Pan Arab - Journals Vol4, No.2 - Endoscopic Management of Cerebrospinal Fluids Leaks

Review Article

Volume 4, No.2

October 2000

Usamah Hadi(1)
Ghassan Kahwagi(1)
Abdel Latif Hamdan(1)
Jamal Shreif(1)
George Haddad(2)

(1)Department of Otolaryngology
      Head & Neck Surgery (2)Department of Neurosurgery        American University of
         Beirut Medical Center
      Beirut
      Lebanon

Correspondence:
Dr. Usamah Hadi
Dept. of Otolaryngology
Head & Neck Surgery
American University of
Beirut Medical Center
Bliss Street
PO Box 11-0236
Beirut
Lebanon
Fax: (961) 1 744 464
E-mail: uhadi@dm.net.lb

Endoscopic Management of Cerebrospinal Fluid Leaks - A Viable Alternative

ABSTACT

Cerebrospinal fluid (CSF) leaks are uncommon, yet serious because they predispose patients to meningitis with its central nervous system complications. Otolaryngologists aided by neurosurgeons are playing an important role in the diagnosis and management of such leaks in the anterior skull base area. Experience with endoscopic techniques and computer guided navigation systems for the management of sinus disease has encouraged the otolaryngologist to address such problems.

We present our series of 9 patients whose fistulae were managed endoscopically. Follow-up ranged from 4 months to 6 years. Control of CSF leak was achieved in 88.8% of our cases with the first attempt. One patient had an unsuccessful repair that was managed via an intracranial approach. The aetiology, diagnostic workup and location of the CSF leak in the anterior skull base will be reviewed. The endoscopic surgical technique employed, as well as the types of grafts used for the repair will be discussed and elaborated upon.

Keywords: Cerebrospinal fluid leak, endoscopy and radiology.

INTRODUCTION

The recent advances and technological improvements in endoscopic instrumentation and the availability of CT guided navigational systems have expanded the role of the otolaryngologist in both diagnosis and management of cerebrospinal fluid (CSF) leaks. Anterior and middle skull base defects associated with CSF leak were previously regarded as primarily neurosurgical problems. Although, intracranial repair is still favoured at many institutions despite its inherent morbidity; extracranial transnasal endoscopic repair of CSF leak is gaining acceptance as a viable alternative to treating such ailments.

We present a prospective review of 9 patients who underwent transnasal endoscopic repair of CSF fistulae during the past 7 years at our institution.

MATERIALS AND METHOD

From 1993 to 1999, 9 patients underwent endoscopic repair of CSF rhinorrhea at our institution. There were 7 males and 2 females, ranging in age from 30 to 65 years. Follow-up data ranged from 4 months to 6 years. Several parameters were evaluated. These included the aetiology, duration of CSF leakage, preoperative complications, defect size and location, diagnostic tests, types of grafts that were used, use of intrathecal fluorescein, lumbar drain and types of nasal packs used.

Results
The causes of the CSF fistulae were:

1.	Iatrogenic in 5 cases, 4 of which were postendoscopic sinus surgery (ESS) and one was posttranseptal transphenoid resection of pituitary tumour.
2.	Traumatic in 3 cases.
3.	Spontaneous in one case.

The duration of CSF leak varied in the iatrogenic group. Of the 4 patients who underwent functional endoscopic sinus surgery (FESS), complicated by CSF leak, 2 were repaired immediately intraoperatively and 2 other patients had a delayed repair of their leak.

The fifth case had a leak following a transsphenoidal resection of a pituitary tumour. Several attempts by the neurosurgeon failed to stop the leak, that was later successfully closed using endoscopic techniques (Table 1). The duration of the leak in the remaining group ranged from several months to more than 2 years in the traumatic group. In the spontaneous category the leak was present for 6 months and the patient survived two episodes of meningitis with no sequelae.

Table 1 - Patient's profile and management

Name	Sex/Age	Duration of Leaks	Post Fess	Trauma	Rhinorrhea	Site of Skull Base Violation	Radiology: CT Scan, Metrizamide	Endoscopic Management	Follow-up
N.A (1)	M/65	2y	Y	I	Lt.	Fovea ethmoidalis	Localised Site	Fascia, Fibrin glue, muscle gelfoam & nasal packs	No recurrence after 6 years
I.S (2)	M/30	1y	N	A	Lt.	Fovea ethmoidalis	Localised Site	Fascia, Fibrin glue, muscle graft	No recurrence
K.H (3)	M/42	1m	S/P TSSP Tumour exc.	I	Bil.	Sphenoid	Localised Site	Attempts at closure failed TF, TM, FAT, FG	No recurrence
S.K (4)	M/40	Intra-op	Y	I	Rt.	Post-ethmoid	NA	Intra-op repair TF, TM, FG	No recurrence
I.A (5)	M/34	Intra-op	Y	I	Lt.	Anterior fovea ethmoidalis	NA	Local mucosal flap Free & fibrin glue	No recurrence
A.M (6)	M/38	5d	Y	I	Rt.	Post-ethmoid 7-8 mm bony defect	NA	Fascia, TM, TG	No recurrence
I.D (7)	F/51	1y	N	A	Lt.	Cribriform & ethmoid fovea	Localised Site	Endoscopic TF, TM Fibrin glue & Nasal pack	Recurrence 1y Lt. Frontal craniotomy
A. D (8)	F/50	4m	N	F	Lt.	Cribriform & ethmoid fovea	Localised Site	TF, TM, FG & Packs Hypertensive crisis	No recurrence after 11 months
O. A (9)	M/30	6m	N	Sp.	Lt.	Fovea ethmoidalis & LLLC	Localised Site	TF, TM, FG & Packs	No recurrence after 11 months

Legend: TM - Temporalis muscle; TF - Temporalis fascia; NA- Not available; I - Iatrogenic; A-Accident; F-Fall; S - Spontaneous

Table 2 - Distribution of skull base defect vs. aetiology

Group Category	Anterior Fovea Ethmoidalis	Posterior Fovea Ethmoidalis	Anterior Ethmoidalis & Cribriform Fossa	Sphenoid	Side Involved
Iatrogenic	3	1		1
Post-Traumatic			3
Spontaneous	1
Right					2 of 9
Left					6 of 9
Bilateral					1 of 9

The commonest location of the skull base defect was in the fovea ethmoidalis. Two patients had, in addition, leaking sites in the area of the cribriform plate. One case had a leak from the sphenoid sinus. When the side of injury was classified on the basis of aetiology, there appeared to be a tendency towards left sided injury in 6 out of 9 cases, two had a right sided injury and one was bilateral, in the sphenoid sinus (Table 2).

DISCUSSION

Increasing incidences of CSF leaks from neurosurgical and otolaryngology procedures, as well as from traumatic accidents, have stimulated the development and refinement of endoscopic procedures. Despite the fact that CSF leaks are relatively uncommon, and though it seems to be a straightforward problem, it remains to be one of the most challenging issues to both the otolaryngologist and the neurosurgeon.

The actual risk for a patient with CSF leak to develop meningitis ranges from 4-50% as reported by various series.(2,11) Post-meningitic sequelae can be devastating to the patient and the use of prophylactic antibiotics to reduce its risk has not been effective.(10,11) Thus, faced with a persistent leak it becomes of paramount importance to confirm the leak and localise the defect in the skull base.

Various laboratory and diagnostic tests are available for the clinician to confirm a CSF leakage and should be undertaken when the history and physical examination are not highly suggestive of the condition. Checking for glucose content by using dextrostix and urostix are sensitive, yet they frequently give false positive results in up to 45% of cases.(16,22) Beta-2 transferrin testing is highly sensitive and is a specific method for documenting the nature of rhinorrhea, yet it's not available for routine usage.(9,21)

Radiologic diagnostic testing plays a crucial role in mapping breaches in the skull base. In the present study, localising the defect was achieved using 2 mm coronal CT scans of the paranasal sinuses and the cribriform plate. CT scan cisternogram with metrazimide was used in most of the cases. This modality was capable of demonstrating the exact route of CSF flow through the fistula site in all of our cases.(13,14) Nabawi, was successful in demonstrating the site of CSF leak by CT - metrizamide in approximately 70 to 85% of his cases.(13)

MRI was performed in one of our patients with spontaneous CSF leak. It is valuable in delineating neural tissue herniation, small encephalocele, arachnoids villi, and dilated ventricles better than CT scan.

CT scan and CT cisternogram may at times fail to localise the fistula site, especially in patients who suffer from slow, low and/or intermittent leaks. Despite the low resolution images of volume leaks nuclear radioisotope scans in such conditions can be helpful in showing laterality of the leak or the cranial fossa of origin.

All defects were closed using extracranial intranasal endoscopic techniques. The site of the anterior skull base defect was usually identified using either pre-operative CT scan alone (Fig. 1 and 2) or with metrizamide contrast material (Fig. 3). It gave precise delineation of the bony cranial base anatomy as well as the related paranasal sinus anatomy. MRI scans were obtained in the case of spontaneous CSF leakage to exclude any associated intracranial pathology (Fig. 4). The biochemical confirmation of rhinorrhea was usually evaluated using glucose testing after retrieval of the specimen in a few cases. Beta-2 transferring testing was not performed.


Figure 1 — A 2 mm coronal CT scan post ESS revealing a major defect in the fovea ethmoidalis (arrow).		Figure 2 — A 2 mm coronal CT scan of the sinuses revealing a bony defect in the left fovea ethmoidalis close to the lateral lamella of the lamina cribosa (arrow).

Figure 3 — A 2 mm coronal CT scan with metrizamide contrast medium revealing a right fovea ethmoidalis crack with spillage of metrizamide into the ethmoid sinus (arrow head). Pooling of metrizamide dye in the left olfactory groove (arrow).

Suspicion of intra-operative leak can usually be confirmed by elevating the head of the patients by asking the anaesthesiologist to perform the valsalva manoeuvre (increasing intrathoracic pressure) that would cause a transient elevation of intracranial pressure, and may thereby accelerate CSF leakage. Intrathecal flourescein injection (0.1 ml of 10% flurosurin, mixed with 10 cc of CSF leak and reinjected) was used initially in two patients to delineate the site of the leak prior to surgical exploration of the skull base. This has been abandoned in the remaining patients for fear of its neurotoxicity. Indwelling lumbar drain has also been deserted due to its potential risks for complication.

The repair of CSF leak in the fovea ethmoidalis is tailored to the site of defect. The need for complete ethmoidectomy and/or sphenoethmoidectomy may or may not be required. Once the defect is localised endoscopically, with or without the aid of fluorescein injected intrathecally, the skull base is exposed by removing few millimetres of mucosa around the defect. This allows the free graft, whether temporalis fascia or

Figure 4 — T2-weighted MRI images of paranasal sinuses revealing presence of CSF in the left ethmoid cavity in the patient with spontaneous leak (arrow).

nasal/turbinate mucosa to adhere directly to bone. In bony defects larger than 1 cm, a piece of septal bone and/or conchal bone can be used to reinforce the defect. Composite graft with bony/mucosa such as middle turbinate or septum can be rotated to close defects in the cribriform plate or fovea ethomidelis respectively

Free mucosal grafts, temporalis fascia and muscle with fibrin glue were used in all our patients. In addition, abdominal fat was employed in securing an adequate seal in the sphenoid sinus. The mucosa free graft is positioned extracranially with its mucosal surface facing the nasal cavity. Fibrin glue was used to hold the graft in position. Surgicel and gelfoam were placed for further support. Vasoline packs and/or nasal balloons were used to support the gelfoams. Lumbar drain was placed for two of our patients for 2 to 5 days, together with complete bed rest and head elevation to 300 for 1 week.

We had a high success rate in our services from the first attempt (8/9). None had a second endoscopic trial at closure of a recurrent leak. In the unsuccessful case, where the leak reoccurred one year later, the neurosurgeon decided to close the defect using a left frontal craniotomy approach.

Our overall successful first repair rate of 88.8% compares favourably with other reported series.(3,17) Variable success rates from craniotomies to repair such fistulae have been reported to be as low as 60% by Park, et al., Arabi and Hubbard have demonstrated that management of CSF leaks using intracranial routes might be associated with failure rate ranging between 20-40%.(8,18) High success rate has been reported using other extracranial non-endoscopic techniques.(3,4,19) Dodson, et al., reported on the largest series of endoscopic repairs, achieving control of CSF leak in 75.9% after the first endoscopic attempt and up to 86.2% after a second attempt.(6)

Autologous tissues, such as free mucosal grafts, temporalis fascia, muscle and/or abdominal fat were the preferred grafts used in our patients. Stored, processed, radiated or otherwise commercially prepared tissue were not used. Vascularised tissues such as septal and middle turbinate mucosal flaps and pericranial flaps can be used when feasible. In rare circumstances when autologous tissue or vascularised flaps are inadequate, as in previous trauma to sinuses or excessive scarring, then biomedical materials can be used to obliterate the fistula site. Hydroxyapatite (HA1) in a cement form can be used to cover large cranial base defects.(5,7) The use of fibrin glue in our series has been very helpful in stabilising the fascial and mucosal grafts and/or flaps. It provides adequate support for the re-constructed site obviating the need for prolonged nasal packing or insertion of lumbar drain.

Intranasal endoscopic repair of CSF leak continues to reflect a high success rate with a minimal degree of morbidity compared to intracranial approaches. We feel it should be the preferred approach to all patients with anterior skull base defect.

CONCLUSION

The excellent visualisation and atraumatic surgical techniques of endoscopic sinus surgery have been applied to the management of 9 cases of CSF leaks. In all cases, the exact site of the lesion was identified endoscopically as well as radiologically.

Three cases had fractures of their anterior cranial fossa following head trauma. Other cases had violation of the cribriform plate and the fovea ethmoidalis as a complication of endoscopic sinus surgery for chronic sinusitis.

The CSF leaks were sealed using free mucosal grafts as well as temporalis fascia and muscle. Fibrin glue was applied in all cases to achieve adequate adhesion of the grafted material to the traumatic site in the floor of the anterior cranial fossa.

We believe the application of endoscopic techniques for intranasal closure of CSF leaks offers a number of advantages over other routine extranasal or intracranial approaches for closure of CSF leaks.

REFERENCES

1.	Arabi B, Leibrock LG: Neurosurgical approaches to cerebrospinal fluid rhinorrhea. Ear Nose Throat J 1992, 71: 300-305
2.	Brisman P, Hughes JEO, Mount LA: Cerebrospinal fluid rhinorrhea. Arch Neurol 1970, 22: 245-252
3.	Burns JA, Dodson EE, Ross CW: Transnasal endoscopic repair of cranionasal fistulae: A refined technique with long-term follow up. Laryngoscope 1996, 106: 1080-1083
4.	Calcaterra TC: Extracranial surgical repair of cerebrospinal rhinorrhea. Ann Otol 1980, 9: 108-116
5.	Costantino PD, Friedman CD, Jones K, et al: Hydroxyapatite cement: I. Basic Chemistry and histologic properties: Arch Otolaryngol Head and Neck Surg 1991, 117: 379-384
6.	Dodson EE, Gross CW, Swerdloff JL, et al: Transnasal endoscopic repair of cerebrospinal fluid rhinorrhea and skull base defects. A review of twenty-nine cases. Otolaryngol Head & Neck Surg 1994, 111: 600-605
7.	Friedman CD, Costantino PD, Jones K, et al: Hydroxyapatite cement: II. Obliteration and reconstruction of the cat frontal sinus. Arch Otolaryngol 1991, 117: 385
8.	Hubbard JL, McDonald TC, Pearson BW, et al: Spontaneous cerebrospinal fluid rhinorrhea: evolving concepts in diagnosis and management based on the Mayo Clinic experience from 1970 to 1981. Neurosurg 1985, 16: 314-321
9.	Irjala KI, Svonpaa J, Laurent B: Identification of CSF leakage by immunofixation. Arch Otolaryngol 1979, 105: 447-448
10.	Klastersky J, Sadeghi M, Brihaye J: Antimicrobial prophylaxis in patients with rhinorrhea or otorrhea: A double blind study. Surg Neurol 1976, 6: 111-114
11.	MacGee EE, Cauthren JC, Brackett CE: Meningitis following acute traumatic cerebrospinal fluid fistula. J Neurosurg 1970, 33: 312-316
12.	Mattox DE, Kennedy DW: Endoscopic management of cerebrospinal fluid leaks and cephaloceles. Laryngoscope 1990, 100: 857-862
13.	Nabawi P, Mafree M, Philips J, et al: The success rate of metrizamide CT cisternography in the evaluation of cerebrospinal fluid (CSF) rhinorrhea. Comput Radiol 1982, 6: 343-354
14.	Naldich TP, Moarm CJ: Precise anatomic localization of atraumatic sphenoethmoidal cerebrospinal fluid rhinorrhea by metnizamide CT cisternography. J Neurosurg 1980, 53: 222-228
15.	Oberascher G: Cerebrospinal fluid otorrhea - new trends in diagnosis. Am J Otol 1988, 9: 102-108
16.	Ommaya AK, Dichiro G, Baldwin M, et al: Non-traumatic cerebrospinal fluid rhinorrhea. J Neurol Neurosurg Psych 1968, 31: 214-225
17.	Papay FA, Maggiano H, Dominquez S, Hassenback SJ, Levine HL, Lavertu P: Rigid endoscopic repair of paranasal sinus cerebrospinal fluid fistulas, Laryngoscope 1989, 99: 1195-1201
18.	Park JI, Strelzow V, Friedman WH: Current management of cerebrospinal fluid rhinorrhea. Laryngoscope 1983, 93: 1294-1300
19.	Persky MS, Rothstein SG, Breda SD, Cohen NL, Cooper P, Ransohoff J: Extracranial repair of cerebrospinal fluid otorhinorrhea. Laryngoscope 1991, 101: 134-136
20.	Schaefer SD, Diehl JT, Briggs WH: The diagnosis of CSF rhinorrhea by metrizamide CT scanning. Laryngoscope 1980, 90: 871-875
21.	Skedros DG, Cass SP, Hirsch BE, Kelly HR: Beta - 2 transferrin assay in clinical management of cerebral spinal fluid and perilymph fluid leaks. J Otolaryngol 1993, 22: 341-344
22.	Spetzler RF, Zabramski JM: Cerebrospinal fluid fistulae. Their management and repair in Youmans JR, (Ed). Neurological Surgery 1990, WB Saunders, Philadelphia.