Materials and Methods

Animals—Six adult (3- to 7-year-old) neutered mixed-breed cats (4 females and 2 males) were assigned to the study, after approval by the Institutional Animal Care Committee (protocol No. 124/2005-CEEA). Mean ± SD weight of the cats was 3.9 ± 0.5 kg. Before commencement of the study, all cats were vaccinated, dewormed, and assessed as clinically normal on the basis of results of a physical examination, serum biochemical analyses, and a CBC. In addition, PCR testing for FeLV and FIV was performed to ensure that all cats included in the study were negative for those infections. Cats were group housed in a large room (in accordance with established standards³⁴), provided with fresh water ad libitum, and fed a commercial dry cat food with supplemental canned food provided periodically.

Anesthesia and instrumentation—Twelve hours prior to each of the 2 study days, each cat was weighed and moved to a single cage in a quiet environment; food was withheld, but water was provided. For induction of anesthesia, the cat was placed in an acrylic chamber into which 5% isoflurane^a in 100% oxygen (5 L/min) was delivered with a calibrated out-of-circuit vaporizer^b and a pediatric circle circuit. The time required for chamber induction (ie, loss of spontaneous movement) was recorded in each study day. Subsequently, the cat was removed from the chamber, and isoflurane was administered via face mask until orotracheal intubation with a cuffed endotracheal tube could be achieved. Thereafter, the cat was positioned in lateral recumbency and mechanically ventilated^c (pressure-controlled mode) with respiratory rate and inspiratory peak airway pressure adjusted to maintain PvCO₂ and ETCO₂ values at 30 to 40 mm Hg throughout the anesthetic episode. A thin plastic catheter was placed through the lumen of the endotracheal tube so that the tip was at the distal end of the tube within the thoracic trachea; the other end, which exited at the connection of the endotracheal tube with the breathing circuit, was connected to a sidestream infrared gas analyzer^d for continuous measurements of ETCO₂ and ET_ISO. The analyzer was calibrated with a reference sample provided by the manufacturer^e at the beginning and midpoint of each study day procedures.

After induction of anesthesia in each cat, the instrumentation phase was completed in approximately 30 minutes while ET_ISO was maintained between 2.0% and 2.3%. A 24-gauge catheter was inserted in a cephalic vein for administration of lactated Ringer's solution by use of a peristaltic pump^f or remifentanil hydrochloride^g CRI (0.25, 0.5, and 1.0 μg/kg/min) diluted with saline (0.9% NaCl) solution by use of a syringe pump.^h All infusions were delivered at rate of 3 mL/kg/h; the infusion of lactated Ringer's solution started after placement of the catheter and was substituted with remifentanil during the respective CRI for the duration of the infusion. A jugular vein was percutaneously catheterized with a 20-gauge catheter for collection of venous blood samples in heparinized syringes for measurement of venous blood pH, PvCO₂, and bicarbonate concentration by use of a blood gas analyzerⁱ with correction for body temperature (°C).^j A thermistor was advanced to the thoracic portion of the esophagus for monitoring of body temperature, which was maintained at 37.5° to 38.5°C by use of a forced warm-air blanket.^k A lead II ECG^j was used to continuously monitor heart rate and rhythm. To determine SAP, an ultrasonographic Doppler^l flow probe was placed over the caudal branch of a saphenous artery with an occluding cuff (width, 40% of the limb circumference) placed above the tarsal joint. A pulse oximeter probe^j was placed on the tongue for arterial SpO₂ monitoring.

MAC determination—Determinations of MAC were completed by use of a previously described technique.³⁵ A supramaximal noxious stimulus was applied by means of two 24-gauge stainless steel needles that were positioned subcutaneously 3 cm apart in the middle third of an antebrachium. The needles were connected to an electrical stimulator^m (50 V; 50 cycles/s; pulse duration, 10 milliseconds). The pattern of electrical stimulation consisted of a sequence of 4 stimuli at 5-second intervals: 2 single stimuli were followed by 2 stimuli that were each continuously applied for 3 seconds. The whole sequence of nociceptive stimulation was completed unless a positive motor response was observed during the process. A positive motor response was defined as a gross purposeful movement of head or limbs, with the exception of the forelimb that underwent electrical stimulation. A negative response included the lack of movement of head and limbs, muscle rigidity, shivering, tail movement, coughing, swallowing, or an increase in spontaneous respiratory efforts during controlled ventilation.

After instrumentation of each cat and a period of equilibration of at least 15 minutes at an ET_ISO of 1.8% to 2.2% (similar to reported values of isoflurane MAC in cats^36–38), ET_ISO was increased or decreased by 0.1% after a positive or negative response, respectively, was obtained. The ET_ISO was kept constant for at least 15 minutes at each set value before delivery of the electrical stimulus, and the process of increasing or decreasing concentrations was repeated until a positive response was followed by a negative response or vice versa. The MAC was calculated as the mean of 2 successive ET_ISO values associated with 1 positive and 1 negative response. Minimum alveolar concentration values were corrected to sea level by use of a formula as follows: (barometric pressure of location/760 mm Hg) X obtained MAC value. The mean barometric pressure obtained from the official city meteorologic station for the altitude at which the experiment was completed (ie, 785 m above sea level) corresponded to 680 mm Hg.

Experimental protocol—Each cat was anesthetized 2 times; there was an interval of at least 2 weeks between the anesthetic procedures. The study was designed so that each cat acted as its own control animal. On the first study day, MAC_Basal was determined. On the second study day, isoflurane MAC was recorded during administrations of 3 CRIs of remifentanil; 0.25, 0.5, and 1.0 μg of remifentanil/kg/min were each administered (in ascending order) without a loading dose. Each remifentanil CRI was maintained for 30 minutes before the first stimulus was applied. After each MAC determination, the remifentanil CRI was discontinued and lactated Ringer's solution (3 mL/kg/h) was infused for 15 minutes before the beginning of the next CRI. After the last MAC determination (ie, determination of MAC_R1.0), MAC_Control was determined after a 15-minute washout period to verify whether isoflurane MAC had returned to the values recorded during baseline conditions (ie, MAC_Basal). During the interval after MAC_R0.5 determination and before MAC_R1.0 determination, the needle electrodes used for nociceptive stimulation were repositioned 1 cm from their previous locations (maintaining a distance of 3 cm between them) to avoid local desensitization.³⁹

The duration of each MAC determination (MAC_Basal, MAC_R0.25, MAC_R0.5, MAC_R1.0, and MAC_Control) on both study days was recorded and considered the time from the beginning of the ET_ISO equilibration period to the last electric stimulus applied to obtain the MAC value. The duration of each remifentanil CRI was also recorded. On both study days, cardiopulmonary and blood gas analysis data were collected immediately before application of each noxious stimulus; values corresponding to each MAC level were calculated as the arithmetic mean of the value obtained at the negative and positive response.

After the final MAC determination on either study day (ie, MAC_Basal or MAC_Control), all catheters were removed and meloxicamⁿ (0.2 mg/kg, SC) was administered; isoflurane administration was discontinued, and the time to extubation, determined by the presence of palpebral reflex and resumption of normal ventilation, was recorded. On the second study day (the remifentanil treatment day), the time to walking with ataxia and time to complete recovery (ie, apparently normal ambulation), determined from the time of extubation, were also recorded.

Statistical analysis—An ANOVA for repeated measurements was used to compare MAC, heart rate, inspiratory peak airway pressure, PvCO₂, venous blood pH and bicarbonate concentration, and ETCO₂ values obtained in the 5 experiments (ie, the values recorded on the first study day [basal values] and on the second study day [values for the 3 remifentanil CRIs and control values]). To compare variables such as esophageal temperature, SAP, respiratory rate, SpO₂, and durations of MAC determination and remifentanil CRIs, a Friedman test was used. Differences were considered significant at a value of P < 0.05. A Shapiro-Wilk test was used for assessment of data normality, and a Levene test was used for the assessment of variables’ variance equality among the experiments. Thus, for variables for which normality was not rejected, data were analyzed via an ANOVA and findings are expressed as mean ± SD. Alternatively, a Friedman test was used, and results are expressed as median (lower quartile to upper quartile). Comparisons among basal, remifentanil CRIs, and control data were performed by use of the CI interval for the mean (95% CI) or median (96.9% CI).

Results

Determinations of MAC and assessments of cardiopulmonary and blood gas variables were completed successfully for all cats in the present study. Values of MAC_Basal and MAC_Control were not significantly different (Table 1). Mean value of isoflurane MAC during each remifentanil CRI was significantly lower than MAC_Basal; however, there was no significant difference among the MAC_R0.25, MAC_R0.5, and MAC_R1.0 values and no significant difference between each of those values and MAC_Control. Times for all MAC determinations were similar.

Unlike findings for the other assessed variables, the effect of remifentanil infusions on PvCO₂ and SAP was significant (Table 2). During the 1.0 μg/kg/min infusion of remifentanil, SAP was significantly increased, compared with the value during the control experiment. During the 0.5 μg/kg/min infusion of remifentanil, PvCO₂ was significantly decreased, compared with the value during the 1.0 μg/kg/min infusion of remifentanil.

Time for induction of anesthesia with isoflurane was 7.6 ± 1 minutes and 7 ± 2 minutes for the first and second study days, respectively, with additional 3 ± 1 minutes of isoflurane administration via face mask on both study days. On the second study day, the durations of remifentanil administrations were not significantly different among the 3 CRI rates (Table 3).

Three of the 6 cats during the first study day (isoflurane alone) and 4 of the 6 cats during the second study day (isoflurane and remifentanil) had a rough recovery from anesthesia, characterized by thrashing and paddling inside the cage for a few minutes at the time of extubation, after which they all became calm. Isoflurane administration was discontinued 14.3 ± 8 minutes and 14.5 ± 2 minutes, respectively, after the last MAC determination (MAC_Basal and MAC_Control) in each study day. Time to extubation was 1.7 ± 1.2 minutes and 1.5 ± 0.5 minutes for the first and second study days, respectively. On the second study day, time to walking with ataxia and time to complete recovery (ie, apparently normal ambulation) were 12.4 ± 1.5 minutes and 24.5 ± 5.2 minutes, respectively.

Discussion

Constant rate infusions of remifentanil cause a maximum isoflurane MAC decrease of 65% and 91% in rats and humans, respectively.^11,15 In dogs that received remifentanil via CRI, enflurane MAC was decreased by approximately 63% and the dose estimated to cause a 50% reduction was 0.72 μg/kg/min.⁸ However, in a clinical study,⁴⁰ infusions of remifentanil at rates of 0.1 and 0.25 μg/kg/min in dogs decreased isoflurane requirements in a dose-dependent manner by as much as 50% during orthopedic procedures. The MAC_Basal in the cats of the present study (1.66%) was similar to values determined in other studies^{1,9,10,20,38,41} for this species, although reported values range from 1.2% to 2.2%. The percentage MAC reduction (compared with MAC_Basal) attributable to remifentanil in cats in the present study (23% to 30% for the 3 remifentanil CRIs) was less than the percentage MAC reductions reported for other species.^8,11,15 However, the percentage MAC reduction among the 3 CRI doses was similar, despite administration of doses in ascending order on the same study day. Opioid-induced MAC reductions in horses, swine, and cats are of less magnitude than those in dogs, rats, monkeys, and humans.^{7,9–13,16–19,42}

Doses selected for investigation in the present study were based on remifentanil infusion rates of 0.2 to 0.23 μg/kg/min that, when combined with propofol (0.3 mg/kg/min), appear to provide adequate clinical anesthesia in cats undergoing ovariohysterectomy.⁴³ Three doses were investigated: 0.25, 0.5, and 1.0 μg of remifentanil/kg/min. The lowest dose was used in the aforementioned clinical study⁴³; the 0.5 and 1.0 μg/kg/min CRIs were included to determine whether a ceiling effect could be achieved because a study⁸ in dogs revealed a ceiling enflurane MAC reduction (63%) at a CRI of 1.0 μg of remifentanil/kg/min.

Ceiling effects for MAC reduction induced by the phenylpiperidine opioid derivatives fentanyl,^16,17 sufentanil,^6,14 alfentanil,^7,9 and remifentanil⁸ have been reported. In the present study, a 4-fold difference in the CRI dose of remifentanil (0.25 vs 1.0 μg/kg/min) did not affect the degree of MAC reduction, indicating that a ceiling isoflurane-sparing effect was already achieved with the lower CRI. On the basis of findings of the present study, the specific infusion rate at which this ceiling effect is achieved in cats cannot be determined.

The value of MAC_Control (determined after cessation of high-dose CRI and subsequent 15-minute washout period) was not significantly different from MAC_R0.25, MAC_R0.5, or MAC_R1.0. In our study, MAC_Control was determined to ensure that MAC had returned to baseline value after completion of all 3 remifentanil CRIs; if MAC_Control differed significantly from MAC_Basal, then likely a cumulative effect of repeated infusions of remifentanil had developed or a modified response to the repeated electrical stimulation during MAC determinations had occurred. With regard to a cumulative drug effect, the elimination half-life of remifentanil must be considered. The elimination half-life of remifentanil in isoflurane-anesthetized cats is 15.7 minutes,³² which is longer than the value in dogs (5.6 minutes)³³ and similar to the value in humans (10 to 20 minutes).^26,44 The relatively short elimination half-life of remifentanil in cats³² suggests that this drug would be rapidly eliminated after a prolonged infusion regimen, such as the one used in the present study. However, elimination half-life has little value for describing the disposition of drugs, such as remifentanil, that are administered via CRI because of the multicompartment model.^45,46 The context-sensitive half-time, defined as the time required for plasma concentration to decrease 50% after termination of an infusion designed to maintain a constant concentration,^45,47,48 incorporates the multicompartmental behavior of drugs following IV administration and better reflects the rapidity of drug elimination during the course of prolonged IV infusion regimens.^45,46 Remifentanil's context-sensitive half-time, determined to date only in humans, is extremely short (approx 3 minutes)^25–27,49 and is independent of the duration of the infusion, indicating that this drug has no cumulative effects in humans.^25–27 In a study⁸ in dogs, the enflurane MAC determined 30 minutes after discontinuation of a CRI of remifentanil was no different from the pretreatment baseline MAC value, indicating no cumulative drug effect for this species. It is possible that the longer halflife of remifentanil in cats results in a cumulative effect because the MAC_Control value in our study was not significantly different from MAC_R0.25, MAC_R0.5, or MAC_R1.0, even though assessment of MAC_Control was started after a 15-minute washout period and the median time to MAC determination was 55 minutes. Because a ceiling effect was achieved for MAC values associated with the 3 CRI doses, it is also possible that all doses were high and facilitated the cumulative effect, which did not allow for complete return to MAC_Basal value at the end of the second study day.

Albeit a nonsignificant difference, MAC_Control was 0.1% to 0.4% lower than MAC_Basal in 4 of the 6 cats in the present study. This could be explained by a modification in the motor response as a result of repeated nociceptive stimulation. Desensitization of the skin caused by repeated electrical stimulation has been reported as a possible source of bias during MAC determination.³⁹ In our investigation, needle positions were changed during the interval between MAC_R0.5 and MAC_R1.0 determinations in an attempt to avoid skin desensitization. If the skin is desensitized by repeated stimulation, purposeful movement could not occur for a given end-tidal anesthetic concentration, and this phenomenon could cause MAC values to be artificially low. Also, we cannot rule out the influence of a small sample size in our study as a cause of a type II error (ie, a false-negative result)⁵⁰ and the increased variability in individual isoflurane MAC values (especially with regard to MAC_Control values), despite the use of a standard method. However, MAC determinations have been commonly conducted by use of similar study designs.^8,9,11,20,38

As for other opioids,^18,42,51 the mechanism by which remifentanil induces MAC reduction most likely involves its antinociceptive effects, although sedation can also contribute to this reduction through CNS depression. In most species, opioids cause sedation, but CNS stimulation is common in cats, horses, and swine.^9,13,18,42 This CNS-stimulating effect can minimize the MAC-sparing effect of opioids^13,18 and may result in increased sympathetic nervous system activity.⁵²

Signs of CNS stimulation associated with opioids have included spontaneous movements in alfentaniltreated horses during anesthesia, with halothane concentrations equivalent to those at which horses did not respond to the noxious stimulation during MAC determinations.¹³ These types of movement were also observed in some cats in the present study, which supports the hypothesis that remifentanil caused some degree of CNS stimulation during isoflurane anesthesia that may have opposed the MAC-sparing effect of remifentanil in this species. Indeed, in species such as humans and dogs, pure OP3 opioid receptor agonists cause sedation in addition to analgesia, and both effects may contribute to the comparatively greater reduction in MAC achieved with remifentanil in those species—63% reduction in enflurane MAC in dogs⁸ and 91% reduction in isoflurane MAC in humans.¹⁵

In the cats of the present study, heart rate and SAP increased by 26% and 23%, respectively, during remifentanil CRI. Although opioid-related decreases in these variables are often reported for other species,^6,11,40,42 cats and horses frequently develop cardiovascular stimulation because of an increase in sympathetic outflow.^{13,18–20,22,52} In isoflurane-anesthetized cats, administration of the phenylpiperidine opioid alfentanil results in cardiovascular stimulation, which is paralleled by increases in plasma concentrations of cathecolamines.²² It has been suggested that the cardiovascular stimulation associated with the use of some opioids in cats is probably attributable to a central stimulatory effect that causes increases in sympathetic nervous system activity.^22,52 Finally, the use of lower ET_ISO during remifentanil CRIs also resulted in less cardiovascular depression and less effect on measured variables.

For the cats in our study, the time to recovery from isoflurane anesthesia was similar to that reported by other authors.² The observed characteristics of anesthetic recovery were probably attributable to residual isoflurane effects and were not associated with the use of remifentanil. The excitement observed in 4 of 6 cats during recovery from isoflurane-remifentanil anesthesia was also evident during the recovery phase in some of those cats (3/6) when they were anesthetized with isoflurane alone.

Results of the present study indicated that CRI of remifentanil reduced the isoflurane MAC in cats, with a ceiling effect at the tested doses, and that no major cardiovascular effects developed in association with these infusion protocols. On the basis of these findings and considering the doses evaluated, remifentanil CRIs at rates > 0.25 μg/kg/min did not achieve increasingly higher MAC reductions; however, whether CRIs at rates < 0.25 μg/kg/min could result in a comparable MAC reduction was not determined in our study. Further clinical and experimental studies involving a greater number of cats and lower doses of remifentanil administered via CRI are recommended to determine the optimal CRI dose of remifentanil for MAC reduction in isoflurane-anesthetized cats.