Pharmacologic Factors Influencing Sensory Evoked Responses

There are multiple drugs used in the perioperative period that can influence the ability to accurately monitor SERs ( Table 38-9 ). An excellent review^[120] provides a detailed

1533

The volatile anesthetics isoflurane, sevoflurane, desflurane, enflurane, and halothane have similar effects in different degrees on all types of SERs. VEPs are the most sensitive to the effects of volatile anesthetics, and BAEPs are the most resistant to anesthetic-induced changes. Spinal

TABLE 38-10 -- Guidelines for choosing anesthetic techniques for procedures when sensory evoked potentials are monitored

Intravenous agents have significantly less effect than equipotent doses of inhaled anesthetics.

Combinations of drugs generally produce additive effects.

Subcortical (spinal or brainstem) sensory evoked potentials are very resistant to the effects of anesthetic drugs. If subcortical responses provide sufficient information for the surgical procedure, anesthetic technique is not important, and effects on cortically recorded responses may be ignored.

and subcortical SSEP responses are significantly less affected than the cortical potentials.^{[121] [122] [123]}

SSEPs, because they are the most widely used intraoperative SER technique, are the most completely studied with respect to the effects of anesthetic drugs. The effects of the volatile agents on cortical SSEPs are dose-dependent increases in latency and conduction times and a decrease in amplitude of cortically but not subcortically recorded signals.^{[121] [122] [123] [124] [125]} When comparing the different volatile agents, studies have reported conflicting results.^{[121] [123]} For example, one study^[123] suggests that halothane has a greater impact on cortical SSEPs than isoflurane or enflurane, whereas another published report^[121] supports a greater effect produced by enflurane and isoflurane than halothane. None of these differences is clinically important, and they may be ignored by the practicing clinician. Among the newer agents, desflurane and sevoflurane appear to have qualitatively and quantitatively similar effects on SERs as isoflurane.^{[126] [127] [128] [129] [130]} Up to 0.5 to 1 MAC of any of the potent inhaled agents in the presence of nitrous oxide is compatible with monitoring of cortical SSEPs ( Fig. 38-13 Fig. 38-14 Fig. 38-15 ) in neurologically normal patients.^{[121] [125]} Neurologically impaired patients may show a significantly greater sensitivity to inhaled agents, even to the point of not tolerating any recordable level of inhaled agent (see Fig. 38-19 ). In general, however, better monitoring conditions are obtained with narcotic-based anesthetics with less than 1 MAC total (nitrous oxide plus potent agent) end-tidal inhaled anesthetic concentration.

The volatile anesthetics result in increases in latency of BAEPs without significantly affecting the amplitude.^{[125] [131] [132] [133]} However, volatile anesthetics cause increases in latency and decreases in amplitude in the early (middle latency) cortical responses after auditory

1534

Figure 38-13 Representative somatosensory evoked potentials for cortical responses (C3, C4-FPz) at various minimum alveolar concentrations of isoflurane. (Adapted from Peterson DO, Drummond JC, Todd MM: Effects of halothane, enflurane, isoflurane, and nitrous oxide on somatosensory evoked potentials in humans. Anesthesiology 65:35, 1986.)

Use of the volatile anesthetics during monitoring of VEPs results in dose-dependent increases in latency with or without changes in amplitude.^{[125] [135] [136] [137] [138]} Isoflurane results

Figure 38-14 Representative somatosensory evoked potentials for cortical responses (C3 or C4-FPz) at various minimum alveolar concentrations of enflurane. (Adapted from Peterson DO, Drummond JC, Todd MM: Effects of halothane, enflurane, isoflurane, and nitrous oxide on somatosensory evoked potentials in humans. Anesthesiology 65:35, 1986.)

Figure 38-15 Representative somatosensory evoked potentials for cortical responses (C3, C4-FPz) at various minimum alveolar concentrations of halothane. (Adapted from Peterson DO, Drummond JC, Todd MM: Effects of halothane, enflurane, isoflurane, and nitrous oxide on somatosensory evoked potentials in humans. Anesthesiology 65:35, 1986.)

Figure 38-16 Influence of isoflurane alone on brainstem auditory evoked potentials in a typical subject. Latency of peaks III and IV to V increased at 1.0% but stabilized with increasing anesthetic depth. (Adapted from Manninen PH, Lam AM, Nicholas JF: The effects of isoflurane-nitrous oxide anesthesia on brainstem auditory evoked potentials in humans. Anesth Analg 64:43, 1985.)

1535

Figure 38-17 Brainstem auditory evoked potential recording obtained in one patient at different concentrations of inspired enflurane. (Adapted from Dubois MY, Sato S, Chassy J, et al: Effects of enflurane on brainstem auditory evoked responses in humans. Anesth Analg 61:898, 1982.)

Although volatile anesthetics cause significant changes in the SER waveforms, it is possible to provide adequate monitoring intraoperatively in the presence of anesthetic doses of volatile anesthetics. Doses of agents causing significant depression of the response to be monitored must be prevented. In our experience, end-tidal concentrations of inhaled agents totaling more than 1.3 MAC have a dose-related, increasing probability of obliterating cortical SSEPs, even in neurologically normal patients. Equally important, anesthetic concentration should not be

Figure 38-18 Waveforms of visual evoked responses from one patient were elicited when awake and anesthetized at three end-expired levels of halothane. Four separate tracings obtained during each condition have been superimposed. (From Uhl RR, Squires KC, Bruce DL, et al: Effect of halothane anesthesia on the human cortical visual evoked response. Anesthesiology 53:273, 1980.)

As with the volatile anesthetics, nitrous oxide causes different effects on the SERs, depending on the sensory system monitored. It causes decreases in amplitude without significant changes in latency in SSEPs when used alone or when added to a narcotic-based or volatile anesthetic.^{[121] [122] [139]} The addition of nitrous oxide to a maintenance volatile anesthetic during the monitoring of BAEPs causes no further change.^[131] Likewise, use of nitrous oxide

1536

The effects of barbiturates on SERs have been studied in animal models and in humans. Increasing doses of thiopental in patients result in progressive dose-dependent increases in latency and decreases in amplitude of SSEPs and in progressive increases in the latency of wave V in BAEPs. The changes in SSEPs are more pronounced than are the changes in BAEPs, and waveforms beyond the initial primary cortical response are quickly obliterated. This finding is consistent with the theories that barbiturates affect synaptic transmission more than axonal conduction. Early waveforms in SERs result primarily from axonal transmission, and later waves depend on multisynaptic pathways in addition to axonal transmission. At doses of thiopental far in excess of those producing an isoelectric EEG, adequate monitoring of early cortical and subcortical SSEPs and BAEPs was preserved.^[140] Other barbiturate compounds show similar effects. Somatosensory and auditory SERs were never obliterated, even at doses above those causing complete suppression of spontaneous electroencephalographic activity.^[141] This observation is important, especially when attempting to monitor the adequacy of CBF during cerebrovascular surgery when the patient has been given large, "protective" doses of barbiturates. The EEG is isoelectric and not helpful for monitoring. The early cortical SSEP waveforms are, however, still preserved and may be very helpful in determining adequacy of CBF. Preserved ability to monitor SSEPs in head-injured patients receiving therapeutic thiopental infusions has been demonstrated.^[142] VEPs are much more sensitive to barbiturates. Small barbiturate doses obliterate all except the earliest waveforms. The early potentials persisted with increases in latency, even in response to very high pentobarbital doses.^[143] Except for VEPs, adequate perioperative monitoring of SERs is possible, even in the presence of high-dose barbiturate therapy, as long as the effects of the drug (i.e., increased latency with moderately decreased amplitude) are considered.

After bolus administration and intravenous infusions, etomidate causes increases in latency of all waves and prolongation of central conduction time in SSEPs. Unlike virtually all other commonly used anesthetics, etomidate causes increases in amplitude of the cortical SSEPs.^{[144] [145]} This effect may be caused by an alteration in the balance of inhibitory and excitatory influences or an increase in the irritability of the CNS. This effect seems to be present in the cortex but not in the spinal cord.^[145] Etomidate infusions have been used to enhance SSEP recording in patients when it was not possible to obtain reproducible responses at the beginning of intraoperative monitoring because of the patients' pathologic findings ( Fig. 38-19 ). After baseline responses that could not be monitored, etomidate augmentation of the SSEP allowed adequate monitoring and detection of intraoperative events leading to compromise of the spinal cord.^[145] The effects of etomidate on BAEPs are dose-dependent increases in latency and decreases in amplitude that are not clinically significant.^[146]

Figure 38-19 Effects of etomidate on somatosensory evoked potentials. A, The tracings were obtained from a mildly mentally impaired patient with severe kyphoscoliosis during the early maintenance phase of anesthesia using isoflurane and fentanyl. B, The tracings were obtained after discontinuing isoflurane and institution of an etomidate infusion at 20 µg/kg/min. Notice the dramatically increased amplitude and clarity of the signal in the cortical channels (arrows), which are both recorded with the same amplification scale.

Droperidol in premedicant doses has various effects on SSEPs. In most patients, decreases in amplitude and loss of late waves were observed, but in a few patients, increases in amplitude were seen. In all patients, conduction time was prolonged,^[147] but effects were not clinically significant. Benzodiazepines also can cause changes in SERs.^{[148] [149]} Diazepam causes increases in latency and decreases in amplitude of SSEPs, increases in latency in the cortical response after auditory stimulation, and no change in BAEPs.^{[148] [149]} Midazolam causes decreases in amplitude without changes in the latency of SSEPs. ^[144]

In general, opioids cause small, dose-dependent increases in latency and decreases in the amplitude of SSEPs. These changes are not clinically significant. Effects on amplitude are more variable than the latency increases.^{[150] [151]} Even at large doses of fentanyl (≤60 µg/kg), reproducible

1537