An Unusual Fatal Complication of
Low Basilar Trunk Aneurysm
Surgery: Isolated Prepontine
Tension Pneumocephalus
Servet Inci, M.D., Bayram Çirak, M.D., and Vural Bertan, M.D.
Department of Neurosurgery, School of Medicine, University of Hacettepe, Ankara, Turkey

Inci S, Çirak B, Bertan V. An unusual fatal complication of low
basilar trunk aneurysm surgery: isolated prepontine tension
pneumocephalus. Surg Neurol 1999;52:485–9.
OBJECTIVE

A case of postoperative tension pneumocephalus after
low basilar trunk aneurysm clipping is presented. To our
knowledge, this is the first case of isolated prepontine
tension pneumocephalus.
BACKGROUND

A 63-year-old woman was admitted for repair of a basilar
aneurysm that had caused a subarachnoid hemorrhage.
She was cooperative and partially oriented. According to
Hunt & Hess classification, she was considered Grade III.
METHOD

The aneurysm was clipped, using a right lateral suboccipital craniectomy with the patient in the sitting position. In the early postoperative period, she had no new
neurological deficit. However, 2 hours later the patient
became lethargic and unresponsive to verbal commands.
Emergency CT scan revealed an isolated prepontine tension pneumocephalus with prominent posterior displacement of the pons. She was immediately taken back to
surgery. Upon incision of the dura mater, air could be
heard escaping under pressure from the posterior fossa
cavity. The clip was in its proper position and all arteries
were patent. Spontaneous respiration and pupil reflexes
returned soon after surgery, but she remained unconscious and died 3 days later.
CONCLUSION

We believe that this death was directly attributable to the
tension pneumocephalus and the distortion of the pons.
Postoperative prepontine tension pneumocephalus, although this is an extremely rare condition, should be
considered if a patient deteriorates after basilar aneurysm surgery in the sitting position. © 1999 by Elsevier
Science Inc.
KEY WORDS

Complication, pneumocephalus, prepontine cistern, sitting
position.

Address reprint requests to: Dr. Servet Inci, Emek Mahallesi, 4. Cadde,
No: 70/8, 06510 Ankara, Turkey.
Received May 19, 1998; accepted June 10, 1998.
© 1999 by Elsevier Science Inc.
655 Avenue of the Americas, New York, NY 10010

ostoperative tension pneumocephalus is a serious and potentially life-threatening condition. Pneumocephalus occurs frequently after posterior fossa surgery, especially when performed in
the sitting position, but in the vast majority of
cases, air is localized in the supratentorial spaces.
To our knowledge, a case of isolated prepontine
tension pneumocephalus has not been reported
previously. In this report, an unusual fatal case of
prepontine tension pneumocephalus after low basilar trunk aneurysm surgery carried out with the
patient in the sitting position is presented, and the
causes of postoperative tension pneumocephalus
are discussed.

P

Case Report
A 63-year-old right handed woman was referred to
Hacettepe Neurosurgery Clinic for repair of a basilar aneurysm that had caused a subarachnoid hemorrhage 10 days previously. At the time of admission to our hospital, this patient had a mild left
hemiparesis, left 6th nerve palsy and stiff neck but
she was cooperative and partially oriented. According to Hunt & Hess classification [10] she was considered Grade III. Physical examination was within
normal limits. There was no pathology on preoperative computed tomography (CT). Cerebral angiography revealed a large low basilar trunk aneurysm
(Figure 1). Surgical intervention was planned and its
risks were explained to the patient.
FIRST OPERATION
This procedure was performed on September 3,
1996, using a right lateral suboccipital craniectomy
with the patient in the sitting position. Hyperventilation, 20mg furosemide and 30g mannitol (0.5g/kg)
were used to reduce brain bulk. Throughout sur0090-3019/99/$–see front matter
PII S0090-3019(99)00040-3

486 Surg Neurol
1999;52:485–9

1

Vertebral angiogram showing a large low basilar
trunk aneurysm.

gery, body temperature, central venous pressure,
intraarterial blood pressure, O2 saturation, and endtidal CO2 partial pressure were monitored continuously. Ventilation was controlled to maintain PaCO2
levels of 28 –30 mmHg. Anesthesia was maintained
with isoflurane, muscle relaxant and 60% nitrous
oxide in oxygen. A small amount of cerebrospinal
fluid (CSF) was drained by opening the right pontocerebellar cistern. The right vertebral artery was
followed and the vertebrobasilar junction was seen.
There was no CSF in the prepontine cistern. The
aneurysm was found against the clivus and in front
of the pons. Because of very dense adhesions in the
prepontine cistern, the base of the aneurysm was
dissected with difficulty away from the basilar artery and the neck was identified. To gain better
access to the base of the aneurysm, the 7th and 8th
nerves were retracted inferiorly and the 5th nerve
superiorly. At last the aneurysm, which projected
anterolaterally, was clipped using a McFadden clip.
The dome of the aneurysm was punctured and approximately 2 cc of blood was aspirated. After hemostasis and papaverin application, the dura mater
was closed. The nitrous oxide was terminated 20
minutes before this procedure. Total anesthesia
time was 5 hours and 30 minutes. The patient was
extubated after surgery in the operating room and
returned to the intensive care unit breathing spontaneously. In the early postoperative period, she
was alert, cooperative and partially oriented. She
had no new neurologic deficit except a right 7th
nerve palsy. Two hours after surgery, the patient
became lethargic and unresponsive to verbal commands. Emergency CT scan revealed a posterior
fossa tension pneumocephalus, mainly in the prepontine cistern, with prominent posterior displacement of the pons (Figure 2). Only a minimal amount

Inci et al

Postoperative CT scan showing prepontine tension
pneumocephalus with prominent distortion of the
pons.

2

of air had passed into the supratentorial spaces.
There was no bleeding or hydrocephalus. Meanwhile the patient rapidly became comatose and unresponsive to pain with fixed miotic pupils. There
was no time for control angiography. Because of
nearby vital structures such as the pons and basilar
artery, percutaneous needle aspiration could not be
performed and she was immediately taken back to
surgery.
SECOND OPERATION
The previous scalp incision was reopened. The
dura mater was extremely tight. Upon incision of
the dura mater, air could be heard escaping under
pressure from the posterior fossa cavity. The surgical field was re-explored. The clip was in its
proper position. Both vertebral arteries, the basilar
artery and its main branches were patent and there
was no vasospasm. The prepontine cistern was
opened to the interpeduncular cistern by microdissection. After dural closure, the surgical cavity was
filled with normal saline. Her spontaneous respiration and pupil reflexes returned soon after surgery
but she remained unconscious and died 3 days
later. Unfortunately, postmortem examination
could not be obtained.

Discussion
We think that the cause of death was not a vascular
insult such as a slipped clip or vasospasm. These
probabilities were eliminated in the second operation. Of course, we cannot totally exclude the possibility of disturbance of microcirculation. But if
microvascular structures were damaged during the
first operation, the patient would not have awak-

Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Exitus
Needle asp.
Needle asp.
Shunt was externalized
Twist drill asp
Needle asp.
Needle asp
Conservative
Reexploration
Twist drill asp.
Conservative
V/A Shunt
Reexploration
2 hours postop.
6 hours postop.
Immediately postop.
2 hours postop.
Intraoperative
Intraoperative
Immediately postop.
Immediately postop.
Immediately postop.
4 hours postop.
6 hours postop.
2 hours postop.
56, M
8, F
18, M
9, F
63, F
5, M
56, M
61, M
52, M
77, M
65, M
62, F
Molez & Mezzadri [19]
Present case

1980
1980
1981
1982
1982
1982
1983
1983
1983
1992
1996
Luenda et al [14]
Grundy & Spetzler [8]
Friedman et al [7]
Artu [2]
Thiagarajah et al [28]
MacGillivray [16]
Toung et al [29]

*This patient died 2 days later because of cerebellar hematoma.
Tm, tumor; Postop, postoperative; Asp, aspiration.

“
“
“
“
“
“
“
“
“
“
“
“

⫹
⫹
?
⫹
⫹
⫹
⫹
⫹
⫹
⫺
⫹
⫹

“
“
“
“
“
“ & intraventricular
Bifrontal subdural
“
“
“
4th ventricle
Prepontine

Good
?
Good*
Needle asp.
?
Needle asp.
Immediately postop.
Immediately postop.
1 hour postop.

Cerebellar tm
Cervical spondylosis
Glossopharyngeal
neuralgia
Trigeminal neuralgia
Brain stem tumor
Acoustic neuroma
Cerebellar AVM
Acoustic neuroma
Cerebellar tumor
Cerebellar tumor
Cerebellar tumor
Cerebellar tumor
Acoustic neuroma
Cerebellar hematoma
Basilar aneurysm
66, M
43, M
69, F
1976
1976
1979
Kitahata & Katz [12]
Hullet & Laing [9]
Lunsford et al [15]

YEAR
AUTHORS

Sitting
“
“

⫹
⫹
⫹

Bifrontal subdural
“
“

TREATMENT
AIR

LOCATION OF

DIAGNOSIS

POSITION

TIME OF THE
SYMPTOMATIC
PNEUMOCEPHALUS
USE OF
NO
AGE &
SEX

Clinical Summary of Patients Operated in the Sitting Position with Symptomatic Tension Pneumocephalus of Early Onset

1

ened from the first anesthesia. However, she had no
new deficits except for a right 7th nerve palsy. Furthermore, spontaneous respiration and pupil reflexes did not return immediately after the second
operation, because no correctable lesion was found
at the second operation except pneumocephalus.
We believe, therefore, that death was directly attributable to the tension pneumocephalus and the
distortion of the pons.
Pneumocephalus was first demonstrated on skull
x-ray films by Luckett [13], in 1913. Since that time,
a variety of etiological factors for the development
of pneumocephalus have been described, including
trauma [26], tumor [4], infection caused by gascontaining organisms [27], and some therapeutic
maneuvers [3,11,22]. Markham [18], in his extensive
review, noted that trauma is by far the leading
cause. In his series, a surgical procedure (mainly
neurosurgical) was the cause of only 3.7% of 295
cases of pneumocephalus.
Although the sitting position provides optimal
operating conditions for neurosurgeons in posterior fossa surgery, it is associated with some potential complications. All complications related to the
sitting position have been reported by Albin et al
[1] and Standefer et al [25]. Air embolism and postural hypotension are well-known complications.
This position is also the most common predisposing cause of postoperative tension pneumocephalus. Tension pneumocephalus as a complication of
the sitting position was first described by Kitahata
and Katz [12], in 1976. In the same year Hullet and
Laing [9] also reported a case. We reviewed the
pertinent literature and found only 11 similar case
reports [2,7,8,9,12,14 –16,19,28,29] (Table 1). All
these cases present some common characteristics:
1) all cases were operated in the sitting position; 2)
symptoms started within the first 6 hours; 3) air
was localized within the supratentorial subdural
(bifrontal) spaces in all cases except one; 4) nitrous
oxide was used in most cases; and 5) all patients
promptly improved after simple air aspiration except our case.
Two mechanisms have been theorized in the development of pneumocephalus. First, the “ball
valve” mechanism requires air under high pressure
to be forced intracranially. A prerequisite is a defect
in the skull base and/or dura mater. Second, during
the operation, gravity allows the brain to settle in
the cranium, CSF drains out and air enters through
the surgical wound to occupy the vacuum thus
created at the top. This mechanism is described as
an “inverted soda-pop bottle effect” by Lunsford et
al [15]. This mechanism is enhanced by hyperventilation and dehydrating agents such as mannitol

Surg Neurol 487
1999;52:485–9

OUTCOME

Prepontine Tension Pneumocephalus

488 Surg Neurol
1999;52:485–9

because by reducing brain bulk they increase aircontaining spaces. The “inverted bottle” mechanism is, of course, responsible for pneumocephalus
in our case. Under normal conditions, air passes
into the supratentorial spaces via subarachnoid
and subdural pathways, but in our case air remained within the prepontine cistern. This is difficult to explain. We postulate that, because of dense
adhesions and fibrin due to old hemorrhage in the
pontocerebellar and prepontine cisterns, air could
not pass into supratentorial cisterns and subdural
spaces. So air was trapped in the posterior fossa
cavity, particularly in the prepontine cistern. After
closure of the dura, the neural tissues (especially
the pons and cerebellum) re-expanded because of
rehydration, cessation of hyperventilation, and the
rebound effect of mannitol. Because of the critical
location of the air, even a slight increase in the
volume of neural tissue caused increased pressure
and the pneumocephalus became symptomatic.
Furthermore, air expanded by body temperature
also contributes to increasing intra cavity pressure.
Finally, cisternal air under tension produced severe
neurological deficits by displacement and/or distortion of the pons.
Nitrous oxide anesthesia has also been implicated in the development of tension pneumocephalus during craniotomy [2,17,21]. Because the solubility of nitrous oxide in blood is 34 times more than
that of nitrogen [17], nitrous oxide readily diffuses
into sealed pockets of air, and rapid increase in
intracranial pressure may occur in the presence of
pneumocephalus [2,6,24]. This phenomenon has
been shown to occur in animals and humans during
air encephalography [23]. But intracranial pressure
returns to the original levels within 10 minutes after
discontinuing nitrous oxide [23]. In our case, nitrous oxide was discontinued at least 20 minutes
before the dura mater was closed. In addition, in
Case 3 presented by Toung et al [29], when nitrous
oxide was not used at all, tension pneumocephalus
still occurred. We believe, therefore, that nitrous
oxide probably does not play a major role in the
development of tension pneumocephalus, but it
may be a contributing factor. However, it has been
suggested that nitrous oxide should be discontinued before dural closure to reduce the incidence of
tension pneumocephalus [2,17,5].
Neurologic symptoms caused by tension pneumocephalus may range from minimal confusion to
death but most of the time it is asymptomatic [30].
According to a prospective study of DiLorenzo et al
[5], the incidence of postoperative pneumocephalus was 100% and only two cases (6.6%) became
symptomatic. This incidence is only 3% in the series

Inci et al

of Standefer et al [25]. In symptomatic cases, meningeal irritation, confusion, disorientation, disturbance of consciousness, and coma are seen frequently. The diagnosis is quickly established on CT
scans that can detect volumes of air as small as 0.5
cc and that can differentiate between epidural, subdural, subarachnoid and intraparenchymal air in
most cases [20]. In symptomatic cases, simple needle aspiration is sufficient for treatment [12,14,15,
28]. We could not perform this procedure because
of nearby vital structures such as the basilar artery
and pons. The goal of treatment is to reduce the
tension within the cranium rather than to eliminate
the intracranial air.
In conclusion, postoperative prepontine tension
pneumocephalus, although this is an extremely rare
condition, should be considered if a patient deteriorates after basilar aneurysm surgery in the sitting
position.
REFERENCES
1. Albin MS, Babinski M, Maroon JC, Janetta PJ. Anesthetic management of posterior fossa surgery in the
sitting position. Acta Anesth Scand 1976;20:117–28.
2. Artu AA. Nitrous oxide plays a direct role in the
development of tension pneumocephalus intraoperatively. Anesthesiology 1982;57:59 – 61.
3. Ash K, Cannon J, Biehl D. Pneumocephalus following
attempted epidural anesthesia. Can J Anesth 1991;38:
772– 4.
4. Clark JB, Six EG. Epidermoid tumor presenting as
tension pneumocephalus. Case report. J Neurosurg
1984;60:1312–14.
5. DiLorenzo N, Caruso R, Floris R, Guerrisi V, Bozzao L,
Fortuna A. Pneumocephalus and tension pneumocephalus after posterior fossa surgery in the sitting
position: a prospective study. Acta Neurochir 1986;
83:112–15.
6. Domino KB, Hemstad JR, Lam AM, Laohaprasit V,
Mayberg TA, Harrison SD, Grady MS, Winn HR. Effect
of nitrous oxide on intracranial pressure after cranialdural closure in patients undergoing craniotomy. Anesthesiology 1992;77:421–5.
7. Friedman GA, Norfleet EA, Bedford RF. Discontinuance of nitrous oxide does not prevent tension pneumocephalus. Anest Analg 1981;60:57– 8.
8. Grundy BL, Spetzler RF. Subdural pneumocephalus
resulting from drainage of cerebrospinal fluid during
craniotomy. Anesthesiology 1980;52:269 –71.
9. Hullet WB, Laing JW. Tension pneumocephalus and
the sitting position. Anesthesiology 1976;43:578.
10. Hunt WE, Hess RM. Surgical risk as related to time of
intervention in the repair of intracranial aneurysms.
J Neurosurg 1968;28:14 –20.
11. Katz Y, Markooits R, Rosenberg B. Pneumocephalus
after inadvertent intrathecal air. Anesthesiology
1990;73:1277–9.
12. Kitahata LM, Katz JD. Tension pneumocephalus after
posterior fossa craniotomy, a complication of the
sitting position. Anesthesiology 1976;44:448 –50.
13. Luckett WH. Air in the ventricles of the brain, follow-

Prepontine Tension Pneumocephalus

ing a fracture of the skull. Report of a case. Surg
Gynecol Obstet 1913;17:237– 40.
14. Leunda G, Cabezudo JM, Areito, Vaquero J, Gilsanz
F. Subdural tension pneumocephalus after posterior fossa operation. Is the inverted bottle phenomenon the only causative factor? Surg Neurol 1981;
15:303–5.
15. Lunsford LD, Maroon JC, Sheptak PE, Albin MS. Subdural tension pneumocephalus. Report of two cases.
J Neurosurg 1979;50:525–7.
16. MacGillivray RG. Pneumocephalus as a complication
posterior fossa surgery in the sitting position. Anaesthesia 1982;17:722–5.
17. Mallamo JR, Hubbard RB, Boone SC, Reisner LS,
Pister JD. Expansion of air-filled subdural space
during nitrous oxide anesthesia. Radiology 1975;
115:369 –72.
18. Markham JW. The clinical features of pneumocephalus based upon a survey of 284 cases with report of 11
additional cases. Acta Neurochir 1967;16:1–78.
19. Molez DG, Mezzadri JJM. Postoperative fourth ventricle tension pneumocephalus. J Neurol Neurosurg
Psychiatr 1992;55:511–2.
20. Osborn AG, Daines JH, Wing SD, Anderson RE. Intracranial air on computerized tomography. J Neurosurg
1978;48:355–9.
21. Pandit UA, Mudge BJ, Keller TS, Samra SK, Kilaru P,
Pandit SK, Cohen PJ. Pneumocephalus after posterior
fossa exploration in the sitting position. Anaesthesia
1982;37:996 –1001.
22. Ruge J, Cerullo L, McLone D. Pneumocephalus in patients with CSF shunts. J Neurosurg 1985;63:532– 6.
23. Saidman LJ, Eger EI. Change in cerebrospinal fluid
pressure during pneumoencephalography under nitrous oxide anesthesia. Anesthesiology 1965;26:67–
72.
24. Skahen S, Shapiro HM, Drummond JC, Todd MM, Zelman V. Nitrous oxide withdrawal reduces intracranial
pressure in the presence of pneumocephalus. Anesthesiology 1986;65:192–195.
25. Standefer M, Bay JW, Trusso R. The sitting position in
neurosurgery: a retrospective analysis of 488 cases.
Neurosurgery 1984;14:649 –58.
26. Stuedel WI, Hacker H. Prognosis, incidence and management of acute traumatic intracranial pneumocephalus: a retrospective analysis of 49 cases. Acta
Neurochir 1986;80:93–9.
27. Taguchi Y, Sato J, Nakamura N. Gas-containing brain

Surg Neurol 489
1999;52:485–9

abscess due to fusobacterium nucleatum. Surg Neurol 1981;16:408 –10.
28. Thiagarajah S, Frost EAM, Singh T, Shulman K. Cardiac arrest with tension pneumocephalus. Anesthesiology 1982;56:73–5.
29. Toung T, Donham RT, Lehner A, Alano J, Campbell J.
Tension pneumocephalus after posterior fossa craniotomy: report of four additional cases and review of
postoperative pneumocephalus. Neurosurgery 1983;
12:164 – 8.
30. Toung TJ, McPherson RW, Ahn H, Donham RT, Alano
J, Lond D. Pneumocephalus: effects of patient position on the incidence and location of aerocele after
posterior fossa and upper cervical cord surgery.
Anesth Anal 1986;65:65–70.

COMMENTARY

The authors describe the highly unusual and fatal
complication of prepontine tension pneumocephalus after surgery for a basilar trunk aneurysm in the
sitting position.
This obviously rare complication is very well documented and supported by the CT scan picture
provided in this manuscript as well as the intraoperative findings. It is quite likely that this was the
reason for the deterioration of the patient. As the
authors mention, tension pneumocephalus is rare,
because there is passage of air from the posterior
fossa to the supratentorial compartments. They hypothesize that, in this case, the air was trapped by
the dense adhesions to the cistern.
Unfortunately, very little can be done to avert this
complication in the future, except perhaps not using nitrous oxide at all. Another option may be to
use an alternative approach, such as a skull base
petrosal approach, to reach this aneurysm in a supine position.
Fady T. Charbel, M.D.
Department of Neurosurgery
University of Illinois at Chicago
Chicago, Illinois