Hypoxic-Ischemic Encephalopathy

Ericka L. Fink

Mioara D. Manole

Robert S. B. Clark

KEY POINTS

Hypoxic-ischemic encephalopathy (HIE) can be a consequence of multiple and diverse etiologies; the most common encountered in the PICU include cardiac arrest, severe shock, and stroke.

Poor outcomes, defined as death or severe neurologic disability, after cardiac arrest are disappointingly high. Children surviving with HIE have a wide spectrum of disability.

Characteristic cerebral blood flow (CBF) patterns occur after cardiac arrest and reperfusion. These include the “no-reflow,” “hyperemic,” “delayed hypoperfusion,” and “recovery” phases. CBF data in animal models suggest that the depth and duration of “delayed hypoperfusion” correlates with the duration of ischemia.

Brain regions that are selectively vulnerable to HIE include the CA1 region of the hippocampus, cortical layers 3 and 5, basal ganglia, and cerebellar Purkinje cells.

High-quality cardiopulmonary (and cerebral) resuscitation and postresuscitation care are essential. Postresuscitation care should focus on the prevention of secondary brain insults (normoxia, normothermia, normocarbia, normotension).

There are no targeted therapies to prevent HIE, but early after resuscitation a therapeutic window exists for the testing of promising interventions such as hypothermia and other physiologic-based experimentally proven therapies.

Clinical hypoxic-ischemic encephalopathy (HIE) is a complex spectrum of brain injury due to diverse etiologies causing hypoxemia, and the brain injury results from hypoxemia and/or ischemia with or without reperfusion. In newborns, infants, and children, the management and prognostication of HIE is complicated by variability in developmental stage at the time of insult, regional vulnerability, and etiology. Consequently, gaining consensus on guidelines and standards for care remain elusive. Taken together, these daunting hurdles deserve special scientific attention, because HIE is a major cause of morbidity and mortality in pediatric patients, and carries a heavy burden for the patient and caregivers not only in terms of disability and quality of life in survivors but also in high costs associated with lifelong healthcare assistance. To date, there is no efficacious treatment for HIE. However, there has been progress in mitigating HIE, as demonstrated with postresuscitative hypothermia use for adults after ventricular tachycardia/fibrillation (VT/VF)-induced cardiac arrest and for neonates after birth asphyxia, and the early application of thrombolytics in adults after embolic stroke. The challenges remain to develop and optimize neuroprotective strategies to further improve outcomes for children at risk of HIE.

CORE PATHOPHYSIOLOGY AND PATHOBIOLOGY

Prolonged hypoxemia and/or ischemia lead to brain injury by triggering multiple pathologic cascades that can lead to cell death and loss of neurologic function. Interwoven with the pathologic cascades are endogenous neuroprotective responses to minimize damage. These cellular responses—with similarities to other types of acute brain injury, such as traumatic brain injury, central nervous system (CNS) infections, stroke, and status epilepticus—are covered in detail in Chapter 56. However, a few fundamental aspects of particular relevance to HIE warrant further comment here.

Global and Focal Hypoxic-Ischemic Encephalopathy

HIE can result from multiple etiologies which cause either global or focal disease. Global HIE can result from respiratory arrest, cardiac arrest, strangulation, poisoning, sagittal

sinus thrombosis, or profound shock states. Cardiac arrest due to asphyxia is the most common cause of global HIE in infants and children, while cardiac arrest due to arrhythmias is the most common cause in adults. Focal HIE can result from embolic or thrombotic stroke, or intracerebral hemorrhage. There are commonalities and important differences between HIE resulting from global and focal insults.

As outlined in Chapter 56, the principal commonality is the interruption of oxygen and blood flow to cells in the brain. Whether it is global due to cardiac arrest or focal due to vascular interruption of local cerebral blood flow (CBF), all parenchymal cells (neurons, astrocytes, oligodendrocytes, or microglia) will not tolerate prolonged durations of ischemia. The degree of brain damage is directly and strongly correlated with the duration of no-flow, and, as such, the clinical outcome is strongly and inversely correlated with the duration of no-flow. Thus the common clinical goal is to restore CBF as rapidly as possible, whether this is by return of spontaneous circulation (ROSC), in the case of cardiac arrest, or revascularization, in the case of stroke.

Final common pathways after critical global or focal ischemia include excitotoxicity from membrane failure with increased release and decreased reuptake of excitatory amino acids such as glutamate and glycine, oxidative stress,
mitochondrial dysfunction, energy failure, and initiation of cell death cascades (Fig. 65.1) (1). Neuronal and glial cell death can result from necrotic, apoptotic, and more recently described autophagic pathways (refer to Chapter 56). Each mechanism of cell death has a characteristic morphologic appearance and temporal pattern that can be observed at the microscopic and ultrastructural levels (2). Necrosis is a process characterized by immediate mitochondrial energy failure, leading to cellular swelling, loss of membrane integrity, and a prominent inflammatory response in surrounding tissues. Apoptosis is an energy-requiring process generally requiring new protein synthesis. Enzymatic degradation of cytoskeletal proteins results in cell soma and nuclear shrinkage, and nuclear deoxyribonucleic acid (DNA) is characteristically fragmented via endonucleases. In contrast to necrosis, apoptosis produces minimal inflammation. Autophagic stress can also result in cell death. Autophagy is an adaptive response to starvation, and results in autodigestion of cellular proteins and organelles to feed the cell. Triggering of autophagy after acute insults could potentially be beneficial or detrimental, likely depending on the degree and duration of injury. The role of autophagy after cerebral ischemia is only recently being investigated.

FIGURE 65.1. Cellular mechanisms resulting in cell death and hypoxic-ischemic encephalopathy. Prominent contributory mechanisms include excitotoxicity, disturbances in calcium homeostasis, oxidative stress, energy failure, and release of substances triggering cell death pathways. Gly, glycine; NMDA, N-methyl-D-aspartate; IP3, inositol 1,4,5-triphosphate; NO, nitric oxide; NOS, nitric oxide synthase; EndoG, endonuclease G; ATP, adenosine triphosphate. PCP, Phencyclidine; PLA, phospholipases A; PLC, Phospholipase C; DAG, diacyl-glycerol; SMA (second mitochondria-derived activator of caspases); C/Diablo, NAD, Nicotinamide adenine dinucleotide. From: Robert S. B. Clark.

The obvious pathophysiologic difference between global and focal critical ischemia is whether systemic hypoxemia and ischemia occurred before, during, or after focal insults. Multiorgan system failure, particularly cardiovascular failure, can profoundly contribute to the pathologic evolution in the postresuscitative phase, making cardiovascular and systemic stabilization a priority after cardiac arrest. Induction of both pathologic and protective hypoxia-inducible cell-signaling pathways can occur prior to the insult in patients with cardiac arrest. These pathways can be triggered in all regions of the brain rather than in those downstream of vascular occlusion in cardiac arrest compared with stroke. Thus, cell-signaling pathways, both bad and good, represent global therapeutic targets after hypoxia-ischemia from cardiac arrest.

Cardiac Arrest

Epidemiology of Pediatric Cardiac Arrest

In one multicenter study, the incidence of out-of-hospital cardiac arrest per 100,000 person-years ranged from 73 for infants, 8 in children aged 1-11, 6 in those aged 12-19 years, to 127 in adults (3). In contrast, the incidence for in-hospital pediatric cardiac arrest was 1.06 per 1000 hospital admissions in a single-center study (4). Some pediatric centers have implemented special medical emergency teams that respond to prearrest and arrest events to address the frequency of in-hospital cardiac arrests and mortality (5).

Table 65.1 outlines several clinical studies examining cardiac arrest in children. In contrast to adults, where the majority of cardiac arrests are due to cardiac arrhythmia and intrinsic heart disease, the majority of cardiac arrests in children are due to asphyxia (6,7). This includes in-hospital and out-of-hospital arrests. Asphyxial cardiac arrest, accounting for ˜80%-90% of pediatric cardiac arrests, begins with respiratory failure, followed by hypoxemia, hypercarbia and acidosis, hypotension, pulseless electrical activity (PEA), then, ultimately, asystole. The most common clinical entities associated with asphyxial cardiac arrest in an outpatient setting include sudden infant death syndrome, pneumonia, aspiration, and submersion injury (8) (Table 65.2). Notably, many of these events may be preventable. The most common etiologies in inpatients include respiratory failure and congenital heart disease complications (9). Cardiac arrhythmias account for ˜10%-20% of pediatric cardiac arrest, with higher frequency occurring in inpatients. Reentrant arrhythmias, Wolff- Parkinson-White syndrome, long QT syndrome, congenital heart disease with postoperative arrhythmia, electrical injury, physical exertion, and trauma are the most frequent causes of arrhythmia in this cohort. Both adults and children who present with VT/VF as an initial rhythm have a higher incidence of good outcome compared with asphyxial arrest. For children with out-of-hospital cardiac arrest presenting

with VT/VF, the survival to hospital discharge is significantly higher versus those presenting with asystole (30% vs. 5%, respectively) (10). This outcome, although less pronounced, appears similar for in-hospital cardiac arrest as children with VT/VF-associated cardiac arrest had the highest survival to hospital discharge (35%), followed by PEA/asystole (27%), and finally late onset VT/VF (11%) (11). Late onset VT/VF is considered a reperfusion arrhythmia (12). The incidence of asphyxial versus VT/VF cardiac arrest reverses with increasing age group. A recent study showed that 7.6% of children aged 1-7 years versus 27% aged 3-18 years presented with VT/VF (13).

TABLE 65.1 REVIEW OF PEDIATRIC CARDIAC ARREST LITERATURE

▪ AUTHORS (YEAR)	▪ OH/IH	▪ N	▪ AGE^a	▪ MALE (%)	▪ ROSC (%)	▪ VT/VF (%)	▪ SURVIVAL TO D/C (%)	▪ GOOD OUTCOME (%)^b	▪ PREDICTORS OF GOOD OUTCOME	▪ PREDICTORS OF POOR OUTCOME
*Retrospective Studies* de Mos et al. (2006)	IH	91	4 y		82	4	25	43	ECMO within 24 h	Renal failure; epinephrine gtt; calcium bolus during CPR
Engdahl et al. (2003)	OH	98	1 y	52		8	5	60
Gerein et al. (2006)	OH	503	5.6 y	58	5	4	2
Hickey et al. (1995)	OH	41	43 mo	68	56	22	27	73	ROSC in field; awake in ED
Horisberger et al. (2002)	IH OH	89	6 mo		87(IH) 57(OH)		58	79		OH arrest; longer ROSC; trauma etiology
Kuisma et al. (1995)	OH	79	2.9 y	53	10	4		80	Witnessed arrest; near-drowning; bystander CPR; rapid ROSC
Parra et al. (2000)	IH	32	3.5 mo		63		42	73	Mechanical support
Pitetti et al. (2002)	OH	189	41 mo	64		4	2.6	0	Sinus rhythm in ED; fewer epinephrine doses
Ronco et al. (1995)	OH	63	14 mo	65	39		9.5	17		First ED rhythm non-VT/VF
Schindler et al. (1996)	OH	80	1 y	50		63	7.5	0	Rapid ROSC	ROSC > 20 min; >2 doses epinephrine
Slonim et al. (1997)	IH	205			24		13.7			Trauma etiology; longer ROSC; higher PRISM III
Suominen et al. (1997)	OH	50	1.2 y		26	8	16	12	CPR < 15 min
*Prospective Studies* Lopez-Herce et al. (2005)	OH	95	63 mo	64	47	9	28	82		ROSC > 20 min
Moler et al. (2009, 2011) and Meert et al. (2009)	IH/OH	353 138	0.9 y 2.9 y	57 69	^c	10 7	49 38	77 24	Both pupils reactive; postoperative CPR	More epinephrine doses; atropine, calcium, or sodium bicarbonate administered; preexisting diseases; electrolyte imbalance or drowning/asphyxia as etiology; low blood pH; duration of CPR
Nadkarni et al. (2006)	IH	880	5.6 y	54	63	14	27	65
Sirbaugh et al. (1999)	OH	300	9 mo	60	11	3	2	17		No ROSC in field
Young et al. (2004)	OH	599		58	29	9	8.6	31	Witnessed arrest; PEA or VF as first rhythm	>3 doses epinephrine; ROSC > 30 min; first rhythm asystole
^aMean or median age, depending on the study. ^bGood outcome among survivors, defined as good outcome, mild disability, or unchanged from baseline at hospital discharge. ^cROSC was a study inclusion criteria. OH, out-of-hospital; IH, in-hospital; ROSC, return of spontaneous circulation; VT/VF, ventricular tachycardia/fibrillation; D/C, discharge; ECMO, extracorporeal membrane oxygenation; CPR, cardiopulmonary resuscitation; ED, emergency department; PEA, pulseless electrical activity.

TABLE 65.2 COMMON ETIOLOGIES OF OUT-OF-HOSPITAL CARDIAC ARREST IN INFANTS AND CHILDREN, ADAPTED FROM YOUNG ET AL. (2004) (7)

▪ ETIOLOGY	▪ FREQUENCY (%)	▪ SURVIVAL TO HOSPITAL DISCHARGE (%)
Sudden infant death syndrome	23	0
Trauma	20	5
Respiratory	16	21
Submersion	12	17
Cardiac	8	8
Central nervous system	6	3
Burn	1	0
Poisoning	1	17
Other	10	10
Unknown	3	6

Sudden cardiac arrest in children is rare (between 0.8 and 6.2 per 100,000 per year) but it is an important cause in the outpatient setting, especially in athletes (14). Hypertrophic cardiomyopathy accounts for approximately one-half of sudden cardiac deaths in children, and most occurred in previously well children (15). Other predisposing conditions include additional structural cardiac anomalies, electrical abnormalities in structurally normal hearts, and ingestion of medications or illicit substances. Most of these conditions are associated with ventricular tachyarrhythmias.

FIGURE 65.2. Typical patterns of cerebral blood flow (CBF) after cardiac arrest. There are differences between CBF after asphyxial arrest, more predominant in children, and cardiogenic arrest, more prominent in adults. Four typical phases of CBF have been described: no-flow (Phase I), hyperemia (Phase (II), hypoperfusion (Phase III), and recovery (Phase IV). ROSC, return of spontaneous circulation; VF, ventricular fibrillation. From: Robert S. B. Clark.

Outcome

Mortality after cardiac arrest in children is very high, and despite the potential for plasticity in the developing brain, neurologic

outcomes can be dismal. These poor outcomes may be related to the prearrest hypoxemia and brain hypoperfusion prior to no-flow ischemia that is seen during asphyxial cardiac arrest (Fig. 65.2). Over half of the children who survive cardiac arrest develop some degree of HIE. A large meta-analysis showed that ROSC from all causes of cardiac arrest was achieved in 13% of pediatric patients. In this study, location of the cardiac arrest had a large impact on survival (24% for children with in- hospital cardiac arrest and 9% for out-of-hospital cardiac arrest) (7). Similarly, in a multicenter, prospective, observational study of children, the in- and out-of-hospital cardiac arrest survival was 49% and 38%, respectively (6). The leading cause of death was neurologic in 69% of the out-of-hospital arrest cohort and cardiovascular in the in-hospital arrest cohort (Table 65.1).

As mentioned earlier, deaths after successful ROSC are most often due to brain death, multi-organ system failure, or withdrawal of support (usually based on projected neurologic outcome). Patients that survive often develop HIE and can have single or multiple neurologic disabilities of varying severity that evolve after the insult. Common manifestations of HIE include cerebral palsy, mental retardation, dysphagia, cortical vision impairment or blindness, hearing impairment or deafness, microcephaly, temperature instability, chronic ventilator dependence, and seizures. Infants after hypoxic-ischemic injury may not demonstrate the clinical neurologic impairments
of HIE until later when developmental milestones are not met and should have long-term developmental screening. Although outcome after cardiac arrest is poor for any age, a Japanese study of out-of-hospital cardiac arrest found that adolescents, who were more likely to have a witnessed arrest and shockable rhythm, had a superior 1-month survival rate (11%) compared with younger children (1%-2%) (16).

Of the children who attain ROSC, prediction of survival and neurologic outcome after cardiac arrest is difficult at best. Precise data do not exist to aid in the delivery of timely and reliable information for clinicians and families involved in patient care or end-of-life decisions. The best scoring system, and timing of its performance, to definitively determine a child’s outcome after cardiac arrest is not known. The Pediatric Overall Performance Category (POPC) score and Pediatric Cerebral Performance Category (PCPC) score can be used to quantify outcome in patients after brain injury in a simplified fashion (17). It is defined as follows: 1, normal; 2, mild; 3, moderate disability; 4, severe disability; 5, coma or vegetative state; and 6, dead. The POPC and PCPC provide a gross estimate of the degree of disability in patients after brain injury, and although they are used to assess therapeutic efficacy in clinical trials, they are not designed to discern specific neurologic diagnoses.

Data are available from large studies on the ability of clinical, radiologic, and/or other testing to predict poor neurologic outcome (Table 65.1). Survival rate to hospital discharge for children after out-of-hospital cardiac arrest is 9%, with twothirds of survivors having neurologic impairment (7). Over half of patients are <1 year of age, with half of these being newly born and having the best survival rate (36%). Males comprise 50%-60% of pediatric cardiac arrest patients, and survival rates are equal between genders. The need for three or more doses of epinephrine or >30 minutes of resuscitation is associated with poor neurologic outcome in survivors (7). Patients with witnessed cardiac arrest have better outcomes than those with unwitnessed cardiac arrest, but the use of prehospital cardiopulmonary resuscitation (CPR) may not make a difference in overall survival rate. Based on data from the US National Registry of CPR (now called Get with the Guidelines), survival rate to hospital discharge was higher for in-hospital cardiac arrest patients (27%) versus out-of-hospital arrest patients (8%), with 65% of those survivors having good neurologic outcome. In that study, the mean time to initiation of CPR was <1 minute and was provided by trained healthcare personnel (12).

In a study of patients having cardiac arrests within a PICU, a hospital discharge rate of 13.7% was reported. For CPR durations of <15 minutes, 15-30 minutes, and >30 minutes, the survival rates were 19%, 12%, and 6%, respectively. Only 2 of 35 (5.7%) patients who had a cardiac arrest prior to their PICU cardiac arrest survived to discharge. Severity of illness, as measured by the Pediatric Risk of Mortality III score, was found to be a significant predictor of survival (18). Another study of cardiac arrest in PICU patients had a similar hospital discharge rate previously reported for inpatient pediatric cardiac arrest, and found that preexisting renal failure and requirement for epinephrine predicted poor outcome (19). One pediatric cardiac ICU reported successful CPR in 63% (24/38) of patients, with 20 having ROSC and 4 placed on mechanical cardiopulmonary support. All 4 of the mechanically supported patients and 10 ROSC patients survived to hospital discharge (37%) and 11 remained alive at 6 months, 3 with neurologic impairment (20). Using a large multicenter registry of in-hospital cardiac arrest, children with surgicalcardiac disease were found to have the best survival to discharge after cardiac arrest (37%), followed by those with medical cardiac disease (28%) and noncardiac disease (23%) (21).

A case series using extracorporeal support in the context of CPR (ECPR) showed that its implementation during CPR has utility for in-hospital cardiac arrest with surprisingly positive outcomes. Although patients had prolonged resuscitation times, good neurologic outcome was still possible, and survival was better in patients with isolated cardiac disease (22). A larger, multicenter, cohort of children provided with ECPR after in-hospital cardiac arrest showed that 44% survived to hospital discharge and 99% of the survivors had favorable outcome (23). Having a cardiac etiology was associated with improved chance of survival.

Observational studies address the ability of clinical examination, laboratory findings, brain electrical activity, or radiologic tests to predict outcome. Absent pupillary and motor response had good predictive value when measured several days after ROSC (24). Children with an EEG that exhibited discontinuous activity and/or was nonreactive had worse outcome in the first 72 hours after ROSC (25). In children who drowned, presence of any hypoxic-ischemic abnormality on brain computed tomography (CT) at any time was associated with death or persistent vegetative state (26). Serum biomarkers of brain injury such as neuron specific enolase (NSE) from neurons and S100b from astrocytes show utility in preliminary studies(26a,101). Children with basal ganglia lesions on conventional brain magnetic resonance imaging (MRI) or in the brain lobes on diffusion-weighted MRI had poor outcome (27). Additional large studies with prospective data are needed to address the prognostic utility of clinical tests and markers in infants and children. The use of hypothermia has confounded much of the outcome prediction data gleaned from adults with cardiac arrest; however, somatosensory-evoked potentials (SSEPs) have demonstrated good reliability (28).

CBF after Cardiac Arrest in Experimental Models

Age-related changes in CBF during normal development and key facets of CBF regulation are presented in Chapter 55 and serve as important background information for the sections that follow. The temporal pattern of CBF after cardiac arrest has been ascertained predominately from experimental models using adult animals and VT/VF cardiac arrest, or global cerebral ischemia paradigms such as aortic or carotid occlusion. Recent studies describing CBF after asphyxial cardiac arrest in models of “pediatric”-relevant ages reveal a distinct pattern of cerebral reperfusion. These differences may be due to distinct pathophysiologic processes in asphyxial versus VF cardiac arrest and also due to dissimilarities in hemodynamic response in the early stages of postnatal brain development compared with the developed brain (29). Asphyxial cardiac arrest produces a physiologically different milieu compared with VT/VF arrest, because of the hypoxemia, acidosis, hypercarbia, and hypotension that precede cardiac arrest (Fig. 65.2). It has been reported that an asphyxial arrest of 7 minutes’ duration produces severe impairment of neurologic function, comparable to a VT/VF arrest of 10 minutes’ duration (30). Furthermore, asphyxial cardiac arrest, in comparison with VT/VF cardiac arrest of similar duration, in dogs resulted in more prevalent and prominent microinfarcts and petechial hemorrhage (31). CBF during and after cardiac arrest is phasic in nature. During cardiac arrest, there is global ischemia, with no- or low-flow if CPR is being provided. Immediately following reperfusion, there can be heterogenous return of CBF in spite of normal or increased cerebral perfusion pressure (CPP, the difference between mean blood pressure and mean intracranial pressure), typically correlating with the duration of ischemia. Studies in adult animal models of cardiac arrest demonstrate that CBF in the postarrest period can be divided into four phases (Fig. 65.2): multifocal (regional) no-reflow (Phase I), hyperemia (Phase II), delayed hypoperfusion (Phase III), and restitution of normal

blood flow (Phase IV) (32,33,34). Studies in pediatric asphyxial cardiac arrest demonstrate that CBF varies for different brain regions within each phase. In the setting of pediatric asphyxial cardiac arrest, progressively increased insult durations diminish the intensity of the hyperemia phase in all regions, with hypoperfusion predominating at longer insult durations (29).

The existence of a multifocal no-reflow phase is somewhat controversial, initially described by Ames et al. (32) in 1968, in a global cerebral ischemia model. Lin (35) and Fischer and Hossmann (33) demonstrated that the hetereogenous brain reperfusion, or the “no-reflow” phenomenon, can also occurr after VF cardiac arrest. This phase is characterized by regions of the brain that fail to reperfuse after ROSC. These areas are more extensive with increasing duration of cardiac arrest, and are observed at a macroscopic and microscopic level. At a microscopic level, there are areas of “no reflow” interspersed with areas of restored blood flow and microinfarcts. 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 CBF after cardiac arrest is characterized by increased global CBF immediately after ROSC and is often referred to as the “hyperemia” phase. This phase is present for ˜15-30 minutes after ROSC. Global CBF during this phase is typically two to three times higher than baseline CBF. Kagstrom et al. (34) demonstrated that during the hyperemia phase, local CBF was extremely heterogeneous, with some areas of no-flow present. Some consider that this hyperemic phase is essential for neuronal functional recovery (36). This opinion is based on studies showing that hypertensive reperfusion improves neurologic outcome, whereas postarrest hypotension worsens neurologic outcome (36,37,38). In addition to hypertensive reperfusion, hemodilution and thrombolysis in this phase have been shown to improve outcome in animal models (39,40). Following resuscitation from asphyxial cardiac arrest in young rats, this phase is characterized by marked region-dependent and insult duration-dependent CBF alterations. Differing degrees of hypoperfusion are present in the cortex immediately after resuscitation, and the hypoperfusion increases proportionally with the duration of asphyxia. Hyperemia is present in the hippocampus, amygdala, and thalamus for 5-30 minutes after moderate durations of asphyxial cardiac arrest, and it is most pronounced in the thalamus, where CBF is two times higher than baseline CBF. Importantly, regional hyperemia is absent after prolonged asphyxial cardiac arrests, with CBF values similar to baseline values (29).

The third phase of CBF after cardiac arrest begins ˜15-30 minutes after ROSC, can persist for several hours, and is often referred to as the “delayed hypoperfusion” phase. During this phase, there is regional heterogeneity of the CBF, with areas of high, normal, and low flow (32,33

Only gold members can continue reading. Log In or Register to continue