ST Segment Elevation at the Surface of a Healed Transmural Myocardial Infarction in Pigs
Conditions for Passive Transmission From the Ischemic Peri-Infarction Zone
Background Ischemia of the myocardium surviving an infarction induces ST segment elevation in infarct-related ECG leads. In cases with no viable tissue, ischemia adjacent to the infarction could induce a similar ECG pattern if there is ST segment potential transmission through the necrotic scar. We analyzed whether acute ischemia adjacent to a healed infarction with no viable tissue may induce ST segment elevation on the surface of the necrotic scar.
Methods and Results Epicardial ST segment changes elicited during 30 minutes of acute reocclusion of the left anterior descending (LAD) coronary artery 2 cm above the first diagonal branch were analyzed by 32-channel mapping in 18 chloralose-anesthetized open-chest pigs with 1-month-old anterior infarctions induced by permanent ligature below the first diagonal branch (group 1). The effect of a previous infarction on the magnitude of ischemic ST segment changes was assessed by similar mapping in 21 control pigs submitted to a LAD ligature 2 cm above the first diagonal branch (group 2, n=11) or just below this branch (group 3, n=10). Myocardial perfusion after coronary ligature was estimated in 7 pigs with chronic infarction and in 3 control pigs by mapping of myocardial technetium-99m–methoxyisobutyl isonitrile (99mTc-MIBI) activity in transmural samples underlying each epicardial electrode. The width of cell layers surviving the infarction was measured and their viability after 60 minutes of coronary reocclusion was assessed by intracellular glycogen staining. Reocclusion of the LAD induced parallel ST segment elevation at the peri-infarction zone and at the necrotic scar, although in the latter region the changes were less marked (maximal ST segment, 8.4±3.0 mV versus 2.7±1.8 mV, ANOVA, P<.001). ST segment elevation inside the scar was greater at the margins (3.9±1.8 mV) than at sites 20 mm toward the center (2.8±1.7 mV, P=.003). The necrotic area was virtually devoid of surviving cells except for a 0.22±0.04-mm-wide subendocardial band that continued to show a positive intracellular glycogen reaction after the second LAD ligature. Acute ischemia adjacent to the infarction (group 1) induced lower ST segment elevation than acute ischemia at a comparable cardiac region in noninfarcted pigs (group 2) (ANOVA, P=.02), despite the fact that these areas developed similar underperfusion after coronary occlusion (percent MIBI activity of that in normal myocardium, 7±8 versus 7±6, P=NS). ST segment changes in group 2 pigs were comparable to those induced in group 3 pigs with a 2-cm-lower coronary occlusion.
Conclusions Acute ischemia adjacent to a chronic infarction induces ST segment elevation at the surface of the scar despite the virtual absence of viable tissue within the infarction. Data suggest a passive ST segment potential transmission through the infarction. Moreover, ischemia adjacent to a chronic infarction induces lower ST segment elevation than ischemia not adjacent to a necrosis. The mechanisms accounting for these regional differences are probably independent of collateral myocardial perfusion and ischemia extension.
The mechanism responsible for ST segment elevation in infarct-related ECG leads is not entirely delineated. It is generally accepted that left ventricular dysfunction or aneurysm causes exercise-induced ST segment elevation in leads with Q waves in most patients with previous myocardial infarction.1 2 3 However, other investigators have considered this ECG pattern indicative of ischemia of residual viable myocardium and potentially a clinical marker to identify patients who will benefit from myocardial revascularization.4
Hypothetically, in cases with no viable tissue within the infarction, the development of ischemia at the adjacent normal myocardium could induce ST segment elevation inside the necrotic area provided that the scar might have the capacity to passively transmit ST segment potential changes generated at the acute ischemic peri-infarction zone.
To test the hypothesis that ST segment elevation can be transmitted through a necrotic scar, we performed a correlative anatomopathological and an ECG study in a porcine model of acute ischemia adjacent to a chronic anteroseptal infarction free of viable tissue. ST segment elevation in the areas of interest was measured selectively by epicardial mapping.
We also analyzed whether the magnitude of ST segment elevation induced by peri-infarction ischemia is influenced by the presence of a necrotic scar. The electrophysiological properties of the myocardium adjacent to a necrosis may be altered as a result of infarction-induced local autonomic denervation5 6 7 and abnormal cellular gap junction distribution.8
Data from this study were obtained from 39 pigs (15 to 20 kg) that were survivors of an initial series of 52 pigs submitted to two operations. The first intervention was done under sterile conditions and consisted of a left lateral thoracotomy at the level of the fifth intercostal space. General anesthesia was induced after administration of metomidate (4 mg/kg IV) followed by sodium thiopental (30 mg/kg IV). The rib was removed to gain better access to the underlying pleura and pericardium, which were opened. In 18 pigs, the left anterior descending (LAD) coronary artery was looped with a Mersilene 5/0 snare just below the first diagonal branch and was permanently ligated (group 1). In the remaining 21 pigs, the LAD was not occluded. The chest was immediately closed in layers, and the pleural air was aspirated. Cardiac rhythm was monitored during the first 2 hours of coronary occlusion to treat malignant ventricular arrhythmias by external electric DC countershock. Ventricular fibrillation occurred in 33% of the occluded pigs. The animals were allowed to recover and received standard postoperative care, including analgesics and antibiotics during 48 hours. Lidocaine (100 mg IM) was administered prophylactically only in those pigs undergoing acute coronary occlusion. One month after coronary ligature, 11 pigs of this group died (6 of them died during the first 48 hours), and 2 others died during anesthesia for the second intervention. Therefore, a total of 39 pigs (18 infarcted and 21 controls) entered in this study.
The second operation was done 1 month after the first thoracotomy and consisted of a midsternotomy under intravenous α-chloralose anesthesia (100 mg/kg IV). A moderate pericardial inflammatory reaction was observed in most cases. The pericardium was carefully detached, and the free margins were sutured to the borders of the sternotomy to cradle the heart. The course of the LAD was identified, and a Mersilene 5/0 snare was positioned around this artery. The ligature was placed 2 cm above the first coronary occlusion in the infarcted pigs (group 1) and at a similar level in 11 out of 21 control noninfarcted pigs (group 2). In the remaining 10 control pigs, the ligature was placed just below the first diagonal branch (group 3). Aortic blood pressure was measured through a saline-filled cannula introduced into the left femoral artery. At regular intervals, blood gases were measured and corrections were made to keep them within normal limits. Pigs were handled in accordance with the position of the American Heart Association and European Community Rules on Research Animal Use. This study was approved by the ethics committee of our institution.
During the second intervention, a rubber membrane containing 32 electrodes, spaced 5 mm and arranged in three columns, was positioned parallel to the LAD (Fig 1⇓) and sutured to the epicardium. The membrane covered an area extending from the ventricular apex (site of chronic infarction in group 1) toward the proximal myocardium that became acutely ischemic during the study. To record the extracellular DC potentials, we used a 32-channel differential amplifier system, and the signals were digitized at 500 Hz and stored in a computer. Nonpolarizable electrodes were made with 0.5-mm-diameter polyethylene tubs containing a cotton thread imbibed with isotonic saline solution; the electrodes were connected to the amplifiers through a chloride silver interface. The 0-mV potential reference was taken at the mediastinal fat.
In each electrogram, we measured displacement of the TQ-ST segment from the zero-potential level induced during acute coronary occlusion. In this model, acute ischemia elicited TQ segment depression and ST segment elevation.9 Data were expressed as the total TQ+ST displacement (millivolts) to be comparable to the ST segment displacement recorded by conventional clinical ECG.
To correlate the morphologic changes in epicardial electrograms with the standard surface ECG, a 12-lead recording was obtained during coronary reocclusion in seven pigs. In these cases, the ECG was recorded after closure of the midsternal thoracotomy.
Mapping of Myocardial Technetium-99m–Methoxyisobutyl Isonitrile Activity
Although coronary collateral flow is virtually absent in healthy porcine hearts,10 we considered the possibility that hearts with chronic infarction may develop collateral coronary perfusion at the adjacent normal myocardium, which, in turn, could affect the ST segment changes induced by ischemia at the peri-infarction area. Myocardial technetium-99m–methoxyisobutyl isonitrile (99mTc-MIBI) activity underlying each epicardial electrode was mapped to compare the degree of underperfusion in the peri-infarction zone (group 1 pigs) with that achieved in a similar anatomic region made ischemic by a primary proximal LAD ligature (group 2 pigs). Myocardial perfusion assessed by 99mTc-MIBI activity strongly correlates with myocardial blood flow measured by radioactive microspheres during coronary occlusion in pigs and dogs.11 12 A dose of 1.7±0.2 mCi of 99mTc-MIBI was injected into the right cavity 2 minutes after coronary occlusion in seven pigs of group 1 and in three pigs with a LAD ligature like that in group 2. After 30 minutes of ischemia, a bolus of 5 mL fluorescein (20%) was injected into the left atrium, and the heart was removed and cut from the apex to the AV ring in three 5-mm-thick parallel slices, each containing one column of electrodes. Position of each electrode relative to normal and acute ischemic area was determined by delineating fluorescein-stained and -unstained regions (Fig 2⇓). In pigs of group 1, the infarct scar was identified after incubation of the slices with triphenyl tetrazolium at 37°C (unreactive white area). The myocardium underlying each epicardial electrode was divided into transmural sections, resulting in a total of 32 sections for each pig (Fig 2⇓). Three additional samples from remote normal left ventricular myocardium were taken as controls. In five cases, all transmural biopsies were subdivided into epicardial, midwall, and endocardial thirds. Gamma scintillation counting of the myocardial samples was performed with an LKB Wallac model 1282 Universal Gamma Counter, and values were corrected for tissue weight. To facilitate comparisons between pigs, myocardial 99mTc-MIBI activity was expressed as percentage of that in remote normal myocardium. Scintigraphy of selected slices was performed with an Elscint Apex SP4 gamma camera before these preparations were cut into transmural segments.
Hearts from pigs not injected with 99mTc-MIBI were removed at the end of the study. The anatomic position of the membrane containing the epicardial electrodes was indicated by leaving fine epicardial suture snares under the corners of the membrane. Locations of the first and last electrodes in each column were indicated by an ink mark; locations of the remaining electrodes were indicated by dividing these columns at 5-mm intervals. After 72 hours of immersion in 95% ethanol solution, the left ventricle was sliced in parallel sections, each containing a column of 10 to 11 electrodes. The entire slices were included in paraffin and stained with Masson’s trichrome and hematoxylin-eosin stains. In these preparations, we measured under a microscope the thickness (millimeters) of subepicardial and subendocardial layers of cells surviving at the center and at the margins of the necrotic scar.
To determine whether cells surviving inside the necrotic region become ischemic after a proximal reocclusion of the LAD, we performed intracellular glycogen staining 60 minutes after coronary reocclusion in three additional infarcted pigs like those of group 1. These hearts were cut in four 5-mm-thick slices parallel to the LAD. Transmural sections (1 to 2 cm long) of these preparations were taken at the center of the scar, at the acute ischemic peri-infarction zone, and at the remote normal myocardium. They were immediately frozen with liquid nitrogen and processed for periodic acid–Schiff (PAS) reaction. This reaction was evaluated under ×400 magnification. In the pig heart, this ischemic period produces a complete and homogeneous depletion of intracellular glycogen at the central ischemic area associated with tissue lactate accumulation and severe creatine phosphate (CP) and ATP depletion.13
The LAD was ligated after baseline recording of aortic pressure, conventional ECG, and epicardial mapping. During 30 minutes of acute ischemia, we recorded blood pressure, ECG leads I and II, and epicardial electrograms every minute. A 12-lead ECG was obtained in seven pigs 30 minutes after coronary reocclusion. Scintigraphic and anatomic studies were performed at the end of the study.
Among the total electrode population, we purposely identified those electrodes over the area of transmural necrosis by direct anatomic examination. The remaining electrodes were considered to be located on the acute ischemic region if they showed ST segment elevation greater than 2 mV or if were placed over the fluorescein-unstained region. Electrodes depicting isoelectric or reciprocal ST segment changes or being placed over the fluorescein-stained region were considered to contact the normal myocardial region.
Regional differences in the magnitude and time course of the ST segment elevation and on the myocardial 99mTc-MIBI activity were statistically evaluated by applying the ANOVA test for repeated measures. A P value <.05 was considered significant.
All pigs developed a transmural myocardial infarction with a sharply demarcated border zone 1 month after LAD occlusion. Within the necrotic scar, there was a homogeneous subendocardial layer of surviving cells (Fig 3A⇓) that measured 0.22±0.04 mm. In contrast, the epicardium showed inconstant and thinner rims (0.05±0.01 mm, P<.001) of surviving cells. After coronary reocclusion, the subendocardial cells subsisting inside the infarction exhibited a positive intracellular glycogen staining (Fig 3B⇓) comparable to that shown by normal cells beyond the peri-infarction ischemic myocardium (Fig 3C⇓, right). A normal PAS reaction was also disclosed by the subendocardial cells of the acute ischemic peri-infarction area (Fig 3D⇓).
ST segment changes induced by coronary reocclusion in pigs of group 1 were analyzed in 319 electrodes (155 at the surface of the scar and 164 over the ischemic peri-infarction zone). At baseline, pigs with chronic infarction showed QS complex with flat or negative T wave in electrodes at the ventricular apex, whereas pigs with no previous infarction showed rS complex with positive T wave in comparable apical sites (Fig 1A⇑). On coronary reocclusion, electrodes at the ischemic peri-infarction area recorded ST segment elevation and large R waves (Fig 1C⇑). Despite the lack of viable tissue within the infarction, electrodes overlying the infarct scar recorded ST segment elevation on previous QS complex (Fig 1C⇑). As Fig 4⇓ shows, ST segment elevations at the necrotic region and the adjacent ischemic myocardium show comparable time courses, but the ST segment changes were smaller inside the necrosis (ANOVA, P<.001). ST segment elevation inside the scar was greater in electrodes placed toward the margin of the scar than at sites located 20 mm toward the center of the necrotic region (3.9±1.8 versus 2.8±1.7 mV, P=.003).
Ischemia adjacent to a chronic infarction induced lower ST segment elevation (ANOVA, P=.02) than ischemia at a comparable cardiac region in group 2 pigs with no previous infarction (Fig 5⇓). These data were obtained by matching 111 electrodes at the peri-infarction zone in the first consecutive 11 pigs of group 1 with 111 electrodes (identical number of electrodes per case) at a comparable cardiac region in the 11 pigs of group 2. To rule out the possibility that these regional differences occurred because group 1 pigs had a smaller acute ischemic area than group 2 pigs, we analyzed the ST segment changes induced in areas of a different extension. As Fig 6⇓ shows, group 2 pigs with a LAD occlusion 2 cm above the first diagonal branch show similar ST segment elevation as pigs of group 3 with a 2-cm-lower LAD occlusion.
The surface ECG mostly correlates with the morphological changes in epicardial electrograms. As Fig 7⇓ shows, coronary reocclusion induced ST segment elevation in infarct-related lead V1 and monophasic potentials in leads V3 and V4 overlying the peri-infarction zone.
Mapping of Myocardial 99mTc-MIBI Activity
Myocardial samples from the ischemic (fluorescein-unstained) region in control pigs and from the acute ischemic peri-infarction zone in pigs with chronic infarction showed a comparable transmural drop in 99mTc-MIBI activity to levels lower than 10% of that in normal myocardium (Fig 8⇓ and the Table⇓). The zone in which 99mTc-MIBI activity changed from normal to severely low scintigraphic activity had an average length of 10 mm (Figs 2⇑ and 8⇓), and no regional differences between the center and the margins of the ischemic area were found (Fig 8⇓). Inside the ischemic myocardium, there was no correlation between the magnitude of ST segment elevation and myocardial 99mTc-MIBI activity under the corresponding recording site. Samples from the infarct scar also showed a severe drop in myocardial 99mTc-MIBI activity to levels below 10% of that in normal myocardium (Fig 8⇓) with no major transmural gradients (the Table⇓).
ST Segment Changes Over the Infarct Scar
The major finding of this study was that acute ischemia adjacent to a healed infarction virtually free of viable tissue induces ST segment elevation in electrodes at the surface of the necrotic area. Because ST segment changes recorded on the necrotic region were in the same direction as those originated by ischemia at the borders of the infarction, at least two causal mechanisms should be considered.
First, ST segment elevation over the infarct scar might be originated by ischemia of cells surviving within the infarction. Although we have observed a 0.22-mm-wide layer of cells surviving in the endocardium, the present model is characterized by its paucity in the amount of tissue subsisting 1 month after permanent coronary occlusion. It is unlikely that these cells became significantly ischemic to induce ST segment changes in the overlying epicardium because they maintained a positive glycogen reaction 60 minutes after coronary reocclusion. When ischemia is severe enough to induce the rapidly evolving ST segment changes and monophasic potentials seen in the ischemic peri-infarction area, cells underlying these recording sites show glycogen depletion 60 minutes after coronary occlusion. In a previous correlative metabolic and ECG analysis in swine, we confirmed that glycogen depletion 60 minutes after coronary occlusion and local ST segment changes, as seen in the present study, indicate severe ischemia because local tissue samples promptly became depleted of CP and ATP.13 Thus, because the time course of the ST segment elevation over the scar paralleled that in the peri-infarction ischemic zone, the subendocardial cells also should have shown glycogen depletion. Studies indicate that subendocardial cells survive the infarction because they can obtain oxygen and metabolic substrate from the blood cavity. Transmembrane action potentials (TAPs) recorded from endocardial cells in acute ischemic14 and in 2- to 4-month-old infarcted regions15 of isolated, coronary-perfused cat15 or rabbit14 left ventricular preparations show that on interruption of coronary perfusion (simulating ischemia) but with tissue bath superfusion maintained with normal oxygenated Tyrode’s solution, TAP variables (resting potential, TAP duration, and TAP amplitude) are relatively unaffected. In the acute ischemic areas, this protection extended across a subendocardial zone of 40 to 60 cell layers.14 At the same time, both studies reported that these cells depict ischemic alterations in intracellular potentials when the superfusate was changed from oxygenated solution to hypoxic solution with maintenance of coronary perfusion, thereby indicating that survival of subendocardial cells does not depend primarily on coronary perfusion. Moreover, occlusion of a coronary artery that provides collateral flow to a 2-week-old infarction in dogs did not reduce the subendocardial blood flow further.16 In our model, the paucity of collateral perfusion to the peri-infarction zone, as indicated by the 99mTc-MIBI data, further suggested that coronary perfusion is not a prerequisite for subendocardial cell survival.
In the absence of ischemia of viable tissue, a second mechanism for the elevation of the ST segment at the surface of the scar should be passive propagation from the ischemic peri-infarction area. Direct evidence of such a mechanism has not yet been reported, but indirect observations support this contention. Specifically, the upright ST segment deviation and its attenuation along the length of the scar seen in this study are in accordance with observations on electrotonic transmission made during propagation of TPA through a zone of block (gap) created by isotonic sucrose perfusion in isolated canine false tendon fibers17 and during discontinuous transmission in chronically infarcted canine myocardium.18 Driven action potentials at the proximal segment of the gap were conducted through the gap by electrotonic displacement of the membrane potential, giving rise to upright “foot potentials” of a progressively decreasing amplitude along the length of the unexcitable segments. Large R waves (first upright component of monophasic potentials) were not recorded in electrodes overlying the necrotic region. The lack of transmission of these relatively fast upright deflections does not rule out the concept of electrotonic transmission because rectangular current pulses transmitted through a zone of unexcitable gap elicit upright foot potentials without propagation of the first rapid component of the pulse.17
Electrotonic transmission through a depressed tissue will depend on the length and resistance of the unexcitable segment and on the amplitude of the signal entering the gap, among others. Although acute ischemia largely increases extracellular and intracellular resistance,19 epicardial mapping of specific electric impedance of normal and infarct tissue in sheep hearts with healing aneurysms20 showed that in contrast with acute ischemia, 1- and 6-week-old infarct aneurysms disclosed a specific impedance 40% lower than that of the noninfarcted region. If the latter measurements were applicable to our model, one would expect that the conditions for electrotonic transmission through a healed infarction may indeed be present. However, further studies are required to directly demonstrate electrotonic transmission of the ST segment potential through an infarct scar.
ST Segment Changes at the Peri-Infarction Area
This study has shown that acute ischemia surrounding a healed infarction induces lower ST segment elevation than acute ischemia at a comparable cardiac region not adjacent to an infarction.
The mechanism causing a lower-than-expected ST segment elevation in pigs with reinfarction is not known. We could speculate that pigs with a proximal reocclusion of the LAD develop a smaller acute ischemic area and hence less ST segment elevation than pigs with a first proximal occlusion of the LAD. Our data indicate that the size of the ischemic area is not a major determinant of the magnitude of ST segment elevation, as demonstrated by the lack of significant ST segment differences between pigs with a LAD occlusion 2 cm above (group 2) or just below (group 3) the first diagonal branch. Coronary occlusion at either of these two sites is known to originate 40% differences in the extension of the ischemic area in pigs.7
Although coronary collateral circulation is sparse in swine,10 21 animals with chronic infarction might have developed coronary collaterals at the peri-infarction zone large enough to attenuate the severity of underperfusion and perhaps the magnitude of local ST segment changes induced by coronary reocclusion. Data from our study rule out this possibility because ischemic myocardium adjacent to and not adjacent to an infarction disclosed a severe and comparable drop in myocardial 99mTc-MIBI activity. During coronary occlusion, myocardial perfusion assessed by recording 99mTc-MIBI activity strongly correlates with myocardial blood flow measured by radioactive microspheres.11 The area of underperfusion in pigs with and without previous infarction was homogeneous and sharply demarcated, indicating that pigs have almost no collateral circulation, which agrees with previous observations.10 21
Attenuation of ST segment elevation in acute ischemic areas surrounding a healed infarction may be due to the fact that myocardial infarction causes sympathetic and parasympathetic denervation in these regions.7 The observation that pigs with myocardial sympathetic and parasympathetic denervation induced by pericoronary application of phenol show lower ST segment elevation during acute LAD ligature than pigs nondenervated with phenol supports this neurogenic mechanism.22
Electron microscopy studies have shown abnormalities in the number and distribution of gap junctions at the border zones adjacent to infarct scars in humans and dogs8 and in hypertrophied myocardial cells.23 Although the potential influence of these alterations on the genesis of ST segment changes has not been delineated, abnormalities in these cellular structures may alter the passive myocardial conduction properties and, in turn, propagating-action potentials.24 In this regard, it is known that morphological alterations in action potential cause the ST elevation seen during acute ischemia.9 13 25
At least two findings support the hypothesis that ischemic electrophysiological alterations in swine may be extrapolated to humans: the resemblance of the evolving ECG changes induced by acute ischemia in humans with those induced in pigs25 and the fact that both species develop transmural infarction with almost no viable tissue after permanent interruption of coronary blood flow. Our data suggest that patients with a healed transmural infarction with no significant viable tissue might develop ST segment elevation in infarct-related ECG leads as a consequence of ischemia at the peri-infarction zone and passive transmission of the ST segment potential changes along the necrotic area.
The observation that ischemia adjacent to a healed infarction induces lower ST segment elevation than ischemia at a comparable cardiac region in noninfarcted pigs may also have its clinical counterpart. Theoretically, patients with acute ischemia at the margins of a chronic infarction (reinfarction) could disclose lower ST segment elevation than patients with a first acute ischemic episode in a comparable same cardiac region.
This study was supported by grants from Fondo de Investigaciones Sanitarias de la Seguridad Social (92/0734 and 93/0450), Madrid, and from Fundació Uriach 1838, Barcelona, Spain. We thank Prof Michiel J. Janse (Amsterdam, The Netherlands) for critical review of the manuscript. We appreciate the technical assistance provided by Jaume Candell-Riera, MD, Joan Castell, MD, and Amparo García, MD, during the scintigraphic studies.
- Received July 8, 1994.
- Revision received September 14, 1994.
- Accepted September 28, 1994.
- Copyright © 1995 by American Heart Association
Lahiri A, Balasubramanian V, Millar Craig MW, Crawley J, Raftery EB. Exercise-induced ST segment elevation: electrocardiographic, angiographic, and scintigraphic evaluation. Br Heart J. 1980;43:582-588.
Dunn RF, Bailey IK, Uren R, Kelly DT. Exercise-induced ST-segment elevation: correlation of thallium-201 myocardial perfusion scanning and coronary arteriography. Circulation. 1980; 61:989-995.
Chahine RA, Lowery MH, Bauerlein EJ. Interpretation of the exercise-induced ST-segment elevation. Am J Cardiol. 1993; 72:100-102.
Margonato A, Ballarotto C, Bonetti F, Cappelletti A, Sciammarella M, Cianflone D, Chierchia SL. Assessment of residual tissue viability by exercise testing in recent myocardial infarction: comparison of the electrocardiogram and myocardial perfusion scintigraphy. J Am Coll Cardiol. 1992;19:948-952.
Barber MJ, Mueller TM, Davies BG, Gill RM, Zipes DP. Interruption of sympathetic and vagal-mediated afferent responses by transmural myocardial infarction. Circulation. 1985;72:623-631.
Cinca J, Bardají A, Carreño A, Mont LL, Worner F, Bosch R, Soldevilla A, Soler-Soler J. Low incidence of ventricular arrhythmias induced by ischaemia at the borders of a chronic infarction in a model with local postinfarction denervation. Cardiovasc Res. 1994;28:1018-1024.
Kléber AG, Janse MJ, van Capelle FJL, Durrer D. Mechanism and time course of ST and TQ segment changes during acute regional ischemia in the pig heart determined by extracellular and intracellular recordings. Circ Res. 1978;42:603-613.
Patterson RE, Kirk ES. Analysis of coronary collateral structure, function, and ischemic border zones in pigs. Am J Physiol. 1983;244:H23-H31.
Sinusas AJ, Trautman KA, Bergin JD, Watson DD, Ruiz M, Smith WH, Beller GA. Quantification of area at risk during coronary occlusion and degree of myocardial salvage after reperfusion with technetium-99m methoxyisobutyl isonitrile. Circulation. 1990; 82:1424-1437.
Janse MJ, Cinca J, Moréna H, Fiolet JWT, Kléber AG, de Vries GP, Becker AE, Durrer D. The ‘border zone’ in myocardial ischemia: an electrophysiological, metabolic, and histochemical correlation in the pig heart. Circ Res. 1979;44:576-588.
Wilensky RL, Tranum-Jensen J, Coronel R, Wilde AAM, Fiolet JWT, Janse MJ. The subendocardial border zone during acute ischemia of the rabbit heart: an electrophysiologic, metabolic, and morphologic correlative study. Circulation. 1986;75:1137-1146.
Kimura S, Bassett AL, Cameron JS, Huikuri H, Kozlovskis PL, Myerburg RJ. Cellular electrophysiological changes during ischemia in isolated, coronary-perfused cat ventricle with healed myocardial infarction. Circulation. 1988;78:401-406.
Antzelevitch C, Moe GK. Electrotonically mediated delayed conduction and reentry in relation to ‘slow responses’ in mammalian ventricular conducting tissue. Circ Res. 1981;49:1129-1139.
Spear JF, Michelson EL, Moore EN. Reduced space constant in slowly conducting regions of chronically infarcted canine myocardium. Circ Res. 1983;53:176-185.
Kléber AG, Riegger CB, Janse MJ. Electrical uncoupling and increase of extracellular resistance after induction of ischemia in isolated, arterially perfused rabbit papillary muscle. Circ Res. 1987;61:271-279.
Fallert MA, Mirotznik MS, Downing SW, Savage EB, Foster KR, Josephson ME, Bogen DK. Myocardial electrical impedance mapping of ischemic sheep hearts and healing aneurysms. Circulation. 1993;87:199-207.
Schaper W. The Collateral Circulation of the Heart. Amsterdam: North Holland; 1971:5-18.
Peters NS, Green CR, Poole-Wilson PA, Severs NJ. Reduced content of Connexin43 gap junctions in ventricular myocardium from hypertrophied and ischemic human heart. Circulation. 1993;88:864-875.
Jalife J, Sicouri S, Delmar M, Michaels DC. Electrical uncoupling and impulse propagation in isolated sheep Purkinje fibers. Am J Physiol. 1989;257:H179-H189.