UNIVERSITY OF CINCINNATI
DATE: October 19, 2002
I, Ehab Hamed , hereby submit this as part of the requirements for the degree of: Doctor of Philosophy in: Pharmaceutical Sciences/ Industrial Pharmacy
It is entitled: Application and Evaluation of Extended Release Technology To Loop Diuretics
Approved by:
William Cacini, Ph.D.
Pankaj Desai, Ph.D.
Myron Gerson, M.D.
Ronald Millard, Ph.D.
Adel Sakr, Ph.D.
APPLICATION AND EVALUATION OF EXTENDED RELEASE TECHNOLOGY TO LOOP DIURETICS
A Dissertation submitted to the
Division of Research and Advanced Studies
of the University of Cincinnati
in partial fulfillment of the
requirements for the degree of
DOCTOR OF PHILOSOPHY (Ph.D.)
in the Department of Pharmaceutical Sciences
of the College of Pharmacy
2002
by
Ehab Ahmed Mamdouh Hamed
B.Sc., Assiut University, Egypt, 1996
Committee Chair: Adel Sakr, Ph.D. Abstract
Loop diuretics offer great advantages in treating edematous states associated with congestive heart failure, liver cirrhosis and kidney failure owing to their intense diuretic effect. Evidences suggested the diuretic effect can be exaggerated by careful control of the rate at which loop diuretics are made available to the urinary tubules. If optimally designed, peroral extended release formulation can provide better utilization of the same total dose of loop diuretic, an effect of utmost importance in edematous patients with high resistance to loop diuretics. Bumetanide multiparticulate immediate and extended release formulations were developed and tested in rabbits. A novel multiple response optimization technique based on superimposing contour diagrams was developed and successfully used to optimize bumetanide release. Instability in drug release from multiparticulate formulations after storage warrants in depth investigation of different formulation and processing factors controlling drug release. Curing time, temperature, plasticizer level, coating polymer lipophilicity, and the use of hydrophilic seal coat were explored in this study. The findings proved instability in bumetanide release is attributed to drug migration into the film coat during storage. Careful selection of plasticizer level and curing conditions together with the use of hydrophilic seal coat prevented drug migration and stabilized drug release after storage. When compared to immediate release formulation in rabbits, equivalent amounts of bumetanide were excreted from both formulations yet at different rates. The slow delivery of bumetanide from the extended release formulation improved its diuretic and natriuretic efficiencies within the first day after dosing. The activation of compensatory mechanisms is thought to diminish the response to extended release bumetanide formulation within the second day. While providing comparable diuretic and saliuretic effects to that of immediate release formulation, extended release bumetanide formulation can offer the advantage of avoiding the initial, unpleasant and intense diuretic effect experienced with immediate release formulations.
Dedication
This dissertation is dedicated to my parents in the utmost appreciation and gratitude for their unconditional love, support and continuous encouragement throughout every step in my life. No words or deeds can ever thank them enough for what they blessed me with. Acknowledgements
I would like to thank my dissertation committee: Dr. W. Cacini, Dr. P. Desai, Dr.
M. Gerson, Dr. R. Millard, and Dr. A. Sakr for their valuable guidance throughout the course of my research.
I wish to express special word of thanks to Dr. A. Sakr, my committee chairman for his support both financially and emotionally. Dr. Sakr dealt with me in a fatherly attitude and provided me with constructive criticism that helped me develop the skills required to fulfill my research needs. It was Dr. Sakr's affection and enthusiasm to Industrial Pharmacy research that brought me to this arena and inspired me with the required endurance to face the day-to-day challenges of my research. It is fair to say that without Dr. Sakr's guidance and support, this work would have never been possible.
I would like to thank Dr. R. Millard for his guidance and valuable scientific advice in my animal study. Dr. Millard's shared with me some of his great experience and wisdom that helped me not only in my research studies but also in my life perspective. I find my contact with Dr. Millard very fruitful for the development of my research skills.
I would like also to thank Dr. M. Gerson for his confidence in my abilities and the financial support to pursue my research goals.
I would like to thank Dr. Desai for providing the fluorescence detector needed for my analytical work. I would like to thank Dr. W. Cacini for providing the chart recorder needed for my analytical work.
I would like to thank Dr. M. Kurtzman from the Laboratory Animals Medical
Services (LAMS) for his indispensable help in the animal study.
I would like to thank Dr. H. Amlal from the department of Nephrology for his help with the urine analysis.
I would like to thank Dr. H. Al-Khalidi from Procter and Gamble for his advice regarding the statistical treatment of my data.
I would like to acknowledge the generous donation of Eudragit RS 30 D from
Rohm America. Special thanks go to Mr. A. Honeycheck for his sincere help and support.
Finally, I would like to thank my colleagues in the Industrial Pharmacy Program for their needed friendship, cooperation and understanding. Table of Contents
Page 1. Introduction 14
1.1. Definition 14
1.2. Pharmacodynamic Rationale for Formulating Loop 14 Diuretics as Extended Release Dosage Forms
1.3. Rationale for Selecting Bumetanide as Model Drug for the 20 Study
1.4. Peroral Extended Release Formulation Design Strategy 25
1.4.1. Diffusion-Controlled Extended Release Formulations 25
1.4.1.1. Reservoir Devices 25
1.4.1.2. Matrix Devices 26
1.4.2. Dissolution-Controlled Extended Release Formulations 26
1.4.3. Osmotic-Controlled Extended Release Formulations 27
1.4.4. Extended Release Formulation Based on Ion Exchange 28 Resin
1.5. Multiparticulate Drug Delivery System 28
1.5.1. Definition and Advantages 28
1.5.2. Manufacture of Multiparticulate Drug Delivery Systems 29
1.5.2.1. Extrusion/ Spheronization Process 29
1.5.2.2. The Use of Fluid Bed Equipment in Pelletization and 31 Coating of Multiparticulate Drug Delivery Systems
1.5.2.2.1. Types of Fluid Bed Equipment 32
1.5.2.2.1.1. Top Spray Fluid Bed 32
1.5.2.2.1.2. Bottom Spray Fluid Bed (Wurster Coater) 32
1 1.5.2.2.1.3. Tangential Spray Fluid Bed (Rotary Fluid Bed) 34 1.5.2.3. Coating of Multiparticulate Systems 34
1.5.2.3.1. Coating of Multiparticulate Systems Using Aqueous 35 Polymeric Dispersion
1.5.2.3.2. Factors to be Considered in Coating of Multiparticulate 36 Systems in Fluid Bed Equipment
1.5.2.3.2.1. Film Forming Temperature 36
1.5.2.3.2.2. Plasticizers 37
1.5.2.3.2.3. Fluidization Air Temperature 39
1.5.2.3.2.4 Spray Rate 39
1.5.2.4. Curing of Coated Multiparticulate System 40
1.6. Optimization of Drug Release from Coated Multiparticulate 42 Systems Using Statistical Modeling
2. Hypothesis 45
3. Objective 45
4. Specific Aims 45
5. Experiment and Methodology 46
5.1. Materials and Supplies 46
5.2. Equipment 47
5.3. Software 48
5.4. Methods 49
5.4.1. Part I: Development and Optimization of Bumetanide 49 Extended Release Formulations Using Response Surface Methodology and Multiple Response Optimization
5.4.1.1. Layering of Bumetanide onto Nu-pariels Sugar Pellets 49
5.4.1.2. Development of Bumetanide Spectrofluorimetric Analytical 50 Method
2 5.4.1.3. Validation of Bumetanide Analytical Method 53
5.4.1.4. Content Uniformity Assessment 53
5.4.1.5. Coating of Bumetanide-Loaded Pellets 54
5.4.1.6. Testing the Release of Bumetanide from Coated Pellets 55
5.4.1.7. Testing the Effect of Dissolution Media pH on the Release 58 of Bumetanide from Coated Pellets
5.4.1.8. Testing the Effect of Agitation Speed on the Release of 58 Bumetanide from Coated Pellets
5.4.1.9. Statistical Analysis and Optimization of Bumetanide 59 Release
5.4.1.10 Preparation and Statistical Evaluation of the Designed 60 Formulation
5.4.2. Part II: Study of the Effect of Curing Conditions and 61 Plasticizer Level on The Release of Bumetanide from Coated Pellets
5.4.2.1. Statistical Experimental Design 61
5.4.2.2. Coating of Bumetanide-Loaded Pellets 63
5.4.2.3. Curing of Coated Pellets 63
5.4.2.4. Testing the Release of Bumetanide from Cured Coated 65 Pellets
5.4.2.5. Seal Coating with Hydroxypropyl Methyl Cellulose (HPMC 65 LV 100)
5.4.2.6. Coating using a Mixture of Eudragit RS and Eudragit RL 65
5.4.2.7. Use of Sodium Chloride as a Channeling Agent in the 68 Coating Formulation
5.4.2.8. Study the Effect of Storage on the Release of Bumetanide 71 from Coated and Cured Pellets
5.4.2.9. Statistical Analysis 71
3 5.4.3. Part III: Testing Selected Optimized Bumetanide Extended 72 Release Formulation in Laboratory Animals
5.4.3.1. Selection of Animal Model 72
5.4.3.1.1. Reasons for Selecting Rabbit as the Animal Model 72
5.4.3.2. Animal Study Design 76
5.4.3.3. Rationale for the number of rabbits used 77
5.4.3.4. Rabbits Manipulation 79
5.4.3.4.1 Rabbits Restraint 79
5.4.3.4.2. Drug Administration 80
5.4.3.4.3. Sample Withdrawal 80
5.4.3.5. Analysis of Blood and Urine Sample for Bumetanide 81 Contents
5.4.3.5.1. Development of HPLC Analytical Method 81
5.4.3.5.2. Validation of Bumetanide HPLC Analytical Method 82
5.4.3.6. Analysis of Urine Samples for Sodium and Potassium 83 Contents
5.4.3.7. Determination of Urine Osmolarity 83
5.4.3.8. Statistical Analysis 84
6. Results and Discussion 85
6.1. Part I: Development and Optimization of Bumetanide 85 Extended Release Formulations using Response Surface Methodology and Multiple Response Optimization
6.1.1. Development and Validation of Bumetanide 85 Spectrofluorimetric Analytical Method
6.1.2. Content Uniformity Assessment 87
6.1.3. Testing the Release of Bumetanide from the Coated 87 Pellets
4 6.1.4. Testing the Effect of Dissolution Media pH on the Release 88 of Bumetanide from coated Pellets
6.1.5. Testing the Effect of Agitation Speed on the Release of 89 Bumetanide from Coated Pellets
6.1.6. Statistical Analysis 94
6.1.6.1. Multiple Regression and Model Building 94
6.1.6.2. Optimization 98
6.2. Part II: Study of the Effect of Curing Conditions and 112 Plasticizer Level on the Release of Bumetanide from Coated Pellets
6.2.1. Coated Pellets Plasticized with 10 % Triethyl Citrate 113
6.2.2. Coated Pellets Plasticized with 15 % Triethyl Citrate 122
6.2.2.1. Investigation of the Release Mechanism 129
6.2.3. Coated Pellets Plasticized with 15 % Triethyl Citrate 131
6.2.3.1. Investigation of the Release Mechanism 132
6.2.4. Effect of Curing Method on Bumetanide Release from 141 Coated Pellets
6.2.5. The Theory of Over Curing 142
6.2.5.1. Effect of HPMC Seal coat on Bumetanide Release after 146 Curing
6.2.5.2. Effect of Incorporating Eudragit RL into the Film Coat on 162 bumetanide release after curing
6.2.6. Effect of Storage on Bumetanide Release from Coated 165 Pellets
6.2.7. Final Optimization of Bumetanide Release 169
6.3. Part III: Testing Selected Optimized Bumetanide Extended 175 Release Formulation in Laboratory Animals
5 6.3.1. Validation of Bumetanide HPLC Analytical Method 175
6.3.2. Comparison of the Rabbit Response to Immediate and 176 Extended Release Bumetanide Formulations
6.3.2.1. Urine Output 176
6.3.2.2. Urinary Excretion of Bumetanide 180
6.3.2.3. Sodium and Potassium Excretion 185
6.3.2.4. Urine Osmolarity 190
7. Summary and Conclusions 192
REFERENCES 196
6 List of Tables
Page Table 1: The desired pharmacokinetic and biopharmaceutical 22 characteristics for drug candidate suitable for extended release formulation as well as the recorded bumetanide characteristics
Table 2: Film forming temperatures of Eudragit RL/S as aqueous 38 dispersions with added plasticizer
Table 3: Bumetanide formulation used in the layering process 51
Table 4: Wurster fluid bed equipment processing conditions for 52 the bumetanide layering process
Table 5: Central composite statistical design used for the study 56
Table 6: Wurster fluid bed equipment processing conditions for 57 coating the bumetanide-loaded pellets
Table 7: Full factorial experimental design used in the curing 62 experiment
Table 8: Coating formulations used in the curing experiment 64
Table 9: Formulation used to seal coat the bumetanide-loaded 66 pellets with HPMC prior to coating with Eudragit RS
Table 10: Wurster fluid bed equipment processing conditions for 67 the HPMC seal coating process
Table 11: Coating formulation containing a mixture of Eudragit RS 69 and RL 30 D
Table 12: Coating formulation containing sodium chloride 70
Table 13: The lengths of different segments of the gastrointestinal 74 tract of laboratory animals compared to humans
Table 14: The pH values of different segments of the 75 gastrointestinal tract of human and some laboratory animals
Table 15: Maximum excitation and emission wavelengths of 86 bumetanide in different media
7 Table 16: Model parameters affecting the percent of bumetanide 96 released after 1,4 and 8 hr
Table 17: F2 Test to compare the observed and the predicted 111 target profiles
Table 18: Effect of plasticizer level on reducing the time needed to 140 completely cure the coated pellets at different curing temperatures
Table 19: Effect of HPMC seal coating on the differences in 154 percent drug released from pellets cured for 1 hr at 60 LC and uncured pellets (uncured – cured)
Table 20: Effect of HPMC seal coating on the differences in 156 percent drug released from pellets cured for 4 hr at 60 LC and uncured pellets (uncured – cured)
Table 21: Effect of HPMC seal coating on the differences in 158 percent drug released from pellets cured for 8 hr at 60 LC and uncured pellets (uncured – cured)
Table 22: Effect of HPMC seal coating on the differences in 161 percent drug released from pellets cured for 24 hr at 60 LC and uncured pellets (uncured – cured)
Table 23: Effect of incorporating Eudragit RL on the differences in 167 release between pellets cured for different times at 60 LC and uncured pellets (uncured – cured)
Table 24: Comparison of release profile from mixture of pellets with 174 and without sodium chloride and target release profile
Table 25: Effect of formulation on the urine output (ml/Kg) in rabbits 179 (n=8)
Table 26: Effect of formulation on the urinary excretion of 182 bumetanide (ug/Kg) in rabbits (n=8)
Table 27: Effect of formulation on the urinary excretion of sodium 188 (mmol/Kg) in rabbits (n=8)
Table 28: Effect of formulation on the urinary excretion of 189 potassium (mmol/Kg) in rabbits (n=8)
8 List of Figures
Page Figure 1: Chemical structure of bumetanide (3-(butylamino)-4- 21 phenoxy-5-sulfamoylbenzoic acid)
Figure 2: Fluid bed equipment with the Wurster insert 33
Figure 3: Effect of Eudragit RS load and triethyl citrate level on the 90 release of bumetanide from coated pellets
Figure 4: Effect of Eudragit RS load and triethyl citrate level on the 91 release of bumetanide from coated pellets
Figure 5: Effect of dissolution media pH on bumetanide release from 92 pellets coated with 7.5% Eudragit RS plasticized with 20% triethyl citrate
Figure 6: Effect of basket agitation speed on bumetanide release 93 from pellets coated with 7.5% Eudragit RS plasticized with 20% triethyl citrate
Figure 7: Three-dimension contour diagram illustrating the effect of 99 Eudragit RS load and triethyl citrate level on the release of bumetanide from coated pellets after 1hr
Figure 8: Three-dimension contour diagram illustrating the effect of 100 Eudragit RS 30D load and triethyl citrate level on the release of bumetanide from coated pellets after 4hr
Figure 9: Three-dimension contour diagram illustrating the effect of 101 Eudragit RS 30D load and triethyl citrate level on the release of bumetanide from coated pellets after 8hr
Figure 10: The different dissolution profiles targeted for the in vitro/in 102 vivo correlation study
Figure 11: Two-dimension contour diagrams illustrating the effect of 106 Eudragit RS load and triethyl citrate level on the percent bumetanide release after 1 hr.
Figure 12: Two-dimension contour diagrams illustrating the effect of 107 Eudragit RS 30D load and triethyl citrate level on the release of bumetanide from coated pellets after 4 hr
9 Figure 13: Two-dimension contour diagrams illustrating the effect of 108 Eudragit RS 30D load and triethyl citrate level on the release of bumetanide from coated pellets after 8 hr
Figure 14: Superimposed contour diagrams 109
Figure 15: Comparison of the observed dissolution profile to the 110 target predicted profile
Figure 16: Effect of curing time at 40 LC on the release of bumetanide 115 from pellets coated with 6% Eudragit RS plasticized with 10% triethyl citrate
Figure 17: Effect of curing time on the percent of bumetanide 116 released after 0.5 hr and 4 hr from pellets coated with 6 % Eudragit RS plasticized with 10 % triethyl citrate and cured at 40 L C
Figure 18: Effect of curing time at 50 LC on the release of bumetanide 117 from pellets coated with 6% Eudragit RS plasticized with 10% triethyl citrate
Figure 19: Effect of curing time on the percent of bumetanide 118 released after 0.5 hr and 4 hr from pellets coated with 6 % Eudragit RS plasticized with 10 % triethyl citrate and cured at 50 L C
Figure 20: Effect of curing time at 60 LC on the release of bumetanide 119 from pellets coated with 6% Eudragit RS plasticized with 10% triethyl citrate
Figure 21: Effect of curing time on the percent of bumetanide 120 released after 0.5 hr and 4 hr from pellets coated with 6 % Eudragit RS plasticized with 10 % triethyl citrate and cured at 60 L C
Figure 22: Comparison of the slowest release profiles obtained after 121 curing for 168 hr at different temperatures for pellets coated with 6 % Eudragit RS plasticized with 10 % triethyl citrate
Figure 23: Effect of curing time at 40 LC on the release of bumetanide 123 from pellets coated with 6% Eudragit RS plasticized with 15% triethyl citrate
10 Figure 24: Effect of curing time on the percent of bumetanide 124 released after 0.5 hr and 4 hr from pellets coated with 6 % Eudragit RS plasticized with 15 % triethyl citrate and cured at 40 L C
Figure 25: Effect of curing time at 50 LC on the release of bumetanide 125 from pellets coated with 6% Eudragit RS plasticized with 15% triethyl citrate
Figure 26: Effect of curing time on the percent of bumetanide 126 released after 0.5 hr and 4 hr from pellets coated with 6 % Eudragit RS plasticized with 15 % triethyl citrate and cured at 50 L C
Figure 27: Effect of curing time at 60 LC on the release of bumetanide 127 from pellets coated with 6% Eudragit RS plasticized with 15% triethyl citrate
Figure 28: Effect of curing time on the percent of bumetanide 128 released after 0.5 hr and 4 hr from pellets coated with 6 % Eudragit RS plasticized with 15 % triethyl citrate and cured at 60 L C
Figure 29: Effect of curing time at 40 LC on the release of bumetanide 134 from pellets coated with 6% Eudragit RS plasticized with 20% triethyl citrate
Figure 30: Effect of curing time on the percent of bumetanide 135 released after 0.5 hr and 4 hr from pellets coated with 6 % Eudragit RS plasticized with 20 % triethyl citrate and cured at 40 L C
Figure 31: Effect of curing time at 50 LC on the release of bumetanide 136 from pellets coated with 6% Eudragit RS plasticized with 20% triethyl citrate
Figure 32: Effect of curing time on the percent of bumetanide 137 released after 0.5 hr and 4 hr from pellets coated with 6 % Eudragit RS plasticized with 20 % triethyl citrate and cured at 50 L C
Figure 33: Effect of curing time at 60 LC on the release of bumetanide 138 from pellets coated with 6% Eudragit RS plasticized with 20% triethyl citrate
11 Figure 34: Effect of curing time on the percent of bumetanide 139 released after 0.5 hr and 4 hr from pellets coated with 6 % Eudragit RS plasticized with 20 % triethyl citrate and cured at 60 LC
Figure 35: Effect of curing time at 50 LC in fluid bed equipment on 143 bumetanide release from pellets coated with 6 % Eudragit RS plasticized with 20 % triethyl citrate
Figure 36: Effect of curing method on bumetanide release from 144 pellets coated with 6 % Eudragit RS plasticized with 20% triethyl citrate and cured at 50LC for 1 hr
Figure 37: Effect of storage time at room temperature on bumetanide 148 release from pellets coated with 6 % Eudragit RS plasticized with 20 % triethyl citrate and cured for 1 hr at 60 LC
Figure 38: Effect of HPMC seal coating on bumetanide release 149
Figure 39: Effect of curing time at 60 LC on bumetanide release from 150 pellets seal coated with 1 % HPMC and coated with 6% Eudragit RS plasticized with 20 % triethyl citrate
Figure 40: Effect of Seal coating with 1% HPMC and curing at 60 LC 153 for 1 hr on the release of bumetanide from pellets coated with 6 % Eudragit RS plasticized with 20 % triethyl citrate
Figure 41: Effect of seal coating with 1% HPMC and curing at 60 LC 155 for 4 hr on the release of bumetanide from pellets coated with 6 % Eudragit RS plasticized with 20 % triethyl citrate
Figure 42: Effect of seal coating with 1% HPMC and curing at 60 LC 157 for 8 hr on the release of bumetanide from pellets coated with 6 % Eudragit RS plasticized with 20 % triethyl citrate
Figure 43: Effect of seal coating with 1% HPMC and curing at 60 LC 160 for 24 hr on the release of bumetanide from pellets coated with 6 % Eudragit RS plasticized with 20 % triethyl citrate
Figure 44: Chemical Structure of Eudragit RS and RL 163 Figure 45: Effect of curing time at 60L C on the release of bumetanide 166 from pellets coated with 6 % Eudragit RL/RS (10/90)
12 Figure 46: Effect of HPMC seal coating and storage time at room 168 temperature on bumetanide release from pellets coated with 6 % Eudragit RS plasticized with 20 % triethyl citrate and completely cured
Figure 47: Effect of sodium chloride on the release of bumetanide 172 from pellets coated with 6% Eudragit RS plasticized with 20% triethyl citrate
Figure 48: Release profiles from mixture of pellets coated with 6% 173 Eudragit RS plasticized with 20 % triethyl citrate with and without sodium chloride compared to target profile
Figure 49: Effect of formulation on the cumulative urine output (ml/Kg) 183 and bumetanide excretion (ug/Kg) in rabbits (n=8)
13 1. Introduction
1.1. Definition
The term “loop diuretics” is used to describe a group of molecules that block the reabsorption of sodium, potassium and chloride from the ascending limb of Henle in the kidney nephron. This part of the nephron has a high reabsorptive capacity thus, loop diuretics induce intense diuresis and are referred to as high ceiling diuretics. They act from the luminal side of the urinary tubules therefore, they must be filtered through the glomerulus or excreted through the organic acid transport mechanism in the proximal tubules to exert their effect.
Furosemide (frusemide) was the first loop diuretic introduced to the world market in the 1960s. Other loop diuretics include bumetanide, torsemide, ethacrynic acid, azosemide, muzolimine, piretanide and tripamide. Only furosemide, bumetanide, ethacrynic acid and torsemide are approved in United
States. Loop diuretics are mainly used for edema associated with congestive heart failure, hepatic cirrhosis and renal impairment diseases (Ives, 1998).
1.2. Pharmacodynamic Rationale for Formulating Loop Diuretics as
Extended Release Dosage Forms
Pharmacodynamic models are mathematical models used to describe the time-independent equilibrium relationship between concentration and effect.
The sigmoid Emax model is most commonly used to describe the effect of loop diuretics. The model is usually written as:
14 s s s E = Emax .[ C / ( C 50 + C )] (1)
Where,
E= Effect; C= Concentration (excretion rate in case of diuretics);
Emax = Maximal effect; C50 = Concentration corresponding to effect of 50% of
maximal effect ; S = Number influencing the slope.
To correct for the base line effect equation (1) can be rewritten as
s s s E = E0 + Emax . [ C / ( C 50 + C )] (2)
Where, E0 is the baseline effect.
The model involves the concept of diminishing return where increasing
the drug concentration produces a gradual saturation of the system expressed
as a decrease in effect until a plateau is reached (Holford and Sheiner, 1981).
The diuretic effect and efficiency (effect per unit stimulus i.e. amount of
drug at the receptor site) of intravenous furosemide were studied in humans.
Furosemide was given as a push intravenous dose of 5 mg/Kg to five healthy
volunteers and diuresis was assessed by measuring the urine volume and its
content of sodium, potassium and chloride over a 5-hr period. The calculated
diuretic effect corresponding to maximum diuretic efficiency calculated using the
Sigmoid Emax model was 1.67-2.33 times greater than that obtained in the
study with the intravenous push mode of administration. Therefore, It was
concluded that Intravenous push of furosemide is the least efficient mode of
administration while a controlled input maintaining furosemide excretion at a
rate that maintains maximum diuretic efficiency will give higher urine output.
15 However, the calculations did not consider possible drug tolerance common with loop diuretics (Alvan et al., 1990).
The importance of improving the diuretic response of loop diuretics arises from the fact that edematous patients are more resistant to loop diuretics than normal subjects. In patients with renal impairment diseases, the decrease in response is due to decrease in the amounts of bumetanide excreted into the urinary tubules as well as decrease in the amounts of sodium filtered through the glomerulus (Beerman, 1984; Voelker et al., 1987). In patients with liver cirrhosis, the delivery of loop diuretics to the urinary tubules is similar to normal subjects (Fuller et al., 1981; Schwartz et al., 1993). The exact mechanism for the decreased response to loop diuretics in patients with liver cirrhosis is still unclear (Brater, 1998). Therefore, more frequent dosing of smaller doses is preferred for the therapy of edema in these patients. Similarly, patients with congestive heart failure have comparable delivery of loop diuretics into the urinary tubules to normal subjects however, the response is significantly decreased (Brater et al., 1980; Vargo et al., 1995).
The intravenous push of bumetanide was compared to the slow intravenous infusion in chronic renal insufficient patients who usually show diuretic resistance owing to their reduced glomerular filtration rate (Rudy et al.,
1991). In the study, bumetanide was administered as intravenous push dose of 12 mg divided into two 6-mg doses separated by a 6-hr interval. The same dose of bumetanide was also administered via intravenous infusion over 12 hr.
The results showed comparable amounts of bumetanide excreted in the urine
16 using the two modes of administration. However, the infusion mode induced significantly more saluresis (25 % increase) over the push mode. Moreover, no signs of myalgia, the main side effect of bumetanide, were noticed in patients receiving the infusion while 3 out of 8 patients receiving the push dose experienced myalgia. The increased saliuretic effect with the infusion mode suggested that this pattern of drug administration produced more efficient drug utilization than the intermittent push dosing. Tolerance developed to both modes (Rudy et al., 1991).
The effect of infusion time on the pharmacokinetics and pharmacodynamics of the same total dose of bumetanide was investigated in rabbits (Ryoo et al., 1993). In the study, 2 mg/Kg bumetanide were infused to rabbits over 10 sec (treatment I), 1 hr (treatment II) and 4 hr (treatment III). It was found that the area under the plasma concentration-time curve decreased with treatment II and III compared to treatment I due to an increase in the total drug non-renal clearance. The results were explained in terms of saturable drug metabolism, where the slow infusion did not allow for metabolism saturation with a consequent increase in the drug metabolism. The percent of unchanged bumetanide excreted in urine over 24 hr decreased with treatment II and III (longer infusion times) as a result of the increased metabolism. However, there was an increase in the urine volume output and the urinary excretion of sodium and chloride. The results were attributed to improved diuretic efficiency in case of the slow intravenous infusion compared to the rapid infusion (Ryoo et al.,1993).
17 It has been shown that oral administration of bumetanide and furosemide resulted in a lower urinary excretion of both drugs than that obtained with intravenous administration, yet similar diuretic and saliuretic effects were obtained with both routes of administration (Kaojarene et al., 1982; Lau et al.,
1986).
Two extended release furosemide formulations were compared to immediate release tablets in humans (Alvan et al., 1992). It was shown that the total diuretic/saliuretic effect of the extended release formulations were only marginally lower (19-28 %, p 0.05) than the immediate release in tablets albeit the drug fractions absorbed from the extended release formulations were markedly lowered (39-51 %, p0.05). The decreased furosemide bioavailability was attributed to the limited absorption of the drug as furosemide is absorbed from the stomach and the upper part of the small intestine. It was concluded that extended release furosemide peroral formulations produced better diuretic efficiency than the immediate release tablet. The total diuretic effect was essentially similar yet the extended release formulations had the advantage of avoiding the sudden burst of diuretic effect (Alvan et al., 1992).
A floating monolithic system of furosemide was developed and tested in beagle dogs in an attempt to improve the bioavailability of furosemide from extended release formulations (Menon et al., 1994). The system aimed to increase the residence time of the furosemide tablet in the stomach by decreasing its density relative to the content of the stomach. In their results, furosemide bioavailability increased significantly when it was formulated as
18 floating monolithic formulation compared to the marketed extended release formulations (Menon et al., 1994).
Extended release bumetanide capsules were compared to immediate release tablets in humans (Yagi et al., 1999). Equivalent urine volume outputs were obtained after administration of the extended release capsules and the immediate release tablets. However, the plasma and urine level of bumetanide with the extended release capsules were less than those obtained with immediate release bumetanide tablets (Yagi et al., 1999).
The results suggested that extended release bumetanide formulation offered better utilization of the drug. The extended release granules used for the study had an extended release profile in hydrochloric acid pH=1.2 (about 30 % released in 4 hr). Nevertheless, the granules had sudden eruption of drug release at pH=6.8 (about 60 % drug released in less than 0.5 hr) (Yagi et al.,
1997). Such in vitro release profiles suggested that these granules may have slow bumetanide release in the stomach (pH= 1.5-3.5) but once introduced to the upper part of the small intestine (pH value = 5-7), a sudden burst of drug release would be anticipated. Therefore, the granules used in the study are not expected to have extended release pattern throughout the gastrointestinal tract. Thus, any comparison between extended and immediate release bumetanide formulations based on the study can be misleading.
Based on these data, extended release peroral formulations of loop diuretics are expected to have greater therapeutic effect and efficiency than immediate release formulations. Extended release formulations also have the
19 additional advantage of avoiding an initial intense diuretic effect. Nonetheless,
further systematic studies comparing peroral extended to immediate release loop
diuretics formulations are needed to better understand how the change in the
rate of drug delivery to receptor sites affect the diuretic/saliuretic outputs. Loop
diuretics that can be absorbed from different segments of the gastrointestinal
tract will serve as better candidates for such studies allowing equivalent delivery
to the urinary tubules yet at different rates.
1.3. Rationale for Selecting Bumetanide as Model Drug for the Study
Chemically, bumetanide is anthranilic acid derivative (Figure 1). It is a
white powder having a calculated molecular weight of 364.41. It is highly
lipophilic and practically insoluble in water but soluble in methanol and alkaline
solutions. It has two ionization constants with Pk values of 3.3 and 7.7 (Tata et
al., 1993).
Bumetanide is one of the well-known loop diuretics in the United States
market. It is available under the brand name Bumex from Roche in different strength tablets including 0.5, 1, and 2 mg. It is prescribed once or twice daily for edema associated with congestive heart failure, renal insufficiency or hepatic cirrhosis. There is no extended release bumetanide dosage form available in the market. The desired biopharmaceutical and pharmacokinetic characteristics of drugs that qualify them for extended release formulation (Ritschel, 1989) as well as the reported bumetanide characteristics are shown in Table 1.
20
CH3(CH2)3HN
C6H5O COOH
SO2NH2
Figure 1. Chemical structure of bumetanide (3-(butylamino)-4-phenoxy-5- sulfamoylbenzoic acid)
21 Table 1. The desired pharmacokinetic and biopharmaceutical characteristics for drug candidate suitable for extended release formulation as well as the recorded bumetanide characteristics
Characteristics Desired Characteristics Bumetanide Characteristics
Molecular Less than 1000 M.W.= 364.41 weight and size
Solubility Drug solubility less Solubility in water and alkaline
than 0.1 ug/ml causes solution is greater than 0.1 ug/ml.
variability and lowered The rapid onset of diuresis
bioavailability. following oral administration of
Solubility less than 0.01 bumetanide rules out the potential
ug/ml causes the solubility related problems
bioavailability to be regarding drug absorption in the
dissolution rate limited. stomach.
In general, solubility
should be greater than
0.1ug/ml for pH range
of 1-7.8.
22 Table 1 (Continued). The desired pharmacokinetic and biopharmaceutical characteristics for drug candidate suitable for extended release formulation as well as the recorded bumetanide characteristics
Characteristics Desired Characteristics Bumetanide Characteristics
Absorption and The drug should be Bumetanide was absorbed from pharmacokinetics absorbed from different different segments of the
segments of the gastrointestinal tract of rats.
gastrointestinal tract. Absorption was fast, about 80%
of the eventually absorbed dose
in 60 minutes were absorbed in
the first 20 minutes from different
segments except the large
intestine. Bumetanide was more
absorbed from acidic medium
from all segments (Lee, Lee, and
Kim, 1994).
Stability Drug should be stable Bumetanide was found stable in
in the gastrointestinal different segments of the
tract with minimum gastrointestinal tract of rats. (Lee,
tendency for first pass Lee and Kim, 1994).
effect and metabolism
by liver enzymes.
23 Table 1 (Continued). The desired pharmacokinetic and biopharmaceutical characteristics for drug candidate suitable for extended release formulation as well as the recorded bumetanide characteristics
Characteristics Desired Characteristics Bumetanide Characteristics
Partition coefficient High Very high
Absorption mechanism Diffusion Unknown
Elimination half-life 0.5-8 h 1-1.5h (Marcantonio et al., 1982; Brater et al., 1983) Bioavailability More than 75% 74-95 % (Marcantonio et al.,
1982; Holazo et al., 1984)
Intrinsic absorption Should be higher than Absorption rate constant
-1 rate constant (Ka) release rate (Ka) = 2.72 h (First order)
(Marcantonio et al., 1982;)
Volume of distribution Should not cause the Effective dose is 2 mg and steady state effective dose of the plasma level drug to exceed 500 mg
Therapeutic window The wider the better Patients receiving high
doses of bumetanide (up to
10 mg/ day) experienced no
life-threatening symptoms
(Stone et al., 1981; Koene,
1982).
24 The effect of food on the bioavailability of bumetanide and furosemide was studied in humans (McCrindle et al., 1996). Food caused a significant decrease in the bioavailability of furosemide but did not significantly affect that of bumetanide. The results suggested that bumetanide may not follow the same pattern of absorption as that of furosemide where the drug is only absorbed from the stomach and the upper part of the small intestine (McCrindle et al., 1996).
In conclusion, most of the pharmacokinetic and biopharmaceutical characteristics of bumetanide suggest its suitability for extended release formulations.
1.4. Peroral Extended Release Formulation Design Strategy
Peroral extended release formulations are defined as formulations from which the drug release is controlled over certain period of time. They are intended for administration via the oral route. According to the mechanism of drug release, peroral extended release formulations are classified to:
1.4.1. Diffusion-Controlled Extended Release Formulations
The release of the drug is controlled predominantly by its diffusion through a water insoluble polymeric layer. Drug dissolution also contributes to the release of the drug but to a lesser extent.
1.4.1.1. Reservoir Devices
Extended release formulations where film coating constitutes the main factor in controlling drug release. They have been introduced into the
25 pharmaceutical industry by SmithKline in the late 1940s (Porter and Bruno,
1990). The first application involved the use of pan-coating process to apply various mixtures of fats and waxes to drug-loaded beads. Since then, a variety of coating materials and coating machines have been developed and modified.
Examples of materials used to control drug release include hardened gelatin, methyl or ethyl cellulose, polyhydroxymethacrylate, methacrylate ester copolymers, and various waxes. Ethyl cellulose and methacrylate ester copolymers are the most commonly used systems in the pharmaceutical industry
(Venkatraman et al., 2000).
1.4.1.2. Matrix Devices
Extended release formulations in which the drug is uniformly distributed through the release-controlling element. Two major types of materials are used in the pharmaceutical industry to control the drug release from matrix devices; insoluble plastics and fatty compounds. Examples of insoluble plastics include methylacrylate-methyl methacrylate copolymers, polyvinyl chloride, and polyethylene. Fatty compounds used include carnuba wax and glyceryl tristearate (Venkatraman et al., 2000).
1.4.2. Dissolution-Controlled Extended Release Formulations
Extended release formulations in which drug release is mainly controlled by the slow dissolution or erosion of the release-controlling element in the formulation. The system can be formulated into reservoir type by encapsulating
26 the drug within slowly soluble polymeric membrane in the form of tablets or small beads (spansules ). Another approach is to formulate the drug into a matrix system using hydrophilic swellable polymers as hydroxypropyl cellulose, hydroxypropyl methylcellulose, methylcellulose, and sodium carboxymethylcellulose (Venkatraman et al., 2000). A major draw back in the dissolution-controlled system is the difficulty to maintain a constant drug release rate since the release rate changes as the size of dosage form diminishes with time.
1.4.3. Osmotic-Controlled Extended Release Formulations
They are reservoir systems into which osmotically active agents are incorporated (if the drug itself is not osmotically active) in the formulations. The dosage form is coated with a semipermeable membrane through which hole(s) are carefully drilled. Gastrointestinal fluid diffuses through the membrane and dissolves the osmotic agent(s) creating high osmotic pressure inside the reservoir. Subsequently, water convects out of the reservoir through the hole(s) under the osmotic pressure difference established carrying the drug. Drug release follows zero order kinetics independent of the pH of the gastrointestinal tract. Examples of materials used as semipermeable membranes include polyvinyl alcohol, polyurethane, cellulose acetate, ethyl cellulose, and polyvinyl chloride. The first osmotic system known as elementary osmotic pump was introduce by Theeuwes and Higuchi in the 1970s. Different types of osmotic pumps have been developed and modified to provide zero-order delivery of varieties of drugs (Santus and Baker, 1995).
27 1.4.4. Extended Release Formulation Based on Ion Exchange Resin
Ion exchange resins are water insoluble, cross-linked polymers containing salt-forming group in repeating positions on the polymer chain. Drug is bound to the resin and its release will depend on the ionic environment including pH and electrolyte concentration of the gastrointestinal tract as well as the properties of the resin. Drug is released through an exchange process with an appropriately charged molecules in the gastrointestinal tract followed by diffusion of the drug out of the resin. Most ion exchange resins used contain sulfonic groups, which can exchange cationic drugs with amine functions (Martin, 1993).
1.5. Multiparticulate Drug Delivery System
1.5.1. Definition and Advantages
Multiparticulate drug delivery systems can be defined as drug delivery systems in which each unit dose is comprised of many entities together shape the drug release profile. By definition, they include different size particles including granules, pellets, micro-capsules, micro-particles, nano-capsules, and nano-particles. Nonetheless, the term has been commonly used in the pharmaceutical literature to describe coated pellets with size range between 0.1-
1 mm (Lehmann, 1997).
They have found wide applicability in formulating peroral extended release formulations since they possess several advantages over single-unit dosage forms. When the particles, often referred to as pellets, are administered orally, they disperse freely in the gastrointestinal tract, which invariably maximize drug
28 absorption, reduce peak plasma fluctuation and minimize potential side effects without appreciably lower drug bioavailability. They also reduce variation in gastric emptying rates and overall transit times. Thus, inter and intra-subject variability of plasma profile are minimized. Another advantage over single-dose tablets and capsules is that high local concentration of the drug, which may be irritant, can be avoided. Multiparticulate drug delivery systems are also less susceptible to dose dumping than the reservoir type single unit dosage forms
(Bodmeier, 1997).
In addition, greater flexibility in designing extended release formulations is affordable using multiparticulate drug delivery systems. Different release profiles can be tailored by appropriate selection of mixing ratios of pellets with faster and slower release profiles. Nevertheless, the use of coated multiparticulate as extended release drug delivery systems carries challenges for the formulators.
Owing to the increased surface area of multiparticulate system, the ability of the sprayed polymeric film to uniformly cover the particle surface and hence the reproducibility of drug release becomes crucial. Multiparticulate coating process is usually done in fluid bed coaters however, perforated coating pans are still preferred by some formulators to coat pellets as well as tablets.
1.5.2. Manufacture of Multiparticulate Drug Delivery Systems
Peroral multiparticulate extended release formulations are available as both reservoir and matrix pellets usually filled into capsules or sometimes compressed into tablets. Reservoir multiparticulate extended release
29 formulations are more common in the pharmaceutical industry. Coated pellets are more robust in terms of drug release reproducibility than the small matrix since the distance traveled by the drug through the matrix is relatively small to allow a precise control of drug release.
The manufacture of the reservoir type multiparticulate drug delivery systems involves two distinctive steps, manufacturing the drug-loaded pellets commonly referred to as pelletization and coating the manufactured drug pellets.
The most widely used pelletization processes in the pharmaceutical industry are extrusion/spheronization, solution/suspension layering, and powder layering.
Drug layering from solution, suspension or powder is usually conducted in fluid bed equipment. The drug-loaded pellets can be coated with the release controller using fluid bed or coating pans. Other techniques used to manufacture pellets include spray-drying, spray-congealing and balling but they have not found wide applicability in the pharmaceutical industry (Ghebre-Sellassie, 1989).
1.5.2.1. Extrusion/ Spheronization Process
Extrusion is a method of applying pressure to a wet mass containing the drug and other formulation excipients until the mass flows through orifices of defined opening to produce extrudates of rod shape. Several extrusion devices are used in the pharmaceutical industry including screw, sieve and basket, roll, and ram extruders. The classification is based on the mechanism by which pressure is applied to the wet mass (Hicks and Freese, 1989).
Spheronization is the process by which the extrudates are converted into spherical pellets of certain particle size. The marumerizer "round maker" is the
30 most commonly used spheronizer in the pharmaceutical industry. The equipment includse a fast rotating friction plate fitted into a vertical hallow cylinder into which the extrudates are introduced. The extrudates breakdown into short spherical segments by contact with the rotating friction plate, collision between particles, and collision with wall of the cylinder and the rotating baffles (Hicks and Freese,
1989).
1.5.2.2. The Use of Fluid Bed Equipment in Pelletization and Coating of
Multiparticulate Drug Delivery Systems
Since its first introduction into the pharmaceutical industry by Wurster in
1959 as a coating machine, fluid bed equipment has found wide spread applications in different pharmaceutical manufacturing processes including granulation, drying, pelletization and coating. One of the major advantages of fluid bed equipment is the limited number of processing steps needed as layering/drying and coating/curing can be done in the same equipment. Other advantages included reduced capital cost, labor, processing time, dust production and cross contamination. Nevertheless, associated with fluid bed are a limited numbers of drawbacks including the risk of explosion and the need of skillful and experienced operators.
Fluid bed processes depend on suspending the materials contained in a conical cylinder in a stream of upward moving preconditioned air. The main parts of fluid bed equipment are shown in Figure 2 (page 33). The filtered fluidizing air is dehumidified and then rehumidifed to carry certain moisture content. The air is
31 heated to a predetermined temperature then finally filtered through HEPA filters prior to introduction to the product bed through an air distribution plate especially perforated to control the motions of the materials within the product bed. The liquid, which can be drug-loaded binder solution or suspension or polymeric solution or dispersion is sprayed on the material while fluidized (Olsen, 1989).
According to the pattern of spraying fluid bed equipment fall into three categories: top spray, bottom spray (Wurster) and tangential spray.
1.5.2.2.1. Types of Fluid Bed Equipment
According to the spraying patterns fluid bed equipment can be classified to:
1.5.2.2.1.1. Top Spray Fluid Bed
The spraying nozzle is located on top of the product chamber. The liquid is sprayed in the opposite direction to material motion. It is mainly used for drying with short processing time owing to the increased contact of the dry hot air with the moist fluidized materials. It is also used for granulation with the advantages of minimizing dust production. It produces free-flowing and porous granules suitable for drug with limited solubility (Lipps, 1991). Top spray fluid bed equipment can also be used for aesthetic coating and hot melt layering/coating (Guirgis et al.,
2001).
32
I
H
A G
C
D B E
F
Figure 2. Fluid bed equipment with the Wurster insert; A. air handling unit; B. air distribution plate; C. product chamber; D. partition chamber; E. spray nozzle; F. binder or polymeric solution/dispersion reservoir; G. expansion chamber; H. filter housing; I. exhaust fan.
33 1.5.2.2.1.2. Bottom Spray Fluid Bed (Wurster Coater)
The spraying nozzle is located at the bottom of the product chamber as seen in Figure 2. The liquid is sprayed in the same direction as the material motion in the partition chamber. Such layout allows more contact between the liquid and the solid. It is used in solution/suspension layering and polymeric coating. It offers the advantage of producing uniform coat thickness, which is very crucial for extended release coating to obtain reproducible release profiles
(Hutchings, 1993; McPhillips et al., 1997; Habib and Sakr, 1999).
1.5.2.2.1.3. Tangential Spray Fluid Bed (Rotary Fluid Bed)
The spray nozzle is located on the side of the product bed and liquid is sprayed in a tangential pattern while the materials are in motion. A rotary disc replaced the air distribution plate whose motion together with that introduced by the fluidizing air and the gravitational pull cause the materials to move in rope- like motion. It can be used for spheronization, granulation, pelletization, solution/suspension layering, powder layering, and coating (Turkoglu, 1995).
1.5.2.3. Coating of Multiparticulate Systems
Coating can be conducted by spraying polymeric solution, suspension or molten materials on the particles. Owing to the increased awareness of environmental hazards of organic solvents, coating processes using polymeric solution in organic solvents have been replaced by aqueous based coating systems. Since the introduction of aqueous based polymeric dispersions by the
34 Emulsion Polymer Institute at Lehigh University in 1979 (Vanderhoff and El-Asser
, 1979), the pharmaceutical industry has shifted toward the newly introduced technology. In the last decade, hot melt coating techniques have found some applicability in the pharmaceutical industry. It offers the advantages of avoiding the energy-consuming evaporation step, which shortens the processing time.
(Guirgis et al., 2001) Nevertheless, the complexity of the process and the instability of the coating material upon storage have limited the wide spread applicability of hot melt technology in the pharmaceutical industry. Currently, coating using aqueous polymeric dispersion is considered as the standard practice in the pharmaceutical industry.
1.5.2.3.1. Coating of Multiparticulate Systems Using Aqueous Polymeric
Dispersions
Coating machines used in the pharmaceutical industry include bottom and tangential spray fluid bed equipment and vented coating pans with the former finding wider application in the coating of multiparticulate systems. Fluid bed equipment offers several advantages over coating pans as a multiparticulate systems coating machine. Since fluid bed equipment is a closed system, it allows vigorous fluidization of small particles without the risk of dust production and material loss. Vigorous fluidization is needed to circumvent the high risk of product aggregation during coating. Fluid bed equipment also has an efficient heat transfer system suitable for use with aqueous based polymeric dispersion
35 as it can provide the high energy needed to evaporate the (Williams and Liu,
2000.)
1.5.2.3.2. Factors to be Considered in Coating of Multiparticulate Systems
in Fluid Bed Equipment
1.5.2.3.2.1. Film Forming Temperature
Several theories have been proposed to describe the mechanism of film
formation after spraying the polymeric dispersions. In general, the polymeric
particles deform and fuse together at the surface of the particle to be coated to
form complete and smooth film. Therefore, polymeric particles must be flexible in
order for the film formation process to be complete. Polymer flexibility increases
dramatically when the glass transition temperature (Tg) is exceeded. Glass
transition temperature can be defined as the temperature at which the polymer
changes from the rubbery to the glassy state. Nevertheless, Increasing the bed
temperature far beyond the (Tg) may lead to particle agglomeration. Film formation usually occurs at temperature below the glass transition temperature due to the presence of water and other materials within the film that can increase polymer softness and mobility. Therefore, minimal film forming temperature can be more meaningful than the glass transition temperature in optimizing the coating process. Minimal film forming temperature is defined as the temperature above which a continuous film is formed under distinct drying conditions.
Minimal film forming temperature is lower than the glass transition temperature however, it varies with evaporation rate and hence it is very sensitive to bed
36 condition. Fluid bed temperature should be maintained at 10-20 L C above
minimal film formation temperature (Lehmann, 1997).
1.5.2.3.2.2. Plasticizers
Plasticizers are organic compounds that are used to lower the polymer
glass transition temperature and the minimal film forming temperature allowing
coating process to be conducted at lower temperatures. Plasticizers effects
extend to other film properties as plasticizers decrease the tensile strength and
elastic modulus, increase or decrease film permeability and increase the film
adhesion properties. Plasticizer selection is based on the type of polymer as the
plasticizers have to be compatible with polymer since strong interaction in
between the plasticizer and the polymer is a prerequisite for any effect. For
example, triethyl citrate, a water-soluble plasticizer, is recommended for use with
methacrylate ester copolymer (Eudragit RS and RL). Other plasticizers used
include acetyl triethyl citrate, tributyl citrate, acetyl tributyl citrate, diethyl
phthalate, polyethylene glycols and triacetin. Only triethyl citrate, triacetin and
polyethylene glycols are dubbed water-soluble plasticizers. The effect of addition
of 10 and 20 % plasticizers on the minimal film forming temperature of Eudragit
RS and RL is shown in Table 2 (Lehmann, 1997). The minimal film forming
temperature of the pure polymeric dispersion is around 40-50 LC. As seen in table, different plasticizers reduce the minimal film forming temperature to different extents. Triethyl citrate is the most efficient plasticizer in reducing the minimal film forming temperature.
37 Table 2. Film forming temperatures of Eudragit RL/RS as aqueous dispersions with added plasticizer (Lehmann, 1997)
Water Minimal Film Forming Temperature
Solubility at (LC)
room Eudragit RL 30 D Eudragit RS 30 D
temperature Plasticizer Level Plasticizer Level (%) 10% 20% 10% 20%
Triethyl citrate 6.9 11 0 20 5
Acetyl triethyl citrate 0.72 18 0 23 8
Tributyl citrate 0.002 17 4 22 8
Acetyl tributyl citrate 0.002 22 10 25 10
Triacetin 7.1 8 0 22 3
Diethyl phthalate 0.15 15 0 22 5
Polyethylene glycol 6000 30 39 34 46 45
38 1.5.2.3.2.3. Fluidization Air Temperature
The most suitable inlet air temperature is the one that allows equilibrium between the application of liquid and the subsequent evaporation. However, relatively low temperature should be used to prevent spray drying. The temperature of the inlet air has two functions, to evaporate the solvent and to soften and help the polymeric particles fusion. Bed temperature should be higher than minimal film forming temperature but not too high to cause excessive spray drying and product agglomeration. For example, bed temperature 20LC above minimal film forming temperature can lead to drying of the sprayed droplet and the inability of polymeric particles to fuse together to form a continuous smooth film (Yang and Ghebre-Sellassie, 1990)
1.5.2.3.2.4. Spray Rate
Polymer spray rate is one of the most critical variables in film coating process in fluid bed equipment as it can affect the degree of wetting and droplet size. At a given atomizing air pressure, increasing the liquid spray rate increases the droplet size and the possibility of powder overwetting. Slowing the spray rate may cause electrostatic problems owing to low bed humidity especially at high temperature. Additionally, the decreased humidity can lead to drying of the sprayed droplet with reduced particle fusion (Watano et al., 1995).
39 1.5.2.4. Curing of Coated Multiparticulate System
The formation of a continuous smooth polymeric film after coating using aqueous based polymeric dispersion is challenging. Conflicting results have been reported in the literature regarding the effect of long-term storage on drug release from coated multiparticulate drug delivery systems. Some investigators reported an increase in drug release (Bodmeier and Paeratakul, 1994) and other reported the opposite (Goodhart et al, 1984; Yuen et al, 1993). Therefore, in depth understanding of the processes involved in film formation specially after the completion of the coating cycle is necessary to predict the behavior of the coated dosage form upon storage.
The film formation process necessitates the fusion and coalescence of the sprayed polymeric particles whose rate can varies dramatically depending on the coating conditions and post coating treatment. The rate of water evaporation from the forming polymeric film must be carefully controlled by controlling the bed temperature, amount and humidity of fluidizing air. Gradual water evaporation is important to establish the driving force for polymeric particle coalescence process. The final step in the film formation process is called the further gradual coalescence and can proceed for long and variable periods of times depending on formulation factors as the level of plasticizer used as well as the coating conditions (Amighi and Moes, 1996). Thus, to complete the process of film formation and to avoid further change in drug release upon storage coated multiparticulate dosage forms are cured. Curing can be defined as the input of energy into the film coated system after the desired film coat level is applied, and
40 it is during the curing step that polymeric particles coalescence continues to complete the film formation process. The most commonly used method of curing is heating in ovens or fluid bed equipment.
Curing of extended release formulation is more critical than any other coated formulation owing to the sensitivity of the release profiles to the curing conditions. Plasticizer level, curing time and temperature as well as the humidity level are thought to affect the response of coated pellets to curing (Bodmeier and
Paeratakul, 1994; Hutchings et al., 1994; Amighi and Moes, 1995; Williams and
Liu, 2000). The use of higher curing temperature allows the completion of film formation at shorter curing time (Bodmeier and Paeratakul, 1994; Hutchings et al., 1994). Increasing plasticizer level facilitates the film formation process and can negate the need for longer curing times (Amighi and Moes, 1995). Curing at higher humidity level was found to be more effective in facilitating film formation as water acts as plasticizer enhancing polymeric particles coalescence (Williams and Liu, 2000). Another important factor that can affect the changes in drug release from coated multiparticulate upon storage is the drug physicochemical properties in particular its degree of lipophilicity. Highly lipophilic drug can migrate to the film coat during curing with a subsequent increase in drug release.
Despite the importance of this factor, very limited studies have focused on the effect of lipophilicity on drug release upon curing. More systematic studies are needed to understand how drug lipophilicity affects its release characteristics from coated multiparticulate upon storage.
41 1.6. Optimization of Drug Release from Coated Multiparticulate Systems
Using Statistical Modeling
Designing extended release formulations with the minimum number of trials is very crucial for the pharmaceutical scientists. Statistical methods based on second order polynomial equations have been used in developing and optimizing extended release formulations (Turkoglu and Sakr, 1992; Gohel and
Amin, 1998; Sastry et al., 1998). Polynomial equations relate the drug release process to formulation and/or operational parameters and can be used to predict the optimum formulation and/or operational parameters needed to obtain the desired release. Nevertheless, complex relationships that can’t be modeled using second order polynomial equations are common in the design of extended release formulations. Therefore, artificial neural networks have been used to model the effect of different formulation/operational variables on drug release
(Takahara et al., 1997; Chen et al., 1999; Turkoglu et al., 1999; Takayama et al.,
2000). Artificial neural networks simulate the biological neurons complex networking and can model different kinds of complex relationships (logarithmic, exponential, etc).
The use of carefully designed statistical scheme limits the number of trials while maintaining the predictability of the derived statistical models. The derived models based on polynomial equations or artificial neural networks are used to build response surfaces that can be used to visualize the effect of formulation and/or operational parameters on the release as well as to optimize the release process, a technique commonly referred to as response surface methodology.
42 Many of the pharmaceutical literature use single point optimization where the release process is optimized with regard to the percent of the drug released after a single time point. Single point optimization could lead to misleading results as different release profiles can have the same percent of drug released after single time point. On the other hand, multiple response optimization methods aim to optimize the release process after more than one time point.
Thus, designing extended release formulation using multiple response optimization is more useful and practical than single point optimization.
Unfortunately, very limited number of reports addressed the use of multiple point optimization techniques in the pharmaceutical industry.
In summary, extended release peroral loop diuretics formulations are expected to improve drug utilization and increase the diuretic response. Only furosemide extended release peroral formulations were thoroughly investigated in comparison to immediate release formulations. Despite better diuretic efficiency, the lower bioavailability of furosemide from extended release formulations caused them to have comparable diuretic effect to immediate release formulations. More systematic studies are needed to investigate other loop diuretics whose bioavailabilities won’t hinder the performance of extended release formulations. The use of multiparticulate drug delivery systems is favored in formulating extended release formulations. Nevertheless, the stability of drug release from such systems after storage is still questionable. The effect of drug lipophilicity on the stability of the release is still ambiguous. Therefore, more fundamental research of these systems is needed in order to improve its stability
43 after long-term storage. Finally, optimizing drug release from peroral extended release formulations is of utmost importance for the pharmaceutical industry. The use of single response optimization methods as reported in the literature doesn’t fulfill the true goals of release optimization and multiple response optimization can serve as a better alternative. However, a very limited number of publications have addressed the use of such technique to optimize peroral extended release formulations.
44 2. Hypothesis
2.1. Peroral extended release bumetanide formulations can be developed and
statistically optimized using fluid bed technology and multiple response
optimization techniques.
2.2. Bumetanide extended release formulations have greater diuretic and
saliuretic effects than immediate release formulations.
3. Objective
The objective of the study is to develop and evaluate the pharmacodynamic behavior of extended release bumetanide peroral formulations in comparison to immediate release formulations.
4. Specific Aims
1. Developing immediate and extended release formulations for bumetanide
using fluid bed technology including drug layering and coating.
2. Studying the effect of coating formulation variables on the release of
bumetanide from the coated pellets.
3. Optimizing bumetanide release from the coated pellets using response
surface analysis and multiple-response optimization procedures.
4. Testing immediate and extended release formulations for their diuretic and
saliuretic effects in laboratory animals.
45 5. Experiment and Methodology
5.1. Materials and Supplies
S Acetonitrile optima, Fisher Scientific, Fair Lawn, NJ
S Bumetanide USP reference standard, U.S.P.C., INC., Rockville, MD
S Bumetanide, Laboratori MAG, Milano, Italy
S Flat top microcentrifuge tubes (1.5 ml polypropylene), Fisher Scientific,
Pittsburgh, PA
S Flat top microcentrifuge tubes (2 ml polypropylene), Fisher Scientific,
Pittsburgh, PA
S Glacial acetic acid HPLC grade, Fisher Scientific, Fair Lawn, NJ
S Heparin sodium injection USP (1000 unit/ml), Elkins-sinn, INC, Cherry Hill, NJ
S IL Test flame photometry standard 100 mmol Na/L 100 mmol K/L,
Instrumentation Laboratory Company, Lexington, MA
S Male white new-zealand rabbits, Myrtle's Rabbitry Inc., Thompson Station, TN
S Membrane filters (0.45 um, Type HA), Millipore Corp., Bedford, MA
S Methanol HPLC grade, Fisher Scientific, Fair Lawn, NJ
S Nu-pariels sugar pellets, CHR Hansen ,Vineland, NJ
S Poly-ethylacrylate-methylmethacrylate-trimethylammonioethylmethacrylate
chloride (Eudragit RS 30D), Rohm GmbH, Darmstadt, Germany
S Polyvinylpyrrolidone (Plasdone K29/32), International Speciality Products, ISP
Wayne, NJ
S Potassium phosphate dibasic, Fisher Scientific, Fair Lawn, NJ
46 S Redi-Tip general purpose (101-1000 ul), Fisher Scientific, Pittsburgh, PA
S Sodium chloride injection USP (0.9%), Abbott Laboratories, North Chicago, IL
S Sodium hydroxide, Fisher Scientific, Fair Lawn, NJ
S Sterile Terumo Surflo disposable I.V. catheter (22G X1"), Terumo Medical
Corporation, Elkton, MD
S Sterile Tuberculin disposable latex free syringes (1 ml), Becton Dickinson &
Co., Franklin Lakes, NJ
S Swinex disc filter holders (25 mm), Millipore Corp., Bedford, MA
S Talc powder, J.T. Baker, Phillipsburg, NJ
S Triethyl citrate (Morflex), Morflex Inc., Greensboro, NC
S Water HPLC grade, Fisher Scientific, Fair Lawn, NJ
5.2. Equipment
S The Advanced™ 3300 Micro Osmometer, Advanced Instrument INC.,
Norwood, MA.
S Accumet 1002 pH meter, Fisher Scientific, Fair Lawn, NJ
S Beckman HPLC system consisted of 126 solvent module and 507-2
autosampler coupled with 157 Fluorescence Detector, Beckman Coulter
Instruments INC., Fullerton, CA
S Computer MAX 50 moisture analyzer, Arizona Instrument, Tempe, AZ
S Ultrasonic cleaner, Fisher Scientific, Pittsburgh, PA
S Glatt Laboratory bottom spray fluid bed (Wurster Coater), GPCG-1, Glatt Ait
Technique, Ramsey, NJ
47 S Hitachi f-2500 fluorescence spectrophotometer, Hitachi Instruments Inc.,
Naperville, IL
S Instrumentation Laboratory 943 flame photometer, Instrumentation Laboratory
Company, Lexington, MA
S Isotemp oven, Model 255G, Fisher Scientific, Fair lawn, NJ
S Marathon 21K/BR Centrifuge, Fisher Scientific, Fair Lawn, NJ
S Masterflex digistaltic digital flow controller, Cole-Parmer Instrument Co.,
Chicago, IL
S Sartorius analytical balance, Model 1702, Sartorius Corp., Bohemia, NY
S Sartorius analytical balance, Model A2005-D20, Sartorius Corp., Bohemia,
NY
S Vankel 7000 dissolution apparatus, Vankel Technology Group, Cary, NC
S Varian 4270 Integrator, Varian INC., Palo Alto, CA
5.3. Software
S ECHIP experimental design and optimization software, Echip Inc., Hockessin,
DE
S Microsoft Office 2000, Microsoft Corporation, Cincinnati, OH
S SAS statistica package, SAS Institute Inc., Cary, NC
S ClogP 4.0, BioByte Corp., Claremont, CA
48 5.4. Methods
5.4.1. Part I: Development and Optimization of Bumetanide Extended
Release Formulations Using Response Surface Methodology and Multiple
Response Optimization
5.4.1.1. Layering of Bumetanide onto Nu-pariels Sugar Pellets
Bumetanide was spray-layered onto Nu-pariels sugar pellets using
GPCG1 fluid bed equipment with the Wurster insert. Bumetanide was dissolved in an alkaline buffer prepared using sodium hydroxide and potassium phosphate dibasic (pH=12.0M0.05) together with polyvinylpyrrolidone according to the formulation shown in Table 3 (page 51). The sugar pellets (300 g) were loaded into the fluid bed equipment and the bumetanide solution was sprayed into the fluidized pellets using a peristaltic pump for final bumetanide strength of 1 %w/w.
The fluid bed equipment processing parameters are shown in Table 4 (page 52).
The parameters were selected based on our experience with the equipment and preliminary studies using plain sugar pellets.
After spraying the required amount of bumetanide solution, the pellets were dried in the fluid bed equipment at 60˚C for 10 minutes till final moisture content of 2-3 %w/w. The final moisture content of the pellets was analyzed using MAX 50 moisture analyzer.
49 5.4.1.2. Development of Bumetanide Spectrofluorimetric Analytical
Method
Bumetanide was analyzed in different media including USP purified water, glycine buffer pH= 2.9, USP hydrochloric acid buffer pH= 1.2, acetate buffer pH=
4.5, and USP phosphate buffer pH=6.4 and 6.8. Twenty-five mg of bumetanide reference standard were dissolved in 25 ml methanol. The solution was serially diluted by buffer or USP purified water to obtain the following concentrations:
0.04, 0.1, 0.2, 0.3, 0.5, 1, 2, 4, 10 and 20 ug/ml. Other formulation ingredients including sodium hydroxide, potassium phosphate dibasic, polyvinylpyrrolidone, and Nu-pariels sugar pellets were added to the media in concentrations similar to those used in the formulations with highest percentage of each ingredient.
Bumetanide solutions with the highest concentration in different media were scanned for the maximum excitation and emission wavelengths using the
F-2500 Hitachi fluorescence detector. The relative fluorescence intensity of each bumetanide solution was then measured using the corresponding maximum excitation and emission wavelengths. The calibration curve equations in each media were constructed by regressing the relative fluorescence intensities against the corresponding bumetanide solutions concentrations.
50 Table 3. Bumetanide formulation used in the layering process
Alkaline buffer:
Ingredient Percentage (W/V)
Sodium hydroxide 0.88
Potassium phosphate dibasic 1.36
USP purified water to 100
Bumetanide solution:
Ingredient Percentage (w/v)
Bumetanide 2
Polyvinylpyrrolidone 1
Alkaline buffer to 100
51 Table 4. Wurster fluid bed equipment processing conditions for the bumetanide layering process
Parameters Unit Value
Inlet air temperature LC 84-88
Outlet air temperature LC 57-59
Product bed temperature LC 60M2
Amount sprayed Gm 180
Spray rate gm/min 7
Atomizing air pressure (output) Bar 2
Nozzle diameter Mm 0.8
Filter shaking time Interval Panel Set 4
Filter shaking time duration Panel Set 0.1
Pressure drop (PD) across the bed Kpa 0.5-0.7
PD exhaust filter Kpa 0.1
Supply air pressure Bar 7
Actual air pressure Bar 6
Exhaust air flap Bar 0.4
Drying temperature LC 55-65
Drying time Min 10
Air distribution plate N/A 4700/A-6
Load Gm 300
52 5.4.1.3. Validation of Bumetanide Analytical Method
Accuracy
Three different new bumetanide solutions of known concentrations (0.652,
1.284, and 3.072 ug/ml) were prepared and their relative fluorescence intensities were measured at the corresponding maximum excitation and emission wavelengths. Accuracy expressed as percent recovery was estimated by comparing the concentrations calculated using the calibration equations to the actual concentrations using the following equation:
Recovery (%) = {Recovered (measured) concentration/ actual concentration}*100
Precision
The relative fluorescence intensities of the three different new bumetanide solutions were measured at the corresponding maximum excitation and emission wavelengths each for 3 replicates. The relative standard deviations of the measured concentrations were calculated and used to express precision.
Linearity and Range
The linearity ranges of the bumetanide calibration curves were checked by calculating the correlation coefficient of the regression equation at the different concentrations measured.
5.4.1.4. Content Uniformity Assessment
The bumetanide-loaded pellets were analyzed for their bumetanide contents using Hitachi f-2500 fluorescence spectrophotometer. Ten samples were randomly removed and crushed independently in a mortar, 100 g of the
53 crushed samples were accurately weighed into 500-ml volumetric flasks half filled with USP purified water. The dispersions were sonicated for 10 minutes in FS ultrasonic cleaner and completed to volume with USP purified water. The samples were then filtered through 0.45 um filter. Filtered samples were analyzed for their bumetanide contents in the fluorescence spectrophotometer using an excitation wavelength of 326 nm and an emission wavelength of 407 nm. Products were considered uniform in their drug contents only if they met the
USP XXV uniformity of dosage unit acceptance criteria (USP, 2002). The criteria specify that the percent drug content in each sample must be within 85-115 % of the theoretical label claim and the relative standard deviation (RSD) is less than
6 %.
5.4.1.5. Coating of Bumetanide-Loaded Pellets
The bumetanide-loaded pellets were coated with different Eudragit RS coating loads containing varying percents of triethyl citrate as a plasticizer in
GPCG 1 fluid bed equipment with the Wurster insert attached. Eudragit RS was selected as the release controlling polymeric film due to its popularity in the pharmaceutical industry and our previous experience using this dispersion.
Triethyl citrate was used as a plasticizer in this study since it is recommended by the manufacturer to be used with Eudragit RS dispersion. It is water soluble and easily mixed with Eudragit RS dispersion with minimal mixing time. Eudragit RS load was measured in terms of the percent pellets weight gain of the dry polymer
54 and the level of triethyl citrate was calculated as percent of the polymer dry weight.
Five different levels of Eudragit RS coating loads and five different triethyl citrate levels were selected according to a preplanned central composite statistical design as shown in Table 5. Central composite design is a second order design that comprises of a 2K factorial runs where K represents the number of causal factors (independent variables), (2 * K) axial runs and center runs. For this study, K was equal to two as only two causal factors namely Eudragit RS load and triethyl citrate level were considered. The axial points were selected at an = (F)1/4 to assure design rotatability. The value can be defined as the distance of the axial runs from the center of the design while F equals the number of factorial runs (2K = 4). The fluid bed equipment parameters used in the coating process are shown in Table 6 (page 57).
5.4.1.6. Testing the Release of Bumetanide from Coated Pellets
The coated pellets were tested for their release profiles in USP purified water using Vankel 7000 dissolution apparatus utilizing the USP XXV basket dissolution method (Apparatus 1) at a rotation speed of 50. Samples removed after 0.5, 1, 2, 4, 8, and 12 hr were analyzed for their bumetanide contents using fluorescence spectrophotometer at an excitation wavelength of 326 nm and emission wavelength of 406 nm.
55 Table 5. Central composite statistical design used for the study
Run Run Coded Values Normal Values # Category Triethyl Eudragit RS Triethyl Eudragit Citrate 30D Load Citrate RS 30D Level Level* Load^ 1 1 1 20 7.5 2 Factorial 1 -1 20 2.5 3 Runs -1 1 10 7.5 4 -1 -1 10 2.5 5 1.41 0 22 5 6 Axial -1.41 0 8 5 7 Runs 0 1.41 15 8.525 8 0 -1.41 15 1.475 9 0 0 15 5 10 Center 0 0 15 5 11 Runs 0 0 15 5 12 0 0 15 5
*Triethyl citrate level is measured as a percent of the dry polymer weight.
^Eudragit RS 30D load is measured as the percent pellets weight gain of the dry polymer.
56 Table 6. Wurster fluid bed equipment processing conditions for coating the bumetanide-loaded pellets
Parameters Unit Value
Inlet air temperature LC 55-60
Outlet air temperature LC 53-58
Product bed temperature LC 40M1
Spray rate gm/min 2.5
Atomizing air pressure (output) Bar 2
Nozzle diameter Mm 1.2
Filter shaking time interval Panel Set 4
Filter shaking time duration Panel Set 0.1
Pressure drop (PD) across the bed Kpa 0.5-0.7
PD exhaust filter Kpa 0.1
Supply air pressure Bar 7
Actual air pressure Bar 6
Exhaust air flap Bar About 0.4
Air distribution plate N/A 4700/A-6
Load Gm 250
57 5.4.1.7. Testing the Effect of Dissolution Media pH on the Release of
Bumetanide from Coated Pellets
Bumetanide release from pellets coated with 7.5 %w/w Eudragit RS plasticized with 20 % triethyl citrate was tested in different dissolution media including USP purified water, USP hydrochloric acid buffer pH = 1.2, acetate buffer pH = 4.5 and USP phosphate buffer pH = 6.4. Bumetanide release was also tested using USP hydrochloric acid pH = 1.2 for 2 hr followed by phosphate buffer (pH = 6.8) for 10 hr. Drug release was tested using Vankel 7000 dissolution apparatus utilizing the USP XXV basket dissolution method
(Apparatus 1) at a rotation speed of 50. Samples removed after 0.5, 1, 2, 4, 8, and 12 hr were analyzed for their bumetanide contents using fluorescence spectrophotometer at the corresponding maximum excitation and emission wavelengths.
5.4.1.8. Testing the Effect of Agitation Speed on the Release of
Bumetanide from Coated Pellets
Bumetanide release from pellets coated with 7.5 %w/w Eudragit RS plasticized with 20 % triethyl citrate was tested in USP purified water using
Vankel 7000 dissolution apparatus utilizing the USP XXV basket dissolution method (Apparatus 1) at a rotation speed of 50 and 100. Samples removed after
0.5, 1, 2, 4, 8, and 12 hr were analyzed for their bumetanide contents using fluorescence spectrophotometer.
58 5.4.1.9. Statistical Analysis and Optimization of Bumetanide Release
Different release parameters have been reported in the literature as response variables to describe and optimize the release process. The most commonly used parameters include the percent of drug released at certain time point, the time needed to release a certain percent of the drug, dissolution curve shape factor, dissolution curve slope factor, dissolution rate constant, and the release order (Turkoglu and Sakr, 1992; Sastry et al., 1997; Takahara et al.,
1997). For this study, only the percent of drug released with time was considered, as it is the most meaningful parameter and represents the key parameter for any in vitro/in vivo correlation process (Food and Drug
Administration, 1997a).
The percent of drug released after 1, 4 and 8 hr were selected as the response variables. The three time points were selected to ensure full description of the release profile. The 1-hr time point describes the initial phase of drug release and detects any dose dumping while the 8-hr time point ensures that most of the drug is released in a period of time comparable to the gastrointestinal residence time. Statistical analysis including multiple regression and response surface analysis were conducted using SAS and ECHIP softwares. The developed models were tested for their significance using analysis of variance
(ANOVA). Results were considered significant when the corresponding p-values were less than 0.05.
Two- and three-dimension contour diagrams visualizing the simultaneous effect of the causal factors on the response at each time point were established.
59 Response surface analysis using the contour diagrams was utilized to select the
coating formulation variables required to produce the desired percent of drug
release at each time point. A release profile with 35 % bumetanide released after
1 hr, 65 % released after 4 hr and 80 % drug released after 8 hr was targeted.
Two-dimension Contour diagrams were superimposed and the overall coating
formulation variables that can simultaneously fulfill the release requirements at all
the time points were estimated.
5.4.1.10. Preparation and Statistical Evaluation of the Designed
Formulation
Bumetanide pellets were coated using the estimated coating formulation
variables (Eudragit RS load and triethyl citrate level) expected to fulfill the target
release requirements under the same coating conditions described earlier.
Coated pellets were tested for their release profiles in USP purified water using
USP XXV dissolution apparatus number 1 at a basket rotation speed of 50 rpm.
The observed release profiles from the designed coated pellets were compared
to the target release profile using the Food and Drug Administration
recommended model independent approach utilizing the similarity factor (F2)
(Food and Drug Administration, 1997b). The similarity factor described in equation 3 is a logarithmic reciprocal square root transformation of the sum of squared errors and is a measurement of the similarity in the percent dissolution between two curves.
60 n 2 -0.5 F2 = 50 S log {[ 1+(1/n) t=1 ( Rt - Tt) ] S 100} (3)
Where, n = number of sample points.
Rt = mean percent of drug released from the reference samples at each time point (target percent of bumetanide release at each time point).
Tt = mean percent of drug released from the test samples at each time point
(observed percent of bumetanide released from the designed formulation).
The guidance states that values of F2 greater than 50 (50-100) ensure sameness or equivalence of the two profiles (Food and Drug Administration, 1997b).
5.4.2. Part II: Study of the Effect of Curing Conditions and Plasticizer
Level on the Release of Bumetanide from Coated Pellets
5.4.2.1. Statistical Experimental Design
A full factorial statistical experimental design was implemented to study the effect of plasticizer level, curing temperature and time on the release of bumetanide from coated pellets as shown in Table 7. Plasticizer levels were selected in the range of 10-20 % of the dry polymer weight gain. Plasticizer levels lower than 10 % were not sufficient to enhance polymer flexibility to the level needed to form continuous film and were not used in this study. Plasticizer levels above 20% resulted in excessive agglomeration owing to the rubbery and sticky state of the polymer and were not included in this study. Curing temperatures in the range of 40-60L C were used in the study as they are higher than the glass transition temperature of the polymer at all the plasticizer levels used.
61 Table 7. Full factorial experimental design used in the curing experiment
Independent Variable Levels Used
Plasticizer level (% polymer dry weight) 10, 15, 20
Curing temperature (LC) 40, 50, 60
Curing time (hr) 0, 1, 4, 8, 24, 168
62 Increasing the temperature aimed to accelerate film formation and
coalescence to allow better understanding of the curing process. Curing times
ranging from 1 hr to 168 hr were used in the study. Prolonged curing times at
higher temperature intended to simulate the effect of long storage times at room
temperature to help predict the changes in drug release from the coated dosage
form upon prolonged storage.
5.4.2.2. Coating of Bumetanide-Loaded Pellets
Bumetanide-loaded pellets were coated with different Eudragit RS coating
formulations according to Table 8 using bottom spray fluid bed equipment. The
solid percent in each coating dispersion was maintained at 20 % by the addition
of USP purified water. The fluid bed equipment coating conditions are displayed
in Table 6 (page 57). Polymer weight gain of 6 % w/w was selected based on the
data established to target a release profile extending over 12-hr time period.
5.4.2.3. Curing of Coated Pellets
Coated pellets were cured in ovens as well as in fluid bed equipment. In
oven curing process, pellets coated with 6 % Eudragit RS plasticized with 10, 15,
and 20 % triethyl citrate were added to labeled petri dishes and placed in the
ovens. The ovens temperatures were automatically controlled at 40LC, 50LC, and
60LC. The relative humidity inside the oven was 20 % M 5. Samples were removed after 1, 4, 8, 24, and 168 hr.
63 Table 8. Coating formulations used in the curing experiment
10 % Triethyl citrate
Ingredients Percentage
Eudragit RS 30 D 41.67
Triethyl citrate 1.25
Talc 6.25
USP purified water 50.83
15 % Triethyl citrate
Ingredients Percentage
Eudragit RS 30 D 40.4
Triethyl citrate 1.82
Talc 6.06
USP purified water 51.72
20 % Triethyl citrate
Ingredients Percentage
Eudragit RS 30 D 39.22
Triethyl citrate 2.35
Talc 5.88
USP purified water 52.55
64 Pellets coated with 6 % Eudragit RS plasticized with 20 % triethyl citrate
were also cured in the fluid bed equipment. Spraying was stopped at the end of
coating cycle while the fluidization process was not interrupted. The inlet air
temperature was dropped to 55-60LC for target curing temperature of 50LC.
Samples were removed after 10, 20, 30, and 60 min.
5.4.2.4. Testing the Release of Bumetanide from Cured Coated Pellets
The coated pellets were tested for their release profiles in USP purified water using Vankel 7000 dissolution apparatus utilizing the USP XXV basket dissolution method (Apparatus 1) at a rotation speed of 50. Samples removed after 0.5, 1, 2, 4, 8, and 12 hr were analyzed for their bumetanide contents using fluorescence spectrophotometer.
5.4.2.5. Seal Coating with Hydroxypropyl Methyl Cellulose (HPMC LV 100)
A seal coat of hydroxypropyl methyl cellulose (HPMC LV100) was applied to the bumetanide-loaded pellets prior to coating with Eudragit RS. HPMC was dissolved in USP purified water together with polyethylene glycol 400 according to Table 9. The solution was sprayed on bumetanide-loaded pellets fluidized in the Wurster fluid bed equipment for a target weight gain of 1 %. The fluid bed equipment parameters used in the seal coating process are shown in Table 10.
The seal coated pellets were subsequently coated with 6 % Eudragit RS using 20
% triethyl citrate as plasticizer. The coated pellets were oven cured at 60 LC for
1, 4, 8, 24, and 168 hr.
65 Table 9. Formulation used to seal coat the bumetanide-loaded pellets with HPMC prior to coating with Eudragit RS
Ingredient Percentage
HPMC LV100 2
PEG 400 0.4
USP purified water to 100
66 Table 10. Wurster fluid bed equipment processing conditions for the HPMC seal coating process
Parameters Unit Value
Inlet air temperature LC 84-88
Outlet air temperature LC 57-59
Product bed temperature LC 60M2
Amount sprayed gm 180
Spray rate g/min 7
Atomizing air pressure (output) Bar 2
Nozzle diameter mm 0.8
Filter shaking time Interval Panel Set 4
Filter shaking time duration Panel Set 0.1
Pressure drop (PD) across the bed Kpa 0.5-0.7
PD exhaust filter Kpa 0.1
Supply air pressure Bar 7
Actual air pressure Bar 6
Exhaust air flap Bar About 0.4
Drying temperature LC 55-65
Drying time min 10
Air distribution plate N/A 4700/A-6
Load gm 300
67 5.4.2.6. Coating using a Mixture of Eudragit RS and Eudragit RL
Bumetanide-loaded pellets were coated with 6 % mixture of Eudragit RL and RS (10:90) using triethyl citrate as a plasticizer at a level of 20% polymer dry weight according to the formulation shown in Table 11. USP purified water was added in a calculated amount to produce 20 % w/w solid content. The coating dispersion was sprayed on the bumetanide pellets fluidized in the Wurster fluid bed equipment using the parameters described in Table 6.
5.4.2.7. Use of Sodium Chloride as a Channeling Agent in the Coating
Formulation
Sodium chloride was incorporated into the coating dispersion at a level 5
% of the dry polymer weight gain as shown in Table 12. The dispersion was used to coat bumetanide pellets using the Wurster fluid bed equipment according to the parameters described in Table 6 (page 57). The dispersion was sprayed for a polymer weight gain of 6 % and the coated pellets were oven cured at 60 LC for 1 hr.
68
Table 11. Coating formulation containing a mixture of Eudragit RS and RL 30 D
Ingredient Percentage
Eudragit RS 30 D 35.29
Eudragit RL 30 D 3.92
Triethyl citrate 2.35
Talc 5.88
USP purified water to 100
69 Table 12. Coating formulation containing sodium chloride
Ingredient Percentage
Eudragit RS 30D 38.1
Sodium chloride 0.57
Triethyl citrate 2.29
Talc 5.71
USP purified water to 100
70 5.4.2.8. Study the Effect of Storage on the Release of Bumetanide from
Coated and Cured Pellets
The following bumetanide pellets were stored in petri dishes for 3 months
at room temperature (25 LC):
1) Pellets coated with 6% Eudragit RS plasticized with 20 % triethyl citrate and
cured at 60LC for 1 hr;
2) Pellets seal coated with 1 % HPMC, coated with 6% Eudragit RS plasticized with 20 % triethyl citrate and cured at 60LC for 1 hr.
Drug release from the stored coated pellets was tested in USP purified water using USP dissolution apparatus number 1 (basket type) at a rotation speed of 50 rpm.
5.4.2.9. Statistical Analysis
Analysis of Variance (ANOVA) using SAS software was utilized to assess the effects of different formulation (plasticizer level, seal coating with HPMC, coating with polymeric combination) and curing variables (curing time and temperature) on bumetanide release from the coated pellets. The differences were considered significant at the corresponding time point when the p-values were less than 0.05.
71 5.4.3. Part III: Testing Selected Optimized Bumetanide Extended Release
Formulation in Laboratory Animals
5.4.3.1. Selection of Animal Model
The animal model selected for the study is white male New-Zealand rabbit.
5.4.3.1.1. Reasons for Selecting Rabbits as the Animal Model
The oral potency of bumetanide decreases in the following order: dog, rabbit, mouse, and rat. The decrease in activity was related to the increase in the extent of metabolism since none of bumetanide metabolites carries diuretic activity
(Magnussen and Eilertsen, 1974). Rats, like humans, metabolize bumetanide by oxidative pathways yet to greater extents. (Schwartz, 1981). Furthermore, bumetanide is less bioavailable after peroral administration in rats than in humans (Busch et al., 1979). Formulating bumetanide in a peroral extended release manner can increase its liver metabolism owing to the slow input of drug into systemic circulation with lower ability to saturate the metabolic pathways in the liver. Therefore, rats will not serve as suitable candidates to study peroral extended release bumetanide formulations due to the expected lower systemic bioavailability owing to incomplete absorption and extensive liver metabolism.
Rabbits, on the other hand, metabolize bumetanide to a lesser extent than rat. The exact similarities and differences in bumetanide metabolic pathways between humans and rabbits are still unclear. Nevertheless, when bumetanide was administered to rabbits by slow infusion for 4 hr, more than 60 % of the intravenous administered dose was excreted as unchanged molecules in urine with extensive diuretic response (Ryoo et al., 1993). The results suggested that
72 bumetanide could be excreted as unchanged molecules in the rabbit urine from peroral extended formulations to a level that can warrant significant diuretic response.
Table 13 shows the lengths of different segments of the gastrointestinal tracts of different laboratory animals compared to humans. The lengths of the gastrointestinal tracts of the rabbits are relatively longer than that of rats and closer to that of dogs. The lengths of different segments of the small intestine, where most of the absorption is expected to take place, of both rabbits and dogs are closer to that of humans than rats. In rabbits, 11-mm enteric-coated tablets of barium sulphate were retained in the stomach for 24 hr while 1-mm diameter granules were partially released during this period (Aoyagi, 1986). When compared to other laboratory animals, rabbits showed the slowest transit while dogs and pigs had the most rapid transit from the gastrointestinal tract (Kerarli,
1995). Accordingly, it is expected that extended release formulation prepared in the form of small particles in the micron size can have equivalent residence time in the gastrointestinal tract of rabbits and humans.
Table 14 shows the pH values of the different parts of the gastrointestinal tract in humans and some laboratory animals. The pH of the stomach of rabbits shows the closest resemblance to that of humans. As for the rest of the gastrointestinal tract, small differences can be anticipated between different laboratory animals and humans in terms of pH values.
73 Table 13. The lengths of different segments of the gastrointestinal tract of laboratory animals compared to humans (Kararli, 1995)
Gastrointestinal Segment Rabbits Dogs Rats Humans
(m) (m) (m) (m)
Small Intestine 3.56 4.14 1.38 6.25
Cecum 0.61 0.08 0.06 0.2
Colon 1.65 0.6 0.1 1.5
Total 5.82 4.82 1.54 7.95
74 Table 14. The pH values of different segments of the gastrointestinal tract of humans and some laboratory animals (Kararli, 1995)
Gastrointestinal Segment Rabbits Dogs Rats Humans
Stomach anterior portion 1.9 5.5 5.0 1.5-3.5
Stomach posterior portion 1.9 3.4 3.3
Small intestine portion 1 6.0 6.2 6.5 Duodenum= 5-7
Small intestine portion 3 6.8 6.2 6.7 Jejunum= 6-7
Small intestine portion 5 7.5 6.6 6.8 Ileum= 7
Small intestine portion 7 8.0 7.5 7.1
Cecum 6.6 6.4 6.8
Colon 7.2 6.5 6.6 5.5-7.0
Feces 7.2 6.2 6.9 Rectum = 7
75 Based on the data presented, rabbits are expected to show close resemblance to humans in terms of the gastrointestinal residence time and the pH values of different segments. Rabbits have been used in the literature as models to study the absorption of various drugs from peroral dosage forms.
(Laethem et al., 1995; Bayomi et al., 1998; Hebucky and Tse, 1998; Miyazaki et al., 1999). Therefore, rabbits were selected as animal model for this study.
5.4.3.2. Animal Study Design
Eight male white New-Zealand rabbits, weighing 2.1 M 0.36 Kg were used for the study. Each rabbit received both the immediate and extended release bumetanide formulations according to a randomized two-treatment cross over design with one week wash out period. Bumetanide-loaded pellets with no subsequent coatings were used as the immediate release formulation. Coated bumetanide pellets with a target predetermined release profile were used as the extended release formulation. Calculations were made so that each rabbit receives 1 mg/Kg bumetanide. Each rabbit was used as its own control by collecting and analyzing its urine samples for at least 48 hr prior to the administration of bumetanide formulations.
All animal handling procedures were conducted in accordance with the
Animal Welfare Act (AWA) administered by the United States Department of
Agriculture (USDA) and Animal and Plant Health Inspection Service (APHIS) as well as the Public Health Service Act (PHSA) administered by the Office for
Protection from Research Risks (OPRR). The procedures were formatted in an
76 animal use protocol and approved by the Institutional Animal Care and Use
Committee (IACUC) at the University of Cincinnati.
5.4.3.3. Rationale for the Number of Rabbits Used
To establish bioequivalence the intravenous studies of Ryoo et al. (1993) was used as reference. Ryoo et al. used three groups of rabbits to study the effect of intravenous infusion time on the pharmacokinetics and pharmacodynamics of the same total dose of bumetanide. The same dose of bumetanide was infused over periods of time of 10 seconds, 1 hr and 4 hr to rabbits. Group I receiving the fast infusion (over 10 seconds) and group II receiving the infusion over 1 hr were considered for comparison to our proposed research. Data from Group I of Ryoo et al. were compared to our data from
Group A (rabbits receiving the immediate release bumetanide formulation) Data from group II of Ryoo et al. were compared to our data from Group B (rabbits receiving the extended release bumetanide formulation). In the results of Ryoo et al., the following parameters were recorded:
Urine output (ml) for group I M SD = 373 M 49.7, 8 hr collection.
Urine output (ml) for group II M SD = 922 M 214, 8 hr collection.
Difference in Means = 549 ml
The sample size (number of rabbits per group) can be determined using the following equation (Daniel, 1999):
2 2 2 N = [( Z1-/2 + Z1-B) ] / (4)
77 Using the universally accepted power (1-B) of 80 % (B =0.2) and confidence (1-
Q) of 95 % (Q =0.05), and considering comparing the diuretic effect in terms of
volume of urine, then:
2 2 Z1-/2 = Z0.975 = 1.96; Z1-B = Z0.8 = 0.84; = Variance = (214) = 45796
2 = Difference in group means = (549)2 = 301401
N = Number of animals per group =1.19 2
However, the difference between the immediate and extended release peroral bumetanide formulations was expected to be lower than the values recorded in the literature owing to the fact that immediate release peroral formulation doesn't release the drug immediately to the systemic circulation. The process of drug absorption through the gastrointestinal mucosa may delay the appearance of the drug in systemic circulation. It has been recorded in the literature that the peak level of bumetanide in blood following the oral administration occurs after 1-2 hours. Accordingly, a difference in the urine output of about 275 (549/2) ml between the immediate and extended release bumetanide formulation was expected. Accordingly, the value of N = 4.75.
Therefore, 4 rabbits were initially selected for the study. To assure powerful statistical analysis, the results established during the study were subject to power analysis. The number of rabbits was considered sufficient to draw statistically acceptable conclusion only when the statistical analytical power was above the universally accepted 80 % level. According to the data generated using the first 4 rabbits, the analytical power was less than 80% and therefore, additional four rabbits were included in the study. The data collected from the 8 rabbits were
78 subjected to statistical power analysis and the power was found to be more than
80%.
5.4.3.4. Rabbits Manipulation
Upon receiving, rabbits were allowed 48 hr to acclimatize to the new husbandry environment. Rabbits were then introduced to the specifically designed metabolism cage where urine can be collected separate from feces.
Urine was collected for at least 48 hr before dosing with bumetanide formulation.
Urine samples collected were used to establish baseline in terms of urine volume and its content of sodium and potassium. Rabbits were then fasted for 12 hr prior to dosing with free access to water.
5.4.3.4.1. Rabbits Restraint
The rabbit was placed in the restraint cage where Elizabeth collar was installed around its neck, catheter was introduced into the ear artery and the drug formulation was administered via a gastric gavage. The rabbit was relocated to the metabolism cage immediately after drug administration. The rabbit was placed in the restraint 5 min prior to each blood sampling time and returned to the metabolism cage after the sample was withdrawn. Rabbit maintained fasting conditions for additional 12 hr post dosing with free access to water.
Blood sample withdrawal was not feasible for periods of time more than 6 hr. Excessive blood clotting and vasoconstriction are hypothesized to be the reasons for such effect. The use of Elizabeth collar and serial blood sampling
79 attempts for 24 hr together with 24-hr fasting period added enormous amount of stress on the rabbits. The developed stress can significantly interfere with normal physiological functions of rabbits, which can affect their response to loop diuretics. Therefore, blood sampling was discontinued from the study.
5.4.3.4.2. Drug Administration
A calculated amount of the formulation containing 1 mg/Kg bumetanide was administered orally using 12-french gauge gastric gavage. A wooden speculum with centric hole was introduced to the rabbit mouth. The pellets were added to the gavage followed by flushing with 15 ml of water to assure maximal dose delivery and minimal residual in gavage needle. Slowly pulling the plunger out and assuring no air suction before injecting the content from the syringe confirmed the insertion of the gavage into the stomach. After the administration of the required dose, the gavage was removed slowly then the speculum was removed.
5.4.3.4.3. Sample Withdrawal Blood Samples
1 ml of blood was withdrawn from the rabbit ear artery using the catheter at each time point. Blood samples were collected at the following time points:
0 min, 5 min, 15 min, 30 min, 45 min, 60 min, 90 min, 2hr, 3 hr, 4 hr, 6 hr, 8 hr,
12 hr and 24 hr. Each sample was replaced with an equivalent amount of sterile normal saline. Blood sampling with later discontinued from the study.
80 Urine samples Urine samples were collected from the base of the metabolism cages in the following interval:
0-1, 1-2, 2-3, 3-4, 4-5, 5-6, 6-8, 8-12, 12-16, 16-24, 24-32, 32-40, and 40-48 hrs.
5.4.3.5. Analysis of Blood and Urine Sample for Bumetanide Contents
5.4.3.5.1. Development of HPLC Analytical Method
Bumetanide contents in blood and urine were analyzed using HPLC separation and fluorescence detection according to the method of Ryoo et al.
(1993). The HPLC system consisted of 126 solvent module (pump) and 507-2 autosampler. Blood samples were allowed to coagulate followed by centrifugation at 4000 g for 5 minutes using Marathon 21K/BR Centrifuge. The supernatant (serum) was collected and used for the analysis. Urine samples were centrifuged at 4000 g for 5 minutes and the supernatant was used for the analysis. Serum samples (0.1 ml) and urine samples (0.2 ml) were mixed with acetonitrile (0.25 and 0.5 ml for serum and urine samples respectively) by shaking followed by centrifugation at 10000 g for 10 min. The supernatant was carefully transferred into clean vials from which 20 uL was injected into C18 HPLC reverse phase column (Inertsil SDS 5, 3.9X30mm, 5 um particle size). The mobile phase consisted of methanol : water : acetic acid in the ratio of 70:30:1 and was run at a flow rate of 1 ml/min. Bumetanide was detected using Beckman
157 fluorescence detector using excitation filters of 305-395 nm and emission filter of 350-650 nm wavelengths. Bumetanide peaks had retention time of around 9 min. Chromatograms were recorded using Varian 4270 integrator.
81 5.4.3.5.2. Validation of Bumetanide HPLC Analytical Method
Linearity and Range
Bumetanide reference standard was dissolved in methanol followed by serial dilution in USP purified water. Rabbit serum and urine samples were spiked with minimal amount of bumetanide solutions to obtain the following final concentrations, Serum: 0.01, 0.025, 0.05, 0.1, 0.25, 1, 2 ug/ml; Urine: 0.167,
0.333, 0.833, 1.667, 3.333, 6.667 ug/ml
Serum and urine samples were injected into the HPLC column and analyzed using the developed method. The relative fluorescence intensities were calculated as the areas under the peaks and regressed against the corresponding bumetanide concentrations to calculate the calibration curve equations. The calibration curves were considered linear as long as the values of the correlation coefficient was above 0.99.
Accuracy
Three different new bumetanide solutions of known concentrations (0.05,
0.25, and 1 ug/ml) were prepared by spiking the corresponding amount of bumetanide solution into serum and urine. The samples were injected into the
HPLC column and the relative fluorescence intensities were measured. Accuracy was calculated by comparing the calculated recovered concentrations to the actual concentrations using the following equation:
Accuracy (Recovery) (%) = {Recovered (measured) concentration/ Actual
(expected) concentration}*100
82 Intra- and Inter-day Variability
Three different bumetanide solutions of known concentrations (0.05, 0.25, and 1 ug/m) were prepared by spiking the corresponding amount of bumetanide solution into serum and urine. The samples were prepared and injected into the
HPLC column in two different days. Each sample was injected three times and the relative fluorescence intensity was measured. Intra-day variability was calculated as the relative standard deviation among the three injections on the same day for the three different concentrations. Inter-day variability was calculated as the relative standard deviation among the 6 replications of each sample injected in two different days.
5.4.3.6. Analysis of Urine Samples for Sodium and Potassium Contents
Urine samples were analyzed for their potassium and sodium content using Flame Photometer. The flame was sparked using propane fuel and the equipment was calibrated using standard sodium and potassium solution (100 mmol/L of each). Urine samples were injected into the flame and their contents of sodium and potassium were measured directly. The equipment was calibrated after each 14 samples.
5.4.3.7. Determination of Urine Osmolarity
Urine osmolarity was determined using Advanced Micro Osmometer.
Twenty microliter of the urine sample measured using specially designed pipette was frozen at a specific cooling rate. The equipment determines the freezing
83 point depression of the urine sample relative to pure water and expresses the results as mOsmol/Kg water.
5.4.3.8. Statistical Analysis
The following statistical analysis were conducted:
1. Analysis of Variance (ANOVA) using SAS software to test for carryover and period effects as well as to compare the effects of both formulations to control.
Insignificant carryover effect reflects independence of the response to the second treatment on the response to the first treatment. Any significant carry over effect can confound the results and lead to statistically unacceptable analysis.
Significance period effect reflects the response depends on the sequence of treatments i.e. response to the immediate release formulation will be different based on whether the formulation was administered first or second to the extended release formulation.
2. Paired t-test using SAS software to compare the difference in responses to the immediate and extended release formulations among the rabbits. Paired t- test is preferred in cross-over design as it can limit the effect of inter-subject variability by using the variance of the difference in the calculations rather than the variance among each group. The differences were considered significant at p-values less than 0.05, marginally significant at p-values between 0.05-0.1, and insignificant at p-values above 0.1
84 6. Results and Discussion
6.1. Part I: Development and Optimization of Bumetanide Extended
Release Formulations using Response Surface Methodology and
Multiple Response Optimization
6.1.1. Development and Validation of Bumetanide Spectrofluorimetric
Analytical Method
Bumetanide calibration curves were constructed in USP purified water, glycine buffer pH= 2.9, acetate buffer pH= 4.5, USP hydrochloric acid buffer pH=
1.2, and USP phosphate buffer pH=6.4 and 6.8. Table 15 presents the measured bumetanide maximum excitation and emission wavelengths in different media.
The calibration curves and equations relating bumetanide concentrations to the relative fluorescence intensities in each media were established at the corresponding maximum excitation and emission wavelengths by simple linear regression.
Accuracy
The recovered bumetanide concentrations calculated using the developed calibration equations were within 100M 2% of the theoretical concentrations for all
bumetanide solutions tested.
Precision
The estimated relative standard deviations for the three different
bumetanide concentrations each measured 3 times were less than 1 %.
85 Table 15. Maximum excitation and emission wavelengths of bumetanide in different dissolution media
Media Excitation Emission
Wavelength (nm) Wavelength (nm)
USP purified water 326 407
Glycine buffer, pH= 2.9 336 442
Acetate buffer, pH= 4.5 333 418
USP Hydrochloric acid, pH = 1.2 340 448
USP Phosphate buffer, pH=6.4 324 408
USP Phosphate buffer, pH= 6.8 324 408
86 Linearity and Range
The linearity ranges of the bumetanide calibration curves were checked by
calculating the correlation coefficients over the concentration range used (0.04-
20 ug/ml). The correlation coefficients estimated were above 0.999 for all the
calibration equations established.
6.1.2. Content Uniformity Assessment
It was found that the bumetanide layering process using GPCG 1 fluid bed
equipment produced pellets that are uniform in their drug content. For all the
batches tested the percent of bumetanide in the loaded pellets were within 85-
115 % of the theoretical label claim (1 %w/w) with relative standard deviation less
than 6 %. The use of atomizing air pressure of 200 KPa was enough to disperse
the bumetanide solution into very small droplets that were uniformly distributed
throughout the fluidized sugar pellets. The bed temperature set at 60LC was high enough to allow for pellets drying while bumetanide solution was sprayed with no agglomeration or sticking.
6.1.3. Testing the Release of Bumetanide from the Coated Pellets
According to the central composite statistical design used in this study, four center runs were conducted where the percent Eudragit RS load was 5 % and the triethyl citrate level in the coating dispersion was 15 %. Center runs were necessary to augment the statistical design as they increase the number of levels of each causal factor and provide extra degrees of freedom to test for pure error.
87 Figures 3 and 4 (pages 90 and 91) show the release profiles from the center runs were intermediate among other slower and faster profiles. As seen in Figure 3, pellets coated with higher Eudragit RS load and/or triethyl citrate level had slower release profiles than center runs. On the other hand, pellets coated with lower
Eudragit RS load and/or triethyl citrate level had faster release profiles than the center runs as seen in Figure 4.
Run number 5 with Eudragit RS load of 5 % and triethyl citrate level of 22
% was not successful due to excessive granulation. The run had the highest level of plasticizer and the polymer was expected to have the lowest glass transition temperature. Consequently, the polymer existed in a highly rubbery and sticky state during the coating process that led to excessive granulation and minimal fluidization. The results suggested that increasing the Eudragit RS load and the plasticizer level extended the release of bumetanide from coated pellets.
6.1.4. Testing the Effect of Dissolution Media pH on the Release of
Bumetanide from coated Pellets
Figure 5 (page 92) displays the effect of different dissolution media pH on the release of bumetanide from pellets coated with 7.5 % Eudragit RS plasticized with 20 % triethyl citrate. As seen in figure, bumetanide release was strongly affected by the dissolution media pH where drug release is accelerated in media of higher pH values. The results can be explained by considering bumetanide pH-dependent solubility. Bumetanide is a week acid with pKa of 3.3. At low pH values bumetanide exists as uncharged molecules with low polar and ionic
88 interaction and hence limited water solubility. Increasing the pH value of the media is accompanied by an increase in the percentage of ionized bumetanide molecules with a resultant increase in drug solubility. The higher the pH of the media the more ionized the bumetanide molecules and the higher the solubility.
Therefore, faster release profiles were obtained in media of higher pH values.
6.1.5. Testing the Effect of Agitation Speed on the Release of Bumetanide
from Coated Pellets
Figure 6 (page 93) displays the effect of basket agitation speed on the release of bumetanide from pellets coated with 7.5 % Eudragit RS plasticized with 20 % triethyl citrate. Increasing the agitation speed from 50 to 100 rpm had no significant effect on the release of bumetanide (p-values above 0.05 at each dissolution time point). Agitation speed can affect drug release rate through its effect on the stagnant boundary layer thickness around the coated pellets. The stagnant boundary layer serves as a barrier to drug diffusion and release.
Increasing the agitation speed is expected to decrease the thickness of the boundary layer and enhance the drug release process. The pellets used in the study had relatively small size with a resultant very thin stagnant layer whose thickness was insensitive to agitation speed.
89
110
100
90
80
70
60 leased e 50 % R 40
30
20
10
0 02468101214
Time (hr)
Figure 3. Effect of Eudragit RS load and triethyl citrate level on the release of bumetanide from coated pellets. () 5% Eudragit RS and 15 % triethyl citrate
(center runs); (o) 2.5 % Eudragit RS and 20 % triethyl citrate; () 8.525 %
Eudragit RS and 15% triethyl citrate; () 7.5 % Eudragit RS and 20% triethyl citrate.
90 110
100
90
80
70 d e
s 60 a e l e 50 % R 40
30
20
10
0 02468101214 Time (hr)
Figure 4. Effect of Eudragit RS load and triethyl citrate level on the release of bumetanide from coated pellets. ()5% Eudragit RS and 15 % triethyl citrate
(center runs); (S) 2.5 % Eudragit RS and 10 % triethyl citrate; (o) 7.5 % Eudragit
RS and 10 % triethyl citrate; () 5 % Eudragit RS and 8% triethyl citrate;
(▲)1.475% Eudragit RS and 15 % triethyl citrate.
91
110
100
90
80
70
60 eased l e 50 % R 40
30
20
10
0 02468101214 Time (hr)
Figure 5. Effect of dissolution media pH on bumetanide release from pellets coated with 7.5% Eudragit RS plasticized with 20% triethyl citrate. (□) USP hydrochloric acid buffer USP pH = 1.2; (◊) USP hydrochloric acid buffer for 2 hr then phosphate buffer pH = 6.8; (∆) USP purified water; (○) acetate buffer pH =
4.5; (■) USP phosphate buffer pH = 6.4.
92
110
100
90
80
70
60 leased e 50 % R 40
30
20
10
0 02468101214
Time (hr)
Figure 6. Effect of basket agitation speed on bumetanide release from pellets coated with 7.5% Eudragit RS plasticized with 20% triethyl citrate pellets. (○) 50 rpm; (□) 100 rpm.
93 6.1.6. Statistical Analysis
6.1.6.1. Multiple Regression and Model Building
Table 16 (page 96) describes the model parameters affecting the release from coated pellets after 1, 4, and 8 hr. The model describing the release after 1 hr can be written as:
Y1 = 73.86 – 16.20 * X1 – 20.58 * X2 - 5.18 * X1* X2 (5)
Where,
Y1 =Percent bumetanide released after 1 hr.
X1 = Eudragit RS load; X2 = Triethyl citrate level.
Negative values indicated the percent of bumetanide released after 1 hr decreased as the Eudragit RS load or the level of triethyl citrate in the coating dispersion increased. Statistical analysis using ANOVA revealed significant model fit with p-value = 0.001.
The model describing the release after 4 hr can be written as:
2 Y2 = 91.55 – 9.11* X1– 14.05 * X2 – 5.06 * (X2) (6)
Where, Y2 = Percent bumetanide released after 4 hr.
94 Both Eudragit RS load and triethyl citrate level had significant negative effect on the percent of bumetanide released after 4 hr. Quadratic terms for the triethyl citrate level were also found to be significant. Statistical analysis using
ANOVA revealed a significant model fit with p-value = 0.0005.
The model describing the release after 8 hr can be written as:
Y3 = 96.28 - 6.98 * X1 – 12.41 * X2 (7)
Where, Y3 = Percent bumetanide released after 8 hr.
Again, both Eudragit RS load and triethyl citrate level had significant negative effect on the release of bumetanide from the coated pellets. Neither quadratic nor interaction terms were found to be significant. Statistical analysis using ANOVA revealed significant model fit with p-value = 0.0107. The results suggested that increasing the Eudragit RS load and triethyl citrate plasticizer level significantly extended the release of bumetanide from the coated pellets.
Increasing Eudragit RS load lead to an increase in the thickness of the coat formed around the pellets with a subsequent reduction in the rate at which bumetanide is released.
95 Table 16. Model parameters affecting the percent of bumetanide released after
1,4 and 8 hr
Parameter 1-hr Release 4-hr -Release 8-hr Release
Estimate p-value Estimate p-value Estimate p-value
Intercept 73.86 0.0001* 91.55 0.0001* 96.28 0.0001*
X1 -16.20 0.0001* -9.11 0.0003* -6.98 0.0108*
X2 -20.58 0.0001* -14.05 0.0001* -12.41 0.0028*
2 (X1) -2.78 0.1856 -5.06 0.0194* -4.89 0.1082
2 (X2) 2.49 0.1535 -2.41 0.1048 3.71 0.1294
X1 X2 -5.18 0.0353* -1.02 0.5252 1.46 0.5832
* ANOVA statistically significant at Q = 0.05
X1 = Eudragit RS Load; X2 = Triethyl Citrate Level
96 Plasticizers are usually included in coating formulations to reduce the polymer glass transition temperature and to impart flexibility to the coat, which facilitates the coating process under moderate condition (Lehmann, 1997). In the pharmaceutical literature, controversial results regarding the effect of plasticizers on drug release from coated dosage forms were reported. Plasticizers have been shown to slow the release process in some studies (Schmidt and Niemann,
1993; Hutchings and Sakr, 1994) and increasing the release process in others
(Bodmeier and Paeratakul, 1993; Rangaiah et al., 1997; Siepmann et al., 1999).
Plasticizers can affect the drug release process in two opposite directions.
They can increase drug release by facilitating drug diffusion through polymeric film for both water soluble and insoluble plasticizers and by leaching out and generating channels in the coat for water-soluble plasticizers. On the other hand, plasticizers can decrease the drug release by facilitating the formation of continuous smooth film with minimal flaws.
In the present study, triethyl citrate, a water soluble plasticizer, decreased the release of bumetanide from coated pellets. The results can be explained in terms of the relatively large surface area of the small pellets (mean particle size of about 500 um) which is difficult to be covered by less flexible polymer.
Increasing the level of triethyl citrate increased the flexibility of the polymer and allowed for more uniform coverage of the surface of the pellets with a subsequent slowing of the release process. The results were in agreement with the results published in the literature (Bodea and Leucuta, 1997) where increasing the level of the plasticizer slowed the release process for propranolol
97 pellets coated with low coating loads compared to the large surface area available for coating.
Figures 7-9 (pages 99-101) present the three-dimensional contour diagrams visualizing the simultaneous effect of Eudragit RS load and triethyl citrate level on the percent of bumetanide released after 1, 4 and 8 hr. Increasing both Eudragit RS load and triethyl citrate decreased the percent bumetanide released after 1,4 and 8 hr.
6.1.6.2. Optimization
Two-dimension contour diagrams were used to optimize the bumetanide release in USP purified water. The following release profile was targeted:
35 % of drug released after 1hr, 50 % of drug released after 2 hr, 65 % of drug released after 4 hr, 80 % of drug released after 8 hr, and 90 % of drug released after 12 hr. The release profile was selected as part of a preliminary investigation to conduct an in vitro/in vivo correlation study. The Food and Drug
Administration (FDA) guidance recommended the use of three dissolution profiles with different release rates for the in vitro/in vivo correlation study (Food and Drug Administration, 1997a). The three different dissolution profiles targeted for the in vitro/in vivo correlation are shown in Figure 10 (page 102). The selected dissolution profile for this study represents the intermediate release profile for the in vitro/in vivo correlation study.
98
Figure 7. Three-dimension contour diagram illustrating the effect of Eudragit RS load and triethyl citrate level on the release of bumetanide from coated pellets after 1 hr.
99
Figure 8. Three-dimension contour diagram illustrating the effect of Eudragit RS load and triethyl citrate level on the release of bumetanide from coated pellets after 4 hr.
100
Figure 9. Three-dimension contour diagram illustrating the effect of Eudragit RS load and triethyl citrate level on the release of bumetanide from coated pellets after 8 hr.
101
110 100 90 80 70 60 50 Released
% 40 30 20 10 0 02468101214
Time ( hr)
Figure 10. The different dissolution profiles targeted for the in vitro/in vivo correlation study. (S) Slow release profile; (□) Intermediate release profile; (∆)
Fast release profile.
102 Developing an in vitro/in vivo correlation allows the dissolution process to be considered as a sensitive, reliable, and reproducible surrogate for bioequivalence testing. Once an in vitro/in vivo correlation is established, the effect of any changes in the formulation, manufacturing site, batch size or equipment process on the product performance can be investigated using the in vitro dissolution rather than the expensive and time consuming bioequivalence study (Food and
Drug Administration, 1997a). Bumetanide can serve as a suitable candidate for in vitro/in vivo correlation since it has low solubility but expected to have high permeability through the intestinal mucosa. Therefore, the rate-limiting step for in vivo bumetanide absorption is the dissolution from the corresponding dosage form rather than its permeation through gastric and intestinal mucosa. Thus, it becomes of great importance to set a well-established in vitro dissolution test that can be correlated to the in vivo release process.
The establishment of in vitro/in vivo correlation has not been fully investigated within the scope of this project. Nevertheless, the dissolution profiles selected for the study were planned to fit into the scheme of in vitro/in vivo correlation protocol. The developed work had set the stage for a future complete in vitro/in vivo correlation study to be conducted in laboratory animals or humans.
Figures 11-13 (pages 106-109) represent the two-dimension contour diagrams illustrating the effects of Eudragit RS load and triethyl citrate level on the percent bumetanide released after 1, 4 and 8 hr respectively. The diagrams also express the level of Eudragit RS load and triethyl citrate level that are required to obtain the desired percent of drug release after each time point
103 independently. Eudragit RS load of 0.87 (7.175 % pellets weight gain of the dry polymer) and a triethyl citrate level of 0.96 (19.8 % of the dry polymer weight) were the selected optimum values necessary to obtain bumetanide release of 35
% after 1 hr. On the other hand, Eudragit RS load of 0.76 (6.9 % pellets weight gain of the dry polymer) and triethyl citrate level of 0.93 (19.65 % of the dry polymer weight) were needed to obtain 65 % of bumetanide release after 4 hr.
For the percent bumetanide release after 8 hr, Eudragit RS load of 0.71 (6.8 % pellets weight gain of the dry polymer) and triethyl citrate level of 0.63 (18.2 % of the dry polymer weight) were needed to achieve 80 % release after 8 hr.
Figure 14 (page 109) shows the three two-dimension contour diagrams superimposed on each other. The levels of Eudragit RS load and triethyl citrate that are needed to obtain a release profile optimized at the three different time points (35 % after 1 hr, 65 % after 4 hr and 80 % after 8 hr) were estimated from
Figure 14 as the following:
Eudragit RS load = 0.76 (6.9 % pellets weight gain of the dry polymer).
Triethyl citrate level = 0.87 (19.35 % of the dry polymer weight).
Bumetanide pellets were coated using the estimated Eudragit RS Load and triethyl citrate level. The coated pellets were tested for their bumetanide release profile in water. The release profiles of the coated pellets as well as the predicted target release profile are shown in Figure 15 (page 110). The coated pellets produced similar release profile to the expected target profile. Both the target and the observed profiles were compared using the Food and Drug
104 Administration recommended similarity factor (F2) (Food and Drug
Administration, 1997b). The results of the comparison are shown in Table
17(page 111). Values of F2 were above 50 at all the time points throughout the release process, which indicated equivalence of the release profile of the coated pellets and the target profile. The developed novel multiple response optimization technique and the results of this study were published in the journal of controlled release in May 2001 (Hamed and Sakr, 2001)
105
TECLevel= 0.96 PolymerL= 0.87
Value Low Limit High Limit 35.10 22.96 47.23
Figure 11. Two-dimension contour diagrams illustrating the effect of Eudragit RS load and triethyl citrate level on the percent bumetanide released after 1 hr. The levels of each variable necessary to obtain 35.1% release are the following:
Eudragit RS load = 0.87, triethyl citrate level = 0.96.
106
TECLevel= 0.93 PolymerL= 0.76 Value Low Limit High Limit 65.02 55.27 74.76
Figure 12. Two-dimension contour diagrams illustrating the effect of Eudragit RS load and triethyl citrate level on the percent bumetanide released after 4 hr. The levels of each variable necessary to obtain 65.02 % release are the following:
Eudragit RS load = 0.76, triethyl citrate level = 0.93.
107
TECLevel= 0.71 PolymerL= 0.63 Value Low Limit High Limit 79.89 64.91 94.87
Figure 13. Two-dimension contour diagrams illustrating the effect of Eudragit RS load and triethyl citrate level on the percent bumetanide released after 8 hr. The levels of each variable necessary to obtain 79.89 % release are the following:
Eudragit RS load = 0.63, triethyl citrate level = 0.71.
108
Percent of Bumetanide Released
TEC Level
Figure 14. Superimposed contour diagrams. The overall variables levels that are expected to fulfill the multiple point optimization are: Eudragit RS load = 0.76, triethyl citrate level = 0.87.
109
110 100 90 80 70 60 leased
e 50
% R 40 30 20 10 0 02468101214 Time (hr)
Figure 15. Comparison of the observed dissolution profile to the target predicted profile. (■) observed profile; (□) target predicted profile.
110 Table 17. F2 Test to compare the observed and the predicted target profiles
2 2 Hr Predicted Observed R-T (R-T) Time (R-T) F2
(R) (T) Points
0 0 0
1 35 40.67 -5.67 32.15 1 32.15 61.99
2 50 57.40 -7.4 54.76 2 86.91 58.8
4 65 72.91 -7.91 62.57 3 149.4 57.35
8 80 77.52 2.48 6.15 4 155.63 59.97
12 90 81.55 8.45 71.4 5 227.03 58.34
n 2 -0.5 F2 = 50 S log{[1+(1/n) t=1 (R-T) ] S 100}
F2 Values above 50 ensure sameness or equivalence of dissolution profiles.
111 6.2. Part II: Study of the Effect of Curing Conditions and Plasticizer
Level on the Release of Bumetanide from Coated Pellets
The drug release from pellets coated with aqueous dispersion is strongly affected by variables influencing the coalescence of the polymer particles and hence the film formation process (Harris and Ghebre-Sellassie, 1997). Process variables such as coating temperature (Yang and Ghebre-Sellasie, 1990;
Bodmeier and Paeratakul, 1991; Williams III and Liu, 2000) curing conditions
(Hutchings et al., 1994, Bodmeier and Paeratakul, 1994) and plasticization
(Lippold et al., 1989; Streuernagel, 1997; Amighi and Moes, 1995) have been investigated and are needed to be optimized to obtain a reproducible release profile. Curing of coated pellets can cause a decrease in drug release to a certain end point (Lippold et al., 1989; Gilligan and Li Wan, 1991; Schmidt and Niemann,
1993, Bodmeier and Paeratakul, 1994, Hutchings et al., 1994; Keshikawa and
Nekagami, 1994; Amighi and Moes, 1996; Guma et al., 1997). Several factors are thought to influence the response of drug release profile to curing including coating conditions, plasticizer level, humidity level, dug physicochemical properties, in addition to curing temperature and time. For the proposed study, the effect of plasticizer level, curing temperature and time on bumetanide release profiles were thoroughly investigated. Plasticizer levels of 10, 15 and 20 % of the dry polymer weight were used and curing was conducted at 40, 50, and 60 LC for time periods extending from 1 hr to a week. Other factors including, coating conditions and humidity level were kept constant.
112 6.2.1. Coated Pellets Plasticized with 10 % Triethyl Citrate
At lower plasticizer level of 10 %, curing temperature of 40 LC for 24 hr was insufficient to induce any significant changes in the drug release profiles as seen in Figures 16 and 17 (pages 115 and 116). Figure 17 represents the effect of curing time on the percent of drug released after 0.5 and 4 hr. These time points were selected as they represent the drug release process. The percent of drug released after 0.5 hr represents the early phase of drug release and detects any dose dumping. The percent of drug released after 4 hr can provide sensitive measurement of the difference in release rate between different pellets with different curing conditions. Percent of drug released after longer time as 8 hr is not sensitive measurement of drug release process since by that time most of the drug is already released from different pellets with different curing conditions.
Curing at 40L C for 168 hr significantly reduced the release of bumetanide from the coated pellets nevertheless, it was not enough to produce as slow a release profile as those obtained at higher temperature.
At curing temperature of 50 LC, curing for 24 hr or more resulted in a significant reduction in the release of bumetanide from the coated pellets as seen in Figures 18 and 19 (pages 117 and 118). Curing times less than 24 hr were not sufficient to introduce the thermal energy needed to enhance polymeric particles coalescence and subsequently delay the drug release.
At curing temperature of 60 LC, curing times as low as 1 hr were sufficient to significantly slow the drug release from the coated pellets compared to the uncured pellets as seen in Figures 20 and 21 (pages 119 and 120). The percent
113 of bumetanide released after 0.5 hr nearly leveled off after 8 hr curing however, the percent of drug released after 4 hr continues to decrease, which indicates that curing may have not been completed even after 168 hr.
Figure 22 (page 121) represents the slowest release profiles obtained after curing for 168 hr at different temperatures. Pellets cured at 50 and 60 LC had similar release profiles followed by pellets cured at 40LC. The results suggested that at a low plasticizer level of 10 %, curing temperature of 40 LC is not sufficient to introduce the thermal energy needed to form complete polymer film on the pellets surface within the curing time studied (168 hr). For this study, curing was considered complete when the dissolution profile is no longer sensitive to curing conditions. To complete the curing process and to establish stable release profile, higher curing temperatures are needed for prolonged period of time. The use of high-energy input for prolonged periods of time is costly and can lead to serious drug stability problems therefore not favored in the pharmaceutical industry. In conclusion, using lower plasticizer level of 10 % is not recommended as it requires prolonged curing at higher temperature to attain complete film formation and reproducible release profile.
114
110 100 90 80 70 60 leased e 50
% R 40 30 20 10 0 02468101214
Time (hr)
Figure 16. Effect of curing time at 40 LC on the release of bumetanide from pellets coated with 6% Eudragit RS plasticized with 10% triethyl citrate. (■) uncured; (□) cured for 1 hr; (▲) cured for 4 hr; (∆) cured for 8 hr; (●) cured for 24 hr; (○) cured for 168 hr.
115 110
100
90
80
70
60 leased e 50 % R 40
30
20
10
0 0 20 40 60 80 100 120 140 160 180 Curing Time (hr)
Figure 17. Effect of curing time on the percent of bumetanide released after 0.5 hr and 4 hr from pellets coated with 6 % Eudragit RS plasticized with 10 % triethyl citrate and cured at 40 L C. (S) percent drug released after 0.5 hr; (■) percent drug released after 4 hr.
116 110
100
90
80
70
60 leased e 50 % R 40
30
20
10
0 02468101214
Time (hr)
Figure 18. Effect of curing time at 50 LC on the release of bumetanide from pellets coated with 6% Eudragit RS plasticized with 10% triethyl citrate. (■) uncured; (□) cured for 1 hr; (▲) cured for 4 hr; (∆) cured for 8 hr; (●) cured for 24 hr; (○) cured for 168 hr.
117
110 100 90 80 70 60 leased e 50
% R 40 30 20 10 0 0 20 40 60 80 100 120 140 160 180 Curing Time (hr)
Figure 19. Effect of curing time on the percent of bumetanide released after 0.5 hr and 4 hr from pellets coated with 6 % Eudragit RS plasticized with 10 % triethyl citrate and cured at 50 L C. (S) percent drug released after 0.5 hr; (■) percent drug released after 4 hr.
118 110
100
90
80
70
60 leased e 50 % R 40
30
20
10
0 02468101214
Time (hr)
Figure 20. Effect of curing time at 60 LC on the release of bumetanide from pellets coated with 6% Eudragit RS plasticized with 10% triethyl citrate. (■) uncured; (□) cured for 1 hr; (▲) cured for 4 hr; (∆) cured for 8 hr; (●) cured for 24 hr; (○) cured for 168 hr.
119
110
100
90
80
70
60
50 % Released 40
30
20
10
0 0 20 40 60 80 100 120 140 160 180 Curing Time (hr)
Figure 21. Effect of curing time on the percent of bumetanide released after 0.5 hr and 4 hr from pellets coated with 6 % Eudragit RS plasticized with 10 % triethyl citrate and cured at 60 L C. (S) percent drug released after 0.5 hr; (■) percent drug released after 4 hr.
120
110
100
90
80
70
60 leased e 50 % R 40
30
20
10
0 02468101214 Time (hr)
Figure 22. Comparison of the slowest release profiles obtained after curing for
168 hr at different temperatures for pellets coated with 6 % Eudragit RS plasticized with 10 % triethyl citrate. (S) uncured; (◊) cured at 40 LC; (□) cured at
50LC; (∆) cured at 60 LC hr.
121 6.2.2. Coated Pellets Plasticized with 15 % Triethyl Citrate
For coated pellets cured at 40 LC, increasing the curing time lead to slower release profile with coated pellets cured for 168 hr showing the slowest release profile as seen in Figures 23 and 24 (pages 123 and 124). The percent of bumetanide released after 0.5 hr leveled off after curing for 8 hr as seen in
Figure 24. On the other hand, the percent of bumetanide released after 4 hr continued to decrease but to a lesser extent after 8 hr curing. The results suggested that curing may have not been completed even after 168 hr.
Different behavior was observed when the curing temperature was increased to 50 LC as depicted in Figures 25 and 26 (pages 125 and 126).
Increasing the curing time initially lead to slower release profiles followed by faster rather than slower profiles when the curing process extended beyond 24 hr. The drug release process was at its slowest pace when the pellets were cured for 24 hr. Similar behavior was noticed when pellets were cured at 60 LC with drug release enhanced when curing time extended beyond 8 hr as seen in
Figures 27 and 28 (pages 127 and 128). It was concluded that curing was completed after 24 hr at 50 LC and 8 hr at 60 LC. The increase in drug release after prolonged curing time can be attributed to drug migration into the film coat during curing as discussed under section 6.2.5.
122
110
100
90
80
70
60 leased e 50 % R 40
30
20
10
0 02468101214 Time (hr)
Figure 23. Effect of curing time at 40 LC on the release of bumetanide from pellets coated with 6% Eudragit RS plasticized with 15% triethyl citrate. (■) uncured; (□) cured for 1 hr; (▲) cured for 4 hr; (∆) cured for 8 hr; (●) cured for 24 hr; (○) cured for 168 hr.
123 110
100
90
80
70
60 leased e 50 % R 40
30
20
10
0 0 20 40 60 80 100 120 140 160 180 Curing Time (hr)
Figure 24. Effect of curing time on the percent of bumetanide released after 0.5 hr and 4 hr from pellets coated with 6 % Eudragit RS plasticized with 15 % triethyl citrate and cured at 40 L C. (S) percent drug released after 0.5 hr; (■) percent drug released after 4 hr.
124
110
100
90
80
70
60 leased e 50 % R 40
30
20
10
0 0 2 4 6 8 10 12 14 Time (hr)
Figure 25. Effect of curing time at 50 LC on the release of bumetanide from pellets coated with 6% Eudragit RS plasticized with 15% triethyl citrate. (■) uncured; (□) cured for 1 hr; (▲) cured for 4 hr; (∆) cured for 8 hr; (●) cured for 24 hr; (○) cured for 168 hr.
125
110
100
90
80
70
60 leased e 50
% R 40
30
20
10
0 0 20 40 60 80 100 120 140 160 180 Curing Time (hr)
Figure 26. Effect of curing time on the percent of bumetanide released after 0.5 hr and 4 hr from pellets coated with 6 % Eudragit RS plasticized with 15 % triethyl citrate and cured at 50 L C. (S) percent drug released after 0.5 hr; (■) percent drug released after 4 hr.
126
110
100
90
80
70
60 leased e 50 % R 40
30
20
10
0 02468101214 Time (hr)
Figure 27. Effect of curing time at 60 LC on the release of bumetanide from pellets coated with 6% Eudragit RS plasticized with 15% triethyl citrate. (■) uncured; (□) cured for 1 hr; (▲) cured for 4 hr; (∆) cured for 8 hr; (●) cured for 24 hr; (○) cured for 168 hr.
127
110
100
90
80
70
60 ased e l e 50 R % 40
30
20
10
0 0 20 40 60 80 100 120 140 160 180 Curing Time (hr)
Figure 28. Effect of curing time on the percent of bumetanide released after 0.5 hr and 4 hr from pellets coated with 6 % Eudragit RS plasticized with 15 % triethyl citrate and cured at 60 L C. (S) percent drug released after 0.5 hr; (■) percent drug released after 4 hr.
128 6.2.2.1. Investigation of the Release Mechanism
Different research groups have investigated drug release mechanisms from coated multiparticulate systems. The systems used included coated spherical drug matrices (Husson et al., 1991), coated spherical drug pellets with or without the addition of pore formers into the coat (Gunder et al., 1995; Frohoff-
Hulsmann et al., 1999; Lippold et al., 1999) and coated spherical system into which an osmotic agent has been incorporated (Schulze and Kleinebudde,
1997). In another approach, statistical models using Gompertz equation and logistic function have been used to describe the release of drugs from coated pellets (Hutchings and Sakr, 1994).
In most of the systems used, the pellets were comprised mainly of drug with or without the addition of water insoluble filler as microcrystalline cellulose.
In our study drug pellets were manufactured by spray-layering the drug into sugar pellets commercially known as Nu-pariels. Nu-pariels sugar pellets are comprised mainly of lactose, a water-soluble osmotic agent. The release mechanism from pellets manufactured by spray layering the drug onto sugar pellets is expected to differ from pellets manufactured by different methods.
Similarity in drug release mechanisms is only expected between our pellets and those into which osmotic agents were incorporated. Drug release from coated pellets into which osmotic agents were incorporated followed zero order kinetics
(Schulze and Kleinebudde, 1997).
To accurately investigate drug release mechanism from coated pellets, only the completely cured pellets were considered. Incomplete curing entails
129 irregular film coat with time-dependent structural changes that can confound any accurate estimation of drug release mechanisms. For coated pellets plasticized with 15 % triethyl citrate, curing was considered complete after 168 hr at 40 LC,
24 hr at 50 LC, and after 8 hr at 60 LC. Zero order kinetics are expected to control the drug release from coated pellets where osmotic agents have been incorporated. Drug release following zero order kinetics can be modeled as the following:
Mt = K t (8)
Where, Mt = the amount of drug released at time t; K= the drug release rate.
To investigate whether the drug release mechanism followed zero order kinetics, the percent of drug released at different time points (0.5 h to 8 h) were regressed over time. The time range was selected to avoid the initial time period
(0.5 h) where steady state diffusion or convection are not yet established. On the other hand, the percent of bumetanide released after 8 h exceeded more than
65% of the total dose incorporated in most of the coated pellets studied, thus bumetanide release after 8 h is expected to deviate from the zero order behavior since the concentration of the drug inside the pellets may drop to less than saturation.
For the completely cured pellets, very high correlation coefficients were calculated when the percent of drug released were regressed over time (0.974 for the pellets cured at 40 LC for 168 hr, 0.998 for pellets cured for 24 hr at 50 LC, and 0.996 for pellets cured 8 hr at 60 LC). The high correlation coefficient
130 suggested that the release followed zero order kinetics. Water is expected to diffuse through the polymeric film into the pellets dissolving lactose and other water-soluble ingredients. The water-soluble plasticizer leaches out of the film coat creating channels through which water carrying the drug convects under the osmotic pressure difference established resulting in a zero order release process.
6.2.3. Coated Pellets Plasticized with 20 % Triethyl Citrate
The release profiles from pellets cured at 40 LC are displayed in Figures
29 and 30 (pages 134-135). Curing can be considered complete after 4 hr.
Pellets cured for 4 and 8 hr had almost identical release profiles while pellets cured for longer periods had faster release profiles. For pellets cured at 50 and
60 LC, curing was complete in 1 hr. Pellets cured for more than 1 hr had faster release profiles as seen in Figures 31-34 (pages 136-139). Increasing the curing temperature accelerated the polymeric particle coalescence process with a subsequent reduction in the time needed to obtain completely coalesced and smooth film to only 1 hr at temperature of 50 and 60 LC.
In addition to curing temperature, the level of the plasticizer used dramatically influenced the rate at which polymeric particles coalesced. Table 18
(page 140) shows the time required to completely cure the coated pellets at different curing conditions. Curing was considered completed when the drug release process became insensitive to curing time and before the release process began to accelerate. The time needed to completely cure the coated
131 pellets at certain curing temperature decreased as the level of plasticizer increased. For curing temperature of 40 LC, the time needed to completely cure the coated pellets decreased from more than 168 h at plasticizer level of 10% to almost 168 h at plasticizer level of 15% and to less than 8 h at plasticizer level of
20%. At curing temperature of 50 LC, the time needed to completely cure the coated pellets decreased from more than 168 h at plasticizer level of 10% to less than 24 h at plasticizer level of 15% and to 1 h at plasticizer level of 20%. Finally, at curing temperature of 60 LC, the time needed to completely cure the coated pellets decreased from more than 168 h at plasticizer level of 10% to less than 8 h at plasticizer level of 15% and to 1 h at plasticizer level of 20%.
Increasing plasticizer level has been reported to enhance polymeric coalescence and allows curing to be complete within shorter curing times
(Bodmeier and Paeratakul, 1994; Amighi and Moes, 1995). The findings can be explained in terms of the increased polymer flexibility with increasing the plasticizer level which allows for more polymeric particles deformation and fusion at shorter curing times or lower curing temperature.
6.2.3.1. Investigation of the Release Mechanism
Curing was considered complete after 4 hr at 40 LC, 1 hr at 50 and 60LC.
Drug release from the completely cured pellets was better modeled using the square-root-of-time relationship universally known as Higuchi’s model. Higher correlation coefficient were estimated when the data were fit into Higuchi’s model
132 than zero order kinetic model (0.995 for pellets cured for 4 hr at 40 LC, 0.992 for pellets cured for 1 hr at 50 LC, and 0.994 for pellets cured for 1 hr at 60 LC).
Higuchi’s model was developed to describe the release of drug from plastic, insoluble, and non-swellable matrices though which the drug is uniformly distributed (Desai et al., 1965). The model was established using thin slabs of drug-carrying matrices and can be written as:
1/2 Q = (D Cs/ (2A- Cs) t) (9)
Where, Q= The amount of drug released per unit area at time t; D= diffusion coefficient of drug in the release medium; = porosity of the polymeric matrix
Cs = solubility of the drug in release medium; = tortuosity if the matrix; A = concentration of the drug in the matrix (g/ml) (Desai et al., 1965)
133 110
100
90
80
70
60 leased e 50 % R 40
30
20
10
0 02468101214
Time (hr)
Figure 29. Effect of curing time at 40 LC on the release of bumetanide from pellets coated with 6% Eudragit RS plasticized with 20% triethyl citrate. (■) uncured; (□) cured for 1 hr; (▲) cured for 4 hr; (∆) cured for 8 hr; (●) cured for 24 hr; (○) cured for 168 hr.
134
110
100
90
80
70
60 leased e 50
% R 40
30
20
10
0 0 20 40 60 80 100 120 140 160 180 Curing Time (hr)
Figure 30. Effect of curing time on the percent of bumetanide released after 0.5 hr and 4 hr from pellets coated with 6 % Eudragit RS plasticized with 20 % triethyl citrate and cured at 40 L C. (S) percent drug released after 0.5 hr; (■) percent drug released after 4 hr.
135 110
100
90
80
70
60 leased e 50 % R 40
30
20
10
0 02468101214 Time (hr)
Figure 31. Effect of curing time at 50 LC on the release of bumetanide from pellets coated with 6% Eudragit RS plasticized with 20% triethyl citrate. (■) uncured; (□) cured for 1 hr; (▲) cured for 4 hr; (∆) cured for 8 hr; (●) cured for 24 hr; (○) cured for 168 hr.
136 110
100
90
80
70
60 leased e 50 % R 40
30
20
10
0 0 20 40 60 80 100 120 140 160 180 Curing Time (hr)
Figure 32. Effect of curing time on the percent of bumetanide released after 0.5 hr and 4 hr from pellets coated with 6 % Eudragit RS plasticized with 20 % triethyl citrate and cured at 50 L C. (S) percent drug released after 0.5 hr; (■) percent drug released after 4 hr.
137
110
100
90
80
70
60 leased e 50 % R 40
30
20
10
0 0 2 4 6 8 10 12 14
Time (hr)
Figure 33. Effect of curing time at 60 LC on the release of bumetanide from pellets coated with 6% Eudragit RS plasticized with 20% triethyl citrate. (■) uncured; (□) cured for 1 hr; (▲) cured for 4 hr; (∆) cured for 8 hr; (●) cured for 24 hr; (○) cured for 168 hr.
138 110
100
90
80
70
60 leased e 50 % R
40
30
20
10
0 0 20 40 60 80 100 120 140 160 180 Curing Time (hr)
Figure 34. Effect of curing time on the percent of bumetanide released after 0.5 hr and 4 hr from pellets coated with 6 % Eudragit RS plasticized with 20 % triethyl citrate and cured at 60 L C. (S) percent drug released after 0.5 hr; (■) percent drug released after 4 hr.
139 Table 18. Effect of plasticizer level on reducing the time needed to completely cure the coated pellets at different curing temperatures.
Plasticizer Level Curing Temperature 40 LC 50 LC 60 LC
10% 168 168 168
15% 168 24 O X O 8 8 O X O 4
20% 8 O X O 4 1 1
140 The equation predicts a linear relationship between the amount of drug released (or the percent of drug released) and the square root of time. The fact that bumetanide release from the coated pellets was better modeled using
Higuchi’s equation suggested that bumetanide may have migrated into the film coat during curing. As a result, the predominant release pathway changed from the water-carried convection through pores within the polymeric film seen with coated pellets plasticized with 15 % triethyl citrate to diffusion through tortuous and porous polymeric film. Using high level of water-soluble plasticizer of 20 % lead to softer polymeric film at the curing temperatures used which facilitated the diffusion of bumetanide into the film coat during curing. When the pellets were introduced into liquid media the water-soluble plasticizer leached out creating tortuous paths within the film coat through which bumetanide diffused out. The release of bumetanide from the hydrated, tortuous and porous plastic films was closely related to the system used to establish Higuchi’s equation hence, the release was better modeled using the square-root-of time relationship.
6.2.4. Effect of Curing Method on Bumetanide Release from Coated Pellets
Bumetanide-loaded pellets coated with 6 % Eudragit RS plasticized with
20 % triethyl citrate were cured at 50 LC in both oven and fluid bed equipment.
Samples cured in the fluid bed equipment were removed after 10, 20, 30, and 60 min. The release profiles from the cured pellets are shown in Figure 35 (page
143). Increasing the curing time slowed the release of bumetanide from the coated pellets owing to the enhanced polymeric particle coalescence. Figure 36
141 (page 144) displays the release profiles from coated pellets cured for 1 hr at 50
LC in both the oven and the fluid bed equipment. As seen in the figure, no difference in release profile was noticed between the pellets cured using both methods. Statistical analysis using ANOVA revealed no significance difference in drug release at each time point. The results suggested that the tumbling effect induced by fluidization in the fluid bed did not interfere with the polymeric particle coalescence that took place on the surface of the coated pellets. Consequently, coated pellets can be cured in the fluid bed equipment at the end of the coating cycle with no need for using different heating equipment. The use of limited number of processing steps and equipment is favored in the pharmaceutical industry. Excessive handling of processed materials increases labor, cost of production, and the risk of cross contamination and batch errors.
6.2.5. The Theory of Over Curing
The increase in bumetanide release observed with coated pellets plasticized with 15 and 20% triethyl citrate after curing for prolonged periods of times can be attributed to over curing. Over curing can be defined as curing coated dosage forms for a period of time beyond that needed for optimal film formation with a subsequent increase in drug release. A very limited number of publications addressed the phenomenon of over curing and two theories have been reported to explain the increase in drug release after prolonged curing.
142
110
100
90
80
70
60 leased e 50
% R 40
30
20
10
0 02468101214 Time (hr)
Figure 35. Effect of curing time at 50 LC in fluid bed equipment on bumetanide release from pellets coated with 6 % Eudragit RS plasticized with 20 % triethyl citrate. (■) uncured; (□) cured for 10 min; (▲) cured for 20 min; (∆) cured for 30 min; (S) cured for 60 min.
143
110
100
90
80
70
60 eased 50 Rel % 40
30
20
10
0 02468101214 Time(hr)
Figure 36. Effect of curing method on bumetanide release from pellets coated with 6 % Eudragit RS plasticized with 20% triethyl citrate and cured at 50LC for 1 hr. (■) uncured; (▲) cured in fluid bed equipment ; (∆) cured in oven.
144 The first theory proposed that the use of excessive thermal energy lead to water and plasticizer loss from the film coat. Water acts as a plasticizer allowing more flexible polymeric chains movement, which lead to enhancement of polymeric particles coalescence and hence smooth film formation. The loss of water and plasticizer increases the polymer brittleness and can lead to the formation of microvoids in certain regions of the film, which can act as alternative passage for drug release resulting in faster release profiles. Moreover, the microvoids can continue to grow and increase in number with a subsequent formation of cracks (Guma et al., 1997).
The second theory to explain over curing suggested that drug migration to the film coat during curing is responsible for the increase in drug release
(Bodmeier and Paeratakul, 1994). Bumetanide is a very lipophilic material with a calculated log octanol/water partiotion coefficient of 3.9 (compared to 2.4 for benzene). The rubbery state of the polymeric coat existing at higher curing temperature (50L and 60 LC) can facilitate drug migration. On the other hand, increasing the plasticizer level with the subsequent decrease in the polymer glass transition temperature can exacerbate the drug migration process.
Therefore, over curing was seen at shorter curing time and lower temperature with coated pellets plasticized with 20 % vs. 15 % triethyl citrate (after 1 hr curing at 50 L C for coated pellets plasticized with 20 % triethyl citrate vs. 168 hr for coated pellets plasticized with 15 % triethyl citrate; after 1 hr curing at 60 L C for coated pellets plasticized with 20 % triethyl citrate vs. 24 hr for coated pellets plasticized with 15 % triethyl citrate).
145 To investigate more into the causes of over curing and how to minimize or avoid it, two different approaches were implemented. The first depends on the use of a sealing hydrophilic coat to separate the drug from Eudragit RS film coat.
The second depends on reducing the lipophilicity of the coat by mixing Eudragit
RS with more hydrophilic polymer as Eudragit RL.
6.2.5.1. Effect of HPMC Seal Coat on Bumetanide Release after Curing
As mentioned earlier, owing to the high lipophilicity of bumetanide it tends to migrate to the lipophilic film coat during curing with a subsequent increase in the release. Bumetanide migration is expected to continue even at room temperature particularly for film coat plasticized with higher level of triethyl citrate due to the increased polymer chain flexibility which lead to an increase in the diffusion coefficient of bumetanide through the film coat. Consequently, instability in drug release profile are expected with faster release profiles obtained after prolonged storage times at room temperature.
Figure 37 (page 148) compares the dissolution profiles from pellets coated with 6 % Eudragit RS plasticized with 20 % triethyl citrate and completely cured at 60 LC for 1 hr at time zero and after storage at room temperature for 3 months.
As seen in figure, bumetanide release was faster after storage for 3 months. The increase in bumetanide release was expected due to bumetanide migration into the film coat at room temperature. One of the approaches used to minimize bumetanide migration during curing and hence stabilize the release profile after
146 storage for prolonged period of time is to seal coat the pellets with a hydrophilic material prior to coating with Eudragit RS.
Hydroxypropyl methyl cellulose (HPMC) was selected to seal coat the pellets since it is a hydrophilic polymer with enough spreadability to allow for film coating besides it is expected not to change the release profile if applied in lower level. The lower viscosity brand (HPMC 100 LV) was used to avoid any complication in the spraying process that may arise from using higher viscosity brands and the pellets were seal coated with 1 % w/w HPMC. Figure 38 (page
149) displays the release profiles from bumetanide pellets with and without
HPMC seal coat. As seen in the figure, HPMC imparted no effect on the release of bumetanide from the pellets and is expected not to interfere with drug release through the coated pellets except through its effect on separating the drug layer from the polymeric film coat.
Bumetanide pellets seal coated with HPMC and coated with 6 % Eudragit
RS plasticized with 20% triethyl citrate were cured at 60 L C for 1, 4, 8, and 24 hr.
The higher plasticizer level and curing temperature were selected to accelerate the drug migration process in order to increase the sensitivity of the measurement of the effect of HPMC seal coat on the drug migration process.
Figure 39 (page 150) presents the dissolution profiles obtained from pellets seal coated with HPMC and cured for different times at 60 LC. As seen in figure, similar pattern to that obtained with pellets without the HPMC seal coat was observed. The curing process can be considered complete after 1 hr with pellets cured for longer periods displaying faster rather than slower release profiles.
147
110
100
90 80
70 60
eased l
e 50
R
% 40 30 20
10 0
02468101214 Time (hr)
Figure 37. Effect of storage time at room temperature on bumetanide release from pellets coated with 6 % Eudragit RS plasticized with 20 % triethyl citrate and
.months 3 (ڤ) ;cured for 1 hr at 60 LC. (■) zero time
148 110
100
90
80
70
60 leased e 50 % R 40
30
20
10
0 0 102030405060 Time (min)
Figure 38. Effect of HPMC seal coating on bumetanide release. (■) pellets with
.pellets with 1 % HPMC seal coat (ڤ) ;no HPMC seal coat
149
110
100
90
80
70
60 leased e 50 % R 40
30
20
10
0 0 2 4 6 8 10 12 14
Time (hr)
Figure 39. Effect of curing time at 60 LC on bumetanide release from pellets seal coated with 1 % HPMC and coated with 6% Eudragit RS plasticized with 20 %
cured for 1 hr; (▲) cured for 4 hr; (∆) cured for (ڤ) ;triethyl citrate. (■) uncured
8 hr; (●) cured for 24 hr.
150 To focus on the effect of HPMC seal coat on bumetanide release after curing, comparison of the release profiles of uncured and cured pellets with and without HPMC seal coat must be made. Figure 40 (page 153) shows the release profiles from pellets with and without the HPMC seal coat and cured for 1 hr at 60
LC. The effect of curing is more significant for pellets with the HPMC seal coat than pellets without the HPMC seal coat. The release of bumetanide from uncured pellets with the HPMC seal coat tends to level off after 8 hr at 70 % drug release. The exact reasons for this unusual behavior are still unclear. Curing, however, improved the drug release from the coated pellets. Therefore, only the percent of drug released within the first 8 hr will be used for comparison.
Table 19 (page 154) presents a comparison of the differences in the percent drug released at different time points between uncured and cured pellets
(1 hr curing at 60L C) for pellets with and without the HPMC seal coat. As seen in Table 19, curing lead to a decrease in drug release from pellets with and without the HPMC seal coat as reflected in the positive signs of the difference in percent drug released with time. Nonetheless, the decrease in release after curing is more significant for pellets seal coated with HPMC (p-value less than
0.05). The results can be explained by considering the separation of the drug layer from the film coat by the hydrophilic HPMC layer with a subsequent reduction in drug migration during curing. Minimizing the drug migration lead to a slower release profile after curing.
The release of bumetanide from pellets seal coated with HPMC was better modeled using zero order kinetics than the Higuchi’s model (correlation
151 coefficient of 0.990). The findings suggested that the drug release pathway changed from diffusion through tortuous and porous film membrane predominant in pellets without the HPMC seal coat to the osmotically derived water-carried convection. Such change in drug release mechanism is expected when the drug layer is separated from the film coat.
Extending curing time to 4 hr caused the drug release from pellets seal coated with HPMC to be faster than pellets cured for 1 hr as seen in Figure 41
(page 155). Table 20 (page 156) presents a comparison of the differences in the percent drug released with time between uncured and cured pellets (4 hr curing at 60 L C ) for pellets with and without the HPMC seal coat. Curing had more significant effect on reducing bumetanide release from pellets seal coated with
HPMC than pellets without the seal coat. Nevertheless, the effect is less pronounced than that obtained after curing for 1 hr, which suggested that some bumetanide may have migrated to the wall during curing even after seal coating with HPNC.
When the curing time was increased to 8 hr, no statistically significant difference could be seen in the decrease in drug release induced by curing between pellets seal coated with HPMC and those without the seal coat as seen in Figure 42 and Table 21 (pages 157 and 158). The results suggested that at prolonged curing time as 8 hr, the HPMC seal coat was not enough to prevent drug migration into the film coat. Subsequently, there was an increase rather than a decrease in drug release after curing as reflected in the negatives values in Table 21.
152
110
100
90
80
70
60 leased e 50 % R 40
30
20
10
0 02468101214
Time (hr)
Figure 40. Effect of seal coating with 1% HPMC and curing at 60 LC for 1 hr on the release of bumetanide from pellets coated with 6 % Eudragit RS plasticized with 20 % triethyl citrate. (__■__) no HPMC seal coating no curing; (---■---) no
(--ڤ--) ;HPMC seal coating no curing (__ڤ__) ;HPMC seal coating and curing
HPMC seal coating and curing.
153 Table 19. Effect of HPMC seal coating on the differences in percent drug released from pellets cured for 1 hr at 60 LC and uncured pellets (uncured – cured)
Release Time Pellets without Pellets with HPMC P-Value
Point HPMC Seal Coat Seal Coat
(uncured - cured) (uncured - cured)
0.5 4.42* 23.97 0.0017
1 10.20* 36.51 0.0002
2 15.10* 45.38 0.0002
4 12.76* 35.31 0.0002
8 -2.49* 6.22 0.0056
* ANOVA statistically significant from pellets with the HPMC seal coat (Q = 0.05)
154 110
100
90
80
70
60 leased e 50 % R 40
30
20
10
0 02468101214 Time (hr)
Figure 41. Effect of seal coating with 1% HPMC and curing at 60 LC for 4 hr on the release of bumetanide from pellets coated with 6 % Eudragit RS plasticized with 20 % triethyl citrate. (__■__) no HPMC seal coating no curing; (---■---) no
(--ڤ--) ; HPMC seal coating no curing (__ڤ__) ;HPMC seal coating and curing
HPMC seal coating and curing.
155 Table 20. Effect of HPMC seal coating on the differences in percent drug released from pellets cured for 4 hr at 60 LC and uncured pellets (uncured – cured)
Release Time Pellets without Pellets with HPMC P-Value
Point HPMC Seal coat Seal coat
(uncured – cured) (uncured – cured)
0.5 -2.18* 11.49 0.0131
1 0.76* 16.89 0.0120
2 4.33* 24.65 0.001
4 5.25* 17.10 0.0417
8 -1.34 -4.51 0.2794
* ANOVA statistically significant from pellets with the HPMC seal coat (Q = 0.05)
156 110
100
90
80
70
60 eased
Rel 50 % 40
30
20
10
0 02468101214 Time (hr)
Figure 42. Effect of seal coating with 1% HPMC and curing at 60 LC for 8 hr on the release of bumetanide from pellets coated with 6 % Eudragit RS plasticized with 20 % triethyl citrate. (__■__) no HPMC seal coating no curing; (---■---) no
(--ڤ--) ; HPMC seal coating no curing (__ڤ__) ;HPMC seal coating and curing
HPMC seal coating and curing.
157 Table 21. Effect of HPMC seal coating on the differences in percent drug released from pellets cured for 8 hr at 60 LC and uncured pellets (uncured – cured)
Release Time Pellets without Pellets with HPMC P-Value
Point HPMC Seal coat Seal coat
(uncured – cured) (uncured – cured)
0.5 -5.39 -4.35 0.3103
1 -2.63 -4.13 0.1994
2 -1.17 4.74 0.06
4 1.05 3.12 0.2665
8 -9.97 -10.49 0.6879
158 Increasing the curing time to 24 hr lead to an opposite effect with pellets seal coated with HPMC showing more increase in drug release after curing than pellets without the seal coat as seen in Figure 43 and Table 22. The exact reasons beyond this unexpected effect are still to be investigated. The solubilizing effect of polyethylene glycol (PEG 6000) incorporated in the HPMC seal coat can not be ruled out as a reason for the increased release. However, further investigation into the exact mechanism of the release enhancement under such prolonged curing conditions is still needed.
In conclusion, seal coating with 1 % HPMC successfully separated bumetanide from the polymer film coat with a subsequent decrease in drug migration during curing. HPMC seal coat of only 1 % weight gain can maintain complete drug separation from the film coat after curing at 60 LC for 1 hr only.
Longer curing time lead to drug migration into the film coat with a subsequent increase in drug release. Higher level of HPMC is needed to maintain the stability of drug release profile after prolonged curing time. Curing at 60 L C for 1 hr was considered the optimum curing conditions for pellets coated with 6 % Eudragit
RS plasticized with 20 % triethyl citrate and seal coated with 1 % HPMC.
159
110
100
90
80
70
60 leased e 50 % R 40
30
20
10
0 02468101214 Time (hr)
Figure 43. Effect of seal coating with 1% HPMC and curing at 60 LC for 24 hr on the release of bumetanide from pellets coated with 6 % Eudragit RS plasticized with 20 % triethyl citrate. (__■__) no HPMC seal coating no curing; (---■---) no
(--ڤ--) ; HPMC seal coating no curing (__ڤ__) ;HPMC seal coating and curing
HPMC seal coating and curing.
160 Table 22. Effect of HPMC seal coating on the differences in percent drug released from pellets cured for 24 hr at 60 LC and uncured pellets (uncured – cured)
Time Point Pellets without Pellets with HPMC P-Value
HPMC Seal coat Seal coat
(uncured – cured) (uncured-cured)
0.5 -21.02* -43.45 0.0015
1 -22.19* -39.89 0.00002
2 -19.2* -25.54 0.0013
4 -14.02* -17.76 0.0172
8 -16.50* -21.98 0.0088
12 -20.67* -22.51 0.0109
* ANOVA statistically significant from pellets with the HPMC seal coat (Q = 0.05)
161 6.2.5.2. Effect of Incorporating Eudragit RL into the Film Coat on Bumetanide Release after Curing
Eudragit RS and RL are methacrylate ester copolymers comprised of three different monomers namely ethyl acrylate (EA), methylmethacrylate (MMA), and trimethylammonioethyl methacrylate chloride (TAMCI). The chemical structures of Eudragit RS and RL are shown in Figure 44. They differ in the percent of TAMCI in the polymeric chains with Eudragit RS carrying a ratio of
EA:MMA:TAMCI of 2:1:0.1 and Eudragit RL carrying a ratio of 2:1:0.2. TAMCI introduces hydrophilic properties into the polymer and therefore Eudragit RL is more hydrophilic and permeable than Eudragit RS (Lehmann, 1997).
Combinations of Eudragit RS and RL can be adjusted to obtain different drug release profiles from coated tablets, capsules, granules, pellets and crystals.
The purpose of this study was to investigate the effect of increasing polymeric film hydrophilicity on bumetanide migration into the film coat during curing. The incorporation of Eudragit RL into the film coat is expected to increases the hydrophilicity of the coat, which is hypothesized to limit drug migration during curing. Bumetanide pellets were coated with the polymeric combination plasticized with 20% triethyl citrate and cured at 60 LC. The
Plasticizer level and the curing conditions were selected as they represent the conditions which favor drug migration and hence provide a sensitive indication of any drug migration impediment effect. Eudragit RL was mixed with Eudragit RS in a ratio of 10:90. Pellets were cured at 60 LC for 1, 4, 8, and 24 hr.
162
R2 CH3
CH C CH C CH 2 2 2
COOR1 + - COOCH2—CH2—N (CH3)3Cl
Ethyl acrylate; R1 = CH2—CH3 , R2 = H
Methyl methacrylate; R1 = CH3 , R2 = CH3
Figure 44. Chemical structure of Eudragit RS and RL
163 Figure 45 (page 166) presents drug release profiles from pellets coated with 6% Eudragit RL/RS and cured for different times. Curing had a significant effect on reducing bumetanide release from the coated pellets after 1 hr.
Increasing curing time beyond 1 hr lead to a faster rather than slower release profile, a similar pattern to that seen with pellets coated with Eudragit RS only.
Nevertheless, the magnitude of increase in drug release with increasing curing time was lower in case of pellets coated with polymeric combination than pellets coated with Eudragit RS alone.
To compare the effect of incorporating Eudragit RL on the dissolution profiles of cured pellets, the percent of drug released after 0.5 and 1 hr were used for the analysis since they represent the most sensitive dissolution parameters to changes in polymeric film formation and drug migration. Table 23
(page 167) presents the differences in release after 0.5 and 1 hr between uncured and cured pellets coated with Eudragit RS alone and combination of
Eudragit RL/RS. As seen in Table 23, for pellets cured for 1 hr, incorporating
Eudragit RL into the film coat resulted in a more significant decrease in release than that observed with pellets coated with Eudragit RS alone (p-values of
0.0005 and 0.004 at time points of 0.5 and 1 hr respectively).
Pellets coated with Eudragit RS alone and cured for 4 hr or more had faster release than the uncured samples as reflected by the negative signs in
Table 23. On the other hand, pellets coated with polymeric combination maintained slower release profile than the uncured samples even after 8-hr curing. Moreover, the extent to which the drug release was enhanced from
164 pellets coated with the polymeric combination after curing for 24 hr was significantly lower than that obtained with pellets coated with Eudragit RS alone
(p-values of 0.032 and 0.003 for dissolution time points of 0.5 an 1 hr respectively).
The results indicated that increasing the hydrophilicity of the film coat limited the increase in drug release upon curing. The findings suggested that the increase in drug release after curing can be attributed to drug migration to the film coat during curing which was minimized by increasing the coat hydrophilicty.
6.2.6. Effect of Storage on Bumetanide Release from Coated Pellets
The following pellets were stored at room temperature for 3 months:
1) Pellets coated 6% Eudragit RS plasticized with 20 % triethyl citrate and cured at 60LC for 1 hr; 2) Pellets coated with 6% Eudragit RS plasticized with 20 % triethyl citrate, seal coated with 1 % HPMC and cured at 60LC for 1 hr.
Figure 46 (page 168) presents a comparison of the release profiles from pellets coated with 6 % Eudragit RS plasticized with 20% triethyl citrate and completely cured at time zero and after storage for three months. As seen in Figure 46, bumetanide release was accelerated at the end of the storage time in comparison to time zero. The results suggested that bumetanide continued to migrate to the polymer film after curing and during storage. The relatively low temperature used (25 LC) was enough to allow sufficient flexibility within the polymer chains to permit for bumetanide molecules partition and migration through the film coat.
165
110
100
90
80
70
60 leased e 50 % R 40
30
20
10
0 02468101214
Time (hr)
Figure 45. Effect of curing time at 60L C on the release of bumetanide from
cured for 1 (ڤ) ;pellets coated with 6 % Eudragit RL/RS (10/90). (■) not cured hr; (▲) cured for 4 hr; (∆) cured for 8 hr; (●) cured for 24 hr.
166 Table 23. Effect of incorporating Eudragit RL on the differences in release between pellets cured for different times at 60 LC and uncured pellets (uncured – cured)
Curing Time Release Pellets Coated with Pellets coated with P-Value
(hr) Time Point Eudragit RS Eudragit RL/RS
(uncured – cured) (uncured – cured)
1 0.5 4.42* 37.27 0.0005
1 10.20* 24.97 0.004
4 0.5 -2.18* 14.58 0.049
1 0.76* 15.19 0.014
8 0.5 -5.39* 14.91 0.023
1 -2.63* 13.62 0.033
24 0.5 -21.02* -6.82 0.032
1 -22.19* -5.92 0.003
* ANOVA statistically significant from pellets coated with Eudragit RL/RS.
167
110
100
90
80
70
60 leased e 50 % R 40
30
20
10
0 02468101214 Time (hr)
Figure 46. Effect of HPMC seal coating and storage time at room temperature on bumetanide release from pellets coated with 6 % Eudragit RS plasticized with
20 % triethyl citrate and completely cured. (■) HPMC seal coating, zero time;
;No HPMC seal coating, zero time (ڤ) ;HPMC seal coating, 3-month storage (▲)
(∆) No HPMC seal coating, 3-month storage.
168 The finding revealed that instability in drug release from such dosage form is inevitable during the normal shelf-life period of two years.
On the other hand, bumetanide pellets coated with 6 % Eudragit RS plasticized with 20 % triethyl citrate, seal coated with 1 % HPMC and completely cured showed no effect of storage on drug release as seen in Figure 46. The
HPMC seal coat successfully separated bumetanide layer from the polymer film coat and maintained this separation during the three-months storage period.
6.2.7. Final Optimization of Bumetanide Release
As mentioned earlier, three different release profiles were targeted to provide enough data to establish in vitro/in vivo correlation. The intermediate release profile was selected to be administered to laboratory animals to test the hypothesis that extended release bumetanide formulations have better diuretic effect than immediate release formulations. Nevertheless, preliminary data from rabbits to which immediate release bumetanide formulations were administered showed two serum concentration peaks. Rabbits are known to consume their feces, a behavior referred to as coprophagy (Suckow and Douglas, 1997). It is our belief that such behavior caused the serum double-peak phenomenon.
Unabsorbed unchanged bumetanide may have been excreted in the feces and recirculated orally when the rabbits ingested their feces.
169 Elizabeth collar was installed around the rabbit neck to prevent any recirculation of fecal materials. The use of Elizabeth collar abolished the second bumetanide serum peak. The results confirmed that the second serum peak was due to recirculating the feces rather than enterohepatic recirculation. The results suggested that the gastrointestinal residence time for the rabbits may be shorter than expected (8 to 12hr) since the second serum peak was seen after 8-10 hr.
Therefore, faster extended release profile was targeted to assure that most of bumetanide dose will be released after 6-8 hr to achieve better comparison with the immediate release formulation. The new dissolution profile targeted had the following release characteristics, 30% after 0.5 hr, 45 % after 1 hr, 65% after 2 hr, 75% after 4 hr, 90% after 8hr, and 95% after 12hr.
Channeling agents were considered to increase the release of bumetanide from the coated pellets to achieve the new release profile. Channeling agents are water soluble materials that are included into the film coat to be leached out upon contact with aqueous media creating channels through which drug is release. The use of channeling agents to modify drug release from coated dosage forms has been reported in the literature (Kallstrand and Ekman, 1983;
Ghebre-Sellasaie, 1987, Benedikt et al., 1988; Bodmeier and Paeratakul, 1991).
Different water-soluble channeling agents were used including Hydroxypropyl methyl cellulose (Gunder et al., 1995; Harris and Ghebre-Sellasie, 1997;
Steuernagel, 1997; Frohoff-Hulsmann et al., 1999) sodium chloride (Tirkkonen and Paronen,1992) and calcium phosphate dibasic (Bodmeier and Paeratakul,
1990).
170 Sodium chloride was chosen as a channeling agent for this study in a level of 5 % of the dry polymer weight gain. Drug release profiles from pellets coated with 6 % Eudragit RS plasticized with 20 % triethyl citrate with and without sodium chloride are shown in Figure 47. The addition of sodium chloride significantly accelerated bumetanide release from the coated pellets after a lag time of 1 hr. The lag time is required to dissolve and leach sodium chloride out of the coat creating the necessary channels for drug diffusion.
As seen in Figure 47, the target drug release profile is intermediate between the release profiles of pellets with and without sodium chloride. A mixture of pellets coated with 6% Eudragit RS plasticized with 20% triethyl citrate with and without sodium chloride in the ratio of 1:1w/w was prepared and tested for its release profile in USP purified water. Figure 48 presents the release profiles from mixture of pellets with and without sodium chloride in comparison to the target release profile. The release profile obtained from the mixture of pellets almost coincided on the target release profile as seen in Figure 48 and Table 24.
Mixture of pellets with and without sodium chloride was considered as the final formulation to be administered to laboratory animals as the extended release bumetanide formulation.
171
110
100
90
80
70
60
eeleased 50
% R 40
30
20
10
0 02468101214
Time (hr)
Figure 47. Effect of sodium chloride on the release of bumetanide from pellets coated with 6% Eudragit RS plasticized with 20% triethyl citrate. (■) 0% sodium chloride; (□) 5% sodium chloride; (○) target.
172
110
100
90
80
70
60 leased e 50 % R 40
30
20
10
0 02468101214
Time (hr)
Figure 48. Release profiles from mixture of pellets coated with 6% Eudragit RS plasticized with 20 % triethyl citrate with and without sodium chloride compared to target profile. (□) coated pellets; (○) target.
173 Table 24. Comparison of release profile from mixture of pellets with and without sodium chloride and target release profile
Release Time Point Target Mixture of Pellets
0.5 30 29.73
1 45 42.95
2 65 64.77
4 75 79.62
8 90 88.08
12 95 94
174 6.3. Part III: Testing Selected Optimized Bumetanide Extended Release
Formulation in Laboratory Animals
6.3.1. Validation of Bumetanide HPLC Analytical Method
Linearity and Range
The calibration equations relating bumetanide concentrations in both serum and urine samples to relative fluorescence intensities were found linear
(correlation coefficient 0.996) over the following concentration ranges:
Serum: 0.05- 2 ug/ml ; Urine: 0.167-6.667 ug/ml
Accuracy
Three different new bumetanide concentrations in serum and urine (0.05,
0.25 and 1 ug/ml) were measured using the developed calibration equations.
Accuracy was estimated by comparing the measured concentrations to the expected concentrations using the equation:
Accuracy (Recovery) (%) = {recovered (measured) concentration/ actual
(expected) concentration}*100
For the three different concentrations measured accuracy was in the range of 90-
115%.
Intra- and Inter-day Variability
For Intra-day variability, three different new bumetanide concentrations in serum and urine (0.05, 0.25 and 1 ug/ml) were measured in three replicates within the same day. Intra-day variability was calculated as the relative standard
175 deviations of the three replicates of each concentration and was found less than
2 %.
For inter-day variability, the three different bumetanide concentrations in serum and urine (0.05, 0.25 and 1 ug/ml) were measured in three replicates in two different days. Inter-day variability was calculated as the relative standard deviations of the six replicates of each concentration. The inter-day variability calculated was less than 6 % except for the lowest concentration of 0.05 ug/ml where the inter-day variability was 22 %.
To avoid any confounding effect of the analytical procedures on the results of the analysis owing to the high inter-day variability reported at low concentration, bumetanide samples in urine were analyzed within one day.
6.3.2. Comparison of the Rabbit Response to Immediate and Extended
Release Bumetanide Formulations
6.3.2.1. Urine Output
Table 25 (page 179) presents comparison of the immediate and extended release bumetanide formulations in term of urine output (ml/Kg). Statistical analysis using ANOVA revealed insignificant carry over effect (p-value of 0.7240) or period effect (p-value of 0.3666). Insignificant carryover effect reflects independence of the response to the second treatment on the response to the first treatment. Insignificant period effect infers that the response is independent of the sequence of treatments i.e. response to the immediate release formulation did not change based on whether the formulation was administered first or
176 second to the extended release formulation. Paired t-test and ANOVA were used to compare the responses of immediate and extended release formulations to each other and to control.
For this study, urine was collected from rabbits for at least 48 hr prior to dosing with bumetanide and was used as control. Both immediate and extended release bumetanide formulations significantly increased the urine output within the first day after dosing compared to control (average of 23.0 M 2.8 ml/Kg/day for the control, 66.3 M 3.8 ml/Kg/day after the immediate release formulation with p-value = 0.0001; 57.2 M 4.9 ml/Kg/day after the extended release formulation with p-value = 0.0009).
As seen in Table 25, comparable diuretic effect was produced by both the immediate and extended release formulations within the first day after dosing (p- value = 0.3570). When the urine flow rate was calculated during the first and second 12 hr, extended release formulation maintained constant flow rate
(average of 2.0 M 0.5 ml/hr/Kg during the first 12-hr period compared to 2.5 M 0.5 ml/hr/Kg within the second 12-hr period with p-value = 0.6606). For the immediate release formulation, there was a sharp decrease in urine flow rate within the second 12-hr period (average of 3.9 M 0.2 ml/hr/Kg within the first 12-hr compared to 1.6 M 0.3 ml/hr/Kg during the second 12-hr period with p-value =
0.0115).
The diuretic effect during the second day after dosing with the immediate release formulation was significantly less than control (average of 7.5 M 1.6 ml/Kg compared to 21.9 M 3.3 ml/Kg for control with p-value = 0.0006). For the
177 extended release formulation, comparable diuretic effect to control was observed on the second day (average of 25.7 M 5.4 ml/Kg compared to 21.9 M 3.3 ml/Kg for the daily control with p-value = 0.7891). During the second day after dosing, more urine was produced with the extended release formulation compared to the immediate release formulation (average of 25.7 M 5.4 ml/Kg/day for the extended release formulation vs. 7.5 M 1.6 ml/Kg/day for the immediate release formulation with p-value = 0.0642 reflecting border line significance).
The two different bumetanide formulations had different patterns of urine flow during the 48-hr period of urine collection. For the immediate release formulation, 89.8 % of the total urine output was collected during the first day after dosing and only 10.2 % was collected in the following day. On the other hand, 69.0% of the total urine output was collected within the first day after dosing with the extended release formulation while 31.0 % was collected in the second day. When comparing the overall diuretic effect, both the extended and the immediate release bumetanide formulations produced equivalent diuretic effects (average of 73.8 M 4.3 ml/Kg/day for the immediate release formulation vs. 82.9 M 5.2 ml/Kg/day for the extended release formulation with p-value =
0.4383).
It was concluded that although a different pattern of urine flow was obtained after the immediate and extended release formulations, both formulations had an overall statistically comparable diuretic effect. The extended release formulation tends to maintain a more extended pattern of urine flow than the immediate release formulation.
178 Table 25. Effect of formulation on the urine output (ml/Kg) in rabbits (n=8)
Control Immediate Extended Release Release
0-24 hr Average M standard error 23.0 M 2.8 66.3 M 3.8 57.2 M 4.9 % of the total response 89.8 69.0 p-Value (comparison to control) 0.0001 0.0009 p-value (comparison to ER) 0.3570 24-48 hr Average M standard error 21.9 M 3.3 7.5 M 1.6 25.7 M 5.4 % of the total response 10.1 31.0 p-Value (comparison to control) 0.0006 0.7891 p-value (comparison to ER) 0.0642 Total Response (0-48 hr) Average M standard error 73.8 M 4.3 82.9 M 5.2 p-value (comparison to ER) 0.4383
179 6.3.2.2. Urinary Excretion of Bumetanide
It has been reported that extended release furosemide formulations had marginally lower diuretic/saliuretic effect (19-28 %) than the immediate release formulations albeit the drug fractions absorbed from the extended release formulations were markedly lowered (39-51 %). The decreased furosemide bioavailability was attributed to the limited absorption of the drug (Alvan et al., 1992). Since bumetanide and furosemide share some structural similarity that can reflect on the absorption from the gastrointestinal tract, it becomes of great importance to measure the percent of bumetanide that reached the urinary tubules following the oral administration of immediate and extended release formulations. Furthermore, estimation of the amounts of bumetanide excreted in urine with time can help us better understand how the change in drug delivery rate can affect the pharmacodynamic response to loop diuretics.
Table 26 (page 182) presents the amounts of bumetanide excreted in the urine after the oral administration of immediate and extended release formulations. During the first day, more bumetanide was excreted in the urine following the immediate release formulation (average of 145.8 M 7.5 ug/Kg/day for the immediate release vs. 101.4 M
13.9 ug/Kg/day for the extended release, p-value = 0.0387). During the second day, an opposite effect was noticed with the extended release formulation delivering more bumetanide into the urine (average of 79.1 M 14.9 ug/Kg/day for the extended release vs. 23.0 M 4.5 ug/Kg/day for the immediate release, p-value = 0.0215).
180 Comparing the overall amount of bumetanide excreted in 48 hr, the two formulations were equivalent in delivering the same total amount of bumetanide into the urinary tubules (average of 168.8 M 8.7 ug/Kg/day for the immediate release corresponding to 16.9 % of the total dose/Kg vs. 180.5 M 18.1 ug/Kg/day for the extended release corresponding to 18.1 % of the total dose/Kg with p-value = 0.2045).
Figure 49 (page 183) presents the effect of formulation on the cumulative urine output and bumetanide excretion. As seen in the figure, the cumulative amount of urine excreted after the immediate release formulation ascended faster than that excreted after the extended release formulation within the first 16 hr. Nevertheless, the urine excretion after the extended release formulation continue to increase while that after the immediate release formulation almost reached a plateau. Consequently, at the end of the 48-hr period of urine collection, equivalent amounts of urine were excreted from both formulations. The same pattern was noticed with the cumulative amounts of bumetanide excreted from both formulations. The amount of bumetanide excreted from the immediate release formulation rose faster than that of the extended release formulation within the first 24 hr after dosing. Bumetanide excretion from the extended release formulation continued to rise at almost the same rate throughout the following
24 hr while that excreted from the immediate release formulation rose at slower rate. At the end of the 48-hr urine collection period, equivalent amounts of bumetanide were excreted from both formulations.
181 Table 26. Effect of formulation on the urinary excretion of bumetanide (ugl/Kg) in rabbits
(n=8)
Immediate Extended P-value Release Release 0-24 hr Average M standard error 145.8 M 7.5 101.4 M 13.9 0.0387 % of the total dose 14.6 10.1 24-48 hr Average M standard error 23.0 M 4.5 79.1 M 14.9 0.0215 % of the total dose 2.3 7.9 0-48 hr Average M standard error 168.8 M 8.7 180.5 M 7.5 0.2045 % of the total dose 16.9 18.1
182 200
180
160 ) g 140 ug/K
ion ( 120
100
anide excret 80
60 Cumulative urine output (ml/Kg) and bumet 40
20
0 0 4 8 12 16 20 24 28 32 36 40 44 48 52 Time (hr)
Figure 49. Effect of formulation on the cumulative urine output (ml/Kg) and bumetanide excretion (ug/Kg) in rabbits (n=8). (__▲__) urine output after the immediate release formulation; (__∆__) urine output after the extended release formulation; (---■---)
bumetanide (---ڤ---) ;bumetanide excreted from the immediate release formulation excreted from the extended release formulation.
183 The results suggested that extended release bumetanide formulation maintained the delivery of bumetanide to the systemic circulation and subsequently the urinary tubules within a rate that slowly declined over the first 48 hr after drug administration.
The immediate release formulation, as expected, delivered a relatively sudden push of bumetanide into the systemic circulation and urinary tubules followed by a sharp decline in bumetanide excretion.
Extended release bumetanide formulation had better diuretic efficiency within the first day after dosing. The extended release formulation had comparable diuretic effect to that of the immediate release formulation despite less bumetanide was excreted in the urine during the first day after dosing. The diuretic efficiency of the extended release formulation decreased significantly within the second day after dosing. Bumetanide was excreted into the urinary tubules from the extended release formulation at a relatively high level during the second day however, no improvement in diuretic response compared to control could be noticed. For the immediate release formulation, less urine output was obtained during the second day compared to control. The results can be explained by considering that during the first day after drug administration water and electrolyte loss can trigger compensatory mechanisms including renin-angiotensin- aldosterone and sympatho-adrenal systems system with a resultant increase in sodium and water reabsorption from the distal and collecting tubules and a decrease in the response.
The activation of compensatory mechanism following single dose administration of loop diuretics has been reported in the literature and described as acute tolerance
184 (Hammarlund and Paalzow, 1985; hammarlund et al., 1985; Li et al., 1986; Cook and
Smith, 1987). When there is inadequate replacement of urinary loss, body mechanisms are rapidly brought into play within a single dose of diuretic to conserve body fluids.
Activation of the renin-angiotensin-aldosterone system and the sympathetic nervous system has been proposed to play a role in the mediation of the development of acute tolerance to loop diuretics. However, the relative contribution of these systems to the development of acute tolerance is not clear (Wakelkamp et al, 1996).
6.3.2.3. Sodium and Potassium Excretion
Tables 27 and 28 (pages 188 and 189) present the urinary excretion of sodium and potassium following both immediate and extended release formulations. The amounts of sodium excreted within the first day after dosing with both formulations were significantly more than control (average of 2.1 M 0.2 mmol/Kg/day for control; 6.4 M 0.2 mmol/Kg/day for the immediate release formulation with p-value = 0.0001; 5.4 M 0.4 mmol/Kg/day for the extended release formulation with p-value = 0.0009). Conversely, the amounts of potassium excreted after dosing with both formulations did not significantly differ from control (average 4.9 M 0.8 mmol/Kg/day for control; 3.6 M 0.2 mmol/Kg/day for the immediate release formulation with p-value of 0.1991; 3.7 M 0.5 mmol/Kg/day for the extended release formulation with p-value of 0.3213).
On the second day after dosing, immediate release formulation had lower sodium but comparable potassium excretion to control (average of 0.3 M 0.05 mmol/Kg/day for sodium compared to 2.3 M 0.2 mmol/Kg/day for control with p-value = 0.0006; average
185 of 3.1 M 0.7 mmol/Kg/day for potassium compared to 5.0 M 0.4 mmol/Kg/day for control with p-value = 0.1628). For the extended release formulation, less sodium was excreted during the second day compared to control (average of 0.8 M 0.2 mmol/Kg/day compared to 2.3 M 0.2 mmol/Kg/day for control with p-value = 0.0070). However comparable amounts of potassium were excreted compared to control (average of 5.8 M
1.4 mmol/Kg/day compared to 5.0 M 0.4 mmol/Kg/day for control with p-value = 0.6918).
Immediate and extended release bumetanide formulations did not differ statistically in the amount of sodium and potassium excreted during the first two days after oral administration (for sodium excretion, p-value = 0.1064 for the first day and 0.1776 for the second day, for potassium excretion, p-value = 0.8316 for the first day and 0.1065 for the second day).
The results are in agreement with other reports in the literature where intravenous bumetanide induced significant increase in the amount of sodium excreted while exerted no effect on the amount of potassium excreted in rabbits (Ryoo et al.,
1993; Yoon et al., 1995). The insensitivity of potassium excretion to bumetanide was attributed to the constant rate of potassium excretion in the distal tubules (Ryoo et al.,
1993). In contrast to rabbits, bumetanide has been shown to increase both sodium and potassium excretion in humans. However, the increase in sodium excretion is higher than that of potassium (Marcantonio et al.; 1982). Sodium rather than potassium excretion has been reported to parallel water diuresis in humans particularly during the short period of time after dosing (Marcantonio et al., 1982)
186 The results suggested that extended release bumetanide formulation had better saliuretic efficiencies during the first day since comparable amounts of sodium and potassium were excreted despite less bumetanide was available in the urinary tubules.
The results are in agreement with that reported in the literature comparing immediate to extended release furosemide formulations (Beermann, 1982). Similar saliuretic effects were obtained from both formulations despite furosemide was less available to the urinary tubules from extended release formulation relative to the immediate release formulation (Beermann, 1982).
During the second day after dosing, more bumetanide was available to the urinary tubules from the extended release formulation however, it induced comparable sodium and potassium excretion to that obtained from the immediate release formulation. Again, the results can be explained by considering the activation of renin- angiotensin-aldosterone system during the second day with a subsequent increase in sodium and water reabosrption.
In conclusion, extended release bumetanide formulation managed successfully in presenting the drug to the urinary tubules in an extended manner. The extended delivery of bumetanide improved its diuretic/saliuretic efficiencies during the first day of dosing. The diuretic/saliuretic efficiencies of the extended release formulation decreased significantly in the second day after dosing due to the activation of compensatory mechanisms. Overall, bumetanide was equally available to the urinary tubules from both the extended and immediate release formulations with equivalent diuretic and saliuretic effects.
187 Table 27. Effect of formulation on the urinary excretion of sodium (mmol/Kg) in rabbits
(n=8)
Control Immediate Extended Release Release 0-24 hr Average M Standard error 2.1 M 0.2 6.4 M 0.2 5.4 M 0.4 % of the total Response 95.5 87.1 p-Value (comparison to control) 0.0001 0.0009 p-value (comparison to ER) 0.1064 24-48 hr Average M standard error 2.3 M 0.2 0.3 M 0.05 0.8 M 0.2 % of the total response 4.5 12.9 p-value (comparison to control) 0.0006 0.0070 p-value (comparison to ER) 0.1776 Total Response (0-48 hr) Average M standard error 6.7 M 0.3 6.2 M 0.7 p-value (comparison to ER) 0.8012
188 Table 28. Effect of formulation on the urinary excretion of potassium (mmol/Kg) in rabbits (n=8)
Control Immediate Extended Release Release 0-24 hr Average M standard error 4.9 M 0.8 3.6 M 0.2 3.7 M 0.5 % of the total response 53.7 38.1 p-value (comparison to control) 0.1991 0.3213 p-value (comparison to ER) 0.8316 24-48 hr Average M standard error 5.0 M 0.4 3.1 M 0.7 5.8 M 1.4 % of the total response 46.3 61.9 p-Value (comparison to control) 0.1628 0.6918 p-value (comparison to ER) 0.1065 Total Response (0-48 hr) Average M standard error 6.7 M 0.8 9.5 M 1.7 p-value (comparison to ER) 0.1358
189 6.3.2.4. Urine Osmolarity
The urine osmolarity was determined using Advanced Micro Osmometer and reported as mOsmol/Kg water. The average osmolarity of control urine samples collected for 48 hr prior to dosing with bumetanide formulations was 1131 mOsmol/Kg water. The urine osmolarity decreased significantly during the first day after dosing with both immediate and extended release bumetanide formulations (average of 483 mOsmol/Kg water after the immediate release formulation with p-value =0.0001 compared to control; average of 445 mOsmol/Kg water after the extended release formulation with p-value =0.0001 compared to control). The two formulations were equivalents in reducing the urine osmolarity (p-value = 0.4555). The results suggested that bumetanide induced more diuresis rather than saliuresis during the first day.
On the second day after dosing with the immediate release bumetanide formulation, hyper-osmotic urine was collected (average of 1606 mOsmol/Kg water compared to 1131 mOsmol/Kg water for the control with p-value = 0.0608 reflecting border line significance). The extended release formulation, on the other hand, maintained iso-osmotic urine excretion compared to control (average of 1065 mOsmol/Kg water compared to 1131 mOsmol/Kg water for the control with p-value =
0.6229).
The results are in agreement with our previous discussion that compensatory mechanisms triggered by water and electrolyte loss during the first day lead to more water reabsorption from the urinary tubules during the second day. Subsequently, the urine collected during the second day after dosing with the immediate release
190 formulation was hyper-osmotic compared to normal urine. For the extended release formulation, bumetanide continued to be excreted during the second day at a level that maintained water excretion at normal level despite the activation of compensatory mechanisms and therefore, the urine collected during the second day was iso-osmotic with normal urine collected prior to dosing with bumetanide formulation.
191 7. Summary and Conclusions
In the first part (part I), bumetanide extended release peroral formulations were developed and tested in vitro. Fluid bed technology including drug layering on sugar pellets followed by polymeric film coating using methacrylate ester copolymer (Eudragit
RS) were utilized. The effects of Eudragit RS load and triethyl citrate plasticizer level on bumetanide release from the coated pellets were investigated. Central composite statistical design was implemented to allow statistically sound analysis using the minimal number of trials. Both Eudragit RS load and triethyl citrate level had significant impairment effect on bumetanide release from the coated pellets in water.
Increasing Eudragit RS load increased the coat thickness and reduced the rate at which bumetanide was released. Triethyl citrate level can affect bumetanide release from two opposite perspectives. Triethyl citrate is a water-soluble plasticizer that leaches out form the film coat upon contact with aqueous media creating channels through which drug release can be accelerated. On the other hand, plasticizer increases the polymer flexibility by lowering the glass transition temperature and hence improves the film formation process with a subsequent decrease in drug release. The net effect of plasticizer depends on the size of the dosage forms to be coated. For the small size pellets used in this study, the effect of plasticizer on film formation is more predominant than its ability to create channels during dissolution and therefore, a decrease in drug release was noticed when the plasticizer level was increased.
Single point optimization techniques have been reported in the pharmaceutical literature to optimize extended release formulations. However, the technique is not
192 efficient in optimizing the release process over an extended period of time since completely different release profiles can share a common single release time point.
Multiple response optimization technique offers the advantages of optimizing the release process at different release time points. In this study, a novel multiple response optimization technique based on superimposing two-dimension contour diagrams was successfully developed and used to optimize bumetanide release from coated pellets.
The percent of drug released after 1,4, and 8 hr were simultaneously optimized by careful selection of Eudragit RS load and triethyl citrate level using the superimposed contour diagrams. The developed optimized formulation had a release profile that was similar to the predetermined target profile according the Food and Drug Administration favored similarity factor comparison algorithm.
Bumetanide release from the coated pellets was unstable after storage at room temperature for three months. An increase in drug release was seen at the end of the storage period. Drug release instability from coated pellets has been reported in the literature with researchers reporting both decrease and increase in drug release after storage. Coated pellets are often cured to complete the film formation process and avoid any drug release instability after storage. Nevertheless, different drugs respond in different fashion to curing conditions. Therefore, a systematic study was planned to investigate the effect of curing conditions and plasticization on the release of bumetanide from the coated pellets (Part II).
Plasticizer level was found to have a dramatic effect on the response of drug release to curing conditions. Lower plasticizer level of 10 % was not enough to complete
193 the film formation process even after curing for 168 hr at 60 LC. Increasing the plasticizer level to 15 and 20 % facilitated the polymeric particle coalescence process and allowed for complete film formation within shorter curing time and at lower curing temperature. Drug release from the coated pellets initially decreased with increasing curing time. Prolonged curing lead to faster rather than slower release profiles. Drug release from coated pellets plasticized with 15 % triethyl citrate followed zero order kinetics while the drug release from the coated pellets plasticized with 20 % triethyl citrate followed Higuchi’s model. The results suggested that bumetanide migration into the film coat during curing or storage is the cause of the increase in drug release.
To investigate more into bumetanide migration and release instability after storage, bumetanide pellets were seal coated with Hydroxypropyl methyl cellulose prior to coating with Eudragit RS. The use of hydrophilic seal coat caused the drug release to follow zero order kinetics reflecting complete separation of the drug layer from the polymeric film coat. Coated pellets with the HPMC seal coat maintained stable release profile after storage for 3 months at room temperature. The findings confirmed the hypothesis that bumetanide migration into the film coat during curing and storage is the main cause of drug release instability.
Bumetanide extended release formulations were tested in laboratory animals in comparison to immediate release formulations (Part III). The extended release formulation managed to maintain an extended in vivo delivery of bumetanide to the urinary tubules for 48 hr. Conversely, the immediate release formulation delivered an initial push of bumetanide followed by sharp decline in bumetanide delivery. Overall,
194 equivalent amounts of bumetanide were made available to the urinary tubules within 48 hr after dosing by both formulations yet at different rates.
Comparable diuretic and saliuretic effects were noticed with both formulations within the first day after dosing albeit less bumetanide was available in the urinary tubules from the extended release formulation. The slow delivery of bumetanide from the extended release formulation improved its diuretic and saliuretic efficiencies (effect per unit stimulus). On the second day, more bumetanide was excreted from the extended release formulation however there was no increase in the urine volume compared to control. The decrease in response to bumetanide within the second day is thought to be attributed to the activation of compensatory mechanism including rennin- angiotensin-aldosterone and sympathetic systems in response to water and electrolyte depletion in the first day. It was concluded that extended release bumetanide formulation had similar diuretic and saliuretic effects but better efficiencies than immediate release formulation in laboratory rabbits. While providing comparable diuretic and saliuretic effects, extended release formulation can offer the advantage of avoiding the initial, unpleasant and intense diuretic effect experienced with the immediate release formulations. The data provide sufficient basis to warrant further investigation of the extended release bumetanide formulation in humans.
195 REFERENCES
Alvan, G., Helleday, L., Lindholm, A. Sanz, E., Villen, T., 1990. Diuretic effect and diuretic efficiency after intravenous dosage of frusemide. Br. J. clin. Pharmacol. 29, 215-219
Alvan, G, Paintaud, G., Eckernas, S.A., Grahnen, A., 1992. Discrepancy between bioavailability as estimated from urinary recovery of frusemide and total diuretic effect. Br. J. clin. Pharmacol. 34, 47-52.
Amighi, K, Moes, A.J., 1996. Influence of plasticizer concentration and storage conditions on the drug release rate from Eudragit RS 30 D film coated sustained release theophylline tablets. Eur. J. Phar. Biopharm. 42, 29-35
Aoyagi, N., 1986. Comparative studies of griseofulvin bioavilability among man and animals. Thesis, Kyoto University. In: Kararli, T., 1995. Comparison of the gastrointestinal anatomy, physiology and biochemistry of humans and commonly used laboratory animals. Biopharm. Drug Disp. 16, 351-380.
Bayomi, M. A., Al-Angary, A. A., Al-Meshal, M. A., Al-Dardiri, M. M., 1998. In vivo evaluation of arteether liposomes. Int. J. Pharm. 175, 1-7.
Beerman, B., 1982. Kinetics and dynamics of furosemide and slow-releasing furosemide. Clin. Pharmacol. Ther. 32, 585-59.
Beermann, A., 1984. Aspects of pharmacokinetics of some diuretics. Acta Pharmacol Toxicol. 54(Suppl.), 17-29.
196 Benedict, G., Steinijans, V.W., Dietrich, R., 1988. Galenical development of a new sustained release theophylline pellet formulation for once-daily administration. Drug Res. 6, 725-730
Bodea, A., Leucuta, S.E., 1997. Optimization of propranolol hydrochloride sustained release pellets using a factorial design. Int. J. Pharm. 154, 49-57.
Bodmeier, R., Paeratakul, O., 1990. Theophylline tablets coated with aqueous latexes containing dispersed pore formers. J. Pharm. Sci. 79, 925-928.
Bodmeier, R., Paeratakul, O., 1991. Process and formulation variables affecting the drug release from chlorpheniramine maleate-loaded beads coated with commercial and self-prepared adueous ethyl cellulose pseudolatexes. Int. J. Pharm. 70, 59-68.
Bodmeier, R., Paeratakul, O., 1993. Dry and wet strengths of polymeric films prepared from an aqueous colloidal polymer dispersion, Eudragit RS30D. Int. J. Pharm. 96, 129- 138.
Bodmeier, R., Paeratakul, O., 1994. The effect of curing on drug release and morphological properties of ethylcellulose pseudolatex-coated beads. Drug Dev. Ind. Pharm. 20, 1517-1533.
Bodmeier, R., 1997. Tableting of coated pellets, review. Eur. J. Pharm. Biopharm. 43, 1-8.
Brater, D.C., Chennavasin, P., Seiwell, R., 1980. Furosemide in patients with heart failure: shift in dose-response curves. Clin. Pharmacol. Ther. 28, 182-186.
197 Brater, D. C., Chennavasin, P., Day, B., Burdette, A., Anderson, S., 1983. Bumetanide and furosemide. Clin. Pharmacol. Therap. 34, 207-213.
Brater, D.C., 1998. Diuretic therapy. New Eng. J. Med. 339, 387-395.
Chen, Y., McCall, T.W., Baichwal, A.R., Meyer, M.C., 1999. Application of artificial neural networks and pharmacokinetic simulation in the design of controlled release dosage forms. J. Contr. Rel. 59, 33-41.
Cook, J.A., Smith, D.E., 1987. Development of acute tolerance to bumetanide: Bolus injection studies. Pharm. Res. 4, 379-384.
Daniel, W.W., 1999. Biostatistics: A Foundation for Analysis in the Health Science, 7th Edition. John Wiley & Sons, Inc., New York, pp. 268-270.
Desai, S.J., Simonelli, A.P., Higuchi, W.I., 1965. Investigation of factors influencing release of solid drug dispersed in inert matrices, J. Pharm. Sci. 54, 1459-1464.
Food and Drug Administration (FDA), 1997a. Guidance for Industry, Extended Release Oral Dosage Forms: Development, Evaluation, and Application of In Vitro/In Vivo Correlations.
Food and Drug Administration (FDA), 1997b. Guidance for Industry, Dissolution Testing of Immediate Release Solid Oral Dosage Forms.
Frohoff-Hulsnmann, M.A., Schmitz, A., Lippold, B.C., 1999. Aqueous ethyl cellulose dispersions containing plasticizers of different water solubility and hydroxypropyl methyl cellulose as coating material for diffusion pellets. I: Drug release rates from coated pellets. Int. J. Pharm. 177, 69-82.
198 Fuller, R., Hoppel, C., Ingalls, S.T., 1981. Furosemide kinetics in patients with hepatic cirrhosis with ascites. Clin. Pharmacol Ther. 30, 461-467.
Ghebre-Sellasie, I., Gordon, R.H., Nesbitt, R.U., Fawzi, M.B., 1987. Evaluation of acrylic based modified release film coatings. Int. J. Pharm. 37, 211-218.
Ghebre-Sellasie, I., 1989. Pellets, a general overview. In: Ghebre-Sellasie, I. (Ed), Pharmaceutical Pelletization Technology. Mercel Dekker Inc., New York , pp.1-15.
Gilligan, C.A., LiWan, P.A., 1991. Factors affecting drug release from a pellet system coated with an aqueous colloidal dispersion. Int. J. Pharm. 73, 51-68.
Gohel, M.C., Amin, A.F., 1998. Formulation optimization of controlled release diclofenac sodium microspheres using factorial design. J. Contr. Rel. 51, 115-121.
Goodhart, F.W., Harris, M.R., Murphy, K.S., Nesbitt, R.U., 1984. An evaluation of aqueous film-forming dispersions for controlled release. Pharm. Technol. 4, 64-71.
Guirgis, H., Broegmann, B., Sakr, A., 2001. Hot melt technology: effect of manufacturing method on tablet characteristics. Pharm. Ind., 63, 499-507.
Guma, N.C., Kale, K., Morris, K.R., 1997. Investigation of film curing stages by dielectric analysis and physical characterization. J. Pharm. Sci. 86, 329-334.
Gunder, W., Lippold, B.H., Lippold, B.C., 1995. Release of drugs from ethyl cellulose microcapsules (diffusion pellets) with pore formers and pore fusion. Eur. J. Pharm. Sci. 3, 203-214.
Habib, W., Sakr, A., 1999. Development and human in vivo evaluation of a colonic drug delivery system”, Pharm. Ind. 61, 1145-1149.
199 Habucky, K. and Tse, F. L., 1998. Disposition of SDZ-ENA-713, an etylcholinesterase inhibitor, in the rabbit. Biopharm. Drug Disp. 19, 285-290.
Hamed, E., Sakr, A., 2001. Application of multiple response optimization technique to extended release formulations design. J. Contr. Rel. 73, 329-338.
Hammarlund, M, Odlind, B., Paalzow, L.K., 1985. Acute tolerance development to the diuretic effect of furosemide in human: pharmacokinetic-pharmacodynamic modeling. J. Pharmacol. Exp. Ther. 233, 477-453.
Hammarlund, M, Paalzow, L.K., 1985. Acute tolerance development to the diuretic effect of furosemide in the rat. Biopharm. Drug Dsipos. 6, 9-21
Harris, M.R., Ghebre-Sellasie, I., 1997. Aqeous polymeric coating for modified -release pellets. In: McGinity, J.W. (Ed.), Aqueous Polymeric Coatings for Pharmaceutical Dosage Forms. Mercel Dekker Inc., New York, pp.63-79.
Hicks, D.C., Freese, H.L., 1989. Extrusion and spheronization equipment. In: Ghebre- Sellasie, I. (Ed.), Pharmaceutical Pelletization Technology. Mercel Dekker Inc., New York , pp.71-101.
Holazo, A. A., Colburn, W. A., Gustafson, J. H., Young, R. L., Parsonnet, M., 1984. Pharmacokinetics of bumetanide following intravenous, intramuscular and oral administration to normal subjects. J. Pharm. Sci. 73,1108-1113.
Holford, N.H.G., Sheiner, L.B., 1981. Understanding the dose-effect relationship. clinical application of pharmacokinetic- pharmacodynamic models. Clin. Pharmacokinet. 6, 429- 453.
200 Husson, I., Leclerc, B., Spenlehauer, G., Veillard, M, Couarraze, G, 1991. Modelling of drug release from coated pellets with an insoluble polymeric membrane. J. Contr. Rel. 17, 163-174.
Hutchings, D., 1993. A Comparative Evaluation of Plasticizers for an Ethylcellulose Latex Coating System: Aspects of Formulation and Process Development. Ph.D. Dissertation, College of Pharmacy, University of Cincinnati.
Hutchings, D., Kuzmak, B, Sakr, A., 1994. Processing considerations for an EC latex coating system: Influence of curing time and temperature. Pharm. Res. 11, 1474-1478.
Hutchings, D., Sakr, A, 1994. Influence of pH and plasticizers on drug release from ethylcellulose pseudolatex coated pellets. J. Pharm. Sci. 83, 1386-1390.
Ives, H.E., 1998. Diuretic Agents. In: Katzung, B.G. (Ed.), Basic & Clinical Pharmacology, 7th Edition. Lange Medical Books/McGraw-Hill, New York, pp.242-261.
Kallstrand, G, Ekman, B., 1983. Membrane-coated tablets: a system for the controlled release of drugs. J. Pharm. Sci. 72, 772-775.
Kaojarene, S., Day, B., Brater, D.C., 1982. The time course of delivery of frusemide into urine: an independent determination of overall response. Kidney Int. 22, 69-74.
Kararli, T., 1995. Comparison of the gastrointestinal anatomy, physiology and biochemistry of humans and commonly used laboratory animals. Biopharm. Drug Disp. 16, 351-380.
Keshikawa, T., Nakagami, H., 1994. Film formation with coating systems of aqueous suspensions and latex dispersions of ethylcellulose. Chem. Pharm. Bull. 42, 656-662.
201 Koene, R. A., 1983. High-dosage bumetanide in patients with chronic renal failure and severe nephrotic syndrome. Nether. J. Med. 26, 262-265. In: Ward, A., Heel, R. C., 1984. Bumetanide: review of its pharmacodynamic and pharmacokinetic properties and therapeutic use. Drugs 28, 426-464.
Kolis, S.J., Williams, T.H., Schwartz, M.A. Identification of the urinary metabolites of 14C-bumetanide in the rat and their excretion by rats and dogs. Drug Metab. Dispos. 4, 169-176.
Laethem, M. E., Belpaire, F. M., Wijnant, P., Bogaert, M. G., 1995. Effect of probenecid on the enantioselective pharmacokinetics of oxprenolol and its glucuronides in the rabbit. Pharm. Res. 12, 1964-1967.
Lau, H.S.H., Hyneck, M.L., Berardi, R.R., Swatz, R.D., Smith, D.E, 1986. Kinetics, dynamics, and bioavailability of bumetanide in healthy subjects and patients with chronic renal failure. Clin. Pharmacol .Ther. 39, 635-645.
Lee, S.H., Lee, M.G. Kim, N.D., 1994. Pharmacokinetics and pharmacodynamics after intravenous and oral administration to rats: absorption from various GI segments. J. Pharm. Biopharm. 22, 1-17.
Lehmann, K.O.R., 1997. Chemistry and Application Properties of Polymethacrylate Coating Systems. In: McGinity, J.W. (Ed.), Aqueous Polymeric Coatings for Pharmaceutical Dosage Forms, Second Edition. Mercel Dekker, Inc., New York, pp. 1- 77.
Li, T., lee, M.G., Chiou, W.L., 1986. Effect of rate and composition of fluid replacement on the pharmacokinetics and pharmacodynamics of intravenous furosemide. J. Pharmacokinet. Biopharm. 14, 495-509.
202 Lippold, B.H., Sutter, B.K., Lippold, B.C., 1989. Parameters controlling drug release from pellets coated with aqueous ethyl cellulose dispersion. Int. J. Pharm. 54, 15-25.
Lippold, B.C., Gunder, W., Lippold, B.H., 1999. Drug release from diffusion pellets coated with aqueous ethyl cellulose dispersion aquacoat ECD-30 and 20 % dibutyl sebacate as Plasticizer: partition mechanism and pore diffusion. Eur. J. Pharm. Biopharm. 47, 27-32.
Lipps, D.M., 1991. Application of Response Surface Methodology to Wet Granulation in Tablet Technology. Ph.D. Dissertation, University of Cincinnati College of Pharmacy.
Lipps, D.M., Sakr, A., 1994. Characterization of wet granulation process parameters using response surface methodology.1. Top spray fluidized bed.J. Pharm. Sci. 83, 937- 946.
Magnussen, M.P., Eilertsen, E., 1974. .Species differences in the diuretic activity and metabolism of bumetanide. Naunyn. Schmiedebergs. Arch. Pharmacol. 282 (Suppl.), R61. In: Schwartz, M. A., 1981. Metabolism of bumetanide. J. Clin. Pharmacol. 21, 555-563.
McCrindle, J.L., Likam, T.C., Barron, WA.,W.,Prescott, L.F., 1996. Effect of food on the absorption of frusemide and bumetanide in man. Br. J. Clin. Pharmacol. 42, 743-746.
McPhillips, A., Uraizee, S., Ritschel, W., Sakr, A., 1997. Evaluation of fluid-bed applied acrylic polymers for the targeted peroral delivery of insulin. STP Pharm. Sci, 7, 476-482.
203 Marcantonio, L. A., Auld, W. H., Skellern, G. G., Howes, C. A., Murdoch, W. R., Purohit, R., 1982. The pharmacokinetics and pharmacodynamics of bumetanide in normal subjects. J. Pharmacokinet. Biopharm.10, 393-409.
Martin, A., 1993. Physical Pharmacy: Physical Chemical Principles in the Pharmaceutical Sciences. 4th Edition. Lea & Fabiger, Philadelphia, p.532.
Menon, A., Ritschel, W.A., Sakr, A., 1994. Development and evaluation of a momolithic floating dosage form for furosemide. J. Pharm Sci. 83, 239-245.
Miyazaki, S., Aoyama, H., Kawasaki, N., Kubo, W., Attwood, D., 1999. In situ-gelling gellan formulations as vehicles for oral drug delivery. J. Contr. Rel. 60, 287-295.
Olsen, K.W., 1989. Fluid Bed Equipment. In: Swarbrick, J. (Ed.), Pharmaceutical Pelletization Technology. Mercel Dekker Inc., New York, pp.39-71
Porter, S.C., Bruno, C.H., 1990. Coating of pharmaceutical dosage forms. In: Lieberman, H.A., Lachman, L., Schwartz, J.B. (Eds.), Pharmaceutical Dosage Forms, Tablets, Vol. 3, Second Edition. Marcel Dekker Inc., New York , pp.77-158.
Rangaiah, K.V., Chattaraj, S.C., Das, S.K., 1997. Effect of solvents, temperature, and plasticizer on film coating of tablets. Drug Dev. Ind. Pharm. 23, 419-423.
Ritschel, W. A., 1989. Biopharmaceutic and pharmacokinetic aspects in the design of controlled release peroral drug delivery systems. Drug Dev. Ind. Pharm. 15, 1073-1103.
Rudy, D.W., James, R.V., Greene, P.K., Esparza, A., Brater, D.C., 1991. Loop diuretics for chronic renal insufficiency: A continuous infusion is more efficacious than bolus therapy. Ann. Int. Med. 115, 360-366.
204 Ryoo, S.H., Lee, M.G., Lee, M.H., 1993. Effect of intravenous infusion time on the pharmacokinetics and pharmacodynamics of the same total dose of bumetanide. Biopharm. Drug Disp. 14, 245-255.
Santus, G., Baker, R.W., 1995. Osmotic drug delivery: a review of the patent literature. J. Contr. Rel. 35, 1-21.
Sastry, S.V., Khan, M.A., 1998. Aqueous-based polymeric dispersion: Face-centered cubic design for the development of atenolol gastrointestinal therapeutic system. Pharm. Dev. Tech. 3, 423-432.
Schmidt, P.C., Niemann, F., 1993. MiniWiD-Coater. Part 3. Effect of application temperature on the dissolution profile of sustained-release theophylline pellets coated with Eudragit RS-30D. Drug Dev. Ind. Pharm. 19, 1603-1612.
Schulze, P., Kleinebudde, P., 1997. A new multiparticulate delayed release system. Part I: Dissolution properties and release mechanism. J. Contr. Rel. 47, 181-189.
Schwartz, M. A., 1981. Metabolism of bumetanide. J. Clin. Pharmacol. 21, 555-563.
Schwartz, S., Brater, D.C., Pound, D., Greene, P.K., Kramer, W.G., Rudy, D., 1993. Bioavailability, pharmacokinetics and Pharmacodynamics of torsemide in patients with cirrhosis. Clin. Pharmacol. Ther. 54, 90-97.
Siepmann, J, Lecomte, F., Bodmeier, R., 1999. Diffusion-controlled drug delivery systems: Calculation of the required composition to achieve desired release profiles. J. Cont. Rel. 60, 379-389.
205 Steuernagel, C.R., 1997. Latex emulsion for controlled drug delivery. In: McGinity, J.W. (Ed.), Aqueous polymeric Coatings for Pharmaceutical Dosage Forms. Mercel Dekker Inc., New York, pp.1-61.
Stone, W.J., Bennet, W.M., Cutler, R.E., 1981. Long term bumetanide treatment of patient with edema due to renal disease. cooperative studies. J. Clin. Pharmacol. 21, 587-590.
Suckow, M.A., Douglas, F.A., 1997. The Laboratory Rabbit. CRC Press, New York, p. 20.
Takahara, J., Takayama, K., Nagai, T., 1997. Mutliobjective simultaneous optimization technique based on an artificial neural network in sustained release formulations. J. Contr. Rel. 49, 11-20.
Takayama, K., Morva, A., Fujikawa, M., Hattori, Y. , Obata, Y., Nagai, T., 2000. Formula optimization for theophylline controlled-release tablets based on artificial neural network, J. Contr. Rel. 68, 175-186.
Tata, H.G., Venkataramanan, R., Sahota, S.K., Florey, K. 1993. Bumetanide. In: Brittain, H.G. (Ed.), Analytical Profiles of Drug Substances and Excipients, Vol 22. Academic Press, New York, pp. 107-144.
The United States Pharmacopeia (USP XXV), 2002. Uniformity of Dosage Units. United States Pharmacopeial Convention, Inc., Rockville, MD, pp.2000-2002. Theeuwes, F., 1975. Elementary osmotic pump. J. Pharm. Sci. 64, 1987-1991.
Tirkkonen, S., Paronen, P., 1992. Enhancement of drug release from ethylcellulose microcapsules using solid sodium chloride in the wall. Int. J. Pharm. 88, 39-51.
206 Turkoglu, M., Sakr, A., 1992. Mathematical modeling and optimization of a rotary fluidized-bed coating process. Int. J. Pharm. 88, 75-87.
Turkoglu, M., He, M, Sakr, A., 1995. Evaluation of rotary fluidized bed as a wet granulation equipment. Eur. J. Pharm. Biopharm. 41, 388-394.
Turkoglu, M., Aydin, I., Murray, M., Sakr, A., 1999. Modeling of a roller-compaction process using neural networks and genetic algorithms. Eur. J. Pharm. Biopharm. 48, 239-245.
Vanderhoff, J.W., El-Asser, M., U.S. Patent 4,177,177, 1979.
Vargo, D.L., Kramer, W.G., Black, P.K., Smith, W.B., Serpas, T., Brater, D.C., 1995. Bioavailability, pharmacokinetics and pharmacodynamics of torsemide and furosemide in patients with congestive heart failure. Clin Pharmacol. Ther. 57, 601-609.
Venkatraman, S., Davar, A., Chester, A., Kleiner, L., 2000. An overview of controlled release systems. In: Wise, D.L. (Ed.), Handbook of Pharmaceutical Controlled Release Technology. Marcel Dekker Inc., New York, pp.431-465.
Voelker, I.R., Cartwright-Brown, D., Andreasen S., 1987. Comparison of loop diuretics in patients with chronic renal insufficiency. Kidney Int. 32, 572-578.
Wakelkamp, M., Alvan, G., Gabrielsson, J., Paintaud, G., 1996. Pharmacodynamic modeling of furosemide tolerance after multiple intravenous administration. Clin. Pharmacol. Ther. 60, 75-88.
Watano, S., Wada, I., Miyanami,K., 1995. Modeling and simulation of drug release from granules coated with an aqueous based system of acrylic copolymer. Chem. Pharm. Bull. 43, 877-880.
207 Williams III, R.O., Liu, J., 2000. Influence of processing and curing conditions on beads coated with an aqueous dispersion of cellulose acetate phthalate, Eur. J. Pharm. Biopharm. 49, 243-252.
Yagi, N., Kiuchi, T., Satoh, H., Ishikawa, Y., Takada, M., Sekikawa, H., 1997. Preparation of sustained release granules for bumetanide. Chem. Pharm. Bull. 45, 918- 922.
Yagi, N., Kiuchi, T., Satoh, H., Ishikawa, Y., Takada, M., Sekikawa, H., 1999. Bioavailability and diuretic effect after administration of retarded capsules of bumetanide in human subjects. Biol. Pharm. Bull. 22, 275-280.
Yang, S.T., Ghebre-Sellasie, I., 1990. The effect of product bed temperature on the microstructure of aqueous based controlled release coating. Int. J. Pharm. 60, 109-124.
Yoon, W.Y., Lee, S.H., Lee, M.G., 1995. Effect of the rate and composition of fluid replacement on the pharmacokinetics and pharmacodynamics of intravenous bumetanide. J. Pharm. Sci. 84, 236-242.
Yuen, K.H., Deshmukh, A.A., Newton, J.M., 1993. Development and in-vitro evaluation of a multi-particulate sustained release theophylline formulation. Drug Dev. Ind. Pharm. 19, 855-874.
208