowell model		Korsmeyer–Peppas model		Release mechanism
K0	R2	K1	R2	KH	R2	N	R2	N	R2	Fickian diffusion
0.328	0.83	− 0.002	0.93	1.87	0.97	0.0023	0.94	0.516	0.96

Fig. 7.

Drug release and kinetic models

Stability studies

Stability studies [ICH guideline Q1A (R²)] of the optimized formulation F9 were performed for the determining factors, viz., entrapment efficiency, particle size, and physicochemical properties. As observed from the records (Table 5), sample analysis after the interval of 7 days, for 1 month instead of 6 months as mentioned in Q1A, R² showed no significant change in all determining factors suggesting the stability of LDC nanoparticles (Bala et al. 2004).

Table 5.

Stability study of optimized formulation

Stability study data
Time (week)	Conditions	Size (nm)	Percent entrapment efficiency
0	Room temperature 25 °C ± 2 °C; 65% ± 5% RH	103.4	78.65
1		110.2	75.24
2		113.6	72.17
3		116.3	70.54
4		120.2	70. 31
0	Accelerated temperature 40 °C ± 2 °C; 75% ± 5% RH	103.4	78.65
1		104.6	77.21
2		115.3	72.25
3		123.5	70.25
4		126.4	68.21

In vitro anticancer study (end point measurement)

It was found that the percentage viability of the U-373 cell significantly decreases with an increase in the concentration of LDC in a different form. The results recorded in Fig. 8 states that the pure drug shows pharmacologically cytotoxic activity on the cancer cell and depends on the dose, i.e., percentage viability of cancer cell line was about 96.2 ± 1.2% at the initial dose of 100 µg/mL, while the percentage viability of the cell declined to 63.1 ± 0.52% at higher concentration (1000 µg/mL). However, LDC shows significantly cytotoxic activity than pure drug at the much lower concentration required to produce a cytotoxic effect on cancer cells. The percentage viability of the U-373 cell line decreased from 84.12 ± 1.8 to 28.1 ± 0.4 at the concentration to 1000 µg/mL, respectively. It indicates that the cytotoxic effect on cancer cell line by LDC was enhanced on comparison with pure drug and and there was an almost twofold increase in the anticancer activity of LDC compared to the pure drug after 24 h exposure. The percentage viability of cancer cells is about 28.1 ± 0.4 at the concentration of 5 Fu at 1000 µg/mL (Beduneau et al. 2007).

Fig. 8.

In vitro anticancer study sample 1: LDC nanoparticles of 5-FU, sample 2: blank stearic acid nanoparticles, std. drug: 5 FU and adriamycin (ADR)

Sterility test

Sterility study is the decisive parameter to ensure the sterility of the finished product. As patients are treated with clinically prepared formulations, these need to comply with the pharmacopeia requirement offering the highest purity. LDC nanoparticles were evaluated for the sterility test as per the official compendia. The growth or turbidity was observed in the standard sample of B. Subtilis, S. aureus, and C. albicans. To achieve the official limits and high degree of purity, a sterility test was conducted to confirm its application in the design of the dosage form. Results represented no growth or turbidity in the test sample of the optimized formulation for negative control and control for needle and syringe. The growth or turbidity was observed in the standard sample of B. subtilis, S. aureus, and C. albicans (Dikpati et al. 2012).

Pharmacokinetic study

The pharmacokinetic behavior was assessed by injecting 5-FU-loaded LDC and pure drug (10 mg/kg) through the I.V. route to the U-373 tumor-bearing mice. Table 6 represents the plasma profile (5-and LDC nanoparticles) and pharmacokinetic parameters, which reveal fast removal (1 h) of the free drug solution from the systemic circulation. However, I.V. injection represented negligible blood concentration showing a biphasic pattern with a rapid elimination phase with the half-life (t_1/2) of 1.33 ± 0.27 h and volume of distribution (Vd) of 5.24 ± 0.29 mL. The value of AUC, AUMC, and MRT for free 5-FU significantly showed lower systemic circulation profile when compared to LDC as suggested by ANOVA (p < 0.005) (Prabhjot et al. 2014; Pan et al. 2020).

Table 6.

Pharmacokinetic parameters of free 5-FU, LDC into the tumor-bearing mice

S.N.	Pharmacokinetic parameters	Units	Free 5-FU	LDC
1	AUC	µg/mL/h	8.37 ± 0.04	19.37 ± 0.09
2	AUMC	µg/mL/h	9.65 ± 0.12	230 ± 0.11
3	Cmax	µg/mL	18.8 ± 0.98	36.2 ± 1.85
4	Vd	mL	5.24 ± 0.29	2.4 ± 0.24
5	t1/2	H	1.33 ± 0.27	11.48 ± 0.21
6	Ke	h−1	0.11 ± 0.01	0.01 ± 0.025
7	Cl	mL/min	2.173 ± 0.05	0.012 ± 0.11
8	MRT	H	1.02 ± 0.11	12.10 ± 0.44

When drug-loaded nanoparticles were compared with 5-FU for pharmacokinetic profile after I.V. injection to animal models, it indicated a significantly greater response than a pure drug concerning area under curve (AUC), area under the first moment curve (AUMC), MRT and t_1/2. The values reported for LDC (AUC = 19.37 ± 0.09 µg/mL h and VD 2.4 ± 0.24 mL) and pure drug (AUC = 8.37 ± 0.04 µg/mL h and VD = 5.24 ± 0.29 mL) indicate higher concentrations of LDC in systemic circulation, while pure 5-FU was found to be largely available in tissue rather than blood circulation. The t_1/2 for LDC represents approximately rise by ninefold, while MRT (12.10 ± 0.44 h) denotes 12-fold increase than pure 5, indicating a prolonged circulation of LDC. The present dose of LDC nanoparticles containing drug in plasma after 24 h studies proves that the 5-FU encapsulated inside the nanoparticles confined the drug in the systemic circulation (Fukumoto et al. 2001).

Cellular cytotoxicity and plasma distribution study

Cellular cytotoxicity of LDC nanoparticles at various drug concentrations against blank stearic acid nanoparticles and the pure 5-FU drug was studied; the result showed that LDC NPs have a good ability to kill the viable cells as compared to blank nanoparticles and pure drug (Fig. 9). In the results reported, it was found that the free 5-FU rapidly appeared in blood within 2 h after I.V. injection of tumor-bearing mice with maximum concentration (13.34 ± 0.98 µg/mL), which demonstrated a decline phase with increase in time. The plasma level (5-FU nanoparticles) after 2 h was found to be 25.23 ± 0.71 µg/mL, indicating a twofold increases compared to free drug. The 5-FU nanoparticles were found to be present in the blood circulation for 24 h and are represented in Fig. 10. A remarkable enhancement of nanoparticles in blood circulation (12.10 ± 0.44) was observed than that of the pure drug indicating 12-fold increases (Calvo et al. 1997).

Fig. 9.

Cell viability study of LDC nanoparticles, pure drug and test control

Fig. 10.

Distribution profile of pure 5-FU and LDC in plasma

Distribution of LDC particles in the brain

Concentrations of 5-FU in the brain for different formulations are shown in Fig. 11. Free 5-FU concentration in the brain was maximum (5.24 ± 0.01 μg/g) after 3 h, while for the optimized formulation of LDC it was twofold greater, estimated at 11.52 ± 0.32 μg/g. The optimized formulation of LDC estimated more AUC in the brain in comparison with other free 5-FU (p < 0.05). The decline phase was observed with increase in time, wherein negligible concentration was observed after 24 h for free 5-FU (0.03 ± 0.12 μg/g) and LDC (0.56 ± 0.2 μg/g). The results depict that LDC nanoparticles showed a higher concentration of drug in the brain as compared with the free 5-FU (p < 0.05) (Hu et al. 2002).

Fig. 11.

Distribution profile of pure 5-FU and LDC in brain

Conclusion

LDC nanoparticles of 5-FU were successfully prepared by the high-pressure homogenization technique using stearic acid as the lipid which successfully conjugated with the 5-FU. The formulation was optimized by response surface methodology using Design-Expert software, and the optimized batch (F9) was characterized for various parameters such as particle size and entrapment efficiency. In vivo biodistribution studies revealed that there was an increase in the permeation of the LDC nanoparticles in the brain compared to simple drug solutions. LDC showed AUC = 19.37 ± 0.09 µg/mL h and VD 2.4 ± 0.24 mL, while pure drug showed AUC = 8.37 ± 0.04 µg/mL h and VD = 5.24 ± 0.29 mL which indicates the higher concentrations of LDC in systemic circulation than pure 5-FU. The t_1/2 for LDC was found to be 11.48 ± 0.21 h, which is a rise by 9-fold as compared with pure 5-FU, while MRT (12.10 ± 0.44 h) denotes 12-fold rise than pure 5-FU indicating a prolonged circulation of LDC. Free 5-FU concentration in the brain was maximum (5.24 ± 0.01 μg/g) 3 h after administration of pure 5-FU, while for the optimized formulation of LDC it was twofold greater (11.52 ± 0.32 μg/g). SEM image showed that most of the nanoparticles have spherical-shaped morphology. In in vitro studies of optimized LDC, nanoparticles showed an initial burst release, followed by a sustained release and the best fit model for release kinetics was the Higuchi model. In vitro evaluation and cell line study show that LDC NPs seem to be a sustained dosage form of 5-FU for brain cancer treatment. The formulation passed the sterility test. The developed formulation has been stable for 3 months at room temperature.

Funding

The authors of the paper have no direct financial relationship with any commercial identity mentioned in this paper.

Compliance with ethical standards

Conflict of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.