Development of bio-relevant dissolution method for prognosis of In-vivo performance of Dipyridamole from modified release capsules

In-vitro biorelevant dissolution test method was developed for Dipyridamole in modified release multi-particulate dosage form, to simulate the product drug release after oral administration to human.  The Dipyridamole concentration in blood plasma achieved after oral administration under pre-prandial (fasting) condition were used for deriving the target dissolution profile deconvoluting the plasma concentration using numerical deconvolution technique. The fraction of drug absorbed was considered as the target dissolution profile. The drug product was tested by using the dissolution method recommended by office of generic drugs, and the dissolution results observed are not comparable with the target dissolution profile.  Dissolution method was developed using reciprocating cylinder, Bio-Dis (apparatus -3 as per USP), by simulating the pre-prandial conditions. A full factorial design of experiment was carried out to achieve the target dissolution profile. Media volume and dips per minute (DPM) are considered as main factors, in design of experiment. The dissolution results achieved with media volume 250ml and 10DPM were found to be comparable with target dissolution profile and observed with the F2 value (similarity factor) of 92%. The developed dissolution method demonstrates a very good in-vitro/in-vivo correlation under pre-prandial condition, and shall be used as a predictive in-vitro tool for evaluation Dipyridamole extended release capsules.

ABSTRACT In-vitro biorelevant dissolution test method was developed for Dipyridamole in modi ied release multiparticulate dosage form, to simulate the product drug release after oral administration to human. The Dipyridamole concentration in blood plasma achieved after oral administration under preprandial(fasting) condition were used for deriving the target dissolution proile by deconvoluting the plasma concentration using numerical deconvolution technique. The fraction of drug absorbed was considered as the target dissolution pro ile. The drug product was tested by using the dissolution method recommended by of ice of generic drugs, and the dissolution results observed are not comparable with the target dissolution pro ile. Dissolution method was developed using reciprocating cylinder, Bio-Dis (apparatus -3 as per USP), by simulating the pre-prandial conditions. A full factorial design of experiment was carried out to achieve the target dissolution pro ile. Media volume and dips per minute (DPM) are considered as main factors, in design of experiment. The dissolution results achieved with media volume 250ml and 10DPM were found to be comparable with target dissolution pro ile and observed with the F 2 value (similarity factor) of 92%. The developed dissolution method demonstrates a very good in-vitro/in-vivo correlation under pre-prandial condition, and shall be used as a predictive in-vitro tool for evaluation Dipyridamole extended release capsules.

INTRODUCTION
Dipyridamole(DPM) is a nucleoside transport inhibitor and antiplatelet agent. Chemical name is 2,2',2",2"'- [(4,4d] pyrimidine-2,6-diyl) dinitrilo] -tetraethanol. Dipyridamole is intensely yellow, crystalline powder or needles. Very soluble in methanol, in alcohol, and in chloroform, slightly soluble in water, very slightly soluble in acetone and in ethyl acetate. The antiplatelet mechanism of action involves by inhibiting the uptake of adenosine into platelets, endothelial cells and erythrocytes. The rate of inhibition occurs in a dose dependent manner at therapeutic concentrations (0.5-1.9 µg/mL). (Dipyridamole, 2008;Zhang et al., 1997) . Aggrenox capsules is the combination of Dipyridamole in a multiparticulate extended release form and Aspirin in immediate release tablet form illed in a hard gelatin capsule. The present study is aimed to develop the dissolution method for Dipyridamole part.   Average weight of illed pellets 504.0mg 7 Average weight of illed tablet 110mg 8 Thickness of tablet 3.50mm 9 Capsule shell size and lock length 23.30mm (size "0EL")  Dissolution is a measurement criteria to evaluate the rate of drug release from inished product. The rate of dissolution shall be extrapolated to rate of drug absorption into human body to predict the invivo performance of medicinal product. The correlation between drug release rate in dissolution media and drug absorption rate in human body is termed as In-vitro/In-vivo Correlation (IVIVC) or In-vitro/In-vivo Relationship (IVIVR) FDA (1997).
The regulatory guidelines are acknowledging biopharmaceutics based dissolution study results, for waving the bio study for formulation containing multiple strengths. The in-vivo predictive dissolution method requires most of the time the dissolution media and residence time by re lecting gastro intestinal conditions. Several dissolution media were developed by simulating human gastro intestinal system, like fasted state simulated intestinal The agitation speed disrupts the structure to have faster erosion of pellets.
Note : Risk assessment measured in 3 categories, low (1), medium (2) & high (3). The Risk number is the multiplication of all the three. The risk number more than 9 will be considered for DOE study luid (FaSSIF) and fed state simulated intestinal luid (FeSSIF). (Galia et al., 1998;Jantratid et al., 2008Jantratid et al., , 2009). The dissolution system can serve a strong indicator based on appropriate IVIVC/R . While developing the dissolution method to simulate the in-vivo performance of drug product, the dissolution media selection for the particular period of time should match with the environment of gastro intestinal track, to maintain sink condition. Segment of human gastro intestinal track and time at each pH condition is presented in Table 1.
In general the pharmacopoeial dissolution speci ica-tion are having the limit for maximum extent of drug release, and the dissolution media recommended do not represent the gastrointestinal (GI) tract environment in a comprehensive manner. For development of bio-relevant dissolution method, bio-relevant dissolution media plays a crucial role, which was developed by closely representing pre-prandial condition and post prandial condition (Klein, 2010;Vertzoni et al., 2004).
USP Apparatus 3 -Reciprocating cylinder is used for solid dosage form drug products, mostly nondisintegrating single units (e.g. tablets), and multi-    ple units (e.g. encapsulated beads). Highly recommended to use for extended release products. USP Apparatus -3 is programmable to run dissolution in various media with different speed at set time intervals. pH changes shall be simulated to gastro intestinal physiology.
A Level-A In-vitro/in-vivo correlation is established by comparing the percentage or fraction of drug dissolved, which was achieved from dissolution results, to the percentage or fraction of drug absorbed, which was achieved by deconvoluting the plasma drug concentration by numerical deconvolution method (Cardot and Davit, 2012).

Figure 3: Main effect of DPM and media volume on dissolution pro ile and interaction effect
ume at 2 levels, the response was evaluated at four time points for dissolution. (Wei and Löbenberg, 2006;Jantratid et al., 2009). The factor levels and response to be measured were presented in Table 2 & the detailed compositions has been presented in Table 3.
Preparation of Fated state simulated gastric luid (FaSSGF) : 0.16 g Lecithin was dissolved in 1.6 ml of Dichloromethane, and added to 5 liters of puri ied water. 0.42 g of Sodium taurocholate was added to the above solution and stirred for 45 minutes. 1 g Pepsin and 20 g of NaCl was added to the above solution, heated at 40 • C, using hot plate under continuous stirring for 30 minutes. The pH was adjusted to 1.6 using 1N HCl. The volume was made up to 10 liters.
Preparation of blank Fasted state simulated intestinal luid (FaSSIF) pH 6.5, pH 7.0 & pH 7.5: 19.77g of sodium dihydrogen phosphate monohydrate, 1.7g of sodium hydroxide pellets and 30.93g of sodium chloride were dissolved in 5L of puri ied water, by stirring for 30 minutes. The pH was adjusted to exactly pH 6.5 or pH 7.0 or pH 7.5 using 1N sodium hydroxide solution or 1N Hydrochloric acid solution.
Preparation of Fasted state simulated intestinal luid(FaSSIF) pH 6.5 , pH7.0 & pH 7.5: 3.3 g sodium taurocholate was dissolved in approximately 500 mL of the blank FaSSIF of specifc pH solution. 10g of Lecithin was dissolved in 100ml of methylene chloride by mixing to the achive the concentartion on 100mg/ml. 11.8 mL of a Methylene chloride solution containing 100 mg/mL lecithin was added to balnk FaSSIF, and stirred well for 15 minutes. A milky emulsion obtained. The solution was introduced into rotavapor, and methylene chloride was evaporated by heating at 40 • C, under vaccum with the RPM of 50. The result was a clear, micellar solution having no perceptible odor of Methylene chloride. After cooling to room temperature, the weight of the solution was checked again. The water lost to evaporation was replaced with demineralized water to obtain a total weight. Finally, the volume was made upto 2L using blank FaSSIF.
Preparation of Simulated colonic luid pH 5.8: 1.44g of dibasic sodium phosphate, 8g of sodium chloride, 0.2g of potassium chloride and 0.24g of monobasic potassium phosphate in were dissolved in 1litre of puri ied water. pH was adjusted to pH5.8 using 1N sodium hydroxide solution or 1N Hydrochloric acid solution. (Marques et al., 2011) Samples were with withdrawn at the following time interval 60,105,120,150,180,240,360,480,600 and 720 min, and replaced with fresh dissolution media. The temperature in the vessels were main-tained at 37±0.5 •C. The sample volume withdrawn was approximately 5 mL, and analysed using HPLC.

RESULTS AND DISCUSSION
The RLD was checked for physical characterization, and results were presented in Table 4.
Deconvolution of Pre-prandial in-vivo data: The mean plasma drug concentration of Dipyridamole obtained from Aggrenox at pre-prandial condition were deconvoluted using wagner nelson numerical deconvolution method. The target dissolution pro ile was derived from fraction of drug absorbed, and the results are presented in Table 5. (Aggrenox, 1999).
The deconvoluted data indicates that 90% of drug is absorbed in 12 hours, which directs the simulated dissolution to be performed for 12 hrs, using appropriate dissolution sink conditions.
In-vitro dissolution of Aggrenox capsules in OGD recommended dissolution media, and the study on effect of RPM Analytical method was developed for evaluation of dissolution, and the standard chromatogram is presented in Figure 1.
A comparative dissolution pro ile of Aggrenox capsules in OGD recommended dissolution media and target dissolution pro ile, along with the effect of RPM on dissolution pro ile are presented in Table 6 and Figure 2 and compared for similarity factor with target dissolution pro ile.
A comparative dissolution pro ile of Aggrenox capsules in The of ice of generic drugs recommended dissolution media was evaluated, for dissolution of Aggrenox capsules. The target pro ile achieved from deconvoluted data was not comparable with the dissolution pro ile achieved even by varying the RPM. The similarity factor (F 2 ) values observed also below 50%. Hence, it was decided to develop a biopredictive dissolution method to simulate the invivo performance of drug product.

Development of bio-relevant dissolution method :
The dissolution method was aimed to develop by using QBD approach. The target pro ile was de ined as deconvoluted dissolution pro ile Kortejärvi et al. (2002) (Kortejärvi et al.,2002;Paul et al., 2008;) Initial risk assessment of CQA (Dissolution pro ile) on variables are, the residence time, molar concentration and pH of buffer, DPM and media volume, in the biorelavance media dissolution method development. The buffer pH & residence time are ixed based on the research work carried out by (Dressman and Reppas, 2000). Hence, two factors were evaluated in this study, risk assessment was performed for both parameters, the RPN number was presented in Table 7 and the study design was estab-lished using minitab.
A full factorial study was performed, for media volume 2 levels were evaluated and , for DPM 4 levels were evaluated. The response was considered a dissolution with 4 time points of 1 hour, 4hrs , 8hrs & 12 hrs. Signi icant factors for dissolution 1hr, 4hrs, 8 hrs and 12 hours are presented below the in the Figure 3 and Table 8 , on main effects and interaction effects.
For all DOE data analysis, the commonly used alpha of 0.05 was chosen to differentiate between signi icant and not signi icant factors. The ANOVA results are presented in Table 9.
The model summary is observed with 100%. The design response was optimised to the target proile using design of experiment study, and the lower, upper level ranges predicted are presented in Table 10. The predicted DPM and media volume required for the product is 10 DPM and 250ml, with the composite desirability of 0.798. The design of target and ranges recommended for the response by minitab was presented below in Table 10, and the response optimisation was presented in Figure 4.
Establishment of the IVIV-R: A comparative dissolution pro ile using USP apparatus -3 and target pro ile is presented in  Kakhi et al., 2013) The similarity factor (F 2 )observed between the fraction of drug absorbed and fraction of drug dissolved is 92. The level A In-vitro/in-vivo correlation was established for Aggrenox under pre-prandial condition, and the R 2 value observed was 0.999, indicates very good predictive capability of the relationship.
Hence, the dissolution method developed using USP apparatus 3 (reciprocating cylinder) at 10 DPM in 250 mL of change over media simulating preprandial condition shall be used as a predictive dissolution method, based on established IVIV-R.

CONCLUSIONS
The inished product is meeting to the acceptance criteria as per OGD recommended dissolution media, and agitation speed is not having any impact on dissolution pro ile. The F2 value observed was 82 & 87 at 75 RPM & 125RRPM respectively in comparison to 100RPM. However, the F 2 value with target dissolution pro ile deconvoluted from in-vivo was observed below 50. A suitable bio-relevant dissolution media was developed, and the developed dissolution media has achieved the level -A IVIVC, with the R 2 value of 0.999, which meets USFDA and EMA requirements.