

ORIGINAL ARTICLE 

Year : 2021  Volume
: 11
 Issue : 5  Page : 216224 

Enhancement of solubility of paclitaxel by applying factorial design
Pragya Baghel^{1}, Amit Roy^{1}, Shekhar Verma^{2}, Trilochan Satapathy^{2}, Sanjib Bahadur^{1}
^{1} Department of Pharmaceutics, Columbia Institute of Pharmacy, Raipur, Chhattisgarh, India ^{2} Department of Pharmaceutics, Pt. Deendayal Upadhyay Smriti Health Science and Aayush University, Raipur, Chhattisgarh, India
Date of Submission  06Nov2020 
Date of Decision  07Sep2021 
Date of Acceptance  10Sep2021 
Date of Web Publication  30Oct2021 
Correspondence Address: Dr. Pragya Baghel Columbia Institute of Pharmacy, Tekari, Near Vidhan Sabha, Raipur  493 111, Chhattisgarh India
Source of Support: None, Conflict of Interest: None
DOI: 10.4103/cmrp.cmrp_51_20
Background: The poor aqueous solubility and low bioavailability of antineoplastic drugs restrict the oral route for the cancer therapy. However, specific nanotechnologies developed to address this issue. The selfmicro emulsifying drug delivery system (SMEDDS) is an imperative tool in solving a lipophilic drug's low solubility and bioavailability issue. Paclitaxel (PTX) is an antineoplastic drug used to treat various kinds of cancer, especially ovarian and breast cancers. It has low aqueous solubility (0.3 μg/ml in water), resulting in poor bioavailability. This study focuses upon formulating SMEDDS incorporating PTX to increase the drug's aqueous solubility. Castor oil (Surfactant), Tween 80 (surfactant), and Transcutol (cosurfactant) were used to formulate SMEDDS. Two independent factors, concentrations of oil and surfactant, were chosen. Two dependent factors, emulsification time and in vitro drug release, were chosen. 3^{2} factorial design analyses give a mathematical expression, which helps to study the effect of an independent factor on a dependent factor. All the nine formulations B1B9 evaluated for globule size, zeta potential, and present drug content. Results: The globule size found is 127–217 nm range, which indicates the formation of the transparent microemulsion; zeta potential fields between −4.29 ± 1.1 and −33.05 ± 4.5 indicate suitable stable formulation. Nearly 63%–84% drug content suggests that the drug, incorporated into prepared SMEDDS. Statistical analysis shows that an increasing amount of surfactant decreases emulsification time; this may also reduce the average droplet size of the resultant microemulsion. With the increase in the concentration of tween 80, PTX release also increased. Rapid and more extent of PTX released from formulated SMEDDS indicate that the aqueous solubility of PTX has increased. B9 formulation is the optimized formulation based on higher per cent cumulative drug release 98% at the end of 60 min, 58% of pure PTX powder solubilized in the dissolution medium. Conclusion: In vitro drug release study proved that the prepared SMEDDS has acceptable properties of immediate release dosage forms. The enhanced solubility of PTX will increase its bioavailability.
Keywords: Antineoplastic, bioavailability, experimental design, lipophilicity, paclitaxel, selfemulsification
How to cite this article: Baghel P, Roy A, Verma S, Satapathy T, Bahadur S. Enhancement of solubility of paclitaxel by applying factorial design. Curr Med Res Pract 2021;11:21624 
How to cite this URL: Baghel P, Roy A, Verma S, Satapathy T, Bahadur S. Enhancement of solubility of paclitaxel by applying factorial design. Curr Med Res Pract [serial online] 2021 [cited 2021 Dec 8];11:21624. Available from: http://www.cmrpjournal.org/text.asp?2021/11/5/216/329704 
Introduction   
Variation in drug absorption and the high tendency of drug efflux through pglycoprotein transporters in the lumen leads to poor bioavailability of lipophilic drugs.^{[1]} Physicochemical properties such as high lipophilicity and molecular weight repeatedly reduce the aqueous solubility and thus dissolution rates of such medications. Data report that 40% and more of such drugs formulated comprise of minimum solubility and oral bioavailability.^{[2]} To overcome the problems mentioned above such as dissolution rate and aqueous solubility techniques introduced and formulated like prodrugs, solid dispersions, inclusion complex structure formation, nanostructures and use of various chemical entities such as surfactants, solubilizing agents and many more to enhance the availability of drugs in the biological system.^{[3]} Microemulsions investigated as the potential new colloidal carrier for lipophilic drugs. Microemulsions, when compared to the conventional emulsion, have a few advantages. It possesses excellent thermodynamic stability, improved capability to the solubilized drug, and the capacity to enhance oral bioavailability and protect the drug against enzymatic hydrolysis. The microemulsion also has some limitations. One of which is poor organoleptic properties. Its high lipid content sometimes leads to poor patient compliance.^{[4]} Selfmicroemulsifying drug delivery system (SMEDDS) is a lipidbased system explicitly designed to improve the oral bioavailability of lipophilic drugs. A few studies report that poorly aqueous soluble drug, when incorporated into SMEDDS, has enhanced the bioavailability.^{[5]} There are several techniques for developing novel drug delivery systems. Researchers have attempted lipidbased delivery of lipophilic drugs such as cyclosporine and concluded that cyclosporine is a potential candidate for such delivery. Lipidbased formulation, especially SMEDDS, is one of the promising techniques to address issues solubilityrelated bioavailability. SMEDDS has gained exposure for their ability to increase solubility and bioavailability.^{[6]} Paclitaxel (PTX) is a diterpenoid (naturally present) used to treat many cancer cells including those of breast, lung and ovarian. It is known to be an effective molecule among antineoplastic agents. The major ratelimiting step in the bioavailability of PTX is its aqueous solubility (0.3 μg/ml).^{[7],[8],[9]} Many researchers attempt the conventional 'trial and error' process to develop SMEDDS. However, this method is timeconsuming and demands a workforce. Solutions for such issues, many statistical toolbased experimental designs help optimize SMEDDS.^{[10],[11],[12]}
Thus, this research aims to develop and optimize PTX loaded SMEDDS using factorial design to enhance the aqueous solubility of PTX. After characterization of the SMEDDS, in vitro drug release of pure PTX and SMEDDS was compared. Increased drug release from SMEDDS compared to pure PTX will reveal that the solubility of PTX has increased.
Methods   
Material
Received PTX as a gift sample from Neon Laboratories, Mumbai, as a gift sample. Castor oil, Tween 80 and Transcutol were obtained from the Loba Chemie Pvt. Ltd., Mumbai. All other chemicals used were bagged in from SigmaAldrich, Mumbai. All chemicals were of analytical grade.
Construction of ternary phase diagram
Different ratios of Tween80 (surfactant), Transcutol (cosurfactant) and castor oil were mixed to prepare other systems. We took 10 mL of 0.1N hydrochloric acid in a beaker and then added to it a mixture of tween 80, Transcutol and castor oil was mixed. We mixed the contents of this beaker using a magnetic stirrer at 37°C.
Formulation of selfmicroemulsifying drug delivery system by applying 3^{2} factorial design
We selected 3^{2} full factorial designs to study the effect of independent variables, amount of surfactant and oil, on dependent variables, emulsification time and per cent drug release of liquid SMEDDS using Design Expert 7.0 version, StatEase, USA. The factorial design utilises various interactive and polynomial terms to evaluate dependent responses statistically.
In this design, we evaluated two factors at three levels. The two independent variables were selected, namely tween 80 (surfactant) denoted in the table as (X1) and castor oil (oil) represented in the table as (X2). [Table 1] shows the development of batches of SMEDDS by 3^{2} factorial designs. The table also shows the coded level and actual level of each factor. All the nine possible formulations went through experimental trials, as shown in [Table 2].  Table 1: Development of batches of selfmicroemulsifying drug delivery system by 32 factorial designs
Click here to view 
 Table 2: Formulation of selfmicroemulsifying drug delivery system of paclitaxel along with batch code
Click here to view 
Method of preparation of selfmicroemulsifying drug delivery system
We formulated all nine formulations containing different concentrations of castor oil (oil), tween 80 (surfactant) and Transcutol (cosurfactant) by the following method. [Table 2] shows the quantity of each ingredient to be incorporated in preparation. [Figure 1] shows the chart of the preparation method of SMEDDS.
Evaluation of prepared selfmicroemulsifying drug delivery system
Determination of optical transparency
All the formulations (B1–B9) were inspected visually in the presence of good light. We observed them against a black and white bright background. We checked them for clarity and transparency.^{[13]}
Robustness to dilution study
We studied robustness to dilution by diluting the formulations (B1–B9) to 50, 100 and 1000 times with dissolution media, i. e., 0.1N HCl solution. The diluted samples were stored for 24 h and observed for any sign of phase separation or precipitation.^{[14]}
Rheological measurement
We performed the rheological measurement of formulations (B1–B9) with the Brookfield viscometer (RV II) with a spindle no. 3. 100 ml of the formulation was taken in a beaker and allowed to settle at 25°C for 10 min before viscosity measurement. Finally, we recorded the dial reading and calculated viscosity.^{[15]}
Determination of emulsification time
The emulsification time of formulations (B1–B9) was determined using dissolution apparatus II (paddle type). In brief, we added 1 mL of each formulation to 500 mL of 0.1N HCl solution at 37°C. Gentle agitation was provided by dissolution paddle rotating at 75 rpm. We assessed the emulsification time visually and then noted the time.^{[16]}
Determination of percentage drug content
1 ml of SMEDDS formulations B1B9 diluted up to a concentration of 10 μg/ml. Then, absorbance was measured at 228 nm using a ultraviolet (UV) spectrophotometer. We calculated the per cent drug consent.^{[17]}
Globule size and polydispersity index determination
Globule size and the polydispersity index (PDI) were estimated using Zetasizer 300HSA (Malvern instrument, Malvern, UK). We diluted 0.1 ml of each SMEDDS formulation 1:1000 v/v with 0.1N HCl. After 5 min, the filtrate was filled to polystyrene cuvette and was placed in a cuvette holder. The lid was closed. Then, globule size and PDI were determined.^{[18],[19]}
Zeta potential determination
Zeta potential was estimated using Zetasizer 300HSA (Malvern instrument, Malvern, UK). All nine formulations were suitably diluted to 1:1000 v/v with the 0.1N HCl solution. The zeta cuvette was filled with filtrate and placed in a cuvette holder. The lid was closed, and we determined the zeta potential.^{[19],[20]}
In vitro drug release study
In vitro drug release was performed using the dialysis membrane method. At first, the membrane was clamped in an open glass tube for drug release and considered a donor compartment. We used 500 ml of 0.1N HCl solution as a dissolution medium, taken in the receiver compartment, then 5 ml of SMEDDS formulation was diluted with 5 ml of 0.1N HCl solution and filled into the glass tube. The glass tube (donor compartment) edge was just touching the receiver compartment. We kept the dissolution test medium 500 mL of 0.1N HCl solution (hydrochloric acid) at 37°C ± 0.5°C. Magnetic stirrer and beads maintained the rotation of 75 rpm for 60 min. We withdrew 1 ml sample from the receiver compartment at fixed intervals, and this was replaced with a fresh medium immediately to maintain the sink condition. Similarly, the complete procedure took 10 mg of pure PTX powder sample for an in vitro release study. Samples were analysed using UVvisible spectrophotometer at 228 nm. The amount of drug release was determined using PCP Disso software developed by Bharti Vidyapeeth Poona College of Pharmacy, Pune.^{[13]}
Results   
Ternary phase diagram
Oil, surfactant and cosurfactant were used in different combinations for the construction of the phase diagram. We recorded visual observations of the emulsification. According to the grade discussed in the method, samples were categorized with no phase separation upon storage for 72 h at room temperature. We selected quality A and B systems as most desired due to the formation of transparent microemulsion and smaller particle size. [Table 3] represents the grade according to visual appearance of the microemulsion. [Figure 2] shows the ternary phase diagram.
Optical transparency
Liquid SMEDDS are generally transparent, homogenous singlephase systems; hence, they were characterised for their optical transparency. All the Liquid SMEDDS were transparent and appeared like a homogenous singlephase liquid when observed for visual clarity against the light.
Robustness to dilution
This study results in the determination of the effect of various dilutions on the formulation. We evaluated the influence of dilution; larger dilutions may better mimic conditions in the stomach following oral administration of SMEDDS.
Rheological measurement
We determined the viscosity of all formulations. The results indicated that viscosity increased with the increase in the concentration of tween 80 (surfactant).
Emulsification time
The result of emulsification time is in [Table 4]. We observed that an increase in the amount of surfactant decreases emulsification time. The batch B7 has less emulsification time.  Table 4: Results of viscosity (cP), emulsification time (second) and percentage drug content
Click here to view 
Per cent drug content
The results of per cent drug content are shown in [Table 4].
Globule size and polydispersity index measurement by zetasizer
The globule size within the range of 100–250 nm usually forms clear transparent microemulsion after the dilution of SMEDDS. The globule size of all the formulations, B1– B9, was found within the range of 100–250 nm.
Zeta potential
Zeta potential is the charge present on the mobile surface of the emulsion. Zeta potential is directly related to the stability of the emulsion. We used Zetasizer to determine zeta potential and found the zeta potential of SMEDDS formulations B1B9 in the range of4 to33 mV. The Zeta potential of all the SMEDDS formulations B1B9 is in [Table 4].
In vitro drug release
In vitro drug release study indicated that an increase in the amount of surfactant increases drug release. We found maximum drug release in the B9 batch. It may be because of the optimum amount of oil for selfemulsification. In vitro drug release studies are shown in [Figure 3] (a. B1B5 and b. B6B9 and PTX).  Figure 3: %CDR of the different formulations, (a. B1B5 and b. B6B9 and PTX
Click here to view 
Optimization
Traditional designing of pharmaceutical formulations depends on a timeconsuming approach of changing one variable, which does not consider the combined effect of independent variables. Factorial design is an essential tool that helps in understanding the complications of pharmaceutical formulations. The results can be expressed either as simple linear or secondorder polynomial equations to evaluate the responses obtained after experiments statistically.
Simple liner equation
Y_{i} = b_{0}+b_{1}×_{1}+b_{2}×_{2}
Or
Second order polynomial equation
The mathematical equation expresses the relationship between dependent and independent variables. Statistical software Design Expert v7.0, StatEase, USA, was used to generate mathematical expressions. The dependent variable is measured as it gets affected during the experiment. The dependent variable responds to the independent variable. An independent variable might affect the outcome of the dependent variable. The concentration of tween 80 (X1) and castor oil (X2) was selected as independent variables, and the dependent variables were emulsification time and per cent cumulative drug release of SMEDDS. Analysis of variance (ANOVA) was used for data analysis, statistically. The reported data were subjected to treatment by 3D response surface methodology, and we identified a study of the interaction of tween 80 (X1) and castor oil (X2) on dependent variables.
Evaluation and interpretation of research findings are most important, and the P value serves a valuable purpose in these findings. Coefficients with more than one factor term represent the interaction terms, and coefficients with higherorder terms indicate the linear nature of the relationship. Overall, all the variables caused a significant change in the responses. ANOVA and multiple regression analysis used StatEase Design Expert 7.0 software.
The fitted equation of full and reduced models relating the experimental responses with the transformed factors is in [Table 5]. One can draw the conclusion using a polynomial equation. One should also consider the magnitude of the coefficient of the polynomial equation and its sign (positive or negative) for concluding. For all the responses, the level of significance of coefficients b_{12}, b_{11} and b_{22} was more than 0.05 and hence insignificant. In contrast, the level of significance of coefficients b_{0}, b_{1} and b_{2} was significant at P < 0.05. We have omitted insignificant levels of the coefficient from full models and then retained substantial levels of the coefficient to generate reduced models. We tested the reduced model in portions to determine if the insignificant coefficients, b_{12}, b_{11} and b_{22}, significantly predict dependent responses.^{[21]}
The summary of the models tested in portion as shown in [Table 6] at α =0.05, the calculated values of F is less than its critical values. One can conclude that the terms X1 × 2, , and do not contribute significantly to the prediction of dependent responses. The regression analysis of the response Y1 showed that the coefficient b1 bears a negative sign, whereas b2 bears a positive sign. One can conclude that increasing the concentration of X1 (Surfactant) is expected to decrease the response value, and increasing the concentration of X2 (Castor oil) is expected to increase the response value. The regression analysis of the response Y2 showed that the coefficient b_{1} and b_{2} bears a positive sign.{Table 6}
Analysis of variance for analysis for emulsification time
The Model F value of 615.57 implies the model is significant. There is only a 0.01% chance that a 'Model Fvalue' this large could occur due to noise.
Values of 'Prob > F' <0.05 specifies significant model terms.
As per this case, A and B are significant model terms.
The final equation in terms of the fundamental factor is.
Emulsification time = 47.333.80 × X_{1 }+_{ }12.63 × X_{2}.
The above equations are in format, Y = b_{0} + b_{1 }×_{ }1 + b_{2 }×_{ }2. Where Y is the dependent variable, b0 is the arithmetic mean response of the nine runs and b_{i} (b_{1}, b_{2}) is the estimated coefficient for the corresponding factor Xi (X1, X2), which represents the average results of changing one factor at a time from its low to high value. The interaction term (X1 × 2) depicts the changes in the response when two factors are simultaneously changed.
Analysis of variance for analysis for % CD release of selfmicro emulsifying drug delivery system
The Model Fvalue of 42.48 implies the model is significant. There is only a 0.01% chance that a 'Model Fvalue'' this large could occur due to noise.
Values of ''Prob > F'' <0.0500 specifies significant model terms.
As per this case, A and B are significant model terms.
The final equation in terms of the fundamental factor is.
% drug release of SMEDDS = 90.27 + 6.34 × X_{1 }+_{ }1.92 × X_{2}.
The above equations are in format, Y = b_{0}+b_{1}×1+b_{2}×_{2}. Where,
Y = dependent variable,
b0 = arithmetic mean response of the nine runs,
and bi (b_{1}, b_{2}) = estimated coefficient for the corresponding factor Xi (X1, X2), which indicates the results in an average of changing one factor at a time ranging from its low to high value. The interaction term (X1 × 2) depicts the changes in the response when two factors are simultaneously changed. The summary of regression analysis of measured response is shown in [Table 5] and the summary of models tests in portions shown in [Table 6].
Analysis through model graphs
 Effect of experimental variables on the emulsification time
 [Figure 4] shows the contour plot and surface plot for response Y1 (emulsification time) which illustrates the effect of tween 80 and castor oil on the emulsification time of SMEDDS. It demonstrates that the emulsification time of SMEDDS depends on factor A (tween 80) and factor B (castor oil). From the plot, we observed that both factor A and factor B have the most significant effect on the emulsification time of SMEDDS
 Effect of experimental variables on the % CD release SMEDDS.
 Figure 4: Contour plot (a) and surface plot (b) for response Y_{1} (emulsification time)
Click here to view 
[Figure 5] shows the contour plot and surface plot for response Y2 (% CD release) which illustrates the effect of tween 80 and castor oil on the % CD release of SMEDDS. It demonstrates that the % CD release of SMEDDS depends on factor A (tween 80) and factor B (castor oil). From the plot, we observed that both factor A and factor B have the most significant effect in % CD release of SMEDDS.  Figure 5: Contour plot (a) and surface plot (b) for response Y_{2} (% Cumulative drug release)
Click here to view 
Discussion   
Ternary phase diagram
From the results, we deduced that an increase in surfactant concentration increased the clarity of the produced emulsion. [Figure 2] shows corresponding ternary phase diagrams of different combinations. It offers various grades of SMEDDS along with recorded observation.
Optical transparency
No traces of undissolved drugs or other solid ingredients were found, indicating optically transparent formulation.
Robustness to dilution
There were no phase separation and precipitation observed in any of the diluted formulations after 24 h.
Rheological measurement
The viscosity of all formulation is shown in [Table 4].
Emulsification time
It is possible that it has least globule size and rapid dissolution.
Per cent drug content
The per cent drug content of all the formulations were in the range of 63%–83.7%.
Globule size and polydispersity index measurement by Zetasizer
From the obtained result of globule size, one can conclude that all the formulations were forming clear transparent microemulsion after dilution. [Table 4] shows PDI and Globule size data. PDI of all the formulations ranges from 0.153 to 0.376, indicating the uniform distribution of the formulations globules.
Zeta potential
The optimum range of zeta potential obtained is due to charges on the microemulsion, which makes SMEDDS formulation charged. We found the zeta potential value in a range of stable formulation.
In vitro drug release
An increase in the proportion of oil with surfactant increases the formation of smaller sized droplets. The maximum surface area of the formulation was available to contact the dissolution medium, which resulted in an increased maximum release. We showed the per cent cumulative drug release (%CDR) of the different formulations in [Figure 3] (a. B1B5 and b. B6B9 and PTX). From [Figure 3]b, one can easily conclude that the solubilisation of PTX powder in the dissolution medium was very slow. However, compared to this, PTX incorporated in SMEDDS show high dissolution, which is due to an increase in solubility of PTX in the dissolution medium.
Optimization
We can conclude that the concentration of X1 (Surfactant) is expected to increase the value of the response, and increasing the concentration of X2 (Castor oil) is expected to increase the value of the response.
Analysis of variance for analysis for emulsification time
The summary of regression analysis of measured responses is shown in [Table 5], and the summary of model tests in portions is shown in [Table 6].
Analysis of variance for analysis for % CD release of selfmicroemulsifying drug delivery system
The summary of regression analysis of measured response is shown in [Table 5], and the summary of model tests in portions shown is in [Table 6].
Analysis through model graphs
Effect of experimental variables on the emulsification time
From the plot, it was observed that both factor A and factor B have most significant effect in emulsification time of SMEDDS.
Effect of experimental variables on the % CD release selfmicroemulsifying drug delivery system
From the plot, it was observed that both factor A and factor B have most significant effect in % CD release of SMEDDS.
Conclusion   
SMEDDS increases the solubility of low aqueous soluble and high lipophilic drugs, which enhances the drug's availability in the formulation. Castor oil, tween 80 and Transcutol were selected as oil, surfactant and cosurfactant, respectively, depending on the highest solubility of PTX in it rather than water. The statistical optimization technique is used to study the effect of variables on the resulting characteristics factorial design approach. Nine preliminary batches obtained by using 3^{2} factorial design leads to optimized concentration of the factors. The dilution study indicates formulation was stable over the storage period. Optical transparency study for all batches was found satisfactory, indicating precise, transparent formation. Emulsification occurs in less time, indicating the ability to emulsify rapidly. We found the per cent drug content within pharmacopoeial limits and obtained an excellent cumulative drug release in all batches. After the application and treatment of dissolution data into various kinetic models, the release behavior follows the Higuchi kinetics of drug release. ANOVA test shows the independent factors have a significant influence on dependent variables and response. Amongst all formulations, batch B9 was selected as an optimised formulation based on lesser emulsification time and higher per cent cumulative drug release. Droplet size found in the nanometer range provides a maximum surface area, improving the dissolution and bioavailability. The PDI and zeta potential indicate the stability of SMEDDS.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
References   
1.  Mei L, Zhang Z, Zhao L, Huang L, Yang X, Tang J, et al. Pharmaceutical nanotechnology for oral delivery of anticancer drugs. Adv Drug Deliv Rev 2013;65:88090. 
2.  Ding Z, Wang L, Xing Y, Zhao Y, Wang Z, Han J. Enhanced oral bioavailability of celecoxib nanocrystalline solid dispersion based on wet media milling technique: Formulation, optimization and in vitro/in vivo evaluation. Pharmaceutics 2019;11:E328. 
3.  Sawatdee S, Atipairin A, Sae Yoon A, Srichana T, Changsan N, Suwandecha T. Formulation development of albendazoleloaded selfmicroemulsifying chewable tablets to enhance dissolution and bioavailability. Pharmaceutics 2019;11:E134. 
4.  Quan G, Niu B, Singh V, Zhou Y, Wu CY, Pan X, et al. Supersaturable solid selfmicroemulsifying drug delivery system: Precipitation inhibition and bioavailability enhancement. Int J Nanomedicine 2017;12:880111. 
5.  Qiao J, Ji D, Sun S, Zhang G, Liu X, Sun B, et al. Oral bioavailability and lymphatic transport of pueraria flavoneloaded selfemulsifying drugdelivery systems Containing sodium taurocholate in rats. Pharmaceutics 2018;10:E147. 
6.  Elfiyani R, Amalia A, Pratama SY. Effect of using the combination of tween 80 and ethanol on the forming and physical stability of microemulsion of Eucalyptus Oil as antibacterial. J. Young Pharm 2017;9:14. 
7.  Choi YH, Han HK. Nanomedicines: Current status and future perspectives in aspect of drug delivery and pharmacokinetics. J Pharm Investig 2018;48:4360. 
8.  Sun Y, Yu B, Wang G, Wu Y, Zhang X, Chen Y, et al. Enhanced antitumor efficacy of Vitamin E TPGSemulsified PLGA nanoparticles for delivery of paclitaxel. Colloids Surf B Biointerfaces 2014;123:71623. 
9.  Choi JS, Cho NH, Kim DH, Park JS. Comparison of paclitaxel solid dispersion and polymeric micelles for improved oral bioavailability and in vitro anticancer effects. Mater Sci Eng C Mater Biol Appl 2019;100:24759. 
10.  Cho HJ, Lee DW, Marasini N, Poudel BK, Kim JH, Ramasamy T, et al. Optimization of selfmicroemulsifying drug delivery system for telmisartan using BoxBehnken design and desirability function. J Pharm Pharmacol 2013;65:144050. 
11.  Holm R, Jensen IH, Sonnergaard J. Optimization of selfmicroemulsifying drug delivery systems (SMEDDS) using a Doptimal design and the desirability function. Drug Dev Ind Pharm 2006;32:102532. 
12.  Kamboj S, Rana V. Qualitybydesign based development of a selfmicroemulsifying drug delivery system to reduce the effect of food on Nelfinavir mesylate. Int J Pharm 2016;501:31125. 
13.  Parakh DR, Patil MP, Sonawane SS, Kshirsagar SJ. Application of factorial design approach in development and evaluation of self microemulsifying drug delivery system (SMEDDS) of mebendazole. J Pharm Investig 2016;47:50719. 
14.  Shen H, Zhong M. Preparation and evaluation of selfmicroemulsifying drug delivery systems (SMEDDS) containing atorvastatin. J Pharm Pharmacol 2006;58:118391. 
15.  Kim DS, Cho JH, Park JH, Kim JS, Song ES, Kwon J, et al. Selfmicroemulsifying drug delivery system (SMEDDS) for improved oral delivery and photostability of methotrexate. Int J Nanomedicine 2019;14:494960. 
16.  Kommuru TR, Gurley B, Khan MA, Reddy IK. Selfemulsifying drug delivery systems (SEDDS) of coenzyme Q10: Formulation development and bioavailability assessment. Int J Pharm 2001;212:23346. 
17.  Wei L, Sun P, Nie S, Pan W. Preparation and evaluation of SEDDS and SMEDDS containing carvedilol. Drug Dev Ind Pharm 2005;31:78594. 
18.  Deshmukh A, Kulkarni S. Solid selfmicroemulsifying drug delivery system of ritonavir. Drug Dev Ind Pharm 2014;40:47787. 
19.  Goddeeris C, Cuppo F, Reynaers H, Bouwman WG, Van den Mooter G. Light scattering measurements on microemulsions: Estimation of droplet sizes. Int J Pharm 2006;312:18795. 
20.  Venkatesh G, Majid MI, Mansor SM, Nair NK, Croft SL, Navaratnam V . In vitro and in vivo evaluation of selfmicroemulsifying drug delivery system of buparvaquone. Drug Dev Ind Pharm 2010;36:73545. 
21.  Bahadur S, Roy A, Baghel P, Chanda R. Formulation of glipizide tablets using fenugreek seed mucilage: Optimization by factorial design. Asian J Pharm 2016;10:S6628. 
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5]
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5]
