Intraoperative recurrent laryngeal nerve monitoring DALE H. RICE. MD. and BARBARA CONE-WESSON. PhD. Los Angeles, California
Intraoperative nerve monitoring has become common for surgical procedures in which cranial or peripheral nerves may be compromised. Intraoperative monitoring of reo current laryngeal nerve function can be accomplished by recording electromy· ographlc activity from fine-wire electrodes placed In the vocalls muscle. The technique and Instrumentation for this are adapted from those used In Intraoperative facial nerve monItorIng for acoustIc neuroma excisIon. Specifically. a commercially available In· strument. the XOMED·NIM provides the capability for monitoring the vocalls muscle electromyogram by means of visual and auditory display. It also provides the cepeblllty for performing evoked electromyographlc tests of nerve Integrity. Intraoperative monitoring of the recurrent laryngeal nerve during thyroidectomy may assist In the more precise dissection of the nerve as well as In verification of nerve Integrity during the operative procedure. thus reducing the risk of Injury. (OTOLARYNGOL HEAD NECK SURG 1991;105:372.)
BACKGROUND Intraoperative nerve monitoring has become common for surgical procedures during which cranial!') or peripheral" nerves may be compromised. We have been impressed by the published reports regarding the intraoperative monitoring of facial nerve integrity."!" We believed that the methods used for facial nerve monitoring could also be applied to the recurrent laryngeal nerve (RLN). In particular, we wished to monitor the recurrent laryngeal nerve during thyroidectomy, a surgical procedure for which the incidence of nerve compromise may reach 13%.11 The patient with thyroid disease is at risk for intraoperative laryngeal nerve injury because of several factors, including the variable anatomy of the recurrent and superior laryngeal nerves, and changes in the normal anatomy resulting from extensive disease, recurrent disease, or radiation therapy. We wished to determine if intraoperative RLN monitoring (IRM) could be used to guide identification of the nerve fibers from other confounding tissues, as well as provide information about nerve integrity during and after surgical maneuvers. From the Department of Otolaryngology-Head and Neck Surgery, University of Southern California Medical School, Los Angeles. Received for publication March 16. 1990; revision received March 7, 1991; accepted March 21, 1991. Reprint requests: Dale H. Rice, MD, Department of OtolaryngologyHead and Neck Surgery, Los Angeles County and University of Southern California Medical Center, Box 795, 1200 North State SI., Los Angeles. CA 90033. 23/1129787
372
Several methods for IRM have been explored in the past, including palpation" or recording of laryngeal muscle movement during RLN electrical stimulation! 3; also, monitoring glottic pressure change as a result of RLN stimulation." Davis et al." recorded electromyagraphic (EMG) activity from laryngeal muscles during surgery, and Lipton et a1.16 and Daube and Harner described how monitoring the spontaneous and evoked laryngeal muscle EMG could be used to monitor RLN integrity during thyroid surgery. Our rationale and procedure for IRM were based on those used for intraoperative monitoring of facial nerve function. The first procedure involves recording the ongoing EMG from musculature innervated by a cranial or peripheral nerve. The EMG can be "played back" using an acoustic monitor from the EMG amplifier to give the surgeon a real-time acoustic signal related to nerve function." The acoustic signal generated by the EMG can indicate baseline or spontaneous levels of muscle activity and will also signal bursts or trains" of EMG activity that may be interpreted in the context of surgical events. That is, changes in the background EMG or prolonged trains or bursts may indicated nerve compromise and warn the surgeon to modify or refrain from surgical manipulation of the nerve." A second method involves evoking EMG activity on the compound motor action potential (CMAP), by presenting an electrical stumulus pulse directly to the nerve by means of a monopolar or bipolar":" probe. The evoked EMG activity can be used to distinguish nerve from
Volume 105 Number 3 September 1991
invading tumor or other tissue, guide the anatomic dissection," and provide a record of neural stimulability throughout the operation. The evidence of stimulability may prove to be prognostic of functional outcome. K A dedicated instrument for nerve monitoring is now commercially available, the XOMED-Treace Nerve Integrity Monitor (NIM, Xomed-Treace, Jacksonville, Fla.). The instrument can be used to monitor ongoing EMG over continuously updated sampling epochs. The NIM provides separate acoustic feedback signals for each of two recording channels. An integrated pulse. generator provides a constant-current or constant-voltage stimulus pulse for evoking EMG activity. Another advantage of the instrument is that the recording epoch can be delayed after presentajtion of the electrical signal for evoked EMG. This reduces the amount of stimulus artifact in the resultant recording. Our experience using this instrument for IRM during thyroidectomy is described in the Methods section. We wish to provide some useful procedures for undertaking IRM, procedures we discovered by trial and error. Certain factors that make IRM somewhat less advantageous than intraoperative facial nerve monitoring will also be described. Our experiences have raised questions about the underlying stimulus-response characteristics of the RLN and the muscles it innervates, which will be discussed subsequently.
ME11tODS Patients
We have used IRM for five patients undergoing thyroidectomies for differentiated carcinoma. All patients had complete otolaryngologic evaluation before surgery. All had clinically normal vocal cord function before and after surgery. Anesthesia
The use of neuromuscular blocking agents must be avoided because these may alter or suppress the neuromuscular events being monitored. It must be recognized, however, that succinyl-choline is used routinely to allow nontraumatic intubation. The anesthesiologist must be informed about the monitoring procedures so that judicious decisions can be made about the type and quantity of anesthetic agents to be administered. Recording Electrodes-Preparation and Placement
Preoperative preparation of several sets of recording electrodes is advised. EMG paired hook-wire electrodes are used for the recordings. The electrodes are compOsed of insulated stainless steel wires, with the terminal 3 to 5 mm stripped of insulation. The ends of
Intraoperative recurrent laryngeal nerve monitoring
373
the wires are bent 180 degrees to aid in insertion and retention of the electrode in muscle tissue. Each pair should be threaded through a 20-cm 22-gauge CHlBA type aspiration needle with the hooked ends advanced just beyond the needle tip. The electrodes may be gas sterilized after they are positioned in the aspiration needle. One pair of electrodes is placed in each thyroarytenoid muscle in the mid-body of the vocal fold, to a depth of approximately 6 to 8 mm. The electrodes are placed under direct laryngoscopic examination. We have found that electrode placement is facilitated by use of the aspiration needle, in comparison to attempts with a long-handled instrument passed through the laryngoscope. Once the electrodes are placed they should be taped to the face to avoid displacement. They are connected to the patient-recorder interface cable using 0.75 mm insulated electrodes that have an alligator clip and 2-mm pin connector at either end. The use of further extension cables for the electrodes is discouraged because this can lead to excessive artifact in the recording. The alligator clip electrodes and their leads should also be taped to the operative table underneath the drapes. This, too, will help guard against electrode disconnection. A standard EEG surface electrode is used for a ground, and is placed on the forehead. The site of placement is prepared by light scrubbing with an abrasive cleaner. A liberal amount of electrode conductive paste is used and the electrode is held in place with surgical tape. When a monopolar stimulus probe is used, an anode electrode must be placed at the back of the shoulder. The anode electrode and its site are prepared in the same fashion as the ground electrode. It should be noted that we tried various alternative locations for anode placement, including sites closer to and farther from the cathode (i.e., the site at which the monopolar probe is applied). The shoulder site placement yielded more reliable recordings of evoked CMAP activity in comparison to the alternative placements. Electrode impedances should be checked after placement and monitored throughout the procedure to assure electrode integrity. Electrode impedances for the intramuscular electrodes may be quite high, between 40 and 50 kO, but they tend to fall to 15 and 25 kO during the dissection. Recording Parameters
Spontaneous EMG activity varies during the operative procedure, requiring ongoing adjustment of the monitor's sensitivity. The range of baseline activity we
374
OlolaryngologyHead and Neck Surgery
RICE and CONE-WESSON
observed was between 40 and 160 j.L V. During the initial stages of dissection for thyroidectomy, we monitor baseline activity at IS-minute intervals and observe the ongoing EMG levels for 2 minutes, using the 60-msec sampling epoch at each interval. We use these levels to determine the "event threshold" level; that is, the EMG level at which an acoustic alarm will be sounded. With the XOMED-NlM, the adjustment of the "event threshold" sensitivity is crucial, because the acoustic alarm signaling increased EMG activity is triggered by EMG events exceeding this level. We adjust the event threshold level to twice the value of the baseline EMG. When the dissection has proceeded to the stage at which the RLN will soon be exposed or manipulated, then the monitoring is constant. We used an ongoing 60-msec sampling epoch for this stage of the procedure. With the event threshold trigger levels appropriately adjusted, any undue manipulation of RLN will incite activity of the vocalis muscle, and the increased EMG will trigger the acoustic alarm, which alerts the surgeon. This can be useful in the identification of the nerve and for the dissection of the nerve- particularly in patients with significant disease in the vicinity of RLN. "Bursts" or "trains" of EMG activity from facial musculature have been reported during facial nerve dissection or manipulation." The bursts are associated with direct manipulation or fluid irrigation in the vicinity of the nerve or electrocautery." Trains can be associated with nerve traction, but may also be present in the absence of surgical provocation. 17 We found train and burst activity evident in the EMG response of laryngeal musculature during surgical manipulation. The diagnostic and prognostic significance of this activity has not yet been determined.
Evoked EMG Nerve integrity can also be determined by direct electrical stimulation of the recurrent laryngeal nerve and observation of the resultant increased EMG activity. Constant-current or constant-voltage stimuli may be used'? with the stimulus presented by means of an insulated flush-tip monopolar (Prass Flush-Tip Probe, Xorned-Treace) or bipolar (Kartush Facial Stimulator, EI Med, Addison, III.) probe. We have been able to evoke activity more consistently and more discretely with the bipolar probe; that is, use of the monopolar probe resulted in more false-positive responses in comparison to the bipolar probe. The bipolar stimulus probe confines the stimulus to a small area and thereby improves the spatial resolution so that neural and nonneural tissue may be distinguished. The stimulus pulse is brief-lOa j.Lsec-and is presented at a rate of four per second. We have observed
variability in the stimulus level needed to evoke a response. With constant voltage stimuli, levels of 0.88 to 1.2 volts elicit a synchronous response. For constantcurrent stimuli, levels of 0.2 to 0.5 rnA are sufficient. The latency of the response appears to be brief, and it is crucial that stimulus artifact be recognized and reduced without compromising the recorded EMG event. With the NIM system used in this investigation, this is achieved by setting a I.S-msec delay between the stimulus and onset of recording. Even this may be too long when working close to the cricothyroid joint. We were able to evoke an EMG response from each patient during the dissection and also after the thyroid tissue had been removed. The RLN was stimulated at 1em or more below Berry's ligament. The evoked EMG response amplitude was large-at least three or four times greater than baseline activity. The stimulus appeared to produce an all-or-nothing response, but we were unable to derive stimulus vs. response amplitude functions with the intraoperative monitoring protocol. The evoked EMG response was evident from the acoustic signal produced by the EMG amplifier, which is synchronous with the stimulus pulse (i.e., acoustic signals produced at four times per second). The response can also be observed on the display screen of the monitor as increased bursts of EMG activity coincident with the stimulus pulses. The stimulability of the nerve indicates that the nerve fibers innervating the muscle are intact. It appears that the close anatomic relationship between the stimulus probe and the recording electrodes, as well as a short response latency, make isolation of the electrical signal difficult. The use of a bipolar stimulus probe, with the stimulus confined between the cathode and anode, yielded better definition of neural from non-neural tissue. We tested for "false-positive" stimulus artifacts by comparing amplifier output when neural tissue was stimulated to that when adjacent non-neural tissue was stimulated. When artifact was present, touching nonneural tissue with the probe would result in synchronous output. For each stimulus level used we performed a control test for non-neural tissue. Responses were judged as present only when control trials were free of artifact. The bipolar probe produced fewer false-positive responses in comparison to the monopolar probe. The bipolar stimulus probe confined the stimulus to the small area between the anode and cathode, resulting in improved spatial resolution for distinguishing neural vs. non-neural tissue.
DISCUSSION This description of our experience is intended to provide the surgeon with procedures that are of immediate
Volume 105 Number 3 September 1991
use when IRM is attempted. In our community, facial nerve integrity monitoring has become the standard of practice for neuro-otologic surgery. It is not our contention, however, that IRM is needed for every thyroidectomy, because the morbidity of the RLN, in the main, is low. II Also, the procedures and dedicated instruments for facial nerve monitoring may need to be modified for IRM to be performed. Research on the threshold, latency, and amplitude characteristics of the evoked EMG is needed to define optimum IRM parameters. Safe levels of constant-current or constant-voltage stimuli need to be defined for this branch of the vagus nerve. The dynamics of the neurotonic response for the RLN are probably quite different from those for the facial nerve, and recording parameters must be modified to address those differences. The effect of anesthetic agents on the RLN and the muscles it innervates must also be investigated. Most laryngeal surgeons are well aware that neuromuscular blocking agents will often leave laryngeal closure reflexes intact. Systematic observation of EMG bursts and trains in the surgical context is needed. Whether these are spontaneous events or are evoked by surgical maneuvers has not been well defined. Their use as diagnostic or prognostic indicators must be evaluated. We look forward to undertaking these investigations and will report our results as they are completed. REFERENCES I. Grundy BL. Electrophysiologic monitoring: Electroencephalography and evoked potentials. In: Newfield P, Cottrel JE, eds. Handbook of neuroanesthesia: clinical and physiologic essentials. Boston: Little Brown and Co., 1983:28-59. 2. Levine RA, Ojemann RG, Montgomery WW, McGaffigan PM. Monitoring auditory evoked potentials during acoustic neuroma surgery: insights into the mechanism of hearing loss. Ann Otol Rhinol Laryngol 1984;93:116-23. 3. Nuwer MR. Evoked potential monitoring in the operating room. New York: Raven Press, 1986 . 4. Nelson KR. Use of peripheral nerve action potentials for intraoperative monitoring. Neurol Clin 1988;6:917-33.
Intraoperative recurrent laryngeal nerve monitoring 375
5. Prass RL. Acoustic EMG monitoring: a prototype of intraoperative cranial nerve monitoring. Short course presented at the American Academy of Neurology, New York: 1987. 6. Prass RL, Luders H. Acoustic (loudspeaker) facial electromyographic monitoring. Part I: Evoked electromyographic activity during acoustic neuroma resection. Neurosurgery 1986; 19:392400. 7. Prass RL, Kinney SE, Hardy RW, Hahn JF, Luders H. Acoustic (loudspeaker) facial EMG monitoring. Part II: Use of evoked EMG activity during acoustic neuroma resection. OTOl.ARYNGOL HEAD NECK SURG 1987;97:541-51. 8. Harner SG, Daube JR, Ebersol MJ, Beatty CWo Improved preservation of facial nerve function with use of electrical monitoring during removal of acoustic neuromas. Mayo Clin Proc 1987;62:92-102. 9. Kartush JM. Electroneurographyand intraoperative facial monitoring in contemporary ncurotology. Paper presented at the American Neurotology Society, Palm Beach, Fla., 1988. 10. Benecke JR Jr, Caler HB, Chadwick G. Facial nerve monitoring during acoustic neuroma removal. Laryngoscope 1987;97:697700. II. Van Heerden JV, Groh MA, Grant ES. Early postoperative morbidity after surgical treatment of thyroid carcinoma. Surgery 1987;101:224-7. 12. James AG, Crocker S, Woltering E, et al. A simple method for identifying and testing recurrent laryngeal nerve. Surg Gynecol Obstet 1985;161:186-7. 13. Spahn JG, Bizal J, Ferguson S, et al. Identification of the motor laryngeal nerves-a new electrical stimulation technique. Laryngoscope 1981;91:1865-8. 14. Hvidegaard T, Vase P, Dalsgaard SC, et al. Endolaryngeal devices for perioperative identification and functional testing of the recurrent laryngeal nerve. OTOLARYNGOL HEAD NECK SURG 1984;92:292-4. 15. Davis WE. Rea JL, Templer J. Recurrent laryngeal nerve localization using a microlaryngeal electrode. OTOLARYNGOL HEAD NECK SURG 1979;87:330-3. 16. Lipton RJ, McCaffrey TV, Litchy WJ. Intraoperative electrophysiologic monitoring of laryngeal musele during thyroid surgery. Laryngoscope 1988;98: 1292-6. 17. Niparko JK, Kileny PRo Intraoperative monitoring offacial function. Semin Hearing 1988;9:127-39. 18. Kartush JM, Niparko JK, Bledsoe SC, Graham MD, Kaminke JL. Intraoperative facial nerve monitoring: a comparison of stimulating electrodes. Laryngoscope 1985;95: 1536-40. 19. Prass RL, Luders H. Constant-current versus constant-voltage stimulation. J Neurosurg 1985;62:622-3.
Intraoperative nerve monitoring has become common for surgical procedures in which cranial or peripheral nerves may be compromised. Intraoperative monitoring of reo current laryngeal nerve function can be accomplished by recording electromy· ographlc activity from fine-wire electrodes placed In the vocalls muscle. The technique and Instrumentation for this are adapted from those used In Intraoperative facial nerve monItorIng for acoustIc neuroma excisIon. Specifically. a commercially available In· strument. the XOMED·NIM provides the capability for monitoring the vocalls muscle electromyogram by means of visual and auditory display. It also provides the cepeblllty for performing evoked electromyographlc tests of nerve Integrity. Intraoperative monitoring of the recurrent laryngeal nerve during thyroidectomy may assist In the more precise dissection of the nerve as well as In verification of nerve Integrity during the operative procedure. thus reducing the risk of Injury. (OTOLARYNGOL HEAD NECK SURG 1991;105:372.)
BACKGROUND Intraoperative nerve monitoring has become common for surgical procedures during which cranial!') or peripheral" nerves may be compromised. We have been impressed by the published reports regarding the intraoperative monitoring of facial nerve integrity."!" We believed that the methods used for facial nerve monitoring could also be applied to the recurrent laryngeal nerve (RLN). In particular, we wished to monitor the recurrent laryngeal nerve during thyroidectomy, a surgical procedure for which the incidence of nerve compromise may reach 13%.11 The patient with thyroid disease is at risk for intraoperative laryngeal nerve injury because of several factors, including the variable anatomy of the recurrent and superior laryngeal nerves, and changes in the normal anatomy resulting from extensive disease, recurrent disease, or radiation therapy. We wished to determine if intraoperative RLN monitoring (IRM) could be used to guide identification of the nerve fibers from other confounding tissues, as well as provide information about nerve integrity during and after surgical maneuvers. From the Department of Otolaryngology-Head and Neck Surgery, University of Southern California Medical School, Los Angeles. Received for publication March 16. 1990; revision received March 7, 1991; accepted March 21, 1991. Reprint requests: Dale H. Rice, MD, Department of OtolaryngologyHead and Neck Surgery, Los Angeles County and University of Southern California Medical Center, Box 795, 1200 North State SI., Los Angeles. CA 90033. 23/1129787
372
Several methods for IRM have been explored in the past, including palpation" or recording of laryngeal muscle movement during RLN electrical stimulation! 3; also, monitoring glottic pressure change as a result of RLN stimulation." Davis et al." recorded electromyagraphic (EMG) activity from laryngeal muscles during surgery, and Lipton et a1.16 and Daube and Harner described how monitoring the spontaneous and evoked laryngeal muscle EMG could be used to monitor RLN integrity during thyroid surgery. Our rationale and procedure for IRM were based on those used for intraoperative monitoring of facial nerve function. The first procedure involves recording the ongoing EMG from musculature innervated by a cranial or peripheral nerve. The EMG can be "played back" using an acoustic monitor from the EMG amplifier to give the surgeon a real-time acoustic signal related to nerve function." The acoustic signal generated by the EMG can indicate baseline or spontaneous levels of muscle activity and will also signal bursts or trains" of EMG activity that may be interpreted in the context of surgical events. That is, changes in the background EMG or prolonged trains or bursts may indicated nerve compromise and warn the surgeon to modify or refrain from surgical manipulation of the nerve." A second method involves evoking EMG activity on the compound motor action potential (CMAP), by presenting an electrical stumulus pulse directly to the nerve by means of a monopolar or bipolar":" probe. The evoked EMG activity can be used to distinguish nerve from
Volume 105 Number 3 September 1991
invading tumor or other tissue, guide the anatomic dissection," and provide a record of neural stimulability throughout the operation. The evidence of stimulability may prove to be prognostic of functional outcome. K A dedicated instrument for nerve monitoring is now commercially available, the XOMED-Treace Nerve Integrity Monitor (NIM, Xomed-Treace, Jacksonville, Fla.). The instrument can be used to monitor ongoing EMG over continuously updated sampling epochs. The NIM provides separate acoustic feedback signals for each of two recording channels. An integrated pulse. generator provides a constant-current or constant-voltage stimulus pulse for evoking EMG activity. Another advantage of the instrument is that the recording epoch can be delayed after presentajtion of the electrical signal for evoked EMG. This reduces the amount of stimulus artifact in the resultant recording. Our experience using this instrument for IRM during thyroidectomy is described in the Methods section. We wish to provide some useful procedures for undertaking IRM, procedures we discovered by trial and error. Certain factors that make IRM somewhat less advantageous than intraoperative facial nerve monitoring will also be described. Our experiences have raised questions about the underlying stimulus-response characteristics of the RLN and the muscles it innervates, which will be discussed subsequently.
ME11tODS Patients
We have used IRM for five patients undergoing thyroidectomies for differentiated carcinoma. All patients had complete otolaryngologic evaluation before surgery. All had clinically normal vocal cord function before and after surgery. Anesthesia
The use of neuromuscular blocking agents must be avoided because these may alter or suppress the neuromuscular events being monitored. It must be recognized, however, that succinyl-choline is used routinely to allow nontraumatic intubation. The anesthesiologist must be informed about the monitoring procedures so that judicious decisions can be made about the type and quantity of anesthetic agents to be administered. Recording Electrodes-Preparation and Placement
Preoperative preparation of several sets of recording electrodes is advised. EMG paired hook-wire electrodes are used for the recordings. The electrodes are compOsed of insulated stainless steel wires, with the terminal 3 to 5 mm stripped of insulation. The ends of
Intraoperative recurrent laryngeal nerve monitoring
373
the wires are bent 180 degrees to aid in insertion and retention of the electrode in muscle tissue. Each pair should be threaded through a 20-cm 22-gauge CHlBA type aspiration needle with the hooked ends advanced just beyond the needle tip. The electrodes may be gas sterilized after they are positioned in the aspiration needle. One pair of electrodes is placed in each thyroarytenoid muscle in the mid-body of the vocal fold, to a depth of approximately 6 to 8 mm. The electrodes are placed under direct laryngoscopic examination. We have found that electrode placement is facilitated by use of the aspiration needle, in comparison to attempts with a long-handled instrument passed through the laryngoscope. Once the electrodes are placed they should be taped to the face to avoid displacement. They are connected to the patient-recorder interface cable using 0.75 mm insulated electrodes that have an alligator clip and 2-mm pin connector at either end. The use of further extension cables for the electrodes is discouraged because this can lead to excessive artifact in the recording. The alligator clip electrodes and their leads should also be taped to the operative table underneath the drapes. This, too, will help guard against electrode disconnection. A standard EEG surface electrode is used for a ground, and is placed on the forehead. The site of placement is prepared by light scrubbing with an abrasive cleaner. A liberal amount of electrode conductive paste is used and the electrode is held in place with surgical tape. When a monopolar stimulus probe is used, an anode electrode must be placed at the back of the shoulder. The anode electrode and its site are prepared in the same fashion as the ground electrode. It should be noted that we tried various alternative locations for anode placement, including sites closer to and farther from the cathode (i.e., the site at which the monopolar probe is applied). The shoulder site placement yielded more reliable recordings of evoked CMAP activity in comparison to the alternative placements. Electrode impedances should be checked after placement and monitored throughout the procedure to assure electrode integrity. Electrode impedances for the intramuscular electrodes may be quite high, between 40 and 50 kO, but they tend to fall to 15 and 25 kO during the dissection. Recording Parameters
Spontaneous EMG activity varies during the operative procedure, requiring ongoing adjustment of the monitor's sensitivity. The range of baseline activity we
374
OlolaryngologyHead and Neck Surgery
RICE and CONE-WESSON
observed was between 40 and 160 j.L V. During the initial stages of dissection for thyroidectomy, we monitor baseline activity at IS-minute intervals and observe the ongoing EMG levels for 2 minutes, using the 60-msec sampling epoch at each interval. We use these levels to determine the "event threshold" level; that is, the EMG level at which an acoustic alarm will be sounded. With the XOMED-NlM, the adjustment of the "event threshold" sensitivity is crucial, because the acoustic alarm signaling increased EMG activity is triggered by EMG events exceeding this level. We adjust the event threshold level to twice the value of the baseline EMG. When the dissection has proceeded to the stage at which the RLN will soon be exposed or manipulated, then the monitoring is constant. We used an ongoing 60-msec sampling epoch for this stage of the procedure. With the event threshold trigger levels appropriately adjusted, any undue manipulation of RLN will incite activity of the vocalis muscle, and the increased EMG will trigger the acoustic alarm, which alerts the surgeon. This can be useful in the identification of the nerve and for the dissection of the nerve- particularly in patients with significant disease in the vicinity of RLN. "Bursts" or "trains" of EMG activity from facial musculature have been reported during facial nerve dissection or manipulation." The bursts are associated with direct manipulation or fluid irrigation in the vicinity of the nerve or electrocautery." Trains can be associated with nerve traction, but may also be present in the absence of surgical provocation. 17 We found train and burst activity evident in the EMG response of laryngeal musculature during surgical manipulation. The diagnostic and prognostic significance of this activity has not yet been determined.
Evoked EMG Nerve integrity can also be determined by direct electrical stimulation of the recurrent laryngeal nerve and observation of the resultant increased EMG activity. Constant-current or constant-voltage stimuli may be used'? with the stimulus presented by means of an insulated flush-tip monopolar (Prass Flush-Tip Probe, Xorned-Treace) or bipolar (Kartush Facial Stimulator, EI Med, Addison, III.) probe. We have been able to evoke activity more consistently and more discretely with the bipolar probe; that is, use of the monopolar probe resulted in more false-positive responses in comparison to the bipolar probe. The bipolar stimulus probe confines the stimulus to a small area and thereby improves the spatial resolution so that neural and nonneural tissue may be distinguished. The stimulus pulse is brief-lOa j.Lsec-and is presented at a rate of four per second. We have observed
variability in the stimulus level needed to evoke a response. With constant voltage stimuli, levels of 0.88 to 1.2 volts elicit a synchronous response. For constantcurrent stimuli, levels of 0.2 to 0.5 rnA are sufficient. The latency of the response appears to be brief, and it is crucial that stimulus artifact be recognized and reduced without compromising the recorded EMG event. With the NIM system used in this investigation, this is achieved by setting a I.S-msec delay between the stimulus and onset of recording. Even this may be too long when working close to the cricothyroid joint. We were able to evoke an EMG response from each patient during the dissection and also after the thyroid tissue had been removed. The RLN was stimulated at 1em or more below Berry's ligament. The evoked EMG response amplitude was large-at least three or four times greater than baseline activity. The stimulus appeared to produce an all-or-nothing response, but we were unable to derive stimulus vs. response amplitude functions with the intraoperative monitoring protocol. The evoked EMG response was evident from the acoustic signal produced by the EMG amplifier, which is synchronous with the stimulus pulse (i.e., acoustic signals produced at four times per second). The response can also be observed on the display screen of the monitor as increased bursts of EMG activity coincident with the stimulus pulses. The stimulability of the nerve indicates that the nerve fibers innervating the muscle are intact. It appears that the close anatomic relationship between the stimulus probe and the recording electrodes, as well as a short response latency, make isolation of the electrical signal difficult. The use of a bipolar stimulus probe, with the stimulus confined between the cathode and anode, yielded better definition of neural from non-neural tissue. We tested for "false-positive" stimulus artifacts by comparing amplifier output when neural tissue was stimulated to that when adjacent non-neural tissue was stimulated. When artifact was present, touching nonneural tissue with the probe would result in synchronous output. For each stimulus level used we performed a control test for non-neural tissue. Responses were judged as present only when control trials were free of artifact. The bipolar probe produced fewer false-positive responses in comparison to the monopolar probe. The bipolar stimulus probe confined the stimulus to the small area between the anode and cathode, resulting in improved spatial resolution for distinguishing neural vs. non-neural tissue.
DISCUSSION This description of our experience is intended to provide the surgeon with procedures that are of immediate
Volume 105 Number 3 September 1991
use when IRM is attempted. In our community, facial nerve integrity monitoring has become the standard of practice for neuro-otologic surgery. It is not our contention, however, that IRM is needed for every thyroidectomy, because the morbidity of the RLN, in the main, is low. II Also, the procedures and dedicated instruments for facial nerve monitoring may need to be modified for IRM to be performed. Research on the threshold, latency, and amplitude characteristics of the evoked EMG is needed to define optimum IRM parameters. Safe levels of constant-current or constant-voltage stimuli need to be defined for this branch of the vagus nerve. The dynamics of the neurotonic response for the RLN are probably quite different from those for the facial nerve, and recording parameters must be modified to address those differences. The effect of anesthetic agents on the RLN and the muscles it innervates must also be investigated. Most laryngeal surgeons are well aware that neuromuscular blocking agents will often leave laryngeal closure reflexes intact. Systematic observation of EMG bursts and trains in the surgical context is needed. Whether these are spontaneous events or are evoked by surgical maneuvers has not been well defined. Their use as diagnostic or prognostic indicators must be evaluated. We look forward to undertaking these investigations and will report our results as they are completed. REFERENCES I. Grundy BL. Electrophysiologic monitoring: Electroencephalography and evoked potentials. In: Newfield P, Cottrel JE, eds. Handbook of neuroanesthesia: clinical and physiologic essentials. Boston: Little Brown and Co., 1983:28-59. 2. Levine RA, Ojemann RG, Montgomery WW, McGaffigan PM. Monitoring auditory evoked potentials during acoustic neuroma surgery: insights into the mechanism of hearing loss. Ann Otol Rhinol Laryngol 1984;93:116-23. 3. Nuwer MR. Evoked potential monitoring in the operating room. New York: Raven Press, 1986 . 4. Nelson KR. Use of peripheral nerve action potentials for intraoperative monitoring. Neurol Clin 1988;6:917-33.
Intraoperative recurrent laryngeal nerve monitoring 375
5. Prass RL. Acoustic EMG monitoring: a prototype of intraoperative cranial nerve monitoring. Short course presented at the American Academy of Neurology, New York: 1987. 6. Prass RL, Luders H. Acoustic (loudspeaker) facial electromyographic monitoring. Part I: Evoked electromyographic activity during acoustic neuroma resection. Neurosurgery 1986; 19:392400. 7. Prass RL, Kinney SE, Hardy RW, Hahn JF, Luders H. Acoustic (loudspeaker) facial EMG monitoring. Part II: Use of evoked EMG activity during acoustic neuroma resection. OTOl.ARYNGOL HEAD NECK SURG 1987;97:541-51. 8. Harner SG, Daube JR, Ebersol MJ, Beatty CWo Improved preservation of facial nerve function with use of electrical monitoring during removal of acoustic neuromas. Mayo Clin Proc 1987;62:92-102. 9. Kartush JM. Electroneurographyand intraoperative facial monitoring in contemporary ncurotology. Paper presented at the American Neurotology Society, Palm Beach, Fla., 1988. 10. Benecke JR Jr, Caler HB, Chadwick G. Facial nerve monitoring during acoustic neuroma removal. Laryngoscope 1987;97:697700. II. Van Heerden JV, Groh MA, Grant ES. Early postoperative morbidity after surgical treatment of thyroid carcinoma. Surgery 1987;101:224-7. 12. James AG, Crocker S, Woltering E, et al. A simple method for identifying and testing recurrent laryngeal nerve. Surg Gynecol Obstet 1985;161:186-7. 13. Spahn JG, Bizal J, Ferguson S, et al. Identification of the motor laryngeal nerves-a new electrical stimulation technique. Laryngoscope 1981;91:1865-8. 14. Hvidegaard T, Vase P, Dalsgaard SC, et al. Endolaryngeal devices for perioperative identification and functional testing of the recurrent laryngeal nerve. OTOLARYNGOL HEAD NECK SURG 1984;92:292-4. 15. Davis WE. Rea JL, Templer J. Recurrent laryngeal nerve localization using a microlaryngeal electrode. OTOLARYNGOL HEAD NECK SURG 1979;87:330-3. 16. Lipton RJ, McCaffrey TV, Litchy WJ. Intraoperative electrophysiologic monitoring of laryngeal musele during thyroid surgery. Laryngoscope 1988;98: 1292-6. 17. Niparko JK, Kileny PRo Intraoperative monitoring offacial function. Semin Hearing 1988;9:127-39. 18. Kartush JM, Niparko JK, Bledsoe SC, Graham MD, Kaminke JL. Intraoperative facial nerve monitoring: a comparison of stimulating electrodes. Laryngoscope 1985;95: 1536-40. 19. Prass RL, Luders H. Constant-current versus constant-voltage stimulation. J Neurosurg 1985;62:622-3.
