Quality Target Product Profile (QTPP) of the proposed drug delivery system

QTPP is a needed document that enables the formalization and tracing of the evolution of the knowledge acquired throughout the drug lifespan. Based upon the Reference Listed Drug (RLD), requirements and targets were set up for QTPP^7).

Critical Quality Attributes (CQA's) of proposed drug delivery system

Relevant CQAs can be known by a frequent process of quality risk management and experimentation which assesses the quality of the drug product.

Risk Assessment by using QTPPs and CQAs

A risk assessment of the drug substances attributes was made to assess the impact of each attribute on the drug product CQAs.

Design of Experiment (DoE)

DoE is a prevailing skillful experiment, conducted in randomized order. The number of experiments required depends upon the selected design which curtails the number of trials. Optimization was executed by two-level, three-factor (2³) full factorial design using Design Expert 11 software. Independent variables (HPMC K100M, Carbopol 934P, and Ethylcellulose (EC)) and dependent variables (Floating time, percentage cumulative drug release at 2 h, 12 h, and t50%) were selected for optimization based on risk assessment of data⁸.

Drug-excipient interaction compatibility studies

An FT-IR spectrum was recorded by the KBr pelletization method (FT-IR Bruker model). 2 mg of the pure drug(s) (Domperidone, Famotidine) were dispersed in Potassium bromide powder (KBr) and by pressurizing 6 tons the pellets were prepared and scanned separately. Individually the spectra were compared for validation of common peaks. The same procedure was adopted for FTIR study of drug(s) with excipients

5 mg drugs (pure) and drug(s) with excipients were placed into DSC aluminum pan (DSC-60, Shimadzu) separately and then scanned at the temperature range of 25-290⁰C of heating increment of 15 ⁰C/min, then the obtained thermographs of the drug(s) (pure) and with excipients was equaled for interaction⁹.

Preparation of Domperidone and Famotidine Effervescent Floating tablets

All the ingredients were collected and weighed precisely. In addition to the previously formulated APIs and Carbopol mixture, the geometric mixture of HPMC K100, Ethyl cellulose (EC), sodium bicarbonate, and filler was added and mixed well along with Talc (2%) and magnesium stearate (1%) then passed through the 60# sieve. The lubricated mixture was compressed using round tooling on Rimex-I rotary tablet machine.

Design Space

In terms of ranges of material attributes and process parameters, Design space was defined. The linear formula is as given below¹⁰.

Y = b₀+ b₁X_1.+ b₂X₂+b₃X₃+b₄X₄+…………..b_nX_n

Where Y is considered as a response, b₀ is the constant, b₁ b_{2 …….}b_n is the variables X₁ X₂........X_n is representing the effect every ordered between -1 and +1.

Control strategy

The parameters and attributes were included in control strategy such as drug type, pharmaceutical materials, and components, operating conditions of facilities and equipment, in-process controls, specifications of finished products, allied methods¹¹.

Standard and blank solution preparation

The standard solution was prepared by dissolving the drug(s) Domperidone (20mg) and Famotidine (20mg) in a buffer (15ml) and acetonitrile (10ml) and transferred into a 25 ml volumetric flask and sonicated for 5min, then filtered through 0.45µm membrane filter. The concentration of the solution was found to be 1.6mg/ml. The blank solution which was used as the mobile phase was prepared as similar to standard without the pure drug(s) in the ratio of 60:40% v/v mixtures of Phosphate buffer and acetonitrile¹⁵.

Sample preparation

The collected dissolution samples equal to 10mg of Domperidone and 20mg of Famotidine were transferred to a 25 ml volumetric flask. The sample was dissolved and diluted with the mobile phase and subjected to sonication for 10min, filtered through a 0.45µm membrane filter.

Chromatographic Conditions: The HPLC separation and quantification were carried out using a stationary phase composed of Analytical column-Phenomonex-Gemini, C- 18 (250x4.6mm id,5μ). The mobile phase composed of potassium dihydrogen phosphate buffer-acetonitrile-methanol (40:40:20) was pumped at a flow rate of 1ml per min, and the run time was 7 minutes. The mobile phase was filtered through a 0.45μm membrane filter and degassed using an ultrasonic bath. The system was equilibrated and saturated with the mobile phase for half an hour before the injection of the solutions. All determinations were carried out at room temperature. The absorbance was measured at 284nm and 260nm. 20µL was the Injection volume. The peak areas were taken to quantitatively determine the amount of drug(s) present in the sample. Further based on the amount of drug(s) %, cumulative % drug release was calculated¹⁶.

Ethical clearance

The study was performed after obtaining approval from the Institutional Animal Ethics Committee (IAEC). A total of 6 rabbits were considered and from that 2 were used for in vivo floating behavior and 4 were used for pharmacokinetic studies.

In vivo floating behavior

In vivo floating behavior of optimized floating formulations was studied in healthy rabbits, weighing 1.5 - 2.5 kg. The study was conducted with the prior approval of the Institutional Animal Ethical Committee. The study is constructed on the principle of monitoring radiological activity. Animals are housed individually in polypropylene cages and maintained under standard conditions. A total of 2 rabbitsTotal 2 rabbits were involved in the study. The rabbits were fasted for 12h before initiating the study. First, an X-ray was taken from the rabbits to ensure the absence of radio-opaque material in the stomach. During the study, animals were not allowed for food but the water was provided ad libitum. Radiopaque formulations were prepared by incorporating 500 mg of barium sulfate into the optimized formulation. At varying time intervals, X-ray photographs of the gastric region were taken for monitoring the floating behavior of formulations¹⁷.

Experimental design

Four healthy rabbits with a bodyweight range of 1.5-2.5 kg were selected through physical examination. An open label, balanced, randomized, Two-period, Two-treatment, Two -sequence, single-dose crossover study design in which four healthy rabbits received one treatment (product) each with a washout period of 7 days such that all products are tested in all the four healthy rabbits during the study¹⁸. Each healthy rabbit receive the following treatment in random order given in (Table 1)

Drug administration

The animals were fasted overnight before administering the dose. After collecting the zero hour blood sample (blank), a standardized diet was given. The healthy rabbits were fixed in an individual rabbit holder with its head protruding outside. Then a mouth gag was introduced in between the two jaws and held in position by holding the upper jaw and lower jaw with a left hand and fixed formulations were administered through the oral route with the help of a pet piller. An I.V. cannula, 20-G KitKath (Hindustan Syringes &Medical Devices, Faridabad, India) was inserted into one marginal ear vein for blood collection before drug administration¹⁹.

Blood sample collection

During each period, 0.5 ml venous blood samples were collected from the marginal ear veins of each rabbit in AcCuvet tubes (Quantum Biologicals Pvt. Ltd., Chennai, India) containing K₃EDTA. Blood Samples were collected at predetermined time intervals of 0, 0.50, 1.00, 1.50, 2.00, 2.50, 3.00, 4.00, 6.00, 8.00, 12.00 and 24.00 h (11 intervals). The collected plasma was immediately separated by centrifugation at 7500 rpm for 15 min from the blood samples and stored in frozen conditions at -20°C before analysis. The concentration of each drug in rabbit plasma samples was measured by the HPLC method.

Determination of pharmacokinetic parameters

Pharmacokinetic parameters such as peak plasma concentration (C_max), time at which C_max occurred (T_max), area under the curve (AUC), elimination rate constant (K_el), biological half-life (t_½), absorption rate constant (K_a) and mean residence time (MRT) were calculated in each case using the data by KineticaTM 2000 software (Inna Phase Corporation, U.S.A) using non-compartmental approach²⁰.

HPLC Method

The components were separated using Phenomenex-C18 (4.6 mm id, 250 mm, 5μm) column using Methanol: 0.1% orthophosphoric acid in water (55:45% v/v) as mobile phase at a flow rate of 1.0 ml/min. The eluents were detected at 284nm (Domperidone), 260nm (Famotidine) using a UV detector.

Rehabilitation

Animals were kept for a washout period of 1 week and rehabilitees. No anesthesia was given and euthanasia was not applicable, as there was no sacrificing of animals to relieve pain and suffering.

Response (R₁) analysis: Floating time (Y₁)

Responses ranged from 5 min to 18 min and the ratio of maximum to a minimum is 3.6 that indicated no need for power transformation. The final equations of the reduced model contain only significant factor terms corresponding to the response analyzed.

Floating time (Y₁) = 12.20 + 1.00 X₁ + 1.01 X₂ + 5.24X₃ +0.81 X₁X₂ +0.66 X₂X₃

From the ANOVA results, it was found that the main factors affecting the floating time were X₁, X₂, X_3, X₁ X₂, X₂ X₃. All factors had a synergistic effect. The adjusted F- value is 4.81 and the model p-value is 0.0119. The curvature term was removed to simplify the analysis. In the Fit Statistics section, the Predicted R² of 0.9789 is in reasonable agreement with the Adjusted R² of 0.9900. The Adequate precision ratio is 6.882 which was considered an adequate signal. Contour plots and 3D plot showed that, to maximize floating time (more than or up to 12 h) the concentrations of HPMC K100 (X₁) and Carbopol 934P (X₂) should be maintained at high levels i.e., 43.8mg (X₁), 45.0mg (X₂). At this concentrations, the floating time was 12.1 h. The perturbation plot showed that as As the concentration of HPMC K100, Carbopol 934P, and ethyl cellulose increased, the Floating time also increased. The Contour plot and the 3D plot is indicated in (Figure 2(a)) and (Figure 3(a))

Response (R₂) Analysis: Drug release at 2 h (Y₂)

Responses range from 10.5% to 39.7% and the ratio of maximum to a minimum is 3.78 that indicated no need for power transformation.

Drug release at 2 h (Y₂) = 24.08 - 2.54 X₁ + 1.37X₂ - 6.34 X₃ + 3.18X₁X₃ - 1.76 X₂X₃

From the ANOVA results, it was found that the major factors affecting drug release at 2 hwere X₁, X₂, X_3, X₁X₃, X₂ X₃. Factors X₁ and X₃ exhibited a antagonistic effect while all other factors had a synergistic effect, The adjusted F- value was 5.01 and the model p-value was 0.0089. The lack of fit F-value of 275.96 implied that it was significant. In Fit Statistics section, the Predicted R² of 0.5272 is in reasonable agreement with the adjusted R² (0.3276). Adequate precision ratio was 7.436 indicating an adequate signal. Contour plots and 3D plot showed that in order to maintain drug release NMT 20%, HPMC K100M and ethylcellulose should be maintained at low levels and carbopol 934P should be maintained at high levels. By maintaining the concentrations of HPMC K100M, Carbopol 934P and Ethylcellulose at 45mg (X₁), 45mg (X₂) and 15mg (X₃) then the drug release was NMT 20%. The perturbation plot showed that an increase in the concentration of HPMC K100M and ethylcellulose led to decrease in drug release and an increase in the concentration of carbopol 934P resulted to increase in drug released. Contour plot and 3D plot is indicated in (Figure 2(b)) and (Figure 3(b))

Response (R₃) Analysis: Drug release at 12h (Y₃)

Responses ranged from 86.4% to 101.3% and the ratio of maximum to a minimum is 1.17 that indicated no need for power transformation.

Drug release at 12 h (Y₃) = 97.53 + 1.14 X₁ + 1.63 X₂ -1.94 X₃ - 1.04 X₁X₂ + 1.98 X₁X₃+ 0.8813 X₂ X₃

From the ANOVA results, it was found that the major factors affecting the Drug release at 12 h were X₁, X₂, X_3, X₁X_2, X₁ X₃, X₂ X₃. Factor X₃ exhibited an antagonistic effect while the remaining factors had a synergistic effect. The Adjusted F- value was 4.33 and the model p-value was 0.0218. The Lack of Fit F-value of 3.43 implies that it was significant. In the Fit Statistics, the Predicted R² of 0.9361 is in reasonable covenant with the Adjusted R² of 0.9610. The Adequate Precision ratio was 4.548 indicatingan adequate signal. Contour plots, 3D plot have shown that to maintain drug release NLT 80% Ethylcellulose should be maintained at low levels and HPMC K100M and Carbopol 934P should be maintained at high levels. By maintaining the concentrations of HPMC K100M, Carbopol 934P, and Ethylcellulose at 45mg (X₁), 45.0mg (X₂), 15mg (X₃) then the drug release is NLT 80%. The perturbation plot shows that as an increase in the concentration of ethylcellulose the drug release is decreased and an increase in the concentration of HPMC K100M and Carbopol 934P the drug release also increased. The Contour plot and 3D plot are indicated in (Figure 2(c)) and (Figure 3(c)).

Response (R₄) Analysis: Time taken to release 50 % drug (Y₄)

Responses range from 3.8 h to 7 h and the ratio of maximum to a minimum is 1.84 that indicates no need for power transformation.

Drug release at 12 h (Y₄) = 5.34 + 0.3750 X₁ - 0.4750 X₂ + 0.2125 X₃ + 0.0375 X₁X₂ + 0.1500 X₁X₃+ 0.5500 X₂X₃

From the ANOVA results, it was found that the major factors affecting the floating time were X₁, X₂, X_3, X₁X₂, X₁ X₃, and X₂ X₃. Factor X₂ had an antagonistic effect while the other factors had a synergistic effect. The adjusted F- value is 14.57 and the model p-value was less than 0.0001. The Lack of Fit F-value 10.55 implicates that Lack of Fit is significant. In Fit Statistics, the Predicted R² of 0.6318 is an injudicious agreement with the Adjusted R² of 0.8190 i.e. the difference is less than 0.2. This indicated there was no large block effect on the model so no need for model reduction and response transformation. The Adequate Precision ratio was 11.350 indicated an adequate signal. Contour plots, 3D plot showed that to maintain extended drug release (50%) HPMC K100M and Ethylcellulose should be maintained at high levels and Carbopol 934P should be maintained at low levels. By maintaining the concentrations of HPMC K100M, Carbopol 934P and Ethylcellulose at 43mg (X₁), 44.8mg (X₂), 15mg (X₃) then the drug release (50%) was achieved at 5.6 h. As perturbation plot showed that as an increase in the concentration of HPMC K100M and Ethylcellulose the drug release increased and an increase in the concentration of Carbopol 934P the drug release also decreased. Contour plot and the 3D plot is indicated in (Figure 2(d)) and (Figure 3(d))

Figure 2:  Contour plots of a) Response (R₁), b) Response (R₂), c) Response (R₃), d) Response (R₄) 

From all the responses, the ANOVAresults exhibited significance hence were selected for further analysis. The model was used to navigate the design space. All factor graphs and interaction graphs showed LSD bars around the predicted mean, which were good quick pair wise comparisons. From the Half-normal plot and Pareto Chart, significant factors were identified i.e., above t-value. The normal probability plot residuals followed a normal distribution i.e., followed a straight line as shown in (Figure 4).

Figure 3:  3D surface response plots of (a) Response (R₁), (b) Response (R₂), (c) Response (R₃), (d) Response (R₄) all four responses. 

Figure 4:  Plots exhibiting significance at p <0.05 (a) Half normal plot (b) Pareto chart 

It was evident from the plots of Residuals versus Predicted, Residuals Versus Run that error terms were not associated with one another and it adequately showed that the model fits the data obtained from Software. Predicted Versus Actual plot data points exhibited a split by the 45-degree line indicated significance. Cook's distance provides valuable information regarding the measurement of regression effect changes if the case is deleted, large values should be identified which were produced by recording errors and should be corrected. In this case, no large values were observed it indicated the data without any errors. All were shown in (Figure 5).

Figure 5:  Plots exhibiting significance at p <0.05 (a) Normal plot (b) Residual versus Run (c) Box-Cox plot (d) Cook's distance 

To optimize all responses with different targets, a multi-criteria decision approach like a numerical optimization technique by desirability function and graphical optimization technique by the overlay plot were used and represented in (Figure 6). The optimized results gave a solution with theoretical target profile characteristics which were shown in (Table 7) relative errors (%) between the predicted and experimental values for each response were calculated and values were found to be within 1.018. the experimental values were in agreement with the predicted value confirming the predictability and validity of the model.

Figure 6:  Overlay plot- graphical optimization technique. 

In vivo Buoyancy Studies

The X-ray images were taken after administration of barium sulfate loaded optimized floating tablets to the rabbits under fasting conditions. (Figure 7) shows X-ray images at different time intervals i.e, 0h, 6h, and 12h in the rabbit. It was observed that the tablet remained in the gastric region for more than 12h.

Figure 7:  X-ray images of buoyant behavior in rabbit (a) at 0h (b) at 6h(c) at 12h 

Pharmacokinetic assessment

The mean plasma concentration of the Domperidone and Famotidine both test and reference in (Table 8). (Figure 8) shows the comparative plasma profiles of Famcool-D Marketed formulation (R) with Domperidone and Famotidine. The mean peak plasma concentration (C_max) of the test (T) formulation was less when compared with the reference (R) formulation. The increase in t_max clearly indicated the drug availability for a prolonged period. After reaching the t_max the drug started elimination and plasma concentration gradually decreased. In the case of the test (T) formulation, the t_max achieved slowly and the drug availability was found for a long time. The increase in AUC_0-t was observed in the test (T) formulation, which was clearly indicated the drug availability for a long duration. The plasma elimination half-life (t_1/2) of the reference (R) and the test (T) formulations were significantly different. Thus the prolonged t_1/2 is another indication of the in vivo performance of the optimized floating tablets. The overall C_max, t_max, AUC_o-t, and t_1/2 were completely different between both the test and reference formulation. Therefore the developed formulations released the drug for a prolonged period of time.

Figure 8:  Plasma concentration time profile of (a) Domperidone and (b) Famotidine. All values were expressed as Mean±SD, n=4