Chapter 48 Middle Fossa Approach
The middle fossa approach for vestibular nerve section was reported in 1904; however, hammer and chisel were used at that time, which put the facial nerve at risk.1 The middle fossa approach did not have widespread application until refined by the senior author (W.F.H.) in 1961.2 The approach was used initially for decompression of the internal auditory canal (IAC) in cases of extensive otosclerosis. That therapy was later abandoned, but it became evident that this approach was suitable for removal of acoustic tumors.
Initially, the middle fossa approach was used for tumors of all sizes. Further experience showed that it was most suitable for small tumors,3–5 however, and that preservation of hearing and facial nerve function was possible in a significant proportion of operated patients.6 The middle fossa approach was used infrequently until the development of gadolinium-enhanced magnetic resonance imaging (MRI). With this development, a larger number of acoustic tumors are diagnosed when they are small and before hearing has been significantly affected, making an attempt at hearing preservation desirable.
The middle fossa approach provides complete exposure of the contents of the IAC, allowing removal of laterally placed tumors without the need for blind dissection.7 This exposure ensures total removal and is well suited for the removal of very small acoustic tumors.8 The facial nerve can be located in its bony canal, allowing positive identification in a location not involved by tumor.
The middle fossa approach is technically difficult because of the lack of robust landmarks and the limited exposure. Bleeding in the posterior fossa can be difficult to control because of the limited access. Because of its location in the superior aspect of the IAC, the facial nerve is subjected to more manipulation in this approach than in other approaches.9,10 In the past, facial nerve results in middle fossa cases have not been as good as results from the translabyrinthine approach for similar-sized tumors.11 The routine use of the facial nerve monitor has helped improve these results, however, and now the results are comparable between the two approaches.12
Several authors use an extended middle fossa approach for large tumors.13–15 The tentorium is divided to give wider access to the posterior fossa. Some authors also perform a labyrinthectomy to enlarge the exposure when hearing preservation is not attempted.16–18
The primary indications for the middle fossa approach are a small acoustic tumor, in the IAC or with moderate extension into the cerebellopontine angle, and good preoperative hearing. In contrast, the retrosigmoid approach is used for patients with good hearing and small tumors located mostly in the cerebellopontine angle, without extension into the lateral aspect of the IAC. For hearing conservation surgery, we use the arbitrary audiometric criteria of speech reception threshold of better than 50 dB and speech discrimination score of better than 50%, although these indications must be individualized to the needs of the patient.19
Some patients have a subjective assessment about the usefulness of the preoperative hearing, such as the ability to use the telephone with the involved ear or the ability to localize a sound source. These subjective valuations should also be taken into consideration when determining hearing preservation candidacy. Some authors advocate attempting hearing preservation in the removal of small acoustic tumors if any measurable preoperative hearing exists.20 Patients older than 65 years do not tolerate the middle fossa approach as well as younger patients because of the fragility of the dura and retraction of the temporal lobe.
Several preoperative factors may predict postoperative hearing preservation. The most obvious is tumor size. Intuitively, the smaller the tumor, the easier it is to remove, and the more likely that hearing will be saved. This trend has been substantiated by several authors.11,21,22 Some authors have also found that the better the preoperative hearing, the more likely it will be preserved,23,24 whereas others have failed to identify such a relationship.11,22,25 Also, an intact preoperative stapedial reflex has been associated with successful postoperative hearing preservation.23
Several authors have reported a relationship between preoperative auditory brainstem response (ABR), and audiometry and hearing preservation.17,22 In one report, hearing was preserved in 78% of patients with an interaural wave V latency difference of 0.4 ms or less.11 For latency differences of 0.5 to 2 ms, the hearing preservation rate decreased to 58%. In patients with no response on the ABR, postoperative measurable hearing remained in only 50%. Patients with a more normal preoperative ABR result apparently have a greater success rate for postoperative hearing preservation. This result may reflect less tumor involvement of the cochlear nerve. Others do not find preoperative ABR to be predictive of hearing outcome.26,27
Tumors arising from the superior vestibular nerve have a higher rate of hearing preservation than tumors arising from the inferior vestibular nerve. Acoustic tumors developing in the inferior portion of the IAC may involve the cochlear nerve earlier and more severely.27,28 In a series of middle fossa acoustic tumor removals, 68% of patients whose tumors were found intraoperatively to arise from the superior vestibular nerve had measurable hearing preservation, whereas only 43% of patients whose tumors originated from the inferior vestibular nerve had measurable postoperative hearing.11 This difference was statistically significant.
Preoperative electronystagmography may predict tumor origin and hearing preservation. The caloric response reflects superior vestibular nerve function. In the presence of a small acoustic neuroma, a normal caloric response indicates an inferior vestibular nerve tumor, whereas a decreased response suggests a tumor arising from the superior vestibular nerve. Of a group of 54 patients who had preoperative electronystagmography, hearing was preserved in 64% with hypoactive caloric responses, whereas postoperative hearing remained in only 45% of patients with normal caloric responses.11 The association of normal caloric tests with nonpreservation of hearing has also been reported by others,10,17 although some series have not found this correlation.29 Vestibular evoked myogenic potentials may prove useful in determining the nerve of tumor origin. Early experience indicates that a combination of normal electronystagmography and a reduced vestibular evoked myogenic potentials predicts the inferior vestibular nerve as the nerve of origin.
For intracanalicular tumors, the radiographic appearance may predict success at hearing preservation. Small tumors that enlarge the IAC have a poorer prognosis for hearing preservation (Jackler RK, personal communication, 1990). In our experience, these small tumors that expand the canal are very adherent to the cochlear nerve, which adversely affects the hearing outcome. Fast spin echo MRI provides ultra-high-resolution images of IAC anatomy.30 With this imaging technique, it is possible to determine the nerve of origin for small tumors. Tumors that are impacted into the lateral end of the IAC, especially tumors that impinge on the cochlear nerve canal, have a lower rate of hearing preservation.31
After a thorough discussion of the relevant anatomy and the necessity to treat acoustic tumors, the options regarding surgical approaches to remove acoustic tumors are described to the patient. For patients with small tumors and good preoperative hearing, the issues of hearing preservation are discussed. We tell such patients that there is approximately a 50% chance of saving hearing. It is also important for patients to understand that hearing rarely improves after tumor removal.32 If the preoperative electronystagmography and ABR data are available and are favorable (see earlier), patients are informed that the prognosis for hearing preservation is above average.
The patient is told that there is approximately a 90% chance that normal or near-normal facial nerve function will be obtained in the long-term, but that there is a 20% to 30% chance of having temporary facial paresis in the early postoperative period. Although the facial nerve results for either the middle fossa or retrosigmoid approach are excellent, our best and most consistent facial nerve results for small tumors occur with the translabyrinthine approach.
Patients with preoperative tinnitus are counseled that the problem will likely get better, but probably will not disappear. Patients with no preoperative tinnitus have approximately a 25% chance of developing it postoperatively.33 Other important possible but rare complications are discussed, including cerebrospinal fluid leak, meningitis, serious brain complications, death, and blood transfusion options. The patient can donate 1 U of autologous blood before surgery, although transfusion is rarely needed.
Recuperation can take weeks to months, and most patients return to work within 4 to 6 weeks. The patient should expect to be dizzy postoperatively, and the rapidity of the central compensation greatly influences the time course of the recuperation.
Intraoperative furosemide and mannitol are given to allow easier temporal lobe retraction. A single dose of corticosteroid such as dexamethasone is routinely used intravenously at the beginning of surgery. This single dose of steroid does not seem to affect wound healing adversely. Long-acting muscle relaxants are avoided during surgery so as not to interfere with facial nerve monitoring. Preoperative antibiotics are administered. (Chapter 1 of this text details surgical site preparation and draping, along with the instruments used, including the House-Urban middle fossa retractor.)