Evolution of Acute Myocardial Injury

The temporal evolution and extent of irreversible myocardial injury after coronary occlusion is quite variable and strongly dependent on the magnitude of residual coronary flow, the hemodynamic determinants of oxygen consumption, and any endogenous protection afforded by preconditioning.^1,2 Reimer and Jennings demonstrated that in the absence of significant collaterals, irreversible injury begins approximately 20 minutes after coronary occlusion³ and progresses in a wavefront from the subendocardium to the subepicardium (Fig. 36-1).⁴ This susceptibility of the subendocardium to injury reflects its higher oxygen consumption and the redistribution of collateral flow to the outer layers of the heart by the compressive determinants of flow at reduced coronary pressure.⁵ In experimental infarction, the entire subendocardium is irreversibly injured within 1 hour of occlusion, and the transmural progression of infarction is largely complete within 4 to 6 hours. Factors that increase myocardial oxygen consumption (e.g., tachycardia, hypertension) or reduce oxygen delivery (e.g., anemia, hypotension) accelerate the progression to irreversible injury. In contrast, repetitive reversible ischemia prior to occlusion can limit irreversible injury through a preconditioning effect.¹

Figure 36-1 Wavefront of necrosis in infarction. Total coronary occlusions of less than 20 minutes do not cause irreversible injury but can cause myocardial stunning, as well as precondition the heart and protect against recurrent ischemic injury. Irreversible myocyte injury begins after 20 minutes and progresses as a wavefront from subendocardium to subepicardium. After 60 minutes, the inner third of the LV wall is irreversibly injured. After 3 hours, there is a subepicardial rim of tissue remaining, with the transmural extent of infarction completed between 3 to 6 hours after occlusion. The most important factor delaying the progression of irreversible injury is the magnitude of collateral flow, which is primarily directed to the outer layers of the heart.

(Modified from Kloner RA, Jennings RB: Consequences of brief ischemia: Stunning, preconditioning, and their clinical implications: Part 1, Circulation 104:2981–2989, 2001, Reprinted with permission.)

Residual Coronary Flow Limits Infarction

The magnitude of residual coronary flow, either due to subtotal coronary occlusion or the presence of intercoronary collaterals, is the most important determinant of the actual time course of irreversible injury. Ultimate infarct size as a percent of the area at risk of ischemia during a total occlusion is inversely related to collateral flow.⁶ Subendocardial flow of as little as 30% of resting values can prevent infarction during ischemic periods exceeding an hour. More moderate subendocardial ischemia (e.g., flow reduced by 50%) can persist for hours without resulting in significant irreversible injury.⁷ This phenomenon explains how the signs and symptoms of ischemia can be present for prolonged periods without producing significant myocardial necrosis. It also accounts for the observation that late coronary reperfusion can salvage myocardium beyond the 6-hour limit predicted from experimental models.

Mechanisms of Myocyte Death

During prolonged ischemia with myocardial infarction, cell death arises from two distinct mechanisms. Oxidative stress associated with reperfusion immediately causes myocyte necrosis, characterized by sarcolemmal disruption with the leakage of cell contents into the extracellular space. The associated damage is further amplified by leukocyte-mediated injury arising from reperfusion that can variably produce intravascular occlusion and the “no-reflow phenomenon.” After initial survival from ischemia and reperfusion, there is a second wave of myocyte loss due to programmed cell death or apoptosis. This is an energy-dependent and coordinated involution of critically damaged cells that circumvents the inflammation associated with necrotic cell death. The relative importance of each mechanism of cell death in myocardial infarction continues to be controversial.

Stunned Myocardium

There is a temporal delay in the recovery of function after single episodes of ischemia lasting more than 2 minutes. This regional dysfunction occurs despite the restoration of perfusion, with the time course of recovery determined by the duration and severity of the antecedent ischemia. Heyndrickx et al. were the first to demonstrate that regional dysfunction could persist for hours following a 15-minute occlusion, despite complete reperfusion (Fig. 36-3),⁸ a phenomenon that Kloner and Braunwald subsequently called stunned myocardium.⁹ The physiologic hallmark of stunned myocardium is a dissociation of the usual close relation between subendocardial flow and function, where regional dysfunction is present despite normal levels of resting perfusion. Stunned myocardium can also occur after demand-induced ischemia. For example, in the presence of a coronary stenosis, exercise can result in depressed regional function that persists for hours after perfusion is restored, and repetitive ischemia can lead to cumulative stunning.¹⁰ From a clinical perspective, the identification of stunned myocardium is important, since unlike other dysfunctional states, regional function will spontaneously normalize in the absence of recurrent ischemia. Furthermore, contractile reserve is invariably present with inotropic stimulation (e.g., dobutamine infusion), usually normalizing regional function (see Table 36-1). In contrast to animal models, acutely stunned myocardium in patients is not always a “pure entity” and frequently coexists with irreversibly injured and viable, chronically dysfunctional myocardium (i.e., chronically stunned and hibernating myocardium, vide infra).

Figure 36-3 Experimental findings in stunned myocardium. A, Myocardial stunning following a brief total occlusion. During a 15-minute total occlusion (OCCL), wall thickening measured by ultrasonic crystals (WT) is dyskinetic, with systolic thinning. After reperfusion (R), function is initially depressed but normalizes after 24 hours. B, Myocardial stunning following a prolonged partial occlusion. During acute ischemia (red circles), there was perfusion-contraction matching or “short-term hibernation.” With reperfusion (blue squares), wall thickening remains depressed and only returns to normal after several days.

(A, Modified from Heyndrickx et al.⁸⁴ and republished with permission from the American Physiological Society. B, From Matsuzaki M, Gallagher KP, Kemper WS, et al: Sustained regional dysfunction produced by prolonged coronary stenosis: Gradual recovery after reperfusion, Circulation 68:170–182, 1983. Reprinted with permission.)

Short-Term Hibernation During Prolonged Moderate Ischemia

When coronary pressure distal to a stenosis falls below the lower limit of autoregulation, flow reserve is exhausted, resulting in subendocardial ischemia during which reductions in subendocardial flow are closely coupled to reductions in regional contractile function.¹¹ When moderate ischemia persists, the close matching between perfusion and contraction results in a new steady-state with reduced regional oxygen consumption and energy utilization, a phenomenon termed short-term hibernation.^7,11 Despite persistent hypoperfusion, the balance between oxygen supply and demand is reestablished, as reflected by regeneration of creatine phosphate and ATP as well as the resolution of lactate production (Fig. 36-4).¹² Importantly, this ability to restore an energetic balance allows myocyte viability to be maintained for a much longer period of time than when flow is absent. Reperfusion of short-term hibernation also leads to stunning, which may take up to a week to resolve (see Fig. 36-3).¹³

Figure 36-4 Metabolic matching during short-term hibernation. During moderate ischemia, contractile function falls to a new steady state and is accompanied by adaptations in myocardial energy metabolism. These are characterized by an initial fall in creatine phosphate (PCr) and adenosine triphosphate (ATP). The transmural metabolite changes are greatest in the subendocardium (red squares), reflecting the transmural variations in flow. With persistent ischemia, metabolism stabilizes with a regeneration of PCr that prevents further ATP depletion. Although not shown, tissue lactate is transiently increased but normalizes with persistent ischemia.

(Adapted from Pantely GA, Malone SA, Rhen WS, et al: Regeneration of myocardial phosphocreatine in pigs despite continued moderate ischemia, Circ Res 67:1481–1493, 1990.)

While short-term hibernation was originally hypothesized to be the mechanism of chronic hibernating myocardium, it is now clear that short-term hibernation is an extremely tenuous condition, with small increases in myocardial oxygen demand precipitating recurrent ischemia and a rapid deterioration in function and metabolism.⁶ Thus, the reversible nature of short-term hibernation is limited by the severity and duration of ischemia, with irreversible injury frequently occurring after more than 12 to 24 hours.¹⁴

Chronically Stunned Myocardium

Basic studies have clearly shown that repetitive episodes of myocardial ischemia with^17–19 and without a residual stenosis^20,21 can result in chronic regional dysfunction with normal resting perfusion. In experimental animal models with a progressive stenosis, chronically stunned myocardium precedes the development of hibernating myocardium.^18,22 Although resting flow is normal in chronically stunned myocardium, the regional uptake of the glucose analog, fluorine-18 2-fluorodeoxyglucose (FDG), is variably increased and appears to be related to the physiologic severity of a coronary stenosis. For example, pigs chronically instrumented with a fixed-diameter left anterior descending (LAD) coronary artery stenosis have regional dysfunction with normal resting perfusion at both 1 and 2 months after initial instrumentation (Fig. 36-6).¹⁸ Regional FDG uptake is initially normal, but it becomes increased as the reduction in flow reserve progresses, and this persists with the transition to hibernating myocardium.

Figure 36-6 Physiologic progression from chronically stunned to hibernating myocardium. Transmural microsphere flow measurements from pigs instrumented with a progressive LAD stenosis at rest and during adenosine vasodilation are shown, along with transmural variations in FDG uptake (fasting conditions) obtained by ex vivo counting. Angiographic stenosis severity and anterior wall motion score (3 normal, 2 mild hypokinesis, 1 severe hypokinesis) at each time point are summarized below the graphs. As stenosis severity increases over time, there is a reduction in vasodilated flow (adenosine) to the LAD region, reflecting a functionally significant coronary lesion. Initially (1 and 2 months after instrumentation), anterior hypokinesis develops, with normal resting flow consistent with chronically stunned myocardium. After 3 months, the stenosis frequently progresses to total occlusion with collateral-dependent myocardium and a critical impairment in coronary flow reserve. At this time, resting flow to the inner two-thirds of the LAD myocardium becomes reduced, compared to normally perfused remote myocardium. The latter findings are consistent with hibernating myocardium and develop without evidence of infarction in this animal model. The temporal progression of abnormalities demonstrates that chronic stunning precedes the development of hibernating myocardium. In contrast to short-term hibernation resulting from acute ischemia, the reduction in resting flow is a consequence rather than cause of the contractile dysfunction. FDG, fluorine-18 2-fluorodeoxyglucose; LAD, left anterior descending coronary artery.

(Modified from Fallavollita JA, Canty JM Jr: Differential 18F-2-deoxyglucose uptake in viable dysfunctional myocardium with normal resting perfusion: Evidence for chronic stunning in pigs, Circulation 99:2798–2805, 1999.)

From a clinical perspective, it is important to recognize that viable, chronically dysfunctional myocardium with normal resting perfusion is much more common than hibernating myocardium,^23,24 and it can develop independently of ischemia. For example, in patients with ischemic cardiomyopathy, regional dysfunction with normal resting perfusion may occur distant from a large dysfunctional region (viable or infarcted) as a result of pathologic remodeling. Dysfunction in this remodeled myocardium will not improve with regional revascularization, since it is not a direct result of ischemia. Thus, normal resting perfusion does not necessarily translate into reversible dyssynergy, and it is not surprising that noninfarcted but remodeled segments sometimes have persistent dysfunction after coronary revascularization.²⁵

Chronic Hibernating Myocardium

Hibernating myocardium is characterized by contractile dysfunction with reduced resting flow, in the absence of acute ischemia or significant necrosis. It is common in patients with ischemic cardiomyopathy,²⁶ where it frequently coexists with nontransmural infarction and can also occur in the absence of heart failure or global LV dysfunction. The best clinical example of isolated hibernating myocardium is in patients with chronic coronary occlusions and collateral-dependent myocardium (Fig. 36-7).^27,28 A similar constellation of findings has been reproduced in pigs in the absence of infarction or heart failure, which has facilitated the investigation of the underlying adaptations to ischemia that result in hibernating myocardium.²² Studies examining these key physiologic adaptations are summarized later.

Figure 36-7 Clinical imaging of chronic LAD occlusion and collateral-dependent hibernating myocardium. The RAO tracings of the left ventriculogram show anterior akinesis (upper left). Transaxial PET scans illustrate ¹³NH₃ flow measurements at rest (lower left) and following pharmacologic vasodilation with dipyridamole (lower right). Resting flow is reduced compared to normal regions in the same heart. The perfusion deficit during vasodilation is even more severe, and quantitative analysis demonstrates that LAD flow reserve is critically impaired. Viability is identified by increased FDG uptake (following an oral glucose load) in the anterior wall (upper right). FDG, fluorine-18 2-fluorodeoxyglucose; LAD, left anterior descending coronary artery.

(Modified from Vanoverschelde J-LJ, Wijns W, Depre C, et al: Mechanisms of chronic regional postischemic dysfunction in humans: New insights from the study of noninfarcted collateral-dependent myocardium, Circulation 87:1513–1523, 1993.)

The unpredictable progression of clinical coronary disease precludes defining the temporal evolution of hibernating myocardium in patients, and it was originally thought that hibernating myocardium arose from a primary reduction in flow in a fashion similar to experimental models of prolonged moderate ischemia and short-term hibernation.⁷ While this is a plausible mechanism for the development of hibernating myocardium in association with an acute coronary syndrome, more recent studies have highlighted the role of the physiologic severity of a coronary stenosis and repetitive ischemia as the major determinants of the progression from stunned to hibernating myocardium. This paradigm developed from animal studies with a slowly progressive stenosis that clearly demonstrated that regional dysfunction with normal resting flow (chronically stunned myocardium) preceded the development of hibernating myocardium, with the down-regulation in resting flow associated with a critical impairment in subendocardial flow reserve (see Fig. 36-6).^15,18,22,29 This progression is not merely a function of time, since the transition from chronically stunned to hibernating myocardium can be experimentally produced in as little as 1 week after placement of a critical stenosis which exhausts coronary flow reserve.¹⁹ A common conclusion of all serial studies is that reduced resting flow in hibernating myocardium is a result rather than a cause of the contractile dysfunction. As will be discussed, the down-regulation of resting flow in hibernating myocardium appears to be a physiologic marker reflecting numerous intrinsic metabolic adaptations of the heart to repetitive episodes of ischemia.

Metabolic and Energetic Adaptations to Chronic Ischemia

Myocardial Glucose Uptake

Acute myocardial ischemia is obviously associated with acute reductions in oxidative metabolism and an increased dependence on anaerobic glycolysis. This is facilitated by the mobilization of glycogen stores, as well as an increase in myocardial glucose uptake secondary to the translocation of the glucose transporter GLUT4 from intracellular vesicles. This underlies the basis of “hot spot” imaging with the glucose analog FDG for exercise-induced ischemia and acute coronary syndromes³⁰ and is likely responsible for the enhanced FDG uptake in viable dysfunctional myocardium following dobutamine infusion.³¹ Chronically dysfunctional myocardium also demonstrates enhanced basal glucose utilization,³² and this adaptation can occur in chronically stunned myocardium.²¹ Although the mechanism by which glucose utilization is increased in viable dysfunctional myocardium is still under investigation, studies in chronically instrumented pigs have shown an increase in membrane-bound GLUT4, likely due to an up-regulation in p38 mitogen-activated protein kinase.³³ Preliminary studies had suggested a role for up-regulation of glucose transporters in chronic ischemia³⁴; however, maximal FDG uptake during insulin stimulation was not altered in pigs with hibernating myocardium (Fig. 36-8).³⁵ A small decrease in maximal insulin-stimulated FDG uptake has been reported in humans.³⁶

Figure 36-8 Differential FDG uptake in hibernating myocardium in the fasting and insulin-stimulated state. Data summarize fasting and maximal insulin-stimulated glucose uptake estimated in pigs and humans with a chronic coronary occlusion and collateral-dependent myocardium. Whereas regional FDG uptake was increased in the fasting state, quantitative values were well below those attained after insulin infusion. Despite regional increases in the fasting state, maximum insulin-stimulated FDG uptake is not increased in hibernating myocardium. FDG, fluorine-18 2-fluorodeoxyglucose.

Oxygen Consumption

Since myocardial oxygen extraction is near maximal at rest, resting oxygen consumption must also be reduced in hibernating myocardium, and this has been clearly documented using invasive techniques in animals (Fig. 36-9).³⁷ Other investigators have noninvasively documented regionally reduced oxygen consumption in viable, chronically dysfunctional myocardium using PET and carbon-11 acetate.²⁷ Most but not all^32,38 studies have shown an energetic balance despite hypoperfusion and some ability to increase metabolism during submaximal stress. For example, in pigs with hibernating myocardium, there was net lactate uptake and normal venous pH during submaximal increases in heart rate, despite the fact that flow and function were lower than in normal myocardium.³⁷ In humans, coronary venous lactate levels³⁹ and myocardial tissue lactate concentrations in biopsy specimens were normal.⁴⁰ Both basic and clinical studies have also demonstrated preserved creatine phosphate/ATP ratios^41,42 and ATP/adenosine diphosphate (ADP) ratios in hibernating myocardium.^40,42 Absolute ATP levels are mildly reduced in hibernating myocardium,⁴¹ but these alterations are also present in the remote, normally perfused myocardium,^42,43 similar to those occurring in normally perfused periinfarct regions of pigs with myocardial infarction.

Figure 36-9 Attenuated functional and metabolic responses to submaximal increases in myocardial metabolism in swine with hibernating myocardium. Graphs depict steady-state physiologic parameters after β-adrenergic stimulation with submaximal epinephrine infusion (E₁-E₄) to increase heart rate up to approximately 150 beats/min. In hibernating myocardium, LAD subendocardial-segment shortening, flow, and oxygen consumption were depressed at rest compared to normal control animals. At each level of β-adrenergic stimulation, regional function, flow, and oxygen consumption were reduced compared to normal myocardium. Nevertheless, there was no functional or metabolic evidence of ischemia. LAD segment shortening did not deteriorate, and there was no lactate release or reduction in anterior interventricular vein pH. These findings indicate that the adaptations in hibernating myocardium allow it to operate at a lower point on the relation between myocardial oxygen consumption and external workload. This intrinsic adaptation protects the heart against the development of an acute supply/demand imbalance during submaximal increases in workload, at the expense of reduced resting function and oxygen consumption. LAD, left anterior descending coronary artery.

(Reprinted from Fallavollita JA, Malm BJ, Canty JM Jr: Hibernating myocardium retains metabolic and contractile reserve despite regional reductions in flow, function, and oxygen consumption at rest, Circ Res 92:48–55, 2003.)

Pathologic Remodeling in Hibernating Myocardium

Myocardial biopsy specimens obtained from patients at the time of coronary bypass surgery have demonstrated a broad array of cellular changes that could contribute to the phenotype of hibernating myocardium. The heterogeneity of these findings has led to controversy over whether the myocyte pathologic alterations represent an adaptive or maladaptive response to chronic repetitive ischemia. Many studies have supported adaptation, with a reversion to a fetal cellular phenotype and a protein expression pattern that would serve to protect hibernating cardiac myocytes from ischemic stress.^47,48 In contrast, other investigators have found a degenerative phenotype characterized by progressive cell death and marked fibrosis.^49,50 The range of findings reported in human biopsies is illustrated in Figure 36-10. The factors that determine a path of progressive structural degeneration, as opposed to successful adaptation to chronic repetitive ischemia, are currently unknown. Nevertheless, degenerative changes appear to be more common when global LV dysfunction and heart failure are present, and adaptation is more common in stable patients with only regional cardiac dysfunction. This suggests that adaptive changes may be importantly modified by neurohormonal activation. The subsequent discussion of potential mechanisms will primarily focus on observations from animals and humans where hibernating myocardium occurs regionally in the absence of heart failure.

Figure 36-10 Diversity of pathologic findings from transmural biopsies of humans with hibernating myocardium. A, Adapted myocyte phenotype. In patients selected so as to avoid the confounding effects of superimposed infarction, Borgers et al. have reported ultrastructural characteristics that suggest reversion to a fetal phenotype.⁴⁸ The major electron microscopic features (EM; right photomicrographs) are myofibrillar loss and mini-mitochondria. At the light microscopy level, there is increased glycogen content and minimal interstitial fibrosis (Light/PAS; left photomicrographs). Necrosis, apoptosis, and autophagy are absent. B, Degenerative myocyte phenotype. In consecutive unselected patients with viable dysfunctional myocardium and heart failure, the amount of fibrosis associated with reversible dyssynergy is substantially greater. Myocardial biopsies reveal apoptotic (panel A) and autophagic myocyte cell death, as well as interstitial remodeling with increased collagen (panel B) and fibronectin (panels C and D) that averages as much as 30% of the biopsy. The physiologic and clinical variables determining the progression to each of the two phenotypes is currently unclear.

(A, Adapted from Vanoverschelde J-L, Wijns W, Borgers M, et al: Chronic myocardial hibernation in humans. From bedside to bench, Circulation 95:1961–1971, 1997. B, Panel A adapted from Elsasser A, Vogt AM, Nef H, et al: Human hibernating myocardium is jeopardized by apoptotic and autophagic cell death, J Am Coll Cardiol 43:2191–2199, 2004. Panels B-D adapted from Elsasser A, Schlepper M, Klovekorn WP, et al: Hibernating myocardium: An incomplete adaptation to ischemia, Circulation 96:2920–2931, 1997.)

Myofibrillar Loss and Glycogen Accumulation

Biopsies from patients with hibernating myocardium reveal several prototypical features that include myofibrillar loss or myolysis, increased glycogen content, and small mitochondria.⁵¹ While highly affected areas can be identified, less than 40% of the myocytes in a hibernating region exhibit these morphologic changes, so normal myocytes still predominate. In addition, when the myofibrillar volume fraction was quantified, there was less than 10% loss of myofibrils (Fig. 36-11).¹⁹ Many of the cellular changes reported are consistent with a reversion to a fetal myocyte phenotype.⁴⁷ The extent that this reflects dedifferentiated adult myocytes versus regenerating myocytes from resident cardiac stem cells is unknown. Nevertheless, these cellular changes can develop rapidly and be recapitulated in animal models in a period of as little as 2 weeks.^19,52,53 They are also seen in disease states such as atrial fibrillation, dilated cardiomyopathy, and in the border zone between normal and infarcted segments of the heart, reviewed elsewhere.¹⁶

Figure 36-11 Myocyte cellular changes in swine with hibernating myocardium in the absence of heart failure. Swine with hibernating myocardium from a chronic LAD occlusion and collateral-dependent myocardium without infarction or heart failure develop an adapted phenotype. Like the results described by Borgers et al., there is increased reticular connective tissue that is only about 2% greater than values in the remote normally perfused region (left column). Hearts appear grossly normal, but there is cellular hypertrophy in the hibernating region that reflects the consequences of apoptosis-induced myocyte loss (second column). The electron microscopic characteristics of hibernating myocardium are similar to humans with an adapted phenotype and demonstrate myofibrillar loss (Myolysis), numerous small mitochondria, and increased glycogen content (right column). Interestingly, while these are markedly different from normal hearts, biopsies of remote nonischemic segments show similar morphologic changes. Global ultrastructural changes are also seen for other structural proteins.⁵² These data indicate that the “cellular hibernating phenotype” is not directly related to ischemia, nor is it the cause of regional contractile dysfunction. This accounts for its presence in disease states not associated with myocardial ischemia. LAD, left anterior descending coronary artery.

(Adapted from Canty JM Jr, Fallavollita JA: Chronic hibernation and chronic stunning: A continuum, J Nucl Cardiol 7:509–527, 2000.)

Myocardial Fibrosis, Cell Death, and Myocyte Hypertrophy

In patients with single-vessel disease and hibernating myocardium, there is no necrosis, and interstitial fibrosis averages less than 15% of the tissue volume.⁵¹ The limited number of myocytes in a clinical biopsy precludes accurate quantification of cell death, but porcine studies have identified apoptotic nuclei in about 1 to 2 per 10,000 myocyte nuclei. Although hibernating myocardium appears grossly normal, the chronic nature of apoptosis results in an approximately 30% regional loss of myocytes, with compensatory cellular hypertrophy occurring to maintain normal wall thickness.⁵⁴ In contrast to some clinical studies that have suggested that cell death and fibrosis inexorably progress over time,^55,56 pigs with hibernating myocardium demonstrate stable function and connective-tissue staining for at least 2 months.^57,58 Some clinical studies have also demonstrated stability rather than deterioration in function and viability.⁵⁹ The intrinsic molecular adaptations discussed later eventually serve to abrogate myocyte apoptosis (Fig. 36-12). Nevertheless, myocyte loss and/or remodeling, even in the absence of fibrosis, may ultimately limit functional recovery after revascularization.^60,61

Figure 36-12 Apoptosis and regional myocyte loss in hibernating myocardium. Data from hibernating LAD regions are compared to remote regions as well as LAD regions from normal animals. A, The progression from chronically stunned to hibernating myocardium is accompanied by regional apoptosis-induced myocyte loss. Apoptosis is increased approximately sevenfold, yet only affects about 1 in 5000 myocyte nuclei. Nevertheless, since this is a chronic phenomenon, there is an approximately 30% reduction in LAD myocyte nuclear number. This occurs without significant fibrosis, as reflected by the fact that the myocyte area remains nearly normal. B, Temporal progression of myocyte apoptosis in swine with viable chronically dysfunctional myocardium. As resting dysfunction develops in response to repetitive ischemia, myocyte apoptosis increases and peaks at 3 months with the transition to hibernating myocardium. While the physiologic features of hibernating myocardium remain unchanged up to 5 months, the intrinsic adaptations to ischemia limit further apoptosis. LAD, left anterior descending coronary artery.

(Adapted from Lim et al.,⁵⁴ Fallavollita et al.,⁸⁵ and Suzuki et al.⁵⁸)

When biopsies are obtained from patients with ischemic cardiomyopathy and heart failure, reversible dyssynergy can be demonstrated, with fibrosis comprising as much as 30% of the sample.⁴⁹ In this setting, myocyte death is more evident, with autophagy and even areas of patchy necrosis.⁵⁰ The fibrosis may be exacerbated by the elevation of pro-inflammatory cytokines associated with active interstitial remodeling.^62–64 Owing to the heterogeneous nature of the histologic changes in ischemic cardiomyopathy and the frequent association with patchy infarction, it is difficult to determine whether there is myocyte adaptation superimposed on nontransmural infarction or an incomplete adaptation to chronic ischemia in hibernating myocardium. In addition, interstitial fibrosis following repetitive ischemia and stunning may actually be reversible after reperfusion in some experimental models,⁶⁵ similar to the reduction in scar volume after myocardial infarction.⁶⁶

Molecular Remodeling in Hibernating Myocardium

Recent studies have elucidated the molecular mechanisms responsible for adaptation in hibernating myocardium using animal models that recapitulate the major physiologic features found in patients. In general, these adaptations serve to preserve myocyte viability at the expense of a depressed contractile function. Collectively, the cellular changes that occur in response to regional ischemia in the setting of single-vessel disease are very similar to the global molecular changes found in hypertrophied, nonischemic myocytes in patients with heart failure.

Mitochondrial Proteins and Metabolism

A key target of the intrinsic adaptations of the myocyte to repetitive ischemia appears to be mitochondrial function. Proteomic analysis demonstrates numerous down-regulated proteins, including the entry points to oxidative metabolism and members of the mitochondrial electron-transport chain (Fig. 36-13).⁶⁷ Collectively, these changes reduce mitochondrial respiration and regional oxygen consumption at any external workload³⁷ and protect myocytes from oxidative injury following simulated ischemia in vitro. One potential mechanism is through the up-regulation of uncoupling protein 2.⁶⁸ The mitochondrial adaptations eventually prevent progressive apoptosis-induced myocyte loss at the expense of a reduced contractile function and metabolism.

Figure 36-13 Mitochondrial protein remodeling in animals with persistent hibernating myocardium. There were no differences in proteomic profiles between the LAD region of normal sham animals compared to remote myocardium. In contrast, animals with hibernating myocardium developed a down-regulation in mitochondrial proteins, including some of the entry points to oxidative metabolism and selected components of the electron-transport chain. These changes did not reflect mitochondrial loss or increased connective tissue. Like the physiologic changes, most of the mitochondrial proteomic changes persisted in animals with hibernating myocardium that were studied 5 months after instrumentation. Exceptions included long-chain acyl-CoA dehydrogenase and the ATP-synthase β-chain, which tended to normalize. While not shown, increases in stress proteins remained persistently elevated, but increases in cytoskeletal proteins normalized in animals with longer-term hibernation (5 months). LAD, left anterior descending coronary artery.

(Adapted from Page B, Young R, Iyer V, et al: Persistent regional downregulation in mitochondrial enzymes and upregulation of stress proteins in swine with chronic hibernating myocardium, Circ Res 102:103–112, 2008.)

Inhomogeneity in Sympathetic Innervation and β-Adrenergic Signaling

Although contractile reserve to β-adrenergic stimulation is present in hibernating myocardium, compared to normal, the response is attenuated,³⁷ which likely accounts for the underestimation of viability with inotropic stimulation as compared to nuclear imaging.⁷³ Despite a critical limitation in subendocardial flow reserve, increases in regional function can occur in the absence of acute metabolic ischemia during steady-state submaximal stimulation.³⁷ At a cellular and molecular level, there are profound regional alterations in myocardial sympathetic innervation, with attenuation of both pre- and post synaptic adrenergic function. Previous studies in isolated vesicular preparations have demonstrated a reduction in cyclic adenosine monophosphate (cAMP) production, as well as a shift from a high- to low-affinity β-receptor subtype.⁷⁴ In addition, there is partial sympathetic denervation⁷⁵ associated with profound reductions in presynaptic sympathetic norepinephrine uptake, as assessed using iodine-131 metaiodobenzylguanidine (MIBG)⁷⁶ or carbon-11 metahydroxyephedrine (HED) with PET imaging (Fig. 36-14).⁷⁷ The magnitude of these changes is similar to those reported in infarcted myocardium and lead to profound inhomogeneity in myocardial sympathetic innervation. They are functionally significant and attenuate in vivo responses to sympathetic nerve stimulation as well as exogenous β-adrenergic stimulation.⁷⁵ The temporal development of these changes and their potential reversibility are currently under study. As will be discussed, inhomogeneity in sympathetic nerve function in hibernating myocardium may identify a myocardial substrate at risk of lethal ventricular arrhythmias, even in the absence of infarction.⁷⁸

Figure 36-14 PET tomograms of carbon-11 metahydroxyephedrine (HED) uptake in pigs with hibernating myocardium. In normal pigs (left panel), HED uptake was homogeneous. The middle panel shows HED uptake in hibernating myocardium. Even though there was no infarction or pathologic fibrosis, there was a prominent reduction in HED uptake in the anterior wall. The right panel summarizes findings in a pig with a chronic myocardial infarction, demonstrating the well-known reduction in sympathetic nerves typical of irreversible injury. These findings indicate that there are profound abnormalities in myocardial sympathetic innervation in hibernating myocardium that can contribute to inhomogeneity in myocardial electrophysiologic responses during stress.

(From Luisi AJ Jr, Suzuki G, deKemp R, et al: Regional ¹¹C-hydroxyephedrine retention in hibernating myocardium: Chronic inhomogeneity of sympathetic innervation in the absence of infarction, J Nucl Med 46:1368–1374, 2005. Reprinted with permission of the Society of Nuclear Medicine.)

Blunted Contractile Reserve in Viable Myocardium

It is clear that contractile reserve during β-adrenergic stimulation has a high positive predictive value for functional recovery. However, the molecular alterations in hibernating myocardium that are responsible for the adaptation to chronic ischemia attenuate inotropic responsiveness. For example, the blunted contractile response in pigs with hibernating myocardium^37,79,80 reflects a reduction in cAMP production due to a shift in the β-adrenergic receptor to a low-affinity subtype.⁷⁴ In addition, reductions in the SR calcium-handling proteins reduce calcium uptake and release for any level of inotropic stimulation.^19,69 These intrinsic responses in hibernating myocardium may underlie the clinical observation that contractile reserve generally underestimates the frequency of viable dysfunctional myocardium.⁷³

Roles of Myocyte Loss and Cellular Remodeling in Limiting Functional Recovery

When resting perfusion is normal, the myocardium is always viable. Nevertheless, contractile dysfunction does not always improve after revascularization of these segments. There are several potential explanations. First, some of these segments may actually exhibit myocyte loss and molecular remodeling typical of hibernating myocardium, but routine clinical imaging may be too insensitive to detect the modest (~20%) relative reductions in flow identified in experimental studies. In addition, the assessment of resting perfusion may not detect infarction confined to the subendocardium, where small amounts of fibrosis can disproportionately reduce regional contraction.

Even when myocardial scarring is carefully excluded with gadolinium magnetic resonance imaging (MRI),²⁵ up to 25% of noninfarcted segments fail to improve after revascularization. The reasons for persistent dysfunction in the absence of recurrent ischemia or scar are not entirely clear but likely reflect the importance of irreversible myocyte remodeling. In support of this, pigs with hibernating myocardium and apoptosis-induced myocyte loss exhibit persistent dysfunction following revascularization, in the absence of any infarction. Similar myocardial changes can result in irreversible LV dysfunction after surgical intervention for advanced valvular regurgitation. This cellular remodeling is likely a major contributor to the LV dysfunction in advanced ischemic cardiomyopathy, where roughly 80% of the left ventricle is viable and less than 20% irreversibly scarred. Thus, developing approaches to identify myocyte remodeling in the absence of fibrosis may be useful in identifying the likelihood of reversible dyssynergy.

The Impact of Hibernating Myocardium on Prognosis (See Chapters 37 and 38)

While nonrandomized clinical studies have demonstrated the profound negative effect of hibernating myocardium on survival and the amelioration of this risk with revascularization,²⁶ cause-specific mortality remains undefined. Studies have primarily focused on the potential reversibility of myocardial dysfunction and heart failure; however, some data suggest that the major impact of hibernating myocardium may relate to an increased risk of sudden death.⁸¹ Basic studies support this contention⁷⁸ and have demonstrated inhomogeneity in cellular myocyte remodeling and regional inhomogeneity in presynaptic and postsynaptic sympathetic nerve function, which can lead to a substrate that is electrically unstable. Indeed, sudden death from ventricular tachycardia/ventricular fibrillation (VT/VF) develops in pigs with chronic collateral-dependent myocardium in the absence of heart failure (Fig. 36-15). This has given rise to the notion that although adapted to ischemia, the myocyte cellular alterations associated with hibernating myocardium lead to vulnerability to lethal arrhythmias. Since asymptomatic myocardial ischemia is so common, it is plausible that similar neural and myocyte remodeling underlie the progression of coronary artery disease and development of sudden death.⁸² Thus, while the amount of hibernating myocardium can predict the likelihood of improvement from the standpoint of myocardial dysfunction and heart failure, even small amounts may increase the risk of sudden death. This hypothesis is the basis of an ongoing prospective study to determine whether the presence of hibernating myocardium and/or regional inhomogeneity in sympathetic innervation can predict an increased risk of sudden death in patients who are candidates to receive an implantable cardiac defibrillator (Fig. 36-16).⁸³

Figure 36-15 Increased risk of arrhythmic death in pigs with hibernating myocardium. Kaplan-Meier survival analysis demonstrated a progressive reduction in survival due to sudden death. There was no mortality in sham-operated control animals. Using implantable Reveal-Plus loop recorders or telemetry, the mechanism of sudden death was almost always ventricular tachycardia (VT) degenerating into ventricular fibrillation (VF) (inset). Postmortem analysis showed that more than 90% of the animals developing sudden death had no pathologic evidence of acute or healed infarction. Total coronary occlusion and physiologic features consistent with hibernating myocardium could be demonstrated several weeks prior to sudden death in a subset of animals. These data indicate that the myocardial adaptive response to ischemia may be a two-edged sword. While protecting myocytes from acute ischemia, they may lead to a substrate characterized by electrical instability and a high risk of lethal ventricular arrhythmias.

(From Canty JM Jr, Suzuki G, Banas MD, et al: Hibernating myocardium: Chronically adapted to ischemia but vulnerable to sudden death, Circ Res 94:1142–1149, 2004.)

Figure 36-16 Patterns of viability and sympathetic denervation in patients with ischemic heart failure. Like pigs, humans exhibit variable amounts of viable but denervated myocardium. The upper images are from a patient with a large HED defect that is completely viable, demonstrating a mismatch between viability and denervation. The lower images show a similar HED defect in a patient with a myocardial scar. Here the denervation is matched with a reduction in FDG. The hypothesis that PET imaging can be used to predict the risk of sudden cardiac death in patients who are candidates for primary prevention with an ICD is being tested in the PAREPET study (Prediction of Arrhythmic Events with Positron Emission Tomography). FDG, fluorine-18 2-fluorodeoxyglucose; HED, carbon-11 hydroxyephedrine; ICD, implantable cardioverter-defibrillator; PET, positron emission tomography.

(From Fallavollita JA, Luisi AJ Jr, Michalek SM, et al: Prediction of arrhythmic events with positron emission tomography: PAREPET study design and methods, Contemp Clin Trials 27:374–388, 2006.)