which the nose is turned and flexion of the
other arm; extension of the head may cause ex-
tension of the arms and relaxation of the legs,
while flexion of the head leads to the opposite
response) can usually be elicited. As with de-
corticate posturing, fragments of decerebrate
posturing are sometimes seen. These tend to
indicate a lesser degree of injury, but in the
same anatomic distribution as the full pattern.
It may also be asymmetric, indicating the asym-
metry of dysfunction of the brainstem.
Although decerebrate posturing usually is
seen with noxious stimulation, in some patients
it may occur spontaneously, often associated
with waves of shivering and hyperpnea. De-
cerebrate posturing in experimental animals
usually results from a transecting lesion at the
level between the superior and inferior colli-
culi.
140
It is believed to be due to the release of
vestibulospinal postural reflexes from fore-
brain control. The level of brainstem dys-
function that produces this response in humans
may be similar, as in most cases decerebrate
posturing is associated with disturbances of
ocular motility. However, electrophysiologic,
radiologic, or even postmortem examination
sometimes reveals pathology that is largely
confined to the forebrain and diencephalon.
Thus, decerebrate rigidity is a clinical finding
that probably represents dysfunction, although
not necessarily destruction extending into the
upper brainstem. Nevertheless, it represents a
more severe finding than decorticate postur-
ing; for example, in the Jennett and Teasdale
series, only 10% of comatose patients with head
injury who demonstrated decerebrate postur-
ing recovered.
139
Most patients with decere-
brate rigidity have either massive and bilateral
forebrain lesions causing rostrocaudal deteri-
oration of the brainstem as diencephalic dys-
function evolves into midbrain dysfunction (see
Chapter 3), or a posterior fossa lesion that
compresses or damages the midbrain and ros-
tral pons. However, the same pattern may oc-
casionally be seen in patients with diffuse, but
fully reversible, metabolic disorders, such as
hepatic coma, hypoglycemia, or sedative drug
ingestion.
138,141,142
Extensor posturing of the arms with flaccid
or weak flexor responses in the legs is typically
seen in patients with injury to the lower
brainstem, at roughly the level of the vestibular
nuclei. This pattern was described in the 1972
edition of this monograph, and has since been
repeatedly confirmed. The physiologic basis of
this motor pattern is not understood, but it may
represent the transition from the extensor pos-
turing seen with lower midbrain and high pon-
tine injuries to the spinal shock (flaccidity) or
even flexor responses seen from stimulating
the isolated spinal cord.
FALSE LOCALIZING SIGNS
IN PATIENTS WITH
METABOLIC COMA
The main purpose of the foregoing review of
the examination of a comatose patient is to dis-
tinguish patients with structural lesions of the
brain from those with metabolic lesions. Most
patients with structural lesions require urgent
imaging. Patients with metabolic lesions often
require an extensive laboratory evaluation to de-
fine the cause. When focal neurologic findings
are observed, it becomes imperative to deter-
mine whether there is a destructive or compres-
sive process that may become life threatening
or irreversibly damage the brain within a matter
of minutes. On the other hand, even when there
is no focal or lateralizing finding to suggest a
structural lesion, it is important to know which
signs point to specific metabolic causes, such
as hypoglycemia or sepsis, that must be sought
urgently. Therefore, the physician should be-
come familiar with the few focal neurologic
findings that are seen in patients with diffuse
metabolic causes of coma, and understand their
implications for the diagnosis of the metabolic
problem.
Respiratory Responses
The range of normal respiratory responses
includes the Cheyne-Stokes pattern of breath-
ing, which is seen in many cognitively normal
people with cardiac or respiratory disorders,
particularly during sleep.
43–45
Sleep apnea
must also be distinguished from pathologic
breathing patterns. Patients with severe sleep
apnea may stop breathing for 10 seconds or so
every minute or two. Their color may become
dusky during the oxygen desaturation that ac-
companies each period of apnea.
Kussmaul breathing, in which there are
deep but slow rhythmic breaths, is seen in
Examination of the Comatose Patient
75
patients with coma due to an acidotic condi-
tion (e.g., diabetic ketoacidosis or intoxication
with ethylene glycol). The low blood pH drives
the deep respiratory efforts, which reduce the
PCO
2
in the blood, thus producing a com-
pensatory respiratory alkalosis. This must be
distinguished from sepsis, hepatic encephalop-
athy, or cardiac dysfunction, conditions that of-
ten cause a primary respiratory alkalosis, with
compensatory metabolic acidosis.
143–145
The
nature of the primary insult is determined by
whether the blood pH is low (metabolic aci-
dosis with respiratory compensation) or high
(primary respiratory alkalosis).
Pupillary Responses
A key problem with interpreting pupillary re-
sponses is that either metabolic coma or di-
encephalic level dysfunction may cause bilat-
erally small and symmetric, reactive pupils.
Thus, a patient with small pupils and little in
the way of focal neurologic impairment may
still have impairment that can be attributed to
either a diencephalic lesion or to symmetric
forebrain compression (e.g., by bilateral sub-
dural hematomas). As a result, it is generally
necessary to do an imaging study (see below)
within the first few hours in most comatose
patients, even if the cause is believed to be
metabolic.
Very small pupils may be indicative of pon-
tine level dysfunction, often indicating an
acute destructive lesion such as a hemorrhage.
However, similar pinpoint but reactive pupils
may be seen in opiate intoxication. Hence, in
patients who present with pinpoint pupils and
coma, it is necessary to administer an opiate
antagonist such as naloxone to reverse poten-
tial opiate overdose. (Because an opioid antag-
onist can elicit severe withdrawal symptoms
in a physically dependent patient, the drug
should be diluted and delivered slowly, stop-
ping as soon as one notes the pupils to enlarge
and the patient to arouse. See Chapter 7 for
details.)
Unreactive pupils usually indicate structural
disease of the nervous system, but pupils may
become unreactive briefly after a seizure.
When a patient is seen who may have had an
unobserved seizure within the past 30 minutes
or so, it is necessary to re-examine the patient
15 to 30 minutes later to make sure that the
lack of pupillary responses persists. Signs of
major motor seizure, such as tongue biting or
incontinence, or a transient metabolic acidosis
are helpful in alerting the examiner to the
possibility of a recent seizure. In addition, be-
cause the seizure usually results in the release
of adrenalin, the pupils typically are large after
a seizure.
Very deep coma due to sedative intoxication
may suppress all brainstem responses, includ-
ing pupillary light reactions, and simulate brain
death (see Chapter 6). For this reason, it is
critical to do urinary and blood toxic and drug
screening on any patient who is so deeply co-
matose as to lack pupillary responses.
Ocular Motor Responses
Typical oculocephalic responses, as seen in a
comatose patient with an intact brainstem, are
not seen in awake subjects, whose voluntary
eye movements supersede the brainstem ves-
tibular responses. In fact, brainstem oculoce-
phalic responses (as if the eyes were fixed on a
point in the distance) are nearly impossible for
an awake patient to simulate voluntarily, and
therefore are a useful differential point in iden-
tifying psychogenic unresponsiveness. On the
other hand, oculocephalic responses may be-
come particularly brisk in patients with hepatic
coma.
Certain drugs may eliminate oculocephalic
and even caloric vestibulo-ocular responses.
Acute administration of phenytoin quite often
has this effect, which may persist for 6 to 12
hours.
146
Occasionally, patients who have in-
gested an overdose of various tricyclic antide-
pressants may also have absence of vestibulo-
ocular responses.
147
Patients in very deep
metabolic coma, particularly with sedative
drugs, may also eventually lose oculovestibular
responses.
Ophthalmoplegia is also seen in combination
with areflexia and ataxia in the Miller Fisher
variant of Guillain-Barre´ syndrome. While
such patients usually do not have impairment
of consciousness, the Miller Fisher syndrome
occasionally occurs in patients who also have
autoimmune brainstem encephalitis (Bicker-
staff’s encephalitis), with impairment of con-
sciousness, and GQ1b autoantibodies.
148
In
such cases, the relationship of the loss of eye
movements to the impairment of conscious-
76
Plum and Posner’s Diagnosis of Stupor and Coma