Thermoregulatory Vasoconstriction Impairs Active Core Cooling

With Institutional Review Board approval and informed consent, we studied 26 ASA Physical Status 1-3 patients undergoing elective neurosurgery. They were sequentially assigned to one of two anesthetic regimens: (1) isoflurane and nitrous oxide (n= 13) or (2) propofol and fentanyl (n = 13).

On arrival to the operating suite, 10 ml/kg of unwarmed intravenous fluid was administered. Most patients were given 50 mg prednisolone the morning of surgery. Without premedication, anesthesia was induced in the first group of patients with fentanyl (up to 250 micro gram) and sodium thiopental (4-6 mg/kg). The other patients were given a comparable amount of fentanyl and propofol (2-3 mg/kg). Intubation of the trachea was facilitated by administration of vecuronium bromide (0.1 mg/kg). Mechanical ventilation was maintained with a circle system having a fresh gas flow of 6 1/min, adjusted to maintain end-tidal P^CO2near 35 mmHg. No airway heating or humidification was provided.

In patients assigned to inhalational anesthesia, anesthesia was maintained with 70% N²O and isoflurane at an end-tidal concentration of 0.5-0.8%. No additional thiopental or opioid was administered. In patients assigned to total intravenous anesthesia, anesthesia was maintained with propofol (6-8 mg *symbol* kg sup -1 *symbol* h sup -1) and fentanyl (4 micro gram *symbol* kg sup -1 *symbol* kg sup -1 *symbol* h sup -1) without nitrous oxide.

Supplemental vecuronium was administered as needed to maintain one or two twitches in response to supramaximal stimulation of the ulnar nerve at the wrist. At least 10 ml *symbol* kg sup -1 *symbol* h sup -1 of fluid was given intravenously, and blood products were replaced to maintain the hematocrit between 25-32%. Fluids were not warmed. Per surgical routine, most patients were given 1 g/kg mannitol shortly after induction of anesthesia.

The patients were covered with a single layer of surgical draping during induction of anesthesia. Starting immediately after induction of anesthesia, active cutaneous cooling was initiated using a prototype forced-air system (Augustine Medical, Eden Prarie, MN). [3]This device provides 1,000 1/min air at 14-15 degrees Celsius into a full-body disposable convective cover. The cover was positioned directly above the patient's anterior skin surface and covered with a single cotton blanket. The arm used to test vasoconstriction was excluded from the forced-air cover. Active cooling continued until core temperature approached 32 degrees Celsius. When surgery was complete, patients were actively rewarmed, again using forced air (Bair Hugger, Augustine Medical). [11-12].

Core temperature, before induction of anesthesia, was measured at the tympanic membrane. The aural probe was inserted until patients felt the thermocouple touch the tympanic membrane; appropriate placement was confirmed when they easily detected a gentle rubbing of the attached wire. The probe was then taped in place, the aural canal occluded with cotton, and the external car covered with a gauze pad. Tympanic membrane temperatures correlate well with distal esophageal temperatures during anesthesia. [6,13]After induction of anesthesia, core temperature was recorded from the distal esophagus. Mean skin temperature was calculated from four sites: 0.3 (T^chest+ T^arm) + 0.2 (T^thigh+ T^calf). [14].

Fingertip blood flow was evaluated using forearm minus fingertip skin-surface temperature gradients; there is an excellent correlation between skin-temperature gradients and volume plethysmography. [15]The gradients were recorded from an arm not having an intravenous cannula or blood pressure cuff. A skin temperature gradient of 0 degree Celsius coincides with the core temperature plateau [10]; consequently, we considered this gradient to indicate significant vasoconstriction. (A gradient of 4 degrees Celsius, which we have used previously, indicates intense vasoconstriction. However, in this study, we were more interested in the core temperature plateau that starts when the gradient reaches 0 degree Celsius.) The distal esophageal temperature triggering significant vasoconstriction identified the thermoregulatory threshold. To facilitate comparison with previous studies, [4-6]we also recorded the core temperature triggering a skin-temperature gradient of 4 degrees Celsius (intense constriction). All temperatures were measured using Yellow Springs Instruments thermistors (Yellow Springs, OH).

Heart rate was monitored continuously using three-lead electrocardiography. Blood pressure was determined oscillometrically at 5-min intervals. Respiratory gas concentrations were quantified using a calibrated end-tidal gas analyzer (Drager, Lucbeck, Germany). All other data were recorded at 15-min intervals, starting immediately before induction of anesthesia (control values).

Morphometric data and core temperatures were compared using two-tailed, unpaired t tests. Isoflurane concentration, propofol dose, vasoconstriction threshold, time of constriction, time required for patients to reach target core temperatures, mean skin temperature, heart rate, and arterial blood pressures also were compared using two-tailed, unpaired t tests. Tumor locations were compared using Fisher's exact test. All values are expressed as mean plus/minus SD; differences were considered significant when P < 0.01.

No complications specifically related to deliberate hypothermia (e.g., cardiac arrhythmias, hemodynamic instabilities) were detected. Age, gender, weight, and height did not differ significantly in the two anesthesia groups (Table 1). We administered isoflurane at an end-tidal concentration of 0.6 plus/minus 0.2%; propofol was administered at a rate of 7.5 plus/minus 1.2 mg *symbol* kg sup -1 *symbol* h sup -1. Specific neurosurgical procedures (e.g., location) and indications (tumor vs. aneurysm) were comparable in the patients given each type of anesthesia.

Vasoconstriction was present in most patients in the immediate preoperative period, but vasodilation was observed in every case shortly after induction of anesthesia. In 6 of 13 patients given isoflurane, vasoconstriction (skin-temperature gradient = 0 degree Celsius) was observed at a core temperature of 34.4 plus/minus 0.9 degree Celsius and a mean skin temperature of 28.9 plus/minus 1.2 degrees Celsius, 1.7 plus/minus 0.5 h after induction of anesthesia. Similarly, in 7 of the 13 patients given propofol, vasoconstriction occurred at a core temperature of 34.5 plus/minus 0.9 degree Celsius and a mean skin temperature of 28.7 plus/minus 1.4 degrees Celsius, 1.6 plus/minus 0.6 h after induction of anesthesia. Patients reaching a gradient of 0 degree Celsius proceeded to a gradient of 4 degrees Celsius in 44 plus/minus 12 min, when the core temperature decreased 0.4 plus/minus 0.2 degree Celsius. Core cooling rates were comparable in each anesthetic group (Figure 1).

Patients remaining vasodilated cooled faster than those in whom vasoconstriction was observed. Consequently, patients given both types of anesthesia were subdivided on the basis of vasomotor responses. Specifically, we compared patients who at some time reached a gradient greater or equal to 0 degree Celsius (which we considered evidence of vasoconstriction) with those in whom vasoconstriction was never detected. Similar amounts of fluid and blood were given to the patients in each group, and ambient temperatures did not differ significantly. Morphometric characteristics also were similar (Table 2). Preoperative and immediate postoperative blood pressures and heart rates did not differ significantly in the two groups.

Core cooling rates in the patients in whom vasoconstriction was observed and those in whom it was not (Figure 2), and times required to reach core temperatures of 33 degrees Celsius and 32 degrees Celsius differed significantly in the two groups (Table 3). Intraoperative vasoconstriction was not correlated with the indication for surgery (tumor vs. aneurysm) or the location within the brain (Table 4).

Core body temperature normally is precisely regulated by effective thermoregulatory responses, which are initiated by small thermal perturbations. Heat stress provokes sweating [16]and active precapillary vasodilation [17]; cold stress initiates, in turn, arteriovenous shunt vasoconstriction, [18]nonshivering thermogenesis (in infants), [19-20]and shivering. [21]The interthreshold range is defined by core temperatures between the sweating and vasoconstriction thresholds not triggering autonomic thermoregulatory responses. [22]This range usually is only [nearly equal] 0.2 degree Celsius, [23]but typical doses of all general anesthetics so far tested increase the range 10-20-fold. Volatile anesthetics increase the sweating threshold. [24,25]but the interthreshold range is augmented mostly by a reduction in the vasoconstriction threshold. [4,8]In contrast, propofol increases the interthreshold range by reducing the vasoconstriction threshold without much increases the sweating threshold. [8].

Based on previous studies [5,6,8,26]and the age of the patients, [27]we expected the relatively low anesthetic doses given our patients to reduce the vasoconstriction threshold 2-3 degrees Celsius. Consistent with this experience, patients who vasoconstricted did so at core temperatures near 34.5 degrees Celsius. But surprisingly, vasoconstriction was not observed in half the patients in each anesthetic group, even at core temperatures of 32 degrees Celsius. Although thermoregulatory responses may be especially impaired in patients undergoing neurosurgery, the absence of concurrent non-neurosurgical control patients does not permit us to make any firm conclusion in this regard.

Nonshivering thermogenesis probably contributes little in unanesthetized adults [28,29]and does not occur during anesthesia. [30]However, thermoregulatory vasoconstriction reduces cutaneous heat loss [9]and constrains metabolic heat to the core thermal compartment. [10]Consequently, vasoconstriction is effective in minimizing further core hypothermia. In one study, for example, core temperature--which was decreasing at a rate of 1 degree Celsius/h before constriction--decreased only an additional 0.4 degree Celsius in the subsequent 3 h. [10]Similarly, core temperature decreased far more in patients in whom leg vasoconstriction was prevented by combined epidural/general anesthesia than in comparable patients given only general anesthesia. [31].

Based on the typical pattern of core hypothermia in adult surgical patients, we have termed the intraoperative period after reemergence of thermoregulatory vasoconstriction as the "plateau phase." [22]Typically, little further core hypothermia is observed once vasoconstriction is triggered in adults. [10,31]However, observed core temperature changes following vasoconstriction depend on a number of factors, including environmental temperature [2,32]and the patient's age and morphometric characteristics. [33]Core temperature, for example, may increase in hypothermic infants once vasoconstriction (and perhaps nonshivering thermogenesis) is triggered.* Similarly, infants in a warm environment become hyperthermic when orthopedic tourniquets constrain metabolic heat to the core thermal compartment. [34].

As might be expected, cure temperatures were comparable in all patients before vasoconstriction. In contrast to the typical adult pattern, patients in whom vasoconstriction was observed continued to become hypothermic during neurosurgery. Cutaneous blood traverses arteriovenous shunts perfusing distal tissues [18]and capillaries that supply the remaining skin. [35]During the initial phase of anesthesia, shunt flow is near maximal, whereas capillary flow remains low. [36]During cold stress, vasoconstriction decreases shunt flow more than tenfold with a smaller decrease in capillary flow. [37](In contrast, heat stress increases capillary flow to as much as 7.5 l/min. [38,39]) Core cooling in our patients continued despite vasoconstriction, in part because temperature in the microenvironment under the forced-air cooling cover was quite cold. Equally important, much of the cooling was directed toward the trunk, which has limited ability to vasoconstrict.

Core temperatures in the constricted and dilated patients were similar before vasoconstriction. Consequently, the time required to reach 34 degrees Celsius did not differ significantly in the two groups. Although intraoperative thermoregulation did not produce a core temperature plateau, it was nonetheless effective, and hypothermia developed at a considerably slower rate in vasoconstricted patients than in those remaining dilated. The patients in whom vasoconstriction was observed therefore required nearly an hour longer to reach target temperatures of 33 degrees Celsius and 32 degrees Celsius. This difference certainly could be clinically important. However, its significance in individual patients will depend on the target core temperature and the time required to reach critical portions of the operation. As in our previous investigation, [10]a skin-temperature gradient of 0 degree Celsius (beginning of vasoconstriction) better identified effective thermoregulation than a gradient of 4 degrees Celsius (intense constriction).

Thermoregulation may be the most common trigger for intraoperative peripheral vasoconstriction, but it is not the only one. Inadequate anesthesia, alpha-adrenergic drugs, and vascular volume depletion also cause constriction. Furthermore, thermoregulation and volume depletion synergistically increase vasoconstriction. [40]Patients undergoing neurosurgery--in whom fluid administration is often restricted--are thus at particular risk for hypovolemia-induced vasoconstriction. Intraoperative vasoconstriction, unresponsive to additional anesthetic, may be relieved by administration of fluid.

Because vasoconstriction significantly increases the time required to reach therapeutic target core temperatures, clinicians may find it helpful to monitor cutaneous vasomotion. The decrease in arteriovenous shunt flow is far more dramatic than the reduction in capillary flow. Consequently, fingers and toes are optimal monitoring sites. Thermoregulatory vasoconstriction can be monitored using a variety of methods, including volume plethysmography, [41]laser Doppler flowmetry, [42,43]and skin-temperature gradients. [4-6]Forearm minus fingertip skin-temperature gradients have the advantage of being inexpensive and easy to implement. There is an excellent correlation between skin-temperature gradients and volume plethysmography (which generally is considered the "gold standard"). [15]Calf minus toe gradients also are reliable. [6]Gradients will be accurate, however, only if measured on an extremity adequately shielded from active cooling.

In summary, patients in whom vasodilation was maintained throughout surgery required nearly an hour less to reach core temperatures of 33 degrees Celsius and 32 degrees Celsius than did those in whom vasoconstriction was observed. This difference could be clinically important, depending on the target temperature and the time required to reach critical portions of the operation. Skin-temperature gradients are an inexpensive and easy method of evaluating vasomotion. Administration of sufficient anesthesia usually will prevent thermoregulatory vasoconstriction.

The authors thank Gerhard Pavecic and Gepa-Med, Austria, for their assistance.

*Bissonnette B: Unpublished data. 1991.