is reached. The hyperpneic phase usually lasts
longer than the apneic phase (Figure 2–5).
This rhythmic alternation in Cheyne-Stokes
respiration results from the interplay of normal
brainstem respiratory reflexes.
42–45
When the
medullary chemosensory circuits sense ade-
quate oxygen and carbon dioxide tension, they
reduce the rate and depth of respiration, caus-
ing a gradual rise in arterial carbon dioxide ten-
sion. There is normally a short delay of a few
seconds, representing the transit time for fresh
blood from the lungs to reach the left heart
and then the chemoreceptors in the carotid
artery and the brain. By the time the brain be-
gins increasing the rate and depth of respira-
tion, the alveolar carbon dioxide has reached
even higher levels, and so there is a gradual
ramping up of respiration as the brain sees a
rising level of carbon dioxide, despite its ad-
ditional efforts. By the time the brain begins
to see a fall in carbon dioxide tension, the levels
in the alveoli may be quite low. When blood
containing this low level of carbon dioxide
reaches the brain, respiration slows or may even
cease, thus setting off another cycle. Hence,
the periodic cycling is due to the delay (hys-
B
C
D
E
A
B
C
D
E
A
1 min
Figure 2–5. Different abnormal respiratory patterns are associated with pathologic lesions (shaded areas) at various levels
of the brain. Tracings by chest-abdomen pneumography, inspiration reads up. (A) Cheyne-Stokes respiration is seen with
metabolic encephalopathies and with lesions that impair forebrain or diencephalic function. (B) Central neurogenic
hyperventilation is most commonly seen in metabolic encephalopathies, but may rarely be seen in cases of high brainstem
tumors. (C) Apneusis, consisting of inspiratory pauses, may be seen in patients with bilateral pontine lesions. (D) Cluster
breathing and ataxic breathing are seen with lesions at the pontomedullary junction. (E) Apnea occurs when lesions en-
croach on the ventral respiratory group in the ventrolateral medulla bilaterally. (From Saper, C. Brain stem modulation of
sensation, movement, and consciousness. Chapter 45 in: Kandel, ER, Schwartz, JH, Jessel, TM. Principles of Neural Sci-
ence. 4th ed. McGraw-Hill, New York, 2000, pp. 871–909. By permission of McGraw-Hill.)
50
Plum and Posner’s Diagnosis of Stupor and Coma
teresis) in the feedback loop between alveolar
ventilation and brain chemoreceptor sensory
responses.
The Cheyne-Stokes respiratory cycle is not
usually seen in normal individuals because the
circulatory delay between a change in alveolar
blood gases and carbon dioxide tension in the
brain is only a few seconds. Even as circulatory
delay rises with cardiovascular or pulmonary
disease, during waking the descending path-
ways that prevent posthyperventilation apnea
also ensure the persistence of respiration even
during periods of low metabolic need, thus
damping the oscillations that produce Cheyne-
Stokes respiration. However, during sleep or
with bilateral forebrain impairment, due either
to a diffuse metabolic process such as uremia,
hepatic failure, or bilateral damage such as ce-
rebral infarcts or a forebrain mass lesion with
diencephalic displacement, periodic breathing
may emerge.
43–45
In patients with heart fail-
ure, the transit time for blood from the lungs
to reach the carotid and cerebral chemorecep-
tors can become so prolonged as to produce
a Cheyne-Stokes pattern of respiration, even
in the absence of forebrain impairment. Thus,
Cheyne-Stokes respiration is mainly useful as
a sign of intact brainstem respiratory reflexes
in the patients with forebrain impairment, but
cannot be interpreted in the presence of sig-
nificant congestive heart failure.
HYPERVENTILATION IN
COMATOSE PATIENTS
Sustained hyperventilation is often seen in pa-
tients with impaired consciousness, but is usu-
ally a result of either hepatic coma or sepsis,
conditions in which circulating chemical stim-
uli cause hyperpnea, or a metabolic acidosis,
such as diabetic ketoacidosis (see Chapter 5).
Other patients have meningitis caused either
by infection or subarachnoid hemorrhage,
which stimulates chemoreceptors in the brain-
stem,
46
probably by altering CSF pH.
Some patients hyperventilate when intrin-
sic brainstem injury or subarachnoid hemor-
rhage or seizures cause neurogenic pulmonary
edema.
47
The ventilatory response is driven by
pulmonary mechanosensory and chemosensory
receptors. The pulmonary congestion lowers
both the arterial carbon dioxide and the oxygen
tension. Stimulation of pulmonary stretch re-
ceptors is apparently sufficient to cause reflex
hyperpnea, as oxygen therapy sufficient to raise
the arterial oxygen level does not always cor-
rect the overbreathing.
Another small group of patients has been
identified who have hyperventilation associ-
ated with brainstem gliomas or lymphomas.
48,49
These patients have spinal fluid that is acellu-
lar, but generally acidotic compared to arterial
pH. In others, the lumbar CSF may have a nor-
mal pH, but it is believed that the tumor causes
local lactic acidosis, which may trigger brain che-
moreceptors to cause hyperventilation (Figure
2–5).
It is theoretically possible for an irritative
lesion in the region of the parabrachial nu-
cleus or other respiratory centers to produce
hyperpnea.
37
The diagnosis of such true ‘‘cen-
tral neurogenic hyperventilation’’ requires that
with the subject breathing room air, the blood
gases show elevated arterial oxygen tension,
decreased carbon dioxide tension, and an ele-
vated pH. The cerebrospinal fluid likewise must
show an elevated pH and be acellular. The re-
spiratory changes must persist during sleep to
eliminate psychogenic hyperventilation, and
one must exclude the presence of stimulating
drugs, such as salicylates, or disorders that stim-
ulate respiration, such as hepatic failure or un-
derlying systemic infection. Cases fulfilling all
of these criteria have rarely been observed,
50,51
and none that we are aware of has come to post-
mortem examination of the brain.
APNEUSTIC BREATHING
Apneusis is a respiratory pause at full inspira-
tion. Fully developed apneustic breathing, with
each cycle including an inspiratory pause, is
rare in humans, but of considerable localizing
value. Experiments in animals indicate that ap-
neusis develops with injury to the pontine re-
spiratory nuclei described above, and experi-
ence with rare human cases would support this
view
52,53
(see Figure 2–5).
Clinically, end-inspiratory pauses of 2 to 3
seconds usually alternate with end-expiratory
pauses, and both are most frequently encoun-
tered in the setting of pontine infarction due
to basilar artery occlusion. However, apneustic
breathing may rarely be observed in metabolic
encephalopathies, including hypoglycemia, an-
oxia, or meningitis. It is sometimes observed
Examination of the Comatose Patient
51