Formulation, in vitro characterization and optimization of taste-masked orally disintegrating co-trimoxazole tablet by direct compression

Introduction Orally disintegrating tablet (ODT) is a dosage form that overcomes the problem of swallowing which is prevalent in about 35% of the general population. Co-trimoxazole (CTX) is given for patients with HIV for the prophylaxis of opportunistic infection (OI), commonly for pneumocystis carinii pneumonia. It was reported that CTX was associated with a 25–46% reduction in mortality among individuals infected with HIV in sub-Saharan Africa. Esophageal candidiasis which usually comes along with HIV/AIDS is one of AIDS defining illness affecting up to 1 in 5 of people with AIDS. This opportunistic illness is manifested by painful or difficulty of swallowing. In this respect, CTX ODT offer the advantages of both liquid dosage forms in terms of easy swallowing thereby improve patient compliance and solid dosage forms in terms of dose uniformity, stability, lower production, and transportation costs. The objective of this study was to formulate, characterize and optimize CTX ODT which could overcome swallowing problem and improve patient compliance. Co-trimoxazole ODTs were prepared by direct compression technique using a semi synthetic super disintegrant (crospovidone) along with other excipients. Two taste masking techniques were employed, addition of sweetening agent, and solid dispersion by using a pH sensitive polymer, Eudragit E-100 at different ratios (1:1, 1:2 and 1:3). Taste masking was determined by comparing taste threshold value and in vitro drug release. Preliminary study was used to investigate the effect of crospovidone, compression force (CF) and Hydroxypropyl cellulose (HPC) on disintegration time, friability and wetting time (WT). Factorial design was used as it enables simultaneous evaluation of formulation variables and their interaction effect. From the preliminary study, the factors that were found significant were further optimized using central composite design. Design-Expert 8.0.7.1 software was employed to carry out the experimental design. The bitterness threshold concentration of Trimethoprim was found to be 150 μg/ml and the in vitro drug release of the three batches of drug to polymer ratio (F1:1, 1:2 and 1:3) was 2.80±0.05, 2.77±0.00 and 2.63±0.00 respectively. From the optimization study, the optimal concentration for the superdisintegrant was 8.60% w/w and a CF of 11.25 KN which gave a rapid disintegration and WT of 13.79 and 23.19 seconds respectively and a friability of 0.666%. Conclusion In this study, co-trimoxazole ODT was formulated successfully. Central composite design was effectively used to model and optimize friability, DT and WT. The method was found effective for estimating the effect of independent variables on the dependent variables by using polynomial equation and surface plots. Optimization of the response variables was possible by using both numerical and graphical optimization and the predicted optimal conditions were confirmed experimentally and were found to be in good agreement within 5% of the predicted responses. The results of the study showed that CTX ODT had significantly rapid disintegration, less than 1% friability and enhanced dissolution profiles. The successful formulation of CTX ODT can solve difficulty of swallowing of conventional tablets for some group of patients which are unable to swallow solid oral dosage form.

If the data are held or will be held in a public repository, include URLs, accession numbers or DOIs. If this information will only be available after acceptance, indicate this by ticking the box below. For example: All XXX files are available from the XXX database (accession number(s) XXX, XXX.).
• If the data are all contained within the manuscript and/or Supporting Information files, enter the following: All relevant data are within the manuscript and its Supporting Information files.
• If neither of these applies but you are able to provide details of access elsewhere, with or without limitations, please do so. For example: Data cannot be shared publicly because of [XXX]. Data are available from the XXX Institutional Data Access / Ethics Committee (contact via XXX) for researchers who meet the criteria for access to confidential data.
The data underlying the results presented in the study are available from (include the name of the third party • All relevant data are within the manuscript and its Supporting Information files and contact information or URL). This text is appropriate if the data are owned by a third party and authors do not have permission to share the data.

• * typeset
Additional data availability information: Tick here if your circumstances are not covered by the questions above and you need the journal's help to make your data available.

Introduction
Oral route is the preferred method of drug administration for systemic effect, and tablets are the most popular of all dosage forms existing today due to their convenience of self-administration, ease and accuracy of dosing, and most importantly patient compliance (1,2). Besides, as solid oral delivery systems do not require sterile conditions, they are less expensive to manufacture and provide better stability for the drugs as compared with liquid dosage forms (3,4). For these reasons most therapeutic agents considered for systemic drug delivery have a tendency to be administered via the oral route (5). But, the oral tablets are associated with some problems; particularly for those segments of the population which are unable to swallow solid oral dosage 3 forms . The most vulnerable groups under this category are pediatric patients: because of   underdeveloped muscular and nervous control and geriatric patients as a result of changes in   various physiological and neurological conditions associated with aging (6,7). However, difficulty in swallowing or dysphagia was also reported to affect nearly 35% of the general population. It is also a pertinent finding with a number of pathological conditions, including stroke, parkinson's disease, neurological disorders and AIDS (8). In addition to dysphagic patients, those who are traveling with no or little access to water are similarly affected, potentially limiting patients not to take orally administered conventional tablets or capsules (9).
In a nutshell, difficulty of swallowing appears to impact patients compliance or adherence to their prescribed medication (10). With the aim of facilitating the ease of oral medication administration and increasing patient compliance, researches have been directed at developing innovative dosage forms for oral administration. Among the dosage forms developed for this purpose, the ODTs have been considered by product development scientists as the best solution for patient suffering from dysphagia (11).
ODT is a solid dosage form containing a medicinal substance which is designed to disintegrate within 30 seconds or less when placed upon the tongue and get in contact with saliva without the need of water (12).
ODTs offer the advantages of both solid dosage forms and liquid dosage forms (5). They provide better stability for drugs as conventional tablet formulations and they are easy to swallow like the liquid dosage forms. (13,14). Because of the ability of ODTs to present solids in the form of suspension or solution even when placed in the mouth under limited bio-fluid without the need of water, it is found to be crucial for busy people who do not always have access to water (15).
Besides, ODTs avoid the risk of chocking (Physical obstruction) that are usually encountered during administration of conventional oral formulations, thus providing improved safety (16).
The ODTs are far better than the oral liquid dosage forms in terms of accurate dosing and easy handling by patients (8). In addition, they are easy to transport and bring less handling and transportation costs due to their small volume and weight. Their production costs are less, which makes them more affordable than standard liquid formulations (13,17).

4
Co-trimoxazole (CTX) is a fixed-dose combination of two antimicrobial drugs namely, sulfamethoxazole (SMX) and trimethoprim (TMP) in 5:1 ratio. TMP is an odorless white powder with a bitter taste. It is slightly soluble in water and some organic solvents like alcohol (18).
SMX on the other hand is a white to off-white crystalline powder which is soluble in ethanol and methanol and poorly soluble in water (0.5 g/L) (19). Based upon the biopharmaceutical classification system both TMP and SMX are class II drugs; which implies low aqueous solubility and high membrane permeability (20).
CTX is a broad-spectrum antimicrobial agent used to treat the urinary, respiratory, and gastrointestinal tract infections (21,22). The antibiotic is categorized as a vital medicine and is being used in HIV programs. It is included in the opportunistic infection (OI) standard treatment guideline (STGs) and the general treatment guidelines of Ethiopia. It is also included in all three editions of the essential medicines list of Ethiopia (23).
Besides, owing to its prophylactic effect on several secondary infections in people living with HIV/AIDS such as pneumocystis carinii pneumonia (PCP), WHO incorporates CTX as an integral component of the HIV chronic care package and many countries including Ethiopia are using this drug for this purpose (24,25). When it is taken regularly as prophylaxis, it reduces mortality and specifically lowers the risk of OIs in adults and children living with HIV (23). It was reported that prophylactic treatment of OIs with CTX had showed a substantial reduction of morbidity and mortality especially in resource-limited areas. For instance, CTX had showed a 25-46% reduction of mortality among individuals infected with HIV in sub-Saharan Africa even in areas with high bacterial resistance to the antibiotic (26). However, esophageal candidiasis which is on the list of AIDS defining illness, aff ect up to 1 in 5 AIDS patients and are responsible for painful or difficulty of swallowing (27). The dysphagia associated with this comorbidity can affect the compliance of drugs including CTX given for these patients.

Taste masking by solid dispersion method
Taste masking was performed according to the method described by Sharma et al., (2010).
Different ratios (1:1, 1:2, and 1:3) of TMP (bitter drug) and powdered Eudragit E-100 were mixed using mortar and pestle. The mixtures were transferred into a stainless steel vessel and 10% ethanol (10 mL) was added to each mixture (drug polymer blend). Each mixture was then stirred constantly using a magnetic stirrer until a thick gel was formed. Incorporated ethanol was removed by evaporation at 40°C in a thermostat oven for 8 hours, and the solidified gel was then 6 crushed into particles using mortar and pestle. Finally, each crushed drug -EudragitE100 blend were allowed to pass through a 45 mesh sieve and stored appropriately (in a closed cabinet to protect it from direct sun light exposure) for further use. In addition to Eudragit E100, the sweetener saccharin sodium was also employed as a taste masker at 1% w/w while formulating CTX ODT.

Selection of best taste masked drug-polymer blend
The best taste masked candidate was selected based on taste threshold concentration, which is the minimum concentration among a range of dilutions of a substance at which the volunteer just starts feeling the bitter taste (30).

Determination of taste threshold value
The threshold of bitterness concentration for the solid dispersion was determined by a panel of 6 healthy human volunteers from whom written consent was taken. Prior to the commencement of taste masking evaluation, ethical clearance was obtained from Mekelle University Health Sciences College Ethical Review Committee (MUHSCERC). The aim of the study was then clarified to the study participants, confidentiality was ensured, and written consent was sought.
Accordingly, 6 (3 males and 3 females) healthy and non-smoking participants with the age of >18 years were involved in taste masking evaluation study.
The following procedure was followed to determine the bitter taste threshold concentration/value of TMP (30). Different concentrations of the drug samples (100, 150, 200, 250 and 300 µg/ml) were prepared in phosphate buffer pH 6.8. The study participants were requested to place 1mL of each sample (starting from the lowest to highest concentration) at the middle of their tongue and spit out after 30 seconds. There was a 10 minute interval before they took the next higher dose.
The smallest concentration at which volunteers felt the bitter taste was recorded and taken as taste threshold value.

In vitro release test at salivary pH
The drug release from the taste masked granules was determined by in vitro test evaluation at salivary pH (30). TMP-Eudragit E100 complex containing equivalent to 20 mg of TMP was placed in 10 mL of phosphate buffer pH 6.8 and shaken for 30 seconds. The amount of drug released was then analyzed by using HPLC (Agilent 1260 infinity, USA). The ratio of drug 7 polymer complex that yielded drug release values just below the taste threshold concentration was selected to be used for taste masking purpose. Where, CI is Carr's index and HR Hausner's ratio.

Angle of repose:
The angle of repose was determined from the dimensions of the powder pile, which was formed when 30 g of powder mass was allowed to flow through a funnel. The orifice of the funnel was fixed at 2 cm top of the powder pile as it is being formed in order to minimize the impact of falling on the tip of the cone. The blend was then allowed to flow through the funnel freely onto the surface. The diameter of the formed powder cone was measured and the angle of repose was calculated from the formed powder pile by using Eq. 2.5 θ = −1 ( ) Eq.2.5 8 Where, H = is height of the formed pile, R = radius of the pile

Preparation of ODTs
Orally disintegrating co-trimoxazole tablets were prepared by direct compression (DC) method.
All the ingredients were allowed to pass through 45 meshes or (354 μm) sieve separately. All except magnesium stearate and colloidal silicon dioxide were mixed in a polybag for five minutes. The blends were then lubricated with magnesium stearate and colloidal silicon dioxide and further mixed in a polybag for three minutes. Finally, each batch of the blend which was ready for compression was converted into tablets using 10 mm flat round punch. The steps for manufacturing of co-trimoxazole ODT tablet by the DC method were as follows: Weighing Sieving Blending Lubrication Compression

Evaluation of tablets
The prepared tablets were evaluated for the following characteristics.

Friability Testing
Pre-weighed sample of 20 tablets was placed in a friability tester (ERWEKA, TAR20, Germany) and rotated at 25 rpm for 4 minutes. The tablets were then dedusted and reweighed, and the friability percentage was computed according to the formula given below.

Hardness test
The hardness of 6 tablets from each sample was determined by using hardness tester (CALIVA, THT2, England). The tablets were placed in the space provided and the crushing strength (in kilo Newton) that caused each tablet to break was recorded.

Tablet thickness
The thickness of six tablets from each sample was also measured using hardness tester (CALIVA, THT2, England).

Wetting time
Wetting time test was conducted as per the method described by (Tabbakhian et al. 2014 andKumar 2016). A piece of double folded tissue paper was placed in clean and dry petri dish plates 9 containing 6 mL of water. Six tablets from each formula were carefully placed individually on the paper and the time elapsed to completely wet or the time taken by the water to reach the upper surface of the tablet was taken as the wetting time (WT).

Disintegration time
Disintegration time test was carried out according to (USP, 2007). Six tablets from each formulation were randomly selected and placed in a disintegration tester (GB Caleva Ltd., Model; DIST2, England) filled with 900 ml distilled water and maintained at 37 ± 2 o C. The time required for complete disintegration of the tablets with no palpable mass remaining in the apparatus was recorded as the disintegration time (DT).

Calibration curve
Stock solution of SMX and TMP was prepared by transferring 100 mg of SMX and 20 mg of infinity, USA). The peak area versus concentration of solutions were plotted to obtain the calibration curve.

Dissolution Studies
An in vitro dissolution study for each batch of tablets was conducted using USP Apparatus II, paddle method, (Pharma test, PTWS, Germany). The paddle was adjusted to rotate at 75 rpm.

Drug content determination
Drug content was determined by following the method described in (USP, 2007). The assay was prepared by weighing finely powdered 20 tablets. An accurately weighed portion of the powder, containing equivalent to 160 mg of SMX and 32 mg of TMP, was transferred to a 100-mL volumetric flask. Methanol (50 mL) was added and the mixture was then sonicated, with intermittent shaking, for 5 minutes. The solution was equilibrated to room temperature, diluted with methanol to volume and mixed thoroughly, then filtered. After filtration, 5 ml of clear filtrate was transferred to a 50 mL volumetric flask, and diluted with the mobile phase to volume.
Standard preparation (20 µL) and the prepared assay were separately injected into the chromatograph, the chromatograms were recorded, and the responses for the major peaks were measured. The quantities (mg) of TMP and SMX were calculated by using the formula given

Eq. 2.7
Where C is the concentration (mg/mL), of the appropriate USP reference standard in the standard preparation; and rU and rS are the responses of the corresponding analyte obtained from the assay preparation and the standard preparation, respectively.

HPLC condition
The amount of TMP and SMX dissolved in the dissolution study was determined by employing HPLC system. A sample of 5 mL was taken at each time interval from the dissolution medium and put into a volumetric flask (25 mL) and diluted to volume with the mobile phase. From these (20 µL) of the assay sample was injected into the chromatograph after being filtered through 0.45 µm membrane filter and were analyzed using HPLC (Agilent 1260 infinity, USA). All chromatographic analyses were carried out at 25°C. The compounds were separated using Chromatographic system: the liquid chromatograph was equipped with a 254-nm detector and a 3.9 mm × 300 mm column that contains packing L1. The flow rate was 1.5 mL per minute.

Experimental design for optimization of CTX ODT
In this study, Design Expert Software (trial version 8.0.7.1) was used for data analysis. Before the optimization study, preliminary screening was carried out. In the preliminary study a 2 level Composition of the preliminary experiment and experimental levels are shown in   Formulation  F1H  F2H  F3H  F4H  F5L  F6L  F7L  F8L  ingredients  Trimethoprim  20  20  20  20  20  20  20  20  Sulfamethoxazole  100  100  100  100  100  100  100  100  Eudragit E100  20  20  20  20  20  20  20  After the preliminary experiment, the factors that were found to be significant (CF and crospovidone concentration) were optimized by employing a CCD with five coded values as shown in Table 2.2. For a 2 factor study in CCD, the total number of experiments to be performed in the design are generally given as sum of 2 n factorial runs, 2n axial runs, and nc center runs (2 n + 2n + nc), where n is the number of factors. Therefore, for n=2, the total number of experiments would be 13: (2 2 + (2×2) + 5) five level for each factor (33). The 13 experiments were carried out to find the optimum area, at which the desired responses were achieved.

Bitterness threshold concentration of co-trimoxazole
From the different concentrations of TMP standard solutions, four of the participants felt the bitterness at the concentration of 200 μg /mL; whereas, the remaining two participants felt it at 150 μg /mL. Therefore, 150 μg/mL was taken as the bitterness threshold concentration of TMP.

In vitro taste-masking evaluation:
SD was the method used for taste masking. This technique, in which the drug is molecularly dispersed within the polymer matrix have shown effectiveness for masking of drugs with unpleasant taste (34). In addition, SD with Eudragit E100, had showed better dissolution profile than the marketed formulations when diclofenac was used as a model drug (35). Since TMP has poor aqueous solubility, this method was selected as it can be used for both taste masking and solubility enhancement. SD only masks the taste, but does not impart sweetness or palatability for the formulation. The sweetener in contrast provides further sweetness and palatability which can easily be taken by patients particularly pediatrics. For this reason the sweetener (Saccharin sodium) was used to augment taste masking. Among the formulations, formulation 1:1 (drug: polymer) was regarded the most cost effective formulation because its drug release was small as compared with bitterness threshold, and uses the minimum amount of the polymer accompanied with low cost of polymer as compared to the other formulations. Hence, it was incorporated as one formulation ingredient for the screening and optimization studies.
Drug release evaluation of the three batches of drug-polymer blends is presented in This marked difference might be attributed to the nature of taste masker Eudragit E100, a pH sensitive polymer, which is insoluble at salivary pH (6.8-7.2) as well as in weakly acidic buffer solutions up to pH 5 (36). In addition, a decrease in solubility of TMP as a function of pH was also reported by Sayar et al (2008), who observed a decrement of solubility from 154.1 to 55.1 mg/mL following a change in the pH of the solutions from 1.2 to 6.8.

Preliminary studies
Preliminary screening was carried out to identify the most critical variables that could have significant impact on the response variables. Based on the review of different literatures, super disintegrant concentration, binder concentration and CF were identified as factors. In the current study, CF and crospovidone were found as the two most significant factors which had an impact on DT, WT and friability.
For the production of tablets DC was employed as it has low cost of production, uses conventional equipment and commonly available excipients. In addition, it provides high mechanical integrity of tablets (8). The popularity of DC has increased due to the introduction of superdisintegrants and a better understanding of their properties since the rate of disintegration and hence the dissolution is principally affected by them (37).
In order to formulate ODTs, superdisintegrants are the most required formulation ingredients.
They are used at low level in solid dosage forms, typically 1-10% w/w relative to the total weight of the dosage unit (38). There are many in the market but crospovidone is the most 15 commonly used superdisintegrant. It has highly porous particles which do not form complexes with drugs and pose no compatibility problem. Crospovidone has an efficient disintegrant action at low concentration usually in the range of 2-5%. However at concentration up to 10%, it has a little effect on the flow properties of some other excipients and drugs (39). Hence, in this study, 2% and 10% were selected as the low-level and high-level concentrations of the polymer, respectively.
A large number of efforts have been applied to modify conventional tableting formulation and/or the process in order to produce ODTs with rapid DT while maintaining sufficient mechanical strength. CF is known to affect tablet properties (40). Increasing CF generally results in tablets with lower porosity and higher mechanical strength but longer DT. An optimum CF is required to combine these opposing characteristics (41) In this study hydroxypropyl cellulose (HPC) was used as a directly compressible binder. In formulation of tablets, HPC is primarily used in the concentrations range of 2-6% w/w in either wet-granulation, dry or DC processes (45). Since tablet formulation with HPC was characterized by high plastic deformation and good compactibility (44), which is needed for direct compressible binders, 2 and 5% were employed as low and high level concentration respectively.

16
Each of the preliminary batches was evaluated for pre-compression parameters and the results are presented below in Table 3.2.

Powder characteristics of the preliminary formulations
As displayed in Table 3

Tablet characteristics of the preliminary formulations
As shown in Table 3.3, the hardness of the tablets was in the range between 4.73 to 10.80 kgf.
Weight variation was found to be in the acceptable range. For tablets with an average weight of 130-324 the acceptable range is below 7.5% of RSD (USP, 2007). Based on this, all the batches had fulfilled this specification. The content uniformity for Co-trimoxazole tablets according to (USP, 2007) necessitates not less than 93.0% and not more than 107% of the stated amount.
Accordingly, all the preliminary batches passed the drug content specification. Weight variation= mean ± RSD All other results are mean ± SD, TMP= trimethoprim, SMX= sulfamethoxazole

Effect of formulation variable on friability
Achieving percentage friability within limits for an ODT is challenging to the formulator since all methods of manufacturing of ODT are responsible for increasing the percent friability values.
Thus, it is necessary that this parameter should be evaluated (46).
Tablet formulations are required to meet the pharmacopoeial specification for friability, which is less than 1% weight loss (USP, 2007). The effect of the independent variables on friability is displayed in Table 4.4. As can be seen from this table, the friability values were in the range between 0.78% and 2.08%. These varying results of friability indicated that this outcome variable was strongly affected by one or more of the factors (P<0.05). In this case, it was significantly affected by the CF (P=0.0019). When the CF was held at high value, the friability of the tablets became less than 1% while, those batches that were compressed at low level of CF failed the specification. It is generally known that an increase in CF results in an increase in tablet hardness, which in turn decreases tablet friability (40). The other two factors didn't show any significant impact on the friability of the tablet (P > 0.05).

Effect of formulation variable on wetting time
The WT of ODTs is another relevant parameter which needs to be evaluated. This is because it gives an insight about the disintegration property of the ODTs (47). In addition, it mimics the action of saliva in contact with tablets (48).
WT was significantly affected with both crospovidone concentration and CF (P < 0.0001 and P < 0.05) respectively. When the CF was increased, the WT also increased and vice versa. For instance, when the CF was at high level (F2H), the WT was 29 seconds, but when the CF went to its low level keeping the other two constant (F6H) the WT was 16 seconds. On the other hand, crospovidone concentration had opposing effect on the WT. If we took F1H and F2H, we saw a decrement of WT from 64 to 29 seconds by increasing crospovidone concentration and keeping the other two constant.

Effect of formulation variables on disintegration time
DT is a defining characteristic for ODT, which is required to disintegrate rapidly in the oral cavity. Therefore, it is reasonable to include this parameter in screening study as response variable.
From the results of the preliminary study, crospovidone was found to significantly affect the DT of the tablets (P < 0.05). When crospovidone concentration was taken from low to high keeping the other factors constant the disintegration decreased from 40 to 14 seconds (in F1H and F2H) and 26 and 15 (F3H and F4H) respectively.
As described in the above paragraph and depicted in Table 3.5, CF of and crospovidone concentration were the two statistically significant factors. As a result, they were selected for further optimization.

Optimization
After the important factors had been identified, the next step was to determine the settings for these factors that result in the optimum value of the responses. From the preliminary experiment, CF and crospovidone concentration were identified as significant factors for further optimization.
Thus their effect on friability, DT and WT were further studied using CCD.
In today's highly competitive environment, a business is not affordable by trial and error of experiments. This happens due to rapid market and technological changes, customers need a quality product with a lower cost. As a result, researchers should ensure research outputs be delivered to the market ahead of competitor with minimum resources (49).
Optimization is a procedure that utilizes available resources to get the best possible results. (50).  (Table 3.6).

Characterization of the powder blend
The bulk density and the tapped density of the formulations ranged from 0. 54

Characterization of tablets
As shown in Table 3

Calibration curve
The peak area obtained for TMP and SMX were plotted against concentration (

In vitro drug release
The results of drug release profiles of TMP and SMX from different ODT formulations are illustrated in Figure 3.3 and 3.4 respectively. The in vitro drug release pattern was not evaluated for F5, F6 and F12 because they did not fulfill the pharmacopoeial specification for friability.
As shown in

Evaluation of tablet for response variables
Results of the response variables for the 13 optimization formulation of co-trimoxazole ODTs is displayed in Table 3.9. For model selection different parameters were evaluated such as: reasonable agreement between adjusted R square and predicted R-square (within 0.20 of each other based on the software suggestion), not aliased, p-value of the model term which is less than 0.05, and the lack of fit pvalue not-significant, meant to be greater than 0.05 (57). The fit summary for the response variables is displayed in Table 4.10. Based on the fit summary, linear model was suggested for DT with a model P value of 0.0015 and lack of fit P value 0.0809 and the quadratic model was suggested for both friability, with a model P< 0.0001 and lack of fit P = 0.0590, and WT with a model P value and lack of fit P value 0.0035 and 0.0552 respectively. Hence, based on the fit summary output, linear model for DT and quadratic model for both friability and WT were selected as best fit models.

25
The goodness of fit of the model was also checked by coefficient of determination (R 2 ) (33). The R-squared values for friability and WT were found to be 0.9897 and 0.9785 respectively. These    As illustrated in Table 4.11, Values of "Prob > F" less than 0.0500 indicate model terms are significant. In this case X1, X 2 1 are significant model terms with a P value of < 0.0001 for both the main and the quadratic effect. But crospovidone concentration (X2), the interaction effect (X1X2) and X 2 2 are insignificant model terms. Therefore, backward elimination procedure was applied to reduce insignificant terms so as to increase the model's predictive efficiency.
Model term refinement should always be attempted before interpreting a satisfactory model. This process might increase both values of explained (R 2 ) and predicted variation. Refinement is primarily achieved through exclusion of the factors that are found to be insignificant in the coefficient plot (60) After eliminating the insignificant model terms the ANOVA for friability was assessed and the result is depicted in Table 3.12. As displayed from the table below, the reduced model became more significant (F = 353.40, p < 0.0001) than the original model (F = 134.7, p < 0.0001) indicating the model predictive efficiency was improved. In both cases, the models were significant with (p < 0.0001) but the F-value was greater for the reduced model which specifies higher significance of the corresponding variable to cause an effect. The reduced model also showed a slight improvement in the adjusted R 2 and high improvement for the predicted R 2 which went from 0.9373 to 0.9608. a decreasement of predicted residual error sum of squares (PRESS), was also seen from 0.37 to 0.23.  Table 4.14 X1, X2, X 2 2 were significant model terms for WT. Where as, the interaction effects of X1 and X2, and the quadratic effect of of CF (X 2 1) on WT were insignificant model terms. As a result, model reduction was done to improve prediction efficiency.
As shown from the reduced model ANOVA (Table 3.15) the P value and F value of the reduced model was more significant. Though the models' P-value for both reduced and un reduced models was the same (P< 0.0001) , the F value changed from 63.75 to 132.98 which implies higher effect of each variable on the response. In addition The "Lack of Fit P-value" of 0.1084 implies the Lack of Fit is not significant and there is a 10.84% chance that a "Lack of fit" this large could occur due to noise. As we need the model to fit, a higher non-significant lack of fit (0.1084) is better than a lower non significant lack of fit (0.0552) since.   In addition to the above parameters which were used for model selection and checking its adequacy, diagnostic checking tests enables researchers to evaluate adequacy of the models (61).
The normal probability plot of the residuals is used to check the normality assumption, if the assumption is true this plot will resemble a straight line. On the other hand a plot of the residual values versus the predicted response values is used to verify the absence of constant error. A random scattering of the residual values indicates that no correlation exists between the observed variance and the response (58,62).
By applying the diagnostic plots provided by the software, such as normal probability plots of the studentized residuals, and the residuals versus the predicted, the model adequacy can be further confirmed (33). Figure from 3.5 to 3.10 shows the normal probability plots of the residuals and the plots of the residuals versus the predicted response for friability, DT and WT.     The assumption of constant variance was also tested through the plots of residuals versus the predicted response ( Fig. 3.6, 3.8 and 3.10). There was no obvious pattern and the residual was scattered randomly on the display. So, the assumption of variance homogeneity was satisfied in this work. Therefore, the models were reliable to describe and adequate for their respective responses.

The mathematical regression models
In order to determine the levels of factors which yield optimum values of responses, mathematical relationships were generated between the dependent and independent variables. Significance of quadratic terms could signal that the relation is non-linear. A positive quadratic term could suggest that the relation is exponential. The small positive coefficient of the Quadratic terms (X 2 1, +0.44), shows a direct weak relationship with tablet friability. The effect of the quadratic term of the superdisintegrant concentration (X 2 2) on WT is strong with a positive sign +7.82 which implies a direct relationship.

Contour and response surface plot analysis
The relationship between independent and dependent variables was graphically represented by 3D response surface and 2D contour plots generated by their respective model (Fig.3.11 to Fig.  3.13). These plots are very useful to show interaction effects of the factors on the responses and are able to show effects of two factors on the response at a time (65).    As depicted above by the surface and contour plots (Fig. 3.13 A & B) the WT decreases as the superdisintegrant concentration increases and as the CF decreases. But at higher super disintegrant concentration there was a slight difference in the WT of tablets even at lower and higher CF. Yet, when the two factors are compared, the supper disintegrant concentration has greater effect on wetting time than CF which can be further confirmed by the regression coefficients which were found to be -22.8 and +5.93 respectively. According to the plots, the most suitable conditions for minimum WT were found to be at the lowest CF (coded X1 = -1) and at the highest concentration of superdisintegrant (coded X2 = +1).

Simultaneous optimization of response variables
After generating the polynomial equations that relate the dependent and independent variables, the formulation was optimized for all the three responses simultaneously. The final optimal experimental parameters were obtained using both numerical and graphical optimization techniques. The software is capable of generating models and calculate the optimum conditions by making some adjustments or compromise to get a combination of factor levels that jointly optimize a set of responses by satisfying the requirements (i.e. optimization criteria) for each of them i.e. multiple response optimization (66). The constraints on the response variables the was in range of 0.5 to 0.8% for friability, 5 to 15 seconds for DT and 12.3-25 seconds for WT.

Numerical optimization
In numerical optimization, the desired goals for each factor and response can be chosen. The constraints or defined criteria for factors and responses during numerical and graphical optimization are presented in Table 3.18. The numerical optimization feature in the design expert package finds one point or more in the factors domain that would maximize this objective function. The process involves combining the goals into an overall desirability function (D).
The desirability approach is recommended due to its simplicity, availability in the software and the flexibility it gives in weighting and giving importance for individual response (67). The desirability function transforms each estimated response, Yi, into a unit less utility bounded by 0 < di < 1, where a higher di value indicates that response value Yi is more desirable, and when di = 0 it means a completely undesired response. The variance of desirability value between 0 and 1 depends on the proximity of the outputs towards the target (68,69). The overall desirability function D(x) for (X1, X2) is displayed both in ramp and 3D plot in

Confirmation test
To confirm the validity of the obtained predicted optimal point, confirmation experiments were carried out at the optimal combinations of the factors in triplicate (X1= 11.25KN and X2= 8.60%). Tablets were evaluated for the outcome variables, (Friability, DT and WT) and the experimental outcomes were compared with the model predicted responses ( Accordingly, if the percent error is found within 5 percent deviation from the actual result; we confirm the validity of the response model and we can say optimization processes is capable and reliable to optimize the response variables (33,70).

Powder characteristics of the three optimized formulations
After mixing all the excipients and the active ingredients according to the final optimized formula, each of the three batches powder blend was characterized for their flow property and the result is depicted in (Table 3.20) which showed good to excellent flow.

Evaluation of the optimized co-trimoxazole ODT tablets
As displayed in Table 3.21, tablets of all the 3 formulations were evaluated for thickness, diameter, hardness, and drug content and all fulfilled the pharmacopoeial specification. The assay results of the optimized batches were found within a narrow range around the label claim evidenced by a small standard deviation, which ensures the consistency of the dosage units. In addition to characterizing the optimized formulation for the above parameters, the release profile of the drug was also determined, Figure 3.17 and 3.18 and Table 3. 22. The results showed that the optimized co-trimoxazole ODTs exhibited good in vitro drug release profile.  As displayed from the dissolution test results, it can be observed that all the three batches showed a mean release of more than 70% before 60 min which satisfies the tolerance limit.
These results showed that the optimized CTX ODT had good in vitro drug release profile.

Conclusion
Preparation of taste masked orally disintegrating co-trimoxazole tablet was possible by the DC method using crospovidone as a super disintegrant. The taste masking was successful with a combination of sweetening agent, saccharin sodium and solid dispersion with the solvent evaporation technique. All the formulation blends showed good flow properties such as angle of repose, bulk density, tapped density which revealed that they could be prepared by DC method without flow problem. Besides the prepared tablets showed good post compression parameters.
Co-trimoxazole ODT formulation has been developed and optimized successfully using central composite design. The method was found effective for estimating the effect of two main independent variables (CF and crospovidone concentration) by using polynomial equation and surface plots. Optimization of the three response variables was possible by using both numerical and graphical optimization and the predicted optimal conditions were confirmed experimentally and were found to be in good agreement within 5% of the predicted responses.
The results of this study showed that co-trimoxazole ODT had rapid disintegration, optimum percentage friability which was less than 1 % and enhanced dissolution profiles. The successful formulation of co-trimoxazole ODT can solve difficulty of swallowing of conventional tablets for some group of patients which are unable to swallow solid oral dosage form. In addition, the cost of production of CTX suspension, which requires sterile condition, transportation cost due to bulkiness of bottles could also be avoided.