I T is generally assumed that the evoked potentials

Transcription

1 J Neurosurg 59: , 1983 uditory evoked potentials recorded from the cochlear nucleus and its vicinity in man GE R. M~LLER, PH.D., ND PETER J. JNNETr, M.D. Division of Physiological coustics, Department of Otolaryngology, and Department of Neurological Surgery, University (~f Pittsburgh School of Medicine, Pittsburgh, Pennsylvania u- Intracranial responses from the auditory nerve and the cochlear nucleus were recorded from patients undergoing neurosurgical operations during which these structures were exposed. Responses to stimulation of the ipsilateral ear with short tonebursts from the vicinity of the cochlear nucleus show a large surface-negative peak, the latency of which is close to that of peak III in the auditory brain-stem evoked potentials recorded from scalp electrodes. There was also a response to contralateral stimulation, smaller in amplitude and with a longer latency. It is concluded that the cochlear nucleus is the main generator of peak IIl responses, and that structures of the ascending auditory pathway that are more central than the cochlear nucleus are unlikely to contribute to wave III of the auditory brain-stem evoked potentials. KEY WORDS cochlear nucleus evoked potentials auditory nerve brain-stem auditory evoked potentials I T is generally assumed that the evoked potentials that can be recorded from the scalp are the result of a sequential activation of the various nerve tracts and nuclei that constitute the ascending auditory pathway. Defining the origin of brain-stem auditory evoked potentials (BEP) has been the aim of extensive recent studies, i Recordings made simultaneously from the scalp and the ear, either from the promontorium by electrocorticography (ECoG) 3"21 or from the ear canal close to the tympanic membrane, 1 have clearly shown that peak I of the scalp recording occurs simultaneously with the initial negative peak in the compound action potential that can be recorded from the ear. This is convincing evidence that peak I in the BEP originates in the auditory nerve. We have previously shown that potentials recorded from the intracranial part of the auditory nerve in man have latencies that are longer than those of peak I of the BEP recorded at the same time. ls'19 The latencies of the potentials recorded from the eighth nerve depend upon the location of the recording electrode on the nerve. 13 These results are taken as evidence that the auditory nerve in man is the generator of not only the first peak but also of the second peak of the auditory evoked potential.l 1-13,15,17 20 This has been attributed to the fact that the auditory nerve in man is about 25 mm in length 8 and that its conduction velocity is only about 20 m/sec. 9 Recordings obtained from the auditory nerve and various structures of the ascending auditory pathway in patients who underwent neurosurgical operations for trigeminal neuralgia and hemifacial spasm showed that the inferior colliculus in man is most likely not the generator of peak V. ~4 In addition, the inferior colliculus is likely to be the main generator of the slow negative potentials that appear with a latency of 7 to 8 msec (SN~o), 2 and of peaks VI and VII. When the stimuli are relatively long tonebursts, in some cases a slow negative potential that resembles the envelope of the stimulus could be seen to follow the initial triphasic pattern of the eighth nerve potential.13 Recordings from the nuclei of the ascending auditory pathway are characterized by slow potentials and only after removal of these slow potentials by highpass filtering does the pattern of peaks become visible, t~'~2'~4 These recent studies have changed our understanding of the neural generators: instead of the simplistic interpretation that prevailed during the first decade that these potentials were studied to aid in the diagnosis of otoneurological problems, z3 a much more complex interpretation now seems necessary) 7 Several of these findings were confirmed by others who studied intracranial recordings from human subjects. 4'22 It also emerges from recent studies that each of the J. Neurosurg. / l~)lume 59 / December,

2 . R. MNler and P. J. Jannetta peaks in the BEP, except peaks I and II, receives contributions from more than one source, ~ ~.~2 and that the crossed and uncrossed pathways combine and contribute to the potentials recorded from the scalp in a complex way. 4'~ In the present study, we have compared BEP recorded from the scalp with recordings from different locations on the brain stem near the cochlear nucleus and the auditory nerve. The cochlear nucleus in man has a slightly different anatomical location that it has in the small animals that are usually used in auditory neurophysiological experiments. The cochlear nucleus in man is buried deep under the cerebellar peduncle. It is therefore not usually exposed in neurosurgical operations. We have made recordings from the vicinity of the cochlear nucleus by placing the recording electrode in the lateral recess during operations for cranial nerve dysfunction where access to the fifth, seventh, or eighth nerve is gained through retromastoid craniectomy. In one patient operated on for a tumor in the fourth ventricle, we recorded from a location medial to the cerebellar peduncle. Clinical Material and Methods The results presented are based on recordings from more than 25 patients* undergoing neurosurgical operations to relieve trigeminal neuralgia, hemifacial spasm, dizziness, and tinnitus by microvascular decompression of the appropriate cranial nerve through retromastoid craniectomy, s-7 One patient was operated on for a tumor in the fourth ventricle. Recordings were made using the technique described previously;~l.~2.j4.j8.~9 therefore, only a brief description of the methods will be given here. Teflon-insulated silver wires with a cotton wick sutured to the uninsulated tip were used to record from the auditory nerve, the surface of the brain stem near the entrance of the eighth nerve, the lateral recess of the fourth ventricle, and the cerebellar peduncles near the cochlear nucleus. The BEP were recorded from the scalp using needle electrodes placed on the vertex just above the pinna (corresponding to the mastoid recording in conventional recordings of BEP), and a common reference electrode was placed over the left clavicle. The recorded potentials were amplified by an amplifier? set at bandpass of 3 to 3000 Hz and with an amplification of x The potentials recorded intracranially were amplified using the same type of amplifier with identical filter settings but at x 2000 amplification. The potentials were recorded on magnetic tape using a Vetter Type B FM tape recorder. The sound stimuli were * The technique used was approved by the Human Use Committee of the University of Pittsburgh School of Medicine, and patients gave informed consent. Some of the data were obtained during monitoring of auditory nerve function during microvascular decompression operations. J6 t mplifier, Type 511J, manufactured by Grass Instruments Co., Quincy, Massachusetts. Hz tonebursts of 1- or 5-msec duration with a rise and fall time of 0.2 msec (to 90% and 10% maximal amplitude, respectively) presented at interstimulus intervals of 110 msec. The potentials recorded from scalp electrodes (BEP) were filtered digitally using filters with zero phase-shift, ~~176 and the same filtering was also in some instances applied to the potentials recorded intracranially in order to remove slow potentials. Results In Fig. 1, the potentials recorded from the eighth nerve (Fig. I) are compared to those recorded from the lateral recess of the fourth ventricle (Fig. 1B). Recording of BEP (Fig. 1C) was done simultaneously from electrodes placed on the scalp on the vertex and just above the pinna. It is seen that the recording from the eighth nerve is characterized by an initial positive deflection followed by a sharp negative peak, then a positive peak and then a second negative peak. There is a slow negative deflection, the duration of which approximates that of the stimulus (5 msec). The large negative peak has a latency close to that of peak II of the BEP. This is in agreement with what we have shown earlier. ~8,~ The recording from the lateral recess (Fig. 1B) shows a small initial positive deflection followed by a slow negative potential on which two or three peaks are riding. The main negative peak in the recording from B C I 0 I RV I I I OlO I0 FIG. 1. Recordings from the eighth nerve (), lateral recess of the fourth ventricle (B), and differentially from electrodes placed on the vertex at a location just above the ipsilateral pinna (C). In the intracranial recordings negativity is shown upward, and in the scalp recordings vertex negativity is shown as an upward deflection. The stimulus was 2000-Hz tonebursts of 5-msec duration, and the intensity was 95 db sound pressure level J. Neurosurg. / Volume 59 / December, 1983

3 uditory evoked potentials from cochlear nucleus in man the lateral recess has the same latency as peak III in the BEP. These recordings were obtained in a patient operated on for hemifacial spasm. The stimuli were 2000-Hz tonebursts of 5-msec duration presented at interstimulus intervals of 110 msec. The recording from the eighth nerve was made by placing an electrode close to but not in direct contact with the nerve. The recordings from the scalp shown in Fig. 1C were filtered digitally, whereas the recording bandwidth for the potentials recorded intracranially (Fig. 1 and B) was 3 Hz to 3 khz and not filtered digitally. Figure 2 shows the same data as Fig. 1 after the potentials recorded intracranially were subjected to the same digital filtering in order to suppress slow components and emphasize the peak pattern. It is seen that the recordings from the auditory nerve are changed little, but that the recordings from the lateral recess after filtering show a clear pattern of peaks. This latter recording is dominated by a large peak with the same latency as peak III of the BEP, and a much smaller peak is seen with the latency of the large negative peak recorded from the eighth nerve. This indicates that the recording electrode, when placed in the lateral recess, was close to the source of peak III while it was far from the generators of earlier peaks. The fact that the main peak in the recording from the lateral recess has the same latency as peak IlI of the BEP recorded simul- taneously indicates that the main source of peak IlI is the cochlear nucleus. The potentials generated by the cochlear nucleus can also be identified by recording from different locations on the auditory nerve (as seen in Fig. 3, B, and C). This graph also shows the BEP recorded using scalp electrodes, one placed on the vertex and one placed just above the pinna (Fig. 3). It is seen that the potentials recorded from the eighth nerve near the porus acusticus (Fig. 1) are characterized by a large negative peak, the latency of which is slightly shorter than that of peak II in the BEP recorded simultaneously. Recording from a position on the nerve halfway between the porus acusticus and the brain stem reveals a potential with a similar shape but with a slightly longer latency and a slightly smaller amplitude. Recording from a location on the brain stem near the entrance of the eighth nerve shows a potential with a different shape. The early B c # / /I /I III FIG. 2. Same data as in Fig. 1 but recordings taken after digital filtering to remove slow potentials. I It III 200nV I 1 I I I I I I 0 I B 9 I0 FIG. 3. Recordings from different locations on the eighth nerve: near the porus acusticus (); halfway between the porus acusticus and the brain stem (B); at the brain stem (C); and recorded differentially between a location just above the pinna and the vertex (D). The stimulus was 2000-Hz tonebursts of l-msec duration and the intensity was 100 db sound pressure level. The recordings were obtained from a patient operated on to relieve dizziness and tinnitus by microvascular decompression of the eighth nerve. V J. Neurosurg. / Volume 59 / December,

4 . R. Moller and P. J. Jannetta negative peak has a small amplitude and the second negative peak has an increased amplitude which dominates the recorded potentials. The latency of this large peak is equal to that of peak III of the BEP. The reason that the amplitude of the second peak increases when the electrode is moved to the brain stem is assumed to be that it is coming closer to the source of that peak, which is likely to be the cochlear nucleus. ccess directly to the cochlear nucleus is usually not possible because in man it is located behind the large cerebellar peduncle. However, in one patient operated on for a tumor located on the roof of the fourth ventricle we recorded from the medial side of the cerebellar peduncle at a location which is presumably close to the cochlear nucleus. recording from the medial part of the cerebellar penduncle near the floor of the fourth ventricle is seen in Fig. 4. It is seen that the recording is dominated by a peak with the same latency as peak III in the BEP recorded simultaneously from the scalp (Fig. 4C), but the recording also clearly shows a peak with the same latency as peak II. This initial deflection in the intracranial recording is a positive peak, and is most likely generated by the auditory nerve fibers entering the cochlear nucleus, and thus the proximal part of the auditory nerve. lso seen in the recordings from the lateral recess is a peak with a latency that coincides with peak V in the BEP recorded from the scalp. This peak may be generated by the fiber tract of the lateral lemniscus that contains fibers originating in the ipsilateral cochlear nucleus. Figure 4B shows the same recording as in Fig. 4 after digital filtering to make the peaks appear more clearly. The initial positive peak is most likely generated by the auditory nerve fibers that terminate in the cochlear nucleus, and the main negative peak represents activity in the cochlear nucleus. Recordings from the same location on the cerebellar penduncle during contralateral stimulation reveal potentials of much lower amplitude, as shown before (Fig. 5) and after (Fig. 5B) filtering. lso shown are the potentials recorded from the vertex (Fig. 5C). The potentials recorded intracranially are characterized by a positive deflection, with a latency that is slightly longer than that of peak III in the BEP, followed by a smaller positive deflection. There are also small peaks with shorter latencies which indicate that these potentials recorded contralaterally cannot be regarded as generated exclusively by a specific nucleus, but likely also receive contributions from several distant sources such as the auditory nerve, cochlear nucleus, and more central structures. The fact that the BEP recorded from the scalp in IpV B # IO,uV o.5~v C FIG. 4. Recordings made from the cerebellar penduncle medially near the floor of the fourth ventricle in response to ipsilateral stimulation before () and after (B) digital filtering. Recordings from the vertex are shown in C. The stimulation was 2000-Hz tonebursts of 5-msec duration at 95 db. The results were obtained in a patient operated on for a tumor in the fourth ventricle. C 20On I V FIG. 5. Recordings from the same location and the same patient as in Fig. 4, but with contralateral stimulation J. Neurosurg. / Volume 59 / December, 1983

5 uditory evoked potentials from cochlear nucleus in man the conventional way is the difference between the potentials at the vertex and at the mastoid, and that the potentials at these two locations have slightly different latencies, makes the peaks of the BEP recorded in the conventional way (namely, differentially recorded between the vertex and mastoid) differ slightly from those of the potentials recorded from either one of these two locations. When a comparison is made between the farfield potentials and the near-field potentials recorded intracranially, it may therefore be more appropriate to compare the potentials recorded intracranially to potentials recorded from either the mastoid or the vertex using a noncephalic reference.~7 The way potentials that are generated by local sources are transformed into far-field potentials that can be recorded on the scalp is a function of several factors about which we have insufficient information. These sources are usually assumed to be dipoles. The potentials that are recorded at a large distance from the source are therefore not only related to the size of the potentials recorded from a position close to the source and the distance to the source, but they also depend I,oo v \ upon the extension of the source. We have compared the potentials recorded at the auditory nerve and those recorded a short distance from the nerve to the far-field potentials. Figure 6 shows an example of such recordings: Fig. 6 shows tracings obtained by recording from the eighth nerve near the porus acusticus, the tracings in Fig. 6B were obtained by placing an electrode on the Gelfoam packing placed over the nerve during closing, the recordings in Fig. 6C were obtained simultaneously from a location on the scalp just above the pinna, and Fig. 6D shows the conventional BEP obtained by differential recordings from the vertex and a position just above the pinna using a noncephalic reference. It is seen that the potentials decrease rapidly in amplitude as the electrode is moved away from the nerve, and that the shapes of the recorded potentials resemble those recorded near the brain stem with at least three peaks easily discernible. The latency of the first peak is slightly longer than that of the potentials recorded from the nerve near the porus acusticus. The two first negative peaks in the potentials recorded from the scalp at a location just above the pinna have latencies that are very similar to those seen in Fig. 6B. These two peaks occur at about the same latencies as peaks II and III of the BEP recorded differentially. The potentials re- corded from the scalp are only about one-fourth of the amplitudes of the potentials recorded intracranially. These results show that peaks II and III have a relatively simple relationship to the potentials that can be recorded intracranially, but that peaks IV, V, and VI are likely to have several different sources. I I ~ I I Vl I I I TIME IN MILLISECONDS FIG. 6. Recordings from the eighth nerve near the porus acusticus (), from the Gelfoam packing placed on the nerve during closing (B), simultaneously from a location immediately above the pinna (over the clavicle) with a noncephalic reference (C); and brain-stem auditory evoked potentials recorded differentizlly (D). The stimulus was 2000-Hz tonebursts of 5-msec duration at 100 db. The patient was operated on for hemifacial spasm. nv Discussion The results of the present study strongly indicate that the main generator of wave III of the BEP in man is the ipsilateral cochlear nucleus. This is in agreement with the assumption made on the basis of results of earlier studies showing that both peaks I and II are generated by the auditory nerve. ~-~3J8'~9 Recordings from two different locations that are assumed to be close to the cochlear nucleus (namely, at the brain stem at the entrance of the eighth nerve and on the medial side of the cerebellar penduncle near the floor of the fourth ventricle) are similar in that the main components of the potentials had latencies that were very close to those of peak III of the BEP recorded from scalp electrodes. Our experiences in recording from the eighth nerve and the brain stem in more than 100 patients show that it is of utmost importance that the recording electrode is accurately placed. The nuclei and fiber tracts of the auditory pathway in man occupy a much smaller space than they do in the small animals traditionally used in auditory research. Slight misplacements of the recording electrode result in recording of only far-field potentials. The fact that the recordings of the present study show J. Neurosurg. / Volume 59 / December,

6 . R. MOiler and P. J. Jannetta strong components at specific latencies indicates that the recordings were obtained from locations that were near one of the neural generators of these potentials. natomical evidence strongly suggests that this particular neural generator is located in the cochlear nucleus. Thus, although it was not possible to place the electrode on the cochlear nucleus under direct visual control, we are confident that the electrode was placed very close to the cochlear nucleus when it was positioned near the entrance of the eighth nerve into the brain stem or on the cerebellar penduncle at the floor of the fourth ventricle. The fact that there is a response from the cochlear nucleus to contralateral stimulation, although of a smaller amplitude, is in agreement with the results of animal experiments. That this response has latencies that have no clear relation to the peaks of the BEP makes it unclear what contribution this source makes to the BEP. References 1. Coats C, Martin JL: Human auditory nerve action potentials and brain stem evoked responses. Effects of audiogram shape and lesion location. rch Otolaryngol 103: , Davis H, Hirsh SK: The audiometric utility of brain stem responses to low-frequency sounds. udiology 15: , Gersdorff MCH: Simultaneous recordings of human auditory potentials: transtympanic electrocochleography (ECoG) and brainstem-evoked responses (BER). rch Otorhinolaryngol 234:15-20, Hashimoto I, lshiyama Y, Yoshimoto T, et al: Brain-stem auditory-evoked potentials recorded directly from human brain-stem and thalamus. Brain 104: , Jannetta PJ: Hemifacial spasm, in Samii M, Jannetta PJ (eds): The Cranial Nerves. Berlin/Heidelberg/New York: Springer-Verlag, 1981, pp Jannetta PJ: Neurovascular compression in cranial nerve and systemic disease. nn Surg 192: , Jannetta PJ: Treatment of trigeminal neuralgia by suboccipital and transtentorial cranial operations. Clin Neurosurg 24: , Lang J: Facial and vestibulocochlear nerve, topographic anatomy and variations, in Samii M, Jannetta PJ (eds): The Cranial Nerves. Berlin/Heidelberg/New York: Springer-Verlag, 1981, pp Lazorthes G, Lacomme Y, Gaubert J, et al: La constitution du nerf auditif. Presse Med 69: , M~ller R: Improving brain stem auditory evoked potential recordings by digital filtering. Ear Hear 4: , Moller R, Jannetta PJ: uditory evoked potentials recorded intracranially from the brain stem in man. Exp Neurol 78:t44-157, M011er R, Jannetta PJ: Comparison between intracra- nially recorded potentials from the human auditory nerve and scalp recorded auditory brain stem responses. Stand udioi 11:33-40, Moller R, Jannetta PJ: Compound action potentials recorded intracranially from the auditory nerve in man. Exp Nenrol 74: , M~ller R, Jannetta PJ: Evoked potentials from the inferior colliculus in man. Electroencephalogr Clin Neurophysiol 53: , Moiler R, Jannetta PJ: Interpretation of brain stem auditory evoked potentials: results from intracranial recordings in humans. Scand udiol 12: , Moiler R, Jannetta PJ: Monitoring auditory functions during cranial nerve microvascular decompression operations by direct recording from the eighth nerve. J Neurosurg 59: , Moiler R, Jannetta PJ: Neural generators of the brain stem auditory evoked potentials (BEP) in man studied in intracranial recordings, in: Proceedings of the Second International Evoked Potentials Symposium. Cleveland, Ohio, 1983 (In press) 18. Moiler R, Jannetta PJ, Bennett M, et al: Intracranially recorded responses from the human auditory nerve: new insights into the origin of brain stem evoked potentials (BSEPs). Electroencephalogr Clin Neurophysiol 52: 18-27, M011er R, Jannetta P J, Moiler MB: Neural generators of brain stem evoked potentials. Results from human intracranial recordings. nn Otolaryngol Laryngol 90: , M011er R, Moiler MB, Millner D: computer system for auditory evoked responses, in Shriver B, Walker T, Grams R, et al (eds): Proceedings of the Fourteenth Hawaii International Conference on System Sciences. North Hollywood, Calif." Western Periodicals, 1981, Vol 2. pp Portmann M, Cazals Y, Negrevergne M, et al: Transtympanic and surface recordings in the diagnosis of retrocochlear disorders. rch Otolaryngol 89: , Spire JP, Dohrmann G J, Prieto PS: Correlation of brain stem evoked response with direct acoustic nerve potential, in Courjon J, Maugui+re F, Revol M (eds): Clinical pplications of Evoked Potentials in Neurology. dvances in Neurology, Vol 32. New York: Raven Press, 1982, pp Thornton RD: Interpretation of cochlear nerve and brain stem evoked responses, in Naunton RF, Fernandez C (eds): Evoked Electrical ctivity in the uditory Nervous System. New York: cademic Press, 1978, pp Manuscript received pril 18, This work was supported by the Deafness Research Council. ddress reprint requests to: age R. M011er, Ph.D., Division of Physiological coustics, Department of Otolaryngology. University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania J. Neurosurg. / Volume 59 / December, 1983