Depression of spinal alpha-motor neurons apparently plays a role in the surgical immobility induced by isoflurane. Using the noninvasive technique of F-wave analysis, the authors tested the hypothesis that depressed motor neuron excitability is an effect common to other clinically relevant inhaled anesthetics.
The authors measured F-wave amplitude in rats anesthetized with desflurane, enflurane, halothane, or sevoflurane. Each animal received one anesthetic at five equipotent anesthetic concentrations (0.6, 0.8, 1.2, and 1.6 minimum alveolar concentration [MAC] and 0.8 MAC with 65% N2O). F waves were detected as late potentials in electromyographic responses evoked in the intrinsic muscles of the hind paw after monopolar stimulation of the ipsilateral posterior tibial nerve.
All tested inhaled anesthetics depressed F-wave amplitude but not M-wave (orthodromic, early muscle activation) amplitude, and increased M-F latency in a dose-dependent manner. At 1.0 MAC, the estimated F/M ratio was 70 +/- 13% SD of that at baseline (0.6 MAC). Nitrous oxide added to 0.8 MAC of the potent vapors depressed F/M ratio by 63 +/- 17%.
All anesthetics tested appeared to depress the excitability of spinal motor neurons. This effect may contribute to surgical immobility, and its magnitude is comparable at equipotent concentrations of agents. The authors hypothesize that this effect is due to hyperpolarization, although, currently, there is insufficient information to discriminate between pre- and postsynaptic mechanisms.
Key words: Anesthetic mechanisms. Anesthetic potency. Anesthetics, volatile: desflurane; enflurane; halothane; sevoflurane. Measurement techniques: electromyography. Spinal cord: motor neurons.
ANESTHESIA-INDUCED immobility apparently is mediated by spinal alpha-motor neurons and is unaltered by acute, hypothermic, cervical spinal cord transection. [1]The excitability of spinal alpha-motor neurons can be assessed noninvasively by detecting the presence of recurrent, evoked electromyographic signals known as F waves. [2,3]Previously, we demonstrated that isoflurane depresses the amplitude of F waves at concentrations that inhibit movement in response to tail-clamp stimulation. [4]This finding suggested that anesthetic-induced immobility may be due, in large part, to reduced excitability of motor neurons. In the current study, we tested the generality of this possible mechanism of anesthetic-induced immobility in the other commonly used volatile anesthetics.
Materials and Methods
Experimental Preparation
With approval of the University of California, San Francisco Committee on Animal Research, we studied young adult ([nearly equal] 3 months), Sprague-Dawley rats whose characteristics are described in Table 1. Animals were allowed food and water ad libitum until the study. Animals were divided into four groups that were distinguished by which anesthetic (desflurane, enflurane, halothane, or sevoflurane) was administered. Anesthesia was induced by inhalation of the selected agent in air, and tracheas were intubated with a 2" (5.1 cm) 16G intravenous catheter. Anesthesia was maintained with the selected agent in oxygen. Mechanical ventilation was adjusted to achieve peak inspiratory airway pressures less or equal to 20 mmHg and end-tidal carbon dioxide concentration of 35+/-5 mmHg. Airway gas concentrations were monitored continuously with an infrared analyzer (CapnoMac Ultima, Datex lnstrumentarium, Helsinki, Finland) using an expiratory limb dead space sampling point previously described. [5]Normal rectal temperature (37.5 degrees C) was maintained using a heat lamp and warming blanket. We injected lactated Ringer's solution subcutaneously to maintain adequate hydration.
Anesthetic Protocol
From previous experience with this rat strain (unpublished data, 1994), we initially defined 1.0 MAC as 8.0% desflurane, 2.2% enflurane 1.1% halothane, and 2.4% sevoflurane. After tracheal intubation, we applied a prospectively randomized sequence of four concentrations of agent (0.6, 0.8, 1.2, and 1.6 MAC) with equilibration for an appropriate time (Table 1) at each step before data acquisition. Minimum alveolar concentration was determined in each animal after F-wave data collection as described previously. [4]The concentration given as the first step was repeated as the last step to assess the possible effect of increasing anesthetic duration. Finally, all animals were equilibrated to 0.8 MAC of the potent agent combined with 65% N2O.
Rats in the enflurane anesthesia group also were exposed to 1.6 MAC enflurane during hyperventilation (PETCO2of [nearly equal] 20 mmHg) in an effort to elicit neuronal irritability and electroencephalographic epileptiform activity.
Electrophysiologic Data Acquisition
The right tibial nerve was stimulated with a cathodal needle electrode in the popliteal fossa driven by a constant current stimulator in external command mode (CCIU-8, FHC, Brunswick, ME). The stimulus duration was fixed at 500 micro second, and the intensity was varied (to assure the absence of confounding "H" waves [4,6]) using a prospectively randomized, intermixed sequence of the following current levels (0.20, 0.65, 1.10, 1.55, and 2.00 mA). A new randomization sequence for stimulus intensity was used at each agent concentration. Stimulation occurred at 5.0+/-0.l-s intervals (using random variation to avoid phase-locked artifact). Evoked electromyographic activity was recorded from needle electrodes in the intrinsic muscles between the third and fourth digits of the right foot, and were referenced to a ground electrode in the right heel. Ten artifact-free traces were obtained at each level of stimulus current.
The electromyographic data were amplified approximately 1,330x by a Grass 7P511 amplifier (Grass Instruments, Quincy, MA) and band-pass filtered (10 Hz-10 kHz), then digitized with 16-bit resolution at a rate of 20 kHz. The digitized electromyographic waveform data were displayed, analyzed, and stored on a PowerMac 7100 (Apple Computer, Cupertino, CA). Stimulator control and all aspects of data acquisition and analysis were automated and controlled by a LabView (Version 3.1.1, National Instruments, Austin, TX) program written for this study by one of the authors (IJR).
Data Analysis
Each animal generated at least 500 evoked electro-myographic waveform traces. Each trace contains two distinct components, the M and F waves (Figure 1). The M wave represents muscle depolarization from orthodromic motor axon conduction of the stimulated action potential. The F wave represents centrally mediated, recurrent motor activity after antidromic conduction of the stimulated action potential. F-wave amplitude measurement can be normalized for subject and electrode placement by calculating the F/M amplitude ratio. After manual verification of the exclusion of artifact-contaminated traces, the analysis software determined the maximal peak-to-peak amplitude in the range of 0.9-3.0 msec as the M-wave amplitude, and that in the range of 5.5-9.0 msec as the F-wave amplitude in each trace. The latency between the peak of the M and the F waves in each trace also was recorded; however, inherent equipment limitations prevented measurement of the stimulus M wave latency. At each equilibrated agent concentration, the measured M- and F-wave amplitudes and latency for each of the five stimulus levels were averaged.
Statistical Analysis
The observed wave amplitudes and latencies were compared within and across anesthetic agents using one-way analysis of variance (Statview 4.1, Abacus, Berkeley, CA). The correlation of F/M ratio and movement response to stimulus was assessed using logistic regression (JMP 3.01, SAS Institute, Cary, Nc). Significance was assumed where P less or equal to 0.05.
Results
Measured MAC in the sample populations were within 4% of the a priori target values (Table 1), except for halothane, where the measured MAC was 15% less than the target level. For the purposes of comparison between drug groups, this difference was judged small and, therefore, was ignored.
Electrophysiology
(Figure 1) is typical electromyographic activity evoked after peripheral nerve stimulation. M-wave amplitude was not altered by exposure to changing concentration of any of the four experimental anesthetics, suggesting no anesthetic-induced depression of neuromuscular transmission at the doses tested (Table 2). The F-wave amplitude and the F/M ratio revealed significant alpha-motor neuron depression with increasing dose of each anesthetic tested (Figure 2, Table 2). The latency between paired M and F waves increased by 0.6, 1.7, or 1.0 msec when anesthetic dose increased from 0.6 to 1.6 MAC during administration of desflurane, enflurane, or halothane, respectively. Sevoflurane did not alter this latency.
The F/M ratio was similar for all agents at 0.6 MAC, but differed at higher concentrations, as illustrated in Figure 2and Table 3. At 0.8 MAC, halothane produced more depression than the other agents, significantly different from desflurane. At higher concentrations, halothane continued to produce smaller F/M ratios than desflurane. This finding may be due to the lower-than-predicted MAC in the study population. At 1.2 and 1.6 MAC, enflurane produced F/M ratios smaller than those produced by desflurane or sevoflurane.
Depression of the F wave (or F/M ratio) predicted surgical immobility. Logistic regression demonstrated that when the F/M ratio was reduced to approximately 70% of the value at baseline (0.6 MAC) (mean F/M 17.3, range 14-22.2, depending on anesthetic; Table 4), the probability of movement response to tail clamp stimulation was reduced from 100% to 50%.
Nitrous oxide (65%) added to 0.8 MAC of any tested volatile agent depressed the F-wave amplitude and the F/M ratio (Table 5). Nitrous oxide, however, did not increase M-F latency, except in the case of enflurane.
Electroencephalographic surveillance during high-dose (1.6 MAC) enflurane anesthesia revealed burst suppression or isoelectric cortical activity and no detectable F-wave activity with, or without, the presence of hypocapnia. There was no evidence of epileptiform activity or increased spinal motor neuron excitability due to hypocapnia.
Discussion
The results of this study confirm the earlier finding that spinal alpha-motor neuron depression occurs during general anesthesia [2,4]and extends this finding by demonstrating it to be a phenomenon common to all of the clinically relevant inhaled anesthetics.
Collection of F-wave signals is contaminated occasionally by the presence of monosynaptic reflex arc-mediated electromyographic activity (H-reflex), which superficially resembles the F wave. H-reflex waves have maximal amplitude at low stimulus intensities and diminish in amplitude with increasing stimulation. M and F waves increase in amplitude with increasing stimulus intensity. [4,6]Stimuli at 0.2 and 0.65 mA were used to detect H-reflex activity, but otherwise were not included in the data analysis. H-reflex activity was sought before formal data collection, and, in animals where such activity occurred, the recording electrode positions were adjusted until only F-wave activity was seen.
The dose-dependent increase in latency between the appearance of the M and F waves at doses greater than 1.0 MAC is a new finding. This phenomenon could be due to a number of possible mechanisms, including: 1) motor neuron hyperpolarization increasing the time required to depolarize and generate a recurrent potential, or 2) increased motor axonal conduction time. An inherent limitation of our data acquisition system prevented accurate determination of the stimulus-to-M-wave latency, which would have been an independent assessment of axonal conduction speed. Previous measurement of axonal conduction speed showed little sensitivity to volatile anesthetics administered at the concentrations studied here. [7,8].
Some differences were identified in the F-wave responses to the four tested anesthetics. Two findings were prominent: enflurane produced significantly more F-wave depression above 1.0 MAC than the other agents, and sevoflurane, unlike the other agents, did not increase M-F latency. A number of prior investigators have described quantitative, if not qualitative, differences in the electrophysiologic effects of different volatile anesthetics on axons, [8-10]but it may be premature to ascribe our results to these differences.
The current study differs from our previous reports in two aspects of stimulation. First, we used a constant current generator to provide a more consistent stimulus than the constant voltage generator we used previously. Second, the new stimulator allowed computer control of the output current. Therefore, stimuli of all required intensities could be randomized and completely intermixed, which reduced the potential for bias due to stimulation sequences of blocks of fixed intensity.
In summary, F-wave analysis suggests spinal alpha-motor neurons as sites of anesthetic action common to inhaled halogenated anesthetics. The apparent reduction of motor neurons excitability could explain the ability of anesthetics to create surgical immobility. We hypothesize that the observed changes in F-wave amplitude are possibly due to motor neuron hyperpolarization, [11]but whether hyperpolarization is due to presynaptic changes (i.e., increased inhibitory afferent traffic), or to postsynaptic changes (i.e., increased potassium or chloride membrane conductance) is unknown. Other mechanisms, such as increased membrane permeability, also may contribute to decreased excitability without necessarily altering the membrane potential. Based on the results of this study, measurement of spinal motor neuron membrane potential and impedance will provide insight into the mechanisms of anesthetic effects on cells that appear to mediate clinically relevant actions.
The authors thank Winifred von Ehrenburg, for editorial assistance, and Abbott Laboratories, for the donation of sevoflurane for this study.