Anesthesia for Intraoperative Neurophysiologic Monitoring of ...

Anesthesia for Intraoperative Neurophysiologic Monitoring ofthe Spinal Cord

*Tod B. Sloan and †Eric J. Heyer

*University of Texas Health Science Center at San Antonio, Texas; and †Columbia University, New York, New York, U.S.A.

Summary: Intraoperative neurophysiologic monitoring (INM) using somatosensoryand motor evoked potentials (MEPs) has become popular to reduce neural risk and toimprove intraoperative surgical decision making. Intraoperative neurophysiologicmonitoring is affected by the choice and management of the anesthetic agents chosen.Because inhalational and intravenous anesthetic agents have effects on neural synapticand axonal functional activities, the anesthetic effect on any given response willdepend on the pathway affected and the mechanism of action of the anesthetic agent(i.e., direct inhibition or indirect effects based on changes in the balance of inhibitoryor excitatory inputs). In general, responses that are more highly dependent on synapticfunction will have more marked reductions in amplitude and increases in latency as aresult of the synaptic effects of inhalational anesthetic agents and similar effects athigher doses of intravenous agents. Hence, recording cortical somatosensory evokedpotentials and myogenic MEPs requires critical anesthetic choices for INM. Themanagement of the physiologic milieu is also important as central nervous systemblood flow, intracranial pressure, blood rheology, temperature, and arterial carbondioxide partial pressure produce alterations in the responses consistent with the supportof neural functioning. Finally, the management of pharmacologic neuromuscularblockade is critical to myogenic MEP recording in which some blockade may bedesirable for surgery but excessive blockade may eliminate responses. A close workingrelationship of the monitoring team, the anesthesiologist, and the surgeon is key to thesuccessful conduct and interpretation of INM. Key Words: Anesthesia—Motorevoked potentials—Somatosensory evoked potentials—Muscle relaxation.

Sensory evoked potentials can be elicited during sur-gery and can be used for intraoperative neurophysiologicmonitoring. However, they differ in their sensitivity toanesthetic agents depending on the neurologic pathwaysinvolved and the specific anesthetic agents used (Kingand Sloan, 2001; Moller, 1995; Nuwer, 1986; Sloan1997, 2002b). For example, visual evoked potentials areextremely sensitive to interference by anesthetics,whereas brainstem auditory evoked potentials are essen-tially unaffected by anesthetics. Somatosensory evoked

potentials (SSEPs) are intermediate in sensitivity, andtranscranially elicited motor evoked potentials (MEPs)are particularly challenging because of marked interfer-ence by anesthetic agents.Injury to the central nervous system can occur at

multiple levels and may involve multiple pathways. Be-cause the spinal MEP and SSEP pathways are served bydifferent arterial supplies (the anterior and posteriorspinal arteries respectively), it is possible that injury tothe motor tract can occur during spinal surgery whenmonitoring with SSEPs alone. Because of this possibil-ity, MEPs were developed to monitor motor pathways ofthe anterior spinal cord. However, it is unusual for motortract injury to arise when SSEPs remain unchanged(Nuwer et al., 1995). To assess central nervous system

Address correspondence and reprint requests to Dr. Tod Sloan,Anesthesiology 7838, University of Texas Health Science Center, 7703Floyd Curl Drive, San Antonio, Texas, USA 78284-3900; e-mail:[email protected].

Journal of Clinical Neurophysiology19(5):430–443, Lippincott Williams & Wilkins, Inc., Philadelphia© 2002 American Clinical Neurophysiology Society

430

integrity more completely, both SSEPs and MEPs aremonitored together. However, monitoring the spinal cordwith MEPs is particularly challenging because anestheticagents easily abolish the potentials elicited.The purpose of this review is to describe the anes-

thetic issues that must be considered when usingMEPs to monitor the central nervous system. Wedescribe the anesthetic effects, physiologic effects,and then the anesthetic considerations required tooptimize MEP monitoring. Because SSEPs are almostalways monitored in conjunction with MEPs (Legatt,2002), the effects of anesthetics on these potentials arealso discussed.

OVERVIEW OF ANESTHETIC EFFECTS ONSENSORY AND MOTOR PATHWAYS

Mechanism of Action of Anesthetic Agents

The impact of anesthetic agents on neurophysio-logic monitoring increases with the number of syn-apses in the pathway monitored because all anestheticagents produce effects by altering neuronal excitabil-ity through changes in synaptic function or axonalconduction (Sloan, 2002a). An increased number ofsynapses in the pathway may account for why corti-cally generated evoked potentials are more sensitive toanesthetics than subcortical signals such as P14, N18/P31, N34 signals. These differences may be importantespecially when there is a loss of cortical signalswithout changes in their subcortical counterparts. Byexamining the specific anesthetic agents, we can gainsome insight into their mechanisms of effect andtherefore methods for the choice of agents when mon-itoring is performed.All anesthetics are either gaseous (inhaled) or dissolvable

in fluids (intravenous). The gaseous agents are either “po-tent” inhalational agents like halothane, enflurane, isoflu-rane, sevoflurane, and desflurane, which are effective atrelatively low concentrations (!10%), or the less “potent”inhalational agent nitrous oxide, which is used at higherconcentrations (usually more than 50%). These potent in-halational agents produce their actions either by alteringspecific receptors (Flood and Role, 1998) or through non-specific effects on cell membranes that alter the conforma-tion structure of the receptor or ion channel at the protein–lipid interface (Sloan, 2002a). For years it has been well-known that the potency of volatile anesthetics varies withlipophilicity, suggesting that the mechanism of anesthesiadepends on changes in the membranes of tissue such asalterations in synaptic function.

The intravenous sedative agents used in anesthesia inter-act at a number of different receptors. For example, barbi-turates, etomidate, althesin, propofol, and benzodiazepinesact at specific receptors, primarily by enhancing the inhib-itory effects of gamma-aminobutyric acid. These agents areknown to bind specifically to the gamma-aminobutyricacid-a receptor, where activation increases chloride conduc-tance, hyperpolarizing the membrane and producing synap-tic inhibition (Sloan, 2002a).Some intravenous agents also work by blocking the

excitatory effects of glutamic acid (Macdonald andBarker, 1979; Sloan, 2002a). There are now at least fourknown types of receptors in this system named by thebinding of specific substances: N-methyl-D-aspartate,!-amino-3-hydroxy-5-methyl-isoxazole propionic acid,quisqualate, and kainate. Ketamine appears to have itsmajor action by inhibiting the N-methyl-D-aspartate re-ceptor, thereby reducing sodium flux and intracellularcalcium levels (O’Shaughnessy and Lodge, 1988). Sim-ilarly, most anesthetic agents interact at multiple recep-tors in numerous pathways such that the effect on evokedresponses likely varies with the spectrum of specificreceptors and pathways affected. For example, barbitu-rates upregulate the N-methyl-D-aspartate receptor andinteract competitively with some of the other bindingsites (Sloan, 2002a).Some anesthetics also activate opioid receptors (mu,

kappa, and delta). Opioids depress electroexcitability byincreasing inward K " current and depressing outwardNa " current via a G-protein mechanism linking thereceptors to the ion channels (Nestler and Aghajanian,1997). This mechanism is distinct from that of volatileand intravenous anesthetics and may explain the fact thatopioids are unlike any other agent.Finally, anesthetic agents act at the “n-type” acetyl-

choline receptors found at the neuromuscular junction.This is the primary site of action of the neuromuscularblocking agents. The potent inhalational anesthetics alsointeract with this receptor at the neuromuscular junction;however, their effect appears to be minimal for alteringelectrophysiologic recordings.The effects of anesthetic agents are thus the result of

direct inhibition of synaptic pathways or the resultof indirect action on pathways by changing the balanceof inhibitory or excitatory influences. Thus, the specificelectrophysiologic responses that rely extensively onsynaptic function will be most susceptible to anesthetics.Although most anesthetics depress evoked response am-plitude and increase latency, some anesthetic agents(etomidate and ketamine) enhance both SSEP and MEPamplitudes, perhaps by attenuating inhibition (Sloan,1997).

431ANESTHESIA AND MEP

J Clin Neurophysiol, Vol. 19, No. 5, 2002

Therefore, anesthetic agents will affect pathways con-taining synapses and this effect will be more prominentas the number of synapses increases. For this reason,SSEP responses generated in cortical structures will bethe more affected (i.e., more sensitive to many anesthet-ics than those generated at more peripheral sites such asthe spinal cord or brainstem, where fewer synapses areinvolved) (Sloan, 1997) (Figs. 1 through 3). Also for thesame reason, effects of anesthetic agents on EEG parallelthe effects of anesthesia on cortical evoked potentialsbecause both depend on cortical synapse activity (Sloan,1999). Because most anesthetic drugs produce a dose-dependent depression of the EEG, they produce de-creased amplitude and increased latency of corticalSSEPs (Sloan, 1997).The motor pathways and MEPs are susceptible to

anesthetic agents at three sites. The first is in the motorcortex. Stimulation of neurons associated with move-ment, such as the pyramidal cells, is either by directstimulation of these cells, which leads to the productionof “D waves,” or indirect stimulation via internuncialneurons, which leads to production of “I waves” (Day etal., 1987b, 1989). Both of these responses are seen inepidural recordings from the epidural space in the spine(Fig. 4). Direct stimulation can be accomplished bytranscranial electrical or intense magnetic stimulation.The D waves produced are relatively unaffected byanesthetics because no synapses are involved in their

production or propagation down the spinal cord. I wavescan be generated by either transcranial electrical ortranscranial magnetic stimulation. They require synapticactivity for their production because they arise fromstimulation of internuncial neurons. Therefore, they areaffected markedly by anesthetics.The second site is in the anterior horn cell. At this

location the D and I waves summate. If they are able tobring the cell to threshold, the anterior horn cell depo-

FIG. 1. Cortical somatosensory evoked potential recordings taken in aketamine- and isoflurane-anesthetized baboon after median nerve stim-ulation. Readily apparent is the marked amplitude reduction and in-crease in latency of the prominent response at 20 to 25 msec withincreasing end-tidal concentrations of isoflurane. Also notable is thenoise in the response at 0.3% from muscle activity with light anesthesiain the absence of pharmacologic muscle relaxation. Recordings takenusing a Biologic Navigator (Biologic Inc., Mundelein, IL, USA) fromC3' to Fz using an amplification of 75,000 and a bandpass filtration of10 to 500 Hz. A total of 250 stimuli at 5.7 Hz were averaged for eachwaveform.

FIG. 2. Somatosensory evoked potential responses recorded over thecervical spine in a ketamine- and isoflurane-anesthetized baboon aftermedian nerve stimulation. Notable is the lack of effect on amplitudeand latency on the major peak at approximately 6 msec that is seen inthe cortical recordings (see Fig. 1). The same recording parameters asFig. 1 were used.

FIG. 3. Somatosensory evoked potential responses recorded by anepidural electrode at the same time as the responses acquired in Fig. 1.Notable is the lack of effect on amplitude and latency on the peaksoccurring at 10 to 15 msec that is seen in the cortical recordings.Recordings were taken using a Biologic Navigator (Biologic Inc.,Mundelein, IL, USA) from a midthoracic bipolar epidural recordingelectrode placed percutaneously using an amplification of 2,000 and abandpass filtration of 3 to 1,500 Hz. A total of 250 stimuli at 5.7 Hzwere averaged for each waveform.

432 T. B. SLOAN AND ERIC J. HEYER


larizes, producing a peripheral nerve action potential.Stimulation of multiple neurons produce a compoundmuscle action potential (CMAP; Fig. 5). Partial synapticblockade at the anterior horn cell, induced by anesthetics,may make it more difficult to reach threshold (Sloan,2002b). The combined effect of anesthetics in blockingthe generation of I waves in the cortex, and synaptictransmission at the spinal cord anterior horn cell reducefurther the probability of generating a CMAP. At higheranesthetic doses, an even more profound synaptic blockat the anterior horn cell may inhibit synaptic transmis-sion regardless of the composition of the descendingspinal cord volley of activity. These effects are greatestwhen transcranial magnetic stimulation is used to gener-ate motor responses and are less with transcranial elec-trical stimulation because of the relative dependence ofthe former on internuncial synaptic depolarization.The third site is at the neuromuscular junction. Fortu-

nately, with the exception of neuromuscular blockingagents, anesthetic drugs have little effect at the neuro-muscular junction.

Action of Specific Anesthetic Agents

Gaseous Agents: Halogenated Inhalational Agents

Perhaps the most common anesthetics in use today arethe halogenated inhalational agents (desflurane, enflu-

rane, halothane, isoflurane, and sevoflurane). All halo-genated inhalational agents produce a dose-related in-crease in latency and reduction in the amplitude of thecortically recorded SSEPs. This effect varies from isoflu-rane (most potent), to enflurane (intermediate), to halo-thane (least potent) (Sloan, 1997). Studies with sevoflu-rane and desflurane suggest that they are similar toisoflurane at a steady state, but because of their morerapid onset and offset of effect (both are more insolublethan isoflurane), they may appear to be more potentduring periods when concentrations are increasing.Halogenated inhalational agents may reduce SSEP

amplitude and prolong latency recorded from the cortexto a greater degree than when these responses are re-corded from electrodes located at the cervicomedullaryjunction, spinal cord, and peripheral nerve.Motor evoked potentials recorded in muscle (myogen-

ic) are abolished easily by halogenated inhalationalagents (see Fig. 5). Single-pulse stimulation myogenictranscranial MEPs appear to be so easily abolished byinhalational agents that they are often unrecordable in thepresence of these agents. When recordable, MEPs mayoccur only at low concentrations (e.g., #0.2 to 0.5%isoflurane). This effect of inhalational agents is likely theresult of depression of synaptic transmission either in theanterior horn cell synapses on motor neurons or in thecortex on the internuncial synapses with a loss of Iwaves. Two observations support the effect of anesthet-ics on motor neurons. Halogenated inhalational agents

FIG. 5. Motor evoked potential responses recorded from the hypothenarmuscles at the same time as the responses acquired in Fig. 4. Readilyapparent is the marked amplitude reduction and increase in latency of theprominent response at approximately 14 msec with increasing end-tidalconcentrations of isoflurane. Recordings taken using a Biologic Navigator(Biologic Inc., Mundelein, IL, USA) from subdermal needle electrodesplaced percutaneously in the hypothenar space using an amplification of5,000 and a bandpass filtration of 3 to 1,500 Hz. A single supramaximalstimulus was used at Cz to Fz using a Digitimer D180A.

FIG. 4. Motor evoked potential responses recorded by an epiduralelectrode in a ketamine- and isoflurane-anesthetized baboon aftertranscranial electrical stimulation. Present is a D wave at approximately4 msec followed by a series of I waves. Notable is the lack of effect onthe amplitude or latency of the D wave and the reduction in number,amplitude, and latency of the I waves as the end-tidal isofluraneconcentration is increased. Recordings taken using a Biologic Naviga-tor (Biologic Inc., Mundelein, IL, USA) from a midthoracic bipolarepidural recording electrode placed percutaneously using an amplifi-cation of 2,000 and a bandpass filtration of 3 to 1,500 Hz. A singlesupramaximal stimulus was used at Cz to Fz using a Digitimer D180A.



reduce the H-reflex (Mavroudakis et al., 1994). The Dwaves recorded in the epidural space, before synapsingin either the anterior horn or the neuromuscular junction,are highly resistant to the effects of inhalational agentsand are easily recordable at high, volatile anestheticconcentrations (Gugino et al., 1997) (see Fig. 4).Because the D wave is resistant to anesthetic depres-

sion, the anesthetic effect at the anterior horn cell can beovercome at low anesthetic concentrations by high-fre-quency (multiple-pulse) transcranial stimulation (trainsof stimuli with interstimulus intervals of 2 to 5 msec)(Kawaguchi et al., 1996; Pechstein et al., 1998). Underthese circumstances the multiple D waves (and I waves ifproduced) summate at the anterior horn cell to generatea peripheral nerve action potential and subsequent myo-genic response (Figs. 6 and 7). However, the best anes-thetic plan is to avoid inhalational agents for optimalMEP monitoring even when using high-frequency stim-ulation technique (Pechstein et al., 1998).Studies with direct spinal or epidural stimulation show

minimal effects of anesthesia on neurogenic or myogenicresponses (Nagle et al., 1996). However, this stimulationactivates both sensory and motor pathways. Therefore,the concern is that the antidromically activated actionpotentials in the sensory fibers would reflexly stimulatemotor neurons at lower spinal levels, bypassing thedescending motor tracts, but producing a myogenic re-

sponse nevertheless. Mochida et al. (1995) studied theresponses in the peripheral nerve and muscle after epi-dural stimulation in the cat and found that the type ofspinal cord stimulation and the anesthetic used may alterthe balance of sensory and motor contributions to theperipheral nerve and muscle response from direct spinalstimulation. Recent studies suggest that with isofluraneanesthesia, the motor component is blocked preferen-tially, perhaps by interaction at the synapses in theanterior horn cell or by differential effects on conductionin the spinal tracts in humans (Deletis, 1993). Thesestudies do not, however, clearly allow a recommendationof anesthesia that will promote preferentially the moni-toring of motor pathways with direct spinal, epidural, orperispinal stimulation. For this reason, intraoperativeneurophysiologic monitoring methods that rely on stim-ulation of the spinal cord may not monitor the motorpathways reliably.

Nitrous Oxide

Nitrous oxide reduces SSEP cortical amplitude andincreases latency when used alone or when combinedwith halogenated inhalational agents or opioid agents.When compared at equipotent anesthetic concentrations,nitrous oxide produces more profound changes in corti-cal SSEPs and muscle recordings from transcranial mo-

FIG. 6. Motor evoked potential responses recorded by an epiduralelectrode in a ketamine- and isoflurane-anesthetized baboon aftertranscranial electrical stimulation. Note that the effect of adding asecond stimulation pulse to the cortex increases the number of waves inthe epidural recording. Recordings taken using a Biologic Navigator(Biologic Inc., Mundelein, IL, USA) from a midthoracic bipolar epi-dural recording electrode placed percutaneously using an amplificationof 2,000 and a bandpass filtration of 3 to 1,500 Hz. A single- ortwo-pulse (interstimulus interval, 2 msec) supramaximal stimulus wasused at Cz to Fz using a Digitimer D185.

FIG. 7. Motor evoked potential responses recorded from the hypothe-nar muscles from a baboon with 1% isoflurane anesthesia when noMEP response is present when only one cortical stimulation pulse isused. A small response is present with two pulses and a substantialresponse occurs with three or more pulses. Recordings taken using aBiologic Navigator (Biologic Inc., Mundelein, IL, USA) from amidthoracic bipolar epidural recording electrode placed percutaneouslyusing an amplification of 2,000 and a bandpass filtration of 3 to 1,500Hz. A single- or multiple-pulse (interstimulus interval, 2 msec) supra-maximal stimulus was used at Cz to Fz using a Digitimer D180A.



tor stimulation than any other inhalational anestheticagent (Sloan, 1997). Like halogenated agents, effects onsubcortical and peripheral sensory responses and onepidurally recorded MEPs are minimal.Despite the depressant effect of nitrous oxide, it has

been used with all types of evoked potential recordings,particularly when combined with opioids (“nitrous–nar-cotic” anesthetic technique). Nitrous oxide may be “con-text sensitive” in that the actual effects may vary depend-ing on the other anesthetics already present. As withsevoflurane and desflurane, nitrous oxide is relativelyinsoluble, such that anesthetic effects can change rapidlywhen concentrations are varied intraoperatively.

Intravenous Anesthetic Agents

Most anesthesia techniques use a mixture of differentanesthetic agents. For example, inhalational agents areoften supplemented with opioids or intravenous seda-tives (e.g., benzodiazepines, etomidate, droperidol, orpropofol). If the inhalational agents need to be avoidedcompletely, then analgesic and sedative intravenousagents can be combined to produce a total intravenousanesthetic (TIVA).

Intravenous Analgesic Agents: Opioids

The effects of the opioid analgesics (alfentanil, fenta-nyl, remifentanil, and sufentanil) on SSEPs and MEPsare less than with inhalational agents, making opioidsimportant components of anesthetic planning for evokedpotential monitoring. They produce minimal changes inspinal or subcortical SSEP recordings. There is somedepression of amplitude and an increase of latency in thecortical responses, especially for the late cortical peaks(latencies more than 100 msec). As with systemic opi-oids, the spinal application of morphine or fentanyl forpostoperative pain management produces minimalchanges in the SSEP and fails to alter the H-reflex(Sloan, 1997). Studies with myogenic responses fromMEPs with electrical and magnetic stimulation showonly mild amplitude decreases and latency increases withthe opioids. In fact, fentanyl may reduce backgroundspontaneous muscle contractions and associated motorunit potentials, further improving myogenic responses.

Special Intravenous Analgesic Agents: Ketamine

The effects of ketamine on evoked responses alsodiffer from those of inhalational agents. Ketamine can

enhance cortical SSEP amplitude and MEP amplitude inmuscle and spinal recorded responses after spinal stim-ulation (Kano and Shimoji, 1974; Schubert et al., 1990).This latter effect on muscle responses may be mediatedby the same mechanisms that potentiate the H-reflex.However, effects on subcortical and peripheral SSEPresponses are minimal, as are the effects on myogenicMEPs (Glassman et al., 1993). Because of these effects,ketamine is a desirable agent for monitoring responsesthat are usually difficult to record under anesthesia (e.g.,myogenic MEPs). However, increases in intracranialpressure with intracranial pathology and hallucinationsmay limit the use of ketamine in certain patients.

Sedative–Hypnotic Drugs

Intravenous sedative agents are used frequently toinduce or supplement opioid- and ketamine-based anal-gesia for TIVA to ensure adequate sedation, anxiolysis,and amnesia. Although ketamine produces some disso-ciative effects in addition to analgesia, supplementationof ketamine with sedative medications can reduce therisk of hallucinations.

Droperidol

When combined with fentanyl (“neurolept anesthe-sia”), droperidol appears to have minimal effects onmyogenic MEPs (Sloan, 1997). Recent concerns aboutprolonged QT changes in the electrocardiogram andarrhythmias may reduce the use of this drug.

Barbiturates

Thiopental remains a popular drug for induction ofanesthesia, although transient decreases in amplitude andincreases in latency of cortical response occur immedi-ately after induction. Longer latency cortical waves aremost affected, whereas minimal effects are seen on thesubcortical and peripheral responses. Studies with an-other barbiturate, phenobarbital, demonstrate that theSSEP is virtually unaffected at doses that produce coma(Drummond et al., 1985), and changes are not seen untildoses are sufficient to produce cardiovascular collapse(Sloan, 1997). Barbiturates are not used commonly dur-ing recording of MEPs because the CMAP responses areunusually sensitive to barbiturates. Furthermore, the ef-fect appears to be quite prolonged. In one study, theinduction bolus eliminated the CMAP from the MEP fora period of 45 to 60 minutes (Glassman et al., 1993).



Benzodiazepines

Midazolam has desirable properties of amnesia andhas been used for monitoring cortical SSEPs. Midazo-lam, in doses consistent with induction of anesthesia (0.2mg/kg) and in the absence of other agents, produces amild depression of cortical SSEP (Sloan et al., 1990) andminimal effects on subcortical and peripheral compo-nents. As with thiopental, midazolam produces pro-longed, marked depression of MEPs, suggesting that italso may be a poor induction agent for MEP recording.

Etomidate

Like ketamine, etomidate increases the amplitude ofcortical SSEP components after injection withoutchanges in subcortical and peripheral sensory responses(Sloan, 1997). This amplitude increase appears coinci-dent with myoclonus seen after administration of thedrug, suggesting a heightened cortical excitability. Asustained amplitude increase with constant drug infusionhas been used to enhance SSEP cortical recordings thatotherwise could not be monitored (Sloan, 1988) and toenhance amplitude in MEPs (Kochs et al., 1986). Studieswith MEPs have suggested that etomidate is an excellentagent for induction and monitoring of CMAP responses.Of several intravenous agents studied, etomidate had theleast degree of amplitude depression after induction

doses or continual intravenous infusion. Thus, etomidatehas been used for induction of anesthesia and as acomponent of TIVA, combined with opioids.

Propofol

Propofol induction produces amplitude depression incortical SSEPs with rapid recovery after termination ofinfusion (Kalkman et al., 1992b). When the SSEP isrecorded in the epidural space, propofol has no notableeffect (Angel and LeBeau, 1992). Studies with electricalor magnetic elicited MEPs have demonstrated a depres-sant effect on response amplitude, consistent with acortical effect (Kalkman et al., 1992a) (Fig. 8). Althoughpropofol does not appear to enhance cortical responses,rapid metabolism allows rapid adjustment of the depth ofanesthesia and effects on evoked responses. As a com-ponent of TIVA, infusions of propofol have been com-bined with opioids and have produced acceptable condi-tions for monitoring of cortical SSEPs and myogenicMEPs (Calancie et al., 1998; Pechstein et al., 1998).Because of this ability to titrate the dose rapidly, it hasbecome an extremely popular agent in TIVA.

Muscle Relaxants

Because muscle relaxants have their major site ofaction at the neuromuscular junction, they have littleeffect on electrophysiologic recordings such as SSEPs,which do not derive from muscle activity. However,profound neuromuscular blockade will prevent recordingof the CMAP during myogenic MEP monitoring. Fur-thermore, partial neuromuscular blockade has the benefitof reducing a substantial portion of patient movement,which accompanies the testing, and facilitates surgicalprocedures when muscle relaxation is needed for retrac-tion of tissues.Two methods are used customarily to assess the de-

gree of neuromuscular blockade. The method that bestquantitates the blockade involves measuring the ampli-tude of the CMAP produced by supramaximal stimula-tion of a peripheral motor nerve (M wave) called T1 andcompares the amplitude with the baseline amplitudebefore adding the blocking agent. When neuromuscularmonitoring is conducted this way, successful monitoringof myogenic responses has been accomplished at T1between 5% and 50% of baseline (Sloan, 2002b).A second technique is referred to as the train-of-four

(TOF) response because it involves examining the mus-cle response when four peripheral motor nerve stimuliare delivered at a rate of 2 Hz. In this technique theamount of acetylcholine released decreases with each

FIG. 8. Motor evoked potential responses recorded from the hypothe-nar muscles in a ketamine-anesthetized baboon after transcranial elec-trical stimulation. Responses displayed are those following variousintravenous doses of propofol. Readily apparent is the marked ampli-tude reduction and increase in latency of the prominent response atapproximately 14 msec similar to the effect of increasing isofluraneconcentrations (Fig. 5). Recordings taken using a Biologic Navigator(Biologic Inc., Mundelein, IL, USA) from subdermal needle electrodesplaced percutaneously in the hypothenar space using an amplificationof 5,000 and a bandpass filtration of 3 to 1,500 Hz. A single supra-maximal stimulus was used at Cz to Fz using a Digitimer D180A.



stimulation such that its effectiveness to compete withthe neuromuscular blocking agent is reduced with eachstimulation. Quantitatively the TOF can be measured bycomparing the amplitude of the M wave of the fourthtwitch (T4) with that of the first (T1) in the T4:T1 ratio.Practically, the number of visible twitches produced isusually recorded with declining numbers of twitches asthe blockade increases. Acceptable CMAP monitoringhas been conducted with two of four twitches remaining.For comparison of the two techniques, only one responseof four is present when T1 is less than 10%, two twitchesat 10 to 20%, and three twitches at 20 to 25% of thebaseline T1 response.When using a neuromuscular blockade, tight control

of the blockade is necessary so that excessive blockadedoes not eliminate the ability to record MEPs, therebymimicking neural injury. For this reason, most cliniciansuse drug infusions, some with closed-loop control sys-tems, to monitor the twitch and to control the infusion. Inparticular, the avoidance of drug boluses is important toprevent wide changes in relaxation, especially excessivedegrees of block.Although recording of myogenic MEP responses is

possible with partial neuromuscular blockade, the ampli-tude of the CMAP will be reduced by the blockade.Studies suggest the actual reduction is nonlinear, with theM wave being reduced more than the MEP (Sloan,2002b). Several studies have demonstrated that the MEPamplitude at a T1 reduced to 20% of baseline correlateswith an MEP reduction of 50 to 60% (Kalkman et al.,1992a; Sloan and Erian, 1993a, b) (Fig. 9).Furthermore, when mechanical measurements are

used (TOF), MEPs have been observed when no me-chanical response was observed (Kalkman et al., 1992a).This nonlinearity can be explained by the observationthat the mechanisms of muscle activation differ betweenthe M response from supramaximal peripheral nervestimulation (as used for measuring neuromuscular block-ade) and the MEP response from transcranial stimula-tion. The MEP response is much larger because centrallyapplied pulses lead to repetitive activation of spinalmotor neurons, with attendant spatial and temporal sum-mation (Day et al., 1987a). For this reason, MEPs aremore robust than the M response during blockade andmay not be abolished as markedly as the M response(T1).In general, the goal with transcranial MEP testing is to

have sufficient blockade such that patient movementwith stimulation is not distracting or hazardous duringthe surgery (particularly when the microscope is used).Furthermore, some muscle relaxation may be necessaryto allow surgical manipulation of structures adherent to

muscles. In other circumstances, muscle relaxation maybe necessary to reduce the noise in recording electrodes(e.g., SSEPs recorded near neck or facial muscles [seeFig. 3] or neurogenic responses recorded epidurally [Fig.

FIG. 10. The effect of muscle relaxation on motor evoked potentialresponses recorded by an epidural electrode in a ketamine- and isoflu-rane-anesthetized baboon after transcranial electrical stimulation. No-table is the paraspinous muscle activity evoked by stimulation (lowertrace) that interferes with the recording of the D and I waves recordedwhen muscle relaxation is present (upper trace). Recordings takenusing a Biologic Navigator (Biologic Inc., Mundelein, IL, USA) froma midthoracic bipolar epidural recording electrode placed percutane-ously using an amplification of 2,000 and a bandpass filtration of 3 to1,500 Hz. A single supramaximal stimulus was used at Cz to Fz usinga Digitimer D185.

FIG. 9. This graph shows the relative mean amplitude ($ standarderror of the mean) of motor evoked potentials (MEPs) recorded in thehypothenar muscles of the monkey after transcranial magnetic stimu-lation versus the relative amplitude of the M wave produced bystimulation of the median nerve at the wrist at various degrees ofneuromuscular blockade with vecuronium and atracurium (pooleddata). Notable is that the relative MEP amplitude is consistently largerthan M wave (indicated by the dashed line), suggesting that neuromus-cular blockade has a more profound block on peripheral stimulation(used to measure the degree of blockade) than on central stimulation (asused for MEP monitoring). Data for this plot was taken, and combined,from data in Sloan and Erian (1993a, b).



10]) or over peripheral nerves, or to reduce muscleartifacts that may be interpreted as neural responses (e.g.,paraspinous muscle responses seen in epidural record-ings of MEP). A T1 blockade of 10 to 20% of baselineappears to accomplish this goal adequately (as men-tioned earlier, this corresponds to two of four twitches ina TOF).For monitoring, the blockade needs to be such that the

remaining MEP amplitude is distinguishable reliablyfrom the background noise. Currently, most authorsrecommended a blockade with T1 between 10 and 20%of baseline or one to two twitches in a TOF. A moreprofound block reduces the MEP excessively and a lessprofound block is associated with excessive patientmovement. Because of varying muscle sensitivity tomuscle relaxants, the neuromuscular blockade should beevaluated in the specific muscle groups used for electro-physiologic monitoring.As a consequence of the amplitude reduction, the

ability to record with partial neuromuscular blockadewill be dependent on other factors that reduce the myo-genic response amplitude, such as anesthesia or neuro-logic disease. Hence, amplitude reduction with initiallysmall responses, or with anesthetic choices that reduceamplitude markedly, may make the use of a blockademore difficult. Fortunately the CMAP amplitude is usu-ally quite large. It is also important to recognize that theuse of amplitude criteria for warning of impending neu-rologic injury may not be possible because inevitablefluctuations in the degree of blockade may obscure theapplication of strict criteria. For this reason, some intra-operative neurophysiologic monitoring of MEPs useonly the presence or absence of a response rather thanamplitude criteria or attempting to “calibrate” the MEPamplitude based on the size of the M wave response.Insufficient data are available for determining the

expected impact of neuromuscular blocking agents whenintraoperative neurophysiologic monitoring is used forevaluating electromyographic responses from mechani-cal irritation or spontaneous activity (such as with facialnerve monitoring or during cauda equina or spinal nerveroot procedures). In these cases concern has been raisedthat mechanical stimulation of the nerve root may be aweak stimulus or that partial paralysis may reduce theability to record these responses. In addition, patientmotion is less of an issue. Motion, if it occurred, couldprovide direct positive feedback to the surgeon as heworks around the nerve or when he intentionally stimu-lates to check nerve integrity. In these cases patientmovement is anticipated by the surgeon (as opposed tounexpected motion during spinal or other surgeries whenthe stimulation is under the control of the monitoring

team). Similar circumstances apply to the surgical testingof pedicle screw or nerve root testing during spinalsurgery, and many monitoring teams elect to avoid mus-cle relaxants. With respect to the spinal procedures, if apartial neuromuscular blockade is used, the surgeon maybe able to stimulate nearby nerve roots (e.g., adjacentlevels of the spine) to identify whether recording ispossible to confirm the probability of the monitoringbeing useful.

Nonanesthetic Intraoperative Factors AffectingMonitoring

In addition to changes resulting from the surgicalmanipulation of the nervous system and anesthetic ef-fects, the physiologic milieu plays an important role inneuronal functioning. In addition, the interaction of thisphysiology and the surgical manipulation may determineneuronal survival.

Blood Flow

Maintenance and control of the patient’s blood pres-sure is part of the anesthetic management of the patienthaving surgery. Numerous studies have demonstrated athreshold relationship between regional cerebral bloodflow and cortical evoked responses (Sloan, 1997). Thecortical SSEP remains normal until blood flow is re-duced to approximately 20 mL/min/100 g. At morerestricted blood flows of between 15 and 18 mL/min/100g of tissue, the SSEP is altered and lost. As with anes-thetic effects, subcortical responses appear less sensitivethan cortical responses to reductions in blood flow.Local factors may produce regional ischemia not pre-

dicted by systemic blood pressure. For example, duringspinal surgery, the effects of hypotension may be aggra-vated by spinal distraction, such that an acceptable limitof systemic hypotension cannot be determined withoutmonitoring (Brodkey et al., 1972; Dolan et al., 1980;Gregory et al., 1979). Other examples of regional effectsinclude peripheral nerve ischemia from positioning, tour-niquets, or vascular interruption; spinal cord ischemiafrom aortic interruption or mechanical distortion; carotidartery interruption; vertebrobasilar insufficiency aggra-vated by head extension; cerebral artery constriction byvasospasm; and cerebral ischemia resulting from retrac-tor pressure.Motor evoked potentials and SSEPs are both sensitive

to spinal cord events produced by vascular ischemia(aortic cross-clamping) or mechanical compression (epi-dural balloon). However, because these tracts are re-moved topographically from one another, MEPs and



SSEPs may show differential sensitivity to an ischemicevent. The SSEP and the myogenic MEP are sensitive tospinal cord ischemia associated with thoracic aorticclamping and reductions in spinal cord blood flow.

Intracranial Pressure

Elevated intracranial pressure is associated with re-ductions in amplitude and increases in latency of corti-cally generated SSEPs (Mackey–Hargardine and Hall,1985). Increased intracranial pressure, probably by virtueof its affect on cortical structures, produces a pressure-related decrease in cortical SSEP responses, with loss ofbrainstem responses as uncal herniation occurs. ForMEPs, a gradual increase in the onset of the responseoccurs until a response can no longer be produced.

Blood Rheology

Because changes in hematocrit can alter both oxygencarrying capacity and blood viscosity, the maximumoxygen delivery is often thought to occur in a midrangehematocrit (30 to 32%). Evoked response changes withhematocrit are consistent with this optimum range. In astudy of upper extremity SSEPs in the baboon, Nagao etal. (1978) observed an increase in amplitude with mildanemia, an increase in latency at hematocrits of 10 to15%, and further latency changes and amplitude reduc-tions at hematocrits less than 10%. These changes werepartially restored by an increase in the hematocrit. Sim-ilar MEP data are not available.

Ventilation

Hypoxemia can also cause evoked potential deteriora-tion (Grundy et al., 1981) before other clinical parame-ters are changed. Alterations in carbon dioxide areknown to alter spinal cord and cortical blood flow. Themost notable changes in cortical SSEPs occur when thecarbon dioxide tension is extremely low, suggestingexcessive vasoconstriction may produce ischemia (car-bon dioxide tensions ! 20 mm Hg). This effect has beensuggested to contribute to alterations in SSEPs duringspinal surgery (Grundy et al., 1982) and may be expectedto produce some MEP changes.

Temperature

Hypothermia-associated alterations in SSEPs andMEPs have been observed (Sloan, 1997). Like anestheticeffects and ischemia, changes are more prominent at thecephalic end of long neural tracts or in components ofresponses associated with multiple synaptic elements.

Hence, SSEP responses recorded from peripheral nervesare affected minimally, whereas those produced by cor-tical structures are affected markedly. Motor evokedpotential responses will have an increased onset withhypothermia from slowed conduction.Changes in regional temperature can also occur, re-

sulting in evoked response alterations that would not bepredicted otherwise based on unchanged body (core)temperature. For example, cold irrigation solutions ap-plied to the spinal cord, brainstem, or cortex causeevoked response changes routinely. Similarly, extremitycooling (as from cold intravenous solutions) can alter theSSEP originating from stimulation to a nerve from thatlimb.With hypothermia, the MEP demonstrated a gradual

increase in onset as esophageal temperature decreasedfrom 38°C to 32°C. An increase in stimulation thresholdwas also observed at lower temperatures. This is consis-tent with both cortical initiation and peripheral conduc-tion being affected by the drop in temperature (Sloan,1991).

Other Physiologic Variables

Changes in a variety of other physiologic variablesmay produce alterations in the evoked responses duringsurgical monitoring. Marked reductions in total bloodvolume and increases in superior vena caval pressureduring cardiopulmonary bypass have been associatedwith SSEP changes. Other changes in the neurochemicalenvironment (e.g., changes in glucose, sodium, potas-

TABLE 1. Anesthetic sensitivity matrix

Anestheticeffect

Sensitive toanestheticagents

Relatively insensitive toanesthetic agents

Insensitive tomusclerelaxants

Group I;corticalSSEP*;corticalAEP*

Group II; epidural, andperispinal SSEP andMEP; far-fieldsubcortical SSEP;sensory cranial nerve(BAEP)*

Sensitive tomusclerelaxants

Group III;TranscranialMEP

Group IV; pedicle screwstimulation, spinalreflex testing, motorcranial nerve (e.g.,facial nerve), andSpinal MEP‡

* Cortical, epidural, perispinal, and subcortical somatosensory evokedpotential (SSEP) and motor evoked potential (MEP) refer to the location ofthe recording electrodes.‡ Spinal MEP refers to the proximal site of stimulation within the

spinal cord with recording in the muscle.AEP, auditory evoked potentials; BAEP, brainstem auditory evoked

potentials.



sium, and other electrolytes) could also result in evokedresponse changes.

ANESTHETIC CHOICE AND MANAGEMENT

The choice of anesthesia for a surgical procedure inwhich electrophysiologic monitoring is to be performeddepends on many factors. Foremost in the mind of theanesthesiologist is the safety and comfort of the patient.Some anesthetic choices will be dependent on the med-ical problems of the patient and some choices will bedependent on the specific surgical needs. Depending onthe electrophysiologic monitoring modalities, constraintsmay occur on the choice of anesthetic agents or theirdoses.The most effective approach to choose a suitable

anesthetic technique is based on preoperative discussionof the case. In general, the modalities can be divided intofour groups based on whether the responses recorded aresensitive or insensitive to anesthesia (primarily inhala-tional agents) and whether they are helped or hindered bymuscle relaxants (Table 1). As is often the case whenmultiple modalities are to be monitored, the most restric-tive aspects of each will define the anesthetic group,which determines the initial anesthetic approach.Group I responses (sensitive to anesthetic agents but

insensitive to muscle relaxants) represent the largestgroup of the more commonly recorded cortical sensoryevoked responses For monitoring during these modali-ties, the inhalational agents, if used, are usually restrictedto less than 0.5 MAC of these agents. Current studiessuggest that desflurane and sevoflurane may be desirablebecause their insolubility allows them to attain steadystate faster with a change in concentration.Some controversy exists regarding the use of nitrous

oxide. As an agent that depresses the evoked potential byreducing amplitude and prolonging latency, it is oftenacceptable when combined solely with intravenousagents. However, when combined with the potent inha-lational agents, it appears to produce a deleterious syn-ergistic effect that produces more depression than wouldbe predicted individually. Because nitrous oxide is rela-tively insoluble, the anesthetic effects can change rapidlywhen concentrations are varied. Increases in amplitudemay be seen when nitrous oxide is discontinued abruptly(possibly obscuring a drop in amplitude, signaling neu-rologic deterioration, which occurred during theprocedure).Fortunately, group I responses are less affected by

intravenous anesthetic agents. For example, opioidsshow mild depression of amplitude and increase in la-tency in cortical responses such that opioid analgesia is

used commonly during recording of cortical SSEPs as acomponent of a total intravenous technique or as asupplement to a reduced concentration of inhalationalagent. Ketamine is an example of an agent that permitsrecording of evoked responses and, in fact, may increasethe cortical response amplitude (Schubert et al., 1990).Droperidol and midazolam can be used for sedation

before induction of anesthesia. Droperidol appears tohave minimal effects on response when combined withopioids. Midazolam (0.2 mg/kg) produces a mild depres-sion of cortical SSEPs and minimal effects on subcorticaland peripheral SSEPs. Because of midazolam’s excellentamnestic qualities, an infusion can be used to maintainsupplemental hypnosis during opioid or ketamine anal-gesia. Thiopental, propofol, and etomidate are used forinduction of anesthesia. Thiopental can be used forinduction with these responses, but newer agents havesupplanted its use in TIVA. Although propofol inductionproduces amplitude depression in cortical SSEPs, withrapid recovery after termination of infusion (Sloan,1997), it has become a popular component of TIVAbecause its rapid metabolism allows titration. Like ket-amine, etomidate produces an amplitude increase ofcortical components after injection (Kochs et al., 1986)with no changes in subcortical and peripheral sensoryresponses, making it a valuable component of TIVA aswell. Although muscle relaxants have no direct effect onSSEPs, they may actually improve the quality of therecordings because electromyographic interference is re-duced in electrodes near muscle groups (see Fig. 3).Pathways that are less dependent on synaptic function,

on which the effects of anesthetic agents are far lessmarked, generally characterize group II responses.Hence, most inhalational and intravenous anesthetics canbe used. As a consequence of the difference betweengroups I and II, substantial interest has been in monitor-ing from recording electrodes placed in the spinal bonyelements, in the subdural or epidural space, and fromrecording electrodes placed subcortically as an alterna-tive to cortical SSEP recording for spine surgery. Prob-lems with extracortical recordings have included markedvariability resulting from motion and dislodgment by thesurgeon as well as lost vigilance of events at locationsother than the spine (e.g., cortical ischemia). The epi-dural technique has become commonplace in Japan andEurope, and despite its invasive nature, this techniqueappears remarkably safe and is preferred by some au-thors (Erwin and Erwin, 1993; Jones et al., 1983).Perispinal electrodes can also be used for recording afterperispinal, epidural, and cortical stimulation (Stephen etal., 1966).Hence, it is common practice to combine recordings



from the subcortical locations to supplement corticalrecordings during spine surgery. As such, if anesthetic orphysiologic shifts alter the cortical response, the subcor-tical response may assist in determining whether thechange is the result of adverse circumstances in thesurgery site or the result of changes in the corticalregions. If the cortical responses are the only responsesmonitored, this differentiation may be difficult and inap-propriate conclusions may be drawn.Stimulation of the spinal cord and recording from the

spinal cord or recording of the peripheral nerve fromspinal stimulation appears to be a group II response. Asdiscussed earlier, the anesthetic choice may affect therelative contributions of motor and sensory components.Group III responses are clearly the most challenging

for the anesthesiologist, because the need to limit musclerelaxation and avoid inhalational agents requires totalintravenous anesthesia. As such, the major drawback oftranscranial MEP has been the effects of anesthesia.Because the effects of opioids are minimal, opioid-basedanesthesia is often used when myogenic transcranialMEPs are monitored. Fentanyl may reduce backgroundspontaneous muscle contractions and associated motorunit potentials, which may improve muscle recordings.Ketamine may also produce an increase in the amplitudeof muscle- and spinal-recorded responses after spinalstimulation (Kano and Shimoji, 1974), making it a de-sirable agent.Thiopental and midazolam produce CMAP depression

at doses less than those affecting the SSEP and lastingfor a long period of time after bolus induction (e.g., 45minutes), making them less desirable agents. Propofolhas been used in MEP when the recordings are epidural;however, it needs careful titration to avoid depression ofthe myogenic MEP responses. Thus, when most anes-thetic protocols can be used for epidural recordings,anesthesia techniques using etomidate, ketamine, propo-fol, and opioids are the most popular for muscle responserecording.Some controversy exists regarding the allowable de-

gree of muscle relaxation. As discussed earlier, someindividuals raise concern that any degree of relaxationmay obscure electromyographic recordings elicited bymechanical irritation of the nerves or when pathologyexists and may reduce amplitude in patients with smallresponses such that MEP may not be monitored. On theother hand, the patient motion that occurs with transcra-nial motor cortex stimulation can be excessive, and somedegree of relaxation is therefore usually desirable. Fur-thermore, muscle relaxation may be of help to removemuscle artifacts in epidural recordings. This suggests thatno generalizations can be made and each patient must be

analyzed individually. Tightly controlled relaxation isneeded to prevent shifts in amplitude (or loss of theresponse) when relaxation is used and monitoring of thisrelaxation in the muscles being monitored is desirable.Current studies with high-frequency trains of transcra-

nial stimuli (two to five pulses with an interstimulusinterval of 2 to 5 msec) produces a more robust stimu-lation. Perhaps these techniques will permit the use ofmore anesthetics for transcranial-stimulated MEP. Al-though high-frequency stimulation would appear to al-low the use of agents that reduce amplitude and prolonglatency (notably low-dose inhalational agents), the au-thors of clinical studies using this technique currentlyrecommend TIVA (Sloan, 2002b). Clearly this is an areaof anesthesia and monitoring that awaits pharmacologicadvances to allow a wider application of this monitoringtechnique.Lastly, group IV responses are usually recorded easily

because, although muscle relaxation is limited, the free-dom to use inhalational agents makes anesthesia lesschallenging. Typical responses here are stimulation ofcranial or peripheral nerves and recording of peripheralmuscle responses (e.g., pedicle screw testing duringspinal instrumentation). As discussed earlier, contro-versy surrounds the use of muscle relaxants. Some au-thors (Sloan, 2002b) have indicated that spontaneousactivity from nerve irritation is difficult to detect duringcontrolled relaxation. Small amplitude responses of in-jured or poorly functioning nerves are particularly diffi-cult to detect, such that many authors recommend avoid-ing muscle relaxants in these cases. Stimulation of spinalcord proximal to areas of risk to produce myogenicMEPs is insensitive to the anesthetic technic (i.e., inha-lational or TIVA) but sensitive to the degree of musclerelaxation.Thus, based on the monitoring modalities involved,

the basic anesthetic constraints can be identified. Usuallythe initial anesthetic choice becomes apparent, as men-tioned earlier, and an initial maintenance anesthesia isdelivered after induction. Before the period of criticalmonitoring, the monitoring team and the anesthesiologistneed to assess the quality of the responses. If the anes-thetic choice has allowed adequate responses, then themajor goal of anesthesia is to maintain a steady state ofanesthetic drugs so that fluctuations do not result inresponse variations that might obscure or simulate indi-cations of neural compromise. If the electrophysiologicdata are inadequate the monitoring team should evaluatewhether the monitoring technique can be adjusted toallow adequate responses. For example, changes in re-cording electrodes, filter settings, and stimulation ratemay improve the responses. Furthermore, shifting to a



technique that is less susceptible to anesthetics maypermit success (e.g., monitoring using subcortical SSEPsinstead of cortical SSEPs).If the electrophysiologic technique cannot be adjusted,

the anesthesiologist needs to determine whether a changein the anesthetic technique is possible to allow betterrecording. With group I modalities, the elimination ofnitrous oxide and potent inhalational agents may permitevoked potential recording. If this is not successful, theanesthesiologist may wish to consider the use of etomi-date or ketamine to enhance the cortical responses. Asimilar situation applies to group III responses, in whichthe usual choice of TIVA may need to be adjusted toinclude etomidate or ketamine. For responses affected bymuscle relaxation (groups III and IV), careful evaluationof the relaxation is necessary.In addition to maintaining a steady anesthetic state, the

anesthesiologist should also seek to maintain a steadyphysiologic milieu. By integrating the electrophysiologicfindings with physiologic and anesthetic monitoring, theanesthesia team can provide as optimal a neural environ-ment as possible for the patient. Finally, by integratingour knowledge of the surgical procedure and these phys-iologic and anesthetic effects, the anesthesiologists canparticipate actively in the electrophysiologic monitoringteam to interpret changes in the monitored responses andto assist surgeons in their operative decision making.

CONCLUSION

Although the choice and management of anesthesia isimportant to the success of intraoperative monitoring, thecritical component of teamwork underlies the entireprocess. In fact, the entire operative team needs toremain focused on their interdependence in creating themost effective outcome for the patient. The interdepen-dence of the surgeon and anesthesiologist is well known,and the interdependence of the surgeon with the moni-toring team is logical. For example, the surgeon relies onthe monitoring team to be his “electrophysiologic eyesand ears” whereas the monitoring team relies on thesurgeon to keep them informed of the changing operativeareas of risk so those eyes and ears can remain focusedon the most appropriate area of neurologic risk.However, often forgotten is the mutual interdepen-

dence of the monitoring team and the anesthesiologist.The monitoring team relies on the anesthesiologist toprovide a physiologic and anesthetic milieu that is sup-portive of monitoring and to inform the monitoring teamwhen untoward anesthetic and physiologic changes mayresult in changes in the monitoring that may obscureeffective neurologic vigilance. In return, the anesthesiol-

ogist relies on the monitoring team to be her “neurophys-iologic eyes and ears” to detect untoward physiologiccompromise reflected in the monitoring and should sug-gest that the physiologic management be adjusted toimprove the neural environment.Clearly, a better understanding of the effects of anes-

thesiology and physiology allows the monitoring team towork more effectively with the anesthesiologist. Like-wise, a better understanding of electrophysiologic mon-itoring allows the anesthesiologist to integrate the resultsin decision making and physiologic management, im-proving patient care.

REFERENCES

Angel A, LeBeau F. A comparison of the effects of propofol with otheranesthetic agents on the centripetal transmission of sensory infor-mation. Gen Pharmacol 1992;23:945–63.

Brodkey JS, Richards DE, Blasingame JP, et al. Reversible spinal cordtrauma in cats: additive effects of direct pressure and ischemia.J Neurosurg 1972;37:591–3.

Calancie B, Harris W, Broton JG, et al. “Threshold-level” multipulsetranscranial electrical stimulation of motor cortex for intraopera-tive monitoring of spinal motor tracts: description of method andcomparison to somatosensory evoked potential monitoring. J Neu-rosurg 1998;88:457–70.

Day BL, Dressler D, Maertens de Noordhout A, et al. Electric andmagnetic stimulation of human motor cortex: surface EMG andsingle motor unit responses. J Physiol 1989;412:449–73.

Day BL, Rothwell JC, Thompson PD, et al. Motor cortex stimulation inintact man. Brain 1987a;110:1191–209.

Day BL, Thompson PD, Dick JP, Nakashima K, Marsden CD. Differ-ent sites of action of electrical and magnetic stimulation of thehuman brain. Neurosci Lett 1987b;75:101–6.

Deletis V. Intraoperative monitoring of the functional integrity of themotor pathways. Adv Neurol 1993;63:201–14.

Dolan EJ, Transfeld EE, Tator CH, et al. The effect of spinal distractionon regional blood flow in cats. J Neurosurg 1980;53:756–64.

Drummond JC, Todd MM, U HS. The effects of high dose sodiumthiopental on brainstem auditory and median nerve somatosensoryevoked responses in humans. Anesthesiology 1985;63:249–54.

Erwin CW, Erwin AC. Up and down the spinal cord: intraoperativemonitoring of sensory and motor spinal cord pathways. J ClinNeurophysiol 1993;10:425–36.

Flood P, Role LW. Neuronal nicotinic acetylcholine receptor modula-tion by general anesthetics. Toxicol Lett 1998;100–111:149–53.

Glassman SD, Shields CB, Linden RD. Anesthetic effects on motorevoked potentials in dogs. Spine 1993;18:1083–9.

Gregory PC, McGeorge AP, Fitch W, et al. Effects of hemorrhagichypotension on the cerebral circulation. II. Electrocortical func-tion. Stroke 1979;10:719–23.

Grundy BL, Heros RC, Tung AS, et al. Intraoperative hypoxia detectedby evoked potential monitoring. Anesth Analg 1981;60:437–9.

Grundy BL, Jannetta PJ, Procopio PT, et al. Intraoperative monitoringof brain-stem auditory evoked potentials. J Neurosurg 1982;57:674–81.

Gugino LD, Aglio LS, Segal NE, et al. Use of transcranial magneticstimulation for monitoring spinal cord motor paths. Semin SpineSurg 1997;9:315–36.

Jones SJ, Edgar MA, Ransford AO, Thomas NP. A system for theelectrophysiological monitoring of the spinal cord during opera-tions for scoliosis. J Bone Joint Surg Br 1983;65:134–9.

Kalkman CJ, Drummond JC, Kennelly NA, et al. Intraoperative mon-itoring of tibialis anterior muscle motor evoked responses to



transcranial electrical stimulation during partial neuromuscularblockade. Anesth Anal 1992a;75:584–9.

Kalkman CJ, Drummond JC, Ribberink AA. Effects of propofol,etomidate, midazolam, and fentanyl on motor evoked responses totranscranial electrical or magnetic stimulation in humans. Anes-thesiology 1992b;76:502–9.

Kano T, Shimoji K. The effects of ketamine and neuroleptanalgesia onthe evoked electrospinogram and electromyogram in man. Anes-thesiology 1974;40:241–6.

Kawaguchi M, Sakamoto T, Ohnishi H, et al. Intraoperative motorevoked potentials induced by direct electrical stimulation of themotor cortex under isoflurane and sevoflurane. Anesth Analg1996;82:593–9.

King J, Sloan TB. Evoked potentials. Intraoperative monitoring of thespinal cord and peripheral nerves. In: Dumitru D, ed. Electrodi-agnostic medicine. 2nd ed. St. Louis: Mosby, 2001.

Kochs E, Treede RD, Schulte am Esch J. Increase of somatosensori-cally evoked potentials during induction of anaesthesia with eto-midate. Anaesthetist 1986;35:359–64.

Legatt AD. Current practice of motor evoked potential monitoring: resultsof a survey. J Clin Neurophysiol 2002;19:454–460.

Macdonald RL, Barker JL. Anticonvulsant and anesthetic barbiturates:different postsynaptic actions in cultured mammalian neurons.Neurology 1979;29:432–47.

Mackey–Hargadine JR, Hall JW III. Sensory evoked responses in headinjury. Central Nerv Syst Trauma 1985;2:187–206.

Mavroudakis N, Vandesteene A, Brunko E. Spinal and brain-stemSEPs and H reflex during enflurane anesthesia. Electroencepha-logr Clin Neurophysiol 1994;92:82–5.

Mochida K, Shinomiya K, Komori H, et al. A new method of multisegment motor pathway monitoring using muscle potentials aftertrain spinal stimulation. Spine 1995;20:2240–6.

Moller AR, ed. Intraoperative neurophysiologic monitoring. Luxem-bourg: Harwood, 1995.

Nagao S, Roccaforte P, Moody RA. The effects of isovolemic hemodi-lution and reinfusion of packed erythrocytes on somatosensoryand visual evoked potentials. J Surg Res 1978;25:530–7.

Nagle KJ, et al. Intraoperative monitoring of motor evoked potentials:a review of 116 cases. Neurology 1996;47:999–1004.

Nestler EJ, Aghajanian GK. Molecular and cellular basis of addiction.Science 1997;278:58–63.

Nuwer MR, ed. Evoked potential monitoring in the operating room.New York: Raven Press, 1986.

Nuwer MR, Dawson EG, Carlson LG, Kanim LE, Sherman JE. Somato-sensory evoked potential spinal cord monitoring reduces neurologicdeficits after scoliosis surgery: results of a large multicenter survey.Electroencephalogr Clin Neurophysiol 1995;96:6–11.

O’Shaughnessy CT, Lodge D. N-methyl-D-aspartate receptor-mediatedincrease in intracellular calcium is reduced by ketamine andphencyclidine. Eur J Pharmacol 1988;153:201–9.

Pechstein U, Nadstawek J, Zentner J, et al. Isoflurane plus nitrous oxideversus propofol for recording of motor evoked potentials afterhigh frequency repetitive electrical stimulation. Electroencepha-logr Clin Neurophysiol 1998;108:175–81.

Schubert A, Licina MG, Lineberry PJ. The effect of ketamine onhuman somatosensory evoked potentials and its modification bynitrous oxide. Anesthesiology 1990;72:33–9.

Sloan T. Anesthetics and the brain. Anesthesiol Clin North Am 2002a;20:1–27.

Sloan TB. Evoked potentials. Anesthesia and motor evoked-potentialsmonitoring. In: Deletis V, Shills J, eds. Neurophysiology inneurosurgery. San Diego: Academic Press, 2002b.

Sloan T. Anesthetic effects on electrophysiologic recordings. J ClinNeurophysiol 1999;15:217–26.

Sloan T. Evoked potentials. In: Albin MS, ed. A textbook of neuroan-esthesia with neurosurgical and neuroscience perspectives. NewYork: McGraw–Hill, 1997:221–76.

Sloan T. Mild hypothermia alters cortical magnetic evoked potentials.Anesth Analg 1991;72:S260.

Sloan TB, Erian R. Effect of atracurium induced neuromuscular blockadeon cortical motor evoked potentials. Anesth Analg 1993a;76:979–84.

Sloan TB, Erian R. Effect of vecuronium induced neuromuscularblockade on cortical motor evoked potentials. Anesthesiology1993b;78:966–73.

Sloan TB, Fugina ML, Toleikis JR. Effects of midazolam on mediannerve somatosensory evoked potentials. Br J Anaesth 1990;64:590–3.

Sloan TB, Ronai AK, Toleikis JR, et al. Improvement of intraoperativesomatosensory evoked potentials by etomidate. Anesth Analg1988;67:582–5.

Stephen JP, Sullivan MR, Hicks RG, Burke DJ, Woodforth IJ, Craw-ford MR. Cotrel–Dubousset instrumentation in children usingsimultaneous motor and somatosensory evoked potential monitor-ing. Spine 1966;21:2450–7.



Anesthesia for Intraoperative Neurophysiologic Monitoring of ...

Documents

Transcript of Anesthesia for Intraoperative Neurophysiologic Monitoring of ...