The term “reperfusion injury” refers to cellular damage that occurs during the reperfusion phase after an episode of ischemia. If reperfusion occurs after a short period of ischemia, all cells are salvaged ( Figure 24-1A ). However, as the duration of ischemia increases, cells become irreversibly injured, and the territory of cell death increases in size over time ( Figure 24-1B ). Sorting out damage that occurs primarily because of reperfusion rather than during the preceding ischemic phase can be difficult, because it is impossible to have reperfusion without ischemia. For this reason, the resulting myocardial damage is often referred to as an ischemia/reperfusion injury. Evidence that reperfusion itself is harmful to the heart comes from studies that showed that certain phenomena first manifest during the reperfusion phase (no-reflow, ventricular arrhythmias) and then worsen as the phase of reperfusion progresses. In addition, studies of therapies administered only at the time of or solely during the reperfusion phase that result in some benefit also provide evidence that reperfusion itself has detrimental effects. Examples include the administration of oxygen radical scavengers at reperfusion, which results in improved function of stunned myocardium; the initiation of hypothermia at 30 minutes after reperfusion that results in a reduction of the zone of no-reflow ; and administering a pharmacologic agent to the patient undergoing postconditioning at reperfusion, which results in a reduction of myocardial infarct size. In this chapter, we describe the pathobiology and clinical manifestations of reperfusion injury, as well as therapeutic strategies for which studies are completed or ongoing. The pathobiology of infarct healing is described in Chapter 4 . Adverse remodeling of the myocardium after myocardial infarction (MI) is addressed in Chapter 36 .
Pathological and Clinical Components of Reperfusion Injury
There are four basic components of “reperfusion injury”: (1) stunned myocardium; (2) reperfusion arrhythmias; (3) the no-reflow phenomenon, which is also known as microvascular obstruction; and (4) lethal myocardial cell injury caused by reperfusion. In our opinion, there is little doubt that the first three phenomena are caused by reperfusion itself. However, we still question the significance and importance of lethal myocardial cell injury caused by reperfusion and address this uncertainty later in this chapter (see the section on Lethal Myocardial Cell Injury Caused by Reperfusion ).
Stunned myocardium refers to myocardium that has been subjected to a period of ischemia that results in reversible injury (i.e., the cells become ischemic, but are not necrotic) in which there is prolonged but transient contractile dysfunction after reperfusion. In early studies in canine models of proximal coronary artery occlusion followed by reperfusion, 5- to 15-minute episodes of ischemia resulted in regional wall motion abnormalities that persisted for several hours to days, despite the absence of cell death. Hence, in such experiments, the myocardium behaved as if it were “stunned” by a brief episode of ischemia followed by reperfusion. The entire hypoperfused area of myocardium is described as the area at risk ( Figure 24-1A ). However, when occlusion of the coronary artery is prolonged, irreversible cellular injury occurs, which results in infarction ( Figure 24-1B ).
The mechanism of stunning is believed to involve damage from the release of reactive oxygen species and from calcium overload that occurs in the early stages of reperfusion, which results in reduced responsiveness of the contractile apparatus to calcium. Calcium uptake by the sarcoplasmic reticulum appears to be impaired in stunned myocardium and might contribute to contractile dysfunction. The observation that therapy with oxygen radical scavengers at reperfusion improves function supports the concept that stunning is a form of functional reperfusion injury. Clinical examples of stunned myocardium include: (1) slow recovery of salvaged myocardium in the outer wall of the ventricle after thrombolytic therapy or percutaneous coronary intervention (PCI) therapy for ST-elevation MI (STEMI); (2) slow recovery of ventricular function after cardiopulmonary bypass procedures; (3) persistent regional wall motion abnormalities after exercise-induced ischemia or prolonged angioplasty balloon inflation during PCI; and (4) slow recovery of function in those with Takotsubo cardiomyopathy.
Our laboratory and others have observed that reperfusion, after a brief period of ischemia (only 5 minutes in the anesthetized rat model), results in a barrage of ventricular arrhythmias, including polymorphic ventricular tachycardia and ventricular fibrillation. The release of toxic oxygen radicals at the time of reperfusion, as well as electrolyte disturbances (including sodium and calcium overload of cardiomyocytes) may contribute to the onset of these arrhythmias. In addition, during the first seconds of reperfusion, inhomogeneity of action potential amplitude and duration in the previously ischemic zone and border zone might contribute to reentry-related arrhythmias. Certain therapies introduced at the time of reperfusion, such as postconditioning (transient episodes of brief coronary reocclusion and reperfusion) markedly diminish reperfusion-induced arrhythmias. Reperfusion arrhythmias, including accelerated idioventricular rhythms, ventricular tachycardia, and ventricular premature beats, have been observed in patients after reperfusion therapy for STEMI (see Chapter 28 ). Bursts of reperfusion-induced ventricular arrhythmias may be associated with larger infarcts and might represent a biomarker of reperfusion injury. Reperfusion arrhythmias may also pose a potentially unrecognized problem in the setting of reversal of coronary vasospasm (such as in Prinzmetal angina), which is, on rare occasion, followed by sudden death.
No-Reflow Phenomenon and Microvascular Obstruction
The no-reflow phenomenon refers to the inability to perfuse portions of the myocardium because of microvascular obstruction after successful reopening of a closed proximal epicardial coronary artery. The fluorescent dye, thioflavin S, is used in animal models to show the anatomic no-reflow zone ( Figure 24-2 ). In a study performed in anesthetized canines in the mid-1970s, we observed that when the proximal circumflex coronary artery was occluded for 90 minutes, followed by reperfusion, that thioflavin S, when injected into the vasculature, failed to penetrate the subendocardium of the posterior left ventricular wall. An electron microscopic examination ( Figure 24-3 ) of these nonfluorescent zones of no-reflow revealed capillaries with focal endothelial swelling, or “blebs” that appeared to be obstructing blood flow. Areas of microvascular hemorrhage with extrusion of red blood cells into the interstitial space adjacent to endothelial gaps were also observed. Occasionally, capillaries appeared to be compressed by adjacent swollen cardiomyocytes, and damaged vessels appeared to be plugged by fibrin tactoids, platelets, and neutrophils. It is likely that the cause of no-reflow is structural damage to the microvasculature within the MI zone ( Figure 24-4 ). The size of the zone of no-reflow within the infarct is correlated with the duration of coronary occlusion ( Figure 24-1B ). We showed that streptokinase, tissue plasminogen activator, or dabigatran could not prevent no-reflow. The size of the no-reflow zone expanded over several hours after reperfusion of the proximal, patent coronary artery. The growth of the anatomic no-reflow zone, which was assessed by injecting fluorescent dye into the vasculature, was accompanied by a deterioration of regional myocardial blood flow within this zone. These findings suggest that whatever damage occurs to the microvasculature happens partially during the reperfusion phase, and therefore, no-reflow is a form of reperfusion injury.
An influx of electrolytes, fluid, calcium, and reactive oxygen species during the early minutes of reperfusion might result in damage to endothelial cells, including focal and diffuse edema, and eventual rupture of endothelial cells that then blocks flow. The zone of no-reflow rapidly expands outward within the infarcted region, finally leveling off at approximately 2 to 8 hours. In general, we observed that the zone of no-reflow is contained within the MI zone (myocyte cell death). We do not believe that no-reflow contributes directly to myocyte cell death. However, experimental and clinical studies increasingly suggest that the no-reflow phenomenon is an important marker of prognosis. The larger the extent of no-reflow, the more myocardial infarct expansion and left ventricular remodeling was likely to occur. This observation makes intrinsic sense; if blood elements cannot access the zone of necrotic debris, healing of the scar is impaired.
No-reflow is now well documented in patients and is clinically manifest in up to 30% of those undergoing reperfusion therapy for acute MI. Several imaging techniques, including nuclear imaging (see Chapter 32 ), echo-contrast (see Chapter 31 ), and magnetic resonance imaging (see Chapter 33 ) have visualized anatomic zones of no-reflow or microvascular obstruction in patients. Similar to the findings in experimental animal studies, the zones of no-reflow observed in humans are typically subendocardial to mid-myocardial and appear to be confined to the necrotic zone. As in animal studies, the presence of no-reflow in patients is associated with greater left ventricular dilation and adverse left ventricular remodeling, as well as a worse clinical outcome, including higher mortality rates and more congestive heart failure. Some studies have shown that the size of the no-reflow zone is a marker of poor clinical outcome, independent of the size of the MI.
No-reflow in humans is complicated by the fact that during PCI, atherosclerotic fragments and thrombi can break off and cause distal embolization, which contributes to additional distal microvascular obstruction. No-reflow may be observed in the catheterization laboratory even during routine nonemergent coronary angioplasties or stenting procedures. In the catheterization laboratory, sluggish flow in an otherwise patent epicardial coronary artery and reduced myocardial blush grade are signs of low or no-reflow and predict a poor outcome.
Lethal Myocardial Cell Injury Caused by Reperfusion
The fourth, most important, but still controversial component of reperfusion injury, is lethal myocardial cell injury caused by reperfusion itself. In this construct, cardiomyocytes are injured, but still alive at the end of ischemia. However, once reperfusion occurs, the cardiomyocytes become irreversibly injured and die because of the act of reperfusion. Potential mechanisms in which reperfusion could cause a reversibly injured cardiomyocyte to become irreversibly injured (dead) are listed in Table 24-1 . In particular, the generation of oxygen-free radicals and the opening of the mitochondrial permeability transition pore during reperfusion are believed to be important contributors to ischemia/reperfusion injury and have become targets for therapy (see the section on Prevention and Management of Reperfusion Injury ). The mitochondrial permeability transition pore is a large-conductance megachannel that is closed under physiological conditions, but it opens in response to changes in the mitochondrial membrane potential, reactive oxygen species, and increased calcium concentration. Opening of the pore is believed to be a key event in the initiation of apoptotic cell death.
Support for the concept that lethal cardiomyocyte reperfusion injury is a real phenomenon comes largely from studies in which therapies given right at or shortly before reperfusion further reduce MI size above and beyond reperfusion alone. However, few clinical trials have shown such benefit. Worsening of chest pain or reelevation of ST segments on the electrocardiogram at reperfusion are sometimes interpreted as a clinical manifestation of reperfusion injury.
Prevention and Management of Reperfusion Injury
The prevention and management of ischemia/reperfusion injury are discussed in the following. Interventions to achieve these goals may be divided into those that are directed at managing the clinical manifestations of reperfusion injury, such as supporting contractility in the setting of myocardial stunning, treating reperfusion arrhythmias or limiting infarct size, and those interventions that are aimed at the prevention of the reperfusion injury itself. At present, all therapies in this latter category are investigational only.
General Interventions to Manage Complications of Ischemia/Reperfusion Injury
Stunned myocardium can be prevented by avoiding the initial episode of ischemia. Stunning has been successfully treated by administering oxygen radical scavenging agents at the time of reperfusion. In addition, stunned myocardium can be overcome by using inotropic drugs. Because stunned myocardium is a transient phenomenon and responds to inotropic drugs, it can usually be managed in the clinical setting with appropriate titration of such drugs.
In general, reperfusion arrhythmias that occur in the setting of MI are amenable to therapy, and in most instances, are not a major clinical problem. Studies in our laboratory have shown that reperfusion-induced ventricular arrhythmias are responsive to treatment with lidocaine, sotalol, ranolazine, and postconditioning.
Interventions Aimed at Reducing Microvascular Obstruction
In general, we have observed that most agents that reduce MI size also reduce the size of the no-reflow zone within the ischemic risk zone. A few therapies have been notable for their ability to reduce no-reflow, even when administered relatively late during the occlusion or even after reperfusion has begun.
When hypothermia was introduced early during a coronary artery occlusion/reperfusion protocol in the rabbit, it reduced both MI size and the extent of the anatomic zone of no-reflow. When hypothermia was introduced 5 minutes before or 5 minutes after reperfusion, it failed to reduce MI size, but still reduced no-reflow. In one study, we introduced hypothermia as late as 30 minutes after reperfusion. Again, it had no effect on MI size, but it still reduced the area of no-reflow.
Agents Influencing Free Radicals
Bendavia (Stealth Biotherapeutics, Newton, Mass.), a mitochondrial targeting peptide given at reperfusion, resulted in a nonsignificant trend toward reducing infarct size, but it also reduced the anatomic zone of no-reflow for any ischemic risk zone size. The oxygen radical scavenging agents, superoxide dismutase plus catalase, which were given at reperfusion in our canine model of 2 hours of proximal coronary artery occlusion and 4 hours of reperfusion, did not reduce MI size, but it did reduce microvascular injury and also reduced low reflow. These studies all show that administering therapy at or close to the time of reperfusion, or even after reperfusion reduces no-reflow and microvascular injury, provides additional support for the concept that no-reflow is a true form of reperfusion injury.
Various therapies that have shown some success in improving low or no-reflow after cardiac catheterization include vasodilators (e.g., nitroprusside, adenosine, and calcium channel blockers) such as verapamil, diltiazem, and others.
Interventions Aimed at Preventing Reperfusion Injury That Causes Cellular Necrosis
The primary treatment of acute STEMI should be early and complete reperfusion of the occluded culprit epicardial coronary artery (see Chapter 13 ). Reperfusion salvages tissue, has a substantial impact on infarct size, and improves clinical outcome. However, if the act of reperfusion also kills some population of cardiomyocytes that are hovering between life and death, then efforts to reduce this reperfusion injury might further improve outcome ( Figure 24-5 ). Numerous clinical trials have attempted to further reduce MI size by administering adjunctive therapies near the time of reperfusion to mitigate reperfusion injury. There is a long list of the agents that showed promise in preclinical trials but that failed in clinical trials. We have previously described these trials, and discussed reasons why these trials may have failed.