Anesthesia Considerationsfor Neurophysiologic Monitoring using the ProPep Nerve Monitoring System™ duringda Vinci® ProstatectomyBecause the ProPep Nerve Monitoring System is measuring stimulated electromyographic (EMG) signals emanating from themuscles in which the nerves of interest terminate, it is important the muscles not be paralyzed during that portion of the surgerywhen neurophysiologic monitoring is being performed. As a result, there are a number of anesthesia considerations that need tobe kept in mind to optimize the validity and quality of the neurophysiologic readings. Please note that all decisions regardinganesthesia are the responsibility of the attending licensed medical practitioner administering anesthesia. It is important that thesurgeon discuss these issues preoperatively with the attending licensed medical practitioner administering the anesthesia.Caution: The use of paralyzing anesthetic agents will significantly reduce, if not completely eliminate, EMG responsesto direct or passive nerve stimulation. Whenever nerve paralysis is suspected, consult the attending licensed medicalpractitioner administering the anesthesia.Before the Start of the Surgery: - A conversation between the Surgeon and the attending medical practitioner administering the anesthesia should take place to discuss: o At what point during the surgery will the monitoring occur; o How will the physician alert the medical practitioner administering the anesthesia that the portion of the case requiring monitoring is approaching and how much lead time would the medical practitioner administering the anesthesia like to be given. This is important information that will allow the medical practitioner administering the anesthesia to ensure the muscle relaxants have worn off adequately so that the surgeon can obtain the best opportunity for recording useful and valid responses during the monitoring process.During The Surgery: - Only short acting muscle relaxants should be used. - Muscle relaxants should be dosed incrementally. o The goal is to keep the patient at 3-2 well defined twitches during the neurophysiologic monitoring. - The surgeon will communicate with the medical practitioner administering the anesthesia when they are approximately 20 minutes (or the previously agreed upon time) away from performing the neurophysiologic monitoring. o This will allow adequate time for the neuromuscular blockade to wear off sufficiently giving the surgeon the best opportunity for optimal responses during the monitoring process.Additional Considerations: - The medical practitioner administering the anesthesia should be prepared to use pressure control ventilation as a means to improve the ability to ventilate the patient when they are becoming “light” on muscle relaxants. - During the period of reduced neuromuscular blockade, the stability of the surgical view for the operating surgeon can be improved by reducing the drive to breath using: o over-ventilation to reduce CO2 o narcotics.ProPep Surgical wishes to thank Dr. Paul Playfair - Chief of Anesthesia at Westlake Medical Center – Austin, TX for hiscontributions to this protocol.References: Attached you will find references that address anesthesia considerations during neurophysiologic monitoring inmore depth. Please refer to the highlighted sections for considerations specific to the mode of neurophysiologic monitoring theProPep Nerve Monitoring System employs.V2.1_12 April 2012
IntraoperativeNeurophysiological Monitoring Second Edition Aage R. Møller, PhD University of Texas at Dallas Dallas, TX
ContentsPreface ..................................................................................................................................... vAcknowledgments ..................................................................................................................vii1 Introduction ..................................................................................................................... 1SECTION I: PRINCIPLES OF INTRAOPERATIVE NEUROPHYSIOLOGICAL MONITORING2 Basis of Intraoperative Neurophysiological Monitoring .................................................. 93 Generation of Electrical Activity in the Nervous System and Muscles .......................... 214 Practical Aspects of Recording Evoked Activity From Nerves, Fiber Tracts, and Nuclei ............................................................................................ 39References to Section I .......................................................................................................... 49SECTION II: SENSORY SYSTEMS5 Anatomy and Physiology of Sensory Systems ................................................................ 556 Monitoring Auditory Evoked Potentials ......................................................................... 857 Monitoring of Somatosensory Evoked Potentials ......................................................... 1258 Monitoring of Visual Evoked Potentials ....................................................................... 145References to Section II ....................................................................................................... 147SECTION III: MOTOR SYSTEMS9 Anatomy and Physiology of Motor Systems ............................................................... 15710 Practical Aspects of Monitoring Spinal Motor Systems ............................................... 17911 Practical Aspects of Monitoring Cranial Motor Nerves ............................................... 197References to Section III ...................................................................................................... 213SECTION IV: PERIPHERAL NERVES12 Anatomy and Physiology of Peripheral Nerves ........................................................... 22113 Practical Aspects of Monitoring Peripheral Nerves ..................................................... 229References to Section IV ..................................................................................................... 233SECTION V: INTRAOPERATIVE RECORDINGS THAT CAN GUIDE THE SURGEON IN THE OPERATION14 Identification of Specific Neural Tissue ....................................................................... 23715 Intraoperative Diagnosis and Guide in Operations ..................................................... 251References to Section V ...................................................................................................... 273 ix
x ContentsSECTION VI: PRACTICAL ASPECTS OF ELECTROPHYSIOLOGICAL RECORDING IN THE OPERATING ROOM16 Anesthesia and Its Constraints in Monitoring Motor and Sensory Systems ................. 27917 General Considerations About Intraoperative Neurophysiological Monitoring .......... 28318 Equipment, Recording Techniques, Data Analysis, and Stimulation ........................... 29919 Evaluating the Benefits of Intraoperative Neurophysiological Monitoring .................. 329References to Section VI ..................................................................................................... 339Appendix ............................................................................................................................. 343Abbreviations ...................................................................................................................... 347Index ................................................................................................................................... 349
280 Intraoperative Neurophysiological Monitoringand so forth will cause increased central conduc- interaction with the NMDA receptor) and it couldtion time (CCT) for somatosensory evoked provoke seizure activity in individuals withpotentials (SSEPs) and essentially make it epilepsy but not in normal individuals. Ketamineimpossible to elicit motor evoked potentials by has been reported to increase cortical somatosen-single-impulse stimulation of the motor cortex sory evoked potential (SSEP) amplitude and to(transcranial magnetic or electrical stimula- increase the amplitude of muscle and spinaltion). This unfortunate effect is present even at recorded responses following spinal stimulationlow concentrations. and it could potentate the H reflex. Ketamine has minimal effects on muscle responses evoked byIntravenous Anesthesia transcranial cortical stimulation. Because of that, Some intravenous agents have almost always ketamine combined with opioids has become abeen used together with inhalational agents, valuable adjunct during some TIVA techniquesbut, recently, the TIVA regimen has become for recording muscle responses. The fact that ket-increasingly prevalent. One reason for that is amine could cause severe hallucinations post-that the inhalational agents, including nitrous operatively and increase intracranial pressure hasoxide, are obstacles when electromyographic reduced its use in anesthesia.(EMG) responses are to be monitored in con- Opioids provide analgesia but do not pro-nection with transcranial stimulation of the vide sufficient degrees of sedation, relief ofmotor cortex. It is an advantage that the mech- anxiety, and loss of memory during operationsanism of action of intravenous agents appears (amnesia). Hence, TIVA usually includes someto be different from that of inhalational agents sedative–hypnotic agents such as barbituratesin such a way that benefits monitoring EMG (thiopental) and benzodiazepines such as mida-and of MEPs (see Chap. 10). zolam. Propofol is an agent that is in increasing use because it provides excellent anesthesia Analgesia. Achieving analgesia (pain relief) and limited effect on MEPs.is a primary component of anesthesia, and for Barbiturates that are often used for inductionmany years, opioids have been used in the of general anesthesia have effects similar toanesthesia regimen together with agents such that of inhalation agents on evoked potentials.as inhalation agents for achieving unconscious- For example, muscle responses to transcranialness (3). One of the oldest synthetic opioids is stimulation are unusually sensitive to barbitu-fentanyl, but now several different agents with rates and the effect lasts a long time, makingsimilar action are in use for that purpose, such barbiturates a poor choice in connection withas alfentanil, sufentanil, and remifentanil. Mus- monitoring MEPs.cle responses evoked by transcranial cortical Etomidate is another popular agent to bestimulation (electrical and magnetic) are only used in intravenous anesthesia. It enhancesslightly affected by opioids. The effects of opi- synaptic activity at low doses; thus, opposite tooids can be reversed by administering nalox- the action of barbiturates and benzodiazepines,one, suggesting that the effect is related to it might produce seizures in patients withµ-receptor activity. Intravenous sedative agents epilepsy when given in low doses (0.1 mg/kg)are frequently used to induce or supplement and it might produce myoclonic activity atgeneral anesthesia, particularly with opioids induction of anesthesia. The ability to enhanceor ketamine, when inhalational agents are not neural activity or reduce the depressant effectsutilized. of other drugs has been used to enhance the Ketamine is a valuable component of anes- amplitude of both sensory and motor evokedthetic techniques allowing recording responses responses. The enhancing of evoked activitythat might be depressed by other anesthetics. occurs at doses similar to those that produce theKetamine could heighten synaptic function desired degree of sedation and loss of recall ofrather than depress it (probably through its memory when used in TIVA.
Chapter 16 Anesthesia 281 Benzodiazepines, notably midazolam, are of anesthetic agent; for instance, it is not possi-often used in connection with TIVA in many ble to record EMG potentials if the patient iskinds of operations because they provide excel- paralyzed, as is the case for many commonlylent sedation and they suppress memories used anesthesia regimens. Recording of corti-(recall). Benzodiazepines can also reduce the cal evoked potentials is affected by most of therisk of hallucinations caused by ketamine. agents commonly used in surgical anesthesia. Monitoring motor evoked responses elicited byMuscle Relaxants transcranial magnetic or electrical stimulation Muscle relaxants are usually not regarded as of the motor cortex requires special attentionanesthetics but often combined with agents on anesthesia and the use of a special anesthe-(intravenous or inhalation) that produce uncon- sia regimen is necessary.sciousness and freedom of pain. Muscle relax-ants are part of a common anesthesia regimen–– Recording of Sensory Evoked Potentialsso-called “balanced anesthesia” (neurolept It is advantageous to reduce the use ofanesthesia)––that includes a strong narcotic for halogenated agents and nitrous oxide in anes-analgesia plus a muscle relaxant to keep the thesia when cortical evoked potentials arepatient from moving, together with a relatively monitored. Monitoring of short-latency sen-weak anesthetic such as nitrous oxide. sory evoked potentials is not noticeably Muscle relaxants used in anesthesia are of two affected by any type of inhalation anesthesia;different types, each affecting muscle responses therefore, short-latency sensory evoked poten-differently: one blocks transmission in the neuro- tials should be used whenever possible formuscular junction (muscle endplate) and the intraoperative monitoring instead of corticalother type depolarizes the muscle endplate, evoked potentials. Auditory brainstem responsesthereby preventing it from activating the muscle. (ABRs), which are short-latency evoked poten-The oldest neuromuscular blocking agent is tials, are practically unaffected by inhalationcurare, but that has been replaced by a long anesthetics and can be recorded regardless ofseries of steroid-type endplate blockers with the anesthesia used. Short-latency componentsdifferent action durations. Pancuronium bro- of SSEPs are not affected by inhalation anes-mide (Pavulon®) was one of the earliest of this thetics, but only upper limb SSEPs haveseries and the effects of pancuronium bromide clearly recordable short-latency components.last more than 1 h when a dose that causes total Short-latency SSEPs evoked by stimulation ofparalysis is administered. Other and newer drugs the median nerve are suitable for monitoringof the same family have a shorter duration of the brachial plexus and the cervical portion ofaction (about 0.5 h for vecuronium bromide, the spinal cord, but they are not useful for mon-[Norcuron®] and atracurium [Tracurium®]). itoring the spinal cord below the C6 vertebra or The most often used muscle-relaxing agent for monitoring central structures such as thethat paralyzes by depolarizing the muscle end- somatosensory cortex. Therefore, it is usuallyplate is succinylcholine. The muscle-relaxing the long-latency components, which are gener-effect of succinylcholine lasts only a very short ated in the cortex, that are used for intraopera-time. tive monitoring of SSEP. The general effect of anesthetics is a lower- ing of the amplitude and a prolongation of the EFFECTS OF ANESTHESIA latency of an individual component of theON RECORDING NEUROELECTRICAL recorded potentials (4) (see Chap. 7, Fig. 7.10). POTENTIALS The effect is different for different components of the evoked potentials, as the potentials are Successful neurophysiological monitoring affected by inhalation anesthetics or barbitu-often depends on the avoidance of certain types rates to varying degrees (5) and the effect varies
282 Intraoperative Neurophysiological Monitoringfrom patient to patient, with children being gen- the use of such “reversing” agents is that a fairerally more sensitive than adults (6). amount of muscle response (10–20%) has Because these components are affected by returned before reversing is attempted. It is alsoinhalation anesthetics it is important to discuss important to note that such reversing does notwith the anesthesiologists in order to select a immediately return the muscle function to nor-type of anesthesia that allow such monitoring. mal, as the effect of the muscle relaxant will last for some time.Recording of EMG Potentials When muscle relaxation is not used during Response from muscles (electromyographic an operation, the patient could have noticeable[EMG] potentials or mechanical response) can- spontaneous muscle activity, which increasesnot be recorded in the presence of muscle the background noise level in recordings of dif-relaxants. It is usually necessary to use a mus- ferent kinds of neuroelectrical potential. This iscle-relaxing agent for intubation. When EMG important when monitoring of evoked poten-recordings are to be done during an operation, tials of low amplitude, such as ABR, is to beit is suitable to use succinylcholine together done. The resulting background noise will pro-with 3 mg of d-tubocurarine (curare) or short- long the time over which responses must beacting endplate blockers, such as atracurium averaged in order to obtain an interpretable(Tracurium) or vecuronium bromide (Norcuron) recording. The muscle activity often increasesduring intubation. This will allow monitoring of as the level of anesthesia lessens. If the musclemuscle potentials 30–45 min after the adminis- activity becomes strong, it might be a sign thattration of the drug, providing that only the min- the level of anesthesia is too low. Early infor-imal amount of the drug is given and that it is mation about such increases in muscle activitygiven only once for intubation. is naturally important to the anesthesiologist so If a short-acting endplate-blocking agent is that he/she can adjust the level of anesthesiaused, it is important to be aware that the para- before the patient begins to move sponta-lyzing action disappears gradually and at a rate neously. In this way, electrophysiological mon-that differs from patient to patient. The rate at itoring can often provide valuable informationwhich muscle function is regained depends on to the anesthesiologist, because if anesthesiathe age, weight, and so forth of the patient, what becomes light, spontaneous muscle activity fre-other diseases might be present, and what other quently manifests in the recording of evokedmedications might have been administered. potentials from scalp electrodes a long time During the time that the muscle-relaxing before any movement of the patient is noticed.effect is decreasing, stimulation of a motor To do that, the output of the physiological ampli-nerve with a train of electrical shocks (such as fier must be watched continuously to detect anythe commonly used “train of four” test) will muscle activity.give rise to a relatively normal muscle contrac- Intraoperative monitoring that involvestion in response to the initial electrical stimu- recording EMG potentials from muscles islus, but the response to subsequent impulses becoming more and more common in thedecreases and will be less than normal. complex neurosurgical operations that can The effect of muscle relaxants of the endplate- now be performed and demands on theblocking type can be shortened (“reversed”) by selection of an appropriate anesthesia regimenadministering agents such as neostigmine, which have, therefore, increased. A close collaborationinhibits the breakdown of acetylcholine and between the anesthesia team and the neuro-thereby makes better use of the acetylcholine physiologist in charge of intraoperativereceptor sites that are not blocked by the muscle neurophysiological monitoring can often solverelaxant that is used. However, a prerequisite for such problems.
Husain 00 1/17/08 11:51 AM Page iii A Practical Approach to Neurophysiologic Intraoperative Monitoring Edited by Aatif M. Husain, MD Department of Medicine (Neurology) Duke University Medical Center Durham, North Carolina New York
Husain 04 1/17/08 11:55 AM Page 55 4 Anesthetic Considerations Michael L. James T he practice of anesthesia has histori- cally relied on the induction of a reversible state of amnesia, analgesia, and consists of four basic stages: premedication, induction, maintenance, and emergence. Prior to entering the operating suite, “premedica- motionlessness. With the improvement of med- tions” may be administered to prepare the ical technology, advancement of knowledge, patient for the perioperative period. Usually and practice of evidence-based medicine, mod- this takes the form of mild sedation for anxiol- ern anesthesiology comprises a great deal more. ysis, analgesics for preprocedural pain, antihy- It has become the role of the anesthesiologist pertensives, antiemetics for patients with a during surgical, obstetrical, and diagnostic pro- high likelihood of postoperative nausea and cedures to provide anesthesia, optimize proce- vomiting, antisialagogues to facilitate intuba- dural conditions, maintain homeostasis, and, tion, etc. In the operating room the historic should it be necessary, manage cardiopul- principles of anesthesia are still the foundation monary resuscitation. Additionally, anesthesi- of practice, and analgesia (i.e., painlessness), ology has found itself branching out into amnesia (i.e., memory loss), motionlessness, chronic and acute pain treatment as well as the and hemodynamic stability can be obtained intensive care unit. Obviously there has been an and maintained by a variety of means. expansion of expectations for the practice of Commonly, general anesthesia is induced anesthesia over the last few decades; however, through the administration of a large bolus ultimately, anesthesiology is the practice of dose of an intravenous sedative-hypnotic (e.g., manipulating a patient’s neurologic system and propofol). A dose of intravenous opioid (e.g., physiology to effect some beneficial end. fentanyl) and a paralytic agent (e.g., vecuro- nium) may be given at this time as well to facil- itate endotracheal intubation. After induction, anesthesia maintenance usually consists of PRINCIPLES OF ANESTHESIA some amount of inhaled volatile anesthetic There are four basic types of “anesthesia”: agent (e.g., isoflurane) in a mix of oxygen and general anesthesia, regional anesthesia, local either air or nitrous oxide and some dose of anesthesia, and sedation. For the purposes of intravenous opioid. The amount of volatile neurophysiologic intraoperative monitoring agent is quantified in terms of mean alveolar (NIOM), general anesthesia (the creation of concentration (MAC). MAC is expressed as a reversible coma) is nearly always required and percentage of inhaled gas and is defined as the 55
Husain 04 1/17/08 11:55 AM Page 56 56 • S E C T I O N I : B a s i c P r i n c iples alveolar partial pressure of a gas at which 50% is arguably the most important factor, and a of patients will not move with a 1-cm abdom- great deal of human physiology is influenced inal surgical incision. However, in practice the by actions of the anesthesiologist. The manner necessary amount of volatile agent is deter- in which these physiologic functions are mined by effect. It is during anesthesia mainte- manipulated often directly determines meas- nance that NIOM occurs (as does the surgical urable neurophysiologic function. Further, it procedure). After the procedure is finished, the is reasonable to assume that physiologic func- expectation is that the anesthetic coma will be tion determines, in large part, the survivability completely reversible, and the patient must of nerves and their supporting structures. emerge from anesthesia without experiencing lasting effects from the agents. Emergence is Temperature usually accomplished by reversing any residual neuromuscular blockade and allowing the It is well established that temperature patient to eliminate volatile agent via breath- plays a significant role in nerve function, espe- ing. Volatile anesthetic agents are minimally cially in the axon. Changes of a fraction of a metabolized and largely removed from the degree can drastically alter latencies and body in the same manner they were intro- amplitudes of neuronal potentials with corti- duced: ventilation. cal structures being more affected than In terms of NIOM, special considerations peripheral nerves (2). Relative hypothermia for general anesthesia are discussed later; how- produces changes that invariably present as ever, it is important for neurophysiologists and slowed latencies from slower nerve conduc- technologists to have a clear expectation of the tion. In addition there are predictable, charac- step-by-step nature whereby anesthetic and teristic effects of profound hypothermia that, surgical procedures are undertaken, and it is at least initially, begin with slowing to a delta important to remember that the operating frequency (3). The opposite is true with rela- room is generally a highly active environment tive hyperthermia for both evoked potentials with people, monitors, equipment, and electri- (EPs) and electroencephalograms (EEGs). It is cal cords all moving about at once. Any change important to note that regional temperature in the NIOM may be due to many factors, not changes are invariably difficult to predict, for the least of which is the surgical procedure, and a variety of reasons. General anesthesia causes every attempt should be made to regain fading an overall cooling effect in the body core due or lost waveforms, as permanent loss may indi- to peripheral vasodilatation, which is usually cate postoperative impairment (1). Therefore opposed by active surface warming and the entire process becomes most efficient when warmed intravenous fluids. Additionally, cold each individual in the room understands all the and/or warm irrigants are nearly always steps, including those of every other individual, applied to the surgical field. As a result, the required to prepare for, perform, and enable extremities, brain, and spinal cord are being emergence from a procedure in an environment heated or cooled depending on where they lie of open communication and respect for each in relation to warmed air blankets, intra- other’s responsibilities. venous fluid lines, the surgical field, etc. Therefore, unless it is individually measured, the actual temperature of a given region is impossible to know, but the potential effects NONPHARMACOLOGIC FACTORS: should be kept in mind during the course of ANESTHETIC CONSIDERATIONS monitoring. It is very common for patients to Physiologic function of the human body experience a decrease in core body tempera- plays a major role in neuronal functioning; it ture for the first 15 minutes after anesthetic
Husain 04 1/17/08 11:55 AM Page 57 CHAPTER 4: Anesthetic Considerations • 57 induction. With active warming during the tionally), patient positioning, tourniquets, administration of most anesthetic agents— vasospasm, vascular ligation, etc. Anecdotally, unless the surgical procedure requires an some have reported discovering incidental alternative strategy—the patient’s temperature ulnar nerve ischemia secondary to compres- will then be kept greater than 36°C by the sion during routine monitoring for spinal anesthesiologist. fusion. When the compression was released, the nerve potentials returned to normal. Blood Flow Ventilation Logic dictates that ischemic nerves do not function normally; therefore measurable neu- Optimal neural functioning depends on ral potentials would become abnormal. In fact maintenance of a homeostatic extracellular it has been demonstrated that somatosensory environment. Hypo- or hypercapnea can alter evoked potentials (SEPs) can be lost when cellular metabolism by changing the acid-base cerebral blood flow falls below 15 status of the individual. In general individuals mL/min/100 g (2). This can be assumed to be tolerate relatively profound acid-base true for the spinal cord and peripheral nerves derangements, especially upward trends in as well. Unfortunately, it is difficult to actually pH. Unless the pH of a patient drops below measure blood flow to any given structure, so 7.2, neuronal mechanisms are maintained. systemic blood pressure is often used as a sur- Additionally, there is a suggestion that rogate. Furthermore, systemic blood flow extremes in hypocarbia (< 20 mmHg partial does not necessarily dictate regional blood pressure) can alter SEP monitoring (5). flow, especially in the brain, which makes it Alternatively, profound hypoxia is poorly tol- even more difficult to predict. Monitors are erated, especially in the surgical setting of becoming available that purport to quantify ongoing blood loss and potential hypotension. regional blood flow (e.g., cerebral oximetry, microdialysis), but a discussion of these is Hematology beyond the scope of this chapter. Essentially then, there are two main considerations for Like hypoxia, profound anemia can con- the neurophysiologist: systemic hypotension tribute to neural dysfunction. Normally, ane- and decreased regional blood flow. When pro- mia is well tolerated to levels of hemoglobin found, systemic hypotension results in glob- less than 7 g/dL. However, in the surgical set- ally reduced blood flow, which translates into ting of possible large volume blood loss, tissue ischemia of varying degrees based hypotension, and hypoxia, it is generally largely on autoregulation. For example, dur- accepted that hemoglobin levels should be ing spinal surgery, controlled deliberate kept above 8 g/dL and may require optimizing hypotension is often requested of the anesthe- at 10 g/dL. At approximately 10 g/dL of siologist so as to assist in controlling blood hemoglobin, oxygen delivery appears to be loss; however, surgical traction and hypoten- maximized and transfusion above this thresh- sion can aggravate each other with deleterious old does not appear to improve augmenta- effects to the patient, and NIOM can assist in tion. There are animal data that support this determining the acceptable limit of systemic supposition in SEP monitoring (6). hypotension (4). There are many examples of causes of decreased regional blood flow, and Intracranial Pressure almost all are due to some interruption in blood supply either due to compression from Increase in intracranial pressure is a rela- surgical instruments (intentionally or uninten- tively well documented cause of shifts in cor-
Husain 04 1/17/08 11:55 AM Page 58 58 • S E C T I O N I : B a s i c P r i n c iples tical responses of EPs and prolongation of EFFECTS OF SPECIFIC motor evoked potentials (MEPs), presumably ANESTHETIC AGENTS due to compression of cortical structures. In general the anesthesiologist and neuro- There is a pressure-related increase in latency physiologist are constantly at odds in that and decrease in amplitude of cortical SEPs nearly all anesthetic agents, given in high and as intracranial pressure becomes patho- enough doses, cause depression of NIOM logic, uncal herniation occurs with subsequent potentials. However, with open communica- loss of subcortical SEP responses and brain- tion and mutual understanding of each other’s stem auditory evoked potentials (BAEPs) (7). activities, NIOM can be successful with nearly Alleviation of this pressure can return EPs to any anesthetic technique. The crucial concept normal. is that any change in either anesthetic or NIOM must be communicated to the team, so Other Factors that every person in the operating room is act- ing under appropriate assumptions. Neuronal function depends on mainte- nance of a homeostatic intra- and extracellu- lar environment determined by potassium, Inhalation Agents calcium, and sodium concentrations. It is log- Despite being the oldest form of anesthe- ical to assume that alteration in these concen- sia, the exact mechanism of action of inhala- trations would result in dysfunction and tion agents remains unclear. Inhalation possible changes in measurable neuronal anesthetics consist of two basic gases avail- potentials. The concentration of these ions is able in the United States: halogenated agents largely in the control of the anesthesiologist, (halothane, isoflurane, sevoflurane, desflu- and maintenance within ranges of normal val- rane) and nitrous oxide. Doses of gas are ues is necessary. In addition, profound hyper- given as percentage of inhaled mixture, and or hypoglycemia should be avoided, as either effective doses are expressed as some amount extreme can result in cellular dysfunction; of MAC. As discussed before, one MAC of an although there is no evidence that they result agent is sufficient to prevent 50% of patients in intraoperative changes in NIOM, there are from moving to the stimulation of surgical data to suggest that both can lead to poor out- incision (Tables 4.1 and 4.2). comes (8). TABLE 4.1 Effects of Inhaled Agents on Evoked Potentials BAEP SEP MEP Agents Latency Amplitude Latency Amplitude Latency Amplitude Desflurane Inc 0 Inc Dec Inc Dec Enflurane Inc 0 Inc Inc Inc Dec Halothane Inc 0 Inc Dec Inc Dec Isoflurane Inc 0 Inc Dec Inc Dec Sevoflurane Inc 0 Inc Dec Inc Dec Nitrous oxide 0 Dec 0 Dec Inc Dec Inc = increased; Dec = decreased; 0 = no change.
Husain 04 1/17/08 11:55 AM Page 59 CHAPTER 4: Anesthetic Considerations • 59 TABLE 4.2 Effects of Anesthetics Agents on Electroencephalogram INCREASED FREQUENCY SUPPRESSED Barbiturate (low dose) Barbiturates (high dose) Benzodiazepine Propofol (high dose) Etomidate Benzodiazepine (high dose) Propofol Ketamine Halogenated agents (< 1 MAC) INCREASED AMPLITUDE ELECTROCEREBRAL SILENCE Barbiturate (moderate dose) Barbiturates Etomidate Propofol Opioid Etomidate Halogenated agents Halogenated agents (1–2 MAC) (> 2 MAC except halothane) Inc = increased; Dec = decreased; 0 = no change. Halogenated Agents ever, cord stimulation results in stimulation The halogenated agents consist of the his- of the sensory and motor pathways, and toric agent halothane, which is still used in halogenated gases preferentially block the most countries outside the United States, and motor responses (10). Therefore it is impor- the modern agents consisting of isoflurane, tant to remember that NIOM utilizing spinal sevoflurane, and desflurane. Each has its own cord stimulation may not reliably monitor MAC, onset and offset times, and metabolism motor function in the presence of halo- based on the inherent properties of the gas. genated gases. For this and reasons men- Their use results in a dose-related decrease in tioned above—namely, easy ablation when amplitude and slowing of latency of SEPs, MEP monitoring is essential—halogenated with the least effect seen in peripheral and gases should usually not be part of the anes- subcortical responses (2). BAEPs are mini- thetic regimen when using this modality. mally affected by halogenated anesthetics at The EEG is affected but usually without usual doses but can be ablated at high doses. hindrance to monitoring. All halogenated MEPs are enormously affected by the use anesthetics produce a frontal shift of the of halogenated agents and can be entirely rhythm predominance when used at induction ablated even with doses of 0.5 MAC. It doses (two to three times MAC doses). The appears that this effect occurs proximal to the gases then produce a dose-dependent reduc- anterior horn cell due to evidence that waves tion in frequency and amplitude. It is impor- recorded distal to the anterior horn cell and tant to note that both isoflurane and proximal to the neuromuscular junction desflurane can produce burst suppression and remain recordable even at high doses of anes- electrocerebral silence at clinical doses. For thetic (9). MEP monitoring may also occur practical purposes, however, all halogenated through spinal or epidural stimulation with agents can be used for maintenance anesthesia minimal effect on recorded responses; how- when NIOM requires EEG monitoring.
Husain 04 1/17/08 11:55 AM Page 60 60 • S E C T I O N I : B a s i c P r i n c iples Nitrous Oxide tions vary rapidly with inhaled concentra- Nitrous oxide is similar to halogenated tions, so that if NIOM is problematic and anesthetic agents and causes a dose-related needs maximizing intraoperatively, discontin- decrease in amplitude and prolongation of uance of nitrous oxide will quickly result in latency of cortical SEPs and ablation of MEPs. the its elimination from the brain and body. This effect seems somewhat limited in subcor- tical and peripheral potentials of the SEPs. At Intravenous Agents equipotent doses to halogenated agents, nitrous oxide may, in fact, cause greater EP Intravenous anesthetic agents are gener- depression (2). Additionally, nitrous oxide has ally used to induce anesthesia and afterwards somewhat indeterminate effects on the EEG to supplement inhalation maintenance anes- that is highly dependent on other agents and thesia. Most modern anesthetic techniques doses being used simultaneously. The effects consist of a variety of agents, intravenous and on the EEG are not wholly predictable, but inhaled; nearly always an intravenous opioid generally, there is frontally dominant high-fre- is administered to augment other agents for quency activity and posterior slowing. Despite either tracheal intubation at induction or this, a frequent anesthetic technique used dur- intense surgical stimulation exceeding a stable ing NIOM is a “nitrous-narcotic” technique. maintenance anesthesia. If halogenated agents The modern version of this technique consists are contraindicated or NIOM becomes prob- of a high-dose remifentanil infusion (0.2 to lematic with their use, a complete anesthetic 0.5 µg/kg/min) with 60% to 70% inhaled can consist of intravenous drugs, or total fraction of nitrous oxide. A high, but con- intravenous anesthesia (TIVA). TIVA exists in stant, amount of nitrous oxide is delivered many forms. The most common regimen is with varying amounts of remifentanil based based on continuous propofol infusion and on surgical stimulation. As long as the per- supplementation with intravenous opioid. centage of inhaled nitrous oxide is held con- However, all manner of TIVAs have been stant, this practice generally allows recordable described, including the use of ketamine, bar- responses for most NIOM except transcranial biturate, midazolam, dexmedetomidine, etc., MEPs, although even then 50% to 60% with drug selection depending on utilizing nitrous oxide may be used. The benefit of specific attributes of an agent to effect a spe- using nitrous oxide is that brain concentra- cific outcome (Tables 4.2 and 4.3). TABLE 4.3 Effects of Intravenous Agents on Evoked Potentials BAEP SEP MEP Agents Latency Amplitude Latency Amplitude Latency Amplitude Barbiturate Low dose 0 0 0 0 Inc Dec High dose Inc Dec Inc Dec Inc Dec Benzodiazepine 0 0 Inc Dec Inc Dec Opioid 0 0 Inc Dec 0 0 Etomidate 0 0 Inc Inc 0 0 Propofol Inc 0 Inc Dec Inc Dec Ketamine Inc 0 Inc Inc 0 0 Inc = increased; Dec = decreased; 0 = no change.
Husain 04 1/17/08 11:55 AM Page 61 CHAPTER 4: Anesthetic Considerations • 61 Barbiturates drug can slow SEP latencies and decrease Some of the oldest intravenous anesthet- amplitudes (12). Furthermore, as with most ics include barbiturates (e.g., thiopental, pen- other anesthetics, even small doses of benzo- tobarbital, phenobarbital, methohexital). diazepines (1 to 2 mg) can lead to a marked These drugs exist in alkaline salt solution and reduction in MEP responses. However, owing exert their mechanism of action at the GABAA to relatively rapid metabolism of single receptor. Of these, thiopental remains in com- adminstration, if small doses of midazolam mon use, in certain surgical cases, as an induc- are given preoperatively, their effects on tion agent and as a means of achieving NIOM are usually minimal. Of note, benzodi- neuroprotection through “burst suppression.” azepines are anticonvulsants and will all pro- Additionally, methohexital is frequently used duce slowing of the EEG into the theta range; to facilitate electroconvulsive therapy (ECT). however, at small doses they create beta- However, much like halogenated agents, bar- rhythm predominance in frontal leads, which biturates will produce EEG slowing and, at is also seen with chronic oral administration. higher doses, burst suppression and electro- cerebral silence. There appears to be little Propofol class effect of barbiturates on SEPs, with each Propofol remains one of the most com- agent producing somewhat different results. mon agents used for the induction of anesthe- Thiopental produces transient decreases in sia and is the most common agent used for amplitude and increases in latency with bolus maintenance anesthesia during TIVA. It is dosing for induction, but phenobarbital pro- packaged in a lipid-soluble solution and its duces little effect until doses causing cardio- site of action is also at the GABA receptor. vascular collapse are reached (11). As with Owing to rapid redistribution after dosing, inhaled agents, SEP cortical potentials seemed propofol is easily titratable to the desired to be most affected, with relative sparing of effect, which makes it very useful for TIVA subcortical and peripheral responses. In con- techniques. Induction doses of propofol (2 to trast, whether with low-dose continuous infu- 5 mg/kg) cause amplitude depression of EEG, sion or single-bolus dosing, MEP responses SEP, and MEP responses, as does high-dose can be entirely abolished with the use of bar- continuous infusion (80 to 100 µg/kg/min). biturates. Any anesthetic given for a surgical However, there is generally rapid recovery procedure requiring MEP monitoring should after termination if long infusion times (>8 exclude the use of barbiturates in any form hours) are avoided (13). In recording SEPs or unless their use (i.e., neuroprotection) super- MEPs from the epidural space, there seems to sedes the benefit from MEP monitoring. be limited effect of the drug on the EPs; this seems to hold true for recordings from the Benzodiazepines scalp or peripheral muscle as well (14). Midazolam is a common intravenous Propofol is also notable as an agent that can benzodiazepine used in preoperative areas produce burst suppression and electrocerebral prior to transfer to the operating suite. silence on the EEG. Despite profound EEG Benzodiazepines also have their site of action suppression at high dose, propofol retains its at the GABA receptor and have the desirable relatively quick termination, allowing for an effects of amnesia, sedation, and anxiolysis. In awake, alert, and neurologically testable general, single one-time doses of midazolam patient at the end of a surgical procedure. given prior to induction have little effect on NIOM during critical portions of the proce- Opioids dure. However, induction doses of midazolam Intravenous opioids represent a critical (0.2 mg/kg) or continuous infusions of the mainstay in the practice of modern “balanced”
Husain 04 1/17/08 11:55 AM Page 62 62 • S E C T I O N I : B a s i c P r i n c iples anesthesia to control perioperative pain. ing NIOM. Additionally, the use of ketamine Nearly all general anesthetics will have some can produce larger amplitudes, with mild form and dose of intravenous opioid as a cen- slowing into the theta range on the EEG, and tral component. Intravenous opioids in current there is anecdotal evidence that ketamine use during the perioperative period include may be proconvulsant. The downside to ket- morphine, hydromorphone, fentanyl, alfen- amine use (and the reason ketamine fell out tanil, sufentanil, and remifentanil; they are of favor prior to the last 5 years) is the occur- administered for various indications and at a rence of emergence delirium and dissociative wide variation in dosing regimens. All intra- hallucinations. Additionally, increase in venous opioids have almost no effect on intracranial pressure from enhanced cerebral NIOM even at very high doses, making them blood flow due to ketamine makes it of lim- of essential importance during anesthesia for ited use in neurosurgical patients with procedures requiring NIOM. Even when given intracranial hypertension as well as in some in the epidural or intrathecal space, they have other patient populations. Ketamine has been minimal effect on EPs (2). It has been noted found particularly useful as a low-dose infu- that generous application of opioids can result sion (10 to 20 µg/kg/min) to supplement a in improved MEP monitoring owing to the propofol/opioid TIVA technique in proce- reduction of spontaneous muscle contraction dures that require anesthetic-sensitive NIOM and lowering of the MAC for other anesthetic (e.g., MEP). The addition of low-dose keta- agents. With regards to the EEG, opioids pro- mine to a propofol-based TIVA allows for a duce a mild slowing into the delta range with- substantial reduction in propofol infusion out effect on amplitude. Opioids will not doses and enhancement of EP responses while produce burst suppression or an isoelectric minimizing the undesirable side effects of ket- EEG even at the highest doses. Of particular amine. For procedures requiring NIOM tech- importance, the development of remifentanil niques that are highly sensitive to the effects has revolutionized opioid use in TIVA. of anesthetics (e.g., transcranial MEP), the Remifentanil is an ultra-short-acting opioid use of ketamine in the anesthetic armamen- with a half-life on the order of 5 minutes tarium should be considered. regardless of dose. This allows for very rapid titration of analgesia with little or no effect on Etomidate emergence times, thus permitting high-dose Etomidate represents another contradic- opioid TIVA to minimize the dose of an asso- tion to the general rule that anesthetic agents ciated sedative-hypnotic. cause EP depression. Induction doses and con- tinuous intravenous infusion enhance both Ketamine MEP and SEP recordings (16). Etomidate has Ketamine is one of the older anesthetic been used in the past as a component of TIVA agents and has undergone a recent resurgence during procedures that require anesthetic-sen- of use owing to the finding that it helps to sitive NIOM (e.g., transcranial MEPs). alleviate postoperative pain and chronic pain Etomidate is also somewhat contradictory in states. Ketamine influences a variety of recep- its EEG effects; at low doses it may be some- tors and has the unique characteristic among what proconvulsant, and it is occasionally anesthetic agents of enhancing EP responses, used for ECT or epilepsy surgery; although at especially in the cortex and spinal cord (15). higher doses it may produce burst suppres- Whether given as single bolus at induction or sion. However, among its many unpleasant as continuous infusion, ketamine can increase side effects, concerns have been raised regard- EP amplitude in SEP, MEP, and BAEP record- ing etomidate-induced adrenal suppression, ing, making it an attractive agent for use dur- which can occur with even single-bolus induc-
Husain 04 1/17/08 11:55 AM Page 63 CHAPTER 4: Anesthetic Considerations • 63 tion doses (0.2 to 0.5 mg/kg). Increased mor- nerve stimuli. MEP monitoring is acceptable tality has been seen with prolonged infusion when neuromuscular blockade is maintained of etomidate, mainly in the intensive care set- at a TOF of two responses. In using MEP ting (17). Nevertheless, etomidate remains monitoring, it is important for the neuro- valuable in cases where NIOM responses are physiologist and surgeon to know whether difficult to obtain and otherwise may not be the patient is paralyzed. If the patient is not recordable. paralyzed, MEP stimulation must be done at times when patient movement is acceptable. Dexmedetomidine If the patient is paralyzed, there are likely to Dexmedetomidine is a relatively new be brief periods when MEP responses are not agent used in human anesthesia. This selective recordable owing to intense paralysis; it is alpha-2 agonist has seen widespread use in then imperative to communicate when a neu- veterinary medicine and has found its way romuscular blocking agent is redosed. into intensive care units and operating rooms However, either practice, paralysis or not, is because of its desirable effects of sedation, acceptable; the main principle is, again, effec- analgesia, and sympatholysis without respira- tive and open communication with all parties tory depression. Though increasing ancedotal in the surgical suite. reports are emerging, there are limited data on the effects of dexmedetomidine on NIOM; however, animal data suggest that there is lit- ANESTHETIC TECHNIQUES tle effect (18). It may be used as a low-dose infusion (0.2 to 0.5 µg/kg/hr) to augment any A variety of anesthetic techniques are anesthetic technique, and it allows for the use acceptable for use during NIOM; the type of of considerably less volatile or intravenous anesthetic should be tailored to the type of anesthesthesia or opioid. Its definitive role in NIOM and the requirements of the surgical anesthetic techniques for highly sensitive procedure. There are, however, a few general NIOM remains to be determined. principles. First, the least amount of anes- thetic agent necessary should be utilized as long as there is little possibility of awareness Paralytics or discomfort on the part of the patient. The Neuromuscular blockers exert their effect liberal use of opioids can allow for a signifi- by blocking acetylcholine at the nicotinic cant decrement in MAC. Second, the more receptor in the neuromuscular junction. They stable an anesthetic dose can remain for the have no effect on monitoring modalities that duration of the case, the less likely that the are not derived from muscle activity (e.g., anesthetic agent might be contributing to EEG, BAEPs, and SEPs). They will com- intraoperative changes in NIOM waveforms. pletely negate MEP monitoring if intense neu- Supplementation of baseline anesthetic drugs romuscular blockade is utilized. However, with opioids or less NIOM-offending agents employing partial blockade will allow sub- can be made at times of more intense surgical stantial reduction in patient movement with stimulation. Overall, there are essentially four testing, improved surgical retraction, and classes of NIOM based on how easily the favorable MEP monitoring. There are many monitoring technique is ablated by anesthetic ways to monitor the amount of neuromuscu- agents. As the relative sensitivity of NIOM to lar blockade; the most common is the “train anesthesia increases, the anesthetic technique of four” (TOF) technique. It consists of meas- should be adjusted to maximize the least uring muscle responses, or compound muscle offending agents. Each group and its anes- action potentials, after four 2-Hz peripheral thetic implications are discussed below.
Husain 04 1/17/08 11:55 AM Page 64 64 • S E C T I O N I : B a s i c P r i n c iples Relative Insensitivity Sensitivity to Anesthetics without Sensitivity to Paralysis NIOM that is relatively insensitive to anesthetic agents in general includes BAEPs NIOM that is not negated by neuromuscu- and SEPs recorded from the epidural space. lar blockade but is sensitive to anesthetic agents With these monitoring methods, nearly all includes SEP monitoring. Care must be taken anesthetic practices can be used with the to minimize offending anesthetic agents and understanding that the general objective is to optimize non-anesthetic variables (i.e., temper- maintain a constant level of anesthesia supple- ature). Generally, volatile or intravenous anes- mented with intermittent opioid dosing to thesia is acceptable if relatively low doses are control increased surgical stimulus. Of course maintained (0.5 MAC for anesthetic gases or the least amount of anesthetic necessary to less than 80 µg/kg/min of propofol). The use of ensure amnesia and analgesia should be used. neuromuscular blockade in this situation Generally all patients have baseline EPs, so allows for a modest decrement in anesthetic that once in the operating room, deviation dose, as patient movement and relaxation then from that baseline can be assessed. If needed, become improbable. However, care must be anesthetic level or technique can then be taken that anesthetic dose is not so low as to adjusted to refine NIOM recordings. permit patient recall or discomfort. Sensitivity to Paralytics Relative Sensitivity Forms of NIOM that are sensitive to neu- The need for MEP monitoring can initiate romuscular blockade include all monitoring some of the more challenging anesthetic issues. that requires elicitation of muscle action Designing an anesthetic technique to optimize potentials (i.e., electromyography, MEP, MEP monitoring adds to an already complex spinal reflex testing, etc.). For these cases, if surgical procedure. A TIVA technique that lim- very fine control of the amount of neuromus- its the amount of sedative-hypnotic agent (i.e., cular blockade can be maintained through propofol, barbiturate) is usually required. vigilant monitoring and drug dosing, neuro- Limiting the dose of sedative-hypnotic to muscular blocking agents can be employed. allow for optimal response recording of Otherwise they should be entirely avoided NIOM requires the use of a second agent, usu- once the patient has been intubated. In fact, ally opioid, to supplement and augment the there are some practices that utilize intraoper- anesthetic properties. For instance, using a ative neuromuscular blockade reversal when propofol-based anesthetic requires the addi- critical monitoring periods approach. In gen- tion of opioid, ketamine, or dexmedetomidine eral, with the exception of MEP recording, infusion to allow a much smaller dose of which is exquisitely sensitive to anesthetic propofol to be administered. Additionally, if technique, other forms of anesthetic agents neuromuscular blockade is used, it must be are acceptable. For cases that rely on an tightly controlled so that profound paralysis unparalyzed patient, relatively “deep” anes- does not preclude MEP responses from the thesia (e.g., high doses of anesthetic agents) muscles. It is not uncommon for the patient to can be used to offset lack of patient paralysis, be unparalyzed during critical monitoring por- allowing optimal surgical conditions of immo- tions of the procedure. Therefore the anesthe- bility and relaxation while maintaining the siologist is often faced with an unparalyzed integrity of NIOM. However, the general patient, whose monitoring requires relatively principle of stable, though relatively high, low doses of an anesthetic, and whose airway anesthetic dose should be maintained. and accessibility is often remote. One current
Husain 04 1/17/08 11:55 AM Page 65 CHAPTER 4: Anesthetic Considerations • 65 practice is to utilize high-dose remifentanil in wave V significant? Neurology 2005;65: infusion to supplement a low-dose propofol- 1551–1555. ketamine based anesthetic. This allows very 2. Sloan TB. Evoked potentials In: Albin MS, ed. low dose propofol (20 to 30 µg/kg/min), which A Textbook of Neuroanesthesia with has minimal effects on MEP responses, to be Neurosurgical and Neuroscience Perspectives. New York: McGraw-Hill, 1997:221–276. offset by low-dose ketamine (10 to 20 3. Stekker MM, Escherich A, Patterson T, et al. µg/kg/min), which enhances MEP responses, Effects of acute hypoxemia/ischemia on EEG and an amount of remifentanil that keeps the and evoked responses at normothermia and patient motionless and relaxed. hypothermia in humans. Med Sci Monit 2002;8:CR223–CR228. 4. Dolan EJ, Transfeldt EE, Tator CH, et al. The CONCLUSIONS effect of spinal distraction on regional spinal cord blood flow in cats. J Neurosurg 1980;53: In developing an anesthetic plan, the type 756–764. of NIOM is often as important a consideration 5. Grundy BL, Heros RC, Tung AS, et al. as the type of surgical procedure. The crucial Intraoperative hypoxia detected by evoked factor for a successful procedure is open and potential monitoring. Anesth Analg 1981;60: candid communication between the operating 437–439. room staff, neurophysiologist, anesthesiologist, 6. Nagoa S, Roccaforte P, Moody RA. The and surgeon. The majority of problems with effects of isovolemic hemodilution and reinfu- sion of packed erythrocytes on somatosensory intraoperative monitoring arise when operating and visual evoked potentials. J Surg Res 1978; room communication does not allow for each 25:530–537. individual to have a clear understanding of the 7. Mackey JR, Hall JW III. Sensory evoked actions of each of the other members. When responses in head injury. Central Nerv Syst everyone involved in the procedure is knowl- Trauma 1985;2:187–206. edgeable about reasonable expectations and 8. McGirt MJ, Woodworth GF, Brooke BS, et al. aware of the current situation, the patient bene- Hyperglycemia independently increases the fits from an operating team that is poised and risk of perioperative stroke, myocardial infarc- fluid in its execution. With that understanding, tion, and death after carotid endarterectomy. it is imperative for the neurophysiologist to Neurosurgery 2006;58:1066–1073. understand the limitations produced by an 9. Gugino LD, Aglio LS, Segal NE, et al. Use of anesthetic and for the anesthesiologist to under- transcranial magnetic stimulation for monitor- stand the effects of certain medications on mon- ing spinal cord motor paths. Semin Spine Surg itoring. Without that fundamental knowledge, 1997;9:315–336. 10. Deletis V. Intraoperative monitoring of the there can be little coordinated activity between functional integrity of the motor pathways. the two parties, resulting in ineffective monitor- Adv Neurol 1993;63:201–214. ing. However, with the knowledge of the basic 11. Drummond JC, Todd MM, U HS. The effects effect of a given anesthetic agent on monitoring of high dose sodium thiopental on brainstem modalities, nearly any anesthetic technique can auditory and median nerve somatosensory be administered safely and effectively with all evoked responses in humans. Anesthesiology types of monitoring. 1985;63:249–254. 12. Sloan TB, Fugina ML, Toleikis JR. Effects of midazolam on median nerve somatosensory evoked potentials. Br J Anaesth 1990;64: REFERENCES 590–593. 1. James ML, Husain AM. Brainstem auditory 13. Kalkman CJ, Drummond JC, Ribberrink AA. evoked potential monitoring: when is change Effects of propofol, etomidate, midazolam,
Husain 04 1/17/08 11:55 AM Page 66 66 • S E C T I O N I : B a s i c P r i n c iples and fentanyl on motor evoked responsesto 16. Kochs E, Treede RD, Schulte AM, Esch J. transcranial electrical or magnetic stimulation Increase in somatosensory evoked potentials in humans. Anesthesiology 1992;76:502–509. during induction of anaesthesia with etomi- 14. Angel BA, LeBeau F. A comparison of the date. Anaesthetist 1986;35:359–364. effects of propofol with other anesthetic 17. Ledingham IM, Watt I. Influence of sedation agents on the centripetal transmission of sen- on mortality in critically ill multiple trauma sory information. Gen Pharmacol 1992;23: patients. Lancet 1983;1:1270. 945–063. 18. Li BH, Lohmann JS, Schuler HG, Cronin AJ. 15. Schubert A, Licina MG, Lineberry PJ. The Preservation of the cortical somatosensory- effect of ketamine on human somatosensory evoked potential during dexmedetomidine evoked potentials after high frequency repeti- infusion in rates. Anesth Analg 2003;96: tive electrical stimulation. Electroencephalogr 1150–1160. Clin Neurophysiol 1998;108:175–181.