Use of bone wax in the prevention of cerebrospinal fluid fistula in acoustic neuroma surgery

The reported incidence of postoperative cerebrospinal fluid (CSF) fistula following surgery for acoustic neuroma varies widely throughout the literature. In a review of neurotologist/neurosurgeon teams, Hoffman1 found a CSF leak rate of 17% by the translabyrinthine approach and of 16% in the retrosigmoid group. Reported incidence of CSF leak in the literature ranges from 8% to 20% in translabyrinthine procedures, from 0% to 27% in retrosigmoid approaches, and from 0% to 12% in middle fossa craniotomy. Techniques for closure of surgically exposed air cell tracts include the use of fat, fascia, muscle, bone wax, bone pate, and tissue adhesives.2 The numerous techniques described in the literature to prevent CSF leak in acoustic neuroma surgery are a testament to the difficulty of the problem and to the lack of universal success with any single method. Consistent prevention of postoperative CSF fistula must reliably produce a water-tight barrier in all air cell tracts that could allow passage of CSF beyond the surgical defect. Our single-surgeon experience suggests that the judicious use of bone wax effectively prevents postoperative CSF otorhinorrhea in both the translabyrinthine and retrosigmoid approaches. In our earlier experience, the transtemporal defect was closed with strips of fat as per the Hitselberger-House technique.3 In our more recent translabyrinthine experience, we adapted thorough bone waxing as an adjunct to the Hitselberger-House fat graft technique. Some transtemporal drill-out details changed as the bone wax technique evolved. During mastoidectomy, the aditus ad antrum is left as small as possible while still allowing direct visualization of Bill's bar. Minimal exposure of the incus protects it from dislocation by the waxing process. The anterior curl of the lateral semicircular canal is left intact, narrowing the space between the facial nerve and the incus. The relatively small resulting mastoid aditus is waxed in continuity with the zygomatic root and more lateral facial recess air cells. Opening the middle ear and facial recess is avoided, and the eustachian tube is not addressed. The retrofacial-hypotympanic air cells are waxed in continuity with the posterior exposure of the sinus tympani (when present) and the air cells inferior to the internal auditory canal. The air cells superior to the internal auditory canal are waxed independently. Sufficient waxing and water-tight closure of air cell tracts is accomplished by pressing a smooth cover of bone wax into position with neurosurgical cottonoid strips. Abdominal fat graft strips, folded over the sigmoid sinus, obliterate the posterior temporal bone defect. During retrosigmoid craniotomy, perisigmoid air cells are occluded with bone wax. After tumor removal the defect in the porous acusticus is waxed, creating a continuous smooth surface of wax over the entire drilled bone surface of the petrous bone, whether or not specific air cells were seen. Infection prevention protocol in this series included perioperative antibiotics for 24 hours, copious Bacitracin irrigation (50,000 U/L) just before waxing air cells, and meticulous control of surgical site sterility. Between 1995 and 1997, we used these techniques in 62 patients. No leaks were encountered in the 27 patients whose tumors were removed via the translabyrinthine approach. One of 35 patients undergoing retrosigmoid craniotomy developed a CSF leak which responded promptly to lumbar drainage (overall leak rate, 1.6%; translabyrinthine, 0%; retrosigmoid, 2.8%). Because of low CSF leak rates using bone wax in retrosigmoid tumor approaches, the senior author (L.J.B.) began applying bone wax to exposed air cell tracts in translabyrinthine surgeries. The problem was to avoid placing bone wax in areas where it could extrude. The concept maximized exposure necessities while preventing CSF from entering the middle ear and eustachian tube through exposed air cells. From the translabyrinthine defect, CSF may access air cell tracts around the labyrinth, the mastoid aditus, the zygomatic root, the facial recess, and the retrofacial-hypotympanic tracts. Air cell tracts around the internal auditory canal may connect to the peritubal air cells.1 Inadvertent subluxation of the stapes can also result in CSF drainage through the oval window into the middle ear.4, 5 Montgomery6 first described the use of fat to fill the translabyrinthine defect. Later, House et al.3 used strips of fat to strategically line the cavity. CSF penetration through the fat graft may allow drainage through the incision or the nose.7 CSF leakage through the wound responds well to adjustment and tightening of the mastoid dressing in most cases. The preventable focus of CSF leak is each and every exposed air cell. Other authors have extended dissection to the middle ear and/or the eustachian tube to create a barrier anterior to the primary dissection route, using fat,8, 9 muscle,10 and/or proplast.11 Closure of the eustachian tube has been attempted with bone pate,12 fibrin glue,13 Gore-Tex,14 hydroxyapatite,15 muscle, scarring,10 and even bone wax,16 all with varying degrees of success. These approaches may remove the incus with the potential risks of tympanic membrane and stapes disruption. Some authors remove the posterior canal wall to access the eustachian tube directly. Closure of the external auditory meatus then becomes necessary.8 Since peritubal cells entering directly into the eustachian tube are present in as many as 65% of histologic sections, obliteration of the lateral orifice of the eustachian tube may not prevent CSF rhinorrhea.17 Management of the peritubal air cells may require extensive drilling. All of these methods add time to an already lengthy procedure. By comparison, obliteration of the surgically traversed air cells where they are encountered is a simple, direct, and efficient means by which to bypass these difficulties. Use of bone wax is more commonly described in the retrosigmoid approach.9, 18 Tisseel,19 fibrin glue,13, 20 and hydroxyapatite21 have all been used in a similar fashion. The incidence of CSF leak has not been shown to be significantly different when using fibrin glue compared with controls.13 Critics of the use of bone wax allege that it creates an inadequate seal and is prone to infection.22 Management of air cell tracts in both translabyrinthine and retrosigmoid approaches has been highly effective and well tolerated in our series. There were no wound infections, and only two patients developed evidence of aseptic meningitis which responded well to brief steroid use. The literature does report removal of infected bone wax from a mastoidectomy cavity 18 years after placement.10 By design, the amount of bone wax used in this technique is significantly less than the amount required to obliterate the entire mastoid cavity. In most cases the actual amount used is less than that required in retrosigmoid surgery, where the use of bone wax is more widely accepted. By blocking only exposed air cells in a sharpened translabyrinthine dissection, wax use is limited and extrusion has not been a problem. The concept of CSF leak control in the retrosigmoid approach is to stop CSF leaks before they enter the residual temporal bone air cell system. Application of the same concept to the translabyrinthine route appears highly successful. Bone wax application to all encountered air cell tracts partitions the anterior temporal bone from the surgical pathway. Obliteration of an intentionally smaller mastoid aditus ad antrum exposure limits disruption of the ossicles and tympanic membrane and avoids wax as a potentially exposed middle ear foreign body. Use of bone wax for the prevention of CSF fistula in acoustic neuroma surgery appears to be a safe, effective, and time-efficient adjunct not only for retrosigmoid, but also for translabyrinthine approaches.