Neurocirculatory Responses to Sevoflurane in Humans: A Comparison to Desflurane

With approval from the Institutional Review Board and after informed consent, 22 healthy, normotensive, male volunteers, aged 19-32 yr, were studied. The administration of sevoflurane was part of an Abbott Laboratories (Abbott Park, IL)-sponsored study carried out with an investigational new drug exemption issued from the Food and Drug Administration. Volunteers were nonsmokers, free of systemic illness, and not taking medications or illicit drugs and had fasted for at least 12 h before testing. Volunteers were studied while supine after ingesting 30 ml of nonparticulate antacid (Bicitra). HR was monitored from leads II and V5 on the electrocardiogram. A 20-G catheter was inserted into the radial artery for determination of blood pressure (BP). An 18-G catheter was inserted into a forearm vein and 5 ml/kg of 0.9% saline was administered over 10 min and maintained at 1 ml *symbol* kg sup -1 *symbol* h sup -1.

Forearm blood flow was measured with a Hg-in-Silastic, temperature-compensated, strain-gauge plethysmograph system in which an upper arm cuff intermittently inflated to 60 mmHg (10 s on and 10 s off) while the venous outflow from the hand was arrested. [14]Forearm vascular resistance was calculated as the ratio of mean arterial pressure (MAP) to forearm blood flow.

Plasma norepinephrine (Nepi) concentrations were determined in duplicate from arterial blood by high-pressure liquid chromatography with electrochemical detection. The minimal detectable concentration of Nepi was 35 pg/ml, and the minimal detectable change of Nepi was 15 pg/ml, based on the variability of repeated measurements from pooled plasma. Blood samples were collected in chilled syringes, transferred to glass tubes containing EDTA and reduced glutathione, and immediately spun at 4 degrees Celsius and 20,000 rpm for 10 min. The plasma was pipetted into polypropylene containers and stored at -70 degrees Celsius until analysis.

As described previously, [15]the common peroneal nerve of the right leg, just distal to the knee, was located via external stimulation, and a needle was positioned within the peroneal nerve, guided by the response to small electrical impulses applied to the needle (1 Hz, 0.02 mA). When the needle impaled a nerve fascicle supplying muscle, a distinct muscular contraction occurred. For the purpose of this study, skin nerves were excluded because they influence only about 4% of the cardiac output, are not under baroreceptor control, and have mixed effector sites (skin blood vessels, sweat glands, and piloerector muscles). Characteristic spontaneous bursts of muscle sympathetic neural activity (SNA) were sought by subtle advancement or withdrawal of the needle within the muscle nerve fascicle. Once an acceptable recording was obtained, the subjects remained relaxed and quiet with an immobile leg so as not to alter the location of the needle within the nerve.

Once an acceptable sympathetic nerve recording was obtained, a 10-min quiet rest period was observed followed by 5 min of hemodynamic (HR, BP, forearm blood flow) and neural (SNA) measurements. Arterial blood was obtained for blood gas and Nepi analysis. The face mask was gently placed, a priming dose of vecuronium (0.01 mg/kg) was given, and 100% Oxygen²was administered for a period of 5 min. End-tidal carbon dioxide and anesthetic concentrations after induction were monitored by an Ohmeda 5250 Infra-red Respiratory Gas Monitor (Madison, WI). Twelve subjects were randomized to receive sevoflurane, and nine received equipotent concentrations of desflurane. Anesthesia was established with propofol (2 mg/kg) and neuromuscular blockade accomplished with vecuronium bromide (0.15 mg/kg). Ventilation was controlled via the mask and without an oral airway for 12 min at a fresh gas flow rate of 6 l/min in a partial rebreathing system. Precisely 2 min after injection of propofol, the sevoflurane vaporizer was activated at a setting of 1%. In the two subsequent 1-min periods, the vaporizer was increased to 2% and 3% and maintained at 3% while end-tidal carbon dioxide concentrations were kept at awake values. In subjects receiving desflurane, a similar sequence of anesthetic administration was employed by delivering 3%, 6%, and 9% inspired concentrations (1.0 MAC for sevoflurane and desflurane in this age group is 2.4 and 7.2%, respectively). On completion of the 12-min period after anesthetic induction, the trachea was intubated by a skilled anesthesiologist (C.W.L.) while data were continuously collected. Ventilation was mechanically controlled to maintain end-tidal carbon dioxide at awake levels. To reduce the end-tidal concentration to 0.41 MAC, the vaporizer setting was reduced to 1% sevoflurane or 3% desflurane for an additional 20-min period of anesthesia, providing an interval of 32 min postinduction before steady-state hemodynamic and neural data were collected. This reduced the likelihood that the initial administration of propofol and the neuroendocrine responses during the induction period influenced steady-state measurements.

Hemodynamic and neural measurements and blood sampling were carried out 10 min after end-tidal sevoflurane or desflurane had reached the desired concentration. The order of anesthetic administration for determination of steady-state responses was 1%, 2%, and 30% sevoflurane and 3%, 6%, and 9% desflurane. In addition, continuous data were recorded during the first 5 min (transition) after rapidly increasing the delivered anesthetic concentration. The transition was carried out simply by advancing the sevoflurane vaporizer setting from 2% to 3% or the desflurane vaporizer from 6 to 9%. An over-pressure paradigm, i.e., advancing the vaporizer beyond the desired concentration to achieve a more rapid steady-state, was not employed. End-tidal carbon dioxide was maintained constant throughout the experimental protocol and confirmed by arterial blood gas analysis at each steady-state measurement period. Blood sampling for Nepi determination was carried out at conscious baseline and at steady-state periods of 0.41, 0.83, and 1.24 MAC.

To determine the peak change in HR and MAP during intubation, the 20 s immediately before intubation was averaged, and the maximum HR and MAP that occurred in the 60 s after intubation were determined individually. The SNA occurring from baseline (preintubation) up until the time of peak MAP was averaged to determine the maximal SNA response. Consecutive hemodynamic and neural measurements were compared with analysis of variance for repeated measures, and post hoc analyses were performed with Dunnett's t tests. Peak changes in variables during induction and intubation were analyzed with unpaired Student's t tests. Probability values that were less than 0.05 were considered sufficient to reject the null hypothesis.

The age, height, weight, and body mass index of the subjects did not differ between groups. There were no differences in hemodynamic measurements and SNA between groups at conscious baseline (Figure 1). Responses to anesthetic induction with propofol and the subsequent 10-min administration via mask of sevoflurane or desflurane are depicted in Figure 1and Figure 2. Propofol administration resulted in significant reductions in SNA and MAP and a significant increase in HR, and these responses did not differ between groups. The inhalation via mask of sevoflurane in incremental steps from 1% to 2% and 2% to 3% (inspired) did not lead to significant changes in SNA or HR. There was a gradual and significant reduction in MAP during the 10-min administration period (Figure 1and Figure 2). In contrast, the step increase in the inspired concentration of desflurane from 3% to 6% and 6% to 9% led to significant increases in SNA and HR and a biphasic response in MAP. The peak absolute increases in these variables from the propofol baseline are depicted in Figure 2.

Laryngoscopy and subsequent tracheal intubation began 10 min after beginning the administration of the volatile anesthetic via mask. At this time, the end-tidal desflurane concentration had achieved 7.9% (approximately 1.1 MAC; Figure 1), whereas the end-tidal concentration of sevoflurane was 2.1% (approximately 0.88 MAC). The subsequent hemodynamic responses to laryngoscopy and tracheal intubation are depicted in Figure 3. The HR and MAP responses did not differ between groups, whereas the SNA response was less in the desflurane group.

The steady-state responses to sevoflurane and desflurane are depicted in Figure 4. There were similar and progressive decreases in MAP and forearm vascular resistance with increasing MAC multiple of each anesthetic. There were no changes in HR with increasing concentration of sevoflurane or at 3% and 6% desflurane, whereas 9% desflurane resulted in a significant increase in HR compared to 6% desflurane. Sevoflurane did not alter central venous pressure (CVP), SNA, or plasma Nepi concentrations at any end-tidal concentration level; however, desflurane resulted in significant and progressive increases in CVP, SNA, and Nepi with increasing MAC multiples (P < 0.05; Figure 4).

The minute-by-minute responses to the rapid increase in the inspired concentration of desflurane and sevoflurane ("transition") are depicted in Figure 5. The increase was initiated from either 2% or 3% sevoflurane. Rapid increases in the inspired concentration of sevoflurane did not alter HR or SNA but reduced MAP, whereas rapid increases in the inspired concentration of desflurane were associated with large and significant increases in SNA, HR, and MAP.

The major findings of this study were as follows: (1) sevoflurane anesthesia was not associated with significant increases in HR or sympathetic nerve activity (SNA) at any steady-state period of anesthesia, whereas higher concentration of the inspired desflurane was associated with increases in these variables; (2) in contrast to desflurane, a rapid increase in the inspired concentration of sevoflurane was not associated with increases in SNA, HR, or BP.

Neurocirculatory variables at rest did not differ between groups (Figure 1and Figure 4). Propofol was used to induce anesthesia and resulted in a significant reduction in SNA and MAP and an increase in HR consistent with our previous observations with propofol. [16]Moreover, we have shown that the sympathoinhibitory effects of propofol reduce the neurocirculatory activation associated with the administration of desflurane. [17].

The 10-min administration of sevoflurane by mask after propofol administration did not alter SNA, HR, or CVP, but there was a gradual decrease in MAP that probably represents a direct effect of sevoflurane on vascular smooth muscle. In contrast, the administration of desflurane via mask after propofol resulted in substantial increases in HR, CVP, and sympathetic outflow and a biphasic MAP response. Although the stimulus for this neurocirculatory activation is unknown, we speculate that the response to the markedly different pungencies of these anesthetics might explain the difference.

During the administration of these anesthetics via mask, the alveolar (end-tidal) rate of rise of sevoflurane was slower than that of desflurane despite identical fresh gas flows and identical (with respect to MAC-equivalent) increases in the delivered concentrations of these two gases. The slightly slower increase in the end-tidal concentration of sevoflurane might be due in part to the slightly greater blood-gas partition coefficient of sevoflurane compared to desflurane (0.60 vs. 0.42) and perhaps due, to a small extent, to the greater degree of metabolism of sevoflurane (3% vs. 0.02%). The more gradual increase in the alveolar concentration of sevoflurane resulted in achievement of a lower MAC-equivalent of sevoflurane at the time of tracheal intubation. This is a likely explanation for the larger sympathetic consequences of laryngoscopy and tracheal intubation in the sevoflurane group. Despite this difference, the peak MAP response to intubation in the sevoflurane group was not significantly different from the desflurane group.

At 0.41 and 0.83 MAC, neither sevoflurane nor desflurane was associated with changes in HR from the respective conscious baseline level. However, both anesthetics produced similar progressive decreases in BP and forearm vascular resistance without significant changes in SNA or Nepi concentrations. This suggests that both agents have important direct effects on vascular smooth muscle such that even relatively small concentrations of these agents lead to reductions in vascular resistance and BP. Although sevoflurane has not been compared to desflurane in other adult human studies, a study in pediatric patients concludes that sevoflurane maintains BP better than does desflurane. [9]The majority of animal studies with sevoflurane have employed isoflurane as the reference anesthetic. [10,12,18]In dogs, either a similar [10]or a larger [18]BP-lowering effect of sevoflurane has been noted in comparison to isoflurane.

Similarly, the vascular resistance effects of sevoflurane have been compared only to isoflurane, and conclusions have been mixed, depending on the species studied. [10,12,13,19]One approach in rats employed equipotent concentrations of either sevoflurane or isoflurane to lower MAP to 50 mmHg. [12]At this level of hypotension, systemic vascular resistance was maintained better in the rats receiving sevoflurane than in those receiving isoflurane.

There was a progressive and significant increase in CVP with increasing concentrations of desflurane but no change in CVP in subjects receiving sevoflurane. The increase in CVP could be due to several effects of desflurane that include pulmonary hypertension, [20]myocardial depression, [21]or venoconstriction. In chronically instrumented dogs, the myocardial depressant effects of sevoflurane and desflurane were indistinguishable. [11].

At the 1.24 MAC steady-state measurement period, there was a significant increase in basal HR, CVP, SNA, and Nepi in subjects receiving desflurane but no change in these variables in subjects receiving sevoflurane. Moreover, the rapid increase in the inspired concentration of desflurane from 6% to 9% resulted in large increases in SNA, HR, and MAP, whereas a similar rapid increase in the inspired concentration of sevoflurane was uneventful. Previous work also has indicated that the rapid increase in the inspired concentration of desflurane results in 10- to 15-fold increases in plasma epinephrine and vasopressin. [7]These data reveal a substantial effect of desflurane that initiates a significant neurocirculatory response. The initial concern that sevoflurane also might produce tachycardia and increase sympathetic outflow was not substantiated in the current study. Prior data from pediatric patients aged 3-12 yr [9]and adult patients, ASA physical status 1 and 2, undergoing elective surgery and receiving sevoflurane [8]indicated that HR was increased 5-10 bpm in tracheally intubated patients before surgical stimulation. The only clear evidence for sevoflurane-mediated tachycardia during steady-state periods of anesthesia has been in chronically instrumented dogs. [10,11]However, these data must be interpreted cautiously because of species differences in the effects of volatile anesthetics on the autonomic nervous system. For example, numerous studies in dogs have failed to demonstrate an effect of desflurane on the autonomic nervous system, [22-24]whereas in humans, the administration of desflurane has been associated consistently with large increases in sympathetic outflow, catecholamines, vasopressin, HR, and BP. [5-7,25]Other work has indicated that the neurocirculatory excitation associated with rapid increases in the inspired concentration of desflurane lessens with repeated exposures. [26]Because of the initial large increase in the desflurane concentration on induction, it is conceivable that the neurocirculatory response recorded during the transition period might have been less than maximal because it was the second application of a large increase in the inspired concentration to the desflurane-treated subjects.

At steady-state measurement periods, the comparisons of the neurocirculatory responses to the anesthetics were unquestionably valid. However, comparisons are limited partially during periods of rapid changes in the inspired concentration. This limitation is due to the fact that the alveolar concentration (as a percent of minimum alveolar concentration) at intubation and the rate of rise of the alveolar concentration during induction and transition periods were greater with desflurane. However, it is unlikely that these small differences explain the lack of sympathetic activation with sevoflurane, because earlier data indicate that a more rapid increase in the delivered concentration of sevoflurane has not been associated with neurocirculatory excitatory hemodynamic responses, [9,27]whereas even smaller rates of increasing the inspired concentration of desflurane have led to substantial neurocirculatory excitation. [17,25].

In summary, the neurocirculatory effects of sevoflurane anesthesia were unremarkable. In contrast to desflurane, the administration of sevoflurane via mask and the rapid increase in the inspired concentration of sevoflurane did not lead to changes in HR, MAP, sympathetic outflow, or norepinephrine concentrations. At steady-state periods, sevoflurane and desflurane were similar in their ability to reduce MAP and forearm vascular resistance, whereas higher concentrations of desflurane were associated with increased HR, CVP, SNA, and Nepi. These data, along with other findings that sevoflurane does not produce coronary steal, [28]suggest that the likelihood of untoward cardiovascular events with sevoflurane anesthesia might be negligible.