Full text of DOI:10.1097/00005537-200206000-00008

The Laryngoscope
Lippincott Williams & Wilkins, Inc., Philadelphia
© 2002 The American Laryngological,
Rhinological and Otological Society, Inc.
Management of Multiple Spontaneous Nasal
Meningoencephaloceles
Rodney J. Schlosser, MD; William E. Bolger, MD
Objectives/Hypothesis: Multiple spontaneous na-
sal meningoencephaloceles in the same patient are
rare lesions. Although many skull base defects occur
after prior trauma or surgery, otolaryngologists must
be aware of the potential for spontaneous encephalo-
celes. We present our experience with this unusual
condition and discuss its pathophysiology and unique
management issues. Study Design: Retrospective.
Methods: Review of medical records, radiographic im-
ages, and cerebrospinal fluid pressures. Results: We
identified 5 patients with multiple, simultaneous,
spontaneous encephaloceles: 4 patients with 2 en-
cephaloceles and 1 patient with 3 encephaloceles (11
in all). Locations of the 11 encephaloceles were sphe-
noid lateral recess (6), frontal sinus with supraorbital
ethmoid extension (2), ethmoid roof (1), frontal sinus
(1), and central sphenoid (1). Three patients had bi-
lateral sphenoid lateral recess encephaloceles, ac-
counting for all six in that location. All four patients
with available radiographic studies demonstrated
empty sella turcica. Surgical approaches included en-
doscopic transpterygoid approach to the lateral sphe-
noid recess (3), endoscopic approach to ethmoid and
central sphenoid (3), and osteoplastic flap with fron-
tal sinus obliteration (2). We had 100% success at lat-
est endoscopic follow-up (mean period, 17 mo). Three
patients had postoperative lumbar punctures with
mean cerebrospinal fluid pressure of 28.3 cm water
(range, 19–34 cm; normal range, 0–15 cm). Conclu-
sions: Multiple spontaneous encephaloceles can be
managed safely and successfully using endoscopic
and extracranial approaches. A high index of suspi-
cion for this diagnosis must be maintained, especially
in patients with radiographic evidence of laterally
pneumatized sphenoid sinuses or empty sella. Spon-
taneous encephaloceles and cerebrospinal fluid leaks
represent a form of intracranial hypertension. Key
Words: Spontaneous encephalocele, endoscopic sinus
surgery, cerebrospinal fluid leak, empty sella.
Laryngoscope, 112:980–985, 2002
INTRODUCTION
Nasal meningoencephaloceles (often shortened to en-
cephalocele or cephalocele) are herniations of the menin-
ges and brain parenchyma into the nose. Because of the
associated skull base defect and cranionasal fistula, cere-
brospinal fluid (CSF) leaks and the risk of meningitis can
accompany the condition when the meningeal layer is
breached.
Encephaloceles are uncommon and can arise from
congenital, traumatic, or spontaneous origins. More expe-
rience has been reported with CSF leaks because they
represent a smaller cranionasal fistula and are more com-
mon than encephaloceles. Most CSF leaks are caused by
skull base fractures or surgical trauma. Spontaneous or
nontraumatic encephaloceles or CSF leaks have been the
least common in most series, accounting for only 3% to 5%
of all CSF leaks.1–4 Historically, spontaneous cranionasal
fistulas were divided into high-pressure leaks, typically
caused by an intracranial tumor and hydrocephalus, and
otherwise idiopathic leaks thought to be in the normal-
pressure category. Our experience is that these spontane-
ous, idiopathic encephaloceles and CSF leaks are probably
a result of elevated CSF pressures exerting hydrostatic
forces on anatomically fragile areas of the skull base.
Extracranial endoscopic repair of encephaloceles and
CSF leaks has been shown to have a high success rate
without the morbidity of traditional intracranial ap-
proaches.2–4 Patients with spontaneous fistulas represent
a unique patient population because of the underlying
pathophysiology of their condition and the overall atten-
uation of their skull base. We present our series of pa-
tients with multiple spontaneous nasal encephaloceles
and discuss diagnosis, management, and probable patho-
physiology of the condition.
Presented at the Meeting of the Eastern Section Meeting of the
Triological Society, Philadelphia, PA, January 27, 2002.
From the Department of Otorhinolaryngology—Head and Neck Sur-
gery, University of Pennsylvania, Philadelphia, Pennsylvania, U.S.A.
MATERIALS AND METHODS
Editor’s Note: This Manuscript was accepted for publication Febru-
ary 6, 2002.
Send Correspondence to William E. Bolger, MD, Department of
Otorhinolaryngology, University of Pennsylvania, 5 Ravdin/Silverstein,
3400 Spruce Street, Philadelphia, PA 19104, U.S.A. E-mail:
william.bolger@uphs.upenn.edu
Patient Data
We retrospectively reviewed medical records and imaging
studies of all patients with multiple spontaneous nasal meningo-
encephaloceles and collected demographic and treatment
information.
Laryngoscope 112: June 2002
980
Schlosser and Bolger: Multiple Encephaloceles

can be removed, leading to intracranial herniation. Conversely, if
the drain height is too high, air may flow retrograde into the
intrathecal space and lead to pneumocephalus. Nursing staff
must understand that the rate of drainage is to be adjusted by
changing the height of the drain, not by clamping it temporarily.
Preoperative Management
All patients were evaluated with preoperative computed
tomography (CT) and magnetic resonance imaging (MRI) scans to
assist in localization of the skull base defects. When clear rhinor-
rhea was present, nasal drainage was sent for ␤-2 transferrin
analysis to confirm the diagnosis. Additional radiographic studies
such as radioactive cisternograms or CT cisternograms were ob-
tained when additional information was needed to identify the
sites of the encephaloceles. Preoperative prophylactic antibiotics
were not used.
Surgical Techniques
Patients were treated with a variety of surgical techniques
based on the location of their encephaloceles as outlined in Table
I. All skull base defects were repaired using a multilayer closure
consisting of cortical bone grafts (typically, mastoid bone) and
fascia grafts as described later in this section.
Lumbar Drain
All procedures were started by placing a lumbar drain be-
fore the induction of general anesthesia. This was performed with
the patient awake and in a sitting position to fill the lumbar
cistern with CSF. Prior experience placing lumbar drains with
the patient in the decubitus position has shown us that, at times,
patients with long-term CSF leaks can have a relative depletion
of CSF, and it can be difficult to withdraw CSF and subsequently
place the drain. Intrathecal fluorescein (0.1 mL 10% solution
diluted in 10 mL CSF) was slowly injected through the spinal
needle just before drain insertion. This aids in precise identifica-
tion of the encephalocele and associated bony defect. Use of flu-
orescein also assists in confirmation of watertight closure at the
end of the procedure. Informed consent for the use of intrathecal
fluorescein was obtained in all patients. The risks, benefits, and
lack of U.S. Food and Drug Administration (FDA) approval for
intrathecal use were disclosed.
Once the drain is in place, it is clamped and general anes-
thesia is induced. Care is taken during induction to avoid mask-
ing the patient and using positive-pressure ventilation because
this carries the risk of causing pneumocephalus through the
cranionasal fistula. Once the skull base defect is exposed, just
before bone graft placement, 15 to 25 mL CSF is removed over a
15-minute period to aid in reduction of the encephalocele. The
lumbar drain is kept open and draining 5 to 10 mL per hour for
the remainder of the procedure and for the initial postoperative
period. This is especially useful in avoiding large spikes in intra-
cranial pressure (ICP) should the patient cough or strain during
extubation or have nausea and vomiting during the postoperative
period.
Transnasal endoscopic repair was performed as previ-
ously described,3 whenever possible, depending on the location
of the encephalocele. The endoscopic approach frequently re-
quires a total sphenoethmoidectomy and possible middle tur-
binectomy to properly identify and expose the skull base defect.
All endoscopic procedures are started by topical decongestion
with oxymetazoline (0.05%), followed by injections using 1%
lidocaine with 1:100,000 epinephrine. Injections are performed
in the lateral nasal wall, turbinates, and transoral and trans-
nasal sphenopalatine blocks, depending on the site of the en-
cephalocele and the approach needed. The nose is irrigated
with clindamycin solution to decrease bacterial contamination
within the operative field. Encephaloceles are reduced or am-
putated using monopolar and bipolar cautery. Meticulous he-
mostasis is required, to avoid intracranial hemorrhage. Once
the encephalocele is adequately reduced and several millime-
ters of mucosa surrounding the bony defect is removed, recon-
struction of the defect is started. A postauricular approach is
used to harvest a thick layer of temporalis fascia and the outer
layer of cortical bone overlying the mastoid air cells. The repair
is performed by gently elevating the dura above the bony skull
base defect and then using the bone graft in an underlay
fashion. This requires great care in bone graft design and
placement because the entire skull base in these patients is
attenuated and can easily fracture or break off, creating an
even larger defect. In addition, these defects often have com-
plex three-dimensional configurations, requiring precise
sculpting of the bone graft. Once the underlay bone graft is of
satisfactory size and shape, it is soaked in clindamycin solution
to minimize any intracranial seeding of bacteria. After position-
ing the bone graft, the temporalis fascia is placed in an overlay
fashion. The fascia graft is intentionally kept thick, in contrast to
the thin fascia graft used for tympanoplasties, to decrease the
chance of having a small pinhole develop. The fascia graft is
followed by multiple layers of absorbable packing.
Diligent care of the lumbar drain is required during the
immediate postoperative period. The drain should be kept open at
all times to prevent elevations in ICP from pushing against the
graft and dislodging it. Adjustments in the height of the drain are
used to keep the drainage rate between 5 and 10 mL CSF per
hour. If the drain height is too low, an excessive amount of CSF
TABLE I.
Patient Data.
Encephalocele
Location
Follow-up
(mo)
Empty
sella
Patient
LA
Operation
BMI (kg/m2)
29.8
ICP (cm H2O)
19
L frontal/SOE
L lateral sphenoid
R lateral sphenoid
OPF
12
ϩ
Transpterygoid
Observation
EH
SS
FL
IY
R posterior ethmoid
L frontal
Endoscopic repair
OPF
30
15
5
ϩ
N/A
ϩ
32.8
N/A
Refuses
Lost
32
L lateral sphenoid
R lateral sphenoid
Transpterygoid
Observation
R frontal/SOE
L central sphenoid
Endoscopic repair
Endoscopic repair
36.0
32.9
R lateral sphenoid
L lateral sphenoid
Transpterygoid
Observation
22
ϩ
34
SOE ϭ supraorbital ethmoid; OPF ϭ osteoplastic flap and obliteration.
Laryngoscope 112: June 2002
Schlosser and Bolger: Multiple Encephaloceles
981

Encephaloceles located in the lateral recess of the sphenoid
sinus required an endoscopic transpterygoid approach.5 This ap-
proach begins with a total sphenoethmoidectomy and maxillary
antrostomy. The posterior wall of the maxillary sinus is then
removed, and the pterygopalatine fossa is entered. The internal
maxillary artery and its branches are identified and transposed
inferiorly or clipped and divided to expose the deeper areas of the
pterygopalatine fossa. The vidian nerve and the maxillary divi-
sion of cranial nerve V are dissected free and preserved if possi-
ble. The anterior wall of the sphenoid sinus is drilled away to gain
access to the lateral recess of the sphenoid sinus. This provides
the exposure needed to reduce the encephalocele and repair the
skull base defect with bone and fascia grafts as described earlier
in this section.
An external approach using an osteoplastic flap and ipsilat-
eral frontal sinus obliteration was used for encephaloceles with
significant extension into the frontal sinus. All visible mucosa is
removed from the frontal sinus, followed by meticulous drilling
with a diamond burr to remove mucosal remnants. Underlay bone
and overlay fascia grafts are placed before obliterating the sinus
with abdominal fat.
three separate encephaloceles, and there were 11 en-
cephaloceles in the current series (Table I). Four of five
patients were women. Average age was 49.2 years (age
range, 39–60 y) with an average of 0.6 prior attempted
repairs (range, 0–2). Three of the women had body mass
index (BMI) greater than 30 (obese), and the fourth
woman had a BMI of 29.8 (overweight). Presenting symp-
toms included CSF rhinorrhea (4 patients), seizure (2
patients), meningitis (1 patient), headache (1 patient), and
dizziness (1 patient). Three encephaloceles in the ethmoid
roof or central sphenoid were repaired using the standard
endoscopic approach (Fig. 1). Three were located in the
lateral sphenoid recess and were repaired using the endo-
scopic transpterygoid approach (Fig. 2A and B). Two were
approached using an osteoplastic flap with frontal sinus
obliteration, and three encephaloceles were asymptomatic
without CSF leak and are being followed with serial MRI
scans every 6 to 12 months. Complications occurred in
only one patient and consisted of decreased sensation in
the distribution of the intraorbital nerve and decreased
sense of smell. To date, average follow-up on all patients is
17 months after encephalocele repair with 100% success.
Radiographic studies demonstrated an empty sella in all
four patients with available scans (Fig. 2C and D). No
patients had radiographic or clinical evidence of hydro-
cephalus or dilated ventricles. Postoperative CSF pres-
sures were measured in three of five patients after suc-
cessful repairing of the leaks. The mean ICP was 28.3 cm
water (range, 19–34 cm; normal range, 0–15 cm). One
patient refused lumbar puncture and another has been
lost to follow-up because of military service.
Postoperative Care
All patients were kept on a regimen of strict bedrest while
lumbar drains were in place (typically, for 2 or 3 d postoperative-
ly). After the lumbar drains were removed, patients gradually
resumed ambulation and light activity for 6 weeks after surgery.
Intravenous antibiotics, typically ceftriaxone because of its CSF
penetration, were used while patients were in the hospital. Oral
antibiotics were used for 2 or 3 weeks after discharge.
Patients were seen every 1 or 2 weeks postoperatively. Con-
servative endoscopic debridement was performed to maintain
patency of sinuses surrounding the defect and to avoid stasis of
secretions and bacterial infections. By 6 weeks after surgery,
most patients had returned to relatively normal activity levels
and little packing remained.
CASE REPORT
The case of one patient (L.A.) demonstrates some of the
challenges encountered in treating this population. This 60-year-
old woman was diagnosed with seizure disorder in 1979. She
developed meningitis in 1997, and a right-sided sphenoid CSF
RESULTS
Five patients had multiple simultaneous encephalo-
celes treated between 1996 and 2001. One of the 5 had
Fig. 1. Coronal computed tomography
(CT) scan (A) and T2-weighted magnetic
resonance imaging (MRI) scan (B) dem-
onstrating a right-sided supraorbital en-
cephalocele. The same patient also had
a large left-sided sphenoid meningoen-
cephalocele (C and D) repaired during
the same operation.
Laryngoscope 112: June 2002
982
Schlosser and Bolger: Multiple Encephaloceles

Fig. 2. Coronal CT scans demonstrating
a left-sided sphenoid lateral recess en-
cephalocele (A) that is being observed
and a right-sided sphenoid lateral recess
encephalocele (B) after endoscopic
transpterygoid repair. Neo-osteogenesis
and residual soft tissue fibrosis after
obliteration of the right-sided lateral re-
cess are evident. Coronal (C) and sagit-
tal (D) MRI scans of the same patient
demonstrate the empty sella (*). The
sella turcica is filled with cerebrospinal
fluid, and the residual pituitary gland is
compressed.
leak was diagnosed. She underwent a sublabial approach by an
outside neurosurgeon with closure of her leak. Nearly 3 years
later, she presented with left-sided CSF rhinorrhea and under-
went a transnasal and external ethmoidectomy approach com-
bined with a lumbar–peritoneal shunt to close the leak. However,
the leak persisted, and she presented to our institution with
left-sided rhinorrhea and a ␤-2 transferrin–positive test result for
CSF. Imaging studies demonstrated bilateral encephaloceles of
the lateral sphenoid recesses, a left-sided frontal/supraorbital
ethmoid encephalocele, and an empty sella. She underwent an
osteoplastic flap procedure, repair of skull base defect with mas-
toid bone graft and temporalis fascia, and left-sided frontal sinus
obliteration. Postoperative radionucleotide cisternogram demon-
strated a functioning lumbar–peritoneal shunt. One month later,
she underwent a successful left-sided transpterygoid approach to
close the lateral sphenoid recess encephalocele. Follow-up studies
have included a normal Schirmer test result for lacrimation,
normal-sized ventricles on CT, a lumbar puncture with opening
pressure of 19 cm H2O (with the functioning shunt), and a small,
asymptomatic stable encephalocele in the right-sided lateral re-
cess of the sphenoid. She has not exhibited any further CSF leaks
or neurological symptoms.
tization of the bony sinus walls creates sites of inherent
structural weakness within the skull base. Hydrostatic
pulsatile forces within the CSF may then lead to the
formation of small pits within the bone at the sites of
arachnoid villi. It is known that certain patients have
impaired CSF absorption and thus have accentuated
spikes in ICP. These significant fluctuations in the peak
CSF pressure at the site of the bony weakness permit
subsequent herniation of dura and brain tissue with the
development of encephaloceles and CSF leaks at these
points of attenuation.
In examining the presence of potential bony defects
as the first condition needed for CSF leak/encephalocele,
Shetty et al.7 studied the pneumatization of the sphenoid
sinus in 11 patients with spontaneous CSF leaks. The
authors found that 10 of 11 (91%) had pneumatization of
the lateral recesses in comparison with 23% to 43% of
normal subjects.8 In addition, 63% of their patients dem-
onstrated arachnoid pits within the bony skull base.
Ohnishi9 has shown that the ethmoid roof is dehiscent
14% of the time. Detailed studies on the pneumatization of
the frontal sinuses and associated potential sites of skull
base defects have not been made.
Given the prevalence of significant paranasal sinus
pneumatization with associated potential bony skull base
defects in the general population, patients with spontane-
ous CSF leaks/encephaloceles also need to have significant
variations in CSF pressure. Normal CSF pressures vary
significantly depending on age, activity, and time of day.
Normal intracranial pressures increase nearly three times
from infancy to adulthood.6 Pressures are also known to
rise from a level of approximately 15 mm Hg in normal,
awake patients to as high as 25 mm Hg during rapid-eye-
movement sleep.10 Thus, diagnostic lumbar punctures
performed routinely in awake patients represent only one
data point and do not detect fluctuations that occur
DISCUSSION
Spontaneous or nontraumatic, “normal pressure” en-
cephaloceles are uncommon entities. Three conditions are
necessary to have a cranionasal fistula: a bony defect,
rupture of the dural/arachnoid membrane, and a pressure
differential across the defect. The pathophysiology of
spontaneous encephaloceles or CSF leaks is unclear, but a
variety of explanations have been offered. Congenital
skull base defects are unlikely because pneumatization of
the frontal sinus and lateral sphenoid sinuses does not
occur until later in childhood. Another proposed theory is
that focal atrophy of the normal contents of the cribriform
plate or sella turcica may create potential spaces that can
be filled with enlarged pouches of normal arachnoid space
extensions.6 It seems more likely that significant pneuma-
Laryngoscope 112: June 2002
Schlosser and Bolger: Multiple Encephaloceles
983

throughout the night and during activities such as sneez-
ing or coughing. In addition, measures of ICP taken in
patients with active CSF leaks are probably artificially
lowered because of the chronic loss of CSF. For these
reasons, we have begun to consider the role of obtaining
postoperative ICP data after successfully repairing active
spontaneous CSF leaks. This may be especially useful in
patients with symptoms possibly attributable to elevated
intracranial pressures, such as headaches and pulsatile
tinnitus. Measuring the ICP postoperatively allows us to
obtain a more realistic idea of the patient’s true CSF
pressure, rather than obtaining pressures preoperatively
or intraoperatively that would be falsely lowered because
of the active leaks.
Cerebrospinal fluid pressures have demonstrated
transient elevations in other disease states, and there is
probably a spectrum of mild to severe disease similar to
that seen with vascular hypertension or glaucoma. Pa-
tients with benign intracranial hypertension (BIH) and
empty sella syndrome (ESS) may have an altered physi-
ology similar to that seen in patients with spontaneous
CSF leaks or encephaloceles. Patients with BIH are typi-
cally obese women who present with elevated CSF pres-
sures, papilledema, and headaches. Clark et al.11 reported
four patients treated for BIH who subsequently developed
spontaneous CSF rhinorrhea. All leaks occurred at the
cribriform plate, and the authors postulated that exagger-
ated CSF pulsatile flow leads to expansion and eventual
rupture of the arachnoid sleeve surrounding olfactory fil-
aments passing through the cribriform plate. Although
the four female patients in our series did not present with
typical BIH symptoms, all four were overweight or obese,
according to their BMI. It is possible that these patients do
not exhibit the symptoms of BIH when they are actively
leaking CSF and do not develop elevated ICP with asso-
ciated findings. It is our impression that they seem to
develop BIH symptoms after their CSF leaks have been
successfully patched.
Empty sella syndrome occurs when a weakened
sellar diaphragm permits herniation of the subarachnoid
space and its contents down into the sella turcica. This can
lead to compression of the pituitary gland and give the
radiological appearance of an empty sella. Although ESS
occurs most commonly after necrosis of a pituitary tumor,
it can be idiopathic and may represent a normal variant,
especially in elderly patients.12,13 Autopsy studies have
shown that 5.5% to 26% of people have a rudimentary
sellar diaphragm,12,14 but radiographic studies demon-
strate only a 5% to 6% incidence of empty sella in other-
wise normal patients.7,15 Primary, idiopathic ESS proba-
bly represents a milder form of intracranial hypertension
and may even occur with normal fluctuations in CSF
pulsations. Only 8% to 15% of patients with ESS have
BIH12,16 but, conversely, up to 94% of patients with BIH
can have ESS.17
Elevated ICP was accompanied by an absent or blunted
nocturnal prolactin level in every patient. Only 3 of 11
patients had normal intracranial pressures at all times,
and all of these had normal nocturnal prolactin spikes.
Davis and Kaye19 also used continuous ICP monitoring in
a patient with ESS and spontaneous CSF rhinorrhea to
show a slightly elevated ICP at baseline, but considerable
peaks. Garcia-Uria et al.20 reported on seven patients
with spontaneous CSF rhinorrhea and ESS. Six of seven
were women with no signs of elevated ICP.
In addition to a weakened sellar diaphragm, patients
with ESS appear to have impaired circulation of CSF.
Abnormal CSF absorption occurs in 80% to 84% of pa-
tients with ESS. This probably occurs because of blockage
of the arachnoid villi over the convexities.21,22 Clinical
signs of papilledema and CSF leaks appear to correlate
with actual increases in ICP and may be relieved after
shunting procedures.18 Patients with ESS and no clinical
signs of elevated ICP may simply have a milder form of
intracranial hypertension that we are unable to detect.
Spontaneous cranionasal fistulas represent a unique
and challenging population. In contrast to patients with
traumatic or postsurgical CSF leaks, patients with spon-
taneous encephaloceles and CSF leaks usually have
broadly attenuated skull bases with large defects, proba-
bly because of fluctuations in their CSF pressure as de-
scribed earlier in the present study. For these reasons, we
are more aggressive in the placement of lumbar drains
during the initial perioperative period and in limiting
their activity postoperatively. Although only one patient
in our series had a shunt in place, additional consideration
may be given to long-term shunt placement in an attempt
to minimize large fluctuations in ICP. Recently, we have
started using acetazolamide for 6 weeks after surgery to
decrease CSF production and hydrostatic pressures dur-
ing the early healing process.
Treatment of patients with nasal encephaloceles and
CSF leaks has changed since the early 1980s. Successful
endoscopic repair of these lesions allows the otolaryngol-
ogist to treat most of these defects without the morbidity
of intracranial procedures. Traditional neurosurgical ap-
proaches in this population have demonstrated good ini-
tial success of 89% but a recurrence rate of up to 50% with
some cases recurring years after their initial repair.15
This emphasizes the need for long-term follow-up for these
patients because of their potential to develop subsequent
leaks at the repair site or other areas of the attenuated
skull base. The issue also arises in these patients as to
which encephalocele to repair first and whether all en-
cephaloceles need to be repaired. It was our treatment
philosophy to repair all actively leaking encephaloceles to
avoid the risk of meningitis or pneumocephalus. Encepha-
loceles were also repaired if they were a likely source of
symptoms such as seizures. We typically only repaired one
encephalocele during each surgery because of the signifi-
cant length of time needed for this complex operation. The
one exception was the repair of two encephaloceles in
patient F.L., who had an active leak from a right-sided
supraorbital encephalocele and a large left-sided enceph-
alocele filling the entire sphenoid (Fig. 1). The repair of
the right-sided leaking defect was made relatively quickly,
The majority of patients with ESS do not have typical
signs or symptoms of elevated intracranial pressures,12,18
but Maira10 performed continuous ICP monitoring on 11
patients with primary ESS. Three patients had elevated
ICP while awake, and an additional five patients had high
intracranial pressures while in rapid-eye-movement sleep.
Laryngoscope 112: June 2002
984
Schlosser and Bolger: Multiple Encephaloceles

5. Bolger WE, Osenback R. Endoscopic transpterygoid approach
to the lateral sphenoid recess. Ear Nose Throat J 1999;78:
36–43.
6. Ommaya AK, Di Chiro G, Baldwin M, Pennybacker JB. Non-
traumatic cerebrospinal fluid rhinorrhoea. J Neurol Neu-
rosurg Psychiatry 1968;31:214–225.
7. Shetty PG, Shroff MM, Fatterpekar GM, Sahani DV, Kirtane
MV. A retrospective analysis of spontaneous sphenoid si-
nus fistula: MR and CT findings. AJNR Am J Neuroradiol
2000;21:337–342.
8. Bolger WE, Butzin CA, Parsons DS. Paranasal sinus ana-
tomic variations and mucosal abnormalities: CT analysis
for endoscopic sinus surgery. Laryngoscope 1991;101:
56–64.
and adequate grafting material had been harvested for
repair of the second encephalocele; therefore, we opted to
perform both repairs at the same time. We have elected to
observe three small, nonleaking encephaloceles in the se-
ries and follow them with serial imaging. It is not yet clear
whether patching a defect at one site increases the likeli-
hood of developing a subsequent CSF leak at another site.
We are also investigating long-term intracranial pres-
sures in this population by obtaining opening pressures
during lumbar punctures after successful patching and
after patients have been given adequate time to achieve a
homeostatic state. This may enable us to more accurately
determine the need for long-term shunts or acetazolamide
in these patients.
9. Ohnishi T. Bony defects and dehiscences of the roof of eth-
moid cells. Rhinology 1981;19:195–202.
10. Maira G, Anile C, Cioni B, et al. Relationships between in-
tracranial pressure and diurnal prolactin secretion in pri-
mary empty sella. Neuroendocrinology 1984;38:102–107.
11. Clark D, Bullock P, Hui T, Firth J. Benign intracranial hy-
pertension: a cause of CSF rhinorrhoea. J Neurol Neuro-
surg Psychiatry 1994;57:847–849.
12. Bjerre Per. The empty sella: a reappraisal of etiology and
pathogenesis. Acta Neurol Scand 1990;82(Suppl 130):
6–24.
13. Hardjasudarma M, White KE, Nandy I, Burns PL. Sellar
emptiness on routine magnetic resonance imaging. South
Med J 1994;87:340–343.
14. Brodsky L. Spontaneous cerebrospinal fluid otorrhea and
rhinorrhea coexisting in a patient with meningitis. Laryn-
goscope 1984;94:1351–1354.
15. Garcia-Uria J, Ley L, Parajon A, Bravo G. Spontaneous ce-
rebrospinal fluid fistulae associated with empty sellae: sur-
gical treatment and long-term results. Neurosurgery 1999;
45:766–774.
16. Kaye AH, Tress BM, Brownbill D, King J. Intracranial pres-
sure in patients with the empty sella syndrome without
benign intracranial hypertension. J Neurol Neurosurg Psy-
chiatry 1982;45:209–216.
17. Zagardo MT, Cail WS, Kelman SE, Rothman MI. Reversible
empty sella in idiopathic intracranial hypertension: an in-
dicator of successful therapy? AJNR Am J Neuroradiol
1996;17:1953–1956.
CONCLUSION
Patients with spontaneous encephaloceles and CSF
leaks represent a challenging population given the under-
lying nature of their disease. The pathophysiology of this
disease probably begins with impaired CSF absorption
that enhances the pulsatile nature of the intracranial
pressures. These elevated forces widely attenuate the
skull base and lead to subsequent herniation of the intra-
cranial contents through inherently weakened areas, such
as the sellar diaphragm or widely pneumatized sinuses.
Radiographic evidence of empty sella and extensive pneu-
matization of the paranasal sinuses should raise the sus-
picion that multiple spontaneous cranionasal fistula may
be present. Endoscopic repair and postoperative manage-
ment in these patients is more difficult than in patients
with encephaloceles or CSF leaks from other causes, and a
comprehensive treatment plan may eventually require
shunting or acetazolamide to prevent further recurrences.
Long-term follow-up and investigation of the intracranial
pressures in this population will enable us to better care
for these patients.
18. Spaziante R, de Divitiis E, Stella L, Cappabianca P, Genovese
L. The empty sella. Surg Neurol 1981;16:418–426.
19. Davis S, Kaye AH. A dynamic pressure study of spontaneous
CSF rhinorrhea in the empty sella syndrome. J Neurosurg
1980;52:103–105.
20. Garcia-Uria J, Carrillo R, Serrano P, Bravo G. Empty sella
and rhinorrhea: a report of eight treated cases. J Neuro-
surg 1979;50:466–471.
BIBLIOGRAPHY
1. Zlab MK, Moore GF, Daly DT, Yonkers AJ. Cerebrospinal
fluid rhinorrhea: a review of the literature. Ear Nose
Throat J 1992;71:314–317.
2. Lanza DC, O’Brien DA, Kennedy DW. Endoscopic repair of
cerebrospinal fluid fistulae and encephaloceles. Laryngo-
scope 1996;106:1119–1125.
3. Mattox DE, Kennedy DW. Endoscopic management of cere-
brospinal fluid leaks and cephaloceles. Laryngoscope 1990;
100:857–862.
21. Brismar K, Bergstrand G. CSF Circulation in subjects with
the empty sella syndrome. Neuroradiology 1981;21:
167–175.
22. Maira G, Anile C, De Marinis L, Mancini A, Barbarino A.
Cerebrospinal fluid pressure and prolactin in empty sella
syndrome. Can J Neurol Sci 1990;17:92–94.
4. Burns JA, Dodson EE, Gross CW. Transnasal endoscopic
repair of cranionasal fistulae: a refined technique with
long-term follow-up. Laryngoscope 1996;106:1080–1083.
Laryngoscope 112: June 2002
Schlosser and Bolger: Multiple Encephaloceles
985