Effect of Fentanyl on the Minimum Alveolar Concentration of Isoflurane in Swine

This study was approved by the Cornell University Animal Care and Use Committee. Unpremedicated, 4-5-month-old, female Yorkshire-cross swine (n = 16), weighing 27 plus/minus 2.2 kg (mean plus/minus SD), were anesthetized via mask with 4-5% isoflurane (Aerrane, Anaquest, Madison, WI) in 100% Oxygen²delivered at 5 l/min. Each pig was orotracheally intubated, and the concentration of isoflurane was reduced to 2-3% in oxygen delivered at 2 l/min. Animals were placed in lateral recumbency. Atropine (0.04 mg/kg, intramuscular) was administered to decrease secretions and protect against fentanyl-induced bradycardia. The lungs were mechanically ventilated to maintain normocapnia (Pa^CO235-45 mmHg). Two catheters were inserted percutaneously into auricular veins, one for administering fluids and the other for infusing fentanyl or saline. A catheter was inserted percutaneously into a femoral artery for continuous monitoring of systemic arterial blood pressures and for obtaining arterial blood samples. A lead II electrocardiogram was used to monitor heart rate and rhythm, and an agent-specific, photo-acoustic gas monitor (Type 1304, Bruel and Kjaer, Naerum, Denmark) measured inspired and expired gases, including end-tidal isoflurane and carbon dioxide. Lactated Ringer's solution was infused intravenously throughout the study to maintain a systemic mean arterial blood pressure greater than 60 mmHg. A rectal temperature probe monitored body temperature, which was maintained at 38-39 degrees Celsius by using a warm water-circulating heating pad to warm the swine or a fan and isopropyl alcohol to cool the swine.

Using a random number table, 16 swine were divided equally into four groups. Three of these groups were randomly assigned to one of three different protocols. Each protocol consisted of two fentanyl infusions as follows: 50 micro gram *symbol* kg sup -1 *symbol* h sup -1 followed by 100 micro gram *symbol* kg sup -1 *symbol* h sup -1, 50 micro gram *symbol* kg sup -1 *symbol* h sup -1 followed by 200 micro gram *symbol* kg sup -1 *symbol* h sup -1, or 100 micro gram *symbol* kg sup -1 *symbol* h sup -1 followed by 200 micro gram *symbol* kg sup -1 *symbol* h sup -1. Thus, eight swine received each dosage. Before each infusion, a loading dose of fentanyl was administered, as described below. The fourth group, the control group, received infusions of normal saline to evaluate the effect of time on MAC^ISO.

For each group, baseline MAC^ISOwas determined before the pigs received fentanyl. The first infusion was administered for 60 min, after which MAC^ISOwas redetermined. After MAC^ISOwas redetermined, the second 60-min infusion was administered, and MAC^ISOwas determined again.

Fentanyl was administered as fentanyl citrate (Sigma, St. Louis, MO) dissolved in 0.9% normal saline. A loading dose for each maintenance infusion rate was determined using standard pharmacokinetic equations, where loading dose = V^D*symbol* C^Pand maintenance infusion rate = CL *symbol* C^P; V^D= volume of distribution, C sub P = plasma fentanyl concentration, and CL = clearance of fentanyl. V sub D and CL were set at 216 ml/kg and 2.42 ml *symbol* kg sup -1 *symbol* min sup -1, respectively, based on previously reported pharmacokinetic data of fentanyl in swine. [20]The initial loading dose was calculated for the appropriate maintenance infusion rate using these data, whereas the second loading dose and maintenance infusion rate were corrected for the theoretical plasma fentanyl concentration already present because of the first fentanyl infusion (see Table 1for specific loading doses for each group). To blind the investigators to the type (saline vs. fentanyl) and dosage of solution being infused, loading doses for the first and second infusions were standardized to 60-ml volumes, and both infusion volumes were standardized for administering the test solution at 100 ml/h. For each infusion, plasma fentanyl samples were obtained after 30 min and when MAC^ISOwas determined to ascertain whether blood fentanyl concentrations were at steady-state.

At the beginning and end of the study, blood was obtained for measuring hematocrit and total solids. During baseline MAC^ISO, each pig received only lactated Ringer's solution. During the first and second infusions, the test solution and lactated Ringer's were administered concurrently, the latter for maintaining blood pressure.

The method used for determining MAC^ISOhas been described previously. [23,24]Briefly, an end-tidal isoflurane concentration was maintained for a minimum of 15 min such that the difference between the inspired and expired concentration of isoflurane was no more than 0.1%. This was done to ensure steady-state equilibration of alveolar isoflurane with that in arterial blood (brain). At the end of each 15-min equilibration period, end-tidal gases, heart and respiratory rates, blood pressures, fluid requirements, and temperatures were recorded. Each pig was stimulated with a hemostat clamped to full ratchet lock on a rear dewclaw, and the limb was moved cranially and caudally for a minute using the hemostat. If there was no purposeful response to stimulation, the end-tidal isoflurane concentration was reduced 10-20%, 15 min was allowed for equilibration, and the stimulus was repeated. Additional decrements of the end-tidal isoflurane concentration were made until there was purposeful movement. The isoflurane concentration was increased 10%, and after the equilibration period, the stimulus was repeated. The isoflurane concentration halfway between that allowing and that preventing movement, to the nearest 0.05%, was the MAC^ISOfor that pig. Once MAC^ISOwas determined, end-tidal carbon dioxide was recorded, and an arterial blood sample was simultaneously collected for analysis of plasma fentanyl, pH, and blood gas tensions. All animals were allowed to recover from anesthesia.

Blood gas samples were stored on ice and analyzed within 30 minutes of sampling on an automated blood gas analyzer (ABL2, Radiometer, Copenhagen, Denmark). Results were corrected for body temperature. Hematocrits were determined by centrifugation of blood in a microhematocrit centrifuge, and total solids were determined with a refractometer. Total solids is clinically thought to be equivalent to total protein, but a refractometer does not distinguish between total protein and small amounts of other particles, which may alter the reflectance of a solution. Plasma samples and solution samples from the first and second infusions were stored at -30 degrees until assayed for fentanyl concentration using a solid-phase radioimmunoassay kit (Coat-A-Count fentanyl, Diagnostic Product, Los Angeles, CA). These assays measure fentanyl base, which is 64% of the weighed fentanyl citrate used in the infusions. All samples were compared with results obtained from a second commercial radioimmunoassay kit (Research Diagnostics, Flanders, NJ). In addition, samples from the stock fentanyl solution, from commercially available medical fentanyl citrate (fentanyl citrate injection, Abbott, Abbott Park, IL), from both assay kit standards, and from randomly selected experimental plasma samples were assayed for fentanyl by infusion atmospheric pressure ionization mass spectrometry. This was done to verify that the stock chemical compound had equal fentanyl activity as the commercially available compound, and that both kits were accurate in measuring fentanyl.

This study was designed as an incomplete block design. The mean percent reductions in MAC^ISOand the absolute changes in MAC^ISOfrom baseline associated with each fentanyl dosage (50, 100, and 200 micro gram *symbol* kg sup -1 *symbol* h sup -1) were compared using analysis of variance for incomplete block designs. [25]The analysis was carried out using the general linear models procedure in SAS. [26]Statistical significance was conservatively set at P less or equal to 0.01 to adjust for multiple comparisons. [27]Because there was some concern that the order of treatments might influence the results, mean percent reduction in MAC^ISOwas compared within treatment groups, where data existed, using Student's t test. [28]The association between the MAC^ISOand plasma fentanyl concentration in each pig was assessed with Pearson's correlation coefficient and simple linear regression. [28].

Measurements of potentially confounding factors (temperature, pH, Pa^O2, or Pa^CO2) were compared across the treatment groups (i.e., fentanyl dosages) using analysis of variance. Tukey's method for multiple comparisons was used to evaluate pairwise comparisons. [25]Comparisons of continuous variables between two groups were compared with the Student's t test (when the groups were independent) and with the paired t test (when the data were dependent).

Results are expressed as mean plus/minus SEM. Variables known to affect minimum alveolar concentration (temperature, pH, Pa^O2, or Pa^CO2) were unchanged and within normal limits throughout the study (Table 2). Each animal received 15 plus/minus 1 ml *symbol* kg sup -1 *symbol* h sup -1 of fluids during the 5.3 plus/minus 0.3-h study. Although not statistically significant, blood pressure tended to increase as fentanyl dosages increased among the groups (Table 2). Within treatment type, there was no significant difference in the percent reduction in MAC^ISOamong pigs receiving 100 micro gram *symbol* kg sup -1 *symbol* h sup -1 fentanyl as the first treatment and those receiving this dose preceded by 50 micro gram *symbol* kg sup -1 *symbol* h sup -1. Similarly, there was no difference between pigs receiving 200 micro gram *symbol* kg sup -1 *symbol* h sup -1 preceded by 50 micro gram *symbol* kg sup -1 *symbol* h sup -1 compared to those receiving this dose preceded by 100 micro gram *symbol* kg sup -1 *symbol* h sup -1. All groups had obtained a steady-state level of plasma fentanyl as evidenced by comparing middle- and end-of-infusion plasma fentanyl concentrations (Table 3).

Previous research has indicated that minimum alveolar concentration does not vary over time. [23]Data from the control group (no fentanyl) confirmed this, and these data were not included in further comparisons in the effectiveness of different fentanyl infusion dosages. There was, however, wide variation in the value of MAC^ISOamong these four animals (Figure 1), and one of these had an unexplained 28% decrease in MAC^ISOduring the second infusion. Despite this individual variation, MAC^ISOdid not change statistically throughout the study for these four animals.

Mean plasma fentanyl concentrations and all MAC^ISOdata are provided in Table 3. The baseline MAC^ISOwas not different between groups or compared to controls. There was a dose-related decrease in MAC^ISOas the dose of fentanyl increased (Figure 2). The MAC^ISOat each dosage of fentanyl was statistically different from the other two dosages. Among individual animals, the percent reduction in MAC^ISOfor swine receiving 50 micro gram *symbol* kg sup -1 *symbol* h sup -1 ranged from 10% to 37%; for those receiving 100 micro gram *symbol* kg sup -1 *symbol* h sup -1, the reduction in MAC^ISOranged from 11% to 48%; and for 200 micro gram *symbol* kg sup -1 *symbol* h sup -1, it ranged from 25% to 63%.

There was a statistically significant (P less or equal to 0.01) negative correlation when MAC^ISO(r = -0.59; Figure 3) was plotted against the plasma fentanyl level for each animal. One animal with a blood fentanyl concentration of 89 ng/ml had a relatively large MAC^ISOand might be considered to be an outlier. This animal received 100 micro gram *symbol* kg sup -1 *symbol* h sup -1 fentanyl followed by 200 micro gram *symbol* kg sup -1 *symbol* h sup -1 and had high MAC^ISOvalues for baseline and both infusions (2.68%, 2.20%, and 2.00%, respectively). When all three MAC^ISOdeterminations from this pig are removed, the MAC sub ISO versus plasma fentanyl becomes even more highly correlated (r = - 0.74).

Fentanyl is often used in anesthetic protocols for research animals, particularly swine. Prior studies using fentanyl in swine, however, did not focus on its anesthetic efficacy. [1,3-14]Information from these studies is not useful because anesthetic dosages were not provided [9,13]or neuromuscular blockers were used. [1-9,11-14]Neuromuscular blockade makes depth of anesthesia difficult to determine, especially when hemodynamic variables and other monitored signs are experimentally manipulated, and are therefore unreliable. The goal of this study was to measure the extent to which fentanyl reduced the MAC^ISOin swine, thus providing information on its potency in this species. The lowest fentanyl infusion used in this study was based on the largest fentanyl infusion reported for swine. [1].

In this study, the mean control value for MAC^ISO, 2.19%, was similar to previous reports of 2.04% [29]and 2.2% [24]for swine. There was a decrease in MAC^ISOwith increasing concentrations of plasma fentanyl. The lowest infusion rate, 50 micro gram *symbol* kg sup -1 *symbol* h sup -1, produced a 24% decrease in MAC^ISO, whereas the largest infusion rate, 200 micro gram *symbol* kg sup -1 *symbol* h sup -1, produced only a 45% reduction in MAC^ISO. In a similar study in humans, investigators reported that plasma fentanyl concentrations as little as 3 ng/ml resulted in a 63% decrease in MAC^ISO. [22]In this pig study, plasma fentanyl concentrations of 59 ng/ml, 20 times greater than in humans, resulted in only a 45% decrease in MAC^ISO. In dogs, plasma fentanyl concentrations of 30 ng/ml resulted in a 65% reduction in the minimum alveolar concentration of enflurane, [21]but in the current study, 26 ng/ml plasma fentanyl concentration resulted in only a 30% reduction in MAC^ISO. It appears that the minimum alveolar concentration-sparing effect of fentanyl and its potency in swine is not as great as it is in humans and dogs. Although it is known that the efficacy of opioids varies from species to species, the extremely large differences between swine and humans were not expected. These results, however, are supported by an earlier study in which morphine was determined to be a much less potent analgesic in swine than in dogs or primates. [15]Because both drugs are opioid agonists, it seems plausible that swine require a larger dose of fentanyl to achieve equipotency. A clear ceiling effect or plateau was not observed within the dose range tested (Figure 2). Earlier studies in other species, however, provide evidence of a ceiling effect after opioid administration. [16,18,21,22]The reasons for this difference could not be determined from our results. One explanation is that these swine were not given a large enough dose of fentanyl to show a plateau in reduction of MAC^ISO.

The reason for the smaller reduction of MAC^ISOcompared to other studies is unclear. Differences in acid-base status, temperature, carbon dioxide, electrolytes, and hemodynamics can affect minimum alveolar concentration, but these variables remained within normal range during our study and in the other comparative studies. Minimum alveolar concentration also may vary depending on differences in application of the supramaximal stimulus (e.g., duration, moving the leg or not), evaluation of responses to the stimuli, or the type of inhalant anesthetic being tested. However, it seems unlikely that minor differences in these factors would account for the large differences observed between species. Species differences in opioid receptor type, number or distribution may account for some of the differences observed. [19,30]Although minimal data are available, fentanyl and morphine appear to have different pharmacokinetics in swine than in dogs and, potentially, other species. [15,20,31].

Another potential cause for differences in the reduction of MAC sub ISO compared to other studies is that our swine may have developed acute tolerance to the fentanyl. Tolerance after a continuous infusion or multiple intermittent boluses has been reported as a potential reason for differences in studies evaluating the anesthetic and analgesic effects of fentanyl in dogs. [16,32]Evidence of acute tolerance was not apparent in this study, however, because the same decrease in MAC^ISOwas found for comparable plasma fentanyl concentrations, regardless whether that level was attained during the first or second infusion.

One further explanation for study differences may be the apparent hysteresis that has been reported to occur between plasma fentanyl concentration and its physiologic effects. [22]That is, there is a lag before fentanyl reaches its site of action. Therefore, if minimum alveolar concentration is determined after a single bolus or a short infusion and before equilibrium has occurred between plasma and receptors, a smaller reduction in minimum alveolar concentration may result at that plasma concentration. In this study, this possibility was eliminated by administering fentanyl as an infusion instead of as a bolus, and steady-state concentrations had been reached by 30 min into the 60-min infusion.

It is important to note that this study did not evaluate the deeper levels of anesthesia necessary to provide adequate surgical anesthesia. The contributing effects of premedicants or combinations of fentanyl with other anesthetics were not examined. Some studies have used between 20 and 50 micro gram *symbol* kg sup -1 *symbol* h sup -1 fentanyl as the primary anesthetic, in combination with only premedicants and paralysis. [1,3]Our data show, however, that a fentanyl infusion of 50 micro gram *symbol* kg sup -1 *symbol* h sup -1 is not adequate as a sole anesthetic in swine because it reduces MAC^ISOby only 25%. Furthermore, fentanyl at 200 micro gram *symbol* kg sup -1 *symbol* h sup -1 reduces MAC^ISOan average of 45%. Therefore, none of the dosages tested in this study should be expected to produce acceptable, complete anesthesia in swine. Adequate anesthesia may be obtained when fentanyl is combined with other anesthetics, but other drugs must be used according to pharmacologic data for swine and not based on human dosages.

In conclusion, results from the current study indicate that fentanyl infusions of 50-200 micro gram *symbol* kg sup -1 *symbol* h sup -1 reduce the minimum alveolar concentration of isoflurane in swine but not to the same degree as has been found in other species. Therefore, fentanyl dosages currently used in research involving swine should be evaluated carefully by the investigator to ensure minimal distress to the animal and to prevent collecting misleading data. Protocols should be designed with the knowledge that dosages up to 200 micro gram *symbol* kg sup -1 *symbol* h sup -1 contribute approximately a 50% MAC equivalent. Extreme care should be taken when applying anesthetic protocols developed for humans to other species, such as swine.

The authors acknowledge the patience and persistence of the Diagnostic Laboratory and Dr. Thomas Reimers, for the blood fentanyl results, and Dr. Tim Wachs, for mass spectrometric fentanyl concentration assays. The authors also thank Jeanne Brohart, for technical expertise; Denise Hine, for secretarial assistance; and Jerry Decker, VMTH chief pharmacist, for drug preparation throughout this project.