INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

CONGESTIVE HEART FAILURE is the leading cause of cardiovascular morbidity in the United States, and once it develops, the 6-yr mortality rate is greater than 60% (13, 17, 23). Understanding the pathogenesis of congestive heart failure is crucial to develop innovative therapies for this disease. Because it is difficult to study the development of heart failure in patients, many experimental models of heart failure have been developed. Currently, the most extensively used experimental models of heart failure are the myocardial infarction (5, 21, 25, 26, 33) and the rapid ventricular pacing (1, 2, 7, 36) models. More recently, transgenic murine models of heart failure also have been described (3, 9, 11, 14).

Myocardial infarction usually can be produced by either extravascular, i.e., occluder or ligature (5, 25, 33), or intravascular, i.e., microembolism (21, 26), coronary artery occlusion. However, it is difficult to achieve stable congestive heart failure by causing an abrupt extravascular coronary artery occlusion, particularly in larger species, because either acute cardiogenic shock or nonischemic compensatory changes occur. Although intravascular microembolization can induce progressive myocardial injury that leads to heart failure, the procedure requires complicated multiple intracoronary injections, and the exact level or site of coronary artery occlusion is difficult to control, often resulting in considerable variability among animals. The major problem with the rapid pacing-induced heart failure model is that the biochemical and hemodynamic alterations revert nearly to normal values soon after pacing is ceased (15, 18), suggesting that the mechanisms of this model are similar to those of reversible dilated cardiomyopathy (4, 16, 22) rather than to those of the irreversible failing heart in humans. Finally, although transgenic murine models may mimic the changes that occur at the cellular and molecular level during heart failure, precise and direct hemodynamic measurements in the conscious state are limited.

The goal of the present investigation was to develop a novel animal model that more closely reflects the pathophysiological process of heart failure in humans. Because the anatomy of the coronary circulation and the myocardial metabolic characteristics of pigs are similar to those of humans (10, 35), and because congestive heart failure in humans is generally caused by myocardial ischemia that results from coronary artery disease, we used coronary artery occlusions to initiate the development of heart failure in the pig. To produce a modestly sized myocardial infarction without acute mortality, a distal site of the left circumflex coronary artery was occluded. Approximately 48 h later, a second proximal coronary artery occlusion was performed at the origin of the left circumflex coronary artery to enlarge the infarct. Even in the presence of the larger sequential infarcts, however, resting global left ventricular (LV) and systemic hemodynamic function were maintained, possibly via compensatory responses in the nonischemic myocardium. It is known that exercise can unmask underlying abnormalities of LV function caused by myocardial ischemia (8, 28, 32), and intermittent tachycardiac stress advances the development of heart failure in patients with ischemic heart disease, which is ameliorated by a -adrenergic receptor blockade (6, 12). Therefore, we used repeated rapid ventricular pacing after the two sequential coronary artery occlusions to promote the development of congestive heart failure. To determine whether heart failure induced by the combined interventions is different from that induced by the rapid pacing alone, the same protocol but without antecedent myocardial infarctions was investigated in a separate group of pigs. All of the experiments were performed after the pigs had recovered from surgery and while they were conscious to avoid potential influences from surgical injury and anesthesia. Cardiac and systemic hemodynamics were measured directly from chronically implanted instrumentation.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Implantation of instrumentation. Twenty-four farm pigs (Yorkshire-Cross) of either sex and weighing 32.1 ± 0.8 kg were trained to enter and rest comfortably in a Panepinto sling (Charles River Laboratories, Wilmington, MA) daily for ~1-2 h. After at least 1 wk of training, the pigs were scheduled for surgical instrumentation (Fig. 1A). Before surgery, each pig was sedated with ketamine hydrochloride (10-12 mg/kg im). After the pigs were endotracheal intubated and ventilated with a respirator (North American Drager, Telford, PA), we maintained general anesthesia with isoflurane (1.5-2.0 vol% in O₂). Using sterile surgical technique, we performed a left thoracotomy at the fifth intercostal space. Catheters made of Tygon tubing (Norton Performance Plastics, Akron, OH) were implanted in the descending aorta and left atria to measure their respective pressures. A solid-state miniature pressure gauge (Konigsberg Instruments, Pasadena, CA) was implanted in the left ventricular (LV) chamber to obtain LV pressure and the rate of change of LV pressure (LV dP/dt). A pacing lead (model 5069, Medtronic, Minneapolis, MN) was attached to the right ventricular free wall, and stainless steel pacing leads were attached to the left atrial appendage. One pair of piezoelectric ultrasonic dimension crystals were implanted on opposing anterior and posterior endocardial regions of the LV to measure the short-axis internal diameter. Proper alignment of the endocardial crystals was achieved during surgical implantation by positioning the crystals to obtain a signal with the greatest amplitude and shortest transit time. The left circumflex coronary artery was isolated, and two hydraulic occluders made of Tygon tubing were implanted proximally and distally to the first obtuse marginal branch. In 23 pigs, a flow probe (Transonic Systems, Ithaca, NY) was placed around the ascending aorta to measure cardiac output. The wires and catheters were externalized between the scapulae, the incision was closed in layers, and air was evacuated from the chest cavity. The animals used in this study were maintained in accordance with the Guide for the Care and Use of Laboratory Animals by the National Research Council (1996), and the studies were approved by the Merck Research Laboratories (West Point, PA) Institutional Animal Care and Use Committee. All pigs were housed individually under conventional conditions, fed commercial pig ration (PMI Feeds, Richmond, IN), and allowed access to water ad libitum.

View larger version (45K): [in this window] [in a new window]   Fig. 1.   A: schematic illustrations of instrumentation used for measurement of cardiac and systemic hemodynamics and of left ventricle (LV) depicting progressive myocardial infarction within the area at risk after two sequential left circumflex coronary artery occlusions (CAO). RV, right ventricle; LA left atrium. B: experimental protocol for heart failure model in conscious pigs.

Experimental measurements. Hemodynamic recordings were made using a data tape recorder (model RD-130TE, TEAC, Montebello, CA) and a multiple-channel oscillograph (model MT95K2, Astro-Med, West Warwick, RI). Aortic and left atrial pressures were measured using strain-gauge manometers (Statham Instruments, Oxnard, CA), which were calibrated in vitro using a mercury manometer, connected to the fluid-filled catheters. The solid-state LV pressure gauge was cross-calibrated with aortic and left atrial pressure measurements. LV dP/dt was obtained by electronically differentiating the LV pressure signal. A triangular wave signal was substituted for the pressure signals to directly calibrate the differentiator (Triton Technology, San Diego, CA). Ascending aortic blood flow was measured using a volume flowmeter (Transonic Systems). Mean arterial pressure, mean left atrial pressure, and mean aortic blood flow (cardiac output) were measured using an amplifier filter (Gould Universal Amplifier, Cleveland, OH). Stroke volume was calculated as the quotient of cardiac output and heart rate. A cardiotachometer (Gould) triggered by the LV pressure signal provided instantaneous and continuous records of heart rate. LV dimension was measured with an ultrasonic transit-time dimension gauge (Triton Technology, San Diego, CA). Total peripheral resistance was calculated as the quotient of mean arterial pressure and cardiac output. LV short-axis end-diastolic dimension (EDD) was defined at the beginning of the upstroke of the LV dP/dt signal. LV end-systolic dimension (ESD) was defined at the time of minimum LV dP/dt. The percent shortening of LV internal diameter, i.e., fractional shortening, was calculated as (EDD  ESD)/EDD × 100. LV velocity of circumferential fiber shortening (V_cf) was calculated from the dimension measurements using the formula: (EDD  ESD) · EDD¹ · ejection time¹ (s). Ejection time was measured as the interval between maximum and minimum LV dP/dt.

Experimental protocol. The protocol is summarized in Fig. 1B. After the pigs had recovered fully from the surgery, i.e., 12-15 days after surgery, all of the experiments were conducted while the pigs were conscious and quietly restrained in a sling. After we made baseline hemodynamic recordings, the pigs were sedated with ketamine hydrochloride (5 mg/kg im). Then the left circumflex coronary artery was occluded distally to the origin of its first margin branch by inflating the implanted hydraulic occluder. During coronary artery occlusion, multiple ventricular premature beats were treated with bolus injections of lidocaine delivered through the left atrium. In 15 pigs, this same procedure was used to occlude the proximal circumflex coronary artery ~48 h after the first occlusion. Right ventricular pacing at a rate of 220 beats/min was initiated using a programmable external cardiac pacemaker (model EV4,543, Pace Medical, Waltham, MA) beginning 1-5 days after the second coronary artery occlusion. The pacing was continued for 1 wk and then terminated for 3 days. This procedure was repeated for another two cycles for a total of three 1-wk pacing periods, each separated by 3 days of rest. In another three pigs, the left circumflex coronary artery was occluded only distally, and then the same pacing protocol was initiated 2 days after the coronary artery occlusion. In an additional six pigs, only the pacing protocol was applied. Hemodynamic recordings were made before and after each 1-wk pacing period (i.e., ~30 min after pacing was stopped) and 7, 14, and 21 days after the final pacing period while the pigs were conscious.

The pacing protocol used in the present study was the result of many feasibility studies. Intermittent, rather than continuous rapid pacing, was used because continuous pacing for less than 2 wk either induced severe LV dysfunction that led to sudden death, or in the surviving animals, the impaired hemodynamics had a clear tendency to recover after cessation of continuous pacing.

At the end of the experiments, the pigs were euthanized with a lethal dose of pentobarbital sodium. Each heart was excised and the ascending aorta was cannulated and retrogradely perfused with 0.9% saline at a driving pressure of 120-160 mmHg. After perfusion, the heart was fixed in 5% Formalin for 2-4 days and then sectioned at the atrioventricular junction. The LV was sliced into six to nine pieces and weighed. Both sides of the individual rings were digitally photographed, and the infarct size was determined by measuring the perimeters of the LV and infarcted region and expressed as a percentage of the LV perimeter (29, 31). Of the 15 pigs with two coronary artery occlusions followed by the repeated pacing, three died at the end of the third week of pacing and one died 2 wk after cessation of the pacing protocol because of severe heart failure. In addition, because of problems with the chronically implanted instrumentation, hemodynamic measurements were not obtained during the latter part of the protocol in these animals, which were euthanized before the end of the protocol. One of the six pigs from the pacing-only group was euthanized at the end of the third week of pacing because of instrumentation failure.

Data analysis. Data before, i.e., baseline, and during the development of heart failure were compared by using the Student's t-test for paired data with a Bonferroni correction. Data from the two protocol groups were compared by using Student's grouped t-test. All values are expressed as means ± SE. Statistical significance was accepted at the P < 0.05 level.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Sequential coronary artery occlusions followed by intermittent rapid pacing. The baseline hemodynamic parameters and LV function in conscious pigs subjected to the sequential coronary artery occlusions followed by the repeated rapid ventricular pacing protocol are shown in Tables 1 and 2, respectively. Tables 1 and 2 also show the changes in these parameters at the end of each of three 1-wk pacing periods and 3, 7, 14, and 21 days after the final pacing, i.e., third week of pacing. In addition to these hemodynamic and LV function changes, characteristic of dilated cardiomyopathy, anorexia, peripheral and pulmonary edema, and reduced physical activity, which are consistent with an advanced stage of congestive heart failure, also were observed in these animals.

Representative waveforms of LV pressure, LV dP/dt, arterial pressure, left atrial pressure, LV internal diameter, ascending aortic blood flow, and heart rate in a conscious pig before and during the development of heart failure are shown in Fig. 2. Clearly, most hemodynamic parameters were affected only slightly after two sequential left circumflex coronary artery occlusions. However, at the end of 1-3 wk of pacing, LV dP/dt and ascending aortic blood flow were decreased markedly, whereas mean left atrial pressure, LV internal diameter, and heart rate were increased significantly. Importantly, these hemodynamic abnormalities persisted after cessation of pacing.

View larger version (121K): [in this window] [in a new window]   Fig. 2.   Representative waveforms of LV pressure, LV rate in pressure change (dP/dt), aortic pressure, left atrial pressure, LV internal diameter, ascending aortic flow, mean ascending aortic flow, and heart rate in a conscious pig before and after 2 sequential CAOs; after 1 and 3 wk of pacing; and 3, 7, 14, and 21 days after cessation of pacing. Note that CAOs did not substantially affect any hemodynamic parameters. However, pacing significantly increased left atrial pressure and LV internal diameter and decreased LV dP/dt and aortic flow. Importantly, even 3 wk after cessation of pacing, severe abnormal hemodynamic changes had not recovered.

The first distal left circumflex coronary artery occlusion did not significantly affect the measured parameters. However, after the second proximal left circumflex coronary artery occlusion, mean left atrial pressure and LV end-diastolic diameter were increased (P < 0.05) by 5 ± 1 mmHg and 3.3 ± 0.5 mm from their baseline values of 5 ± 1 mmHg and 38.7 ± 1.5 mm, respectively, whereas LV dP/dt, V_cf, and cardiac index were reduced (P < 0.05) by 381 ± 96 mmHg/s, 0.37 ± 0.04 s¹, and 15 ± 5 ml · min¹ · kg¹ from the baseline values of 2,778 ± 112 mmHg/s, 1.06 ± 0.05 s¹, and 122 ± 4 ml · min¹ · kg¹, respectively.

Immediately after the first week of pacing, cardiac index, total peripheral resistance, and LV dP/dt were markedly altered compared with after the two sequential coronary artery occlusions but before pacing. Also, LV end-diastolic diameter was increased, and LV systolic pressure and V_cf were decreased to a greater extent at the end of the first week of pacing compared with those after the coronary artery occlusions alone. However, 3 days after the first week of pacing, hemodynamic function, particularly LV dP/dt, cardiac index, and total peripheral resistance, had recovered substantially (Figs. 3 and 4). At the end of the second and third week of pacing, the decrease in cardiac index and the increases in mean left atrial pressure, LV end-diastolic and end-systolic diameters, and heart rate were greater than those at the end of the first week of pacing (Tables 1 and 2), whereas the decreases in LV dP/dt, LV fractional shortening, and V_cf were similar to those at the end of the first week of pacing. Importantly, 3 days after cessation of the second and third weeks of pacing, hemodynamic function had recovered to a lesser extent compared with 3 days after the first week of pacing (Figs. 3 and 4).

View larger version (24K): [in this window] [in a new window]   Fig. 3.   Time course of changes in mean arterial pressure (A), cardiac index (B), mean left atrial pressure (C), and total peripheral resistance (D) before (C) and after three 1-wk periods of pacing (P) separated by 3 days of rest and during a 21-day postpacing recovery period in conscious pigs with and without myocardial ischemia. Pacing significantly increased mean left atrial pressure and total peripheral resistance, whereas significantly decreasing cardiac index in both groups. These changes, however, diminished over 21-day recovery period in pigs subjected to pacing only but persisted in pigs subjected to myocardial ischemia before being paced.

View larger version (24K): [in this window] [in a new window]   Fig. 4.   Time course of changes in LV systolic pressure (A), LV fractional shortening (B), LV dP/dt (C), and LV end-diastolic diameter (D) before (C) and after three 1-wk periods of pacing (P) separated by 3 days of rest and during a 21-day postpacing recovery period in conscious pigs with and without myocardial ischemia. Pacing significantly increased LV end-diastolic diameter and significantly decreased LV fractional shortening and LV dP/dt in both groups. However, during the 21-day recovery period, LV fractional shortening and LV dP/dt returned toward control, i.e., before heart failure in the group without myocardial ischemia. In contrast, in the group with myocardial ischemia and pacing, LV fractional shortening, and LV dP/dt remained depressed.

During the 7-21 days after the final cycle of pacing, i.e., third week of pacing, there was a tendency for mean left atrial pressure, cardiac index, and total peripheral resistance to recover. However, the values of these parameters still were significantly (P < 0.05) different from their baseline values. For example, 14 days after cessation of pacing, mean left atrial pressure and total peripheral resistance were increased by 14 ± 4 mmHg and 0.31 ± 0.05 mmHg · ml¹ · min · kg, respectively, whereas cardiac index was reduced by 37 ± 6 ml/min from the baseline level (Tables 1 and 2). On the other hand, the reductions in LV dP/dt, LV fractional shortening, V_cf, and LV end-diastolic and end-systolic diameters were maintained after pacing was terminated (Tables 1 and 2). Mean arterial pressure was affected slightly throughout the entire periods of hemodynamic measurements. Heart rate was increased significantly at the end of each cycle of pacing and then gradually returned toward the baseline during the 3-day recovery periods. To determine whether the severity of heart failure was related to the time between the second coronary artery occlusion and initiation of the first week of pacing, the hemodynamic data were divided into two groups based on whether pacing was initiated 1 day or 3.8 ± 0.6 days after the second coronary artery occlusion. Figure 5 shows the averaged data of LV dP/dt, mean left atrial pressure, LV end-diastolic diameter, and heart rate 3 days after the second and third cycles of pacing. The increases in mean left atrial pressure, LV end-diastolic diameter, and heart rate in the group in which pacing was initiated 1 day after the second occlusion were significantly greater (P < 0.05) than those observed when the pacing protocol was initiated 3.8 ± 0.6 days after the second occlusion. The decreases in LV dP/dt were similar for these two groups.

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Changes in LV dP/dt (A), mean left atrial pressure (B), LV end-diastolic diameter (C), and heart rate (D) in conscious pigs subjected to pacing that was initiated 1 day or 3.8 ± 0.6 days after 2 sequential CAOs. Data are averages 3 days after the second and final week of pacing.

Body weight, LV weight, and infarct size, expressed as a percentage of the LV perimeter, in pigs subjected either to pacing only or to one or two coronary artery occlusions followed by pacing are shown in Table 3. Both body weight and LV weight were similar among the three groups. In the pigs with two sequential coronary artery occlusions, the infarct size was 33.1 ± 2.2%, which was significantly greater (P < 0.05) than the 17.9 ± 3.2% in the pigs with one coronary artery occlusion. Figure 6 shows representative LV cross sections from one pig subjected to 3 wk of intermittent rapid ventricular pacing follow by 3 wk of recovery and from two other pigs subjected to the same pacing protocol and recovery period but after one or two sequential coronary artery occlusions. The area of thinning of the LV wall in the ischemic region represented ~30% of the total left ventricle in the pig subjected to the two coronary artery occlusions, which was greater than in the pig subjected to only one coronary artery occlusion. In the pig subjected to pacing only, the LV wall was intact.

View larger version (35K): [in this window] [in a new window]   Fig. 6.   Representative LV cross sections from one pig subjected to 3 wk of intermittent rapid ventricular pacing followed by 3 wk of recovery (A) and from two other pigs subjected to same pacing protocol and recovery period but after one (B) or two (C) sequential coronary artery occlusions. Note that area of thinning of LV wall in ischemic region was ~30% of total LV in pig with 2 coronary artery occlusions, which was greater than in the pig with a single coronary artery occlusion.

Rapid pacing compared with coronary artery occlusions followed by rapid pacing. The baseline hemodynamic parameters in conscious pigs subjected to intermittent rapid ventricular pacing with and without previous myocardial ischemia induced by coronary artery occlusions are shown in Tables 1 and 2. There were no baseline differences between the two groups in any of these parameters. The changes in mean arterial pressure, mean left atrial pressure, cardiac index, total peripheral resistance, LV systolic pressure, LV dP/dt, LV fractional shortening, and LV end-diastolic diameter for the two groups are shown in Figs. 3 and 4.

In the pigs subjected to pacing only, the decreases in LV dP/dt, cardiac index, and total peripheral resistance after each of the three 1-wk pacing periods were almost identical to those observed in the pigs with myocardial ischemia followed by pacing (Figs. 3 and 4). However, the increases in mean left atrial pressure, LV end-diastolic diameter, and heart rate and the decreases in V_cf and LV fractional shortening were greater in the pigs with myocardial ischemia followed by pacing (Figs. 3 and 4). After the 3-day rest periods between the pacing periods, LV dP/dt, mean left atrial pressure, cardiac index, and total peripheral resistance were closer to the baseline levels in the pigs subjected to pacing only compared with those in the pigs with myocardial ischemia followed by pacing (Figs. 3 and 4). For example, 3 days after the second pacing period, the changes in mean left atrial pressure (+5 ± 3 mmHg), LV dP/dt (9 ± 7%), LV fractional shortening (19 ± 10%), V_cf (17 ± 7%), and cardiac index (16 ± 7%) were no longer significantly different from baseline for the group subjected to pacing only, although LV end-diastolic (+27 ± 3%) and end-systolic diameter (+33 ± 4%) were still significantly increased (P < 0.05). In the pigs subjected to pacing in the presence of myocardial ischemia, the changes in mean left atrial pressure (+19 ± 2 mmHg), LV dP/dt (34 ± 4%), LV fractional shortening (53 ± 5%), V_cf (48 ± 6%), and cardiac index (35 ± 4%) were still significant 3 days after the second week of pacing.

In the pigs without previous myocardial ischemia, LV dP/dt and mean left atrial pressure had returned to baseline levels within 7 days after the final week of pacing (Figs. 3 and 4). Also, although LV fractional shortening, V_cf, and cardiac index were still decreased, and total peripheral resistance was still increased, these changes were no longer statistically different from their baseline values (Tables 1 and 2). By the 14th day of recovery, the increased LV end-diastolic and end-systolic diameters were no longer significantly different from the baseline levels (Tables 1 and 2). In contrast, each of these parameters remained significantly altered during the 3-wk recovery period after termination of pacing in the pigs with myocardial ischemia before being paced (Tables 1 and 2).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The most common cause of congestive heart failure in patients is thought to be ischemic cardiomyopathy resulting from insufficient myocardial perfusion associated with complex neurohormonal activation, which contributes to abnormal cardiac performance, systemic vasoconstriction, and sodium retention. In the present study, we demonstrated for the first time that serial myocardial infarctions followed by intermittent rapid ventricular pacing results in significant decreases in LV dP/dt, LV fractional shortening, V_cf, and cardiac index and results in significant increases in left atrial pressure, total peripheral resistance, and LV end-diastolic and end-systolic diameters in conscious pigs. Most importantly, the severe impairments in cardiac and systemic hemodynamics persisted after cessation of rapid pacing, a profile unlike that observed in pigs subjected to the same pacing protocol but in the absence of previous myocardial infarctions. The severe, stable hemodynamic changes that were observed indicate that the current model is suitable for the study of chronic congestive heart failure.

One of the unique features of the present model is that the myocardial infarction is produced by two sequential left circumflex coronary artery occlusions performed 48 h apart. This procedure minimizes acute mortality while maximizing the infarct size in a stepwise manner. However, coronary artery occlusions alone, even those resulting in a moderate to large infarct, are insufficient to induce heart failure without subsequent metabolic stress. The results show clearly that the distal coronary artery occlusion did not induce any significant changes in LV function or systemic hemodynamics, and when combined with the proximal, occlusion altered global LV function only modestly as exhibited by a 5 ± 1 mmHg increase in mean left atrial pressure and decreases in LV dP/dt and cardiac index of 14 ± 4% and 13 ± 3%, respectively. Furthermore, the recovery data (Fig. 5) indicate that hemodynamic function was impaired to a greater extent when rapid pacing was initiated 1 day rather than approximately 4 days after the second coronary artery occlusion, suggesting a requirement for close temporal coupling between the repetitive ischemic events and the subsequent myocardial stress of rapid pacing. In addition, the data (Figs. 3 and 4) show that left atrial pressure, LV dP/dt, cardiac output, and total peripheral resistance all tended to recover significantly within 3 days after the first cycle of rapid pacing. Thus it is unlikely that the cardiac and systemic hemodynamic dysfunction would have met the level of heart failure over the period of the present protocol without rapid pacing.

Many previous studies have demonstrated that ligation of the left coronary artery in rats induces left ventricular dysfunction leading to heart failure (25). The difference between these studies and the current study could be attributed to the relatively short life span and high basal heart rate of rats compared with those of pigs. In addition, the infarct size is 35-50% of the left ventricle (24, 25, 31) in rats that develop severe LV dysfunction and survive, which cannot be reproduced acutely in larger species without resultant cardiogenic shock. A few studies have reported that chronic heart failure can be induced 3-4 wk after single-stage left circumflex coronary artery occlusion in pigs (33, 34, 37). In these studies, however, LV dP/dt remained in the normal range after coronary artery occlusion, indicating that myocardial contractility was not depressed (33, 34). In another study (37), only 6 of 18 pigs with coronary artery occlusion were reported to have developed heart failure, and in these pigs LV end-diastolic pressure was increased by only 6 mmHg compared with normal pigs. Thus, based on the hemodynamic data, the pigs from these prior studies cannot be considered to have met the more rigorous criteria for heart failure. Although myocardial infarction without rapid pacing is unable to induce heart failure in pigs in a relatively short period of time, the results of the present study do not eliminate the possibility that heart failure would eventually develop during a longer period of time after myocardial infarction.

It has been demonstrated that the hemodynamic and biochemical changes that occur in pacing-induced heart failure models revert to the baseline levels after cessation of pacing (15, 18). Consistent with these findings, we found that abnormal LV function and systemic hemodynamics returned toward baseline after cessation of pacing in pigs subjected to intermittent rapid ventricular pacing but not in the pigs subjected to the same pacing protocol after serial myocardial ischemia. The different hemodynamic profiles after cessation of pacing suggest that the underlying mechanisms accounting for LV dysfunction following combined myocardial infarction and repeated rapid ventricular pacing differ from those responsible for the changes induced by pacing only. It is conceivable that the nonischemic myocardium can compensate for the loss of regional function after coronary artery occlusions and maintain global LV performance. However, rapid ventricular pacing further increases the energy demand of the nonischemic myocardium possibly beyond the range of compensation, particularly when the infarct is large. Consequently, compensatory mechanisms presumably operating within the border zone may be insufficient, resulting in irreversible myocardial cell damage. Indeed, as our data show, the severe LV dysfunction and systemic vasoconstriction induced by coronary artery occlusions followed by rapid pacing persisted several weeks after cessation of pacing, unlike what was observed with rapid pacing only. It is likely that the superimposition of rapid pacing following the coronary artery occlusions caused the salvageable tissue within the risk area to become further injured.

Several lines of evidence indicate that the amount of tissue that becomes infarcted within the risk area is determined by the balance between the quantity of collateral perfusion and the myocardial energy demand (19, 27). It also is well documented that the infarct size when oxygen consumption is low, i.e., during bradycardia, is significantly smaller than when oxygen consumption is elevated by an increase in heart rate (20). It has also been shown that there is a gradual increase in regional myocardial blood flow to the risk area after coronary artery occlusion, suggesting that even in an intensely ischemic myocardium, a substantial amount of tissue within the area at risk is salvageable (30). In the present study, rapid ventricular pacing was initiated soon after coronary artery occlusion to increase myocardial energy demand in a region where blood flow had been compromised to cause the salvageable tissue within the risk area to become damaged. Indeed, based on the anatomical area perfused by the entire left circumflex coronary artery (35) and the infarct size, expressed as a percentage of the risk area, in the porcine species (29), the infarct size after repeated rapid pacing was greater than expected in the present study.

In summary, two sequential left circumflex coronary artery occlusions in pigs resulted in a stepwise increase in infarct size without acute mortality. However, even after a relatively large myocardial infarction, LV function and systemic hemodynamics were only slighted impaired, suggesting that coronary artery occlusion alone is insufficient to induce chronic heart failure in swine. Intermittent rapid ventricular pacing soon after coronary artery occlusion appears to increase the imbalance between myocardial energy supply and demand both in ischemic border zones and in nonischemic zones. Therefore, unlike the model of heart failure induced by rapid pacing alone, the severe left ventricular dysfunction and peripheral vasoconstriction in the current model did not reverse after cessation of pacing, suggesting that the underlying mechanism is different from that of pacing-induced heart failure.