tioning for extended periods often leads to
prolonged, expensive, and futile procedures ac-
companied by great emotional strain on family
and medical staff. Conversely, the recupera-
tive powers of the brain sometimes can seem
331
astounding to the uninitiated, and individual
patients whom uninformed physicians might
give up for hopelessly brain damaged or dead
sometimes make unexpectedly good recoveries
(see pitfalls, page 338). It is even more impor-
tant to know when to fight for life than to be
willing to diagnose death. (3) Critical care facili-
ties are limited and expensive and inevitably
place a drain on other medical resources. Their
best use demands that one identify and select
patients who are most likely to benefit from
intensive techniques, so that these units are not
overloaded with individuals who can never re-
cover cerebral function.
The cornerstone of the diagnosis of brain
death remains a careful and sure clinical neu-
rologic examination (Table 8–2). In addition, a
thorough evaluation of clinical history, neu-
roradiologic studies, and laboratory tests must
be done to rule out potential confounding vari-
ables. The diagnosis of brain death rests on two
major and indispensable tenets. The first is that
the cause of brain nonfunction must be inher-
ently irreversible. This means that damage
must be due to either known structural injury
(e.g., cerebral hemorrhage or infarction, brain
trauma, abscess) or known irreversible meta-
bolic injury such as prolonged asphyxia. The
second indispensable tenet is that the vital
structures of the brain necessary to maintain
consciousness and independent vegetative sur-
vival are damaged beyond all possible recovery.
The cause of brain damage must be known
irreversible structural or metabolic disease.
This first criterion is crucial, and the diagnosis
of brain death cannot be considered until it is
fulfilled. The reason for stressing this point is
that both in the United States and abroad often
‘‘coma of unknown origin’’ arising outside of a
hospital is due to depressant drug poisoning.
Witnesses cannot be relied upon for accurate
histories under such circumstances because
efforts at suicide or homicide can readily in-
duce false testimony by companions or family.
Even in patients already in the hospital for the
treatment of other illnesses, drug poisoning
administered by self or others sometimes oc-
curs and at least temporarily can deceive the
medical staff. Accordingly, the diagnosis of an
irreversible lesion by clinical and laboratory
means must be fully documented and unequiv-
ocally accurate before considering a diagnosis
of brain death. The ease of being mistaken in
such a diagnosis is illustrated by some of the
results of a collaborative study sponsored sev-
eral years ago by the National Institutes of
Table 8–1 Harvard Criteria for Brain
Death (1968)
1. Unresponsive coma
2. Apnea
3. Absence of cephalic reflexes
4. Absence of spinal reflexes
5. Isoelectric electroencephalogram
6. Persistence of conditions for at least 24 hours
7. Absence of drug intoxication or hypothermia
From Ad Hoc Committee of the Harvard Medical School.
2
Table 8–2 Clinical Criteria for Brain
Death in Adults and Children in the
United States
A. Coma of established cause
1. No potentially anesthetizing amounts of
either toxins or therapeutic drugs can be
present; hypothermia below 308C or other
physiologic abnormalities must be corrected
to the extent medically possible.
2. Irreversible structural disease or a known
and irreversible endogenous metabolic cause
due to organ failure must be present.
B. Absence of motor responses
1. Absence of pupillary responses to light and
pupils at midposition with respect to dilation
(4–6 mm)
2. Absence of corneal reflexes
3. Absence of caloric vestibulo-ocular responses
4. Absence of gag reflex
5. Absence of coughing in response to tracheal
suctioning
6. Absence of sucking and rooting reflexes
7. Absence of respiratory drive at a PaCO
2
that
is 60 mm Hg or 20 mm Hg above normal
baseline values (apnea testing)
C. Interval between two evaluations, by
patient’s age
1. Term to 2 months old, 48 hours
2. >2 months to 1 year old, 24 hours
3. >1 year to <18 years old, 12 hour
4. !18 years old, optional
D. Confirmatory tests
1. Term to 2 months old, two confirmatory tests
2. >2 months to 1 year old, one confirmatory
test
3. >1 year to <18 years old, optional
4. !18 years old, optional
332
Plum and Posner’s Diagnosis of Stupor and Coma
Health.
5
The findings of toxicologic analyses
revealed many more cases in which drug poi-
soning caused deep coma than had been sus-
pected clinically by physicians, not all of whom
had previous experiences with the ubiquity and
subtlety of sedative-induced coma. The most
common underlying causes of brain death are
listed in Table 8–3. Documentation of struc-
tural injury explaining loss of brainstem func-
tion by computed tomography (CT) or mag-
netic resonance imaging (MRI) is possible in
almost all patients. If scans are normal and
clinical history is equivocal for the origin of ce-
rebral demise, an examination of the cerebral
spinal fluid is indicated.
A prospective study
7
evaluated 310 patients
with cardiac arrest or other forms of acute
medical coma who met the clinical criteria of
brain death for 6 hours; none improved de-
spite maximal treatment. Asystole occurred in
all within a matter of hours or days. Jorgenson
and Malchow-Moller
8
systematically examined
the time required for recovery of neurologic
functions in 54 patients following cardiopul-
monary arrest, and plotted these times against
eventual outcomes. For respiratory move-
ments, pupillary light reflexes, coughing,
swallowing, and ciliospinal reflexes, the longest
respective times of reappearance compatible
with any cerebral recovery were 15, 28, 58, and
52 minutes. In other words, if no recognizable
brain function returned within an hour, the
brain never recovered.
Time periods for repeated evaluations of
brain death criteria may vary and are influ-
enced by the etiology of injury. Several guide-
lines suggest a minimum time period of 24
hours over which human subjects must show
signs of brain death following anoxic injury (or
other diffuse toxic-metabolic insult, e.g. air, fat
embolism, endocrine derangement) before the
final diagnosis can be reached.
9
Evaluation
times for identified structural injuries of the
brainstem are typically shorter. Since time is
so strong a safeguard, and few brain-damaged
patients escape receiving at least an initial dose
of a drug (alcohol or sedative outside of hospi-
tal, sedatives or anticonvulsants inside), guide-
lines suggest a 6-hour period of observation be-
fore making a clinical diagnosis of brain death
(https://www.aan.com/professionals/practice/
guidelines/pda/Brain_death_adults.pdf). This
seems a reasonable time interval for cases where
all circumstances of onset, diagnosis, and treat-
ment can be fully identified.
CLINICAL SIGNS
OF BRAIN DEATH
All observers agree that in order to conclude
that the vital functions of the brain have ceased,
no behavioral or clinical reflex responses that
depend on structures innervated from the su-
praspinal nervous system can exist. In a practical
sense, because forebrain function depends on
the integrity of the brainstem, the brain death
examination primarily focuses on functional
brainstem activity (Table 8–2). These obser-
vations may be accompanied by confirmatory
tests providing evidence of absence of cerebral
hemispheric and upper brainstem function,
discussed below.
Brainstem Function
PUPILS
The pupils must be nonreactive to light. In the
period immediately following brain death, the
agonal release of adrenal catecholamines into
the bloodstream may cause the pupils to be-
come dilated. However, as the catecholamines
are metabolized, the pupils return to a midpo-
sition. Hence, although the Harvard criteria re-
quired that the pupils be dilated as well as fixed,
midposition fixed pupils are a more reliable
sign of brain death, and failure of the pupils to
return to midposition within several hours af-
ter brain death suggests residual sympathetic
activation arising from the medulla. The pupils
should be tested with a bright light and the
physician should be certain that mydriatic
Table 8–3 Most Common Etiologies
of Brain Death
1. Traumatic brain injury
2. Aneurysmal subarachnoid hemorrhage
3. Intracerebral hemorrhage
4. Ischemic stroke with cerebral edema
and herniation
5. Hypoxic-ischemic encephalopathy
6. Fulminant hepatic necrosis with cerebral edema
and increased intracranial pressure
From Wijdicks,
6
with permission.
Brain Death
333
agents, including intravenous atropine, have
not been used (although conventional doses
of atropine used in treating patients with car-
diac arrest will not block the direct light re-
sponse). Neuromuscular blocking agents, how-
ever, should not affect pupillary size as nicotinic
receptors are not present in the iris. One
recent report has described an unusual ob-
servation of persistent asynchronous light-
independent pupillary activity (2.5 seconds
constriction/10 seconds dilation) in an other-
wise ‘‘brain-dead’’ patient.
10
OCULAR MOVEMENTS
Failure of brainstem function should be deter-
mined by the inability to find either oculo-
cephalic or caloric vestibulo-ocular responses
(see Chapter 2). In patients in whom a history
of possible trauma has not been eliminated,
cervical spine injury must be excluded before
testing oculocephalic responses. Care should
be taken when performing cold water caloric
testing to ensure that the stimulus reaches the
tympanic membrane. Up to 1 minute of ob-
servation for eye movement should follow ir-
rigation of each side with a 5-minute interval
between each examination.
MOTOR, SENSORY, AND
REFLEX ACTIVITY
The initial Harvard criteria demanded that
there be an absence of all voluntary and reflex
movements, including absence of corneal re-
sponses and other brainstem reflexes; no pos-
tural activity, including decerebrate rigidity;
and no stretch reflexes in the extremities. Re-
flex responses mediated by the brainstem (e.g.,
corneal and jaw jerk reflexes as well as cuta-
neous reflexes such as snout and rooting re-
flexes) must be absent before making the di-
agnosis of brain death. The absence of a gag
reflex should be tested by stimulation of the
posterior pharynx, but may be difficult to elicit
or observe in intubated patients. Additionally,
response to noxious stimulation of the supra-
orbital nerve or temporomandibular joints
11
should be tested during the examination. How-
ever, spinal reflex activity, in response to both
noxious stimuli and tendon stretch, often can
be shown to persist in experimental animals
whose brains have been destroyed above the
spinal level. The same reflexes can be found in
the isolated spinal cord of humans following
high spinal cord transection.
A variety of unusual, spinally mediated move-
ments can appear and persist for prolonged
periods during artificial life support.
12–18
Such
phenomena include spontaneous movements
in synchrony with the mechanical ventilator;
slow body movements producing flexion at the
waist, causing the body to rise to a sitting po-
sition (‘‘Lazarus sign’’); ‘‘stepping movements’’;
and preservation of lower body reflexes.
4
The
consensus view is that in a patient in whom
apneic oxygenation shows no return of breath-
ing, such movements are generated by the spi-
nal cord and the vital functions of the brain-
stem have no chance of recovery, making the
diagnosis of brain death appropriate. It is im-
portant to note that spontaneous hypoxic or
hypotensive events and apnea testing may pre-
cipitate these movements. Surprisingly, exten-
sor plantar responses are not found in brain-
dead patients.
14
Instead, plantar responses are
either flexor, absent, or consistent with undu-
lations of toe flexion.
19
APNEA
Spontaneous respiration must be absent. Most
patients on a mechanical ventilator will have a
PaO
2
above and a PaCO
2
below normal levels.
However, the threshold for stimulation of re-
spiratory movements by the blood gases usu-
ally is elevated in patients in deep coma, some-
times to PaCO
2
values as high as 50 to 55 mm
Hg. As a result, such patients may be apneic
for several minutes when removed from the
ventilator, even if they have a structurally nor-
mal brainstem. To test brainstem function
without concurrently inducing severe hypox-
emia under such circumstances, respiratory
activity should be tested by the technique of
apneic oxygenation. With this technique, the
patient is ventilated with 100% oxygen for a
period of 10 to 20 minutes. The respirator is
then disconnected to avoid false readings and
oxygen is delivered through a catheter to the
trachea at a rate of about 6 L/minute. The
resulting tension of oxygen in the alveoli will
remain high enough to maintain the arterial
blood at adequate oxygen tensions for as long
as an hour or more. The PaCO
2
rises by about
3 mm Hg/minute during apneic oxygenation in
a deeply comatose or clinically brain-dead pa-
tient.
20
Apneic oxygenation of 8 to 10 minutes
334
Plum and Posner’s Diagnosis of Stupor and Coma
thus allows the PaCO
2
to rise without danger
of further hypoxia and ensures that one ex-
ceeds the respiratory threshold. A PaCO
2
that
rises above 60 mm Hg without concomitant
breathing efforts provides unequivocal evi-
dence of nonfunctioning respiratory centers.
The American Academy of Neurology guide-
lines for brain death (https://www.aan.com/
professionals/practice/guidelines/pda/Brain_
death_adults.pdf) accept either a PaCO
2
of
60 mm Hg or a value 20 mm Hg higher than
baseline as the threshold for maximum stim-
ulation of the respiratory centers of the me-
dulla oblongata. Chronic pulmonary disease
producing baseline hypercapnia may compli-
cate the apnea testing and can be identified in
initial blood gas examination by elevated se-
rum bicarbonate concentration. In such cases,
ancillary testing is recommended by current
guidelines. Alternatively, hypocapnia often
arises in the setting of hyperventilation to man-
age increased intracranial pressure (ICP). Since
it is important to start the examination near a
target PCO
2
of 40 mm Hg, hypocapnia should
be corrected by adjusting the minute volume
of ventilation through either a reduction of the
tidal volume or a resetting of the respiratory
rate.
During testing the patient should be ob-
served for respiration defined as abdominal or
chest excursions.
6
If respiration occurs during
apnea testing, it is usually early into the testing.
After 8 minutes have elapsed, arterial blood
gases should be sampled and the ventilator
reconnected. The absence of respiratory move-
ments and rise of PCO
2
past 60 mm Hg indi-
cates a positive apnea test. Alternatively, if
respiratory movements are seen, the test is
negative and retesting at a later time is indi-
cated. Prior to initiating apnea testing the ab-
sence of brainstem reflexes should have al-
ready been established. Additionally, several
other prerequisites must be established, as in-
dicated in Table 8–4. Hypothermia must be
excluded; if core temperatures obtained by
rectal measurement are below 36.58C, the
patient should be warmed with a blanket. A
systolic blood pressure of greater than 90 mm
Hg should be maintained using dopamine
infusion if required. If hypotension (systolic
blood pressure less than 90 mm Hg) arises dur-
ing the examination, blood samples should be
promptly drawn and the ventilator immedi-
ately reconnected. Conversely, any elevation
of blood pressure during testing is evidence of
lower brainstem function. As diabetes insipidus
is a common complication of severe brain in-
juries, this should be recognized if present and
managed. Accordingly, efforts should ensure
euvolemia or positive fluid balance for at least
6 hours prior to testing. Finally, arterial gas
pressures should reflect PO
2
greater than
200 mm Hg and PCO
2
greater than or equal to
40 mm Hg prior to testing as discussed above.
Confirmatory Laboratory Tests
and Diagnosis
When the clinical examination is unequivocal,
no additional tests are required. However, if
there is any question, clinical practice guide-
lines suggest the potential use of four modalities
of confirmatory testing in the determination of
brain death
4,6
: conventional angiography, elec-
troencephalography (EEG)/evoked potential
(EP) studies, transcranial Doppler sonography
(TCD), and cerebral scintigraphy. Consensus
criteria for brain death determination are only
available for EEG/EP studies.
STUDIES TO ESTABLISH CESSATION
OF CEREBRAL BLOOD FLOW:
CEREBRAL ANGIOGRAPHY,
TRANSCRANIAL DOPPLER
SONOGRAPHY, AND CEREBRAL
SCINTIGRAPHY
Cerebral angiography is a widely accepted
procedure for determination of brain death and
can be used to overcome the limitation of the
clinical neurologic examination due to facial
trauma, baseline pulmonary disease, and other
Table 8–4 Prerequisites for Apnea
Testing
1. Core temperature >36.58C or 978F
2. Systolic blood pressure >90 mm Hg
3. Euvolemia. Option: positive fluid balance in the
previous 6 hours
4. Normal PCO
2.
Option: arterial PCO
2
>
40 mm Hg
5. Normal PO
2
. Option: preoxygenation to
obtain arterial PO
2
>
200 mm Hg
Adapted from Wijdicks.
6
Brain Death
335
confounding factors. This procedure also has
the advantage, when positive, of establishing a
structural cause of brain death (i.e., absence of
blood flow to the brain). In cases where the
original cause of cerebral injury is not known,
the absence of blood flow provides the crucial
information necessary to declare brain death
with certainty.
Physiologically, two events may produce fail-
ure of the cerebral circulation. First, a sudden
and massive increase in ICP (e.g., during sub-
arachnoid hemorrhage) may cause it to rise to
the level of arterial perfusion pressure at which
point cerebral circulation ceases. The second,
and probably more common, occurrence is a
progressive loss of blood flow that accompanies
death of the brain. As the dead tissue becomes
edematous, the local tissue pressure exceeds
capillary perfusion pressure, resulting in stasis
of blood flow, further edema, and further vas-
cular stasis. If the respiratory and cardiovas-
cular systems are kept functioning for many
hours or days after brain circulation has ceased,
the brain undergoes autolysis at body temper-
ature, resulting in a soft and necrotic organ at
autopsy referred to by pathologists as a ‘‘res-
pirator brain.’’
21
Demonstration of the failure of intracerebral
filling at the level of entry of the carotid and
vertebral arteries indicates brain death. Re-
cently, magnetic resonance angiography (MRA)
has been reported for diagnosis of brain death,
but this technique is less reliable, as MRA often
fails to demonstrate slow flow. Additional MRI
criteria for brain death include loss of the sub-
arachnoid spaces, slow flow in the intracaver-
nous and cervical internal carotid arteries, loss
of flow void in both small and large intracra-
nial arteries and venous sinuses, and loss of
gray-white matter distinction on T1-weighted
images, but ‘‘supranormal’’ distinction on T2-
weighted images.
22
However, until additional
data are available on the reliability of these in-
dicators for determining brain death, the pres-
ence of complete cessation of brain function
on examination, or complete loss of blood flow,
must remain the gold standards for diagnosis.
In addition, the rarity of ventilators that are
compatible with MR scanners continues to limit
the availability of this mode of diagnosis.
Bilateral insonation of the intracranial ar-
teries using a portable 2-MHz pulsed Doppler
device (transcranial Doppler ultrasonograph
[TCD]) is now a widely used confirmatory test
for brain death.
23
The middle cerebral arteries 18>18> Dostları ilə paylaş: |