Objectives/Hypothesis: Trauma and surgery are the most common causes of cer
ebrospinal fluid (CSF) rhinorrhea. Surgical repair is recommended for patie
nts with CSF leaks that do not respond to conservative measures, traumatic
CSF leaks that require transcranial surgery for associated brain injuries,
and iatrogenic defects that are discovered intraoperatively. The purpose of
our study was to ascertain the outcome after transnasal endoscopic repair
of CSF leaks and to identify factors regarding the patient, CSF fistula, an
d treatment that may influence the results of the repair. Methods: We perfo
rmed a metaanalysis of all studies published in English between 1990 and 19
99 that reported a minimum of five patients with CSF fistulae that were rep
aired using an endoscopic approach. We analyzed data that included type of
graft and technique used during the repair, surgical complications, the use
of packing, and the use of lumbar drains and antibiotics. The success rate
was monitored and correlated with the other variables. The meta-analysis d
atabase was compared with and added to a database comprising our own patien
ts. Results: Fourteen studies comprising 289 CSF fistulae met the inclusion
criteria Endoscopic repair of CSF leaks was successful in 90% (259/289) of
the cases after a first attempt. Seventeen of 30 persistent leaks (52%) we
re closed after a second attempt. Thus ultimately 97% (276/289) of the leak
s were repaired using an endoscopic approach. The success rate of repairs u
sing any of the reported techniques and materials was high and not statisti
cally different. The incidence of major complications such a meningitis, su
bdural hematoma, and intracranial abscess was less than 1% for each complic
ation. Conclusion: The endoscopic approach is highly effective and is assoc
iated with low morbidity. The literature supports the endoscopic approach u
sing a variety of techniques and materials for the repair of CSF leaks.