Brain Energy Failure

The brain depends on large amounts of exogenous substrate (oxygen and glucose) because of its high metabolic demands, limited energy stores, and reliance on oxidative metabolism. Cellular energy depletion is postulated to be the triggering event that initiates the many cascades of injury that occur during ischemia [Krause et al., 1988]. Adenosine 5-triphosphate (ATP) is the energy source that drives all cellular physiologic processes [Bellamy et al., 1996; Sweeney et al., 1995]. One of the most vital processes performed is the preservation of ATP to maintain membrane ionic gradients [Ackerman and Clapham, 1997]. Under normal conditions, maintenance of ionic gradients across the cell membrane accounts for nearly 50 percent of total cellular expenditure [Erecinska and Silver, 1989]. Interruption of cerebral blood flow results in loss of consciousness and reductions in electroencephalographic (EEG) activity within seconds. Within 5–7 minutes, complete energy failure occurs, accompanied by disturbances of neuronal and glial ion homeostasis, including sodium and water influx and efflux of potassium (Figure 76-1). When the extracellular potassium concentration reaches a critical threshold, voltage-gated channels undergo depolarization, precipitating extracellular calcium influx [Sweeney et al., 1995]. If flow remains inadequate and energy failure persists, calcium-mediated events, such as phospholipase and protease activation, can lead to irreversible cellular injury and necrosis, as well as cerebral acidosis with elevated lactate levels and decreased hydrogen ion concentration (pH). If blood flow is restored, recovery of basal cellular metabolism (ATP levels, protein synthesis, oxygen consumption) and normal pH occurs if the ischemic injury is of limited duration [Ljunggren et al., 1974].

Fig. 76-1 Global ischemia results in a cascade of events.

These include reduction in adenosine triphosphate (ATP), loss of cellular ionic gradients, increases in extracellular K⁺ concentration, glutamate release with activation of N-methyl-d-aspartate (NMDA) and non-NMDA receptors, calcium influx and release of intracellular calcium, and activation of secondary messenger systems and a variety of destructive enzymatic processes.

Certain neurons have long been known to be especially vulnerable to global hypoxic-ischemic insults, particularly the hippocampal CA1 and CA4 zones, and cortical layers III and V [Krause et al., 1988; White et al., 1996]. Five minutes of complete global brain ischemia produces cell death in these regions within 48–72 hours. This time interval between ischemia and cell death consists of an early postischemic period when ATP and lactate levels recover, followed, after 24 hours, by a decrease in ATP levels and lactate re-accumulation [Kochanek, 1993; Siesjo et al., 1995].

After anoxia or ischemia, restoration of aerobic metabolism is essential but not sufficient for recovery. Although the restoration of brain perfusion (as occurs with successful CPR) will re-establish energy stores, cell injury and death continue in a process known as reperfusion injury [Hoesch et al., 2008]. Mechanisms of reperfusion injury include lipid peroxidation and other damage caused by oxygen free radicals that accumulate after reintroduction of oxygen to the ischemic region, and neuronal damage mediated by inflammatory cells. Reperfusion is also characterized by continued activation of extracellular glutamate and intracellular calcium-dependent enzyme systems, calcium-dependent gene expression, caspase activation and immune-mediated damage by microglia [Hoesch et al., 2008].

Despite global metabolic recovery, certain neurons progress to cell death. Restoration of blood flow and oxidative metabolism after energy failure in the brain leads to specific cellular and molecular dysfunction because of two related processes – cell necrosis and apoptosis – both of which evolve during and after resuscitation [Hockenbery, 1995; Schnaper, 1994]. These processes, coupled with the need to restore highly integrated functions, explain the unfortunate clinical scenario of the vegetative state despite restoration of normal function in other organ systems. The relation of these two forms of cell death to selective vulnerability of neurons in the brain is beginning to emerge.

Cell necrosis, which is characterized by denaturation and coagulation of cellular proteins, is the basic pattern of pathological cell death and results from progressive reduction in cellular ATP content [Buja et al., 1993; Sweeney et al., 1995]. Necrosis involves progressive derangements in energy and substrate metabolism that are followed by a series of morphologic alterations, including swelling of cells and organelles, development of subsurface cellular blebs, amorphous deposits in mitochondria, condensation of nuclear chromatin, and breaks in plasma and cell organelle membranes [Brierley et al., 1973]. It was traditionally assumed that an ischemic cell death occurred through this process and this selective vulnerability represented a specific predilection for the development of necrosis in certain neurons after transient ischemic insults.

It has been demonstrated, however, that cell death after hypoxic-ischemic insults can occur by a second pathway, apoptosis. Programmed cell death is important during development, in normal physiology, and in the pathogenesis of disease it is called apoptosis [Schnaper, 1994]. Apoptosis has been described as directed cell suicide [Hockenbery, 1995]. Its unique feature is that, although external events may trigger apoptosis, all of the machinery for the process is contained within the cell. Activation of this machinery may result from toxicity of the extracellular milieu, binding to a specific cell receptor, or removal of a factor that prevents apoptosis from occurring. Apoptosis is thus clearly distinct from necrosis. Necrosis occurs despite the cell’s effort to survive. The development of apoptosis involves new protein synthesis and the activation of endonucleases with a resultant characteristic cleavage of DNA at linkage regions between nucleosomes to form fragments of double-stranded DNA [Mackey et al., 1997].

The stimuli triggering apoptosis are not clearly defined, although protease activation or oxidant injury to DNA has been proposed. Reports indicate that cell death in selectively vulnerable brain regions, such as the CA1 region of the hippocampus, after transient global brain ischemia, occurs by an apoptotic mechanism. After a global ischemic insult, DNA fragmentation is most pronounced in neurons of the CA1 region of the hippocampus, which suggests that apoptosis may play a role in both selective neuronal necrosis and delayed neuronal death [Nitatori et al., 1995]. The concept of delayed neuronal death is important because it implies that there is a window of opportunity for treatment after global ischemia.

Apoptosis in the post-ischemic brain is not limited to scattered neuronal death in what has traditionally been deemed selectively vulnerable regions; it is also seen in penumbral regions around evolving cerebral infarctions [Li et al., 1995]. The severity of the ischemic insult and other local factors likely determine whether an injured neuron recovers, undergoes programmed cell death, or dies a necrotic death. It is quite possible, although only a speculative notion, that after cardiopulmonary arrest and resuscitation, a continuum exists in neurons from recovery to necrosis that depends on the duration of the insult, the local milieu, and the given brain region. None the less, in any given brain region, whether neuronal death is produced by necrosis, apoptosis, or both, a highly complex series of events is involved during the arrest and after restoration of spontaneous circulation. Because apoptosis occurs in stages, there exist several potential strategies for reducing cell death [Kochanek et al., 2001]. See chapters 13–15 for additional discussion of these mechanisms.

Calcium-Mediated Injury

Calcium plays a strategic role in regulation of many cellular metabolic processes; therefore, the concentration of cytosolic free calcium is tightly controlled. Hypoxic-ischemic injury interrupts intracellular calcium homeostasis, which results in massive increases in the intracellular concentration of calcium. This calcium accumulation is believed to promote irreversible cellular injury [Bellamy et al., 1996; Lipton and Rosenberg, 1994] (Figure 76-2).

Fig. 76-2 Increases in intracellular Ca²⁺ mediate several enzymatic pathways that cause cell injury.

These include pathways of protein kinases, proteases, phospholipases, nitric oxide synthase, other free radical synthesis, and endonucleases, as well as other second-messenger systems. The net effect is related to the severity and duration of ischemia, and a series of complex events that results in cell necrosis and apoptosis, as well as other forms of partial injury that remain poorly understood. ATP, adenosine triphosphate.

Transient calcium accumulation occurs in all cells during ischemia, but secondary irreversible accumulation occurs in the selectively vulnerable zones many hours later. Electrophysiologic studies demonstrate that delayed neuronal death is preceded by neuronal hyperactivity. It is hypothesized that ischemic and early post-ischemic calcium accumulation leads to complex series of derangements in cellular metabolism [Morgenstern and Pettigrew, 1997; Sweeney et al., 1995]. The intracellular accumulation of calcium:

The disturbance of intracellular calcium homeostasis is recognized as a final common pathway of neuronal death, either necrotic or apoptotic. These conditions, in concert with excessive release of calcium-dependent excitatory neurotransmitters (glutamate, aspartate), lead to uncontrolled excitotoxicity and cell death.

Excitotoxic Injury

During hypoxic-ischemic damage, pathologically prolonged membrane depolarization occurs and, in certain neuronal populations, leads to excessive release of neurotransmitters into the synaptic cleft [Lipton and Rosenberg, 1994; Rogers and Kirsch, 1989]. The effect of these neurotransmitters is prolonged by failure of the ATP-dependent presynaptic reuptake mechanisms. Glutamate and aspartate are the major excitatory amino acid neurotransmitters in the mammalian central nervous system (CNS); both also have neurotoxic properties [Rogers and Kirsch, 1989]. Hypoxia-induced neuronal death is mediated by synaptic activity. Inhibition of synaptic glutamate release or blockade of glutamate receptors may prevent hypoxic neuronal injury. Glutamate is the major neurotransmitter in the selectively vulnerable zones and accumulates extracellularly in these regions after hypoxic or ischemic insults. The mechanisms by which glutamate may harm neurons during ischemia and reperfusion are becoming more clearly defined [Kalia et al., 2008].

Glutamate is released at the presynaptic terminal in response to neuronal stimulation and acts by binding to postsynaptic dendritic receptors. Two main classes of excitatory neurotransmitter receptors have been identified. One class consists of the ligand-gated ion channels (ionotropic receptors) and includes N-methyl-d-aspartate (NMDA); alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA), or quisqualate; and kainate receptor types [Rogers and Kirsch, 1989; Sweeney et al., 1995]. Toxicity caused by NMDA receptor activation is usually rapid, whereas AMPA or kainate receptor-mediated cell death is slower to develop [Dugan and Choi, 1994]. The current understanding of the glutamate receptor subtypes is that the NMDA receptor is more complex and contains more modulatory sites than does the kainate receptor [Lipton and Rosenberg, 1994]. The other class of excitatory neurotransmitter receptors includes the metabotropic receptors that are coupled with G proteins and modulate intracellular second messengers such as calcium, cyclic nucleosides, and inositol triphosphate. Seven subtypes of the metabotropic glutamate receptor have been identified [Nakanishi, 1992].

When activated, the ionotropic glutamate receptors open sodium channels and may also have an important role in initiation and propagation of membrane depolarization and spreading depression. With inotropic receptor activation, rapid excitatory amino acid-mediated calcium accumulation occurs [Sweeney et al., 1995]. In the presence of ischemia, this calcium accumulation is exacerbated by cellular energy failure, which disables the Na⁺/K⁺-ATPase membrane pump and results in further calcium accumulation. Altered calcium homeostasis leads to activation of many deleterious processes, including activation of phospholipases, proteases, endonucleases, protein kinases, and calmodulin-regulated enzymes such as nitric oxide synthase. There is evidence that nitric oxide is an important mediator of glutamate neurotoxicity [Dawson, 1994]. Re-establishment of the energy supply can reverse these changes. Delayed glutamate-related neuronal injury is most likely caused by activation of inotropic receptors and subsequent calcium influx.

Intracellular free calcium may increase during ischemia either by influx of extracellular calcium or by release of bound or sequestered intracellular calcium [Morgenstern and Pettigrew, 1997]. Cerebral ischemia is accompanied by a significant decrease in extracellular calcium concentration, which is consistent with an intracellular shift of calcium. Energy failure during ischemia results in membrane depolarization, which allows calcium influx via voltage-sensitive calcium channels and causes glutamate release. Glutamate can promote calcium influx by three different mechanisms [Haun et al., 1996]. The first and most obvious mechanism is by opening the NMDA receptor-gated calcium channel. Second, stimulation of non-NMDA receptors opens sodium channels and results in massive influx of sodium and subsequent membrane depolarization, which then allows calcium entry via voltage-sensitive calcium channels. Third, the ion channel gated by non-NMDA receptors allows direct influx of calcium. Three additional possibilities exist for calcium entry during ischemia:

Massive influx of sodium during ischemia creates conditions that inhibit or even reverse sodium/calcium exchange, leading to further calcium influx. It is also possible that calcium enters through areas of the cell membrane that have been damaged by the ischemic insult. Release of bound or sequestered calcium likely occurs through two different mechanisms. First, hydrogen ion, which accumulates during ischemia, can directly displace bound calcium. Second, sequestered intracellular calcium may be mobilized by inositol-1,4,5-triphosphate, which is produced by glutamate-stimulated hydrolysis of polyphosphoinositides, mediated by the metabotropic receptor [Berridge and Irvine, 1984; Sugiyama et al., 1987].

Although most work has centered on the role of calcium influx, some investigations have focused on the importance of the release of sequestered calcium. Mitani et al. [1993] demonstrated that two-thirds of the increase in the intracellular concentration of calcium seen in cultures of hippocampal neurons subjected to glucose and oxygen deprivation is caused by release of sequestered calcium, whereas only one-third is caused by extracellular calcium influx. Although the relative contribution of influx and release from internal stores remains unclear, there is accumulating evidence that release from internal stores does play a role.

Activation of Intracellular Enzymes

The marked increase in intracellular calcium concentration activates at least four classes of enzymes [Morgenstern and Pettigrew, 1997; Sweeney et al., 1995]. Phospholipases break down the lipid cellular membrane, releasing arachidonic acid that generates prostaglandins and free radicals. Protein kinases activate enzymes in an unordered manner, including nitric oxide synthase and xanthine oxidase, which are also generators of free radicals. Proteases begin the uninhibited breakdown of the cytoskeleton, and endonucleases initiate DNA fragmentation.

Protease activation may play a central role in mediating both necrosis and programmed cell death. With regard to necrosis, numerous calcium-dependent enzymes become activated during ischemia and produce important neuronal structural injury. One class of calcium-dependent proteases, calpains, has received the greatest amount of attention. Calpains are cytosolic thiol proteases that degrade numerous cytoskeletal proteins such as neurotubules and neurofilaments, as well as activate protein kinase C and phospholipases. Inhibition of calpain activation has produced marked reduction in ischemic brain injury, particularly after focal ischemia [Bartus et al., 1994]. Proteases may also play a pivotal role in the initiation of programmed cell death.

There is growing evidence that some neuronal death after ischemia is mediated by the activity of cysteine-requiring aspartate-directed proteases (caspases), the protease responsible for apoptosis in mammals [Hotchkiss et al., 2009].

Phospholipase Release of Free Fatty Acids

Metabolism of membrane phospholipids through activation of phospholipases is postulated to play a key role in the pathophysiology of ischemic brain injury (Figure 76-3).

Fig. 76-3 Membrane phospholipids are hydrolyzed by two pathways: the PLA₁/PLA₂ pathway and the PLC/DAG lipase pathway.

Arachidonic acid, produced by phospholipid hydrolysis, is the substrate for eicosanoid synthesis and oxygen radical production. DAG, diacylglycerol; FFA, free fatty acids; IP3, inositol 1,4,5-triphosphate; PC, phosphatidylcholine; PE, phosphatidylethanolamine; PI, phosphatidylinositol; PLA₁, phospholipase A₁; PLA₂, phospholipase A₂; PLC, phospholipase C; PS, phosphatidylserine.

Free fatty acids are released from neuronal membranes during ischemia; the amount of release is proportional to the duration of ischemia. Free fatty acid release is the only known cerebral metabolic indicator that continues to increase in proportion to the duration of ischemia after completion of energy failure [Shin et al., 1983]. Free fatty acids are released by two distinct but related processes. First, phosphatidylinositol is hydrolyzed by phospholipase C with the production of diacylglycerol and inositol phosphates [Abe et al., 1987]. Phospholipase C-mediated hydrolysis begins during the initial moments of ischemia and is related to the magnitude of neurotransmitter receptor stimulation. Diacylglycerol is then hydrolyzed by lipases to free fatty acids, predominantly arachidonic acid and stearic acid. Second, other brain glycerophospholipids are hydrolyzed by phospholipase A₂, which is activated by increases in intracellular calcium concentration. The process of free fatty acid release and metabolism is not a generalized process in the neuronal membrane but is concentrated in synaptic regions and is thus related to excitotoxicity.

The free fatty acids released have potential detrimental effects through at least three mechanisms. First, free fatty acid metabolism via the cyclo-oxygenase pathway contributes to oxygen radical production during reperfusion [Kontos, 1987]. Second, free fatty acid and diacylglycerol directly increase membrane fluidity, inhibit ATPases, increase neurotransmitter release, promote brain edema, and uncouple oxidative phosphorylation. Third, enzymatic oxidation of arachidonic acid during reperfusion by cyclo-oxygenase, lipoxygenase, or cytochrome P-450 produces a large number of bioactive lipids (prostaglandins, thromboxanes, leukotrienes, and hydroxyl acids), many of which have detrimental effects. Ischemia induces a pro-inflammatory state that increases tissue vulnerability to further injury on reperfusion [Collard and Gelman, 2001].

Activation of Nitric Oxide Synthesis

Nitric oxide plays a multifaceted role in the brain as a neurotransmitter and a regulator of cerebral blood flow [Bhardwaj et al., 1997; Dawson and Dawson, 1995; Dawson et al., 1992]. Sites of nitric oxide production include neurons, vascular endothelium, perivascular neurons, and astrocytes [Bhardwaj et al., 1997]. If present in abnormally high concentrations, nitric oxide may exert neurotoxic effects [Dawson, 1994; Moncada and Higgs, 1993].

Nitric oxide is produced in a reaction catalyzed by nitric oxide synthase, in which oxygen and l-arginine are converted into l-citrulline and nitric oxide. Three isoforms of nitric oxide synthase have been described [Samdani et al., 1997] (Table 76-3). There are two constitutive isoforms: neuronal and endothelial. An inducible isoform has also been described and has been demonstrated in macrophages, microglia, and astrocytes. The constitutive enzymes are calcium-calmodulin-dependent enzymes. Inducible nitric oxide synthase is calcium-independent, is induced by endotoxin and cytokines, and, when stimulated, produces large amounts of nitric oxide in a sustained manner [Clark et al., 1996]. Neurons containing nitric oxide synthase are present throughout the brain, with the highest concentration in the cerebellum and the lowest concentration in the medulla.

Table 76-3 Nitric Oxide Synthase (NOS) Isoforms

In some neurons, nitric oxide synthase activity is regulated by the NMDA receptor. NMDA receptor stimulation results in calcium influx and activation of nitric oxide synthase. Nitric oxide is produced and then diffuses to target cells and stimulates guanylate cyclase, which leads to the production of cyclic guanosine monophosphate. Cyclic guanosine monophosphate then produces the physiologic effect (e.g., vasorelaxation, cell signaling). Inducible nitric oxide synthase is expressed by several cell types (including macrophages, microglia, and astrocytes) in response to stimulation by cytokines. Nitric oxide produced by this nitric oxide synthase isoform is involved in cell-mediated cytotoxicity [Samdani et al., 1997]. The mechanism of this cytotoxicity is not yet fully understood but may involve inhibition of key enzymes necessary for DNA replication and mitochondrial energy production.

As part of the excitotoxic cascade of injury, glutamate release activates both endothelial and neuronal nitric oxide synthase, which increase nitric oxide production during focal and global ischemia [Bari et al., 1997; Samdani et al., 1997]. During ischemia, endothelial nitric oxide synthase production may be protective by increasing cerebral blood flow, whereas during ischemia, and more so during reperfusion, neuronal nitric oxide release synthase production may cause additional neurotoxicity. Nitric oxide produced under conditions of ischemia contributes to the cytotoxicity of glutamate, presumably through hydroxyl radical production (Figure 76-4). Nitric oxide contributes to hydroxyl radical production through its reaction with the superoxide radical. Nitric oxide and superoxide radical react to form peroxynitrite, which then decomposes to yield nitrogen dioxide and the toxic hydroxyl radicals. Several days after focal ischemia, activation of inducible nitric oxide synthase from phagocytic cells may contribute to delayed injury. The majority of evidence suggesting a role for nitric oxide has been demonstrated in models of focal ischemia. Less convincing but definite evidence exists for a potential neurotoxic role during global ischemia. Although one of the major effects of nitric oxide is to increase cyclic guanosine monophosphate, multiple mechanisms have been proposed to define its neurotoxic effects.

Fig. 76-4 Excessive nitric oxide is formed on sustained glutamate stimulation of N-methyl-d-aspartate receptors (NMDA-R).

Nitric oxide (NO) freely diffuses to adjacent target neurons, in which it combines with the superoxide anion (O2⁻), produced by mitochondria and xanthine oxidase, to yield the peroxynitrite anion (ONOO⁻), which is an extremely potent oxidant. ONOO⁻ also is protonated and decomposes to the hydroxyl (OH) free radical and the nitrogen dioxide (NO₂) free radical, which are potent activators of lipid peroxidation. Damaged DNA activates the nuclear enzyme poly-adenosine diphosphate (ADP)-ribose synthetase (PARS). PARS transfers ADP-ribose units to nuclear proteins by using nicotinamide adenine dinucleotide (NAD) as the source of ADP-ribose. For every mole of NAD consumed in this reaction, it takes four energy equivalents of adenosine triphosphate to regenerate NAD from nicotinamide. PARS can transfer more than 100 ADP-ribose units/protein in a matter of seconds. This rapid consumption of energy can deplete a cell of its energy stores. It is hypothesized that, if there is sufficient DNA damage and free radical production, the activation of PARS initiates a futile cycle, resulting in the complete depletion of cellular energy stores, impairment of the ability to regenerate those energy stores, and, subsequently, cell death. CaM, calmodulin; l-arg, l-arginine; l-cit, l-citrulline; NOS, nitric oxide synthase; ONOOH, peroxynitrous acid.

An established pathway of nitric oxide-mediated neuronal cell death is nitric oxide activation of the nuclear enzyme poly-adenosine diphosphate (ADP)-ribose-synthetase (PARS) [Dawson and Dawson, 1995; Samdani et al., 1997]. Nitric oxide activates PARS by damaging DNA. PARS participates in DNA repair by catalyzing the transfer of ADP-ribose units from nicotinamide adenine dinucleotide (NAD) to nuclear proteins. For every mole of ADP-ribose transferred, 1 mol of NAD is consumed, and four free energy equivalents of ATP are necessary to regenerate NAD. Therefore, over-activation of PARS can rapidly deplete cellular energy stores.

Depending on its source, nitric oxide may be toxic or protective in the brain under ischemic conditions. Overproduction of nitric oxide from either neuronal or inducible nitric oxide synthase leads to neurotoxicity; however, nitric oxide production from endothelial nitric oxide synthase protects brain tissue by maintaining regional cerebral blood flow. Studies emphasize the necessity for developing truly selective inhibitors for neuronal and inducible nitric oxide synthase to protect the brain adequately from ischemic injury caused by overproduction of nitric oxide and simultaneously to maintain or enhance regional cerebral blood flow [Samdani et al., 1997].

Formation of Oxygen Radicals

Toxic oxygen radical species, produced during post-ischemic reperfusion, have been implicated as important contributors to reperfusion injury and delayed cell death [Siesjo et al., 1989]. Oxygen free radicals are not one specific compound. They are a group of substances that are formed during ischemia, when oxygen becomes unavailable as the terminal electron acceptor in the electron transport chain. A free radical is any molecule that has an unpaired electron in its outermost orbit; such molecules are short-lived and highly reactive. They include nitric oxide, superoxide anion, hydrogen peroxide (not itself a free radical), hydroxyl radical, and peroxynitrate. The lone electron results in molecular instability and the tendency to initiate and propagate chain reactions.

Oxygen free radicals are generated during arachidonic acid metabolism to prostaglandins, as a byproduct of xanthine oxidase-catalyzed production of uric acid through catecholamine oxidation, through a mitochondrial leak during the oxidation of hemoglobin, and through the action of nitric oxide synthase [Kontos, 1987; Siesjo et al., 1995; Traystman et al., 1991] (Figure 76-5). Free radicals are quite destructive to cellular components such as membrane lipids, especially in the presence of iron. Iron is normally transported in the blood, tightly bound to transferrin and stored inside the cell bound to ferritin. In ischemic conditions with accompanying acidosis, iron may be displaced from its normal binding sites and can catalyze reactions that promote oxygen radical formation [Komara et al., 1986]. Most commonly implicated is the Haber–Weiss/Fenton reaction, whereby potent hydroxyl radicals are produced from hydrogen peroxide in the presence of free iron.

Fig. 76-5 Sources and physiologic effects of oxygen free radicals.

Reperfusion of ischemic neurons and glia produces several highly destructive free radicals, including nitric oxide (NO), peroxynitrite (ONOO⁻), hydrogen peroxide (H₂O₂), and superoxide (O₂⁻). These free radical species peroxidize cell membranes, alter the blood–brain barrier, and disrupt DNA.

The impact of the liver and the intestines in producing dysfunction in other organ systems needs further clarification. The liver and intestines have a great deal of xanthine dehydrogenase, which during ischemia can be converted to xanthine oxidase [Bellamy et al., 1996; Nielsen et al., 1997]. The latter could enter the systemic circulation, generate free radicals, and cause additional dysfunction and injury at distant sites.

The brain may be particularly vulnerable to free radical injury for several reasons. One is the high concentration of polyunsaturated fatty acids, especially arachidonic acid. As noted previously, free fatty acids are released throughout ischemia. On exposure to oxygen radical species, these free fatty acids are vulnerable to lipid peroxidation [Krause et al., 1988]. Cerebrospinal fluid has low concentrations of iron-binding proteins; therefore, iron released from injured neurons or glia is likely to contribute to these peroxidation reactions. Byproducts of these reactions – for example, malondialdehyde and conjugated dienes – have been used as markers of the extent of lipid peroxidation after brain injury [Schmidley, 1990]. Lipid peroxides accumulate in the selectively vulnerable zones during reperfusion after transient forebrain ischemia [Bromont et al., 1989; Komara et al., 1986]. The peroxides do not accumulate during the ischemic period itself or in areas that are not reperfused, and thus are implicated in reperfusion injury [Oliver et al., 1990].

Lipid membranes are a natural target of free radicals, especially in the brain, because they are abundant, and their polyunsaturated nature makes them easy to oxidize [Schmidley, 1990]. The effects of free radicals on membranes include changes in membrane fluidity and alteration of ion channels and transport proteins. In addition to disruption of the cell membrane, fragmentation of the mitochondrial membrane results in a decrease in cellular energy production.

The ability of the cell to defend itself against free radicals is limited, and free radical scavenging enzymes, such as superoxide dismutase and catalase, may be overwhelmed after an ischemic area is reperfused [Vannucci and Perlman, 1997]. Surprisingly, restoration of blood flow to hypoxic-ischemic areas is potentially detrimental because the influx of oxygen can be used as a source of oxygen free radicals through the processes noted previously. Stimulated by high tissue carbon dioxide tension, low oxygen tension, and low pH, restored blood flow to an ischemic area is often increased above normal (luxury perfusion). In experimental preparations, oxygen free radicals are detected primarily during reperfusion. After luxury perfusion, there is prolonged hypoperfusion and an associated decrease in cerebral metabolism. In animal experiments, these decreases are not seen with pretreatment with oxygen free radical scavengers.

Inflammatory Reaction

There is growing evidence that acute inflammation contributes to neuronal damage after ischemia and reperfusion [Hoesch et al., 2008; Adrie et al., 2002]. The mechanisms for inflammatory injury can be divided into three broad categories: adverse effects on blood rheology; vasculature injury; and cytotoxic injury to neurons [Doherty and Hutchison, 2009]. They are also discussed in more detail in Chapter 15.

Genetic Damage and Regulation

Two aspects related to global asphyxial injury and genetic regulation are emerging as critical components of the injury cascade. The first concerns mechanisms associated with injury and repair of the genetic material itself, and the second concerns regulation of multiple genes and the proteins that they in turn regulate in response to hypoxia-ischemia.

Studies have demonstrated that ischemia and reperfusion can induce gene damage in the CNS. Reactive oxygen species generated by cerebral oxidative stress interact with nucleic acids and cause the formation of oxidative DNA and RNA lesions that result from DNA base modifications or single-stranded breaks in neurons or astrocytes [Liu and Arora, 2002; Liu et al., 2001]. Oxidative DNA lesions cause a change in coding properties during DNA and RNA synthesis (replication and transcription), or may terminate chain elongation during transcription and translation. Either process can affect protein synthesis.

The development of microarray systems for gene expression profiling since the mid-1990s has provided insights into the mechanisms of the brain’s response to global hypoxic injury [Jin et al., 2001; Papadopoulos et al., 2000; Tang et al., 2002]. Future studies should not only increase the understanding of the molecular basis of ischemic injury, but also suggest potential therapies. Because investigators in such studies frequently use several thousand gene transcripts, using different animal models with varying degrees of severity and duration of global hypoxia-ischemia, as well as varying time periods of reperfusion, it will be some time before a unified or general understanding of the underlying mechanisms of injury are more clearly appreciated.

Early during hypoxia-ischemia, overall gene expression in the brain is reduced or ceases to maintain energy metabolism to support essential cell activities. Immediate early genes undergo rapid induction after global ischemia [Kogure and Kato, 1993]. Some of the immediate early gene products that have been identified include c-fos, c-jun, and the zinc finger family of proteins; heat shock proteins; and amyloid precursor protein. Immediate early genes are transcription factors that bind to target genes and alter their expression. These stress response proteins may directly affect neuronal death and survival after ischemia. However, despite much research into their expression, it remains unclear which ones contribute to the outcome of hypoxic-ischemic injury [Papadopoulos et al., 2000]. This confusion occurs partly because some are upregulated, whereas others are downregulated, and their respective neurotoxic or neuroprotective roles remain to be defined.

The heat shock proteins are an early expressed group; they are molecular chaperones, which bind intracellular proteins and prevent their inappropriate folding [Papadopoulos et al., 2000]. Hsp-70 is markedly induced by hypoxia and is neuroprotective. Hsp-32 is another heat shock protein that may be neuroprotective after hypoxia. It encodes heme-oxygenase-1, which, after induction, has been found to protect neurons in models of transient forebrain ischemia, possibly by producing bile pigments that are strong antioxidants, as well as carbon monoxide, which can cause vasodilatation and inhibits platelet aggregation.

Growth factors are also released within hours after ischemia and may also serve a neuroprotective role. They are polypeptides and the major types of growth factors include nerve growth factor; brain-derived neurotropic factor; neurotrophin-3, 4, 5; and basic fibroblast growth factor [Papadopoulos et al., 2000]. Nerve growth factor, brain-derived neurotropic factor, and neurotrophin-3 bind receptors on the cell surface that have tyrosine kinase activity, and activation of tyrosine kinase is associated with neuronal survival.

Hours later, cytokines, adhesion molecules, isoforms of nitric oxide synthase, and gene products involved in apoptosis are upregulated. Some of these genes protect the cells from dying; some promote the death of injured cells; and others have no known effects on cell survival. Cytokines are produced by activated lymphocytes, macrophages, and astrocytes, and include tumor necrosis factor-α, transforming growth factor-β, and interleukin-1β. Cytokines participate in the inflammatory response and accelerate entry of inflammatory cells from the circulation into the brain. Certain cytokines (e.g., tumor necrosis factor-α and interleukin-1β) may also induce the expression of glycoproteins known as adhesion molecules on endothelial cells, which also facilitate recruitment of inflammatory cells into the ischemic brain [Papadopoulos et al., 2000].

Neuronal degeneration can also be promoted by induction of apoptosis genes or genes that cause a stress to the cells, such as those related to free radical production or to the production of various nitric oxide synthase isoforms [Koistinaho and Hokfelt, 1997]. Cells that undergo apoptosis shrink down and ultimately are phagocytosed by neighboring cells. Genes that accelerate apoptosis include members of the bax and bad families, p53, and Fas; those that inhibit apoptosis are associated with bcl-2, bcl-x (L), and DAD1 families [Papadopoulos et al., 2000]. When apoptosis is initiated, cytochrome c is released from damaged mitochondria and, in concert with apoptotic protease-activating factor, induces members of the caspase family, including interleukin-1-converting enzyme. As apoptosis proceeds, nuclear components and the cell skeleton are destroyed, detach other cell constituents, and are phagocytosed.

Two studies demonstrated both the elegance of the application of microarray technology and the dilemmas in synthesizing the results. In one study, 15 minutes of global cerebral ischemia in rats was associated with 1.7-fold or greater increased expression of 57 genes and 1.7-fold or greater decreased expression of 34 genes studied up to 72 hours after hypoxia [Jin et al., 2001]. Induced genes included those involved in protein synthesis, pro-apoptotic genes, anti-apoptotic genes, injury-response genes, receptors, ion channels, and enzymes. Transcriptional induction of several genes was also reported, as was co-induction of several groups of related genes (e.g., vascular endothelial growth factor and its receptor, neuropilin-1). A second study examined 8000 transcripts in rat brains 24 hours after exposure to 6 hours of hypoxia (8 percent oxygen) and found very little gene induction (15 upregulated genes and 11 downregulated genes) [Tang et al., 2002]. In contrast, the same investigators found that, in a model of permanent focal ischemia, 415 genes were upregulated and 158 were downregulated. Presumably, the differences between global hypoxia and ischemia were attributable to greater tissue injury in the ischemic model, but there is no apparent explanation of the differences in the response to global hypoxia in these two groups of studies.

Cerebral Blood Flow and Metabolism After Resuscitation

The pioneering studies, in which global cerebral blood flow and cerebral metabolic rate for oxygen were measured in animal models of global ischemia or cardiac arrest, focused on the early post-resuscitation period. In their classic study, Snyder et al. [1975] demonstrated that, after 15 minutes of global brain ischemia in dogs, cerebral blood flow transiently increased to levels well above baseline. After 15–30 minutes, cerebral blood flow progressively decreased to a level below normal for the rest of the monitoring period (90 minutes). This pattern of early transient post-ischemic hyperemia and subsequent delayed post-ischemic hypoperfusion has been observed almost universally in global cerebral ischemia models, including asphyxia-induced cardiac arrest. The level of hyperemia and subsequent hypoperfusion varies in relation to the duration of the insult. Although these phases of increased and decreased cerebral blood flow characterize the net global effect, regional cerebral blood flow is often heterogeneous, particularly during post-ischemic hypoperfusion, when areas of decreased and increased perfusion may coexist. Metabolism, as assessed by cerebral metabolic rate for oxygen, is reduced during the early post-ischemic period and then progressively recovers to a level that varies, depending on the model used and the duration of ischemia. In some models, including ventricular fibrillation in dogs, significant recovery of cerebral metabolic rate for oxygen may occur during the first few hours, despite persistent post-ischemic hypoperfusion, creating the potential for a secondary ischemic insult during reperfusion. Whether this increase in cerebral metabolic rate for oxygen represents appropriate synaptic activity, seizures, or changes in the basal metabolic rate is uncertain. In other models, global cerebral blood flow and cerebral metabolic rate for oxygen are matched during the first few hours after ischemia. In such cases, post-ischemic cerebral blood flow appears to be determined by the metabolic needs of the brain, and delayed hypoperfusion may not be a significant cause of neuronal injury. Thus, delayed hypoperfusion is likely the result of brain injury rather than the cause [Michenfelder and Milde, 1990].

Of historic interest is the concept of “no reflow,” which was suggested in the 1960s by Ames et al. [1968]. These investigators noted that, after a period of global cerebral ischemia, reflow could not be re-established. At a microscopic level, areas of “no reflow” are interspersed with areas of restored blood flow and microinfarcts [Fink et al., 2008]. Brain regions that are selectively vulnerable include the thalamus, amygdala, hippocampus, and striatum. It is hypothesized that vasospasm, perivascular edema, and increased blood viscosity play a role in the development of this “no reflow” phenomenon.

The second phase of cerebral blood flow is characterized by increased global cerebral blood flow, and is often referred to as the “hyperemic” phase. Some believe that this hyperemic phase is essential for neuronal functional recovery [Fink et al., 2008]. This opinion is based on studies that show that hypertensive reperfusion improves neurologic outcome, whereas post-arrest hypotension worsens neurologic outcome. In addition to hypertensive reperfusion, hemodilatation and thrombolysis in this phase have been shown to improve outcome in animal models [Fink et al., 2008].

Drugs such as nimodipine can increase cerebral blood flow during the early post-ischemic hypoperfusion phase after global cerebral ischemia. Cerebral metabolic rate for oxygen recovery generally is not increased by treatment, even though nimodipine improves neurologic outcome after global ischemia in some models. Safar et al. [1996] reported that a multifaceted flow-promotion treatment strategy to increase cerebral blood flow and reduce cerebral metabolic rate for oxygen early after ventricular fibrillation in dogs improved outcome. This improvement was accomplished with cardiopulmonary bypass, mild hypothermia, hemodilution, and transient hypertension.

Studies in which investigators measured cerebral blood flow and cerebral metabolic rate for oxygen in humans after cardiac arrest shed interesting light on these data from animals. Whereas cerebral blood flow and cerebral metabolic rate for oxygen typically are measured early (0–2 hours after arrest) in experimental animal studies, such measurements in humans are obtained beginning 6–12 hours after arrest. Beckstead et al. [1978] measured global cerebral blood flow and cerebral metabolic rate for oxygen in 25 adults after cardiac arrest. In all, 21 of these patients had a vegetative outcome. In agreement with the experimental animal studies, the researchers observed hypoperfusion and hypometabolism with coupling of cerebral blood flow and cerebral metabolic rate for oxygen. During the subsequent 2 days, however, global cerebral blood flow increased to levels equal to or greater than normal, whereas metabolism failed to recover. Thus, in humans with poor neurologic outcome after cardiac arrest, absolute or relative delayed hyperemia followed the hypoperfusion phase. Results in two more studies in adults have confirmed these findings. Using xenon-133 washout, Cohan et al. [1989] observed that patients regaining consciousness had increased cerebral blood flow (hyperemia) within 24 hours. Similarly, Love et al. [1989] used stable xenon-enhanced computed tomography (CT) and discovered that the combination of diffuse hyperemia and loss of reactivity of cerebral blood flow to changes in PaCO₂ was seen 100 hours after arrest in adults who never regained consciousness.

Thus, immediately after cardiac arrest accompanied by restoration of systemic hemodynamic stability, transient global brain hyperemia occurs and is followed by a period of patchy hypoperfusion. The magnitude and duration of these alterations in flow appear to be related to the duration of the insult. In patients with good outcomes, global cerebral blood flow recovers over a subsequent 24–72 hours, and CO₂ reactivity remains intact. In patients who do not regain consciousness or progress to brain death, absolute or relative cerebral blood flow hyperemia develops, with impaired CO₂ reactivity [Prough and Zornow, 1997]. It must be recognized that the measurement of metabolism in these studies has traditionally been cerebral metabolic rate for oxygen. Bergsneider et al. [1995] used positron emission tomography (PET) in humans after traumatic brain injury and reported the occurrence of delayed hyperglycolysis in some comatose patients. Studies of cerebral glucose utilization with PET after cardiac arrest are needed.

Results from clinical studies of asphyxia-induced cardiac arrest in children are scarce and somewhat conflicting with regard to the prognostic implications. Normal values of post-cardiac arrest cerebral blood flow with loss of responsivity to PaCO₂ were observed in children with poor outcome at 24 hours after asphyxia-induced cardiac arrest by strangulation, as studied with stable-xenon CT [Ashwal et al., 1991]. Similarly, in studies of subjects between 24 and 48 hours after near-drowning, Ashwal et al. [1990] observed low cerebral blood flow and no relation between cerebral blood flow and PaCO₂ in the seven nonsurvivors, which again suggests loss of cerebral blood flow reactivity to changes in PaCO₂. In that study hyperemia was not routinely observed in either vegetative survivors or children who died, but only a single cerebral blood flow measurement was made. Beyda [1987] obtained serial measurements of post-cardiac arrest cerebral blood flow with xenon-enhanced CT in a series of children who had suffered asphyxia-induced cardiac arrest from submersion accidents. Children with good neurologic outcomes had slightly decreased cerebral blood flow values at 12 hours, which increased to normal during the subsequent 24–60 hours. In these children, cerebral blood flow reactivity to CO₂ was intact. Children with an eventual vegetative outcome or brain death exhibited hyperemia with loss or attenuation of CO₂ reactivity. This hyperemia progressed to low or normal flow over the next 12–72 hours in children with vegetative outcome, and progressed to low and then no flow, with the development of brain death.

Sudden Infant Death Syndrome

Sudden infant death syndrome (SIDS) is defined as the sudden death of an infant younger than 1 year that remains unexplained after a thorough case investigation, including performance of a complete postmortem examination, examination of the death scene, and review of the medical history [Willinger et al., 1991]. The addition of requirements for examining the death scene and reviewing the history give greater precision to the diagnosis, but it remains one of exclusion, based on the absence of any specific findings on routine postmortem examination. In many unresolved cases, the history, investigation, or postmortem examination reveals information that excludes the diagnosis of SIDS but does not explain the cause of death; suspected abuse, neglect, or accidental suffocation; vomiting or diarrhea without evidence of infection; unreliable information; or findings such as mild bronchopneumonia, supraglottitis, or maternal drug use are not sufficient to explain death [Gilbert-Barness and Barness, 1995]. Even though almost all cases of sudden unexpected death in infants continue to be ascribed to SIDS, the uncritical certification of the syndrome as a diagnosis, without a postmortem examination, death scene investigation, and careful history, prevents accurate diagnosis of the cause of death [Bass et al., 1986; Valdes-Dapena, 1995].

Epidemiology

Incidence and age at death

SIDS is the leading cause of postneonatal infant death in the United States and is the third leading cause of infant mortality overall [Kinney and Thach, 2009]. Although SIDS may occur during the first month of life, most neonatal deaths can be accounted for by specific perinatal factors, including prematurity, infection, and congenital diseases.

The incidence of SIDS peaks between 2 and 4 months of age; 95 percent of all deaths from the syndrome occur by 6 months of age [Hunt, 1992]. SIDS, therefore, is primarily a disorder that strikes children between 1 and 6 months. It is so rare outside those age limits that the diagnosis in such cases should be made with caution, and only after all possible causes, including metabolic diseases, are excluded [Gilbert-Barness and Barness, 1995]. The age-at-death distribution of SIDS is the most consistent, unique, and provocative characteristic identified [Peterson, 1988].

Identification of risk factors

In recent years, SIDS has been substantially demystified by major advances in our understanding of its relation to sleep and homeostasis, environmental and genetic risk factors, and biochemical and molecular abnormalities [Kinney and Thach, 2009]. The most important advance has been the discovery that the prone (on the stomach) sleep position more than triples the risk of SIDS, which in the early 1990s led to national and international campaigns advocating a supine sleep position for infants [Willinger et al., 1994].

Considerable investigation has been carried out to identify other factors that may contribute to SIDS. Unfortunately, no epidemiological differences have been of sufficient sensitivity and specificity to permit prospective identification of future SIDS victims [Hoffman and Hillman, 1992]. One way of conceptualizing the emerging multidisciplinary data is with the Triple-Risk Model, first proposed in 1994 [Filiano and Kinney, 1994]. Like other models it emphasizes the interaction of multiple factors in the pathogenesis of SIDS [Kinney and Thach, 2009]. According to this model, SIDS occurs when three factors present simultaneously. The first factor is an underlying vulnerability in the infant; the second is a critical developmental period; and third, an exogenous stressor (Box 76-1).

Box 76-1 Components of the Triple-Risk Model of Sudden Infant Death Syndrome

Intrinsic Risk Factors

Genetic

Male sex

Polymorphism in the gene encoding the promoter region of the serotonin transporter

Black or Native American race or ethnic group

Developmental

Prematurity

Environmental

Perinatal exposure to smoking

Parental smoking, ethanol intake, or drug use

Socioeconomic disadvantage

Extrinsic Risk Factors

Sleeping on the side or in the prone position

Soft bedding

Bed sharing, co-sleeping

Mild infections, including colds

(Modified from Kinney HC, Thach BT. The sudden infant death syndrome. N Engl J Med 2009;361:795–785.)

Risk factors for SIDS can be divided into extrinsic and intrinsic categories [Paterson et al., 2006]. Extrinsic risk factors are physical stressors that would place a vulnerable infant at risk for asphyxia or other homeostatic derangement [Kinney and Thach, 2009]. Such extrinsic factors include prone and side-sleeping positions, bed clothes that cover the head, sleeping on sofas or other soft furniture in which the infant could become wedged, a high ambient temperature in the sleeping environment, soft bedding, and bed sharing [Kinney and Thach, 2009].

Sleeping in the prone position by infants has been one of the most consistent risk factors for SIDS; an avoidance of prone sleeping has been the focus of several national campaigns to prevent the syndrome. The evidence that prone sleeping is a risk factor for SIDS has been so compelling that the American Academy of Pediatrics Task Force on Infant Positioning and SIDS [1992] recommended the supine sleeping position for infants. The fear that aspiration would occur if an infant vomited or regurgitated while sleeping in the supine position appears to be unwarranted; in fact, the incidence of death from aspiration has actually decreased among infants sleeping in the supine position [Gilbert-Barness and Barness, 1995]. Although the lateral position for sleeping has also been recommended, positioning infants on their sides may not be as safe, because the infant may roll over into the prone position.

The recommendation for sleep in a nonprone or supine position is for healthy infants only. Gastroesophageal reflux and certain upper airway anomalies that predispose to airway obstruction and perhaps some other illness may be indications for a prone sleeping position [AAP, 1996].

Although the incidence of prone sleep position is currently 20 percent or less, 30–50 percent of infants with SIDS are still found in the prone position [Hauck and Tanabe, 2008; Mitchell et al., 2008].

Scientific studies have demonstrated that bed sharing by mother and infant can alter and synchronize sleep patterns of the two. These studies have led to speculation in the lay press that bed sharing, sometimes referred to as co-sleeping, may also reduce the risk of SIDS. Although bed sharing may have certain benefits (such as encouraging breastfeeding), no scientific studies demonstrate that bed sharing reduces the incidence of SIDS. Conversely, there are studies suggesting that bed sharing, under certain conditions, may actually increase the risk of the syndrome [AAP, 1997].

In addition to extrinsic factors related to external events around the time of death, intrinsic factors are postulated to affect the underlying vulnerability of the infant and thus increase the risk of SIDS. Intrinsic risk factors can be subdivided into developmental factors, such as prematurity, and putative genetic factors, such as familial SIDS, male sex (by 2:1 ratio), and race or other ethnic group [Kinney and Thach, 2009].

Certain genetic polymorphisms have been associated with SIDS [Kinney and Thach, 2009]. Polymorphisms associated with SIDS have been reported in a variety of genes involved in autonomic function, neurotransmission, energy metabolism, and response to infection [Hunt, 2005; Weese-Mayer et al., 2007]. In addition, the vulnerable infant’s response to environmental factors may actually reflect aberrant intrinsic responses. For that reason, events and environmental conditions extrinsic to the infant, such as poverty, adverse prenatal response to certain substances, and postnatal exposure to cigarette smoke may trigger intrinsic responses in the vulnerable infant [Kinney and Thach, 2009].

For example, prenatal exposure to alcohol and cigarette smoke may have a direct effect on neurotransmitter systems that are critical to homeostatic control in the developing human brain [Duncan et al., 2008]. A number of epidemiologic studies have documented that smoking during pregnancy and after birth are two major and independent risk factors for SIDS [Perkin, 2004]. With the reduction in the incidence of infants being put to sleep prone, maternal smoking has become the major modifiable risk factor for SIDS [Horne et al., 2002].