Isoflurane Produces Marked and Nonlinear Decreases in the Vasoconstriction and Shivering Thresholds

With approval of the Committee on Human Research at the University of California, San Francisco, we studied eight male volunteers having the following morphometric characteristics: age 28+/-5 yr; height 174 +/- 7 cm; and weight 66+/-8 kg. None was obese, was taking medication, or had a history of thyroid disease, dysautonomia, or Raynaud's syndrome.

Volunteers were each evaluated on four days:(1) a target end-tidal isoflurane concentration of 0.55%, (2) a target concentration of 0.70%. (3) control (no anesthesia) and a target end-tidal concentration of 0.85%, and (4) a target end-tidal concentration of 1.00%. Thresholds were thus determined without anesthesia and at an isoflurane concentration of 0.85% on a single day. We were able to combine the two determinations and still retain good circadian control because vasoconstriction and shivering were easy to obtain without anesthesia. The treatment order was randomly assigned and at least two days were allowed between each study day. All studies were conducted in the spring of 1995.

Volunteers fasted 8 h before arriving at the laboratory, were minimally clothed, and rested supine in a 22–23 degrees C room during the protocol. Studies were scheduled so that thermoregulatory responses in each volunteer were triggered at similar times to minimize circadian fluctuations. A catheter was inserted into a left forearm vein for fluid administration. Lactated Ringer's solution was initially infused at [nearly equal] 100 ml/h. Throughout the protocol, arms were protected from active warming and cooling to avoid locally mediated vasomotion. [7]However, all other skin below the neck was similarly manipulated.

Anesthesia was induced without any premedication by infusion of propofol ([nearly equal] 2 mg/kg) and incremental concentrations of isoflurane. A bolus of lactated Ringer's solution ([nearly equal] 10 ml/kg) was administered during induction of anesthesia; subsequently, fluid was again administered at a rate of [nearly equal] 100 ml/h. The volunteers breathed spontaneously, either via a face mask or laryngeal mask; however, ventilation was assisted when necessary to maintain end-tidal P^CO2near 37 mmHg.

We took care throughout the protocol to minimally stimulate the volunteers. For example, ambient lighting was dimmed and extraneous noise was avoided. Skin temperature changes were restricted to less or equal to 3 degrees C/h because this rate is unlikely to trigger dynamic thermoregulatory responses. [8]And finally, we handled the face mask gently and avoided moving the volunteers during anesthetic administration.

Skin and core temperatures were gradually decreased, using a circulating-water mattress and forced-air cooler until significant vasoconstriction was achieved. [9]The thermoregulatory study ended each day when shivering was detected, but the volunteers subsequently participated in a brief study evaluating the effects of alfentanil administration on pupillary responses (unpublished data).

Core temperature was recorded from the tympanic membrane (Mallinckrodt Anesthesiology Products, St. Louis, MO). Tympanic membrane temperature and distal esophageal temperatures correlate well under the circumstances of this study. [6]Mean skin-surface temperature and cutaneous heat transfer were calculated from measurements at 15 area-weighted sites. [10,11]Temperatures were recorded at 1-min intervals from thermocouples connected to Iso-Thermex thermometers having an accuracy of 0.1 degree C (Columbus Instruments, Columbus, OH).

Absolute right middle fingertip blood flow was quantified using venous occlusion volume plethysmography at 5-min intervals. [12]A sustained decrease in fingertip blood flow to < 1.0 ml/min identified the beginning of vasoconstriction. Heart rate and blood pressure were determined oscillometrically at 5-min intervals (Modulus CD, Ohmeda, Salt Lake City, UT). End-tidal isoflurane and P^CO2were measured using an Ohmeda Rascal monitor.

Electromyographic activity was used to quantify shivering. After mild skin abrasion and degreasing, silver/silver chloride monitoring electrodes were positioned to record the electrical activity of the right pectoralis and rectus abdominis, and the quadriceps bilaterally. The active electrodes were positioned 4 cm apart and oriented in the direction of the muscle fibers. [13]After appropriate amplification (Model P511, Grass Instruments, Quincy, MA), the signals were recorded on a thermoelectric printer having a linear resolution to 1,000 Hz (Dash-4, Astro-Med, West Warwick, RI). Onset of shivering was subsequently determined by an investigator blinded to treatment and core temperature: synchronous waxing-and-waning activity sustained for 5 min identified significant shivering. [14].

Hemodynamic responses, ambient temperature, and end-tidal isoflurane and P^CO2on each study day were first averaged within each volunteer, the resulting values were then averaged among volunteers. Results for each study day were compared using repeated-measures analysis of variance and Dunnett's tests for comparison to control (0% isoflurane).

The cutaneous contribution to vasoconstriction and shivering is linear. [15]We thus used measured skin and core temperatures in degrees centigrade at each threshold to calculate the core-temperature threshold that would have been observed had skin been maintained at a single designated temperature:Equation 1where the fractional contribution of mean skin temperature to the threshold was termed beta. T (^core)(calculated) thus equals the measured core temperature, T^core, plus a small correction factor consisting of beta/(1-beta) multiplied by the difference between actual (T^skin) and designated [T (^skin)(designated)] skin temperatures. We have previously described the derivation, validation, and limitations of this equation. [2]We used a beta of 0.2 for vasoconstriction and shivering. [15]The designated skin temperature was set at 34 degrees C, a typical intraoperative value.

Response thresholds, the vasoconstriction-to-shivering range (difference between the vasoconstriction and shivering thresholds), hemodynamic, and respiratory data were compared using repeated-measures analysis of variance and Dunnett's test for comparison to control. First-order regression and residual analysis were used to evaluate linearity of the threshold dose-dependence. Results are presented as mean +/- SD; P < 0.05 was considered statistically significant.

There were no statistically significant or clinically important differences in ambient temperature, relative humidity, heart rate, or end-tidal P^CO2on the five study days. Mean arterial blood pressure was reduced during isoflurane administration, but the decrease was not clinically important (Table 1).

Isoflurane administration produced a dose-dependent reduction in the vasoconstriction and shivering thresholds, decreasing each [nearly equal] 4.6 degrees C at an end-tidal concentration of 1%. The vasoconstriction-to-shivering range was 1.5+/-0.8 degree C without isoflurane, and did not change significantly during isoflurane administration (Figure 1and Table 2). Residual analysis indicated that the vasoconstriction and shivering thresholds decreased nonlinearly during isoflurane administration (Figure 2).

The normal vasoconstriction-to-shivering range (difference between the vasoconstriction and shivering thresholds) is [nearly equal] 1 degree Celsius. [8]Propofol, [2]desflurane, [5]alfentanil, [3]and clonidine [16]each reduce the vasoconstriction and shivering thresholds comparably. That is, the vasoconstriction-to-shivering range remains normal, although both thresholds are reduced by drug administration. Isoflurane, like the other sedatives and general anesthetics, reduced the vasoconstriction and shivering thresholds comparably. These data suggest that the major cold-defenses are similarly integrated and generally respond synchronously. The only apparent exception to this rule is meperidine, which reduces the shivering threshold twice as much as the vasoconstriction threshold. [17]This special antishivering activity of meperidine [18,19]may be mediated at the level of the spinal cord by the drug's kappa-agonist activity. [20].

Isoflurane produced a nonlinear reduction in the vasoconstriction threshold, a result contrasting with with that of our previous report. [6]We must therefore ask why the results from these two investigations are divergent. The major reason appears to be that the methods in the two studies differed significantly in terms of thermoregulatory sophistication and control of potentially confounding factors.

An important difference between our current and previous evaluations of the vasoconstriction threshold during isoflurane anesthesia is that the original study was conducted in patients during surgery. Surgical stimulation almost inevitably confounds intraoperative studies. Painful stimulation during 0.8 minimum alveolar concentration enflurane anesthesia increases the vasoconstriction threshold, [21]although the dose-dependence of this effect has yet to be established. In addition, we have noticed that even mild stimulation, such as adjusting a face mask, can trigger vasoconstriction at very low anesthetic concentrations. Therefore, it seems likely that the confounding effect of painful stimulation is itself nonlinear. To the extent that this proves true, dose-response curves developed in volunteers should be extrapolated to surgical patients with caution.

An additional difference is that the current study evaluated relatively low isoflurane concentrations whereas most patients in our previous investigation were given higher concentrations. Isoflurane concentrations less or equal to 1% were used in the current protocol because it would have been difficult to determine the shivering thresholds at higher concentrations. A consequence of this design, however, is that the two studies evaluate somewhat different aspects of the dose-response curve.

Patients in our previous study [6]were randomly assigned to the highest or lowest isoflurane concentrations they could tolerate while maintaining adequate blood pressures. This scheme differs from a full randomization in which specific isoflurane concentrations are prospectively assigned, and may cause a systematic bias. That is, patients who require less than typical amounts of anesthesia will tolerate the lowest concentrations but likely have substantial thermoregulatory impairment. In contrast, those patients requiring more than typical amounts of anesthesia will tolerate high concentrations with perhaps little associated thermoregulatory inhibition. The result is a potential “linearizing” effect whereby thresholds at low anesthetic concentrations are artifactually reduced and those at high anesthetic concentrations are artifactually elevated.

Because both skin and core temperatures contribute to central control of vasoconstriction and shivering, thermoregulatory response thresholds would ideally be determined at a given mean skin temperature. At the time of our original study, [6]minimally invasive human methods for keeping sentient skin temperature constant while significantly varying core temperature did not exist. Consequently, we recorded mean skin temperature, but simply considered the vasoconstriction threshold to be the core temperature triggering vasoconstriction. Even now, no method has been developed that allows skin temperature to be maintained over a range of core temperatures sufficient to test isoflurane concentrations between 0 and 1%. However, the contribution of skin temperature to central control was recently shown to be [nearly equal] 20% in humans. [8]Using this value (and Equation 1) allows us to compensate for changes in skin temperature and report thresholds at a constant skin temperature.

The thresholds we currently report were corrected for skin temperature using a designated value of 34 degrees C. When we applied a similar compensation to our previous data, it had virtually no effect at low anesthetic concentrations because measured skin temperatures were near 34 degrees C. However, calculated thresholds were [nearly equal] 0.5 degree C less than observed core temperatures at the higher anesthetic concentrations because skin temperatures in these patients was 2–3 degrees C less than 34 degrees C. Failure to correct for the effects of skin temperature on thermoregulatory control thus produced a distinct linearizing bias: thresholds at low anesthetic concentrations were unchanged, but those at the highest concentrations were artifactually increased (compared with corrected values). These results emphasize the importance of controlling skin temperature during thermoregulatory protocols, or of compensating for experimentally induced changes when control is not possible.

Conversely, our method of calculating core temperature thresholds at designated skin temperature requires the assumption that the cutaneous contribution to control of vasoconstriction and shivering (beta) remains constant. However, beta has not been determined during volatile anesthesia. To the extent that beta varies with isoflurane dose, our results will thus be erroneous. The coefficient of cutaneous contribution has also not been determined for the range of skin temperatures used in this study, another potential source of error.

In summary, the vasoconstriction-to-shivering range remains unchanged by isoflurane administration. In this regard, the effects of isoflurane are similar to those of desflurane, propofol, and alfentanil. We previously reported that isoflurane administration linearly decreases the vasoconstriction threshold in surgical patients. This study was only the third specifically quantifying thermoregulatory response thresholds during anesthesia, and showed that isoflurane--like halothane and nitrous oxide/fentanyl--markedly reduces the core temperature triggering autonomic cold defenses. Our current data, however, indicate that the dose-dependence for vasoconstriction is nonlinear, with isoflurane reducing the threshold disproportionately at higher anesthetic concentrations. Differing dose-dependence in the two studies may result from nonlinear enhancement of thermoregulatory responses by surgical stimulation. However, differences more likely result from limitations of the previous study design, including a sub-optimal randomization scheme and failure to correct for the effects of skin temperature on central thermoregulatory control.

Mallinckrodt Anesthesiology Products (St. Louis, MO) donated the thermocouples used in the study.