Audiometric Evaluation

The majority of patients with an acoustic neuroma will demonstrate abnormal findings on audiologic testing. A complete audiologic test includes pure-tone audiometry, acoustic reflex testing with a measurement of reflex decay and speech reception audiometry. Prior to the availability of ABR testing, several additional tests were used to distinguish between cochlear and retrocochlear lesions. These tests included tone decay testing, the suprathreshold adaptation test, the performance intensity function for phonetically balanced words test, the short-increment sensitivity index and the alternate binaural loudness balance test. For the most part, these audiologic tests have been replaced by ABR testing. In one series, an extensive audiologic test battery without ABR testing detected a mixed retrocochlear loss in 69 percent and pure retrocochlear findings in only 27 percent of patients with acoustic schwannomas.

Pure- Tone Audiometry

The unilateral SNHL associated with acoustic schwannomas can be detected with pure-tone audiometry. By varying the sound intensity, thresholds can be obtained at octave intervals from 250 to 8000 Hz. This test requires the cooperation of the patient, who must indicate every time he is able to hear the pulsed tone, no matter how faint or short it may be. In a study of 66 patients with acoustic schwannomas, the audiologic profiles revealed either a high-frequency hearing loss (56 percent), complete deafness (17 percent), flat loss (14 percent) or normal hearing (8 percent). The incidence of normal audiograms has been reported elsewhere to be 3 to 6 percent. Isolated low-frequency pure­tone loss is a rare occurrence in patients with acoustic schwannomas.  Audiometry seems to be a poor predictor of tumor size. In their series of 300 patients with acoustic schwannomas, Thomsen and Tos found no significant difference between the degree of pure­tone loss and the size of tumors less than 4 cm. A strong correlation was found, however, between the extent of hearing loss and tumor size greater than 4 cm.

Acoustic Reflex Testing

The acoustic reflex is a response of the stapedius muscle to the presentation of a loud sound. The reflex arc begins with the stimulation of the ganglion cells in the cochlea that terminate in the cochlear nucleus. The second-order neurons project ipsilaterally and contralaterally from the anterior ventral cochlear nucleus to the superior olivary complex. Third-order neurons connect with efferent motor fibers to the stapedius muscle within the facial nucleus. In acoustic reflex testing, a sufficiently loud tone is introduced into the ear, which results in the bilateral contraction of the stapedius muscle. The stapedius muscle acts to stiffen the ossicular chain and the tympanic membrane. The response of the stapedius muscle to the acoustic reflex can be detected by a pressure­sensitive probe that is placed in the ear canal and connected to an impedance audiometry bridge. This test does not require the subjective response of the patient.

Acoustic tumors may interfere with this reflex arc. Consequently, the acoustic reflex may be absent or may rapidly decrease in intensity when a prolonged stimulus is presented (reflex decay). Absent acoustic reflexes were noted in 76 percent of patients with acoustic schwannomas and significant reflex decay was found in 62 percent in one series. Other studies have reported the sensitivity for this test to be 85 to 96 percent. Unfortunately, the high sensitivity of this simple diagnostic test corresponds with poor specificity that limits its usefulness.

Speech Discrimination

Patients with acoustic tumors will frequently complain of difficulty understanding speech, especially over the telephone. By measuring the patient's response to a list of familiar words, the audiologist can confirm and quantify the difficulty in speech discrimination. Failure to comprehend less than 90 percent of the presented words is considered an abnormal result. Poor speech discrimination that is out of proportion to the degree of pure-tone loss is the audiologic hallmark of an acoustic neuroma. This phenomenon is most likely due to the fact that up to 75 percent of auditory fibers may be damaged before a demonstrable change of pure-tone threshold is evident. The actual reported incidence of abnormal speech discrimination in the literature, however, varies from 20 to 72 percent. The sensitivity of this test is even lower in acoustic neuroma patients with normal or symmetric pure-tone audiometry findings. Abnormal speech discrimination can be found in only 5 to 22 percent of these cases. Thus, speech discrimination testing is an unreliable tool to rule out a retrocochlear lesion. The rollover phenomenon, defined as the decay of the speech discrimination score with increased stimulus intensity, is of historical interest only. When present, this finding is indicative of a retrocochlear lesion: however, the sensitivity of this test is very low.

Auditory Brain Stem Response

The ABR is the most useful audiometric test in the diagnosis of acoustic schwannomas. This technique measures the electrical potential from mastoid and vertex electrodes in the first 15 ms following an acoustic stimulus. Repetitive stimuli and sampling allow the evoked potentials to be separated from the background noise. In normal patients it is possible to define seven waves in the ABR. These waves are labeled with sequential Roman numerals and are thought to represent successive tracts and synapses within the auditory pathway, The largest and most reproducible of these waves is wave V. The size and latencies of these waves are dependent on the stimulus intensity, such that a 3.0-ms shift in wave V occurs between threshold and a 60-decibel (dB) hearing level stimulus. Thus, a correction factor must be applied to account for the degree of hearing loss and patients with severe hearing loss may not produce recognizable waveforms. Because of this limitation. ABR testing is not practical in patients with hearing loss exceeding 70 to 80 dB in the 1000- to 4000-Hz range. Wave latency is also affected by conductive hearing loss, which may limit the interpretation of data in patients with middle ear pathology.

Stretching of the cochlear nerve by an acoustic tumor produces a delay of the ABR waves. Several criteria have been applied to ABR audiometry in order to detect abnormalities in the latency of wave V. The absolute latency of wave V has been compared with normative data, The normal latency for wave V is between 5.0 and 5.7 ms, but the large variability of this value in normal patients has limited its clinical use, The interwave period between wave I and wave V may be used to detect a retrocochlear lesion. However, many patients with SNHL or acoustic tumors may not have a detectable wave I or may have a delay in both wave I and V, with a normal interwave latency. The technique most commonly used for detection of acoustic tumors is to compare the waveforms from the suspected ear with the contralateral side. Because the patient acts as his own control, this method reduces the effects of normal variability. The maximum interaural latency difference between waves I and V in the normal population is no more than 0.2 ms. When this criterion is applied. ABR audiometry is an extremely sensitive and specific test for patients suspected of having an acoustic neuroma.

Most authors agree that ABR testing is able to detect acoustic schwannomas in over 90 percent of cases, Two recent studies have compared ABR audiometry to Gd-MRI. which is able to detect extremely small intracanalicular acoustic schwannomas, These results suggest that many small tumors may be missed by ABR audiometry. The sensitivity of ABR audiometry is generally greater than 92 percent, even with a high percentage of small tumors. The difference between studies may be due to population differences (e.g., differences in tumor size) or interpreter variability. Audiologists may disagree on the interpretation of an ABR in up to 22 percent of cases. The incidence of false-negative results on ABR audiometry correlates inversely with the extent of hearing loss. Therefore, ABR testing is quite sensitive in acoustic neuroma patients with significant hearing loss, as long as the loss does not exceed 70 to 80 dB. On the other hand, an acoustic neuroma cannot be adequately ruled out in patients with a strong clinical history with ABR audiometry alone. Because of the possibility of a false-negative ABR, many authors recommend Gd-MRI in high-risk patients even if screening ABR is normal.

There are comparatively few studies that examine the specificity of ABR audiometry in the diagnosis of acoustic neuroma. Selters and Brackmann found the specificity of ABR testing to be 89 percent when the threshold of interaural latency difference is set at 0.2 ms. A slightly lower specificity (80 percent) has been reported elsewhere using this same criterion and a specificity of 77.8 percent has been reported using absolute-latency criteria. The incidence of false-positive (retrocochlear) ABR findings corresponds to the degree of hearing loss and seems to be highest in patients with mixed cochlear and retrocochlear hearing loss as compared to those with vertigo, unilateral tinnitus, disequilibrium or facial nerve problems.

Vestibular Evaluation

Electronystagrnography

Electronystagmography (ENG) is an examination of eye movements during several manoeuvres that elicit inappropriate eye movements or nystagmus. The electrical potential between the cornea and the retina creates an electric field in the front of the head that changes as the eyes rotate in their orbits. This electric field can be detected by electrodes placed on either side of the eyes. Patients perform a series of tasks, including tracking a moving target, looking far to one side or the other (extreme lateral gaze), watching an optokinetic stimulus (stripes traveling across the visual field) and placing the head in different positions. The most useful test in patients suspected of having an acoustic neuroma is the caloric stimulation test as introduced by Barany and modified as the bi­thermal caloric test by Fitzgerald and Hallpike in 1942. In this test, the external ear is irrigated with warm and cold water. A positive test is defined as a caloric response that is reduced by 30 percent or more when compared to the caloric response of the contralateral ear. Because each vestibular organ is stimulated separately, this test is able to differentiate the side of a vestibular lesion. Caloric irrigation stimulates the lateral semicircular canal; therefore, this technique is only sensitive to lesions that affect the superior vestibular nerve.

Small acoustic schwannomas will manifest an ipsilateral reduced caloric response, whereas larger tumors may demonstrate more centralized findings such as failure of fixation suppression, slowing of optokinetic nystagmus, saccadic pursuit and bilateral horizontal gaze nystagmus. In early studies of acoustic neuroma patients, the incidence of abnormal ENG was reported to be from 80 to 99 percent. Because most patients in these series had abnormal ENG findings, it was generally believed that most tumors originated from the superior vestibular nerve. A review of both the surgical and temporal bone data, however,  reveals that these lesions are equally distributed between the two vestibular nerves.

Linthicum et al. found the overall sensitivity of ENG for tumors of the superior vestibular nerve to be 97 percent, compared to 60 percent for tumors of the inferior vestibular nerve. They also reported a reduced vestibular response in only 43 percent of patients with small acoustic schwannomas. If one assumes that small acoustic schwannomas only affect the vestibular nerve from which they originate, these data support the idea that roughly half of acoustic schwannomas are located within the superior vestibular nerve.

Rotational Testing

The sinusoidal harmonic acceleration (SHA) test simultaneously stimulates both lateral semicircular canals by placing the patient in a rotating chair. The patient's horizontal eye movements are recorded by the same methods as with ENG. The major advantage of this technique is the precise control of the stimulus, whereas the inability to localize the side of a lesion is its major disadvantage. In the diagnosis of acoustic schwannomas, the sensitivity of SHA testing is 67 percent and the specificity is 86 percent. Therefore, SHA testing is not considered to be cost effective either alone or in conjunction with ABR audiometry.

Posturography

A recent addition to the armamentarium of vestibular tests, platform posturography is a sensitive test of a variety of central and peripheral balance disorders. This test measures the patient's balance on a platform with six conditions designed to manipulate visual and/or proprioceptive input. This technique is extremely useful in the monitoring of patients during vestibular rehabilitation, but its use in the diagnosis of acoustic schwannomas has not been established.

Diagnostic Efficiency

A diagnostic strategy for patients suspected of having an acoustic neuroma should be based on a cost-effective approach that compares the expense of appropriate screening tests against the cost of delayed diagnosis. Late diagnosis and the corresponding resection of larger tumors is associated with higher surgical morbidity. The average cost of surgical treatment may increase by fourfold in the presence of major complications. In a study of 66 acoustic neurinomas treated in England, Moffat et al. weighed the cost of early diagnosis of small- to medium-size acoustic schwannomas against the increased morbidity and cost of diagnosis after these tumors have grown to a large size. The authors estimate that the financial burden to society of one patient with a "fair" surgical result from a large tumor is equivalent to the cost of resecting 39 small tumors with good surgical results. They conclude that the early diagnosis of acoustic schwannomas with the appropriate use of screening tests is justified in both financial and human terms.

All patients with unilateral auditory symptoms (nonpulsatile tinnitus or hearing loss) should be evaluated with a complete history. a thorough neuro-otologic examination and standard audiometry. Depending on the history and the results of these initial studies, the patient should then be evaluated by either ABR testing or MRI. Because of its rather high rate of false-negative results. ENG is not considered to be cost effective as a screening test for acoustic neuroma. ENG may be used to determine if the tumor originates from the superior versus inferior vestibular nerve. Small tumors of the inferior vestibular nerve may have a normal ENG, whereas superior vestibular nerve tumors usually cause a reduced vestibular response to caloric testing.

The standard diagnostic approach for screening patients with a possible acoustic neuroma has been an initial audiogram followed by ABR audiometry in those patients with asymmetric SNHL or an abnormal speech discrimination score. Gd-MRI was, thus, reserved for patients with an abnormal ABR. However. Welling et al. proposed an approach to the diagnosis of acoustic neurinomas that groups patients according to the relative probability of having a tumor. Patients with unilateral asymmetric SNHL, unilateral tinnitus, and decreased speech discrimination were considered to have a high probability of having an acoustic neuroma (>30 percent). The probability of acoustic neuroma in patients with sudden SNHL or unexplained unilateral tinnitus was estimated at an intermediate level (5 to 30 percent). Low-risk patients included those with isolated vertigo, historically explained unilateral hearing loss or tinnitus or symmetric hearing loss. Welling et al advocate an initial audiometric evaluation followed by Gd-MRI for patients in the intermediate- and high-risk groups and ABR testing for patients with a low probability of having a tumor. In contrast. Kotlarz et al. claim that the prevalence of acoustic neurinomas in the suspected population (patients with asymmetric hearing loss or unexplained otologic complaints) is only 5 to 7 percent. They argue that an initial ABR study with subsequent MRI for patients with abnormal ABR findings is the most cost-effective diagnostic strategy. For patients with a relatively low risk of acoustic neuroma (5 to 7 percent) this strategy should accurately rule out acoustic neuroma in 99 percent of cases with a normal ABR. This relationship is primarily determined by the tumor prevalence in a given population. Therefore, the predictive value of a negative ABR falls dramatically as the risk of an acoustic tumor rises. Although MRI is more expensive than ABR testing, it is clearly the test of choice for patients with a high risk of acoustic neuroma.

When an old fashion technology start to show its inferiority, the discussion start about the cost-effectiveness.  At the present time,  MRI is relatively low costing and it gives the best morphologic picture . The other otologic diagnostic armamentarium remain in use to evaluate the functional state of all the involved structures. We are now living in the era of IT  and the surgeon must have the most detailed information about the patient.

Unusual Tumors of the Cerebellopontine Angle

Although acoustic neuroma is the most frequently encountered neoplasm in the CPA, the clinician must also consider the possibility of an "unusual" tumor in the differential diagnosis of CPA masses. Accurate diagnosis of the CPA lesion may aid in the selection of an appropriate surgical approach and in preoperative counseling. Of the roughly 8 percent of CPA tumors that are not acoustic schwannomas, meningiomas are the most common, comprising about 3 percent of all CPA tumors. Tumors of epithelial rests cells, commonly referred to as epidermoid tumors or primary cholesteatomas, account for approximately 2.5 percent of CPA tumors and facial nerve schwannomas constitute roughly 1 percent. The remaining 1.5 percent is composed of a long list of rare tumors, including other cranial nerve schwannomas, arachnoid cysts, cholesterol granulomas, cavernous hemangiomas, lipomas,  chordomas, chondrosarcomas, choroid plexus papillomas, metastatic tumors, giant cell tumors, dermoid tumors, teratomas, medulloblastomas, hemangioblastomas, brain stem gliomas, arteriovenous malformations and aneurysms of the basilar artery or its branches.

It is impossible to distinguish nonacoustic neuroma tumors on the basis of symptomatology alone. The presenting symptoms and signs of these patients are generally similar to those of patients with acoustic schwannomas, although they may exhibit more variability than the latter. Like acoustic schwannomas, the common symptoms are hearing loss, vestibular dysfunction and headache; but the incidence of "atypical" symptoms is much higher in the nonacoustic neuroma group. In a review of 34 CPA tumors other than acoustic schwannomas. 50 percent had normal pure-tone hearing and vestibular function tests. and the incidence of cranial nerve symptoms other than those of the cochleovestibular or trigeminal nerves was 24 percent. The efficacy of audiovestibular testing may also vary with the tumor type. Meningiomas generally present with audiovestibular symptoms. However, the pure-tone averages are better with meningiomas as a group than with acoustic schwannomas. Epidermoid tumors are also characterized by minimal pure-tone hearing loss along with poor speech discrimination. Likewise, ABR testing is reported to be less sensitive with uncommon tumors. Abnormal ABRs have been found in only 83 percent of patients with facial nerve neuromas and 75 percent with meningiomas or epidermoid tumors.  These audiometric findings, however, cannot reliably distinguish any of these tumors from acoustic schwannomas.

Recent reviews of rare CPA tumors have shown hearing loss to be the predominant symptom in patients with arachnoid cysts, cavernous hemangiomas, lipomas and glossopharyngeal neuromas. In contrast, auditory or vestibular symptoms were less common in patients with trigeminal schwannomas. cholesterol granulomas, choroid plexus papillomas and chondrosarcomas or chordomas. Primary central nervous system tumors and metastatic tumors within the CPA may present with hearing loss as well, although the progression of symptoms is usually more rapid and other neurological findings are common. Furthermore, neuro­epithelial tumors may also produce bilateral ABR abnormalities.

The incidence of facial nerve dysfunction is more variable in the nonacoustic tumor group. Facial twitch or tic is considered an early diagnostic feature of both epidermoid tumors and facial nerve schwannomas.  Facial paralysis may be present in 46 percent of all facial nerve schwannomas, although these tumors may grow quite large within the CPA before facial weakness develops. Although electromyography and acoustic reflex testing may demonstrate early facial nerve involvement in patients with normal clinical function, the utility of these tests in the diagnosis of a facial nerve neuroma is uncertain. Epidermoid tumors and cavernous hemangiomas also have a higher incidence of progressive facial paralysis than acoustic schwannomas.

The fact that many unusual tumors produce symptoms and clinical findings that are indistinguishable from those of acoustic neurinomas underscores the importance of complete imaging studies. It is difficult to differentiate uncommon CPA tumors from acoustic schwannomas based on the auditory, vestibular or facial nerve findings alone. The majority of these tumors, however, have characteristic findings on MRI scans. As demonstrated with lipomas of the lAC, these features may only be evident when a complete MRI study with nonenhanced T1 and T2 images is performed in addition to the Gd-MRI. A gadolinium-enhanced T1-weighted study is an excellent screening examination, but a full MRI series is optimal when a tumor is detected. Moreover, complementary CT or magnetic resonance angiography may provide additional information when an unusual tumor type is suspected.

Conclusion

The past decade has brought tremendous technical advances in the diagnosis of acoustic schwannomas. These advances are reflected in the tumor size at the time of diagnosis. A recent study revealed a decrease in the average size of tumors diagnosed at a major neuro-otologic practice from 27.9 mm prior to 1975 to 16.5 mm in the period from 1988 to 1989. Despite these improvements in diagnostic technology, the average duration of symptoms prior to diagnosis is still greater than 4 years. Based on  estimates of tumor growth, this delay may result in a 10- to 16-mm expansion of an existing tumor. The majority of tumors are diagnosed in the early cisternal stage ( 1 to 3 cm), but from 7 to 16 percent reach a large size (> 3 cm) prior to diagnosis. This size is large enough to cause significant brain stem compression and complicate resection. Diagnostic delay may be due to the patient's failure to seek medical attention or to the clinician's failure to appropriately screen patients at risk. The economic and personal costs of delayed diagnosis are substantial. It is thus incumbent on each clinician to institute an appropriate, cost-effective diagnostic strategy that will provide early detection of these tumors