Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Protein binding displacement interactions between propafenone, 5-hydroxypropafenone and other antiarrhythmic… Wang, Jing 1990

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1991_A6_7 W38.pdf [ 5.43MB ]
Metadata
JSON: 831-1.0098731.json
JSON-LD: 831-1.0098731-ld.json
RDF/XML (Pretty): 831-1.0098731-rdf.xml
RDF/JSON: 831-1.0098731-rdf.json
Turtle: 831-1.0098731-turtle.txt
N-Triples: 831-1.0098731-rdf-ntriples.txt
Original Record: 831-1.0098731-source.json
Full Text
831-1.0098731-fulltext.txt
Citation
831-1.0098731.ris

Full Text

P R O T E I N B I N D I N G D I S P L A C E M E N T I N T E R A C T I O N S B E T W E E N P R O P A F E N O N E , 5 - H Y D R O X Y P R O P A F E N O N E A N D O T H E R A N T I A R R H Y T H M I C D R U G S ; H P L C A N A L Y S I S O F I N D O C Y A N I N E G R E E N by JING WANG B.Sc. (Pharm), B e i j i n g M e d i c a l U n i v e r s i t y , 1987 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENT FOR THE DEGREE OF MASTER OF SCIENCE i n THE FACULTY OF GRADUATE STUDIES ( F a c u l t y o f P h a r m a c e u t i c a l S c i e n c e D i v i s i o n o f P h a r m a c e u t i c s and B i o p h a r m a c e u t i c s ) We a c c e p t t h i s t h e s i s as c o n f o r m i n g t o t h e r e q u i r e d s t a n d a r d THE UNIVERSITY OF BRITISH COLUMBIA MAY, 1990 (5) C o p y r i g h t by J i n g Wang, 1990 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada DE-6 (2/88) i i ABSTRACT Propafenone (PF) is a class I antiarrhythmic agent used to treat ventricular and superventricular tachyarrhythmias. After oral administration, PF undergoes extensive metabolism. 5-hydroxypropafenone is one of the major active metabolites of PF. In this thesis, the displacement of PF and 5-hydroxypropafenone in purified human aj-acid glycoprotein (AAG), albumin, serum and whole blood was examined using equilibrium dialysis. Free and bound PF and 5-hydroxypropafenone concentrations were analyzed by gas-chromatography with electron-capture detection (GC-ECD). Several antiarrhythmic agents (e.g. propranolol, verapamil, lidocaine, phenytoin and quinidine) and acetyl salicylic acid (ASA) were used as displacing agents. These displacing agents were chosen as specific probe for binding to recognized sites of the plasma protein molecules. The binding of 5-hydroxypropafenone to isolated human AAG phosphate buffer solution was studied over the concentration range of 50-20,000 ng/ml. One class of binding site was found. Saturation of binding sites occurred at concentrations of 5-hydroxypropafenone greater than 2,000 ng/ml. In purified solutions of human AAG or albumin, propranolol, verapamil, ASA and quinidine did not displace PF from its binding sites to a statistically significant degree. Phenytoin displaced PF to a statistically significant degree from its binding sites in albumin but not in AAG solution. Lidocaine caused a statistically significant increase in PF free fraction in solution of either AAG or albumin. 5-Hydroxypropafenone free fraction was not affected by verapamil ASA and propranolol, but increased with the addition of lidocaine in purified human AAG solution. In purified human albumin solutions, ASA caused a significant increase in 5-hydroxypropafenone, while no displacing effect was observed with either lidocaine, propranolol or verapamil. In human serum, quinidine, phenytoin and lidocaine caused a statistically significant increased PF free fraction. Neither ASA, quinidine nor phenytoin caused a statistically significant displacement of 5-hydroxypropafenone from its binding site in serum. Quinidine also displaced PF from its binding sites in whole blood while no displacing effect was observed with phenytoin. Quinidine and lidocaine were the most potent displacing agents for PF protein binding. They exhibited the displacement through competitive inhibition in purified human AAG. None of the displacing agents used seemed to be a potent inhibitor of 5-hydroxypropafenone protein binding in serum. Since PF is a drug known to exhibit very significant first-pass metabolism, we wished to prepare to study certain pharmacological interventions which might alter hepatic blood flow, and, hence PF clearance. Before such a study of PF could be conducted, we required a reliable measure of hepatic blood flow. The dye indocyanine green (ICG) affords such an estimate. We, therefore, set out to duplicate and i v refine on existing HPLC ICG assay. In early attempts, the detection limit of our HPLC analysis of ICG in human serum was 0.80 //g/ml. The analysis method of ICG in rat plasma gave an apparently good on-column detection limit (3.8 ng) using current detection technology available in our laboratory. However, we s t i l l could not detect the low concentration ICG samples as reported. This is, in part, due to the low injection volume that can be accommodated by our microbore HPLC system and more importantly the limitation of maximum detection wavelength available (600 nm). Fluorometric of ICG is not feasible and the chemical nature of this compound limit the use of solvent-solvent extraction method for sample preparation. With the manipulation of the sample extraction procedure, such as the use of solid phase extraction, or modification of our current liquid sample introduction configuration devices, it might be possible to develop a better sample preparation procedure in order to detect the low concentration which most likely will be encountered during the course of animal and/or human experimentation with ICG. V TABLE OF CONTENTS page ABSTRACT i i LIST OF TABLES x LIST OF FIGURES xi LIST OF ABBREVIATIONS xiv ACKNOWLEDGEMENTS xvi 1. INTRODUCTION 1 1.1 Pharmacology of Propafenone 1 1.2 Therapeutic Uses 2 1.3 Pharmacokinetics of Propafenone 2 1.3.1 Absorption 2 1.3.2 Distribution , 3 1.3.3 Metabolism and Elimination 3 1.3.4 Protein Binding of Propafenone 8 1.4 Pharmacological Effect of 5-Hydroxypropafenone 9 1.5 Protein Binding of 5-Hydroxypropafenone 10 1.6 Propafenone and Food Interaction 10 1.7 Effect of Alteration of Protein Binding and Blood Flow on Total Body Clearance 11 1.7.1 Effect of Binding Displacement on High Clearance Drugs 13 1.7.1.1 Displacement Mechanisms: Drugs and Endogenous Substrate 14 1.7.1.2 Selection of Protein Solutions as Drug Binding Displacement Media 15 1.7.1.3 Quantitative Analysis of Drug Protein Binding 16 1.7.2 Effects of an Increase in Liver Blood Flow on the Bioavailability of High Clearance Drugs 18 1.8 Indocyanine Green Clearance as an Indicator of Hepatic Blood Flow 19 1.9 Rationale 23 1.9.1 Rationale for Protein Binding Displacement Study 23 1.9.2 Rationale for Reproduction of the HPLC Analysis Method of ICG in Serum Samples 26 1.10 Objectives 26 v i 2. EXPERIMENTAL 28 2.1 Materials and Supplies 28 2.1.1 Drugs, Metabolites and Internal Standards 28 2.1.2 Chemicals and Reagents 28 2.1.3 Proteins 29 2.1.4 Solvents 29 2.1.5 Gases 29 2.1.6 Equil ibrium Dialysis Device 29 2.1.7 Other Supplies 30 2.2 Columns 30 2.2.1 GLC Column 30 2.2.2 HPLC Column 30 2.3 Equipment 31 2.3.1 Gas-Liquid Chromatography 31 2.3.2 High-Performance Liquid Chromatography 31 2.3.3 Spectrometer 31 2.3.4 Miscellaneous 32 2.4 Preparation of Stock and Reagent Solutions 32 2.4.1 Drug, Metabolites and Internal Standards for GLC Analysis 32 2.4.2 Drugs for Equilibrium Dialysis 33 2.4.3 Purified Protein Solutions 35 2.4.4 Drug and Internal Standard Solutions for HPLC Analysis 35 2.4.5 Reagent Solutions 36 2.5 Equilibrium Dialysis Procedure 37 2.6 Protein Binding of 5-Hydroxypropafenone to Purified Human AAG 38 2.7 Rosenthal Analysis of the Binding of 5-Hydroxypropafenone to Purified Human AAG 38 2.8 Scatchard Plot of 5-Hydroxypropafenone Binding to AAG 38 2.9 Free Fraction vs Total Concentration Plot of 5-Hydroxypropafenone Binding to AAG 39 2.10 Protein Binding Displacement Studies 39 2.10.1 Effect of Buffer Strength on the Protein Binding of Propafenone and 5-Hydroxypropafenone 39 2.10.2 Effect of Anticoagulant on the Protein Binding of Propafenone and 5-Hydroxypropafenone 40 2.10.3 Protein Binding Displacement Study Using Human AAG and Albumin as the Biological Fluids 40 2.10.4 Protein Binding Displacement Study Using Human Serum and Whole Blood as the Biological Fluids 41 2.10.5 Characterization of the Nature of Binding v i i Displacement Interaction between PF and Lidocaine and PF and Quinidine in Purified Human AAG 41 2.10.5.1 Binding of PF to AAG over the Therapeutic Concentration Range of PF 41 2.10.5.2 Characterization of the Nature of the Displacement Interaction: Competitive or Non-competitive 42 2.10.6 Measurement of Albumin Concentration in Purified Albumin Solution 42 2.11 Analysis of Propafenone and 5-Hydroxypropafenone 42 2.12 HPLC Analysis of Indocyanine Green (ICG) 43 2.12.1 Reproduction of Reported HPLC ICG Assay Methods 43 2.12.2 Modification of the Analysis Method of Indocyanine Green of Christie (1986) and Burns (1989) 44 2.12.3 Comparison of the Chromatogram of Serum Spiked with Indocyanine Green with That of Serum Blank 46 2.12.4 Modification of the Chromatographic Conditions to Minimize the Blank Interference 46 2.12.5 Extraction Efficiency 46 2.12.6 Post Extraction Stabi l i ty 47 2.12.7 Animal Preparation 47 2.12.8 Analysis of Plasma Samples 48 2.12.9 The Potential to Detect Low Concentration Samples with the New Method 49 2.12.10 Solid Phase Extraction Method 49 2.13 Data Analysis 50 2.13.1 Stat is t ica l Analysis 50 2.13.2 Protein Binding Data 50 2.13.3 Indocyanine Green Plasma Clearance 50 3. RESULTS 52 3.1 Volume Shift and pH Change of Equilibrium Dialysis Samples 52 3.2 Binding of 5-Hydroxypropafenone to Purified Human AAG 52 3.2.1 Rosenthal Analysis 52 3.2.2 Scatchard Plot of 5-Hydroxypropafenone Binding to Human AAG 52 3.2.3 Free Fraction versus Total Concentration Plot 54 3.3 Protein Binding Displacement Effect 54 3.3.1 Effect of Buffer Strength on the Binding of PF and 5-Hydroxypropafenone 54 3.3.2 Effect of Anticoagulant on the Protein Binding of PF and 5-Hydroxypropafenone 58 3.3.3 Protein Binding Displacement Studies using Purified Human Protein Solutions (AAG and Albumin) as the Biological Fluids 63 3.3.3.1 Protein Binding Displacement of Various Drugs on PF Binding to Purified Human v i i i AAG 63 3.3.3.2 Protein Binding Displacement Effect of Various Drugs on the Binding of PF to Purified Human Albumin 63 3.3.3.3 Protein Binding Displacement Effect of Various Drugs on the Binding of 5-Hydroxypropafenone to Human AAG 68 3.3.3.4 Protein Binding Displacement Effect of Various Drugs on the Binding of 5-Hydroxypropafenone to Human Albumin 68 3.3.4 Protein Binding Displacement Effect Using Human Serum as the Biological Fluid 73 3.3.5 Protein Binding Displacement Effect on PF Free Fraction Using Human Whole blood as the Biological Fluid 73 3.3.6 The Nature of the Displacement Interaction between PF and Quinindine and PF and Lidocaine Using Purified Human AAG as the Biological Fluid 77 3.3.6.1 Rosenthal Plot of the Binding of PF to Purified AAG over the Therapeutic Concentration Range of PF 77 3.3.6.2 The Nature of the Displacement of PF from Human AAG by Quinidine and Lidocaine 77 3.4 HPLC Analysis of Indocyanine Green (ICG) 80 3.4.1 HPLC Analysis of Indocyanine Green According to the Method of Rappaport and Thiessen 80 3.4.2 HPLC Analysis of Indocyanine Green with Fluorometric Detection 80 3.4.3 HPLC Analysis of ICG with Diode Array Detection 81 3.4.4 Comparison of Sample and Blank Chromatograms Using a Modified Method of Christie (1986) and Burns (1989) 87 3.4.5 ICG Extraction Efficiency 87 3.4.6 Post Extraction Stability 90 3.4.7 Plasma Data of ICG in Rat 90 3.4.8 The Potential to Detect ICG in Low Concentration Samples with the New Method 94 3.4.9 Solid Phase Extraction 94 4. Discussion 96 4.1 Protein Binding Technique 96 4.2 Protein Binding of 5-Hydroxypropafenone 99 4.3 Displacing Effect of Verapamil, Propranolol, ASA, Phenytoin, Lidocaine and Quinidine 101 4.3.1 Verapamil 101 4.3.2 Propranolol 102 4.3.3 Acetyl salicylic Acid (ASA) 103 4.3.4 Phenytoin 104 4.3.5 Quinidine 106 i x 4.3.6 Lidocaine 107 4.4 Clinical Significance of the Protein Binding Displacement Effect 108 4.5 HPLC Analysis of ICG 109 4.5.1 HPLC Analysis of ICG with Fluorometric Detection 109 4.5.2 HPLC Analysis of ICG with Diode Array Detection 110 4.5.2.1 Sample Preparation 110 4.5.2.2 HPLC Analysis of ICG Plasma Concentration in Rat 111 4.5.2.3 Summary 112 5. Summary and Conclusions 115 6. References 119 X LIST OF TABLES page 1. Methods for the determination of hepatic blood flow. 20 2. Free fraction of PF in a solution of AAG or albumin dissolved in 0.067 M or.0.1 M phosphate buffer. 57 3. Free fraction of 50HPF in a solution of AAG or albumin dissolved in 0.067 M or 0.1 M phosphate buffer. 59 4. Free fraction of PF in whole blood using heparin or EDTA as the anticoagulant. 60 5. Free fraction of 50HPF in whole blood using heparin or EDTA as the anticoagulant. 62 6. Extraction efficiency of ICG in serum. 87 7. Peak height ratios of ICG/diazepam before and after serum extraction. 90 8. Free fraction of 50HPF and PF in AAG, albumin, serum and whole blood. 100 9. PF free fraction before and after treatment with differing displacing agents 116 10. 50HPF free fraction before and after treatment with differing displacing agents 117 x i LIST OF FIGURES Page 1. Chemical structure of propafenone. 1 2. Relationship between steady-state mean plasma concentration and total daily dose. The dashed line indicates the response expected if the relationships were linear. Points and bars indicate mean and SEM. 4 3. Chemical structure of PF major metabolites: 5-hydroxypropafenone and N-depropylpropafenone 6 4. Chemical structure of indocyanine green (ICG). 22 5. A scheme of extraction procedure of ICG in human serum. 45 6. A scheme of extraction procedure of ICG in rat serum. 49 7. Relationship between the ratio of bound concentration/free concentration and bound concentration of 5-hydroxypropafenone in purified human AAG (Rosenthal plot), the inserted graph shows the relationship between bound and total concentration of 5-hydroxypropafenone. 53 8. Relationship between [r'/bound concentration] and r' of 5-hydroxypropafenone in purified human AAG (Scatchard plot); concentration of AAG was used in the calculation of [r ' ] . 55 9. Relationship between the ratio of free concentration/total concentration (free fraction) and total concentration of 5-hydroxypropafenone over the concentration of 20-20,000 ng/ml. 56 10. Protein binding displacing effect of propranolol, verapamil and ASA on the free fraction of PF in purified human AAG dissolved in 0.067 M phosphate buffer. The pH in each treatment group is 7.4, the value is the mean + s.d. and n is the number of replicates. 64 11. Protein binding displacing effect of phenytoin, quinidine and lidocaine on the free fraction of PF in purified human AAG dissolved in 0.1 M phosphate buffer. The pH in each treatment group is 7.4 and the value is the mean ± s.d., * indicates statistically significant difference with two sample t-test combined with Bonferroni inequality at p < 0.05, n is the number of replicates. 65 12. Protein binding displacing effect of propranolol, verapamil and ASA on the free fraction of PF in purified human albumin dissolved in 0.067 M phosphate buffer. The pH in each treatment group is 7.4, the value is the mean ± s.d. and n is the number of replicates. 66 x i i 13. Protein binding displacing effect of phenytoin, quinidine and lidocaine on the free fraction of PF in purified human albumin dissolved in 0.1 M phosphate buffer. The pH in each treatment group is 7.4 and the value is the mean ± s.d., * indicates statistically significant difference with two sample t-test combined with Bonferroni inequality at p < 0.05, n is the number of replicates. 67 14. Protein binding displacing effect o f propranolol, verapamil and ASA on the free fraction of 5-hydroxypropafenone in purified human AAG dissolved in 0.067 M phosphate buffer. The pH in each treatment group is 7.4, the value is the mean ± s.d. and n is the number of replicates. 69 15. Protein binding displacing effect of lidocaine on the free fraction of 5-hydropropafenone in purified human AAG dissolved in 0.1 M phosphate buffer. The pH in each treatment group is 7.4 and the value is the mean ± s.d., * indicates statistically significant difference with two sample t-test at p < 0.05, n is the number of replicates. 70 16. Protein binding displacing effect of verapamil, ASA and propranolol on the free fraction of 5-hydropropafenone in purified human albumin dissolved in 0.067 M phosphate buffer. The pH in each treatment group is 7.4 and the value is the mean ± s.d., * indicates statistically significant difference with two sample t-test combined with Bonferroni inequality at p < 0.05, n is the number of replicates. 71 17. Protein binding displacing effect o f lidocaine on the free fraction of 5-hydropropafenone in purified human albumin dissolved in 0.1 M phosphate buffer. The pH in each treatment group is 7.4 , the value is the mean ± s.d. and n is the number of replicates. 72 18. Protein binding displacing effect o f DPH (phenytoin), quinidine and lidocaine on the free fraction o f PF in human serum. The pH in each treatment group is 7.4 and the value is the mean ± s.d., * indicates statistically significant difference with two sample t-test combined with Bonferroni inequality at p < 0.05, n is the number of replicates. 74 19. Protein binding displacing effect of phenytoin, ASA and quinidine on the free fraction of 5-hydropropafenone in human serum. The pH in each treatment group is 7.4, the value is the mean ± s.d. and n is the number of replicates. 75 20. Protein binding displacing effect of quinidine and phenytoin on the free fraction of PF in human whole. The pH in each treatment group is 7.4 and the value is the mean ± s.d., * indicates statistically significant difference with two sample t-test combined with Bonferroni inequality at p < 0.05, n is the number o f replicates. 76 21. Relationship between the ratio of bound concentration/free x i i i concentration and bound concentration of PF in purified human AAG (Rosenthal plot), the inserted graph indicates the relationship between bound and total concentration of PF. 78 22. Relationship between the reciprocal of the moles of PF bound per mole of AAG and the reciprocal of unbound PF without displacing agent ; in the presence of quinidine ; in the presence of lidocaine . 79 23. Excitation and emission (at excitation wavelength of 260 nm) spectra of ICG using pH 6.0 phosphate buffer as part of the mobile phase. 82 24. A representative chromatogram that compares fluorometric detection with UV detection. 83 25. Excitation spectrum of ICG using pH 4.0 citrate buffer as part of the mobile phase. 84 26. A representative chromatogram obtained using the modification method (section 2.12.2). 85 27. A representative standard curve obtained using the modification method (section 2.12.2). 86 28. Representative chromatograms of serum (0.25 ml) spiked with ICG and serum blank. 88 29. Representative standard curve obtained using 0.25 ml serum. 89 30. A representative chromatogram in rat plasma. 91 31. A representative standard curve obtained in rat plasma. 92 32. Plasma concentration-time curve of ICG in rat plasma. 93 33. A representative chromatogram obtained using solid phase extraction method. 95 34. Ultraviolet/visible spectrum of ICG in distilled water (10 /zg/ml). 113 x i v L I S T OF ABBREVIAITIONS AAG a p a c i d g l y c o p r o t e i n AUC a r e a u n d e r t h e c o n c e n t r a t i o n versus t i m e c u r v e A U C o r a i a r e a u n d e r t h e c o n c e n t r a t i o n versus t i m e c u r v e a f t e r o r a l a d m i n i s t r a t i o n AV a t r i o v e n t r i c u l a r C m a x peak c o n c e n t r a t i o n o f d r u g i n s e r u m CI c l e a r a n c e C l i n t i n t r i n s i c c l e a r a n c e C V . c o e f f i c i e n t o f v a r i a t i o n D d o s e [D] f r e e d r u g [DP] d r u g - p r o t e i n c o m p l e x E e x t r a c t i o n r a t i o ECD e l e c t r o n - c a p t u r e d e t e c t o r F s y s t e m i c a v a i l a b i l i t y GLC g a s - l i q u i d c h r o m a t o g r a p h y GLC-ECD g a s - l i q u i d c h r o m a t o g r a p h y - e l e c t r o n - c a p t u r e d e t e c t i o n HDL h i g h d e n s i t y l i p o p r o t e i n HP H e w l e t t - P a c k a r d HPLC h i g h - p e r f o r m a n c e l i q u i d c h r o m a t o g r a p h y ICG i n d o c y a n i n e g r e e n I.D. i n t e r n a l d i a m e t e r i . v . i n t r a v e n o u s LDL l o w d e n s i t y l i p o p r o t e i n X V M.W. molecular weight n number of homogeneous binding sites 50HPF 5-hydroxypropafenone [P] free protein (unbound) PF propafenone PTFE polytetrafluoroethylene r correlation coefficient rpm revolutions per minute s.d. standard deviation t time t h a l f - l i f e t m a x time to reach peak concentration T dosing interval TEA triethyl amine UV ultraviolet VLDL very low density lipoprotein x v i ACKNOWLEDGEMENTS I would like to sincerely thank Dr. James Axel son for his supervision, guidance and financial support. I would like to thank my committee members Drs. Frank Abbott, Jack Diamond and John Sinclair for their advice and support through out my study. Thanks to Johannes Schenk and Zhen Yu for their help in animal experiments. Special thanks to George Tonn, Mathew Wright, Swamy Yeleswaram, Sue Panesar and Grace Chan for their critical advice in my experiments and in the writing of the thesis and their constant support and encouragement. My appreciation to Judit Orbay and Andras Seitz for their help in experiments and friendship. Thanks to Sonia Chan for helping me through the days when I was a teaching assistant. Thanks to my fellow graduate students Marion Wong, Seema Gadkari, Heyi Liu and Yongjiang Hei for their friendship and help. Special thanks to my husband, Futong Cui, and my family, for their support and understanding. This project was supported by The Canadian Heart and Stroke Foundation. I would finally like to gratefully thank the B.C. and Yukon Heart and Stroke Foundation for the financial Traineeship support. X V I 1 This thesis is dedicated to father, for giving me the courage to believe in my capabi l i t ies , . t o my dear s ister , for her continuous support and encouragement, to husband, Futong, for his unfailing encouragement and understanding and to the memory of my mother. 1 1. I N T R O D U C T I O N 1.1 Pharmacology of Propafenone Propafenone (PF) (Figure 1, marketed as a racemate) is a type IC antiarrhythmic agent with weak B-blocking act iv i ty and calcium channel blocking act iv i ty (Harron and Brogden, 1987). Like other type IC antiarrhythmic agents, PF exerts its pharmacological effects through the depression of sodium influx (Vaughan Williams, 1970). PF also causes a dose-dependent decrease in the maximum rate of depolarization and conduction velocity in tissue depolarized by the fast sodium channel (Dukes and Vaughan Williams, 1984). Compared with propranolol, the B-blocking act iv i ty of PF is re lat ive ly small (McLeod et al., 1984). In patients with ventricular arrrhythmia, PF caused significant prolongation of PR and QRS intervals; increases in atroventricular nodal conduction time and His-Purkinje conduction time (Hartel, 1985; Nauardla et al., 1984; Connolly et al., 1983a,b). Haemodynamical ly , PF did not change mean arterial pressure or heart rate at concentrations ranging from 956 to 1564 /zg/ml (Shen et al., 1984). - C O - C H 2 - C H 2 - C 6 H 5 0-CH 2 -CH-CH 2 -NH-C 3 H 7 Figure 1. Chemical structure of propafenone (PF) 2 1.2 Therapeutic Uses Propafenone has been used in the treatment of chronic ventricular arrhythmias, atrial and AV nodal junctional re-entrant tachycardias, and Wolff-Parkinson-White tachycardias (Harron and Brogden, 1987). It is effective for the acute management of atrial f ibrillation (Vita et al., 1989) and prevention of recurrent atrial f ibrillation (Kerr et al., 1988). It is more effective than the traditional antiarrhythmic agents: quinidine, disopyramide, tocainide and is comparable to lidocaine, flecainide, and metoprolol in efficacy and has a low incidence of adverse effects (Harron and Brogden, 1987). 1.3 Pharmacokinetics of Propafenone 1.3.1 Absorption The bioavailability of PF after oral administration is quite low (4.8% with a 150 mg dose and 12% with a 300 mg dose ) due to extensive first-pass metabolism (Hollmann et al., 1983). It is rapidly absorbed after oral administration and dose-dependent increases in both peak plasma concentration and AUC have been reported (Hollman et al., 1983; Connolly et al., 1983b; Frabetti et al., 1986; Zoble et al., 1989). After administration of three different doses of PF (150, 300, 450 mg) to 19 healthy subjects, Hollmann (1983) found disproportional increases in peak plasma concentration ( 139.3, 384.0, 827.0 /ig/ml). A similar relationship was found in arrhythmia patients (Connolly et al., 1983). A 3 fold increase in dose from 300 to 900 mg resulted in a 10 fold 3 elevation of steady-state PF concentrations (Figure 2). This non-linear relationship between dose and plasma concentration was most likely due to the saturation of the cytochrome P-450 isoenzyme(s) which was the same cytochrome P-450 responsible for debrisoquine's 4-hydroxylation (Siddoway et a/ . , 1987). Further study using human kidney donor livers indicated that 5-hydroxypropafenone was formed by the cytochrome P-450 isozyme which participated in the polymorphic oxidation of bufuralol (Kroemer et al., 1989). 1.3.2 Distribution The tissue distribution of PF and 5-hydroxypropafenone in one patient was determined in autopsy samples 24 hours after death (Latini, 1987). The highest concentration of PF was found to be in the lung (12 ug/q wet tissue), followed by the heart (6.2 fig/g) and liver (4 fig/g). The metabolite, 5-hydroxypropafenone, tissue concentration was lower than the parent compound in most organs except for the liver (7.2 fig/g). However, the metabolite/parent drug ratio was close to 1 in the heart (compare with 0.1 - 0.4 in plasma) indicating that there was greater partitioning of 5-hydroxypropafenone into the myocardium than for the parent drug, propafenone, and thus that antiarrhythmic activity of this metabolite was possible. 1.3.3 Metabolism and Elimination Propafenone undergoes extensive first-pass metabolism after oral administration (Harron and Brogden, 1987). In humans, less than 1% of 4 CONNOLLY et al. E 0 300 600 900 DAILY DOSE PROPAFENONE (mg) Figure 2. Relationship between steady-state mean plasma concentration and total daily dose. The dashed l ine indicates the response expected i f the relationships were l inear. Points and bars indicate mean and SEM. 5 the parent drug was excreted in urine or feces after administration of deuterium-labelled PF (Hege et al., 1984). Fifty three percent of the dose was recovered in the feces as metabolites. In patients with frequent ventricular ectopy who underwent chronic treatment with PF (300 mg, b . i . d . for 2 weeks), the major metabolites were 5-hydroxypropafenone and N-depropylpropafenone (the structures are shown in Figure 3) (Kates et al., 1985). At steady-state, the plasma concentrations of 5-hydroxypropafenone and N-depropylpropafenone are about 23% and 17% of the corresponding propafenone concentration, respectively. The effect of enzyme induction on the pharmacokinetics of PF and its major metabolite 5-hydroxypropafenone was studied by Chan (1989). Phenobarbital treatment (100 mg daily at bedtime for 23 days) induced hepatic microsomal enzymes and enhanced the extent of the first-pass metabolism of PF in eight healthy non-smoking and eight healthy heavy cigarette smoking Caucasian males (age 20-45 y) , respectively. A significant increase in C I o f PF after phenobarbital treatment was observed. On the other hand, phenobarbital treatment has no effect on 5-hydroxypropafenone elimination, which indicates that the enzyme responsible for the metabolization of PF to 5-hydroxypropafenone is not subject to enzyme induction caused by phenobarbital. When compared to the non-smokers, heavy cigarette smokers appeared to show an increase in the clearance of PF, however, a general conclusion that smoking induced the metabolism of PF was d i f f i c u l t to address. It was, therefore, concluded that while phenobarbital induced the metabolism of PF in smokers, there was no effect of smoking per se on the metabolism of PF. 6 <^~\-CO-CH 2-CH 2-C 6H 5 O-CHo-CH-CHo-NH-CoH-7 OH propafenone <QKO-CH2-CH2-C6H5 0-CH2-CH-CH?-NH? OH C N-depropyl propafenone HO J^^ -CO-CH2-CH2-C6H5 O-CH0-CH-CH0-NH-C0H7 OH * ' 5-hydroxy propafenone Figure 3. Chemical structures of PF major metabolites: 5-hydroxypropafenone and N-depropylpropafenone 7 The pharmacokinetics of PF is time-dependent as studied by Giani et al. (1988). Giani et al. (1988) compared the ratio of the plasma AUCs of 5-hydroxypropafenone and PF after a single dose of PF to that seen after one month multiple dose therapy of PF (300 mg t . i .d . ) in 10 patients with stable, frequent, premature ventricular beats. They found that the AUC ratio of metabolite/PF decreased from 0.63 after the single dose experiment to 0.32 following one month therapy and also the elimination half-life was almost doubled (6.7 h following one month therapy and 3.5 h after single dose). They suggested that this decrease in PF clearance was due partly to a drop in the rate of metabolism of PF to 5-hydroxypropafenone, as 5-hydroxypropafenone did not accumulate in the same magnitude as PF. The mean total body clearance of PF after single i.v. dose of 70 mg to 8 healthy subjects before and after 12 day's application of PF 150 mg b . i .d . were 1.14 L/min and 1.10 L/min, respectively (Hollmann et a/ . , 1983). In 10 patients with superventricular tachycardia given 2.3 ± 0.2 mg/kg i.v., total body clearance was 1.03 L/min (Arboix et al., 1985). The volume of distribution of PF after i.v. administration was « 3 L/kg in healthy subjects (Hollmann et al., 1983) and 1.6 L/kg in patients with arrhythmias (Arboix et al., 1985). These values are greater than the actual body volume suggesting considerable tissue binding. The mean °f ^ i" healthy volunteers was found to be approximately 4.6 h (range 2.3-9.5 h) after an oral single dose of 300 8 mg, and 2.8 h (range 2.1-4.1 h) after a 70 mg i.v. (Hollmann et a/ . , 1983). In patients displaying episodes of paroxysmal supraventricular tachycardias, the elimination half-life (ti/2) of PF w a s approximately 80 min (range 41-148 min) after a single intravenous dose of 2.3 ± 0.2 mg/kg (Arboix et al., 1985). In patients with chronic, stable cardiac dysrhythmias, the t]/2 of PF was 3.6 h after a single 900 mg oral dose (Keller, 1978). In another study, the tj/2 of" PF w a s 3 - 8 n (range 1.9-7.9 h) and 3.75 h (range 1.65-7.8 h) after a single dose of 150 mg or 300 mg, respectively (Frabetti et al., 1986). Individual differences in drug metabolism may contributed to the variation seen in the tj/2 of PF. The effective plasma concentration of PF at steady-state shows great intersubject variability. The clinical therapeutic range for this drug is 0.06-1 //g/ml (Latini et al., 1990). To obtain greater than 70% suppression of ventricular ectopic depolarizations, the plasma concentration ranges from 42 to 1801 ng/ml between subjects (Siddoway et al., 1987). Changes in PR and QRS in the ECG are related to plasma steady-state concentrations (Frabetti et a7.,1986; Zoble et al., 1989). Taking the non-linear relationship between dose and plasma concentration into account, it is obvious that there is a disproportional relationship between dose and pharmacological response with PF. 1.3.4 Protein Binding of Propafenone Propafenone binds extensively to both plasma and tissue proteins. More than 95% of PF binds to serum proteins (Seipel and Breithardt, 1980), and the binding is proportional to the concentration of AAG 9 (Gill is et al., 1985). Gill is et al. (1985) have studied the binding of PF and other antiarrhythmic agents to purified human AAG. Using radio-labelled drugs, two classes of binding sites for PF over the concentration range of 0.009-81 ug/m\ were demonstrated. PF also bound to a greater extent than lidocaine, verapamil and propranolol to AAG. In vitro, in normal serum PF bound independent of concentration within the range of 0.5-1.5 ^g/ml and the binding was concentration dependent at PF concentration greater than 0.2 uq/m\ (Chan et al., 1989). The free fraction of PF decreased approximately 50% in uraemic serum. A good correlation (r=0.8302) between AAG concentration and the PF binding ratio was also reported. The binding of PF in whole serum and isolated protein solutions of AAG, albumin, HDL, LDL and VLDL was examined extensively by Tonn (1990). The binding of PF was found to be pH dependent in serum and in solutions of albumin and AAG, respectively. Moderate binding to albumin and extensive binding to AAG were found. The binding to AAG correlated well with the binding observed in serum. The degree of binding was similar in control serum and lipoprotein deficient serum. The binding of PF to HDL and LDL was greater than serum and non-saturable binding was found in solutions of LDL and VLDL. 1.4 Pharmacological Effect of 5-Hydroxypropafenone 5-Hydroxypropafenone, the major metabolite of PF, has been found to be pharmacologically active in animal models (von Philipsborn et al., 1984; Valenzuela et al., 1987; Thompson et al., 1988). In the intact animals (aconitine induced arrhythmia rats and coronary ligation induced infarction arrhythmic dogs), 5-hydroxypropafenone was more potent as an 10 antiarrhythmic agent than the parent compound (von Philipsborn et al., 1984). This phenomenon was also observed by Malfatto et a7.(1989) in coronary ligation induced infarction arrhythmic dogs. 1.5 Protein Binding of 5-Hydroxypropafenone The in vitro protein binding of the active metabolite has been studied extensively in our laboratory. In human serum, 5-hydroxypropafenone displays concentration independent binding with a mean free fraction of 0.184 over the concentration range of 0.12 to 35.3 jug/ml (Tonn, 1990). The binding of 5-hydroxypropafenone was found to be pH dependent in serum and in a solution of AAG, but not in a solution of albumin. Moderate binding of 5-hydroxypropafenone to albumin was observed. A significant decrease in 5-hydroxypropafenone binding to AAG was observed upon the addition of PF. Decreased binding was also found in albumin upon the addition of a physiological concentration of AAG. Removal of lipoprotein in serum resulted in increases in the free fraction of 5-hydroxypropafenone as compared to control serum. The binding of 5-hydroxypropafenone to HDL, LDL, VLDL was similar to PF. 1.6 Propafenone and Food Interaction Food induced changes in the pharmacokinetics of high clearance drugs such as propranolol (Mclean et al., 1981) and metoprolol (Melander et al., 1977) have been well documented. This food effect was partially explained by a food induced change in splanchnic blood flow and intrinsic drug metabolism (Mclean et al., 1978; Svensson et al., 1983; 11 Svensson et al., 1984). When Axelson et a / . (1987) studied the effect of food on the pharmacokinetics of propafenone given as a single oral dose, they found that a highly significant increase in the b ioavai labi l i ty of this drug (> « 150% increase) in the fed vs fasted states, characterized by increased C m a x and A U C o r a i , with no change in t m a x in most subjects studied. The effect of food on this drug was part ia l ly contributed by a transient increase in l iver blood flow induced upon food intake and the possible interference with hepatic clearance induced by food components. 1.7 Effect of Alteration of Protein Binding and Blood Flow on Total Body Clearance Hepatic blood flow, plasma protein binding and hepatic intr ins ic clearance are the three important factors which determine the disposition of drugs that undergo extensive first-pass metabolism. The elimination of most drugs from the body involves the processes of both metabolism (biotransformation) and excretion. For many drugs that undergo extensive metabolism, i.e., high clearance drugs, the renal clearance is negligible, and the total body clearance is approximately equal to hepatic clearance: Cljg « C l ^ . When the l iver is considered as a wel l -st irred compartment with concentration of drug in the l iver in equilibrium with the effluent blood, the relationship among hepatic clearance, protein binding and hepatic blood can be described by the following equation (Pang and Rowland, 1977; Pang, 1980): c 1 i n t f B C1H = QH (1) % + c l i n t f B 12 where CI^  is hepatic clearance; is hepatic blood flow; C l i n t is the intrinsic clearance which is used to describe the ability of the liver to remove drug in the absence of flow limitations and it reflects the inherent ability of the liver to irreversibly remove drug by all pathways; fg is the free fraction of the drug. For high clearance drugs, (fy « C I j n t fg, therefore CI^  « Q .^ In this case, drugs are so avidly extracted by the liver that binding to plasma proteins does not limit their removal, thus total body clearance is not affected by alterations in plasma protein binding. However, the hepatic clearance is hepatic blood flow dependent (Gibaldi and Perrier, 1982). For low clearance drugs, » C l ^ n t fg, and CI^  « C l ^ fg. This indicates that the intrinsic elimination processes responsible for the removal of these drugs are very weak, such that only unbound drug in the plasma can be cleared. Hepatic extraction and consequent hepatic clearance is limited by the extent of plasma protein binding and independent of hepatic blood flow (Gibaldi and Perrier, 1982). Steady-state plasma concentrations, volume of distribution and elimination half-life are affected by protein binding displacement from blood and/or tissue proteins (D'arcy and McElany, 1982 and 1983; Mackichan, 1984 and 1989). The relationship between clearance and steady-state plasma concentration can be described by the following equations: 13 F D C s s = - - - - - - (2) T C 1 T B Where T is the dosing interval and CIjg is the total body clearance. 1.7.1 Effect of Binding Displacement on High Clearance Drugs As mentioned before, for high clearance drugs, the hepatic clearance is approximately equal to total body clearance. Also, the hepatic clearance is limited by hepatic blood flow and is not affected by the degree of protein binding. Therefore, the total body clearance of high clearance drugs would remain the same upon a change in the degree of protein binding induced by displacement. Since the steady-state total plasma drug concentration is determined by the total body clearance under fixed dosing regimen (Equation 2), changes in the degree of protein binding as a result of binding displacement has no effect on steady-state total drug concentration. However, the steady-state free drug concentration is affected by protein binding according to the equation: C s s f r e e = fg C s s . Binding displacement will lead to an increase in free fraction and the consequent increase in free drug concentration. This is of clinical significance for a number of reasons. It is well believed that only the free drug is the pharmacological active species as free drug can pass through cell membrane, reach the receptor sites and exert pharmacological activity (Goldstein, 1949). Accordingly, the pharmacological response is likely to be better 14 correlated with free drug concentration than total concentration. For drugs with narrow therapeutic indices, fluctuations in free drug concentration can lead to either toxic effects or subtherapeutic drug levels depending upon circumstances. When therapeutic drug monitoring is based on total drug concentration, the change in free drug concentration may well be undetected in the case of high clearance drugs. Polydrug therapy is frequently used in the treatment of disease, therefore, drug binding displacement interactions can, in some instances, be important (Mackichan, 1989). It is, thus, important to study the displacement effect on free drug concentration in order to modify dosage regimens. It is also important to characterize drug plasma protein binding in order to predict the possible alteration in binding and, hence the effect, in the presence of coadministered drugs, of various disease states which alter plasma protein concentration. 1.7.1.1 Displacement Mechanisms: Drugs and Endogenous Substrate Displacement involving drugs can be competitive and non-competitive. If two drugs are bound at the same site on a macromolecule, there will be a simple competition between the drugs for these sites; compounds with a high affinity will displace drugs with a lower affinity at the same binding sites. The displacement of one drug by another will depend on the displacer's concentration. Non-competitive displacement can occur if one drug changes the physical chemistry of the macromolecule to which it binds and induces tertiary conformational changes in the macromolecule itself. 15 Endogenous substances which can be elevated during certain disease states may compete with drugs for plasma binding sites. It has been shown that administration of heparin alters the plasma binding of phenytoin (Craig et al., 1976), quinidine (Kessler et a?., 1979) and propranolol (Wood et al., 1979 a,b). The increased free fatty acids level as a result of in vitro hydrolysis of plasma triglycerides by lipases after in vivo administration of heparin is responsible for the observed changes in binding. Although such an interaction is unlikely to give rise to adverse effects in patients, such changes in binding may distort pharmacokinetic data derived from free drug serum levels in subjects using a heparin lock for blood sampling. 1.7.1.2 Selection of Protein Solutions as Drug Binding Displacement Media Albumin and AAG are, quantitatively speaking, the most important proteins in drug binding in human blood. Isolated protein solutions can give detailed information about binding, such as binding affinity, capacity and binding sites. Compared with isolated protein solutions, serum and whole blood contain more ligand binding macromolecules and, therefore, possess higher buffer capacity upon binding displacement particularly when the ligand binding is nonspecific. Therefore, it is better to use isolated AAG and albumin solutions to define the displacement interaction mechanism(s) and using serum as well as whole blood to find the clinically important displacing agents. Only when the free fraction of high clearance drug is increased, by the displacing agent in serum or whole blood, will the interaction be of potential clinical significance, since drug and/or its metabolite(s) exists in whole blood in physiological conditions. 1.7.1.3 Quantitative Analysis of Drug Protein Binding Drug protein binding is one type of a 1igand-receptor interaction. For reversible drug protein binding, equilibrium exists among free drug [D], unbound protein [P] and drug protein complex [DP] as shown below: Where KI and K2 are the absolute association and dissociation constants, respectively. At equilibrium, the relationship between [D], [P] and [DP] can be described by the following equation: Where Ka is the apparent or intrinsic association constant. Using r as the molar ratio of drug bound to total protein, 7.e. r = [DP]/([DP] + [P]), substituting [P] with ([Pt] - [DP]) ( [Pt] is the total protein concentration), the following equation is derived: [D] + [P] = * : [ D P ] (3) K2 Ka = K1/K2 = [DP]/([D][P]) (4) Ka[D][Pt] [DP] = (5) 1 + Ka[D] If more than one number of binding sites exist per protein molecule, the following equation is used: 17 nKa[D][P] [DP] = (6) 1 + Ka[D] This is the equation to describe the simplest one ligand binding to one group of mutually noninteracting binding sites. With the mathematical manipulation of Equation 6, a few equations can be derived: 1 1 1 — - = + — (7) r n[D]Ka n r/[D] = nKa - rKa (8) [DP]/[D] = n[Pt]Ka - [DP]Ka (9) Equation 7 describes the Klotz reciprocal plot (Svensson et al., 1986; Klotz, 1983), while equation 8 is used to create the Scatchard plot (Scatchard, 1949), equation 9 is used to produce the Rosenthal plot (Rosenthal, 1967). From these plots, the number of binding sites and the intrinsic association constant can be determined. The Klotz reciprocal plot is similar to the Lineweaver-Burke plot (Svensson et al., 1986; Klotz, 1983) used in enzyme kinetics. In the presence of a pure competitive inhibitor for protein binding, the slope of this plot will be altered by a factor of (1 + [I]/Ki). Where [I] is the concentration of inhibitor and Ki is the inhibitory constant. The Ki value can thus be determined from the following equation: 18 Slope = (Uninhibited Slope) ( 1 + [I]/Ki) (10) 7 .e. PHSlope) Ki = (11) Slope - Uninhibited Slope Where slope and uninhibited slope are the slopes in the Klotz reciprocal plots in the presence and absence of inhibitor, respectively (Engel, 1988). 1.7.2 Effect of an Increase in Liver Blood Flow on the Bioavailability of High Clearance Drugs Increases in blood flow can be classified into two categories: constant increases and transient increases. The pharmacokinetic changes as a consequence of constant increase in liver blood flow includes: a) , reduction in tj/2 due *° the augmentation in total body clearance (Cljg) (assuming the volume of distribution does not change); b) . alteration in AUC o r a ^ due to the following reasons: Equation 12 describes the relationship between oral bioavailability (AUC o r a-|) and systemic availability (F) and total body clearance (Clyg), where the systemic availability can be expressed by Equation 13 A U C ora l = ( F °ose) /Cl (12) F = 1-E = 1 -Cl i n t / (Q + C l i n t ) . (13) 19 As the Cljg increases, F increases, depending on the magnitude of changes, the AUC o r a- | may increase, decrease, or not change (Gibaldi and Perrier, 1982; McLean et al., 1978). If the increase in liver blood flow is transient, as seen after food intake, the magnitude of increase in systemic availability is greater than total body clearance, and a net increase in AUC o r a-| will be observed (Gibaldi and Perrier, 1982; McLean et al., 1978). 1.8 Indocyanine Green Clearance as an Indicator of Hepatic Blood Flow One possible partial explanation for the interaction between food and PF has been a transient increase in Q ,^ although it also was suggested that food components interferred with hepatic clearance, as well (Axelson et al., 1987). However, no measurement of was made in this study. There are two main obstacles in the measurement of liver blood flow in human: the dual nature of hepatic supply: one by hepatic artery, one by portal vein, and the relative inaccessibility of the liver circulation. Before 1945, all the hepatic blood flow measurements made were highly invasive. In 1945, Bradley and his colleagues employed the Fick principle to measure liver blood flow in man using the dye: bromosulphthalein (BSP), a substance which is removed by hepatocytes into the bile. Thereafter, a series of techniques has been developed. The methods can be classified into three categories: clearance techniques, such as indocyanine green clearance, indicator dilution techniques, such as injection of radio labelled ( 1 5 Cr) red blood cells, and Doppler techniques which using ultrasound to measure blood flow 20 (Bradley et a l . , 1974; Ohnhaus, 1979; Baker, 1978). Table 1 l ists the available methods. Table 1. Methods for the determination of hepatic blood flow CLEARANCE TECHNIQUES Hepatocyte excretory clearance Bromsulthphalein Indocyanine green 1 3 1I-Rose Bengal Hepatocyte metabolic clearance Ethanol Galactose Reticuloendothelial particle clearance •^P-chromic colloid 198Au S^Fe-saccharate 99mTc-sulfur colloid ^^I-heat denatured serum albumin Inert gas 133Xe INDICATOR DILUTION TECHNIQUES serum albumin ^^Cr-red blood cells ULTRASOUND TECHNIQUE (Adapted from Bradley et a l . , 1974) The clearance technique is one of the most convenient methods to measure blood flow. In this technique, there are several requirements for the test substance or indicator. It should: 1. distribute only within the vascular space, 2. only be removed from the body by the liver with an extraction 21 ratio ~1, 3. not participate in enterohepatic circulation, 4. not be toxic at the doses employed, 5. and its concentration in blood should be measurable with sufficient accuracy (Paumgartner, 1975). Indocyanine green (ICG), a tricarbocyanine dye (Figure 4), is one of the substances that can be used for measurement of blood flow. It was introduced into clinical medicine by Fox et al.(1957a; 1957b; 1960). Theoretically, based on its molecular weight of 775 and its two polar sulphonic acid groups, it was seemed very likely that ICG undergoes biliary excretion into feces (Levine, 1978). Experimentally, Wheeler et al. in 1958 found that more than 97% of the administered dose of ICG can be recovered in unaltered form in the bile of dog. In patients with uncomplicated gall bladder disease undergoing cholecystectomy, Meijer found that the recovery of unchanged ICG in bile over 18 h was 80% of the dose after i.v. injection of either 0.5, 1.0 or 2.0 mg/kg (Meijer et al., 1988). Extrapolation of biliary excretion rate curves after that period indicated that at least 5% of the dose would be excreted subsequently, which means that recovery was almost complete. Under these experimental circumstances, urinary excretion and metabolism did not occur to a significant extent with biliary excretion accounting for the largest proportion of the elimination process for the dye. The volume of distribution was close to plasma volume indicating that the distribution of ICG was primarily within the vascular space. Thus, ICG appears to be eliminated from blood almost exclusively by the liver. Uptake of ICG by peripheral tissue, kidneys, or lungs is negligible in 22 Figure 4. Chemical structure of indocyanine green (ICG). 23 either human or experimental animals (Cherrick et al., 1960; Humton et a7.,1960; Ketterer et al., 1959; Leevy et al., 1963; Rapaport et al., 1959; Reubi et al., 1966; Wheeler et al., 1958). The toxicity of ICG is relatively low. Given in high doses up to 5 mg/kg body weight, no significant toxic effects were observed in human (Leevy et al., 1967). Therefore, it has been concluded that the disappearance of ICG from blood is a reflection of hepatic removal and the dye has been determined to be safe for use in the evaluation of liver function and blood flow. 1.9 Rationale 1.9.1. Rationale for protein binding displacement study Hypothesis: Displacement of PF and 5-hydroxypropafenone from plasma proteins such as AAG and albumin by other highly bound antiarrhythmic drugs and ASA may result in an increase in free drug concentration with a resulting increase in therapeutic effect and/or toxicity. Propafenone is highly protein bound in serum of normal healthy volunteers (Chan et al., 1989) and the binding is proportional to the concentration of AAG (Gill is et al., 1985). Like most antiarrhythmic drugs, PF undergoes extensive hepatic metabolism with less than 1% excreted unchanged in urine and feces (Hege et al., 1984). The high total systemic clearance (1.14 L/min in healthy volunteers and 1.03 L/min in arrhythmic patients) numerically approaches liver blood flow 24 (1.5 L/min ) (Hollmann et al., 1983; Arboix et a/ . , 1985 ), which, together with the extensive metabolism, results in the characterization of PF as a liver blood flow-dependent, highly extracted and high clearance drug. Moreover, PF appears to show a steep dose-response relationship (Connolly et al., 1983b; Siddoway et al., 1984a). This indicates that a small change in drug concentration will result in greater change in response when compared with drugs that display normal dose-response relationships. To date, good correlation between serum concentration and therapeutic response of propafenone has not been developed and large individual variations in therapeutic concentration of PF exist in patients (Harron and Brogden, 1987). Fluctuation in PF free drug concentration and the antiarrhythmic activity of 5-hydroxypropafenone and possible other metabolites most likely contribute to the observed variation in therapeutic drug concentrations. The antiarrhythmic activity of 5-hydroxypropafenone in various experimental models has been well documented. These models include isolated cardiac strips and coronary ligation in dogs (Von Philipsborn et al., 1984; Thompson et al., 1985; Delgado et al., 1987), and isolated rat heart preparation (Oti-Amoako et a7.,1990). In human, comparison of the pharmacodynamic effects of intravenous and oral propafenone showed that the presence of relatively large amounts of 5-hydroxypropafenone after oral PF administration prolonged further the PQ interval and QRS duration than intravenous PF administration where the amount of 5-hydroxypropafenone was negligible (Haefeli et al., 1990). Moreover, poor metabolizer patients who are not able to form substantial amounts of 5-hydroxypropafenone require higher propafenone concentrations for 25 arrhythmia suppression than extensive metabolizers with high plasma levels of this metabolite (Siddoway et a/ . , 1987). Although there is no data available to display the direct electrophysiologic effect of 5-hydroxypropafenone in human, it contributes to the observed pharmacological effect. Therefore, it is important to study its protein binding behavior to obtain a better understanding of the dose-response relationship observed after PF administration. A fluctuation in PF free drug concentration could be induced by dose-dependent protein binding of the drug itself, or by displacement interactions with coadmininstered drugs. Quite frequently, polytherapy is used in clinical treatment of dysrhythmias. Among all the possible drugs coadministered with PF, phenytoin, lidocaine, propranolol, quinidine and verapamil are the most likely potent displacers due to the high plasma protein binding characteristics of these drugs. Lidocaine is approximately 25-50% bound to AAG, phenytoin 93% to albumin, propranolol 55% to albumin and 58% to AAG, quinidine 60-90% to albumin and AAG, and verapamil is 90% bound to plasma protein (Edcardsson and Olsson, 1987; Evans et al., 1986). ASA binds to albumin and results in permanent acetylation of the albumin molecule (Hawkins et al., 1968). Also, since ASA is such a commonly used drug, it seemed appropriate to study the displacement effect of this drug on the binding of both PF and 5-hydroxypropafenone. The therapeutic concentrations for these drugs in /zg/ml are: 1.4-6.0 for lidocaine, 10-18 for phenytoin, 0.05-1 for propranolol, 2-5 for quinidine, 0.06-0.2 for verapamil (Latini et al., 1990) and 50-400 for ASA (Evans et al., 1986). The maximum therapeutic concentrations for these displacers will be used. Purified human AAG 26 and albumin, serum and whole blood will be used as equilibrium dialysis media due to the fact that different media exhibits different "buffering" capability with respect to displacement. 1.9.2. Rationale for Reproduction of the HPLC Analysis Method of ICG in Serum Samples Several HPLC methods for the measurement of ICG concentrations in serum are available with relative high sensitivity (Rappaport and Thiessen, 1982; Hollins et al., 1987; Dorr et al., 1989; Donn et al., 1984; Christie et al., 1986; Burns et al., 1989). It will be our objective to establish an HPLC measurement method which would be suitable for the measurement of ICG serum concentration during the course of experiments designed to determine hepatic blood flow under different experimental conditions. 1.10 OBJECTIVES 1. To study the effect of buffer strength on the free fraction of PF and 5-hydroxypropafenone in solutions of purified human albumin, AAG and whole blood. 2. To study the effect of heparin and EDTA on the free fraction of PF and 5-hydroxypropafenone. 3. To characterize the binding of 5-hydroxypropafenone to purified human AAG. 4. To study the displacing effects of phenytoin, lidocaine, propranolol, quinidine, verapamil and ASA in purified human AAG, albumin, serum and whole blood. To characterize the nature of the protein binding displacement in purified human AAG solution. To reproduce and attempt to improve upon an HPLC assay method for ICG. 28 2. EXPERIMENTAL 2.1 Materials and Supplies 2.1.1 Drugs, Metabolites and Internal Standards Propafenone hydrochloride, 5-hydroxypropafenone hydrochloride, L i -1115 hydrochloride (internal standard for PF quantitation), L i-1548 (internal standard for 5-hydroxypropafenone quantitation) and verapamil hydrochloride were the kind gifts of Knoll Pharmaceuticals Canada Inc. (Markham, Ontario, Canada). ASA, lidocaine, phenytoin sodium salt, quinidine sulphate, propranolol hydrochloride and indocyanine green (ICG) (90% pure) were purchased from Sigma Chemical Co. (St. Louis, MO, U.S.A.). Diazepam was obtained from the Vitamin and Chemical Divison, Hoffmann-La Roche Limited (Montreal, Canada). 2.1.2 Chemicals and Reagents Triethyl amine (TEA) (Sequanal™ grade) and heptafluorobutyric anhydride (HFBA) were obtained from Pierce Chemical Co. (Rockford, IL, U.S.A.). ACS reagent grade sodium hydroxide was purchased from Fisher Scientific Co. (Fair Lawn, NJ, U.S.A). ACS reagent grade monopotassium phosphate, disodium phosphate, sodium carbonate, sodium chloride and potassium carbonate were purchased from BDH Chemicals (Toronto, Ontario, Canada) while ACS reagent grade hydrochloric acid was obtained from American Scientific and Chemical (Seattle, WA, U.S.A.). 29 2.1.3 Proteins Purified human AAG, and albumin were purchased from Sigma chemical Co. (St. Louis, MO, U.S.A.). 2.1.4 Solvents Toluene, isopropyl alcohol (distilled in glass) and HPLC grade acetonitrile (ultraviolet (UV) cutoff 190 nm) were obtained from Caledon Laboratories Ltd. (Georgetown, Ontario, Canada). HPLC grade dichloromethane and methanol were obtained from BDH Chemicals (Toronto. Ontario, Canada). Deionized distil led water and HPLC grade water were obtained via a Milli-R0^ Water System (Millipore Corp., Bedford, MA, U.S.A.). 2.1.5 Gases Ultra high purity (UHP) hydrogen and argon/methane (95:5) were purchased from Matheson Gas Products Canada Ltd. (Edmonton, Alberta, Canada). Nitrogen, U.S.P. and medical air (breathing grade) were obtained from Union Carbide Canada Ltd. (Toronto, Ontario, Canada). 2.1.6 Equilibrium Dialysis Device Plexi-Glass^ dialysis cells (1.0 ml) were used for equilibrium dialysis. Cellophane dialysis membrane "sacks" (molecular weight cutoff of 12,000 Daltons) were obtained from Sigma Chemical Co. (St. Louis, U.S.A.). 30 2.1.7 Other Supplies Pyrex^ disposable screw-top glass culture tubes (10 ml),, with polytetrafluoroethylene (PTFE) lined screw caps were purchased from Canlab (Vancouver, B.C., Canada). Veinsystems™ Butterfly^-19 INT Cannula were purchased from Abbott Laboratories, Ltd. (Montreal, Canada). Vacutainer^ blood collection tubes were obtained from Becton Dickinson Canada Inc. (Mississauga, Ontario, Canada). Syringe sample f i l ter and C-18 extraction columns (100 mg/1 ml) were purchased from Phenomenex (Torrance, California, U.S.A.). 2.2 Columns 2.2.1 GLC Column A bonded-phase fused-silica capillary column, 25 X 0.31 mm I.D. was used for all GLC analyses. The stationary phase of the column was cross-linked 5% phenylmethylsilicone with a film thickness of 0.52 fim and a phase ratio of 150 (Hewlett-Packard, Palo Alto, CA, U.S.A.). 2.2.2 HPLC Column An ODS-Hypersil reverse phase column, 100 mm X 2.1 mm I.D. with s i l ica particle size of 5 fim (Hewlett Packard, Toronto, Ontario, Canada), was used for HPLC analyses of ICG. 31 2.3 E q u i p m e n t 2.3.1 G a s - L i q u i d C h r o m a t o g r a p h y GLC a n a l y s e s were p e r f o r m e d on a Model 5830A H e w l e t t - P a c k a r d (HP) g a s - l i q u i d c h r o m a t o g r a p h , e q u i p p e d w i t h a Model 18835B s p l i t / s p l i t l e s s c a p i l l a r y i n l e t s y s t e m , a ^ N i e l e c t r o n - c a p t u r e d e t e c t o r , a Model 18850A GC t e r m i n a l f o r c h r o m a t o g r a p h i c peak i n t e g r a t i o n a n d a Model 7671A a u t o m a t i c s a m p l e r . A s p l i t l e s s i n j e c t i o n mode u s i n g a f u s e d - s i l i c a i n l e t l i n e r was e m p l o y e d . T h e r m o g r e e n T ^ LB-2 s e p t a ( S u p e l c o , I n c . , B e l e f o n t e , PA, U.S.A.) c h a r a c t e r i z e d b y v e r y l o w - b l e e d a t h i g h i n l e t t e m p e r a t u r e s were u s e d . The s e p t u m was c h a n g e d r o u t i n e l y t o p r e v e n t l e a k a g e f o l l o w i n g r e p e a t e d p u n c t u r i n g d u r i n g a u t o m a t i c i n j e c t i o n s . 2.3.2 H i g h - P e r f o r m a n c e L i q u i d C h r o m a t o g r a p h y A HP Model 1090 l i q u i d c h r o m a t o g r a p h e q u i p p e d w i t h a HP Model 1040A d i o d e - a r r a y UV d e t e c t o r and a Model 310 HP c o m p u t e r f o r d a t a a n a l y s i s was u s e d . 2.3.3 S p e c t r o p h o t o m e t e r A H e w l e t t P a c k a r d 8452A d i o d e a r r a y s p e c t r o p h o t o m e t e r was u s e d f o r q u a n t i t a t i o n o f a l b u m i n . 32 2.3.4 Miscellaneous Other equipment used were: Eppendorf micropipettes, a vortex mixer (type 37600, Sybron), an incubation oven (isotemp, model 350, Fisher Scientific Industries, Springfield, MA, U.S.A.), an IEC model 2K centrifuge (Damon/IEC Division, Needham Hts., MA, U.S.A.), and a METTLER AE 163 balance (Mettler Instrumetente AG, CH-8606 Greifensee, Switzerland). 2.4 Preparation of Stock and Reagent Solutions 2.4.1 Drug, Metabolites and Internal Standards for GLC Analysis Propafenone hydrochloride was accurately weighed and dissolved in deionized disti l led water using sequential dilution to a final concentration of 100 ng/ml (11.07 mg of PF hydrochloride is equivalent to 10 mg of PF free base). 5-Hydroxypropafenone hydrochloride was accurately weighed and dissolved in methanol:deionized distil led water (1:9) using sequential dilution to a final concentration of 100 ng/ml (11.02 mg of 5-hydroxypropafenone hydrochloride is equivalent to 10 mg of 5-hydroxypropafenone free base). Li-1115 hydrochloride was accurately weighed and dissolved in deionized distil led water using sequential dilution to a final concentration of 200 mg/ml (11.11 mg of Li-1115 hydrochloride is equivalent to 10 mg of Li-1115 free base). 33 Li-1548 hydrochloride was accurately weighed and dissolved in deionized distilled water using sequential dilution to a final concentration of 200"ng/ml (11.06 mg of Li-1548 hydrochloride is equivalent to 10 mg of Li-1548 free base). All stock and working solutions were protected against light by wrapping the glass containers with aluminum f o i l . Solutions were stored at 4°C after preparation for up to a maximum of four months. 2.4.2 Drugs for Equilibrium Dialysis Propafenone and 5-hydroxypropafenone hydrochloride were accurately weighed and dissolved in pH 7.4 isotonic phosphate buffer (0.067 M and 0.1 M) to produce a concentration of 10 //g/ml for use in the protein binding displacement study. The solutions were stored at 4°C after preparation for up to 4 months. 5-Hydroxypropafenone hydrochloride (4.96 mg) was accurately weighed and dissolved in pH 7.4 isotonic phosphate buffer (0.1 M) to produce a stock solution of 45 //g/ml. Solutions ranging from 50-40,000 ng/ml were obtained from this stock solution using sequential dilution. These solutions were used for characterization of binding of 5-hydroxypropafenone to purified human AAG. Verapamil hydrochloride was accurately weighed and dissolved in pH 7.4 isotonic phosphate buffer to produce a concentration of 10 //g/ml (1.09 34 mg of verapamil hydrochloride is equivalent to 1 mg of verapamil free base). Propranolol hydrochloride was accurately weighed and dissolved in pH 7.4 isotonic phosphate buffer to a concentration of 10 //g/ml (1.14 mg of propranolol hydrochloride is equivalent to 1 mg of propranolol free base). Lidocaine (6 mg) was accurately weighed and dissolved in a few drops of 1 N HCl and diluted with pH 7.4 isotonic phosphate buffer to a concentration of 60 //g/ml. Quinidine sulfate was accurately weighed and dissolved in 0.1 M pH 7.4 isotonic phosphate buffer to a concentration of 50 //g/ml (5.755 mg of quinidine sulfate is equivalent to 5 mg of quinidine free base). These solutions (verapamil, propranolol, lidocaine and quinidine) were stored in light-tight containers at 4°C for up to 6 weeks. Aspirin was accurately weighed and dissolved in pH 7.4 isotonic phosphate buffer (containing either PF or 5-hydroxypropafenone) to a concentration of 400 //g/ml (20 mg aspirin was weighed). Solutions were prepared fresh prior to dialysis. Phenytoin sodium was accurately weighed and dissolved in pH 7.4 isotonic phosphate buffer containing PF or 5-hydroxypropafenone to a concentration of 18 //g/ml (1.956 mg of phenytoin sodium salt is equivalent to 1.8 mg of phenytoin free acid). 35 The final solutions for the displacement studies were prepared by diluting the respective stock solutions with pH 7.4 isotonic buffer. These solutions include PF combined with propranolol; 5-hydroxypropafenone with propranolol; PF with lidocaine; 5-hydroxypropafenone with lidocaine; PF with quinidine; 5-hydroxypropafenone with quinidine. In all the solutions used for displacement studies, the concentrations of PF and 5-hydroxypropafenone were 1 /zg/ml and 0.5 #g/ml, respectively. The concentrations of propranolol, verapamil, lidocaine and quinidine were 1.0, 0.2, 6.0 and 5 //g/ml, respectively. 2.4.3 Purified Protein Solutions AAG was accurately weighed and dissolved in pH 7.4 isotonic phosphate buffer to a final concentration of 90 mg/dL. Albumin was accurately weighed and dissolved in pH 7.4 isotonic phosphate buffer to a final concentration of 4.5 g/dL. 2.4.4 Drug and Internal Standard Solutions for HPLC analysis Indocyanine Green was accurately weighed and dissolved in HPLC grade methanol to a final concentration of 1 ug/m]. Diazepam was accurately weighed and dissolved in either HPLC grade methanol (I.S.a) or HPLC grade acetonitrile (I.S.b) to a concentration of 10 ug/m]. 36 2.4.5 Reagent Solutions Triethylamine (0.03%, v/v) was prepared by diluting TEA (75 a]) with toluene (250 ml). Four or five pellets of NaOH were added to this solution to keep the solution dry. Sodium hydroxide (NaOH) 1 M and 5 M solutions were prepared by dissolving NaOH pellets in deionized disti l led water. Sodium carbonate (Na2C03) 0.1 M solution and potassium carbonate (K2CO3) 5 M solution were prepared by dissolving Na2C03 and K2CO3 powder in deionized disti l led water. Hydrochloric acid (HC1) 1 M was prepared by diluting concentrated (37%) ACS reagent grade HC1 in deionized distil led water. Isotonic phosphate buffer (pH 7.4, 0.067 M) was prepared by accurately weighing 1.80 g of monopotassium phosphate (KH2PO4); 7.40 g disodium phosphate (Na2HP04); 4.20 g of sodium chloride (NaCl) and dissolving in deionized distil led water to a final volume of 1 l i ter. Isotonic phosphate buffer (pH 7.4, 0.1 M) was prepared by dissolving 1.8 g of NaCl, 11.4 g of Na2HP04 and 2.7 g of KH2P04 in deionized distilled water to a final volume of 1 l i ter . Phosphate buffer (pH 6.0, 0.067 M) was prepared as follows: 4.70 g of 37 Na2HP04 and 4.60 g of KH2PO4 were accurately weighed and dissolved in deionized disti l led water to a final volume of 500 ml, respectively. The final buffer solution was obtained by combining these two solutions in a ratio of 9:1 (KH2P04 : Na2HP04). Phosphate buffer (pH 6.0, 0.005 M) was prepared by dissolving 0.605 g of KH2P04 and 0.079 g of Na2HP04 to one l i ter of HPLC grade water. Citrate buffer (pH 4.0, 0.005 M) was prepared by dissolving 0.596 g of c i tr ic acid and 0.270 g of disodium phosphate in one l i ter of HPLC grade water. Citrate buffer (pH 4.4, 0.005 M) was prepared by dissolving 0.52 g of c i tr ic acid and 0.62 g of disodium phosphate in one l iter of HPLC grade water. 2.5 Equilibrium Dialysis Procedure The equilibrium dialysis was performed in the following manner: The cellophane dialysis membrane was boiled in distilled water for 1 hour and then soaked in phosphate buffer (pH 7.4) for a minimum of 1 hour, maximum of 2 days (kept at 4°C) before mounting into the dialysis cells. Extreme care was taken in order not to touch the surface of the membrane during the process. Equal volumes (0.8 ml) of drug (in isotonic phosphate buffer) and study fluid (protein solution, serum or blood) were added to the respective sides of the dialysis cells, separated by the cellophane membrane. The unit was dialysed at 37°C for 8 hour to reach equilibrium. During 38 equilibrium dialysis, the cell was rotated automatically by a motor rotator at a speed of 14 rpm. Before and after equilibrium dialysis, both pH and volume of the sample were measured in order to examine volume shift or pH change. Aliquots of buffer and biological fluid were then subjected to GLC-ECD analyses. 2.6 Protein Binding of 5-Hydroxypropafenone to Purified Human AAG The characterization of the protein binding of 5-hydroxypropafenone to purified human AAG was studied by dialyzing various concentrations of 5-hydroxypropafenone against the solution of AAG. 5-Hydroxypropafenone and AAG were dissolved in pH 7.4 isotonic phosphate buffer (0.1 M). The concentrations of 5-hydroxypropafenone used were: 50, 100, 150, 200, 500, 1,000, 5,000, 10,000, 20,000, 30,000, 40,000, 45,000 ng/ml. Five replicate measurements were made at each concentration. 2.7 Rosenthal Analysis of the Binding of 5-Hydroxypropafenone to Purified Human AAG The characterization of the binding parameters of 5-hydroxypropafenone to purified AAG were examined using Rosenthal analysis (Rosenthal, 1967). The molar concentration of bound 5-hydroxypropafenone was plotted against the concentration ratio of bound to free. 2.8 Scatchard Plot of 5-Hydroxypropafenone Binding to AAG A Scatchard plot was also used to interpret the binding data, where 39 the moles of drug bound per mole of protein (r') was plotted against [r'/free drug concentration] (Scatchard, 1949). From this plot, the classes of binding sites can be obtained. 2.9 Free Fraction vs Total Concentration Plot of 5-Hydroxypropafenone Binding to AAG The 5-hydroxypropafenone binding data was further treated by plotting the free fraction vs total concentration of 5-hydroxypropafenone. The free fraction is the concentration ratio of free drug to total drug. Both total and free drug concentration were obtained by GLC analysis of 5-hydroxypropafenone in the AAG solution and buffer solution after equilibrium dialysis. 2.10 Protein Binding Displacement Studies In the protein binding displacement studies, PF and 5-hydroxypropafenone were dissolved in phosphate buffer, respectively, and spiked with various displacing agents or their phosphate buffer solutions subsequently. The combined solution was then dialysed against various biological fluids (purified AAG, albumin, serum or whole blood). The dialysis was performed in either three or four replicates. 2.10.1 Effect Of Buffer Strength on the Protein Binding of Propafenone and 5-hydroxypropafenone The influence of ionic strength of phosphate buffer on the binding of 40 PF and 5-hydroxypropafenone to human AAG, albumin, or whole blood was studied at buffer concentration of 0.067 M and 0.1 M. The free fraction of PF and 5-hydroxypropafenone at each buffer concentration was determined and compared. 2.10.2 Effect of Anticoagulant on the Protein Binding of Propafenone and 5-hydroxypropafenone The effect of anticoagulant on the protein binding of PF and 5-hydroxypropafenone in whole blood was determined as follows: Whole blood containing either heparin or EDTA was dialysed against pH 7.4 isotonic phosphate buffer solutions containing PF or 5-hydroxypropafenone. Propafenone or 5-hydroxypropafenone concentrations were measured by GLC analyses. The free fraction was calculated by dividing the free concentration of PF or 5-hydroxypropafenone by the corresponding total concentration. The effect of anticoagulant on the protein binding was observed by comparing the free fraction of PF and 5-hydroxypropafenone in heparinized whole blood to those obtained by using EDTA. 2.10.3 Protein Binding Displacement Study Using Human AAG and Albumin as the Biological Fluids The displacing agents included: ASA, propranolol, verapamil, phenytoin, quinidine and lidocaine. In those studies involving ASA, propranolol and verapamil, the concentration of phosphate buffer was 0.067 M, other displacement studies were conducted at a buffer concentration of 0.1 M. The free fractions of PF and 5-hydroxypropafenone after 41 displacement were compared to those obtained without addition of the various displacing agent. 2.10.4 Protein Binding Displacement Study Using Human Serum and Whole Blood as the Biological Fluids Blood (the blood was anticoagul ated with heparin and EDTA, respectively) and serum were obtained from healthy, young, male Caucasian volunteers. The heparinized whole blood was collected using the green top Vacutainers^ which contain 143 USP units of Lithium heparin per tube. The whole blood anticoagulated with EDTA 'was collected using lavender top VacutainersR which contain 0.048 ml of 15% EDTA (K3) (7.2 mg) and 0.01 mg of potassium sorbate. When human serum was used as the biological fluid, the displacers used were: phenytoin, ASA and quinidine for 5-hydroxypropafenone and phenytoin, quinidine and lidocaine for PF. When whole blood was used as the biological fluid, the displacers used were quinidine and phenytoin for PF. The free fractions were compared before and after the addition of each displacing agent. 2.10.5 Characterization of the Nature of Binding Displacement Interaction between PF and Lidocaine and PF and Quinidine in Purified Human AAG. 2.10.5.1 Binding of PF to AAG over the Therapeutic Concentration Range of PF This study was carried out at PF concentrations of 0.1, 0.2, 0.4, 0.6, 0.8, 1.0, 1.5 jug/ml. The binding of PF alone, PF spiked with 42 lidocaine (6 //g/ml) or quinidine (5 /zg/ml), were dialysed against an equal volume of AAG solution. 2.10.5.2 Characterization of the Nature of the Displacement Interaction: Competitive or Non-competitive The binding data obtained from the above three studies were analyzed by plotting the reciprocal of the moles of bound drug per mole of protein versus the reciprocal of the concentration of unbound drug (moles). The nature of the displacement was then determined from the plot (Webb, 1963). 2.10.6 Measurement of Albumin Concentration in Purified Albumin Solution The concentration of albumin in purified albumin solutions was measured by the method of Lowry, et al. (1951) using bovine serum albumin as standard protein. The analytical wavelength used was 650 nm. The concentration of albumin used for the determination of PF and 5-hydroxypropafenone free fraction in 0.067 M and 0.1 M isotonic phosphate buffer were 3.80 and 3.90 g/dL, respectively. The concentration of albumin used for the study of the displacement effects of ASA, propranolol and verapamil on the free fraction of PF and 5-hydroxypropafenone is 3.64 g/dL. 2.11 Analysis of Propafenone and 5-Hydroxypropafenone The concentration of PF and 5-hydroxypropafenone in both buffer and biological fluid compartments of the dialysis cell were analyzed by the GLC-ECD methods of Chan et a7., 1987; 1989. 4 3 2.12 HPLC Analysis of Indocyanine Green (ICG) 2.12.1 Reproduction of Reported HPLC ICG Assay Methods Rappaport and Thiessen reported the first HPLC analysis method for ICG (1982). The method involved precipitation of plasma samples with acetonitrile in a volume ratio of 1.6:1 (acetonitrile:plasma). After centrifugation at 1,700 X g for 2 min, the supernatant was injected into the HPLC and detected by UV absorption at 254 nm. The detection limit was 0.4 /xg/ml using 0.25 ml of plasma. However, there was no report about the volume injected and the on-column detection limit was not known. In order to increase the assay sensitivity, two other HPLC methods with fluorescence detection were also tried on our system. one fluorometric determination method tried was reported by Hoi 1 ins et a7.(1987). The sample preparation procedure and the chromatographic conditions were the same as that of Rappaport and Thiessen. The excitation emission maxima were 780 nm and 810 nm, respectively. An excitation spectrum and an emission spectrum at an excitation wavelength of 260 nm were obtained within our fluorometric wavelength range. The detection of ICG was tried at various excitation and emission wavelength pairs. The method of Dorr and Pollack (1989) was also tried on our system. They utilized the same sample preparation procedure as that of Rappaport and Thiessen and the mobile phase was composed of pH 4.0 phosphate buffer (containing 1% triethylamine) and acetonitrile in a ratio of 2:1). The excitation and emission wavelengths were 214 nm and 370 nm, respectively. 44 The on-column detection limit was 3 ng with 20-200 //l injected. An excitation spectrum at these experimental conditions was obtained and detection of ICG at the reported excitation and emission wavelength were attempted.-2.12.2 Modification of the Analysis Method of Indocyanine Green of Christie (1986) and Burns (1989) The phase separation method with ammonium sulphate reported by Christie et al. (1986) was used in our first assay of ICG. A slight modification of the procedure reported by Burns et al. (1989) was applied in our sample preparation. The following procedure was used for the preparation of a standard curve: Standard curve range was from 0.1 to 1 fig (0.1, 0.2, 0.4, 0.6, 0.8 and 1.0 ug). The corresponding volumes of standard ICG solution (1 /ig/ml) were added to each tube. The solution was evaporated in the water bath of 40°C under After evaporation, 100 /zl of serum was added to each tube and set for half an hour with frequent vortexing to get complete protein binding. Saline (0.35 ml, diazepam solution (75 fi\) (I.S.a) and acetonitrile (0.35 ml) were then added to the samples in sequence. The mixture was then vortexed and centrifuged at 2,500 X g for 10 min to precipitate protein. The supernatant was transferred and mixed with ammonium sulphate, the mixture was again vortexed and centrifuged. The organic phase was then passed through a syringe sample f i l ter , the filtrate (100 fi\) was diluted with 100 u} of pH 6.0 phosphate buffer, and 80 /zl of this solution was injected into the HPLC. The mobile phase contained 40 % 45 of acetonitrile and 60 % of pH 6.0 phosphate buffer and methanol in a ratio of 50:4. The ICG and diazepam (internal standard) was measured by diode array detection (DAD) at 216 nm. A scheme of the procedure is shown in Fig 5. ICG methanol standard solution evaporation under No ,in water bath (30°C) ICG residue 0.1 ml serum vortexing for 30 min ICG serum solution 0.35 ml saline 75 a\ diazepam solution(I.S.a) 0.35 ml acetonitrile vortexing centrifuging (10 min) discard precipitant solvent layer ammonium sulphate vortexing centrifuging discard aqueous layer organic layer | syringe sample filtration fi ltrate I diluted with equal volume of pH 6.0 phosphate buffer t inject 80 /il to HPLC Figure 5. A scheme of extraction procedure of ICG in human serum 4 6 2.12.3 Comparison of the Chromatograph of Serum Spiked with Indocyanine Green with that of Serum Blank The chromatograph of serum spiked with ICG and serum blank (volume 0.25 ml) were obtained following the same preparation procedure and compared. 2.12.4 Modification of the Chromatographic Conditions to Minimize the Blank Interference The modifications involved were: changing the composition of mobile phase; changing pH 6.0 phosphate buffer in the mobile phase to pH 4.0 citrate buffer; changing the flow rate. None of these changes could completely eliminate interference peaks in the chromatogram, but changing the mobile phase composition from 40 % to 39 % of acetonitrile and flow rate from 0.5 ml/min to 0.2 ml/min gave the best minimization of blank interference. These HPLC conditions were then used for studying extraction efficiency and post extraction stability. 2.12.5 Extraction Efficiency Different concentrations of ICG methanol solutions were made (40, 32, 24 and 16 /zg/ml). Each of these solution (25 fi\) was added to each of the two sets of tubes. To one set, 0.25 ml serum was added to each tube followed by extraction with acetonitrile. Immediately before filtration, internal standard solution (60 jul) was added to each tube. To the set devoid of serum, methanol (40 ji\) and internal standard solution (60 //l) 47 were added. The samples were then subjected to the same dilution procedure. The amounts of ICG recovered was calculated from the calibration curve of ICG in methanol and compared with the amount added. 2.12.6 Post Extraction Stability Aliquots of the extracted serum sample were stored in black vial holders and injected at varies times (10 times, 48 h) after preparation. Peak height ratios (ICG/diazepam) and chrbmatograms obtained after each injection were used as indicator of ICG stability. Constant peak height ratio and chromatogram indicated that there were no ICG degradation products and ICG was stable up to that injection time. 2.12.7 Animal Preparation In order to test the ut i l i ty of the assay method to analyze the plasma concentration of ICG, a i.v. injection was made into a rat (Sprague-Dawley). The rat was anesthetized with methohexital sodium ( i .p . ) . Two cannulas (PE50) were surgically implanted: jugular vein and carotid artery. After the rat regained consciousness, ICG solution (0.3 ml, 5 mg/ml) was injected to the jugular vein followed by a catheter flush with 0.3 ml heparin saline (100 iu/ml). Blood (0.2-0.3 ml) was taken from the carotid artery at 1, 3, 5, 8, 10, 15, 30, 45 and 60 min after injection. The blood samples were centrifuged at 2,000 X g for 10 min and plasma was immediately collected and ICG concentration analyzed the following day. 48 2.12.8 Analysis of Plasma Samples Due to the small volume of plasma sample obtained from rat (100 fi]), the assay had to be carried out at microassay level. SIickseal micro-centrifuge tubes were used for sample preparation. As the centrifuge in our laboratory did not have the appropriate rotor to hold these small vials, another thermostatic centrifuge was used. The maximum centrifugal force in this centrifuge was 17,000 X g. The plasma sample was deproteinized by the procedure described before. After centrifugation at 17,000 X g for 3 min, the supernatant was clear such that direct injection on to the HPLC was possible. In order to detect low concentrations, instead of diluting the sample with normal saline (followed by deproteinization with acetonitrile), as before, the sample was precipitated with acetonitrile directly. The supernatant (10 fi\) was injected into the HPLC. The mobile phase was the same as before, but with a flow rate of 0.5 ml/min. A scheme of the sample preparation procedure is shown in Figure 6. A standard curve was prepared using peak height ratio, the plasma concentration of the samples were determined from this standard curve. 49 ICG plasma sample at various time after ICG injection discard precipitant ICG standard methanol solution Jevaporate under N 2 at room temp ICG residue 50 fi\ blank rat plasma ICG plasma solution / 80 fi\ acetonitrile contain diazepam (I.S.b) centrifuge at 17,000 X g, 3 min inject 10 fi\ supernatant to HPLC Figure 6. A scheme of the extraction procedure of ICG in rat plasma 2.12.9 The Potential to Detect Low Concentration Samples with the New Method One possible way to increase sensitivity for samples of low concentration is to increase the injection volume. This approach, however, leads to higher background contributions to the chromatogram which can possibly interfere with drug measurement. With this new method increasing the injection volume from 10 to 50 /zl was tried. 2.12.10 Solid Phase Extraction Method Donn et a7.(1984) used solid phase extraction to prepare samples for 5 0 HPLC assay of ICG. Plasma samples were first deproteinized by acetonitrile, diluted with pH 4.4 citrate buffer, and then loaded onto a C18 extraction column (100 mg/lml). After washing with deionized distilled water twice, ICG was eluted by methanol and injected into the HPLC. The same method was tried on our system. 2.13 Data Analysis 2.13.1 Statistical Analysis The free fractions of PF and 5-hydroxypropafenone after spiked with various drugs were compared to the corresponding control values. For comparison of two treatment groups, two sample t-test at a significance level of 5% (p<0.05) was used (Zar, 1984). For comparison of multiple treatment groups, two sample t-test combined with Bonferroni inequality at p < 0.05 was used (Meddis, 1984). 2.13.2 Protein Binding Data For both the Rosenthal and Scatchard plots, the computer programs AUT0AN (Sedman and Wagner, 1976) and N0NLIN (Metzler, 1974) were used to obtain the best f i t of the data. Linear regression was used to test the linearity of the data. 2.13.3 Indocyanine Green Plasma Clearance Indocyanine plasma concentration-time profile was analyzed by AUT0AN 51 and two compartment model was the best f i t for the data. The AUC™ was calculated by trapezoidal method and the plasma clearance was obtained using the following equation: CI = Dose / AUC°°. 52 3.RESULTS 3.1 Volume Shift and pH Change of Equilibrium Dialysis Samples Both volume and pH of samples on each side of the dialysis cell were measured before and after dialysis. No significant volume shift and pH change was observed in the dialysate using the two sample t-test at p < 0.05. The volume shift was within 5% of the original volume (0.80 ml, the volume of the fluid on each side of the dialysis cell before dialysis) and the pH change was within 1.3% of the original pH (7.40). 3.2 Binding of 5-Hydroxypropafenone to Purified Human AAG 3.2.1 Rosenthal Analysis The binding ratio (bound concentration/free concentration) of 5-hydroxypropafenone in purified human AAG was plotted versus the bound concentration using the method of Rosenthal (1967) (Figure 7). A linear relationship was found with r = -0.951. A single class of binding sites was observed with a intrinsic association or affinity constant of K = 1.614 X 10^  M~l. The estimation of the affinity constant was also obtained by using the computer program EnzfitterR (Leatherbarrow, 1986). The affinity constant obtained this way was 1.531 X 10^  M" 1 . 3.2.2 Scatchard Plot of 5-Hydroxypropafenone Binding to Human AAG The binding data plotted as r' (the moles of drug bound per mole of Q) 3.00 2.50 2.00 1.50 c o 1.00 .00. 0.50 0.00 i CO o 15.00 12.00 L i n e a r r e g r e s s i o n Y=-0.161X---+ 2.209 r=-0.951 o c 8 6.00 h C Z3 m 3.00 0.00 * 0.00 10.00 20.00 30.00 40.00 50.00 Total cone. M X 10 6 60.C 0.00 3.00 6.00 9.00 [Bound] X 10 6 12.00 15.00 F i g u r e 7. R e l a t i o n s h i p between the r a t i o o f bound c o n c e n t r a t i o n / f r e e c o n c e n t r a t i o n and bound c o n c e n t r a t i o n o f 5- h y d r o x y p r o p a f e n o n e in p u r i f i e d human AAG ( R o s e n t h a l p l o t ) , t h e i n s e r t e d graph shows the r e l a t i o n s h i p between bound and t o t a l c o n c e n t r a t i o n o f 5-hydroxypropafenone. 5 4 protein) versus the ratio of r'/free 5-hydroxypropafenone concentration, is shown in Figure 8. The concentration of AAG (M.W. = 44,100) was used in the calculation of r' where a linear relationship was found with a r = -0.935. One type of binding site was observed over the concentration range studied. 3.2.3 Free Fraction versus Total Concentration Plot The free fraction of 5-hydroxypropafenone vs total concentration of 5-hydroxypropafenone is shown in Figure 9. At lower concentrations (50-800 ng/ml), the binding was independent of concentration with a mean free fraction of 0.3066 ± 0.025 (CV. = 8 %). At higher concentrations (2000 to 20,000 ng/ml), binding was found to be dependent on concentration and a linear relationship between free fraction and total concentration was observed (r = 0.9644). 3.3 Protein Binding Displacement Effect 3.3.1 Effect of Buffer Strength on the Binding of PF and 5-Hydroxypropafenone Table 2 shows the free fraction of PF in a solution of purified human AAG or albumin dissolved in varying ionic strengths of phosphate buffer, the pH of which was 7.4. In purified human AAG solution, the mean free fraction in 0.067 M phosphate buffer was 0.075 ± 0.01, which was significantly greater than the mean free fraction in 0.1 M phosphate buffer of 0.057 ± 0.004 (two sample t-test with p < 0.05). In purified human 180.00 r 150.00 120.00 -Q 9 0 . 0 0 60.00 30.00 -0.00 0.00 150.00 300.00 450.00 r X 10 6 600.00 750.00 F i g u r e 8. R e l a t i o n s h i p between [ r ' / b r j u n d c o n c e n t r a t i o n ] and r ' o f 5 - h y d r o x y p r o p a f e n o n e i n p u r i f i e d human AAG ( S c a t c h a r d p l o t ) ; c o n c e n t r a t i o n o f AAG was u s e d i n t h e c a l c u l a t i o n o f [ > ' ] • c n Figure 9. Relationship between the ratio of free concentration/total concentration (free fraction) and total concentration of 5 -hydroxypropafenone over the concentration of 20-20000 ng/ml. Table 2. Free fraction of PF in a solution of AAG or albumin dissolved in 0.067 M or 0.1 M phosphate buffer sample PF free fraction PF free fraction in AAG in albumin buffer cone. 0.067M 0.1M 0.067M 0.1M 1 0.089 0.055 0.427 0.339 2 0.079 0.054 0.421 0.344 3 0.064 0.064 0.457 0.344 4 0.076 - 0.421 0.330 mean 0.075* 0.057* 0.431* 0.339* ± s.d. ± 0.01 ± 0.004 ± 0.017 0.006 C.V.% 14.1 6.5 4.0 1.9 albumin 3.8 g/dL 3.9 g/dL cone. concentration of AAG: 90 mg/dL initial cone, of PF: 1 /zg/ml pH of phosphate: 7.4 *: statistically significant differences with two sample t-test at p < 0.05 58 albumin solution, the mean free fraction in 0.067 M and 0.1 M phosphate buffer were 0.431 ± 0.017 and 0.339 ± 0.006, respectively. The free fraction in 0.067 M phosphate buffer was significantly greater than in 0.1 M phosphate buffer (two sample t-test with p < 0.05). Table 3 shows the free fraction of 5-hydroxypropafenone in a solution of purified human AAG or albumin dissolved in either 0.067 M or 0.1 M phosphate buffer (pH = 7.4). In purified human AAG solution, the mean free fraction in 0.067 M phosphate buffer was 0.321 ± 0.029, which was not different from the mean free fraction in 0.1 M phosphate buffer of 0.321 ± 0.028. In purified human albumin solution, the mean free fraction in 0.067 M phosphate buffer of 0.271 ± 0.017 was significantly lower than the mean free fraction of 0.354 ± 0.013 in 0.1 M phosphate buffer (two sample t-test with p < 0.05). In summary, buffer strength did affect the binding of PF in purified human AAG and albumin solutions and 5-hydroxypropafenone in purified human albumin, but had no apparent effect on the binding of 5-hydroxypropafenone in purified human AAG solution. Therefore, comparison of the effect of different displacing agents on the binding of PF and 5-hydroxypropafenone had to be conducted at the same phosphate buffer concentrations. 3.3.2 Effect of Anticoagulant on the Protein Binding of PF and 5-Hyd roxypropafenone Table 4 shows the free fraction of PF in whole blood using heparin and EDTA as anticoagulants. PF was dissolved in 0.067 M and 0.1 M isotonic Table 3. Free fraction of 50HPF in a solution of AAG or albumin dissolved in 0.067 M or 0.1 M phosphate buffer sample 50HPF free fraction 50HPF free fraction in AAG in albumin buffer cone. 0.067M 0.1M 0.067M 0.1M 1 0.306 0.308 0.256 0.339 2 0.302 0.351 0.302 0.360 3 0.354 0.338 0.354 0.363 4 - 0.289 - -mean 0.326 0.321 0.271* 0.354* ± s.d. ± 0.029 ± 0.028 ± 0.017 0.013 C.V.% 9.0 8.9 6.3 3.7 albumin 3.8 g/dL 3.9 g/dL cone. concentration of AAG: 90 mg/dL initial cone, of 50HPF: 0.5 //g/ml pH of phosphate: 7.4 *: statistically significant differences with two sample t-test at p < 0.05 Table 4. Free fraction of PF in whole blood using heparin or EDTA as the anticoagulant sample PF free fraction PF free fraction in "heparin" blood in "EDTA" blood buffer cone. 0.067M 0.1M 0.067M 0.1M 1 0.084 0.069 0.048 0.040 2 0.069 0.064 0.048 0.048 3 0.083 0.051 0.049 0.056 4 - 0.043 0.054 0.044 mean 0.079* 0.057 0.050* 0.047 ± s.d. ± 0.008 ± 0.012 ± 0.003 0.007 C.V.% 10.7 20.9 5.8 14.6 initial cone, of PF: 1 //g/ml pH of phosphate: 7.4 *: statistically significant difference with two sample t-test at p < 0.05 61 phosphate buffer (pH7.4), respectively. In 0.067 M phosphate buffer, the mean free fractions were 0.078 ± 0.008 in heparinized whole blood and 0.050 ± 0.003 in whole blood using EDTA as the anticoagulant. The two values were significantly different from each other (two sample t-test with p < 0.05). In 0.1 M phosphate buffer, the mean free fraction in heparinized whole blood was 0.057 ± 0.012, and was not significantly different from the value of 0.047 ± 0.007, the free fraction in whole blood using EDTA as the anticoagulant (two sample t-test with p < 0.05). However, there was a tendency for the free fraction in heparinized blood to be higher than in whole blood when EDTA was used as the anticoagulant. Table 5 shows the free fraction of 5-hydroxypropafenone in whole blood using either heparin or EDTA as the anticoagulant. 5-Hydroxypropafenone was dissolved in 0.067 M and 0.1 M phosphate buffer before dialysis, respectively. In 0.067 M phosphate buffer, the mean free fraction in heparinized whole blood was 0.174 ± 0.012 which was not significantly different from a mean free fraction of 0.163 ± 0.01 in whole blood using EDTA as the anticoagulant. In 0.1 M phosphate buffer, the mean free fraction in heparinized whole blood was 0.162 ± 0.011 and was not different from mean free fraction in whole blood using EDTA as an anticoagulant (0.163 ± 0.007, two sample t-test with p < 0.05). In conclusion, the binding of PF in whole blood was affected by the use of anticoagulant while the binding of 5-hydroxypropafenone was not. Therefore, comparision of the effect of binding displacers in whole blood must be conducted using the same anticoagulant. Table 5. Free fraction of 50HPF in whole blood using heparin or EDTA as the anticoagulant sample 50HPF free fraction 50HPF free fraction in "heparin" blood in "EDTA" blood buffer cone. 0.067M 0.1M 0.067M 0.1M 1 0.156 0.171 0.184 0.149 2 0.156 0.155 0.183 0.174 3 0.161 0.161 0.169 0.170 4 0.178 0.166 0.159 0.156 mean 0.163 0.163 0.174 0.162 ± s.d. ± 0.01 ± 0.007 ± 0.012 0.011 C.V.% 6.4 4.3 6.9 7.2 initial cone, of 50HPF: 0.5 ^g/ml pH of phosphate: 7.4 *: statistically significant difference with two sample t-test at p < 0.05 63 3.3.3 Protein Binding Displacement Studies using Purified Human Protein Solutions (AAG and Albumin) as the Biological Fluids 3.3.3.1 Protein Binding Displacement Effect of Various Drugs on PF Binding to Purified Human AAG Neither propranolol, verapamil nor ASA displace PF from its binding sites as shown in Figure 10. Similarly, phenytoin and quinidine did not displace PF from its binding sites, however, when lidocaine was used as the displacing agent, statistically significant increases in PF free fraction were observed (Figure 11, two sample t-test combined with Bonferroni inequality at p < 0.05). The PF free fraction seemed to increase with the addition of ASA, although the apparent difference did not reach statistical significance due to the large variation in the free fraction in the solution of PF spiked with ASA. The maximum displacing effect was observed with lidocaine (124% increase in PF free fraction). 3.3.3.2 Protein Binding Displacement Effect of Various Drugs on the Binding of PF to Purified Human Albumin Similar to the effects observed in AAG solution, propranolol, verapamil and ASA did not cause a statistically significant increase in PF free fraction (Figure 12, two sample t-test combined with Bonferroni inequality at p < 0.05). Phenytoin and lidocaine caused statistically significant increases in PF free fraction (Figure 13). Although the increase in PF free fraction caused by quinidine seemed to be quite large (70%, Figure 13), it failed to reach statistical significance due to the 64 PF free fraction in pure AAG solution < 0.075 ± 0.01 (n-4) -s CD c+ a CO Z3 c+ cn -i o c X J P F + P r o p r a n o l o l »- H 0.091 ± 0.005 (n-3) 0.088 ± 0.008 (n=3) P F + ASA 0.095 ± 0.024 (n=4) Figure 10. Protein binding displacing effect of propranolol, verapamil and ASA on the free fraction of PF in purified human AAG dissolved in 0.067 M phosphate buffer. The pH in each treatment group is 7.4, the value is the mean ± s.d. and n is the number of replicates. 65 PF free fraction in pure AAG solution P F 0.057 ± 0 . 0 0 4 (n-3) s CD W r+ 3 n> =3 P F + Phenytoim- -* 0.083 ± 0.016 (n-4) -5 O c X3 P F + Quinidine ^ 0.105 ± 0.016 (n-3) P F + L idocaine 0.129 ± 0.013* (n-4) Figure 11. Protein binding displacing effect of phenytoin, quinidine and lidocaine on the free fraction of PF in purified human AAG dissolved in 0.1 M phosphate buffer. The pH in each treatment group is 7.4 and the value is the mean ± s.d., * indicates statistically significant difference with two sample t-test combined with Bonferroni inequality at p < 0.05, n is the number of replicates. 66 PF free fraction in pure albumin solution PF ^ 0.431 ± 0.017 (n=4) CD 0» r + . 3 CD =3 r + CD -5 O c •a PF + Propranolol H 0.452 ± 0.038 (n=4) 0.405 ± 0.015 (n=3) P F + A S A H0.443 ± 0.047 (n=4) Figure 12. Protein binding displacing effect of propranolol, verapamil and ASA on the free fraction of PF in purified human albumin dissolved in 0.067 M phosphate buffer. The pH in each treatment group is 7.4, the value is the mean ± s.d. and n is the number of replicates. 67 PF free fraction in pure albumin solution. PF V 0.339 ± 0.006 (n-4) PF + Phenytoin H 0.649 ± 0.055* (n=3) PF + Quinidine -«0.577 ± 0.1824 (n=4) PF + Lidocaine +0.333 ± o.oos* (n=4) Figure 13. Protein binding displacing effect of phenytoin, quinidine and lidocaine on the free fraction of PF in purified human albumin dissolved in 0.1 M phosphate buffer. The pH in each treatment group is 7.4 and the value is the mean ± s.d., * indicates statistically significant difference with two sample t-test combined with Bonferroni inequality at p < 0.05, n is the number of replicates. 68 large variation within the group. The maximum displacing effect was observed with phenytoin (91 % increase), followed by lidocaine (13%). 3.3.3.3 Protein Binding Displacement Effect of Various Drugs on the Binding of 5-hydroxypropafenone to Human AAG 5-Hydroxypropafenone was dissolved in a phosphate buffer solution and subsequently spiked with various displacing agents (Figures 14 and 15) before dialysis against a purified human AAG solution; the free fraction of 5-hydroxypropafenone increased significantly by 47%, in lidocaine treatment group (two sample t-test at p < 0.05, Figure 15). Verapamil, ASA and propranolol did not cause any significant changes in 5-hydroxypropafenone free fraction (Figure 14, two sample t-test combined with Bonferroni inequality at p < 0.05). 3.3.3.4 Protein Binding Displacement Effect of Various Drugs on the Binding of 5-hydroxypropafenone to Human Albumin ASA significantly displaced 5-hydroxypropafenone from its binding sites on purified human albumin by 47% (Figures 16, two sample t-test combined with Bonferroni inequality at p < 0.05). Lidocaine (Figure 17), verapamil and propranolol (Figure 16) did not cause any statistically significant changes in 5-hydroxypropafenone free fraction (two sample t-test combined with Bonferroni inequality at p < 0.05). 69 50HPF free fraction in pure AAG solution 5 0 H P F -<0.321 ± 0.029 (n-3) 50HPF+VerapaTTTi -"0.304 ± 0.068 (n=3) 5 0 H P F + ASA 0.476 ± 0.087 (n=3) 5 0 H P F + P r o p r a n o l o l -t0.355 ± 0.029 (n=3) Figure 14. Protein binding displacing effect of verapamil, ASA and propranolol on the free fraction of 5-hydroxypropafenone in purified human AAG dissolved in 0.067 M phosphate buffer. The pH in each treatment group is 7.4, the value is the mean ± s.d. and n is the number of replicates. 70 50HPF free fraction in pure AAG solution -s CD & r + 3 CD c+ CO ~i o cz -o 50HPF -*0.321 ± 0.028 (n-4) 50HPF + Lidocaine ^ 0.472 ± 0.03* (n-4) Figure 15. Protein binding displacing effect of lidocaine on the free fraction of 5-hydroxypropafenone in purified human AAG dissolved in 0.1 M phosphate buffer. The pH in each treatment group is 7.4 and the value is the mean ± s.d., * indicates statistically significant difference with two sample t-test at p < 0.05, n is the number of replicates. 71 50HPF free fraction in pure albumin soluti on CO r + 3 ro -s o c -o 5 - O H P F H0.271 ± 0.017 (n-3) 5 0 H P F + V e r a p a m i l 0.281 ± 0.045 (n-3) 5 0 H P F + A S A H0.398 ± 0.028* (n-3) 5 0 H P F + P r o p r a n o l o l 0.304 ± 0.01 (n-3) Figure 16. Protein binding displacing effect of verapamil, ASA and propranolol on the free fraction of 5-hydroxypropafenone in purified human albumin dissolved in 0.067 M phosphate buffer. The pH in each treatment group is 7.4 and the value is the mean ± s.d., * indicates statistically significant difference with two sample t-test combined with Bonferroni inequality at p < 0.05 and n is the number of replicates. 50HPF free fraction in pure albumin solution 50HPF 0.354 ± 0.013 (n-3) 50HPF + L i d o c a i n e 4 * ° - 3 4 2 ± o.oi (n-4) 17. Protein binding displacing effect of lidocaine on the free fraction of 5-hydroxypropafenone in purified human albumin dissolved in 0.1 M phosphate buffer. The pH in each treatment group is 7.4, the value is the mean ± s.d. and n is the number of replicates. 73 3.3.4 Protein Binding Displacement Effect Using Human Serum as the Biological Fluid In human serum, quinidine, phenytoin and lidocaine significantly increased the free fraction of PF as shown in Figure 18 (two sample t-test combined with Bonferroni inequality at p < 0.05). The maximum increase in the free fraction was observed with lidocaine (580%), followed by quinidine (210%) and phenytoin (150%). Figure 19 demonstrates that neither phenytoin, ASA nor quinidine displaces 5-hydroxypropafenone from its binding sites in human serum (two sample t-test combined with Bonferroni inequality at p < 0.05). 3.3.5 Protein Binding Displacement Effect on PF Free Fraction using Human Whole Blood as the Biological Fluid Since quinidine and phenytoin caused a significant increase in the free fraction of PF in serum, a further study was performed to determine the displacing ability of these two agents in whole blood. Statistically significant increase in PF free fraction was observed in quinidine treatment group and phenytoin did not cause statistically significant increase in PF free fraction (Figure 20, two sample t-test combined with Bonferroni inequality at p < 0.05). PF free fraction in serum P F 0.017 ± 0.003 (n=4) P F + Q u i n i d i r V e - H0.052 ± 0.009* (n=4) P F + D P H HO.041 ± 0.007* (n=4) P F + L i d o c a i n e nO.114 ± 0.006* (n=4) Figure 18. Protein binding displacing effect of DPH (phenytoin) quinidine and lidocaine on the free fraction of PF in human serum. The pH in each treatment group is 7.4 and the value is the mean ± s.d. , * indicates statistically significant difference with two sample t-test combined with Bonferroni inequality at p < 0.05 and n is the number of replicates. 75 50HPF free fraction in serum 50HPF HO.189 ± 0.015 (n-4) CD 3 CD 3 £75 - s o 0.182 ± 0.009 (n-4) 50HPF+Quinidine -10.211 ± o . o i (n-3) 5 0 H P F + A S A HO.224 ± 0.045 (n-4) Figure 19. Protein binding displacing effect of phenytoin, ASA and quinidine on the free fraction of 5-hydroxypropafenone in human serum. The pH in each treatment group is 7 .4, the value is the mean ± s .d. and n is the number of replicates. 76 PF Free Fraction in Heparinized Whole blood PF H0.057 ± 0.012 (n-4) CD a 3 CD 13 r-t-O o c T3 PF+Quinidine -.0.101 ± o.oi* (n-4) PF+Phenytoin -<0.079 ± 0.01 (n-4) Figure 20. Protein binding displacing effect of quinidine and phenytoin on the free fraction of PF in human whole blood. The pH in each treatment group is 7.4 and the value is the mean ± s.d., * indicates statistically significant difference with two sample t-test combined with Bonferroni inequality at p < 0.05 and n is the number of replicates. 77 3.3.6 The Nature of the Displacement Interaction between PF and Quinidine and PF and Lidocaine using Purified Human AAG as the Biological Fluid 3.3.6.1 Rosenthal Plot of the Binding of PF to Purified AAG over the Therapeutic Concentration Range of PF The Rosenthal plot of the binding of PF to purified human AAG over the therapeutic concentration range of PF is shown in Figure 21. A linear relationship exists between the binding ratio and bound concentration of PF (r = -0.953) and indicating that PF binds to one class of binding sites over the therapeutic concentration range with an intrinsic association or affinity constant of K = 3.44 X 106 M " 1 . 3.3.6.2 The Nature of the Displacement of PF from Human AAG by Quinidine and Lidocaine A linear relationship between the reciprocal of the concentration of unbound PF and the reciprocal of the moles of PF bound per mole of AAG over the concentration range was observed (Figure 22). In the presence of quinidine and lidocaine, less PF was bound to AAG (Figure 22). Common ordinate intercepts indicate that both quinidine and lidocaine compete with PF for the same binding site on AAG. The inhibitory constant (Ki) of quinidine and lidocaine against the binding of PF by human AAG was estimated using the equation Ki = i / (Kp/K - 1), where i equals the molar concentration of inhibitor, K and Kp are the slopes of the line in the absence and presence of the inhibitor. The Ki value for both quinidine and lidocaine was approximately 12 nM. Therefore, the nature of the displacing 28.00 24.00 20.00 £ 16.00 Li_ c 12.00 O CD 8.00 4.00 0.00 1.20 ^0 .90 cn 3 § 0.60 • o ~u c ° 0.30 0.30 0.60 0.90 Total cone, ug/ml 1.20 L i n e a r r e g r e s s i o n : Y=-0.344X + 24.909 r=-0.953 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 Bound M X 10 7 40.00 F i g u r e 21. R e l a t i o n s h i p between t h e r a t i o o f bound c o n c e n t r a t i o n / f r e e c o n c e n t r a t i o n and bound c o n c e n t r a t i o n o f PF i n p u r i f i e d human AAG ( R o s e n t h a l p l o t ) , t h e i n s e r t e d g r a p h i n d i c a t e d t h e r e l a t i o n s h i p between bound and t o t a l c o n c e n t r a t i o n o f PF. —i co 160 r 20 --20 Figure 22. 1/PF (moles'1) Relationship between the reciprocal of the moles of PF bound per mole of AAG and the reciprocal of unbound PF without displacing agento-o ; in the presence of quinidine a-a ; in the presence of lidocaine A - A . 80 effect exhibited by quinidine and lidocaine in purified AAG was competitive inhibition of binding of PF to AAG. 3.4 HPLC Analysis of Indocyanine Green (ICG) 3.4.1 HPLC Analysis of Indocyaine Green according to the Method of Rappaport and Thiessen The HPLC separation and measurement method was attempted on a Hewlett Packard Model 1090 HPLC equipped with a Model 1040A Diode-Array Detector. Unfortunately, the detection l imit attained by Rappaport and Thiessen (0.4 /zg/ml) was not obtained, in large part due to the small volume of sample which can be injected into our microbore HPLC system. Also, the contamination due to incomplete protein precipitation in the injection solution produced a high column pressure which reduced the l i f e time of the column dramatically. The protein contamination present could not be completely eliminated using centrifugation prior to sample introduction into the HPLC due to the limited range of rpm available using our bench top centrifuge. 3.4.2 HPLC Analysis of Indocyanine Green with Fluorometric Detection The fluorometric method reported by Holl ins' (1987) offered better sensit iv i ty than other existing methods such as that of Rappaport and Thiessen (1982) . Although the maximum wavelength in our system was 600 nm, we obtained an excitation spectrum under similar experimental conditions to that published by Hollins so that the assay was conducted at different excitation and emission wavelengths. An excitation spectrum 81 obtained on our HPLC system using the same sample preparation procedure and HPLC conditions as Hoi Tins' (1987) is shown in Figure 23. The emission spectrum at an excitation wavelength of 260 nm was also shown in the same figure. Excitation wavelengths between 210-260 nm gave the highest response. Various excitation wavelengths within this range were tried on our system. The fluorometric detection of ICG was compared with Diode Array UV detection at 216 nm, simultaneously. Unfortunately, fluorometric detection did not show better sensitivity than the corresponding Diode Array detection. A representative chromatogram is shown in Figure 24. Furthermore, changing the HPLC conditions such as mobile phase composition, flow rate did not improve separation or sensitivity. Another excitation spectrum using the method reported by Dorr and Pollack (1989) is shown in Figure 25 . In this method pH 4.0 citrate buffer was used as part of the mobile phase. The reported excitation and emission wavelengths (214 and 370 nm, respectively) were tried on our system, and there was no separation between the solvent front peak and ICG. 3.4.3 HPLC Analysis of ICG with Diode Array Detection A representative chromatogram obtained using the modification method described before (section 2.12.2 ) is shown in Figure 26. A representative standard curve is shown in Figure 27. 190 2 4 0 2 9 0 3 4 0 3 9 0 4 4 0 4 9 0 5 4 0 Excitation wavelength (nm) Figure 23. Excitation o-o and emission A -A (at excitation wavelength of 260 nm) spectra of ICG using pH 6.0 phosphate buffer as part of the mobile phase. co ro 73 Id-II <J) l 00:1 90 80 701 60 501 40 30 20 : 1 0 : L C LOT 0" R 254 ,4 FTLD 550 , i 00 of JW.D o f JH.D •ICG with UV detection ICG with fluorometric detection 4 6 T i me ( m i n . ) 1 0 Figure 24. A representative chromatogram that compares fluorometric detection with UV detection. 3 0 - p 2 8 -2 6 -2 4 -2 2 -2 0 -1 8 -1 6 -1 4 -1 2 -1 0 -8 -6 -4 -2 -0 — 0 5 0 100 150 2 0 0 2 5 0 3 0 0 3 5 0 4 0 0 4 5 0 5 0 0 5 5 0 6 0 0 exc i ta t ion wave leng th ( n m ) Figure 25. Excitation spectrum of ICG using pH 4.0 citrate buffer as part of the mobile phase. co 4=> 85 ins-31-2 7 -CL cu N Figure 26. A representative chromatogram obtained using the modification method (section 2.12.2). 1.000 o 0 . 8 0 0 C L 0J N O ^ 0 . 6 0 0 o o L i n e a r r e g r e s s i o n : Y=0.933X - 0.017 r=0.9999 D a CD 0 . 4 0 0 < 0 . 2 0 0 0 . 0 0 0 0 . 0 0 0 0 . 2 0 0 0 . 4 0 0 0 . 6 0 0 0 . 8 0 0 1.000 A m o u n t of [CG a d d e d (ug) Figure 27. A representative standard curve obtained using the modification method (section 2.12.2). K 2 0 0 co o-i 87 3.4.4 Comparison of Sample and Blank Chromatograms Using a Modified Method of Christie (1986) and Burns (1989) When using 0.1 ml serum spiked with ICG and diazepam, there was no interference from blank. However, when 0.25 ml serum was used, there was a small peak in the blank which interfered with the diazepam peak as shown in Figure 28. This interference peak was minimized by changing the mobile phase composition from 40% to 39% of acetonitrile and flow rate from 0.5 ml/min to 0.2 ml/min. Also, instead of using area ratio, peak height ratio was used for standard curve and calculations. A standard curve obtained under these HPLC conditions is shown in Figure 29. 3.4.5 ICG Extraction Efficiency The extraction efficiency of indocyanine green using the HPLC conditions described in section 2.12.5 is shown in Table 6. The mean recovery is 73%. Table 6. Extraction efficiency of ICG in serum Amount added Amount recovered Recovery 0.40 0.27 68 0.60 0.45 75 0.80 0.58 72 1.00 0.76 76 mean 73 D t C Ft 2 i 6', 4 1 6 , 4 S 5 " 5 5 0 , 1 0 0 o f b f SERUM'S 0 D Top: serum s p i k e d w i t h ICG Bottom: serum b l a n k 4 G Time (min.) 1 0 1 2 Figure 28. Representative chromatograms of serum (0.25 ml) spiked with ICG and serum blank. 1.000 r o .OOO ' 1 1 1 1 ' 1 0.000 0.200 0.400 0.600 0.800 1.000 1.200 Amount of ICG added (ug) Figure 29. Representative standard curve obtained using 0.25 ml serum. 90 3.4.6 Post Extraction Stability Table 7 shows the peak height ratios of ICG/diazepam 48 h after preparation. The ratios were not significantly different from each other and indicated that ICG was stable for at least 48 h after preparation when stored at 4°C and protected from light. Table 7. Peak height ratios of ICG/diazepam at various times after serum extraction Time after extraction h Peak height ratio ICG/diazepam 0 1.51 1.0 1.52 2.5 1.53 4.5 1.48 5.5 1.47 20.5 1.52 24.0 1.50 26.0 1.52 28.0 1.52 44.5 1.48 48.0 1.48 3.4.7 Plasma Data of ICG in Rat A representative chromatogram obtained using the sample preparation procedure in section 2.12.8 is shown in Figure 30. A standard curve and the plasma concentration of ICG vs time curve are shown in Figure 31 and 32, respectively. The calculated ICG plasma clearance was 9.216 ml/min, and was in agreement with the literature value (10.17 ml/min for a 300 g body weight rat, Burns et al., 1989). The on-column detection with our method was 3.8 ng (1.1 uq/m\ using 50 ul plasma). The method of Burns et <L £ 58' 40-30-20" 10-0-V 0 LC LC R 2 16,4 P 2 16,4 5 5 0 , 1 0 0 55 0,100 CD <_> B rt Q . QJ Nl (O •r— Q A O f o f RBDI.D RflTBLRNK. D Top: rat plasma sample Bottom: rat plasma blank T i me ( m i n . ) Figure 30. A representative chromatogram of ICG in rat plasma obtained using the extraction procedure shown in Figure 6. «5 Amount of ICG added (ug) Figure 31. A representative standard curve obtained in rat plasma 94 al., 1989 reported a detection limit of 30 ng/ml. Since there was no volume of injection reported, the on-column detection limit was not known. 3.4.8 The Potential to Detect ICG in Low Concentration Samples with the New Method When the sample injection volume was increased from 10 to 50 /il in the attempt to detect low concentration sample as reported by Burns et a7.(1989), the resolution of the ICG peak from plasma debris peak diminished and the method could no longer be applied. This led to a search for another method of analysis for ICG. 3.4.9 Solid Phase Extraction A representative chromatogram is shown in Figure 33. Although this sample preparation procedure allowed higher injection volume, the separation s t i l l was not complete. D £ LC R 2 16,4 I 00 90: 80: 70 601 50 401 30 20 l 0 0 LC n 0 216 5 5 0 , 1 0 0 5 5 0 , 1 0 0 ICG o f JN2 0R13R.D o f J N 2 0 R 1 4 R . D Diazepam 2 3 T1 me (min. ) Top: serum spiked with ICG Bottom: serum blank Figure 33. A representative chromatogram obtained using solid phase extraction method 96 4. DISCUSSION 4.1 Protein Binding Technique Several techniques have been used to study the nature of ligand-protein interaction. Commonly employed methods are equilibrium dialysis, ultrafiltration, ultracentrifugation and gel filtration for in vitro protein binding studies (Kurz et a7., 1977); and, most recently, microdialysis has been developed and used as an in vivo method in rats (Scott et al., 1990). Among these methods, equilibrium dialysis and ultrafiltration are those most frequently used in the research laboratory setting. Equilibrium dialysis is the "classical" method to determine protein binding (Kurz et a7., 1977). Equilibrium dialysis has the advantage that the free and bound amount of drug are in direct contact during the dialysis process and, therefore, the dynamic equilibrium of binding will not be affected by the experimental procedure (Kurtz et a7., 1977). In addition, equilibrium dialysis is inexpensive, has good reproducibility and gives accurate quantitative data. However, the total drug concentrations before and after dialysis are not the same due to the free passing of unbound drug across the membrane and/or the adsorption of drug to the dialysis device. In ultrafiltration, the protein-free ultrafiltrate is separated from the protein containing phase by forcing it through a semipermeable membrane or f i l ter either by pressure or centrifugal force. Ultrafiltration is less time consuming and applicable to highly concentrated protein solutions or tissue homogenates. However, 97 appreciable binding of drug to the membrane will interfere with protein-binding determinations since only the free drug in the ultrafiltrate is measured and the non-specific binding to the membrane can not be corrected. In addition, a sieve effect, v iz . , increasing protein concentration due to filtration of small molecules and water, and the narrowing of pores in the membrane, due to possible binding of protein molecules to membrane material, make the method less reliable (Kurtz et al., 1977) In ultracentrifugation, the separation of the small free drug molecule from the higher molecular weight drug-protein complex is accomplished by centrifugal force. Ultracentrifugation offers an alternative choice to ultrafiltration when membrane binding problems are encountered. However, when the drugs bind to low-density lipoproteins, it is difficult to separate the free drug from the drug-protein complex by the centrifugal force as the low density drug-1ipoprotein complex tends to stay at the surface of solvent media (Rowland, 1980). Also, free drug molecules with high molecular weight will tend to move to the bottom of the centrifugation tube due to the influence of centrifugal force. This will lead to higher binding values as more drug is confined in the drug-protein complex compartment (Kurz et al., 1977). Relative to other methods, ultracentrifugation is expensive, complicated and its theoretical basis is not well established (Lindup, 1975). In gel f i ltration, bound and unbound amounts of a drug are separated provided that the protein molecules and the protein drug complex cannot penetrate the gel pores. The method can be applied to 98 drugs of relatively high molecular weight but is not suitable for drugs with low affinity to protein, also relatively large volumes of protein solution are needed to obtain correct binding values. The microdialysis method gives equivalent in vitro protein binding values as compared to ultrafiltration (Herrera et al., 1990). The in vivo application of this method provides several advantages over the in vitro method. Samples can be collected at high rates from several sites, without fluid loss, with a single animal (Scott et al., 1990). The method is also associated with some disadvantages. It is invasive, for in vivo use, as a small probe has to be implanted in the tissue of interest (Scott et al., 1989, 1990), the small volume of sample available requires much higher sensitivity of drug assay method. So far, the method is limited in animal study only. Propafenone exhibits extensive adsorption (16.2%) to the ultrafiltration device, thus this method is not suitable for PF protein binding studies as it leads to an underestimation of free drug concentration or free fraction (Chan et al., 1989). Although the adsorption of PF to the equilibrium dialysis cell is similar to that of ultrafiltration device, it does not affect the calculation of free PF, since samples from both sides of the cell are assayed for PF concentration after equilibrium. Therefore, equilibrium dialysis is the method of choice for PF protein binding study. There are several complications in the use of equilibrium dialysis in addition to non-specific binding (binding of drug to the dialysis 99 cell or membrane). These include volume shift (Miller et al., 1981) and pH change (Lui et al., 1986) after equilibrium; protein leakage into the dialysate due to poor membrane quality, improper pore size and/or poor technique in mounting the dialysis membrane; bacterial growth due to dialysis temperature and long dialysis time (Lui et al., 1986; Ilett et al., 1975). In the present PF studies, volume and pH before and after each dialysis were measured, there was no significant change in either volume or pH. The protein leakage was also tested by adding 3% trichloroacetic acid to a sample of the buffer side of the dialysis cell; absence of precipitation was used as an indicator of no protein leakage. The possibility of bacterial growth was not studied since the dialysate appeared to be nonturbid and the dialysis time (8 hours) was not sufficiently long enough to encourage bacterial growth. 4.2 Protein Binding of 5-Hydroxypropafenone In purified isolated human AAG solution, 5-hydroxypropafenone binds to one class of binding sites with affinity constant of 1.6 X 10^  (section 3.2.1). During displacement experiments which conducted at similar initial concentrations (0.5 jug/ml), the free fraction of 5-hydroxypropafenone is similar in AAG and albumin solution (Table 8), whereas, the free fractions of PF are significantly different in AAG and albumin. This may indicate that at very low concentrations, the binding of 5-hydroxypropafenone is not specific to either AAG or albumin. Also, the free fractions in serum and blood are much lower than that in isolated protein solutions, which further indicates that the binding of 5-hydroxypropafenone to blood protein might be nonspecific. If the 100 Table 8. Free fraction of 50HPF and PF in AAG, albumin, serum and whole blood 50HPF free fraction PF free fraction AAG 0.321 ± 0.028 0.057 ± 0.004 Albumin 0.354 ± 0.013 0.339 ± 0.006 serum 0.189 ± 0.015 0.017 ± 0.003 blood 0.163 ± 0.007 0.057 ± 0.012 101 binding is specific to a certain protein, one would expect that the free fraction would be the same in either isolated protein solution or mixture of a number of proteins, as long as the binding sites on that particular protein are not saturated. But in the serum and whole blood the free fraction of 5-hydroxypropafenone is approximately half of that seen in the pure isolated protein solution. The nonspecific binding characteristic of 5-hydroxypropafenone was also characterized by Tonn (1990). 4.3 Displacing Effect of Verapamil, Propranolol, ASA, Phenytoin, Lidocaine and Quinidine 4.3.1 Verapamil Verapamil is a calcium channel blocking drug which has been used extensively as an antiarrhythmic agent to control superventricular tachycardia (Krikler and Spurrell, 1974). Verapamil is approximately 90% bound to plasma proteins in a concentration independent manner over the concentration range from 35 to 1,557 ng/ml suggesting a great excess of binding sites (Keefe et al., 1981). When verapamil (200 ng/ml) was used as a displacing agent for the protein binding of PF and 5-hydroxypropafenone in an solution of purified isolated AAG (Figures 10 and 14) or albumin (Figures 12 and 16), no displacing effect was observed. One possible explanation could be the concentrations of PF (1 //g/ml) and 5-hydroxypropafenone (0.5 //g/ml) were sufficiently low that even with the addition of verapamil, there were still enough binding sites to accommodate each drug in the presence of each other. It is 102 also possible that the binding affinity of verapamil is too low to displace PF or 5-hydroxypropafenone from their binding sites, or verapamil binds to different sites on AAG and albumin from those of PF and 5-hydroxypropafenone. Therefore, no displacement interaction could be observed. Higher concentrations of PF and 5-hydroxypropafenone could have been used, but the results would be less meaningful clinically since the concentration used were maximal therapeutic concentrations encountered during routine clinical use of PF. 4.3.2 Propranolol Propranolol is an beta-adrenergic receptor blocking agent frequently used as an antiarrhythmic drug. In human serum, propranolol binds mainly to AAG (Sager et al., 1978; Piafsky et al., 1978). In our study, when propranolol was used as a displacer on PF binding to AAG and albumin, no binding interaction was observed {i.e. no change in PF free fraction was seen with the addition of propranolol, Figures 10 and 12). There was also no change in 5-hydroxyprdpafenone free fraction in AAG or albumin before and after the in vitro solution was spiked with propranolol (Figures 14 and 16). The lack of displacing effect of propranolol on PF binding was probably due to the very low concentration of propranolol used (1 ug/ml), nonsignificant binding of both PF and propranolol to albumin and lower binding affinity of propranolol to AAG than PF. Both PF and propranolol are basic compounds. They both bind to AAG with similar affinity to one class of binding sites (in a magnitude of 10° M"1) (Gill is et al., 1985). However, the binding affinity to a second class of binding sites is much lower for propranolol (« 10^  M"*) than that for PF (« 10^). This, together with the low concentration of propranolol used ( 1 //g/ml) indicates that propranolol can not displace PF from its binding sites on AAG and therefore, no apparent effect on PF free fraction in purified human AAG solution would be observed. As both PF and propranolol are not bound specifically to albumin due to their chemical nature (basic compound), the lack of displacing effect of propranolol in purified human albumin solution is not surprising. Since protein binding of 5-hydroxypropafenone is nonspecific to either AAG or albumin, the observed no change in 5-hydroxypropafenone free fraction in purified isolated human albumin solution upon the addition of propranolol is anticipated. Higher concentration of propranolol might exhibit displacing interaction, but a plasma concentration of 1 //g/ml is the maximum therapeutic concentration for this drug. 4.3.3 Acetyl salicylic Acid (ASA) ASA has been widely used as an antipyretic, analgesic and anti-inflammatory agent for over 100 years. Salicylate serum concentrations of up to 100 //g/ml are required for effective analgesia, whereas serum levels of 300-400 //g/ml are expected in the management of acute rheumatic fever (Evans et al, 1986). The displacing effect of ASA on the protein binding of phenytoin has been well documented (Ehrnebo and Odar-Cederlof, 1977; Odar-Cederlof and Borga, 1977; Leonard et al., 1981; Lunde et al., 1971; Paxton, 1980). ASA caused permanent acetylation of the lysine residues of human albumin under physiologic 104 conditions in vitro (Pinckard et al., 1973). In our displacement study, ASA did not cause any change in PF free fraction in isolated AAG and albumin (Figures 10 and 12), no change in 5-hydroxypropafenone free fraction in a solution of AAG (Figure 14) and serum (Figure 19) was observed, whereas an increased free fraction of 5-hydroxypropafenone in a solution of albumin was found (Figures 16). Due to the specific binding of PF to AAG, the lack of a displacing effect of ASA on the binding of PF to AAG is expected. Compared with the concentration of ASA and PF, the relatively large concentration of albumin is probably the main reason for the observation of a stable PF free fraction. It is also possible that PF binding sites in albumin are not around the lysine residue area and therefore, acetylation of albumin molecules exerted by ASA has no effect on PF free fraction in purified isolated human albumin solution. Therefore, the displacing interaction between PF and ASA is not important. Although ASA caused significant increase in 5-hydroxypropafenone free fraction, this interaction is not significant in serum, it is, therefore, unlikely to be important clinically. 4.3.4 Phenytoin Phenytoin is used to treat seizure disorders and cardiac arrhythmia. When it is used as an antiarrhythmic agent, the therapeutic concentration range is between 10-18 /xg/ml (Bigger et al., 1968). In plasma, phenytoin is 90% bound and it is bound mainly to albumin (Evans et al., 1986). The binding can be displaced by weak acids, such as phenylbutazone, salicylic acid, valproic acid and sulfisoxazole (Evans et al., 1986). When phenytoin was used as displacing agent in our 105 study, the free fraction of PF increased 45% in purified human AAG (Figure 11), 91% in albumin (Figure 13), 150% in human serum (Figure 18), and 39% in whole blood (Figure 20), although the increases observed in AAG and whole blood did not reach a level of statistical significance (two sample t-test combined with Bonferroni inequality at p < 0.05). On the other hand, the free fraction of 5-hydroxypropafenone in serum was not affected by the addition of phenytoin (Figure 19). One reason could be that the concentration of 5-hydroxypropafenone is too low (half of that of PF), and even with the addition of phenytoin, there are s t i l l enough binding sites to adapt both 5-hydroxypropafenone and phenytoin, therefore, no displacing effect could be observed. It may also be due to the nonspecific binding of 5-hydroxypropafenone to albumin that even if there is displacement of 5-hydroxypropafenone from albumin binding sites by phenytoin, the free 5-hydroxypropafenone binds subsequently to other proteins existing in serum. The binding affinity of phenytoin to human albumin is high (1.3 X 104 M" 1 , Levy et al., 1982), it is thus easy to understand that the extent of increase in free fraction of PF is higher in purified isolated human albumin solution than that in AAG solution. The large increase in PF free fraction in serum (150%, Figure 18) is almost equal to the summation of the individual increase in free fraction in AAG (45%, Figure 11) and albumin (91%, Figure 13), which indicates that PF does not bind to other serum proteins, such as lipoprotein to a significant degree. This is in agreement with the observation that the degree of binding of PF is similar in normal serum to lipoprotein deficient serum (Tonn, 1990). The statistically nonsignificant increase in PF free fraction in whole blood could be due to the uptake of free PF by the red blood cells as PF exits in these 106 cells (Tonn, 1990) Higher concentration of 5-hydroxypropafenone could be used, but 0.5 jug/ml is close to the clinical situation as the maximum therapeutic concentration for PF is only 1 ug/m]. 4.3.5 Quinidine Quinidine is a well-known antiarrhythmic agent. In human plasma, quinidine is approximately 90% protein bound (Reidenberg and Affrime, 1973), and it binds mainly to albumin and low and high density lipoproteins (Nilsen, 1976; Nilsen and Jacobsen, 1975). When quinidine was used as the displacing agent in our study, the free fraction of PF increased significantly in serum (210%, Figure 18) and whole blood (78%, Figure 20). no change in 5-hydroxypropafenone free fraction in serum was observed (Figure 19). Further study indicates that quinidine competes with PF for binding sites in purified isolated human AAG solution (Figure 22). Since quinidine binds mainly to albumin and low and high density lipoprotein (Nilsen, 1976; Nilsen and Jacobsen, 1975), it is thus expected that the free fraction of PF should also increase in a solution of AAG or albumin. In fact, the free fraction of PF did increase in AAG or albumin in our study (84% in AAG, Figure 11; 70% in albumin, Figure 13), but it failed to reach statistical significant due to the large variation within the treatment group and the conservative statistical test used. Again, the uptake of free PF by the red blood cells may contribute to the lower degree of increase in PF free fraction in whole blood (78%) than in serum (210%). 5-hydroxypropafenone displays concentration independent binding in serum over the concentration range from 0.12 to 35.3 Mg/ml (Tonn, 1990), which suggests 107 the large excess of binding sites. 5-hydroxypropafenone also binds to lipoprotein to a significant degree as increased free fraction of 5-hydroxypropafenone in lipoprotein deficient serum compared to normal serum is observed (Tonn, 1990). It is thus not difficult to understand that there is no significant displacement interaction between quinidine and 5-hydroxypropafenone in human serum. 4.3.6 Lidocaine Lidocaine is considered the parenteral drug of choice for the acute treatment or prevention of ventricular arrhythmias associated with acute myocardial infarction or cardiovascular surgery and ventricular tachycardia. The binding of lidocaine is correlated well with AAG concentration (Shand, 1984). Basic drugs such as amitriptyline, chlorpromazine, disopyramide, imipramine, meperidine, nortriptyline, propranolol and quinidine did not displace lidocaine from serum AAG binding sites to a great extent. However, lidocaine was displaced by bupivacaine (Goolkasian et al., 1983). When lidocaine is used as a displacing agent for the binding of PF, significant increases in PF free fraction were observed in AAG (124%, Figure 11), albumin (12%, Figure 13) and serum (580%, Figure 18). The extent of increase was much higher in AAG and serum than in albumin. Due to the specific binding of PF and lidocaine to AAG, greater displacing ability in AAG than in albumin is expected. Further study indicates that lidocaine competes with PF for binding sites in purified isolated human AAG solution (Figure 22). The binding affinities of PF and lidocaine to solution of isolated human AAG are similar (1.7 X 105 for PF and 1.4 X 105 for lidocaine, Gill is et 108 al., 1985), but the molar concentration of lidocaine is much higher than that of PF (approximately 9 times higher) in our displacement study. This might contribute to the observed increase in PF free fraction in AAG as the drugs compete for available binding sites. The pKa of lidocaine is 7.9 (Courtney, 1980) and it is less basic than PF (pKa = 9.0, Kohlhardt and Seifert, 1980). It is thus possible that the binding affinity of lidocaine to albumin is higher than that of PF, and displaces PF from its binding sites in albumin causing a subsequent increase in free fraction of PF in albumin. The large, statistically significant increase in PF free fraction in serum (5.8 fold) implies that this displacement interaction might be clinically important. 4.4 Clinical Significance of Displacing Effect Combinations of drugs are often necessary in the treatment of diseases. For drugs that bind extensively to both plasma and tissue proteins, such as PF, the binding displacement interaction may not be clinical important, as the displaced free drug can be quickly distributed in tissues and reach a new equilibrium. Since the volume of distribution is large, the free drug concentration in the plasma would be the same or at least approaches the pre-interaction level (D'Arcy and McElnay, 1982; McElnay and D'Arcy, 1983). Although phenytoin and quinidine display PF from its binding sites significantly in serum, or whole blood, there might be no significant change in free PF concentration in vivo (when the drugs are given orally) due to the rapid distribution in tissues, and even there is, it will be transient in nature. However, the binding displacement interaction between PF and 109 lidocaine could be clinically important. Since lidocaine is given intravenously and if it is given by bolus injection, displacement of PF from its binding sites will be much faster than distribution and re-equalization of displaced free PF, and accumulation of free PF may occur. This increased free PF plasma concentration may lead to increased drug reaching receptors and/or entering such a compartment as the CSF and, hence, toxic effects. 4.5 HPLC Analysis of ICG 4.5.1 HPLC Analysis of ICG with Fluorometric Detection The two reported HPLC analysis methods of ICG with fluorometric detection (Hollins et al., 1987; Dorr and Pollack, 1989) were tried on our system without success. The excitation and emission maxima used in the method of Hollins' were 780 and 810 nm, respectively. With 200 fi\ injection, plasma concentrations as low as 3.2 ng/ml can be detected by this method. We could not reproduce this method as our fluorometric detector wavelength range is from 190 to 600 nm. Although we did try to get an excitation spectrum and a corresponding emission spectrum (Figure 23) on our system, the fluorometric response was much lower than the corresponding DAD response (Figure 24). Dorr and Pollack used 214 nm as excitation wavelength and 370 nm as emission wavelength, however, we could not get any detector response at this excitation and emission wavelength pair. Therefore, we concentrated our effort on HPLC analysis of ICG with DAD detection. 110 4.5,2 HPLC Analysis of ICG with Diode Array Detection 4.5.2.1 Sample Preparation Direct deproteinization of serum sample with acetonitrile before injection into HPLC as reported by Rappaport and Thiessen is not feasible with our system, as the microbore system we have is easily contaminated by debris contained in serum and the 5 am, 2.3 i.m.column is easily plugged and produces high column back pressure. The chemical nature of ICG (Figure 4) limited the use of solvent-solvent extraction by either acidification or alkalization. The phase separation method with ammonium sulphate as reported by Christie et al. (1985) is one way to minimize protein contamination. The method reported by Burns et al. (1989) involved phase separation and was employed in the sample preparation procedure described in Experimental section 2.12.2 (Figure 5). The utilization of syringe sample filter made the sample cleaner and allowed it to be injected into our HPLC system. This procedure was used to develop a standard curve (using peak area ratio) ranging from 0.1 ug to 1.0 ug (Figure 27). With the use of 0.1 ml serum in this procedure, a 0.1 ug detection limit corresponds to a serum concentration of 1 /ig/ml. This detection limit is not adequate for our future use. When the serum volume was increased to 0.25 ml, blank interference with internal standard was encountered (Figure 28). This is probably due to the limited centrifugation speed (3,000 rpm) of our bench top centrifuge. The same problem was displayed in the method of Rappaport and Thiessen (1982), where the interference peak from rabbit blank plasma sample contributes to 5% of the internal standard peak height. I l l With modification of the chromatographic conditions, the blank interference was minimized and a second standard curve was developed using peak height ratio(Figure 29). With the increase in serum volume, it was possible to detect serum ICG concentration of 0.8 /zg/ml. The extraction efficiency of ICG was 73% (section 3.5.5) and the prepared sample was stable for at least 48 hours when stored at 4°C in black vial holder (section 3.5.6), it seems possible to use this method for our future HPLC analysis of ICG in human serum with -automatic sample injection. 4.5.1.2 HPLC Analysis of ICG Plasma Concentration in Rat Due to the small volume of plasma (100 #1) which can be obtained from rat during frequent serial sampling, micro-centrifuge tubes normally have to be used for plasma sample preparation. Since our centrifuge did not have the necessary rotor to accommodate the small tubes, another high speed centrifuge (17,000 rpm) was used. With the use of high speed centrifugation, the supernatant obtained was so clear that direct injection into HPLC was possible. Therefore, another sample preparation procedure described in Experimental section 2.12.8 (Figure 6) was used. The detection limit obtained with this method was 1.1 fig/m\ using 50 fi\ of plasma. This detection limit was much higher than 30 ng/ml, using 35 /zl plasma, as reported by Burns et al. (1989). The difference in detection wavelength (they used 784 nm and we used 216 nm) was the principal reason for the observed high detection limit of this method. It is known that the maximum absorbance wavelength for ICG is between 780 - 806 nm depending on the solvent media (Owen, 1973). 112 Moreover, the relative difference in absorbance is 0.16 between wavelength 216 and 784 nm (Figure 34). Therefore, the low detection wavelength might be the main reason for the high detection limit we obtained. We could not use 784 nm as our detection wavelength as our DAD detector does not cover the visible wavelength range. Another factor contributing to the failure of the method could be the difference in injection volume. In this study, 10 /zl was injected into the HPLC system while the volume used by earlier investigation was not reported. An increase in the injection volume from 10 fi\ to 50 #1 in our method leads to loss of separation due probably to the low capacity of our 10 cm, 5 (im particle size, 2.3 i.m. column. The differences in sample preparation procedure might also contribute to the high detection limit. Since a clean sample was obtained with the utilization of high speed centrifuge in our method, phase separation using ammonium sulphate in sample preparation employed by Burns et al. was not used. 4.5.2.3 Summary The HPLC analysis system we have is a microbore system and the maximum detection wavelength is 600 nm. Since the maximum absorption wavelength of ICG is around 800 nm (Landsman et al, 1976), better sensitivity would be obtained if this wavelength could be used as the detection wavelength in our HPLC system. Using the DAD, we were forced to use 216 nm as the detection wavelength, which is the maximum absorption wavelength in the range of 190-600 nm. The reference wavelength used is 550 nm. This may contribute partly to the observed lower sensitivity as there might be some absorption at the reference 0.3008? 0.2t070-i=c Q.130S2 0.12Q3S-0.OGO1? 0.0000-200 / N 300 ' — i — 1 — 1 — 1 — 1 — i — 1 — ' — • — 1 — r — 400 500 500 UfiVELENGTH  700 soo A n n o t a t e d W a v e l e n g t h s : 1 : W a v e l e n g t h = 2 1 6 2 : W a v e l e n g t h = 7 8 4 R e s u l t = 0 . 1 3 4 8 1 1 R e s u l t = 8 . 2 9 3 2 4 3 Figure 34. Ultraviolet/visible spectrum of ICG in distilled water (10 /zg/ml) 114 wavelength. The low sample capacity of the micropore system limits the increase in injection volume, detection of low concentration samples by increasing injection volume is sharply limited by column capacity. However, with the use of long column, for example, 25 cm instead of 10 cm as we did, a higher volume injection may be possible. Although ICG is a weak acid with pKa = 3.27 (Bjornsson et al., 1982), solvent-solvent extraction for ICG sample preparation is unlikely due to the physico-chemical properties of ICG. The two sulfonate groups of ICG and its high water solubility (Figure 4) limits the use of acid or base in the attempt to get free ICG acid. Solid phase extraction probably represents the best choice for ICG sample preparation. Although the separation is not complete at the HPLC conditions used, it might be possible to get better separation with modifications. 115 5. SUMMARY AND CONCLUSIONS 1. 5-Hydroxypropafenone binds to one class of binding site in purified human AAG solution. The displacing effect of various agents on the binding of PF and 5-hydroxypropafenone to various protein solutions is summarized in Table 9 and 10, respectively. The results can be concluded as follows: 2. Verapamil did not displace either PF or 5-hydroxypropafenone from their binding sites in either purified human AAG or albumin solution. 3. Propranolol did not affect the binding of PF or 5-hydroxypropafenone in solutions of AAG and albumin. 4. ASA caused significant increase in 5-hydroxypropafenone free fraction in a solution of albumin but not AAG; ASA did not affect PF binding. 5. Lidocaine caused significant increase in PF free fraction in either purified human AAG, albumin or serum. It is a potent displacing agent for PF protein binding. It also affected the binding of 5-hydroxypropafenone in a solution of AAG but not in a pure albumin solution. 6. Phenytoin affected the binding of PF in solution of albumin and serum, but not in AAG and whole blood. Table 9 PF free fraction before and after treatment with differing displacing agents biological fluid AAG albumin serum bl ood buffer conc.(M) 0.067 0.1 0.067 0.1 0.1 0.1 treatment control 0.075 ± 0.01 0.057 ± 0.004 0.431 ± 0.017 0.339 ± 0.006 0.017 ± 0.003 0.057 ± 0.012 ASA 0.095 ± 0.024 -0.443 ± 0.047 - - -propranolol 0.091 ± 0.005 -0.452 ± 0.038 - - -verapamil 0.088 ± 0.008 -0.405 ± 0.015 - - -1idocaine -0.129 ± 0.013* -0.383 ± 0.008* 0.114 ± 0.006* -quinidine -0.105 ± 0.016 -0.577 ± 0.182 0.052 ± 0.009* 0.101 ± 0.01* phenytoin -0.083 ± 0.016 -0.649 ± 0.055* 0.041 ± 0.007* 0.079 ± 0.01 *: statistical significance using two sample t-test combined with Bonferroni inequality at p < 0 .05 . Table 10 50HPF free fraction before and after treatment with differing displacing agents biological fluid AAG albumin serum buffer conc.(M) 0.067 0.1 0.067 0.1 0.1 treatment control 0.321 ± 0.029 0.321 + 0.028 0.271 ± 0.017 0.354 ± 0.013 0.189 ± 0.015 ASA 0.476 ± 0.087 - 0.398 ± 0.028* - 0.224 ± 0.045 propranolol 0.355 ± 0.029 - 0.304 ± 0.01 - -verapamil 0.304 ± 0.068 - 0.281 ± 0.045 - -1idocaine 0.472 ± 0.03* - 0.342 ± 0.01 0.114 ± 0.006 quinidine - - - -0.211 ± 0.01 phenytoin - - - -0.182 ± 0.009 *: statistical significance using two sample t-test combined with Bonferroni inequality at p < 0.05. 118 serum, but not in AAG and whole blood. 7. Quinidine affected the binding of PF in serum and blood, however, the increase in PF free fraction in solution of isolated human AAG or albumin failed to reach statistical significant. It did not affect the binding of 5-hydroxypropafenone. 8. The nature of the displacing effect of lidocaine and quinidine in AAG is competitive inhibition. 119 REFERENCES Arboix M., Puigdemont A., Moya A. and Cinca J., Pharmacokinetics of intravenous propafenone in patients with episodes of paroxymal supraventricular tachycardia, Meth. Find Exptl. Clin. Pharmacol. 7:435-438, 1985. Axelson J.E., Chan G.L.-Y., Kirsten E.B., Mason W.D., Lanman R.C. and Kerr C.R., Food increases the bioavailability of propafenone, Br. J. Clin. Pharmac. 23:735741, 1987. Baker D.W., Forst F.K. and Daigle R.E., Doppler principles and technique in ultrasoud:its application in medicine and biology, Fry F.J. ed. Elsevier, Amsterdan, ppl61-287, 1978. Bigger T, Schmidt D.H. and Kutt H., Relationship between the plasma level of diphenylhydantoin sodium and its cardiac antiarrhythmic effects, Circ, 38:363-374, 1968. Bjornsson O.G., Murphy R. and Chadwick V.S., Physicochemical studies of indocyanine green (ICG): absorbance/concentration relationship, pH tolerance and assay precision in various solvents, Experientia 38:1441-1442, 1982. Bradley, E.L., Measurement of hepatic blood flow in man, Surgery 75:783-789, 1974. Bradley S.E., Ingelfinger F.J., Bradley G.P. and Curry J.J., The estimation of hepatic blood flow in man, J. Clin. Invest. 24:890-897, 1945. Burns E., Ball C.E., Christie J.P., Broadhead G.D., Tucker G.T. and Bax N.D.S., Direct and indirect measurement of the hepatic extraction ratio of indocyanine green in the rat. Clin. Sci. 76-503-508, 1989. Chan G.L.-Y., Propafenone pharmacokinetics: GLC-ECD analysis; metabolic induction by phenobarbital in non-smoking and smoking healthy volunteers; protein binding; pharmacodynamics in patients, PhD. Thesis, University of British Columbia, 1989. Chan G.L.-Y., Axelson J.E., Abbott F.S., McErlane K.M. and Kerr C.R., Determination of propafenone in biological fluids by fused-silica capillary gas-chromatography using electron-capture detection, J. Chromatogr., 417:295-308, 1987 Chan G.L.-Y., Axelson J.E., Abbott F.S., McErlane K.M. and Kerr C.R., Determination of 5-hydroxypropafenone in biological fluids by capillary gas chromatography using electron-capture detection, J. Charmatogr., 495:349-353, 1989. Chan G.L.-Y., Axelson J.E., Price J.D.E., McErlane K.M. and Kerr C.R., In vitro protein binding of propafenone in normal and uraemic human sera, Eur. J. Clin. Pharmacol., 36: 495-499, 1989. 120 Cherrick G.R., Stein S.W., Leevy CM. and Davidson C.S., Indocyanine green: observations on its physical properties, plasma decay and hepatic extraction, J. Clin. Invest., 39:592, 1960. Christie J.P., Bax N.D.S., Lennard M.S. and Tucker G.T., Indocyanine green kinetics in man: a reappraisal, Br. J. Clin. Pharmacol., 21:568p-569p, 1985. Connolly S.J., Kates R.E., Lebsack C.S., Echt D.S. and Mason J.W., Clinical efficacy and electrophysiology of oral propafenone for ventricular tachycardia, Am. 0. Cardiol., 52:1208-1213, 1983a. Connolly S.J., Kates R.E., Lebsack C.S., Harrison D.C. and Winkle R.A., Clinical Pharmacology of Propafenone, Circ, 68:589-597, 1983b. Connolly S.J., Lebsack C, Winkle R.A. Harrison D.C. and Kates R.E., Propafenone disposition kinetics in arrhythmia patients, Clin. Pharmacol. Ther. 36:163-168, 1984. Courtney K.R., Structure-activity relationship for frequency-dependent sodium channel block in nerve by local anesthetics, J. Pharmac. Exp. Ther., 213:114, 1980. Craig W.A., Evenson M.A. and Ramgopal V., The effect of uremia, cardiopulmonary bypass and bacterial infection on serum protein binding; in Benet (Ed.) The effect of disease states on drug pharmacokinetics, pp. 125-136 (American Pharmaceutical Association, Washington, D.C, 1976). D'Arcy P.F. and McElnary J.C, Drug interactions involving the displacement of drugs from plasma protein and tissue binding sites, Pharmac. Ther., 17:211-220, 1982. Donn K.H., Powell J.R., Rogers J.F. and Plachetka J.R., Lack of effect if histamine Hg-receptor antagonists on indocyanine green disposition measured by two methods, J. Clin. Pharmacol. 24:360-370, 1984. Dorr M.B. and Pollack G.M., Specific assay for the quantitation of indocyanine green in rat plasma using high-performance liquid chromatography with fluorescence detection, J. Pharm. Sci. 78:328-333, 1989. Dukes I.D. and Vaughan Williams E.M., The multiple modes of action of propafenone, Eur. Heart J., 5:115-125, 1984. Edvardsson N. and Olsson S.B., Clinical value of plasma concentrations of antiarrhythmic drugs, Eur. Heart J., 8 (Supp.A):83-89, 1987. Ehrnebo M. and Odar-Cederlof I., Distribution of pentobarbital and diphenylhydantoin between plasma and cells in blood: effect of salicylic acid, temperature and total drug concentration, Eur. J. Clin. Pharmacol., 11:37-42, 1977. Evans W.E., Schentag J.J. and Jusko W.J. (Eds), Applied 121 pharmacokinetics: principles of therapeutic drug monitoring, 2nd Edition, 1986. Fox I.J., Brooker L.G.S., Heseltine D.W., Essex H.E. and Wood E.H., A tricarbocynanine dye for continuous recording of dilution curves in whole blood independent of variations in blood oxygen saturation, Proc. Mayo. clin. 32:478-1957a. Fox I.J. and Wood E.H., Indocyanine green: Physical and physiologic properties, Proc. Staff Meet. Mayo Clin. 35:732-744, 1960. Fox I.J. and Wood E.H., Applications of dilution curves recorded from the right side of the heart or venous circulation with the aid of a new indicator dye, Proc. Mayo. Clin. 32:541, 1957b. Frabetti L., Marchesini B., Capucci A., Cavallini C, Gubelli S. and Magnani B., Antiarrhythmic efficacy of propafenone: evaluation of effective plasma levels following single and multiple doses, Eur. J. Clin. Phamacol. 30:665-671, 1986. Fraser D.G., Ludden T.M., Evens R.P. and Sutherland E.W., Displacement of phenytoin from plasma binding sites by salicylate, Clin. Pharmacol. Ther., 27:165-169, 1980. Gathje J., Steuer R.R. and Nicholes K.R.K., Stability studies on indocyanine green dye, J. Appl. Physiol. 29:181-185, 1970. Giani P., Landolina M., Giudici V., Bianchini C, Ferrario G., Marchi S., Riva E. and Latini R., Pharmacokinetics and pharmacodynamics of propafenone during acute and chronic administration, Eur. J. Clin. Pharmacol. 34:187-194, 1988. Gibaldi M. and Perrier D., Pharmacokinetics, 2nd Edition, Marcel Dekker, Inc., New York and Basel, 1982. Gill R.W., Measurement of blood flow by ultrasound: accuracy and sources of error, Ultrasound in Med. and Biol., 11:625-641, 1985. Gil lis A.M., Yee Y-G and Kates R.E., Binding of antiarrhythmic drugs to purified human ai-acid glycoprotein, Biochem. Pharmacol. 34:4279-4282, 1985. Goldstein A, The interaction of drugs and plasma proteins, Pharmacol. Rev., 1:102-165, 1949. Goolkasian D.L., Slaughter, R.L., Edwards D.J. and Lalka D., Displacement of lidocaine from serum ai-acid glycoprotein binding sites by basic drugs, Eur. J. Pharmacol. 25:413-417, 1983. Haefel i E.W., Vozeh S., Ha H-R and Follath F., Comparison of the pharmacodynamic effects of intravenous and oral propafenone, Clin. Pharmacol. Ther., 48:245-254, 1990. Harron D.W.G. and Brogden R.N., Propafenone: A review of its 122 pharmacodynamic and pharmacokinetic properties, and therapeutic use in the treatment of arrhythmias, Drugs 34:617-647, 1987. Hartel G., Efficacy of oral propafenone in chronic ventricular arrhythmias: a placebo controlled cross-over exercise study, Eur. Heart J. 6:123-129, 1985. Herrera A.M., Scott D.O. and Lunte C.E., Microdialysis sampling for determination of plasma protein binding of drugs, Pharma. Res., 7:1077-1081, 1990. Hege H.G., Hollmann M., Kaumeier S. and Lietz H., The metabolic fate of zH-labled propafenone in man, Eur. J. Drug Metab. Pharmacok. 9:41-55, 1984. Henry L.A., Dunlop A.W., Mitchell S.N., Turner P. and Adams P., A model for the pH dependence of drug-protein binding, J. Pharm. Pharmacol., 33:179-182, 1981. Hollins B., Noe B. and Henderson J.M., Fluorometric determination of indocyanine green in plasma, Clin. Chem., 33:765-768, 1987. Hollmann M., Brode E., Hotz D., Kaumeier S. and Kehrhahn O.H., Investigations on the pharmacokinetics of propafenone in man, Arzneim.-Forsch./Drug Res. 33:763-770, 1983. Hunton D.B., Bollman J.L. and Hoffman H.N., Studies of hepatic function with indocyanine green, Gastroenterology 39:713, 1960. Ilett K.F., Hughes I.E. and Jellett L.B., The use of kanamycin in equilibrium dialysis at 37°C, J. Pharm. Pharmacol., 27:861-863, 1975. Jusko W.J. and Gretch M., Plasma and tissue binding of drugs in pharmacokinetics, Drug Met. Rev., 5:43-140, 1976). Kates R.E., Yee Y.G. and Winkle R.A., Metabolite cumulation during chronic propafenone dosing in arrhythmia, Clin. Pharmacol. Ther., 37:610-614, 1985. Keefe D.L., Yee Y.-G. and Kates R.E., Verapamil protein binding in patients and in normal subjects, Clin. Pharmacol. Ther., 29:21-26, 1981. Kerr C.R., Klein G.J., Axelson J.E. and Cooper J.C, Propafenone for prevention of recurrent atrial fibrillation, Am. J. Cardiol. 61:914-916, 1988. Kessler K.M., Leech R.C. and Spann J.F., Blood collection techniques, heparin and quinidine protein binding, Clin. Pharmacol. Ther., 25:204-210, 1979. Ketterer S.G. and Wiegand B.D., The excretion of indocyanine green and its use in the estimation of hepatic blood flow, Clin. Res. 7:71, 1959. Klotz I.M., Ligand-receptor interactions: what we can and cannot learn 123 from binding measurements, Trends Pharmacol. Sci., June:253-255, 1983. Kohlhardt M. and Seifert C, Naunyn-Schmiedeberg's Archs Pharmac 31P:55, 1980 Krikler D.M. and Spurrell R.A.J., Verapamil in the treatment of paroxysmal supraventricular tachycardia, Postgrad. Med. J., 50:447-453, 1974. Kroemer H.K., Mikus G., Kronbach T., Meyer U.A. and Eichelbaum M., In vitro characterization of the human cytochrome P-450 involved in polymorphic oxidation of propafenone, Clin. Pharmacol. Ther., 45:28-33, 1989. Kunin CM., Clinical pharmacology of the new penicillins. II. Effect of drugs which interfere with binding to serum proteins, Clin. Pharmacol. Ther., 7:180-188, 1966. Kurz H., Trunk H. and Weitz B., Evaluation of methods to determine protein-binding of drugs, Arzneim.-Forsch./Drug Res., 27(11), Nr. 7: 1373-1380, 1977. Landsman M.L.J., Kwant G., Mook G.A. and Zijlstra W.G., Light-absorbing properties, stability, and spectral stabilization of indocyanine green, J. Appl. Physiol. 40:575-583, 1976. Latini R., Maggioni A.P. and Cavalli A., Therapeutic drug monitoring of antiarrhythmic drugs: rationale and current status, Clin. Pharmacokinet. 18:91-103, 1990. Latini R., Riva E., Cazzaniga M.G., Maggioni A.P. and Volpi A., Distribution of propafenone and its active metabolite, 5-hydroxypropafenone, in human tissues, Am. Heart J., 113(3):843-844, 1987. Leatherbarrow, R.J., Enzfitter, A non-linear regression data analysis program for the IBM PC, Leatherbarrow R.J., Ed. Manual, Elsevier-Biosoft, 1986. Leevy CM., Bender J., Silverberg M. and Nyalor J., Physiology of dye extraction by the liver: comparative studies of sulfobromophthalein and indocyanine green, Ann. N.Y. Acad. Sci. 111:161, 1963. Leevy CM., Smith F., Longueville J., Paumagartner G. and Howard M.M., Indocyanine green clearance as a test for hepatic function. Evaluation by dichromatic ear densitometry, J. Amer. Med. Ass., 200:236, 1967. Levy R.H. and Lai A.A., Valproate absorption, distribution and excretion. In: Woodbury D.M., Penry J.K. and Pippenger C.E., eds., Antiepileptic drugs, New York: Raven Press, 1982. Leonard R.F., Knott P.J., Rankin G.D., Robinson D.S. and Melnick D.E., Phenytoin-salicylate interaction, Clin. Pharmacol. Ther., 29:56-60, 1981. 124 Levine W.G., Biliary excretion of drugs and other xenobiotics, Ann. Rev. Pharmacol. Toxicol. 18:81-96, 1978. Lindup W.E., Drug-albumin binding, Biochem. Soc. Trans., 3:635-640, 1975. Lindup W.E., Plasma protein binding of drugs-some basic and clinic aspects, Prog. Drug Metab., 10:141-186, 1987 Lowry O.H., Rosebrough N.J., Farr A.L. and Randall R.J., Protein measurement with the folin phenol reagent, J. Bio. Chem., 193:265-275, 1951. Lui C. and Chiou W.L., pH shift during equilibrium dialysis, Biopharm. Drug. Dispos., 7:309-314, 1986. Lunde P.K.M., Rane A., Yaffe S.J., Lund L. and Sjoqvist F., Plasma protein binding of diphenylhydantoin in man: interaction with other drugs and the effect of temperature and plasma dilution, Clin. Pharmacol. Ther., 11:846-855, 1971. MacKichan J.J., Pharmacokinetic consequences of drug displacement from blood and tissue protein, Clin. Pharmacokin., 9:32-41, 1984. MacKichan J.J., Protein binding drug displacement interactions: fact or fiction ? Clin. Pharmacoki.16:65-73, 1989. Malfatto G., Zaza A., Forster M., Sodowick B., Danilo P. Jr. and Rosen M.R., Electrophysiologic, inotropic and antiarrhythmic effects of propafenone , 5-hydroxypropafenone and N-depropylpropafenone, J. Pharm. Exp. Ther., 246:419-426, 1988. McElnay J.C. and D'Arcy P.F., Protein binding displacement interactions and their clinical importance, Drugs, 25:495-513, 1983. McLean A.J., Isbister C, Bobik A. and Dudley F.J., Reduction of first-pass hepatic clearance of propranolol by food, Clin. Pharmac. Ther., 30:31-34, 1981 McLean A.J., McNamara P.J., duSouich P., Gibaldi M. and Lalka D., Food, splanchnic blood flow, and bioavailability of drugs subject to first-pass metabolism, Clin. Pharmacol. Ther., 24:5-10, 1978. Mcleod A.A, Stiles G.L. and Shand D.G., Demonstration of beta adrenoceptor blockade by propafenone hydrochloride: clinical, pharmacologic, radioligand binding and adenylate cyclase activation studies, J. Pharmacol. Exp. Ther., 228:461-466, 1984. Meddis R.(Ed), Statistics using ranks: A unified approach, Basil Blackwell Inc., New York, p.288-290, 1984. Meijer D.K.F., Weert B. and Vermeer G.A., Pharmacokinetics of billiary excretion in man. VI. Indocyanine green, Eur. J. Clin. Pharmacol., 125 35:295-303, 1988. Melander A., Daniel son K., Scherster B. and Wahlin E., Enhancement of the bioavailability of propranolol and metoprolol by food, Clin. Pharmac. Ther., 22:108-112, 1977. Metzler CM., Elfring G.L. and McEwen A.J., A package of computer programs for pharmacokinetic modelling, Biometrics 30:253-267, 1988. Muller-Peltzer H., George G.and Neugebauer G., Beta-blocking and electrophysiologic effects of propafenone in volunteers, Eur. J. Clin. Pharmacol. 25:831-833, 1983. Naccarella F., Bracchetti D., Palmieri M., Marchesini B. and Ambrosioni E., Propafenone for refractory ventricular arrhythmias correlation with drug plasma levels during long-term treatment, Am. J. Cardiol. 54:1008-1014, 1984. Nilsen O.G., Serum albumin and lipoproteins as the quinidine binding molecules in normal human sera, Biochem. Pharmacol. 25:1007-1012, 1976. Nilsen O.G. and Jacobsen S., The binding of quinidine to protein fractions of normal human sera, Biochem. Pharmac, 24:995, 1975. Odar-Cederlof I. and Borga 0., Impaired plasma protein binding of phenytoin in uremia and displacement effect of salicylic acid, Clin. Pharm. Ther., 20:36-47, 1977. Olanow C.W., Finn A.L. and Prussak C, The effects of salicylate on pharmacokinetics of phenytoin, Neurology, 31:341-342,1981. Ohnhaus E.E., Methods of the assessment of the effect of drugs on liver blood flow in man, Br. J. Clin. Pharmac. 7:223-229, 1979. Owen V.M.J., Laboratory note estimation of indocyanine green concentrations, Clin. Biochem. 6:132-135, 1973. Pang K.S. and Rowland M., Hepatic clearance of drugs. I. theoretical considerations of a "well-stirred" model and "parallel tube" model. Influence of hepatic blood flow, plasma and blood cell binding, and the hepatocellular enzymatic activity on hepatic drug clearance, J. Pharmacokin. Biopharm., 5:625-653, 1977. Pang K.S., Hepatic clearances of drugs and metabolites, TIPS, 1:247-251, 1980. Paumgartner G., The handling of indocyanine green by the liver, Schweiz. Med. wschr., 105(4):1-30, 1975. Paxton J.W., Effects of aspirin on salivary and serum phenytoin kinetics in healthy subjects, Clin. Pharmacol. Ther., 27:170-178, 1980. Piafsky K.M., Borga 0., Odar-Cederlof I., Johansson C. and Sjoqvist F., Increased plasma protein binding of propranolol and chlorpromazine 126 mediated disease-induced elevations of plasma ai-acid glycoprotein, N. Engl. J. Med., 299:1435-1439, 1978. Pinckard R.N., Hawkins D.and Farr R.S., The influence of acetyl salicylic acid on the binding of acetrizoate to human albumin, Ann. N. Y. Acad. Sci., 226:341-354, 1973. Rappaport P.L. and Thiessen J.J., High-pressure liquid chromatographic analysis of indocyanine green, J. Pharm. Sci. 71:157-161, 1982. Rapapdrt E., Ketterer S.G. and Wiegand B.D., Hepatic clearance of indocyanine green, Clin. Res. 7:289, 1959. Reidenberg M.M. and Affrime M., Influence of disease on binding of drugs to plasma proteins, Ann. N. Y. Acad. Sci., 226:115, 1973. Reubi F.C., Grossweiler N. and Gurtler R., Renal circulation in man studied by means of a dye-dilution method, Circ. 33:426, 1966. Rosenthal H.E., A graphic method for the determination and presentation of binding parameters in a complex system, Ana. Biochem. 20:525-532, 1967. Rowland M., Plasma protein binding and therapeutic drug monitoring, Ther. Drug Monit., 2:29-37, 1980. Sager G., Nilsen O.D. and Jacobsen S., Variable binding of proranolol in human serum, Biochem. Pharmacol. 28:905-911, 1979. Scatchard G., The attractions of proteins for small molecules and ions, Ann. N.Y. Acad. Sci.51:660-672, 1949. Scott D.O., Sorensen L. and Lunte C.E., In vivo microdialysis sampling coupled to liquid chromatography for the study of acetaminophen metabolism, J. Chromatogr., 506:461-469, 1990. Scott D.O., Bell M.A. and Lunte C.E., Microdialysis-perfusion sampling for investigation of phenol metabolism, J. Pharm. Biomed. Ana., 7:1249-1259, 1989. Sedman A.J. and Wagner J.G., A decision making pharmacokinetic computer program, Publication distribution service, 615 East University Avenue, Ann Arbor, Michigan 48106, (AUT0AN), 1976. Seipel L. and Breithardt G., Propafenone: a new antiarrhythmic drug, Eur. Heart J., 1:309-313, 1980. Shand D.G., ai-Acid glycoprotein and plasma lidocaine binding, Clin. Pharmacokin. 9(Suppl. 1):27-31, 1984. Shen E.N., Sung R.J., Morady F., Schwartz A.B. and Scheinman M.M., Electrophysiologic and haemodynamic effects of intravenous propafenone in patients with recurrent ventricular tachycardia, J. Am. College Cardiol. 3:1291-1297, 1984. 127 Siddoway L.A., Thompson K.A., McAllister C.B., Wang T., Wilkinson G.R., Roden D.M. and Woosley R.L., Polymorphism of propafenone metabolism and disposition in man: clinical and pharmacokinetic consequences, Circ, 75:785-791, 1987. Somberg J.C, Tepper D. and Landau S., Propafenone: a new antiarrhythmic agent, Am. Heart J. 115:1274-1279, 1988. Steinberg I.Z. and Schachman H.K., Ultracentrifugation studies with absorption optics. V. Analysis of interacting systems involving macromolecules and small molecules, Biochemistry, 5:3728-3747, 1966. Svensson C.K., Edwards D.J., Mauriello P.M., Barde S.H., Foster A.C, Lane R.A., Middelton E. Jr. and Lalka D., Effect of food on hepatic blood flow: Implications in the 'food effect' phenomenon, Clin. Pharmac. Ther., 34:316-323, 1983. Svensson C.K., Mauriello P.M., Barde S.H., Middelton E. Jr. and Lalka D., Effect of carbohydrate on estimated hepatic blood flow, Clin. Pharmac Ther., 35:660-665, 1984 Svensson C.K., Woodruff M.N. and Lalka D., Influence of protein binding and the use of unbound (free) drug concentration, In: Evans W.E., Schentag J.J. and Jusko W.J. (eds): Applied pharmacokinetics: principles of therapeutic drug monitoring, Applied Therapeutics, San Francisco, CA, pp.187-219, 1986. Thompson K.A., Iansmith D.H.S., Siddoway L.A., Woosley R.L. and Roden D.M., Potent electrophysiologic effects of the major metabolites of propafenone in canine purkinje fibers, J. Pharmacol. Exp. Ther. 244:950-955, 1988. Tillement J-P, Albengres E. and Urien S., Plasma binding of drugs, Pharmacy International, 1:64-65, 1980. Tonn G.R., The in vitro characterization of the drug-protein binding of racemic propafenone, and its active metabolite 5-hydroxypropafenone in human serum, and in solutions of isolated human serum proteins, M.Sc Thesis, University of British Columbia, 1990. Tozer T.N., Gambertoglio J.G., Furst E.D., Avery D.S. and Holford N.H.G., Volume shifts and protein binding estimates using equilibrium dialysis: application to prednisolone binding in humans, J. Pharm. Sci., 72:1442-1446, 1983. Vaugham Williams E.M., Classification of antiarrhythmic drugs, In: Sandoe E., Flensted-Jensen E. and Olesen K.H. (eds.): Cardiac arrhythmias, Sodertalje Ad Astra, Sweden, pp449-473, 1970. Valenzuela C, Delgado C. and Tamargo J., Electrophysiological effects of 5-hydroxypropafenone on guinea pig ventricular muscle fibers, J. Cardiovas. Pharmacol., 10:523-529, 1987. 128 Vita J.A., Friedman P.L., Cantillon C. and Antman E.M., Efficacy of intravenous propafenone for the acute management of atrial fibrillation, Am. J. Cardiol. 63:1275-1278, 1989. Von PhilipsbornG., Gries J., Hoffman H.P., Kreiskott H. and Kretschmar R., Pharmacological studies with propafenone and its main metabolite 5-hydroxy-propafenone, Arzneim.-Forsch./Drug Res., 34:1489-1497, 1984. Webb J.L., Enzyme and metabolic inhibitors, Volume 1, 1963. Wheeler H.O., Cranston W.I. and Meltzer J.I., Hepatic uptake and biliary excretion of indocyanine green in the dog, Proc. Soc. Exp. Biol. (N.Y.) 99:11, 1958. Wood M., Shand D.G. and Wood A.J., Altered drug binding due to use of indwelling heparinized cannulas (heparin lock) for sampling, Clin. Pharmacol. Ther., 25:103-107, 1979a. Wood M., Shand D.G. and Wood A.J.propranolol binding in plasma during cardiopulmonary bypass, Anesthesiology 51:512-516, 1979b. Zar J.H. (ed.), Biostatistical analysis, Prentice-Hall, Inc., Englewood Cliffs, N.J., second edition, p.126-131, 1984. Zoble R.G., Kirsten E.B., Brewington J. and the Propafenone research group Tampa, FT a. Pharmacokinetic and pharmacodynamic evaluation of propafenone in patients with ventricular arrhythmia, Clin. Pharmacol. Ther. 45:535-541, 1989. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0098731/manifest

Comment

Related Items