ORIGINAL PAPER

Endoscopic endonasal approaches to anterior skull base

defects in pediatric patients

Davide Locatelli & Federico Rampa & Ilaria Acchiardi &

Maurizio Bignami & Andrea Pistochini &

Paolo Castelnuovo

Received: 22 December 2005# Springer-Verlag 2006

Abstract

Introduction We studied 12 pediatric patients with congen-

ital or acquired anterior skull base defects. All subjects

underwent surgery owing to progressive symptoms. The

endoscopic endonasal approach is a new method in the

treatment of this pathology in children.

Materials and methods Twelve children had surgery to

correct anterior skull base defects: seven patients with a

spontaneous anterior basal meningoencephalocele and five

with posttraumatic cerebrospinal fluid (CSF) leakage. The

defects were repaired using the endoscopic endonasal

approach, which combined with the fluorescein diagnostic

test, detects the exact location of the skull base defect.

Different closure techniques were used to obtain a

permanent graft, depending on the type, location, and size

of the defect. An intraoperative fluorescein test confirmed

the absence of CSF leakage after surgery.

Results The follow-up period ranged from 3 to 72 months.

Symptoms resolved in all patients after surgery and none of

them experienced complications or recurrence of CSF

leakage. Postoperative magnetic resonance scans showed

that the defect had successfully been repaired in all patients.

Discussion The surgical treatment of skull base defects in

children reduces life-threatening risks, which include

infections, CSF leaks, and enlargement or trauma of the

sac. The endoscopic technique minimizes surgical scars and

has little impact on brain tissue. The endoscopic endonasal

approach to the anterior skull base helps to preserve the

physiology of the nose and sinuses and reduces the impact

on the still developing splanchnocranium in pediatric

patients. It ensures a definitive repair of the defect and

requires a very short inpatient period.

Keywords Endoscopy . Endonasal surgery .

Meningocele . Rhinorrea . Skull base .

Congenital defect . Cerebrospinal fluid leaks . Fistula

Introduction

Anterior skull base defects in children can be congenital

(meningoencephaloceles) or posttraumatic (skull base frac-

tures). Anterior basal meningoencephaloceles are rare,

representing 10–20% of all craniospinal dysraphisms [1]

with an incidence of 1:5,000 live births, as reported by

Suwanwela and Suwanwela [2]. Nasoethmoidal encephalo-

celes protrude through a defect between the nasal bone and

nasal cartilage; the most common symptoms are progres-

sive nasal obstruction with mouth breathing, snoring, and

nasal discharge due to the projection of the herniated mass

into the nasopharynx [1]. Transsphenoidal encephaloceles

are associated with more serious abnormalities of the

midline of the brain and sometimes involve vital structures

[3, 4].

Posttraumatic cerebrospinal fluid (CSF) leaks in chil-

dren, with an incidence of 0.2–0.3% of all head traumas in

Childs Nerv Syst

DOI 10.1007/s00381-006-0114-7

D. Locatelli (*) : F. Rampa : I. Acchiardi

Department of Neurosurgery, IRCCS Policlinico S. Matteo,

University of Pavia,

27100 Pavia, Italy

e-mail: dlocatelli@smatteo.pv.it

M. Bignami : P. Castelnuovo

Department of Otolaryngology, Ospedale di Circolo,

University of Insubria,

Varese, Italy

A. Pistochini

Department of Otolaryngology,

University of Sassari,

Sassari, Italy

childhood, are less common than in adults. The greater

plasticity of the bone and the incomplete development of

the paranasal sinuses mean that the traumatic event can be

absorbed. Posttraumatic cerebrospinal fluid leaks occasion-

ally cause intermittent rhinorrhea, and the only symptom

may be recurrent episodes of meningitis.

Early surgical repair is necessary to avoid the risk of

infection and progressive breathing impairment in congen-

ital meningoencephaloceles.

We report our results on 12 children with anterior skull

base defects, seven with congenital meningoencephaloceles

and five with posttraumatic lesions and CSF leaks. All

operations were performed using an endonasal endoscopic

technique.

Materials and methods

We operated on 135 patients with anterior skull base defects

between May 1995 and May 2004. Endonasal endoscopic

repair was carried out in 121 cases, and a combined

intracranial–extracranial approach was used in 14. The

present study included 12 children, six males and six females,

whose ages ranged from 1 to 18 years (mean age 8 years).

Seven children, three males and four females, had a

congenital meningoencephalocele, which was clinically

revealed by nasal obstruction (four cases) or recurrent

meningitis (three cases). The malformation involved the

cribriform plate in five cases, the ethmoid and the sphenoid

sinus in one case, respectively (Table 1). Five children,

three males and two females, sustained head injury and

CSF leakage after being involved in a car accident (three

cases), after being kicked by a horse (one case), and after

heading a ball (one case). The lesion was confined to the

ethmoid in one patient but was multiple in the remaining

cases: two children had lesions in the ethmoid and

cribriform plate, one in the ethmoid and the sphenoid

sinus, and one in the ethmoid and the posterior wall of the

frontal sinus. Rhinorrhea was the initial symptom in these

children and was associated with headache in one case

(Table 2). Two of these patients had previously been treated

at other institutions: a 12-year-old girl for a meningo-

encephalocele in the sphenoid sinus and a 12-year-old boy

for posttraumatic CSF leakage.

Diagnostic tests included laboratory tests, endoscopic

evaluation, computed tomography (CT) and magnetic

resonance (MR) imaging studies.

Beta-2-transferrin detection in fluid samples collected

from the nose was used as a diagnostic test when active

rhinorrhea was present [5].

We performed endoscopic endonasal evaluation to

visualize CSF leaks or monolateral nasal obstruction, as

well as to rule out any other nasal anatomical variations

(Fig. 1).

Spiral CT, with a bone and high-definition multiplanar

reconstruction (MPR) algorithm, was performed with the

Table 1 Symptoms presented by patients with congenital skull base defects

Case Age Sex Onset sign Location of the skull

defect

Side MEP Previous attempts to

repair the CSF leak

FU

1 1 F Nasal obstruction Cribriform plate Left Yes No 19

2 1 M Nasal obstruction Cribriform plate Left Yes No 12

3 3 M Nasal obstruction Cribriform plate Right Yes No 38

4 4 F Nasal obstruction Cribriform plate Right Yes No 69

5 8 M Recurrent meningitis (four episodes) Cribriform plate Left Yes No 12

6 14 F Recurrent meningitis (three episodes) Ethmoid Right Yes No 30

7 12 F Recurrent meningitis (four episodes) Sphenoid, pterygoid Left Yes Yes 13

FU Follow-up (months), MEP meningoencephalocele

Table 2 Symptoms presented by patients with posttraumatic anterior skull base defects

Case Age Sex Etiology Onset sign and

symptom

Location of the skull

defect

Side Previous attempts

to repair the CSF leak

FU

1 4 F Horse kick Rhinorrhea Ethmoid Left No 63

2 11 M Header Rhinorrhea Ethmoid, cribriform plate Left No 71

3 12 M Car accident Rhinorrhea Ethmoid, sphenoid sinus Right Yes 29

4 18 M Car accident Rhinorrhea Ethmoid, cribriform plate Right No 81

5 17 F Car accident Rhinorrhea with headache Ethmoid, posterior wall

of frontal sinus

Left No 14

FU Follow-up (months)

Childs Nerv Syst

patient in a supine position. This provided detailed studies

in the axial, coronal, and sagittal planes, which localized

the bony defect. CT cisternography was not performed.

MRI with 3-mm slices was carried out with T1- and T2-

weighted sequences and in the fluid attenuated inversion

recovery sequence when necessary. Contrast enhancement

was used to better define the content of the sac in the event

of masses projecting into the nasal cavity.

All posttraumatic patients had a preoperative fluorescein

test to confirm CSF leakage and, when possible, to pinpoint

the lesion, in accordance with Stammberger [6]; we also

took CSF samples for laboratory tests.

All patients underwent endoscopic endonasal repair of

the skull base defect.

The surgical procedure was performed under general

anesthesia and blood pressure was controlled. After

inducing anesthesia, the intrathecal contrast was injected,

following Stammberger’s guidelines, before graft place-

ment in the definitive surgical position [6, 7].

All surgery was performed with the patient lying on his

back with his head slightly extended and turned to the right,

facing the surgeon; head clamps were not used. The child’s

face was cleansed and covered with a sterile cloth with a hole

for the nose. Tampons soaked with topical anesthetic and

vasoconstrictor drugs were inserted into both nostrils using

bayonet forceps [8, 9]. The tampons were removed after 5 to

10 min, and local anesthetic drugs were then injected

submucosally, under endoscopic vision, to facilitate the

detachment of the nasal mucosa from cartilage or bone surface.

The direct paraseptal endonasal approach was performed

in five patients with a meningoencephalocele in the

cribriform plate. In these cases, the middle turbinate had

been pushed aside by the protruding mass and was thus

conserved. A specific blade was used to separate the

mucous membranes of the septal and lateral walls as far as

the subperiosteal plane to reach the pedicle of the lesion. The

extruding dural wall was dissected from the bone and

mucoperichondrium, fully exposing the defect in the bone;

Table 3 Surgical repair of congenital anterior skull base defects

Case Location MEP Previous attempts to

repair the CSF leak

Graft Graft material FU

1 Cribriform

plate

Yes No Combined Dural substitute + cartilage (underlay) and mucoperichondrium

(overlay)

19

2 Cribriform

plate

Yes No Composite Septal mucoperiostium + cartilage 12

3 Cribriform

plate

Yes No Combined Dural substitute + cartilage + septal mucoperiostium 38

4 Cribriform

plate

Yes No Free

simple

Abdominal fat 69

5 Cribriform

plate

Yes No Combined Middle conca bone + dural substitute + mucosa of conche bullosa 12

6 Ethmoid Yes No Combined Dural substitute + middle conca bone + dural substitute (underlay)

and middle turbinate mucoperiostium (overlay)

30

7 Sphenoid,

pterygoid

No Yes Combined Dural substitute + cartilage (underlay) and mucoperichondrium

(overlay)

13

FU Follow-up (month)

Fig. 1 a T1 coronal MR

study showing a congenital

meningoencephalocele in

the right nasal fossa, which

dislocates the nasal

structures. b Endoscopic

appearance

Childs Nerv Syst

the meningeal layer was then opened and the contents of the

sac, if judged to be still functional, were replaced inside the

skull base using an appropriate spatula. We used a combined

graft to repair the defect in four cases and a free simple graft

with abdominal fat in one patient (Table 3).

An ethmoidectomy was carried out when the defect was

localized in the ethmoid or if the defect involved both the

cribriform plate and the ethmoid. The mucoperichondrium

surrounding the defect was detached, and bipolar forceps

were used to reinsert the meningeal protrusion. A combined

graft was used in these cases (Table 4).

The transethmoidal–pterygoidal–sphenoidal approach

was used in one case to correct a meningoencephalocele

of the lateral sphenoid sinus. The ethmoidectomy was

combined with the excision of the middle turbinate and a

wide antrostomy to identify the posterior wall of the

mascellar sinus and the pterygoid that was finally drilled

out to expose the lateral wall of the sphenoid sinus. An

intraoperative fluorescein test confirmed the absence of

CSF leakage in all cases. The grafts were then fixed into

position with fibrin glue and held in place by hemostatic

fibrillary sponges or Vaseline gauzes [10, 11]. Nasal

packing was left in place for about 2 days.

All patients underwent postoperative endoscopic evalu-

ation and MR imaging 1 month after surgery. The MR test

was performed under sedation (depending on the patient’s

age), during which endoscopic endonasal evaluation was

also performed.

Results

Seven children had congenital defects of the anterior skull

base. On endoscopic endonasal evaluation, the four

children with monolateral nasal obstruction had a smooth,

whitish-gray mass protruding into the nasal cavity. Of the

three patients with recurrent meningitis, one had a small

pulsating mass in the anterior portion of the olfactory

groove, another did not present abnormalities, while the

patient who had previously undergone surgery at another

hospital had a pulsation in the pterygoid area, which

synchronized with his heart beat.

The five children with posttraumatic skull base lesions

presented with monolateral rhinorrhea. On endoscopic endo-

nasal evaluation, four patients were found to have a normal

anatomy, while one patient had a meningoencephalocele.

Beta 2 transferrin was tested in all patients with active

rhinorrhea; all tested cases were positive.

Complete imaging studies were performed in all cases:

CT delineated bony structures while MR identified the

content of the sac and checked for the presence of other

brain abnormalities [12].

An intraoperative fluorescein test allowed us to pinpoint

the leak in 11 of the 12 children [9]: endoscopic endonasal

exploration was performed on all our patients. The direct

paraseptal endonasal approach was used in five patients with

a meningoencephalocele in the cribriform plate. An ethmoi-

dectomy was carried out in one patient with a small

ethmoidal meningoencephalocele and in five patients with

posttraumatic CSF leaks (the defect was confined to the

ethmoid in the 4-year-old girl, while the remaining four

cases presented multiple defect sites). The transethmoidal–

pterygoidal–sphenoidal approach was used in one patient

with a meningoencephalocele of the lateral sphenoid sinus.

Intraoperative fluorescein tests confirmed that the defect had

successfully been repaired in all cases with no CSF leakage.

None of our patients developed complications or

required repeat surgery.

Table 4 Surgical repair of posttraumatic anterior skull base defects

Case Location of the skull

defect

Side Graft Graft material Previous attempts to repair the

CSF leak

FU

1 Ethmoid Left Combined Middle turbinate mucoperiostium + middle

conca bone

No 63

2 Ethmoid, cribriform

plate

Left Composite Dural substitute + middle turbinate No 71

3 Ethmoid, sphenoid

sinus

Right Combined Middle turbinate + abdominal fat Yes 29

4 Ethmoid, cribriform

plate

Right Combined Dural substitute + middle turbinate No 81

5 Ethmoid, posterior

wall of frontal sinus

Left Combined (Obliterative technique) dural substitute,

abdominal

fat, and turbinate bone in the sphenoid

No 14

Free simple graft of mucoperiostium from

the middle

turbinate in the pterygoid

FU Follow up (months)

Childs Nerv Syst

Mean hospitalization was about 1 week and mean

duration of follow-up was 36 months, with a range of 12

to 72 months. Normal MRI confirmed the total resolution

of symptoms (Fig. 2).

An outpatient endoscopic endonasal evaluation 1

month after the operation revealed that the skull

base defect had fully healed and that nasal mucosa had

returned to normal (Fig. 3). Breathing had also normalized.

Discussion

The diagnosis and treatment of anterior skull base defects is

a challenging problem [8]: the clinical suspicion of

meningoencephaloceles in children arises in the event of

monolateral nasal obstruction, while rhinorrhea and recur-

rent meningitis may suggest the presence of CSF leakage.

Endonasal evaluation is one of the first diagnostic tests

we routinely perform, using 2.7-mm diameter endoscopes

with 30° lenses. This procedure is easy to perform during

the MRI study and avoids having to resedate patients; it

allows us to evaluate the morphology of the nasal cavity

and to directly visualize active rhinorrhea or the presence of

a pulsating mass projecting into the nose.

The B2 transferrin dosage in the liquid sample collected

from the nose has a sensitivity of 97.7% [13–15]; this test is

rapid and requires just few milliliters of liquid [15–17]. One

limit of this test is that rhinorrhea must be active, and even

in cases of persistent rhinorrhea, it is not always easy to

collect the fluid sample, especially from pediatric patients.

Nowadays, spiral CT is the best imaging study for

localizing skull base defects as it is possible to obtain very

thin slices (0.5 mm) with a bone algorithm. The examina-

tion is very short so pediatric patients do not have to be

anesthetized. High-definition MPR allows for a complete

study in the axial, coronal, and sagittal planes in a supine

position; cistern CT, with the patients lying face down, is

used to localize the CSF leak in children with active

rhinorrhea after dye injection. CT evaluation of the

nasofrontal region is often difficult in the neonate because

the anterior skull base is largely cartilaginous at birth. It is

crucial to understand the CT appearance of this region in

the normal neonate and infant.

CT may be inadequate because small defects can be

missed through the partial volume effect, while the

detection of a small opening is not necessarily related to a

connection between the nasal mass and the intracranial

compartment. In difficult cases, acquisition of CT images

Fig. 3 a Endoscopic endonasal

repair of a meningoencephalo-

cele. b Outpatient endoscopic

evaluation after 3 months

Fig. 2 a Preoperative T1

sagittal MR study of a

congenital meningoencepha-

locele. b Postoperative T1

sagittal MR study after

endonasal endoscopic repair

Childs Nerv Syst

through the defect after intrathecal contrast administration

may determine soft tissue continuity with the subarachnoid

space; sagittal reconstruction of 1-mm images is optimal

[12]. One-millimeter axial and coronal CT is performed.

The coronal section has to be perpendicular to the hard

palate to show the entire temporal bone and exclude a

defect in the tegmen timpani or roof of the mastoid: in fact,

CSF flow from the middle ear or the mastoid through the

eustachian tube could simulate rhinorrhea. The CT pattern

can indicate whether there is a protrusion of cerebral tissue

into a meningeal sac or a tumor growing at the base. High-

resolution, 3-mm, axial, coronal, and sagittal MRI in T1-

and T2-weighted sequences is mandatory to formulate a

differential diagnosis. MRI is more sensitive than CT in

detecting a continuity between the nasal and cranial cavities

when there is a lesion with a very narrow neck; the

examination is also valuable in differentiating herniated

brain from inflammatory nasal masses, nasal gliomas, or

nasal dermal sinus cysts [1]. The signal intensity of nasal

gliomas is similar to that of both epidermoid tumors and

brain tissue; therefore, anatomic factors are more important

than signal characteristics in distinguishing between these

malformations. Meningoceles have a typical low signal in

T1 and a hyperintense signal in T2-weighted sequences.

Meningoencephaloceles have a typical brain-like signal of

the content in T2 sequences. Negative contrast enhance-

ment (gadolinium DPTA) rules out tumors. During neuro-

radiological studies, the middle and posterior cerebral fossa

are checked to exclude other congenital defects [18].

It is important to avoid any kind of biopsy during this

diagnostic phase; in the event of encephaloceles, this

manoeuvre could have serious consequences, namely,

cerebral injury, infections or rhinorrhea [19, 20].

All outpatients with posttraumatic CSF fistulas had a

preoperative fluorescein test: the fluorescein dyes the

cerebrospinal fluid, which becomes directly visible through

the endoscope, identifying the exact site of the leak.

Identification is aided by the use of a blue optical filter

system introduced into the light source for the endoscopic

equipment [9]. The fluorescein test was unsuccessful in

only one child with associated spinal trauma that may have

limited the normal diffusion of the dye. The fluorescein test

is essential to localize multiple CSF leaks. If the test is

positive, then surgery is mandatory.

The intraoperative fluorescein test was performed on all

our patients, allowing us to visualize the exact site of the

CSF rhinorrhea. No complications arose after injecting the

dy,e and the test appeared to be safe if performed following

the Graz school guidelines [7].

The direct paraseptal approach is performed in the event

of a meningoencephalocele in the olfactory groove that

lateralizes the middle turbinate; no other surgical procedure

on the ethmoid is carried out in these cases. In fact, if the

space on the cribriform plate is large enough to position the

graft, the middle turbinate is preserved and an autologous

graft is harvested from the nasal septum. The first step in

this surgical procedure is to reach and isolate the pedicle.

We found that a small neck is easier to repair not only

because of its size but also because the sac does not allow

important parts of the overlying central structures to

herniate. It is important to identify the correct layer to

separate the dura from the bone and perform the anatomical

dissection. The dura is often tightly adherent to the

underlying mucosa. The dura mater must be detached

thoroughly from the surrounding bone to fix the graft and

prevent recurrences during the child’s growth. The intra-

dural detachment of cerebral tissue adherence is obtained,

and a dural substitute is positioned using the underlay

technique under the bone (intracranially) and over the dura.

Thanks to the submucous access, the endoscope makes it

possible to dissect scarring tissue smoothly, minimizing

impact on the brain and conserving nasal and sinus mucosa.

An ethmoidectomy, with the excision of the middle

turbinate, is performed when the ethmoidal roof has to be

fully exposed to localize the CSF leak. This method is used

mainly to repair posttraumatic CSF leaks. Traumatic CSF

leaks involve multiple defect sites and the entrapment of

the dura mater in the rim of the fracture. In these cases, the

intraoperative fluorescein test is valuable to localize all the

defects, and the ethmoidal roof has to be drilled with a

diamond burr to completely expose the leak. The middle

turbinate is generally excised and used to harvest the bone

or the mucoperichondrial graft. The tear in the dura mater is

fully exposed, and all mucosa must be carefully removed

from the surrounding bone to fix the mucoperiosteum or

mucoperichondrial graft (free simple graft) over the bone

(overlay technique) and to prevent recurrences during the

child’s growth.

The transethmoidal–pterygoidal–sphenoidal approach is

used for defects in the lateral wall of the sphenoid sinus. We

used this approach in one patient, who had previously

undergone surgery at another hospital, to remove a left

parasphenoidal CSF cyst, position a subdural–peritoneal

drain, and carry out extracranial plastic repair of the skull

base of the middle fossa. Two CSF leaks were detected after

the pterygoid and anterior wall of the sphenoid sinus had

been fully exposed: one in the paralateral left side of the

sphenoid sinus, probably because the original graft onto

the skull base of the middle fossa had shrunk, and one in the

pterygoid due to abnormal pneumatization. After drilling

the bone to expose the two defects, the leak in the sphenoid

sinus was repaired with a dural substitute, abdominal fat, and

turbinate bone, while the leak in the pterygoid was repaired

with a free simple graft of mucoperiostium from the middle

turbinate. We believe that the choice of graft is essential for a

successful duraplasty. We generally use an autologous graft

Childs Nerv Syst

harvested from the nose to achieve a highly compatible

duraplasty. The mucoperiostium, from the middle turbinate,

and the mucoperichondrium, from the nasal septum, are

fashioned and positioned to fit the defect. Particular care

must be taken when implanting the graft: the mucosal part to

the nasal cavity and the periosteum to the skull base. It is

crucial not to obstruct the natural ostia of the paranasal

sinuses so as to respect and preserve normal sinus function.

Harvesting the graft from the nose avoids incising the

abdomen to obtain abdominal fat. If the area in the olfactory

groove is wide enough to position the graft, as in children

with a meningoencephalocele in the cribriform plate that

pushes the middle turbinate to one side, the middle turbinate

is preserved. The graft taken from the nasal septum is then

positioned with the underlay technique, and the mucoperi-

chondrium is used as overlay. Different nasal tissues and

other autologous tissue can be used to close the skull base in

particular situations [7, 21, 22].

We fixed the grafts with fibrin glue and hemostatic

sponges and did not observe any cases of graft displacement.

The absence of fluorescein flow through the graft during

the endoscopic test, using the Valsalva maneuver, reassures

the surgeon that the fistula was effectively repaired.

The main alternative to the endoscopic endonasal

technique is the transcranial approach.

A transcranial approach through a unilateral or bicoronal

bone flap allows the neck of the protruding brain tissue to be

isolated but requires a wider exposure with the retraction of

one or both frontal lobes. The intracranial repositioning of

the extruded tissue is difficult, and the repair of the dural

defect requires a larger graft. It is also difficult to control the

entire skull base defect without visualizing it directly. If an

endonasal approach is not used, the meningocele sac has to

be left in place, possibly causing breathing impairment and

requiring a second surgical procedure.

The limits of the extracranial endoscopic endonasal

approach depend on the characteristics of the endoscope:

loss of the third dimension in the surgical field and a long

learning curve. Failure of an endoscopic approach may

relate to the inability to successfully localize the defect,

inadequate preparation of the defect area for the graft

placement, the nature and amount of CSF leakage, the

nature and size of the graft, as well as graft displacement or

incomplete apposition of the graft to the skull base [23, 24].

The extension of the lesions can also be a limit for this

approach: fractures extending cranially or laterally out of

endoscopic view require a combined approach. A necrotic

or fibrotic meningoencephalocele or a large CSF-filled sac

may both be treated endoscopically: in the latter case by

draining and reducing the dimension of the sac and by

debriding the sac with its nonfunctional content in the

former. Large hernias with facial deformities require

combined transcranial endonasal approaches.

The goal of the endoscopic transnasal technique in

pediatric patients with a transethmoidal defect is to ensure a

stable duraplasty with minimally invasive surgery, avoiding

craniotomy. This less invasive technique is well-tolerated

by the child as clinical recovery is rapid and it requires only

a short stay in hospital.

Outpatient endoscopic evaluation, carried out 1 month

after the surgical procedure, usually demonstrates that nasal

mucosa has returned to normal without scars.

In conclusion, in children, the endoscopic endonasal

approach achieves the definitive repair of an anterior skull

base defect and preserves the normal function of the nose and

air sinuses with a very short inpatient period. It also reduces

the impact on the still-developing splanchnocranium.

References

1. Macfarlane R, Rutka JT, Armstrong D (1995) Encephaloceles of

the anterior cranial fossa. Pediatr Neurosurg 23:148–158

2. Suwanwela CI, Suwanwela NA (1972) Morphological classifica-

tion of sinciptial encephaloceles. J Neurosurg 36:201–211

3. Kai Y, Nagahiro S, Yoshioka S (1996) Application of the skull

base technique to the repair of transsphenoidal meningoencephalo-

celes. Pediatr Neurosurg 25:54–56

4. Cassiano RR, Jassir D (1999) Endoscopic cerebrospinal fluid

rhinorrhea repair: is a lumbar drain necessary? Otolaryngol Head

Neck Surg 121(6):745–750 (Dec)

5. Meco C, Oberascher G, Arrer E, Moser G, Albegger K (2003)

Beta-trace protein test: new guidelines for the reliable diagnosis of

cerebrospinal fluid fistula. Otolaryngol Head Neck Surg 129

(5):508–517 (Nov)

6. Stammberger H (1997) Surgical occlusion of cerebrospinal

fistulas of anterior skull base using intrathecal sodium fluorescein.

Laryngorhinootologie 76:595–607

7. Wolf G, Greistorfer K, Stammberger H (1997) Der endoskopishe

Nechweis von Liquorfisteln mittels der Fluoreszeintechnik

(in German). Laryngo- Rhino- Otol 76(10):588–594 (Oct)

8. Castelnuovo P, Mauri S, Locatelli D (2001) Endoscopic repair of

cerebrospinal fluid rhinorrhea: learning from our failure. Am J

Rhinol 15:333–342

9. Locatelli D, Castelnuovo P, Santi L, Cerniglia M, Maghnie M, Infuso

L (2000) Endoscopic approaches to the cranial base: prospective and

realities. Childs Nerv Syst 16(10–11):686–691 (Nov)

10. Bachert C, Verhaeghe B, van Cawenberge P, Daele J (2000)

Endoscopic endonasal surgery (EES) in skull base repair and CSF

leakage (Review). Acta Otorhinolaryngol Belg 54(2):179–189

11. Bibas AG, Skia B, Hickey SA (2000) Transnasal endoscopic

repair of cerebrospinal fluid rhinorrhea. Br J Neurosurg 14(1):

49–52 (Feb)

12. Barkovich AJ (ed) (2000) Congenital malformation of the brain

and skull. In: Pediatric neuroimaging, 3rd edn. Lippincott

Williams & Wilkins, Philadelphia, pp 271–276 (chapter 5)

13. Oberascher G. 1988. A modern concept of cerebrospinal fluid

diagnosis in oto- and rhinorrhea. Rhinology 26(2):89–103 (Jun)

14. Vernocchi A (2004) Beta 2 transferrina. In: Neurotraumatology

symposium of the 53th national Congress of the Italian society of

neurosurgery, Milan, November 21–24, 2004

15. Skedros DG, Cass SP, Hirsch BE, Kelly RH (1993) Beta-2

transferrin assay in clinical management of cerebrospinal fluid and

perilymphatic fluid leaks. J Otolaryngol 22:314–344

Childs Nerv Syst

16. Meurman O, Irjala K, Suonpaa J, Laurent B (1979) A new method

of identification of cerebrospinal fluid leakage. Acta Otolaryngol

87(3–4):366–369 (Mar–Apr)

17. Ryall RG, Peacock MK, Simpson DA (1992) Usefulness of beta-

2-transferrin assay in the detection of cerebrospinal fluid leaks

following head injury. J Neurosurg 77(5):737–739 (Nov)

18. Schick B, Draf W, Kahle G, Weber R, Wallenfang T (1997) Occult

malformations of the skull base. Arch Otolaryngol Head Neck Surg

123:77–80

19. Dodge HW Jr, Love JG, Kernohan JW (1959) Intranasal

encephalomeningoceles associated with cranium bifidum. Arch

Surg 79:75–84

20. Schmidt PH, Luyendijk W (1974) Intranasal meningoencephalo-

cele. Arch Otolaryngol 99:402–405

21. Stankiewicz JA (1991) Cerebrospinal fluid fistula and endoscopic

sinus surgery. Laryngoscope 101(3):250–256

22. Wigand ME (1981) Transnasal ethmoidectomy under endoscopic

control. Rhinology 19(1):7–15 (Mar)

23. Lanza DC, O’Brien DA, Kennedy DW (1996) Endoscopic repair

of cerebrospinal fluid fistulae and encephaloceles. Laryngoscope

106(9 Pt 1):1119–1125 (Sep)

24. Lindstrom DR, Toohill RJ, Loehrl TA, Smith TL (2004) Manage-

ment of cerebrospinal fluid rhinorrhea: the Medical College of

Wisconsin experience. Laryngoscope 114(6):969–974 (Jun)

Childs Nerv Syst