The preparation of co-crystal using ketoprofen and fumaric acid was optimized using 2-factor, 3 level CCD. The concentration of ketoprofen (X1) and concentration of fumaric acid (X2) were designated as formulation variables whereas the % drug release and solubility (µg/ml) were picked as response variables. The TDC of different batches of ket-FACo was found to be between 94.48 to 97.81 %, thus depicting that handsome amount of drug has been loaded and also no physical changes were observed during stability studies and even after six months. In different batches of ket-FACo solubility in phosphate buffer (PBS pH-7.4) values range from 29.08 to 10.63 µg/ml whereas ketoprofen solubility was found to be 3.64 µg/ml.
As presented in Table 1, solubility of ket-FACo varied in the range of 30.68-57.44µg/ml. The pure ketoprofen dispensed a solubility of 11.24 µg/ml in water at room temperature.
Table 1. Formulation parameters and responses for central composite experimental design.
Batch
|
Conc. Of Ketoprofen (mg)
|
Conc. Of Fumaric acid (mg)
|
Solubility (µg/ml)
|
Solubility in PBS pH-7.4 (µg/ml)
|
∆G (KJ/MOL)
|
% Drug release in 60 min.
|
Drug content (% )
|
F1
|
254.29
|
116.07
|
30.68±0.01
|
10.63±0.32
|
-3.49
|
62.21±0.03
|
95.89±0.06
|
F2
|
508.58
|
116.07
|
42.04±0.03
|
18.36±0.12
|
-1.92
|
70.01±0.05
|
95.66±0.01
|
F3
|
254.29
|
232.14
|
48.3±0.01
|
23.39±0.02
|
-3.04
|
71.94±0.02
|
95.33±0.04
|
F4
|
508.58
|
232.14
|
57.44±0.05
|
29.08±08
|
-2.68
|
83.68±0.01
|
97.81±0.01
|
F5
|
254.29
|
174.10
|
45.97±0.2
|
21.59±0.22
|
-2.54
|
70.52±0.06
|
96.39±0.06
|
F6
|
508.58
|
174.10
|
51.08±0.3
|
28.46±0.13
|
-3.19
|
80.59±0.07
|
95.60±001
|
F7
|
381.43
|
116.07
|
36.47±0.1
|
15.16±0.41
|
-3.32
|
66.52±0.02
|
95.04±0.02
|
F8
|
381.43
|
232.14
|
53.66±0.09
|
28.52±0.17
|
-2.93
|
79.71±0.04
|
97.47±0.07
|
F9
|
381.43
|
174.10
|
49.08±0.08
|
27.46±0.27
|
-3.36
|
75.54±0.05
|
97.25±0.03
|
F10
|
381.43
|
174.10
|
48.14±0.4
|
26.94±0.31
|
-3.08
|
75.04±0.2
|
95.53±0.05
|
F11
|
381.43
|
174.10
|
49.51±0.04
|
27.67±0.29
|
-3.04
|
75.59±0.3
|
94.88±0.01
|
F12
|
381.43
|
174.10
|
50.23±0.06
|
26.54±1.1
|
-3.15
|
73.52±0.1
|
94.48±0.06
|
F13
|
381.43
|
174.10
|
50.16±0.03
|
26.83±0.11
|
-3.12
|
73.02±0.03
|
96.35±0.08
|
Ketoprofen
|
|
|
11.24±0.03
|
3.64±0.14
|
|
23.05±0.01
|
|
All values are expressed as mean ± S.D., n=3.
Table 1 shows the results of solubility of different batches of Ket-FACo organized according to design protocol. The responses produced were fitted into several polynomial models using CCD. The response solubility was fitted greatest into quadratic model with none transformation of the data. The polynomial models for the responses solubility (Y1) can also be expressed by the equation (8) with determination correlation (R2) of 0.9659.
Y1 = 49.36+4.27X1+8.37X2-0.5550X1X2-0.6822X12-4.14X22 (8)
Table 2 summarizes the results of ANOVA on the solubility and % drug release response surface model, demonstrated that model was found significant with lack of fit as non-significant. Adequate precision of solubility is found to be 28.33 indicates an adequate signal. The adequate precision measuring signal to noise ratio (greater than 4) is desirable. Fig. 1 (a) show the collective effect of concentration of ketoprofen and fumaric acid on solubility. It may be reckoned from the plots that a curvilinear relationship exists between independent and dependent variables. It is also inferred from the plot that higher level of ketoprofen and fumaric acid results in increase in solubility. However, the effect of the concentration of fumaric acid (X2) seems to be more pronounced as compared to the concentration of ketoprofen (X1). This increase in solubility may be due to formation of soluble complex between ketoprofen and fumaric acid. Fumaric acid presents higher solubility than the drug that comes out of the crystal lattice. The drug in co-crystal get supersaturated in aqueous medium and possesses more energy as compared to crystalline phase, thereby, exhibit marked increase in solubility than the pure drug.
In vitro drug release
Results of in-vitro drug release (Table 1) revealed that 62.21 to 83.68% and 23.50% of ketoprofen got released from different batches of Ket-FACo and pure drug solution respectively, in 1 h study. This rise in percentage drug release from Ket-FACo as compared to pure drug may be associated to the solubility data. As previously mentioned that the fumaric acid form the soluble complex with the drug and thereby increase the drug wettability that leads to a better solubility and further heads towards better rate of drug release. The adjusted polynomial equation obtained for the in-vitro drug release (Y2) is shown in equation 9 with determination correlation of adjusted R² of 0.958 and predicted R2 of 0.905. The predicted is in reasonable agreement with the adjusted R² with a difference less than 0.2.
Y2=74.84+4.93X1+6.10X2+0.985X1 X2 -0.0336X12 -2.47X22 (9)
Table 2 recapitulating the results of ANOVA of in-vitro release data on response surface model fitted best in quadratic model (after none transformation of the data). The responses observed were fitted into different polynomials models using the experimental design. Adequate precision of in-vitro drug release is found to be 25.57 indicates an adequate signal.
Table 2. Model summary statistics.
Model
|
Lack of Fit
|
Response factor(Y)
|
F-value
|
Prob.>F
|
R2
|
Adeq.Prec.
|
C.V (%)
|
F-value
|
Prob.>F
|
Y1
|
69.00
|
<0.0001
|
0.9659
|
28.33
|
2.79
|
4.09
|
0.1035
|
Y2
|
56.65
|
<0.0001
|
0.975
|
25.57
|
1.60
|
0.937
|
0.501
|
To attain stability a natural tendency to acquire minimum Gibbs energy is always there. All the values of ΔG are negative (Table 1) at all levels of carrier demonstrating spontaneity of drug solubilization process.
Optimization
The optimization equations 8 and 9, involving the response and independent factors were assembled based on a quadratic model. To the responses i.e. solubility and in-vitro drug release the desirability function was applied with constraints to obtain the higher level of both, the batch F4 comes out to be optimized batch. In this fashion, the formulation containing fumaric acid (228.82mg) as coformer and drug content (508.58 mg) with addition of ethanol, established the maximum desirability, was organized and evaluated.
The mathematical optimization tool with desirability method was employed to prepare co-crystal. The constraints of maximum solubility and maximum % release was imposed on independent variables for optimization. The parameters recommended by the design were concentration of ketoprofen (508.58 mg) & concentration of fumaric acid (228.82 mg) that provide co-crystal with solubility of 56.06 µg/ml (predicted value 56.63 µg/ml) and % drug release 83.35% (predicted value 84.22%). The closer agreement between predicted and observed values indicated the high prognostic ability of the model. Fig. 2 shows the in vitro release profile of ketoprofen as pure drug and optimized batch (F4) of co-crystal formulation.
The release rate data of ketoprofen from co-crystal and from drug solution was fitted into several kinetic models to estimate release kinetics and mechanism of drug release. The release rate data was found to be put best into Higuchi model (with R2 =0.985) of release kinetics. Further, the value of n=0.475 (0.43<n<0.85), release exponent of Korsemeyer and Peppas equation, indicated that the release of ketoprofen from co-crystal occurs by diffusion and erosion of the matrix.
Fourier Transform Infrared Spectroscopy (FT-IR) analysis
FTIR is an excellent analytical technique to study the deviations in the position caused by the vibration modes of the functional groups. This technique reveals the shift in characteristic peaks of drug and coformer due to co-crystal formation involving H bonding between the corresponding functional groups. The spectra of ketoprofen fig. 3 (a) showed characteristic absorption band at 2979.27 cm-1 due to –CH stretching. The peak appearing at 1697.76 cm-1 can be ascribed to -C=O stretching of acid while peak appearing at 1655.77 cm-1 is due to -C=O stretching of ketone. The absorption bands at 1598.67 cm-1 (-C=C= stretching), 1442.21 cm-1 (-C=O stretching of aromatic ring), 1420.59 cm-1 (-C-H deformation of -CH3 asymmetrical) and 1370.04 cm-1 (-C-H deformation of -CH3 symmetrical) also appeared. The FTIR spectra of fumaric acid exhibit peaks at 3084.26 cm-1, 1684.31 cm-1, 1422 cm-1 ascribed to the -O-H stretching, -C=O stretching vibration and –C-C aromatic stretching, respectively. The FTIR spectra of ket-FACo displayed that -C=O stretching of –COOH group of ketoprofen get shifted from 1697.76 cm-1 to 1667.30 cm-1 and the peak due to -C=O stretching of fumaric acid at 1655cm-1 got disappeared. Therefore, FT-IR analysis confirms the interactions occurring between ketoprofen and fumaric acid. These interactions are essentially hydrogen bonds between the carboxylic group of the fumaric acid and the main functional groups of ketoprofen (-C=O and -O-H) which are able to generate supramolecular heterosynthons.
Powder X-ray diffraction analysis (PXRD)
The powder X-ray diffraction (PXRD) pattern of ketoprofen, fumaric acid and optimized batch of ket-FACo illustrated in fig. 3 (b). The diffraction peaks (and Miller indices) at 2θ of 12.74 (100), 18.52 (200), 22.85 (211), 24.00 (221), 26.32 (222), 28.920 (300), 36.71 (322) and 22.86 (211), 28.92 (300), 29.49 (310), and 30.03 (311) showed crystalline structure of ketoprofen and fumaric acid respectively. The major diffraction peaks at 2θ, 18.57 (200), 23.27 (211), 28.01 (222) and 29.28 (300) were also observed in PXRD spectra of co-crystal that portrayed crystalline nature of resultant product. The distinctive PXRD pattern of the ket-FACo was distinguishable from ketoprofen and fumaric acid, this outcome specifies the formation of a new crystal phase [62].
Differential scanning calorimetry (DSC)
DSC thermograms [fig. 4 (a)], reported that pure ketoprofen showed a sharp endotermic peak at 94.5º that corresponds to its melting point. The peak at 280.4ºC attributed to melting point of fumaric acid. In the thermogram of the prepared co-crystal, peaks were found to be displaced and difference in intensity is also observed from that of its constitutional components indicating the occurrence of weak cohesive forces that bonded together by reversible hydrogen bonding, suggesting the development of co-crystal formation. The thermal behavior of the ket-FACo was prominent, with a different melting transition from that seen with either of the constitutional components; this recommends the formation of a new phase.
Scanning electron microscopy (SEM)
The SEM image of the optimized batch of ket-FACo [fig. 4(c)] depicted good crystalline characteristics. This crystalline character was reinforced by the XRD data, as discussed earlier. The voids over the surface of the co-crystal may brace the imbibition of the solvent and biological fluids and thereby proliferating the solubility and bioavaibility of ketoprofen as estimated.
Nuclear magnetic resonance (NMR) Spectroscopy
NMR spectroscopy is used to characterize the co-crystal by studying the chemical environment of their nuclei and hydrogen bonding and it also offers valuable information regarding interactions. In the NMR pattern of ket-FACo, the carbonyl carbon of ketoprofen corresponding to 196.46 ppm and140.13 ppm has shifted to 196.10 ppm and 142.22 ppm, respectively. A deviation in the carbonyl carbon of carboxylic group in fumaric acid shifted from 166.41 ppm and 140.51 ppm to 175.52 ppm and 137.45 ppm, respectively (Fig. 5). This suggests an interaction between alcoholic group of fumaric acid and –COOH group of ketoprofen in ket-FACo.
Computational studies
The constituents of the co-crystal interact through weak non covalent interactions (NCI) [63] and in order to determine the points of contacts between the ketoprofen and coformer fumaric acid, molecular electrostatic surface potential (MESP) analysis was performed and the extreme positive and negative values from MESP are displayed in fig. 6.
The magnitude of these values signifies that both H-bond donor and acceptor are present in the crystals of both compounds. MESP of ketoprofen exhibits both positive (+53.53 kcal/mol) and negative (-37.40 and -32.94 kcal/mol) extreme values thus proving that it can form intermolecular as well as intramolecular hydrogen bonds during process of co-crystalization. Similarly, fumaric acid displays positive (+69.87 and +69.85 kcal/mol) and negative (-33.5 and -33.46 kcal/mol) extreme values which confirms its hydrogen bonding capability. It has stronger hydrogen bond donor ability due to higher positive extreme value and on the other hand ketoprofen has higher hydrogen bond acceptor ability due to higher negative extreme value. According to Etter's rule [64], there is more probability of interactions between most polar parts of the molecules in a co-crystal. Therefore, these two compounds will pair in the co-crystal through hydrogen bonding. This pairing of these molecules was accomplished by combining two molecules and energy minimization of combined form. The interaction diagram of the molecules in combination is shown in fig. 7 (a, b).
Reduced density gradient (RDG) analysis [fig. 7 (a)] exhibits formation of two hydrogen bonds between two fumaric acid and ketoprofen. Two hydroxyl groups made these two hydrogen bonds (shown as blue colored discs) with two carbonyl oxygen atoms of ketoprofen. Further, van der Waals interactions can be observed between double bond region of fumaric acid and phenyl ring of ketoprofen (shown as green and brown color). The results of RDG calculations are also in line with the MESP predictions. The hydrogen bonding interactions between these two molecules were further confirmed by Hirshfeld surface mapped by electron density with promolecular approximation analysis [fig. 7 (b)]. This calculation shows three regions of high electron density; two are the same as found in RDG analysis while the third is between carbonyl oxygen of fumaric acid and phenyl hydrogen of ketoprofen. These observations are also in agreement with the findings of MESP analysis. The positive parts of one molecule are interacting with negative parts of another molecule.
The geometry of the both molecules individually as well as in combined form was minimized using MOPAC [65] with latest PM7 method choosing value of gnorm as 0.01 after optimization with molecular mechanics method. Molecular energy and other various properties of the minimized discrete molecules and co-crystal were calculated with Firefly [66] by density functional theory (DFT) taking 6-31G* basis set in B3LYP method. The MESP calculation, RDG analysis and Hirshfeld surface mapped by electron density with promolecular approximation [67] calculations were performed with Multiwfn 3.8 [68] and visualization was done with the help of VMD [69].
Biological evaluation of the co-crystal
In-vitro anti-inflammatory activity
The % protection from denaturation of protein is comparably plotted at different concentration of optimized batch of formulation and pure drug ketoprofen (fig. 8). Egg albumin protein denaturation method displayed concentration dependent anti-inflammatory activity by protecting the protein. Half maximal inhibitory concentration (IC50) values of ketoprofen and optimized formulation was calculated by nonlinear regression analysis. The IC50 values for pure drug ketoprofen and optimized formulation was observed to be 556.11 µg/ml and 327.33 µg/ml respectively. Thus it can be inferred that optimized co-crystal formulation is additionally effective as compared to pure drug in generating anti-inflammatory response.
In-vivo anti-inflammatory activity
The improvement in activity of ketoprofen and co-crystal formulation was comparatively assesed by the increase in paw volume of control groups. The paw edema volume (before and after drug administration) and % inhibition of edema at different time interval was convinced and displayed in Table 3. The ketoprofen and co-crystal showed inhibition of paw edema as 49.34±0.18% and 60.39±0.15 at the end of 5 h, respectively thus demonstrating quick onset of action by co-crystal in contrast with the pure drug ketoprofen.
Statical Analysis:- Data was compared by ANOVA followed by Tukey’s test. The p value is <0.0005 is considered as significant.
Table 3. Effect of ketoprofen and optimized batch of ket-FACo formulation on the paw edema induced by carrageenan in Wistar rats.
Time (min)
|
Paw volume
|
(mm)
|
|
Inibition (%)
|
|
|
Control
|
Pure drug
|
Co-crystal
|
Pure drug
|
Co-crystal
|
60
|
4.25±0.07
|
4.12±0.03*
|
4.02±0.09*
|
1.05±0.03
|
5.17±0.08#
|
120
|
4.54±0.10
|
3.86±0.06*
|
3.53±0.04*
|
10.97±0.01
|
16.65±0.09#
|
180
|
4.96±0.014
|
3.48±0.04*
|
3.22±0.06*
|
29.83±0.03
|
35.08±0.27#
|
240
|
5.49±0.06
|
3.31±0.05*
|
3.04±0.04*
|
39.70±0.10
|
44.62±0.03#
|
300
|
6.11±0.022
|
3.09±0.03*
|
2.42±0.035*
|
49.34±0.18
|
60.39±0.15#
|
All values are expressed as mean ± S.D., n=6. *Significant (p<0.05) compared to control. #Significant (p<0.05) compared to pure drug (ketoprofen).
Analgesic activity
The results of the % analgesic activity (PAA) of test, reference and control group are shown in Table 4. The PAA (equation 8) was comparatively evaluated for Ket-FACo and pure drug based on its potential to suppress pain. Ket-FACo showed significant effect in enhancing the pain thershold to a certain extent when compared to that of drug (ketoprofen), thus, stipulating that an improvement in solubility further tweaked the pharmacological response.
Table 4. % Analgesic effect of ketoprofen and optimized batch of Ket-FACo by tail flick method in mice.
Treatment
|
PAA
|
0h
|
1h
|
2h
|
3h
|
4h
|
5h
|
Standard
(ketoprofen)
|
0.06±0.020
|
31.30±0.71*
|
35.18±0.12*
|
42.36±0.17*
|
50.93±0.14*
|
71.96±0.17*
|
Test
(Ket-FACo)
|
0.09±0.021
|
32.14±0.24*
|
50.44±0.14*#
|
60.23±0.18*#
|
65.41±0.12*#
|
75.68±0.22*#
|
Control (vehicle)
|
0.01±0.012
|
0.91±0.19
|
1.32±0.02
|
0.83±0.10
|
1.1±0.20
|
0.021±0.03
|
All values are expressed as mean ± S.D., n=6.*Significant (p<0.05) compared to control.#Significant (p<0.05) compared to pure drug (ketoprofen).