Editorials

Advances in the management of CSF leaks

New techniques will improve the management of unilateral clear nasal discharge

Important advances have been made in the diagnosis, localisation, and surgical management of cerebrospinal fluid (CSF) rhinorrhoea. CSF leaks have been associated with about a 10% risk of developing meningitis per year.¹ It is important to diagnose the cause of unilateral clear rhinorrhoea and to differentiate unilateral autonomic rhinitis from the rupture of a mucus retention cyst (the contents of which are light straw coloured) or a CSF leak. CSF leaks can occur spontaneously although there may be a history of trauma or surgery. A specimen of the discharge must be sent for analysis of  2 transferrin by immunofixation; this test has a high specificity and has superseded all other diagnostic techniques.² The glucose oxidase test has poor predictive value and should no longer be used.³ Unilateral autonomic rhinitis can look like CSF rhinorrhoea, and it is essential that fluid be sent for  2 transferrin analysis before surgery is contemplated.⁴

Localisation of the site of any suspected defect is best detected by high definition coronal computed tomograpy of the anterior skull base.⁵ If this fails to localise the site of a defect, a T2 weighted magnetic resonance image may help, and this has superseded computed cisternography.⁶ In a small proportion of patients, the site of the leak may remain unclear or there may be a suspicion that there is more than one defect in the skull base. Under these circumstances a diagnostic or peroperative lumbar puncture using fluorescein dye will help to localise the area. The use of dilute fluorescein and a "pencil point" epidural needle has minimised the complications associated with this technique.⁷

Endoscopic closure has revolutionised the surgical management of CSF rhinorrhoea and has reduced the morbidity associated with it; a recent meta-analysis reviewed the use of endoscopic repair.⁸ The sense of smell is almost always preserved using this technique but it is usually lost when a transcranial approach is used.⁸ When endoscopic closure is used the length of time spent in hospital is usually restricted to 36 hours, and a craniotomy is avoided. Nasal endoscopic repair has a success rate of 90% at the first attempt and 97% after a second attempt.⁸ Morbidity is minimal.

Although an intracranial approach has the advantage of allowing for the resection of any coexisting intracranial pathology, the success rates achieved using anterior craniotomy are less than 75%. ^{9 10} A frontal craniotomy often results in a loss of the sense of smell and uncommonly, but importantly, may be complicated by postoperative intracerebral haemorrhage, cerebral oedema, epilepsy, frontal lobe dysfunction with memory and concentration deficits, and osteomyelitis of the frontal bone flap. In addition, this technique requires the patient to spend five to seven days in hospital, results in hair loss along the incision line, and the patient must not drive until judged to have recovered from the operation. An extradural approach is, however, needed for defects of the posterior wall of the frontal sinus or defects larger than 5 cm because these cannot be managed endoscopically.

If a CSF leak occurs in conjunction with a tumour of the skull base or a severe fracture of the skull base, a craniotomy and removal of the posterior wall of the frontal sinus and its lining along with the anterior wall of the frontal sinus with split calvarial bone is indicated.¹¹ The use of a pericranial flap reduces the postoperative incidence of CSF leaks.

No prospective randomised study has been done comparing these techniques. However, on the basis of series reports^12-14 and a meta-analysis,⁸ the differences in morbidity and success rates between the techniques make transnasal endoscopic repair the treatment of choice for most CSF leaks from the anterior cranium and sphenoid sinus.

N S Jones, professor in otorhinolaryngology. 

Department of Otorhinolaryngology, University Hospital, Nottingham NG7 2UH (nick.jones{at}nottingham.ac.uk)

Daniel G Becker, assistant professor. 

Department of Otorhinolaryngology, Head and Neck Surgery, University of Pennsylvania Health System, 3400 Spruce Street, Philadelphia, PA 19104, USA (beckerd{at}mail.med.upenn.edu)