Diagnostic accuracy of brainstem auditory evoked potentials during microvascular decompression Parthasarathy D. Thirumala, MD, MS Gregory Carnovale Miguel E. Habeych, MD, MPH Donald J. Crammond, PhD Jeffrey R. Balzer, PhD
Correspondence to Dr. Thirumala: [email protected]
ABSTRACT
Objective: The primary aim of the study was to assess the sensitivity and specificity of intraoperative monitoring in predicting postoperative hearing loss during microvascular decompression (MVD).
Methods: The study was designed as an examination of the diagnostic accuracy of brainstem evoked potentials compared with reference standard for nonserviceable hearing loss (Class C/D), which is a change of more than 50 dB on pure tone threshold, and change of speech discrimination score of more than 50. All patients underwent surgery and audiograms at a University of Pittsburgh Medical Center facility in the study period 2005–2012. All participants received a pre- and postaudiogram within 90 days before or after the operation. During the operation, participants received intraoperative monitoring with a supervising physician. A total of 238 patients were selected. Brainstem auditory evoked potentials (BAEPs) were indexed into categories of change based on their maximum change and response at the end of surgery. Differences in hearing outcome by BAEP change were analyzed. Results: Age and sex did not affect outcomes. Patient outcome was affected by condition. The BAEP categories significant changes, transient loss, and persistent loss show a sensitivity/specificity of 0.905/0.701, 0.667/0.903, and 0.429/0.972, respectively. The receiver operating characteristic curve has an area under the curve of 0.870 with a 95% confidence interval of 0.783 to 0.957. Conclusions: Loss of wave V during MVD is a specific indicator of postoperative hearing loss. The current alarm criteria used to warn the surgeon is a sensitive indicator of impending postoperative hearing loss. Classification of evidence: This study provides Class IV evidence that in patients undergoing MVD, intraoperative BAEPs accurately identifies those who will have postoperative hearing loss. Neurology® 2014;83:1747–1752 GLOSSARY AAO-HNS 5 American Academy of Otolaryngology–Head and Neck Surgery; AUC 5 area under the curve; BAEP 5 brainstem auditory evoked potential; CI 5 confidence interval; CN 5 cranial nerve; GN 5 geniculate neuralgia; GPN 5 glossopharyngeal neuralgia; HFS 5 hemifacial spasm; HL 5 hearing loss; MVD 5 microvascular decompression; PTA 5 pure tone audiometry; ROC 5 receiver operating characteristic; SDS 5 speech discrimination score; TGN 5 trigeminal neuralgia.
Supplemental data at Neurology.org
Microvascular decompression (MVD) effectively treats for trigeminal neuralgia (TGN), hemifacial spasm (HFS), geniculate neuralgia (GN), and glossopharyngeal neuralgia (GPN) by relieving the vascular pressure on the cranial nerve (CN) at the root exit zone.1–3 Retraction at the cerebellopontine angle to decompress the CN is the most common mechanism of postoperative sensorineural hearing loss (HL).4 HL occurs at a rate of 2% to 20% in HFS, 1% to 23.8% in TGN, 1.5% to 4.2% in GPN, and 22% in GN.2,5–7 The wide variations in the incidence of HL could be secondary to the criteria used to define HL. We utilized the American Academy of Otolaryngology–Head and Neck Surgery (AAO-HNS) classification system8 to evaluate HL after MVD,8 where HL corresponding to a 50% decrease in speech discrimination scores (SDS) and/or a 50-dB increase in pure tone audiometry (PTA) scores provides a comprehensive From the Departments of Neurologic Surgery (P.D.T., M.E.H., D.J.C., J.R.B.) and Neurology (P.D.T.), University of Pittsburgh Medical Center; and Department of Neuroscience (G.C., J.R.B.), University of Pittsburgh, PA. Go to Neurology.org for full disclosures. Funding information and disclosures deemed relevant by the authors, if any, are provided at the end of the article. © 2014 American Academy of Neurology
1747
assessment of hearing function. Intraoperative monitoring of the brainstem auditory evoked potentials (BAEPs) of CN VIII has been shown to reduce postoperative CN VIII morbidity.9 To prevent HL, the surgeon is generally warned before the wave V latency increases to 0.5 or 1.0 millisecond, or amplitude decreases more than 50% compared with baseline values.10,11 This criterion is somewhat arbitrary and is not predictive of hearing dysfunction after MVD.10,11 Although the trend in changes in wave V during cerebellopontine angle surgery is significant in patients with HL,12 we clearly do not understand the predictive value of waveform V changes in HL. The primary aim of our study was to evaluate the diagnostic accuracy using sensitivity and specificity of changes in the latency and amplitude of wave V of BAEPs and to predict HL after MVD. METHODS Study population. A cross-sectional study was conducted from data collected in consecutive patients with TGN, HFS, GN, or GPN who underwent MVD at a University of Pittsburgh Medical Center facility between 2005 and 2012. All patients who underwent MVD received preoperative baseline and continuous intraoperative BAEP monitoring including preand postoperative audiometric examination. The data were collected for each patient after the procedure and an audiometric evaluation were completed. The data analysis was planned after the collection of the data. Patients without full intraoperative monitoring and/or pre- and postoperative audiometric data who underwent MVD were excluded from the study. The study evaluated the diagnostic accuracy of BAEPs to predict postoperative HL. The study provides Class IV evidence for utilizing BAEPs in patients undergoing MVD.
Standard protocol approvals, registrations, and patient consents. The study was approved by the University of Pittsburgh institutional review board for retrospective review of data on human subjects (MOD08120394-04/PRO08120394).
Audiologic investigations and HL criteria. We considered tone audiometry consisting of PTA scores and SDS to be the indicator of auditory function.13,14 Postoperative HL status was assessed using the 1995 AAO-HNS classification system.8 Class B referred to useful or serviceable HL and Class C/D was nonserviceable HL that was not amenable to hearing aids.8 We performed an otoneurologic examination on all patients preoperatively (median 1 day, range 1–49 days) and postoperatively (median 7 days, range 1–90 days). This consisted of an audiogram with measurement of pure tone thresholds (air and bone conduction for octave frequencies 250–8,000 Hz) as previously described.6 All audiograms were performed at the Eye and Ear Institute, University of Presbyterian Hospital, by an audiologist. Intraoperative monitoring. Recordings. All patients underwent intraoperative monitoring using BAEP recordings in the following manner. Independent alternate ear stimulation of at least 1748
Neurology 83
November 4, 2014
17.5 Hz was performed in the right and left side with at least 85 dB nHL (normal HL), using alternating rarefaction and condensation clicks. The contralateral ear was stimulated with white noise at 65 dB nHL. We observed the data for 12 milliseconds, and averaging at least 256 responses. BAEP recording electrodes were positioned as follows: (channel 1) vertex to left ear mastoid Cz/A1; (channel 2) vertex to right ear mastoid Cz/A2; and (channel 3) vertex to cervical C2 (Cz/Cv2). Amplifier bandpass was 100 to 1,000 Hz for all channels. After induction of anesthesia and positioning the patient, the baseline responses were established. Continuous real-time BAEPs were obtained from the operative side in all patients. The contralateral ear BAEP responses were recorded at baseline. Analysis of neurophysiologic parameters. The absolute latencies of waves I, III, and V (LwI, LwIII, and LwV, respectively) of the BAEPs were identified based on previously described methods.15 The amplitudes of waves III and V (AwIII and AwV, respectively) were measured with exclusion for wave I (AwI), which could be masked by stimulation artifact. Wave V of BAEP is the main focus of the analysis, because it is the most robust waveform that can be recorded consistently during MVD. The BAEP waveforms were analyzed at the following times during the procedure: (1) baseline recording after the patient was intubated and positioned; (2) change START, defined when a first persistent decrease in amplitude of wave V (AwV) .50% and/or increase in LwV .0.5 millisecond compared with baseline was observed; (3) change MAX, defined as the time when a maximum change in the AwV and/or LwV compared with baseline was observed (this included instances when no waveforms were identifiable [loss of response]); and (4) a final recording at the time of skin closure. If loss of response occurred, amplitude was recorded as zero and latency was recorded as 12 milliseconds, which is the maximum observation interval. Physician oversight and interpretation were performed using a combined on-site and remote model utilized by the University of Pittsburgh Medical Center.16 The BAEP waveforms were identified and evaluated independently by research students, who were supervised by the corresponding author.
Group discrimination. Predictive value including sensitivity and specificity of changes in BAEPs to identify HL was calculated after categorizing patients into no change, significant change, transient loss, and persistent loss based on changes in BAEPs during MVD. BAEP waveform changes in latency and amplitude are dynamic during MVD, secondary to the degree of retraction and/or compression affecting the auditory nerve or its vasculature. Hence, we thought that categorizing the changes that incorporated the current alarm criteria may be helpful in understanding the value of BAEPs during MVD. Loss of amplitude of 100% and/or prolongation in wave V latency greater than 12 milliseconds was regarded as loss of waveforms. The loss of wave V of BAEP was further subdivided into persistent loss and transient loss based on whether the waveforms improved after the loss during decompression. Transient loss was regarded as wave V peaks that recovered to baseline in amplitude and latency before the end of surgery. Persistent loss BAEPs did not recover before the end of surgery. Patients who had consistent decrease in amplitude of more than 50% of wave V and/or persistent increase in the absolute latency of the peak of wave V that was $0.5 milliseconds and did not have loss of waveform during MVD were reported as significant change.11,17 “Consistent changes” represent 2 consecutive epochs of latency and or amplitude changes in BAEPs compared with baseline responses in more than 2 consecutive epochs. Two consecutive epochs were used to eliminate technical issues as the cause of waveform change, such as electrical interference from
Figure 1
Patient selection
Patient selection methods with index test, brainstem auditory evoked potential category, reference standard, and ontological examination are shown. BSER 5 brainstem evoked response; MVD 5 microvascular decompression; UPMC 5 University of Pittsburgh Medical Center.
cauterizing or drilling. BAEPs in patients whose wave V changes did not meet these criteria were categorized as no change.
Statistical methods. Mean and SD of age, percentage of female patients, and percentage incidence of HL were calculated for all MVD procedures and by condition. A distribution of PTA scores and SDS was given for all patients, by HL category and by condition displayed in a box plot. Previous MVD studies have reported that a significantly higher number of patients who had permanent loss of wave V had new, postoperative HL.12 To compare our data with previous studies on the degree of change of
Table 1
Descriptive statistics All patients
HFS
TGN
GPN
GN
No. of patients
238
123
100
8
7
Age, y, mean (SD)
52.41 (12.61)
52.29 (11.98)
53.04 (13.68)
48.25 (8.84)
50.28 (12.2)
Female, n (%)
162 (68)
76 (62)
73 (73)
7 (87.5)
5 (71)
HL, n (% of category)
21 (8.8)
18 (15)
3 (2)
0
0
Abbreviations: GN 5 geniculate neuralgia; GPN 5 glossopharyngeal neuralgia; HFS 5 hemifacial spasm; HL 5 hearing loss; TGN 5 trigeminal neuralgia. Patients are separated by condition. HL corresponds to postoperative hearing C/D.
wave V, a test of group discrimination was performed. To account for a small sample, a Fisher exact test was performed, which is nonbiased for sample size. The test determines whether at least one of the categories is significantly different from the others. We calculated the sensitivity and specificity of BAEP categories using the reference test (otoneurologic examination using the 1995 AAO-HNS classification system). BAEP categories include no change, significant change, transient loss, and persistent loss. A true-positive result was regarded as BAEPs with the degree of change tested or more severe and experiencing HL C/D. A false positive was regarded as BAEPs with the degree of change tested or more severe and experiencing HL A/B. A false negative was BAEPs with less severe change than the degree of change tested with HL C/D. A true negative was BAEPs with less severe change than the degree of change tested with HL A/B. A 95% confidence interval (CI) was calculated for all sensitivities and specificities. To further evaluate the diagnostic accuracy of change in BAEPs and postoperative HL, we used a receiver operating characteristic (ROC) curve with BAEP categories used as cutoff values. ROC was chosen because the area under the ROC curve (AUC) has been used as a measure of prediction accuracy for a variety of biomedical applications. ROC AUC, significance, and 95% CI were calculated. Standard error was calculated using a nonparametric distribution. The significance for all tests was set at p , 0.05. All statistics were completed using SPSS version 20 (IBM Corp., Armonk, NY). Neurology 83
November 4, 2014
1749
Table 2
Number of patients with changes and hearing loss BAEP change No significant change
Significant change
Loss of waveforms
Transient loss
Persistent loss
Total
No. of patients
144
49
35
20
15
238
Patients with hearing loss
2
5
14
5
9
21
Sensitivity (95% CI)
0.905 (0.696–0.986)
0.667 (0.430–0.854)
0.667 (0.430–0.854)
0.429 (0.217–0.660)
Specificity (95% CI)
0.701 (0.635–0.761)
0.903 (0.856–0.939)
0.903 (0.856–0.939)
0.972 (0.941–0.990)
Positive predictive value (95% CI)
0.184 (0.117–0.276)
0.4 (0.243–0.578)
0.4 (0.243–0.578)
0.6 (0.328–0.825)
Negative predictive value (95% CI)
0.986 (0.946–0.997)
0.965 (0.927–0.984)
0.965 (0.927–0.984)
0.946 (0.905–0.970)
Abbreviations: BAEP 5 brainstem auditory evoked potential; CI 5 confidence interval. Sensitivity, specificity, and positive and negative predictive value for hearing loss are compared by BAEP category.
RESULTS Participants. Figure 1 displays patient selection. A total of 238 patients who underwent MVD at the University of Pittsburgh Medical Center during the study period (figure 1) were eligible to participate. Descriptive statistics for number of patients with each condition, age with SD, sex, and HL are provided in table 1.
BAEP and ontological results. The probability of HL
when there was no significant change (no change in wave V) was 1.3%, compared with the probability of HL of 10.2% of patients when a significant change in the BAEP was reported, 25% of patients with
Figure 2
ROC curve
Receiver operating characteristic (ROC) curve to evaluate the changes in brainstem auditory evoked potential and its relationship with hearing loss. Points on the ROC curve correspond to specificities and sensitivities given in table 2. Note that diagonal segments are produced by ties. 1750
Neurology 83
November 4, 2014
transient loss, and 60% of patients with persistent loss. Box plots display the distribution of SDS and PTA scores by condition (see figures e-1 and e-2 on the Neurology® Web site at Neurology.org). Results of categorical discrimination. The Fisher exact
statistic had a value of 46.319, with a statistical significance ,0.001 to significantly discriminate various BAEP changes and the incidence of HL. Table 2 displays specificities and sensitivities of various categories of changes in BAEPs. The cutoff for significant change was sensitivity of 90.5% (69.6%–98.6%) and specificity of 70.1% (63.5%–76.1%) (table 2). The cutoff for transient loss was sensitivity of 66.7% (43.0%–85.4%) and specificity of 90.3% (85.6%–93.9%) (table 2). The cutoff for persistent loss was a sensitivity of 42.9% (21.7%–66.0%) and specificity of 97.2% (94.1%–99.0%) (table 2). The ROC curve has an AUC of 0.870, significance ,0.001, and 95% CI of 0.783–0.957, as displayed in figure 2. The cutoff points in the ROC curve correspond to the sensitivity and specificity of index categories, visually defined by figure 3, where a more severe change in BAEP corresponds to a more positive result. DISCUSSION Our study results indicate that loss of wave V during MVD is a highly specific predictor of postoperative HL. This was seen in patients with transient as well as persistent loss of wave V responses. The primary mechanism for CN VIII injury during MVD has been identified as nerve stretching or compression during cerebellar retraction, but can also be caused by damage to CN vasculature providing blood flow to the cochlear nerve or middle ear.18 During MVD, manipulation of the auditory nerve can result in neurapraxic nerve injury and significant changes in waves II and III of the BAEP. However, AwII and AwIII of the BAEP are typically small during intraoperative recording and, hence, are often difficult to reliably record in the operating room during MVD. Change in wave V reflects damage to ascending converging afferents whose synaptic
Figure 3
Highlights the various changes of BAEPs including the significant changes, transient, and complete loss of responses during microvascular decompression
Timeline is shown on the time bar to highlight its relationship with the procedure. BAEP 5 brainstem auditory evoked potential.
integration causes a smaller postsynaptic potential due to a reduction in the overall strength of inputs and their synchrony, which cumulates over several synapses. Even though the MVD-related nerve injury affects the proximal segment of the auditory pathway, monitoring wave V during MVD surgery is a robust and useful measure of conduction along the brainstem auditory pathway. A significant loss of AwV could be correlated to mechanical deformation or ischemia degree of severity applied to CN VIII.18,19 Increase in pressure or lack of blood flow leads to loss of amplitude and is accompanied by histologic changes of nerve injury.18,19 However, improvement in the amplitude of a nerve response after complete loss as seen in BAEPs during MVD can occur without permanent injury to the nerve.19 Most frequently this can occur when the stretch on CN VIII is transient or removed in a timely manner. Hence, loss of wave V response of BAEPs might be a proxy of imminent significant histologic change, and conditions leading to this state during MVD should be avoided. To devise alarm criteria in order to predict and prevent HL during MVD, a sensitive predictor of impending HL is needed. The currently adopted arbitrary “significant changes” include a decrease in amplitude of .50% of wave V and/or persistent
increase in the absolute latency of the peak of wave V that is $1 millisecond.11,20 In our study, significant changes in patients noted during MVD were highly sensitive for identifying HL. We believe this is the first study to evaluate the sensitivity of alarm criteria in intraoperative neurophysiologic monitoring involving BAEPs to evaluate and predict impending HL. Our study revealed 2 patients without significant changes in wave V of BAEPs during MVD who experienced postoperative HL suggesting mechanisms of underlying HL unrelated to CN VIII apraxia and/or injury. Primate studies on the effect of focal brainstem ischemia show that a decrease in brainstem blood flow increases the latency of BAEP waveforms.21 Sensorineural HL may occur as a result of factors unrelated to a conduction block of the auditory nerve or cochlear damage, such as damage to higher order auditory tracts or postoperative brainstem ischemia that does not manifest during the course of surgery.22 In the absence of any signs of clinical brainstem ischemia, or significant change or loss of BAEPs during MVD, it is possible that damage to the cochlea may occur in the immediate postoperative phase. Although this study represents the largest crosssectional diagnostic review of BAEP wave V changes correlated to postoperative HL after MVD, it is not without Neurology 83
November 4, 2014
1751
limitations. The study does not represent a random sample of all MVD procedures because some patients elected to undergo postoperative audiograms only if significant hearing problems were identified postoperatively. Patients who did not have audiometric testing probably had no postoperative hearing change, which most likely overestimates the incidence of HL after MVD. Loss of wave V during MVD is a specific indicator of postoperative HL. The current alarm criteria used to warn the surgeon is a sensitive indicator of impending postoperative HL. HL infrequently occurs when there is no change in BAEPs during MVD. Although stretch-induced injury is common in patients undergoing MVD, other mechanisms of injury need to be considered and evaluated during the procedure.
7.
8.
9.
10.
11. AUTHOR CONTRIBUTIONS Parthasarathy Thirumala: drafting/revising the manuscript, study concept or design, analysis or interpretation of data, accepts responsibility for conduct of research and will give final approval, statistical analysis. Gregory Carnovale: drafting/revising the manuscript, study concept or design, analysis or interpretation of data, accepts responsibility for conduct of research and will give final approval, acquisition of data, statistical analysis, study supervision. Miguel Haybeych: drafting/revising the manuscript, study concept or design, accepts responsibility for conduct of research and will give final approval, acquisition of data, study supervision. Donald Crammond: drafting/revising the manuscript, study concept or design, analysis or interpretation of data, accepts responsibility for conduct of research and will give final approval, acquisition of data, study supervision. Jeffrey Balzer: drafting/revising the manuscript, study concept or design, analysis or interpretation of data, accepts responsibility for conduct of research and will give final approval, acquisition of data.
12.
13.
14.
15. STUDY FUNDING No targeted funding reported.
16. DISCLOSURE The authors report no disclosures relevant to the manuscript. Go to Neurology.org for full disclosures.
Received March 19, 2014. Accepted in final form August 4, 2014. REFERENCES 1. Acevedo JC, Sindou M, Fischer C, Vial C. Microvascular decompression for the treatment of hemifacial spasm: retrospective study of a consecutive series of 75 operated patients—electrophysiologic and anatomical surgical analysis. Stereotact Funct Neurosurg 1997;68:260–265. 2. Patel A, Kassam A, Horowitz M, Chang YF. Microvascular decompression in the management of glossopharyngeal neuralgia: analysis of 217 cases. Neurosurgery 2002;50:705–710. 3. Ramnarayan R, Mackenzie I. Brain-stem auditory evoked responses during microvascular decompression for trigeminal neuralgia: predicting post-operative hearing loss. Neurol India 2006;54:250–254. 4. Little JR, Lesser RP, Lueders H, Furlan AJ. Brain stem auditory evoked potentials in posterior circulation surgery. Neurosurgery 1983;12:496–502. 5. Rizvi SS, Goyal RN, Calder HB. Hearing preservation in microvascular decompression for trigeminal neuralgia. Laryngoscope 1999;109:591–594. 6. Shah A, Nikonow T, Thirumala P, et al. Hearing outcomes following microvascular decompression for hemifacial spasm. Clin Neurol Neurosurg 2012;114:673–677.
1752
Neurology 83
November 4, 2014
17.
18.
19.
20.
21.
22.
Rupa V, Saunders RL, Weider DJ. Geniculate neuralgia: the surgical management of primary otalgia. J Neurosurg 1991;75:505–511. Committee on Hearing and Equilibrium guidelines for the evaluation of hearing preservation in acoustic neuroma (vestibular schwannoma). American Academy of Otolaryngology– Head and Neck Surgery Foundation, INC. Otolaryngol Head Neck Surg 1995;113:179–180. Wilkins RH, Radtke RA, Erwin CW. Value of intraoperative brainstem auditory evoked potential monitoring in reducing the auditory morbidity associated with microvascular decompression of cranial nerves. Skull Base Surg 1991;1:106–109. American Clinical Neurophysiology Society. Guideline 11: Guidelines for Neurophysiologic Intraoperative Monitoring. Milwaukee: American Clinical Neurophysiology Society; 2009. Available at: http://www.acns.org/practice/guidelines. Martin WH, Stecker MM. ASNM position statement: intraoperative monitoring of auditory evoked potentials. J Clin Monit Comput 2008;22:75–85. James ML, Husain AM. Brainstem auditory evoked potential monitoring: when is change in wave V significant? Neurology 2005;65:1551–1555. Sindou MP. Microvascular decompression for primary hemifacial spasm: importance of intraoperative neurophysiological monitoring. Acta Neurochir 2005;147:1019–1026. Polo G, Fischer C, Sindou MP, Marneffe V. Brainstem auditory evoked potential monitoring during microvascular decompression for hemifacial spasm: intraoperative brainstem auditory evoked potential changes and warning values to prevent hearing loss—prospective study in a consecutive series of 84 patients. Neurosurgery 2004;54:97–104. Chiappa KH. Brainstem Auditory Evoked Potentials: Methodology, 3rd ed. New York: Lippincott-Raven Publishers; 1997. Thirumala PD, Kassasm AB, Habeych M, et al. Somatosensory evoked potential monitoring during endoscopic endonasal approach to skull base surgery: analysis of observed changes. Neurosurgery 2011;69:ons64–ons76. Committee on Hearing and Equilibrium guidelines for the evaluation of results of treatment of conductive hearing loss. American Academy of Otolaryngology–Head and Neck Surgery Foundation, Inc. Otolaryngol Head Neck Surg 1995;113:186–187. Ogata K, Naito M. Blood flow of peripheral nerve effects of dissection, stretching and compression. J Hand Surg Br 1986;11:10–14. Wall EJ, Massie JB, Kwan MK, Rydevik BL, Myers RR, Garfin SR. Experimental stretch neuropathy: changes in nerve conduction under tension. J Bone Joint Surg Br 1992;74:126–129. American Clinical Neurophysiology Society. Guideline 11C: Recommended Standards for Intraoperative Monitoring of Auditory Evoked Potentials. Milwaukee: American Clinical Neurophysiology Society; 2009. Available at: http://www.acns.org/practice/guidelines. Baik MW, Branston NM, Bentivoglio P, Symon L. The effects of experimental brain-stem ischaemia on brain-stem auditory evoked potentials in primates. Electroencephalogr Clin Neurophysiol 1990;75:433–443. Lin CD, Wei IH, Tsai MH, et al. Changes in guinea pig cochlea after transient cochlear ischemia. Neuroreport 2010;21:968–975.
Diagnostic accuracy of brainstem auditory evoked potentials during microvascular decompression Parthasarathy D. Thirumala, Gregory Carnovale, Miguel E. Habeych, et al. Neurology 2014;83;1747-1752 Published Online before print October 8, 2014 DOI 10.1212/WNL.0000000000000961 This information is current as of October 8, 2014 Updated Information & Services
including high resolution figures, can be found at: http://www.neurology.org/content/83/19/1747.full.html
Supplementary Material
Supplementary material can be found at: http://www.neurology.org/content/suppl/2014/10/08/WNL.0000000000 000961.DC1.html
References
This article cites 19 articles, 4 of which you can access for free at: http://www.neurology.org/content/83/19/1747.full.html##ref-list-1
Subspecialty Collections
This article, along with others on similar topics, appears in the following collection(s): Evoked Potentials/Auditory http://www.neurology.org//cgi/collection/evoked_potentials-auditory Trigeminal neuralgia http://www.neurology.org//cgi/collection/trigeminal_neuralgia
Permissions & Licensing
Information about reproducing this article in parts (figures,tables) or in its entirety can be found online at: http://www.neurology.org/misc/about.xhtml#permissions
Reprints
Information about ordering reprints can be found online: http://www.neurology.org/misc/addir.xhtml#reprintsus
Neurology ® is the official journal of the American Academy of Neurology. Published continuously since 1951, it is now a weekly with 48 issues per year. Copyright © 2014 American Academy of Neurology. All rights reserved. Print ISSN: 0028-3878. Online ISSN: 1526-632X.
Correspondence to Dr. Thirumala: [email protected]
ABSTRACT
Objective: The primary aim of the study was to assess the sensitivity and specificity of intraoperative monitoring in predicting postoperative hearing loss during microvascular decompression (MVD).
Methods: The study was designed as an examination of the diagnostic accuracy of brainstem evoked potentials compared with reference standard for nonserviceable hearing loss (Class C/D), which is a change of more than 50 dB on pure tone threshold, and change of speech discrimination score of more than 50. All patients underwent surgery and audiograms at a University of Pittsburgh Medical Center facility in the study period 2005–2012. All participants received a pre- and postaudiogram within 90 days before or after the operation. During the operation, participants received intraoperative monitoring with a supervising physician. A total of 238 patients were selected. Brainstem auditory evoked potentials (BAEPs) were indexed into categories of change based on their maximum change and response at the end of surgery. Differences in hearing outcome by BAEP change were analyzed. Results: Age and sex did not affect outcomes. Patient outcome was affected by condition. The BAEP categories significant changes, transient loss, and persistent loss show a sensitivity/specificity of 0.905/0.701, 0.667/0.903, and 0.429/0.972, respectively. The receiver operating characteristic curve has an area under the curve of 0.870 with a 95% confidence interval of 0.783 to 0.957. Conclusions: Loss of wave V during MVD is a specific indicator of postoperative hearing loss. The current alarm criteria used to warn the surgeon is a sensitive indicator of impending postoperative hearing loss. Classification of evidence: This study provides Class IV evidence that in patients undergoing MVD, intraoperative BAEPs accurately identifies those who will have postoperative hearing loss. Neurology® 2014;83:1747–1752 GLOSSARY AAO-HNS 5 American Academy of Otolaryngology–Head and Neck Surgery; AUC 5 area under the curve; BAEP 5 brainstem auditory evoked potential; CI 5 confidence interval; CN 5 cranial nerve; GN 5 geniculate neuralgia; GPN 5 glossopharyngeal neuralgia; HFS 5 hemifacial spasm; HL 5 hearing loss; MVD 5 microvascular decompression; PTA 5 pure tone audiometry; ROC 5 receiver operating characteristic; SDS 5 speech discrimination score; TGN 5 trigeminal neuralgia.
Supplemental data at Neurology.org
Microvascular decompression (MVD) effectively treats for trigeminal neuralgia (TGN), hemifacial spasm (HFS), geniculate neuralgia (GN), and glossopharyngeal neuralgia (GPN) by relieving the vascular pressure on the cranial nerve (CN) at the root exit zone.1–3 Retraction at the cerebellopontine angle to decompress the CN is the most common mechanism of postoperative sensorineural hearing loss (HL).4 HL occurs at a rate of 2% to 20% in HFS, 1% to 23.8% in TGN, 1.5% to 4.2% in GPN, and 22% in GN.2,5–7 The wide variations in the incidence of HL could be secondary to the criteria used to define HL. We utilized the American Academy of Otolaryngology–Head and Neck Surgery (AAO-HNS) classification system8 to evaluate HL after MVD,8 where HL corresponding to a 50% decrease in speech discrimination scores (SDS) and/or a 50-dB increase in pure tone audiometry (PTA) scores provides a comprehensive From the Departments of Neurologic Surgery (P.D.T., M.E.H., D.J.C., J.R.B.) and Neurology (P.D.T.), University of Pittsburgh Medical Center; and Department of Neuroscience (G.C., J.R.B.), University of Pittsburgh, PA. Go to Neurology.org for full disclosures. Funding information and disclosures deemed relevant by the authors, if any, are provided at the end of the article. © 2014 American Academy of Neurology
1747
assessment of hearing function. Intraoperative monitoring of the brainstem auditory evoked potentials (BAEPs) of CN VIII has been shown to reduce postoperative CN VIII morbidity.9 To prevent HL, the surgeon is generally warned before the wave V latency increases to 0.5 or 1.0 millisecond, or amplitude decreases more than 50% compared with baseline values.10,11 This criterion is somewhat arbitrary and is not predictive of hearing dysfunction after MVD.10,11 Although the trend in changes in wave V during cerebellopontine angle surgery is significant in patients with HL,12 we clearly do not understand the predictive value of waveform V changes in HL. The primary aim of our study was to evaluate the diagnostic accuracy using sensitivity and specificity of changes in the latency and amplitude of wave V of BAEPs and to predict HL after MVD. METHODS Study population. A cross-sectional study was conducted from data collected in consecutive patients with TGN, HFS, GN, or GPN who underwent MVD at a University of Pittsburgh Medical Center facility between 2005 and 2012. All patients who underwent MVD received preoperative baseline and continuous intraoperative BAEP monitoring including preand postoperative audiometric examination. The data were collected for each patient after the procedure and an audiometric evaluation were completed. The data analysis was planned after the collection of the data. Patients without full intraoperative monitoring and/or pre- and postoperative audiometric data who underwent MVD were excluded from the study. The study evaluated the diagnostic accuracy of BAEPs to predict postoperative HL. The study provides Class IV evidence for utilizing BAEPs in patients undergoing MVD.
Standard protocol approvals, registrations, and patient consents. The study was approved by the University of Pittsburgh institutional review board for retrospective review of data on human subjects (MOD08120394-04/PRO08120394).
Audiologic investigations and HL criteria. We considered tone audiometry consisting of PTA scores and SDS to be the indicator of auditory function.13,14 Postoperative HL status was assessed using the 1995 AAO-HNS classification system.8 Class B referred to useful or serviceable HL and Class C/D was nonserviceable HL that was not amenable to hearing aids.8 We performed an otoneurologic examination on all patients preoperatively (median 1 day, range 1–49 days) and postoperatively (median 7 days, range 1–90 days). This consisted of an audiogram with measurement of pure tone thresholds (air and bone conduction for octave frequencies 250–8,000 Hz) as previously described.6 All audiograms were performed at the Eye and Ear Institute, University of Presbyterian Hospital, by an audiologist. Intraoperative monitoring. Recordings. All patients underwent intraoperative monitoring using BAEP recordings in the following manner. Independent alternate ear stimulation of at least 1748
Neurology 83
November 4, 2014
17.5 Hz was performed in the right and left side with at least 85 dB nHL (normal HL), using alternating rarefaction and condensation clicks. The contralateral ear was stimulated with white noise at 65 dB nHL. We observed the data for 12 milliseconds, and averaging at least 256 responses. BAEP recording electrodes were positioned as follows: (channel 1) vertex to left ear mastoid Cz/A1; (channel 2) vertex to right ear mastoid Cz/A2; and (channel 3) vertex to cervical C2 (Cz/Cv2). Amplifier bandpass was 100 to 1,000 Hz for all channels. After induction of anesthesia and positioning the patient, the baseline responses were established. Continuous real-time BAEPs were obtained from the operative side in all patients. The contralateral ear BAEP responses were recorded at baseline. Analysis of neurophysiologic parameters. The absolute latencies of waves I, III, and V (LwI, LwIII, and LwV, respectively) of the BAEPs were identified based on previously described methods.15 The amplitudes of waves III and V (AwIII and AwV, respectively) were measured with exclusion for wave I (AwI), which could be masked by stimulation artifact. Wave V of BAEP is the main focus of the analysis, because it is the most robust waveform that can be recorded consistently during MVD. The BAEP waveforms were analyzed at the following times during the procedure: (1) baseline recording after the patient was intubated and positioned; (2) change START, defined when a first persistent decrease in amplitude of wave V (AwV) .50% and/or increase in LwV .0.5 millisecond compared with baseline was observed; (3) change MAX, defined as the time when a maximum change in the AwV and/or LwV compared with baseline was observed (this included instances when no waveforms were identifiable [loss of response]); and (4) a final recording at the time of skin closure. If loss of response occurred, amplitude was recorded as zero and latency was recorded as 12 milliseconds, which is the maximum observation interval. Physician oversight and interpretation were performed using a combined on-site and remote model utilized by the University of Pittsburgh Medical Center.16 The BAEP waveforms were identified and evaluated independently by research students, who were supervised by the corresponding author.
Group discrimination. Predictive value including sensitivity and specificity of changes in BAEPs to identify HL was calculated after categorizing patients into no change, significant change, transient loss, and persistent loss based on changes in BAEPs during MVD. BAEP waveform changes in latency and amplitude are dynamic during MVD, secondary to the degree of retraction and/or compression affecting the auditory nerve or its vasculature. Hence, we thought that categorizing the changes that incorporated the current alarm criteria may be helpful in understanding the value of BAEPs during MVD. Loss of amplitude of 100% and/or prolongation in wave V latency greater than 12 milliseconds was regarded as loss of waveforms. The loss of wave V of BAEP was further subdivided into persistent loss and transient loss based on whether the waveforms improved after the loss during decompression. Transient loss was regarded as wave V peaks that recovered to baseline in amplitude and latency before the end of surgery. Persistent loss BAEPs did not recover before the end of surgery. Patients who had consistent decrease in amplitude of more than 50% of wave V and/or persistent increase in the absolute latency of the peak of wave V that was $0.5 milliseconds and did not have loss of waveform during MVD were reported as significant change.11,17 “Consistent changes” represent 2 consecutive epochs of latency and or amplitude changes in BAEPs compared with baseline responses in more than 2 consecutive epochs. Two consecutive epochs were used to eliminate technical issues as the cause of waveform change, such as electrical interference from
Figure 1
Patient selection
Patient selection methods with index test, brainstem auditory evoked potential category, reference standard, and ontological examination are shown. BSER 5 brainstem evoked response; MVD 5 microvascular decompression; UPMC 5 University of Pittsburgh Medical Center.
cauterizing or drilling. BAEPs in patients whose wave V changes did not meet these criteria were categorized as no change.
Statistical methods. Mean and SD of age, percentage of female patients, and percentage incidence of HL were calculated for all MVD procedures and by condition. A distribution of PTA scores and SDS was given for all patients, by HL category and by condition displayed in a box plot. Previous MVD studies have reported that a significantly higher number of patients who had permanent loss of wave V had new, postoperative HL.12 To compare our data with previous studies on the degree of change of
Table 1
Descriptive statistics All patients
HFS
TGN
GPN
GN
No. of patients
238
123
100
8
7
Age, y, mean (SD)
52.41 (12.61)
52.29 (11.98)
53.04 (13.68)
48.25 (8.84)
50.28 (12.2)
Female, n (%)
162 (68)
76 (62)
73 (73)
7 (87.5)
5 (71)
HL, n (% of category)
21 (8.8)
18 (15)
3 (2)
0
0
Abbreviations: GN 5 geniculate neuralgia; GPN 5 glossopharyngeal neuralgia; HFS 5 hemifacial spasm; HL 5 hearing loss; TGN 5 trigeminal neuralgia. Patients are separated by condition. HL corresponds to postoperative hearing C/D.
wave V, a test of group discrimination was performed. To account for a small sample, a Fisher exact test was performed, which is nonbiased for sample size. The test determines whether at least one of the categories is significantly different from the others. We calculated the sensitivity and specificity of BAEP categories using the reference test (otoneurologic examination using the 1995 AAO-HNS classification system). BAEP categories include no change, significant change, transient loss, and persistent loss. A true-positive result was regarded as BAEPs with the degree of change tested or more severe and experiencing HL C/D. A false positive was regarded as BAEPs with the degree of change tested or more severe and experiencing HL A/B. A false negative was BAEPs with less severe change than the degree of change tested with HL C/D. A true negative was BAEPs with less severe change than the degree of change tested with HL A/B. A 95% confidence interval (CI) was calculated for all sensitivities and specificities. To further evaluate the diagnostic accuracy of change in BAEPs and postoperative HL, we used a receiver operating characteristic (ROC) curve with BAEP categories used as cutoff values. ROC was chosen because the area under the ROC curve (AUC) has been used as a measure of prediction accuracy for a variety of biomedical applications. ROC AUC, significance, and 95% CI were calculated. Standard error was calculated using a nonparametric distribution. The significance for all tests was set at p , 0.05. All statistics were completed using SPSS version 20 (IBM Corp., Armonk, NY). Neurology 83
November 4, 2014
1749
Table 2
Number of patients with changes and hearing loss BAEP change No significant change
Significant change
Loss of waveforms
Transient loss
Persistent loss
Total
No. of patients
144
49
35
20
15
238
Patients with hearing loss
2
5
14
5
9
21
Sensitivity (95% CI)
0.905 (0.696–0.986)
0.667 (0.430–0.854)
0.667 (0.430–0.854)
0.429 (0.217–0.660)
Specificity (95% CI)
0.701 (0.635–0.761)
0.903 (0.856–0.939)
0.903 (0.856–0.939)
0.972 (0.941–0.990)
Positive predictive value (95% CI)
0.184 (0.117–0.276)
0.4 (0.243–0.578)
0.4 (0.243–0.578)
0.6 (0.328–0.825)
Negative predictive value (95% CI)
0.986 (0.946–0.997)
0.965 (0.927–0.984)
0.965 (0.927–0.984)
0.946 (0.905–0.970)
Abbreviations: BAEP 5 brainstem auditory evoked potential; CI 5 confidence interval. Sensitivity, specificity, and positive and negative predictive value for hearing loss are compared by BAEP category.
RESULTS Participants. Figure 1 displays patient selection. A total of 238 patients who underwent MVD at the University of Pittsburgh Medical Center during the study period (figure 1) were eligible to participate. Descriptive statistics for number of patients with each condition, age with SD, sex, and HL are provided in table 1.
BAEP and ontological results. The probability of HL
when there was no significant change (no change in wave V) was 1.3%, compared with the probability of HL of 10.2% of patients when a significant change in the BAEP was reported, 25% of patients with
Figure 2
ROC curve
Receiver operating characteristic (ROC) curve to evaluate the changes in brainstem auditory evoked potential and its relationship with hearing loss. Points on the ROC curve correspond to specificities and sensitivities given in table 2. Note that diagonal segments are produced by ties. 1750
Neurology 83
November 4, 2014
transient loss, and 60% of patients with persistent loss. Box plots display the distribution of SDS and PTA scores by condition (see figures e-1 and e-2 on the Neurology® Web site at Neurology.org). Results of categorical discrimination. The Fisher exact
statistic had a value of 46.319, with a statistical significance ,0.001 to significantly discriminate various BAEP changes and the incidence of HL. Table 2 displays specificities and sensitivities of various categories of changes in BAEPs. The cutoff for significant change was sensitivity of 90.5% (69.6%–98.6%) and specificity of 70.1% (63.5%–76.1%) (table 2). The cutoff for transient loss was sensitivity of 66.7% (43.0%–85.4%) and specificity of 90.3% (85.6%–93.9%) (table 2). The cutoff for persistent loss was a sensitivity of 42.9% (21.7%–66.0%) and specificity of 97.2% (94.1%–99.0%) (table 2). The ROC curve has an AUC of 0.870, significance ,0.001, and 95% CI of 0.783–0.957, as displayed in figure 2. The cutoff points in the ROC curve correspond to the sensitivity and specificity of index categories, visually defined by figure 3, where a more severe change in BAEP corresponds to a more positive result. DISCUSSION Our study results indicate that loss of wave V during MVD is a highly specific predictor of postoperative HL. This was seen in patients with transient as well as persistent loss of wave V responses. The primary mechanism for CN VIII injury during MVD has been identified as nerve stretching or compression during cerebellar retraction, but can also be caused by damage to CN vasculature providing blood flow to the cochlear nerve or middle ear.18 During MVD, manipulation of the auditory nerve can result in neurapraxic nerve injury and significant changes in waves II and III of the BAEP. However, AwII and AwIII of the BAEP are typically small during intraoperative recording and, hence, are often difficult to reliably record in the operating room during MVD. Change in wave V reflects damage to ascending converging afferents whose synaptic
Figure 3
Highlights the various changes of BAEPs including the significant changes, transient, and complete loss of responses during microvascular decompression
Timeline is shown on the time bar to highlight its relationship with the procedure. BAEP 5 brainstem auditory evoked potential.
integration causes a smaller postsynaptic potential due to a reduction in the overall strength of inputs and their synchrony, which cumulates over several synapses. Even though the MVD-related nerve injury affects the proximal segment of the auditory pathway, monitoring wave V during MVD surgery is a robust and useful measure of conduction along the brainstem auditory pathway. A significant loss of AwV could be correlated to mechanical deformation or ischemia degree of severity applied to CN VIII.18,19 Increase in pressure or lack of blood flow leads to loss of amplitude and is accompanied by histologic changes of nerve injury.18,19 However, improvement in the amplitude of a nerve response after complete loss as seen in BAEPs during MVD can occur without permanent injury to the nerve.19 Most frequently this can occur when the stretch on CN VIII is transient or removed in a timely manner. Hence, loss of wave V response of BAEPs might be a proxy of imminent significant histologic change, and conditions leading to this state during MVD should be avoided. To devise alarm criteria in order to predict and prevent HL during MVD, a sensitive predictor of impending HL is needed. The currently adopted arbitrary “significant changes” include a decrease in amplitude of .50% of wave V and/or persistent
increase in the absolute latency of the peak of wave V that is $1 millisecond.11,20 In our study, significant changes in patients noted during MVD were highly sensitive for identifying HL. We believe this is the first study to evaluate the sensitivity of alarm criteria in intraoperative neurophysiologic monitoring involving BAEPs to evaluate and predict impending HL. Our study revealed 2 patients without significant changes in wave V of BAEPs during MVD who experienced postoperative HL suggesting mechanisms of underlying HL unrelated to CN VIII apraxia and/or injury. Primate studies on the effect of focal brainstem ischemia show that a decrease in brainstem blood flow increases the latency of BAEP waveforms.21 Sensorineural HL may occur as a result of factors unrelated to a conduction block of the auditory nerve or cochlear damage, such as damage to higher order auditory tracts or postoperative brainstem ischemia that does not manifest during the course of surgery.22 In the absence of any signs of clinical brainstem ischemia, or significant change or loss of BAEPs during MVD, it is possible that damage to the cochlea may occur in the immediate postoperative phase. Although this study represents the largest crosssectional diagnostic review of BAEP wave V changes correlated to postoperative HL after MVD, it is not without Neurology 83
November 4, 2014
1751
limitations. The study does not represent a random sample of all MVD procedures because some patients elected to undergo postoperative audiograms only if significant hearing problems were identified postoperatively. Patients who did not have audiometric testing probably had no postoperative hearing change, which most likely overestimates the incidence of HL after MVD. Loss of wave V during MVD is a specific indicator of postoperative HL. The current alarm criteria used to warn the surgeon is a sensitive indicator of impending postoperative HL. HL infrequently occurs when there is no change in BAEPs during MVD. Although stretch-induced injury is common in patients undergoing MVD, other mechanisms of injury need to be considered and evaluated during the procedure.
7.
8.
9.
10.
11. AUTHOR CONTRIBUTIONS Parthasarathy Thirumala: drafting/revising the manuscript, study concept or design, analysis or interpretation of data, accepts responsibility for conduct of research and will give final approval, statistical analysis. Gregory Carnovale: drafting/revising the manuscript, study concept or design, analysis or interpretation of data, accepts responsibility for conduct of research and will give final approval, acquisition of data, statistical analysis, study supervision. Miguel Haybeych: drafting/revising the manuscript, study concept or design, accepts responsibility for conduct of research and will give final approval, acquisition of data, study supervision. Donald Crammond: drafting/revising the manuscript, study concept or design, analysis or interpretation of data, accepts responsibility for conduct of research and will give final approval, acquisition of data, study supervision. Jeffrey Balzer: drafting/revising the manuscript, study concept or design, analysis or interpretation of data, accepts responsibility for conduct of research and will give final approval, acquisition of data.
12.
13.
14.
15. STUDY FUNDING No targeted funding reported.
16. DISCLOSURE The authors report no disclosures relevant to the manuscript. Go to Neurology.org for full disclosures.
Received March 19, 2014. Accepted in final form August 4, 2014. REFERENCES 1. Acevedo JC, Sindou M, Fischer C, Vial C. Microvascular decompression for the treatment of hemifacial spasm: retrospective study of a consecutive series of 75 operated patients—electrophysiologic and anatomical surgical analysis. Stereotact Funct Neurosurg 1997;68:260–265. 2. Patel A, Kassam A, Horowitz M, Chang YF. Microvascular decompression in the management of glossopharyngeal neuralgia: analysis of 217 cases. Neurosurgery 2002;50:705–710. 3. Ramnarayan R, Mackenzie I. Brain-stem auditory evoked responses during microvascular decompression for trigeminal neuralgia: predicting post-operative hearing loss. Neurol India 2006;54:250–254. 4. Little JR, Lesser RP, Lueders H, Furlan AJ. Brain stem auditory evoked potentials in posterior circulation surgery. Neurosurgery 1983;12:496–502. 5. Rizvi SS, Goyal RN, Calder HB. Hearing preservation in microvascular decompression for trigeminal neuralgia. Laryngoscope 1999;109:591–594. 6. Shah A, Nikonow T, Thirumala P, et al. Hearing outcomes following microvascular decompression for hemifacial spasm. Clin Neurol Neurosurg 2012;114:673–677.
1752
Neurology 83
November 4, 2014
17.
18.
19.
20.
21.
22.
Rupa V, Saunders RL, Weider DJ. Geniculate neuralgia: the surgical management of primary otalgia. J Neurosurg 1991;75:505–511. Committee on Hearing and Equilibrium guidelines for the evaluation of hearing preservation in acoustic neuroma (vestibular schwannoma). American Academy of Otolaryngology– Head and Neck Surgery Foundation, INC. Otolaryngol Head Neck Surg 1995;113:179–180. Wilkins RH, Radtke RA, Erwin CW. Value of intraoperative brainstem auditory evoked potential monitoring in reducing the auditory morbidity associated with microvascular decompression of cranial nerves. Skull Base Surg 1991;1:106–109. American Clinical Neurophysiology Society. Guideline 11: Guidelines for Neurophysiologic Intraoperative Monitoring. Milwaukee: American Clinical Neurophysiology Society; 2009. Available at: http://www.acns.org/practice/guidelines. Martin WH, Stecker MM. ASNM position statement: intraoperative monitoring of auditory evoked potentials. J Clin Monit Comput 2008;22:75–85. James ML, Husain AM. Brainstem auditory evoked potential monitoring: when is change in wave V significant? Neurology 2005;65:1551–1555. Sindou MP. Microvascular decompression for primary hemifacial spasm: importance of intraoperative neurophysiological monitoring. Acta Neurochir 2005;147:1019–1026. Polo G, Fischer C, Sindou MP, Marneffe V. Brainstem auditory evoked potential monitoring during microvascular decompression for hemifacial spasm: intraoperative brainstem auditory evoked potential changes and warning values to prevent hearing loss—prospective study in a consecutive series of 84 patients. Neurosurgery 2004;54:97–104. Chiappa KH. Brainstem Auditory Evoked Potentials: Methodology, 3rd ed. New York: Lippincott-Raven Publishers; 1997. Thirumala PD, Kassasm AB, Habeych M, et al. Somatosensory evoked potential monitoring during endoscopic endonasal approach to skull base surgery: analysis of observed changes. Neurosurgery 2011;69:ons64–ons76. Committee on Hearing and Equilibrium guidelines for the evaluation of results of treatment of conductive hearing loss. American Academy of Otolaryngology–Head and Neck Surgery Foundation, Inc. Otolaryngol Head Neck Surg 1995;113:186–187. Ogata K, Naito M. Blood flow of peripheral nerve effects of dissection, stretching and compression. J Hand Surg Br 1986;11:10–14. Wall EJ, Massie JB, Kwan MK, Rydevik BL, Myers RR, Garfin SR. Experimental stretch neuropathy: changes in nerve conduction under tension. J Bone Joint Surg Br 1992;74:126–129. American Clinical Neurophysiology Society. Guideline 11C: Recommended Standards for Intraoperative Monitoring of Auditory Evoked Potentials. Milwaukee: American Clinical Neurophysiology Society; 2009. Available at: http://www.acns.org/practice/guidelines. Baik MW, Branston NM, Bentivoglio P, Symon L. The effects of experimental brain-stem ischaemia on brain-stem auditory evoked potentials in primates. Electroencephalogr Clin Neurophysiol 1990;75:433–443. Lin CD, Wei IH, Tsai MH, et al. Changes in guinea pig cochlea after transient cochlear ischemia. Neuroreport 2010;21:968–975.
Diagnostic accuracy of brainstem auditory evoked potentials during microvascular decompression Parthasarathy D. Thirumala, Gregory Carnovale, Miguel E. Habeych, et al. Neurology 2014;83;1747-1752 Published Online before print October 8, 2014 DOI 10.1212/WNL.0000000000000961 This information is current as of October 8, 2014 Updated Information & Services
including high resolution figures, can be found at: http://www.neurology.org/content/83/19/1747.full.html
Supplementary Material
Supplementary material can be found at: http://www.neurology.org/content/suppl/2014/10/08/WNL.0000000000 000961.DC1.html
References
This article cites 19 articles, 4 of which you can access for free at: http://www.neurology.org/content/83/19/1747.full.html##ref-list-1
Subspecialty Collections
This article, along with others on similar topics, appears in the following collection(s): Evoked Potentials/Auditory http://www.neurology.org//cgi/collection/evoked_potentials-auditory Trigeminal neuralgia http://www.neurology.org//cgi/collection/trigeminal_neuralgia
Permissions & Licensing
Information about reproducing this article in parts (figures,tables) or in its entirety can be found online at: http://www.neurology.org/misc/about.xhtml#permissions
Reprints
Information about ordering reprints can be found online: http://www.neurology.org/misc/addir.xhtml#reprintsus
Neurology ® is the official journal of the American Academy of Neurology. Published continuously since 1951, it is now a weekly with 48 issues per year. Copyright © 2014 American Academy of Neurology. All rights reserved. Print ISSN: 0028-3878. Online ISSN: 1526-632X.
