Drastic Decrease in Isoflurane Minimum Alveolar Concentration and Limb Movement Forces after Thoracic Spinal Cooling and Chronic Spinal Transection in Rats

The University of California, Davis Animal Care and Use Committee approved this study. Chronic and acute terminal experiments were conducted on 16 adult Sprague-Dawley rats (10 male and 6 female; weight, 340–620 g). Animals were given free access to food and water and were maintained on a 12 h–12 h light–dark cycle with lights on at 07:00.

Chronic experiments were conducted on eight rats (five male and three female) with a T8 spinal transection and three female sham-operated adult Sprague-Dawley rats (weight, 350–450 g). Anesthesia was induced in an acrylic box with isoflurane (5%), and animals were intubated with a 14-gauge arterial catheter and mechanically ventilated with 100% oxygen and 1.5% isoflurane. Body temperature was monitored and maintained at 37 ± 1°C with an electric heating pad. End-tidal carbon dioxide and anesthetic concentration were also monitored during surgery and each testing period with a calibrated Ohmeda Rascal II analyzer (Helsinki, Finland). Under aseptic conditions, a midline incision was made at the T7–T9 level, and a laminectomy was performed on T8. The cord was then completely transected at T8 with a No. 11 scalpel blade, and after visual inspection to verify complete transection, gel foam was packed between the cut ends of the cord. The muscle and overlying fascia were closed with absorbable sutures, and the skin was closed with 2-0 silk sutures. Immediately before surgery and for the entire 28-day testing period, animals were given the antibiotic enrofloxacin (10 mg · kg⁻¹· day⁻¹subcutaneously; Bayer, Shawnee Mission, KS). Immediately after surgery, rats were given buprenorphine (0.08 mg/kg). Urine was expressed two to three times daily by applying manual pressure to the bladder, and the rats’ hindquarters were thoroughly washed. Saline injections (10 ml subcutaneously) were given during surgery and twice daily during the next week. The animals’ diets were supplemented with Nutri-Cal (Evsco Pharmaceuticals, Buena, NJ) for the first 3 days after surgery, and the animals were given food mash daily (by adding water to dry rat chow) for the entire 28-day postoperative period. Body weight, grooming, appetite, bowel movements, hydration, and urine color were monitored closely each day. For sham-operated animals, surgery was performed as described above, except gel foam was placed in the laminectomized site, over the intact spinal cord at T8. Sham animals were given saline, Nutri-Cal, mash, and enrofloxacin for 5 days, after which special caretaking needs were discontinued.

Before data collection and surgery, rats were acclimated to the TF restrainers and the hind paw withdrawal (HPW) apparatus during three daily 45-min sessions. Two awake baseline TF and HPW latency measurements were taken at 3 days and 1 day before surgery and at postoperative days 3, 7, 14, and 28. Reflex latencies tested under isoflurane (box induction and intubation as described above) were performed immediately before spinal transection surgery and at postoperative days 3, 7, 14, and 28. Body temperature, end-tidal carbon dioxide, and anesthetic concentration were monitored and maintained as described in the section Surgery and Caretaking in Chronic Spinally Transected Animals.

To measure TF latency, rats were placed in a cylindrical acrylic restrainer with the tail protruding out of a hole in the rear gate. The ventral surface of the distal third of the tail was placed over a 0.5-cm hole in an aluminum box, with a 300-W halogen projector bulb housed beneath the hole. Electrical current from a variable alternating current power supply lighted the bulb so that it heated the overlying tail until the rat moved its tail away from the light. The variable power supply was set to yield a TF latency of approximately 4 s. During formal testing, the TF latency was measured by a timer that was started at the onset of tail heating and stopped at the moment of the tail flick. An 8-s cutoff was set to prevent tissue damage. Five TF measurements were made in each rat at each session with a 2- to 4-min time interval between trials. The average of the numeric middle three TF latency measurements was taken for each rat per session.

Thermal HPW latency was measured by placing the rat on a clear glass surface heated to 30 ± 1°C.8An infrared light beam (Plantar test 7370; Ugo-Basile, Verese, Italy) was directed onto the middle portion of the plantar surface of the rat’s hind paw, and the time until the paw reflexively moved away from the beam was measured. The infrared intensity was set to produce a HPW of approximately 6 s. The stimulus was terminated at 18 s to prevent tissue damage. Compared with the TF stimulus, a longer cutoff time for HPW was permitted because the apparatus produces a thermal stimulus that reaches the pain tolerance level more slowly over time (as judged by the experimenter placing his fingertip over the stimulus). Five latency measurements were made per rat per session, with at least 2- to 4-min time intervals between trials. The average of the numeric middle three latency measurements was taken for each rat per session.

Initially, TF and HPW latencies were measured under isoflurane anesthesia at 0.8 MAC. However, it was found that spinally transected animals never exhibited TFs at 0.8 MAC. Therefore, after transection in five animals, we progressively decreased isoflurane concentrations by 0.2% increments until we observed TF latencies within the 8-s cutoff. We discontinued testing if no TF occurred down to 0.5% isoflurane (approximately 0.35–0.45 MAC).

Minimum alveolar concentration determinations were performed immediately before spinal transection or sham surgery and at postoperative days 1, 3, 7, 14, and 28. Anesthesia induction and intubation and monitoring of body temperature, end-tidal carbon dioxide, and anesthetic concentration were conducted as described in the section Surgery and Caretaking in Chronic Spinally Transected Animals.

The MAC values for both tail and forepaw stimulation were determined. For tail stimulation, we applied a supramaximal mechanical stimulus (30-cm hemostat that delivered 1.7 N/mm²) midway down the length of the tail. The clamp was applied and oscillated (rotated back and forth approximately ±30°) at approximately 2 Hz for 1 min or until gross purposeful movement was observed during the 1 min of clamping.9For forepaw stimulation, we applied a supramaximal electrical stimulus (60 mA, 100 Hz; model NS252J; Fisher and Paykel Healthcare, Auckland, New Zealand) delivered through platinum needle electrodes (Grass Instruments, West Warwick, RI) inserted into the ventral forepaw skin. Multilimb movement was interpreted as a positive response, whereas single-limb withdrawals and tonic limb or neck extensions were interpreted as a negative response.9In spinally transected animals, a positive response consisted of bilateral hind limb or forelimb movements in response to tail clamp or forepaw electrical stimulation, respectively, whereas unilateral limb withdrawals and tonic limb extensions were not considered positive movement responses. Depending on the response, the anesthetic concentration was increased or decreased by 0.2%. After an equilibration time of 15–20 min, the clamp or electrical stimulus was reapplied. This process was continued until two anesthetic concentrations were found that just permitted and just prevented movement. The average of these values was the MAC.9 

Tail stimulation was usually ineffective in eliciting movement in spinal transected animals, even at isoflurane concentrations at or below 0.6 MAC. Therefore, in addition, we determined MAC values for hind paw clamping (n = 8; using the same mechanical stimulus used for tail stimulation). If no movement was observed down to 0.5% isoflurane, testing was discontinued, and the rat’s MAC value was designated as 0.4%.

Acute terminal experiments were conducted on five adult male Sprague-Dawley rats (weight, 450–600 g) to investigate the effects of spinal cold block on movement elicited by supramaximal mechanical stimulation. Animals were anesthetized and prepared for surgery as described in the section Surgery and Caretaking in Chronic Spinally Transected Animals. In addition, cannulations of the carotid artery and jugular vein were performed to monitor blood pressure (model PB-240; Puritan-Bennett Corp., Hazelwood, MO) and for fluid administration, respectively. Blood pressure was always maintained above a mean arterial pressure of 75 mmHg, with lactated Ringer’s solution (Abbott Laboratories, Chicago, IL) administered intravenously when necessary. A laminectomy was performed at T8 to permit placement of a cooling probe to the cord. The spinal cold-block method was performed in the same manner as in our previous study,10using a custom-built cooling probe machined from a solid aluminum block. Ethanol was circulated through a copper coil embedded in dry ice and then to the probe using a variable-speed roller pump. We continuously monitored the temperatures of the dorsal and ventral surfaces of the spinal cord (directly below the probe) using two small thermistors (Physitemp, Clifton, NJ). After surgery was completed, the rat’s MAC value was determined before placing the rat in a stereotaxic frame for recording limb forces.11The cord was slowly cooled over a time period of 10–15 min. During spinal cold-block testing, the temperature of the dorsal surface of the cord was maintained between 0° and 3°C.

In chronic spinally transected animals at postoperative day 28 and in acute spinal cold-block experiments, the isometric forces of all four limbs in response to supramaximal tail, hind paw, and forepaw stimulation were measured by attaching each limb to a force transducer (model FT-03; Grass Instruments) via  a 1-0 silk suture. Noxious supramaximal mechanical stimulation (10-s duration) was applied every 3–4 min, and resulting responses were recorded at 0.6 and 0.8 MAC. In cold-block experiments, limb force measurements were taken before spinal cooling (at 0.8 MAC), during spinal cooling (at 0.6 MAC), and after rewarming the spinal cord (at 0.6 MAC). Limb force data were digitized at a rate of 200 Hz and collected on a personal computer using a Powerlab with Chart software (AD Instruments, Grand Junction, CO).

Changes in MAC values across pretransection and posttransection time points were compared using two-factor analysis of variance followed by post hoc  Tukey multiple comparisons. TF and HPW latencies were each compared across time points and between awake and anesthetized states using three-factor analysis of variance followed by post hoc  Tukey multiple comparisons. Limb forces (area under the curve and peak force) were normalized as percent of maximum for each animal and were compared using two-factor analysis of variance with post hoc  Tukey multiple comparisons. Single pairwise comparisons were made using a two-tailed, paired or unpaired t  test where indicated. P  values of less than 0.05 were considered statistically significant.

Before spinal transection, the mean MAC for tail clamping was 1.4 ± 0.1%. Three days after transection, the mean MAC value for tail clamping decreased to less than 0.5% isoflurane. That is, only three of eight rats showed movement in response to tail clamping at 0.5% isoflurane (i.e. , MAC = 0.6%), with the remaining five rats showing no movement down to 0.5%. This was followed by a small but significant partial recovery to mean MAC values of 0.7 ± 0.2% at both 14 days (P = 0.02) and 28 days (P = 0.014) after transection (fig. 1). During the posttransection period, mean hind paw clamp MAC values also showed a partial recovery over the 28-day posttransection period, with increases in MAC values from 0.6 ± 0.2% at 3 days, to 0.9 ± 0.1% at 14 days (P < 0.001), and 0.8 ± 0.1% at 28 days (P = 0.015) after transection (fig. 1). However, the partial recovery of MAC values for tail clamp remained significantly decreased from pretransection values (P < 0.00001). This comparison could not be made for hind paw clamping because pretransection MAC determinations were not conducted.

Mean MAC values for forepaw stimulation (1.4 ± 0.2%) in spinally transected animals remained unchanged throughout the 28-day testing period. In three sham-operated animals, MAC values for both forepaw stimulation (1.3 ± 0.2%) and tail clamping (1.4 ± 0.2%) remained unchanged during the 28-day period.

Despite a large decrease in tail clamp MAC values (to < 36% of control) at 3 days after transection, the TF reflex in awake rats was enhanced, with a significant decrease in latency from 4.2 ± 0.5 s before transection to 2.9 ± 1.0 s (P < 0.015) at 3 days after transection, followed by a significant recovery to 4.4 ± 0.8 s at 28 days after transection (P < 0.008; fig. 2A). Under 0.8 MAC isoflurane (using the pretransection MAC value for tail clamping), TF latency before transection was significantly delayed to 5.7 ± 1.5 s (P < 0.014) compared with latencies measured in awake rats. After spinal transection and under 0.8 MAC isoflurane, only one of eight rats exhibited a TF reflex, which only occurred at 3 days after transection. The remaining animals showed no TF up to the 8-s cutoff at all posttransection time points (fig. 2A). When isoflurane was decreased to 0.5% (approximately 0.35–0.45 MAC) and TF was tested in five animals at 28 days after transection, only two of the five animals exhibited TF reflexes, with latencies of 5.1 and 4.3 s.

In awake animals, HPW latencies remained unchanged throughout all pretransection and posttransection time points. HPW latencies before transection were significantly increased from 6.7 ± 1.0 s in the awake state to 9.4 ± 2.0 s under 0.8 MAC isoflurane (P < 0.021; fig. 2B). Spinal transection caused mean HPW latencies under isoflurane to further increase to 14.8 ± 3.4 s at 3 days (P < 0.001) and 16.2 ± 1.6 s at 7 days after transection (P < 0.0001) compared with pretransection HPW latencies under isoflurane (fig. 2B). This increase was followed by a partial but significant decrease in HPW latency at 14 days (to 11.7 ± 3.7 s; P < 0.003) and 28 days (to 12.9 ± 3.3 s; P < 0.05) compared with latencies measured at 7 days after transection (fig. 2B). Although we did not quantitatively measure the force of thermal HPWs in this testing procedure, we observed that under isoflurane, HPWs were clearly of much lesser magnitude at all posttransection time points compared with those measured before spinal transection. In fact, it was rare that the rat fully displaced its hind paw away from the infrared beam.

By 7 days after transection, all awake rats exhibited clear signs of spasticity, which included movements of tail and hind limbs, as well as flexion of the digits in response to innocuous stroking or manipulation of these appendages. This behavior persisted though the 28-day posttransection time point. Under isoflurane anesthesia, the hind limbs and tail became flaccid despite the presence of spasticity in awake animals.

In acute cold-block experiments, rats exhibited robust and synchronous movement of all four limbs in response to supramaximal noxious stimulation of the hind paw or tail, before cooling the spinal cord (individual example in fig. 3A, left traces). During spinal cold block, the force of hind limb movements in response to hind paw or tail clamping was nearly abolished at isoflurane concentrations down to 0.6 MAC (fig. 3A, middle traces). In these experiments, we could not perform testing at isoflurane concentrations below 0.6 MAC because of ethical concerns. We did not observe spontaneous movements, indicating that these animals were not conscious or perceiving pain. After rewarming of the spinal cord (to 34–36°C), vigorous limb movements in response to noxious tail and hind paw stimulation returned (fig. 3A, right traces). Mean limb forces were significantly depressed under spinal cold block, in response to both noxious tail (P < 0.001; fig. 3B) and hind paw stimulation (P < 0.004; fig. 3C).

In chronic spinal rats, hind limb movement forces elicited by noxious stimuli were nearly abolished, although weak ipsilateral hind limb withdrawals in response to hind paw clamping were still present at 0.8 MAC (individual example in fig. 4A, top traces). However, forelimb movement forces in response to noxious forepaw stimulation were bilateral and robust (fig. 4A, bottom traces). Mean force integrals for tail, hind paw, and forepaw clamping are shown in figure 4B. Because we could not obtain limb force data in these same animals before transection, mean hind limb forces in chronic spinal animals were compared with hind limb forces measured in animals used for the spinal cold-block experiments. In 28-day chronic animals, hind limb forces elicited by hind paw and tail clamp, at both 0.6 and 0.8 MAC, were significantly lower than in separate intact animals (P < 0.007 in all cases, unpaired two-tailed t test). However, the hind limb forces elicited by tail and hind paw clamping in chronic animals were not significantly different from those exhibited by animals during acute spinal cold block (P = 0.78).

The current results extend previous studies regarding supraspinal versus  spinal contributions to the immobilizing action of volatile anesthetics.2,5–7Spinal transection in chronic rats drastically reduced MAC values to 40% or less at 3 days after transection, followed by a partial recovery of MAC (to 60% of control) occurring between day 3 and day 14 after transection (fig. 1) During a 28-day period, TF and HPW latencies were either facilitated or unchanged (fig. 2), indicating that the decreases in isoflurane requirements were not due to a baseline motor depression such as that seen during spinal shock. By 7 days after transection, all animals showed signs of spasticity. However, under isoflurane anesthesia, TF and HPW were greatly depressed or absent. The parallel depression in both MAC and spinal reflexes under isoflurane anesthesia suggest that these two types of motor responses might have some degree of overlap in their respective circuitry or at least share a common site of anesthetic action. This is consistent with our previous studies,12,13which show that hind limb withdrawal force is greatly reduced or abolished in each animal’s immediate isoflurane peri-MAC range. Therefore, hind limb withdrawal seems to be a strong predictor of MAC. However, because reflex latencies remained unchanged over time while MAC showed partial recovery, anesthetic effects on the neuronal plasticity underlying the two types of motor responses might differ. Finally, forces generated by limb movements were profoundly depressed (by ≥ 90%) during acute reversible spinal cold block (fig. 3) and in 28-day chronic spinal rats (fig. 4). During spinal cooling, the forces of limb movements were nearly abolished at 0.6 MAC, and none of these movements would be considered a positive response for the purposes of MAC determination.

Although the current data are consistent with removal of descending supraspinal influences, we cannot exclude the possibility that neuronal plasticity below the transection could have rendered the spinal cord more sensitive to anesthesia, while leaving the reflexes intact in the awake state. Such a change could result in decreased activity of spinal neurons through increased glycinergic activity14or decreased alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid-type glutamate receptor expression.15However, reflex latencies remained unchanged in the awake state, and movement suppression by spinal cooling under anesthesia recovered immediately after rewarming the spinal cord. This would suggest that loss of descending projections might play a relatively greater role in decreasing anesthetic requirements and movement force in the very early stages after spinal cord trauma, with a possibly greater role for plasticity at later stages.

The decreases in MAC and limb forces after spinal transection or cooling were not likely attributed to the animal’s inability to exhibit bilateral limb movements in the awake state. We noted that all awake spinal-transected animals were capable of exhibiting robust, bilateral hind limb movements in response to tail pinching (usually alternating but sometimes a synchronous “kicking” movement). At the 7-day and later time points, all animals exhibited bilateral limb movement in response to nonnoxious manipulation of the limbs, stroking the hindquarters, and general handling. Hence, all rats were quite capable of exhibiting vigorous, bilateral movements in the awake state. Many previous studies have indicated that central pattern generators controlling hind limb stepping are distributed within the lower thoracic–lumbar segments,16–18well below the T8 segment transected or cooled in the current study.

The current results seem to be at odds with one study in which low cervical (C7) hypothermic transection (freeze-lesioning) of the spinal cord caused no significant changes in isoflurane MAC values.2However, single hind limb withdrawals in that study were considered “purposeful” movement, whereas the typical methodology for MAC determination, used in the current study, does not include single limb withdrawals as a positive response. All animals had thermal HPWs (albeit of weak magnitude) at 0.8 MAC, and furthermore, during MAC determination with hind paw clamping, all rats had weak to moderate withdrawals at 80–90% of pretransection MAC values (also see force traces showing hind limb withdrawals at 0.8 MAC in figs. 3A and 4A). Therefore, the current data seem consistent with those of Rampil.2However, the movements in that study were “less vigorous but of the same type,” and it is therefore unclear what types of movements were observed and how they compare with movements typically observed in a MAC study and in the current study. This confusion could stem from the limitations of MAC determination, which classifies movement into an “all-or-none” response based on subjective evaluation of limb and neck movements. The current study has the added advantage of multilimb force measurement, which objectively demonstrates how spinal transection changes the magnitude and pattern of limb movements elicited by a supramaximal noxious stimulus.

The current results are consistent with previous studies from our laboratory and of others’ studies addressing the role of supraspinal sites in anesthetic-induced immobility.5–7In goats, isoflurane MAC was reversibly reduced from 1.4% to 0.8% (in the torso) when isoflurane concentration was selectively decreased in the cranial circulation.5In our recent study,13isoflurane caused a large depression of spontaneous and noxious heat–evoked activity of rostral ventromedial medullary ON cells, which are believed to facilitate nociceptive reflexes through descending projections to the spinal cord.19–21A previous study reported that increases in TF latency in isoflurane-anesthetized spinal animals results from removal of a supraspinal α¹adrenoceptor–mediated descending facilitation.6Because α¹-adrenoceptor activation in the rostral ventromedial medulla excites ON cells, decreases TF latency, and mediates certain states of hyperalgesia,22–25isoflurane could engage supraspinal adrenergic pathways that relay in the rostral ventromedial medulla to excite ON-cells and thus facilitate spinal nociceptive sensorimotor circuits. However, the effects of isoflurane on other descending facilitatory pathways cannot be excluded.

Although the current study did not assess the role of specific descending pathways in the spinal cord, the decrease in movement after spinal transection and cooling would more likely be attributed to interruption of descending fibers in the ventral cord, which primarily mediate facilitation of nociceptive reflexes and locomotion.26–31However, a smaller contribution from dorsolateral pathways cannot be excluded.29For the cold-block experiments conducted in the current study, we found that approximate ventral cord temperatures below approximately 12°C were necessary to prevent movement of the forelimbs in response to noxious mechanical stimulation, indicating that ascending nociceptive transmission in ventrolateral pathways were blocked. When responses to noxious stimuli were tested at slightly higher ventral cord temperatures (fig. 3A), not only did forelimbs exhibit movement, but the pattern of movement changed from none or tonic flexion to the larger, repetitive “gross and purposeful” movements, suggesting that activity descending along ventral pathways was largely responsible for generating this type of movement.

Although previous studies have shown depressant actions of volatile anesthetics on dorsal horn neurons,32–38an increase in the immobilizing potency of isoflurane after spinal transection was probably not a result of changes in dorsal horn neuronal activity. First, spinal transection or selective dorsolateral funiculus blockade causes an increase rather than a decrease in noxious stimulus–evoked discharges in dorsal horn neurons,39–42and we recently confirmed this same effect on dorsal horn neurons under isoflurane anesthesia during spinal cold block.10Second, several of our previous studies have shown that isoflurane has little to no effect on nociceptor-driven responses of dorsal horn neurons in the peri-MAC (approximately 0.8–1.2 MAC) concentration range where noxious stimulus–evoked movement is abolished.10,12,43,44Therefore, both the isoflurane-induced descending facilitation and the direct spinal depressant action of isoflurane seem to target more ventrally located spinal interneurons or motoneurons. Several previous studies have shown depressive effects of volatile anesthetics on motoneurons.45–48 

Changes in spinal cord function after spinal cord injury occur not only from immediate removal of descending influences but also as a result of longer-term neural plasticity that underlies states of motor hyperreflexia/spasticity3,4and, if the lesion is above T7, autonomic hyperreflexia.49In the current study, we found that although MAC values remained significantly reduced during the entire 28-day posttransection period, they nonetheless showed partial recovery between the 3- and 28-day time points. However, TF and HPW latencies in awake animals were unchanged from baseline levels, and TF latencies were only facilitated at the 3-day time point when MAC values were at their lowest. This mismatch might be explained by the two different types of stimuli used for reflex latency measurement (thermal stimulus) and MAC determination (mechanical stimulus) or by differences in stimulus intensity (near-threshold noxious vs.  supramaximal noxious, respectively). Whereas noxious thermal stimuli primarily activate cutaneous C fibers, with a lesser contribution from Aδ fibers,50,51the mechanical stimulus would tend to recruit both Aδ and C nociceptors as well as low-threshold Aβ fibers that innervate both cutaneous and deep tissues. The onset of spasticity in awake animals coincided with partial recovery of MAC values, and thus sensitization of reflex pathways receiving low-threshold primary afferent input might have contributed to MAC increases over time in spinally transected animals. The increases in spinal excitability and MAC values could have resulted from increases in interneuronal or motoneuron excitability52–54or primary afferent sprouting55that coincides with spasticity/hyperreflexia in chronic spinal rats. There are a plethora of molecular and cellular changes that occur after spinal cord injury that could potentially lead to sensitization of spinal pathways, including altered expression of N -methyl-d-aspartate–type glutamate receptors,56neurotrophins,57inflammatory mediators, and voltage-gated sodium channels.58 

It is possible that some of the partial recovery in MAC was attributed to the presence of hyperalgesia after repeated stimulation of the tail and hind paw, because there was a slight albeit insignificant increase in forepaw MAC values over time. However, the lack of changes in reflex latencies and the relatively long testing intervals (1–2 weeks) suggest that confounding influences from repeated stimulation, if any, were minimal.

Isoflurane requirements and limb movement forces are drastically reduced by transecting or cooling the spinal cord, in the absence of spinal shock, and are thus consistent with the notion that the immobilizing potency of isoflurane in rats is determined by the net result of opposing actions from spinal and supraspinal sites. Therefore, future studies should focus on both spinal and supraspinal anesthetic actions and interactions. The potent spinal depressant effect found in the current study invites the development of anesthetics that target the spinal cord without effect on supraspinal facilitatory centers.

The results further suggest that anesthetic requirements are much lower in patients with spinal cord injuries if surgery is performed below the level of the lesion. However, the exact requirements for a spinal cord injured individual probably depend on the extent and location of the spinal injury and the amount of time since the injury. In humans, changes in motor and autonomic reflex excitability can occur months to more than a year after spinal cord injury,59,60far later than the 28-day testing period used in the current study. Therefore, it is not known how these later changes might further influence anesthetic requirements. Future studies are necessary to unveil the specific neuronal populations and molecular mechanisms involved in both supraspinal and spinal actions of volatile anesthetics and how reorganization of neural pathways after spinal cord injury influences anesthetic requirements.

The authors thank Richard Atherley, B.S., Milo Bravo, B.S., Cindy Clayton, D.V.M., and Justine Overman, B.S. (Technicians, Department of Anesthesiology and Pain Medicine, University of California, Davis, California), for their expertise and assistance with experimentation in chronic spinally transected animals; and Earl Carstens, Ph.D. (Professor, Section of Neurobiology, Physiology, and Behavior, University of California, Davis), for the use of his behavioral testing equipment.