Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The actions of calcium antagonists on arrhythmias and other responses to myocardial ischaemia in the.. 1986

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
UBC_1986_A1 C87.pdf [ 16.75MB ]
UBC_1986_A1 C87.pdf
UBC_1986_A1 C87.pdf
Metadata
JSON: 1.0097215.json
JSON-LD: 1.0097215+ld.json
RDF/XML (Pretty): 1.0097215.xml
RDF/JSON: 1.0097215+rdf.json
Turtle: 1.0097215+rdf-turtle.txt
N-Triples: 1.0097215+rdf-ntriples.txt
Citation
1.0097215.ris

Full Text

c THE ACTIONS OF CALCIUM ANTAGONISTS ON ARRHYTHMIAS AND OTHER RESPONSES TO MYOCARDIAL ISCHAEMIA IN THE RAT By MICHAEL JOHN CURTIS B.Sc. (Hon.), Univers i ty of London (Chelsea College) 1979 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES Department of Pharmacology and Therapeutics, Faculty of Medicine We accept th is thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA Ju ly 1986 ®Michael John Cu r t i s , 1986 In p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t of the requirements f o r an advanced degree at the U n i v e r s i t y of B r i t i s h Columbia, I agree t h a t the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and study. I f u r t h e r agree t h a t p e r m i s s i o n f o r e x t e n s i v e copying of t h i s t h e s i s f o r s c h o l a r l y purposes may be granted by the head o f my department o r by h i s o r her r e p r e s e n t a t i v e s . I t i s understood t h a t copying or p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l gain s h a l l not be allowed without my w r i t t e n p e r m i s s i o n . The U n i v e r s i t y of B r i t i s h Columbia 1956 Main Mall Vancouver, Canada V6T 1Y3 0 DE-6 (2/81) - i i - ABSTRACT Studies were carr ied out in order to examine the actions of calcium antagonists in acute myocardial ischaemia and the mechanism(s) responsible for these act ions; the' s c i e n t i f i c hypothesis under test was that calcium antagonism in the vent r ic les i s antiarrhythmic in acute myocardial i s c h - aemia. In addi t ion, experiments were carr ied out to invest igate the ro le of the sympathetic nervous system in arrhythmogenesis. The actions of seven calcium antagonists on responses to myocardial ischaemia were investigated in vivo using the conscious rat preparat ion. I t was found that the drugs with i den t i f i ab le actions in the heart a t t r ibutab le to calcium antagonism possessed antiarrhythmic a c t i v i t y , whereas the drugs producing only systemic vasodi la tat ion were without tangible antiarrhythmic a c t i v i t y . The resu l ts were taken as evidence in support of the main hypoth- esis (above). In no instance did any of the drugs produce consistent dose- dependent - infarct-reducing act ions. It was establ ished from the comparison of the opt ica l enantiomers of verapamil that antiarrhythmic potency corresponded with calcium antagonist potency. I t was also shown that these drugs appeared to have no ef fect on g N a in the heart in vivo (as predicted from work by others, in v i t r o ) . Evidence that arrhythmias were reduced as a resu l t of ef fects on i . si in the ischaemic ven t r i c le was accrued from several s tud ies. In iso la ted Langendorff-perfused rat vent r ic les the calcium antagonist a c t i v i t y of the verapamil enantiomers was great ly potentiated by ra i s ing K + concentration to levels seen during acute myocardial ischaemia, whereas n i fed ip ine , which showed l i t t l e i f any antiarrhythmic a c t i v i t y in v-ivo did not show marked + K -dependent calcium antagonist a c t i v i t y . In a separate ser ies of experiments i t was demonstrated that se r i a l ablat ions in the CNS had profound ef fects on occlusion-induced arrhythmias, but that these ef fects occurred independently of the level of adrenoceptor ac t i va t i on . It was hypothesised that surgery reduced ischaemia-induced arrhythmias, by v i r tue of e i ther i t s ef fects on serum K + concentration or i t s ef fects on the number of c i r cu la t i ng thrombocytes. - iv - TABLE OF CONTENTS CHAPTER Page TITLE PAGE i ABSTRACT i i TABLE OF CONTENTS iv LIST OF TABLES x LIST OF FIGURES xi LIST OF ABBREVIATIONS xiv ACKNOWLEDGEMENTS XV AUTHORISATION FROM ANIMAL CARE UNIT, U. B. C. xvi 1 INTRODUCTION 1 1.1 C l i n i c a l myocardial ischaemia and in fa rc t ion 1 1.1.1 Overview 1 1.1.2 Infarct ion 3 1.1.3 Arrhythmias 4 1.1.4 Therapeutic approaches 14 1.2 Experimental myocardial ischaemia and in fa rc t ion 17 1.2.1 Overview 17 1.2.2 Methods for producing ischaemia 18 1.2.3 Differences between species 24 1.2.3.1 Infarct ion 24 1.2.3.2 Arrhythmias 26 1.2.3.3 Choice of species 28 1.2.4 A h is tory of coronary occlusion in rats 36 1.3 Ventr icu lar arrhythmias in acute myocardial ischaemia 46 1.3.1 The electrophysiology of exc i tab le t issue 47 1.3.2 The electrophysiology of the normal ven t r i c le 54 1.3.3 Elect rophysio log ica l changes caused by myocardial ischaemia 57 1.3.4 Models Of Arrhythmogenesis 60 1.3.4.1 Reentry 60 1.3.4.2 Abnormal Automaticity 62 1.3.4.3 Triggered Automaticity 66 1.3.5 Epicard ia l Act ivat ion Mapping 67 1.3.6 The slow inward current and arrhythmogenesis in acute ischaemia 70 1.3.6.1 Introduction 70 1.3.6.2 How might i s l - contr ibute to arrhythmogenesis? 70 1.4 The pharmacology of calcium antagonists 73 1.4.1 Def in i t ion 73 1.4.2 Pharmacology of phenethylalkylamines and 1,4-dihydropyridines 74 - V - 1.5 Aims of studies 77 1.5.1 The action of calcium antagonists in acute myocardial ischaemia 77 1.5.2 Arrhythmogenesis in acute myocardial ischaemia 78 2 METHODS 80 2.1 Coronary occlusion in rats 80 2.1.1 Overview 80 2.1.2 Preparation 81 2.1.2.1 Occluders 82 2.1.2.2 Leads and l ines 87 2.1.3 Coronary occlusion 91 2.1.3.1 Sequence of events 92 2.1.3.2 Monitoring of responses to occlusion 96 2.1.3.3 Occluded zone (OZ) 97 2.1.3.4 Infarct zone (IZ) 98 2.1.4 Def in i t ion of occlusion-induced arrhythmias 99 2.1.4.1 Introduction 99 2.1.4.2 Premature vent r icu lar contractions (PVC) 102 2.1.4.3 Ventr icu lar tachycardia (VT) 103 2.1.4.4 Ventr icular f i b r i l l a t i o n (VF) 103 2.1.4.5 Other arrhythmias 104 2.1.4.6 Rationale for d e f i b r i l l a t i o n 105 2.1.4.7 Arrhythmia scores 107 2.1.5 ECG changes produced by occlusion 110 2.1.5.1 ' S - T ' segment elevat ion 110 2.1.5.2 Pathological R-waves 114 2.1.5.3 Pathological Q waves 114 2.1.6 ECG changes produced by drugs 115 2.1.6.1 P-R interval 115 2.1.6.2 QRS interval 115 2.1.7 Measurement of serum K + concentration 116 2.1.8 Exclusion c r i t e r i a 117 2.1.9 S t a t i s t i c s 122 2.1.9.1 Normalisation procedures 122 2.1.9.2 Censoring 123 2.1.9.3 S t a t i s t i c a l tests 123 2.2 Calcium antagonist studies in coronary-occluded rats 124 2.2.1 General experimental design 124 2.2.2 Phenethylalkylamines 125 2.2.2.1 Anipamil and Ronipamil 125 2.2.2.2 (+)- And (-)-verapami1 126 2.2.3 1,4-Dihydropyridines 127 2.2.3.1 Felodipine 127 2.2.3.2 Nifedipine and DHM9 , 128 2.3 Arrhythmogenesis and the ro le of the CNS 129 2.3.1 Introduction 129 2.3.2 Preparation 130 2.3.2.1 P i th ing 130 2.3.2.2 Sp ina l i sa t ion 131 2.3.2.3 Decerebration 131 - v i - 2.3.3 Other manipulations 132 2.3.4 S t a t i s t i c s 135 2.4 Prel iminary Screen for drug a c t i v i t y in acute ischaemia 135 2.4.1 Introduction 135 2.4.2 Preparation 135 2.4.3 Experimental endpoints 136 2.4.3.1 Def in i t ions 136 2.4.3.2 Val idat ion of behaviour endpoints 137 2.4.4 Comprison of ( + )-, ( - ) - and (ij-verapami1 137 2.4.5 S t a t i s t i c s 139 2.5 E lec t r i ca l l y - induced arrhythmias in conscious rats 139 2.5.1 Introduction 139 2.5.2 Preparation 140 2.5.3 Experimental endpoints 140 2.5.3.1 Maximum fol lowing frequency 140 2.5.3.2 Threshold voltage 141 2.5.3.3 Threshold pulse width 141 2.5.4 Comparison of (+)- and (-)-verapamil 141 2.6 Haemodynamic ef fects of calcium antagonists in pithed rats 142 2.6.1 Introduction 142 2.6.2 Preparation 143 2.6.3 Variables measured 143 2.6.4 Comparison of (+)- and (-)-verapami1 143 2.6.5 S t a t i s t i c s 144 2.7 Actions of calcium antagonists in perfused ra t ven t r i c les 144 2.7.1 Introduction 144 2.7.2 Perfusion apparatus 145 2.7.3 Spec i f i ca t ions for Langendorff perfusion 146 2.7.4 Variables measured 146 2.7.4.1 Isochoric l e f t vent r icu lar developed pressure 146 2.7.4.2 Coronary blood flow 147 2.7.4.3 Ventr icu lar e x c i t a b i l i t y 147 2.7.5 Comparison of (+)- and (-)-verapami1 147 2.7.6 Comparison of n i fed ip ine and DHM9 148 2.7.7 S t a t i s t i c s 148 2.8 Metabolism of (^O-verapamil in acute myocardial ischaemia 149 2.8.1 Introduction 149 2.8.2 Animals 149 2.8.3 Drug administrat ion and t issue samples 149 2.8.4 Extract ion 150 2.8.5 HPLC 152 2.8.6 Calcu la t ion of (^J-verapami1 concentration 152 3 RESULTS 154 3.1 Metabolism of (^-verapami l in acute myocardial ischaemia 154 3.1.1 Concentration of (^-verapami l in blood matrices 154 3.1.2 Concentration of (±)-verapami1 in the vent r icu lar myocardium 155 - v i i - 3.2 Arrhythmogenesis and the ro le of the CNS 158 3.2.1 Overview 158 3.2.2 Occluded zones (OZ) 158 3.2.3 Arrhythmias 160 3.2.3.1 Arrhythmia scores 160 3.2.3.2 VT and VF 160 3.2.3.3 PVC 160 3.2.4 Haemodynamic var iables 165 3.2.5 ECG changes 165 3.2.6 Thrombocytes, leukocytes and serum K + 168 3.2.7 Summary 168 3.3 Actions of anipamil and ronipamil in acute myocardial ischaemia 170 3.3.1 Overview 170 3.3.2 OZ and in fa rc t zone (IZ) 170 3.3.3 Arrhythmias 171 3.3.4 Mor ta l i t y 175 3.3.5 Haemodynamic var iables 175 3.3.6 ECG changes 177 3.3.7 Plasma concentrations of anipamil 177 3.3.8 Summary 180 3.4 Actions of fe lod ip ine in acute myocardial ischaemia 180 3.4.1 Overview 180 3.4.2 OZ, IZ and morta l i ty 180 3.4.3 Arrhythmias 181 3.4.4 Haemodynamic var iables 186 3.4.5 ECG changes 186 3.4.6 Summary 186 3.5 Actions of verapamil enantiomers in acute myocardial ischaemia 189 3.5.1 Overview 189 3.5.2 Effects of enantiomers before occlusion 190 3.5.2.1 P-R and QRS in terva ls 190 3.5.2.2 Blood pressure and heart rate 190 3.5.3 OZ, IZ and morta l i ty 190 3.5.4 Arrhythmias 195 3.5.4.1 VT and VF 195 3.5.4.2 PVC 195 3.5.4.3 Arrhythmia scores 195 3.5.4.4 Arrhythmias at 24 h 203 3.5.5 Haemodynamic var iables 203 3.5.6 ECG changes 203 3.5.7 Summary 208 3.6 Actions of n i fed ip ine and DHM9 in acute myocardial ischaemia 208 3.6.1 Overview 208 3.6.2 Ef fects of enantiomers before occlusion 208 3.6.2.1 P-R and QRS in terva ls 208 3.6.2.2 Blood pressure and heart rate 210 3.6.3 OZ, IZ and morta l i ty 210 3.6.4 Arrhythmias 210 3.6.5 Haemodynamic var iables 216 - v i i i - 3.6.6 ECG changes 216 3.6.7 Serum K + concentration 218 3.6.8 Summary 218 3.7 Prel iminary Screen for drug a c t i v i t y in acute ischaemia 218 3.7.1 Overview 218 3.7.2 Val idat ion of behaviour endpoints 218 3.7.3 OZ and IZ 219 3.7.4 Mor ta l i t y and morbidity 219 3.7.5 Summary 223 3.8 E lec t r i ca l l y - i nduced arrhythmias in conscious rats 223 3.8.1 Overview and summary 223 3.9 Ef fects of verapamil enantiomers in pithed rats 225 3.9.1 Overview 225 3.9.2 S t a b i l i t y of the preparation 225 3.9.3 Ef fects of verapamil enantiomers 227 3.9.3.1 Blood pressure and heart rate 227 3.9.3.2 P-R and QT in terva ls 227 3.9.4 Summary 227 3.10 Actions of verapamil enantiomers in perfused rat ven t r i c les 232 3.10.1 Overview 232 3.10.2 Isochoric l e f t vent r icu lar developed pressure 233 3.10.3 Ventr icu lar e x c i t a b i l i t y 233 3.10.4 Addit ional comments 237 3.10.5 Summary 237 3.11 Actions of n i fed ip ine and DHM9 in iso la ted perfused vent r i c les 237 3.11.1 Overview 237 3.11.2 Isochoric l e f t vent r icu lar developed pressure 238 3.11.3 Ventr icular e x c i t a b i l i t y 240 3.11.4 Summary 240 4 DISCUSSION 243 4.1 The conscious rat preparation for myocardial ischaemia studies 243 4.1.1 Overview 243 4.1.2 Precis ion of var iables 243 4.1.2.1 Occluded zone and in fa rc t zone 243 4.1.2.2 Arrhythmias 245 4.1.2.3 ECG changes 245 4.1.3 Responses to drugs 246 4.1.3.1 Overview 246 4.1.3.2 Class 1 antiarrhythmics 247 4.1.3.3 Class 2 antiarrhythmics 248 4.1.3.4 Class 3 antiarrhythmics 250 4.1.3 .5 Class 4 antiarrhythmics 250 4.1.3.6 Arachidonic acid metabolites and an t iph log is t i cs 252 4.1.3.7 Other en t i t i es 253 4.1.4 The conscious rat versus other preparations 254 4.1.4.1 Advantages and disadvantages of rats 255 - ix - 4.2 Important determinants of arrhythmogenesis in acute ischaemia 256 4.2.1 Role of OZ 256 4.2.2 Role of the autonomic nervous system 258 4.2.3 Role of K + 261 4.2.4 Role of thrombocytes and leucocytes 263 4.2.5 Role of arachidonic acid metabolites 264 4.2.6 Role of l i p i d metabolites 264 4.2.7 Role of heart rate and blood pressure 265 4.2.8 Role of the fast and slow inward currents 265 4.3 Mechanism of action of calcium antagonists in acute ischaemia 268 4.3.1 Role of calcium antagonism 268 4.3.1 .1 Overview 268 4.3.1 .2 Anipamil versus ronipamil 269 4.3.1 .3 (+)- Versus (-)-verapamil 272 4.3.1 .4 Felod ip ine, n i fed ip ine and DHM9 278 4.3.2 Role of calcium antagonism in the myocardium 282 4.3.3 Role of calcium antagonism in the ischaemic ven t r i c le 289 4.3.4 Role of other pharmacological properties 293 4.3.4 .1 Inh ib i t ion of the fast inward current 293 4.3.4 .2 Blockade of a-adrenoceptors 294 4.3.4 .3 Indirect actions 296 4.4 General conclusions 297 4.4.1 Arrhythmogenesis in acute myocardial ischaemia 297 4.4.2 Action of calcium antagonists in acute myocardial ischaemia 298 5 REFERENCES 301 LIST OF TABLES TABLE 1 Summary of groups in the CNS ablat ion study. 2 Concentration of (^-verapami l in blood matr ices. 3 D is t r ibu t ion of (^J-verapami1 in the ven t r i c l es . 4 Haemodynamic ef fects of anipamil and ronipamil before and af ter occ lus ion. 5 Extent of ischaemia, i n f a r c t i on , and ECG changes after occ lus ion: ef fects of verapamil enantiomers. 6 Arrhythmias and morta l i ty fo l lowing coronary occ lus ion: ef fects of verapamil enantiomers. 7 Extent of ischaemia, i n f a r c t i on , and ECG changes after occ lus ion: ef fects of n i fed ip ine and DHM9. 8 Arrhythmias and morta l i ty fo l lowing coronary occ lus ion: ef fects of n i fed ip ine and DHM9. 9 Pre-drug values for haemodynamic and ECG var iables in pithed ra t s . 10 Incidence of PVC in iso la ted ven t r i c l es : ef fects of verapamil enantiomers. - XI - LIST OF FIGURES FIGURE Page 1 Sequence of events on the day of occlusion fo l lowing connection 93 of leads and l i n e s . 2 Anecdote of ECG changes caused by occlusion in a conscious ra t . I l l 3 Effects of ablat ions in the CNS on occluded zone. 159 4 Ef fects of ablat ions in the CNS on arrhythmia score. 161 5 Ef fects of ablat ions in the CNS on the incidence of vent r icu lar 162 f i b r i l l a t i o n . 6 Ef fects of ablat ions in the CNS on the incidence of vent r icu lar 163 tachycardia. 7 Ef fects of ablat ions in the CNS on the log^o number of PVC. 164 8 Ef fects of ablat ions in the CNS on heart rate and blood pressure 166 before and af ter coronary occ lus ion. 9 Ef fects of ablat ions in the CNS on ECG changes fo l lowing coronary 167 occ lus ion. 10 Changes in serum K + concentrat ion, leukocytes and thrombocytes 169 in pithed rats before and af ter coronary occ lus ion. 11 Ef fect of ronipamil and anipamil on arrhythmia score. 172 12 Ef fect of ronipamil and anipamil on the incidence and duration 173 of various arrhythmias during the f i r s t 4 h fo l lowing coronary occlusion 13 Ef fect of ronipamil and anipamil on occlusion-induced ECG changes 179 (R-wave and 'S -T ' segment e levat ion) . 14 Ef fects of fe lod ip ine on 0Z and IZ, and to ta l and arrhythmia- 182 induced morta l i ty during the 0 - 4 h period fo l lowing occ lus ion. 15 Ef fects of fe lod ip ine on the incidence of vent r icu lar f i b r i l l a t i o n 183 and tachycardia during the 0 - 3 0 min and 0 - 4 h periods fo l lowing occ lus ion . 16 Ef fects of fe lod ip ine on arrhythmia score for the 0 - 3 0 min and 184 0 - 4 h periods fo l lowing occ lus ion. 17 Ef fects of fe lod ip ine on log^o n u m D e r of PVC during the 0 - 3 0 185 min and 0 - 4 h periods fo l lowing occ lus ion. - X I 1 - 18 Ef fects of fe lod ip ine on blood-pressure and heart rate 1 min 187 before and 1 h after occ lus ion. 19 Ef fects of fe lod ip ine on maximum S-T elevat ion and the time at 188 which maximum S-T elevat ion occurred. 20 Changes in P-R and QRS in response to {*)- and (-)-verapami1. 191 21 The ef fects of ( ± ) - and (-)-verapamil on blood pressure and 193 heart rate at 1 min before occ lus ion. 22 The ef fects of (+)- and (-)-verapamil on the incidence of PVC 197 in re la t ion to the time after occ lus ion. 23 The ef fects of (+)- and (-)-verapami1 on the incidence of 199 vent r icu lar tachycardia in re la t ion to the time after occ lus ion. 24 The ef fects of (+)- and (-)-verapamil on the incidence of 200 vent r icu lar f i b r i l l a t i o n in re la t ion to the time after occ lus ion. 25 The ef fects of (+)- and (-)-verapami1 on arrhythmia score. 201 26 Mean percent changes in blood pressure and heart rate from pre- 205 drug values in response to (+)- and (-)-verapami1, before and af ter occ lus ion. 27 The ef fects of (± ) - and (-)-verapami1 on S-T segment elevat ion 206 at various times before and af ter occ lus ion. 28 The ef fects of (± ) - and (-)-verapamil on R-wave amplitude 207 at various times before and af ter occ lus ion. 29 The ef fects of n i fed ip ine , DHM9 and vehic le on P-R and QRS 211 in terva l before occ lus ion. 30 Mean percent changes in blood pressure and heart rate from pre- 212 drug values in response to DHM9 and n i fed ip ine , before and af ter occ lus ion. 31 Ef fects of n i fed ip ine and DHM9 on arrhythmia score. 213 32 Effects of n i fed ip ine and DHM9 on serum K + concentrat ion. 217 33 Ef fects of (+)-, (*)- and (-)-verapami1 on OZ and IZ in the 220 'prel iminary sc reen ' . 34 Ef fects of (+)-, (*)- and (-)-verapami1 on tota l morta l i ty 221 and morta l i ty associated with sudden convulsive-type behaviour during the 0 - 4 h period after occ lus ion, and morbidity during the 5 min period before occ lus ion . - x i i i - 35 Ef fects of (+)- and (-)-verapami1 on mean aor t ic blood pressure 222 heart rate and P-R in terva l in experiments to supplement the 'prel iminary sc reen ' . 36 Ef fects of (+)- and (-)-verapamil on threshold voltage and 224 threshold pulse width for induction of f i b r i l l o - f l u t t e r , and maximum fol lowing frequency in conscious rats subjected to e l e c t r i c a l st imulat ion of the l e f t ven t r i c l e . 37 Time-course of changes in mean aor t ic blood pressure and heart 226 rate in rats fo l lowing p i th ing . 38 Effects of (+)- and (-)-verapamil on mean aor t ic blood pressure 228 heart rate in pithed ra t s . 39 Effects of (+)- and (-)-verapami1 on Q-T and P-R in te rva ls in 229 pithed ra t s . 40 Dose-response curves for the ef fects of (-)- and (+)-verapami1 234 on developed pressure in Langendorff-perfused rat ven t r i c les paced at 300/min. 41 The re la t ionsh ip between negative inotropic potency of (-)- and 235 (+)-verapamil and K + concentration in iso lated Langendorff- perfused ra t ven t r i c l es . 42 The re la t ionsh ip between negative inotropic potency of n i fed ip ine 235 and K + concentration in iso la ted Langendorff-perfused rat ven t r i c l es . 43 The ef fect of d i f ferent buffer K + concentrations on the 241 threshold voltage and pulse width for capture of the l e f t ven t r i c le in the Langendorff-perfused rat ven t r i c le preparat ion. - X I V - LIST OF ABBREVIATIONS action potent ial AP act ion potent ial duration APD arrhythmia score AS calcium Ca conductance g conduction ve loc i ty e current i e f fec t i ve ref ractory period ERP hour(s) h in fa rc t zone IZ length constant x maximum d ias to l i c potent ial MDP membrane potent ial E m minute(s) min occluded zone OZ potassium K premature vent r icu lar contract ion PVC resistance r second(s) sec slow inward current i $ ^ sodium Na sodium current n*Na standard error of the mean s.e.mean time constant tau vent r icu la r f i b r i l l a t i o n VF vent r icu lar tachycardia VT - XV - ACKNOWLEDGEMENTS In order to ensure blindness in experimentation, the fo l lowing people k ind ly assisted with drug d i l u t i on and (occasional ly) with drug administra- t i on : Kathleen Saint , Tony Au and Michael Walker. Experiments c i ted in the text as having been carr ied out by others in the laboratory (measurement of c i r cu la t i ng p la te le ts and leukocytes) were carr ied out by Kathleen Saint . Dr R Wall i s thanked for his guidance with the HPLC experiments. The calcium antagonists used in the experiments were requested from and supplied by the fo l lowing: Dr B. Ljung of Haessle ( fe lod ip ine) ; Drs M. Raschack and R Kretzschmar of Knoll (opt ical enantiomers of verapamil, anipamil and ron ipami l ) ; Dr R. Whiting of Syntex (DHM9); Dr Kazda of Bayer (n i fed ip ine) . I am indebted to the Department of Pharmacology and Therapeutics, The Univers i ty of B r i t i sh Columbia and the B. C. Heart Foundation, a l l of whom provided me with f i nanc ia l support. I would l i k e to thank the fo l lowing people for the i r support and encouragement concerning the experiments which were carr ied out, and for the i r kindness and advice during my time in the department; Dr Bernard MacLeod, Dr Catherine Pang, Dr Morley Sut ter , Carol ine Bruce and Hari Nai r . I must also thank Tannis for putt ing up with my work being scattered across the kitchen tab le , and my bad temper. F i n a l l y , I am indebted to my supervisor, Dr Michael Walker, not only for his encouragement, for thr ightness, and for his example as a person searching for the truth (even in the face of advers i ty ) , but also for presenting me with a problem to so lve , a hypothesis to t es t , an establ ished preparation with which to begin my studies and an essen t ia l l y f ree hand to chose my own d i rec t i on . One could ask for nothing more. - xvi - AUTHORISATION FROM ANIMAL CARE UNIT, U. B. C. The experiments carr ied out in pa r t ia l fu l f i lment of the requirements for the degree of doctor of philosophy, and described in th is t hes i s , were approved as e th ica l by the Animal Care Unit of U. B. C. (reference number 511/W8). - 1 - 1 INTRODUCTION 1.1 C l i n i c a l myocardial-ischaemia-and in fa rc t ion 1.1.1 Overview Myocardial ischaemia is def ined, simply, as an impairment of coronary blood f low, and myocardial in fa rc t ion is defined as the necrot ic changes resu l t ing from myocardial ischaemia ( e . g . , Steadman's Medical Dict ionary, 1972). However, myocardial ischaemia and in farc t ion i s a heterogenous disease, varying from patient to patient in cause, l oca t ion , sever i t y , t ime- course and sequelae (Henderson, 1984; O l i ve r , 1982; Poole-Wilson, 1983; H i l l is and Braunwald, 1977; Maseri et a l 1 9 7 8 ) . I t i s general ly accepted that coronary blood flow insu f f i c iency i s the common event which produces the s igns, symptoms and sequelae of myocardial ischaemia and i n fa rc t i on . Ischaemia may involve par t ia l r e s t r i c t i o n or complete loss of blood f low, and may be permanent or temporary. Temporary ischaemia is associated with reper fus ion, which may occur gradually or abruptly (Fol ts et a l ; , 1982; Maseri et a l ; , 1978). Symptoms of acute myocardial ischaemia include angina pector is (chest pain) and d i zz iness , signs include syncope, vent r icu lar arrhythmias, hypotension, S-T segment a l te ra t ions , a reduction in cardiac output and leakage into the systemic c i r cu la t i on of creat ine phosphokinase, and sequelae include vent r icu lar arrhythmias, i n fa r c t i on , a prominant Q wave in chest lead ECGs, heart f a i l u r e , pulmonary oedema, and hypertension (Henderson, 1984; O l i ve r , 1982). The most common presentation of myocardial ischaemia is angina pecto- r i s . Angina of e f fo r t i s associated with rad iat ing chest pain and S-T seg- ment a l te ra t ions resu l t ing from an increase in myocardial oxygen demand in the set t ing of an inadequate oxygen supply, and r e l i e f may be obtained by reducing sympathetic dr ive to the heart with B-adrenoceptor antagonists, or by d i l a t i ng coronary vessels with organic n i t r i t e s (see Poole-Wilson, - 2 - 1983). Pr inzmetal 's variant form of angina (Prinzmetal e t - a l ; , 1959) may occur at rest in the absence of any apparent sympathetic nervous system-med- iated increase in myocardial oxygen demand. Angina of e f fo r t i s general ly associated with coronary a r te r i osc le ros i s^ whereas Pr inzmetal 's var iant i s not. The l a t t e r i s believed to be associated with coronary vasospasm (Hel lstrom, 1973; 1977), although th is i s not yet f i rm ly estab l ished. Ischaemic heart disease can often be at t r ibuted to coronary a r t e r i o - s c l e r o s i s . Stenosis (narrowing of vessels) leads to a loss of physio logical reserve, resu l t ing in i nsu f f i c i en t oxygen supply during periods of increased demand (angina pec to r i s ) . Sustained stenosis of su f f i c i en t sever i ty may lead to i n fa r c t i on . Coronary a r te r iosc le ros i s is pr imar i ly a large vessel disease (Gensini et a l , 1971), and morta l i ty resu l t ing from acute coronary occlusion is associated with s ingle vessel disease in 84% of cases (Liberthson et al 1982). Thus, the l e f t anter ior descending (LAD), l e f t circumflex and r ight coronary ar ter ies are the most common s i tes of coronary a r t e r i o s c l e r o s i s - i n - duced ischaemia and mor ta l i ty . In patients with occlusive stenosis in one or more coronary ar tery , the most important determinant of ischaemia is coronary artery anatomy, which var ies with age. Functional co l l a te ra l anastamoses are rare in young humans, but co l l a t e ra l s become larger and more numerous with age, perhaps in conjunction with (and as a consequence of) the development of a r te r i osc le ro - s i s (Baroldi and Scomazzoni, 1967; Fu l ton, 1965; Harr is et a l ; , 1969; Gensini and Bruto da Costa, 1972; Newman, 1981). It i s probable that slowly developing coronary s tenos is , from whatever cause, leads to a compensatory development of coronary co l l a te ra l anastamoses, as has been demonstrated in pigs (Schaper, 1971). - 3 - 1.1.2 Infarct ion Most of our knowledge of the in fa rc t ion process and i t s re la t ion to morbidity and surv ival has come from animal experimentation (see below). For th is reason, an exhaustive review of the in fa rc t ion process and i t s prognostic s ign i f icance w i l l not be given here. It i s c l ea r l y undesirable for working muscle t issue to d ie . If s u f f i - c ient myocardial muscle becomes necrot ic and f ibrous then the a b i l i t y of the heart to pump blood w i l l be impaired. Experimental evidence has shown that in fa rc t ion leads to myocyte hypertrophy in surv iv ing t i s sue , but that large in fa rc ts preclude complete funct ional recovery, owing to the surv ival of an i nsu f f i c i en t number of c e l l s ; hyperplasia does not occur (Anversa . e t - a l . , 1984; 1985a; 1985b; 1986). There are two important points which must be examined when in farc t s ize i s considered. F i r s t l y , a c lear d i s t i nc t i on must be drawn between preven- t ion and delay of necros is . It i s conceivable that necrosis could be delay- ed by a drug which causes vasodi la tat ion of co l l a te ra l anastamoses or by a drug which slows myocardial metabolism (reducing the rate of formation of cytotoxic products of anaerobic metabolism). Both types of drug would pre- vent in fa rc t ion provided that the stenosis were resolved (by thrombolysis or bypass surgery). I t i s unfortunate, therefore, that in most c l i n i c a l stud- ies aimed at reducing in fa rc t s ize no del ineat ion of the patient population into those with and those without co l l a te ra l anastamoses i s made, and l i t t l e attempt i s made to ra t i ona l i se the aims of the study in terms of the known pharmacology of the drug under inves t iga t ion . Some drugs have been invest igated for the i r a b i l i t y to reduce in fa rc t s ize in humans on the basis of the i r a b i l i t y to reduce myocardial oxygen consumption or oxygen demand (heart ra te , force of contract ion, e t c . ) . However, i f occlusion is complete, and co l l a te ra l c i r cu la t i on is minimal or - 4 - absent, then no amount of o f f - loading w i l l salvage the ischaemic myocar- dium. Attempting to prevent in fa rc t ion by such a strategy is analogous to attempting to survive in a vacuum by holding one's breath. I t should also be noted that most of the reports in the l i t e ra tu re con- cerning c l i n i c a l in farc t s ize refer to enzymatical ly determined in fa rc t s i z e . There are a var ie ty of techniques for measuring in fa rc t s ize by det- ermining the concentration in the blood of enzymes which are normally only found i n t r a c e l l u l a r ^ , for example creat ine phosphokinase (CPK). I t has been shown that in experimental animals the peak leve ls of serum CPK c o r r e l - ate with in fa rc t s ize determined h i s t o l o g i c a l l y fo l lowing s a c r i f i c e (Shel l et a l ; , 1971). However, i t i s not c lear whether th is re la t ionsh ip s t i l l holds under the inf luence of drugs. It is quite possible that a drug may inf luence CPK without a l te r ing in fa rc t s i z e . The problem c l i n i c a l l y , of course, is that one cannot examine a heart h i s t o l o g i c a l l y unless the patient d ies . Therefore one is dependent in most cases on ind i rec t measures of in fa rc t s i z e . There are angiographic techniques for measuring coronary 133 blood flow in v ivo , e .g . using radioact ive contrast imaging with xenon ( P i t t et a l . , 1969), and ultrasound techniques for imaging an in fa rc t (e .g . Mattrey and Mi t ten, 1984), but these techniques are not normally used in large scale mult icentre c l i n i c a l t r i a l s . 1.1.3 Arrhythmias C l i n i c a l l y , the major type of death in associat ion with myocardial ischaemia i s 'sudden' . Sudden death has been defined as death occurring within 1 h of the patient l as t being seen a l i v e , and has been suggested to occur in approximately 30% of the to ta l patient population with coronary artery disease (Armstrong et a l ; , 1972). Sudden death i s the major cause of death in pat ients with myocardial ischaemia. I t i s estimated that approxi- mately 400,000 such deaths occur per year in the USA alone (Lown, 1982). - 5 - The cause of sudden death is usual ly uncerta in. I t has been shown that approximately 43%of sudden deaths occurring during the 4 weeks fo l lowing the onset of chest pain occur during the f i r s t , hour (Armstrong e t -a l . - , 1972). In add i t ion , i t has been estimated that less than 10% of pat ients dying during the f i r s t hour af ter the onset of chest pain are seen by a physician (Ol iver , 1982). In other words, the majority of patients who die as a resu l t of myocardial ischaemia do so without a de f in i te establishment of the cause of death. In the fol low-up reports of the Framingham study of the epidemiology of sudden death, the fo l lowing comments were made: 'The assignment of sudden deaths to coronary aet iology i s la rge ly by inference, since few other diseases can k i l l in a matter of minutes. Coronary ae t i o lo - gy i s assumed when aor t ic d i ssec t ion , ruptured aneurysm, and pulmonary embo- l ism are excluded c l i n i c a l l y or on postmortem examination. Thus, death within minutes in persons not i l l at the time with a po ten t ia l l y le thal i l l n e s s permits c l a s s i f i c a t i o n as coronary sudden deaths with reasonable cer ta in ty ' (Kannel e t - a l ; , 1984). The cause of sudden death is general ly at t r ibuted to VF (Ol iver , 1972; Campbell, 1983; 1984; e t c . ) . Many c l i n i c i a n s use the terms sudden death and fa ta l VF interchangeably, ( e . g . , O l i ve r , 1982). There are several persuasive reasons in support of t h i s , despite the absence of d i rec t evidence in most cases. F i r s t l y , evidence from work using experimental animals (see exper i - mental sect ion of Introduction) c lea r l y demonstrates that VF occurs during the f i r s t hour af ter coronary occ lus ion, and that th is is the major cause of death at th is time (cardiogenic shock and pulmonary oedema being r e l a t i v e l y rare during the f i r s t hour af ter occ lus ion) . Secondly, i f VF i s the major cause of sudden death then a reduction in ear ly VF should lead to a corres- ponding reduction in sudden death. With regard to th is point , i t was found, in 1956, (Zol l e t - a l . ) that VF in humans could be reverted by applying a DC - 6 - shock to the chest. This procedure was r igourously applied in Seatt le by paramedics in order to d e f i b r i l l ate patients before admission to hosp i ta l , and the resu l t was a reduction in out -of -hospi ta l sudden death of approxim- ate ly 55% (Cobb et a l ; , 1980), confirming that VF and sudden death were re la ted . In add i t ion , an invest igat ion of the ECGs of patients co l laps ing outside of hospital and receiv ing paramedic care within 15 min of col lapse has provided a strong suggestion that sudden death and VF are one and the same; of 426 pat ients , 72 % were in VF when the ECG was f i r s t recorded, and of the remainder, only 1 % were not experiencing vent r icu la r arrhythmias (Liberthson et a l ; , 1982). However, i t i s important to consider that although sudden death is probably caused by VF, there i s no establ ished c r i t e r i o n , in the absence of ECG evidence, for categor is ing a death as hav- ing resulted from VF, and that most sudden death occurs out of hospital (Campbell, 1984) in the absence of ECG monitoring. I r respect ive of the exact incidence of fa ta l VF in c l i n i c a l myocardial ischaemia and i t s contr ibut ion to the body of 'sudden death ' , i t neverthe- less remains that VF can occur during myocardial ischaemia in humans, and that th is event is often fa ta l ( e . g . , Ju l ian e t - a l 1 9 6 4 ) . The natural h is tory of vent r icu lar arrhythmias during myocardial ischaemia and in farc t ion is not well establ ished in humans. This is under- standable since many patients die before admission to hosp i ta l . There is very l i t t l e information concerning the very ear ly phase ( f i r s t few minutes) of acute myocardial ischaemia in humans. Campbell e t - a l . (1981) evaluated 38 previously unmedicated patients admitted to hospital with 'acute myocar- d ia l i n f a r c t i o n ' , and expressed arrhythmia incidence in re la t ion to the onset of the symptoms (not def ined, presumably chest pa in) . I t was found that primary VF (defined as VF occurring in the absence of ' shock ' , heart f a i l u r e or heart block) occurred p r i nc i pa l l y within the f i r s t 4 h af ter the - 7 - onset of symptoms. VF was almost always associated with an i n i t i a l 'R on T' premature vent r icu lar contraction (defined as a QRS complex f a l l i n g within 85% of the prevai l ing QT interval of a normal sinus QRS complex), whereas vent r icu lar tachycardia (VT) was almost never associated with such an occur- rence. VF and R on T premature vent r icu lar contract ions (PVCs) were almost absent af ter 4 h. In contrast , non R on T PVCs and VT (defined as 3 or more consecutive PVCs at a rate of 120/min or more) increased in frequency bet- ween 4 and 12 h af ter the onset of symptoms. It i s of in terest that despite the uncertainty concerning the exact onset of occ lus ion , the degree of s ten- o s i s , the volume of ischaemia, e t c . , in th is study (Campbell et a l ; , 1981), the time d is t r ibu t ion of arrhythmias was not grossly d i f ferent from that reported for dogs (Harr is , 1950), and rats (Johnston et a l . , 1983a). Adgey et a l ; (1971) also examined the natural h is tory of arrhythmias occurring during the f i r s t few hours af ter the onset of symptoms. They examined 284 patients with ECG evidence of 'acute myocardial i n f a r c t i on ' (S-T segment changes or bundle branch block) and found a 31 % incidence of bradyarrhythmia, 25% incidence of PVCs, 10% incidence of VF, 4% incidence of a t r i a l f i b r i l l a t i o n and 0.4% incidence of supraventr icular tachycardia dur- ing the f i r s t hour af ter the onset of symptoms ( i t must be noted that the above does not include the incidence of fa ta l vent r icu lar arrhythmias, which were not analysed). Corresponding incidences during the 3rd and 4th hour were 2, 6, 0.7, 2, 0 and 0%, respect ive ly . The decl ine with time in the incidence of VF corresponds with that reported by Campbell et a l ; (1981). Adgey e t - a l ; (1971) also reported that the incidence of VF 'a f te r 4 h' was low (4%). Since the p o s s i b i l i t y of a second episode of myocardial ischaemia was not explored, i t i s possible that the true incidence of VF occurring af ter 4 h as a resu l t of the i n i t i a l episode of myocardial ischaemia was less than 4%. It is a problem of most c l i n i c a l studies that there i s gener- - 8 - a l l y l i t t l e examination of the time course of ischaemia in re la t ion to the p o s s i b i l i t y of further c r i t i c a l stenosis and occlusion higher up the a r t e r i - osc le ro t i c t ree . Recent c l i n i c a l studies of arrhythmias associated with myocardial ischaemia and in fa rc t ion have concentrated to a cer ta in extent on the desire to predict sudden death/ fatal VF. This is par t ly a resu l t of the hesitance in prescr ib ing antiarrhythmic agents to a l l patients at r i sk of developing myocardial ischaemia and i n fa r c t i on , owing to the occurrence of serious side ef fects with long-term use ( e . g . , Kosowsky et a l ; , 1973; Jel ineck e t a l ; , 1974; Campbell, 1983). In addi t ion, there are no drugs which have been shown to unequivocally reduce c l i n i c a l l i fe - th rea ten ing arrhythmias asso- ciated with myocardial ischaemia and in farc t ion (Campbell, 1983; 1984). Therefore there is not su f f i c i en t j u s t i f i c a t i o n , in terms of cost versus benef i t , for whole-sale prescr ip t ion of a par t i cu la r drug in pat ients at r i sk of developing myocardial ischaemia and i n fa rc t i on . As a resu l t of these considerat ions, corre la t ions between benign arrhythmias and l i fe - th rea ten ing arrhythmias have been sought, under the assumption that a spec i f i c patient population 'at high r i s k 1 of sudden death can be del ineated. In 1967, Lown et a l . proposed that cer ta in arrhythmias (for example, R on T PVCs) const i tuted 'warning arrhythmias' . A r e l a t i o n - ship between R on T PVCs and sudden death had been speculated e a r l i e r (Smirk and Palmer, 1960). For many years, therapeutic decisions were based on the detection of these warning arrhythmias. However, there were contradictory reports ( e . g . , E l -Sher i f e t a l . , 1976; Rabkin et a l ; , 1982) which suggested that 'warning arrhythmias' were just as common in patients not subsequently developing VF as in those who d i d . The la t te r suggestion has been supported by a recent study in which i t was found that a high incidence of complex PVCs ( including R on T) and VT in otherwise healthy subjects was not assoc i - - 9 - ated with sudden death during a 10 year period (Kennedy et a l . - , 1985). The natural h is tory study of Campbell (Campbell et_al±, 1981) showed that whi ls t R on T PVCs were extremely common in patients during the 10 min preceeding VF, they were almost as common in patients not developing VF. Campbell concluded that 'warning arrhythmias' were of no pred ic t ive value for th is reason. In addi t ion, i t i s worth considering that any 'warning' which does not appear un t i l 10 min before a l i fe - th reaten ing event i s of l i t t l e use as a guide to therapy, espec ia l l y out of hosp i ta l . Part of the confusion concerning 'warning arrhythmias' stems from a fundamental phi losphical dichotomy. This concerns the model of arrhythmo- genesis in myocardial ischaemia and in fa rc t ion to which an invest igator subscribes. S p e c i f i c a l l y , i s ischaemia more or less important than i n fa rc - t ion in arrhythmogenesis? In experimental preparations, VF occurs before t issue has become in fa rc - ted (see experimental sect ion in Introduct ion). Therefore, when considering myocardial ischaemia and i n fa r c t i on , the pr inc ipa l prerequis i te for VF would appear to be myocardial ischaemia. In support of th is premise derived from experimental animal studies are the fo l lowing pieces of c l i n i c a l evidence. F i r s t l y , i t has been suggested that most patients who die from sudden death do not have myocardial i n fa rc - t ion when examined postmortem (Lovegrove and Thompson, 1978), ind icat ing that sudden death may have resulted from myocardial ischaemia, but not myo- card ia l i n fa r c t i on . Secondly, i t has been reported that less than 20% of patients resusci tated from VF exhib i t ECG signs of in fa rc t ion such as Q waves in V leads (Cobb e t - a l ; , 1980). This has lead some c l i n i c i a n s (e .g . O l i ver , 1982) to voice the opinion that c l i n i c a l l y , VF i s associated with ischaemia, whereas e l e c t r i c a l i n s t a b i l i t y i s not a feature of in farcted t i ssue , since i t is not v iab le . - 10 - Proponents of the 'warning arrhythmia 1 hypothesis take a d i f ferent view. The ra t iona le for 'warning arrhythmias' i s based on the premise that ' a d i s t i n c t i v e myocardial pathophysiologic derangement long preexists in vict ims of sudden death ' , and that 'because such deaths are the resu l t of VF, i t i s not i l l o g i c a l to assume the abnormality to be e lectrophysio logic and the r i sk indicator to re la te to al tered cardiac rhythm' (Lown, 1982). This philosophy is not supported by the evidence described above. Lown did not include in his scheme any recognit ion of the duration of symptoms (whether the patient is experiencing ischaemia or i n fa r c t i on ) , nor did he examine the re la t i ve importance of re - in fa rc t i on versus re- ischaemia. Lown dismisses the evidence which suggests that 'warning arrhythmias' are of no value in predict ing VF (see Campbell, 1983), and comments that 'the pre- va i l i ng pessimism concerning the p o s s i b i l i t y of protect ing a patient who has experienced VF against recurrence i s unwarranted' (Lown, 1982). Lown may be overstat ing his case, because whereas Campbell considers the ent i re popu- la t ion of myocardial ischaemia and in fa rc t ion pat ients , Lown's studies were car r ied out on a selected subset of myocardial ischaemia/ infarct ion pat ients , namely those patients 'who have experienced malignant vent r icu lar arrhythmias' , and the prognostic c r i t e r i a which were developed therefore only apply to these pat ients . As discussed in deta i l above, th is subset of patients (those who have experienced complex VT) are the very patients who are least l i k e l y to subsequently experience VF (Campbell e t - a l ; , 1981). These are the patients who have e i ther survived the VF occurring during the f i r s t 3 - 4 h of myocardial ischaemia, or passed through th is danger period without experiencing VF. This group, therefore, does not include any of the t ru l y a t - r i sk pat ients . Indeed, pat ients who die of sudden death before receiv ing medical attent ion were not considered in th is a r t i c l e (Lown, 1982). Therefore the statement that ' abo l i t i on of advanced grades of ven- - 11 - t r i c u l a r premature beats prevents the recurrence of po ten t ia l l y le thal arrhythmias' (Lown, 1982) does not apply to myocardial ischaemia/ infarct ion patients as a whole. Campbell, re fer r ing to a l l ischaemia/ infarct ion pat ients , stated that whi ls t many studies have demonstrated that cer ta in drugs reduce R on T PVCs and complex vent r icu lar contract ions, 'no tangible benef i t of suppressing these arrhythmias has been detec ted ' , (Campbell, 1984). Campbell also stated that 'no current strategy for vent r icu lar arrhythmia prophylaxis or treatment in acute myocardial in fa rc t ion i s s a t i s - factory ' (Campbell, 1983). The f a i l u r e to d is t inguish between the various stages and condit ions of myocardial ischaemia and i n fa r c t i on , pa r t i cu la r l y between the i n i t i a l occlusion and the subsequent course of the disease(s) may be responsib le, in part , for the confusion over both prognosis and the assessment of a n t i - arrhythmics. It has been suggested that improvements must be made in th is regard in re la t ion to the design of c l i n i c a l t r i a l s , or 'the treatment of these patients w i l l continue to be made from a posi t ion of abject ignorance' (Bigger, 1984). In summary, there i s no evidence that 'warning arrhythmias' usefu l l y predict sudden death/VF, pa r t i cu la r l y in pat ients who have passed through the f i r s t few hours of myocardial ischaemia. The 'warning arrhythmia' concept is contingent upon d i f ferent arrhyth- mias being part of a continuum. There i s abundant evidence from studies with experimental animals (see experimental sections) which suggests that PVCs, VT and VF do const i tute a continuum, e i ther in non-ischaemic prepa- rat ions (Dresel and Sutter , 1961) or in experimental myocardial ischaemia. Our laboratory has shown that some antiarrhythmics can reduce the incidence of VF without af fect ing PVCs in rats subjected to coronary artery occlusion (Johnston et a l ; , 1983a). With regard to var ia t ions in the natural h is tory of the disease, in par t i cu la r , the incidence of re-ischaemia and r e - i n f a r c t i o n , there is no reason to bel ieve that the arrhythmias resu l t ing from a f i r s t episode of myocardial ischaemia predict in any way the outcome of a subsequent episode of ischaemia. Therefore, in terms of therapy, i t would be f o l l y to withold an e f fec t ive prophylaxis against VF ( i f such a drug were shown to ex is t ) on the basis of an absence of VF (or R on T PVCs) during a f i r s t episode of myocardial ischaemia in one pat ient , while administering that drug to a second patient with a h is tory of VF (or R on T PVCs) during myocardial ischaemia. There are strategies which have been used c l i n i c a l l y to determine which drug should be prescribed in order to prevent subsequent sudden death in patients who have experienced myocardial i n fa rc t i on , based on responses to drugs. An example is programmed e l e c t r i c a l st imulat ion of the r igh t ven- t r i c l e (Fisher et a l ; , 1977). The technique i s based on the p r inc ip le that i f a premature stimulus is del ivered during the terminal phase of r e p o l a r i - sat ion (the 'vulnerable per iod ' ) then VT or f i b r i l l a t i o n may be e l i c i t e d (de Boer, 1921; Wiggers and Wegria, 1940). Essen t i a l l y , the threshold (current or pulse width) for induction of VT or VF is determined before and af ter the administrat ion of various drugs, with the aim of f ind ing a drug which com- p le te ly suppresses the induction of the arrhythmia upon s t imula t ion. I t has been claimed that a pos i t ive response to a drug predicts that treatment with that drug w i l l prevent spontaneous arrhythmias in 90 - 95 % of patients dur- ing the fo l lowing 1 - 2 years (Ruskin et a l • , 1983). Nevertheless, i t should be recognised that 1 5 - 6 0 % of patients in the studies reviewed by Ruskin who did not respond to any drug during e l e c t r i c a l st imulat ion went on to experience no VT or VF during the fol lowing 1 - 2 years, while receiv ing apparently ' i n e f f e c t i v e ' therapy according to the resu l ts of the st imulat ion tes ts . Therefore the 90-95% 'success' rate (Ruskin et a l . , 1983) is perhaps - 13 - misleading. The problem with th is method i s that although i t may be reason- ably e f fec t i ve in demonstrating that a par t icu lar drug may prevent ven t r i cu - la r arrhythmias occurring in the f i r s t 2 years fo l lowing myocardial ischae- mia and i n fa rc t i on , i t does not benef i t the vast majority of pat ients who suf fer sudden death without warning and in whom no programmed st imulat ion has been undertaken. A lso , despite the claimed success of such methods, i t remains that there are no e f fec t i ve agents for preventing sudden death (Campbell, 1983; 1984), and that enormous numbers of people s t i l l die from sudden death/VF. I f one accepts the premise that the arrhythmias occurring during the f i r s t few hours of myocardial ischaemia (Campbell e t - a l ; , 1981; Adgey e t - a l ; , 1971) are the major cause of death in patients with myocardial ischaemia and in farc t ion (Ol iver , 1982), what are the charac te r i s t i cs and importance of the arrhythmias which begin approximately 4 h af ter the onset of symptoms (Campbell et a l . - , 1981)? Essen t i a l l y , la te arrhythmias have only been of in terest c l i n i c a l l y in terms of whether they predict subsequent sudden death (see foregoing paragraphs) or pose a s ign i f i can t health r isk in themselves. There are no c l i n i c a l studies which have attempted to estab l ish the re la t ionsh ip between late vent r icu lar arrhythmias and the onset of ischaemia, the extent of ischaemic muscle mass, the degree of stenosis (residual blood f low) , the frequency and extent of re-ischaemia or the r e l a - t i ve importance of ischaemia versus i n fa rc t i on . The possible ro le of arrhythmias occurring more than 24 h af ter the onset of symptoms in the genesis of sudden death has been discussed, and i t appears that there i s no prognostic value in the analysis of such arrhythmias (Campbell, 1984). In add i t ion , the poor corre la t ion between the i n i t i a l episode of ischaemia and la te arrhythmias is exemplif ied by the observation that although the i n c i - dence of PVCs and VT occurring during the f i r s t 10 h after the onset of - 14 - chest pain correlated with in fa rc t s ize (determined by measuring serum CPK), the rate of PVC measured 1 - 1 0 months af ter the onset of chest pain bore no re la t ion to in farc t s ize in the same patients (Roberts, et a l ; , 1975). This implies that i t i s possible that the so-ca l led la te arrhythmias in humans may be a mixture of arrhythmias associated with an old in farc t (generated by reentry around the i n fa rc t , perhaps; see section on arrhythmogenesis in Introduct ion), arrhythmias associated with pers is t ing par t ia l ischaemia and arrhythmias associated with new bouts of ischaemia. In summary, c l i n i c a l information concerning the aet iology of sudden death and i t s re la t ionsh ip with VF and other vent r icu lar arrhythmias is incomplete, owing to the fact that the majority of patients dying suddenly do so out of hosp i ta l . C l i n i c a l information concerning the natural h is tory of arrhythmias and the i r underlying causes in the min, h, days, weeks and months fo l lowing the onset of symptoms is confusing, perhaps as a resu l t of the desire to estab l ish an ideal protocol for treatment without having to tediously order and c l a s s i f y each condit ion and i t s cha rac te r i s t i cs . How- ever, such an approach i s necessary in order to remove the confusion concer- ning mechanisms of arrhythmogenesis and the contentious aspects of progno- s i s . It i s c l ea r , however, that the current c l i n i c a l approach has not pro- vided any e f fec t i ve prophylaxis against sudden death (Campbell, 1983; 1984; Furberg, 1983b). 1.1.4 Therapeutic approaches Theore t ica l ly the primary therapeutic aim would be to prevent myocardial ischaemia from occurr ing. The Framingham study (Kannel e t a l . - , 1984) suggested that in previously healthy humans, the incidence of sudden death increases with increasing age, t r i p l i n g in males from 2 to 6 per 1000 per- sons between the age groups 45 - 54 and 65 - 74 years. In addi t ion, the frequency in males i s approximately 2 - 3 times that in females. Age and - 15 - sex are not t reatable condit ions at present. However, the Framingham study also iden t i f i ed hypertension, diabetes, c igaret te smoking, obesity and the ra t io of low versus high density l ipoprote in serum cholesterol as important indicators of coronary artery disease and sudden death. While cor re la t ion does not prove cause-and-effect, i t i s general ly considered that a healthy l i f e s t y l e , a balanced d ie t and exercise are associated with a low incidence of coronary artery disease and sudden death. A safe e f fec t i ve prophylaxis against coronary occlusion would be des i r - able. In th i s regard, the resu l ts of a long term t r i a l of asp i r in in appa- rent ly healthy volunteers is expected to be published reasonably soon. In theory, thrombolysis w i l l only be of benefi t in pat ients with e f fec t i ve co l l a te ra l vascu lar isat ion or par t ia l occ lus ion, in whom ischaemia is not complete ( in whom the delay between the onset of symptoms and the i n i t i a t i o n of thrombolysis i s not as c r i t i c a l as i t would be expected to be in patients with complete occlusion in the absence of funct ioning co l l a te ra l anastamo- ses) . Since th is has not been thoroughly invest igated in the many c l i n i c a l studies of thrombolysis (Yusuf et a l ; , 1985), then i t is not possible to comment further in th is regard. Since coronary artery disease remains a s ign i f i can t health r i s k , then i t is expedient to l im i t the sequelae, namely arrhythmias and i n fa rc t i on . There are at least two ways in which attempts can be made to reduce VF and other arrhythmias. F i r s t l y , automatic d e f i b r i l l a t o r s may be implanted. This has been found to be a successful approach (Echt et a l . - , 1985), but since surgery i s required, i t i s desirable to ident i fy patients most at r i sk of sudden death; the implantation of d e f i b r i l l a t o r s is not a pract ice which can be carr ied out on a very large scale at present. In add i t ion , i t has recent ly been suggested that automatic d e f i b r i l l a t o r s can reduce the qua l i ty of l i f e by v i r tue of the anxiety and fear associated with the i r use (Cooper - 16 - et a l ; , 1985). A l te rna t i ve l y , pharmacological prophylaxis could be admini- s tered. Such a strategy would be expected to protect the pat ient without undue inconvenience in terms of side e f f ec t s , inconvenient dosing regimen and frequent fo l low-up. However, a drug su i tab le for th is purpose does not e x i s t , owing to a lack of proven benefi t on one hand, or unacceptable s ide-e f fec ts on the other ( e . g . , Campbell et a l ; , 1984). Along with the search for a cure for cancers, the search for a safe e f fec t i ve agent which may be used in a large number of subjects who might be categorised as at r i sk of coronary artery disease would appear to be the branch of drug research with the greatest potent ial c l i n i c a l impact in terms of prolonging l i f e . In addit ion to arrhythmia prevention, attempts should be made to l im i t i n fa r c t i on . While i t would c l ea r l y be benef ic ia l to prevent in fa rc t ion resu l t ing from coronary occ lus ion, th is is not a r e a l i s t i c goal at present, par t ly because therapy i s usual ly not i n i t i a t e d un t i l hours af ter the onset of symptoms. Experimental studies in animals without e f fec t i ve co l l a te ra l anastomoses have i l l u s t r a t ed that 15 min of ischaemia w i l l lead to i r reve r - s i b l e myocardial c e l l death, even i f complete reperfusion is achieved (Hort and Da Canal i s , 1965b). While unequivocal c l i n i c a l information i s lack ing , i t i s reasonable to suppose that in the absence of extensive co l l a te ra l vascu la r i sa t ion , much the same w i l l occur in humans. Theore t i ca l l y , any measure which reduces in fa rc t s ize i s desirable in order to reduce the l i ke l ihood of "cardiac output f a i l u r e (Maroko e t - a l ; , 1971). However, unless an agent i s given as prophylaxis, i t i s d i f f i c u l t to imagine how i t might prevent i n f a r c t i on , unless i t possesses the capab i l i t y of e i ther converting f ibrous t issue into muscle, st imulat ing myocardial m i tos is , or converting dead c e l l s into l i v i n g c e l l s . At present, no such drugs e x i s t , and prevention of myocardial in farc t ion with a view to l im i t i ng - 17 - pump f a i l u r e and cardiogenic shock (Agress e t a l ; , 1952) has not been demon- strated e i ther c l i n i c a l l y or experimentally (Reimer and Jennings, 1985). However, i t might be hoped that a fas t -ac t ing myocardioplastic growth factor might one day be developed. 1.2 Experimental-myocardial ischaemia and in fa rc t ion 1.2.1 Overview For an animal model of any human disease to be i d e a l , i t should exhib i t the fo l lowing cha rac te r i s t i cs . I t should: a . completely mimic at least one aspect of the disease b. respond in the same manner as humans to drugs c. have the precis ion and accuracy of a good bioassay d. allow various responses to be measured e. be simple and cost l i t t l e in terms of apparatus, t ime, and exper- t i s e . There is no model of myocardial ischaemia and in fa rc t ion which dup l i - cates the human condi t ion. This was the conclusion of the Coronary Heart Disease Task Group Panel Report of 1973 (see Winbury, 1975). In 1978, the Workshop on More Uniform Animal Models and Protocols for Assessment of Interventions to Protect Ischemic Myocardium held at the NIH recommended the study of 3 coronary occlusion preparations, the rat for ' r e l a t i v e l y inexpen- sive and rapid screening of po ten t ia l l y useful t reatments' , the anaesthe- t ised dog ' fo r v e r i f i c a t i o n of ef fect iveness in a more physio logical s i t u - a t ion ' and the conscious dog ' fo r test ing the most promising agents under the most physio logical condi t ions ' (see Reimer et a l ; , 1985). The l a t t e r recommendations may perhaps appear pecul iar when factors other than 'physio- log ica l relevance' are considered (see sect ion dealing with choice of spec ies) . In addi t ion, i t has been suggested that 'no animal heart i s t ru l y comparable to that of man' (Sasyniuk and Na t te l , 1982) implying that r e l a - - 18 - t i ve rather than absolute physio logical relevance i s the issue in question, and as such should not be the primary consideration in the choice of exper i - mental species. 1.2.2 Methods for producing ischaemia Ar te r iosc le ros i s may be induced with a high cholesterol d ie t . However, with the possible exceptions of the hypercholesterolaemic hare (Pearson et a l ; , 1983), mini-p ig (Jacobsson, 1984), and quai l (Cheung et - a l ; , 1983) th is method is not considered to be spec i f i c for the coronary vesse ls . In most inves t iga t ions , generalised stenosis is produced, therefore th is approach i s ra re ly used (Winbury, 1964). Hypercholesterolaemia-induced coronary a r te r iosc le ros i s models are also compromised owing to the necessity that the endpoint (development of in fa rc t ion or arrhythmias) must occur wi thin a prescribed time-frame, or experimentation becomes unacceptably inconvenient. It may be possible to circumvent th is problem by increasing oxygen demand by a t r i a l pacing in the set t ing of a compromised oxygen supply induced by cholesterol feeding, but th is method produces only reve rs ib le , pacing-dependent ischaemia (demonstrated by S-T segment e levat ion) ; arrhyth- mias and in fa rc t ion do not occur (Lee and Baky, 1973). There are several techniques for producing small vessel occlusion and d i f fuse myocardial ischaemia. By in jec t ing starch suspensions (Roos and Smith, 1948), p l as t i c microspheres (Weber et a l . , 1972) or lycopodium spores (Guzman et a l . , 1962), i t i s possible to produce a r te r i o l a r occlusion lead- ing to generalised and d i f fuse myocardial i n f a r c t s . This technique was f i r s t t r i ed using powder, wax, o i l and ink, more than 100 years ago (Panum, 1862, see Tillmanns e t a l ; , 1975). There are many disadvantages of these techniques. F i r s t l y , in order to preclude generalised systemic occ lus ion, i t i s necessary to in jec t the material into a coronary ar tery. This necess- i ta tes e i ther open-chest experimentation, or techn ica l l y demanding coronary catheter isat ion techniques. Secondly, most animals die from cardiac output f a i l u r e (Weber et a l ; , 1972), which is not the major cause of death in c l i n - i ca l myocardial ischaemia (Campbell, 1983; 1984) or in other models of myo- card ia l ischaemia. Th i rd ly , a r t e r i o l a r occlusion and the resul tant d i f fuse ischaemia and in fa rc t ion d i f fe rs extensively from c l i n i c a l occlusion-induced ischaemia which general ly resu l ts from a r te r iosc le ros i s of large coronary ar ter ies (Gensini et a l ; , 1971). While t h i s , in i t s e l f , i s not necessar i ly a disadvantage, the d i f fuse character of the experimental in fa rc t would be expected to be d i f f i c u l t to quant i fy , thereby making i t d i f f i c u l t to deter- mine the ef fect of a drug on in farc t s i z e . In add i t ion , and perhaps more importantly, models of arrhythmogenesis (see elsewhere) are contingent upon the presence of a well defined focus of ischaemia which must be of suf f - i c ien t volume to serve as a substrate for conduction delays and the genera- t ion of in jury currents (Gettes, 1974; Janse, 1982; Kleber e t - a l . , 1978; e t c . ) . These models of arrhythmogenesis would not apply to the d i f fuse ischaemia preparations. Indeed, the d i f fuse ischaemia preparations do not appear to be associated with vent r icu lar arrhythmias at a l l (see Winbury, 1975). There are a var ie ty of techniques for producing occlusion of a major coronary artery by embol isat ion. The 2 major approaches are d i rect produc- t ion of an occlusive embolus and the induction of an occlusive thrombus. Direct embolisation may be induced by in jec t ing mercury (Lluch e t - a l ; , 1969), placing a s ta in less steel cy l inder (Nakhjavan et a l ; , 1968), steel ba l l bearing (R ib ie l ima, 1964) or detachable catheter t i p (Hammer and P i s a , 1962) into a coronary ar tery , or i n f l a t i ng an intracoronary balloon (Corday et a l . , 1974). In add i t ion , i t has been demonstrated that i f a cy l i nd r i ca l magnet i s placed around a coronary ar tery , and small (4 um diameter) iron par t i c les administered, the par t i c les w i l l be captured by the magnet and - 20 - occlude the vessel (Elz inga et - a l , 1969). These techniques are a l l capable of producing complete occ lus ion, although th is must be ve r i f i ed by measuring coronary blood f low. Par t i a l occlusion may be brought about v ia d i rect embolisation by inser t ing a cyl inder of lead f o i l into a coronary ar tery; the cy l inder w i l l pass down the artery for a distance governed by i t s outer circumference, while the degree of stenosis is governed by the inner circum- ference (Johnsrude and Goodrich, 1969). This technique i n i t i a l l y produces a known degree of s tenos is , but the device is thrombogenic, and serves to induce secondary progressive s tenos is . Deliberate thrombogenic techniques include e lec t r i ca l l y - i nduced thrombo- genesis, and placement of a thrombogenic foreign object in a coronary ar t - ery. Placement of an electrode in a coronary artery lumen and another on the chest wall can produce thrombi i f current i s passed between them (Sala- zar , 1961). This study reported d i f fuse thrombi d i s ta l to the intraluminal e lectrode. However, Weiss (1971) managed to produce complete occlusion of the circumflex or LAD artery using s im i la r techniques, and used the prepar- ation to evaluate antiarrhythmic drugs. The e l e c t r i c a l production of throm- bi is used today by Lucchesi for evaluating antiarrhythmic and an t i - i n f a r c t agents in dogs ( e . g . , Patterson et a l • , 1981; 1983). Foreign-body-induced thrombi may be produced by inser t ing thrombogenic objects into coronary vesse ls , for example, magnesium a l l oy or copper hel ices (Kordenat e t - a U , 1972); the resul tant thrombus is general ly located at the s i t e of the i n - se r t i on . The most common means of producing myocardial ischaemia in use today is occlusion of a major coronary artery by cons t r i c t i on . There are a var ie ty of techniques for producing th is end (see below). Occlusion may be gradual or abrupt. Gradual coronary occlusion may be brought about by the use of aneroids - 21 - (Berman et a l ; , 1956), which are r ings of hygroscopic material which swell and progressively reduce the lumen of the artery which they have been placed around. Rings of ge lat in impregnated with diacety l phosphate induce the formation of granulomatous t issue which w i l l gradual ly cons t r i c t the en- c i r c l e d vessel (Asada e t a l . - , 1962), in a manner analogous with aneroids. A s im i la r resu l t can be produced by the use of pneumatic cuffs (Khouri and Gregg, 1967; Bond et a l . - , 1973) or hydraul ic cuffs (Khouri et a l . , 1968; Hood et a l ; , 1970). A major l im i ta t ion of these techniques is the requi re- ment for the placement of a coronary flow-probe in order to estab l ish the extent of occlusion and the moment of complete occ lus ion. A lso , the prob- lems associated with the a r te r iosc le ros i s models ( in addit ion to the problem of generalised systemic a r te r iosc le ros is ) also apply to the gradual coronary occlusion models, namely that production of ischaemia, arrhythmias and in fa rc t ion does not take place within a convenient time period for exper i - mentation. Abrupt, complete coronary artery occlusion is perhaps the simplest and most ra t iona l technique for invest igat ing myocardial ischaemia. This is because two var iables which are d i f f i c u l t to measure and more d i f f i c u l t s t i l l to regulate are removed from the experimental arena. These var iables are the time course of occlusion and the absolute blood f low. The time course of occlusion i s known only for abrupt occlusion techniques, whi ls t bloodflow in an occluded vessel is only known i f occlusion is complete (flow in p a r t i a l l y occluded vessels must be measured d i r e c t l y ) . The major v a r i - able in abrupt, complete occlusion preparations is therefore the extent of co l l a te ra l blood f low. Abrupt occlusion may be produced by clamping an artery with screw clamps (Gregg e t a l -., 1939), bulldog clamps or Goldblat t clamps (Jennings et a l ; , 1960). In small animals (such as rats) with small coronary ar te r ies i t i s possible to cauter ise a major vessel e l e c t r i c a l l y - 22 - (Staab et a l . - , 1977; Prum e t -a l . - , 1984). More simply, a vessel may be 1igated. Coronary l i ga t ion was f i r s t undertaken by Chirac in 1698, who observed a loss of heart movement as a consequence (see Tillmanns e t - a l . , 1975). Interest in the dependence of heart a c t i v i t y on the coronary c i r cu la t i on resumed in the 19th century, when Erichsen (1842) determined the duration of occlusion necessary for producing vent r icu lar s t a n d s t i l l in dogs. Subse- quently, Cohnheim (1881) l igated dog coronary ar ter ies and developed the hypothesis that the coronary c i r cu la t i on i s comprised mainly of end ar ter ies (see Tillmanns e t -a l . - , 1975 for t rans la t ion and d iscuss ion) . In the 20th century, l i ga t ion of a coronary artery together with ECG recording was f i r s t carr ied out in 1918 (Smith), but was rare ly carr ied out again un t i l 1935 (Johnston e t - a l : ) . The reason for the apparent lack of in terest in myocar- d ia l ischaemia in the ear ly part of th is century stems from the general be l ie f at that time that coronary occlusion is a un iversa l ly fa ta l event (see Fye, 1985 for review). In most l i ga t ion techniques a simple s i l k l igature i s used. To gain access to the designated artery and t ighten the l i g a t i o n , experiments were i n i t i a l l y carr ied out using open-chest dogs (Townshend Por ter , 1894). However, for long term experiments on mor ta l i ty , chests were subsequently closed and animals allowed to regain consciousness ( e . g . , Smith, 1918; Le Roy et a l ; , 1942). Most ear ly experiments involved 1-stage coronary l i g a - t i o n . However, an adaptation of a technique designed for par t ia l l i ga t ion of a coronary artery was used by Harris to l iga te coronary ar ter ies in 2 stages (Harr is , 1950). Harr is had observed that abrupt occlusion of the l e f t anter ior descending coronary artery produced VF in approximately 50 % of dogs (Harr is , 1948) within the f i r s t 10 min after l i g a t i o n . Animals s u r v i - ving th is i nsu l t experienced few arrhythmias un t i l approximately 4.5 h af ter l i g a t i o n , whereupon PVCs ensued, increasing in in tens i ty with time; by 8 h, frequent VT developed, pers is t ing for a further 2 - 4 days. However, i t was reasoned that i f the arrhythmias occurring during the f i r s t 10 min of occ lu - sion could be circumvented then the frequency of surv ival would be i n - creased, al lowing for a more deta i led and spec i f i c invest igat ion of the second phase (4.5 - 8 h) and th i rd phase (8 h to 4 days) of arrhythmias. Therefore, an i n i t i a l par t ia l l i ga t i on was undertaken by including a needle in the loop of the l i ga tu re , then removing the needle once the l igature had been t ightened; the lumen of the ar tery was therefore stenosed to the diame- ter of the needle. After 30 min or 1 h, a second l iga ture was tightened at the posi t ion of the f i r s t to f u l l y occlude the vesse l . The resu l t was the el iminat ion of phase-1 arrhythmias (those occurring during the f i r s t 10 min af ter complete occlusion) without change in phase-2 and phase-3 arrhyth- mias. I t i s of in terest to note that th is much-quoted work of Harr is (1950) was in fact pre-empted, to a cer ta in extent, by much e a r l i e r work (Michae- l i s , 1894), in which i t was demonstrated that i f small coronary artery branches were t ied before the larger ar ter ies then cardiac s t ands t i l l was prevented. A s ign i f i can t advance was made by the development of a noose-type device for 'atraumatic ' occ lus ion. Fischer and Edwards (1963) threaded a small polythene tube under a coronary artery and then passed both ends through a larger polythene tube, such that t rac t ion between the tubes would occlude the vesse l . Rushmer e t - a l ; , (1963) used the same p r i n c i p l e , subst i tu t ing a nylon suture for the small polythene tube, in order to produce coronary occlusion in conscious dogs. This technique allowed arrhythmias and i n - fa rc t ion to be invest igated in the absence of anaesthetic and acute surg ica l preparation for the f i r s t t ime. Unfortunately, perhaps, l i t t l e use of th is technique was made for almost 20 years. Coronary occlusion in conscious animals was not invest igated extensively un t i l our laboratory developed a method for coronary occlusion in conscious rats (Au et a l ; , 1979a), using an occluder s im i la r to that used by Rushmer. The use of a 'pre-prepared 1 animal with a loose occluder implanted around a coronary artery and exter- io r ised through the skin for induction of occlusion in the absence of anaesthetic and recent surgery may be an extremely important advance, since recent work has suggested that much of the information provided by i n v e s t i - gations into ischaemia-induced arrhythmogenesis in anaesthetised, acutely prepared animals is misleading (see sect ion dealing with arrhythmogenesis in Discussion). 1.2.3 Differences between species 1.2.3.1 In fa rc t ion . Infarct s ize reduction has been t r a d i t i o n a l l y invest igated in dogs, and therefore the majority of the l i t e ra tu re concer- ning myocardial in fa rc t ion re fers to th is species. However, the v a r i a b i l i t y of in fa rc t s ize in dogs has been suggested to make th is species essen t ia l l y useless for quant i tat ive assessment of in fa rc t s ize and i t s modi f icat ion. This fact was recognised as long ago as 1918, and was at t r ibuted to the presence of large and varied co l l a te ra l anastamoses (Smith, 1918). For th is reason, Johns and Olson (1954) developed the rat preparation for assessing i n fa r c t i on , and concluded that any species with undeveloped co l l a t e ra l vascu lar isat ion (rats and mice) would be sui tab le for evaluating i n f a r c t i on , whereas species with var iable coronary vasculature (hamsters and dogs) and species with extensive co l l a te ra l anastomoses (guinea-pigs) would be unsui t - able for study. Perhaps su rp r i s ing l y , th is work was completely ignored (see sect ion deal ing with coronary occlusion in r a t s ) , and studies of myocardial ischaemia and in farc t ion continued to be carr ied out (almost exc lus ive ly ) using dogs. Indeed, i t was work using dogs which led Braunwald's group (Maroko e t - a l . , 1971) to propose the concept of myocardial salvage. I t was - 25 - suggested that i t should be possible to reduce the ult imate s ize of an in fa rc t by measures 'designed for reduction of myocardial oxygen demands and improvement of coronary per fus ion 1 . It i s perhaps s i gn i f i can t that these invest igators (Maroko et a l • , 1971) did not in fact demonstrate that any manipulation (ouabain, glucagon, propranolol , haemorrhage, methoxamine or isoproterenol) reduced in fa rc t s ize as such. The index of in fa rc t ion used was S-T segment e leva t ion , measured from DC electrograms recorded from the surface of the heart. In only 2 groups was in fa rc t s ize (at 24 h) recorded, and evidence that in fa rc t s i ze was al tered was not convincing. The many experimental and c l i n i c a l studies carr ied out over the fo l low- ing years have not confirmed the hypothesis of Maroko et - al.- (1971) that in fa rc t s ize can be reduced in a po ten t ia l l y c l i n i c a l l y useful manner by drug treatment, under the condit ion of i r reve rs ib le coronary occ lus ion. Although i t is possible to delay both the ECG signs of ischaemia and the development of i n fa r c t i on , the evidence does not suggest that any treatment can prevent death in non-perfused t issue (see Reimer and Jennings, 1984). Cer ta in ly , no treatment has yet been claimed to have prevented i n fa rc t i on . The var iab le nature of the outcome of coronary occlusion in dogs is exemplif ied by a study carr ied out by Sobel and associates (Shell et a l ; , 1971). These authors developed the method of quant i f i ca t ion of in fa rc t s ize based on serum levels of CPK. They found that peak serum CPK correlated l i nea r l y with in fa rc t s ize determined at 24 h by measuring the myocardial content of CPK. However, the actual value of in fa rc t s ize in 22 dogs at 24 h was highly va r iab le , 21.5 ± 18 (mean ± s .d . ) as % vent r icu lar weight, the range of values being 1-55%. These values are typ ica l of published in fa rc t s ize in dogs. For example, Burmeister and Reynolds (1983) reported that the coe f f i c ien t of var ia t ion ( s . d . as a%of the mean) was 23% in mon- grel dogs and 73% in beagles, while Miyazaki et-al . - (1984) reported that - 26 - in fa rc t s i ze varied from 0.6 to 46 %of to ta l vent r icu lar weight, with a 60% coe f f i c ien t of var ia t ion in mongrel dogs. With v a r i a b i l i t y such as t h i s , i t i s d i f f i c u l t to imagine how so many reports of 'myocardial salvage' have been generated and continue to be generated ( e . g . , Jugdutt, 1985; Bednar et a l . , 1985; Tumas et a l . , 1985). Perhaps the many reports of in fa rc t s ize reductions in dogs are the resu l t of sampling er ror , or measurement of 'u l t imate ' i n fa rc t s ize too soon af ter occ lus ion. 1.2.3.2 Arrhythmias. In conscious ra t s , severe vent r icu lar arrhythmias f i r s t occur in the period 4 - 2 0 min af ter occ lus ion , and a second major phase appears af ter 1 h and las ts for 2 - 6 h (Clark e t - a l ; , 1980; Johnston e t - a l ; , 1983a). In dogs, ischaemia-induced arrhythmias were o r i g i n a l l y described as ear ly , occurring during the f i r s t hour, and l a t e , peaking at approximately 24 h af ter coronary occlusion (Harr is , 1950). However, recent ly the ear ly (phase-1) arrhythmias have been subdivided into phase- la , occurring during the f i r s t 1 - 3 min of ischaemia, and phase-lb, occurr ing 5 - 2 0 min after coronary occlusion (Haase and S c h i l l e r , 1969; Meesman, 1982). Phase-la arrhythmias, which usual ly comprise only of PVC, have only been reported in anaesthetised dogs; information concerning the i r occurrence in conscious dogs i s not avai lab le at present. However, the type of anaesthetic used appears to inf luence the occurrence of phase-la arrhyth- mias, in that they are far more frequent in dogs anaesthetised with pento- barbitone than dogs anaesthetised with morphine-chloralose-urethane or ni trous oxide (Meesman, 1982). In add i t ion , phase-la arrhythmias have been reported to occur in pentobarbitone anaesthetised rats (Fagbemi, 1984), but not conscious rats (Johnston et a l . , 1983a). Phase-la arrhythmias have also been reported in pentobarbitone anaesthetised pigs (Bergey e t - a l . , 1982); i t i s not establ ished whether such arrhythmias occur in conscious p igs. - 27 - In dogs 24 h af ter occ lus ion, episodes of VT and mul t i focal PVC are commonly observed (Smith, 1918; Har r i s , 1950). In our laboratory, we have ra re ly observed VT and never observed VF in rats at 24 h, although almost a l l 24 h survivors exhib i t mul t i focal PVCs. The reason for th is di f ference between rats and dogs has not been examined, however i t i s possible to speculate that the presence or absence of co l l a te ra l anastamoses may govern the sever i ty of 24 h arrhythmias. In th is regard, i t has been suggested that abnormal automaticity in p a r t i a l l y ishaemic Purkinje f ib res (maintained by co l l a te ra l perfusion) i s responsible for 24 h arrhythmias in dogs (F r ied - man e t a l ; , 1973; Lazzara e t a l ; , 1973). This idea i s consistent with the mul t i focal nature of these arrhythmias. In r a t s , the re la t i ve lack of funct ional co l l a te ra l anastamoses compared with dogs (Maxwell e t - a l ; , 1984; Winkler e t a l ; , 1984) may not permit surv ival of su f f i c i en t Purkinje t issue to t r igger VT and VF 24 h after occ lus ion. In conscious ra t s , arrhythmias occurr ing during the f i r s t 4 hours af ter occlusion have been described in deta i l on the basis of many experiments in our laboratory (Johnston et a l . , 1983a). Ventr icular arrhythmias are ex- tremely common and include PVC, VT and VF. Sinus bradycardia and a t r i o - vent r icu lar blocks are much less common. Very occas iona l l y , a t r i a l arrhyth- mias ( f i b r i l l o f l u t t e r ) and supraventr icular tachycardia are seen. Spontaneous reversion of VT during acute myocardial ischaemia i s common, both experimentally and c l i n i c a l l y . Spontaneous reversion of VF has been reported in humans (Robinson and Bredeck, 1917; Maseri et - a l . - , 1982), but the incidence i s considered to be low. However, there is l i t t l e object ive evidence to substantiate th i s be l i e f . In conscious rats the incidence of VF i s high (approximately 90 %), and the incidence of spontaneous d e f i b r i l l - ation i s correspondingly high (approximately 60 %). However, the incidence of sustained VF is even higher (approximately 90 %), according to exper i - - 28 - merits from our laboratory (Johnston e t - a l . , 1983a). In dogs, there have been occasional reports of spontaneous reversion of VF ( e . g . , Smith, 1918; Gibson e t - a l . , 1986). In p igs , spontaneous reversion of VF has been observ- ed (unpublished observations from our laboratory) , although th is event i s ra re . In the absence of any evidence that spontaneously revert ing ven t r i c - ular arrhythmias d i f f e r fundamentally in o r ig in maintenance or mechanism from sustained vent r icu lar arrhythmias, there is no reason to bel ieve that spontaneous reversion of VF i s an experimental disadvantage. On the con- t ra ry , our laboratory has shown that spontaneous d e f i b r i l l a t i o n permits the experimental animal to surv ive, al lowing for an increase in the y i e l d of information from each preparation (Johnston e t -a l . - , 1983a). 1.2.3.3 Choice- of -species. The sect ion dealing with c l i n i c a l myocardial ischaemia and in fa rc t ion suggested that most deaths in patients with myocardial ischaemia occur from VF, and that the incidence of VF dec- l ines exponent ial ly with time fo l lowing the i n i t i a l onset of chest pain (Campbell e t - a l ; , 1981). Therefore i t i s of great importance to develop treatments for arrhythmias associated with acute myocardial ischaemia. It has been suggested that a drug which can be taken as prophylaxis by pat ients at high r i sk of myocardial ischaemia to prevent VF would have far more im- pact on morta l i ty than therapy i n i t i a t e d after admission to hospi ta l (Campbell, 1984). This implies that arrhythmias occurr ing during the f i r s t few hours, and pa r t i cu la r l y during the f i r s t few minutes of myocardial ischaemia are the arrhythmias which must be invest igated in th i s regard. C l i n i c a l l y , these acute arrhythmias are general ly missed; a patient has e i ther died from them, been resusci tated by paramedics, or , perhaps, recov- ered spontaneously by the time they are admitted to hosp i ta l . This leads to a dilemma. One may ei ther invest igate ear ly arrhythmias in experimental animals (phase-1 according to Har r i s , 1950) or late (phase-3) arrhythmias. - 29 - Any experimental preparation in which phase-1 arrhythmias are measured cannot be evaluated for i t s ' c l i n i c a l re levance ' , because there is insuf f - i c i en t information concerning these arrhythmias and the i r suscep t i b i l i t y to drugs in humans to provide a useful template. Therefore i t i s desirable to use several animal species for such s tud ies . A l te rna t i ve l y , one may concentrate on phase-3 arrhythmias, under the assumption that drugs which reduce phase-3 arrhythmias w i l l also reduce phase-1 arrhythmias ( in other words that common mechanisms of arrhythmogen- esis operate for phase-1 and phase-3 arrhythmias). Essen t i a l l y , therefore, one must estab l ish one of the two fol lowing c r i t e r i a (depending on which experimental strategy is chosen) in order to j u s t i f y the choice of model and species. F i r s t of a l l , i f i t i s decided to invest igate phase-1 arrhythmias then the only concern should be the l o g i s t i c charac te r i s t i cs of the model. I t must be establ ished that the model provides accurate and precise information (any debate concerning the c l i n i c a l relevance of such a model is redundant). A l t e rna t i ve l y , i f phase-3 arrhythmias are to be studied as a model of phase-1 arrhythmias then i t fol lows that a drug which i s found to i nh ib i t phase-3 arrhythmias should also i nh ib i t phase-1 arrhythmias. Therefore, i t is of absolute necessity to estab l ish that a l l drugs which are e f fec t i ve or ine f fec t ive against phase-3 arrhythmias in the chosen preparation possess the same p r o f i l e of a c t i v i t y against phase-1 arrhythmias in that same prepa- ra t i on . The majority of published reports in which such comparisons have been carr ied out do sa t i s f y th is c r i t e r i o n . It fo l lows, therefore, that the evaluation of drugs for potent ial c l i n i c a l use as prophylaxis against phase- 1 arrhythmias must not be based upon an extrapolat ion from a c t i v i t y against phase-3 arrhythmias. Such a strategy i s analogous to invest igat ing i n te r - ventions for preventing f i r e s in the home by measuring the ef fect of the - 30 - experimental intervention on room temperature. I t i s , however, permissible to invest igate phase-3 arrhythmias for the i r own sake. In th is regard, c l i n i c a l relevance may be tested d i r e c t l y , because phase-3 arrhythmias are easy to study c l i n i c a l l y . However, whether phase-1 or phase-3 arrhythmias are studied may have important c l i n i c a l rami f ica t ions . With further reference to the analogy of the study of the prevention of f i r e s in the home, i f acute out-of -hospi ta l VF is analogous to a f i r e , i t may also be that c l i n i c a l phase-3 arrhythmias are merely analogous to an increase in room temperature. In other words, while the prevention of out of hospital VF (phase-1 arrhythmias) w i l l have enormous consequences in terms of l i ves saved, the prevention of phase-3 arrhythmias may have l i t t l e therapeutic impact. A l l the techniques for producing myocardial ischaemia, with the excep- t ion of e lectrocautery, were developed using dogs. Unt i l 1985, dogs were eas i l y the most common species used in studies of myocardial ischaemia and i n fa rc t i on . Recently, the rat has become popular. Popular i ty , however, does not necessar i ly imply s u i t a b i l i t y . There is no concensus at present concerning the ideal choice of species for invest igat ing myocardial ischae- mia and i n fa r c t i on . I t has been argued that there is no ideal species for such s tud ies, and that the choice depends upon the objects and goals of an invest igat ion (Fozzard, 1975; Harken et a l . , 1981). In contrast with animal models of other diseases, such as bacter ia l i n fec t i on , in which a high degree of s i m i l a r i t y between the animal t issue and human t issue is des i rab le , i t i s possible to argue that animals with hearts s im i la r to the human heart are paradoxical ly unsuitable for studies of myocardial ischaemia and i n fa rc t i on . There are several unrelated reasons for th is viewpoint. F i r s t l y , an animal with a heart which resembles the healthy, young human heart in terms of co l l a te ra l vascu lar isat ion is by de f in i t i on a poor substrate for invest igat ions concerned with the behaviour of the chron ica l l y s ick human heart with extensive atheromae and co l l a te ra l development. Such an animal ought, however, to provide a good model of the young human heart . It has been suggested that young humans are more suscep- t i b l e to fa ta l VF than older humans (Morgan Jones, 1969), and i t has been argued that th is is a resu l t of the re la t i ve lack of co l l a te ra l anastamoses in the younger human (Ol iver , 1982). Pig hearts lack co l l a te ra l anastamoses and have consequently been suggested to resemble young human hearts (Fedor et a l . , 1978; Verdouw et a l . , 1983). The problem in th is instance i s that while such an animal may in theory provide a good model of myocardial ischaemia in otherwise healthy young humans, there i s very l i t t l e in fo r - mation concerning the charac ter is t i cs of th is c l i n i c a l s i t ua t i on . Moreover, there is essen t i a l l y no information concerning the effect iveness of drug therapy in such pat ients . Furthermore, un t i l the improbable eventual i ty of a large scale c l i n i c a l study of the effect iveness of a range of drugs in preventing the signs and sequelae of acute myocardial ischaemia in young healthy humans, then such information w i l l remain unavai lable. In such a circumstance, any argument that an animal with few e f fec t i ve co l l a te ra l anastamoses provides a good model of acute myocardial ischaemia in healthy young humans i s merely hypothet ica l . At the other end of the spectrum, i t has been suggested that a species with extensive or varied co l l a te ra l vascu lar isat ion resembles the mature or e lder ly human heart . The mature human heart may develop co l l a t e ra l anasta- moses in response to the slow developmemt of a r te r i osc le ros i s (Schaper, 1971). In dogs the development of co l l a te ra l anastamoses over a period of weeks fo l lowing coronary artery l i ga t i on may represent such a process (Eckstein e t - a l ; , 1941). Amongst common experimental species, dog hearts usual ly have we l l -deve l - - 32 - oped co l l a te ra l anastamoses and may resemble the mature or e lder ly human heart (Schaper, 1971; 1979). C lea r l y , i f we assume that a l l other var iables are equal , then a species with an extensive coronary co l l a te ra l c i r cu la t i on is preferred when myocardial ischaemia in the mature human ( in whom myocar- d ia l ischaemia occurs more frequently than younger humans) is to be modeled. However, there are other considerat ions. One important considerat ion in any s c i e n t i f i c invest igat ion is the pr ior knowledge of the power of the experiment. I f one is interested in the ef fect of a treatment on a Gaussian d is t r ibu ted var iab le then the sample s i z e , n, i s dependent on two indepen- dent va r iab les , the ef fect iveness of the treatment and the var iance. If two species d i f f e r in terms of the variance of a par t i cu la r va r iab le , then provided there are no other confounding fac to rs , the preferred species w i l l be the one with the lower variance. Consideration of which of the 2 species is the most c lose ly re lated to humans becomes of minor importance. There i s a tendency, nevertheless, for aesthet ic rather than l o g i s t i c considerations to d ic tate the choice of species in myocardial ischaemia and in fa rc t ion s tud ies. The dog, un t i l recent ly , has been the most popular species for i n v e s t i - gating myocardial ischaemia and i n fa rc t i on . For example, a large general review of models of coronary artery disease published in 1975 is almost exc lus ive ly devoted to dog preparations (Winbury, 1975). However, there is an h is tory of evidence which warns that var ia t ion in the coronary anatomy of dogs jeopardises myocardial ischaemia studies by const i tu t ing a large source of variance ( e . g . , Smith, 1918). Paradoxica l ly , the dog has been favoured for th i s very reason; the var iable nature of the coronary anatomy of the dog i s q u a l i t a t i v e l y s im i la r to that of the human (see above). However, the question a r i ses : i s i t desirable to use an animal myocardial ischaemia preparation in which the coronary anatomy pa ra l l e l s the human coronary - 33 - anatomy? In th is regard i t is important to consider the sample s ize necess- ary for a c l i n i c a l study of the effect iveness of a drug in reducing mortal- i t y occurring during acute myocardial ischaemia. The problem in th is instance is that the required information is not ava i l ab le , since the acute phase of myocardial ischaemia occurs out of hospital in most cases, and has resolved or has lead to death (see preceeding sec t ions) . Nevertheless, with regard to the la te r ' pos t - i n fa r c t i on ' per iod, in order to show a s t a t i s t i c - a l l y s i gn i f i can t reduction of morta l i ty of 10 - 20% (from approximately 5 - 15% during the f i r s t year) one requires a study sample of several thou- sands (May, 1983a; 1983b). It i s therefore undesirable to use a preparation which c lose ly mimics the human s i t ua t i on . As Meesman has c l ea r l y demonstra- ted, the major source of variance in dogs with regard to the outcome of myocardial ischaemia is the extent of co l l a te ra l vascu lar iza t ion (Meesman, 1982). Co l la te ra l flow correlates not only with arrhythmias (Meesman, 1982) but also with in fa rc t s i z e . I t has recent ly been shown that in fa rc t s ize in dogs can be predicted on the basis of the extent of co l l a te ra l vascu la r i - sation independently of the s ize of the occluded bed ( ' r i sk zone') (Yel lon e t - a l ; , 1986b). In addi t ion, a recent study showed that whereas simple occlusion of the LAD produced only minor vent r icu lar arrhythmias, LAD occ lu - sion coupled with the obstruct ion of retrograde co l l a te ra l blood flow to the occluded bed resul ted in severe vent r icu lar arrhythmias in 75% dogs (Shoji e t - a l ; , 1986). Therefore, a species with l i t t l e e f fec t i ve co l l a te ra l perfu- s ion , productive of consistent and predictable regions of ischaemia upon coronary occ lus ion, would be more appropriate than the dog, in terms of the precis ion and accuracy for a given to sample s i z e . The two species in which a well defined and reproducible area of ischae- mia fo l lowing coronary occlusion has been documented are the pig and the ra t . Coronary co l l a te ra l anastamoses connecting to the l e f t anter ior - 34 - descending coronary artery in the pig are almost absent (Schaper, 1971), leading to a sharp demarcation between the normal and the ischaemic zones fo l lowing occlusion of th is artery (Kleber e t -a l . - , 1978; Janse e t - a l ; , 1979). Several reports suggest that co l l a te ra l s are neg l ig ib le in the rat ven t r i c le (Johns and Olson, 1954; Selye e t -a l . - , 1960; Maxwell e t -a l . - , 1984; Winkler e t - a l ; , 1984; Schaper et a l . , 1986). This suggestion is substant i - ated by the consistent extent of ischaemia and in fa rc t ion in rat hearts fol lowing coronary occlusion (Johnston e t - a l . , 1983a; Bernauer, 1982; Lepran e t -a l . - , 1983; e t c . ) , which compares favourably with the consistency seen in pig studies (Verdouw e t - a l ; , 1983c; Sjoquist e t - a l ; , 1983; e t c . ) . There is one report , however, which suggests that there is a di f ference between rats and pigs in that gradual stenosis induced over a period of 3 months w i l l induce co l l a te ra l growth in p igs , su f f i c ien t to prevent in fa rc t formation upon complete occ lus ion, whereas rats have end a r te r i es , and are unable to develop co l l a te ra l anastamoses, even fo l lowing slow occlusion (Schaper e t - a l ; , 1986). In summary, various hypotheses concerning the nature of arrhythmias occurring c l i n i c a l l y during and af ter myocardial ischaemia have t ranslated into a template for determining the ' c l i n i c a l relevance' of experimental studies (see Reimer et a l . , 1985). This template may be ser ious ly mislead- ing. It must be remembered that far more i s known about the nature of ischaemia-induced arrhythmias and in fa rc t ion in experimental animals than in humans, pa r t i cu la r l y during the c r i t i c a l f i r s t few hours af ter occ lus ion. C l i n i c a l l y , i t has been suggested that a lack of s c i e n t i f i c basis for the management of vent r icu lar arrhythmias is founded in poorly contro l led c l i n i - cal studies (Campbell, 1983). C lea r l y , the confusion concerning the nature of r i s k , cause and the strategy of management of arrhythmias occurring during acute myocardial ischaemia dictates that the c l i n i c a l concepts of - 35 - arrhythmogenesis should fol low from studies with experimental animal prepar- at ions, rather than govern the acceptab i l i t y of an experimental approach. I r respect ive of the arguments concerning c l i n i c a l relevance of d i f ferent species, un t i l a ser ies of drugs have been character ised for the i r a c t i v i t y against arrhythmias and in fa rc t ion in a range of experimental preparations with establ ished precis ion and accuracy, and un t i l these resu l ts have been compared with the analogous condit ion in humans with c lear d i s t i nc t i on between complete and par t ia l occ lus ion, abrupt and gradual occ lus ion, the time of onset of occ lus ion , large and small r i sk zones, the extent of c o l l a - tera l blood supply, the duration of ischaemia in the case of reperfusion, the degree of reper fus ion, the extent of pre-existant in fa rc t ion and the locat ion of pre-existant i n f a r c t i on , then any discussion of c l i n i c a l r e l e - vance is redundant. At present, most experimental preparations have not been characterised s a t i s f a c t o r i l y , and sources of error variance have not been recognised and el iminated. The foregoing chapters have alluded to one fact which ought to play the key ro le in the choice of experimental preparat ion. That i s , without ischaemia there is no i n fa r c t i on , and no ischaemia/ infarct ion re lated arrhythmias. Therefore, the primary consideration should be the reprodu- c ib le production of ischaemia of known sever i ty . The most ra t iona l means of producing ischaemia of known sever i ty is to abrupt ly, completely and perman- ent ly occlude a coronary artery in a species with minimal co l l a te ra l vascu- l a r i sa t i on and l i t t l e intraspecies v a r i a b i l i t y . In add i t ion , i t would be an advantage i f the species were small (not requir ing techn ica l l y demanding preparative equipment or inconvenient housing), inexpensive (to allow large sample s i z e s ) , robust (to allow rapid recovery from preparative surgery) and 'good natured' (to permit conscious animal experiments). There is a species which f i t s a l l these requirements, the common laboratory ra t . - 36 - 1.2.4 A h is tory of coronary occlusion in rats Coronary occlusion in rats was f i r s t undertaken in 1946 by Heimburger. His experiments were prompted by the desire to invest igate whether the production of adhesion of the pericardium to the epicardium was capable of preventing in farc t ion (by generation of co l l a te ra l anastamoses of ext racar- diac o r i g i n ) . His experiments were a f a i l u r e , in as much as in ject ions into the per icard ia l sac of cod l i v e r o i l , a lcoho l , sodium s a l i c y l a t e , urea, sodium stearate, soap and sodium morrhuate f a i l e d to prevent i n fa r c t i on . Nevertheless, Heimburger establ ished that l i ga t i on of the l e f t coronary artery in rats produced death within 24 h of occlusion in approximately 25 % of animals, and that extensive in fa rc t ion could be produced. The technique used for occlusion was l i ga t ion using a s i l k suture under pos i t i ve pressure anaesthesia. The study was essen t i a l l y qua l i t a t i ve , and no attempt was made to ve r i f y occ lus ion. Before continuing with the discussion of myocardial ischaemia in r a t s , the question of the v e r i f i c a t i o n of occlusion should be considered. V e r i - f i ca t i on of occlusion is an aspect of experimentation which has been neglec- ted throughout the h is tory of myocardial ischaemia and in fa rc t ion s tud ies . Without such ve r i f i ca t i on i t becomes impossible to d is t ingu ish between animals (or humans) in which a treatment has prevented or delayed in fa rc t ion and animals (or humans) in which occlusion was par t ia l or absent, or present but o f fset by co l l a te ra l vascu la r i sa t ion . Although the nature of co l l a te ra l vascu lar izat ion has been extensively documented in pigs and dogs (e .g . Schaper 1971; Meesman, 1982), very l i t t l e work has been undertaken in r a t s . Ear ly work (Johns and Olson, 1954; Selye et a l ; , 1960) showed that the coronary c i r cu la t i on in the ra t heart is very uniform, and that the coronary ar te r ies are end ar te r ies ( in contrast with guinea pigs in which extensive co l l a te ra l vascu lar izat ion is present) . - 37 - Recent radiolabeled microsphere studies in rats have shown that the extent of co l l a te ra l perfusion in the occluded zone (OZ) is neg l i g i b l e , flow being less than 0.01 ml/min/g according to Winkler e t a l ; (1984) and 6.1 ± 0.7% of normal flow according to Maxwell e t a l ; (1984). It should be mentioned that Kannengiesser et a l • (1975) found that residual flow in the OZ in iso la ted rat hearts was 18% of pre-occlusion f low, as assessed by microsphere techniques. This value i s somewhat higher than the values measured in -v ivo by Winkler e t - a l ; (1984) and Maxwell e t - a l . (1984). The higher estimate of flow i s l i k e l y to be the more accurate, owing to the fact that flow measurements are not dependent simply on the a b i l i t y of a bed to trap microspheres, as is general ly assumed; microspheres tend to d is t r ibu te p re fe ren t ia l l y into beds rec iv ing high f low. The resu l t is an under-estimation of flow in beds receiv ing lower f low. I t has been reported that rats do not develop co l l a t e ra l anastamoses during prolonged and gradual s tenos is , in contrast with p igs , and that in farc t ion is inev i tab le upon complete occlusion in r a t s , in contrast with guinea pigs in which in fa rc t ion is prevented by the extensive pre-existant co l l a te ra l vascu lar isat ion (Schaper e t - a l ; , 1986). Bloor et a l ; (1967) have also invest igated the rat coronary c i r c u l a t i o n . Although they did not look for co l l a te ra l anastamoses, they reported that the l e f t ven t r i c le i s supp- l i ed by a s ing le large coronary artery in more than 90 % of rats in an ex- tremely reproducible and consistent manner. Of the remaining 10 % of r a t s , almost a l l received l e f t vent r icu la r perfusion from 2 main ar ter ies a r is ing from separate o s t i a , and that the second artery was usual ly an independent septal ar tery , not supplying the l e f t vent r icu lar f ree w a l l . The work of Heimburger generated no in terest for many years. However, in 1954, Johns and Olson decided to undertake a detai led comparison of several small animals for the i r a p p l i c a b i l i t y to the study of myocardial - 38 - ischaemia and i n fa rc t i on . The experiments were prompted by the view that the species which had been exc lus ive ly used up to that t ime, the dog, was in essence useless for the study of myocardial ischaemia and i n fa r c t i on ; 'as a test procedure, coronary occlusion in the dog produces resu l ts so var iab le as to have only l imi ted s t a t i s t i c a l usefu lness ' . In th is regard, Johns and Olson were among the f i r s t to recognise that l o g i s t i c s play at least as an important part in the choice of a species for study as physio logical iden- t i t y with the human species. Johns and Olson examined the outcome of coro- nary occlusion in over 500 mice, r a t s , hamsters and guinea p igs. The exper- iments were carr ied out in a manner ident ica l with that used by Heimburger, (1946) except that hearts were exposed v ia an in tercosta l i nc i s ion rather than by thoracotomy. In prel iminary experiments (Johns and Olson, 1954), the coronary c i r c u - la t ion was invest igated by in jec t ing lead oxide (as a suspension in melted gelat in) into the ascending aorta ( b r i e f l y clamped to r e s t r i c t flow to the coronary a r t e r i es ) . The coronary c i r cu la t i on of the rat and mouse were described as s im i l a r . Vessels were c l ea r l y v i s i b l e fo l lowing in jec t ion of contrast medium and d issec t ion . The coronary ar ter ies lay jus t beneath the ep icard ia l surface, and were described as predominantly end ar ter ies ( c o l - la te ra l i n te ra r te r i a l anastomoses being infrequent) . In add i t ion , the l e f t anter ior descending coronary artery (LAD) was described as the dominant ar tery , whereas the circumflex artery was described as merely a small branch of the LAD. For th is reason, i t i s appropriate to describe the artery of occlusion in rats (and mice), that artery which emerges from under the a t r i a l appendage, as the l e f t main coronary ar tery , although most i n v e s t i - gators commonly refer to i t as the LAD. Johns and Olson's studies of the guinea pig (1954) are perhaps worth mentioning, since a comparison of the responses of the guinea pig and rat to - 39 - coronary occlusion has been used to invest igate the ro le of co l l a te ra l c i r cu la t i on in l im i t i ng in fa rc t ion (Winkler e t - a l , 1984). Johns and Olson described the guinea pig as having prominent coronary ar ter ies c l ea r l y v i s i b l e on the ep icard ia l surface. The circumflex was well developed, and both th is vessel and the r ight coronary artery were described as the source of la rge, prominent and profuse co l l a te ra l anastomoses feeding into the region supplied by the LAD. In keeping with the extent and the reproduc ib i l i t y of the co l l a te ra l vascu lar isat ion in rats and mice compared with hamsters and guinea p igs , the respect ive incidences of in fa rc t ion were 83, 64, 31 and 25% (Johns and Olson, 1954). Despite the extent of the i r study, Johns and Olson (1954) generated cer ta in gross pieces of misinformation. In pa r t i cu la r , the authors stated that VF was absent in r a t s . However, there is a simple reason for th is misleading statement. The authors did not record the ECG, and VF was d iag- nosed by observing ' s t r i k i n g changes in rate and rhythm' of the heart while the chest was open. Since the chest was closed immediately af ter occ lus ion , and since la ter studies ( e . g . , Johnston e t - a l . , 1983a) have c lea r l y shown that vent r icu lar arrhythmias do not occur unt i l several min af ter occ lus ion , i t i s no wonder that the authors did not ' see ' VF. Nevertheless, the s ta te - ment that ' i n contrast to dogs, mice and rats are able to survive occlusion of the l e f t main coronary ar tery; VF does not occur' (Johns and Olson, 1954) undoubtedly served to dissuade researchers from using rats in the study of myocardial ischaemia and i n fa r c t i on , despite the conclusion that ' t h i s method of coronary occlusion produces a test in fa rc t which is more nearly standard than any current ly a v a i l a b l e ' . Before discussing further developments in the h is tory of coronary occ lu - sion in the ra t , a comment should be made concerning the smallness of the - 40 - coronary ar ter ies of the ra t , as described by Johns and Olson (1954). Neither Johns and Olson, nor Heimburger (1946) mentioned that the i r occ lu - sions must (by v i r tue of the imposs ib i l i t y of d issect ing free almost i n v i s - ib le artery imbedded in the myocardium) have included some myocardium with the l igated ar tery. It i s possible that i f the artery i s missed and only muscle l igated then spurious 'small i n f a r c t s ' may form, confounding analy- s i s . However, according to Heimburger (1946), l i ga t i on of myocardium alone, or the l e f t coronary vein alone produced no s ign i f i can t sequelae ( in fa rc - t i on ) . In other words, unless the LAD is included in the l i g a t i o n , no in fa rc t ion occurs. From 1954 to 1973 very l i t t l e work was done using rat occlusion prepara- t i ons . Bryant e t - a l ; (1958) used Johns and Olson's technique (1954) to invest igate the in fa rc t process by electron microscopy. They found that gross s t ructura l changes in the myocardium did not occur un t i l at least 5 h af ter occ lus ion, by l i gh t microscopy. However, electron microscopy revealed swel l ing of the sarcoplasmic ret iculum and mitochondrial enlargement af ter only 1 h of occ lus ion. By 2 h, deposit ion of l i p i d bodies, uniform enlarge- ment of mitochondria and par t ia l d isrupt ion of myof ibr i ls were present. Animals sac r i f i ced at la te r times showed increasingly severe i n t r a c e l l u l a r d is rup t ion , such as l i p i d body formation in c e l l nuclei (4 h) , loss of de f in i t i on of myofilaments (5 h) general d isarray of i n t r a c e l l u l a r order (24 h) and f i n a l l y extensive i n f i l t r a t i o n of neutrophi ls and macrophages, with associated phagocytosis (48 h) . The authors made no comments on the i r reasons for using r a t s , and gave no ind icat ion of the reproduc ib i l i t y of the extent of i n f a r c t i on . A year l a te r , Kaufman et a l ; (1959) embarked on a ser ies of studies on in fa rc t ion in rats (a ser ies which appears to have terminated af ter 1 pub l i - ca t ion) . Rats were chosen because they were considered to provide ' large - 41 - numbers' and 'uniform s tock ' . Johns and Olson were c i ted as the source of the technique of occ lus ion. These invest igators (Kaufman e t -a l . - , 1959) showed that occlusion caused depletion of succinate dehydrogenase which began 4 h after occlusion and slowly progressed over the fo l lowing 12 h. In confirmation of the resu l ts of Bryant et a l ; , Kaufman e t - a l ; found that no apparent a l terat ions occurred in gross mycardial structure unt i l at least 6 h af ter occ lus ion . However, these authors observed i n f i l t r a t i o n by mononu- c lear leukocytes beginning at 18 h af ter occ lus ion, reaching 'moderate' proportions by 24 h, at a time when Bryant e t - a l ; (1958) detected no such i n f i l t r a t i o n . The scarr ing process began 36 h af ter occlusion with the i n f i l t r a t i o n of f i b rob las t s . Kaufman e t - a l ; (1959) reported that the dec- reases in succinate dehydrogenase and other enzymes with time para l le led those reported fo l lowing occlusion in dogs according to Jennings e t - a l . (1957), and concluded that the rat was a ' sa t i s fac to ry animal for the study of sequential histochemical and morphologic changes in the myocardium'. The fo l lowing year, the much-cited work of Selye et-al .- (1960) was publ ished. These authors re i tera ted the view of Johns and Olson (1954) that the dog, by v i r tue of the var iab le outcome of occlusion in th is species, i s unsuitable for the study of myocardial ischaemia and i n fa rc t i on . However, Selye e t - a l ; (1960) c r i t i c i s e d Heimburger (1946) and Johns and Olson (1954) for the i r 'complicated' procedure, the high post-operative morta l i ty and the f a i l u r e to produce an in fa rc t with every occ lus ion. To be f a i r to Heimburg- er (1946) and Johns and Olson (1954), the f i r s t 2 c r i t i c i sms are completely i n v a l i d . F i r s t l y , i t appears that a l l 3 groups prepared rats using essen t i - a l l y the same technique. The e a r l i e r 2 studies are superior in as much as they both employed a r t i f i c i a l r esp i ra t i on . Selye e t - a l . (1960) used no a r t i f i c i a l r esp i ra t i on , claiming that they could occlude an artery so quick- l y that none was required. However, the speed was dependent on the rather - 42 - crude expedient of ' e x t e r i o r i s i n g ' the heart, i . e . , temporari ly 'popping' i t through the in tercosta l i n c i s i o n . Secondly, although i t i s d i f f i c u l t to make d i rect comparisons owing to the imprecise declarat ions of time of death, i t appears that per i -operat ive morta l i ty was 25% in Heimburger's study (1946), 21% in Johns and Olson's (1954), and 10% in Selye 's (1960), di f ferences which are not pa r t i cu la r l y s t r i k i n g . In addi t ion, no d i s t i n c - t ion was made between death resu l t ing from occlusion and death resu l t ing from bad surgery and anaesthetic overdosage. The myth that rats do not develop VF fo l lowing occlusion (Johns and Olson, 1954) presumably confounded Selye e t - a l i ' s perception of the super io r i ty of the i r technique, g iv ing r i s e to the assumption that any per i -operat ive death occurs as a resu l t of poor technique. Although Selye e t - a l ; (1960) did not record the ECG, they appeared to acknowledge that the extent of occlusion inf luences s u r v i v a l , since they found that the production of 'smal ler ' in fa rc ts by de l ibera te ly occluding the LAD d i s ta l to i t s o r ig in was associated with a marked reduction in mor ta l i ty , whereas l i ga t i on of the both l e f t and r ight coronary ar ter ies in the same rat resul ted in immediate death in 80% of animals. These authors (Selye e t - a l . , 1960) also confirmed the report of Heimburger (1946) that occlusion of the coronary veins alone produced no sequelae (unless the ent i re coronary sinus was occluded, in which case pe r i ca rd i t i s and super f ic - i a l ep icard ia l c a l c i f i c a t i o n developed). The only other point of in terest in Se lye 's work concerns long term su rv i va l . It was found that animals surv iv ing the f i r s t 24 h af ter occlusion general ly a l l survived the fo l low- ing month, death general ly occurring only as a resu l t of 'accidental i n fec - t i ons ' (Selye e t - a l . , 1960). Several pecul iar publ icat ions appeared in the fo l lowing years. The f i r s t report of ECG information fo l lowing occlusion in rats was given in - 43 - 1961, when Normann e t - a l ; , using the technique of Johns and Olson (1954) recorded 5 leads (3 limb and 2 chest) at 2 h af ter occlusion (presumably during br ie f ether anaesthesia). Remarkably, no arrhythmias were observed, although ST segment elevat ion in the anter ior chest lead was apparent, and deep Q waves were seen in the same lead at 24 h and 7 days af ter occ lus ion. A s im i la r study was performed by Zsoter and Bajusz in 1962, and the only arrhythmias seen were a few PVCs in 1 ra t . However in th is study (Zsoter and Bajusz, 1962) the ECG was not recorded un t i l 2 days af ter occ lus ion. No information concerning surv ival af ter occlusion was given. Bajusz, who had previously worked with Selye and Zsoter, re i tera ted the be l i e f that coronary occlusion causes no arrhythmias in r a t s , in his book on in fa rc t ion (Bajusz, 1963). In 1964 and 1965, 4 deta i led and extensive, but completely neglected works were published by Hort 's group (Hort et a l . , 1964; Hort and Da Canal- i s , 1965a; 1965b; Hort 1965). These authors carr ied out coronary occlusion in over 1000 rats by 1 stage and 2 stage l i g a t i o n , with and without reperfu- s ion . According to the English language abstracts fo l lowing the German tex t , the c r i t i c a l duration of ischaemia necessary for production of i r r ev - e rs ib le damage was 13 - 15 min. These authors also stated that in fa rc t ion could only be delineated s u f f i c i e n t l y c l ea r l y for ana lys is , according to the t r iphenyl tet razol ium sta in ing method (Jestadt and Sandr i t ter , 1959), when measured 10 h or more af ter occ lus ion , whereas that t issue destined to become infarcted could be d i f fe rent ia ted from normal vent r icu lar t issue at only 15 min af ter occlusion by l i gh t microscopy, using the c r i t e r i on of 's t reched' versus 'non-streched' muscle f ib res ( ' s t re tched ' being i r r eve rs - ib l y damaged but not yet dead). This c r i t e r i on was suggested to be potent i - a l l y useful in the diagnosis of acute myocardial ischaemia in pat ients who have died from unknown causes (sudden death) without evidence of coronary - 44 - thrombosis or i n fa rc t i on ; th is recommendation has been ignored. The authors concluded that since the standard error of mean in fa rc t s ize was always less than 10% of the mean, then the coronary-1igated rat prepartion was sui tab le for measuring the ef fect of drug treatment on in fa rc t s i z e . In th is regard, these authors (Hort and Da Canal i s , 1965a;1965b) provided sound quant i tat ive evidence of rep roduc ib i l i t y to support the e a r l i e r semiquantitat ive evidence of Heimburger (1946) and Johns and Olson (1954). Despite the c lear exposit ion of ear ly vent r icu lar arrhythmias in dogs fo l lowing coronary occ lus ion, which had been known for decades ( e . g . , Town- shend Por ter , 1894; Smith 1918), i t was not un t i l 1973 that the myth that rats do not experience arrhythmias fo l lowing occlusion was d ispe l led by the simple expedient of monitoring the ECG during and immediately af ter occ lu - s ion . This work was not carr ied out by any of the invest igators who had ea r l i e r proclaimed the rat to be superior to the dog for studying myocardial ischaemia and in fa rc t ion ( e . g . , Heimburger, 1946; Selye e t -a l . - , 1960) but by 2 members of the Hungarian Army Medical Corps (Kenedi and Losconci , 1973a). Kenedi and Losconci recorded the ECG during the 10 min period fo l lowing occlusion (while the ether anaesthesia used in preparation wore o f f ) . Before discussing the i r r esu l t s , i t i s worth mentioning that the technique of Selye e t - a l . (1960) was used in preparat ion, and that in every case the act of ex te r i o r i s i ng the heart through the in tercosta l i nc is ion was reported to produce t ransient i r regu lar VT, a t r ioven t r i cu la r block and sinus brady- card ia ; the technique of Heimburger (1946) as used by Johns and Olson (1954) would not be expected to lead to arrhythmias in th is way, since heart exter- io r i s a t i on was not employed. Kenedi and Losconci (1973a) found that during the f i r s t 10 min af ter occ lus ion , a high incidence of vent r icu lar arrhythmias occurred. General ly, a short run of PVC occurred immediately upon occ lus ion, but these i n i t i a l - 45 - arrhythmias ( la ter ca l led phase-la by Fagbemi, 1985) resolved in a l l cases. However, i t was the opinion of these authors that the run of PVC occurring upon occlusion was a resu l t of the mechanical trauma of preparation rather than ischaemia, because these arrhythmias were s im i la r to the pre-occlusion arrhythmias induced by ex te r io r i s ing the heart through the in tercosta l i n c i s i o n , rather than the arrhythmias which occurred a few minutes after occ lus ion. After an unspecif ied i n t e r v a l , a high incidence of PVC, VT and VF occurred. Unfortunately, the exact incidence and time course of the arrhythmias was not given, although of the 5/20 rats which experienced VF or ' f l u t t e r ' , 2/5 d ied. In addit ion to th is f i r s t exposit ion of arrhythmias associated with acute myocardial ischaemia in r a t s , Kenedi and Losconsci (1973a) also noted that a l l rats surv iv ing 24 h af ter occlusion exhibi ted a deep Q wave, and ST segment elevat ion in lead 1 of the ECG. Arrhythmias induced by coronary occlusion in rats were not invest igated again un t i l 1979, when Szekeres' group (Lepran e t -a l . - , 1979; Siegmund e t - a l ; , 1979a; 1979b) and our laboratory (Au et a l . , 1979a; 1979b) began the i r ser ies of experiments, which have continued throughout the 1980s, along with work by Pa r ra t t ' s group (Clark e t - a l ; , 1980; e t c . ) , Winslow (Kane and Winslow, 1980; e t c . ) , Bernauer (1980; e tc . ) and others. In general , the above Investigators have concentrated on arrhythmias, although our labora- tory and Bernauer always measure the OZ and the extent of i n f a r c t i on . Other researchers have concentrated on in fa rc t s ize reductions in ra t s , and have completely ignored arrhythmias, for example, the Harvard contingent (MacLean e t - a l ; , 1976; Kloner e t - a l . , 1977; Pfef fer e t - a l . , 1979; 1982; 1985), Chiar- i e l l o ' s group (Ch ia r ie l l o et a l . , 1980; 1983; 1984; 1985), Flaim and Ze l i s (1981) and others. In doing so, these authors ignore censoring associated with VF-induced morta l i ty ; in most cases, the numbers of ear ly deaths and the i r causes are not given. - 46 - Independently of whether in fa rc t ion or arrhythmias are the main subject of study, the h is tory and past l i t e ra tu re concerning the use of the rat have been almost un iversa l ly ignored. For example, the o r ig ina l work of Kenedi and Losconci (1973a) has only been appropriately c i ted twice (Clark e t a l , 1980; Abrahamsson and Almgren, 1980). In add i t ion , c red i t is general ly given to Selye (1960) for inventing coronary occlusion in ra t s , whereas in r e a l i t y the technique was developed by Heimburger, 14 years e a r l i e r . 1.3 Ventri eu1ar-arrhythmi as -i n - acute myocardi al i schaemi a This chapter is concerned with the genesis of ischaemia-induced ven t r i - cular arrhythmias. Many of the concepts associated with arrhythmogenesis, such as reentry (Mines, 1913) have been under invest igat ion for many years. An understanding of what l i e s behind the gross pathologic expressions of heart disease such as VF would be expected to ass is t in the predict ion of which type of drug might be of benef i t . However, i t i s also worth consider- ing that the cart often comes before the horse, in that many c l i n i c a l l y useful drugs have been developed on the basis of a purely empirical app- roach. In essence, a mult i layered approach to the problem of ischaemia- in- duced arrhythmias i s ind icated, and i t would be foo l i sh to dismiss an empir- i ca l approach on the grounds that i t does not take su f f i c i en t considerat ion of the underlying pathologic mechanisms. For example, the an t i -u lcer H,, receptor histamine antagonists were developed from the premise that h i s t a - mine was somehow involved in ulcerogenesis v ia a mepyramine-insensitive mechanism. What that mechanism i s , in terms of the biochemical basis for the generation of a gast r ic les ion and i t s precise re la t ionsh ip with stomach pH, remains uncertain. Nevertheless, cimetidine and ran i t i d ine have proven to be e f fec t i ve an t i -u lcer agents. This chapter attempts to i l l u s t r a t e what i s known and what remains to be ascertained with regard to the e lect rophys io log ica l basis of arrhythmo- - 47 - genesis in acute myocardial ischaemia. In add i t ion , an attempt has been made to examine why and how calcium antagonists might be of benefi t in such circumstances. As background information, a br ie f summary of some of the major features of the electrophysiology of exc i table t issue i s also given. 1.3.1 The electrophysiology of exc i table t issue Arrhythmogenesis is an e lect rophysio log ica l phenomenon. In order to understand the electrophysiology of myocardial ischaemia i t i s necessary to understand normal heart e lectrophysiology. Normal vent r icu la r function is dependent upon the propagation of depolar isat ion. The charge i s ca r r i ed , as in the case of conduction in nerves, by ions. The charac te r i s t i cs of con- duction are governed in part by the passive e l e c t r i c a l propert ies of the t i s sue , and in part by the act ive or dynamic propert ies of the t i s sue . The passive e l e c t r i c a l propert ies may be approximated according to the cable equations developed by Kelv in (1855). If the resistance across the c e l l membrane i s high compared with the i n t r a c e l l u l a r and ex t race l lu la r res is tance, then any transmembrane potent ial di f ferences and currents can be considered as functions of longi tudinal distance and time, only. For such an ' i d e a l ' cab le , the change in membrane potent ia l resu l t ing from the focal appl icat ion of current w i l l decay in distance and time according to : d 2 E m / d x 2 . l / r 2 = C m .dE m /d t + i . where x is distance along the cable from the current source, E m i s memb- J m rane po ten t ia l , r 2 i s i n t r ace l l u l a r f l u i d res is tance, C m i s membrane capacitance, t i s time and i . is the component of the membrane current carr ied by ions (Hodgkin and Huxley, 1952a). Rearrangement of the cable equation gives r i se to two membrane constants which are often useful in the understanding of e lectrophysiology. The length constant, x = J r m / r 2 describes the distance along the membrane from a current source at which the resul tant potent ia l change has f a l l en to - 48 - 1/e (approximately 37%) of that at the point of the current source. The time required for V (Em) to decay from i t s i n i t i a l value (VQ) to 63% of th is value gives the product R m C m , commonly ca l led the time constant of the membrane ( tau m ) . This ar ises because the membrane potent ia l at time t a f ter appl icat ion of a current across the membrane i s given by: When t is equal to R m C m , then V. = V m ( l - e - 1 ) , and 1-e" 1 = 0.63. Ill i l l U IT laX The act ive e l e c t r i c a l propert ies of the t issue depend upon the asymmet- r i c a l nature of exc i tab le t i ssue . The oriented enzyme, Na +-K +-dependent + ++ ++ ATPase, the Na -Ca exchanger, i n t r a c e l l u l a r Ca sequestrating proteins such as calsequestr in and other energy-dependent processes serve to generate a chemical potent ial gradient for several charged e n t i t i e s . In addi t ion, Donnan equi l ib r ium, the presence of a r e l a t i v e l y f i xed impermeant body of i n t r a c e l l u l a r anionic pro te in , and the presence of a r e l a t i v e l y high rest ing K + permeabil i ty versus that of other charged e n t i t i e s , leads to the establishment of a membrane po ten t ia l . F i n a l l y , as w i l l be discussed below, i . does not obey Ohm's Law. This v a r i a b i l i t y in ion ic conductance (g.) const i tutes the operational mechanism of the act ive e l e c t r i c a l prop- e r t ies of exc i tab le t issue which leads to propagation. At r es t , t issue in the heart (excluding nodal) can be said to approxim- ate a resistance-capacitance c i r c u i t with a K + bat tery, such that the resu l t ing rest ing membrane potent ial i s given by the Nernst equation: E m = R T / F . l n ( [ K + ] o / [ K + ] . ) + + + where [K ] and [K ] . are the local concentrations of K on the immediate outside and inside of the membrane, respec t ive ly . The Nernst equation for K + gives the reversal potent ial for the K + current , which is approximately -90 mV under normal physio logical condit ions ( [ K + ] Q approximately 2.5 - 4.5 meq/1 and [ K + ] i 120 - 140 meq/1), and - 49 - th is corresponds reasonably well with the rest ing membrane potent ial of most non-nodal heart t i ssue . This model is based on the assumption that at r es t , the K + permeabil i ty of the membrane far exceeds that of other charge-car- rying species, such as Na and Ca . The act ive dynamic propert ies of exc i tab le t issue may be approximated by var ia t ions in the equations which were developed by Hodgkin and Huxley (1952a; 1952b; 1952c; 1952d) to explain the rec t i f y i ng propert ies of e x c i t - able t i ssue . In th is instance, r e c t i f i c a t i o n is the deviat ion of the cur- rent-voltage re la t ionsh ip from l inear (ohmic) as a resu l t of vol tage- dependent changes in ion ic conductance. Rec t i f i ca t i on i s explained by invoking the concepts of separate ion ic currents associated with r e l a t i v e l y ion-se lec t ive channels which may vary with membrane potent ial and/or time (so-ca l led voltage and time dependence). The equations describe the beha- viour of these conductances in re la t ion to membrane potent ia l and time. Rec t i f i ca t ion has been demonstrated in Purkinje f ib res (Deck and Traut- wein, 1964). At membrane potent ia ls pos i t ive to approximately -30 mV, there is an increase in a conductance which i s r e l a t i v e l y se lec t ive for K + . The propert ies of th is conductance resemble those of a conductance described by Hodgkin and Huxley in the squid axon which act ivates slowly fo l lowing step changes in holding po ten t ia l , and is c a l l e d , as a consequence, the delayed r e c t i f i e r (see below). The current associated with th is conductance is car r ied mainly by K + , i s outward-going, causes repo lar iza t ion and i s act ivated by depolar isat ion. There i s a second type of r e c t i f i c a t i o n in Purkinje t issue associated with a K + conductance. This conductance is act ivated by changing holding potent ial in voltage clamped (see below) t issue from -30 mV to more negative values. Since depolar isat ion reduces the conductance for th is K +current (as opposed to the behaviour of the delayed r e c t i f i e r ) , th i s current is - 50 - known as the anomalous r e c t i f i e r current , with 'inward going' r e c t i f i c a t i o n . I t i s perhaps important, at th is stage, to comment on the use of voltage clamping. When invest igat ing current f lowing as a resu l t of a change in membrane po ten t ia l , i t i s necessary to el iminate capacitance current ( i c ) . This is done by maintaining membrane potent ial at a constant value (the holding po ten t ia l ) , because i = C m . d E m / d t . Since the value dE m /dt is zero i f E m i s held (clamped) constant, then the term C m . dE m / d t disappears from the cable equation, leav ing: d 2 E m / d x 2 . l / r 2 = i . In other words, the transmembrane current carr ied by ions becomes a f a i r l y simple function of membrane po ten t ia l , under clamped condi t ions. The p r i n - c ip le of voltage clamping was u t i l i s e d by Hodgkin and Huxley (1952a; 1952b; 1952c; 1952d), in order to measure i - in the squid giant axon. According to Hodgkin and Huxley, r e c t i f i c a t i o n may be described mathe- mat ica l ly by invoking voltage-dependent f i t t i n g parameters which govern a conductance channel. In the case of the delayed r e c t i f i e r , conductance i s governed by the parameter, n. For a population of delayed r e c t i f i e r conduc- tance channels, the to ta l current i s dependent on the proportion of channels in the open s ta te . The var iable n behaves according to the equation: dn/dt = a n ( l - n ) - 8 n ( n ) where c*n and &n are voltage dependent rate constants, n i s the molar f rac t ion of channels in the open state and 1-n is the molar f rac t ion of channels in the closed s ta te . Hodgkin and Huxley found that the i r resu l ts best f i t a model where K + conductance (gK) was proportional to n^. The simplest physical concep- tua l i sa t i on of the above re la t ion is that n is a pa r t i c l e which functions to 'gate ' the conductance channel and ex is ts e i ther in an open state (n) or a closed state (1-n). The rate constant a n refers to the rate of conversion of a channel from the rest ing (closed) state to the depolarised (open) s ta te , and v ice-versa . The voltage-dependence of the channel, which ac- counts for i t s rec t i f y i ng proper t ies, is explained by the voltage-dependence of c*n and e n ; the value of ct n increases upon depolar isat ion, while Bn decreases upon depolar isat ion. Although th is model was developed for nerve, manipulation of the values of a n and e n can provide theoret ica l conductance-voltage re la t ions which correspond well with experimental ly-derived data in Purkinje t issue (Noble, 1960; Noble 1962). I t i s suspected that there are many d i f fe rent conductances in the heart which show a range of propert ies. These propert ies include ion s e l e c t i v i t y , ac t iva t ion by changes in membrane po ten t ia l , inac t iva t ion by changes in membrane po ten t ia l , inact iva t ion with time (see below) and reac t i va t ion . In add i t ion , under pathological condit ions such as myocardial ischaemia, some of these propert ies may be influenced by changes in pH, c y c l i c AMP, tempera- ture, e tc . (see Hauswirth and Singh, 1978). The delayed r e c t i f i e r i s voltage-dependent, since a n and Bp vary in s ize with membrane po ten t ia l . However, the Na + conductance (g^ a) in the squid axon (and the s im i la r ' f as t inward current ' in the ven t r i c le ) shows an addit ional property, time dependence. This refers to the observation that the increase in g ^ g which occurs upon depolar isat ion i s t rans ien t , swi tch- ing off ( inact ivat ing) with time. This phenomenon was explained by the invocation of two gating var iables ( in contrast with the s ing le species for the delayed r e c t i f i e r ) , one of which, m, behaves l i k e n of the delayed r e c t i f i e r , sh i f t i ng to the open channel state upon depolar isa t ion, while the second, h, behaves in the opposite manner, sh i f t i ng to the closed state upon depo lar isa t ion . If e i ther n or h is in the closed s ta te , then that pa r t i c - ular channel i s closed (conductance for that channel is minimal), s ince , in - 52 - analogy with genet ics, the open state is recessive while the closed state i s dominant. In order for g^ a to f i r s t increase upon depolar isat ion, then decrease, i t i s necessary that the rate constant, a m , which governs the rate of conversion of m from the rest ing (closed) state to the open s ta te , be larger than a^, the rate constant for the conversion of h from the rest ing (open) state to the closed s ta te , both of which reactions occur upon depolar isat ion. The in teract ion between the fas t inward current ( i ' N a ) and the delayed r e c t i f i e r in time governs the a b i l i t y of an AP to occur, and in space, to propagate, in many exci table t i ssues . Depolar isat ion serves to increase g^ia v i a ef fects on a m . This e f fect leads to more depolar isa t ion. However, depolar isat ion also serves to reduce g^ a v ia ef fects on c^, and also serves to increase gK v ia ef fects on a n . These l a t t e r ef fects serve to repolar ise the membrane. However, since the maximum value of a m i s much larger than that of a n or a^, then provided depolar isat ion is large enough (at or above ' threshold ' ) and occurs qu ick ly (so as to preclude steady-state elevat ion of leading to ' i nac t i va t i on ' ) then pos i t ive feedback of depolar isat ion w i l l occur before the repo lar iza t ion processes are act ivated, producing the AP. The AP is terminated by a combination of inact iva t ion of g^ a and act ivat ion of the delayed r e c t i f i e r . The assymetry of exc i tab le c e l l s leads to the p o s s i b i l i t y of propaga- t i o n , since K i r cho f f ' s law is sa t i s f i ed by the induction of outward current some distance away from the inward current , l inked to the inward current by longi tudinal ( i n t r ace l l u l a r and ex t race l lu la r ) current , which i s induced in turn by the longi tudinal potent ial gradient resu l t ing from focal depolar isa- t i o n . Ampl i f ica t ion may occur in th is c i r c u i t as a resu l t of s im i la r pro- cesses which lead to the AP, as fo l lows. If wi thin a region of membrane depolar isat ion is s u f f i c i e n t l y la rge, and occurs s u f f i c i e n t l y qu ick ly , then w i l l swamp repolar is ing K + current , leading . to further depolar isa- t ion at points further and further away from the i n i t i a l point of depo lar is - ation in a manner described by the cable equation. This manifests as a wave of depolar isat ion (the 'propagated A P ' ) . The threshold for generation of a propagated AP depends upon x m and t a u m . A propagated AP passes as a wave through exc i tab le t i s sue , leading to depolarisation-1 inked events, such as neurotransmitter re lease, or in the case of vent r icu lar t i ssue , coordinated muscle contract ion. The gating propert ies of the conductance channels confer d i rec t ion to conduction in vent r icu lar t i ssue . This is because the propagating wave of exc i ta t ion leaves behind a band of t issue in which g ^ a i s inac t iva ted. The i nac t i va - t ion of g ^ a i s a consequence of depolar isat ion-induced increases in a ^ . This i n e x c i t a b i l i t y i s known as ' r e f r a c t o r i n e s s ' . The duration of re f rac to - r iness under normal circumstances i s dependent on repo la r i sa t ion , since repo lar isa t ion serves to increase and reduce a ^ , leading to a sh i f t in the equi l ibr ium of h to the open-state. With the passing of t ime, and the re-establishment of po la r i sa t i on , as more and more channels become re -ava i lab le for opening, the t issue i s said to pass from the stage of absolute re f ractor iness to re la t i ve re f rac to r iness , and f i n a l l y to the f u l l y exc i tab le s ta te . The conduction ve loc i t y (e = x / t ) of a propagated AP can be derived in terms of the cable equation: e 2 = ( x 2 . d 2 V / d t 2 ) / ( t a u m . d V / d t + V) This der ivat ion predicts that e is d i r ec t l y proportional to x, and inversely proport ional to the square root of t au m . In addi t ion, e i s proportional to the square root of the maximum r i se rate of the AP (dV/dt ). Under max normal circumstances, d V / d t m 3 v i s the major determinant of ©, with © almost l i n e a r l y re lated to the square root of d V / d t m a x . Manipulations - 54 - which ei ther prevent channels from opening (by f i x i n g the channel in the inact ivated s ta te , preventing act ivat ion or phys ica l l y blocking the chann- e l s ) , or slow the k ine t ics of opening by sh i f t i ng the re la t ionsh ip between am and E m to the r ight (to more pos i t i ve values of Em) w i l l reduce e. In most nerve t issue and normal vent r icu lar t i s sue , dV/dt i s governed Hid A by i ^ a . In ischaemic vent r icu lar t i s sue , i t i s possible that d V / d t m a x i s dependent on currents other than the normal fas t inward current (see below). 1.3.2 The electrophysiology of the normal ven t r i c le Variat ions in the act ive and passive e l e c t r i c a l propert ies of vent r icu la r t issue between d i f ferent regions of the ven t r i c le govern the shape of the propagating AP. These var ia t ions are brought about by var ia t ions in the dimensions of the c e l l s in spec i f i c regions and also var ia t ions in the conductance channels present in the d i f fe rent regions. By manipulation of ex t race l lu la r ion composition (for example, replace- ment of Na + with choline or L i + ) , addit ion of substances shown to se lec - t i v e l y i nh i b i t spec i f i c conductances, such as tetrodotoxin which se lec t i ve l y blocks g N a (Moore et a l . , 1967), and by appl icat ion of various techniques for clamping voltage or current such as the recent ly- introduced patch-clamp technique (Lee e t - a l ; , 1980), i t has been possible to describe some of the conductance systems which contr ibute to the propagated AP of the heart . While i t i s possible to study iso la ted heart c e l l s (despite the i r small s ize) by patch clamping (Lee e t - a l ; , 1979), i t i s d i f f i c u l t to study heart c e l l s in vivo because of movement (the organ beats'.) , and unfavourable c e l l u l a r and i n t e r ce l l u l a r geometry (Attwell e t - a l . , 1979). Nevertheless, although some conductances have yet to be f u l l y character ised, i t i s gener- a l l y accepted that the shape of the vent r icu lar AP is mainly governed by three current systems, i N , the outward repo lar is ing currents (carr ied - 55 - mainly by K + ) and the slow inward current ( i s l -» car r ied mainly by C a + + ) . In vent r icu lar t i s s u e , i ^ a with i t s rapid opening and c los ing k ine t i cs dominates the i n i t i a l r i s e of the AP, and governs the d V / d t m , v of depolar- max isa t ion (Beeler and Reuter, 1970a). This in turn governs the conduction ve loc i t y . The proportion of i* N a channels ava i lab le for opening governs the e x c i t a b i l i t y of the t i ssue . Ventr icular i ^ a i s s im i la r to the Na + current described by Hodgkin and Huxley for the squid giant axon, with minor di f ferences (such as the voltage-dependence of the inh ib i to ry action of tetrodotoxin in the v e n t r i c l e ) . Repolar izat ion is associated with several K + currents including a current s im i la r to the delayed r e c t i f i e r current in the squid axon, ca l led i K 2 (McAl l i s ter and Noble, 1966), which is governed by the gating var iable s (analogous to n ) . Several other outward currents may also contribute to repo la r i za t i on , for example, mixed ion currents (predominantly K + ) ca l led \l a n d \2 ( N o b l e a n d Ts ien, 1969). The gK systems inf luence the AP durat ion, since inh ib i t i on of gK delays repo lar iza t ion by prolonging the t a i l of the AP. I f the to ta l duration of the AP (often measured as the time for 90 % repo la r i sa t ion , or APD90) is increased, then the ref ractory period is increased. This i s because the membrane potent ial must remain s u f f i - c i en t l y negative for a time su f f i c i en t for conversion of the h var iables from the i r closed to the i r open state ( react ivat ion) for e x c i t a b i l i t y to be restored. In th is regard, i f Purkinje t issue i s clamped at -70 mV then the g N a system is 50 ^ inact ivated (Weidmann, 1955a). The th i rd major current in vent r icu lar t issue i s the slow inward current ++ ( i s . j ) . This current is carr ied mainly by Ca , and is the major deter- minant of the plateau of the AP in vent r icu lar t issue (Mascher and Peper, 1969; Beeler and Reuter, 1970a; 1970b; New and Trautwein, 1972). As such, - 56 - serves to govern the AP durat ion, in conjunction with the repo lar is ing outward K + current systems described above. The l a t t e r may themselves be par t ly governed by i ^ , since i ^ causes a t ransient r i se in unbound i n t r ace l l u l a r C a + + concentration which has been shown to enhance some K + conductances (Isenberg, 1977b; 1977c). In the ven t r i c l es , restorat ion of normal e x c i t a b i l i t y af ter the propagation of an AP requires that i .. as well as i ^ a recovers from inac t i va t ion . Since the duration of the AP is governed by i .. to a large extent, then re f ractor iness i s also dependent on i . . Inh ib i t ion of i ' s i w i l l shorten AP duration at 25% r e p o l a r i - sat ion (APD25), but APD90 may ei ther increase or decrease, presumably accor- ding to the dependence of the t a i l of the repo lar is ing current on free i n t r a c e l l u l a r C a + + ( th is may vary from one type of vent r icu lar t issue to another, and may also vary from species to spec ies) . It i s possible that during myocardial ischaemia, i . plays an important ro le in addit ional aspects of conduction in the vent r icu lar t i s sue , as w i l l be discussed in deta i l below. Just as i ^ a possesses two gating var iab les , i . has been shown to be governed by two var iab les , denoted as d and f , analogous to m and h, respec- t i v e l y (Reuter, 1973). The voltage-dependence of a d i s such that the threshold for i . i s approximately -50 mV, s l i g h t l y more pos i t ive than that for . For f , the rate constant for conversion to the open state (Bf) fa r exceeds the rate constant for conversion to the closed state (df) at membrane potent ia ls more negative than -60 mV. The rate constants for ac t iva t ion (a^) and inact iva t ion (a^) are much larger ( ind icat ing slower k ine t ics ) than corresponding rate constants for i^ . There is some species va r i a t i on , a£ being 80-200 msec in cat and dog ven t r i cu la r t issue (McDonald and Trautwein, 1978; Reuter and Scholz, 1977), and 10-30 msec in rat vent r icu lar t issue (Isenberg and Klockner, 1980). In the la t te r - 57 - species, the fast k ine t ics of i $ ^ probably account for the abbreviated plateau of the AP (compared with the cat , dog, pig and human vent r icu lar AP), which i s associated with a shorter APD90 of 100 instead of 300 msec (Langer, 1978), and accords with the high rest ing heart rate of 350-450 beats/min. Since vent r icu lar c e l l s are small compared wi th , for example, the squid giant axon, and are part of a funct ional syncytium, i t i s d i f f i c u l t to invest igate the i r electrophysiology in a manner which provides unequivocal information (Beeler and McGuigan, 1978; McDonald, 1982). Furthermore, beating c e l l s are d i f f i c u l t to impale with microelectrodes. Patch clamp studies are favoured at present for the i r power to reveal the propert ies of vent r icu lar conductances. Nevertheless, as one removes c e l l s or membranes from the i r physio logical environment one may introduce factors which con- found inves t iga t ion . For example, i • i s enhanced by pharmacological and physio logical manipulation which increases c y c l i c AMP (cAMP). It i s thought that cAMP t r iggers phosphorylation of the f ' p a r t i c l e ' , leading to slowing of inac t iva t ion of i . and an enhancement of peak i . (Bean e t - a l . , 1984). Therefore, since the basic propert ies of conductance channels in vent r icu lar t issue remain to a cer ta in extent unclear, i t i s possible to suggest many explanations for the mechanism by which a drug inf luences vent r icu lar e lectrophysiology, whether t issue i s normal or abnormal ( ischae- mic, for example). 1.3.3 E lect rophys io log ica l changes caused by myocardial ischaemia The study of the electrophysiology of myocardial ischaemia is confounded by the dynamic unstable charac te r is t i cs of myocardial ischaemia. The e l ec t - rophysio logical propert ies of the ischaemic ven t r i c le vary from one region to another, and also from time to t ime, as i n t r a - and i n t e r c e l l u l a r biochem- i s t r y var ies from one region to another, and changes from time to time. - 58 - Furthermore, the e lect rophysio log ica l changes vary according to whether blood supply i s completely or p a r t i a l l y reduced. In short , these considera- t ions d ic tate that 'the ischaemic myocardium' i s not a s ing le en t i t y . The electrophysiology of the squid giant axon is considered to be well understood, that of the normal heart less so, and that of the heart under the inf luence of regional ischaemia even l ess . Attempts to corre late micro- scopic changes in electrophysiology during acute myocardial ischaemia with the macroscopic consequences of such changes - arrhythmias - have yet to provide a consensus. This is par t ly a resu l t of the d i f f i c u l t i e s in study- ing the electrophysiology of the heart and par t ly a resu l t of the v a r i a b i l - i t y in experimental preparations and approach. Nevertheless, despite a lack of convincing cor re la t ion between spec i f i c e lect rophys io log ica l changes in myocardial ischaemia and arrhythmias, much has been speculated concerning mechanisms of arrhythmogenesis. Recently, i t was demonstrated, using anaesthetised p igs , that the maxi- mum d i a s t o l i c potent ial (MDP) in subepicardial muscle c e l l s begins to f a l l p rec ip i tous ly within seconds of occlusion (Downar e t - a l ; , 1977b; Kleber et a l 1 9 7 8 ) . ' Simultaneously, the AP amplitude, d V / d t m a x and APD90 a l l decrease. The decrease in dV/dt v and APD90 are considered to be secondary to the f a l l in MDP, s ince , as discussed in deta i l prev iously , steady state d i a s t o l i c depolar isat ion leads to inact ivat ion of g N a » However, i t i s possible that ischaemia d i r ec t l y reduces g ^ a independently of MDP changes, although i t is not necessary to postulate such a mechanism. As a conse- quence of the reduction in APD90, the e f fec t i ve ref ractory period (ERP) decreases, since react ivat ion of g ^ a occurs more qu ick ly . However, th is s i tua t ion is not maintained. As MDP continues to f a l l ERP lengthens, des- p i te a maintained narrow APD90. Since ERP continues beyond the point of - 59 - f u l l r epo la r i sa t i on , th is phenomenon i s known as pos t - repo la r i sa t ion- re f rac - tor iness (Lazzara e t - a l ; , 1978). Since post - repo lar isa t ion- re f rac tor iness can be induced by depolar isat ion alone in iso la ted vent r icu la r t issue (Inoue e t - a l ; , 1984), i t fol lows that i t may be possible to explain th is phenomenon in terms of the normal vent r icu lar conductance channels (see below). I t i s important to note that the e lect rophys io log ica l changes described are not uniform in time or space. For example, adjacent c e l l s with very s imi la r res t ing E m may vary in ERP from 180 to 500 msec in pig ischaemic subepicardial muscle (Downar e t - a l . , 1977b). As a resu l t of the f a l l in d V / d t m a x , e f a l l s , producing charac te r i s t i c slow conduction in the ischaemic t i ssue . In add i t ion , depolar isat ion to between -55 and -60 mV is associated with complete inac t iva t ion of g ^ , and under these circumstances complete conduction block may occur. The phenomena described above were recorded in vivo from subepicardial t issue (Downar et a l . , 1977b; Kleber et a l . , 1978). I t i s not possible to record from deeper layers in -vivo, owing to the l im i ta t ions of i n t r a c e l l u l a r recording electrodes (electrodes are too f r a g i l e to be plunged deep into the myocardium and penetrate c e l l s ) . I t i s possible to record from iso la ted s t r ips of subendocardial t issue in v i t r o , but i t i s unclear to what extent the experimental technique inf luences the var iables under invest igat ion (not least because the t issue is general ly superfused, in which case i t i s d i f f - i c u l t to regulate the experimental ischaemia). A compromise preparation is the iso la ted Langendorff-perfusion rat heart in which careful i nc is ion exposes the sub-endocardial t i ssues , while coronary occlusion may be under- taken to produce regional ischaemia. Experiments carr ied out in our labor- atory using th is preparation have revealed s im i la r changes to those des- cr ibed above for subepicardial pig t issue (Inoue e t -a l . - , 1984). In addi - t i on , i t was shown that the changes were more severe and less revers ib le - 60 - with time in the deeper layer c e l l s versus surface c e l l s . The heterogeneity of ischaemic electrophysiology in time and space was c l ea r l y demonstrated in these experiments (Inoue et a l . , 1984). Qua l i t a t i ve l y , i t i s c lear that the act ive and passive e l e c t r i c a l prop- er t ies of vent r icu lar t issue may change with time in a f a i r l y reproducible manner during myocardial ischaemia, and attempts have been made to re la te th i s information to theoret ica l mechanisms of arrhythmogenesis, including reentry, automaticity and tr iggered automatic i ty. 1.3.4 Models Of Arrhythmogenesis 1.3.4.1 Reentry. A b r ie f descr ipt ion of some models of arrhythmo- genesis fo l lows. Reentry was o r i g i n a l l y described as reexc i ta t ion of the myocardium v ia a c i r cu la r route (Mines, 1913). The prerequis i te for reentry is one-way block of propagation in one limb of the reentrant c i r c u i t . The nature and mechanism of reentry, and the roles of re f rac to r iness , conduction ve loc i t y and duration of the 'wave of exc i t a t i on ' were e x p l i c i t l y described by Mines (1913) using the to r to ise heart , and re i tera ted by Schmitt and Ehrlanger (1928) using a t u r t l e heart preparat ion. The mechanism requires that the normal wave of exc i ta t ion brings Em in the blocked limb of the c i r c u i t c loser to threshold for conduction, such that the normal wave can then travel retrogradely up th is limb and reenter i t s previous pathway. For th is to happen, the t issue of the anterograde pathway must be exc i tab le at the time the wave front reenters. This c r i t e r i on w i l l be met i f one of two mechanisms operate. F i r s t l y , i f the c i r c u i t is long enough i t w i l l allow su f f i c i en t time to elapse for the anterograde pathway to recover e x c i t a b i l - i t y . This type of reentry is ca l led macroreentry. The pathway w i l l be func t iona l l y long i f conduction ve loc i t y is reduced, e i ther in the anterograde or retrograde l imb. A l t e rna t i ve l y , i f the re f rac - tory period in the anterograde pathway i s very short ( e x c i t a b i l i t y being - 61 - restored qu ick l y ) , then the requirement for a prolonged in terva l between exc i ta t ion and re-exc i ta t ion is reduced. In th is case the reentry c i r c u i t may be extremely short , perhaps only a few mm (Sasyniuk and Mendez, 1971). Indeed, transmural (ep i - to endocardial) reentry has recent ly been reported (Kramer et a l ; , 1985). This type of reentry i s ca l led microreentry. Acute myocardial ischaemia provides a l l the theoret ica l requirements for reentry: slow conduction, dispersion of re f rac tor iness and areas of 2-way conduction block around which a reentrant impulse may enc i r c l e , and of 1-way block through which retrograde conduction may occur ( e . g . , Janse, 1982). How does one recognise reentry experimentally ? The techniques employed for invest igat ion in open-chest animals have not been successfu l ly applied to c l i n i c a l or closed-chest animal s tudies. In open-chest animals an array of DC electrodes may be placed on the epicardium for mapping the act ivat ion pathways. By th is method, arrhythmias may be v isua l i sed (see paragraphs dealing with automatici ty, below). I t may be poss ib le , in the fu ture, to prepare experimental animals with in-dwel l ing DC electrode arrays. In the c l i n i c a l s i t ua t i on , or in closed-chest animals, standard 12 lead ECGs do not permit one to d is t inguish between reentry and automaticity (see below). Reentry, t heo re t i ca l l y , may give r i se to PVCs or VT. VT may degenerate to VF, in which case the mechanism of i n i t i a t i o n of the arrhythmia w i l l be los t to ana lys is , since VF is usual ly defined as uncoordinated chaotic e l e c t r i c a l a c t i v i t y (Moe e t - a l ; , 1964), without recognisable QRS complexes (Bigger, 1980). In closed-chest experimental animals undergoing repeated episodes of VT, i t has been suggested that the de l ivery of a premature e l e c t r i c a l stimulus to the ven t r i c le can terminate or i n i t i a t e the arrhythmia, i f the mechanism is reentry (Bigger and Goldreyer, 1970). The premature stimulus presumably e i ther conducts through the one-way blocked t issue to i n i t i a t e VT, or depol- - 62 - ar ises the one-way blocked t issue al lowing anterograde conduction of the normal impulse to terminate VT. The use of premature st imulat ion does not permit the recognit ion of reentrant PVCs, only r e l a t i v e l y sustained arrhyth- mias, such as VT. Trains of high frequency vent r icu lar st imulat ion w i l l often i n i t i a t e or terminate reentrant VT by the same mechanism as premature s t imula t ion. Such st imulat ion w i l l i n i t i a l l y capture the reentrant c i r c u i t , then, as frequency is increased, the ven t r i c le w i l l f a i l to fo l low. Once the stimulus is terminated, sinus rhythm w i l l resume. However, since the underlying condi- t ion is not changed, the arrhythmia w i l l shor t ly resume. Unfortunately, th is response is not unl ike the post-overdrive suppression response seen with automaticity (see below). Therefore, in closed chest animals i t i s d i f f i c u l t to prove reentry as the cause of VT. Furthermore, i t i s imposs- ib le to prove reentry as the underlying t r igger for VF, according to the above considerat ions. In open-chest animals, using ep icard ia l mapping techniques, the s i tua t ion is d i f ferent (see below). 1.3.4.2 Abnormal - Automatici ty. The understanding of abnormal automaticity as a mechanism of arrhythmogenesis requires the understanding of normal automatic i ty. During d ias to le , the membrane potent ial in normal vent r icu la r muscle is almost constant. However, in the sinus node a slow spontaneous d i a s t o l i c depolar isat ion precedes and t r iggers the propagated AP. Other t issues in the heart ( for example the a t r ioven t r i cu la r node and the Purkinje f ib res) also show th is slow depolar isat ion, but d V / d t m a x i s greatest in the sinus node. As a resu l t , the frequency of t r igger ing of propagated APs is highest in the sinus node. Consequently the propagated AP or ig inat ing from the sinus node overdrives the spontaneous depolar isat ions in other regions of the heart , and dictates the heart ra te . This cons t i t - utes the pacemaker property of the sinus node. The current responsible for the spontaneous d i a s t o l i c depolar isat ion i s known as the pacemaker current. In the sinus node the pacemaker current has recent ly been shown to resu l t from the in teract ion of two currents (Shibata and G i l e s , 1985), the decay of a delayed r e c t i f i e r K + current ca l led i K , and the act ivat ion of an ++ inward current carr ied by Ca , ca l led i c . The decay of i K i s h ighly voltage-dependent over the range -80 to -55 mV. The act ivat ion of i ^ a occurs at between -60 and -55 mV with peak current at 0 mV. Abnormal automaticity simply means that heart t issue other than the sinus node has taken over the pacemaker ro le of the sinus node (Hoffman and Dangman, 1982; Sasyniuk, 1984). Usual ly , in order for th is to occur, the rate of f i r i n g of the abnormal pacemaker region must exceed that of the sinus node and overdrive the heart (but see comments concerning parasystole, below). The anatomical source of th is pacemaker a c t i v i t y is known as the ectopic focus. Abnormal pacemaker a c t i v i t y may occur as a resu l t of an increase in the steepness of the slope of the d i a s t o l i c depolar isat ion in latent pacemaker t i s sue , or a reduction in the threshold for generation of the (abnormal) propagated AP. The l a t t e r may resu l t from ei ther a reduction in the absolute threshold in mV, or from a sh i f t in the maximum d i a s t o l i c potent ial (MDP) to a more pos i t ive value. The d i f f i c u l t i e s in recognising reentry in closed-chest animals also apply to the recognit ion of abnormal automatici ty. Automaticity may be associated with intermit tant ex i t block whereby the ectopic impulse f a i l s to propagate (Fisch et a l . - , 1971). This w i l l occur i f the ectopic focus d i s - charges jus t af ter the normal wave of exc i ta t ion has passed by (leaving a band of re f rac tory t i s sue ) . Under such circumstances one w i l l see, perhaps, only PVC in the ECG. A l te rna t i ve l y , the ectopic focus may be protected, by entrance block, from overdrive suppression by the normal wave of exc i ta t ion (Wennemark and Bandura, 1974), in which case automatic arrhythmias may occur - 64 - even when the cycle length of the abnormal pacemaker is longer than the sinus cycle length. Theore t ica l l y , entrance block may be caused by a band of t issue which possesses one-way blocking proper t ies, or by anatomical abnormal i t ies. Conceivably, entrance block may also resu l t from the induc- t ion of a band of ref ractory t issue by the previous ectopic impulse. If the normal sinus frequency and the entrance-block-protected ectopic frequency are s im i l a r , the normal and abnormal pacemakers may al ternate in the i r dominance, producing parasystole. I t has been suggested that abnormal automaticity and reentry may be d i f fe rent ia ted by studying the ef fects of overdrive pacing (see Bigger and Goldreyer, 1970; Vasa l l e , 1977). If an automatic arrhythmia is overdriven by pacing, then fo l lowing termination of overdrive one w i l l see a delay (post-overdrive suppression), then a resumption of the automatic arrhythmia. Reentrant arrhythmias do not show typ ica l post-overdrive suppression. The mechanism of post-overdrive suppression has been studied in dog Purkinje f ib res (Vasal le , 1977). Ear ly during the onset of overdrive pac- ing , MDP f a l l s . Providing that st imulat ion does not exceed a c r i t i c a l frequency, maximum d i a s t o l i c depolar isat ion then r i ses and the slope of the pacemaker potent ia l is reduced. Immediately upon cessation of overdr ive, the ectopic frequency i s lower than i t had been before overdrive pacing, as a resu l t of the ef fect of overdrive pacing on the slope of the pacemaker po ten t ia l . Over a short priod of time the or ig ina l e lect rophys io log ica l charac te r i s t i cs of the ectopic focus re turn, and the arrhythmia resumes. In the case of reentry there i s general ly no post-overdrive suppression, and the reentrant arrhythmia resumes after an unpredictable i n t e r v a l . Neverthe- l e s s , i t i s possible that reentrant arrhythmias may resume, fo l lowing over- d r i ve , af ter an in terval s im i la r to that seen fo l lowing post-overdrive suppression of abnormal automatici ty. Therefore a reentrant arrhythmia may - 65 - be incor rec t ly classed as automatic. In add i t ion , parasysto l ic automati- c i t y , protected from overdrive by entry block, w i l l not exh ib i t post-over- dr ive suppression, and may therefore be incor rec t l y classed as reentrant. As discussed above, overdrive techniques are only useful for assessing r e l a t i v e l y sustained arrhythmias. However, i t i s possible to gain ins ight into arrhythmogenic mechanisms ( including mechanisms of PVC induction) from ECG recording, coupled with simple physio logical manipulations. In th is regard, reentrant PVCs are tr iggered i n i t i a l l y by the normal wave of propa- gat ion. Therefore, by analysing the frequency of PVC in re la t ion to heart ra te , i t may be possible to estab l ish a re la t i onsh ip . Furthermore, heart rate may be slowed or stopped by st imulat ing the vagal e f fe rents , or acce l - erated by administering atropine ( e t c . ) , such that PVC incidence may be measured over a wide range of heart ra te . Theore t ica l l y , PVCs resu l t ing from automaticity should not vary in frequency with changes in heart ra te , whereas the frequency of reentrant PVCs should increase with tachycardia, and f a l l with bradycardia. However, the study of the ef fect of changes in heart rate on PVC incidence i s not an i n f a l l i b l e means of d iagnosis, owing to the p o s s i b i l i t y of entry- and ex i t -b lock . In add i t ion , i t i s necessary to ensure that the manipulation designed to a l te r heart rate does not i n f l u - ence other var iab les , d i r ec t l y or i n d i r e c t l y , leading to an a l te ra t ion in PVC frequency. It i s almost impossible to control for th is p o s s i b i l i t y . I t i s possible to make cer ta in speculations concerning the mechanism of arrhythmogenesis of PVCs by simply observing the ECG without other manipula- t i o n . In th is regard, sustained f ixed coupling of PVCs (of consistent QRS conf igurat ion) with the normal sinus beats is regarded as ind ica t ive of reentry rather than automaticity (Langendorf and P ick , 1967). The reason for th is i s that the frequency of an ectopic arrhythmia would not be expec- ted to be the same as the sinus frequency, whereas a reentrant arrhythmia - 66 - should be l inked to the sinus cycle in a f ixed manner. I t i s questionable whether any of the tests and experiments described are useful for d is t inguish ing between reentry and automaticity during acute myocardial ischaemia, despite the i r possible usefulness in other circumstan- ces. During the development of myocardial ischaemia, the e lect rophysio log- i ca l charac te r i s t i cs of a hypothetical reentrant c i r c u i t w i l l change with time, causing changes in coupling i n t e r v a l . Therefore, reentry in acute myocardial ischaemia may be incor rec t l y described as automaticity i f the c r i t e r i on for diagnosis i s a f i xed coupling i n t e r v a l . In add i t ion , i t i s possible that parasysto l ic (entry-blocked) foc i may be influenced by the wave of normal exc i ta t ion v ia e lec t ro ton ic intercourse ( re f lec t ion) across an area of entry block ( J a l i f e and Moe, 1976). F i n a l l y , in th is regard, the funct ional d i s t i nc t i on between reentry and automatici ty becomes questionable when the phenomenon of tr iggered automaticity i s considered. 1.3.4.3 Triggered Automat-i c i t y . Triggered automaticity has only been demonstrated in v i t r o . The reason for th is is quite simple. In order to demonstrate tr iggered automaticity i t i s necessary to stimulate quiescent t i s sue , and i t is not possible to have quiescent ven t r i c les i n -v i vo . Since tr iggered automaticity i s , by d e f i n i t i o n , automaticity which is i n i t i a t e d by an exogenous source, and the exogenous source theore t i ca l l y includes the normal wave of depo lar isa t ion, then i t is possible that abnormal automatic- i t y in vivo may be tr iggered by the sinus beat (Spear and Moore, 1982). In such a case the frequency of automatic PVC would corre la te with heart ra te . Therefore, i t i s not possible to d is t inguish between automaticity and re - entry on the basis of the re la t ionsh ip between heart rate and PVC frequency unless tr iggered automaticity i s ruled out. However, i t would be necessary to induce asystole in order to test for t r iggered automaticity in v i vo , and th is i s not f eas ib l e . - 67 - 1.3.5 Epicard ia l Act ivat ion Mapping Although i t i s not possible to prove whether an arrhythmia is automatic or reentrant in closed-chest animals, the technique of ep icard ia l mapping allows for v i sua l i sa t i on of conduction pathways in anaesthetised open-chest animals and i n - v i t r o . Epicardia l mapping has been car r ied out in conjunc- t ion with simultaneous i n t r ace l l u l a r recording in acute myocardial ischae- mia, and attempts have been made to re la te the pattern of abnormal e lec t ro - genesis and conduction with the underlying e lectrophysio logy. How do the e lect rophysio log ica l changes in acute myocardial ischaemia corre la te with arrhythmogenic mechanisms v isua l i sed by ep icard ia l mapping ? In coronary-occluded blood-perfused dog and pig hearts, records of reentry, but not automatic i ty, were observed (Janse e t - a l ; , 1980; Janse and Kleber 1981; Janse, 1982). Most of these arrhythmias arose from the normal t issue close to the occluded bed within the f i r s t few minutes of occ lus ion. The time course of arrhythmias corresponded reasonably well with the time course of ischaemia-induced arrhythmias in - vivo in dogs and pigs ( e . g . , Bergey et a l . , 1982; 1984). In teres t ing ly , the i n i t i a l t r igger for the reentrant VT was an impulse which arose from the normal t i ssue . The mechanism proposed to account for th is phenomenon was as fo l lows. By comparing the AP of normal c e l l s with that of ischaemic c e l l s i t was found that many ischaemic c e l l s were in the plateau phase of the cycle when normal c e l l s were at the rest ing repolar ised phase. This s i tua t ion occurred as a resu l t of the heterogeneity of conduc- t ion ve loc i t y in the ischaemic t issue versus the non-ischaemic t i s sue , in associat ion with slow conduction in the ischaemic t i s sue . The consequences of th is s i tua t ion included a large potent ial d i f ference between the normal and the ischaemic t i ssue . It was proposed that a flow of in jury current between the depolarised ischaemic t issue and the polar ised non-ischaemic - 68 - t issue may have been su f f i c i en t to re-exc i te the non-ischaemic t i ssue . A l te rna t i ve l y , since the non-ischaemic t issue could be in the plateau phase of the AP when the ischaemic t issue was repo la r ised , in jury current may have flowed in the opposite d i rec t ion to re -exc i te the ischaemic t i s sue . For th is mechanism to operate, i t i s necessary that a layer of c e l l s ex is ts (separating the current source and sink) which i s inexc i tab le , at least temporari ly, and capable of permitt ing e lec t ro ton ic coupling between the source and sink (Janse and Kleber, 1981). Whatever the mechanism of i n i t i a t i o n , ac t iva t ion maps revealed fragmen- tat ion and fusion of wavefronts, one- and two-way block and reentry. Reent- ry was associated with VT. VF was associated with mul t ip le wavelets t r a v e l - ing slowly along tortuous routes among mul t ip le i s l e t s of conduction block which changed posi t ion and magnitude from moment to moment. VT ei ther terminated as a resu l t of major s ingle wavefronts a r r i v ing at a large area of inexc i tab le t i s sue , or degenerated to VF as major wavefronts s p l i t into subsidiary wavefronts. VF did not terminate spontaneously (within the 10 sec allowed before d e f i b r i l l a t i o n was applied in th i s study). I t was sug- gested that th is was because VF was associated with more wavefronts than VT, therefore i f some wavefronts terminated, others remained to perpetuate the arrhythmia. This is a possible explanation for the higher frequency of spontaneously revert ing VF in small animals such as rats (Kenedi and Loscon- c i , 1973a; Clark et a l . , 1980) compared with larger animals such as dogs (Townshend Por ter , 1894; Smith, 1918). Presumably the larger hearts have more wavefronts in VF than smaller hearts, and consequently express a lower p robab i l i t y of spontaneous d e f i b r i l l a t i o n . I t was found that st imulat ion of the l e f t s t e l l a t e ganglion improved conduction in the ischaemic t i s sue , and increased the incidence of VT and VF. I t i s possible that th is enhanced sympathetic a c t i v i t y increased e x c i - - 69 - t a b i l i t y in the ischaemic t i ssue , f a c i l i t a t i n g the conduction of PVCs and thei r degeneration into VT and VF (Janse and Kleber, 1981). High concentra- t ions of l idocaine increased the proportion of the ischaemic t issue which was inexc i tab le at any given moment, thereby causing termination of VT and prevention of progression to VF. To summarise, of the major theoret ica l mechanisms of arrhythmogenesis, only reentry has been demonstrated in acute myocardial ischaemia. This evidence was provided from epicard ia l mapping experiments. However, reentry appeared to be tr iggered in many cases by an ectopic in jury current a r is ing from normal Purkinje f i b r e s . The prerequis i te for induction of an ectopic impulse appeared to be a dispersion of re f ractor iness between the normal and the ischaemic t i ssue . This heterogeneity was produced by the slow conduc- t ion in the ischaemic t i ssue . Abnormal slow conduction and var iab le conduc- t ion block in the ischaemic t issue appeared to be responsible for the con t i - nuation of the arrhythmias and the i r progression into VT and VF. Conduction block was made worse by l i doca ine , such that arrhythmias terminated soon af ter i n i t i a t i o n , while l e f t s t e l l a t e ganglion st imulat ion improved conduc- t ion (lessened block) and worsened the arrhythmias. Do these phenomena occur in rats fol lowing coronary occlusion ? Our laboratory has demonstrated ef fects of occlusion in -v i t ro on i n t r a c e l l u l a r potent ia ls (Inoue et a l . , 1984) s im i la r to those demonstrated in other species. In add i t ion , ind i rec t evidence in v ivo such as a reduction in ref ractory period (Northover, 1986) and a lack of e f fec t of vagal s t imula- t ion-induced bradycardia on arrhythmia incidence (Mertz and Kaplan, 1982) suggests that the arrhythmias occurring during the 30 min period af ter occlusion are more l i k e l y to be reentrant than automatic. A question which ar ises i s , to what extent does i ' N a contribute to arrhythmogenesis compared with i $ i ? The ef fects of l idocaine and l e f t - 70 - s t e l l a t e st imulat ion (Janse, 1982) suggest that both may be involved. 1.3.6 The slow inward current and arrhythmogenesis in acute ischaemia 1.3.6.1 Introduction As discussed previously , in normal vent r icu lar t issue the i n i t i a l up- stroke of the AP (phase-0) is governed by i N a , and the plateau by i .. The changes in phase-0 and rest ing E m produced by myocardial ischaemia, as described above, may represent one of four e lect rophys io log ica l phenomena. F i r s t l y , the reduction in dV/dt may represent a severely depressed Hid A ^Na' 1 S because t n e depolar isat ion associated with acute myocardial ischaemia w i l l cause inac t iva t ion of i N a . Secondly, the upstroke may represent an enhanced i $ ^ . This i s because depolar isat ion to -60 to -65 mV w i l l bring E m c loser to the threshold for i . without inducing J m s i steady-state inac t iva t ion of i s - (o^ should be s im i la r to that at an MDP of -90 mV), whereas i ^ a w i l l be almost completely inact ivated at th is vol tage. Th i rd ly , the upstroke may involve both depressed i ^ a and en- hanced i $ . j . F i n a l l y , the upstroke may be associated with an abnormal current system which possesses elements of i ^ a and/or i s l - , but perturbed by factors of ischaemia other than depolar isa t ion, such as increases in pH ( I i j ima e t - a l . , 1986) and the products of anaerobic metabolism (Corr et a l ; , 1984). 1.3.6.2 How - might- - j ,. - contr ibute to -arrhythmogenesis ? The e lec - t rophys io log ica l changes produced by acute myocardial ischaemia (Downar et a l . , 1977a; 1977b; Kleber et a l ; , 1978), and the attendant hypothetical arrhythmogenic mechanisms (reentry and automatic i ty, e tc . ) reveal that i $ ^ may play an important ro le in arrhythmogenesis (Hauswirth and Singh, 1978). The pr inc ipa l reasons for such a suggestion are as fo l lows. a . Acute myocardial ischaemia is associated with depolar isat ion (Downar e t - a l . , 1977b; Kleber e t - a l . , 1978; Inoue et a l . , 1984) to E m values close to those producing complete inact iva t ion of i ^ a (Beeler and Reuter, 1970a). However, from the known voltage-dependence of i . (Reuter, 1973), i t would be expected that th is depolar isat ion should not reduce i . , but rather enhance i t by v i r tue of a reduction in threshold for ac t i va t i on . b. Acute myocardial ischaemia is characterised by slow conduction in the ischaemic t issue (Downar et a l ; , 1977b; Kleber e t -a l . - , 1978). This slow conduction can be accounted fo r , almost completely, by the f a l l in dV/dt of the upstroke of the ischaemic AP (Dodge and Cranef ie ld , 1982); max r 3 the increase in longi tudinal i n t e r ce l l u l a r resistance associated with increases in intercalated disc resistance may also contr ibute (Weidmann, 1982). Slow conduction in depolarised Purkinje t issue has been shown to be dependent on C a + + (Cranef ield e t a l . , 1972), From considerations of res t ing E m , as discussed above, the current responsible for the upstroke of the slow conduction AP may be i •. In th is regard, slow conduction in Purkinje t issue in -v i t ro has been shown to be inh ib i ted by (^-verapami l at concentrations below those which i nh ib i t i ^ a (Cranef ield e t a l ; , 1974). c . Acute myocardial ischaemia i s associated with a large increase in ex t race l l u la r K + concentration ( [K + ] Q ) (H i l l and Gettes, 1980; Hirche et a l . , 1980). While th is phenomenon may contr ibute to the depolar isat ion occurr ing in acute myocardial ischaemia and i t s attendant consequences (see above), i t may also serve to release noradrenaline from sympathetic nerve endings (Hirche e t - a l ; , 1985). By act iva t ing adrenoceptors, noradren- a l ine i nh ib i t s inact ivat ion of i s i (Bean et a l . , 1984). This arrhythmo- genic mechanism is highly specula t ive, however, and recent evidence from our laboratory (Botting et a l . , 1983) and elsewhere (Daugherty e t -a l . - , 1986) does not support a major ro le for catecholamines in arrhythmogenesis in acute myocardial ischaemia. Therefore, i f i' s^ is involved in arrhythmo- genesis, th is involvement is essen t ia l l y independent of the modulating inf luence of adrenoceptor act ivat ion and antagonism on i . . d. Although current evidence does not support a ro le for automatici ty in arrhythmogenesis in acute myocardial ischaemia (Janse, 1982; Janse and Kleber 1981; Mertz and Kaplan, 1982; Northover, 1986), th is p o s s i b i l i t y i s not precluded. If abnormal automaticity i s the expression of enhanced latent pacemaker currents in Purkinje t issue (Spear and Moore, 1982), then i t is possible that i . , or a s imi la r current plays an important ro le in arrhythmogenesis. This is because the pacemaker current in the sinus node ++ comprises, in part , of i*£a, a Ca current s im i la r to i $ ^ (Shibata and G i l e s , 1985), and phase-0 depolar isat ion is carr ied almost exc lus ive ly by i s i (Yangihara e t - a l . , 1980; Brown, 1982). e. Act ivat ion mapping experiments have shown that a d ispar i t y in AP phase between the normal and ischaemic t issue generates in jury currents which t r igger reentry (Janse, 1982; Janse and Kleber 1981). The pr inc ipa l determinant of such t r igger ing i s the occurrence of an AP plateau adjacent to exc i table t i ssue . The plateau of the AP in normal (and probably also in ischaemic) vent r icu lar t issue is governed by i • (Reuter, 1973). The AP plateau in non-ischaemic t issue i s much longer than the plateau in ischaemic t issue (Downar et a l ; , 1977b; Kleber et a l . , 1978; Inoue e t - a l ; , 1984). I t fo l lows that the longer the plateau in non-ischaemic t i s s u e , the higher the probab i l i t y of t r igger ing an e lect ro ton ic re -exc i ta t ion of the ischaemic t i s sue . Therefore, i • may be involved in arrhythmogenesis by v i r tue of i t s ro le in creating a dispersion of AP plateau (and consequently a d isper- sion in re f rac tor iness) between the ischaemic and non-ischaemic t i ssue . f . I t i s possible to hypothesize other mechanisms by which i .. i s instrumental in i n i t i a t i n g or maintaining arrhythmias in acute myocardial ischaemia. Only one further p o s s i b i l i t y w i l l be discussed here, and th is concerns the hypothetical mechanism of arrhythmogenesis known as tr iggered automatic i ty. From i n - v i t r o s tud ies , tr iggered automaticity has been l inked to the occurrence of o s c i l l a t o r y a f terpotent ia ls (Cranef ie ld , 1977), also known as ear ly or delayed af ter -depolar isat ions (DADs), depending on whether they occur before or af ter f u l l r epo la r i za t i on , respect ive ly . While there is no evidence to support a ro le for tr iggered automaticity in arrhythmogen- es is in acute myocardial ischaemia, i t has been suggested that DADs and tr iggered a c t i v i t y occur in surv iv ing Purkinje t issue excised from the occluded vent r icu lar regions of dog hearts fo l lowing the development of in fa rc t ion (El Sher i f e t - a l ; , 1982). The involvement of i g i in DADs has been suggested by the observations that DADs are enhanced by ra i s ing ex t ra- c e l l u l a r C a + + , occur in depolarised t issue (espec ia l l y t issue depolarised by cardiac g lycos ides) , and that DADs are inh ib i ted by compounds which i nh ib i t i . (Ferr ier and Moe, 1973). However, other evidence suggests that DADs are generated by the t ransient inward current (Lederer and Ts ien, 1976), a current which appears to be d i s t i nc t from i $ ^ . Recently i t was shown that simulated ischaemia (16.2 meq/1 K + , low pH, lactate and hypoxia) abolished DADs induced by high frequency pacing in Purkinje t issue in v i t r o (Opie e t - a l ; , 1986; Coetzee e t a l . , 1986), sugges- t ing that th is mechanism of arrhythmogenesis in acute myocardial ischaemia may not be important. 1.4- The-pharmaeology-of caleium-antagonists 1.4.1 Def in i t ion Calcium antagonists were o r i g i n a l l y described as drugs which produce the i r pharmacological e f fects predominantly by i nh ib i t i ng vol tage-act ivated 2+ 2 + entry of Ca in a manner which can be inh ib i ted by Ca (Fleckenstein et a l ; , 1969). This de f in i t i on has been extended to include drugs which 2+ 2+ i nh ib i t Ca entry resu l t ing from the opening of Ca channels coupled - 74 - with drug receptors (see Jam's and Tr igg le , 1983), and drugs which i nh ib i t 2+ i n t r a c e l l u l a r Ca -dependent processes such as binding with calmodulin (Rahwan, 1983; Lynch and Rahwan, 1982). The expansion of the de f in i t i on has occurred in conjunction with the introduct ion of new terms such as 'slow channel b lockers ' and 'calcium entry b lockers ' (see Nayler, 1983). The term calcium antagonist i s used here in accordance with F leckenste in 's o r ig ina l de f in i t i on (see above and F leckenste in , 1983 for review). 1.4.2 Pharmacology of phenethylalkylamines and 1,4-dihydropyridines Phenethylalkylamine calcium antagonists include verapamil, ga l lopami l , b e p r i d i l , t i apami l , anipamil and a var ie ty of t r i a l preparations such as D888. 1,4-dihydropyridines include n i fed ip ine , fe lod ip ine , nimodipine, and n i t rend ip ine. Verapamil, gallopamil and n i fed ip ine have been studied in the most deta i l (see Henry, 1979; 1980; Tr igg le , 1981; F leckenste in , 1983; Nayler and Horowitz, 1983). Phenethylalkylamines have been suggested to possess a higher a f f i n i t y for the channel associated with i • when the channel is in the open or inact ivated state (Kohlhardt and Mnich, 1978; Pelzer e t - a l . , 1982; Lee and Ts ien, 1983), whereas 1,4-dihydropyridines are equipotent in rested and open channels (Lee and Tsien, 1983). This di f ference may account for the w e l l - known frequency-dependence of phenethylalkylamines (Sanguinetti and West, 1982) compared with 1,4-dihydropyr idines, which possess l i t t l e or no f r e - quency-dependence, but considerably more rest ing block (Woods and West, 1983; 1985; Hachisu and Pappano, 1983). Both 1,4-dihydropyridine and phen- ethylalkylamine calcium antagonists have been suggested to act by slowing the recovery from inact ivat ion of the i • channel (Lee and Tsien, 1983). Studies with permanently charged analogues of gallopamil (Hescheler et a l ; , 1982), and studies carr ied out using 'sk inned' (plasmalemma-free) smooth-muscle and cardiac t issue (Fleckenste in, 1977; Kreye e t - a l ; , 1983; - 75 - Itoh e t a l ; , 1984) have indicated that both phenethylalkylamine and 1 ,4-d i - hydropyridine calcium antagonists produce the i r pharmacological ef fects v ia an action on the inner surface of the plasmalemma, and that i n t r ace l l u l a r ?+ actions such as inh ib i t ion of Ca binding with calmodulin (S i lver at a l , 1984) do not contr ibute to the ef fect of these drugs on exc i tab le t i s sue . Both phenethylalkylamines and 1,4-dihydropyridines possess a wide spec- trum of pharmacological proper t ies, such as i nh ib i t i on of i ^ a (Bayer et a l . , 1975a; Nawrath e t - a l . , 1981; Yatani and Brown, 1985), i nh ib i t i on of co r t i cos te r iod release from the adrenal cortex (Costa et a l ; , 1983), i n h i b i - t ion of myosin l i gh t chain kinase a c t i v i t y (Movsesian e t - a l . , 1984), e tc . However, consideration of the concentrations required to produce these —6 ef fects (general ly 10 M or more) has lead most invest igators to regard these ef fects as r e l a t i v e l y unimportant compared with ef fects on vascular smooth-muscle and myocardial t issue resu l t ing from inh ib i t i on of vol tage- operated calcium entry (see Nayler and Horowitz, 1983; F leckenste in, 1977; 1983; Henry, 1979; T r igg le , 1981; 1982). It is of in terest to determine the mechanism of any action of a drug. In order to ascribe an ef fect of a drug to calcium antagonism a var ie ty of approaches may be taken, since i t is often impossible to d i r ec t l y measure calcium entry through voltage-operated channels at the same time as other var iab les , espec ia l l y in -v ivo. Of fundamental importance is information concerning the concentration range over which calcium antagonism is produ- ced; i dea l l y one would l i k e to know EC^Q values. For example, i f one wished to know whether the antiarrhythmic actions of verapamil and n i f ed ip - ine were produced by calcium antagonism in the ven t r i c l es , one might app- roach the problem by comparing E D 5 Q values for reductions in arrhythmias with EDgQ (or EC^Q ) values for calcium antagonism in ven t r i c l es . In order to support th is information i t might be considered of in terest to - 76 - invest igate the re la t i ve calcium antagonist potency of these drugs in a var ie ty of t issues under a var ie ty of condi t ions. There are a number of studies which have compared several calcium antagonists for the i r re la t i ve potencies in vascular and cardiac muscle, in vivo and in v i t r o . In order to i l l u s t r a t e some of the problems inherent in determining whether the ef fect of a drug may be at t r ibuted to calcium antagonism, some studies in which n i fed ip ine and (^-verapami l were compared are summarised. EC^Q or ED^Q values were e i ther taken d i r ec t l y from the text of the publ icat ions or approximated by in terpolat ion (or extrapolat ion) of the data presented. 2+ ECgg values for i nh ib i t i on of depolar isat ion/Ca -dependent contrac- t ion in vascular smooth muscle range from 3 x 10 - ^ M (Kenakin and Beek, 1985) to 2.4 x IO" 8 M (Mi l la rd e t - a l . , 1983) for (^)-verapamil, and from 1.5 x 10" 8 M (Mi l la rd e t - a l ; , 1983) to 6.3 x I O - 9 M (Kenakin and Beek, 1985) for n i fed ip ine . In a l l of the studies considered (Fleckenstein-Grun et a l . , 1976; Lee e t - a l ; , 1983; Kenakin and Beek, 1985; M i l l a rd e t - a l ; , 1983; Nakayama et a l ; , 1985) n i fed ip ine was more potent than (±)-verapami1. However, the di f ference in potency varied considerably, from 1.7:1 (Lee et a l ; , 1983) to 350:1 (Fleckenstein-Grun e t -a l . - , 1976). This may re la te to the var ie ty of preparations, test condit ions and methods used. In heart preparations (vent r ic les and a t r ia ) a s im i la r survey reveals an even greater va r ia t i on . EC^Q values for (^)-verapamil have been reported _5 to range from 10 M (Nabata, 1976; Raschack, 1976a; Briscoe and Smith, 1982) to 1.5 x 10~ 8 M (Clarke et a l . , 1984a; 1985). Corresponding E C 5 Q for n i fed ip ine range from 7.2 x 10~ 6 (Briscoe and Smith, 1982) to _ Q 5.8 x 10 M (Nabata, 1976). In a t r i a and vent r i c les n i fed ip ine has been reported to be as much as 62.5 times as potent as (±)-verapamil (Raschack, 1976a), but other invest igators have found (^)-verapami1 to be up to 5 times as potent as n i fedip ine (Clarke e t - a l ; , 1985). Ni fedipine was found to be - 77 - more potent than (±)-verapamil in 4 studies (Clarke e t - a l ; , 1984; 1985; Lee et a l ; , 1983; M i l l a rd e t -a l . - , 1983), whereas (±)-verapami 1 was found to be more potent than n i fedip ine in 7 studies (Raschack, 1976a; Nabata, 1976; Briscoe and Smith, 1982; Winslow e t - a l . , 1983; Kenakin and Beek, 1985; Nakayama et a l . , 1985). The data d iscussed above i l l u s t r a t e s that i t i s dangerous to re l y on published reports of the potency and re la t i ve potency of calcium antagonists with regard to the invest igat ion of the mechanism of action of calcium antagonists (as ant iarrhythmics, for example). 1.5 Aims of-studies 1.5.1 The action of calcium antagonists in acute myocardial ischaemia The pharmacologist functions to examine and describe the propert ies of drugs, and then account for the mechanism(s) underlying these e f fec ts . In acute myocardial ischaemia, the ef fects of calcium antagonists (Flecken- s t e i n , 1969) are not well character ised. The f i r s t report of the actions of a calcium antagonist in acute myocar- d ia l ischaemia appeared in 1968, when Kaumann and Aramendia demonstrated that 0.79 mg/kg (±)-verapamil abolished arrhythmias and prevented death when administered 10 min before coronary occlusion in anaesthetised dogs. Since th is time there have been a number of reports confirming the antiarrhythmic a c t i v i t y of (^-verapami 1 and d i l t iazem in dogs, although n i fedip ine has general ly not been found to be antiarrhythmic ( e . g . , Guelker et a l ; , 1983; Kobayashi et a l . , 1983; Clusin e t - a l . , 1984). Most of these studies used low doses of calcium antagonists (0.1 - 0.5 mg/kg (±)-verapamil , 0.04 - 0.08 mg/kg n i fed ip ine ) . The resu l ts were highly var iab le between studies, and a c lear exposi t ion of the antiarrhythmic actions of calcium antagonists, and possible mechanism(s) of action remained e lus i ve . In the development and character isat ion of the conscious rat preparation - 78 - by our laboratory, i t was found that (^-verapami 1 and quinidine appeared to have antiarrhythmic a c t i v i t y (Johnston e t - a l ; , 1983a). I t was decided, therefore, to invest igate in more deta i l the classes of antiarrhythmics of which these drugs represent, beginning with c lass-4 (calcium antagonist) antiarrhythmics such as (^-verapami 1 (Singh and Vaughan Wi l l iams, 1972), using the conscious rat preparat ion. The experiments with calcium antagonists described in the Methods sec- t ion were designed to answer the fo l lowing questions: a. Do calcium antagonists reduce arrhythmias induced by acute myocar- d ia l ischaemia in conscious ra ts? b. Do calcium antagonists reduce in fa rc t s ize fo l lowing permanent coronary occlusion? c. Do these ef fects ( i f any) occur as a resu l t of calcium antagonism? d. What factors determine these e f fec ts? It was expected that by carrying out experiments b l ind and assembling dose-response curves for a range of va r iab les , using several preparations ( in vivo and in -v i t ro ) i t would be possible to provide unequivocal answers to questions a and b. In add i t ion , by considering the dose-dependence of the ef fects of d i f ferent types of calcium antagonists on the same var iab les , i t was hoped that c ircumstant ial evidence in support of the hypothesis described in question c, or unequivocal disproof of th is hypothesis would be provided. By considering the overal l resu l ts of the experiments, specula- t i ve hypotheses concerning question d were expected to be generated. 1.5.2 Arrhythmogenesis in acute myocardial ischaemia The introduct ion has broached the subject of arrhythmogenesis in acute myocardial ischaemia. An understanding of the mechanism(s) of arrhythmoge- nesis i s a necessary adjunct to the understanding of the mechanism of action of benef ic ia l drugs. The experiments carr ied out as part of th is thesis - 79 - were not designed to provide a de f i n i t i ve statement concerning the determin- ants of arrhythmogenesis in acute myocardial ischaemia. However, in view of the con f l i c t i ng evidence concerning the ro le of the autonomic nervous system in arrhythmogenesis, and the re la t ionsh ip between i ^ and adrenoceptors (Bean et a l ; , 1984), an attempt was made to resolve th i s par t i cu la r question by carry ing out a series of graded ablat ions in the CNS. In r a t s , evidence concerning the ro le of the sympathetic nervous system in arrhythmogenesis is contradictory. Campbell and Parrat t (1983) found that a var ie ty of s-adrenoceptor antagonists reduced arrhythmias induced by occlusion in anaesthetised ra t s . Marshall e t - a l ; (1981a) found that isopre- nal ine infusion increased the sever i ty of arrhythmias while adrenaline and noradrenaline reduced the sever i t y . However, in conscious ra t s , our labora- tory found that propranolol , labeta lo l and chemical sympathectomy did not inf luence arrhythmias (Botting et a l ; , 1983). Therefore experiments were carr ied out in order to answer the fo l lowing questions: a. Does the autonomic nervous system play an independent ro le in arrhythmogenesis fol lowing coronary occlusion in rats? b. Is the ro le of the autonomic nervous system of su f f i c i en t magnitude for i t to const i tute a major determinant of arrhythmogenesis? The use of se r i a l ablat ion in the CNS coupled with se lec t ive replacement of catecholamines was chosen as a unique method for answering the questions out l ined without resor t ing to the use of drugs as too l s , a pract ice which often involves un jus t i f ied assumptions concerning the s p e c i f i c i t y and se lec - t i v i t y of drug ac t ion. If the autonomic nervous system (pa r t i cu la r l y the sympathetic nervous system) plays a neg l ig ib le ro le in arrhythmogenesis i t is important to estab l ish th is f ac t . Such a resu l t has important consequen- ces, since i t would be d i f f i c u l t , in th is event, to j u s t i f y the use of B - a d - renoceptor antagonists as antiarrhythmics in the prophylaxis of sudden death. - 80 - 2 METHODS 2.1 Goronary-occlusion-in rats 2.1.1 Overview A br ie f h i s t o r y . o f coronary occlusion in rats was given in the Intro- duct ion. A conscious animal preparation was developed by our laboratory in order to remove the po ten t ia l l y confounding inf luences of anaesthesia and recent major surgery from the experimental arena (Johnston et a l ; , 1983a). I t i s our opinion that when invest igat ing the actions of drugs or the nature of a disease process, the i n i t i a l experiments should be car r ied out using conscious unrestrained preparations, f ree from ' a r t i f i c i a l ' constraints such as anaesthesia. In the case of the conscious rat model of myocardial ischaemia, th is approach is consistent with the fact that the human con- d i t ions of myocardial ischaemia general ly occur in the absence of anaesthetic and concurrent surgery. The choice of animal species in ischaemia studies and the human spectrum of disease were discussed in some deta i l in the Introduct ion. It was stressed that c l i n i c a l l y , information concerning the c r i t i c a l f i r s t h af ter the onset of symptoms (when sudden death and vent r icu la r arrhythmias are at a premium), and information concerning the re la t ionsh ip between the extent of ischaemia, i t s time-course and i t s sequelae ( p a r t i c u l a r l y arrhythmias) are both incomplete at best. Therefore, arguments concerning c l i n i c a l r e l e - vance are essen t ia l l y po in t less , since no human template of myocardial ischaemia and in fa rc t ion e x i s t s . The philosophy behind the use of the conscious rat preparation i s based on a simple premise, namely that the pr inc ipa l aim of a myocardial ischaemia model should be the reproducible and unequivocal production of myocardial ischaemia of known sever i ty and reproducible sequelae, in a manner which can be manipulated simply, whereby the preparation serves as a bioassay. The - 81 - ra t may be the only species which meets a l l these requirements (dogs, ham- sters and guinea-pigs are too va r iab le , pigs are too la rge, primates are too expensive and mice are too smal l ) . The fo l lowing chapters attempt to co l la te a l l the methods which have been described, in part , in various publ icat ions from our laboratory (Au e t - a l ; , 1979a; 1979b; Johnston et a l ; , 1983a; 1983b; Bott ing et a l ; , 1983; Cur t is et a l . , 1984; 1985a; 1985b; 1986a; 1986b; Cur t is and Walker, 1986a; 1986b), in an e x p l i c i t manner which would permit the reader to reproduce the techniques without further reading or ass is tance. 2.1.2 Preparation Before preparat ion, a l l instruments, leads, l ines and occluders were placed in a bath of 70% ethanol in d i s t i l l e d water for ant isept ic purposes. In addit ion hands were ca re fu l l y washed in soapy water and r insed in 70% ethanol. As a further precaution against i n fec t i on , f inger na i l s were cut as short as poss ib le . A l l experiments were carr ied out using male Sprague Dawley or Wistar r a t s , 230 - 350 g. A glass gas ja r was equi l ibrated with 5 % halothane in oxygen del ivered v ia a vapourisor (F luotec) . For the l as t 300 - 400 prep- arat ions, a humidi f ier was included in the anaesthetic c i r c u i t . This com- prised simply of a conical f lask containing a saturated sa l ine solut ion through which the anaesthetic gas was bubbled. The sa l ine prevented the growth of algae and yeasts for periods of at least 9 months. Each rat was placed in the j a r 2 - 3 min af ter halothane had been introduced. Experience had shown that i f a rat was placed in the ja r before equ i l ib ra t ion with halothane, or i f a lower concentration of halothane ( 1 - 3 %) was used, then rats passed through the exc i ta tory plane of anaesthesia more s lowly, and exhibi ted coprophagic behaviour, which compromised the subsequent stage of preparation ( in tubat ion) . - 82 - Induction of anaesthesia took approximately 1 min, but rats were kept in the gas chamber for a further 1 - 2 min to induce deep anaesthesia (exempli- f i ed by slow, shallow resp i ra t i on ) . Following induct ion, each rat was in tu - bated as fo l lows. The mouth was f ixed open by temporari ly attaching the upper jaw to the bench, and a 14 gauge human intravenous catheter (Jelco) was inserted into the trachea with the aid of a paediat r ic laryngoscope. The sharp point of the metal insert of the catheter had been previously blunted and smoothed for atraumatic locat ion of the catheter. The vocal cords, v i s i b l e as two white bands on the l e f t and r igh t side of the entrance to the trachea, were used as landmarks. The paediat r ic laryngoscope had been f i l e d on one side in order to permit atraumatic entry into the small buccal cav i ty of the ra t s . Intubation was ve r i f i ed by observing the condensation produced when the exhaled a i r was directed onto the surface of the bench. The rats were transferred to a small square metal operating tab le , maintained on 1% halo- thane, and prepared with an occluder, int ravascular l ines and ECG leads, as fo l lows. Anaesthesia was adjusted during preparation such that movement was just prevented. 2.1.2.1 Occluders. In contrast with ea r l i e r work in rats in which coronary occlusion was brought about using s i l k thread (Heimburger, 1946; Johns and Olson, 1954; e t c . ) , our laboratory uses a snare device made from nylon and polythene (Au et a l ; , 1979a; 1979b; Johnston et a l ; , 1983a). The manufacture of the occluder has never been described in d e t a i l . Essen t i a l l y , a guide was made from an 11 cm length of PE10 polythene tubing. One end was f la red by b r ie f exposure to heat from a soldering i r on , wh i ls t 1 cm from the other end, a f lange was made by b r i e f l y melting the tubing by ro ta t ing i t in f ront of a je t of hot a i r . The hot a i r j e t was also used in the manufacture of blood pressure and intravenous l ines (see below), and was created simply - 83 - by passing pressurised a i r through a th in copper tube over a Bunsen burner. The occluder was made complete by threading a 5.0 gauge atraumatic nylon suture (Ethicon) through the polythene guide such that the needle end of the suture appeared at the f la red end of the guide. The occluder was implanted as fo l lows. A 1 cm skin i nc i s ion was made over the 4th to 6th r ibs on the l e f t thorax. This was enlarged by blunt d issect ion (Spencer Wells forceps) . The forceps were then inserted under the Pectoral is muscle, which was gently separated from the underlying Rectus Abdominus, exposing the Intercostal muscles beneath. A r t i f i c i a l resp i ra t ion was immediately i n i t i a t e d . Using a Palmer pump, or equivalent, at a stroke volume of 4 ml per ra t and a stroke rate of 54/min, the anaesthetic regimen described above was de l ivered, and contro l led using the same c r i t e r i a des- cr ibed above. Voluntary resp i ra t ion was switched to a r t i f i c i a l by the simple red i rec t ion of gas flow through the plumbing of the anaesthesia set-up. The 5th or 6th in tercosta l space was then punctured using the Spencer Wells forceps. This inc is ion was enlarged by blunt d i ssec t ion . The choice of 5th versus 6th in tercosta l space was a rb i t ra ry , and general ly not noted. If the heart was exposed unfavourably for placement of the occluder, then a second in tercosta l i nc is ion was made. This was necessary in less than 5 % of prep- arat ions. The in tercosta l inc is ion was widened by inser t ing 4 aluminium ret ractors attached to braided s i l k thread or lengths of PE90 polythene tubing. The aluminium ret ractors were placed ca re fu l l y in order to avoid damaging the l e f t lung, and were f ixed by attachment to the small s ta in less steel operating table (v ia s l i t s in i t s corners). The ret ractors could be eas i l y moved and reposi t ioned. In la ter experiments, a l l re t ractors were made with polythene attachments, since these were easier to replace ( i f they broke during surgery) than the braided s i l k attachments. By applying gentle pressure under the r igh t thorax, and replacing 1 of - 84 - the ret ractors under the thymus (which i s large and over l ies the heart in most r a t s ) , i t was a simple matter to expose the l e f t ven t r i c le and a t r i a l appendage. Using a pair of small blunt forceps, the th in per icard ia l mem- brane was retracted at a posi t ion overlying the junct ion between the l e f t a t r i a l appendage and the l e f t ven t r i c l e , at which point i t could be manipu- lated e a s i l y . Using a second pair of small blunt forceps, a small tear in the pericardium was created. The aluminium r i b ret ractors were then moved, one after another, by inser t ing the i r t ips through the small per icard ia l tear , such that the pericardium was now included with the retracted t i ssue . As a r e s u l t , the heart was l i f t e d toward the in tercosta l i nc is ion by th is per icard ia l c rad le , f a c i l i t a t i n g subsequent surgery. Without d isturbing the heart by gripping i t between the thumb and fo re- f inger (Johns and Olson, 1954), or ex te r io r i s ing i t through the in tercosta l i nc is ion (Selye e t - a l . , 1960), the nylon suture of the occluder was sewn under the l e f t coronary artery (general ly referred to as the LAD, see Intro- duction) by inser t ing the needle into the l e f t ven t r i c le under the overhang- ing l e f t a t r i a l appendage, and bringing i t out high on the pulmonary conus. The a t r i a l appendage was displaced s l i g h t l y for th is purpose by gentle manipulation with a small pair of blunt forceps. The LAD is described as i n v i s i b l e to the naked eye in rats by some Authors (Heimburger 1946; Johns and Olson 1954). These Authors located the LAD using the highly v i s i b l e coronary veins as landmarks. In our experiments, although the veins were found to be v i s i b l e , i t was usual ly possible to see the artery quite c l ea r l y as a thin pink l i ne emerging from under the a t r i a l appendage. The s t i t ch of the suture was made wide, such that the needle entered and l e f t the myocar- dium approximately 2 mm ei ther side of the ar tery . As Heimberger (1946) and Selye et al.- (1960) have pointed out, unless the artery i s included in the l i ga t i on there are no sequelae of consequence. This whole process usual ly - 85 - produced no blood l oss . Occasional ly ( in less than 10% of preparations) there was a l i t t l e bleeding, amounting to less than 1 ml of blood. If th is occurred, 5 min was allowed for c l o t t i n g , then the thoracic cav i ty was gently cleared with gauze in order to restore a blood-less f i e l d . Following the sewing of the suture under the l e f t coronary ar tery , the suture was sewn through the f la red end of the guide tubing. Therefore, the suture emerged from the guide tubing, passed into the myocardium, under the LAD, out of the myocardium and back through the t i p of the guide tubing, making a loose loop or snare. The needle of the suture was then cut o f f , and the remaining length of nylon ca re fu l l y melted down to form a small b a l l , which prevented the suture from being pul led back through the f la red t i p of the guide tubing. The nylon suture extended the f u l l length of the guide tubing. The s ize of the loop at the end proximal to the heart was adjusted by pu l l ing on the suture at the end d i s ta l to the heart . I t was found in some ear ly experiments that occlusion sometimes resul ted in part of the thymus being caught in the occluder. This could be completely prevented by making the loop smal l , such that the f la red t i p of the guide tubing was posit ioned adjacent to , or jus t underneath the a t r i a l appendage. Once the s ize of the loop had been adjusted, the d i s ta l length of suture was melted down to a small ba l l adjacent to the d i s ta l t i p of the guide tubing, in a manner s im i la r to that by which the proximal end of the suture had been a f f i xed . The Pectoral is muscle was sutured l i g h t l y to the Rectus Abdominus muscle with s i l k . The loose ends of the s i l k suture were then t ied to the occluder to f i x i t in posi t ion in the thorax. In a l l but the l as t 50 - 60 prepara- t i ons , the pneumothorax was evacuated at the time the chest was closed by inser t ing a length of PE90 polythene tubing through the in tercosta l i n c i s i o n , and applying negative pressure to the thorax as the chest was c losed. This - 86 - precaution was recent ly successfu l ly replaced by hyper inf la t ion of the lungs by t rans ien t ly increasing stroke volume to 6 ml per rat as the chest was c losed. This modif icat ion reduced the time taken to prepare the r a t s . As soon as the chest was c losed, the halothane anaesthetic was replaced with 100 %oxygen. The occluder was then fed subcutaneously to the subscap- ular region using a s ta in less steel t rocar . The occluder was ex ter io r ised between the shoulder blades, and a s i l k suture was attached to the occluder jus t below the flange at the d i s ta l end ( re la t i ve to the heart ) . This piece of s i l k was designed to serve as a ' f l a g ' to ass is t in the locat ion of the occluder on the day of occ lus ion. The length to which the occluder projected from the subscapular region was l imi ted to no more than 1 cm, in order to reduce the l i ke l ihood of the occluder being chewed or tugged at by the rat during the week of recovery from surgery. This was arranged by pu l l ing on the occluder with a pair of small forceps. In la te r experiments the occluder was allowed to s l i p jus t beneath the sk i n , with only the s i l k ' f l a g ' exposed. The rat was then turned onto i t s back, and the chest wound was i n f i l t r a t e d with C ica t r i n (bac i t rac in , neomycin and streptomycin) powder, and Marcaine (1% bupivacaine). The skin over the Pectoral is muscle was closed with a purse-str ing suture of s i l k . The rat was disconnected from the respi ratory pump, and allowed to breath spontaneously. The in terva l between c los ing the thoracic i nc i s ion and c los ing the skin i nc i s ion was general ly approximately 2 - 3 min, and since the rat was receiv ing only oxygen during th is t ime, su f f i c i en t halo- thane was expired to permit the respi ratory centre to override the a r t i f i c i a l respi ratory rhythm, such that once the rat was disconnected from the r e s p i - ratory pump, i t immediately began to breath spontaneously. However, the rat general ly remained immobile for a further 2 - 5 min. During th is t ime, C ica t r i n and Marcain were i n f i l t r a t e d into the subscapular wound, and the - 87 - other ex ter io r ised leads and l ines (see below) were t i d i ed up. The locat ion of the occluder was the las t stage in preparat ion, but was described f i r s t since i t is the focus of the experiment. Af ter the induction of anaesthesia, the f i r s t stage of preparation was the locat ion of the blood pressure and intravenous l i n e s . 2.1.2.2 Leads -and -1ines. Blood pressure and intravenous l ines were made from PE polythene tubing, using the hot a i r welding technique described for manufacture of the occluder. A 14 cm length of PE50 tubing was welded to a 10 cm length of PE10 tubing by ro tat ing and melting the i r t i p s , then pressing them together. This process was carr ied out af ter the lengths of tubing had been threaded onto a length of th in wire, in order to prevent the welding process from occluding the lumen. The seal between the tubings was tested by c los ing the open end of the PE10 tubing, attaching a needle and a i r - f i l l e d syringe to the open end and applying pressure to t h i s , ostens ib ly closed system, while i t was submerged in d i s t i l l e d water. An incomplete seal was revealed by bubbles emerging from the junct ion of the tubings. A cur l was put into the PE10 end of the l i n e , by looping i t around a glass rod and submerging i t in bo i l ing water for 3 sec. The loop was f ixed by sub- merging i t in ice cold water. The l ines were implanted using a modif icat ion of the method developed by Weeks (Weeks and Jones,1960; Weeks, 1981). A midl ine laparotomy was per- formed, extending from approximately 1 cm poster ior to the xiphoid process to the region overly ing the b i furcat ion of the common i l i a c vesse ls . Using f ingers only, the connective t issue overly ing the abdominal aorta and vena cava was c leared. A s i l k thread was then placed round these vesse ls , using a pair of small forceps, in the standard manner. The thread was usual ly posit ioned in the region of the b i furcat ions of the renal veins and a r te r i es , at which point the small forceps passed under the aorta and vena cava most - 88 - e a s i l y . A more poster ior approach occasional ly produced haemorrhaging from the vena cava. Whilst th is was rare ly f a t a l , and usual ly resolved without sequelae by the time of occ lus ion ' and s a c r i f i c e , i t was deemed inconvenient (5 min was allowed for c lo t t i ng i f ever haemorrhaging occurred, and t h i s , na tu ra l l y , delayed surgery). Therefore in most of the las t 500 - 600 prepa- rat ions the aorta and vena cava were accessed from the former, anter ior approach. Once the aorta and vena cava had been located, these vessels were sepa- rated using small forceps. Again, th is procedure was most eas i l y accom- pl ished when the more anter ior approach was made. The thread was divided such that tension on one or the other of the resul tant t i es would stop flow in ei ther the aorta or the vena cava. The l ines themselves were then located in the rat as fo l lows. Using the same trocar which would subsequently be used for locat ing the occluder, a channel was created through the Psoas muscle and subcutaneously to the subscapular region. With the trocar in p lace, two l ines were threaded into th is channel, and the trocar was then removed, leaving the l ines in p lace. A syringe f u l l of sal ine (without heparin) was attached to each of the l ines at the i r ex i t point between the shoulder blades (note that th is ex i t point was used subsequently for e x t e r i - o r i s ing the occluder) . The abdomen was then retracted on the l e f t side (using two ret ractors in the same manner as for the placement of the occ lu - der) . Both the l ines at the end proximal to the blood vessels (the PE10 ends) were severed, using a pair of sharp s c i s s o r s , to produce a pointed bevel to f a c i l i t a t e the subsequent inser t ion into the vesse ls . Usual ly , the vena caval l i ne was located f i r s t . The a i r in the l i ne was displaced with s a l i n e , and the l i ne was l i g h t l y held with a pair of small forceps. The vena cava was temporari ly occluded with the s i l k thread, and a t iny hole was created - 89 - d is ta l to the occlusion (proximal to the hear t ) , just anter ior to the renal ve ins. The l i ne was then inserted into the hole, and the occlusion removed. The process was repeated for the placement of the aor t ic l i n e , with the exception that the hole, d i s ta l to the occluder, was also d i s ta l to the heart . In both the case of the vena caval and also the aor t ic l i n e , the PE10 tubing was inserted 2 - 3 cm into the vesse ls , point ing towards the heart . The l i n e s , and the i r placement, were designed in order that blood press- ure recording and drug administrat ion could be undertaken v ia funct ioning (non-occluded) vesse ls . The holes in the vessels were made using a 27 gauge hypodermic needle which had been bent to an angle of approximately 120 ° . The hole was of a smaller diameter than the PE10 tubing, such that the l ines could be moved f ree ly without blood leakage. In add i t ion , the small diameter of the PE tubing did not occlude the aorta or vena cava; th is was obviously c r u c i a l , since these were recovery r a t s . The aorta and vena cava were used because they are large, eas i l y accessib le vesse ls , su i tab le for cannulation with chronic indwell ing non-occluding l i n e s . The abdomen was dusted with C i c a t r i n , and the body wall closed with a running s t i t c h . The skin was then c losed, also with a running s t i t c h , and the wound was i n f i l t r a t e d with C ica t r i n and Marcaine. Following placement of the occluder by the method described above, the ex ter io r ised ends of the cannulae were treated as fo l lows. S i l k f lags were attached to each cannula in the same way that a f l ag was attached to the occluder. Approximately 0.3 ml of sa l ine was then injected into the vena caval l i n e , which was abruptly clamped with a pair of Spencer Wells forceps, the t ips of which had been made atraumatic by encasement in soft polythene tubing. The open end of the exposed PE50 portion of the cannula was then sealed by mel t ing, using an ordinary c igaret te l i gh te r . The same process was car r ied out for the aor t ic - 90 - l i n e . The melted seals were f la red such that the i r diameters were approx- imately double that of the PE50 tubing. The l ines were allowed to extend from between the ra t s ' shoulder blades by about 1 cm. As in the case of the occluder, the l ines sometimes sl ipped beneath the skin during the in terva l between preparation and occ lus ion; the s i l k f lags permitted easy locat ion of the l i n e s , and the i r posi t ion under the skin was considered acceptable, since the cannulae were therefore out of the reach of the ra t s . In general, how- ever, the f l a r e d , melted ends of the cannulae prevented them from s l ipp ing completely under the sk in . The rats appeared to be unconcerned by the presence of the ex ter io r ised cannulae and the occluder. There was an extremely low incidence of in fect ion around the ex i t s i t e (fewer than 1 % of rats had pus exudation from the wound), presumably as a resu l t of the prophylaxis with C i c a t r i n . ECG leads were prepared from tef lon-coated s ta in less steel wi re. Approximately 1 cm of one end of each lead was de-insulated using an ordinary c igaret te l igh te r f lame. The lead was approximately V3. The chest lead was implanted with the occluder, as fo l lows. A t ight ba l l was made in one end of the lead by wrapping i t around a 21 gauge hypodermic needle. This end was then posit ioned underneath the Pectoral is muscle, using the 21 gauge needle as a t rocar . The d i s ta l end of the chest lead was ex ter io r ised with the occluder between the shoulder blades. The large trocar used for th is purpose was directed through the hole made e a r l i e r for the ex te r io r i za t ion of the int ravascular cannulae. The limb leads were implanted on the day of occ lus ion. They were made of the same material as the chest lead, but were posit ioned in a s l i g h t l y d i f fe rent manner (see below). After preparat ion, the animals were placed in indiv idual cages and given tap water and Purina rat chow ad l ib i tum. The body weight of each rat was - 91 - recorded before preparat ion. The procedures described usual ly took approx- imately 25 - 40 min to complete, from induction of anaesthesia to recovery of consciousness. 2.1.3 Coronary occlusion This sect ion deals with the experiment proper, which was car r ied out approximately 7 days after preparation (range 4 - 1 5 days). The sequence of events is described below. The actual process of occlusion was as fo l lows. The occluder was gripped by the Spencer Wells forceps with the atraumatic t i ps (see above) jus t above, and adjacent to the f lange located approximately 1 cm from the end d i s ta l to the heart . The purpose of the flange was to provide a stock for the forceps which would permit the outer guide of the occluder to be held f i rm without excessive clamping, such that the inner snare of the occluder remained free to move within the outer guide. Using f i nge rs , the small bobble on the d i s ta l end of the inner snare of the occ lu - der was pul led from the adjacent outer guide tubing to the extent of approx- imately 2 mm. A second pair of forceps was then used to f i rmly gr ip the exposed inner snare of the occluder. Traction was then applied smoothly between the inner snare and the outer guide tubing to produce occ lus ion. Successful occlusion occurred when the outer guide tubing became cr ink led and i t was no longer possible to move the inner snare in re la t ion to the outer guide. This moment was designated time zero. The atraumatic forceps were clamped down f i rmly on the occluder at th is moment, and the length of exposed inner snare of the occluder was melted down with a soldering iron to form a bobble adjacent to the d i s ta l end of the outer guide tubing, f i x i ng i t in p lace. Time zero was accompanied in most cases by a sudden change in the ECG, but th i s was not taken as a c r i t e r i on for occ lus ion , since i t was possible that drug treatment could delay the ECG changes caused by occ lus ion. Never- - 92 - the less , in the present experiments i t was found that an absence of ECG changes within 15 min of occlusion was associated in every instance with an absence of an occluded zone (measured upon s a c r i f i c e , see below), ind icat ing that occlusion had been unsuccessful. This event occurred in less than 5% of r a t s . 2.1.3.1 Sequence of events. The sequence of events on the day of occlusion is out l ined in Figure 1. Rats were anaesthetised with halothane in the manner described above, and allowed to breathe 1.5% halothane through a small face mask. The limb leads of the ECG (made in the same way as the chest lead) were then placed as fo l lows. Each wire was threaded through a 23 gauge hypodermic needle, such that the non-insulated t ip projected approximately 0.5 cm from the sharp end of the needle. The wire was then bent back c lean ly , forming a barb. The needle, together with the electrode was passed subcutaneously, e i ther to the elbow region of the forelimb (dorsal aspect) for the 2 forelimb leads, or to the r igh t anter ior dorsal f lank region for the r ight hind limb lead. The entry point was the same for a l l three leads, namely approximately 0.5 cm anter ior to the ex i t point of the occluder and intravascular l i n e s . The ECG leads were disconnected from the hypodermic needle by pinching the skin around the needle t i p and ca re fu l l y withdrawing the needle. As the needle was withdrawn, the electrode remained in p lace, owing to the action of the barb. In approximately the f i r s t 50 preparat ions, the limb leads were implan- ted on the day of preparat ion. Although they did not appear to cause d i s - comfort, they were sometimes removed by the rat during the in terva l between preparation and occ lus ion. Therefore, i t was considered sensible to implant limb leads on the day of occlusion in every case. In th is way an element of consistency was introduced, which was absent i f some rats were used with fresh electrodes placed on the day of occ lus ion, and some not. CORONARY OCCLUSION IN CONSCIOUS RATS r o C D -60 Sequence of events on the day of study (time scale in min) 1 day la te r 0 •15 -5,+5 +15 +30 +60 +120 +180 +240 0 30 stab i1 i sa t ion drug admi ni s t r a t i on OCCLUSION DEATH? SACRIFICE measure occlusion measure i n f a rc t Figure 1. Sequence of events on the day of occlusion fol lowing connection of leads and l ines - 94 - The 3 limb leads were connected to one another by simply twist ing the i r exposed ends together, forming a composite whole body lead, which was trimmed to approximately 2 cm length using a pair of s c i s so rs . The te f lon was removed from the end of the composite lead as described above. It should be noted that the anaesthetic mask was temporari ly moved well away from the ra t during th is procedure. The un-insulated leads were connected to 1 of 2 lengths of enamel-coated s ta in less steel wire by twis t ing the wires together. The t ips of the wires had been scraped with a scalpel blade to remove the insu la t ing enamel. The connections were iso la ted by applying a small piece of surg ical tape to each. The tape was attached in such a way that i t could eas i l y be removed at a la ter time. Add i t iona l l y in th i s regard, when the leads and wires were connected, the twist was always c lockwise, f a c i l i t a t i n g subsequent disconnect ion. Whilst the rat was s t i l l anaesthetised with halothane, the blood pressure l i ne was connected to a pressure transducer v ia a length of PE tubing and a blunted 23 gauge needle t i p . The intravenous l i ne was connected to a syringe f u l l of sa l ine in a s im i la r manner. Both l ines were flushed with approx- imately 0.3 ml sa l i ne . The intravenous l i ne was ind is t inguishable from the a r te r i a l l i ne by v isual inspect ion, but once the melted t i p of the l i ne had been snipped o f f , the l ines were eas i l y d is t inguishable in most cases by the fact that blood immediately f i l l e d the aor t ic l i n e , but not the vena caval l i n e . In the infrequent event that blood did not flow back through ei ther l i n e , both l ines were flushed with sa l ine un t i l a d i s t i nc t i on could be made. The aor t i c l i ne was kept open by attaching a leak pump in ser ies with the l i n e , according to the method of Weeks (1981). In at least 75% of cases, negative pressure on the vena caval l i ne caused dark venous blood to appear in the l i n e . In the remaining 25% of cases, the venous l i ne was tested by the in jec t ion of a small amount of adrenaline which e l i c i t e d a small pressor - 95 - response when the l ine was patent. The placement of the ECG leads and the connection of the blood pressure and intravenous l ines to the transducer and syringe of s a l i n e , respect ive ly , took no more than 10 min. Rats regained consciousness upon termination of anaesthetic general ly within 30 sec, and were f u l l y a le r t within approx- imately 2 - 5 min. Blood pressure and the ECG were displayed on a Grass Polygraph (4 channel, model 7 ) , and ECG was also displayed on a delayed loop osci l loscope with a 4 sec delay and a 4 sec real time display (Honeywell Type E for M). The la t te r was pa r t i cu la r l y useful for the diagnosis of arrhythmias (see below). The ECG channel was ca l ib ra ted once per week; l i t t l e adjustment was ever required. The blood pressure channel was c a l i - brated every time i t was used. After approximately 1 h, the administrat ion of drugs, i f any, and coro- nary occlusion were carr ied out while the ra t was f u l l y conscious and not restra ined in any way. The procedure was as fo l lows. Intravenously admin- is tered drugs were injected slowly over a 10 min per iod. A fast (100 mm/sec chart speed) record of ECG and blood pressure was taken before drug admin- i s t ra t i on and 4 min af ter completion of i n j ec t i on . One min af ter the second fas t recording of blood pressure, the occluder was t ightened, as described above. Occlusion did not appear to cause the rats any d i s t ress . No changes in behaviour ever occurred during the f i r s t few min after occ lus ion, unless blood pressure f e l l prec ip i tous ly fo l lowing occlusion ( in which case the animal became subdued), or unless acute pulmonary oedema, (characterised by laboured resp i ra t ion) developed. In general , the f i r s t behavioural response to occlusion was sudden convulsive-type behaviour which occurred in con- junct ion with VF (see below). The rats were monitored for at least 4 h fo l lowing occ lus ion. Fast speed polygraph recordings were made every min for the f i r s t 15 min, every 5 min - 96 - for the fol lowing 15 min, then every 15 min. If the rat survived for 4 h, i t was disconnected from the ECG wires and blood pressure transducer by reversing the processes used for connecting the animal to these recording devices. The animal was conscious and not restra ined during th is process. Usual ly the rat was subdued at th is time as a resu l t of the sequelae of occlusion (see below) and was submissive and compliant enough for th is disconnection to be completed without any d i f f i c u l t y . After 24 h, i f the rat was s t i l l a l i v e , i t was reconnected to the blood- pressure and ECG recorders and care fu l l y monitored for a fur ther 30 min, whereupon i t was s a c r i f i c e d . Reconnection to the recording devices was carr ied out in the same way as the i n i t i a l connection the previous day, with the exception that the rat was not anaesthetised. Instead, i t was placed on the bench and wrapped in a lab coat, with only the subscapular region exposed. Rats, l i k e many rodents, do not struggle when they cannot see. In th is manner the l ines and leads were reconnected without inc ident . I t i s worth mentioning that in ear ly experiments, when the limb leads were implan- ted on the day of preparation along with the occluder and blood pressure l i n e , e t c . , i t was routine to connect a l l rats to the blood pressure and ECG recording devices in th is manner ( i . e . , when the ra t was conscious) on the day of occ lus ion. Upon death or s a c r i f i c e , the heart was excised and the occluded zone and in fa rc t s i ze were determined (see below), and a general postmortem examin- ation was performed (the appearance of the lungs, l i v e r , kidneys, spleen, bladder and snout were noted, and any gross abnormalit ies were recorded). Exclusion c r i t e r i a (see below) were considered at each stage of experimen- ta t i on . 2.1.3.2 Monitoring of - responses -t-o occ lus ion. As described above, the ECG and blood pressure were recorded continuously for 4 h fo l lowing - 97 - occ lus ion , and again at 24 h af ter occ lus ion. Behavioural changes were noted in a general way, and any pecu l i a r i t i es were recorded. Most of the tech- niques described below have been described in br ie f previously (Au e t - a l ; , 1979; Johnston et a l . , 1983a; Cur t is et a l . , 1984; 1985b). A l l information concerning each rat was recorded permanently on ind iv idual analysis sheets. 2.1.3.3 Occluded-zone (OZ). The OZ was recorded for every ra t . In 24 h surv ivors , s a c r i f i c e was undertaken by de l iver ing a blow to the head with a metal cosh. No anaesthetic was used. The heart was excised and per- fused v ia the aorta according to Langendorff (1895) with sa l ine (0.9 %). Once blood was no longer present in the perfusate, the sa l ine was replaced with sa l ine containing Indocyanine (Fast Green dye, BDH) 0.5 g/1. Approx- imately 20 - 50 ml of th is solut ion was allowed to pass through the coronary c i r c u l a t i o n . The del ivery of the 2 so lu t ions , sa l ine and Fast Green dye was contro l led using a T-tube device connected to 2 large reservoi rs containing each so lu t i on , and was regulated with a simple glass 2 by 2-way stopcock. In the case of animals which died overnight, care was taken not to d i s - lodge the s tas is thrombi present in the l e f t ven t r i c l e . Perfusion with approximately 20 ml Fast Green dye sometimes took 5 - 1 0 min in such hearts, compared with the 2 - 3 min in the case of hearts from f resh ly dead animals. However, general ised c lo t t i ng in the coronary c i r cu la t i on had probably not occurred, because the hearts from rats dying overnight provided OZs as well defined and of a s imi la r extent to those from f resh ly dead animals. It i s more l i k e l y that perfusion rate was reduced as a resu l t of the contracture which was a common feature of hearts removed from rats dying overnight s ince , with reference to the ef fect of sys to le , coronary perfusion is highly sens i t i ve to the con t rac t i l e state of the myocardium. Af ter perfusion, the heart was removed and processed. F i r s t of a l l , the a t r i a , aorta and pulmonary vessels were removed and discarded. Delineation - 98 - of the remaining vent r icu lar myocardium into normal and occluded t issue was made by v isual inspect ion. The two zones were almost invar iab ly well d i f fe rent ia ted with a c lear-cut border, as might be expected on the basis of the reported lack of co l l a t e ra l s in rat hearts (Johns and Olson, 1954; Maxwell et a l . , 1984; Winkler et a l . , 1984). Using a small pai r of sc issors the perfused (green) and non-perfused (pink) t issues were separated, l i g h t l y blot ted and weighed. 2.1.3.4 Infarct -zone - ( IZ) . The in fa rc t s ize was only determined for animals which survived for 24 h. In th is regard, i t has been shown that an in terva l of at least 10 h must be allowed af ter occlusion for in farc t s ize (measured in the manner described below) to be s u f f i c i e n t l y developed for quant i f i ca t ion in rats (Hort and Da Cana l is , 1965a). Infarct s ize was determined by a modif icat ion of the method of Jestadt and Sandr i t ter (1959), based on the production of formazan red by the reduction of 2 ,3 ,5 - t r i pheny l - tetrazol ium (TTZ) by H + emanating from NAD- or NADP-1inked dehydrogen- ases. The normal and occluded t issues were cut into s t r i ps approximately 2 mm wide and incubated with TTZ (Fischer S c i e n t i f i c ) 0.1 g in 10 ml phos- phate buffer at 37 °C. The buffer was made by d isso lv ing 25 g Na^HPO^ and 0.5 g NaHgPO^ in 2 1 d i s t i l l e d water (pH adjusted to 8.5 - 8.6 with 0.1 M NaOH). Incubation was carr ied out for approximately 10 min ( incuba- t ion was terminated when the s t r ips of non-occluded t issue had turned deep purple from the formazan red reac t ion , rather than after a spec i f i c t ime). In la ter experiments, the OZ t issue was cut not into s t r i p s , but into 2 discs by s l i c i n g through the subendocardium. The resul tant discs were no more than 2 mm thick (by v isual inspection) at any point . The ra t iona le for th is approach was to invest igate whether the endocardial and ep icard ia l surfaces were salvaged to any extent as has been suggested by Hort and Da Canalis (1965b) and others. Although no quant i ta t ive data was assembled, i t - 99 - appeared that the endocardial surface was usual ly not infarcted (white) at 24 h, and occasional ly parts of the epicard ia l surface were also not white, although a gentle scrape with a pair of sharp sc issors was always su f f i c i en t to remove th is 'salvaged' t i s sue . In addi t ion, by separate weighing of th is t issue i t was found that i t s presence changed the in fa rc t s i ze by only 2 to 3 %. Infarct s ize was quant i f ied by separating the white and purple t issues and weighing them. However, the t issue was f i r s t f i xed for 2-3 days in 10 ml formal sa l ine (made by d isso lv ing 3.56 g NaCl and 125 ml 40% formaldehyde in 375 ml d i s t i l l e d water). I t has been shown that the i n f a r c t , as detected by th is method, corresponds with the h i s t o l o g i c a l l y i den t i f i ed in fa rc t (Fishbein e t a ! . , 1981). 2.1.4 Def in i t ion of occlusion-induced arrhythmias 2.1.4.1 Introduct ion. In order to invest igate the ef fects of treatments on a va r iab le , an unequivocal de f in i t i on of that var iab le i s necessary. In the case of occlusion-induced arrhythmias, i t i s general ly the case that var ia t ions ex is t in de f i n i t i ons . While these dif ferences may not necessar i ly be important in themselves, they make d i rect comparison between studies d i f f i c u l t to in terpret in some cases. Although there is var ia t ion in the d i s t i nc t i on between a run of PVC and VT, and di f ferences in the quant i f i ca t ion of PVC, the main source of contention l i e s in the d i s - t i nc t ion between VT (pa r t i cu la r l y that of the torsade de pointes var iety) and VF. VF was f i r s t described by Erichsen (1842) in terms of the con t rac t i l e behaviour of the myocardium ( th is was more than 60 years before the develop- ment of the ECG). VF has been defined in various ways. Moe e t - a l ; (1964) defined VF as chaot ic , asynchronous f ract ionated e l e c t r i c a l a c t i v i t y . Bigger (1980) gave an operational d e f i n i t i o n , based on the charac te r i s t i c ECG - 100 - pattern of 'the absence of QRS complexes and T waves and the presence of low-amplitude baseline undulat ions ' . These de f in i t ions are not amenable to quant i tat ive d i f f e ren t i a l diagnosis because they are subject ive de f i n i t i ons , and they do not d i f fe ren t ia te between abrupt VF and VF occurr ing after a run of VT. This i s not r e a l l y a serious problem c l i n i c a l l y ; the d i s t i nc t i on between i r regu lar torsade de pointes, vent r icu lar f l u t t e r and VF i s , perhaps academic, since a l l 3 arrhythmias are associated with cardiac output below that necessary for the maintenance of l i f e , and the primary concern i s the prevention of mor ta l i ty . However, more rigorous de f in i t i ons are necessary experimental ly. Abrahamsson's group who induce myocardial ischaemia in rats define VF as asynchronous disorganised e l e c t r i c a l a c t i v i t y of at least 5 sec duration (Abrahamsson e t - a l ; , 1985). While th is de f in i t i on introduces an important caveat, namely that some indicat ion concerning the duration of the phenomenon may be useful in d iagnosis, i t nevertheless f a i l s as a t ru l y object ive d e f i n i t i o n , since how does one define 'asynchronous' and 'd isorganised' ? Opie's group have defined VF as ' to ta l i r r egu la r i t y of morphology of the repe t i t i ve ectopic complexes for at least 6 cyc les ' (Lubbe et a l ; , 1978), introducing once more the notion that VF is only VF i f disorder is present for an arb i t ra ry but spec i f ied durat ion. C lear ly there i s a subject ive element in the diagnosis and de f in i t i on of VF. This is an important point and attempts to deal with th is problem are discussed below. I t i s unfortunate that many researchers do not appear to have a r i g i d set of c r i t e r i a for def in ing arrhythmias. A glance through the l i t e ra tu re w i l l reveal that most researchers do not define what they mean by VT and VF. In pa r t i cu la r , i t i s often unclear by what c r i t e r i a torsade de pointes and VF are d i f f e ren t ia ted , and how many consecutive PVCs const i tute a run of VT. If PVC, VT, torsade de pointes, f l u t t e r and VF represent a continuum, - 101 - as was once believed ( e . g . , Har r i s , 1950), then the d i s t i nc t i on between these arrhythmias would not be ser ious ly important; a drug which reduces PVC would be expected to reduce VT and VF as w e l l . However, there is some indicat ion from studies with animals (Dresel and Sut ter , 1961) and from fol low-up invest igat ions in healthy humans with a high frequency of PVC (Kennedy e t - a l . , 1985), that PVC, VT and VF are not necessar i ly part of a continuum. Therefore the p o s s i b i l i t y ex is ts that a drug may inf luence se lec t i ve l y one type of vent r icu lar arrhythmia but not another. The la t te r issue is neither proven nor disproven at present, and wholesale r e - c l a s s i f i c a t i o n of drugs as ' c l a s s i c a l antiarrhythmic agents' and ' an t i a r rhy thm ic /an t i f i b r i l l a to ry agents ' , with s e l e c t i v i t y for VF versus PVCs (Anderson, 1984) is somewhat premature. I r respect ive of whether ischaemia-induced vent r icu lar arrhythmias repre- sent a continuum or not, i t i s nevertheless important in any experiment to define endpoints. Since arrhythmia de f in i t ions vary from one invest igator to another, a l l one can aim for is internal consistency at present. Before describing the c r i t e r i a which were used for def ining vent r icu lar arrhythmias produced by coronary occlusion in r a t s , the subject ive nature of c l a s s i f i c - at ion of arrhythmias must be re i te ra ted . No matter what de f in i t i on is chosen for VF and VT, there remains an element of sub jec t i v i t y , pa r t i cu la r l y with regard to d i f fe ren t ia t ion between torsade de pointes and VF. A simple way around the problem of sub jec t i v i t y i s to carry out studies using a b l ind and random protocol . In th is manner, inconsistencies in d e f i - n i t ion should be spread randomly and evenly between study groups, and conscious and unconscious bias should be e l iminated. I t must be stressed that attempts were made to carry out a l l the experiments described in a b l ind and random manner, where poss ib le . - 102 - The fo l lowing def in i t ions were used in the diagnosis of arrhythmias fo l lowing coronary occlusion in r a t s . 2.1.4.2 Premature vent r icu lar -cont ract ions-(PVG). PVC were defined as premature QRS complexes occurr ing independently of the P wave. PVC were general ly accompanied by a t ransient drop in aor t ic blood pressure. Owing to the high frequency f i l t e r i n g of the ECG by the Grass polygraph, no attempt to measure coupling in terva ls was undertaken. Measurement of coupling in terva ls i s believed to give an ind icat ion of whether the PVC is reentrant or automatic (see Introduct ion). However, i t was considered that the equip- ment avai lab le was not amenable to the routine measurement of coupling in te rva ls , since measurement of such by hand would be ex t raord inar i l y time consuming. Only s i ng le t s , doublets (bigemini) and t r i p l e t s were counted as PVCs. Longer runs were recorded as VT (see below). I t i s noteworthy that other workers include every def lect ion in a run of VT as a PVC (Kane and Winslow, 1980; Clark et a l . , 1980). This manoeuvre is based on the assump- t ion that PVC and VT are ident ica l in mechanism and drug s e n s i t i v i t y . While our de f in i t i on implies that s i ng le t s , doublets and t r i p l e t s represent the 'same' arrhythmia, we only associate these arrhythmias for ana ly t ica l con- venience; since the incidence of doublets and t r i p l e t s is var iab le and lower than the incidence of s i ng l e t s , a useful invest igat ion of the ef fects of a drug on these ind iv idual types of arrhythmia would require an increment in the study group s i z e . I t was considered that the cut -of f value of 4 PVCs, which was low, removed the p o s s i b i l i t y of spurious amalgamation of arrhyth- mias which are not necessar i ly the product of a common e lect rophys io log ica l mechanism (namely short runs of VT and long runs of PVCs). In other words, in the absence of c lear evidence concerning the point at which a run of PVCs ceases to be PVCs and becomes VT, i t was decided to a r t i f i c i a l l y segregate PVCs and VT. E s s e n t i a l l y , since one cannot be sure whether VT and PVCs are - 103 - expressions of the same arrhythmia or not, i t i s sensible to separate them for analysis by generating an arb i t ra ry cut -o f f po int . 2.1.4.3 Ventr icu lar- tachycard ia-(VT). VT was defined as a run of 4 or more consecutive PVC. No res t r i c t i on was made on the associated ra te . This de f in i t i on d i f f e rs from that of other workers who measure ischaemia- induced arrhythmias in r a t s . VT has been defined as 7 or more PVCs at a rate of > 600/min (Kane e t - a l ; , 1980), heart rate exceeding 500 beats/min (Lepran et a l ; , 1981b), 7 or more PVCs, no l im i ta t ion on rate (Fagbemi and Par ra t t , 1981b; McLennan et a l . , 1985) and 5 or more PVCs (Mertz and Kaplan, 1982; Manning e t - a l ; , 1984; Daugherty e t - a l . , 1986). C l i n i c a l l y , VT has been defined as 3 or more consecutive PVCs (Lown et - a l ; , 1973), and th is de f in i t i on has been adopted by Verdouw's group in the i r work with pigs (Verdouw et a l . , 1978). 2.1.4.4 Ventr icular - f i b r i l l a t i o n - (VF). VF was defined as disorder in the ECG accompanied by a precip i tous f a l l in blood pressure. In add i t ion , since d e f i b r i n a t i o n was undertaken in a l l rats experiencing 10 sec of VT or VF, a time constraint was imposed in order to d i f fe ren t ia te between 'primary' VF and VF 'secondary' to VT, namely that disorder had to have begun within 10 sec of the las t sinus beat for the arrhythmia to be classed as VF, other- wise the arrhythmia was c l a s s i f i e d as VT. Disorder was defined as an i r r e g - ular cycle length with no i den t i f i ab le QRS complex. As discussed above, the word 'd isorder ' lends an element of sub jec t i v i t y to the d e f i n i t i o n . Bias was eliminated as far as possible by b l ind and random experimental design. In the introduction to th is subsect ion, a var ie ty of operational d e f i n i - t ions of VF were given, and i t was suggested that the timing and durat ion.of 'd isorder ' needs to be def ined, as well as the actual nature of ' d i so rde r ' . Some Authors have avoided the question of the de f in i t i on of VF al together. Winslow i n i t i a l l y reported the incidence of VF in rats fo l lowing occlusion - 104 - (Kane and Winslow, 1980), but in la te r studies (Marshall e t a l . - , 1981a; 1981b; 1981c; Marshall and Winslow, 1981; e tc . ) VT and VF were combined, and the arrhythmia was ca l led ' f i b r i 1 l o f l u t t e r ' . Northover has adopted a s imi la r s t rategy, avoiding 'a rb i t ra ry and subject ive d i s t i n c t i ons ' between VT and VF by simply recording the time spent in the 'combined forms' (Northover, 1985). Other groups have taken a completely d i f ferent approach, def in ing VF in absolute terms. For example, Harron et a l . (1985) defined VF as a run of ectopic beats with a rate of 720/min or more. It is evident that there i s a lack of consistency between research groups in terms of de f in i t i on of VF. The method used in the current experiments is simply a de f in i t i on of VF which was f e l t to be reasonably consistent with c l i n i c a l de f i n i t i ons , and more importantly, amenable to routine use, i n te r - na l l y consis tent , but not obsessively unequivocal to the point of absurdity (see the de f in i t i on used by Harron e t - a l ; , 1985, above). Despite the d i f ferent c l a s s i f i c a t i o n s used, the resu l ts obtained by various groups with Na + channel blocking drugs (see Discussion) are extremely consis tent , a l l studies demonstrating a n t i f i b r i l l a t o r y a c t i v i t y for quinidine and Org-6001, for example, suggesting that the v a r i a b i l i t y in arrhythmia c l a s s i f i c a t i o n appears not to compromise inves t iga t ions , in terms of the recognit ion of a n t i f i b r i l l a t o r y drugs. However, the v a r i a b i l i t y in control incidence of VF within research groups suggests that there may be an element of internal inconsistency. This may re la te to the fact that apart from ourselves, research groups do not appear to carry out the i r studies using b l ind and random protocols. This may give r i s e to some misleading information, pa r t i cu la r l y concerning drugs with weak antiarrhythmic act ions, as a resu l t of the loss of precis ion of endpoints. 2.1.4.5 Other-arrhythmias. Although VF, VT and PVC are the arrhythmias of major in terest and importance, a l l arrhythmias fol lowing - 105 - occlusion are recorded. The incidence of non-ventr icular (and non-nodal) arrhythmias i s , however, extremely low in control r a t s , and i t is not considered worthwhile invest igat ing them in a quant i tat ive manner. A t r i o - vent r icu la r (AV) blocks occur from time to t ime, pa r t i cu l a r l y in associat ion with pulmonary oedema-induced gasping. Anecdotal observations have strongly suggested that gasps in rats with severe pulmonary oedema are coupled, 1:1 with Moebitz Type-2 AV block. Presumably the AV block i s the resu l t of a vagal r e f l ex . A l l types of AV block and d issoc ia t ion have been observed. Third degree AV block is usual ly only seen in rats experiencing pers istent and severe respiratory d is t ress (exudation of sputum), and according to the exclusion c r i t e r i a (see below) are not usual ly included in s tud ies . Sinus bradycardia and tachycardia do not occur very often during the f i r s t 4 h af ter occ lus ion, but both have been observed in 24 h survivor r a t s . A t r i a l arrhythmias are extremely ra re . It i s d i f f i c u l t to d is t inguish between a t r i a l f l u t t e r and f i b r i l l a t i o n using the V3 lead, but some sort of a t r i a l ' f i b r i l l o f l u t t e r 1 has been observed ( in fewer than 1% of r a t s ) . 2.1.4.6 Rationale for- - d e f i b r i l l a t i o n . VF is a serious arrhythmia which is l i fe - th reaten ing i f i t does not spontaneously rever t . It is well establ ished that rats can spontaneously d e f i b r i l l ate ( e . g . , Johnston e t - a l . , 1983a). However, between approximately 25% (Kenedi and Losconci , 1973a) and 95% (Siegmund et a l . , 1979b) of control rats which experience VF die from VF (values being dependent on whether conscious or anaesthetised animals are used, and on how VF is def ined). Therefore, i t i s expected in any study that the sample s ize w i l l vary with t ime, as animals d ie , and w i l l a lso vary between groups. This means f i r s t l y that ult imate group s izes w i l l vary, compromising the value of s t a t i s t i c a l t es ts , secondly that group s ize may be so small in controls that meaningful comparisons cannot be made, and t h i r d l y - 106 - that much information concerning the time course and the in ter re la t ionsh ips of var iables is los t as animals die during the course of the study per iod. This censoring can only be eliminated i f the animals are a l l kept a l i ve by d e f i b r i l l a t i n g VF as i t occurs. In a l l rats an attempt was made to revert a l l episode of VT and VF l a s t - ing longer than 10 sec by thump-version. VT was reverted as well as VF because many (but not a l l ) episodes of VT reduce blood pressure to close to zero in a manner analogous to that seen with VF; i t was considered that in order to preclude the generation of a rb i t ra ry de f in i t i ons concerning the ' seve r i t y ' of VT, a l l VT would be reverted af ter 10 sec in the manner used for VF. The procedure was as fo l lows. After 10 sec of continuous VF or VT the rat was l i f t e d out of the home cage by the t a i l . Within a few seconds convulsive-type behaviour (see sect ion concerning prel iminary screen) ceased and syncope ensued, whereupon the chest was f l i c ked with the index f inger . Usual ly only 1 or 2 f l i c k s were required to revert the arrhythmia to sinus rhythm. If reversion did not occur immediately then the f l i c k s were continued. Reversion was d iag- nosed by the combination of a sudden return to consciousness, a sudden increase in aor t i c blood pressure and the termination of the arrhythmia (as seen on the ECG record) . Because 10 sec was the maximum time allowed before d e f i b r i l l a t i o n was i n i t i a t e d , d i f fe ren t ia t ion between VT and VF was only attempted during the 10 sec period of the arrhythmia. Of course, i f VT were allowed to continue for longer than 10 sec, i t may possibly degenerate to VF. Therefore our method of diagnosing VF may lead to an underestimation of VF compared with other workers who do not use d e f i b r i l l a t i o n . To our knowledge, only 1 other group attempts thump-version, Charnock's group (McLennan e t - a l . , 1985), using our technique. - 107 - Although thump version great ly reduces censoring due to arrhythmia- induced mor ta l i ty , i t introduces a new form of censoring, namely that h i t t i ng the chest may 'change th ings ' compared with rats which have not experienced VF. This p o s s i b i l i t y was examined by our laboratory (unpublished obser- vations) and i t was found that d e f i b r i l l a t i o n reduced the incidence of mor ta l i ty from VF, but did not inf luence the incidence of VF i t s e l f . In other words, d e f i b r i l l a t i o n does not inf luence the l i ke l ihood of subsequent VF. Therefore i t was considered that the benef i ts of thump-version out- weighed the possible disadvantages. Thump-version creates 2 classes of VT and VF, that which spontaneously reverts to sinus rhythm within 10 sec of onset (SVT and SVF) and that which does not spontaneously revert before thump-version (NVT and NVF). The generation of subclasses of VT and VF may be i r re levant in terms of ana lys is ; while we continue to note the incidence and log^Q number of episodes of VT, SVT, NVT, VF, SVF and NVF, we have found that drugs which inf luence SVF also inf luence NVF, and drugs which inf luence SVT also inf luence NVT. Therefore i t i s reasonable to assume that SVT and NVT are ident ica l in terms of mechanism of generation; the same can be said for SVF and NVF. This is an important point because i t implies that spontaneous reversion of VF to sinus rhythm in rats is not a disadvantage of the preparat ion. Moreover, the occurrence of SVF, and the read i l y revertable nature of NVF (see resu l ts ) means that most animals w i l l survive the period of continuous monitoring such that censoring produced by ear ly morta l i ty i s reduced. 2.1.4.7 Arrhythmia scores. In quanti fying arrhythmias, many workers record the i r incidence, number, type and durat ion. Such information has l imi ted value since the number of episodes of VT and VF are log^ -norma l l y d is t r ibu ted var iables (Johnston et a l ; , 1983a); th is fac t i s not taken into considerat ion by other workers in the f i e l d . In addi t ion, thump-version may - 108 - inf luence the duration of VT and VF compared with the values found by other invest iga tors . In the hands of workers who do not revert VT and VF, the mean duration of VF is dominated by the terminal event. I n tu i t i ve l y , such measures of duration of VF should be log^g normally d is t r ibuted for th is reason. However, correct ion of data i s never undertaken (Clark et a l . - , 1980; Kane and Winslow, 1980; e t c . ) . In our laboratory, i t was found that even with thump-version, the durations of VT and VF were log-^Q normally d i s - t r ibuted (Johnston et a l • , 1983a). Nevertheless, since thump-version c lea r l y censors the durations of VT and VF, i t i s not reasonable to completely re l y upon such var iables for antiarrhythmic quan t i f i ca t ion . There is another reason for th is conclusion; i f a drug abolishes VF then there i s no mathe- matical value for log^g VF durat ion. If an arb i t ra ry value of 1 sec i s given to rats not experiencing VF then the resu l t i s a standard deviat ion of zero in the group in which VF is absent, creat ing a variance inhomogeneity which inval idates the use of standard parametric s t a t i s t i c a l tests (although the non-parametric Mann-Whitney U test may be used). In summary, measuring arrhythmia duration alone is not an acceptable means of gauging the sever i ty of arrhythmias. An a l te rnat ive approach is to generate a number scale which can be used to summarise and grade a l l the arrhythmias fo l lowing occlusion in terms of incidence and sever i t y . This scale (arrhythmia score) should accomodate and summarise complex arrhythmia data se ts . Arrhythmia scores should be normally d is t r ibu ted (to permit parametric s t a t i s t i c a l test ing) and l i n e a r l y add i t i ve . The arrhythmia score which has been used for occlusion studies in the conscious rat in our laboratory has been shown to be Gaussian d is t r ibuted in control rats (Johnston et a l ; , 1983a) and amenable to modified t t es t s . - 109 - This score is as fo l lows: 0 = No more than 49 PVCs, 1 = 50 - 499 PVC, 2 = No more than 1 episode of SVT or SVF and/or > 499 PVC, 3 = More than 1 episode of VT and/or VF < 60 sec to ta l durat ion, 4 = VT and/or VF of 60 - 119 sec to ta l durat ion, 5 = VT and/or VF of > 119 sec to ta l durat ion, 6 = Fatal VF occuring 15 min - 4 h after occ lus ion, 7 = Fatal VF occurring 4 min - 14 min 59 sec af ter occ lus ion, 8 = Fatal VF occurring 1 min - 3 min 59 sec af ter occ lus ion, 9 = Fatal VF occurr ing before 1 min af ter occ lus ion . This score has been adopted by Charnock's group for studies in anaesthetised rats (McLennan and Charnock, 1984; McLennan et a l . , 1985; McLennan, 1986). Other arrhythmia scores have also been developed. Martinez and Crampton (1981) have used a score based on the product of arrhythmia duration and arrhythmia type (the la t te r being graded from 1 to 5 according to seve r i t y ) . However, th is score was found not to be amenable to parametric s t a t i s t i c a l t es t i ng , since drug treatments al tered the var iance. Mueller e t - a l ; (1984) have used a simple 3 point score which they sum for each group and submit to the chi test for ana lys is . Woodward's group have used a group arrhythmia score, whereby arrhythmias for a whole group are summed, and a score assigned to the group (Daugherty e t - a l ; , 1986). However, the object of having an arrhythmia score seems to have been forgotten by th is group, since th is par t i cu la r score does not seem to be amenable to any s t a t i s t i c a l t es t . The ideal arrhythmia score should be amenable to parametric s t a t i s t i c a l tests (analysis of variance followed by modified t tests for mul t ip le comparisons, such as Tukey's t e s t ) . In th is manner, arrhythmia score may be plotted against log dose in order to estimate ED™ values for overal l - 110 - antiarrhythmic actions (see sect ion concerned with s t a t i s t i c s ) . At least 6 other arrhythmia scores were invest igated during the course of experimentation, since i t was considered worthwhile to invest igate whether more c lose ly Gaussian-distr ibuted scores could be invented. 2.1.5 ECG changes produced by occlusion Coronary occlusion produces charac te r i s t i c S-T segment elevat ion (Pardee, 1920) and changes in R wave s i z e . Since these e f fec ts are produced by occ lus ion, and are presumably the resu l t of myocardial ischaemia, i t was considered of in terest to measure and record these var iab les . Apart from ourselves (Johnston e t - a l . , 1983a; e t c . ) , only Bernauer (1982; 1983; 1985) has attempted to measure and quantify the ECG changes produced by occlusion in r a t s . The optimum method for measuring and expressing such changes is therefore not wei1 'establ ished. A var ie ty of techniques have been considered as fo l lows. Figure 2 i l l u s t r a t e s the changes in ECG conf igurat ion with time fo l lowing occ lus ion . 2.1.5.1 'S -T ' segment e leva t ion . The posi t ion of the T wave of the ECG in rats i s not as c lear-cut as i t i s in other species. In most chest leads ( including the V3 lead which we use), the T wave i s superimposed upon the terminal portion of the QRS complex, ind icat ing that repo lar iza t ion is beginning before depolar isat ion is complete (Cooper, 1969). In order to measure S-T segment e levat ion , an a rb i t ra ry , but standardised posi t ion for the T wave was determined. Pr io r to occ lus ion, a sample of approximately 50 ECGs recorded at fast chart speed (100 mm/sec) were inspected, and the pos i t ion of the S wave in re la t ion to the Q wave was measured. This was found to be approximately 30 msec (3 mm on the fast speed record) . There- fo re , as an index of S-T segment e levat ion , the height of th is S wave pos i - t ion above i s o e l e c t r i c was measured before (S ) and af ter occlusion (S ). - Ill - KEY: _ ECG lmv '200 msec 1 - HR Time PVC no. After VT no. Occlusion VF no. BP ! 'I U A A 456 beats/mtn 444 beats/mln 453 beats/mln 444 beats/mtn 444 beats/mln -5 m1n -1 m1n 1 min 123 itmHg 123 iraiHg 103 mmHg 2 m1n 3 m1n 97 irniHg 95 umHj l\ h 434 beats/mln 512 beats/mtn 22 PVC 4 m1n 1 VT 5 m1n 1 VF 93 mmHg 105 mmHg "v-A 522 beats/mln 485 beats/mtn 474 beats/m1n 468 beats/mtn 19 PVC 2 PVC 20 mtn 25 min 10 min 15 m1n 83 mmHg 88 mmHg 95 mmHg -I I A VJV 98 mmHg v J \ 474 beats/m1n 500 beats/mtn 519 beats/mtn . 525 beats/mln. 524 beats/mtn. 525 beats/mtn 2 PVC 2 PVC 50 PVC 30 m1n 45 min 1 h 1.25 h 1.5 h 1.75 h 95 mmHg 108 «Hg 105 mmHg 105 nroHg 103 mmHg 105 irniHg I I. ' ' p. 525 beats/m1n 536 beats/m1n 540 beats/mtn 525 beats/mln 544 beats/mtn 503 beats/mtn 16 PVC 69 PVC 24 PVC 36 PVC 2 h 2.25 h 2.5 h 2.75 h 1 VT 3 h 3 VF 103 irniHg 98 irniHg 90 itmHg 85 irniHg 63 PVC 435 PVC 2 VT 3.25 h 29 VT 83 ItmHg 75 nuiHg 522 beats/mtn 517 beats/mtn 512 beats/mln 514 beats/mtn 511 beats/mtn 3.5 h 84 PVC 376 PVC 3.75 h 4 h 70 mmHg 70 limHg 24 PVC 12 VT 4.25 h 65 mmHg 1 PVC 65 mmHg 1 PVC 60 mmHg 505 beats/mtn 1 PVC 4.75 h 60 imHg A., J A A A A - A - A 496 beats/mtn 491 beats/mtn 475 beats/mtn 482 beats/mtn 475 beats/mtn 365 beats/mtn 2 PVC 4 PVC 10 PVC 5.25 h 5 VT 5.5 h 4 VF 52 nmHg 55 mmHg 55 irniHg 5.75 h 3 PVC 6 h 60 mmHg 8 PVC 55 nmHg 24 h 12 PVC/mtn 75 flmHg Figure 2. Anecdote of ECG changes caused by occlusion in a conscious rat - 112 - I soe lec t r i c was defined as the voltage at the foot of the P wave of the preceding beat. Before occ lus ion, values are negative, since the S wave posi t ion is negative to i s o e l e c t r i c . 'S -T ' segment elevat ion begins immedi- a te ly upon occ lus ion, and in order to s impl i fy subsequent ana lys is , a l l negative (pre-occlusion) values are assigned the value zero. 'S -T ' segment elevat ion was expressed in a var ie ty of ways. The var iab le o r i g i n a l l y used in our laboratory was the dimensionless dSTR. This was determined by measuring the R wave amplitude (see below) before occlusion (RQ) and at times af ter occlusion (R t)» and was calculated as ( S t - S Q ) . ( R t / R Q ) . In other words, 'S -T ' segment was corrected for the change in R wave amplitude fo l lowing occlusion as a funct ion of pre- occlusion R wave. Other measures of 'S -T ' segment elevat ion have been subsequently evaluated, as fo l lows; Uncorrected e leva t ion , referred to as ST, calculated as S t mV. Elevat ion corrected for amplidude as a % of R wave amplitude, referred to as ST 2 /R , calculated as ( S t ) 2 / R t mV. Elevat ion as a % of R wave amplitude, referred to as ST%, calculated as (100) (S t ) /R t (dimensionless). The c r i t e r i on for determining which measure was most useful was simply to determine which var iab le produced the largest coe f f i c ien t of var ia t ion (defined as mean divided by standard dev ia t ion) . I t was found during the course of experimentation that ST% was consis tent ly the most precise v a r i - ab le. The coe f f i c ien ts of var ia t ion for ST% were approximately double those found for dSTR. The var iab le dSTR has been shown to be Poisson d is t r ibuted (Johnston e t - a l ; , 1983a). The problem with a l l measures of S-T segment elevat ion is that the var iab le regresses with time in a non-l inear manner. An attempt was made to account for th is by measuring the maximum value of S-T segment - 113 - elevat ion and the time at which th is occurred. Since the development of S-T segment elevat ion is not s t r i c t l y a mono-exponential process (see resu l ts ) i t was decided to measure the maximum value of S-T segment elevat ion (rather than the half-maximum value) . This was carr ied out according to the fo l low- ing c r i t e r i a . S-T segment elevat ion was measured at regular in terva ls af ter occlusion from 100 mm/sec chart speed records. The maximum value of S-T segment e l e - vation was determined by considering each time point consecutively and applying an arb i t rary '5% r u l e ' in the case of ST%, or a '0.05 mV ru le in the case of dSTR. This ru le operates as fol lows (using the der ivat ive ST% as the example). The f i r s t value i s i n i t i a l l y c l a s s i f i e d as the maximum. This is superceded by the f i r s t subsequent value which exceeds i t by 5% or more. In the example below, each successive apparent maximum has been underl ined. The f i na l underlined value, 85% was the value taken as the actual maximum according to the c r i t e r i a out l ined. The time at which the maximum occurred, 120 min, was taken as the 'time of maximum ST%' Time (min) 1 2 5 10 15 30 60 120 180 ST% 6 15_ 17 66 68 80 82 85 88 Of course, i f the re la t ionsh ip between S-T segment elevat ion and time af ter occlusion were a simple saturat ing monoexponential funct ion of the form: 'S -T ' e levat ion = 1/(1 + tau/ t ) where t is time af ter occlusion and tau is the time constant (defined here as the time at which 'S -T ' segment elevat ion is hal f maximal), then tau would be the ideal var iable to record. However, when 'S -T ' e leva- t ion i s plot ted against time ( e . g . , Johnston e t - a l . , 1983a) then i t i s c lear - 114 - that a simple monoexponential equation w i l l not f i t the re la t i onsh ip . Indeed, the re la t ionsh ip does not f i t any simple model. Since the action of drugs on 'S -T ' segment elevat ion was not the primary concern, i t was not considered worthwhile elaborat ing upon the analysis of the var iab le beyond using the semi-empirical approach out l ined above. 2.1.5.2 Pathological- -R-waves. Coronary occlusion in rats produces an i n i t i a l increase in the amplitude of the R wave, followed by a gradual decl ine in amplitude to values smaller than those seen before occ lus ion. An attempt was made to quantify R wave changes using a s imi la r approach to that used in assessing S-T segment e levat ion . R wave amplitude was easier to measure than S-T segment e leva t ion , since the peak of the R wave is c l ea r l y v i s i b l e on fas t chart speed recordings (100 mm/sec). R wave amplitude (mv) was taken as the def lect ion of the peak of the R wave above the i s o e l e c t r i c point . As in the case of S-T segment e levat ion, R wave does not corre la te with time af ter occlusion in a simple manner, and maximum R wave amplitude (using a 0.05 mV rule) and the time at which th is occurred were determined in a manner analogous to the determin- at ion of maximum S-T segment e leva t ion . 2.1.5.3 Pathological - Q waves. The chest lead ECGs of normal rats exh ib i t no Q wave, presumably because the rat heart f ronta l plane axis is perpendicular to the horizontal plane, according to ECG vector analysis (Cooper, 1969). However, coronary occlusion produces a deep Q wave in the chest leads which appears at approximately 2 h af ter occ lus ion, and pers is ts for at least 11 days (Normann et a l ; , 1961; Zsoter and Bajusz, 1962). If lead-I i s recorded, a Q wave appears wi thin 10 min of occlusion in approxim- ate ly 30% of r a t s , long before i t i s present in the chest leads (Kenedi and Losconci , 1973a). Using a V3 lead we ra re ly see a Q wave before 1 h af ter occ lus ion, but i t s presence is almost universal in rats surv iv ing 24 h. - 115 - Recently, attempts have been made to quanti fy Q wave development by measuring the time at which a s ign i f i can t Q wave is present in the ECG. A s ign i f i can t Q wave was defined as a downward def lect ion from i s o e l e c t r i c equal to approximately 10% of the R wave amplitude. 2.1.6 ECG changes produced by drugs Since the ECG i s always recorded before drug administrat ion and again 1 min before occ lus ion , there is an opportunity for assessing the e lec t ro - physio logical e f fects of drugs in the same rats in which ant iar rhythmic /ant i - in fa rc t a c t i v i t y is assessed. This i s achieved by measuring P-R interval and QRS i n t e r v a l . 2.1.6.1 P-R i n t e r v a l . P-R in terva l re f l ec t s the duration of conduc- t ion from high in the atrium through to the AV junct ion and bundle branches (see Horan and Flowers, 1980). Since conduction ve loc i t y in the AV node is much slower than conduction ve loc i t y in the atr ium, in accordance with the lower dV/dt „ of the upstroke of the action potent ia l and the lower e x c i -max t a b i l i t y (Merideth e t - a l . , 1968, and see Introduct ion), then P-R in terva l is essen t ia l l y a re f lec t ion of conduction through the AV node. As such, drugs inf luencing P-R in terva l therefore inf luence AV conduction. The P-R in terva l is f a i r l y easy to record in r a t s , although the foot of the P wave can be somewhat i nd i s t i nc t on occasions, pa r t i cu la r l y when recorded on a Grass Polygraph. P-R interval was measured according to the standard de f in i t i on (from the foot of the upstroke of the P wave to the s ta r t of the QRS). The ef fects of drug treatment on P-R interval were determined by comparing values 15 min before occlusion (pre-drug) with those 1 min before occlusion (4 min af ter drug adminis t rat ion) . 2.1.6.2 QRS i n t e r v a l . The QRS in terva l re f l ec t s vent r icu lar depol- a r isa t ion (see Horan and Flowers, 1980). If the QRS is widened, th is re f l ec ts a d ispers ion, or delay in vent r icu lar conduction, and reduction - 116 - in conduction ve loc i t y . The QRS interval was not measured in the conven- t ional manner (from the beginning of the Q wave to the end of the S) because i t was considered that the posi t ion of the downward going peak of the S was more amenable to measurement than the less wel l -def ined terminal end of the S wave. The ef fects of drug treatment on QRS interval were determined by comparing values 15 min before occlusion (pre-drug) with those 1 min before occlusion (4 min af ter drug adminis t rat ion) . In some experiments, an attempt was made to measure QT i n t e r v a l . The problems involved in measuring QRS were even more of a confounding factor here, since the T wave is the least wel l -def ined of the ECG waves in the ra t . The T wave re f l ec t s vent r icu lar repo la r i sa t i on , and i s therefore dependent on action potent ial durat ion. Therefore, QT in terva l re f l ec ts a combination of vent r icu lar conduction ve loc i t y (governed almost exc lus ive ly by dV/dt ), vent r icu lar action potent ial plateau (governed by i .) and H i d A S i vent r icu lar repo lar isa t ion (governed by inact iva t ion of i $ ^ and act iva t ion of repo lar is ing K + currents, see Introduct ion). It can be seen that QT interval is therefore a rather non-speci f ic va r iab le . In add i t ion , in rats the T wave is superimposed on the QRS (Cooper, 1969; D r i s c o l l , 1980). This makes measurement of QT interval pa r t i cu la r l y d i f f i c u l t . 2.1.7 Measurement of serum K + concentration Serum K + was measured in recent experiments as a consequence of the resu l ts of invest igat ions into mechanisms of arrhythymogenesis (see below). Since i t has been shown that ex t race l l u la r K + r i ses in the ischaemic t issue with a time course corresponding with that of ear ly occlusion-induced arrhythmias (Hirch e t - a l ; , 1980), and since the incidence of arrhythmias fo l lowing myocardial ischaemia is inversely proportional to serum K + c l i n i c a l l y (Nordrehaug and Von der L ippe, 1983; 1985) and experimentally in coronary-1igated rat hearts in v i t ro (Lubbe e t a l . , 1978; Daugherty e t a l ; , - 117 - 1981), i t was decided to monitor serum K + in order to determine whether the antiarrhythmic actions of drugs were re lated to a l tera t ions in serum K + . A blood sample (0.5 - 0.8 ml) was withdrawn from the aor t i c blood press- ure l i ne at approximately 2 h af ter occ lus ion. The sample was spun at 10000 x g for 2 min, using an Eppendorf centr i fuge (Model 3200). The plasma was removed using a small pipette with a rubber bulb attached, and allowed to c l o t . The serum was then pipetted into a 1 ml stoppered vessel (occasion- a l l y a second spin was required to separate the serum from the c l o t ) . The K + concentration was determined in the ana ly t ica l laboratory of the Acute Care Hospital (by the i r techn ic ians) , using a K + - s e l e c t i v e electrode (Kodak Ektachem). 2.1.8 Exclusion c r i t e r i a I t is of paramount importance, when invest igat ing myocardial ischaemia, to ensure that animals are included into the study only when the coronary artery has been occluded and ischaemia has been produced. This is ensured by measuring the 0Z, ex v i vo , according to the technique described above. However, there are other less obvious sources of variance which may jeopar- dise the precis ion and accuracy of an experiment. Over the years, a set of exclusion c r i t e r i a has been developed which is designed to reduce the v a r i - ance not a t t r ibutab le to treatment. It must be stressed that the exclusion c r i t e r i a must be treated in the same manner as a l l other aspects of exper i - mentation and ana lys is , namely that i t must be applied b l i nd l y . The exclusion c r i t e r i a are t ie red in as much as rats must be excluded before, during or af ter occlusion according to a de f in i te chronological sequence. In other words, pre- or post-occlusion exclusion c r i t e r i a must not be applied post-hoc (except when postmortem v e r i f i c a t i o n i s requi red) . - 118 - A. The fo l lowing action is taken in re la t ion to the fo l lowing pre-occlusion abnormal i t ies, with the object of excluding rats in which ei ther the occluder has been acc identa l ly tightened during the in terval between preparation and experimentation, or in which in fect ions are present: a. The presence of a Q wave, as defined above, d ictates obl igatory exc lus ion, on the grounds that a Q wave i s only seen in chest leads in rats fol lowing coronary occlusion (Normann et a l . - , 1961; Zsoter and Bajusz, 1962), and that a Q wave i s associated in our experience only with an in fa rc t or scar t i s sue . b. If more than 5 PVC occur during the 15 min period pr ior to drug administrat ion then the rat should be excluded on the grounds that there is a lesion of some sort in the myocardium which may inf luence the outcome of occ lus ion. c . If there has been > 25% weight loss between preparation and occ lu - s ion , rats are only excluded i f there are other signs of i l l n e s s , such as diarrhoea and/or inflammation associated with surg ical wounds, and/or pre-drug mean a r te r i a l blood pressures of 85 mmHg or less (> 2 s . d . from mean pre-occlusion va lues) . d. I f there are signs of lung i n fec t i on , such as exudate around the snout and/or noisy resp i ra t i on , the rat is excluded. In th is regard, the s e n s i t i v i t y of rats to lung in fect ions i s wel l known. Evans et a l . (1985) reported that male Sprague Dawley rats have often been received from the suppl ier with 'a pulmonary in fect ion characterised by wheezing', and that the morta l i ty in these rats fol lowing coronary occlusion was increased. In ear ly experiments in our laboratory i t was probable that some rats were included despite having an underlying pulmonary i n fec t i on . During the las t 2 years the problem of respi ratory in fec t ion became so severe on occasions that rats died even before preparative surgery had been carr ied out. Mor ta l i ty associated with pneumopathy after occlusion is characterised by extensive exudation of sputum, often bloody, during the f i r s t 5 - 1 5 min af ter occ lus ion, ind ica t ive of severe pulmonary oedema. Postmortem examination reveals mott l ing and occasional ly haemorrhaging of the lungs, and severe pulmonary oedema. Often the thorax is f u l l of serum, and the lungs have a s im i la r appearance to the l i v e r (uniformly red) . Lung swabs and sputum swabs have been examined (by Charles Ford in Medical Micro- bio logy, UBC) and found to cons is tent ly grow monocultures of Psuedo- monas aeruginosa. Current ly , attempts have been made to remove th is problem by s l i g h t l y ac id i f y ing the dr inking water, and al lowing an in terval of at least a week between a r r i va l of the rats from the suppl ier and surg ical preparat ion. These manoeuvres, in conjunction with regular d is in fec t ion of the laboratory with bleach appear to have reduced the incidence of l i fe - th rea ten ing pneumopathy, e. We have had no reason to exclude rats from study on the grounds of low blood pressure. Pa r ra t t ' s group require such exclusion c r i t e r i a since they use acutely prepared anaesthetised animals (Clark et a l . , 1980), which can be obtunded by preparative surgery. The fo l lowing action is taken in the event of the fo l lowing post -occ lu- sion abnormal i t ies, with the object of excluding rats with incomplete or non-existant occlusions and rats which experience reperfusion as a resu l t of the loosening of a defect ive occluder: a . I f there is no increase in R wave, and/or no ST elevat ion fo l lowing occ lus ion, the rat i s excluded: i . i f postmortem examination reveals an unacceptably small OZ or IZ (see below), - 120 - i i . i f the rat dies within 4 hr of occlusion and is found to have in f lamination or scarr ing in the heart , i i i . i f postmortem examination reveals that the occluder i s loose, b. I f , at some time after occ lus ion, the ECG returns to the pre-occ lu- sion conf igurat ion, the rat is excluded: i . i f the OZ and/or IZ are unacceptably small (see below), i i . i f the l i ga to r is found to be loose. c . I f the rat dies within the f i r s t 10 min after occlusion in associat ion with immediate calamatous hypotension, the rat is excluded i f the thorax i s found to contain blood. I t is possible that overtightening of the occluder can cause haemorrhaging from the coronary vesse ls , perhaps as a resu l t of roughness in the region of the f la red outer guide tubing. This par t i cu la r problem is seen in less than 2% of r a t s . d. Animals experiencing fa ta l or non-fatal cardiogenic shock are not excluded from stud ies . Cardiogenic shock has been defined as a f a l l in blood pressure of at least 30%, maintained for at least 30 min, associated with ECG signs of ischaemia, but not caused by arrhythmias (Agress et a l . , 1952). This de f in i t i on adequately describes the syndrome which we recognise as cardiogenic shock in rats subjected to coronary artery occ lus ion . C. The fo l lowing action is taken in the event of the fol lowing postmortem abnormal i t ies, in order to exclude animals with inadequate or abnormal occ lus ions: a. Rats are excluded i f they are found to have an inappropr iately small OZ, defined as < 25 %ventr icular weight (more than 2 s . d . from the mean). - 121 - b. Rats having inappropriately small IZ, defined as < 50% of the OZ are excluded. This c r i t e r i on is somewhate a rb i t ra ry , and i s based on the assumption that since the ra t has no funct ional co l l a te ra l s capable of providing 'myocardial salvage 1 (Johns and Olson, 1954; Selye e t -a l . - , 1960; Maxwell et a l . , 1984; Winkler e t - a l . , 1984; Schaper et a l . , 1986), then the IZ at 24 h should be a reasonably f ixed % of the OZ. This appears to be the case, since to date, our laboratory has not found any drugs which cons is tent ly reduce IZ, which may mean that 'myocardial salvage' is impossible in rats (see Discussion). The arb i t rary nature of th is exclusion c r i t e r i on is not r e a l l y a problem, since no rats have ever been excluded from a study so le l y on the basis of th is c r i t e r i o n . c . I f in farcted t issue or scarr ing in the heart is found in a ra t which dies before 4 h after occ lus ion, the rat i s excluded, on the grounds that in farc t ion is not detectable using TTZ un t i l at least 6 h af ter occ lus ion, and is not well demarcated un t i l at least 10 h af ter occlusion (Hort and Da Cana l is , 1965a). Ear ly scarr ing is i nd i ca - t ive of in farc t ion produced during or shor t ly af ter preparat ion. Less than 1% of rats have been excluded on th i s bas is . d. If pus is found at the occlusion s i t e , the ra t i s excluded. Occasional ly , acutely prepared animals are used for experimentation. The exclusion c r i t e r i a out l ined above are used except where they are obvious- l y inappropriate. In accordance with Par ra t t ' s c r i t e r i a (Clark e t - a l ; , 1980), rats with pre-drug pre-occlusion mean a r te r i a l blood pressures of < 70 mmHg are excluded ( in the present experiments no rats ac tua l ly f a i l ed th is tes t ; pre-occlusion blood pressure was always > 80 mmHg). - 122 - 2.1.9 S t a t i s t i c s In most experiments, a control group of rats was compared with at least two treated groups. The control group was defined as a group of animals treated with drug veh ic le . The var iables compared were e i ther Gaussian (normally) d is t r ibuted or binomial ly d is t r ibuted (Johnston e t a l ; , 1983a). In a l l comparisons, the l im i t of ' s t a t i s t i c a l s ign i f i cance ' was defined as p < 0.05. In accordance with the requirements for undertaking the types of s t a t i s t i c a l tests described below, randomisation to treatment and b l ind analysis of records were carr ied out. Whenever poss ib le , treatment was also given b l i n d . Obviously a procedure such as decerebration or pi th ing (see below) could not be carr ied out b l i n d . The group s i ze was kept to a minimum, and was based loosely on the minimum sample s ize required to reveal a 50 % reduction in VT and VF. The control incidence of VT and VT during the 0 - 4 h period after occlusion in conscious rats i s approximately 90 - 100 % 2 (Johnston et a l . - , 1983a; e t c . ) . In a 1- ta i led chi t es t , the minimum group s ize to reveal a 50 %reduction in VT and VF i s 9 (Mainland et a l ; , 1956). 2.1.9.1 Normalisation procedures. In order to carry out parametric t es t s , such as Duncan's mult ip le range tes t , i t was necessary to ensure that the var iab le was Gaussian d is t r ibu ted . I t has been shown by our laboratory (Johnston et a l ; , 1983a) that many of the var iables measured are log^g- Gaussian d i s t r i bu ted . Therefore, in order to compare means, the log^g values were ca lcu la ted. This manoeuvre was carr ied out for the fol lowing var iab les : PVC, number and duration of VT and VF and time of maximum R wave and S-T segment e levat ion . Correct ions for co-variance by the use of normalization procedures were carr ied out, in cer ta in instances, to improve p rec is ion . For example, i t has been shown that the arrhythmia score (AS) corre lates l i nea r l y with JbT (Johnston e t a l . , 1983a). Therefore AS can be expressed as a function of - 123 - OZ. There i s a major problem assoc i a ted w i th the r e l a t i o n s h i p between AS and OZ, namely tha t AS/JoT becomes meaningless i f AS i s z e r o . Th is i s never the case in c o n t r o l s , but i n d rug - t rea ted groups t h i s i s o f ten the case . An attempt was made to overcome t h i s problem by deve lop ing an ischaemia score (ASMC), which f unc t i ons in the same way as the AS , but i nc ludes scores f o r s i gns o f ischaemia (R wave s i z e , e t c . ) . The f a c t tha t AS c o r r e l a t e s w i th J"5z, but cannot e a s i l y be co r rec ted f o r \J0Z~ was a l s o approached by adding 1 to each AS v a l u e , then c o r r e c t i n g t h i s va lue f o r ^JOZ, g i v i n g the v a r i a b l e (AS + 1)1^01. However, the p r e c i s i o n of OZ i s h i g h , such tha t i n p r a c t i c e and AS are not l i n e a r l y c o r r e l a t e d but are i ns tead s c a t t e r va lues around a s i n g l e p o i n t . Only when smal l OZs are d e l i b e r a t e l y produced ( C u r t i s et a l ; , 1984) i s the r e l a t i o n s h i p between \[0Z~ and AS apparent . When la rge OZs on l y are produced there i s no need to c o r r e c t AS f o r \JoZ. 2 . 1 . 9 . 2 Censo r i ng . Censor ing as a r e s u l t o f death was d i scussed in the s e c t i o n concerned w i th thump-vers ion d e f i b r i l l a t i o n . Censor ing cannot be avoided and i t s e f f e c t can on ly be m in im ised . The type of censor ing which j e o p a r d i s e s an experiment must be i d e n t i f i e d and app rop r ia te a c t i o n taken . The censor ing in t roduced by thump-vers ion was cons idered to j e o p a r d i s e the experiment l e s s than the censor ing assoc i a t ed w i th a l l ow ing animals to d i e . 2 . 1 . 9 . 3 S t a t i s-t-i c a l - t e s t s . For b i n o m i a l l y d i s t r i b u t e d v a r i a b l e s 2 ch i t e s t s were used. Ma in land ' s cont ingency t a b l e s of minimum con t ras t s 2 were used f o r ch i t e s t i n g (Main land e t - a l ; , 1956) . For G a u s s i a n - d i s t r i b u t e d v a r i a b l e s , a n a l y s i s of va r i ance (ANOVA) was c a r r i e d o u t , us ing a U . B . C . s t a t i s t i c s program (Gregg and O s t e r l i n , 1977) . Only i f t reatment c o n s t i t u t e d a s i g n i f i c a n t source of va r i ance accord ing to an F t e s t were means compared (us ing Duncan's m u l t i p l e range t e s t ) . Th is t e s t was app rop r i a t e s i n c e the number of groups exceeded two in every s tudy . In no - 124 - instance was a simple t test used for comparing means, since the t test i s only appropriate when two groups are compared; i f a 1 in 20 probab i l i t y i s set as the l im i t of chance, and 20 groups are compared with a control group, then the probab i l i t y approaches 100% that at least one group w i l l be found to be d i f ferent from cont ro ls , according to the t t es t . Modified t t e s t s , such as Duncan's mult ip le range tes t , avoid th is p i t - f a l l , because the number and type of comparisons are accounted f o r . 2.2 Calcium-antagonist-studies in coronary-^occluded - rats 2.2.1 General experimental design Randomisation to treatment was carr ied out by assembling a table of l i n e s . Each l i ne was assembled by drawing playing cards numbered according to the number of groups in each study. Drug stocks were prepared, and appropriate d i lu t ions made by various members of the laboratory according to spec i f i c ins t ruc t ions . The stock solut ions were coded. Therefore the person preparing a syringe of drug for a par t i cu la r rat was also b l ind to treatment. Lines were chosen in random order by the person preparing the i n j ec t i on . This person prepared a second table in which the ra t code (assigned on the day of preparation) and the coded treatment were noted. In the case of rats treated with unstable drugs ( fe lod ip ine , n i fed ip ine and DHM9), the person making up the drugs was instructed to weigh out a spec i f ied amount of drug for each ra t , and was therefore not b l ind to the treatment. However, the person administering the drug, occluding the coron- ary ar tery , monitoring responses and analysing the records remained b l ind to the treatment. A l l other aspects of study were as out l ined in previous chapters. Up to 4 rats were occluded on each day of experimentation, with a 15 min in terva l between each occ lus ion. In th is manner, rat-1 had passed through the phase of ear ly arrhythmias (Clarke e t a l 1 9 8 0 ; Johnston et a l . , 1983a) - 125 - by the time rat-2 was due for occ lus ion; rat-2 received i t s drug treatment while rat-1 was being monitored for arrhythmias, and so on. For each rat a complete h is tory was kept, both in the laboratory day book and also on indiv idual analysis sheets. The analysis sheets contained a l l information concerning haemodynamic, ECG and arrhythmia data, as wel l as OZ, IZ, serum K + , body weight, date of preparation and occ lus ion, general comments concerning preparat ion, behavioural responses to drug treatment and occlusion and postmortem f ind ings . When the study was complete, the codes were broken and the resu l ts analysed. According to the p r i nc ip le of random- isat ion to l ines i t i s s t a t i s t i c a l l y acceptable to break the code upon the completion of each l i n e . However, th is pract ice was avoided. 2.2.2 Phenethylalkylamines 2.2.2.1 Anipamil -and-Ronipamil. The ef fects of anipamil and ronip- amil on responses to coronary occlusion were invest igated because these anal - ogues of verapamil d i f f e r in respect to the i r calcium antagonist a c t i v i t y , and were therefore considered to be valuble tools for tes t ing the hypothesis that calcium antagonism in the ven t r i c le is antiarrhythmic during acute myocardial ischaemia. Anipamil and ronipamil are analogues of verapamil. In the case of anip- amil the main chain of verapamil has been extended by 10 C-atoms and the methoxy substi tuents removed from the 4 posi t ion on the phenyl r i ngs , leaving l,7-bis-(3-methoxyphenyl)-3-methylaza-7-cyano-nonadecane. Ronipami1 res - embles anipamil except for a lack of methoxy substi tuents on the phenyl r i ngs , i .e i t i s l,7-bisphenyl-3-methylaza-7-cyano-nonadecane. Both drugs have been demonstrated to have anti- ischaemic a c t i v i t y in a var ie ty of preparations (Kovach, 1984; Kretzschmar and Raschack, 1984; Raschack 1984; Urbanics and Kovach 1984). Both anipamil and ronipamil were administered by the p.o. route. The - 126 - doses chosen were 50 and 150 mg/kg, based on prel iminary t o x i c i t y studies in which i t was found that administrat ion of up to 300 mg/kg was non-toxic (higher doses produced death in approximately 50 % of rats given anipamil ; the cause of death was unclear, since i t occurred overnight in every i ns - tance, but was considered to be the resu l t of cardiovascular depression, since the drug caused a progressive decrease in blood pressure with time in these r a t s ) . Both drugs were dissolved in d i s t i l l e d water and administered at a volume of 0.25 ml/kg body weight. Suspensions of 20 or 60 mg/ml of e i ther drug were prepared and gently heated to 60 °C to f a c i l i t a t e d isso- l u t i on , then allowed to cool s l i g h t l y before adminis t rat ion. Control animals received an equivalent volume of d i s t i l l e d water at the same temperature as the drug suspensions. Oral administrat ion was car r ied out 4 h before occlusion during a period of extremely br ie f anaesthesia (1 min) with halo- thane. An in t ragas t r i c tube for drug administrat ion was created by simply f i x i n g a 12 cm length of PE90 tubing to a 3 ml syr inge. The treatments were administered slowly in order to preclude accidental administrat ion into the trachea. The p o s s i b i l i t y of aspirat ion pneumonitis was discounted since rats lack a well-developed vomit re f lex (Briggs and Oehme, 1980). Drugs were made up fresh approximately once per week, re f r igerated and stored in l igh t -proof containers. 2.2.2.2 (-+-)- And (-)-verapami 1. The opt ica l enantiomers of verap- amil were compared for the i r actions on the responses to coronary occlusion in order to test the hypothesis that calcium antagonism in the vent r i c les is antiarrhythmic during acute myocardial ischaemia, and also to examine the complementary hypothesis that the antiarrhythmic act ion of (^J-verapamil in conscious rats (Curt is e t - a l . , 1984) occurred by v i r tue of calcium antagonism in the vent r ic les (see Introduct ion). Both hypotheses predict an a n t i - arrhythmic potency ra t i o equal to the calcium antagonist potency ra t i o in - 127 - the vent r icu la r myocardium, based on the reported potency di f ference between the enantiomers for calcium antagonism ( e . g . , Bayer et a l ; , 1975b; 1975c; Nawrath e t -a l . - , 1981; Ferry e t - a l ; , 1985). Drugs were dissolved in sa l ine and administered at a volume of 0.25 ml/kg by slow i . v . in jec t ion over 10 min, beginning 15 min before coronary occ lu - s ion , according to the protocol out l ined in the general methods sect ion (see Figure 1) . Ten groups of rats (n = 9 per group) were used. Two groups received sa l ine (cont ro ls ) . The remaining groups received ei ther (-)-verap- amil (0.2, 0.6, 2 or 6 mg/kg) or (+)-verapamil (0.4, 4 , 8 or 12 mg/kg). The doses studied were chosen on the basis of the e a r l i e r experiment with (^J-verapamil (Curt is e t - a l . , 1984), in which 6 mg/kg was found to be the EDgQ for reducing arrhythmia score (AS). Drug stocks were made up in advance of the study by d i l u t i o n , and were re f r igerated and stored in l ight -proof vesse ls , coded to ensure double- bl indness. 2.2.3 1,4-Dihydropyridines 2.2.3.1 Felodip ine. Felodipine was evaluated on the basis of i t s s e l e c t i v i t y as a calcium antagonist for the vasculature versus the myocardium ( e . g . , Au and Sutter , 1984). the hypothesis that calcium antagonists reduce ischaemia-induced arrhythmias v ia an action in the ven t r i c les predicts that fe lod ip ine (and other 1.4-dihydropyridine calcium antagonists) w i l l only reduce arrhythmias at high doses. In add i t ion , a calcium antagonist which shows marked s e l e c t i v i t y for the vasculature may prec ip i ta te 1i fe- threatening reductions in blood pressure at doses below those producing blood leve ls su f f i c i en t to reduce i .. in the vent r ic les and reduce the incidence and sever i ty of ischaemia-induced arrhythmias. Therefore, i t was predicted that fe lod ip ine would have l i t t l e i f any antiarrhythmic a c t i v i t y at doses which produced large reductions in blood pressure. - 128 - Eight groups of rats were used (n = 9 per group). In addit ion to the standard large OZ (LOZ) rats (5 groups), 3 groups of small OZ (SOZ) were prepared and subjected to coronary occlusion by Kathy Johnston. The LOZ rats received 1 of 3 doses, 0 .2 , 2.6 or 12.2 ymol/kg, according to the schedule out l ined in the general methods. These doses were 0.08, 1 and 4.68 mg/kg, respec t ive ly , and have been coded as L, I and H (low, intermediate and h igh, respect ive ly) for the purpose of label ing the f igures in the Results sec t ion . Only the lower and higher doses were given to the SOZ ra t s . Felodipine was dissolved in 20 %ethanol in sa l ine and administered at a volume of 0.25 ml/lOOg body weight. Care was taken to protect a l l syringes from d i rec t sunl ight / f luorescent room l i gh t ing before administ rat ion. Twenty seven control rats were used (18 LOZ rats and 9 SOZ r a t s ) . Control rats received 0.25 ml/lOOg body weight of 20%ethanol in sa l i ne . The LOZ and SOZ rats were analysed separately, since treatment was the only known source of variance within the LOZ and SOZ se r i es , but operators were a source of variance between the LOZ and SOZ se r i es . Drugs stocks were made up in advance of the study and were re f r igerated and stored in l ight -proof vesse ls . 2.2.3.2 Ni fedipine and DHM9. These 1,4-dihydropyridines were stud- ied in order to supplement the work with fe lod ip ine . DHM9 has been reported to possess a s e l e c t i v i t y of action for vent r icu lar t issue versus vascular smooth muscle, in contrast with a l l other 1,4-dihydropyridine calcium antag- onists (Clarke e t - a l ; , 1984b). Therefore, according to the hypothesis that calcium antagonism in the vent r i c les is antiarrhythmic during acute myocar- d ia l ischaemia, i t was expected that DHM9 should be more e f fec t i ve in reducing ischaemia-induced arrhythmias than n i fed ip ine at doses producing a s imi la r degree of blood pressure lowering. Five groups of rats were studied (n = 9 per group). Drugs were dissolved in 20%ethanol in sa l ine and administered at 0.25 ml/lOOg body - 129 - weight in accordance with the schedule out l ined in the general methods sect ion. Controls received the ethanol vehic le alone. The doses were 0.5 and 2 mg/kg n i fed ip ine and 5 and 20 mg/kg DHM9. These doses appear to be la rge, but they were based on a prel iminary t o x i c i t y study in which i t was found that conscious rats could to lerate at least 16 mg/kg n i fed ip ine . Although not invest igated in any d e t a i l , i t was considered that the dose- l im i t i ng factor was in fac t the ethanol vehic le rather than the drug i t s e l f . Prel iminary invest igat ions with DHM9 showed that no haemodynamic ef fects occurred at a l l at cumulative doses well in excess of 30 mg/kg. In accor- dance with the general method, stock solut ions were prepared in advance of the study, coded, re f r igerated and stored in l ight -proof containers. Care was taken to prevent exposure of the drug-containing syringes to l igh t before administ rat ion. True blindness could not be achieved in th is study, since the yellow colourat ion of the drugs could be seen in the thin PE tubing entering the subscapular region of the rats at the highest doses (both DHM9 and n i fed ip ine are v i v i d yellow in co lour ) . However, a l l records were analysed b l i n d . 2.3 Arrhythmogenesis and the- ro leof - the-GNS 2.3.1 Introduction The experiments out l ined here were not designed to answer a l l questions concerning arrhythmogenesis in r a t s , but were intended to examine the ro le of the CNS and the sympathetic nervous system. Our laboratory has car r ied out prel iminary experiments for th is purpose in the past (Bott ing et a l ; , 1983). The strategy was to remove, by surgery, the autonomic nervous system in a graded manner in some groups of animals, and replace i t by infusion of catecholamines in others. In addi t ion, the extent of surgery was considered as a possible source of var iance. In conjunction with th is l a t t e r consider- - 130 - a t ion , the ef fects of p i th ing (the most extensive surg ical ablat ion) on possible mediators and modulators of arrhythmias (serum K + , leukocytes and thrombocytes) were determined in a separate ser ies of experiments. A l l groups in th is study consis ted, as usual , of 9 ra t s . Occlusion was car r ied out as described above, and blood pressure was recorded v ia the aor ta , except when stated otherwise. Since there were many groups, and preparation was complicated, a summary table has been prepared which l i s t s a l l the groups (Table 1 ) . 2.3.2 Preparation 2.3.2.1 P i th ing . Male Sprague Dawley rats were l i g h t l y anaesthe- t ised with halothane (4% in oxygen) and intubated. The rats were prepared for occlusion in the normal manner, with an aor t ic blood pressure l i n e , jug- ular intravenous l i ne and Vg ECG leads. These rats were allowed to recover from preparative surgery for approximately 7 days. This group was subse- quently p i thed, as described below, and const i tuted the chron ica l l y prepared (c) pithed (P) group ( i . e . , cP ) . A second group was prepared with occluders, Vg ECG leads, carot id blood pressure l ines and femoral intravenous l ines (by standard ca the te r i - sation techniques using PE tubing). This group const i tuted the acutely prepared (a) pithed (P) group ( i . e . , aP), since they were pithed immediately af ter preparat ion. P i th ing was carr ied out as fo l lows. A s ta in less steel rod (3 mm d i a - meter) was passed through the orb i t and down the spinal cord during l i gh t anaesthesia (1% halothane in oxygen). Immediately afterwards, a r t i f i c i a l resp i ra t ion with 100% oxygen was inst igated (stroke volume 4 ml per 300 g body weight, 54 strokes/min). This a r t i f i c i a l resp i ra t ion regimen has been shown to produce blood gas and blood pH leve ls wi thin the normal range in pithed rats (Milmer and Clough, 1985). The rats were then mounted v e r t i c a l l y - 131 - with the head point ing downward, by securing the t a i l and the p i th ing rod in 2 clamps attached to a re tor t stand. The blood pressure transducer was elevated at th is time in order that i t should remain in the same horizontal plane as the heart . A recta l thermocouple was connected v ia an Indicat ing Contro l ler unit (YSI Model 73ATA) to a 100 W l i gh t bulb, placed 20 cm from the ra t , and a sheet of polythene was placed over the ent i re preparat ion, to maintain body temperatute at 37 - 38 °C. Sal ine (1 ml per 100 g body weight) was then in jected i . v . in an attempt to compensate for the i n i t i a l blood- pressure lowering ef fect of p i th ing . Coronary occlusion was carr ied out 30 min af ter p i th ing . Rats were monitored for 4 h, then sac r i f i ced (exsang- u inat ion) . A l l var iables were measured in the usual way (see general Methods sec t ion) . Infarct s ize (IZ) was not measured, since at least 10 h is required for quant i f iab le in fa rc t development according to the TTZ sta in ing technique (Hort and Da Cana l is , 1965b). 2.3.2.2 Sp ina l i sa t i on . Rats were prepared and occluded in a manner ident ica l with that described for acutely prepared pithed rats (the aP group) with the exception that instead of p i th ing , the rats were sp ina l i sed . This was achieved by inser t ing a steel rod into the sku l l through the foramen magnum at the level of C I , advancing i t r o s t r a l l y and rotat ing i t l a t e r a l l y to mascerate the b ra in . This group const i tuted the acutely prepared (a) sp ina l ised (S) group ( i . e . , aS). As in the case of the aP group, 30 min was allowed between ablat ion in the CNS and occ lus ion. At the time of occ lus ion, spinal ref lexes (foot withdrawal to pinching) were return ing. This group received s im i la r cardiovascular and respi ratory support to the pithed rat groups. This group was also monitored for 4 h then s a c r i f i c e d . 2.3.2.3 Decerebration. Rats were prepared and occluded in a manner ident ica l with the acutely prepared pithed and sp ina l ised groups (aP and aS respect ive ly) with the exception that the ablat ion carr ied out was decere- - 132 - brat ion. Decerebration was achieved by removing the brain ros t ra l to the midcol1icular level using the blunt end of a spatula fo l lowing craniotomy. The empty space was packed with gel foam. A 30 min in terva l was allowed between ablat ion and occ lus ion. In contrast with the previous groups, no cardiovascular or resp i ra tory support was required. After preparation the rats in th is group breathed spontaneously and maintained a high blood press- ure while hor i zon ta l . This group const i tuted the acutely prepared (a) decerebrate (D) group ( i . e . , aD). 2.3.3 Other manipulations The groups described above received ablat ions in the CNS at d i f ferent l eve l s . The pithed groups received complete CNS ab la t ion , the sp ina l ised rats were in tact d i s ta l to C l , and the decerebrate group possessed in tact resp i ra tory and vasomotor centres as well as spinal re f l exes . In a supplementary group of acutely prepared pithed ra t s , an attempt was made to restore the ef fects of the sympathetic nervous system by infusing a mixture of noradrenaline and adrenal ine. This infusion was begun 15 min before occlusion and was designed to elevate blood pressure to leve ls seen in conscious rats (mean of approximately 100 - 110 mmHg; Johnston e t - a l ; , 1983a). The mixture was 4:1 noradrenaline:adrenaline (on a weight b a s i s ) , and the infusion rate was var ied from 0.2 to 5 ug /kg/min noradrenal ine. Infusion volume was kept below 10 ml/kg/h. This group const i tuted the acutely prepared (a) , pithed (P) noradrenaline/adrenaline treated (N) group ( i . e . , aPN). Respiratory and cardiovascular support and a l l other aspects of experimentation were ident ica l with those used for the aP and aS groups. Four groups of ' con t ro l ' rats were used. A standard group of conscious chronical ly-prepared rats were prepared and occluded in the manner described in the general methods sec t ion . This group const i tuted the chron ica l l y prepared (c) conscious (C) group ( i . e . , cC) and served as a control group - 133 - free from recent surgery and ablat ions in the CNS. A second, s imi la r group received an infusion of noradrenaline/adrenaline mixture ident ica l with that administered to the aPN pithed group. This group of conscious rats const i tuted the chron ica l l y prepared (c) conscious (C) noradrenaline/adrenaline (N) treated group (cCN), and served as control for the ef fects of the catecholamine infusion in the absence of ablat ions in the CNS. A th i rd group of rats was prepared in the usual way, but instead of 7 days, these rats were allowed only 1 h to recover from preparative surgery before occ lus ion. This group const i tuted the acutely prepared (a) conscious (C) group ( i . e . , aC), and served as a control for recent surgery in the absence of surgical ablat ion in the CNS. F i n a l l y , a group of rats was prepared for occlusion according to Clark e t - a l ; (1980) using pentobarbitone (60 mg/kg i .p . ) anaesthesia, with the minor exception that our standard occluder (not the s i l k type of Clark e t - a l . ) and our V3 ECG leads (not lead 2) were used. Blood pressure was recorded from the l e f t carot id ar tery , and a cut-down tracheotomy was per- formed (for del ivery of a r t i f i c i a l resp i ra t ion according to the regimen used for the pithed and sp ina l ised groups). Occlusion was carr ied out 30 min af ter preparat ion, and rats were sac r i f i ced 4 h after occ lus ion. This group const i tuted the acutely prepared (a) barbi turate anaesthetised (B) group ( i . e . , aB), and served as a control for recent minor surgery in the absence of surgical ablat ion in the CNS but in the presence of chemical ablat ion in the CNS (anaesthesia). - 134 - Table 1. Summary of Groups in the CNS Ablations Study Acutely Prepared-Groups Code Conscious aC Anaesthetised aB Pithed aP Pithed plus catecholamine infusion aPN Spinal ized aS Decerebrate aD Chronical ly-Prepared Groups Code Conscious cC Conscious plus catecholamine infusion cCN Pithed CP Others Code Isolated perfused hearts I Detai ls of the surg ica l preparation are given in the tex t . - 135 - An addit ional group was included in th is study, a group of iso lated hearts perfused v ia the aorta at 37 °C with a modified Krebs Henseleit buffer containing 5.3 meq/1 K + . This group was prepared and occluded by Reza Tabrizchi according to the method of Kannengeisser et a l . (1975). This group const i tuted the iso la ted heart (I) group, and served as a general reference group. 2.3.4 S t a t i s t i c s Variables were measured and analysed as described for the standard occlusion preparat ion. The f u l l data set for th is study is extremely complicated, and i t is possible to discuss at great length the possible impl icat ions of the r e s u l t s . In order to s impl i fy matters somewhat, s t a t i - s t i c a l s ign i f icance has only been noted between the chron ica l l y prepared conscious group (cC) and the other groups. 2.4 Prel iminary-Screen for drug a c t i v i t y in acute-ischaemia 2.4.1 Introduction Although the conscious ra t preparation for invest igat ion of myocardial ischaemia and in farc t ion is capable of providing up to 10 data points per var iab le per week, i t was decided to develop a prel iminary screen for assessing new drugs with which 4 times as much information could be generated in the same time. This preparation was designed simply to provide i n fo r - mation concerning morta l i ty and in fa rc t s i z e . However, assessment of animal behaviour was also used, in order to attempt to del ineate between morta l i ty resu l t ing from vent r icu lar arrhythmias (VF) and other causes. 2.4.2 Preparation Male Sprague Dawley rats were prepared for occlusion in the manner described in deta i l in preceding chapters. No intravascular catheters and ECG leads were implanted, only the occluder. By th is technique, 16 rats could be prepared d a i l y , with ease. The rats were housed ind i v idua l l y and - 136 - allowed approximately 7 days before being subjected to coronary occlusion in the usual manner. 2.4.3 Experimental endpoints Following occ lus ion, rats were ca re fu l l y monitored by observation alone. The behaviour of each rat was recorded on ind iv idual analysis sheets. A l l observation and analysis was car r ied out b l i n d . 2.4.3.1 Def in i t ions . Certain behaviours were ca re fu l l y categorised for each time in terva l fo l lowing occlusion (the time in te rva ls corresponded with those used for the standard occlusion preparat ion). The time of death was noted, and OZ was measured in the usual manner. In rats surviv ing 24 h, the IZ was also measured. Behaviour was c l a s s i f i e d and assessed according to the fol lowing sub- jec t i ve c r i t e r i a . Morbidity was graded according to the presence or absence of the fo l lowing 3 behaviours, respi ratory d is t ress (which was defined as laboured breathing), head-down posture, and prone posture. These 3 end- points were considered to be signs of cardiogenic shock (Agress e t - a l . , 1952). If nothing more serious than minor panting was present, the rat was classed as normal. Severe vent r icu lar arrhythmias were diagnosed on the basis of sudden 'convuls ive- type ' behaviour. This was defined as sudden f renzied attempts to climb out of the home cage accompanied by convulsive-type limb movements and a sudden blanching of the ears and eyes. This behaviour i s en t i re l y cha rac te r i s t i c of VF (or severe torsade de pointes) las t ing longer than 10 sec in the standard instrumented rat preparat ion, and has never been observed in associat ion with other sequelae of coronary occlusion such as AV block or cardiogenic shock. Occasional ly , rats develop fa ta l pulmonary oedema fol lowing occ lus ion. The associated behaviour is d i f ferent from that caused by VF. Fatal pulmon- - 137 - ary oedema in conscious rats is associated with f renzied behaviour, but i s character ised by the absence of pal ing of the ears and eyes and the presence of expectoration of copious sputum (often bloody). Furthermore, convuls ive- type behaviour resu l t ing from VF las ts for less than 10 sec, since syncope rap id ly fol lows the loss of cardiac output, whereas the f renzied behaviour associated with acute pulmonary oedema general ly las ts for more than 30 sec, since cardiac output is maintained. Acute pulmonary oedema is always f a t a l ; the death throes are associated with serious expectoration and an unpleasant gagging sound. In contrast , sudden convulsive-type behaviour resu l t ing from vent r icu lar arrhythmias is never associated with sudden expectorat ion, and i s not invar iab ly f a t a l ; often a rat w i l l loose consciousness, then suddenly the ears and eyes w i l l turn pink and the rat w i l l regain consciousness without apparent i l l e f fec t . F i n a l l y , acute pulmonary oedema is general ly associated with a highly pathognomic prodroma, involv ing the signs of morbidity out l ined above. Sudden convulsive-type behaviour i s not associated with any prodroma. 2.4.3.2 Val idat ion of behaviour- end-points. The behaviours des- cr ibed above were tested in a b l ind study in which 12 standard f u l l y - i n s t r u - mented rats were prepared and occluded in the usual manner, while an observer recorded behaviour in the manner described. These par t i cu la r ra ts were not d e f i b r i l l a t e d i f VF developed, but were allowed to die (or spontaneously d e f i b r i l l a t e ) and were therefore ident ica l to rats subsequently used for comparing the verapamil enantiomers. 2.4.4 Comprison of (+)-, ( - ) - and (±)-verapamil As part of the invest igat ions of the actions of verapamil in myocardial ischaemia and i n fa r c t i on , a study was carr ied out with the opt ica l enantio- mers and the racema'te, using the above-described technique. Since the prec is ion of th is method was unknown, and since up to 16 rats could be - 138 - occluded and observed per day, i t was decided to use large group s izes (n = 25 per group). Rats were prepared and allowed to recover from surgery as descr ibed. On the day of study, up to 16 rats were stat ioned on a s ing le laboratory bench. Drugs were administered v ia a super f i c ia l t a i l vein using a bu t te r f l y hypo- dermic needle while each rat was temporari ly restra ined in a perspex res t - r a i n e d Drugs were in jected over a 10 min per iod, and 5 min was then allowed before occ lus ion , in a manner analogous to the method used for studies in standard ful ly- instrumented r a t s . The cycle time ( in terva l between success- ive occlusions) was therefore 15 min. This in terva l was designed such that close attent ion could be given to the behaviour of each rat during the c ruc ia l f i r s t 5 - 1 0 min af ter occlusion (when the highest incidence of VF and VT occurs) . Rats were continuously monitored for at least 4 h af ter occ lus ion . The fo l lowing treatments were administered: sa l ine (0.25 ml/100 g body weight), (+)-verapamil (6 mg/kg), (±)-verapamil (6 mg/kg) and (-)-verapamil (6 mg/kg). The dose of 6 mg/kg was chosen in accordance with the previously determined E D ^ Q for (±)-verapamil in ful ly- instrumented conscious rats (Curt is et a l . , 1984). The object of the study was essen t i a l l y ident ica l with the object of the study with the enantiomers in fu l ly- instrumented rats described previously , namely to test the hypothesis that the antiarrhythmic actions of ( ^ - ve rap - amil occur by v i r tue of calcium antagonism. The hypothesis predicts that the incidence of convulsive-type behaviour should be reduced more by (-)-verapamil than by (+)-verapami1, with (±)-verapami1 having intermediate a c t i v i t y . In addi t ion, the experiments were also carr ied out in order to examine and character ise th is new method of assessing drug a c t i v i t y in acute myocardial ischaemia and in fa rc t ion on the basis of a combination of object- - 139 - ive (OZ, IZ and mortal i ty) and subject ive (behavioural) end-points. 2.4.5 S t a t i s t i c s The incidence of morta l i ty and VF are binomial ly d is t r ibu ted in conscious rats fo l lowing occlusion (Johnston e t -a l . - , 1983a). Therefore, the incidence of sudden convulsive-type behaviour, morta l i ty and morbidity were analysed 2 using chi , in the manner described previously . The resu l ts were categor- ised in terms of the f i r s t 0.5 h and the 0.5 - 4 h periods af ter occ lus ion, in accordance with the bimodal d i s t r i bu t ion of ischaemia-induced arrhythmias with time in conscious rats (Johnston e t - a l . , 1983a). 2.5 Electr ical ly-^ induced-arrhythmias- in conscious rats 2.5.1 Introduction The opt ica l enantiomers of verapamil are known to block i N a in ven- t r i c u l a r muscles at high concentrations (Nawrath et a l . - , 1981). These concentrations are 50 - 150 times in excess of those necessary to abol ish i' s l- and c o n t r a c t i l i t y in the normal ven t r i c le (Nawrath et a l . - , 1981). Therefore, the p o s s i b i l i t y that blockade of i ' N a (sodium channel blockade) contr ibutes to the pharmacological actions of (^ -verapami l and i t s enant io- mers in vivo seems highly un l i ke l y . However, th is p o s s i b i l i t y was neverthe- less examined by comparing the actions of the enantiomers for the i r a b i l i t y to inf luence arrhythmias induced by e l e c t r i c a l st imulat ion of the l e f t ven t r i c le in conscious ra t s . I t was previously shown that quinidine i n f l u - enced e l e c t r i c a l l y induced arrhythmias in conscious rats (Curt is e t a l ; , 1984) at the same dose which reduced occlusion-induced arrhythmias (arrhyth- mia score) by 50% (Johnston et a l ; , 1983a), whereas (^J-verapami1 had no such actions at the dose reducing arrhythmia score by 50% (Curt is et a l ; , 1984). Therefore the assessment of e lec t r i ca l l y - i nduced arrhythmias was considered to be useful for d i f f e ren t ia t i ng between Na + channel blockers and calcium antagonists. - 140 - 2.5.2 Preparation Male Sprague Dawley rats were prepared in a manner ident ica l with that for rats used for coronary occlusion s tud ies , with 1 exception. Instead of a coronary occluder, 2 tef lon-coated s ta in less steel wire electrodes were implanted approximately 3 mm apart into the l e f t v e n t r i c l e , in approximately the centre of the l e f t vent r icu lar w a l l . This was achieved by plunging the e thano l - s te r i l i sed leads into the v e n t r i c l e , using a 23 gauge hypodermic needle as a locator . A l l leads and l ines were ex ter io r ised in the subscap- ular reg ion, and approximately 7 days was allowed for recovery from th is preparative surgery. On the day of study, the rats were connected to the standard devices for intravenous drug adminis t rat ion, and blood pressure and ECG recording. 2.5.3 Experimental end-points The var iables described below were each measured 3 times every 5 min. Their measurement has been described previously , but not in great deta i l (Curt is et a l ; , 1984). Stimulation of the l e f t ven t r i c le with square wave pulses was undertaken using a Grass st imulator (Model SD9), which was c a l i - brated using a standard voltmeter (Beckman Model 3020B). A permanent record of the ECG and blood pressure was made using the standard Grass polygraph, while the ECG was also continuously monitored using a delayed loop o s c i l l o - scope (Honeywell Type E for M). Discr iminat ion of end-points was carr ied out using the osc i l l oscope . 2.5.3.1 Maximum- following- frequency. Maximum fol lowing frequency was defined as the frequency at which the ven t r i c le f a i l e d to fo l low the stimulus on a 1:1 bas is . Fa i lu re to fol low was accompanied by a character- i s t i c blood pressure change; the dropped-beat produced a drop in blood pressure, and the subsequent beat produced a large pulse as a resu l t of the increase in f i l l i n g time and end-dioasto l ic pressure. The blood pressure - 141 - record could be used for d iscr iminat ion of the end-point, but in th is ser ies of experiments the osc i l loscope alone was used. Maximum fol lowing frequency was determined at 0.8 msec pulse width and twice threshold vol tage, by slowly increasing the st imulat ion frequency from a baseline of 5 Hz. 2.5.3.2 Threshold vol tage. The threshold voltage for inducing VT was determined at 50 Hz and 0.8 msec by the same p r inc ip le used for deter- mining maximum fol lowing frequency. The high frequency was used in order to maximise the probab i l i t y of de l iver ing a pulse during the vulnerable per iod, the terminal portion of the QT (de Boer, 1921; Wiggers and Wegria, 1940). In addit ion to the appearance in the ECG of VT, the threshold voltage for VT was character ised by a drop in blood pressure. VT was almost always non- sustained. 2.5.3.3 Threshold pulse- width. The threshold pulse width for inducing VT was determined from the osc i l loscope according to the same c r i t e r i a as those used for measuring the threshold vol tage. This var iab le was measured at 50 Hz and twice threshold vol tage. 2.5.4 Comparison of ( + )- and (-)-verapamil In the e a r l i e r study with (±)-verapamil , i t was found that 6 mg/kg had no ef fect on the var iables in question in conscious rats (Curt is et a l . , 1984). Therefore, i t was decided to evaluate only the equivalent of the highest doses of the enantiomers administered to coronary-occluded ra t s , since i t was not expected that the lower doses would inf luence the var iab les . Therefore e i ther 8 or 12 mg/kg ( + )-verapami 1 or 2 or 6 mg/kg (-)-verapami1 (n = 6 per group) were administered v ia the vena caval i . v . l i n e , over a 10 min per iod. The st imulat ion var iables were recorded 1 min before the s ta r t of drug administrat ion and 5 min af ter completion of drug administra- t i o n . The %changes in the var iables were recorded and analysed. The values for the var iables were noted on the ECG t races , and were analysed b l i n d . - 142 - Mean %changes in the var iables were compared by ANOVA and Duncan's mult ip le range tes t . 2.6 Haemodynami c and EGG - ef fects of -ca lc i um- antagon i s t s - i n pi thed-rats 2.6.1 Introduction In previous sect ions, the procedures for assessing the antiarrhythmic act ion of the enantiomers of verapamil were described. The premise was that (-)-verapamil is more potent as a calcium antagonist than (+)-verapamil. It was decided that rather than re ly on l i t e ra tu re values for the re la t i ve potency of the enantiomers of verapamil (Bayer e t - a l ; , 1975b; 1975c; Ferry e t - a l . , 1985; Nawrath et a l . - , 1981; Raschack 1976b; Echizen e t -a l . - , 1985), which give disparate values ranging from 4 to over 100 in favour of ( - ) -verapami l , various attempts should be made to gauge the re la t i ve potency of the enantiomers ourselves. The f i r s t method was to examine the ef fects of the enantiomers on blood pressure, heart ra te , P-R in terva l and QRS i n - terval in conscious ra t s . This par t i cu la r experiment const i tuted a portion of the occlusion study with the enantiomers, since a comparison of values before and af ter drug administrat ion provided th is information. However, i t was decided to supplement th is information with other s tud ies . Ea r l i e r work with pithed rats (see sect ion concerned with arrhythmo- genesis in myocardial ischaemia) had shown that i t was possible to produce a pithed ra t preparation with a high mean blood pressure of at least 60 mmHg, which remained v iable for well over 4 h. I t was decided to use th is prepa- rat ion to invest igate the actions of the verapamil enantiomers in the absence of autonomic tone and re f lexes . This information was considered to be of i n te res t , not least because there appears to be a large d ispar i t y between the potency of (^-verapami l in anaesthetised versus conscious rats (Curt is et a l . , 1984). - 143 - 2.6.2 Preparation The pithed rat preparation was essen t ia l l y the same as that described prev ious ly , with the exception that in every instance blood pressure was recorded from the aor ta, and drugs were administered into the vena cava. Each preparation was allowed to s t a b i l i z e for at least 1.5 h before study, and mean aor t ic blood-pressure and the ECG were pe r iod i ca l l y recorded during th is t ime. 2.6.3 Variables measured Mean blood pressure, heart ra te , P-R in terva l and QT in terva l were recorded. The ECG var iables were measured at a chart speed of 100 mm/sec, using standard ECG c r i t e r i a (Horan and Flowers, 1980) as modified by Dr i sco l l (1980) for the ra t . Reductions in blood-pressure and prolongation of P-R in terva l are ef fects cha rac te r i s t i ca l l y produced by calcium antagonists, and are considered to be the resu l t of calcium antagonism (see Nayler and Horowitz, 1983). 2.6.4 Comparison of (+)- and (-)-verapamil Once mean blood pressure had reached a stable level (no more than a 5 mmHg var ia t ion over a 10 min per iod) , increasing doses of e i ther (+)-verapami1 (0.4, 0.8, 4 and 8 mg/kg) or (-)-verapami 1 (0.02, 0.06, 0.2 and 0.6 mg/kg) were administered (n = 6 per drug). Each successive dose was administered as a slow i . v . in jec t ion over a 10 min per iod. I t was observed in p re l im in- ary experiments that the peak blood-pressure lowering e f fec t of the enantio- mers occurred during the f i r s t 30 sec af ter f i n i sh i