Acoustic Neuroma

The operative mortality rate has dropped dramatically, from 40% in the early 20th century to less than 1%. With current microsurgery, postoperative facial paralysis, once the rule, is now an uncommon permanent sequelae. Attempts at hearing conservation, unimaginable at the beginning of the 20th century, are increasingly successful.

These very dramatic improvements are the result of the convergence of several factors. Vastly improved imaging techniques permitting early diagnosis, adaptation of the microscope, development of facial and auditory nerve–monitoring techniques, improved anesthesia, and improved perioperative management have all contributed to improved outcomes.

Frequency

Clinically diagnosed acoustic neuromas occur in 0.7-1.0 people per 100,000 population. The incidence may be rising, a reflection of the increasing frequency with which small tumors are being diagnosed with the more widespread use of MRI. A 2005 study by Lin et al suggested the prevalence of incidental acoustic neuromas to be 2 in 10,000 people.[4] Careful autopsy studies can detect small vestibular schwannomas in a higher percentage of elderly patients, which suggests that many acoustic neuromas never become clinically apparent.

Most patients diagnosed with an acoustic neuroma have no apparent risk factors. Exposure to high-dose ionizing radiation is the only definite environmental risk factor associated with an increased risk of developing an acoustic neuroma. Multiple studies have determined cell phone use is not associated with an increased risk of developing an acoustic neuroma, although data on the effects of long-term cell phone use are still pending.[5]

Neurofibromatosis type II occurs in individuals who have defective tumor suppressor gene located on chromosome 22q12.2. The defective protein produced by the gene is called merlin or schwannomin. Bilateral acoustic tumors are a principle clinical feature of neurofibromatosis type II, although other manifestations, including peripheral neurofibromata, meningioma, glioma, and juvenile posterior subcapsular lenticular opacities, are often present as well. Many patients with neurofibromatosis type II present in late adolescence or early adulthood but occasionally may present later in the fifth to seventh decade with slowly growing tumors.

The vast majority of acoustic neuromas develop from the Schwann cells investing the vestibular portion of the vestibulocochlear nerve. Less than 5% of tumors arise from the cochlear nerve. The superior and inferior vestibular nerves appear to be the nerves of origin with about equal frequency. Overall, three separate growth patterns can be distinguished within acoustic tumors, as follows: (1) no growth or very slow growth, (2) slow growth (ie, 2 mm/y), and (3) fast growth (ie, ≥ 1.0 cm/y on imaging studies). Although most acoustic neuromas grow slowly, rarely, a tumor may grow quickly and may double in volume within 6 months to a year.

Although some tumors adhere to one or another of these growth patterns, others appear to alternate between periods of no or slow growth and rapid growth. Tumors that have undergone cystic degeneration (presumably because they have outgrown their blood supply) are sometimes capable of relatively rapid expansion because of enlargement of their cystic component. Because acoustic tumors arise from the investing Schwann cell, tumor growth generally compresses vestibular fibers on the surface. Destruction of vestibular fibers is slow; consequently, many patients experience little or no disequilibrium or vertigo. Once the tumor has grown sufficiently large to fill the internal auditory canal, it may continue growth either by expanding bone or by extending into the cerebellopontine angle. Growth within the cerebellopontine angle is generally spherical, which is different than the sessile growth pattern seen in a meningioma of the cerebellopontine angle.

Acoustic tumors, like other space-occupying lesions, produce symptoms by any of four recognizable mechanisms: (1) compression or distortion of the spinal fluid spaces, (2) displacement of the brain stem, (3) compression of vessels producing ischemia or infarction, or (4) compression and/or attenuation of nerves.

Because the cerebellopontine angle is relatively empty, tumors can continue to grow until they reach 3-4 cm in size before they contact important structures. Growth is often sufficiently slow that the facial nerve can accommodate to the stretching imposed by tumor growth without clinically apparent deterioration of function. Tumors that arise within the internal auditory canal may produce early symptoms in the form of hearing loss or vestibular disturbance by compressing the cochlear nerve, vestibular nerve, or labyrinthine artery against the bony walls of the internal auditory canal.

As the tumor approaches 2 cm in diameter, it begins to compress the lateral surface of the brain stem. Further growth can occur only by compressing or displacing the brain stem toward the contralateral side. Tumors greater than 4 cm often extend sufficiently far anteriorly to compress the trigeminal nerve and produce facial hypesthesia. As the tumor continues to grow beyond 4 cm, progressive effacement of the cerebral aqueduct and fourth ventricle occurs with eventual development of hydrocephalus.

Treatment depends on multiple factors including the age and medical status of the patient, tumor size and location, hearing status, and patient preference. In older patients with small tumors, careful observation may be elected consisting of serial MRIs. In older patients with a growing tumor, radiosurgery may be an appropriate option. Young patients, large tumors (greater than 2.5 to 3 cm) and patients with small tumors and intact hearing may choose surgery. See Surgical therapy.

The cerebellopontine angle is a space filled with spinal fluid. It has the brain stem as its medial boundary, the cerebellum as its roof and posterior boundary, and the posterior surface of the temporal bone as its lateral boundary. The floor of the cerebellopontine angle is formed by the lower cranial nerves (IX-XI) and their surrounding arachnoid investments. The flocculus of the cerebellum may lie within the cerebellopontine angle and may be closely associated with cranial nerves VIII and VII as they cross the cerebellopontine angle to enter the internal auditory canal.

The facial nerve arises 2-3 mm anterior to the root entry zone of the vestibulocochlear nerve. The foramen of Luschka (ie, the opening of the lateral recess of the fourth ventricle) is located just inferior and posterior to the root entry zones of the facial and vestibulocochlear nerve. A tuft of choroid plexus can frequently be observed extruding from it. Inferior and a bit anterior to the foramen of Luschka is the olive, and just posterior to the olive lie the rootlets of origin for cranial nerves IX, X, and XI. The hypoglossal nerve exits the brain stem through a series of small rootlets anterior to the olive.

Vascular structures within the cerebellopontine angle

The most important vascular structure within the cerebellopontine angle is the anterior inferior cerebellar artery (AICA). It arises most commonly as a single trunk from the basilar artery but can arise as two separate branches. In rare cases, it originates as a branch of the posterior inferior cerebellar artery (PICA). As the AICA moves from anterior to posterior, it first follows the ventral surface of the brain stem, but within the cerebellopontine angle it takes a long loop laterally to the porus acusticus. In 15-20% of cases, the AICA actually passes into the lumen of the internal auditory canal before turning back on itself toward the posterior surface of the brain stem. (These AICA loops are not symptomatic.) The AICA can thus be divided into the premeatal, meatal, and postmeatal segments.

The main branch of the AICA passes over cranial nerves VII and VIII in only 10% of cases. The remainder of the time, it either passes below the VII and VIII cranial nerves or, in 25-50% of individuals, actually passes between them. Three branches that regularly arise from the meatal segment of the AICA can be identified. Small perforating arteries supply blood to the brain stem. The subarcuate artery passes through the subarcuate fossa into the posterior surface of the temporal bone, and the third regular branch is the internal auditory artery (labyrinthine artery). Cranial nerves VII and VIII receive their blood supply from small branches of AICA.

Two venous structures must be kept in mind during surgical procedures involving the cerebellopontine angle. The petrosal vein (of Dandy) brings returning venous blood from the cerebellum and lateral brain stem to the superior or inferior petrosal sinus. It is generally encountered in the area of the trigeminal nerve anterior to the porus acusticus. The petrosal vein often carries enough venous blood that its obstruction can lead to venous infarction and cerebellar edema, and it should be preserved if at all possible. Additional venous blood reaches the superior petrosal sinus through a series of bridging veins that cross the cerebellopontine angle. Although every attempt should be made to preserve these veins, their sacrifice is generally inconsequential.

The vein of Labbé carries returning venous blood from the inferior and lateral surface of the temporal lobe to the superior petrosal sinus, tentorial venous lakes, or the transverse sinus. Its configuration and anatomy is quite variable. However, obstruction, obliteration, or occlusion of the superior petrosal sinus may, in some cases, result in occlusion of the vein of Labbé. Sudden occlusion of the vein of Labbé carries with it a high risk of venous infarction of the temporal lobe and rapid life-threatening cerebral edema.

Nerves

The facial nerve leaves the brain stem anterior to the foramen of Luschka. As it leaves the brain stem, the fibers are sheathed in oligodendroglia derived from the central nervous system. Within a few millimeters of leaving the brain stem, however, the nerve loses its oligodendroglial ensheathment and becomes ensheathed instead by Schwann cells. Throughout the remainder of its peripheral course, it remains within its Schwann cell investment. It passes directly across the cerebellopontine angle for about 15 mm, accompanied by the vestibulocochlear nerve. It consistently enters the internal auditory canal by crossing the anterior superior margin of the porus acusticus.

The vestibulocochlear nerve arises from the brain stem slightly posterior to the facial nerve. It remains sheathed in oligodendroglia for approximately 15 mm (almost to the point at which it passes into the internal auditory canal). It has the longest oligodendroglial investment of any peripheral nerve. The junction between oligodendroglia and Schwann cells (ie, the Obersteiner-Redlich zone) occurs just medial to the porus acusticus. Because acoustic neuromas arise from Schwann cells, they arise most commonly within the most lateral portions of the cerebellopontine angle or the internal auditory canal.

The nervus intermedius (nerve of Wrisberg) leaves the brain stem together with the vestibulocochlear nerve. At some point within the cerebellopontine angle, the nervus intermedius crosses over to become associated with the facial nerve. It may do so as several separate rootlets. The point where the nervus intermedius crosses to become associated with the facial nerve shows considerably variation, but in 22% of individuals, it is adherent to the vestibulocochlear nerve for 14 mm or more. As the vestibulocochlear and facial nerve reach the porus acusticus (medial opening of the internal auditory canal) they pass together with the nervus intermedius and sometimes a loop of AICA.

Internal auditory canal

The internal auditory canal is approximately 8.5 mm in length (range 5.5-10.5 mm), lined with dura, and filled with spinal fluid. Its medial end is oval in shape and is referred to as the porus acusticus. Its lateral end is a complicated structure referred to as the fundus or lamina cribrosa. The fundus is divided into a superior and inferior half by the transverse crest. The upper half is further subdivided into an anterior and posterior segment by a vertical crest, often referred to as Bill’s Bar, named after William House, who popularized its importance as a surgical landmark. The vertical crest separates the macula cribrosa superior, a series of very small openings through which the terminal fibers of the vestibular nerve pass in order to reach the cupula of the superior semicircular canal, from the meatal foramen, which marks the point at which the facial nerve leaves the internal auditory canal and enters the fallopian canal as the labyrinthine segment.

Because the most lateral portion of the internal auditory canal is 4-5 mm inferior to the level of the geniculate ganglion, the labyrinthine segment of the facial nerve must take a vertically oriented course upward to reach it. The labyrinthine segment may be less than a millimeter wide as it passes between the cochlea and the anterior end of the superior semicircular canal. The inferior portion of the fundus is a single oval-shaped space, the anterior portion of which is occupied by a rounded depression (tractus spiralis foraminosus) filled with small openings to accommodate the terminal branches of the cochlear nerve. The posterior portion is filled with a macula crista inferior through which pass the terminal ends of the inferior vestibular nerve.

Temporal bone

The anatomy of the superior surface of the temporal bone must be mastered if middle fossa approaches are to be undertaken successfully. Laterally, the irregular superior surface of the temporal bone transitions relatively smoothly to the temporal squamosa. The free edge of the tentorium and the superior petrosal sinus attach to the medial edge of the superior surface of the temporal bone. The arcuate eminence, a bony prominence that is perpendicular to the petrous ridge and lies two centimeters medial to the squamous temporal bone, often overlies the superior semicircular canal. The arcuate eminence is often difficult to identify, especially in well-pneumatized temporal bones.

The geniculate ganglion usually lies within the substance of the temporal bone just medial to and a few millimeters anterior to the head of the malleus. The geniculate ganglion may be dehiscent, or alternatively, it may lie several millimeters beneath the superior surface of the bone. The head of the malleus is generally easy to identify if the thin bone of the tegmen tympani is removed so as to enter into the middle ear space. In difficult surgical situations, the head of the malleus can be used to identify the geniculate ganglion. The greater superficial petrosal nerve originates from the geniculate ganglion and courses anteromedially, passing over the superior surface of the temporal bone at the facial hiatus. The facial hiatus is generally 4-8 mm anterior to the geniculate ganglion. The greater superficial petrosal nerve can generally be identified in this area. It can then be followed retrograde to the geniculate ganglion.

The middle meningeal artery and associated veins traverse the foramen spinosum, which is located approximately 1 cm anterolaterally to the greater superficial petrosal nerve. The mandibular division of the trigeminal nerve traverses the foramen ovale, which lies a few millimeters anterior and medial to the foramen spinosum. The horizontal portion of the carotid canal courses through the anterior temporal bone medial to the foramen spinosum and foramen ovale. The cochlea cannot be identified from the surface appearance of the superior temporal bone. It lies just anterior and inferior to the labyrinthine segment of the facial nerve but is deep to the geniculate ganglion.

Few absolute contraindications to the surgical removal of an acoustic tumor exist. Serious medical illness may make surgical removal in some patients too risky. Surgery must often be performed for large tumors with brain stem shift and obstructive hydrocephalus, even in the presence of significant medical illness. The translabyrinthine approach is contraindicated in a patient with chronic otitis media.

The decision to operate should be carefully considered when the tumor is within the internal auditory canal of a patient's only hearing ear. In some cases, observing the tumor until hearing has been lost is best, while in other cases, attempting surgical removal with hearing conservation is more prudent.

Unilateral hearing loss is overwhelmingly the most common symptom present at the time of diagnosis and is generally the symptom that leads to diagnosis. Assume that any unilateral SNHL is caused by an acoustic neuroma until proven otherwise; an MRI scan with gadolinium contrast should be obtained. The tumor can produce hearing loss through at least two mechanisms, direct injury to the cochlear nerve or interruption of cochlear blood supply. Progressive injury to cochlear fibers probably accounts for slow progressive neurosensory hearing loss observed in a significant number of patients with acoustic neuromas. Sudden and fluctuating hearing losses are more easily explained on the basis of disruption of cochlear blood supply.

Consistent with direct injury to cranial nerve VIII, a significant number of individuals with acoustic neuroma have speech discrimination scores reduced out of proportion to the reduction in the pure-tone average—a feature typical for retro-cochlear lesions. This can often be demonstrated through audiological testing by a phenomenon called "rollover," in which speech discrimination scores decrease as the volume of the speech stimulus increases. Such marked reductions in speech discrimination scores are not invariable, however. A normal speech discrimination score does not rule out an acoustic tumor. Patients with acoustic tumors may have normal or near-normal hearing and speech discrimination scores.

A study by Lee et al of over 100 patients with acoustic neuroma found that tinnitus was the most frequent symptom accompanying hearing loss. It was also determined that approximately 65% of patients had asymmetrical hearing and that an association existed between greater severity of deafness and lower speech discrimination scores.[6]

Hearing loss associated with acoustic neuroma can be sudden or fluctuating in 5-15% of patients. Such hearing loss may improve spontaneously or in response to steroid therapy. However, a gadolinium enhanced MRI should be ordered in anyone with a sudden or fluctuating loss even if hearing returns to normal.

Not surprisingly, the discovery of acoustic neuromas in persons with normal hearing has been increasing as gadolinium-enhanced MRI is becoming more common. In addition, tumor size poorly correlates with hearing status as patients with large tumors may have normal hearing and patients with small tumors may be profoundly deaf in the affected ear. The presence of unilateral tinnitus alone is a sufficient reason to evaluate an individual for an acoustic neuroma. Although tinnitus is most commonly a manifestation of hearing loss, a few individuals with acoustic tumors (around 10%) seek treatment for unilateral tinnitus without associated subjective hearing loss.

Vertigo and disequilibrium are uncommon presenting symptoms among patients with an acoustic neuroma. Rotational vertigo (the illusion of movement or falling) is uncommon and is occasionally seen in patients with small tumors. Disequilibrium (a sense of unsteadiness or imbalance), on the other hand, appears to be more common in larger tumors. Overall, if carefully questioned, approximately 40-50% of patients with an acoustic neuroma report some balance disturbance. However, balance disturbance is the presenting symptom in less than 10% of patients. The destruction of vestibular fibers apparently is sufficiently slow as to permit compensation. If patients have persistent imbalance or dizziness, vestibular rehabilitation is often beneficial, even if the tumor remains untreated.

Headaches are present in 50-60% of patients at the time of diagnosis, but fewer than 10% of patients have headache as their presenting symptom. Headache appears to become more common as tumor size increases and is a prominent feature in patients who develop obstructive hydrocephalus associated with a very large tumor.

Facial numbness occurs in about 25% of patients and is more common at the time of presentation than facial weakness. Objective hypoesthesia involving the teeth, buccal mucosa, or skin of the face is associated with larger tumors, but a subjective reduction in sensation that cannot be documented on objective examination occurs commonly with medium-sized and small tumors. Decrease in the corneal reflex generally occurs earlier and more commonly than objective facial hypoesthesia. Although approximately 50-70% of individuals with large tumors have objectively demonstrable facial hypoesthesia, they are often unaware of it, and it is uncommonly the presenting symptom.

The motor fibers in the facial nerve can accommodate substantial stretching as long as it occurs slowly and are much more resistant to injury than sensory fibers of the trigeminal nerve. Facial weakness is sufficiently uncommon (< 1%) that facial weakness associated with a small- or medium-size tumor should raise suspicion that it is not an acoustic neuroma. Other diagnosis should be considered including facial neuroma, hemangioma, meningioma, granuloma, arteriovenous malformation (AVM), or lipoma.

Large tumors (>4 cm) can obstruct the flow of spinal fluid through the ventricular system by distorting and obstructing the fourth ventricle. In the early decades of this century, 75% of patients presented with hydrocephalus. Although hydrocephalus is not as common a presenting symptom as it once was, recognizing its symptoms (headaches, impaired vision, cognitive difficulties, seizures, neurologic deficits) is important to potentially diagnosing a larger, obstructing tumor.

Although a thorough head and neck exam, including otoscopy, is recommended during evaluation of a patient, there are often no specific findings on exam that will detect an acoustic neuroma. Examination is important to help rule out other possible otologic diagnoses that may be more readily detectable on exam and may account for presenting symptoms that are overlapping those of acoustic neuroma. Subtle facial numbness or paresthesias or a lack of a corneal reflex suggest a tumor affecting the trigeminal nerve. Audiometry is recommended as an adjunct during the initial visits of a patient undergoing workup. 

Routine lab studies are generally not required. If microsurgery is considered, a blood type and screen and baseline hemoglobin should be done.

See the list below:

• The definitive diagnostic test for patients with acoustic tumors is gadolinium-enhanced MRI.

  • Well-performed scanning can demonstrate tumors as small as 1-2 mm in diameter. On the other hand, thin-cut CT scanning can miss tumors as large as 1.5 cm even when intravenous contrast enhancement is used.

  • Gadolinium contrast is critical because nonenhanced MRI can miss small tumors.

  • Fast-spin echo techniques do not require gadolinium enhancement and can be performed very rapidly and relatively inexpensively. However, such highly targeted techniques risk missing other important causes of unilateral sensory hearing loss, including intra-axial tumors, demyelinating disease, and infarcts. Nonetheless, these techniques are useful if a tumor is being observed, since the tumor is known to be present, and the size of tumor can be easily measured.

  • MRI is contraindicated in individuals with ferromagnetic implants.

• Fine-cut CT scanning of the internal auditory canal with contrast can rule out a medium-size or large tumor but cannot be relied upon to detect a tumor smaller than 1-1.5 cm.

• If suspicion is high and MRI is contraindicated, air-contrast cisternography has high sensitivity and can detect relatively small intracanalicular tumors. However, this procedure is rarely performed, now being of historical interest.

A variety of audiometric tests were developed in the mid-20th century in an attempt to identify patients with increased likelihood of having an acoustic neuroma. That was a worthwhile undertaking when definitive radiographic imaging consisted of some form of either pneumoencephalography or formal arteriography. Such testing is no longer used. Even the auditory brain stem evoked response (ABR) is now infrequently used as a screening test for acoustic neuroma. ABR screening techniques miss 20-35% of acoustic tumors smaller than 1 cm. Moreover, ABR is likely to miss those tumors in patients with excellent hearing, which are the cases most favorable for hearing conservation procedures.

Two histologic types of tissue have been identified in acoustic tumors. Antoni A tissue consists of elongated spindle cells with a palisading pattern. Antoni B tissue, on the other hand, has a loose spongy texture and markedly reduced cellularity. A given acoustic neuroma may contain areas with both Antoni A and Antoni B tissue. Another histologic feature characteristic of schwannomas are rows of palisading nuclei called Verocay bodies. Although the histologic appearance of acoustic tumors is fairly straightforward, they can occasionally be difficult to distinguish from meningiomas. Immunohistochemical staining can distinguish schwannomas from meningiomas in difficult cases. Schwannomas are immunoreactive to S-100 antibody while meningiomas are immunoreactive to epithelial membrane antibody (EMA).

Koos staging system

No widely accepted staging system exists for acoustic neuromas. However, the Koos staging system, as follows, is often cited in publications[7, 8] :

• Grade 1 - Tumor involvement includes only the internal auditory canal
• Grade 2 - The tumor extends into the cerebellopontine angle but does not contact the brain stem; maximum tumor diameter is 20 mm
• Grade 3 - The tumor fills the cerebellopontine angle without brain stem displacement
• Grade 4 - The brain stem is compressed by the tumor, and the cranial nerves are displaced

Acoustic neuromas are managed in one of the following three ways: (1) surgical excision of the tumor, (2) arresting tumor growth using stereotactic radiation therapy, or (3) careful serial observation.

Observation

Simple observation with routine imaging and audiologic evaluation, without any therapeutic intervention, can be considered in the following groups of patients:

• Elderly patients

• Patients with small and asymptomatic tumors

• Patients with medical conditions that significantly increase the risk of operation

• Patients who refuse treatment

• Patients with a tumor on the side of an only hearing ear

During an observation period, most (70% or more) patients who were eligible for hearing conservation surgery initially lost their eligibility.

There is controversy over whether certain management modalities better preserve hearing over time. A cross-sectional study of long-term hearing preservation following observation, primary microsurgery, and stereotactic radiosurgery found observation to be associated with the highest rate of hearing preservation.[9] Other studies have reported less hearing loss in patients undergoing stereotactic radiosurgery compared with observation, while still other studies have shown no difference.[10]

Telian has analyzed important variables that should be evaluated when observation is considered, including the following: 1) preoperative hearing in both ears, 2) the risk of immediate hearing loss as a consequence of surgery, 3) the risk of facial nerve paralysis, 4) the risk of other surgical complications and their seriousness, 5) the patient's life expectancy, 6) the size of the tumor, 7) tumor growth rate, and 8) patients with neurofibromatosis type 2 (NF2) or bilateral tumors.[11]

The percentage of tumors with growth varies from 30-70% among studies. On average, over a 5-year period, about half of tumors may be expected to grow.[10]  The majority of growth occurs in the first few years of follow-up,[12]  and the occurrence of growth within the first year may be an early indicator of future need for intervention.[13] Studies have indicated that patients who are being observed ultimately require therapeutic intervention at rates ranging from 15-40%.

The risk of hearing loss during the observational period remains a significant consideration. A systemic review of hearing preservation in observed tumors demonstrated that half of patients will have preserved good or serviceable hearing after 5 years.[14] Speech discrimination scores and good hearing at high frequencies predict a higher likelihood for functional hearing preservation.[15] ​

Hunter et al showed correlation between baseline pure-tone audiometry and speech discrimination scores and maintaining serviceable hearing. Based on audiologic testing at the time of diagnosis, their data predicted a two-fold likelihood of developing nonserviceable hearing for every 10 dB increase in pure-tone average and a 1.5-fold likelihood of developing nonserviceable hearing for every 10% decline in speech discrimination score.[16]

There are no universally agreed upon factors that can reliably predict which tumors will grow, though some studies have shown growth to be associated with initial tumor size and certain symptoms. Social factors should also be evaluated to help determine if a patient is an appropriate candidate for observation.

In a study measuring volumetric growth of observed vestibular schwannomas, initial tumor size, nonincidental diagnosis, symptoms of disequilibrium, facial hypoesthesia, and aural fullness were associated with tumor growth.[17]

Ostler et al found that older age, unmarried status, and smaller tumor size on diagnosis were associated with a higher likelihood of choosing observation over treatment. Conversely, patients who lived a longer distance from the treatment center were more likely to choose treatment.[18]

Intervention is indicated when tumors exhibit significant growth, which is generally considered greater than 2 mm per year. Options include either microsurgery or radiosurgery, and optimal therapy is dependent on such factors as tumor size, patient age, and hearing status.[19]

Stereotactic radiosurgery

Stereotactic radiosurgery has emerged as an alternative to microsurgery for selected patients with acoustic neuroma. Stereotactic radiation therapy makes use of one of several radiation sources and is administered using a variety of different machines with proprietary names (eg, Gamma Knife, CyberKnife, Brainlab).

Stereotactic radiosurgery uses radiation delivered to a precise point or series of points to maximize the amount of radiation delivered to target tissues while minimizing the exposure of adjacent normal tissues. It can be delivered as a single dose or as multiple fractionated doses.

The effects of radiation delivered at the current low dose (12-13 Gy) likely prevents further tumor growth by causing obliterative endarteritis of the vessels supplying the tumor. Radiosurgery may affect tumor cells undergoing mitosis by causing double strand DNA breaks. Hansen et al demonstrated acoustic neuroma cells are radioresistant at the current low-dose radiation used with radiosurgery.[20]

A study by Boari et al study of patients with an acoustic neuroma found Gamma Knife radiotherapy for these tumors to be safe and effective, providing control in 97.1% of the study’s patients and tumor volume reduction in 82.7% of them; the mean relative volume reduction in the latter group was 34.1%.[21]

Advantages of radiation therapy include the following:

• Outpatient procedure

• Decreased cost, initially

• Most patients return to work the following day

• Lower immediate posttreatment morbidity and mortality

• Less chance for facial nerve dysfunction

Disadvantages of stereotactic radiation include the following:

• Necessity for regular monitoring and frequent rescanning (in the end, costs associated with long-term monitoring could exceed those of surgery)

• Does not eliminate the tumor and may fail to control tumor growth, sometimes requiring salvage surgery

• Higher incidence of trigeminal nerve injury

• Unknown long-term incidence of secondary malignancies. The best current estimates of developing a secondary malignancy from the radiosurgery are 1 in 1000 patients over 30 years

• Does not address disequilibrium and may lead to long-term balance dysfunction

Stereotactic radiosurgery and fractionated stereotactic radiotherapy have the potential for hearing preservation, at least in the short-term. Hearing preservation is dependent on multiple factors, including tumor size, tumor location, and radiation dose. Most centers use a dose of 12-13 Gy at the 50% isodose line when considering hearing preservation. Hearing preservation is also dependent on the radiation dose to the cochlea, cochlear nerve, and cochlear nucleus. Kim et al recently noted transient volume expansion that is commonly seen after radiosurgery portends the worse prognosis for hearing preservation.[22]  Most patients will eventually lose hearing over a long enough period of time.

Microsurgical removal remains the treatment of choice for tumor eradication. Various surgical approaches can be used to remove acoustic tumors. Each approach is discussed in detail in the following sections.

A prospective, observational study by Nellis et al indicated that in patients with acoustic neuroma, those most likely to pursue treatment with surgical resection rather than active surveillance are persons under age 65 years with medium to large tumors; growing tumors; significant hearing loss; and higher headache severity scores.[23]

In a retrospective study of 4137 patients who underwent vestibular schwannoma surgery, including 519 elderly persons (aged 65 years or older), Sylvester et al found that elderly patients had a higher rate of comorbidities (including diabetes mellitus, hypertension, and pulmonary disease) and in-hospital complications (including acute cardiac events, iatrogenic cerebrovascular infarction/hemorrhage, postoperative bleeding, and mortality), as well as greater lengths of hospital stay (6.5 days vs 5.4 days in nonelderly patients).[24]

Three different approaches are used in the management of acoustic neuromas, the retrosigmoid, translabyrinthine, and middle fossa approaches. All have advantages and disadvantages as indicated below.

Advantages of the retrosigmoid approach

The retrosigmoid approach can be applied to all acoustic tumors and to many other histologic tumor types. It can be used for operations that sacrifice hearing and operations that attempt to conserve hearing. Its only limitation in this respect is its inapplicability for small tumors that occupy the far-lateral portions of the internal auditory canal (ie, the fundus of the canal).

The retrosigmoid approach provides the best wide-field visualization of the posterior fossa. The inferior portions of the cerebellopontine angle and the posterior surface of the temporal bone anterior to the porus acusticus are more clearly observed than via the translabyrinthine approach. Panoramic visualization is especially helpful when displacement of nerves is not predictable, which occurs commonly with meningiomas.

Since destruction of the labyrinth is not required as part of the retrosigmoid approach, hearing conservation surgery can be attempted for tumors up to 2 cm via this approach.

Disadvantages of the retrosigmoid approach

The retrosigmoid approach may require cerebellar retraction or resection. Manipulation of the cerebellum provides opportunities for postoperative edema, hematoma, infarction, and bleeding.

Increased incidence of cerebrospinal fluid leak occurred in some series.

The retrosigmoid approach is associated with greater likelihood of severe protracted postoperative headache.

Advantages of the translabyrinthine approach

The translabyrinthine approach provides the best view of the lateral brain stem facing the acoustic tumor. Retraction of the cerebellum is rarely necessary.

The fundus of the internal auditory canal is completely exposed; the facial nerve can be identified at a location where it is undistorted by tumor growth and compressed into the labyrinthine segment, decreasing the risk of delayed postoperative facial nerve palsy.

Incidence of cerebrospinal fluid leak is decreased in some series.

Disadvantages of the translabyrinthine approach

The translabyrinthine approach always results in profound hearing loss.

The inferior portions of the cerebellopontine angle and cranial nerves are not as well visualized as they are in the retrosigmoid approach. The temporal bone anterior to the porus acusticus is also less well visualized.

A fat graft is required. Removal of fat from the abdomen creates opportunities for donor site complications, including hematoma, bleeding, and infection.

The sigmoid sinus is more vulnerable to injury. Bleeding from the sigmoid sinus can be difficult to control and can significantly increase operative blood loss. If a dominant sigmoid sinus is occluded during the operation, postoperative intracranial pressure elevation or venous infarct can occur.

A high jugular bulb or anteriorly placed sigmoid sinus can substantially compromise the space available for tumor removal. Occasionally, the space is so contracted that another approach has to be selected.

Advantages of the middle cranial fossa approach

The middle cranial fossa approach is the only procedure that fully exposes the lateral third of the internal auditory canal without sacrificing hearing. Another advantage is that the surgery is extradural.

Disadvantages of the middle cranial fossa approach

The facial nerve generally courses across the anterior superior portion of the tumor. Consequently, it is in the way during tumor removal and is more vulnerable to injury. Although long-term facial nerve outcomes are as good with the middle cranial fossa approach as with other approaches, temporary postoperative paresis is more common.

The risk of dural laceration and avulsion becomes increasingly more likely as patients become older. The dura mater in elderly patients is more friable. This becomes especially noticeable during the sixth and seventh decades of life.

The approach provides limited exposure of the posterior fossa. Also, the operation is technically difficult and demanding.

Some patients incur postoperative trismus related to manipulation and/or injury to the temporalis muscle.

The temporal lobe must be retracted, presenting the opportunity for temporal lobe injury, usually in the form of a hematoma that is asymptomatic and, therefore, probably occurs more frequently than is realized. Scattered reports exist of seizure disorder following middle cranial fossa surgery, presumably due to temporal lobe injury.

Approach Selection

A variety of different considerations go into deciding which approach should be used for any individual patient. These variables are detailed below.

Preoperative hearing level

If the patient has no useful hearing, either the translabyrinthine or the retrosigmoid approach is selected, depending upon the experience and training of the surgeon. In most centers performing large numbers of surgeries for acoustic tumors, the translabyrinthine approach is preferred. Opinions vary considerably about what constitutes useful hearing. The 50/50 rule is frequently quoted. The rule suggests that individuals with a pure-tone average greater than 50 dB and speech discrimination less than 50% do not have useful or salvageable hearing. Other surgeons have stricter criteria and consider only individuals with better than a 30-dB pure-tone average and more than 70% discrimination for hearing conservation operations.

Auditory brain stem response

Normal preoperative ABR findings favor hearing conservation. Marked abnormalities of ABR wave morphology or increased wave I-III and I-V latencies make hearing conservation less likely.

Electronystagmography

An abnormal caloric test on electronystagmography (ENG) increases the likelihood of successful hearing conservation surgery. The ENG tests the horizontal semicircular canal, which is innervated by the superior vestibular nerve. A normal ENG finding arguably demonstrates that the superior vestibular nerve is normal. Consequently, the acoustic tumor must have originated from the inferior vestibular nerve, which is directly adjacent to the cochlear nerve. Surgical removal, then, is more likely to directly injure the cochlear nerve or interfere with cochlear blood supply. Vestibular evoked myogenic potential (VEMP) testing is abnormal when the inferior vestibular nerve is affected. As a result, an abnormal VEMP with normal caloric testing on ENG strongly suggests an inferior vestibular nerve tumor with poorer hearing preservation.

Tumor size

Opportunities for hearing conservation decrease as tumors become larger. Hearing is much more difficult to conserve when tumors are 1.5-2.0 cm in diameter than if they are small intracanalicular tumors. Consequently, some surgeons limit hearing conservation surgery to smaller tumors, preferring to use a translabyrinthine approach to maximize the chance of facial nerve preservation for larger tumors.

Tumor position

If hearing conservation is to be attempted and the tumor lies within the lateral portions of the internal auditory canal (fundus), many surgeons prefer a middle fossa approach. The middle fossa approach permits direct exposure of the fundus of the internal auditory canal without sacrificing hearing. The approach is frequently used for any tumor lying completely within the internal auditory canal, although tumors limited to the medial portions of the internal auditory canal can be managed using a retrosigmoid approach. Some surgeons extend the use of the middle fossa technique to include tumors that extend as much as 5-10 mm into the cerebellopontine angle. Division of the superior petrosal sinus may be required to gain sufficient access to the posterior fossa with larger tumors.

Generally, however, tumors that have significant volume medial to the plane of the porus acusticus are removed using a retrosigmoid approach if hearing is to be conserved. If hearing conservation is not an issue, the retrosigmoid approach is sometimes preferred for tumors with significant inferior extension since the lower cranial nerves are better visualized with a retrosigmoid approach. Occasionally, the retrosigmoid approach is combined with a translabyrinthine approach for such large acoustic neuromas.

Relevant anatomy

The following anatomic variations can make the translabyrinthine approach much more difficult and at times impossible.

• High-riding jugular bulb: In some individuals, the jugular bulb may be at the level of the inferior internal auditory canal, making adequate exposure for tumor removal difficult.

• Anteriorly placed sigmoid sinus: In such circumstances, the distance between the sigmoid sinus and the external auditory canal may be a few millimeters or less. Such a dramatic limitation of the space within which the surgeon has to operate not only makes a successful tumor removal much more difficult but puts the facial nerve and the displaced sinus itself at significantly increased risk of injury.

• Contracted sclerotic mastoid: Such mastoid cavities provide little room for tumor removal. Moreover, they are often associated with suppurative otitis media, in itself a contraindication to the translabyrinthine approach.

• Reduced or absent flow in the contralateral sinus: Previous operation, trauma, congenital anomalous development, and previous or concurrent disease can all result in markedly reduced or absent venous outflow through the contralateral sinus. In such cases, consideration may be given to a retrosigmoid approach merely because it reduces the risk of injury to the remaining sinus, occlusion of which would result in catastrophic venous infarction.

Surgeon preference

Some surgeons have more experience and are much more comfortable with one approach relative to another. Generally, such preferences should be followed. However, if hearing conservation is a realistic option using an approach unfamiliar to the primary surgeon, consideration should be given to referring the patient to someone who is familiar with the appropriate approach.

Patient preference

Patient preferences should be carefully considered even when they do not conform to the surgeon's judgment. Some patients are adamant about going to any lengths for hearing conservation even when the treating physician is quite convinced that the patient's hearing is so poor as to be of little or no practical utility. Some patients willingly sacrifice even good hearing if doing so even slightly enhances the possibility of successful facial nerve preservation.

Translabyrinthine approach

The translabyrinthine approach is the most versatile of the three common approaches to the cerebellopontine angle. The main disadvantage is profound deafness in the operated ear due to violation of the membranous labyrinth. In general, even the largest acoustic neuromas can be removed through a translabyrinthine craniotomy. In addition, the facial nerve is found at the fundus of the internal auditory canal where the vertical crest (Bill’s bar) provides a natural plane for facial nerve dissection from the superior vestibular nerve. At the author’s institution, the translabyrinthine approach is preferred with any acoustic neuroma over 2 cm or if the patient has poor preoperative hearing.

The patient is laid supine and a Mayfield head frame may be used. An incision is then made two finger-breadths from the postauricular sulcus. The temporalis muscle and mastoid periosteum are identified. The skin flap is then elevated anteriorly, leaving as much periosteum down as possible. The periosteum is then incised along the linea temporalis and then towards the mastoid tip. This will allow a water-tight second layer for closure to prevent postoperative cerebrospinal fluid leakage. The mastoid periosteum is then elevated from the underlying mastoid bone. Often, the emissary vein is encountered and this can be controlled with bipolar coagulation and/or bone wax.

A wide cortical mastoidectomy is performed. The middle and posterior fossa dura are identified as well as the sigmoid sinus. The bone is removed from these structures to allow retraction of the temporal lobe dura and sigmoid sinus. Next, the antrum, lateral semicircular canal, and vertical facial nerve are identified.

The incus is removed and a facial recess is performed. The tensor tympani tendon is sectioned and the eustachian tube is packed with oxidized cellulose packing. The middle ear space is then packed with temporalis muscle.

A labyrinthectomy is performed and the jugular bulb is identified. The internal auditory canal is subsequently identified and troughs are developed both superiorly and inferiorly around the internal auditory canal until approximately 270° of internal auditory canal is exposed. The remaining bone is then removed from the internal auditory canal and the facial nerve is found as it turns into the labyrinthine segment. The superior vestibular nerve is then followed out to the ampullated end of the superior semicircular canal.

At this point, the transverse crest and vertical crest (Bill’s bar) are identified. The superior vestibular nerve is then reflected inferiorly from the ampullated end of the superior semicircular canal. The facial nerve can often be found superior and medial to this and is confirmed using a facial nerve stimulator. At this point, the tumor is debulked and the facial nerve is located at the origin from the brain stem. Once the tumor is adequately debulked, the acoustic neuroma is then dissected from the facial nerve. Often, the facial nerve is very adherent to the acoustic neuroma around the porus of the internal auditory canal and a small scrap of tumor is left behind to prevent injury to the facial nerve. This scrap of tumor rarely grows, and if growth is detected on serial imaging, radiosurgery can be used to control further growth.

Once the tumor has been removed, the posterior fossa dura is then re-approximated. Fat is harvested from the abdomen and packed into the surgical defect. The periosteal and skin layers are closed in a water-tight fashion. The patient wears a pressure dressing for three days.

Retrosigmoid approach

The patient is placed in the supine position on the operating table, with the head toward the contralateral shoulder. The true lateral or park-bench position is still used by some surgeons because it permits the occiput to be rotated more superiorly. This allows a more direct view of the internal auditory canal.

The operation is performed through either a vertically oriented linear incision or an anteriorly based U-shaped flap. An occipital craniotomy is then performed. Any mastoid air cells are carefully waxed off to prevent postoperative cerebrospinal fluid leak. The dura is opened and the arachnoid incised. The cerebellum frequently falls away from the posterior surface of the temporal bone after the cisterna magna has been opened. Hyperventilation, steroids, and intraoperative diuretics (principally mannitol) are used to reduce intracranial pressure and to provide additional exposure with a limited amount of retraction. Nonetheless, gentle cerebellar retraction is occasionally required, especially with larger tumors.

Once adequate exposure has been obtained, the tumor is clearly visualized along with the brain stem and lower cranial nerves. However, cranial nerves VII and VIII are rarely observed because they are almost always pushed forward and lie across the anterior surface of the tumor, which cannot be visualized. Debulking of the tumor is the next step and must be carefully performed so as to maintain the anterior portions of the capsule in order to prevent injury to cranial nerve VII and/or VIII. Once the tumor has been substantially debulked, the posterior wall of the internal auditory canal can be removed using a high-speed drill.

The labyrinth should be preserved when removing the posterior internal auditory canal, to prevent hearing loss. Portions of the labyrinth are commonly medial to the lateral end of the internal auditory canal. Although no single anatomic landmark is completely reliable for prevention of injury to the labyrinth, the singular nerve and its canal, and the operculum of the vestibular aqueduct, are used as important surgical landmarks. Careful measurements taken from preoperative CT scans can provide useful information during drilling of the posterior wall of the internal auditory canal.

The length of the internal auditory canal varies considerably, and knowing exactly how much posterior canal wall needs to be removed to adequately expose the tumor can help limit inadvertent injury to the labyrinth. Blind extraction of tumor from the internal auditory canal without removing the posterior wall poses a significant risk to the facial and/or auditory nerve integrity and increases the chance of leaving tumor at the fundus. Use of intraoperative angled endoscopes has been reported as an adjunct in performing this phase of the operation.

Every effort should be made to prevent bone dust from entering the subarachnoid space during the intradural drilling of the internal auditory canal. One probable cause for severe and intractable postoperative headache is spillage of bone dust into the subarachnoid space during tumor removal. Surgicel, absorbable gelatin powder, Telfa pads, and/or cottonoid strips are placed around the operative site so that bone dust adheres to them and is removed when they are removed. Once the internal auditory canal is exposed, the dura is opened and the tumor is removed. Although never proven, dissection from medial to lateral is thought to be less traumatic to both the cochlear nerve and to the vascular supply of the inner ear. The vestibular nerves are generally sacrificed, and unless hearing is to be preserved, the cochlear nerve is sacrificed as well.

Eventually, the surgeon is left with the anterior portions of the capsule adhered to the brain stem and cranial nerve VII. As the tumor capsule is carefully removed from the brain stem, the root entry zone of cranial nerve VII can be identified. The capsule is then carefully removed from the facial nerve with as little trauma as possible.

The facial nerve monitor facilitates this portion of the dissection. A meaningful amount of data now shows that results are improved when facial nerve monitoring is employed. A variety of techniques have been used to monitor the cochlear nerve when hearing preservation is desired. The most commonly used method is intraoperative ABR, but it has a number of disadvantages. Most importantly, it requires summing a large number of repetitions in order to extract a response from background noise. Consequently, a delay occurs between surgical manipulations and ABR changes. Direct cochlear nerve monitoring offers the advantage of real-time feedback, but a fully satisfactory method of placing and securing the electrode still is lacking.

Once tumor removal is complete and hemostasis is complete, the dura is closed and the craniotomy defect is repaired, either by replacing the original bone flap or with methyl methacrylate or hydroxyapatite.

Middle cranial fossa approach

Although some surgeons use an extended middle cranial fossa approach for tumors that extend a centimeter or more outside the porus acusticus into the cerebellopontine angle, the middle cranial fossa approach is most frequently used for purely intracanalicular tumors. It is, by consensus, the approach of choice for small tumors that lie within the lateral portions of the internal auditory canal when hearing conservation is desired.

The head must be in the true lateral position. In young individuals with a supple neck, this can often be accomplished by turning the head to the side with the patient in the supine position. But if neck mobility is limited or concern exists that forced head turning will limit posterior fossa circulation or aggravate cervical spine disorders, then a true lateral (park-bench) position should be used.

Exposure must be centered over a vertically oriented line that passes approximately 1 cm anterior to the external auditory meatus. This is most easily accomplished through a linear incision. A posteriorly based U-shaped or curvilinear S-shaped incision can be used if concern exists about scar contracture. Depending upon the incision used, the temporalis muscle is incised or reflected inferiorly. A temporal craniotomy (approximately 5 cm by 5 cm) is performed with its base at the root of the zygoma. The dura is elevated from the floor of the middle cranial fossa, and osmotic diuretics, head elevation, hyperventilation, and steroids are used to limit cerebral edema.

The dura of the temporal lobe is then elevated off the superior surface of the temporal bone. The anterior extent of such elevation is usually the foramen spinosum, but the middle meningeal artery can be divided between clips and elevation continued anteriorly to the foramen ovale if additional exposure is desired. Dural elevation should proceed from posterior to anterior to avoid injury to an exposed greater superficial petrosal nerve or geniculate ganglion. Bleeding from the veins associated with the middle meningeal artery is often quite brisk but can generally be controlled with oxidized cellulose packing and absorbable gelatin foam soaked in thrombin. Medial dissection continues to the free edge of the temporal bone.

The superior petrosal sinus is attached to the posterior surface of the temporal bone but not always at its superior edge. Care must be taken to avoid injuring it. If inadvertent injury occurs, bleeding can generally be controlled with intraluminal oxidized cellulose packing, electrocautery, or hemoclips. When extended middle cranial fossa approaches are employed, the superior petrosal sinus is deliberately divided between clips.

When it can be identified easily, the arcuate eminence is a helpful landmark. Careful drilling can often identify the blue line of the superior canal inferior to it. Because the most difficult exposure to achieve during middle fossa surgery is the lateral posterior end of the internal auditory canal, dissection is performed as close to the superior semicircular canal as possible. The greater superficial petrosal nerve is generally easy to visualize and can be followed retrograde to the geniculate ganglion. It lies approximately 1.0 cm directly medial to the foramen spinosum. Once the area of the geniculate is identified, small diamond burrs are used to completely expose it. If the greater superficial petrosal nerve cannot be located and no other landmarks are available, the middle ear space can be entered from above and the head of the malleus can be identified. The geniculate ganglion lies approximately 2-3 mm anterior and medial to the head of the malleus.

Once the geniculate ganglion has been completely exposed, the labyrinthine portion of the nerve can be identified and followed medially and inferiorly into the internal auditory canal. The labyrinthine portion of the nerve takes a markedly vertical and medial course as it moves from the lateral geniculate ganglion to the proximal fundus of the internal auditory canal, which lies five or more millimeters deep to the geniculate ganglion. Some surgeons prefer to identify the internal auditory canal medially. Once the medial end of the canal is completely identified, they follow the canal laterally to the fundus of the internal auditory canal.

The bone overlying the internal auditory canal should be removed until approximately 270 º of the internal auditory canal is exposed. The most difficult area to expose is the point at which the superior vestibular nerve penetrates the labyrinthine bone to innervate the ampulla; however, exposure in this area is critical if the anatomy of the lateral end of the internal auditory canal is to be well visualized. If the superior vestibular nerve channel is identified, tumor removal is generally successful and relatively straightforward.

Larger tumors frequently have the facial nerve splayed out over the anterior superior portions of the tumor. Tumor removal begins, as with other approaches, by careful debulking. Once the tumor is debulked, enough room is created within the internal auditory canal to carefully remove the tumor capsule from the inferior surface of the facial nerve. Again, care must be taken to avoid torsion or twisting of the nerve during tumor removal.

Once the tumor has been completely removed, the integrity of the facial nerve is tested using the intraoperative facial nerve monitor. Presumably, the monitor has been in use throughout the case. If the facial nerve can be stimulated with low stimulus intensities, chances of good postoperative facial nerve function increase. Fat is then packed into the internal auditory canal after using bone wax to fill obvious air cells to prevent postoperative cerebrospinal fluid leak. The facial nerve monitor generally alerts the physician if fat is being packed in too tightly that the integrity of the facial nerve is being compromised. Retractors are removed, and the temporal lobe dura is allowed to relax. The bone plate is replaced using miniplates, and the wound is closed in multiple layers.

Unless a complication develops, postoperative care is straightforward. The patient is generally kept in the ICU overnight so that rapid intervention is available if postoperative intracranial pressure increases or bleeding occurs. Vestibular rehabilitation should begin on the first postoperative day and continues twice daily throughout the hospital stay. Most patients can be discharged on the third or fourth postoperative day.

A follow-up MRI is obtained within 6-12 months after surgical excision to document the completeness of tumor removal and to serve as a baseline for further follow-up scans. Assuming complete tumor removal, follow-up MRI should be obtained at 5 years and at 10 years. If the findings on the 10-year scan are normal, further imaging should be performed only if clinical circumstances require it. More frequent or surveillance imaging may be considered in cases of questionable radiologic progression, nodular enhancement, or residual tumor.[25]  Risk of recurrence is strongly associated with the presence of nodular enhancement on the first postoperative MRI and incomplete resection.[26]  Postoperative MRI scans must be performed with fat-suppressed techniques if fat was used to obliterate the surgical site.

Vascular injury

Injury to the AICA (much less commonly to the PICA) fortunately occurs very rarely. Although the AICA may be loosely attached to the tumor capsule, separating it from the tumor is generally fairly easy. Sacrificing the AICA itself has variable consequences depending upon the details of individual patient anatomy. It can be catastrophic and lead to devastating neurologic injury or death.

The branches of AICA most vulnerable to injury are, of course, the labyrinthine artery and branches supplying the facial nerve. Some otherwise perplexing cases of postoperative facial nerve weakness may be related to interruption of facial nerve vascular supply due to coagulation of small branches of the AICA. Failure to conserve hearing may be due to the disruption of cochlear blood supply. Because the labyrinthine artery may be intimately associated with the tumor, sacrifice often cannot be avoided. Conservation of the internal labyrinthine artery becomes more difficult as tumor size increases, doubtlessly accounting in some measure for the reduced success in hearing conservation with larger tumors. Neurologic injury or cerebral edema secondary to venous injury usually occurs as a result of injury to the sigmoid sinus itself, the petrosal vein of Dandy, or to the vein of Labbé.

Occlusion of the sigmoid sinus has variable effects depending largely upon the patient's unique venous anatomy. If the contralateral venous outflow tract is patent and communication through the torcula Herophili is adequate, complete occlusion may be asymptomatic. The size of the two sigmoid sinuses is usually asymmetrical, with a greater volume of blood usually flowing through the right-sided sinus. Depending on how much more blood flows through the dominant sinus, occlusion of the dominant sinus can result in catastrophic increases in intracranial pressure, venous infarction, and even death. Because a number of potential collaterals exist between the torcula Herophili and the jugular bulb, occlusion of the sigmoid sinus close to the torcula Herophili is much more likely to have significant adverse effects than its occlusion close to the jugular bulb.

The petrosal vein of Dandy is a single large outflow tract in some patients but consists of series of several large veins in others. Its occlusion can result in edema and infarction of either the temporal lobe or the brain stem. Although neurologic injury secondary to occlusion of the Dandy vein is not inevitable, severe injury can occur, and the petrosal vein should be carefully preserved.

Occlusion of the vein of Labbé results in severe edema of the temporal lobe and temporal infarct. The edema can be sufficiently severe to cause brain herniation and death. The vein of Labbé generally enters the superior petrosal or transverse sinus between the torcula Herophili and the point at which the superior petrosal sinus joins the transverse sinus. Thus, it is generally not directly in the field during acoustic tumor surgery. Occasionally, however, injury to the superior petrosal sinus results in its obliteration, and in some instances, the vein of Labbé is also injured or obstructed. Its presence and importance should be kept in mind during acoustic tumor surgery.

Hemorrhage into the posterior fossa in the immediate postoperative period can produce brain stem compression and death quite rapidly. Death can occur within a few minutes. Rapid neurologic deterioration in the first 24 hours postoperatively should raise suspicion of posterior fossa hemorrhage and mandates rapid and decisive intervention. If time permits, a rapid unenhanced CT scan should be obtained to secure the diagnosis while the operating room is prepared for an immediate return to surgery. If neurologic deterioration is rapid, forgo CT scanning and take the patient directly back to the operating theatre. If deterioration is very rapid with loss of consciousness, decerebrate posturing, and signs of imminent death, open the wound at the bedside to permit a posterior fossa decompression prior to emergent transportation of the patient to the operating room for wound exploration, debridement, and extensive irrigation.

Cerebellar injuries

Injury to the cerebellum was common in the early decades of the century, but its incidence has dramatically diminished in recent decades. Cerebellar injuries still occur but are generally not troublesome. The rotating shaft of the surgeon's burr is often the culprit because surgeons usually look past the shaft to the head of the burr to concentrate on bone removal. The shaft is often outside the surgical field of view. Such small areas of injury rarely have noticeable sequelae. Bleeding can be controlled with oxidized cellulose, cautery, or gelatin sponges, and edema is limited. Direct injury to the cerebellar hemisphere from compression and retraction, intracerebral hemorrhage, infarction due to alteration of the arterial inflow, or venous engorgement with or without infarction can produce severe edema of the entire cerebellum. Brain stem compression and/or intracranial herniation can produce death. Obstruction of the fourth ventricle and cerebral aqueduct can produce significant hydrocephalus.

Management should consist of aggressive use of osmotic diuretics, hyperventilation, and steroids. If medical management is unsuccessful, resection of part of the involved cerebellar hemisphere may be required.

Facial paralysis

In some cases, postoperative facial paralysis is unavoidable. The tumor may simply be attached too intimately to the thin attenuated facial nerve. Sometimes the tumor has enveloped the facial nerve, and tumor removal cannot be accomplished without resection of a portion of the facial nerve.

Eye care is critical to successful management of postoperative facial paralysis. Make liberal use of artificial tears specifically adapted to deal with dry eye (eg, Bion Tears). At night, place ocular lubricants (eg, Lacri-Lube) in the eye. If aggressive use of artificial tears during the day (q15-30min) and ointment at night is insufficient to maintain corneal hydration and exposure keratitis begins to develop, then consider use of an eye patch, placement of a gold weight, and lower lid shortening procedures. Tarsorrhaphy should be used only as a last resort and is rarely required. Coexisting injury to cranial nerve V with corneal hypesthesia or anesthesia vastly increases the problem in management. The lack of corneal sensation provides the patient with no reliable guide as to the severity of corneal epithelial disruption. In such cases, tarsorrhaphy is much more likely to be required.

If facial function does not recover, facial reanimation may be considered. This is commonly performed by anastomosing the distal facial nerve branches to the hypoglossal or mandibular branch of the trigeminal nerve. For long-standing paralysis, the free transfer of a gracilis muscle anastomosed to the masseter branch of the trigeminal nerve may be used.

Cerebrospinal fluid complications

Transient abnormality of cerebrospinal fluid resorption may lead to mild temporary postoperative hydrocephalus. Although postoperative shunting can facilitate controlling cerebrospinal fluid fistula, it is now rarely, if ever required. Even when hydrocephalus is present in the preoperative period, it generally resolves without difficulty in the first few postoperative weeks.

Postoperative meningitis occurs in two forms. Bacterial meningitis is potentially life threatening and occurs in less than 1% of patients in the postoperative period. It can occur within the first 24-36 hours postoperatively, or its appearance may be delayed for a couple of weeks. Once initiated, it can progress very rapidly, and individuals can lapse from a normal level of consciousness into a dense coma in a matter of a few hours.

Consequently, intervention must be rapid. Diagnosis depends upon the presence of fever and, in the alert patient, of headache, photophobia, nuchal rigidity, and decreasing level of consciousness. If meningitis is suspected, perform an immediate lumbar puncture to obtain fluid for culture, but only after a CT scan has excluded the possibility of significant hydrocephalus, which could lead to brain herniation.

Obtain spinal fluid for Gram stain, glucose, protein, and white blood cell count. If the Gram stain is positive, spinal fluid glucose is less than 40 m/dL, or the spinal fluid white blood cell count is higher than 2500 cells/mm3, begin antibiotics immediately pending culture results. If the spinal fluid does not meet any of these criteria, closely observe the patient with the understanding that any deterioration of the condition requires a repeat lumbar puncture for additional spinal fluid.

Aseptic meningitis has been reported in 7-70% of postoperative neurosurgical patients. It shares with bacterial meningitis the clinical signs of increasing headache, fever, nuchal rigidity, and elevation of cerebrospinal fluid pressure. Spinal fluid profile in such patients shows marked elevation of white blood cell count and cerebrospinal fluid protein levels, but cerebrospinal fluid glucose remains within the reference range, and culture results (when they are finally complete) are normal. Corticosteroids are successful in managing aseptic meningitis, and their prompt administration often results in marked decrease in headache and nuchal rigidity within a few hours.

Spinal fluid leak through either the wound or the eustachian tube and middle ear occurs in 2-20% of patients. It can occur after translabyrinthine or retrosigmoid approaches and is less common after middle fossa craniotomy. When it follows retrosigmoid approaches, the path of egress is generally through pneumatized air cell tracts.

A retrospective study by Luryi et al indicated that in patients undergoing acoustic neuroma surgery, a high body mass index (BMI) is a risk factor for cerebrospinal fluid (CSF) leak and the need for revision surgery. The investigators found that 11.6% of patients with a BMI of 30.0 or above suffered a CSF leak, compared with 5.1% of patients with a BMI of less than 30.0.[27]

Cerebrospinal fluid is produced within the ventricular system at a rate of 0.3 mL/min or at about 500 mL/day. It enters the subarachnoid space in the posterior fossa via the midline and lateral foramen of the fourth ventricle. Contamination of the cerebrospinal fluid circulation by blood, bone dust, and necrotic debris at the time of surgery often impairs cerebrospinal fluid absorption directly by mechanical interference in the arachnoid villi or indirectly by inciting an inflammatory response within the subarachnoid space. The syndrome may vary from brief asymptomatic elevation of cerebrospinal fluid pressure to clinically manifested aseptic meningitis (discussed above). Cerebrospinal fluid escaping through the wound can initially be managed by resuturing the wound. Sometimes this results in elimination of the difficulty, while at other times it merely produces cerebrospinal fluid rhinorrhea, as the spinal fluid finds an alternate means of egress.

If the cerebrospinal fluid leak persists for more than 12-24 hours after initiation of conservative management, including pressure dressing and consistent head elevation, then consider reducing the cerebral spinal fluid pressure by 1 of 3 measures, including (1) multiple lumbar punctures, (2) continuous or intermittent drainage via lumbar intradural catheter, or (3) permanent cerebrospinal fluid diversion by means of an indwelling shunt.

When cerebrospinal fluid diversion is selected, the most common method is an indwelling subarachnoid catheter placed into the lumbar subarachnoid space. The drain is opened episodically to remove cerebrospinal fluid and decrease pressure. Some surgeons observe a minimum drainage period of 2 days, others up to 5 days. The general consensus is that if the drain has been in place for more than 5 days, it should be replaced to avoid infection. Variation among surgeons is considerable as to when reexploration is required. Some centers reexplore after 24-48 hours of drainage; other centers use as many as two 5-day trials of continuous lumbar drainage before considering a second operation.

Severe postoperative headache has long been associated with retrosigmoid procedures. This problem appears to have diminished considerably since the introduction of two intraoperative steps: (1) care is taken to avoid contaminating the spinal fluid and subarachnoid space with bone dust, and (2) the bone flap is replaced and any residual bony defect is eliminated with methyl methacrylate or hydroxyapatite. The latter step eliminates the direct attachment of posterior cervical musculature to the dura. When postoperative headaches do occur, they should be managed with relatively high-dose nonsteroidal anti-inflammatory agents and aggressive regimens of manipulative physical therapy.

Tinnitus

Tinnitus becomes worse in 6-20% of individuals after tumor removal. Often the tinnitus remains unchanged. In about 25-60% of patients, tinnitus is eliminated or improved. Although 30-50% of patients who had no preoperative tinnitus develop it in the immediate postoperative period, such tinnitus only rarely becomes troublesome.

A study by Bell et al of 53 patients indicated that in patients who undergo acoustic neuroma resection, the prognosis for tinnitus resolution is worse for those who are younger, whose preoperative hearing was serviceable, and who have residual tumor postoperatively.[28]

On the other hand, a study by Alvarez et al found that younger patients had particularly good results on the Tinnitus Handicap Inventory questionnaire following translabyrinthine removal of vestibular schwannomas. The study also found that patients with the worst preoperative hearing demonstrated the best postoperative outcomes on the questionnaire.[29]

Recurrence/residual tumor

Recurrence is uncommon after acoustic tumor removal. Overall, the recurrence rate is less than 5%. The vast majority of recurrences follow retrosigmoid removal. Presumably, a small amount of tumor is left in the lateral end of the internal auditory canal where intraoperative visualization is difficult. Tumor recurrence may be suspected by recurring headache, altered sensation to the face, or dysarthria and dysphasia if the lower cranial nerves become involved.

Inflammation in the tumor bed may persist for months and even years after acoustic tumor removal, and consequently, areas of contrast enhancement are present on postoperative gadolinium MRI. Distinguishing tumor recurrence from postoperative inflammation can be quite difficult. Tumor recurrences tend to be globular while postoperative inflammatory enhancement tends to be linear. Often, however, one must view serial scans to detect tumor recurrence. Fat suppression techniques are essential for postoperative surveillance to distinguish recurrence from fat packing. Surveillance for postoperative tumor recurrence should persist for 8-10 years postoperatively.

Facial function

Preservation of facial function continues to improve, especially with the widespread use of facial nerve monitoring. However, facial nerve outcomes continue to vary according to tumor size. When tumors are smaller than 1.5 cm, good facial nerve function can be expected (House-Brackmann grade I-II) in greater than 90% of patients.

In addition to tumor size, preoperative electrophysiologic testing can help predict postoperative outcome, although this testing is not commonly used. Demonstrable electrophysiologic abnormalities on nerve conduction studies, electromyography, and blink reflex testing correlate well with postoperative facial nerve deficits. Arriaga has shown that patients with poor facial nerve function at the time of discharge (House-Brackmann V-VI) had a 25% chance of recovery of normal function (House-Brackmann I-II). Less optimistic is the report by Sterkers; in this, patients who had House-Brackmann III function or worse at a 4- to 6-week postoperative evaluation were left with significant deficit and generally had some synkinesis.

Facial nerve paralysis may be delayed and may develop within a few hours to a week or more after acoustic neuroma removal. Incidence of delayed facial palsy varies from 10-30%. The mechanism of action is unclear. Ischemia secondary to vasospasm, vascular injury, traction, nerve edema, stretching, and even a viral reactivation have been proposed. Unlike final facial nerve outcome, incidence of delayed facial paralysis does not appear to be related to tumor size.

The vast majority of individuals who have delayed onset of facial paralysis make complete and total recoveries. If deterioration is severe (more than 3 House-Brackmann grades), some chance of poor long-term outcome exists. Perioperative steroids are widely used in an attempt to enhance both immediate and long-term postoperative facial nerve function, but unequivocal evidence for their effectiveness is lacking. The use of perioperative antiviral agents is used by some centers to prevent delayed paralysis from viral reactivation, as in Bell palsy.

Hearing outcome

The ability to preserve hearing has increased substantially over the last decade or two. Depending on criteria for successful hearing conservation, hearing can be preserved in 30-80% of properly selected patients.

Radiosurgery does not appear to have a significantly higher rate of hearing preservation than does properly conducted surgery when long-term results are compared. Chopra et al demonstrated a hearing preservation rate of 44% at 10-year follow-up in 216 patients receiving Gamma Knife radiosurgery.

Rosenberg et al and Tucci et al have both shown stability of hearing over time after surgery.[30] On the other hand, Shelton et al's study reported significant hearing deterioration in 30-50% of patients who originally had successful hearing preservation.

Rehabilitation

The hearing deficit after the removal of an acoustic neuroma can have a significant impact on quality of life. Rehabilitation options include a contralateral routing of signals (CROS) hearing aid or a bone-anchored hearing aid (BAHA). The BAHA consists of a surgically implanted titanium abutment that osseointegrates into the calvaria. A speech processor is then snapped onto the abutment, allowing sound to transmit through the skull to the normal contralateral ear. Newer BAHAs are placed under the scalp, and the speech processor attaches to the scalp by a magnet. Although these newer systems prevent skin issues that occur with percutaneous BAHAs, they are not as compatible with MRI.

All patients with neurofibromatosis type 2 develop bilateral acoustic neuromas, and most will eventually lose hearing in both ears. The auditory brain stem implant (ABI) can be inserted during tumor removal and provides patients with the benefit of auditory perception to assist with communication.

If a patient experiences facial nerve weakness and not total paralysis after surgical removal, eye care consisting of artificial tears and lubricant often will be sufficient until facial nerve function returns. If the facial nerve is severed intraoperatively, the nerve can be approximated at the time of surgery. If the patient has facial nerve paralysis 1 year after surgery, the chance of further recovery is remote. At this time, facial nerve reanimation procedures should be considered.

In 2018, the Congress of Neurological Surgeons released a series of guidelines on the diagnosis and management of vestibular schwannomas.[31, 32, 33, 34, 35, 36] The group’s guidelines concerning hearing preservation in patients with sporadic vestibular schwannomas include the following[37] :

• Patients whose baseline hearing meets the criteria for American Academy of Otolaryngology-Head and Neck Surgery (AAO-HNS) class A or Gardner-Robertson hearing classification (GR) grade I and who are considering stereotactic radiosurgery utilizing modern dose planning, should be counseled that there is a high probability (>75-100%) of hearing preservation at 2 years, a moderately high probability (>50-75%) of hearing preservation at 5 years, and a moderately low probability (>25-50%) of hearing preservation at 10 years
• Patients with AAO-HNS class A or GR grade I hearing at baseline who are considering microsurgical resection of small to medium-sized sporadic vestibular schwannomas should be counseled that there is a moderately high probability (>50-75%) of hearing preservation immediately following surgery, a moderately high probability (>50-75%) of hearing preservation at 2 years, a moderately high probability (>50-75%) of hearing preservation at 5 years, and a moderately low probability (>25-50%) of hearing preservation at 10 years
• Patients with AAO-HNS class A or GR grade I hearing at baseline should be counseled that, if they undergo management with conservative observation, there is a high probability (>75-100%) of hearing preservation at 2 years and a moderately high probability (>50-75%) of hearing preservation at 5 years; insufficient data were available to determine the probability of hearing preservation at 10 years for these patients