|Year : 2018 | Volume
| Issue : 3 | Page : 247-254
Surgical outcome of endoscopic repair of cerebrospinal fluid rhinorrhea
Mohamed H Mansour1, Ahmed S Abdelgelil2, Mohamed K Ibraheem2, Yahia M Dawood2
1 Department of Neurosurgery, Faculty of Medicine, Al-Azhar University, Cairo, Egypt
2 Department of Otorhinolaryngology, Faculty of Medicine, Al-Azhar University, Cairo, Egypt
|Date of Submission||18-May-2018|
|Date of Acceptance||10-Jan-2019|
|Date of Web Publication||15-Apr-2019|
Mohamed K Ibraheem
MD of Otorhinolaryngology, Cairo
Source of Support: None, Conflict of Interest: None
Background Many surgeons have practiced endoscopic repair of cerebrospinal fluid (CSF) rhinorrhea in past decade and have gained popularity. There are several centers that adopt endoscopic repair as a first-line treatment for CSF rhinorrhea with high success rate.
Objective The aim of this study was to detect the role of the transnasal endoscopic approach in the repair of the CSF rhinorrhea and assess its surgical outcome.
Patients and methods A prospective study of 40 patients having CSF rhinorrhea was conducted. All patients were treated with endoscopic repair for CSF rhinorrhea at Neurosurgery Departments of Al-Azher University and specialized Hospital of Otorhinolaryngology and Neurosurgery from January 2016 to January 2018. The majority of them were men. The defects were closed in multilayers using fascia lata, septal cartilage graft, and nasal mucosa in most patients.
Results The left cribriform plate area was found as the most common site of CSF leak. The spontaneous CSF leak was the most common cause of CSF rhinorrhea in this study, and post-traumatic (iatrogenic) CSF leak was the second most common cause. Our success rate of the first attempt at endoscopic repair was 90%, and in the second attempt was 97.5%. The causes that lead to recurrence of leak were large defect, failure of localization of the defect, and comorbid conditions such as bone erosion and chronic cough that developed postoperatively.
Conclusion Endoscopic endonasal approach is a safe and effective way to repair CSF leaks of different causes. Multilayer repair technique is safe and effective, with low complications and preservation of nasal and neurological functions. However, more studies and long series are recommended.
Keywords: CSF rhinorrhea, endoscopic repair, multilayer repair, transnasal
|How to cite this article:|
Mansour MH, Abdelgelil AS, Ibraheem MK, Dawood YM. Surgical outcome of endoscopic repair of cerebrospinal fluid rhinorrhea. Al-Azhar Assiut Med J 2018;16:247-54
|How to cite this URL:|
Mansour MH, Abdelgelil AS, Ibraheem MK, Dawood YM. Surgical outcome of endoscopic repair of cerebrospinal fluid rhinorrhea. Al-Azhar Assiut Med J [serial online] 2018 [cited 2020 Jul 5];16:247-54. Available from: http://www.azmj.eg.net/text.asp?2018/16/3/247/255857
| Introduction|| |
Cerebrospinal fluid (CSF) rhinorrhea is the leakage of CSF into the nasal cavity. It may be either traumatic or nontraumatic. Traumatic CSF leakage may be accidental, following skull base fractures, or surgical, after skull base surgery. The nontraumatic or spontaneous leakage can be with or without elevated intracranial pressure .
Nontraumatic high-pressure CSF rhinorrhea is mainly caused by tumor obstruction; other causes are benign increased intracranial tension or hydrocephalus. Nontraumatic CSF rhinorrhea may be also with normal intracranial pressure. The spontaneous CSF leaks are idiopathic or unknown in origin .
The anterior cranial base fractures are divided into four grades according to Sakas et al., 1998 classification: grade I, small size defect present at cribriform plate, and the fracture line passes through the cribriform plate; grade II, moderate size defect present at frontoethmoidal region, and the fracture line passes through the medial portion of the anterior cranial fossa floor; grade III, moderate to large size defect present at lateral frontal, and the fracture line passes through the lateral frontal sinus; and type IV, complex, more than one defect, and it is a combination of the previous types .
The diagnosis of CSF leak is at first by history of trauma or spontaneous and confirmed by biochemical analysis of CSF leak to ensure the leak is CSF and not nasal secretion, followed by endoscopic evaluation and radiological investigation to detect the site of defect, strategy of surgery, and the best approach of surgical technique .
The site of skull base defect has an important role in determining the surgical approach. The most common CSF leak sites are the cribriform plate, then the roof anterior or posterior ethmoid, the olfactory groove, the roof or the lateral wall of sphenoid sinus, and the posterior wall of frontal sinus .
The first line of management of CSF leak is conservative treatment, especially postaccidental CSF leaks inform of bed rest, elevation of the head, avoidance of straining activities, fluid restriction, and diuretics. Most acute postaccidental CSF leaks heal with conservative management .
When the patients do not respond to conservative management, surgical repair is indicated. The transcranial approach and the extracranial approach are the different techniques of surgical repair of CSF leak. An external ethmoidectomy or osteoplastic flap and transnasal visualization with microscope or endoscope are different approaches of extracranial surgical repair .
The success rate of intracranial surgical repair by bifrontal craniotomy is 70–80% with high morbidity and many complications. The complications of intracranial approaches are such as anosmia, memory deficits, hemorrhage, cerebral edema, and osteomyelitis of the bone flap ,.
The endoscopic technique has several advantages over the open technique, either transcranial or extracranial approach, such as a direct access, exact identification of the site of the dural tear, and precise placement of the graft. It also allows preservation of the functional anatomy of the nose including smell .
| Aim|| |
The aim of this study was to detect the role of the transnasal endoscopic approach in the repair of the CSF rhinorrhea and its surgical outcome.
Moreover, this study aims to analyze the etiopathogenesis of CSF rhinorrhea, the success rate of endoscopic repair, and the causes of failure after surgery.
| Patients and methods|| |
This study was carried out on 40 patients having CSF rhinorrhea. They were managed at Neurosurgery Departments of Al-Azhar University and specialized hospitals of Otorhinolaryngology and Neurosurgery in the period between January 2016 and January 2018.
All patients suspected to have CSF rhinorrhea were subjected to the following:
- Full history taking including personal history, family history, and past history to determine the onset and duration of symptoms and possible etiological factors (hypertension, meningitis, previous nasal, or neurosurgical operations).
- Endoscopic nasal examination to confirm the diagnosis and exclude the presence of other nasal problems.
- Laboratory and radiological evaluations were performed and included the following:
- Confirmation that a nasal discharge was CSF through:
- Quantitative glucose measurement.
- Cytological and chemical analysis of the fluid.
- Site of identification with the following:
- High-resolution computed tomography (CT) scan of skull base and paranasal sinuses: thin-slice axial and coronal CT 1–2 mm thickness shows the skull base bones of the anterior and the middle cranial fossae, in addition to conventional CT brain and paranasal sinuses to delineate any cranial pathology and defects affecting paranasal sinuses and to assess the trajectory during endoscopic approaches ([Figure 1] and [Figure 3]).
|Figure 1 Computed tomography cisternography coronal cut shows the defect at the cribriform plate (arrow).|
Click here to view
- MRI of the brain in selected cases: T1-weighted and T2-weighted MRI to identify the site of CSF leaks and to exclude any intracranial lesions ([Figure 2] and [Figure 4]).
- CT metrizamide cisternography was done in some cases with vague defect. It has been useful for the diagnosis and localizing the site of CSF leaks.
|Figure 2 Coronal T2 MRI shows hyperintense area owing to defect at the cribriform plate.|
Click here to view
|Figure 3 Computed tomography coronal cut shows the defect at the posterior ethmoid roof.|
Click here to view
|Figure 4 Coronal T2 MRI shows hyperintense area owing to defect at the posterior ethmoid (arrow).|
Click here to view
- The patient was placed in the supine position with the head slightly elevated and turned toward the surgeon.
- All procedures were performed under hypotensive anesthesia to provide clear bloodless surgical field, providing the patient with the least hemodynamic changes and shorter recovery time.
- The lateral side of the thigh was sterilized, and the fascia lata graft was harvested to be ready for use in repair site of the defect.
The extent of the operation was dependent on the site of the leak. The mucosa of suspected site of CSF leak in the skull base was refreshed, and the middle turbinate was excised if needed.
We removed approximately 5–7 mm safety margin of exposed bone with diathermy or small-cupped forceps when the site of the leak was identified to facilitate the adhesion of the graft to the underlying bone. The free graft was prepared to be ∼1–2 mm larger than the skull base defect ([Figure 1],[Figure 3],[Figure 5]).
|Figure 5 Intraoperative endoscopic view shows dura at the site of the defect (postethmoid).|
Click here to view
The first layer was fascia lata graft that was placed through the defect and the exposed surrounding skull base bone according to the size of defect, supported by cartilage grafts from septal cartilage, covered by mucosal pedicle flap of middle turbinate or nasal septum. The next layer was a piece surgicel or gelfoam, which was positioned over the grafts. Tissue adhesives may be positioned over grafts to be anchored to the skull base in some cases ([Figure 6],[Figure 7],[Figure 8]).
- In the large size defect, the graft or dural patches was ‘tucked in’ above the bony skull base, ethmoid, or cribriform plate. This ‘tucking in’ helps to hold the graft in place. Then, it was supported by a mucoperiosteal flap from the middle turbinate to seal CSF leaks. Then free autogenously graft is fixed over the fistula by fibrin glue.
- At the end of operation, packing the nose was done by nonabsorbable packing (merocel tampon or antibiotic soaked gauze) to help apply pressure to the graft site to be removed after 1 week.
- After removal of oral pack, all patients were extubated without coughing in deeper plane as coughing can displace the graft.
- Postoperatively, all patients were advised bed rest in semisitting position. They were also given stool softeners and diuretics to prevent increase intracranial tension. Intravenous antibiotics are continued for 5–7 days. We did not use spinal taps or lumbar drains postoperatively in our study except in one recurrent case. We depend on using mannitol solution, which was given as intravenous drip to lower the CSF pressure in the first 48 h postoperatively, together with keeping the patient in 45° head-up position.
- Merocel tampon was removed 1 weak postoperatively. An endoscopic inspection of the operative site with removal of crusts and secretions was performed leaving the operative site untouched, and all patients were observed for recurrence of CSF rhinorrhea and development of complications.
- The patients were discharged from hospital on day 7 if there are no complications. The patients were advised to avoid nose blowing for several weeks.
- Follow-up of all patients was every 2 months for 12 months by clinical assessment using endoscopic examination and by radiological investigations [Figure 9].
|Figure 6 Intraoperative endoscopic view shows the insertion of fascia lata graft at the site of the defect.|
Click here to view
| Results|| |
This study included 40 patients having CSF rhinorrhea, with 12 (34%) females and 28 (66%) males. The age of patients ranged between 5 and 65 years, with a mean age of 32 years. The etiology of CSF rhinorrhea in our cases was traumatic in 26 (65%) patients and nontraumatic in 14 (35%) patients ([Table 1]).
The traumatic group (N=26) included 16 (40%) patients from surgical trauma and 10 (25%) patients from nonsurgical trauma. In postsurgical cases, the primary pathology was 12 (30%) cases of pituitary lesions treated with endoscopic trans-sphenoidal hypophysectomy, two (5%) cases were after excision of skull base tumors (one case was olfactory neuroblastoma and another case was hemangiopericytoma), and two (5%) cases were after functional endoscopic sinus surgery of extensive fungal sinusitis, as shown in Tables 2 and 3 and Chart 1.
CSF leak was from the left side in 30 (60%) cases and from the right side in 10 (40%) cases. The sites CSF leak were the cribriform plate in 20 (50%) cases, followed by sphenoid sinus in 12 (30%) cases, fovea ethmoidalis in four (10%) cases, posterior ethmoid in two (5%) cases, and frontal in two (5%) cases ([Table 4]) and chart 2.
|Table 4 Sites of cerebrospinal fluid leak in 40 patients with cerebrospinal fluid rhinorrhea|
Click here to view
Multilayer technique was performed in all cases. Ethmoid bone of middle turbinate was used in 18 (45%) cases, whereas septal cartilage was used in 22 (55%) cases.
Postoperative follow-up varied from 2–12 months. Minor complications were recorded in seven (17.5%) cases, two cases with nasal adhesions and five cases with hyposmia.
In this study, there were no major complications such as meningitis, pneumocephalus, intracranial bleeding, or vision problems.
In this study, with the attempts of multilayer endoscopic CSF rhinorrhea correction, 36 (90%) patients had adequate closure of the fistula with no signs of the leak during the period of follow-up and four (10%) cases had recurrences ([Table 5]).
|Table 5 Success rate of endoscopic repair of cerebrospinal fluid rhinorrhea|
Click here to view
The first two cases were owing to surgical trauma with postoperative CSF leak (one case was after trans-sphenoidal hypophysectomy and another case was after endoscopic surgery of extensive fungal sinusitis). The exact site of the defect cannot be identified intraoperatively owing to excessive intraoperative bleeding and fibrosis, so the graft was positioned over the suspected area, and the patient did not follow the postoperative instructions completely.
Another two cases were with skull base tumors causing CSF leak; the lesions were destructive with bone erosion causing large size defect of ∼25 mm, and the patient developed postoperative irritative cough that did not respond to treatment and might be the cause of graft dislodgment.
All these four cases had a successful second endoscopic closure except one case of skull base tumor owing to destructive nature of the tumor causing bone erosion with excessive intraoperative bleeding. This patients developed increased intracranial tension with papilloedema postoperatively, so subarchnoid catheter was done for 5 days. The patient refused further surgery and was lost to follow-up.
In the second attempt endoscopic repair of these four cases, the defects were large size, and we used adhesive dural patches (better than fascia lata) ensured at dural edges between the brain and dura, supported by cartilage grafts covered by mucosal flab. All these layers were augmented by fibrin glue.
In this study, the success rate of endoscopic repair of CSF rhinorrhea at the primary surgery was 90% and with the revision surgery increased to 97.5%.
| Discussion|| |
CSF rhinorrhea is a potentially life-threatening condition. It results from a tear of the dura and fracture of the skull base, and it can lead to serious complications such as meningitis and brain abscess .
The management of CSF rhinorrhea is controversial but can be divided into conservative or surgical treatment. The conservative treatment is advised for the first 1–2 months, and if failed, the surgical treatment must be started. The surgical treatment is divided into intracranial or extracranial approaches. The intracranial approach performed through a craniotomy was historically the first surgical technique used to repair anterior cranial fossa leaks. Extracranial approaches have subsequently become more popular. More recently, the introduction of endoscopic sinus surgery has added a new technique for closure of CSF leaks .
Nasal endoscopy aids in diagnosis and diminishes the need for surgical trauma to the external skin and intervening bone. Thus, intranasal trauma is kept to a minimum, and the bony framework can be preserved.
The endoscopic surgical technique achieves a high success rate with minimum complications. For this reason. the endoscopic transnasal repair of CSF fistula has gained popularity during the past 30 years .
This study included 40 patients having CSF rhinorrhea, with 33 (66%) males and 17 (34%) females. This male predominance was reported by Bhalaya and Joseph (12 : 8), and also by Locatelli and colleagues, who reported a ratio of 72 : 33 ,.
In this study, the high male predominance was in traumatic cases duoto the male occupies the risky job and more liable to trauma because of their higher activity. In addition, women are limited to indoor activities, reducing the chance of injury.
In this study, the cribriform area was the most common site of CSF rhinorrhea, which was found in 20 (50%) cases followed by sphenoid sinus in 12 (30%) cases, fovea ethmoidalis in four (10%) cases, posterior ethmoid in two (5%) cases, and frontal in two (5%) cases. This result is similar to the study of Yadav and colleagues in a study of 227 cases, reporting 184/227 (81.25%) cases in the cribriform area. Moreover, Chowdhury and colleagues reported 6/13 cases in the cribriform area. On the contrary, Hari and colleagues reported that anterior ethmoid was the most common site (40%) followed by cribriform plate (30%) ,.
In this study, the cribriform area was the most common site of CSF leak because it is the thinnest and the weakest area of the anterior skull base .
In this study, we used computed tomography with metrizamide or omnipaque injection through a lumbar puncture to localize the site of the defect in five cases in which the defect is not seen in normal CT scan of skull base.
Metrizamide is a water-soluble substance that is injected intrathecally and should be seen on CT scanning if a CSF leak is present .
Nuss and Constantino  reported 100% success rate in identifying the site of CSF leakage, who artificially increased CSF pressure by using metrizamide injection to convert an inactive leak into an active leak and identify the site of CSF.
In this study, in all cases with CSF rhinorrhea, the multilayer endoscopic technique was used to close them. This technique was used in the management of post-traumatic CSF leaks, after 4–6 weeks of conservative treatment, all cases of surgical trauma was after functional endoscopic sinus surgery and after pituitary surgery.
In this technique, refreshment of the edge of defect was performed, and then multilayered grafting was done by fascia lata or mucosal layer of middle turbinate flap and then supported by surgicel and gelfoam. Lastly, nasal pack was put and removed after 1 week.
This technique is similar to techniques of many authors who described their endoscopic techniques for repair of CSF rhinorrhea placing the intracranial and the extracranial approaches, such as Hardy in 1970, who described the endonasal repair of sphenoid sinus CSF rhinorrhea using a septal mucosal flap .
Maurice et al. covered the cribriform plate by a pedicled flap from the middle nasal turbinate after removing the bone from it .
Casiano described free fascia grafts in conjunction with mucoperiosteal flaps to close CSF fistulae in 20 patients via endoscopic ethmoidectomy .
In this study, the success rate of endoscopic technique of CSF rhinorrhea repair at the first attempt was 90% and with the second attempt increased to 97.5%. This is similar to results of Singh and colleagues who had success rate of 91% in first attempts and 97% in second attempts by using septal cartilage with soft tissue graft, surgical, and gelfoam. Chowdhury et al.  reported success rate of 93.9% in primary surgery and 100% in the revision surgery.
The causes of failure in this study were as follows: the exact site of the defect was difficult to be detected intraoperative owing to excessive intraoperative bleeding and fibrosis, graft dislodgment by irritative cough, large size defect with bone erosion by skull base tumors, or the patients did not follow the postoperative advice completely.
This is similar to many authors, such as Singh and colleagues and Chowdhury and colleagues who reported that the factors related to the potential failure of an endoscopic approach include the following:
- Inability to localize the defect successfully.
- Graft displacement.
- Small graft size not enough to cover the defect.
- Noncompliance of the patient with postoperative instructions.
In this study, in large size defects and in recurrent cases, we used fibrin glue over fascia lata graft . this is similar to Wormald and McDonogh . Who had impressive results of cases treated endoscopically with the bath-plug technique of middle turbinate using two layers of fibrin foam with 100% success.
From this study, in the large size defect, we must use adhesive dural patches (better than fascia lata) ensured at dural edges between the brain and dura, supported by cartilage grafts, and covered by mucosal flab, and all these layers should be augmented by fibrin glue.
In the endoscopic repair, there is no need for metallic reconstruction by miniplate, and all materials used in repair were natural such as cartilage graft, mucosa, and fascia lata, and augmented by fibrin glue.
Patient satisfaction was high because of no external incision, early postoperative recovery, and less postoperative pain.
Finally, the endoscopic technique provides excellent field of vision, excellent illumination with magnification, and well localization of the defect. So, it is a preferred approach in traumatic and spontaneous CSF leaks not associated with increased intracranial tension by space-occupying lesions.
| Conclusion|| |
The endoscopic technique provides excellent field of vision, excellent illumination with magnification, well localization of the defect, and accurate positioning of the graft.
Multilayer repair technique is safe and effective with low complications and preservation of nasal and neurological functions. A high success rate is attached with this technique, with low complications. So, it is a preferred approach in traumatic and spontaneous CSF leaks not associated with increased intracranial tension by space-occupying lesions. However, more studies and long series are recommended.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Chin GY, Rice DH. Transnasal endoscopic closure of cerebrospinal fluid leaks. Laryngoscope 2003; 113:136–138.
Singh R, Hazarika P, Nayak DR, Sarkar A. Endoscopicrepair of cerebrospinal fluid rhinorrhea: Manipal experience. Indian J Otolaryngol Head Neck Surg 2009; 61:14–18.
Sakas DE, Beale DJ, Ameen AA, Whitwell HL, Whittaker KW, Krebs AJ et al.
Compound anterior cranial base fractures: classification using computerized tomography scanningas a basis for selection of patients for dural repair. J Neurosurg 1998; 88:471–477.
Locatelli D, Rampa F, Acchiardi I, Bignami M, De Bernardi F, Castelnuovo P. Endoscopic endonasal approaches for repair of cerebrospinal fluid leaks: nine-year experience. Neurosurgery 2006; 58(Suppl 2):ONS-246–ONS-256.
Sharma DP, Singh D, Sinha S. CSF rhinorrhea: an overview of endoscopic repair. Indian J Neurotrauma 2010; 7:157–162.
Zweig JL, Carrau RL, Celin SE. Endoscopic repair of cerebrospinal fluid leaks to the sino nasal tract: predictors of success. Otolaryngol Head Neck Surg 2005; 123:195–201.
McCormack B, Cooper PR, Persky M, Rothstein S. Extracranial repair of cerebrospinal fluid fistulas: technique and results in 37 patients. Neurosurgery 1990; 27:412–417.
Tolley NS, Brookes GB. Surgical management of cerebrospinal fluid rhniorrhea. J R Coll Surg Edinb 1992; 37:12–15.
Kennedy DW. Functional endoscopic sinus surgery: Anaesthesia, technique and postoperative management. In: Kennedy DW, Bolger WE, Zinreich SJ, editors. Diseases of the sinus, diagnosis and management. Hamilton, London: BC Decker; 2012. 211–221.
Anand VK, Murali RK, Glasgold MJ. Surgical decisions in the management of cerebrospinal fluid rhinorrhoea. Rhinology 1995; 33:212–218.
Stamm AC, Pignatari SN. Transnasal endoscopic-assisted surgery of the skull base. In: Cummings CW, Flint PW, Harker LA, et al., editors. Cummings otolaryngology head and neck surgery. 6th ed. Philadelphia, Pennsylvania: Elsevier Mosby; 2015. 3855–3875
Chowdhury FH, Haque M, Islam S. Endoscopic transnasal repair of CSF rhinorrhea: technique and outcome. Bangladesh J Otorhinolaryngol 2011; 17:14–20.
Yadav J, Parihar V, Janakiram N, Pande S, Bajaj J, Namdev H. Endoscopic management of cerebrospinal fluid rhinorrhea. Asian J Neurosurg 2016; 11:183–193. [Full text]
Citardi MJ. Cerebrospinal fluid rhinorrhea. In: Cummings CW, Flint PW, Harker LA et al.
, editors. Cummings otolaryngology head and neck surgery. 5th ed. Philadelphia, Pennsylvania: Elsevier Mosby; 2010. 1276–1289
Nuss D, Constantino P. Diagnosis and management of cerebrospinal fluid leaks. operative technique of Otolaryngology-head and neck surgery. St Louis, Missouri: Mosby-Year Book; 1996. 79–95
Hardy J. Transsphenoidal hypophysectomy. J Neurosur 1971; 34:582–594.
Maurice V, Ropper AH, Delacy AR, Brown RF. Adams and Victor’s principles of neurology. New York: McGraw-Hill Medical Pub. Division; 2005. 541–543
Casiano R. Endoscopic sinus surgery dissection manual: a stepwise anatomically based approach to endoscopic sinus surgery. New York: Informa Healthcare; 2002.
Wormald PJ, McDonogh M. ‘Bath-plug’ technique for the endoscopic management of cerebrospinal fluid leaks. J Laryngol Otol 1997; 111:1042–1046.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6], [Figure 7], [Figure 8], [Figure 9]
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5]