Anal fissure nanocarrier oflercanidipine for enhancedtransdermal delivery: formulationoptimization, ex vivo and in vivoassessmentNafees Ahmad, Saima Amin, Yub Raj Neupane & Kanchan Kohli†
Hamdard University, Faculty of Pharmacy, Department of Pharmaceutics, New Delhi, India
Objective: Pathogenesis of anal fissure (AF) is associated with raised resting
anal pressures (RAP) involving contraction of smooth muscle. Therefore, the
drug delivery strategy should be customized to reduce this raised RAP. In
this investigation, in order to achieve this task, a transdermal nanoemulsion
(NE) gel of lercanidipine (LER) was developed and optimized to evaluate its
permeation ability and in vivo performance. Further, the same formulation
was explored for droplet size analysis, zeta potential measurement and
stability studies.
Methods: Pseudo-ternary phase diagram was constructed to determine NE
region. The NE was optimized (OPT) by employing three-factor, three-level
Box-Behnken design expert software; the independent variables decided
were composition of oil, Smix and water and dependent variables, that is,
responses were cumulative amount of drug permeated across rat abdominal
skin in 24 h (Q24), steady-state flux (Js) and viscosity. The in vivo efficacy was
assessed by measuring anorectal pressure in male Wistar rats.
Results: The OPT NE formulation, composed of Capryol 90 (12.70% w/w),
Cremophor EL (18.0% w/w), Transcutol HP (18.0% w/w) and water (60.00%)
w/w was found to have permeation flux of 60.27 µg/cm2/h, release of
1699.52 µg/24 h and 491.95 cP viscosity. In addition, a small average droplet
size (82.71 ± 9.96nm) and long-term stability at room temperature (1.666 years)
was observed. The in vivo investigation demonstrated direct evidence on
significant reduction (27.75%) in anorectal pressure over a period of 4 h.
Conclusion: These preliminary finding suggested that NE-based gel system of
LER may provide promising perspective in management of AF.
Anal fissure (AF), a prevailed complication among all age groups with equalextent [1], is a linear, longitudinal split in the lining of the distal anal canal thatextends from underneath the dentate line to the anal verge. Chronicity of this com-plication leads to severe and sharp anal pain during defecation and it might becomea benign disorder. Generally, it is associated with raised resting anal pressures (RAP)(> 90 mmHg) as is evident from previous studies that RAP in patients sufferingfrom AFs is significantly higher than in healthy persons attributable to increasedactivity of internal anal sphincter [2-4]. Therefore, the reduction of these RAP is areasonable therapy for this disease [5].
Smooth muscle relaxation is assumed to be a novel way bywhich > 60% of the AF patients may be cured with the topicaluse of the therapeutic agents as they have advantages over surgi-cal treatment in avoiding long-term complications with noneed of hospitalization [2]. Recently, several pharmacologicalagents have been explored that may effectively cause temporarysphincterotomy and heal most fissures leading to approximatelytwo-thirds of patients avoiding surgery in the current scenario.Also, this became the first option in patients with a high risk ofincontinence [6]. Smooth muscle relaxation has been tried witha variety of therapeutic agents such as glyceryl trinitrate, oraland topical diltiazem, lacidipine and nifedipine [7,8].Owing to the role of calcium ion (Ca2+) in smooth muscle
contraction, calcium channel blockers (CCB) reduce the con-tractility of cardiac and smooth muscle fibers by antagonizingthe cellular influx of Ca2+. This causes reduction in intracellu-lar Ca2+ concentrations and prevents activation of the myosinlight chain kinase required for smooth muscle cell contractiondue to less availability of Ca2+ for bonding with calmodulin.This is how CCB offer a means of reducing mean anorectalpressure and allowing the time for a fissure to heal [9,10]. Top-ical CCB have been claimed to have healed chronic AFs in65 -- 95% of patients [11]. The first clinical study on the effectsof CCB over RAP showed that pressure was reduced with sub-lingual nifedipine in both healthy volunteers and in patientshaving hypertonic sphincters [12]. Preliminary results from amulticenter study showed some benefit from the use of0.2% nifedipine gel twice daily in acute fissures [13]. Lercani-dipine (LER) is a third-generation CCB of the dihydropyri-dine group that selectively inhibits the transmembraneinflux of Ca2+ via L-type calcium channels with a greatereffect on vascular smooth muscle than on cardiac smoothmuscle [14-16].Choice of an appropriate vehicle for transdermal delivery was
an effective technique to maximize the flux across the skin intosystemic circulation [17,18]. Nanoemulsions (NEs) are assumedto be an ideal vehicle for drug delivery across skin as they offerlow skin irritation, large solubilizing capacity and high drug-loading potential, attributing an increased thermodynamicactivity and subsequently increased partitioning toward theskin. In addition, ingredients of NE act as permeation enhancerthat may destroy the structure of stratum corneum (SC) leadingto increases in flux of drug via skin [19-23]. Moreover, it providesall the possible requirements of a liquid system, including easyformation, low viscosity with Newtonian behavior and verysmall droplet size (50 -- 200 nm) that confer NEs large surface-to-volume ratio, favoring close contact with the skin and creat-ing high concentration gradient to improve substrate perme-ation. Also, low surface tension ensures better adherence tothe skin and the dispersed phase may act as a reservoir to trans-port bioactive molecules in a more controlled fashion [24]. How-ever, the low viscosity of NE limits its clinical application due toinconvenience of use. Previous report illustrated that hydrogelssuch as Carbopol 940 and xanthan gum can thicken microe-mulsions without the loss of stability and permeation rates
but still there is lack of information (e.g., stability, permeationrate) regarding the influence of the hydrogel addition on NE,even though some hydrogels as thickeners have recently beenreported to change the rheological properties of NE for topicaldelivery [20,25-27].
Optimization of NE formulations in a systematic mannerfor different variables by applying Box-Behnken design(BBD) was made to reveal any synergism among the selectedvariables. Also, it yields the most promising NE formulationswith advantages of economics in terms of time, money anddevelopmental effort as the formulation of a NE systemdepends on the composition of rational blends of oil, Smixand water [28]. Hence, application of BBD was envisaged asmost exigent for this prospect.
The present study mainly focused on the preparation, opti-mization, characterization, stability, permeation potential andin vivo efficacy of developed NE gel (NEG) formulation.
2. Materials and methods
2.1 MaterialsLER was received as a gift sample from Glenmark Pharma-ceuticals Ltd, batch number A22026013, Mumbai, and certi-fied to contain 99.81% purity. The following excipients weregift sample from Gattefosse (Mumbai, India): Transcutol HP(diethylene glycol monoethyl ether), Capryol 90 (propyleneglycol monocaprylate). Cremophor EL (polyethoxylated cas-tor oil) was obtained from BASF (Mumbai). Polyethylene gly-col (PEG) 400 and diethyl ether were bought from Merck(Mumbai, India). Dimethyl sulfoxide (DMSO), dimethylfor-mamide (DMF) and phosphate-buffered saline (PBS, pH 7.4)were purchased from Sigma--Aldrich. Ketamine, xylazine andsaline water were purchased from local pharmacy shop.Sodium azide and hydroxypropyl methylcellulose (HPMCK4m) were procured as gift samples from Arbro Pharmaceut-icals (New Delhi, India). Carbopol 940 was obtained fromRanbaxy Research Laboratories (Gurgaon, Haryana, India).Standard capsaicin was obtained as gift sample from KentzHealthcare Ltd (Delhi, India). Isopropyl alcohol was pur-chased from SISCO Research Laboratories (Mumbai, India).Deionized water was obtained in the laboratory, using ionicinterchanged columns Milli-Q (Millipore).
2.2 Methods2.2.1 Pseudo-ternary phase diagramSolubility of LER was determined in various oils, surfactantsand cosurfactants of GRAS category and was published inour previous paper [29]. Based on that, pseudo-ternary phasediagrams were constructed by aqueous titration method todetermine the concentration of each component for existingNE region. Surfactant and cosurfactant were mixed in differentvolume fractions to give Smix ratios of 1:0, 1:1, 1:2, 1:3, 2:1,3:1 and 4:1. For each phase diagram, oil and specific Smix ratiowere mixed thoroughly in 16 different combinations from1:9 to 9:1 (1:9, 1:8, 1:7, 1:6, 1:5, 1:4, 1:3.5, 1:3, 1:2.33, 1:2,
1:1.5, 1:1, 1:0.66, 1:0.43, 1:0.25 and 1:0.11) to define theboundaries of phases precisely formed in the phase diagrams.After each 5% addition of the aqueous phase to the oil andSmix mixture, visual observation was made and recorded toassign a transparent and easily flowable oil/water NE [29].
2.2.2 Preparation of full thickness rat skinInstitutional animal ethics committee of Hamdard Universitypermitted the protocols for preparation of full thickness malealbino Wistar rat (200 -- 250 g) skin. After sacrificing the ratswith excess ether inhalation, hair on abdominal surface wasremoved with hair clipper taking extreme precaution not todamage the skin. The shaved skin was then scissored outfrom the animals and subcutaneous tissue was surgicallyremoved while the dermal side was cleaned with cotton swabsoaked in isopropyl alcohol to remove adhering fat [30,31].The prepared skin was washed immediately with PBS,wrapped in aluminum foil and stored at -20�C till further use.
2.2.3 Ex vivo skin permeation studiesFranz diffusion cell with a surface area of 0.785 cm2 was usedfor ex vivo skin permeation studies. The rat skin was mountedin between the donor and receptor chamber of the diffusioncell with SC facing the donor chamber. NE (1 ml) was placedin the donor chamber and 10 ml of PBS (pH 7.4) containing40% (v/v) of PEG 400 was filled in receptor chambercontaining 0.02% w/v of sodium azide to arrest microbialgrowth. PEG 400 was assimilated to maintain sink condi-tions. The contents of receptor chamber was agitated at400 rpm and positioned over a multi-magnetic stirrer (Cin-tex, Mumbai, India). The study was conducted at 37 ± 2�Cand samples of 1 ml were collected at predeterminedtime points and replaced with PBS (pH 7.4) containing40% (v/v) PEG 400. The drawn sample was analyzed forcumulative amount of LER permeated by a validated andwell-calibrated UV spectrophotometric method (ShimadzuUV-1601, Japan) at 240 nm wavelength [30] and concentra-tion was corrected for sampling effects according to followingequation [32,33]:
(1)
A-n
1n
t
t s
n1
1
n 1A
V
V VAA
==⎛
⎝⎜
⎞
⎠⎟⎛
⎝⎜⎜
⎞
⎠⎟⎟
−
−
where An1 is the corrected concentration of the nth sample, An
is the measured concentration of LER in the nth sample, An1-1
is the corrected concentration in the (n - 1)th sample, Cn - 1 isthe measured concentration of the LER in the (n - 1)th sam-ple, Vt is the total volume of the receiver fluid and Vs is thevolume of the sample drawn.
2.2.4 Data analysisThe cumulative amount of drug permeated across a unit areaof skin was plotted as a function of time. Steady-state flux(Jss) and lag time (tlag) were obtained from the slope andthe x-intercept of the linear portion, respectively [34,35].
Permeability coefficient (KP) was calculated by the followingequations:
(2)
K J / CP S d=
where Cd represents drug concentration in donor compart-ment. To compare the permeation enhancement capacitiesof excipients used in NE, enhancement ratio (ER) was calcu-lated as follows:ER = drug permeability coefficient or flux after enhancertreatment/drug permeability coefficient or flux beforeenhancer treatment.
ER indicates the effect of vehicle on permeation ofLER [35]. Data of this investigation was statistically analyzedby applying one-way ANOVA (Tukey’s multiple compari-sons) at a significance level of p < 0.05.
The target flux was calculated using the followingequation [33]:
(3)
Target flux Css A Clt BW= × ×( ) /
where, Css denotes LER concentration at therapeutic level(21.25 µg/l) and Clt is the total body clearance: 0.97 l/h/kg(calculated from volume of distribution, 1.120 l/kg and half-life of 4.3 h) [36], BW is the standard body weight of rats whichwere used in the study (0.250 kg), A represents the surface areaof the Franz diffusion cell (i.e., 0.785 cm2). The calculated tar-get flux value for LER was 6.56 µg cm-2 h-1.
2.2.5 Optimization of NE formulation using BBD
softwareEach NE formulation suggested by BBD software was pre-pared by spontaneous emulsification method in which 0.2%w/w of LER was appropriately dissolved in oil phase and therequired quantity of Smix was added to it with stirring at vor-tex mixer (Remi, India) for 10 min. The aqueous phase wasthen added slowly with continuous stirring for another10 min to get a clear transparent NE. A three-factor, three-level BBD expert (Version 8.0.6.1, Stat-Ease, Inc., Minneap-olis, MN, USA) software was used to statistically optimize theformulation factors. Quadratic response surface model wasexplored along with generation of second-order polynomialequation to evaluate main effects, interaction effects and qua-dratic effects on the amount of LER permeated in 24 h (Q24),flux and viscosity. A design matrix comprising 17 experimen-tal runs was constructed. The nonlinear computer-generatedquadratic model is given as:
(4)Y b b X b X b X b X X b X X
b X X b X b X b
2 1 2 2 3 3 12 1 2 13 1 3
23 2 3 11 12
2 22
= + + + + +
+ + + +
0
333 32X
where Y is the measured response associated with each factorlevel combination; b0 is an intercept; b1 to b33 are regressioncoefficients computed from the observed experimental valuesof Y and X1, X2 and X3 are the coded levels of independentvariables. The terms X1, X2 and X3 (i = 1, 2 or 3) represent
AF nanocarrier of LER for enhanced transdermal delivery
the interaction and quadratic terms, respectively [33,37,38]. Theselected dependent and independent variables are shownin Table 1 along with their low, medium and high levels whichwere determined from pseudo-ternary phase diagrams. Statis-tical validation of the polynomial equations generated bydesign expert was established based on the ANOVA provisionin the software. The models were evaluated in terms of statis-tically significant coefficients and R2 values. Various feasibil-ity and grid searches were conducted to find the optimumparameters.
2.3 Characterization
2.3.1 Percentage transmittanceAbout 1 ml of the NE formulation was diluted 100 timesusing double-distilled water and analyzed by UVs spectropho-tometer (Shimadzu UV-1601, Japan) at 500 nm usingdouble-distilled water as blank [29,39].
2.3.2 Refractive indexRefractive index of NE formulations was determined by anAbbes type refractometer (Precision Standard Testing Equip-ment Corp., Germany) [39].
2.3.3 Drug content determinationWeighed amount of formulations were assayed to determinethe drug content by dissolving it in methanol and stirredby vortex mixer. The solutions were diluted to followLambert--Beer law [29]. The filtered solutions were analyzedby an in-house validated HPLC method. The apparatus con-sisted of Agilent HPLC (1120 series) binary pump systemwith UV detector and a C18 reverse phase column [(AgilentTC C18 (2), 250 � 4.6 mm), particle size 5 µm, Agilent,Switzerland] equipped with a guard column of same packingmaterial with a provision of temperature control and anonline degasser. EZChrom Elite software was used for the col-lection and integration of data. Mobile phase consisted ofmethanol with Millipore water (90:10 v/v) at 1.2 ml/minflow rate, detection at 240 nm with retention time at5.53 min.
2.3.4 Droplet size analysis (particle size distribution)About 0.1 ml of NE formulation was dispersed in 50 ml dis-tilled water and the globule size so dispersed was determinedby dynamic light scattering (DLS) technique using Zetasizer(Zetasizer Ver. 6.01, Malvern Instruments, UK). All measure-ments were done in triplicate using disposable polystyrenecuvettes [29,39,40].
Table 1. Variables and observed responses in BBD for NE formulations.
Formulation
code
Independent
variables
Dependent variables tlag KP ER % DC RI ± SD % T ± SD
2.3.5 Zeta potentialZeta potential was measured by photon correlation spectros-copy using Zetasizer (Nano ZS, Malvern Instruments, UK)which measures the potential ranged from -120 to 120 V.NE formulation (300 mg) was diluted 100 times usingdouble-distilled water and was analyzed for zeta potential.All measurements were carried out at 25�C in triplicate [29,39].
2.3.6 Transmission electron microscopic analysisTransmission electron microscopic (TEM) analysis was doneto determine the shape of the dispersed oil droplets afterdilution of NE with distilled water (1:100). A drop of dilutedNE was applied to a 300 mesh copper grid and left for 1 min.After this, the grid was kept inverted and a drop of phospho-tungstic acid (PTA, 2% w/v) was applied to the grid for 10 s.Excess of PTA was removed by absorbing on a filter paper andthe grid was analyzed by JEM-2100F (JEOL, USA) operatedat 200 kV with AMT image capture engine software [29,39,41].
2.3.7 Viscosity measurementViscosity of the NE (600 µl) formulation was determined atspeed of 70 rpm by R/S CPS plus Rheometer instilled withRHEO3000 software (Brookfield Engineering Laboratories,Inc., Middleboro, MA, USA) over 25 ± 0.5�C with spindleC 50-1 (50 mm diameter). Wait time for the operation was60 s [29,39].
2.4 In vivo efficacy of NEG formulationThe in vivo efficacy of developed formulation of LER wasevaluated in capsaicin-induced AF model [42] (CPCSEAHamdard university, Protocol approval number 808) by mea-suring anorectal pressure over a period of 4 h at differentintervals in anaesthetized (intramuscular injection of ketamine90 mg/kg and xylaxine 9 mg/kg) male Wistar rats [43]. Thediuretic effect of anesthesia was offset by rehydration withintraperitoneal saline (2 ml/h). The rats were divided intofour groups of three animals in each. Rats in group 1 receivedinducer, group 2 received NEG, group 3 received conven-tional gel (CGEL) and group 4 was applied with placeboNEG (PLACEBO). Anorectal pressure was measured via awater-filled balloon made of latex rubber (5 � 10 mm size)connected through catheter to a transducer. The pressuresignal was recorded using Eddy’s instrument. After insertionof the balloon assembly, the system was allowed to reach astable baseline (generally between 1 ± 1.5 h) before initiationof the experiment.
2.4.1 Data treatment and statistical analysisThe area-under-the-contraction-waveforms (AUEC) per min-ute was recorded over 3 consecutive min, and these triplicatevalues were averaged to yield a single value of AUEC. Allthe obtained results were expressed as mean ± SD of threerats in each group. One-way ANOVA was done followed byTukey’s multiple comparison test and the level of significancewas set as p < 0.05 (Graph Prism Stat).
2.5 Stability studies and accelerated stability studyStability studies and accelerated stability studies on OPTNEG was done as per ICH Q1A (R2) guidelines. Stabilitystudy was done to check the stability of developed formula-tion at refrigerated temperature (8 ± 2�C), and room temper-ature (25 ± 2�C). Accelerated stability study was done topredict the shelf-life at room temperature.
3. Result and discussion
3.1 Selection of NE componentsThe screening of Cremophor EL as surfactant (safe, compati-ble, nonionic), Transcutol HP (cosurfactant) and Capryol90 (oil phase) were made to obtain a safe and effective NEformulation by solubility study. In addition, transcutol hasbeen studied previously for dermal and transdermal deliverybecause of its non-toxicity, biocompatibility with the skin,miscibility with polar and non-polar solvents and effectivesolubilizing ability for several therapeutic agents [44]. Further-more, Capryol 90 may solubilize large amount of LER thatcould maintain high concentration gradient toward skin, asprevious reports illustrated that significantly higher perme-ation flux may be achieved due to high solubilizing potentialthat in turn creates higher concentration gradient acrossskin [45,46].
3.2 Pseudo-ternary phase diagramFigure 1 illustrates the pseudo-ternary phase diagram fordifferent Smix ratios to achieve NE region. Dotted area repre-sents the isotropic and low-viscosity single-phase NE region,while remainder of the phase diagram represents the turbidregion, that is, multiphase conventional emulsions.
3.3 Ex vivo skin permeation studiesThe ex vivo skin permeation studies was conceived in a quest tooptimize the NE formulation based on steady-state fluxcalculation. The permeation profiles of LER NEs through ratskin were studied. The calculated permeation parameters(Table 1) such as Q24 (350.42 -- 1786.16 µg/24 h-1), JS(15.88 to 62.3 µg cm-2 h-1) and tlag (0.307 -- 5.898 h) for allexperimental formulations demonstrated that the permeationparameters of LER from NE were markedly influenced bythe composition of NE. The oil phase Capryol 90 selected inthe present work had remarkable influence over permeation.
Formulation BBD F6 illustrated the higher amount of LERpermeated (1786.16 3 µg) with a JS of 62.3 µg cm-2 h-1 and atlag of 4.41 h. LER solution (control) illustrated Q24 of317.81 µg, JS of 11.33 µg cm-2 h-1 and a tlag of 2.27 h. TheER of NE was 1.40- to 5.50-fold higher than control (LERwas dissolved in water containing 5% (v/v) DMF and 40%(v/v) PEG 400). These results demonstrated that NE had apotent enhancement effect on transdermal delivery.
AF nanocarrier of LER for enhanced transdermal delivery
3.4 Optimization of NE formulation by BBD softwareFrom the results of pseudo-ternary phase diagrams, Capryol90 (10 -- 16%), Smix (1:1, 36 -- 44%) and water content(50 -- 60%) that may create hydration effect on the SC toinfluence the permeation ability of formulations [47,48] wereselected as they may form clear NE with large surface areain these specified ranges. Also, they may disrupt thediffusional barrier of the SC by acting as permeationenhancers [49].For estimation of quantitative effects of the different com-
bination of factors and factor levels on Q24, flux and viscosity,the response surface models were calculated with design expertsoftware by applying coded values of factor levels. The modeldescribed may be represented as:
(5)
Y1 Flux 54 2 6 11 9975 X 4 1 812X
1 11938X 1 6 751 2
3
( ) = + − −
+ −
. . .
. .
0 00 0 0
0 0 0XX X
9 75 X X 4 34875X X
12 91363X 27988X
1 2
1 3 2 3
12
22
− −
− −
−
0 0 0
0
0
. .
. .
.558487X32
(6)
Y2 Q 1313 576 337 28375X
13 1275 X 84 1 625X
31
24 1
2 3
( ) = + −
− +
−
. .
. .
00
0 0 0
11 445 X X 49 9475 X X
112 45 X X 296 89675X
69 6
1 2 1 3
2 3 12
. .
. .
.
00 0
000
−
− −
+ 99575X 33 52825X22
32+ .
(7)
Y3 Viscosity 37 922 2 195 X
12 13 X 34 3525 X
7
1
2 3
( ) = + +
− +
+
0 00 00
000 0
. .
. .
77 8925 X X 8 8475 X X
3 6 75 X X 17 8 525X
1 3 1
1 2 1 3
2 3 12
. .
. .
.
0 0
0 0 0 0
0
+
+ +
+ 11525X 5 6475X22
32− .0
3.4.1 Fitting of model to dataFormulation BBD F6 illustrated a significantly higheramount of drug permeation (Y2, Q24) and higher flux (Y1)among the formulations. The responses observed for 17 for-mulations prepared were simultaneously fitted to first-order,second-order and quadratic models and it was observed that
100
80
60
40
20
0
OPT NE
NEG
CGEL
Control
% C
um
ula
tive
am
ou
nt
rele
ase
(mg
/cm
2 )
Time (h)E.
B.
W
1210864200.1 1 10 100 1000 10,000
A. Smix 1:1 Smix 1:2 Smix 2:1 Smix 1:3 Smix 3:1
O W O W O W O W O
C. D.
400,000
300,000
200,000
100,000
0
-200 -100 100 2000
0 5 10 15 20 25 30
Figure 1. Diagram showing pseudo-ternary phase diagram of different ratios for optimization of NE (A) with the
characterization parameters of OPT NE such as (B) droplet size distribution, (C) TEM, (D) zeta potential and (E) comparative
permeation profile of OPT NE with NEG, CGEL and control solution (p < 0.0001).
the best fit model was quadratic model as depicted by thecomparative values of R2, standard deviation and percentagecoefficient of variation (Table 2). The second-order polyno-mial regression equation generated for each response was rep-resented as Equations 5, 6 and 7. A positive value representsan effect that favors the optimization, while a negative valueindicates an inverse relationship between the factor and theresponse. It is evident that the independent variable X3
(water) had positive effect on the responses, Q24 (Y2) andflux (Y1) and viscosity (Y3).
3.4.2 Contour plots and response surface analysisTwo-dimensional contour plots and three-dimensionalresponse surface plots were constructed and the formulationwas OPT by superimposing these plots for the responses(Figure 2A). These graphs were very useful in studying theinteraction effects of the two factors over the responses bykeeping third factor at constant level at a time.
3.4.3 Data analysisUnder the design matrix evaluation for response surfacemodel, no aliases were found for quadratic model. Aliaseswere calculated based on response selection by consideringmissing datapoints, if necessary. The model F-value of85.89 and values of ‘Prob > F’ < 0.0500 indicated that modelterms were significant.
The value of the correlation coefficient (R2) of equationfor flux was found to be 0.9910 -- indicating good fit(Table 2).The ‘Predicted R-Squared’ of 0.8688 was in reason-able agreement with the ‘Adjusted R-Squared’ of 0.9795.‘Adequate Precision’ measures the signal-to-noise ratio.A ratio > 4 is desirable; the ratio of 34.977 (Table 2) indi-cated an adequate signal. This model can be used to navigatethe design space. The flux values measured for the differentformulations illustrated wide variation, that is, values rangedfrom a minimum of 15.88 µgcm-2h-1 (BBD F7) to a maxi-mum of 62.3 µgcm-2h-1 (BBD F6). The results clearly dem-onstrated that the flux value was strongly affected by thevariables selected for the study. The main effects of X1, X2,and X3 represented the average result of changing one vari-able at a time from its low level to its high level. The inter-action terms (X1X2, X1X3, X2X3, X
21, X
22 and X2
3) showedhow the flux changed when two variables were simulta-neously changed. The negative coefficients for all three inde-pendent variables indicated an unfavorable effect on the flux,while the positive coefficients for the interactions between
two variables designated a favorable effect on flux. Amongthe three independent variables, the lowest coefficient valueis for X1
2 (-12.91363), indicating the insignificancy of thisvariable in prediction of flux.
The value of R2 of equation for release was found to be0.9940 -- indicating well fit (Table 2). The ‘PredictedR-Squared’ of 0.9863 was in reasonable agreement with the‘Adjusted R-Squared’ of 0.9162. The model F-value of128.53 and values of ‘Prob > F’ < 0.0500 indicated thatmodel terms were significant. ‘Adequate Precision’, a measurefor the signal-to-noise ratio was found to be 45.766, reflectedan adequate signal. The Q24 measured for the different for-mulations illustrated wide variation (350.42 -- 1786.16 µg).The interaction terms (X1X2, X1X3, X2X3, X2
1, X22 and
X23) demonstrated that how the release altered when two
variables were simultaneously changed. The negative coeffi-cients for all three independent variables designated an unfa-vorable effect on the release, while the positive coefficientsfor the interactions between two variables pointed a favorableeffect on release. Among the three independent variables, thelowest coefficient value is for X1 (-337.28375), thus indicat-ing that this variable was insignificant in prediction of release.
3.4.4 OptimizationThe optimum formulation was selected based on the criteriaof attaining the maximum value of flux, maximum releaseand optimum value of viscosity by applying constraints onY1 (50 £ Y £ 65), Y2 (1300 £ Y £ 1800) and Y3 (350 £Y £ 550). On trading of various response variables andcomprehensive evaluation of feasibility and exhaustive gridsearch, the formulation composition with oil concentrationof 12.70%, Smix 36.00% and water 60.00% was found tofulfill the maximum requisite of an optimum formulationbecause of maximum Q24 (1699.52 µg), flux (60.27µgcm-2 h-1) and optimum viscosity (491.91cP) values. TheOPT formulation was formulated as NEG using (HPMCK4m) at 4% (w/v) and Carbopol 940 (0.75%) as gellingagent in 1:1 ratio.
3.4.5 Validation of applied BBDThe validity of the calculated optimal parameters and pre-dicted response was confirmed by preparing OPT formula-tions according to the above values of the factors (Table 3)and subjected to ex vivo skin permeation studies. The percent-age prediction error was below 5%, indicating that theobserved responses were very close to the predicted values.
Table 2. Summary of results of regression analysis for responses Y1, Y2 and Y3 for fitting to quadratic model.
S No. Quadratic model R2 Adjusted R2 Predicted R2 Adequate precision SD % CV
Percentage prediction error is helpful in establishing the valid-ity of generated equations and in describing the domain ofapplicability of response surface model. Linear correlationplots between the actual and the predicted response variables
are shown in Figure 2B. The linear correlation plots drawnbetween the predicted and experimental values demonstratedhigh values of R2 (Q24, 0.9868; flux, 0.9910; viscosity,0.9766), indicating goodness of fit.
Oil
Oil
Oil
FluxF
lux
Pre
dic
ted
Pre
dic
ted
Pre
dic
ted
Flu
x
Flu
x
Flux Flux
Oil
Sm
ix
Smix
SmixWater Water
SmixW
ater
Wat
er
a. b. c.
e. f. g.
1.0
0.5
0.0
-0.5
-1.0
1.0
0.5
0.0
-0.5
-1.0
1.00.50.0-0.5-1.0
70
600
350
0.9868 0.97662000
0.0
0.9910
Actual
B.
A.
Actual Actual
70
70
10
350.0 600.00.00 2000.07010
1
100
-1 -1
70
70
1
1
0 0
-1 -1
70
70
1
1
0 0
-1 -1
1.00.50.0-0.5-1.0
1.0
0.5
0.0
-0.5
-1.01.00.50.0-0.5-1.0
Figure 2. A. Contour plot showing the effect of (a) oil (X1) and Smix (X2), (b) oil (X1) and water (X3), (c) Smix (X2) and water (X3)
on response Y1 (flux), corresponding 3D response surface plots (d)--(f). B. Linear correlation plots (a, b, c) between actual and
3.5 Percentage transmittancePercentage transmittance of OPT NE close to 100% (Table 4)signified that it was clear and transparent, imparting eleganceand enhancing its acceptability by the patient. Further, it mayalso imply that the size of the globules in the NE formulationapproximates the nanometer range. This, in turn, indicatedthat the drug in the formulation has a large surface area forrelease and absorption in biological media [25].
3.6 Refractive indexRefractive index of the OPT NE formulations was observed(Table 4) as closer as of neat Capryol 90 (1.422 ± 0.002),thereby indicating its thermodynamic stability and isotropicnature. Similarity of the values of refractive index is a signof the uniform NE structure which, in turn, ensure dose uni-formity of the formulation and it may help in minimizingbioavailability variations during course of the therapy [25].
3.7 Droplet size analysis (particle size distribution)The globule size and poly dispersity index (PDI) of OPT NEwas found to be 82.71 ± 9.96 nm and 0.446 ± 0.024 (n = 3,p value < 0.001), respectively (Table 4). Due to their subcellu-lar and submicrometer size, NE is expected to penetratedeeper into tissue through fine capillaries. This would allowefficient delivery of therapeutic agent to target site in thebody [25,26].
3.8 Zeta potentialZeta potential signifies degree of repulsion between neighbor-ing molecules, like charged particles in dispersion; it can berelated to the stability of colloidal dispersions. Zeta potentialof OPT NE was observed to be -16.17 mV (Table 4). Negativevalues of zeta potential illustrated that the formulations were
negatively charged attributable to the anionic groups of thefatty acids and glycols present in the oil, surfactant and cosur-factant. Also, high values of zeta potential of the formulationsdenoted stability of the system. Thus, there are minimal chan-ces of coagulation or flocculation of the system in the biolog-ical environment and during its shelf-life [25].
3.9 TEM analysisSince TEM directly generates images at high resolution andcan capture any concomitant structures as well as microstruc-ture transitions, it is the most vital tool for the study of micro-structures [25,27]. It was evident from TEM that spherical NEdroplets emerged as dark and the surroundings were found tobe bright. The droplet sizes were in proximity with the resultsobtained using DLS.
3.10 In vivo efficacy of NEG formulationA pictorial scheme of the animal preparation is shownin Figure 3A and B and it shows the time course of LER effecton the anorectal pressure after treatments of developed LERNEG. It was evident that at NEG treatment, LER exhibitedimmediate, sustained as well as time-dependent effect, at leastat the dose studied which causes decreases in AUEC, indica-tive of reduction of in vivo anorectal pressure. During treat-ment, the developed LER NEG produced 27.75% reductionin anorectal pressure after 4 h of treatment when appliedintrarectally (about 1 mg/kg of body weight). The study wascarried out only for 4 h, since the anaesthetized preparationusually began to show instability after this duration. It wasalso observed that intrarectal LER infusion did not causeany apparent changes in mean arterial pressure (MAP) duringthe time course of study (Figure 3C, data not shown).
As the PLACEBO did not exhibit significance reduction inanorectal pressure in group receiving capsaicin in DMSO
Table 3. Composition of OPT NE formulations, the predicted and experimental values of response variables and
percentage prediction error.
Composition of OPT
NE (X1, X2, X3) (% w/w)
Responses Experimental value Predicted value Percentage
only, this clearly illustrated that excipient of NEG itself has nosignificant effect over anorectal pressure. The effects of LERNEG were essentially constant from the first point of mea-surement (30 min) to the end of the experiment (4 h). One-way ANOVA indicated that LER NEG produced effects sig-nificantly higher as compared to the, PLACEBO NEG andLER CGEL (p < 0.005) as well (Figure 4).
A limitation of the present study was that the fluid-filledballoon assembly for pressure measurement was relativelylarge, and its lateral apex was placed ~ 5 mm from the anus.Thus, the pressure tracings obtained were unlikely to bederived exclusively from the anal sphincter, and the effectsof LER on in vivo anal sphincter relaxation could not bedirectly obtained.
In conclusion, the pharmacodynamic studies demonstratedthat LER significantly inhibited anorectal contractions in theanaesthetized rat, over at least for 4 h.
3.11 Stability studies according to ICH Q1A (R2)
guidelines for OPT NEG formulationNEs have been known to enhance the physical as well as chem-ical stability of many drugs with no adverse effects [29]. There-fore, an attempt was made in present study to enhance physicalas well as chemical stability of LER using NEG formulation.The LER content in the stability samples after 90 days werefound to be 98.3% at room temperature in NEG. The resultsreveal that LER was chemically stable during the study.
During accelerated stability study, the degradation of LERwas very slow at each temperature which indicated its chemicalstability in the NEG formulation. During stability study, thechanges in observed parameters (viscosity, pH) were not foundto be statistically significant (p > 0.05) which indicated thatOPT formulation were stable. The shelf-life [50] of developedformulation was found to be 1.666 years at room temperature.
4. Conclusion
NE system with powerful permeation ability, good stabilityand in vivo efficacy was investigated for delivery of LERthrough skin. The present study conclusively illustrated theuse of a statistical BBD in prediction of Q24, flux and viscosityfor optimization of NEs. The derived polynomial equationsand contour plots aid in predicting the values of selected inde-pendent variables for preparation of optimum NE formula-tions with desired properties. Result of stability study revealedthat changes in observed parameters were not found to be sta-tistically significant (p > 0.05), which indicated that OPT for-mulation was stable both physically and chemically. Also, theshelf-life of developed formulation was found to be 1.666 yearsat room temperature. The pharmacodynamic studies demon-strated that LER significantly inhibited anorectal contractionsin the anaesthetized rat, at least over a period of 4 h.
So, the developed NEG system of LER might be a promis-ing prospective carrier for transdermal delivery of poorly sol-uble active molecule such as LER.
Acknowledgment
The authors wish to express their appreciation to F Ahmad,HIMSR Jamia Hamdard New Delhi, for providing laboratoryfor pharmacodynamic study. The authors also acknowledgethe support of Gattefosse (Mumbai, India) for providing dif-ferent excipients and A kumar from Aimil instrument fordetermining zeta potential in his laboratory.
Declaration of interest
The authors state no conflict of interest and have received nopayment in preparation of this manuscript.
Bibliography
1. Dykes SL, Madoff RD. Benign anorectal:
Anal fissure. In: Wolff BG, Fleshman JW,
Beck DE, Pemberton JW, Wexner SD,
editors. The ASCRS textbook of colon and
rectal surgery. Springer Science and
Business Media LLC, New York,
2009; p. 178-91
2. McCallion K, Gardiner KR. Progress in
the understanding treatment of chronic
anal fissure. Postgrad Med J
2001;77:753-8
3. Hancock BD. The internal sphincter and
anal fissure. Br J Surg 1977;64:92-5
4. Arabi Y, Alexander-Williams J,
Keighley MRB. Anal pressure in
hemorrhoids and anal fissure. Am J Surg
1977;134:608-10
5. Goligher J. Surgery of the anus, rectum and
colon. 3rd edition. Bialliere, Tindall; 1975
6. Garcıa-Granero E, Munoz-Forner E,
Mınguez M, et al. Treatment of chronic
anal fissure. Cir Esp 2005;78:24-7
7. De Nardi P, Ortolano E, Radaelli G.
Comparison of glycerine trinitrate and
botulinum toxin-a for the treatment of
chronic AF: long-term results.
Dis Colon Rectum 2006;49:427-32
8. Kassai M, Illenyi L, Horvath OP.
Current treatment of anal fissure.
Orv Hetil 2001;142:1565-8
9. Abraham AA, Debrah SA. Current
concepts in anal fissures. World J Surg
2006;30:2246-60
10. Collins EE, Lund JN. A review of
chronic anal fissure management.
Tech Coloproctol 2007;11:209-23
11. Perry WB, Dykes SL, Buie WD, et al.
Practice parameters for the management
of anal fissures. Dis Colon Rectum
2010;53:1110-15
12. Chrysos E, Xynos E, Tzovaras G, et al.
Effect of nifedipine on rectoanal motility.
Dis Colon Rectum 1996;39:212-16
13. Antropoli C, Perrotti P, Rubino M, et al.
Nifedipine for local use in conservative
treatment of anal fissures: preliminary
results of a multicenter study.
Dis Colon Rectum 1999;42:1011-15
14. Bianchi G, Passoni A, Griffini PL.
Effects of a new calcium antagonist on
cardiac contractility of conscious rabbits.
Pharmacol Res 1989;21:193-00
15. Bang LM, Chapman TM, Goa KL.
Lercanidipine: a review of its efficacy in
AF nanocarrier of LER for enhanced transdermal delivery
