@prefix vivo: . @prefix edm: . @prefix ns0: . @prefix dcterms: . @prefix skos: . vivo:departmentOrSchool "Pharmaceutical Sciences, Faculty of"@en ; edm:dataProvider "DSpace"@en ; ns0:degreeCampus "UBCV"@en ; dcterms:creator "Bhattacharjee, Rathindra Chandra"@en ; dcterms:issued "2010-08-25T20:33:05Z"@en, "1988"@en ; vivo:relatedDegree "Master of Science - MSc"@en ; ns0:degreeGrantor "University of British Columbia"@en ; dcterms:description """Propafenone is a new class 1C antiarrhythmic agent with additional calcium antagonistic and beta-blocking activities. Clinically it is effective in the treatment of supraventricular and ventricular tachycardia, atrial and ventricular fibrillation, ventricular premature contractions and for the management of Wolf-Parkinson-White syndrome. In North America it is still an investigational drug. Propafenone is a chiral drug and is used clinically in the racemic form. The enantiomers of numerous chiral drugs have been shown to differ in their disposition kinetics in the body due to their stereoselective pharmacokinetics and/or pharmacodynamic properties. Two enantiomers are thus often considered as two different entities. The relative antiarrhythmic activities of individual enantiomers of propafenone have not been studied, nor their pharmacokinetic parameters have been elucidated. In order to study the possible enantioselective role of propafenone in the body, a stereoselective assay method would be required. The present study describes the development of a sensitive and stereoselective chromatographic assay method for the simultaneous determination of the two enantiomers of propafenone in human plasma. Attempts for direct separation of the enantiomers of propafenone included several GLC and HPLC chiral stationary phases. The chiral stationary phases were a Chirasil-Valʳ GLC stationary phase, a Pirkle 2,4 dinitro-(D)-phenylglycine HPLC stationary phase and a β-cyclodextrin HPLC stationary phase. Unfortunately, these did not resolve the enantiomers of propafenone. Formation of the diastereomers with R(+)-⍺-methyl benzyl isocyanate and racemic propafenone were partially resolved on a reverse phase HPLC using a 5 u, 25 x 0.45 cm i.d. ODS column and methanol/water (70:30) as the mobile phase. However, due to the long retention time (42 min), incomplete resolution (RS=1.15) and poor sensitivity for detection (500 ng of each enantiomer injected) this method was not deemed suitable for the pharmacokinetic studies planned, since the therapeutic plasma concentration range of propafenone is 64-1044 ng/mL. The second chiral derivatizing reagent, 2,3,4,6-tetra-0-acetyl-β-D-glucopyranosylisothiocyanate (GITC), was synthesized in our laboratory. This reagent gave better resolution of the enantiomers (RS=1.4) within 15 minutes with enhanced sensitivity for detection (150 ng of each enantiomer injected). To further optimize the limit of detection for future pharmacokinetic studies of propafenone, R(-)-1 -(naphthyl) ethylisocyanate, a chiral derivatizing agent, was employed. This reagent reacted with racemic propafenone and permitted the resolution of both enantiomers within 24 minutes (R5=l.25) and the minimum level of detection was 100 ng (at the detector) for each enantiomer of propafenone. Using this method, linearity was established over the concentration range, 125-1000 ng for each enantiomer (injected) with a coefficient of determination (r²) of greater than 0.99. Reproducibility and precision of this assay method was obtained with an average coefficient of variability of 4.5% for the R(-) enantiomer and 7.2% for S(+) enantiomer at concentrations of 125-1000 ng/mL. Below the lower quantity, the NEIC-propafenone reaction virtually stopped at the conditions set for derivatization. A similar lack of reactivity at low concentrations was also observed with the GITC-propafenone reaction. The absence of an autocatalysing effect of propafenone at lower nanogram levels, as well as two possible conformational forms of propafenone were also investigated. The existence of two conformational isomers of propafenone, due to intramolecular hydrogen bonding in aprotic solvents, was chromatographically verified. In addition, chromatographic separation of all the proposed conformers was obtained, indicating that enantiomeric separation and quantitation of propafenone enantiomers as their urea derivatives is substantially hindered. To eliminate hydrogen bonding interactions, the carbonyl group of propafenone was blocked with dansylhydrazine and subsequently derivatized with the chiral R(-)NEIC reagent. The HPLC resolution (RS=1.35) of this dual derivative was better than that using the R(-) NEIC reagent alone, and the minimum level of detection was 2.5 ng for each enantiomer. Unfortunately, this procedure still did not provide adequate assay precision and accuracy at the lower levels required for single dose pharmacokinetic studies."""@en ; edm:aggregatedCHO "https://circle.library.ubc.ca/rest/handle/2429/27800?expand=metadata"@en ; skos:note "DEVELOPMENT OF A SENSITIVE AND STEREOSELECTIVE HIGH PERFORMANCE LIQUID CHROMATOGRAPHIC ASSAY METHOD FOR PROPAFENONE ENANTIOMERS IN HUMAN PLASMA RATHINDRA CHANDRA BHATTACHARJEE M.Sc. (Chemistry), University of Dacca, 1973 A THESIS SUBMITTED AS PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF THE FACULTY OF GRADUATE STUDIES (Faculty of Pharmaceutical Sciences, D i v i s i o n of Pharmaceutical Chemistry) We accept t h i s thesis as conforming to the required standard The University of B r i t i s h Columbia January, 1988 ©Rathindra C. Bhattacharjee .1988 BY MASTER OF SCIENCE i n 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 1956 Main Mall Vancouver, Canada V6T 1Y3 Date ^ MM DE-6(3/81) i i ABSTRACT Propafenone is a new class IC antiarrhythmic agent with additional calcium antagonistic and beta-blocking act iv i t ies . C l in i ca l l y i t is effective in the treatment of supraventricular and ventricular tachycardia, atr ia l and ventricular f i b r i l l a t i o n , ventricular premature contractions and for the management of Wolf-Parkinson-White syndrome. In North America i t is s t i l l an investigational drug. Propafenone is a chiral drug and is used c l i n i c a l l y in the racemic form. The enantiomers of numerous chiral drugs have been shown to d i f fe r in their disposition kinetics in the body due to their stereoselective pharmacokinetics and/or pharmacodynamic properties. Two enantiomers are thus often considered as two different ent i t ies . The relative antiarrhythmic act iv i t ies of individual enantiomers of propafenone have not been studied, nor their pharmacokinetic parameters have been elucidated. In order to study the possible enantioselective role of propafenone in the body, a stereoselective assay method would be required. The present study describes the development of a sensitive and stereoselective chromatographic assay method for the simultaneous determination of the two enantiomers of propafenone in human plasma. Attempts for direct separation of the enantiomers of propafenone included several GLC and HPLC chiral stationary phases. The chiral stationary phases were a Chirasi l -Val^ GLC stationary phase, a Pirkle 2,4 dinitro-(D)-phenylglycine HPLC stationary phase and a p'-cyclodextrin HPLC stationary phase. Unfortunately, these did not resolve the enantiomers of propafenone. Formation of the diastereomers with R(+)-or-methyl benzyl isocyanate i i i and racemic propafenone were part ia l ly resolved on a reverse phase HPLC using a 5 u, 25 x 0.45 cm i .d . ODS column and methanol/water (70:30) as the mobile phase. However, due to the long retention time (42 min), incomplete resolution (R$=1.15) and poor sensit iv ity for detection (500 ng of each enantiomer injected) this method was not deemed suitable for the pharmacokinetic studies planned, since the therapeutic plasma concentration range of propafenone is 64-1044 ng/mL. The second chiral derivatizing reagent, 2,3,4,6-tetra-0-acetyl-fi-D-glucopyranosylisothiocyanate (GITC), was synthesized in our laboratory. This reagent gave better resolution of the enantiomers (R$=1.4) within 15 minutes with enhanced sensit iv ity for detection (150 ng of each enantiomer injected). To further optimize the l imit of detection for future pharmacokinetic studies of propafenone, R(-)-1 -(naphthyl) ethylisocyanate, a chiral derivatizing agent, was employed. This reagent reacted with racemic propafenone and permitted the resolution of both enantiomers within 24 minutes (R5=l.25) and the minimum level of detection was 100 ng (at the detector) for each enantiomer of propafenone. Using this method, l inear ity was established over the concentration range, 125-1000 ng for each enantiomer (injected) with a coeff ic ient of determination (r^) of greater than 0.99. Reproducibility and precision of this assay method was obtained with an average coefficient of var iab i l i ty of 4.5% for the R(-) enantiomer and 7.2% for S(+) enantiomer at concentrations of 125-1000 ng/mL. Below the lower quantity, the NEIC-propafenone reaction v i r tua l ly stopped at the conditions set for derivatization. A similar lack of reactivity at low concentrations was also observed with the iv GITC-propafenone reaction. The absence of an autocatalysing effect of propafenone at lower nanogram levels, as well as two possible conformational forms of propafenone were also investigated. The existence of two conformational isomers of propafenone, due to intramolecular hydrogen bonding in aprotic solvents, was chromatographically ver i f ied. In addition, chromatographic separation of a l l the proposed conformers was obtained, indicating that enantiomeric separation and quantitation of propafenone enantiomers as their urea derivatives is substantially hindered. To eliminate hydrogen bonding interactions, the carbonyl group of propafenone was blocked with dansylhydrazine and subsequently derivatized with the chiral R(-)NEIC reagent. The HPLC resolution (R$=1.35) of this dual derivative was better than that using the R(-) NEIC reagent alone, and the minimum level of detection was 2.5 ng for each enantiomer. Unfortunately, this procedure s t i l l did not provide adequate assay precision and accuracy at the lower levels required for single dose pharmacokinetic studies. V TABLE OF CONTENTS ABSTRACT i i TABLE OF CONTENTS v LIST OF FIGURES X LIST OF TABLES x i i LIST OF SCHEMES xi i i SYMBOLS AND ABBREVIATIONS xiv ACKNOWLEDGEMENT xvi i i CHAPTER PAGE 1. INTRODUCTION 1 2. BACKGROUND 2 2.1 Chemistry 2 2.2 Cl in ical Indications of Propafenone 3 2.3 - Comparative Cl inical Tr ia ls against other Antiarrhythmic Agents 4 2.4 Plasma Levels and Therapeutic Effectiveness . . . .5 2.5 Selected Pharmacodynamic Studies 6 2.5.1 Pharmacology 6 2.5.2 Electrophysiological Effects on the Heart 7 2.5.3 Haemodynamic Effects 8 2.6 Arrhythmia Supression 8 2.7 Adverse Effects 9 2.8 Drug Interactions 9 2.9 Metabolism of Propafenone 10 2.10 Pharmacokinetics of Propafenone 11 2.11 Protein-Binding of Propafenone 12 2.12 Stereoselectivity in Drug Action 13 2.12.1 Absorption 14 2.12.2 Presystemic Elimination 14 2.12.3 Distribution 15 2.12.4 Metabolism 17 2.12.5 Renal Clearance 19 2.13 Stereoselective Drug Analysis 20 vi 2.14 Analytical Methods for Propafenone 25 2.14.1 Analytical Methods for R,S Propafenone.25 2.14.2 Analytical Method for Propafenone Enantiomers 26 2.15 Objective of the Study 27 3. EXPERIMENTAL 28 3.1 Supplies 28 3.1.1 Drug and Internal Standards 28 3.1.2 Chemicals and Reagents 28 3.1.3 Solvents 29 3.2 Chromatographic Stationary Phases 29 3.3 Equipment 30 3.4 Stock Solutions 31 3.4.1 R,S Propafenone Hydrochloride 31 3.4.2 R,S Propafenone Free Base 31 3.4.3 R,S-2'-[3-(ethyl ami no)-2(hydroxy) propoxy]-3-phenyl propiophenone (Li-1115) hydrochloride and Desipramine Hydrochloride (Internal Standards) . . 31 3.4.4 (0.2 M) Phosphate Buffer Solution of pH 2.8 31 3.4.5 Trichloroacetic Acid Solution (10% w/v) 31 3.5 Synthesis of 2,3,4,6 Tetra-O-Acetyl-0-D-Glucopyranosylisothiocyanate (GITC) 32 3.5.1 Purif ication of Acetobromo-a-D-glucose 32 3.5.2 Synthesis of GITC 32 3.6 Preparation of the Derivatives of R,S Propafenone .33 3.6.1 Heptafluorobutyric Anhydride (HFBA) Derivative 33 3.6.2 N-Trifluoroacetyl (L) Pro!oylchloride [N-TPC] Derivative 33 3.6.3 /J-Naphthoylchloride Derivative 34 3.6.4 R(+)-a-Methylbenzylisocyanate [R(+)MBIC] Derivative 34 v i i 3.6.5 Isopropylisocyanate Derivative . . . . 34 3.6.6 2,3,4,6-Tetra-O-acetyl-^-D-Glucopyranosylisothiocyanate (GITC Derivative 35 3.6.7 R(-)-l-(Naphthyl)ethylisocyanate [R(-)NEIC] Derivative 35 3.6.8 Propafenone-Dansyl Hydrazine-R(-)NEIC Derivative 35 3.6.9 Derivatization of Propafenone with (+)Naproxen 36 3.7 Preliminary GLC and HPLC Analysis of Propafenone using Chiral and Achiral Stationary Phases . . . . 37 3.7.1 Chiral GLC Phase 37 3.7.2 Achiral GLC Stationary Phase 37 3.7.3 Chiral HPLC Phases 38 3.7.4 Achiral HPLC Phases 39 3.8 HPLC Resolution of the Enantiomers of Propafenone .39 3.8.1 Resolution of the Diastereomeric R(+)MBIC Derivatives 39 3.8.2 Resolution of the Diastereomeric GITC Derivatives 40 3.8.3 Resolution of the Diastereomeric R(-)NEIC Derivatives 41 3.8.4 Resolution of the Diastereomeric Propafenone-Dansylhydrazine-R(-)NEIC Derivatives 41 3.9 Quantitative Analysis of Propafenone Enantiomers . 42 3.9.1 Selection of Internal Standard . . . . 42 3.9.2 Selection of External Standard . . . . 42 3.9.3 Selection of Catalysts 42 3.9.4 Extraction Solvents 43 3.9.5 Plasma Protein Precipitation 43 3.9.6 Optimization of Reaction Conditions for Derivatization 43 3.9.7 Stoichiometric Ratio of R(-)NEIC Reagent and Drug 44 3.9.8 Optimizing the Sensit ivity for Detection 45 3.9.9 Polarimetric Measurement of R(-)NEIC Derivatives of R,S Propafenone . . . . 45 3.9.10 Structural Identity of the Derivatives of R,S Propafenone . . . . 46 3.9.11 Efficiency of Recovery of Propafenone Enantiomers from Plasma 46 3.10 Assay of Propafenone Enantiomers by High-Performance Liquid Chromatography with UV Detection 47 vi i i 3.10.1 Extraction, Derivatization and HPLC Analysis of R,S Propafenone, using R(+)MBIC as a Chiral Derivatizing Reagent 47 3.10.2 Extraction, Derivatization and HPLC Analysis of R,S Propafenone using GITC as Chiral Derivatizing Reagent . . . . 48 3.10.3 Extraction, Derivatization and HPLC Analysis of R,S Propafenone using R(-) NEIC as Chiral Derivatizing Reagent . .48 3.11 Calibration Curve and Precision of Assay of R,S Propafenone R(-)NEIC Derivatives 49 3.12 Reverse-Phase Thin-Layer Chromatographic Analysis of Propafenone-GITC and Propafenone-Dansylhydrazine-GITC Derivatives 50 4. RESULTS AND DISCUSSION 52 4.1 Analytical Development for Chromatographic Resolution of Propafenone Enantiomers 52 4.1.1 Gas-Liquid Chromatographic Studies on Chiral and Achiral Stationary Phases . 52 4.1.2 High-Performance Liquid Chromatographic Studies on Chiral and Achiral Stationary Phases 57 4.2 High-Performance Liquid Chromatographic Resolution of Propafenone Enantiomers . 60 4.2.1 Resolution of R(+)-a-Methylbenzyl -isocyanate [R(+)MBIC] Derivatives of Propafenone Enantiomers 60 4.2.2 Resolution of 2,3,4,6 Tetra-O-Acetyl-/J-D-Gl ucopyranosyl i sothi ocyanate (GITC) Derivatives of Propafenone Enantiomers 62 4.2.3 Resolution of R(-)-l-(Naphthyl)ethyl-isocyanate [R(-)NEIC] Derivatives of Propafenone 62 4.3 Elution Order of R and S Propafenone 65 4.4 Configuration of Structures of R and S Propafenone Derivatives 66 4.5 Sensit ivity of Propafenone and its Derivatives to Detection by HPLC 66 4.6 Mechanism of Diastereomeric Resolution of Propafenone Enantiomers 70 ix 4.7 Intramolecular Hydrogen Bonding and Conformation of Propafenone 74 4.7.1 HPLC Separation of Conformers of Propafenone 74 4.7.2 Reaction of Propafenone Conformers . . . with Chiral Isocyanates and Resolution of the Diastereomers by HPLC 75 4.8 Mechanism of Propafenone Isocyanate Reaction . . . 79 4.9 Dual Derivatization of Propafenone and HPLC Separation of Enantiomers 83 4.10 Kinetics of Derivatization and HPLC Analysis . . . 86 4.11 Stoichiometric Ratio of R(-)NEIC Reagent to Propafenone 86 4.12 Recovery of Propafenone Enantiomers from Plasma . .88 4.13 Linearity and Reproducibility of Assay 90 5. SUMMARY AND CONCLUSIONS 95 6. REFERENCES 97 LIST OF FIGURES FIGURES 1. Structure of Propafenone Enantiomers 2. Representative Chromatograms Showing Unresolved Peaks of R,S Propafenone under Various GLC Conditions • • . 3. Stereochemical Interaction Between Propafenone Enantiomers and Ch i ras i l -Va l K Stationary Phase (A) and Pirkle Statonary Phase (B) 4. Representative Chromatograms Showing Unresolved Peaks of R,S Propafenone by HPLC 5. The General Mechanism of Inclusion-Complexing in /J-Cyclodextrin Stationary Phase (A), and the Inclusion Complex Formation of Propafenone Enantiomers within the /J-Cyclodextrin Cavity (B). . . . 59 6. Reverse Phase HPLC Separation of Enantiomers of Propafenone as their Diastereomeric R(+)-a-Methyl-benzyl Urea (A), 0-D-G1ucopyranosyl Thiourea (B) and R(-)-1-(naphthyl)ethyl Urea (C) Derivatives . . . .63 7. IR and HPLC Analyses of Acetobromo-a-D-glucose and GITC 64 8. UV Absorption of Propafenone and its Derivatives in Different Solvents 67 9. Total-Ion Mass Chromatograms of the Derivatives of S(+)Propafenone (A) and R(-)Propafenone (B) with R(-)-l-(Naphthyl)ethylisocyanate 68 10. EI Mass Spectra of S(+)Propafenone (A) and . R(-)Propafenone (B) as their R(-)-l-(Naphthyl) ethyl isocyanate Derivatives 69 11 Mechanism of Diastereomeric Separation of Propafenone Urea and Thiourea Derivatives 72 xi LISTS OF FIGURES cont'd 12. Structure of the Chiral Derivatizing Reagents 73 13. Representative HPLC Chromatograms of Underivatized (A) and Derivatized (B) Propafenone 76 14. IR Spectrum of Propafenone in Nujol Mull. . . . 77 15. Single vs. Dual Derivatization of Propafenone 82 16. HPLC Separation of the Enantiomers of Propafenone After Dual Derivatization 85 17. Time and Temperature Dependence for Derivative Formation 87 18. Stoichiometric Ratio of Reagent to Drug [R(-)-1-(Naphthyl)ethyl isocyanate/Propafenone] 89 xi i LIST OF TABLES TABLES PAGE 1. Data Representing the HPLC Separations of Enantiomers of Propafenone 61 2. Efficiency of Recovery of Propafenone Enantiomers from Plasma 92 3. Calibration Curve Data and Inter-Assay Var iabi l i ty of Propafenone Enantiomers in Plasma 93 4. Intra-Assay Var iabi l i ty of Propafenone Enantiomers in Plasma 94 xi i i LIST OF SCHEMES SCHEME PAGE 1. Schematic Diagram Showing the Intramolecular Hydrogen Bonding and Conformational Isomers of Underivatized (2,3) and Derivatized (4,5 & 6) Propafenone 78 2. Reaction Scheme for the Derivatization of Propafenone with R(-)-l-(Naphthyl)ethyl-isocyanate 80 3. Chemical Pathway for Dual Derivatization of Propafenone (2 & 3) to the Diastereomeric Urea Derivatives of Dansyl-Propafenone Hydrazone (5). . . 84 4. Assay Procedure for Propafenone Enantiomers by HPLC 91 xi v SYMBOLS AND ABBREVIATIONS AAG aj-acid glycoprotein AV atrio-ventricular CNS central nervous system CV coeff icient of variation DOPA 3-(3,4-dihydroxyphenyl)alanine dp/dt rate of development of le f t ventricular pressure ECD electron capture detector EI electron impact EI-MS electron impact mass spectrum (spectra) Em emission Ex excitation FID flame ionization detector GC/GLC Gas-liquid chromatography GITC 2,3,4,6 tetra-O-acetyl -/J-D-glucopyranosyl isothiocyanate HFBA heptafluorobutyric anhydride HPLC high-performance l iquid chromatography hr hour HV His-ventricular, His-bundle HSA human serum albumin i .d . internal diameter IPA isopropanol, 2-propanol I.S. internal standard XV reaction rate constant methyl benzyli socyanate m i l l i l i t r e molarity, molar micron microgram normality, normal naphthylethyli socyanate nanogram N-trif luoro-acetyl-L-proloylchloride atreoventricular conduction period of electrocardiogram premature ventricular contraction polytetrafluoroethylene intraventricular conduction period of electrocardiogram interval of the electrocardiogram representing ventricular depolarization and repolarization correlation coefficient coeff ic ient of determination reverse phase resolution factor retention time s ino-atr ial + standard deviation xv i TCA trichloroacetic acid TEA triethylamine TLC thin-layer chromatography uv ultraviolet VPB ventricular premature beats xvi i This thesis is dedicated to the loving memory of my father. xvi i i ACKNOWLEDGEMENT I would l ike to express my sincere gratidude to my thesis supervisor Dr. Keith McErlane for his valuable guidance, constant encouragement and support during the course of this study. I gratefully acknowledge the helpful discussion and criticisms from Drs. Frank Abbott, James Axel son and Charles Kerr as the members of my research committee. I want to extend my sincere thanks to Mrs. Barbara McErlane and David Kwok for their technical assistance. Special thanks to my wife Ayetri Bhattacharjee for her boundless patience and constant encouragement throughout the years of my studies. 1 1. INTRODUCTION Propafenone, 2-[3-(propyl amino)-2-(hydroxy)-(propoxy)]- 3-phenylpropiophenone is a new class 1C antiarrhythmic agent. The drug was synthesized in 1970 and its hydrochloride salt has been commercially available in West Germany (Knoll, AG) since 1977 as RytmonormR. Propafenone has also been approved for marketing in Spain, Italy and Portugal for the treatment of cardiac arrhythmias of varying origin. In North America i t is s t i l l an investigational drug. A number of studies of this new antiarrhythmic drug have appeared over the past ten years and have established its pharmacological (Keller et a l . . 1978, Kohlhardt, 1977, Palma et a l . . 1982) electrophysiological (Seipel et a l . . 1977, Walleffe et a l . , 1981) and antiarrhythmic (Durante et a l . , 1980, Meyer-Estorff et a l . , 1978) properties as well as its tox ic i t ies (Kretzschmer et a l . . 1983a, Von. Philipsborn, 1984). The drug may be administered either oral ly or intravenously. With oral administration, dose dependent systemic b ioavai labi l i ty of 12% to 50% has been reported (Hollmann et a l . . 1983, Palma et a l . . 1982, Connolly et a l . , 1983). The mean therapeutic plasma concentrations of propafenone have been reported to be 64 ng/ml to 1044 ng/ml and the mean h a l f - l i f e of the drug is from 3 to 6 hours (Hollmann et aj.., 1983, Connolly et a l . , 1983). Propafenone structurally resembles propranolol for i ts aryloxy propanolamine moiety, with a chiral center in its molecular arrangement. The drug is used as a racemic mixture of R and S enantiomers. About 25% of the drugs in use today contain chiral centres (Simonyi, 1984). The fact that drug enantiomers d i f fer chemically and 2 exhibit d i f fer ing biological ac t iv i t ies , has led to the suggestion that they be considered as different compounds (Simonyi, 1984, Ariens, 1986). The differences in intr ins ic biological act iv i ty of enantiomers are due mainly to enantioselectivity in protein binding, storage, and/or transport processes in the body. Thus, the plasma levels and pharmacokinetics of racemic drugs do not always ref lect those of the individual active enantiomers. Despite this knowledge, most chiral drugs, including propafenone, are used therapeutically as racemic mixtures. The relative act iv i ty of the individual enantiomers of propafenone have not, as yet, been determined nor have their pharmacokinetic parameters have been elucidated. The present study was therefore aimed at developing a sensitive high-performance l iquid chromatographic assay method for the simultaneous determination of both enantiomers of propafenone in human biological f lu ids. 2. BACKGROUND 2.1 Chemistry Propafenone structurally resembles the beta-adrenoceptor blockers which include propranolol. Its molecular formula is C21H27NO3 and its molecular weight is 341.46. Propafenone Hydrochloride (pKa=9) is a crysta l l ine white powder with a melting point of 172°C. It is only s l ight ly soluble in cold water. Propafenone base is also crystal l ine with a melting point of 69°C (observed). 3 3 R , S P R O P R A N O L O L Propafenone has three reactive functional groups. Due to the proximity of the ortho aryl carbonyl group to the active alcoholic and amino hydrogens in the side chain, propafenone is believed to have intramolecular hydrogen bonding, resulting in two conformational isomers. The relative abundance of the two isomers are expected to be dependent on the pH of the solvent and the conformational f l e x i b i l i t y of the two isomers, as has been shown for other such molecular arrangements (Kuhn et a l . , 1964). Except for chemical properties, the thermodynamic and so lubi l i ty parameters of the two conformational isomers are assumed to be almost identical . Propafenone, being a chiral drug, consists of two enantiomers. The absolute configurations of the two enantiomers have very recently been reported by Blaschke et al.,1987, and were determined to be the R(-) and S(+) propafenone respectively. 2.2 Cl in ica l Indications of Propafenone Propafenone has been shown to be an effective antiarrhythmic agent due to its blockade of the fast inward sodium current (Kohlhardt, 1984). 4 In patients with Wolf-Parkinson-White syndrome, Breithardt, et a j . , 1984, reported that propafenone was an effective agent for long-term management of this disease. The drug is useful in decreasing the multiple episodes of recurrent ventricular tachycardia or ventricular f i b r i l l a t i o n (Heger et a l . , 1984). Podrid et a l . , 1982, demonstrated that propafenone suppressed complex ventricular ectopy without depressing the ejection fraction in patients with normal lef t ventricular function, but i t reduced the ejection fraction in those exhibiting myocardial dysfunction. Propafenone was evaluated as a potent antiarrhythmic agent in arrhythmias that appear to depend on sympathetic drive (Coumel et al_., 1984). The drug decreases the anterograde conduction in the accessory pathway as well as in the normal pathway, thereby protecting the heart against rapid ventricular rates during circus movement tachycardia, atr ia l f i b r i l l a t i on or during atr ial f lut ter (Waleffe et a l . . 1981). 2.3 Comparative Cl inical Tr ia ls against other Antiarrhythmic Agents Asshauer and Vorpahl, 1977, compared the oral doses of propafenone (300 mg) with procainamide (500 mg) in patients with stable ventricular extrasystoles. Propafenone was found to s ignif icantly reduce the number of extrasystoles, while they remained refractory with procainamide. Propafenone was established to be superior to disopyramide, perhexiline and quinidine in patients with history of myocardial infarction and ectopic ventricular arrhythmias (De Brito et al.,1979). However, Klempt and his co-workers, 1982, evaluated propafenone (900 mg) against mexiletine (600 mg) and flecainide (400 mg) in patients following 5 myocardial infarction. The percent reduction in single and repetitive ventricular premature beats (VPB) were 80 and 97 for propafenone, 53 and 83 for mexiletine and 94 and 99 for f lecainide. Both flecainide and propafenone lengthened the duration of PQ and QRS intervals seen in the electrocardiogram, while mexiletine did not change these parameters. The subjective side effects of headache and dizziness were similar for a l l three drugs. 2.4 Plasma Levels and Therapeutic Effectiveness The plasma concentration-effect relationship of propafenone on atrioventricular conduction time (PR interval) in patients with cardiac arrhythmias revealed that a significant correlation between serum propafenone concentration and changes in PR interval was present (Keller, et al.,1978). The maximum increase in PR interval was achieved between two to three hours at the highest plasma concentration of propafenone and returned to baseline as the drug concentration f e l l . A similar concentration-effect relationship was observed in the suppression of ventricular premature beats when propafenone concentrations in plasma reached levels of 64 to 1044 ng/mL (Connolly et al.,1983). However, marked inter-subject variations in the effective plasma concentrations were also noted. A potent antiarrhythmic effect has also been demonstrated for 5-hydroxy propafenone, a major metabolite of propafenone, in rats, and was correlated with greater antiarrhythmic response than the parent drug (Kretzschman et a l . , 1983, Siddoway et a l . , 1983b, 1987). 6 2.5 Selected Pharmacodynamic Studies 2.5.1 Pharmacology Racemic Propafenone has been shown to be an effective drug in the treatment of supraventricular and ventricular arrhythmias and ectopic beats (Hollmann et a l , 1983). In both in vitro and in vivo experiments in animal models, i t has been found to be a membrane stabi l iz ing antiarrhythmic drug which inhibits the fast Sodium (Na+) inward current (Ledda et aj_., 1981). The local anesthetic effect of propafenone corresponds to that of procaine, and at higher concentrations calcium antagonistic effects were also noted. It i s , however, 100 times less potent than verapamil (Scholz, 1982) in this later action. The effect of oral and intravenous propafenone in patients with ventricular arrhythmias has been evaluated. Oral propafenone was found to be more effective than the intravenous formulation in that a complete regression of PVC was noted in 85% of cases; whereas the intravenous drug was found to be effective in 50% of cases in suppressing the severe ventricular arrhythmias (Ledain et a l . , 1982). Propafenone contains an aryloxy propanolamine structure, which is common to beta adrenergic blockers. The weak beta sympatholytic effect of the drug has been observed in inhibiting the isoprenaline-induced positive inotropic and chronotropic effects in isolated guinea pig atria (Ledda et aj.., 1981). The degree of inhibition of these beta sympatholytic effects was found to be 50 times greater in tracheal muscles than in myocardium (Scholz, 1982). The beta sympatholytic effect seen in human volunteers was estimated to be 2 to 5% of that of other typical beta blocking agents (Von.Philipsborn, 1984). Propafenone 7 was found to influence exercise-induced tachycardia as well as the resting heart rate, and to a certain extent, the systol ic and diastol ic blood pressure (Muller-Peltzer et a l . , 1983). Anticonvulsant effects have also been demonstrated for propafenone at a dose seven times greater than that required for the suppression of arrhythmias (Kretzschmer et a l . , 1983a). The normal and dose-induced ECG pattern in rabbits indicated that at higher doses some CNS side effects occur, such as dizziness and fatigue (Von. Philipsborn, 1984). 2.5.2 Electrophysiological Effects on the Heart The antiarrhythmic effect of propafenone is based on an inhibition of the rapid transmembrane Na+ influx by depressing the maximal rate of depolarization (V m a x ) in ventricular myocardium and Purkinje fibres without modifying the resting potential (Ledda et a l . , 1981). Propafenone was graded as a class IC antiarrhythmic agent, although i t looses its virtual speci f ic i ty for the Na+ system and produces additional changes in the plateau and repolarization phase of the action potential, consistent with a decrease of slow calcium-induced inward current (Ledda et a l . , 1981). At higher concentrations of the drug, i t has an inhibiting effect on SA node and gains the property of a class 4 antiarrhythmic agent (Kohlhardt, 1984). In patients with ventricular arrhythmias, the drug causes a 16% increase in the PR interval of the electrocardiogram and a prolongation of the QRS interval by 18%. A slight increase of the QT interval has also been reported at higher doses (Salerno et a l . , 1984). The drug thus prolongs the atrioventricular and intraventricular conduction times and le f t ventricular systolic function (Hollmann, 1983). 8 2.5.3 Haemodynamic Effect Bachour et a l . . 1977, f i r s t reported a change in the haemodynamic parameters after intravenous administration of propafenone to patients with coronary heart disease. A significant r ise of right atr ia l and pulmonary artery pressure was observed between five to twenty minutes, with an instantaneous increase in peripheral resistance and decrease in cardiac output. F i l l et a l . , 1977, reported that heart rate and le f t ventricular end-systolic pressure were relat ively unchanged after intravenous administration of propafenone but with a significant decrease in le f t ventricular dp/dt ratio within 10 minutes. In patients with low, grade II - V arrhythmias, propafenone demonstrated a negative chronotropic effect of 16% of the patients. Mexiletine, tocainide and disopyramide showed similar cardiac effects in 17, 19, and 30% of patients, respectively (Wester et a l . , 1982). 2.6 Arrhythmia Suppression The concentration-response relation for suppression of arrhythmias by propafenone shows substantial interindividual var iab i l i ty . Salerno et a l . , 1984, reported minimum effective trough plasma concentrations in patients ranging from 91 to 3271 ng/mL. Similarly, Connolly et aj.., 1983, observed that the effective concentrations ranged from 64 to 1044 ng/ml, and that the highest concentration required in their patients for 90% suppression of ectopic beats was approximately 800 ng/mL. Siddoway and his associates, 1984, reported that the mean concentration ranges for effective dosing were 143 to 1992 ng/mL. 9 2.7 Adverse Effects Propafenone is considered to be relat ively safe as compared to other antiarrhythmic agents. The common side effects after prolonged use of the drug are bitter taste, mild nausea and dry mouth. Moreover, central nervous system side effects, including visual blurring, dizziness and paraesthesias, were also noted when the serum propafenone concentrations were 900 ng/mL or more (Connolly et a l . , 1983). These effects were more pronounced in poor metabolizers than extensive metabolizers of propafenone (Siddoway et a l . , 1987). Cardiovascular side effects are rare but drug-induced transient bradycardia and ventricular f i b r i l l a t i on were observed in a few cases (Beck et a l . , 1980). 2.8 Drug Interactions Limited drug interaction studies have been reported in the l i terature for propafenone. However, Beck et a l . , 1982, reported that co-administration of propafenone and propranolol increased the PR interval by an average of 25%, whereas administration of propafenone and propranolol separately increased the PR interval by 13 and 19%, respectively. This indicates that the additive effect of both substances can lead to an inhibition of atrioventricular conduction and of sinus node function. Similar increases in effect were observed after the co-administration of propafenone and l idof lazine (Beck et a l . , 1980). Due to the high level of protein-binding of propafenone (97%), a possible drug interaction would be expected with warfarin. While this has not been determined, phenprocoumon did not interfere with 10 propafenone disposition. A 37% increase in serum digoxin levels was observed after co-administration of propafenone in d ig i ta l ized patients (Knoll AG report on propafenone, 1984, Belz et a l . , 1982). 2.9 Metabolism of Propafenone Propafenone, l ike most antiarrhythmic drugs, undergoes extensive hepatic oxidative metabolism and less than 1% of the drug is excreted unchanged in urine and bi le (Marchensini et a l . , 1983). The glucuronide and sulphate conjugates of 5-hydroxy propafenone are the main metabolites in plasma, urine and b i le, followed by the conjugates of hydroxy-methoxy propafenone and propafenone i t se l f (Hege et a l . , 1984, Kretzschman et a l . , 1983). 5-hydroxy propafenone was found to be twice as potent as propafenone in its antiarrhythmic efficacy in rats (Hege et aj.., 1984). A minor metabolite, N-depropyl propafenone was also identif ied in plasma, although its pharmacological and antiarrhythmic act iv i t ies are s t i l l uncertain (Philipsborn, 1984, Kates et a l . , 1985). Like lorcainide, hepatic metabolism of propafenone involves a saturable process which accounts for the non-linear characterist ics of drug metabolism (Connolly et a l . , 1983). Siddoway et a l . , 1983b, 1987, c lass i f ied arrhythmic patients as either extensive or poor metabolizers of propafenone on the basis of their ab i l i ty to metabolize debrisoquine by cytochrome P-450 isozyme(s). The steady state concentration of 5-hydroxy propafenone exceeded those of the parent drug in extensive metabolizers, while poor metabolizers accumulated low concentrations of metabolite and accrued higher concentrations of the parent drug. Metabolism of chiral drugs is generally stereospecific and i t could be that, l ike propranolol, the racemic drug propafenone may also 11 exhibit stereospecific metabolism. 2.10 Pharmacokinetics of Propafenone Propafenone is readily absorbed from the gastrointestinal tract and reaches a peak plasma concentration within 2 to 3 hours with dose dependent b ioavai labi l i ty ranging from 12 to 50% (Hollmann et a l . , 1983, and Connolly et al_., 1983). The change in b ioavai labi l i ty with variation of dose is nonlinear which indicates that propafenone is subject to saturable processes during presystemic l i ver passage. Food also increases the bioavai labi l i ty of propafenone. A 147 % increase of b ioavai labi l i ty has been reported by Axelson et a l . , 1987, when propafenone was taken oral ly with a standard breakfast. The disposition kinetics after intravenous administration of the drug shows that propafenone is rapidly distributed with a distribution half-l i f e of 4.7 minutes. The volume of the central compartment is approximately 0.7 to 1.1 L/kg (Hollmann et aj.., 1983). The plasma concentration decays rapidly in a bi-exponential manner. The steady-state volume of distribution is approximately 1.9 to 3 L/kg (Seipel et aj.., 1980). More than 97% of the drug is protein-bound and the largest concentration of propafenone was found in lungs, which is 10-fold higher than that found in heart muscle (Seipel et a l . , 1980). It has been reported that propafenone single-dose pharmacokinetics, based on total serum concentrations, are non-linear (Connolly et a l . , 1984). The same authors reported a 10-fold increase in steady-state serum concentrations for a 3-fold increase in dose from 300 mg to 900 mg of propafenone. The apparent terminal elimination h a l f - l i f e of propafenone ranges from 3.6 to 4.6 hours, however, 12 variations in ha l f - l ives from 1.8 to 17.2 hours have also been reported due to interindividual var iab i l i ty (Keller et_al_., 1978, Hollmann et aj.., 1983, and Connolly et aj.., 1983). In patients with complex ventricular arrhythmias, hal f - l ives ranging from 1.8 to 32.3 hours have also been reported (Siddoway et a l . , 1983). There was also great var iab i l i ty in the clearance following oral administration, with a range of 192 to 4918 mL/min., and dose-corrected mean plasma concentration ranges of 0.13 to 3.32 ng/mL per milligram of dose. Clearance also decreases during long-term therapy (Siddoway et aj_., 1983). Very recently, Siddoway and his associates, 1987, reported that propafenone disposition kinetics d i f fer between poor metabolizers and extensive metabolizers and the authors speculated that the nonlinear relationship between dose and plasma concentration is l i ke ly due to saturation of the cytochrome P-450 isozyme(s) involved. 2.11 Protein-Binding of Propafenone Neumann, (1978) f i r s t studied the plasma protein-binding of ^H-propafenone in man using an equilibrium dialysis technique. At plasma concentrations up to 1600 ng/mL, plasma protein-binding was about 97%. Protein- binding was observed to decrease s l ight ly at higher concentrations of the drug. Similar protein-binding results have been reported in the dog and the rat (Neumann, 1979). Propafenone, in the concentration range of 50 to 1250 ng/mL, equilibrates rapidly between human erythrocytes and plasma in the ratio of 0.66 to 0.74 which indicated that the drug is extensively bound to plasma proteins at therapeutic concentrations (Hollmann et a l . , 1983). Higuichi et a l . , 1985, determined the plasma protein binding of the drug in both in vitro 13 and in vivo and reported that plasma protein- binding of propafenone was 69 to 88% in rats, 94 to 95% in dogs and 76 to 89% in humans. In another study, Steinbach et al_., 1983, reported that about 98% of propafenone is bound to plasma proteins in humans. Basic drugs usually bind to AAG, and this binding can also be stereoselective. The S(+) enantiomer of disopyramide was reported to bind preferential ly with AAG as compared to the R(-) enantiomer (Lima et a l . , 1984,), and S propranolol was noted to bind preferential ly to that of R propranolol (Albani et a l . . 1984). G i l l is et a l . , 1985, reported the presence of two binding sites on AAG for both racemic propranolol and propafenone. From non-linear Scatchard plots the authors pointed out that stereoselective binding characteristics for both drugs are expected on AAG. 2.12 Stereoselectivity in Drug Action An asymmetric molecule and its mirror image are characterized by many identical physical and chemical properties, but when the molecule finds i t s e l f in the chiral environment afforded by the body, physiological manifestations of differences between enantiomers are frequently observed. The enantioselective behavior is due to the association of the chiral drugs to the chiral structures in organisms resulting in diastereomeric properties of drug enantiomers, which contribute to differences in drug transport, binding, drug-receptor interaction and drug biotransformation. Numerous studies have been reported with isolated tissues and intact animals to verify the enantioselective nature of the interaction of drugs with ce l lu lar and tissue compartments (Ariens et a l . , 1983; 14 Smith, 1984; Patil et a l . , 1970). An example is R(-) norepinephrine, which is 300 times more potent than its antipode as an adrenergic receptor agonist on the rabbit aorta, and has higher binding a f f in i ty for the receptors, or preferred intr ins ic act iv i ty at the binding s i te. The differences between the pharmacological act iv i ty of enantiomers may be of pharmacokinetic as well as of pharmacodynamic orig in. However, the relat ive differences in absorption, distr ibution, metabolism, protein- or tissue-binding may confer stereoselectivity of drug enantiomer interactions in a chiral environment. 2.12.1 Absorption Due to similar l i p i d and aqueous so lub i l i t ie s , absorption of drug enantiomers in the body should be identical and follow f i rst -order kinetics - except for the manner in which the drug enantiomers show enantioselectivity and are absorbed by active transport processes. For example, in studies in rabbits by means of an in situ ligated loop and an in-v i tro tissue accumulation technique, Shindo et a]_., 1973, demonstrated that L-DOPA is absorbed from rat intestine by an active transport mechanism, which proceeds independently from the decarboxylation process. Such a carrier-mediated active process of absorption is saturable and leads to dose-dependent absorption of the substrate. Further examples of enantioselective absorption include the L-isomers of methotrexate and ascorbic acid, which are preferential ly absorbed in the body as compared to their antipodes (Wade et a l . , 1973; Handel et a l . . 1984). 2.12.2 Presystemic Elimination 15 Enzymes are asymmetric in nature and exhibit differences in a f f in i t i e s between drug enantiomers for reactive s i tes. The drug enantiomers of high extraction ratio exhibit differences in b ioavai labi l i ty for f irst-pass extraction by both l i ver and gut. It has been found that propranolol undergoes stereoselective presystemic clearance in dogs (Walle and Walle, 1979) and in man (Silber and Reiglemen, 1980; Von Bahr et a l . , 1982) and that intr ins ic clearance of R(+) propranolol exceeds that of the more pharmacologically active S(-) propranolol enantiomer. Similarly, a significant stereoselective hepatic extraction has been demonstrated for metoprolol. The bioavai labi l i ty of the S(-) enantiomer was found 1.4 times higher than that of R(+) enantiomer in extensive metabolizers (Dayer et a l . , 1985). Another example of stereoselective f irst-pass metabolism is verapamil; in which the systemic ava i lab i l i ty of the more active (-) enantiomer was 1/2 to 1/3 that of the (+) enantiomer (Eichelbaum et a l . , 1984). 2.12.3 Distribution Stereoselective distribution may arise due to carrier-mediated transport or selective uptake of drugs by various organs and by binding to tissue and plasma proteins. Selective tissue-binding would increase the elimination h a l f - l i f e of the more highly bound enantiomer and may contribute to differences >in enantiomer concentrations at relevant \\ receptors as well as in plasma. The liver/plasma concentration ratios of S and R phenprocoumon in rats were found to be s ignif icantly dif ferent; the values being 6.9 + 0.5 and 5.2 ± 0.2, respectively (Schmidt and Jahnchen, 1977), indicating a preferred uptake of the pharmacologically more potent enantiomer into the target organ. 16 Cardiac uptake of the beta-adrenergic blocker, propranolol, is enantioselective. Kawashima et a l . , 1976, in a study in rats, reported that the serum and heart concentrations of the individual enantiomers of propranolol were different. After 5 minutes, the serum levels of the pharmacologically active S isomer were three times lower than those of i ts antipode. During the following 4 hours, however, the R form disappeared faster from the serum ( t j / 2 = 24 min.) than the S form ( t ^ = 52 min.). The i n i t i a l lower serum levels of S isomer were interpreted as indicative of a highly selective cardiac uptake. However, most of the propranolol found in heart tissue was the S isomer (82 + 3% up to 90 min. and 100% after 2 hours). It appears, therefore, that the pharmacologically most active enantiomer is taken up by l i ver and rapidly metabolized. These facts are similar to the findings of George et a l . , 1972, in which the plasma hal f - l ives of S and R propranolol in man were 3.2 + 0.42 hours and 2.02 + 0.55 hours respectively. These results also closely parallel those observed in rats, suggesting a comparable and therapeutically important stereoselective uptake by cardiac tissue. Similarly, (-) timolol in the rat was taken up much more avidly by particulate fractions of heart, lung and brain than the (+) enantiomer, which was bound only to non-specific sites and eliminated faster (Tocco et a l . , 1976). Selective transport into rat brain has been documented for the active S enantiomer of a-methyldopa (Ames et a l . . 1977). Due to the chiral nature of human serum albumin (HSA) and o^-acid glycoprotein (AAG), stereoselective interactions of drug enantiomers are to be expected. The stereoselective binding interaction of L tryptophan with HSA has been well documented. L Tryptophan was found to bind to a 17 single s ite with an a f f in i ty of about 100 times than that for D tryptophan (McMenany et a l . . 1958). Studies of the stereoselective binding of drug enantiomers to plasma proteins have been reviewed by Alebic-Kolbah et a l . . 1979. In 1975, Muller and Wolhert found that the S(+) enantiomer of oxazepam hemisuccinate had 35-fold higher binding a f f in i ty to human serum albumin than its antipode. 2.12.4 Metabolism The two enantiomers of a chiral substrate react di f ferently with enzymes, which are themselves chiral and hence enantioselectivity can be expected. Stereoselective metabolism was observed after the sequential administration of R and S enantiomers of mephenytoin to dogs (Kupfer and Bircher, 1979). This study showed that aromatic hydroxylation to form 4-hydroxymephenytoin was highly selective for the S enantiomer. Conversely the R enantiomer, was metabolised to 5-phenyl 5-ethylhydantoin (PEH) via N-demethylation. The long elimination half-l i f e (> 80 hr.) for PEH, coupled with the short elimination h a l f - l i f e (4 hr.) for 4-hydroxy-mephenytoin led to the supposition that the major therapeutic effect from chronic administration of racemic mephenytoin is due to the R enantiomer. Enantioselective biotransformation was observed with the anticoagulant drug warfarin. In man, the biological ly more potent enantiomer, S warfarin, was metabolized and inactivated primarily by oxidation to 7-hydroxy warfarin (Lewis et a l . , 1974). Conversely, a major route of biotransformation for R warfarin was via a highly stereoselective reduction to R/S warfarin alcohol (Lewis et a l . , 1974). 18 The metabolic fate of the sedative hypnotic, glutethimide, presents an even more dramatic example of enantioselective pathways for biotransformation. In the dog, 4-hydroxy-glutethimide appeared to arise exclusively from the R enantiomer, while the 1-hydroxyethyl metabolite appeared to be formed from the S isomer. However, the same degree of stereoselectivity was not maintained in man (Kennedy and Fischer, 1979). Enantioselectivity has been shown in genetic polymorphic oxidation of several chiral drugs. The (-) enantiomer of the beta adrenergic antagonist, bufurolol, was more extensively ring hydroxylated in position 4 than its (+) antipode, while the reverse is true for 1'-hydroxylation. In l i ver microsomes from normal subjects (extensive metabolizers), the (+)/(-) ratio for the rate of 1'-hydroxylation favoured the (+) enantiomer by 2-fold. The metabolism of the anticonvulsant, mephenytoin, offers another example of a stereoselective hydroxylation defect. While in humans, S mephenytoin (t\\/2 = 1 hour) was rapidly p-hydroxylated, the R enantiomer (tj/2 = 70 hours) is slowly N-methylated to PEH, an active metabolite which tends to accumulate due to very slow elimination ( t ^ = 150-200 hour) (Wedlung et a l . , 1985). Propafenone is mainly metabolized by cytochrome P-450 isozyme(s) to y ie ld 5-hydroxy propafenone. This oxidative metabolic route was found to be genetically controlled by an autosomal recessive t ra i t (determined by debrisoquine administration) and a bimodal distribution in patients was noted (Siddoway et a l . , 1987). The poor metabolizers showed higher drug levels than extensive metabolizers. Beta-adrenergic drugs, including bufuralol and metoprolol, showed stereoselective biotransformation and also exhibited genetic polymorphism (Lennard et aj.., 1983). Being structurally related to beta-adrenergic drugs, 19 stereoselective biotransformation of propafenone could also be a logical poss ib i l i ty . The most intriguing aspect of the metabolism of chiral drugs is that of enantiomeric inversion of a number of active derivatives of 2-arylpropionic acids by such processes. It has been observed that one isomer of these compounds becomes configurationally more stable than the other, with a unidirectional configurational inversion (Kripalani et al_., 1976). A representative example is that of the administration of racemic ibuprofen or its enantiomers to several species, including man, in which an extensive inversion of the pharmacologically less potent R enantiomer was observed (Kaiser et a l . , 1976). Additional evidence has been reported for the R to S inversion of other related antiinflamatory drugs such as clidanac (Tamura et a l . , 1981). 2.12.5 Renal Clearance Renal clearance is not usually stereoselective unless a drug is actively secreted and/or reabsorbed. Stereo-selectivity in f i l t r a t i on at the glomerulous would depend mainly on stereoselective differences in the non protein-bound levels of one the enantiomers of the drug (Williams and Lee, 1985). 2.13 Stereoselective Drug Analysis Conventional methods for measuring optical act iv i ty of isomers include polarimetry, optical rotatory dispersion and circular dichroism. Chromatographic methods offer dist inct advantages for the analyses and separation of optical isomers over other methods due to the small sample size required, independence from the magnitude of specif ic rotation, and 20 most importantly, independence from other optical ly active species present (Koniq et a l . , 1977). Both GLC and HPLC have been widely used for the separation of diverse enantiomeric compounds. In general, the resolution of enantiomers by chromatographic means has been achieved by one of the two methods. The f i r s t involves the conversion of the racemate to a mixture of diastereomers using a suitable chemical reaction with a chiral reagent, followed by resolution of the diastereomers on a non-chiral stationary phase (Konig et a l . , 1977 and Parr et a l . , 1971). The second method involves direct separation of the enantiomers on a chiral stationary phase (Pirkle et a l . , 1979). Earl ier attempts for the separation of enantiomers as their diastereomers were reported by Gil-Av et al.., 1966, who resolved a series of racemic secondary alcohols as their (+) l act ic acid derivatives on a GLC capi l lary column coated with polypropylene glycol. Other investigators have used N-trif luoroacetyl-L-proloylchloride (N-TPC) as a chiral derivatizing agent for the GLC assay of amino acids (Westley et al_., 1968, Dabrowiak et al_., 1971), and primary and secondary alcohols. Koreeda et a l . , 1973, used an HPLC technique to resolve the enantiomers of abscisic acid derivatized with (+)-Q'-methoxy-a-trifluoromethylphenyl acetychloride (MTP). Helmchen and Strubent, 1974, showed that diastereomeric amides formed from the reaction of racemic amines with optical ly pure methylmandelylchloride were separable by HPLC. Hermansson et a l . , 1982, employed the anhydrides formed with tert-butoxycarbonyl-L-alanine and tert-butoxy-L-leucine with R,S alanine for the determination of the ratio of the R enantiomer in presence of the S enantiomer of several alkanolamines with beta-adrenergic act iv i ty. 21 With S(-)-l-phenylethylisocyanate as a chiral derivatizing reagent, Thompson and his associates, 1982, were able to separate the enantiomers of propranolol as their urea derivatives on a reverse phase HPLC system. An extension of this method was achieved by Sedman and his associate, 1983, to separate the enantiomers of eleven beta-blocking agents derivatized with 2,3,4-tri-0-acetyl-a-D-arabinopyranosylisothiocyanate (AITC) or 2,3,4,6-tetra-0-acetyl-/?-D-glucopyranosylisothiocyanate (GITC). Recently, Bel anger et a l . , 1985, resolved the enantiomers of mexiletine as their thiourea derivatives with GITC. Despite the fact that diastereomeric separation of enantiomers has been widely used as a technique, there are basic requirements that must be considered. The optical ly active reagent (chiral selector, SE) must be available in the purified form as this has a direct influence on accuracy of the maximum detectable optical purity of the chiral solutes (selectands, SA). The reaction must be mild so that v i r tua l ly no racemization of the chiral centres of the SE and SA occur. Final ly, the derivatization reaction must be quantitative for each enantiomer because the constants (Kj and K2) of the reactions, as shown below, may not be equal. K l (R)-SE + (R)-SA > (R)-SE-(R)-SA K2 (R)-SE + (S)-SA - - > (RJ-SE-(S)-SA Perhaps the most important requirement is the steric conformation of the chiral centre(s) and the distance(s) to the reactive functional group(s) from chiral centers in the SE and SA molecules, since 22 favourable and unfavourable spatial arrangements of the chiral substituents with respect to each other in the resulting diastereomeric (R)-SE-(R)-SA, and (R)-SE-(S)-SA derivatives, ref lect their different 1ipophi l ic i ty or polarity. These arrangements have a direct influence on adequate resolution by diverse chromatographic systems. F inal ly, for sensitive and selective analytical purposes, SE reagents should have a chromophore or fluorophore to enhance the detect ib i l i ty of the derivatives (Linder et a l . 1984). The direct separation of enantiomers, although a valuable technique, is strongly dependant on structural elements of the SE and SA molecules. It involves the formation of transient diastereomeric complexes between the chiral solute and the chiral selector/sorbent stationary phase which does not involve covalent bonds. The relative s tab i l i ty of these diastereomeric complexes results in different rates of elution of the enantiomers. The direct resolution of racemic amino acids was reported by Kotake et al.., 1951. Dalgliesh, 1952, subsequently studied the structural features necessary for resolution to occur with aromatic amino acids using cellulose paper chromatography. He postulated that for direct separation, at least three points of attachment were required. These included the formation of hydrogen bonds, a dipole-dipole interaction, and a 7r-7r stereochemical interaction and/or dispersion forces between the solute and the sorbent. Gil-Av et a l . , 1966, f i r s t introduced the use of an N-trif luoroacetyl-L-isoleucine-lauryl ester chiral stationary phase for the separation of N-trifluoracetyl-amino acid esters on wall-coated capi l lary columns. Recognizing the essential role of the -NH-C0CH-(R)-NH-C0- group (where, R is the chiral center) in these types of peptide ester phases, the 23 search for structural features of amides which might increase their se lect iv i ty and thermal s tab i l i ty , led to the development of the N-lauroyl-L-valyl-tert-butylamide phase and, subsequently, the diamide phases which exhibited greater eff ic iency, higher resolution factors and reduced retention times (Feibush et a l . . 1971,). Frank and his co-workers, 1978, later coupled L-valine-tert-butylamide to the carboxyl groups of the co-polymer of dimethylsiloxane and carboxyalkylmethylsiloxane to produce a highly versati le GLC chiral stationary phase (Chirasil-Val*) which is now commercially available. It has been used to separate enantiomeric drugs and metabolites of amino acids of high and low vo la t i l i t y and some amino alcohols (Frank et a l . , 1978) and primary amines (McErlane et a l . , 1983). In addition, the high thermal s tab i l i ty of Chirasi l -Val^ made i t possible, for the f i r s t time, to employ a mass-spectrometer coupled to a GLC system for the analysis of enantiomers (Frank et a l . . 1978, Weiss et a l . , 1978). Another class of potentially active stationary phases that have shown great promise for enantiomer resolution, are the carbonyl bis-(amino acid)-esters. Enantiomeric amine derivatives could be separated using these phases in l iqu id , solid and mesophase, depending on the operating temperature (Liu et a l . , 1982, Feibush et a l . , 1971, Corbin et a l . , 1970, Lochmuller et a l . , 1974). Chiral resolution of enantiomers by HPLC has included the use of chiral columns such as the Pirkle^ chiral column which is an N-(3,5-dinitrobenzoyl)-D-phenylglycine phase bonded to s i l i c a . This stationary phase is capable of separating the enantiomers of primary amines, amino alcohols, and amino acids (McErlane et a l . . 1987, Pirkle et a l . , 1981). Cyclodextrin bonded stationary phases and protein bonded phases 24 (AAG or bovine serum albumin attached to s i l i c a ) , which are also commercially available, have been used for the HPLC separation of variety of compounds including amides, amino acids and amino alcohols (Armstrong, 1984). The three point interaction for chiral recognition in cyclodextrin phases is achieved through a t i gh t - f i t inclusion complex formation between the solute and stationary phase. The use of ion-pair reagents, such as quaternary ammonium compounds or alkyl sulphonic acids, has been proposed for chiral HPLC applications. The method is based on ion-pair chromatography with a chiral counter ion added to the mobile phase. Petterson and Sch i l l , 1981, demonstrated the usefulness of (+)-10-camphorsulfonic acid as a chiral ion-pair reagent for the direct resolution of amino alcohols of the beta-blocker series. The authors were also able to resolve the ephedrine analogues by ion-pair formation of chiral amines by using a non-chiral l ipophi l i c counter-ion. These ion-pairs could be s tereose lect ive^ extracted and/or partitioned into a chiral stationary phase created by adsorption of a l ipophi l i c chiral mobile phase additive, (R,R-di-N-butyltartarate), on a reverse phase system. Another technique, called ligand exchange chromatography was developed by Davenkov et a l . , 1980, to separate the enantiomers of amino and/or carboxyl compounds. In this method, an optical ly active ligand is bonded covalently to an insoluble carr ier. After charging this adsorbant with a metal ion, the racemate to be resolved forms diastereomeric complexes with the metal-ion and the adsorbent, which lead to resolution of the enantiomers. Other methods reported for the resolution of enantiomers include the use of antiserum with the ab i l i ty to discern the stereochemical 25 differences in drug molecules. Several authors have used this procedure to develop sensitive and stereoselective radioimmunoassay procedures. Findley et a l . , 1981, developed a stereospecific radioimmunoassay for (+) pseudoephedine in humans. Using an identical approach, Midha et aj.., 1983, studied (+) and (-) ephedrine in human plasma after an oral dose of the racemate. 2.14 Analytical Methods for Propafenone 2.14.1 Analytical Method for R,S Propafenone Several HPLC and GLC methods have been reported for the analysis of racemic propafenone in biological f lu ids. The method described by Keller et a l . , 1978, was not optimal since sensit iv i ty was low and large volumes of plasma (4 to 8 mL) were required. The HPLC methods described by Harapat and Kates, 1982, and Brode et a l . , 1982, f u l f i l l e d sensit iv i ty and reproducibil ity requirements for c l in i ca l monitoring. Both methods involved extraction from alkalinized plasma and back extraction from the organic phase to an acid phase. A highly sensitive HPLC method with fluorometric detection has been reported by Brode et a l . , 1984. This method was used to monitor the plasma concentrations of racemic propafenone and its 5-hydroxy metabolite down to 0.2 ng/ml, but large differences in the coefficient of variation at low drug levels were reported. Moreover, the procedure was time-consuming and there was d i f f i cu l t y in removing the excess fluorogenic reagent. GLC methods with electron capture detection are very sensitive. Marchensini and his associates, 1982, developed a GLC method which could monitor propafenone down to 10 ng/mL in plasma. Chan et a l . , 1986, 26 developed a similar GLC method, using HFBA as the derivatizing reagent. Appl icabi l i ty of this method has been reported by Axelson et a l . , 1978, to measure the bioavai labi l i ty of propafenone while the drug was taken with food. This method was also aimed at monitoring the plasma protein-binding of total and unbound fraction of the racemic drug in plasma. 2.14.2 Analytical Method for Propafenone Enantiomers Thus far, one analytical study has been published for the separation of the R and S isomers of propafenone by fractional crysta l l izat ion of the diastereomeric tartar ic acid derivatives (Blaschke et a l . , 1987). The same authors also used (+) phenyl ethylisocyanate reagent to determine the purity of the enantiomers by HPLC. This procedure was used to obtain the pure individual enantiomers for the study of their absolute configuration, but not for their measurement in biological samples. At the present time a stereoselective assay for the enantiomers of propafenone is lacking. Thus, for the elucidation of the pharmacokinetic parameters of propafenone enantiomers, a sensitive and stereoselective assay is required. 27 2.15 Objective of the Study To develop a sensitive and stereoselective chromatographic assay method for propafenone enantiomers in human biological f lu ids. 28 3. EXPERIMENTAL 3.1 Supplies 3.1.1 Drug and Internal Standards R,S-2'-[3-(propyl ami no)-2-(hydroxy)-propoxy]-3-phenylpropiophenone hydrochloride (Propafenone hydrochloride), and R,S-2'-[3-(ethylamino)-2(hydroxy)-propoxy]-3-phenylpropiophenone hydrochloride (Li-1115 hydrochloride) - Knoll Pharmaceuticals, Markham, Ont., Canada. a-Bromonaphthalene - ICN Pharmaceuticals, Inc., Plainview, N.Y. USA. Desipramine hydrochloride - Sigma Chemical Co. ST.Louis, Mo., USA. (+)Naproxen - Syntex Inc. Mississauga, Ont. Canada. 3.1.2 Chemicals and Reagents R(+)-or-Methyl benzylisocyanate, R-(-)-l-(naphthyl)ethyl isocyanate, dansylhydrazine - Aldrich Chemical Co. Milwaukee, Wis. USA. Acetobromo-a-D-glucose - Sigma Chemical Co. Si lver thiocyanate - Eastman Kodak Co. Rochester, N.Y. USA. 2.3,4,6 Tetra-O-acetyl-/J-D-glucopyranosylisothiocyanate (GITC) Synthesized in our laboratory. Phenytoin, dicyclohexylcarbodiimide (DCCI) - Sigma Chemical Co. Urea - Mallinckrodt Inc. Paris, France. Heptafluorobutyric anhydride (HFBA), tr i f luoroacet ic anhydride (TFAA) -Pierce Chemical Company, Rockford, II., USA. Piperidine, pyridine, trichloroacetic acid (TCA) - BDH, Poole, England. Triethylamine, N-trif luoroacetyl-L-proloylchloride - Aldrich Chemical Co. Sodium hydroxide - American Sc ient i f ic and Chemicals, Portland, Or. USA. 29 Mono- and disodium phosphate - Mallinckrodt Chemical Works, St. Louis, Mo. USA. Hydrochloric acid, perchloric acid, sodium sulphate (granular), dimethylformamide - BDH Chemicals, Vancouver, B.C. Canada. 3.1.3 Solvents Acetonitr i le, benzene, chloroform, dichloromethane, n-hexane, methanol, 2-propanol, toluene, water, acetone (HPLC grade) - Fisher Sc ient i f ic , Vancouver, B.C. Canada. 3.2 Chromotographic Stationary Phases 5u Ultrasphere ODS column (25 X 0.46 cm i.d.) - Beckman Instruments Inc., Berkeley, Ca. USA. HPLC disposable guard column (1.5 X 0.32 cm i.d.) with C^g cartridge and reusable holder - Rainin Instrument Co. Inc., Woburn, USA. 3% Si lar 10 C GLC, and 3% OV-17 packed columns (2 m X 2 mm i.d.) coated on Cromosorb W - HP, 80/100 mesh - prepared in our laboratory. SE-30 fused-s i l ica capi l lary column (50 m X 0.25 mm i .d . - Hewlett-Packard, Avondale, Pa., USA. Ch i ra s i l -Va l R fused s i l i c a capi l lary column (50 m X 0.25 mm i.d.) -Applied Science. Rockwood, Ont.,Canada. n Pirkle ionic (type 1-A) hi-chrom reversible HPLC column (25 cm X 4.6 mm i.d.) containing 3,5 dinitrobenzoyl-(D)-phenylglycine bonded to-7-amino silanized s i l i c a - Regis Chemical Co., Morton Grove, II., USA. Cyclobond (I) -/J-cyclocdextrin bonded HPLC column (25 X 0.46 cm i.d.) -Advanced Separation Technologies, Inc. Whippany, N.J. USA. 30 3.3 Equipment A Hewlett Packard model 1082B high-performance l iquid chromatograph equipped with a Hewlett Packard model 79850B data terminal - Hewlett Packard, Palo Alto, Ca. USA. Holochrome variable uv detector - Gil son Medical Electronics, Middletown, Wi, USA. Schoeffel model 970 LC fluorometer - Kratos, Westwood, N.J. USA. A Hewlett Packard model 5830 gas chromatograph, equipped with flame ionization, and electron Capture (6^Ni) detectors - Hewlett Packard. A Hewlett Packard 5700A gas chromatograph; interfaced to a Varian Mat-I l l Mass spectrometer. A Varian model 620/L computer to record electron impact spectra on a chart recorder. A Unicam SP-1000 IR spectrometer, Pye Unicam Ltd. Cambridge, England, UK. A Beckman model 24 uv spectrophotometer, Beckman Sc ient i f ic Instruments Division, Irvine, Ca., USA. A Perkin-Elmer model 141 Polarimeter, Perkin-Elmer and Co. GMBH., Uberlingen, West Germany. Miscellaneous Vortex mixer - Sybron Co., Dubuque, Io, USA. pH meter - Orion Research Inc., Ma., USA. Thomas Hoover Capillary Melting point apparatus - Arthur H. Thomas Company, Philadelphia, Pa., USA. Direct connect universal column pref i l ter - Alltech Associates Inc., Deerfield, II., USA. Mil l ipore f i l t e r s - Mil l ipore Corp., Bedford, Ma. USA. 31 3.4 Stock Solutions 3.4.1 R.S Propafenone Hydrochloride R,S propafenone hydrochloride (equivalent to 0.1 mg/mL of the free base) was prepared in deionized d i s t i l l ed water. The solution was further diluted with deionized d i s t i l l ed water to the desired concentrations. The stock and di lute solutions were kept at 4°C for up to 3 months. 3.4.2 R,S Propafenone Free Base R,S propafenone hydrochloride in water was basified with sodium hydroxide at pH > 12 and extracted into benzene. The organic extract was evaporated to dryness under a gentle stream of clean nitrogen in a 40°C water bath. The dried crystals of propafenone base were accurately weighed and dissolved in known volumes of organic solvents. 3.4.3 R.S-2 7-f3-(ethyl ami no)-2-(hydroxy)-propoxvl-3 phenvlpropiophenone Hydrochloride (Li-1115) and Desipramine Hydrochloride (Internal Standards) Stock solutions equivalent to 0.1 mg/mL of the free base were prepared by dissolving these compounds in deionized d i s t i l l ed water and were kept at 4°C until used. 3.4.4 (0.2 M) Phosphate Buffer Solution of pH 2.8 Monosodium phosphate (27.8 g) and disodium phosphate (28.4 g) were dissolved separately in 1 L of HPLC grade water to give solutions A and B respectively. An aliquot of 93.5 mL of solution A was added to 6.5 mL 32 of solution B in a 100 mL volumetric flask and perchloric acid was added dropwise to the mixture to adjust the pH to 2.8. The buffer was kept at 4°C until used. 3.4.5 Trichloroacetic Acid Solution (10% w/v) Trichloroacetic acid (10 g) was dissolved in deionized d i s t i l l ed water to a f inal volume of 100 mL. 3.5 Synthesis of 2,3,4,6 Tetra-0-Acetvl-fl-D-Glucopyranosylisothiocyanate (GITC) 3.5.1 Purif ication of Acetobromo-a-D-qlucose The commercial product, acetobromo-a-D-glucose, was purif ied by rapidly cooling a hot saturated solution in absolute alcohol. The crysta l l ine material which separated was f i l tered under suction and then was thoroughly washed with l ight petroleum (30°- 60°) to remove traces of alcohol (Ber, 49, 584, 1916). The purif ied material was stored at -20°C until required. 3.5.2 Synthesis of GITC Purified acetobromo-a-D-glucose (1.025 g) was heated at 103 + 2°C in dry, freshly d i s t i l l ed xylene (5 mL) with dried s i lver thiocyanate (1.25 g) under ef f ic ient s t i r r ing. After one hour, additional s i lver thiocyanate (0.85 g) was added in two equal portions at intervals of 30 minutes. The reaction mixture was then f i l tered while hot through a f r i t ted glass funnel by suction. The precipitate of s i lver bromide was washed twice with 5 mL portions of hot xylene. About 20 mL of cold 33 petroleum ether (30°- 60°) was added to the f i l t r a te and the mixture was kept at 4°C for 16 hours. The resulting crystals were recovered and recrystal l ized three times with mixture of xylene-petroleum ether (1:4) and dried over phosphorous pentoxide. The purity of the product was determined by reverse phase HPLC with 65:35 methanol/water as mobile phase. M.P. 110-111°C ( l i t . m.p. 112-114°C, Nimura et al..1982) IR (stretching frequency): N = C = S, 2100 cm\" 1, CH30 - C = 0, 1750 cm\" 1. (Prepared by a modification of the method of Nimura et a l . , 1980) 3.6 Preparation of the Derivatives of R,S Propafenone 3.6.1- Heptafluorobutvric Anhydride (HFBA) Derivative To 0.5 mg of R,S propafenone free base in dry benzene (1 mL) were added 25 uL of HFBA and 1 uL of TEA. The reaction mixture was agitated for 1 minute and then heated at 65°C for 1 hour. After cooling, the reaction mixture was evaporated to dryness under a gentle stream of clean nitrogen at 40°C in a water bath. The residue was reconstituted with 1 mL of dry benzene 3.6.2 N-Trifluoroacetvl (L) Proloylchloride TN-TPC1 Derivative To 1 mg of R,S propafenone free base in dry dichloromethane (1 mL) were added, 10 uL of N-TPC and 5 uL of dry pyridine. The reaction mixture was st irred by vortex for 1 minute and then heated at 45°C for 34 20 minutes. After cooling, the reaction mixture was evaporated to dryness under a stream of clean nitrogen in a 40°C water bath. The residue was reconstituted with 1 mL of dry dichloromethane. 3.6.3 l-Naphthoylchloride Derivative To 0.5 mg of R,S propafenone free base in 1 mL of dry dichloromethane were added, 0.5 mg of /J-naphthoyl chloride and 1 uL of dry TEA. The reaction mixture was heated at 45°C for 20 minutes and then evaporated to dryness under a stream of clean nitrogen at 40°C in a water bath. The residue was reconstituted with 1 mL of methanol. 3.6.4 R(+)-tt-Methylbenzvlisocyanate fR(+)MBICl Derivative To 1 mg of R,S propafenone free base in 1 mL of dry acetonitrilerdimethylformamide (8:2) were added, 5 uL of R(+)MBIC and 1 uL of TEA. The reaction mixture was held at room temperature for 30 minutes and then evaporated to dryness under a stream of clean nitrogen in a 40°C water bath. The residue was reconstituted with 1 mL of chloroform for GLC-FID experiments. For HPLC experiments the derivative was reconstituted with 1 mL of methanol. 3.6.5 Isopropylisocyanate Derivative To 0.5 mg of propafenone free base in 0.5 mL of dichloromethane were added, a 50 molar excess of isopropyl isocyanate and 1 uL of dry TEA. The reaction mixture was stirred by vortex for 1 minute and was heated at 45°C for 20 minutes in a dry heating block. The contents of the tube were then evaporated to dryness under a gentle stream of 35 nitrogen at 40^C in a water bath. The residue was reconstituted with 1 mL of dichloromethane. 3.6.6 2.3.4,6-Tetra-Q-acetvl-fl-D-Glucopyranosvl isothiocyanate (GITC) Derivative To 0.1 mg of R,S propafenone free base was added, a 50 molar excess of 0.5% w/v GITC reagent in dry acetonitrile:dimethylformamide (8:2). A solution of 5 uL of 1% w/v diphenhydantoin (phenytoin) in chloroform, or 1 uL of TEA were added as catalysts. The reaction mixture was st irred by vortex for 1 minute and was held for 30 minutes at room temperature before evaporating to dryness under a stream of clean nitrogen. The residue was reconstituted with 1 mL of methanol. 3.6.7 R(-)-l-(Naphthvl)ethv1isocyanate rR(-)NEICl Derivative To 2 ug of propafenone free base in a 10 mL screw-capped culture tube were added, 100 uL of 0.01% v/v R(-)NEIC in acetonitr i le and 400 uL of dry acetonitr i le containing 0.003 (M) TEA. The reaction mixture was heated at 75°C for 30 minutes in a dry heating block. The tube was cooled, and 1 uL of piperidine was added. The tube was reheated at 75°C for 30 minutes to destroy the excess reagent. After cooling, the mixture was evaporated to dryness and reconstituted in 150 uL of 2-propanol. 3.6.8 Propafenone-Dansy1hvdrazine-R(-)NEIC Derivative To propafenone base (20 - 1000 ng) was added, 100 uL of 0.2% w/v dansylhydrazine in the presence of 0.48% w/v TCA in dry toluene. The reaction mixture was diluted with 400 uL of dry benzene and st irred by 36 vortex for one minute in a screw-capped culture tube. The tube was heated at 75°C for 20 minutes and the excess reagent was destroyed by adding 100 uL of acetone and heating for another 20 minutes at 75°C. After cooling, the contents of the tube were evaporated to dryness under a gentle stream of clean nitrogen in a 40°C water bath. The residue was basified with 2 mL of 2 M sodium hydroxide. The free base was extracted with 6 mL of benzene which was subsequently evaporated to dryness under a stream of clean nitrogen. The residue was reacted with 150 uL of 0.01% v/v R(-)NEIC in dry acetonitr i le at room temperature for 30 minutes. The excess R(-)NEIC reagents were destroyed by adding 1 uL of piperidine and storing the samples at room temperature for 20 minutes. 3.6.9 Derivatization of Propafenone with (+)Naproxen An aliquot of 10 ug of propafenone base and a 50 molar excess of (+) naproxen in 1 mL of dry dichloromethane were allowed to react at room temperature for one hour in the presence of excess of dicyclohexylcarbodimide (DCCI). The reaction mixture was f i l tered and sequentially washed with 5% HC1, 5% NaOH, and water, and reconstituted with 1 mL of 2-propanol. An aliquot of the reaction mixture was injected onto the HPLC column, employing methanol/water (75:25) as mobile phase delivered at a flow rate of 1 mL/min. 37 3.7 Preliminary GLC and HPLC Analysis of Propafenone using Chiral and Achiral Stationary Phases 3.7.1 Chiral GLC Phase Underivatized propafenone did not elute from the chirasi l-. Va l R column. The following derivatization procedure were used to increase the vo l a t i l i t y and / or se lect iv i ty of the chiral phase for direct resolution of various derivatives of propafenone. a) An aliquot of 1 uL of the freshly prepared HFBA derivative of propafenone (0.5 mg/mL) in benzene was injected onto the Ch i ra s i l -Va l R capi l lary column in the GC system coupled with electron capture detector. Chromatographic conditions - column temperature: 220°C, injection temperature: 250°C, carrier gas (He) flow rate: 1.2 mL/min, make-up gas (argon/methane, 95:5) flow rate: 30 mL/min, detector temperature: 350°C, injection mode: sp l i t (1:30). b) An aliquot of 2 uL of the freshly prepared isopropyl isocyanate derivative of propafenone (0.5 mg/mL) in dichloromethane was injected onto the ch i r a s i l - Va l R column in the GC system. The detector used was flame ionization. . Chromatographic conditions - column temperature: 200°C, injection temperature: 250°C, detector temperature: 275°C, carr ier gas (He) flow rate: 1.3 mL/minute, make-up gas (He) flow rate: 20 mL/minute, injection mode: sp l i t (1:30). 3.7.2 Achiral GLC Stationary Phase a) An aliquot of 1 uL of the freshly prepared R(+)MBIC derivative of 38 propafenone (1 mg/mL) in dry chloroform was injected on to SE-30 capi l lary column in the GC system. Chromatographic conditions - column temperature: 190°C, injection temperature: 240°C, detector (FID) temperature: 275°C, carrier gas (He) flow rate: 1.2 mL/minute, make-up gas (N2) flow rate: 30 mL/minute, injection mode: sp l i t (1:20). b) An aliquot of 4 uL of freshly prepared R(+)MBIC derivative of propafenone (1.0 mg/mL) in dry chloroform was injected onto an 0V-17 packed column in the GC. Chromatographic conditions - injection temperature: 300°C, column temperature: 280°C, FID temperature: 275°C, carr ier gas (He) flow rate: 40 mL/minute. c) An aliquot of 1.5 uL of freshly prepared N-TPC derivative of propafenone (1 mg/mL) in dichloromethane was injected onto the SE-30 capi l lary column in the GC. Chromatographic conditions - column temperature, 180°C, injection temperature: 225°C, detector temperature: 275°C, carrier gas (He) flow rate: 1.2 mL/min, make-up gas flow rate: 25 mL/min, injection mode: sp l i t (1:30). 3.7.3 Chiral HPLC Phases a) Propafenone free base (100 ug) was dissolved in 500 uL of methanol and 20 uL of this solution was injected onto a /J-cyclodextrin HPLC column (25 cm X 0.45 cm i .d. ) . Mobile phase: Methanol/water (60:40) delivered at a flow rate of 0.8 mL/min. Detection: 254 nm. 39 b) An aliquot of 10 uL of the freshly prepared 2-naphthoyl chloride derivative of propafenone (0.5 mg/mL) was injected onto a (25 X 0.4 cm i.d.) Pirkle ionic (type- 1A) HPLC column. Mobile phase: 9% 2-propanol in hexane delivered at a flow rate of 1 mL/min. Detection: 254 nm. 3.7.4 Achiral HPLC Phases a) An aliquot of 20 uL of the R(+)MBIC derivative of propafenone (1 mg/mL) in methanol was injected onto a (25 X 0.45 cm i.d.) 5u ultrasphere, ODS HPLC column. Mobile phase:) Methanol/water(70:30), delivered at a flow rate of 1 mL/min. Detection: 254 nm. b) A 20 uL aliquot of the freshly prepared GITC derivative of propafenone (0.1 mg/mL) was injected onto a (25 X 0.45 cm i.d.) 5u ODS column. Mobile phase Methanol/ phosphate buffer (0.2 M) of pH 2.8 (75:25), delivered at a flow rate of 1 mL/min. Detection: 250 nm. 3.8 HPLC Resolution of the Enantiomers of Propafenone 3.8.1 Resolution of the Diastereomeric R(+)MBI Derivatives Duplicate samples, each containing 20 ug of R,S propafenone hydrochloride in 1 mL of water, were placed in 10 mL PTFE-lined screw-capped culture tubes, and 0.5 mL of 3 M NaOH and 6 mL of benzene were added. The tubes were t ightly capped and tumbled for 20 minutes on a rotating tumbler. The tubes were then centrifuged for 10 minutes at 2500 r.p.m. and the upper benzene layers were transferred to clean dry 40 test tubes. The solvent was evaporated to dryness following the procedure described before. The residues were reconstituted in 100 uL of dry dichloromethane and 5 uL of R( + )-a-methylbenzyl isocyanate was added to each tube. The tubes were st irred by vortex for 1 minute and kept at room temperature for 30 minutes. After reaction, 5 uL of HPLC grade water was added to the reaction mixtures to destroy the excess reagent. After a further 20 minutes, 20 uL of the derivative was injected onto the HPLC-0DS column (5u. 25 X 0.45 cm i .d . ) . The mobile phase, methanol/water (71:29) was delivered at a flow rate of 0.8 mL/min. Detection: 254 nm. 3.8.2 Resolution of Diastereomeric GITC Derivatives Duplicate samples, each containing 10 ug of R,S propafenone hydrochloride (1 mL of stock solution) were pipeted into 10 mL PTFE-lined screw-capped culture tubes and 0.5 mL of 3 M NaOH was added to each tube. Extraction was carried out with 6 mL of benzene following the procedure described above. To the residue was added, 95 uL of 0.5% w/v GITC in dry acetonitr i le. The mixture was st irred by vortex and held at room temperature for 30 minutes. To each tube was added, 5 uL of HPLC grade water and the tubes were held at room temperature for a further 20 minutes. A 20 uL aliquot was injected onto the HPLC ODS column (5u, 25 X 0.45 cm i .d. ) . The mobile phase was methanol/phosphate buffer (0.2 M) of pH 2.8 (74:26), delivered at a flow rate of 1 mL/min. Detection: 250 nm. 41 3.8.3 Resolution of the Diastereomeric R(-)NEIC Derivatives Duplicate samples, each containing 1 ug of R,S propafenone hydrochloride (in 1 mL of d i s t i l l ed and deionized water) were placed in 10 mL PTFE-lined screw-capped culture tubes, and 0.5 mL of 3 M NaOH was added. The free base was extracted with 6 mL of benzene as described above. The dried residue was reconstituted with 150 uL of dry acetonitr i le and then reacted with 5 uL of 0.1% v/v R(-)NEIC in dry acetonitr i le. The reaction mixture was st irred by vortex for one minute and then heated for 30 minutes at 75°C. The excess reagent was destroyed with 1 uL of piperidine by allowing the samples to s i t for a further 20 minutes. The reaction mixture was evaporated to dryness and then reconstituted with 150 mL of 2-propanol. An aliquot of this was injected onto the HPLC ODS column (5u, 25 X 0.45 cm i .d . ) . The mobile phase was methanol/water (76:24) delivered at flow rate 1 mL/min. Detection: 230 nm. 3.8.4 Resolution of diastereomeric Propafenone-Dansylhydrazine -R(-)NEIC Derivative A sample of 1 ug of propafenone base (extracted as described in sections 3.8.1 - 3.8.3) was reconstituted with 0.9 mL of dry benzene. To this, 0.1 mL of 0.2% w/v dansylhydrazine in the presence of 0.48% w/v tr ichloroacetic acid (TCA) in toluene was added (Brode et a l . , 1984). The reaction mixture was st irred by vortex for 1 minute and then heated for 20 minutes at 65°C in a dry heating block. After cooling, excess acetone was added and the sample was heated for a further 20 min. at 65°C to destroy the excess dansylhydrazine. The contents of the tube were cooled and 42 evaporated to dryness under clean nitrogen at 40°C in a water bath. The residue was basified with 0.5 mL of 3 M NaOH and extracted with 6 mL of benzene. After evaporating the organic solvent, 100 uL of 0.1% v/v R(-) NEIC in acetonitr i le was added to the residue and the samples were held for 30 min. at room temperature. An additional 0.9 mL of acetonitr i le was added to bring the volume to 1 mL. The tube was st i rred by vortex and 10 to 20 uL was injected onto the HPLC ODS column (5u, 25 X 0.45 cm i .d. ) , Mobile phase, methanol water (74:26) delivered at 1 mL/min. Detection: Fluorescence, Ex: 220 mm. Em: 418 nm (cut-off f i l t e r ) . 3.9 Quantitative Analysis of Propafenone Enantiomers 3.9.1 Selection of Internal Standard 2' -[2-(Hydroxy)-3-(ethyl amino)-propoxy]-3-phenylpropri ophenone, Li-1115 was i n i t i a l l y chosen as an internal standard based on its structural s imi lar ity to that of propafenone. Desipramine, an achiral secondary amine was also chosen as a second internal standard required for complete derivatization. 3.9.2 Selection of External Standard a-Bromonaphthalene was chosen as an external standard to measure the recovery of propafenone from plasma as well as to assess the stoichiometric ratio of reagent required for complete derivatization of propafenone. 3.9.3 Selection of Catalyst Phenytoin and Urea were i n i t i a l l y chosen as Afunct ional catalysts 43 for the propafenone-GITC reaction. Triethyl amine was chosen as a catalyst for the propafenone-R(-) NEIC and Propafenone-HFBA reactions. 3.9.4 Extraction Solvents Four organic solvents: benzene, dichloromethane, a mixture of benzene, dichloromethane and 2-propanol (7:3:1), and chloroform, were evaluated to assess their eff iciency for extraction of propafenone base and the internal standards from plasma. The phase volume ratio of ^org/^aqu w a s ^* 3.9.5 Plasma Protein Precipitation Propafenone is highly bound to plasma proteins and i t was therefore necessary to precipitate the proteins before extraction. For this purpose, 0.4 mL of 10% trichloroacetic acid in water was added to 1 mL of plasma containing propafenone hydrochloride (1 to 2 ug/mL). The plasma was shaken for one minute and then the pH was adjusted above 12 with 0.5 mL of 3 M NaOH, followed by extraction, derivatization and HPLC analysis. 3.9.6 Optimization of Reaction Conditions for Derivatization a) Optimum Reaction Time Optimum reaction times for propafenone and the internal standard was studied with four, t r ip l i ca te samples, each containing 3 ug of R,S propafenone hydrochloride and 1 ug of the internal standard in 1 mL of 44 d i s t i l l ed deionized water. Extraction was carried out as described in section 3.8. Derivatization was carried out separately with R(+)MBIC, GITC and R(-)NEIC at room temperature, at different time intervals of 15, 30, 45 and 60 minutes. A 5 uL portion of 0.002% v/v of a-bromonaphthalene, as external standard was added to each sample and stirred by vortex. The samples were analyzed by a plot of peak height ratio against time. b) Optimum Reaction Temperature for Derivatization of Propafenone with RMNEIC Five sets of tubes, each set in t r ip l i ca te , containing 2 ug of R,S propafenone hydrochloride and 1 ug of the internal standard were prepared in 1 mL of d i s t i l l ed deionized water. Extraction with 6 mL of benzene was carried out as described before. To each tube, 100 uL of 0.01% v/v R(-)NEIC in dry acetonitri le was added. A volume of 50 uL of 0.003 M TEA and 350 uL of dry acetonitri le was added to each tube and the tubes were st irred by vortex. Each set of tubes were held for 30 minutes at room temperature, 45°C, 60°C, 75°C and 90°C. The tubes were cooled, evaporated to dryness, and reconstituted with 100 uL of 2-propanol. An aliquot of 20 uL from each tube was injected onto the HPLC ODS column. The samples were analyzed by plotting peak-height ratio against temperature using a 5 uL portion of 0.002% v/v of a-bromonaphthalene as external standard added to each sample before injection onto the column. 3.9.7 Stoichiometric Ratio of R(-)NEIC Reagent & Drug Five aliquots of t r ip l i ca te samples, each containing 45 2 ug of propafenone base, were subjected to derivatization with a 1.8, 4.5, 9, 45 or 90 molar excess of R(-)NEIC at 75°C for 30 minutes. The samples were evaporated to dryness under nitrogen at 40°C in a water bath and reconstituted with 150 uL of 2-propanol. 5 ul of 0.002% v/v of a-bromonaphthalene was used as the external standard. An aliquot of 20 uL from each tube was injected onto the HPLC ODS column for analysis. 3.9.8 Optimizing the Sensitivity for Detection The uv sensit iv i ty of propafenone was determined in solvent systems of (a) acetonitrile/phosphate buffer (0.2 M) of pH 2.8 (65:35), (b) acetonitr i le, (c) methanol/phosphate buffer (0.2 M) of pH 2.8 (75:25),and (d) methanol/water (75:25),. The uv sens i t iv i ty was also determined for the propafenone GITC and propafenone R(-)NEIC derivatives. Al l uv scans of propafenone and its derivatives were done on a Beckman model 25, uv-spectrophotometer over the range, 180 nm to 350 nm. Respective solvent blanks were used as the reference in each case. 3.9.9 Polarimetric Measurement of R(-)NEIC Derivatives of R,S Propafenone Racemic propafenone hydrochloride (1 mg) was converted to its free base following the procedure described before. The free base was reacted with a 5 molar excess of R(-)NEIC in dry acetonitr i le at 75°C for 30 minutes to y ie ld diastereomeric urea derivatives of the enantiomers. A volume of 20 uL of the diastereomeric mixture was repeatedly injected on to the HPLC ODS column (5u , 25 X 0.45 cm i.d.) using methanol/water (75:25) as the mobile phase delivered at 0.9 46 mL/min. The two diastereomeric fractions from the HPLC eluant, which corresponded to two resolved peaks of the enantiomers of propafenone were separately collected into two clean flasks. The purity of each fraction was ascertained by injection of 200 uL of each fraction onto the same column. The mobile phase was removed under d i s t i l l a t i o n at reduced pressure at room temperature and the residues were reconstituted in equal volumes of chloroform. Optical rotation studies of each fraction were conducted on a Perkin-Elmer Model 141 polarimeter in a 1 mL tube at 25°C. The rotation of the fraction which eluted ear l ier was -22°, and that of the fraction which eluted later was - 5 ° . 3.9.10 Structural Identity of the Derivatives of R.S Propafenone The two diastereomeric fractions containing the enantiomers, isolated as described above, were evaporated to dryness in PTFE-lined screw-capped culture tubes and reconstituted with 2 mL of methanol. A 2 uL aliquot of each fraction was injected onto the GC-MS (HP 5700A GC, interfaced with a Varian Mat-Ill Mass Spectrometer) using EI mode. Conditions - filament current: 300 uA, electron beam energy: 70 ev, ion source pressure: 8 X 10\" 6 torr, column: 3% s i l a r 10 C on chromosorb W-HP (2 m X 2mm i .d. ) , injection port temp.: 250°C, oven temperature: 150°C to 185°C at 8°C/min, column helium flow rate: 20 mL/min. 3.9.11 Efficiency of Recovery of Propafenone Enantiomers from plasma Samples of 1000 ng and 2000 ng (equivalent to the free base) of R,S propafenone hydrochloride were added to two 1 mL aliquots of plasma in two PTFE-lined screw-capped culture tubes. The samples were extracted with benzene as described before. In two separate tubes, 47 identical amounts of R,S propafenone free base were taken and a l l four tubes containing the samples were reacted with R(-)NEIC at 75°C for 30 minutes. The samples were reconstituted with 0.25 mL of 2-propanol and st irred by vortex for 1 minute. To each tube, 5 uL of 0.002% v/v of a-bromonaphthalene were added and the samples were assayed by HPLC. Recovery analyses were done as the percentage of peak height ratios of identical concentrations of enantiomers extracted, to that of the corresponding enantiomers unextracted. 3.10 Assay of Propafenone Enantiomers by High-Performance Liquid Chromatography with UV Detection 3.10.1 Extraction, Derivatization and HPLC Analysis Four aliquots of 1 mL of blank human plasma in t r i p l i c a te , in 10 mL PTFE-lined screw-capped culture tubes were spiked with 250, 500, 1000 and 2000 ng (equivalent to the free base) of R,S propafenone hydrochloride and 1000 ng of Li-1115. Additional d i s t i l l e d water was added to adjust to equal volumes. Plasma proteins were precipitated with 0.4 mL of 10% trichloroacetic acid as described in section 3.9.5. The pH of the samples was adjusted above 12 by the addition of 0.5 mL of 3 M sodium hydroxide and were extracted with 6 mL of benzene. Anhydrous sodium sulphate (1 g) was added to remove trace quantities of water from the organic extract. After centrifugation for 5 minutes at 2500 r.p.m., the organic portions were transferred to clean, dry test tubes. The extracts were evaporated to dryness at 40°C in a water bath and the residue in the culture tubes was reacted with 5 uL of R(+)MBIC in 150 uL of dry acetonitr i le at room temperature for 30 minutes. The excess 48 reagent was destroyed by the addition of 2 uL of water and the samples were held at room temperature for an additional 30 minutes. A 20 uL aliquot was injected onto the HPLC ODS column (25 X 0.45 cm i .d . ) . Mobile phase, methanol/water (70:30) delivered at 1 mL/min. Detection: 254 nm. 3.10.2 Extraction, Derivatization and HPLC Analysis of R,S Propafenone using GITC as a Chiral Derivatizing Reagent Four aliquots of 1 mL of blank human plasma in t r i p l i c a te , in 10 mL PTFE-lined screw-capped culture tubes were spiked with 250, 500, 1000 and 2000 ng (equivalent to the free base) of R,S propafenone hydrochloride and 300 ng of desipramine hydrochloride. Plasma protein precipitation and extraction steps were similar to section (3.10.1). Derivatization was carried by the addition of 150 uL of 0.5% w/v GITC solution in dry acetonitr i le at room temperature for 30 minutes. The excess reagent was destroyed with 2 uL of water and held for an additional 30 minutes at room temperature. A 20 uL aliquot was injected onto the HPLC ODS column. For very low concentrations, (less than 500 ng/mL) the whole volume of derivatized product was injected. Mobile phase: methanol/phosphate buffer (0.2 M) of pH 2.8 (75:25) delivered at 1 mL/min. Detection: 250 nm. 3.10.3 Extraction, Derivatization and HPLC Analysis of R.S Propafenone using R(-)NEIC as a Chiral Derivatizing Reagent Four aliquots of 1 mL of blank human plasma each in t r i p l i c a te , in 10 mL PTFE-lined screw-capped culture tubes were spiked with 250, 500, 1000 and 2000 ng (equivalent to free base) of R,S propafenone 49 hydrochloride and 300 ng of Li-1115 hydrochloride (or, desipramine hydrochloride). Additional d i s t i l l ed water was added to adjust to equal volumes before plasma precipitation. After extraction, as outlined in section (3.10.1), derivatization was carried out with 100 uL of 0.01% v/v R(-)NEIC in 400 uL of dry acetonitri le containing 0.003 M TEA. The tubes were heated at 75°C in a dry heating block for 30 minutes. After cooling to room temperature, 1 uL of piperidine was added to each tube to destroy the excess reagent. The contents of the tubes were evaporated to dryness and reconstituted with 200 uL of 2-propanol. A 20 uL aliquot was injected onto the HPLC ODS column. A larger volume was injected for samples of concentrations below 300 ng/mL of plasma. Mobile phase: methanol/water (76:24) delivered at 1 mL/min. Detection, 230 nm. 3.11 Calibration Curve and Precision of Assay of R,S Propafenone R(-)NEIC Derivatives. Five aliquots containing 250, 500, 1000 and 2000 ng (equivalent to the free base) of R,S propafenone hydrochloride solutions, in t r ip l i ca te , were placed in PTFE-lined screw-capped culture tubes containing 1 mL of blank plasma. To each tube, 300 ng of R,S Li-1115 (internal standard) was added. Additional d i s t i l l ed water was added to adjust to equal volumes. Extraction and derivatization with R(-)NEIC were done as described in section 3.10.3. The cal ibration curves were constructed by plotting peak-height ratios of each enantiomer to those of the known concentrations of propafenone and the internal standard. Inter-assay var iab i l i ty was determined from the t r ip l i ca te preparations, whereas intra-assay var iab i l i ty was determined by 50 t r ip l i ca te injection of four of the samples containing 250, 500, 1000 and 2000 ng of R,S propafenone. 3.12 Reverse Phase Thin Layer Chromatography of Propafenone-GITC and Propafenone-Dansvlhydrazine-GITC Derivatives An aliquot of 0.1 mg of propafenone base was reacted with GITC in 100 uL acetonitr i le at room temperature for 30 min. as described in section 3.6.6. The reaction mixture, propafenone free base in acetonitr i le, and 0.5% GITC in acetonitri le were spotted on a KCjg reverse phase TLC plate and developed with methanol/phosphate buffer (0.2 M) of pH 2.8 (76:24). An unreacted propafenone spot (Rf=0.29, a propafenone-GITC derivative spot (Rf=0.21) and a spot for excess GITC (Rf=0.46) were observed in the reaction mixture. Another test tube containing 0.1 mg of propafenone base was reacted with an excess of dansylhydrazine in the presence of tr ichloroacetic acid in toluene at 75°C for 15 min. After dansylation the reaction mixture was spotted on a KCj 8 TLC plate along with propafenone and dansylhydrazine. Propafenone-dansylhydrazone (Rf=0.53) and excess dansylhydrazine (Rf=0.93) were ident i f ied. No residual propafenone spot was detectable on the plate. After extracting the propafenone-hydrazone derivative with 5 mL of benzene at pH >12, followed by evaporation to dryness as before, the residue was reacted with an excess of 0.5% GITC in 100 mL of acetonitr i le at room temperature for 30 minutes. TLC analysis of the reaction mixture indicated three spots. Among these, one fluorescent spot of propafenone-dansylhydrazine-GITC (Rf=0.2) and excess GITC (Rf=0.45) spot were identi f ied. A third spot (Rf=0.91), which eluted with the solvent 51 front was also fluorescent and was identif ied as the GITC-dansyl hydrazine adduct. 52 4. RESULTS AND DISCUSSION 4.1. Analytical Development for Chromatographic Resolution of Propafenone Enantiomers 4.1.1 Gas-Liquid Chromatographic Studies on Chiral and Achiral Stationary Phases Init ia l GLC experiments for the resolution of enantiomers of propafenone (Fig. 1) were carried out based upon the principles of direct separation of the enantiomers on commercially available chiral stationary phases. A Chirasi l -Val^ capi l lary column, which was coated with opt ical ly active, N-isobutyryl-L-valine tert . butylamide as the stationary phase, was used at i ts maximum recommended temperature of 220°C coupled with an electron capture detector. The heptafluorobutyryl derivatives of propafenone enantiomers eluted in R+c=29.34 minutes without enantiomeric separation (Fig. 2A). However, underivatized propafenone did not elute from the column under the similar conditions but using FID. The trif luoroacetyl derivatives of the enantiomers of propafenone likewise were unresolved under the conditions employed. This chiral phase has been demonstrated to be effective for the resolution of primary amines, amino acids and amino alcohols (McErlane and P i l l a i , 1983, Frank et a l . , 1978) but was not found suitable for the chiral separation of enantiomeric secondary amines such as propafenone. It was concluded that the lack of an amino hydrogen in the substituted amide derivatives of propafenone as depicted in f i g . 3A, reduced the interactions with the chiral stationary phase and thus prevented chiral resolution. Figure 1. Structure of Propafenone Enantiomers 54 cu c o c cu 4-re C L o s-0) c o c cu 4— rt C L o i-m I — A Figure 2. Representative Chromatograms Showing Unresolved Peaks of R,S Propafenone Under Various GLC Conditions. A) HFBA derivative of propafenone chromatographed on Ch i ra s i l -Va l R capi l lary column at 220°C, detection ECD. B) N-TPC derivative of propafenone chromatographed on SE-30 capi l lary column at 180°C, detection FID. C) R(+)MBIC derivative of propafenone chromatographed on SE-30 capi l lary column at 190°C, detection FID. D) R(+)MBIC derivative of propafenone chromatographed on OV-17 packed column at 280°C, detection FID. 55 P r o p a f e n o n e S t a t i o n a r y P h a s e B P r o p a f e n o n e S t a t i o n a r y P h a s e Y« NH ( c o v a l e n t ) Y« 6NH 3 ( i o n i c ) F i g u r e 3 . S t e r e o c h e m i c a l I n t e r a c t i o n Between HFBA D e r i v a t i v e o f P r o p a f e n o n e E n a n t i o m e r s and C h i r a s i l - V a l R S t a t i o n a r y P h a s e (A) and B e t w e e n N a p h t h o y l C h l o r i d e D e r i v a t i v e o f P r o p a f e n o n e and P i r k l e R S t a t i o n a r y P h a s e ( B ) . 56 To increase the hydrogen bonding capacity of propafenone for the Ch i ras i l -Va l R stationary phase, racemic propafenone was reacted with isopropylisocyanate to form the urea derivatives. However, GLC analysis of the enantiomer derivatives on this chiral column s t i l l fa i led to resolve the enantiomers (figures not shown). Alternative GLC methods were chosen based on the principles of diastereomeric separation of the enantiomers of propafenone on suitable achiral phases. The dif ferentia l thermodynamic s tab i l i t y aided by the conformational r i g id i ty of the diastereomer molecules (Rose et a l . , 1966) have been noted to fac i l i t a te resolution of enantiomers. Fig. 2B is a representative chromatogram of the diastereomeric N-tr i f luoro-acetyl-L-proloyl chloride (N-TPC) derivatives of propafenone. An SE-30 capi l lary GLC column was used at a temperature of 180°C coupled with a flame ionization detector. An unresolved peak of the enantiomer derivatives eluted in Rt=12.39 minutes. A few extra peaks appeared which were thought to be due to the reported (Silber et a l . , 1980) degradation of the reagent during storage, even at low temperature. A second reagent that was employed for the diastereomer formation was R(+)a-methylbenzylisocyanate. However, the enantiomers of propafenone, as their diastereomeric urea derivatives, could not be resolved by the SE-30 capi l lary column nor by an OV-17 packed column. The columns were operated at 190°C and 280°C, respectively (Fig. 2C and 2D). The unresolved peaks in the chromatograms eluted in 13.85 and 4.03 min. respectively (Fig. 2C and 2D). Parallel blank experiments, without propafenone, were carried out to identify the propafenone peaks in each of the chromatograms discussed. 57 4.1.2 High-Performance Liquid Chromatographic Studies on Chiral and Achiral Stationary Phases A major emphasis was given to HPLC as a method for the direct separation of propafenone enantiomers. The f i r s t chiral HPLC column employed was a Pirkle (1-A) column, containing N-3,5-dinitrobenzoyl-D-phenylglycine ionical ly bonded to 7-amino si lanized s i l i c a . Due to the strong interaction of the propafenone base with this stationary phase, propafenone did not elute from this column. The /3-naphthoyl chloride derivative of propafenone was chromatographically examined using 9% 2-propanol in hexane as the eluting solvent (Fig. 4A). The polarity of the mobile phase was varied by changing the proportion of 2-propanol in the mixture, however, there was no resolution of the enantiomers with any of the mobile phases employed. In order to alter the interactions, the a-naphthoyl and 2,4 dinitrobenzoyl derivatives were examined. However, no resolution was observed with either of the derivatives. Direct HPLC separation of the propafenone enantiomers was also attempted using a /3-cyclodextrin bonded column and methanol/water (65:35) as the mobile phase, delivered at a flow rate of 0.8 mL/min. In the experimental study performed, as shown in f i g . 4B, no separation of enantiomers was observed by the fact that an unresolved peak of propafenone eluted in 60 min. in the chromatogram. This /J-cyclodextrin stationary phase usually fac i l i tates the separation of solutes, including chiral ones, by the principle of exclusion chromatography. The exact f i t t i ng of the solute in the desired clatherate, hydrogen bonding interactions between the solute and the stationary phase at the mouth of the cyclodextrin cavity and hydrophobic interactions (depicted 58 P r o p a f e n o n e B P r o p a f e n o n e \\ 20 40 60 0 20 40 R E T E N T I O N T I M E ( M I N ) 60 F i g u r e 4 . R e p r e s e n t a t i v e C h r o m a t o g r a m s S h o w i n g U n r e s o l v e d P e a k s o f R , S - P r o p a f e n o n e by HPLC. A) / J - N a p h t h o y l d e r i v a t i v e o f p r o p a f e n o n e c h r o m a t o g r a p h e d on P i r k l e c o l u m n . HPLC c o n d i t i o n s : m o b i l e p h a s e ; 2 - p r o p a n o l / h e x a n e ( 9 : 1 0 0 ) d e l i v e r e d a t 1 mL/min; d e t e c t i o n 254nm. B) U n d e r i v a t i z e d p r o p a f e n o n e c h r o m a t o g r a p h e d on / J - c y c l o d e x t r i n c o l u m n . HPLC c o n d i t i o n s : m o b i l e p h a s e ; m e t h a n o l / w a t e r ( 6 5 : 3 5 ) d e l i v e r e d a t 0.8 mL/min; d e t e c t i o n 254nm. 59 Figure 5 . The General Mechanism of Inclusion-Complexing in 0-Cyclodextrin Stationary Phase (A), and the Inclusion Complex Formation of Propafenone Enantiomers Within the 0-Cyclodextrin Cavity (B). . 60 in Fig.5), leads to enantiomeric resolution. It was thus concluded that the essential interaction of the propafenone enantiomers were absent with the /?-cyclodextrin phase. 4.2 High-Performance Liquid Chromatographic Resolution of the Propafenone Enantiomers 4.2.1 Resolution of R(+)-a-Methv1benzylisocyanate fR(+)MBIC1 Derivatives of Propafenone Enantiomers The f i r s t successful resolution of propafenone enantiomers was achieved by HPLC using commercially available R(+) o-methylbenzylisocyanate [R(+)MBIC] as the chiral derivatizing reagent. Both R,S propafenone and the internal standard, R,S Li-1115 reacted with R(+)MBIC at room temperature in 30 minutes. The resulting diastereomeric urea derivatives were resolved on a 5 u ODS column with a mobile phase of methanol/water (68:28) delivered at a flow rate of 0.8 mL/min. Resolution (R$=1.25) of propafenone enantiomers was obtained in less than 60 minutes. The retention time could be decreased to 40 minutes (Fig. 6A) at the expense of resolution (R$=1.15) using an increased proportion of methanol in the mobile phase of methanol/water (70:30). Addition of tetrahydrofuran and acetonitr i le as organic modifiers to the methanol/water mobile phase abruptly decreased the resolution. Unfortunately, using this derivatizing reagent, the lower detection l imit of each enantiomer was 500 ng/mL in plasma and the method was found not to be suff ic ient ly sensitive (Table 1, page 61) for pharmacokinetic studies. Moreover, the reagent was extremely unstable and degradation of the reagent was observed during a few days of storage Table 1. Data Representing the HPLC Separation of Enantiomers of Propafenone. COLUMN USED SOLVENT SYSTEM ( l m L / m i n ) DETECTION METHOD DERIVATIZING REAGENT RETENTION TIME(MIN) R ( - ) ;B<>) RESOLUTION FACTOR (R A) DETECTION LIMIT OF E N A N T I O M E R SIG./NOISE RATIO ODS 5JU (25 x 0 . 4 5 c m i . d . ) 70:30 METHANOL/ WATER 254 nm R ( + ) M B I C 4 0 ; 42 1 . 1 5 500 ng 3 : 1 ?5:*5 M E T H A N O L / 0.2M PHOS-BUFF .pH2.ti 250 nm G I T C 1 2 ; 1 4 . 5 1.4 150 ng 4:1 76:21* METHANOL/ WATER 230 nm R(-)NEIC 1 9 . 5 ; 2 1 1-25 IOO ng 3 : 1 7 5 : 2 5 METHANOL/ WATER FLUORO. EX .220 nm EM.J+18 nm DANSYL HYDRAZINE & R(-)NEIC 21 .7; 23- '* 1.35 2 . 5 ng 3 : 1 62 under nitrogen at 4°C. Similar information was reported by Gal et a l . , 1981. 4.2.2 Resolution of the 2.3,4,6 Tetra-O-Acetvl-B-D-Glucopvranosvl Isothiocyanate (GITC) Derivatives of Propafenone Enantiomers Prior to conducting this experiment, the chiral derivatizing reagent GITC was synthesized from a-D-bromoglucose. The purity of the compound was checked by reverse phase HPLC and functional group characterization was studied by IR. A characterist ic chromatographic peak and the C=N=S stretching frequency at 2100 cm\" 1 depicted in f i g . 7 indicated that the GITC reagent thus synthesized was pure and no evidence of enantiomeric inversion was observed. The GITC synthesized was allowed to react with R,S propafenone at room temperature in 30 minutes and the resulting diastereomers were resolved on an ODS column (Fig.6B). The solvent system employed was methanol/phosphate buffer (0.2 M) at pH 2.8 (75:25) delivered at flow rate of 1 mL/min. Near baseline resolution (R$=1.4) of the diastereomers was obtained within 15 minutes. The internal standard, desipramine as i ts thiourea derivative, eluted in 24 minutes. The thiourea derivatives, which have maximum molar extinction coefficients at 250 nm (Nimura et a l . , 1980) provided a minimum detection l imit of 150 ng of each enantiomer of propafenone injected onto the column. In addition to effective resolution of the enantiomers, the GITC reagent could be easily handled and was found more stable than R(+)MBIC during storage. J 10 20 30 40 50 —i 30 0 10 20 RETENTION TIME (MIN) i— o 1. R ( - ) PROPAFENONE 2. S(+) PROPAFENONE 3. D E S I P R A M I N E ( I . S . ) k. R ( - ) L i - 1 1 1 5 ( 1 . S . ) 5. S(+) L i - 1 1 1 5 3 10 20 30 —i 40 Figure 6. Reverse Phase HPLC Separation of Enantiomers of Propafenone as their Diastereomeric R(+)-o:-Methylbenzyl Urea (A), /J-D-Glucopyranosyl Thiourea (B) and R(-)-l-(Naphthyl)ethyRJrea (C) Derivatives. Chromatographic Conditions are given in Table 1. Figure 7 . IR and HPLC Analyses of Acetylated a-D-Bromog1ucose and GITC. A) IR spectra of acetylated a-D-bromoglucose in nujol. Ester carbonyl stretching frequency at 1750 cm\" 1. B) IR spectra of acetylated GITC in nujol. Ester carbonyl stretching frequency at 1750 cm\" 1 and N=C=S stretching frequency at 2100 cm\" 1 . C) Chromatogram of acetylated a-D-bromoglucose in HPLC. Column; ODS. Mobile phase; methanol/water (65:35) at 1.2 mL/min. Detection; 254 nm. D) Chromatogram of GITC after third recrysta l l izat ion. 65 4-2.3 Resolution of the Rf-)-1-(Naphthyl )ethv1isocyanate fR(-)NEICl Derivatives of Propafenone To take advantage of increased uv absorption characteristics of naphthyl over the benzyl moiety substituted on the chiral centre of the isocynate reagent, a third chiral reagent, R( - ) - l -(Naphthyl)ethylisocyanate [R(-)NEIC] was chosen for the derivatization of R,S propafenone. This reagent also formed diastereomeric urea derivatives with propafenone. The resolution of the enantiomers (R$ 1.25) was obtained on a 5u ODS column within 24 minutes using methanol/water (76:24) as the mobile phase (Fig. 6C and Table 1). Detection at 230 nm was employed since the derivative provided maximum absorption at this wavelength (Fig. 8E). The minimum detectable quantity was 100 ng/mL in plasma for each enantiomer. When the pH of the mobile phase was adjusted below pH 7 using perchloric acid, the elution time of the diastereomeric derivatives from the ODS column was substantially increased, which was considered to be due to the reduced solvation effect of the acidic mobile phase with the naphthyl derivatives. 4.3 Elution Order of R and S Propafenone In order to determine the elution order of the enantiomers of propafenone as their R(-)NEIC derivatives, the two diastereomeric fractions corresponding to the two peaks on the chromatogram (Fig. 6C) were separately collected in two flasks and the mobile phase was evaporated to dryness as described in section (3.9.9). The elution order of the peaks was determined to be R(-) propafenone, followed by S(+) propafenone. This was confirmed by polarimetric analysis of the 66 relative degree of rotation of each fraction of diastereomer. The rotation of two fractions containing the purif ied diastereomers was -22° and - 5 ° , respectively. According to the ' rule of sh i f t ' reported by Freudenburg, 1933 and Finar, 1969, the more negative rotation was attributed to that from the R(-) propafenone, whereas the S(+) propafenone provided the least negative rotation due to the constant additive rotational contribution from the R(-)NEIC reagent. Similar elution order of R(+) propranolol and S(-) propranolol by reverse phase HPLC has been reported by Thompson et a l . , 1982 using S(-)MBIC as the derivatizing reagent. 4.4 Confirmation of Structures of R and S Propafenone Derivatives The two HPLC fractions corresponding to the derivatives of R(-) and S(+) propafenone were individually collected and subjected to GC-MS analyses in the EI mode as described in section (3.9.10). Both fractions provided identical total mass-ion chromatograms (TIC) (Fig. 9A and 9B) and EI mass spectra from scan 160 (Fig. 10A and 10B). The fragmentation patterns of mass numbers of 77, 127, 155, 182 and 197 corresponded to the fractions of the derivatives, are depicted in f ig 10. Underivatized propafenone apparently did not elute out from the GLC column under identical chromatographic conditions as evidenced by the lack of any peak in the TIC. 4.5 Sensit ivity of Propafenone and its Derivatives Unlike metoprolol, atenolol, or salbutamol, propafenone did not show any fluorescence despite the fact that a fluorogenic aryloxy moiety 67 200 250 300 350 NM Figure 8. Ultraviolet Absorption of Propafenone and its Derivative in Different Solvents. Solvent systems: (A) Propafenone base in acetonitr i le/ phosphate buffer (5 mM) of pH 2.8 (65:35). (B) Propafenone base in acetonitr i le. (C) Propafenone base in methanol/phosphate buffer (0.2 M) of pH 2.8 (75:25). (D) Propafenone hydrochloride in water. (E) R(-)NEIC derivative of propafenone in methanol/water (75:25). 68 SCAN SCAN Figure 9 . Total-Ion-Mass Chromatograms of Derivatives of S'+lProDafenone (A) and R(-)Propafenone (B) with R(-)-1 -(Naphthyl)ethylisocyanate. GC-MS conditions: Filament current,300 uA, Electron beam energy, 70 ev, Ion-source pressure, 8 X 10-6torr,Injection port temperature, 250*C, Column (a 3«/ s i l a r 10 C on chromosorb W-HP, 2m X 2mm i.d.) temperature , 150°to 285°C at 8\"C / min. Column helium flow rate, 20 mL/min. Figure 10 . EI Mass Spectra of S(+)Propafenone (A) and R(-)Propafenone (B) as their R(-)-l-(Naphthyl) ethylisocyanate Derivatives. 70 is common to al l of these compounds. The absence of fluorescence of propafenone was ascertained on a Perkin-Elmer model 650-10S fluorescence spectrophotometer using a variety of solvents. The carbonyl group ortho to the aryloxy group was believed to provide quantum deactivation effects on propafenone thus eliminating any fluorogenic properties. The carbonyl group conjugated to the aryl moiety of propafenone structure did ,however, confer sensit iv ity for uv detection. A comparative study of the shift of the primary uv bands in different solvents is given in Fig. 8. Although the highest uv sensit iv ity for detection at 209 nm could be attained using acetonitrile/phosphate buffer at pH 2.8 (65:35), this did not provide enantiomeric resolution of propafenone derivatized with the chiral isocyanates. All solvent systems for enantiomeric resolutions required methanol in various proportions (Table 1) and hence i t was necessary to work in the range of 248-254 nm. The R(-)NEIC derivative of propafenone provided the highest X m a x at 230 nm, using methanol/water (75:25) as mobile phase and a minimum detection l imit of 100 ng (at the detector) for each enantiomer was observed (Table 1). The (+) naproxen derivative of propafenone, although assumed to provide better sensit iv ity due to its highly conjugated structure, the individual enantiomers could not be resolved by HPLC using methanol/water (76:24) as the mobile phase delivered at 1 mL/min. (figure not shown). 4.6 Mechanism of Diastereomeric Resolution of Propafenone The mechanism of diastereomeric resolution of propafenone is depicted in figure 11. The essential cr i ter ion of resolution of the diastereomers was the formation of intramolecular hydrogen bonds between 71 the hydrogen of the secondary alcoholic group and the oxygen of the ureido carbonyl oxygen of the urea derivative formed, thus giving rise to maximum conformational dyssymmetry in two diastereomeric structures. These two structures however, have different part it ion coeff icients and would enantiomerically separate by reverse phase HPLC. A similar approach has been reported by Thompson et a l . . 1982, for the separation of the enantiomers of propanolol as their S(-)MBIC derivatives by reverse phase HPLC. The same authors also emphasized the role of this hydrogen bonding by the fact that blocking the OH-group by conversion to its TMS ether, abolished the diastereomeric resolution. From the structural consideration of the three kinds of chiral derivatizing reagents employed, which are shown in fig.12, the R(-)NEIC reagent imparted greater conformational restraint than R(+)MBIC and proved to be the better resolving reagent of the two. However, the most bulky acetylglycosyl residue of GITC reagent imparted the greatest conformational r i g id i ty in the structure of the diastereomeric thiourea derivatives of propafenone and therefore provided the maximum resolution. According to Nimura et a l . , 1980, GITC derivatives of chiral amines confer both bulkiness and hydrophobicity, thus giving rise to an increased interaction with an ODS stationary phase and hence better resolution. In l iquid chromatography, the d i f ferent ia l solvation of the solutes by the eluting solvents also plays a major role on the mechanism of separation (Pirkle et a l . , 1981). The GITC residue of the derivative of propafenone was better solvated in the mobile phase employed and provided better resolution within a shorter time (15 minutes, Fig. 6B). Figure 11. Mechanism of Diastereomeric Separation of Propafenone Urea and Thiourea Derivatives 73 R(+)-a-Methylbenzyli socyanate ( R(+)MBIC ) R(-) -1-(Naphthyl)ethylisocyanate ( S(-)NEIC ) 2,3,4,6 Tetra-O-Acetyl /}-D-Glucopyranosyl isothiocyanate ( GITC ) Figure 12. Structure of the Chiral Derivatizing Reagents. 74 4.7 Intramolecular Hydrogen Bonding and Conformation of Propafenone 4.7.1 HPLC Separation of the Conformers Propafenone free base, when chromatographed on the ODS column, using a mobile phase of pH 5.5 containing methanol/phosphate buffer (0.2 M) (75:25), exhibited one major peak with a small shoulder (Fig. 13A). This was considered to be due to the existance of at least two conformational states with almost identical 1 ipophi l ic i ty and solvation properties. Upon examining the molecular model of propafenone i t was clear that the free base of propafenone could exist in two conformational states due to the formation of two alternate intramolecular hydrogen bonds. Due to the proximity of the aryl carbonyl group ortho to the side chain containing the secondary alcoholic and amino groups, two different conformations may result from intramolecular hydrogen bonding. This has been shown in scheme 1, where the conformers #2 and #3 are the propafenone free base in two different conformations, due to alternative intramolecular hydrogen bonds. The predominance of one, which corresponded to the major peak in the chromatogram of f i g . 13A, was due to the greater f l e x i b i l i t y of the cycl izat ion. The ten membered cycl ic form as shown in scheme 1, #3 would be more f lexib le than the nine membered form #2, thus giving rise to the existence of two conformational forms of propafenone. Further evidence of the formation of a hydrogen bond in propafenone was found when a low amplitude broad peak appeared in the range of 3275 to 3525 cm' 1 in the IR spectrum (Fig.14). Similar phenomena of intramolecular hydrogen bonding and the predominance of one conformer over the other 75 has been reported by Kuhn et a l . . 1964 and Murthy et a l . , 1968, for 1,4-diols and other substituted aromatic alcohols and acids. 4.7.2 Reaction of Conformers with Chiral isocyanates and the HPLC Resolution of the Diastereomers Of the two conformers of propafenone, the one with a free N-H group reacted with a chiral isocyanate to form diastereomeric ureide derivatives, where the two chiral centers in the diastereomer are separated by six atoms. The enantiomeric resolution of R(-) and S(+) propafenone as their ureide derivatives would happen when a stable conformational r i g id i ty was attained by the formation of an intramolecular hydrogen bond between the alcoholic OH group and the ureido carbonyl group as discussed in section 4.6. However, an alternative intramolecular hydrogen bond, between the same OH group and the aryl carbonyl group of propafenone was also feasible. The later conformation would negate the diastereomeric resolution of propafenone enantiomers. This has been shown in scheme 1. The conformers #4 and #5 meet the essential c r i te r i a for resolution due to the formation of a very specif ic intramolecular hydrogen bond in each case. The conformer #6 however, is formed by an alternate hydrogen bond between the OH and the aryl carbonyl group and would not follow the c r i t e r i a for resolution. From HPLC studies, both resolved and unresolved peaks at different retention times were found as shown in f i g . 13B. In agreement with scheme 1, two pairs of resolved peaks #4 and #5 with R^. of 24 and 30 min. respectively and one unresolved peak with Rj. of 35 min. were found in the HPLC chromatogram of f i g . 13B. 76 OQ 9 R(-) \\ S(+) / RETENTION TIME(MIN) 10 ao 30 40 Figure 13. Representative Chromatograms of Underivatized (A) and Derivatized (B) Propafenone in HPLC. A) Propafenone free base elutes as one prominent peak (3) with a shoulder (2). HPLC conditions; mobile phase, 25% phosphate buffer (0 2 M) in methanol at pH 5.5, delivered at 1 mL/min. Column; 5u ODS, 25 x 0.45 cm i .d . Detection; 248nm. B) Diastereomeric R(-) NEIC derivatives of propafenone appear as three conformational forms (4, 5, & 6) of which 4 and 5 resolve into R(-) and S(+) isomers. Mobile phase; 20% water in methanol delivered at 1 mL/min. Column; as above. Detection; 230 nm. Figure 14 . Infrared Spectrum of Propafenone in Nujol Mull. o CH, Propafenone.HC1 M H .HCI N>-CMy-C-CHa-N-CHrCHj-CH , PROPAFENONE BASE WITH FREE -NH-GROUP ^/X>-CtCV0-M PROPAFENONE BASE WITH LOCKED -NH-GROUP R ( _ ) - 1 - ( N a p h t h y l j e L h y l -i s o c y a n a t e O* 00 0 IM* H-i-CM, \" O -CMi-d-CH,- N - C H J - C H J - C H J 00 P r o p a f e n o n e u r e a d e r i v a t i v e P r o p a f e n o n e u r e a d e r i v a t i v e v-CH^-A-CM^-N-CHrCMj-CM, H Propafenone urea derivative Scheme 1. Schematic Diagram Showing the Intramolecular Hydrogen Bonding and Conformational Isomers of Underivatized (2,3) and Derivatized (4,5 & 6) Propafenone. 79 4.8 Mechanism of Propafenone Isocyanate Reaction The reaction of an amine with an isocyanate is governed primarily by the basicity or nucleophil icity of the N-H bond to form the substituted ureide. Isocyanate reactivity is decreased for substituents with increasing electron donating effect whereas the amine react iv i ty is increased for substituents with increasing electron donating effect of the substituents (Arnold et a l . , 1957). Briody and Narinesingh, 1971, determined that the reaction kinetics of amines with isocyanates s t r i c t l y followed second order kinetics and that the rate l imit ing step was the slow proton transfer in the intermediate which was aided by a catalyst. The proton exchange mechanism between the catalyst and the amine-isocyanate intermediate has been confirmed by the primary hydrogen isotope effect, using a deuterated catalyst (Briody et aj.., 1971). Catalysis by the reacting amine, as well as by the product ureide, were also studied in such reactions (Satchell and Satchell, 1975). Bifunctional catalysts such as amides, ureides and carboxylic acids with very weak acidity were considered to be very effect ive. These catalysts form cycl ic transition states in aprotic solvents which f ac i l i t a te the slow proton transfer in the intermediate leading to completion of the reaction. In a few cases, triethylamine, as a weak basic catalyst, has been employed but the mechanism of catalysis is s t i l l uncertain (Briody et a l . , 1971, Satchell et a l . , 1975). Propafenone is a secondary amine and was allowed to react with two different chiral isocyanates and a chiral isothiocyanate to form diastereomeric derivatives. A typical reaction scheme of propafenone and R(-)NEIC is depicted in scheme 2. The bipolar intermediate is autocatalysed by the reacting propafenone to the product ureide. ISOCYANATE - + H— N - C 3 H ? CH- N = C - 0 C H , K - N - C 3 H 7 H INTERMEDIATE PRODUCT AM I N K H - N - C - H . PRODUCT UREA PROPAFENONE(AMINE) B WHERE, K, & K_^) K 2 ( 0 R , K 3 ) ti— -CH 2-CHOH-CH -O 0= C - C H a - C H a MECHANISM FOR CATALYST-MEDIATED REACTION AMINE ^ T ~ C H - N = c - o\" C A T A L Y Z E ^ ' CH, I 3 CH- N = C - 0 H ^ 3 ^ C 3 H ? R V PHOUUCTS. , INTERMEDIATE CATALYZED | 1 H / \\ C 3 H ? P5 - / CH- H = C - 0 / II C,H„ H - N - C - N ^ > ' \\ V R CH 3 H PRODUCT UREA ^ H C=»0 CH 3 A^C }H 7 S c h e m e 2. R e a c t i o n S c h e m e f o r t h e D e r i v a t i z a t i o n o f P r o p a f e n o n e w i t h R ( - ) - l n a p h t h y l e t h y l i s o c y a n a t e . 81 However, the reaction of propafenone with isocyanate or isothiocyanate does not go to completion under the conditions employed and therefore a portion of residual propafenone is le f t unreacted. This was shown, with GITC as the derivatizing reagent, by evaluation of the reaction products using reverse phase TLC (Fig. 15). In the presence of excess reagent and at optimum reaction conditions a residual propafenone spot (Rf=0.29) was observed. Moreover, at very low concentrations of propafenone, below a reaction concentration of 1.67 ug/mL, the reaction between propafenone and isocyanate or isothiocyanate was found to be very slow and apparently stopped, since no increase in peak height of the urea derivative was observed between one to sixteen hours in HPLC. The retarding effect of the reaction kinetics can be postulated to be due to two factors. At low concentrations of propafenone, the triggering auto-catalytic effect of the amine would be less prominent and the rate of the reaction would be retarded. However, the essential, and rate determining step, for the completion of the reaction is the proton exchange between the catalyst and the intermediate. Due to the possible intramolecular hydrogen bonding between the amino-hydrogen and the aryl carbonyl group, the exchangeable hydrogen is not readily available to perpetuate the catalysing act iv i ty . Moreover, the exchangeable proton in the intermediate is not suf f ic ient ly lab i le , since i t is derived from propafenone i t se l f . These two factors are believed to retard the isocyanate/ propafenone reaction and v i r tua l ly impede the reaction when the propafenone level f a l l s below a reaction concentration of 1.67 ug/mL. 82 Reverse Phase Thin-Layer Chromatography P R O P A F E N O N E ( P P F ) G I T C R T / 3 0 m i n . 1 DANSYL H Y D R A Z I N E ( D H ) 75° C / 1 5 m i n . P P F - D H P P F P P F - G I T C G I T C DH ( R f = . 9 3 Q i T C ( R f = - 5 3 ) (Rf= . 2 9 ) (Rf= . 2 1 ) (Rf=.M>) R T / 30min. P P F - D H - G I T C (Rf= . 2 ) G I T C DH-GITC (Rf= . i f 5 ) ( R f = . 9 D : : ELI)TING S O L V E N T : 7 6 , 2 4 - M E T H A N O L / P H O S . B U F F E R ( . 2 M ) , p H = 2 . 8 Figure 15, Single vs. Dual Derivatization of Propafenone. 83 4.9 Dual Derivatization of Propafenone and HPLC Separation of Enantiomers To overcome the effect of intramolecular hydrogen bonding between the aryl carbonyl group and the secondary amino hydrogen and to provide the necessary enantiomeric resolution, a dual derivatization technique was adopted (Scheme 3). The aryl carbonyl group was reacted with dansyl hydrazine in an acidic medium and the resulting dansyl hydrazone was further derivatized with a chiral isocyanate to form the diastereomeric derivatives. From reverse phase thin-layer chromatographic experiments, i t was observed that the derivatization of propafenone with GITC would go to completion i f propafenone was pre-reacted with dansyl hydrazine (Fig. 15). Similarly R(-)-NEIC reacted quantitatively with the propafenone-dansylhydrazine derivative to yield the dual derivative which was enantiomerically separated on the ODS column (Fig. 16). The resolution factor (R$=1.35) was s l ight ly better than that of the propafenone-R(-)-naphthyl urea derivatives (R$=1.25) presumably because of better solvation by the mobile phase of methanol/water (75:25). Using fluoroscence detection with excitation at 220 nm and emission at 418 nm, the detection l imit for each enantiomer was 2.5 ng in plasma. However, there were limitations to this method. The excess fluorescent dansylhydrazine from the reaction mixture could not be e f f i c ient ly removed using excess acetone to react with the residual dansyl hydrazine. As a result, the fluorescence detector photomultiplier tube was over excited with the excess derivatizing reagent. Moreover, the assay method could not be optimized to the desired precision level required for pharmacokinetic studies. 84 Scheme 3 . Chemical Pathway for Dual Derivatization of Propafenone (2 & 3) from the Diastereomeric Urea Derivatives of Dansyl-Propafenone Hydrazone (5) 85 « R ( - ) 03 CO M • s ( + ) 10 20 RETENTION TIMB(MIK) 30 igure 16. HPLC Separation of Enantiomers of Propafenone After Dual Derivatization. Chromatographic Conditions are given intable 1. 86 4.10 Kinetics of Derivatization and HPLC Analysis The most effective derivatizing reagents that led to resolution of the enantiomers of propafenone were R(+)MBIC and GITC and R(-)NEIC Therefore, the optimal derivatization conditions for reaction of the reagents with propafenone were studied. The peak-height ratio of R,S propafenone derivatized with R(+)MBIC did not s igni f icant ly change between 30 to 60 minutes of derivatization at room temperature (Fig. 17A). The optimum time for the derivatization of R,S propafenone with GITC was found to be between 20 and 30 minutes (Fig. 17B). Therefore, the reactions were allowed to proceed for 30 minutes at room temperature in further studies. The R(-)NEIC reagent provided constant peak-height ratios from 30 to 60 minutes at room temperature, but when the temperature was varied, keeping the time constant, a slight increase of peak-height ratio was observed between 60 to 75°C (Fig. 17C and 17D). Each of the three kinds of derivatives of propafenone formed with R(+)MBIC, GITC and R(-)NEIC were found to be stable for at least 72 hours at room temperature. 4.11 Stoichiometric Ratio of R(-)NEIC Reagent to Propafenone In establishing the reaction kinetics of propafenone with R(+)MBIC and GITC a large excess of reagents were employed. The excess reagent, when destroyed with water resulted in smaller reagent peaks which did not interfere with the diastereomer peaks of the propafenone enantiomers. R(-)NEIC, on the contrary, being re lat ive ly more 30 4 5 TIME(MINUTE) 30 45 TIME(MINUTE) 60 6 0 1.5 s . En 33 CD M W ? .5 W a . B e-i < a en W i • R(-)Propafenone • S(+)Propafenone 15 30 4 5 TIME(MINUTE) A 60 15 30 45 60 /5 TEMPERATURE(° C) 90 Time and Temperature Dependence for Derivative Formation. A) At room temperature treated with R(+)MBIC. B) At room temperature treated with GITC. C) At room temperature treated with R(-)NEIC. D) At varying temperature for 30 minutes treated with R(-)NEIC. 88 l ipophi l i c exhibited a broad peak due to the excess reagent. In order to optimise the stoichiometric ratio of R(-)NEIC reagent and propafenone, the same amount of R,S propafenone was derivatized with 1.8, 4.8, 9, 45 and 90 molar excess of the reagent (Fig. 18). At concentrations of R(-)NEIC beyond a 45 molar excess, propafenone enantiomeric peaks were masked by the excess reagent peak. The peak height ratios did not s ignif icantly change above a 1.8 molar excess of reagent. For convenience of having a clean and uninterferred chromatogram of propafenone enantiomers, a 2 to 20 molar excess of reagents was used for the analytical assay. 4.12 Recovery of Propafenone Enantiomers from Plasma Propafenone is highly protein bound. The i n i t i a l recovery of total drug from alkalinized plasma (pH >12) was lower than compared to that of water. Plasma protein precipitation with 10% tr ichloroacetic acid prior to basif ication and extraction of the drug into benzene (6 mL) substantially increased the recovery of each enantiomer. The eff iciency of recovery for each enantiomer as i t s R(-) NEIC derivative was calculated from the calibration curve of corresponding pure free base enantiomers derivatized similarly. For the measurement of peak-height ratio for each enantiomer a-bromonaphthalene was employed as an external standard. The recovery was 77.7% for R(-) propafenone and 78.2% for S(+) propafenone for concentrations of 500 ng and 1000 ng/mL (Table 2). Similar recovery values (78%) of racemic propafenone have been reported by Harapat et a l . . 1982, when the drug was extracted into 1% isoamyl alcohol in heptane. More than 90% recovery of racemic propafenone was reported by Chan et a]_., 1987, when the authors 89 25' R(-) propafenone S(+) propafenone 0 1.8 i+.5 9 k$ 90 MOLAR RATIO (REAGENT/DRUG) Figure 18. Stoichiometric Ratio of Reagent to Drug [R(-) -l-(Naphthyl )ethyl i socyanate/Propafenone]. 90 calculated the recovery of the racemic propafenone from the standard curve of propafenone hydrochloride reacted d i rect ly with heptafluorobutyric anhydride without extraction of the free base. 4.13 Linearity and Reproducibility of Assay Among the stereoselective assay methods for propafenone developed, that using R(-)NEIC as the chiral derivatizing reagent (Scheme 4) was found to be the most sensitive. The l inear i ty and precision of the assay were determined by analyzing plasma samples, which had been spiked with each enantiomer over the concentration range of 150 to 1000 ng/mL. The best f i t through the data points for each enantiomer was obtained from linear regression analysis (Table 3). The coeff ic ient of determinations, r^, were 0.9988 and 0.9960 for a l l of the data points for the R(-) and S(+) propafenone respectively. For inter-assay var iab i l i ty , the percent coeff ic ient of variation, which is a measure of reproducibil ity in sample analysis, was within 5% above 250 ng of each enantiomer. Below this quantity the percent coeff icient of var iab i l i ty increased to 8.6% for the R(-) isomer and 14.8% for the S(+) isomer of propafenone, indicating that at lower nanogram levels, the assay was not suf f ic ient ly reproducible. The intra-assay var iab i l i ty determined by t r ip l i ca te injections of each of the samples was within 5% except for one which was 10.4% (Table 4). 91 ANALYTICAL PROCEDURE PROPAFENONE.HC1 (250-2000ng) IN PLASMA (lmL) ADD DESIPRAMINE.HCl/Lilll5(I.S.), 10% TCA (0.4mL) SHAKEN ADD NaOH TO pH 12 AND BENZENE (6 mL) Tumbled for 2 0 min. followed by centrifuge for 10 min. at 2500 r.p.m. UPPER ORGANIC LAYER IS TRANSFERRED TO DRY TUBE AND ADD lgm OF ANHYDROUS SODIUM SULPHATE Tumbled for 5 min.; centrifuged and the Y l i q u i d i s transferred to another dry tube, and evaporated to dryness. ADD 100UL OF 0.01% v/v R(-)NEIC IN DRY ACETONITRILE PLUS 400UL OF DRY ACETONITRILE HEATED AT 7 5°C FOR 3 0 MINUTES COOLED, ADD luL OF PIPERIDINE, HEATED FOR 20 min. AT 7 5°C COOLED, EVAPORATED TO DRYNESS AND RECONSTITUTED WITH 2 00uL OF IPA Y INJECTED ONTO THE HPLC COLUMN Scheme 4. Assay Procedure for Propafenone Enantiomers by HPLC. Table 2. Efficiency of Recovery of Propafenone Enantiomers from Plasma. P e a k - H e i g h t R a t i o s * A D P e a k - H e x g h t R a t i o s Q u a n t i t y o f R ( - ) P P F R ( - ) P P F R e c o v e r y S(+) P P F S ( - ) P P F A fieCovery e n a n t i o m e r u n e x t r a c t e d e x t r a c t e d (%) u n e x t r a c t e d e x t r a c t e d (^) ( n g ) 500 0.436+0.0^ 0.337+0.03 77.13±1 0.426±0.34 0.334+0.2y 77.34+0.5 1000 0.979+0.04 0.768±0.03 78.45±0.52 0.927+0.36 0.733+0.02 79.05+2.1 • m e a n ± s t a n d a r d d e v i a t i o n , n=3 A PPF : Propafenone Table 3. Calibration Curve Data and Inter-Assay Var iab i l i ty of Propafenone Enantiomers in Plasma. W e i g h t o f e a c h e n a n t i o m e r ( n g ) P M k - H e i g h t R a t i o C .V . R ( - ) propafenone/ (%) i n t e r n a l s t a n d a r d A P e a k - H e i g h t R a t i o S ( + ) propafenone/ i n t e r n a l s t a n d a r d A C . V . • A 125 250 500 1000 0.419*0.036 0.738*0.036 1 .413*0.026 2.548*0.067 8.6 0.342*0.051 4.9 0.723*0.297 1.9 1.353* 0.0^9 2.6 2.427* 0.051 14.8 4.1 3.6 2. 1 S l o p e 1.2154 1.1775 I n t e r c e p t 0.1381 0.1076 r 2 0.9988 0.9960 * m e a n * s t a n d a r d d e v i a t i o n ; A A c o e f f i c i e n t o f v a r i a b i l i t y Table 4« Intra-Assay Var iab i l i ty of Propafenone Enantiomers in Plasma Weight of each Peak-Height Ratio C . V A A Peak-Height Ratio C.V. A A enaiatiomer(ng) R(-) propafenone/ (%) s(+) propafenone/ (