Plum and posner’s diagnosis of stupor and coma fourth Edition series editor sid Gilman, md, frcp
Yüklə 9,02 Mb. Pdf görüntüsü
|
|
depressed to essentially functionless levels by anesthetic depressant drugs, yet lose none of its capacity for total recovery when the anesthetic disappears. Several investigators have demon- strated that experimental animals and humans can be resuscitated to full functional activity after periods of deep anesthesia producing hours to days of isoelectric EEG flattening. 101 This tolerability is used clinically in cases of status epilepticus to prevent continuous seizure activity from damaging the brain. A corollary is that in cases of coma due to sedative overdose, the depth and duration of coma are not indic- ative of the potential for recovery of function. In animal experiments, general anesthesia, either before or within a few hours of an is- chemic insult to the brain, protects against brain damage when measured a few days af- ter the insult. However, at 3 weeks there is no difference in the degree of neuronal damage between the anesthetized animals and those treated without anesthesia, 102,103
indicating no protection against the delayed effects of anoxia (see page 219). Clinical experience with barbiturate anes- thesia and drug poisoning indicates that given good medical care, most patients usually survive Multifocal, Diffuse, and Metabolic Brain Diseases Causing Delirium, Stupor, or Coma 205
anesthesia, even profound suppression of neu- ral activity resulting from self-administered bar- biturates or other sedative drugs. Even when coma is so deep that artificial respiration must be provided for several days and the blood pressure supported by vasopressor agents for a week or more, patients can awaken with no apparent or measurable impairment of brain function. Hence, it is critical to determine the presence of sedative overdose when evaluating the prognosis of a patient in coma, even those with other causes of coma. The complete reversibility of anesthetic coma, plus the low metabolic rate that accom- panies deep anesthesia, has inspired efforts to determine whether barbiturate anesthesia can minimize the expected extent of postanoxic is- chemic brain damage. Barbiturates also scav- enge free radicals from reoxygenated tissue, but it remains to be proved that this represents an important biologic function in resuscitation. On the other hand, phenobarbital also induces cytochrome P450, which serves as a source of reactive oxygen species. Whether these oppo- site effects help, hurt, or have no effect on the brain is unclear. 104,105 Of some interest, a ran- domized trial of neonates with hypoxic- ischemic encephalopathy indicates that phe- nobarbital in a dose of 20 mg/kg given intra- venously within 6 hours of birth in term and near-term neonates was associated with a de- crease in lipid peroxides. There was also a de- crease in antioxidant enzymes and antioxidant vitamins in the CSF. A trend suggested that lower levels of lipid peroxides in the CSF were associated with a better outcome. 106 Barbitu-
rate coma is effective in controlling intractable status epilepticus, but its role in any other brain injury is, at this writing, uncertain. 101
Barbitu- rate anesthesia has been applied to patients in coma from head trauma. It lowers ICP, but it is unclear if it affects outcome. 107 MECHANISMS OF IRREVERSIBLE ANOXIC-ISCHEMIC BRAIN DAMAGE
Anoxia, ischemia, and hypoglycemia, although biologically different, 108 can combine under several circumstances to damage the brain. Somewhat different but overlapping patho- logic changes characterize the irreversible brain injury caused by each of these three conditions. Systemic and local circulatory dif- ferences among them influence the exact ge- ography and type of cellular response. Similar changes in the brain mark the postmortem findings of several conditions, including pa- tients dying in coma after fatal status epi- lepticus, carbon monoxide poisoning, or several of the systemic metabolic encephalopathies. Global Ischemia Complete cerebral ischemia, as in cardiac ar- rest in man, causes loss of consciousness in less than 20 seconds. Within 5 minutes, glucose and high-energy phosphate stores are depleted. Following that the patient, even if successfully resuscitated, may be left severely brain dam- aged. This is especially true in elderly pa- tients who most frequently suffer cardiac ar- rest because their brains are more vulnerable to ischemic damage. By definition, during car- diac arrest the CBF falls to zero. Resuscita- tion results in transient hyperemia with in- creased blood flow and oxygen metabolism; subsequently, both decrease in a heteroge- neous fashion. 109 In most patients, when blood flow is re-established, cerebral autoregulation is either absent or the curve is shifted to the right, such that CBF begins to fall at a higher mean arterial pressure than it did before the cardiac arrest. As a result, it is important to maintain normal and perhaps slightly elevated blood pressure after cardiac arrest. Both vascular and neuronal factors play a role in the seemingly brief periods of global ischemia that can damage the brain in clinical circum- stances. Changes to vascular endothelium dur- ing the course of ischemia, as well as additional changes to glial cells (swelling to compress en- dothelial vessels, viscosity changes in blood), may lead to poor perfusion once cardiac func- tion is restored. This so-called ‘‘no-reflow phe- nomenon’’ 110
increases with prolonged duration of ischemia. 110–112 Loss of autoregulation can aggravate edema formation, lead to hemor- rhage, and cause additional neuronal damage, so-called ‘‘reperfusion injury.’’ 113
The combi- nation of the ischemia and its aftermath results in neuronal necrosis, 114
particularly in the hip- pocampus, but if the ischemia is prolonged, elsewhere in the hemispheres as well. Although the exact mechanisms are not un- derstood, it is likely that during the ischemia 206
Plum and Posner’s Diagnosis of Stupor and Coma the loss of high-energy phosphates causes cel- lular depolarization that induces the release of glutamate, which in turn causes entry of toxic levels of calcium into neurons. In the reper- fusion phase, the restoration of oxidative me- tabolism probably produces a burst of excess free radicals that are also cytotoxic. 113
Cardiac arrest can either cause death of neurons, particularly in vulnerable areas asso- ciated with reactive astrocytes, or microinfarcts and areas of pancellular necrosis associated with perivascular diffuse tissue spongiosis. The lat- ter lesions appear in a laminar distribution and are more profound in watershed zones between the major territories of arterial supply. Both types of lesions are more intense and hetero- geneous in patients dying after a period of pro- longed coma. 115
Particularly vulnerable areas include the oc- cipital cortex, the frontoparietal cortex, the hippocampus, the basal ganglia, the thalamic reticular nucleus, Purkinje cells of the cerebel- lum, and the spinal cord (Figure 5–5). Laminar necrosis of the cortex generally involves layers III and V, which contain the greatest numbers of large pyramidal cells. The most vulnerable area is the CA1 region of the hippocampus. Some patients with lesions restricted to the CA1 region who recover from cardiac arrest can develop a residual severe anterograde am- nesia (see Patient 5–6). 116
Focal Ischemia Focal ischemia differs from global ischemia in that it allows for collateral circulation to deliver at least some blood to the areas surrounding the area of no perfusion induced by the vas- cular occlusion. The surrounding area, called the penumbra, 117
suffers low flow but not cel- lular death. It is the goal of the physician treat- ing the patient to try to preserve that area and return its metabolism to normal. Like global ischemia, damage can occur either during the ischemic period or during reperfusion. 118 Schaller and Graf 118 have diagramed a three- peaked curve presenting times at which the penumbra is susceptible to tissue damage. The first occurs during ischemia with damage re- sulting from oxygen depletion, energy failure, depolarization of neurons and synapses, and homeostasis failure. The second occurs after reperfusion with damage caused by excitotox- icity as well as disturbed homeostasis. The third occurs several weeks later with late damage to neurons and glial cells via both necrosis and apoptosis. As indicated above, interventions that appear to ameliorate the first two peaks, such as the use of anesthetic agents at the time of ischemia, do not appear to have any effect on the delayed necrosis. 103
Focal ischemia also differs from global is- chemia in its therapeutic window. The phy- sician has minutes to restore circulation in a patient with cardiac arrest before irreversible brain damage with a significant neurologic def- icit occurs. With focal ischemia there is, by def- inition, collateral blood flow to the surrounding tissue and often an area of partial ischemia, the penumbra that surrounds the area of most in- tense ischemia. The tissue constituting the penumbra may have blood flow below the level at which it functions normally, but yet not so low as to cause immediate infarction. Hence, there is often a window of time that may persist Figure 5–5. Computed tomography scan of a comatose patient after prolonged cardiopulmonary resuscitation. The scan was taken at a time when the patient was deeply co- matose but breathing. The brain is swollen, with cortical sulci virtually obliterated. No signal differentiation can be seen between gray and white matter. The lentiform nuclei are hypointense, suggesting basal ganglia infarction. (Scan courtesy Dr. Sasan Karimi.) Multifocal, Diffuse, and Metabolic Brain Diseases Causing Delirium, Stupor, or Coma 207
for several hours, during which the tissue in the penumbra can be saved; in some cases this will reduce the area of what would otherwise be infarcted tissue to the point where there may be little or no neurologic deficit. The exact time window undoubtedly varies depending on the individual vascular anatomy and the nature of the vascular obstruction, but evidence from trials of thrombolytic therapy indicates that it often persists for as long as 3 hours. The time window may, in fact, be longer, but by 3 hours the risk of a hemorrhage into the infarcted tis- sue becomes greater than the benefit from sal- vaging partially ischemic tissue. 118 Hyperglycemia during reperfusion increases infarct volume and may cause hemorrhage. It also reduces the CBF. The mechanism for this effect is not clear, but it could result from dam- age to endothelium, increased expression of ad- hesion molecules, or glycosylation of critical proteins that lead to vasodilation. Patient 5–6 A 44-year-old woman was found unconscious in her room when her husband returned home. He called for paramedics and she was transported to the hospital, where a diagnosis of carbon monoxide poisoning was made. She had a brief period of cardiac arrest from which she was resuscitated. She remained first unconscious and then poorly re- sponsive for about 10 days before recovering. When she recovered she appeared a little euphoric but was able to relate to her husband and family in perfectly logical fashion. She returned home and tried to go back to work as a college teacher of Spanish. Unfortunately, she rapidly discovered that she could not remember where she had parked her car and could not remember what she was to teach that day, although once she was involved in teach- ing, she was able to do relatively well. With careful preparation of lesson plans in advance and ar- rangements for her car to be in the same place and described to her in writing, she was able to con- tinue to function at the community college. Hypoxia Unlike ischemia and hypoglycemia, hypoxia alone is rarely responsible for brain necrosis. 114
In fact, hypoxic preconditioning of experimen- tal animals by exposure to moderate hypoxia of 8% to 10% oxygen for 3 hours protects against cerebral ischemia delivered 1 or 2 days later. 119 Miyamoto and Auer exposed rats to an arterial PO 2 of 25 torr for 15 minutes and failed to find necrotic neurons. 114
Unilateral carotid ligation (ischemia) caused necrosis even in animals ex- posed to an arterial O 2 of 100 torr. In these experiments, hypoxia exacerbated the effects of ischemia. In most situations in humans, hypoxia leads to either hypotension or cardiac arrest so that hypoxic insults are for the most part a mixture of hypoxic and ischemic injury. Pure hypoxia, such as occurs in carbon monoxide poisoning, is more likely to lead to delayed injury to the subcortical structures of the hemispheres. Typically the damage will occur 1 to several days after the patient awak- ens from the hypoxic episode and involves a characteristic distribution, including the pos- terior hemispheric white matter and basal gan- glia, often leaving the patient blind and with a choreic movement disorder. A similar pattern of brain injury is seen with a variety of mito- chondrial encephalopathies and deficits in car- bohydrate metabolism, suggesting that the in- jury is due to failure of oxidative metabolism. The reason that the injury has a predilection for these sites is unknown, although the neu- rons in the globus pallidus have a particularly high constitutive firing rate, and this may pre- dispose them to hypoxic injury. EVALUATION OF NEUROTRANSMITTER CHANGES IN METABOLIC COMA Several neurotransmitters control arousal, sleep-wake cycles, and consciousness. They are probably also involved in metabolic en- cephalopathies and their role, where known, is discussed in the sections below on specific encephalopathies. 120
Acetylcholine The cholinergic system described in Chapter 1 plays an important role in consciousness. 121
The nicotinic alpha-4-beta-2 receptor is in- hibited by clinically relevant doses of volatile 208 Plum and Posner’s Diagnosis of Stupor and Coma anesthetics and ketamine, although whether the inhibition is clinically relevant is not clear. 122
However, anticholinergic agents that cross the blood-brain barrier can cause mem- ory loss and florid delirium, and anticholiner- gic medications are an independent risk factor for delirium in older medical inpatients. 123
Dopamine Dopamine plays a key role in arousal. A wide range of stimulant drugs (amphetamine, meth- ylphenidate, modafinil) are antagonists of the dopamine reuptake pump, and if mice lack this dopamine transporter, the drugs do not have a stimulatory effect. Patients with Parkinson’s disease have increased sleepiness, as do pa- tients treated with dopamine antagonists. Par- adoxically, D2 agonist drugs can also cause sleepiness. The reason for this puzzling re- sponse appears to be due to the fact that the D2 receptor can be either pre- or postsynaptic. Dopamine has its major stimulatory effects via postsynaptic receptors, but the D2 receptor is also found presynaptically on dopamine terminals, where it down-regulates dopamine release. Thus, D2 agonist drugs reduce endog- enous dopamine release. Interestingly, dopa- mine agonists can cause delirium, whereas dopamine blockers are often used to treat de- lirium. Dopamine antagonists also cause EEG slowing. 124
Dopamine release is increased in hypoxia at a time when acetylcholine release is decreased. 120
Gamma-Aminobutyric Acid As indicated above, the GABA A receptor is a major target of many general anesthetic agents. Benzodiazepines, which are GABA A poten-
tiators, can cause memory loss, delirium, 125
and, rarely, coma. 126
Increased concentra- tions of endogenous GABA agonists, both benzodiazepine-like and non-benzodiazepine- like, have been found in patients with hepatic encephalopathy. 127
A variety of GABA B re- ceptor agonists, such as baclofen, are also se- dating. Gamma-hydroxybutyrate (GHB), which has recently been approved for use in narco- lepsy, binds both to GABA B receptors and probably to specific GHB receptors. This drug causes profound impairment of consciousness and high-voltage delta-wave EEG activity. It has achieved a reputation as a ‘‘date rape’’ drug because in lower doses it causes memory loss and sometimes delirium. Serotonin Several investigators have implicated the evo- lutionary very old serotonin in the pathogen- esis of delirium. Both high and low levels of serotonin have been associated with delir- ium.
128–130 Serotonin levels are dependent on the transport of tryptophan, a large neutral ar- omatic amino acid that crosses the blood-brain barrier. Because several other large amino ac- ids, including isoleucine, leucine, methionine, phenylalanine, and tyrosine, use the same sat- urable carrier, they compete with one another. Thus, changes in the amino acid levels in the plasma affect serotonin metabolism in the brain. For example, recent studies suggest that inges- tion of tryptophane-rich alpha-lactalbumin at bedtime improves morning alertness and brain measures of attention in normal individuals. 131 The effect of serotonin withdrawal is a little less clear. Increased tryptophan uptake results in increased brain serotonin activity in patients with hepatic encephalopathy. 120
Histamine Histamine is now known to play a key role in maintaining a waking state. Histamine neurons in the tuberomammillary nucleus in the hypo- thalamus comprise a major component of the ascending arousal system. Inhibition of the his- tamine neurons with a GABA agonist in cats causes sleepiness, and disinhibition with bi- cuculline causes wakefulness and prevents the sedating effects of anesthetics. Animals with knockouts either of the gene for histidine de- carboxylase, which synthesizes histamine, or the H 1
cortex, are more sleepy and do not respond to other arousing neurotransmitters such as orexin. Those H 1 antagonists that are used to treat allergies and also cross the blood-brain barrier cause considerable sleepiness in hu- mans. H 2
tidine, and famotidine, have, on rare occasions, been associated with delirium, particularly in the elderly. 132,133
This response may be due to Multifocal, Diffuse, and Metabolic Brain Diseases Causing Delirium, Stupor, or Coma 209
nonspecific interaction with other histamine receptor subtypes. Glutamate The most common excitatory neurotransmit- ter in the brain, glutamate is used by almost all neurons involved in thalamocortical and long- range corticocortical transmission. Drugs that block NMDA receptors, which are required for memory phenomena such as long-term poten- tiation (LTP), including ketamine, nitrous ox- ide, and phencyclidine, cause intense delirium. However, these drugs do not reduce activity in the arousal system, and may in fact heighten it. As a result, subjects who have had ketamine often report bizarre and distorted experiences, but may be aware even when they appear not to be. Up-regulation of glutamate neurotrans- mission has been associated with alcohol with- drawal delirium (delirium tremens). 134
Norepinephrine Norepinephrine is used by neurons of the locus coeruleus, which also is a major component of the ascending arousal system. Although ablation of the locus coeruleus has minimal effects on consciousness, due to redundant pathways from other monoaminergic systems, its neurons fire in association with novel stimuli in the environ- ment and are most active during wakefulness. Beta blockers can cause depression, but not impairment of consciousness. Alpha blockers mainly impair consciousness when they cause peripheral vasodilation and orthostatic hypo- tension. CSF norepinephrine is elevated during alcohol withdrawal 135
and may be involved in opiate withdrawal as well; treatment with the alpha-2 agonist clonidine can relieve the with- drawal symptoms. Cocaine, which blocks reup- Yüklə 9,02 Mb. Dostları ilə paylaş:
|