Abstract

Introduction

Low-dose ketamine infusion (blood concentration around 100 ng/mL) during surgery reduces the incidence of postoperative shivering after remifentanil-based anesthesia. We hypothesized that perioperative infusion of very low-dose ketamine (blood concentration around 40 ng/mL) during remifentanil-based anesthesia may also prevent the development of remifentanil-induced shivering during the 2-hour period after the end of anesthesia.

Materials and methods

Fifty female patients scheduled to undergo laparoscopic cystectomy or oophorectomy were assigned to one of two groups: (1) ketamine group, in which the patients received ketamine infusion (0.1 mg/kg/hour) from induction of anesthesia to emergence from anesthesia; and (2) control group, in which the patients received saline infusion from induction up till emergence from anesthesia. Anesthesia was induced and maintained by target-controlled infusion of propofol (estimated blood concentration: 2–4 μg/mL) and infusion of remifentanil, at 0.2–0.3 μg/kg/minute. Patients were observed for shivering from the end of anesthesia to 120 minutes after anesthesia. The time point at which the patient began to shiver was recorded and assigned to one of four time periods: at emergence, from emergence to 30 minutes after anesthesia, from 30 minutes to 60 minutes after anesthesia, and >60 minutes after anesthesia.

Results

During the 120-minute observation period, the number of patients who shivered was higher in the ketamine group than the in control group (18 vs. 8, ketamine group vs. control group, p = 0.01). The time period during which patients began to shiver was different between the two groups (1 patient, 4 patients, and 13 patients vs. 3 patients, 2 patients, and 3 patients at emergence, from emergence to 30 minutes, and from 30 minutes to 60 minutes after anesthesia, respectively; ketamine group vs. control group, p = 0.007).

Conclusion

Intraoperative infusion of very low-dose ketamine during remifentanil-based anesthesia may increase the incidence of postoperative shivering.

Keywords

analgesics; ketamine; opioid; remifentanil; shivering;

1. Introduction

Remifentanil hydrochloride is a new mu-opioid analgesic drug with an extremely rapid clearance. In the postoperative period, patients frequently experience shivering after emergence from remifentanil-based anesthesia. The major cause of postoperative shivering is core hypothermia due to anesthesia-induced heat distribution, evaporation, and radiation during surgery; however, even if the patient's core temperature is carefully maintained, remifentanil-induced shivering still develops.¹ Röhm et al²demonstrated that total intravenous anesthesia with propofol and remifentanil induced a nearly two-fold incidence of postanesthesia shivering than with desflurane-fentanyl anesthesia. The mechanism of remifentanil-induced shivering has not been studied.

Nakasuji et al³ demonstrated that low-dose ketamine infusion (blood concentration around 100 ng/mL) could prevent the development of shivering during the early recovery phase from anesthesia. They concluded that the preventive effect of ketamine on the development of remifentanil-induced shivering was through N-methyl-D-aspartate (NMDA) receptor antagonism during anesthesia. However, during the short observation period in that study, there should be a modulation of thermoregulation by the residual effect of ketamine. To elucidate the mechanism of the preventive effect of ketamine on remifentanil-induced shivering, modulation of thermoregulation by residual ketamine should be excluded. Thus, the blood concentration of ketamine should be nearly zero during the time of evaluation of shivering.

Very low-dose ketamine infusion (blood concentration around 40 ng/mL) enhances or has a synergistic effect on opioid-induced analgesia by inhibiting NMDA receptor activation.⁴ We expected that very low-dose ketamine infusion should also prevent the development of postoperative remifentanil-induced shivering.

The purpose of this study was to determine whether very low-dose ketamine infusion could prevent the development of remifentanil-induced shivering.

2. Materials and methods

Fifty patients scheduled to undergo laparoscopic cystectomy or oophorectomy for benign (non-malignant) ovarian tumor were recruited for this randomized, double-blind, placebo-controlled study. Approval by the institutional review board and written informed consent from each patient were obtained. Exclusion criteria included morbid obesity (body mass index >30) and the presence of febrile symptoms at the time of surgery. Patients were randomly assigned to one of two groups using a computer-generated randomization schedule: (1) ketamine group (n = 25), in which patients received ketamine infusion (0.1 mg/kg/hour) from induction up until emergence from anesthesia; and (2) control group (n = 25), in which patients received saline infusion from induction to emergence from anesthesia. The dose of ketamine was determined in simulation by a target-controlled infusion program [STAMPUMP, Steven Shafer M.D., Anesthesiology Service (112A), PAVAMC, Palo Alto, CA, USA]. According to the protocol stipulated, the assignment was contained in a sealed envelope, and dilution of ketamine or saline was prepared by nurses who were not involved in the anesthesia procedure or evaluation of the patient. The anesthesiologist who managed the anesthesia was blind to the assignment of the patient.

Premedication was omitted. In the operating room, the patient's left antecubital vein was secured for perioperative infusion. Intraoperative monitoring included electrocardiograph, pulse oximetry, and non-invasive blood pressure measurement. Thermistor skin temperature probes (Monatherm, Tyco, Mansfield, MA, USA) were placed at the right index finger and the radial part of the forearm. Skin temperature was measured by the thermistor built in the anesthesia monitor (S5, GE Healthcare, London, UK). Forearm–fingertip temperature gradient > 0 was defined as peripheral vasoconstriction.⁵ Core temperature was measured by a thermocouple electrode (Monatherm, Tyco) inserted into the acoustic meatus. After placement of the patch electrode for brain monitoring (Entropy, GE Healthcare), an epidural catheter was placed through the Th12/L1 interspace for anesthesia and postoperative pain relief.

Anesthesia was induced by target-controlled infusion of propofol set at a plasma concentration of 3 g/mL and constant infusion of remifentanil at 0.25 μg/kg/minute. Ketamine or placebo infusion (saline) was started at the time of anesthesia induction. Rocuronium bromide 0.9 mg/kg was given to facilitate tracheal intubation. Anesthesia was maintained by target-controlled infusion of propofol at 2.0 μg–4.0 μg/mL, and remifentanil, 0.2–0.3 at μg/kg/minute, and 0.75% epidural ropivacaine 8–10 mL. The dose of remifentanil and that of propofol were varied to maintain the value of state entropy (SE), a parameter of the level of sedation,⁶ between 30 and 60 on brain monitoring (Entropy, GE Healthcare). For the patients with hypotension (systolic blood pressure < 80 mmHg), ephedrine 6 mg, was incrementally administered. At the beginning of skin closure, propofol, remifentanil, and ketamine or saline infusion were discontinued, and 8 mL 0.375% ropivacaine followed by continuous infusion of 0.2% ropivacaine at 4 mL/hour, was administered. Patients were extubated in the operating room.

2.1. Measurements

During the surgery, core temperature, forearm temperature, and fingertip temperature were recorded after placement of the epidural catheter (baseline), after intubation (induction), 1 hour after induction of anesthesia (1 hour), and at emergence from anesthesia (end of anesthesia). Ambient temperature was measured by a thermocouple, and was kept at around 25°C. The patient's upper body was warmed by a forced air warming device (Bair Hugger, Arizant Healthcare, Eden Prairie, MN, USA) to maintain the core temperature above 35.5°C. The values of relative entropy (RE: GE Healthcare) and SE were recorded prior to induction of anesthesia (baseline), prior to and after tracheal intubation, and 1 hour after tracheal intubation. After emergence from anesthesia, the presence of shivering, defined as fasciculation in more than one muscle or strong fasciculation activity of the entire body, was assessed and recorded by one investigator (O.M.) who was blind to the patients' assignment at the postanesthesia care unit (PACU) until 1 hour after extubation and at the gynecological ward until 2 hours after extubation. In the PACU and gynecological ward, patients were covered by a warm blanket to maintain body temperature. Patients who shivered for a duration of >30 minutes were given pethidine 50 mg. The time at which the patient began to shiver was recorded and classified into one of four time periods: at emergence from anesthesia, from emergence to 30 minutes after anesthesia, from 30 minutes to 60 minutes after anesthesia, and >60 minutes after anesthesia. In the PACU and gynecological ward, the core temperature was measured 30 minutes, 60 minutes, and 120 minutes after emergence from anesthesia with a near-infrared thermometer (Mimippi, Terumo, Tokyo, Japan) put in the acoustic meatus. Measurement of skin temperature was continued until 30 minutes after emergence from anesthesia in the PACU.

2.2. Statistical analysis

We performed a pilot study in 20 patients (n = 10 in each group). Seven patients (70%) in the ketamine group and three patients (30%) in the control group shivered during the observation period. Using power analyses (alpha = 0.05, beta = 0.8), 24 patients in each group was sufficient to present a statistically significant difference. The endpoint in this study was the number of patients who shivered during the observation period. The numbers of patients who shivered in the ketamine and control groups were compared using Fisher's exact test. The difference in the values of RE as well as SE over time between the ketamine and control groups was compared using repeated measures analysis of variance (ANOVA). The difference in core temperature between the ketamine and control groups, as well as between patients who shivered (shivering patients) and patients who did not shiver (non-shivering patients) was compared using repeated measures ANOVA. The dose of remifentanil was compared between the shivering patients and non-shivering patients using the unpaired t test. Differences were considered to be significant at p < 0.05. We examined the relationship between forearm-fingertip temperature gradient and core temperature at emergence from anesthesia in the shivering patients and non-shivering patients of both the ketamine group and the control group, respectively, and calculated the vasoconstriction threshold using linear regression analysis.

3. Results

All 50 patients could complete the study without complications. There were no significant differences in patients' characteristics between the ketamine and control groups (Table 1). The number of patients who shivered during the observation period was significantly higher in the ketamine group than that in the control group (Table 2). The number of patients who received pethidine was significantly higher in the ketamine group (Table 2). The distribution of the number of patients who began to shiver during various time periods was significantly different between two groups (Table 2). There was no significant difference in core temperature between the ketamine and control groups (Fig. 1A). A significantly higher core temperature was observed in shivering patients only at 2 hours after emergence from anesthesia (Fig. 1B). The values of RE prior to intubation, after intubation, and 1 hour after intubation, as well as the value of SE after intubation were significantly lower in the ketamine group than in the control group (Fig. 2). There was no significant difference in the dose of remifentanil between the patients who shivered and the patients who did not shiver (2.0 ± 0.72 vs. 1.9 ± 0.77, shivering patients vs. non-shivering patients, respectively, p = 0.39). Fig. 3 presents the relationship between the forearm-fingertip temperature gradient and that of core temperature in the control group (Fig. 3A) and in the ketamine group (Fig. 3B). The estimated vasoconstriction threshold was 37.0 ± 0.25°C in shivering patients and 36.7 ± 0.3°C in non-shivering patients of the control group, and 36.5 ± 0.2°C in shivering patients and 37.0 ± 0.15°C in non-shivering patients of the ketamine group (mean ± 95% confidence intervals).

Table 1. Characteristics of the patients.

Download full-size image

Table 2. Number of shivering patients over time after emergence from anesthesia.

Download full-size image

Download full-size image

Fig. 1. (A) Changes in core temperature over time. There were no significant differences between the ketamine and control groups; (B) changes in core temperature over time. The core temperature in shivering patients was significantly higher than that in non-shivering patients 2 hours after emergence from anesthesia.

Download full-size image

Fig. 2. Changes in relative entropy (RE) and state entropy (SE) over time. (A) Values of RE prior to intubation, after intubation and 1 hour after intubation in the ketamine group were significantly lower than those in the control group (p < 0.05); (B) the value of state entropy after intubation was significantly lower in the ketamine group than in the control group (p < 0.05).

Download full-size image

Fig. 3. (A) Correlation between forearm-fingertip temperature gradient and core temperature at emergence from anesthesia, in shivering patients (open circle) and non-shivering patients (filled circle) of the control group. Thick line: non-shivering patients, core temperature (°C) = 36.7 – 0.29. * Forearm-fingertip temperature gradient (°C); r2 = 0.001 (p = 0.21). Thin line: shivering patients, core temperature (°C) = 37.0 – 0.18. * Forearm-fingertip temperature gradient (°C); r2 = 0.45 (p = 0.07). (B) Correlation between forearm-fingertip temperature gradient and core temperature at emergence from anesthesia, in shivering patients (open circle) and non-shivering patients (filled circle) of the ketamine group. Thick line: non-shivering patients, core temperature (°C) = 37.0–0.235. * Forearm-fingertip temperature gradient (°C); r2 = 0.541(p = 0.059). Thin line: shivering patients, core temperature (°C) = 36.5 – 0.17. * Forearm-fingertip temperature gradient (°C); r2 = 0.303 (p = 0.017).

4. Discussion

In the present study, patients who received very low-dose ketamine infusion tended to present shivering. The result in this study, which showed a high number of patients in the ketamine group presented shivering during the observation period, was surprising. Nakasuji et al, demonstrated that low-dose ketamine infusion (blood concentration > 100 ng/mL) during surgery could prevent the development of shivering at the early phase after anesthesia (30 minutes after anesthesia).⁶ Even though the numbers of patients who shivered in the two groups were similar within 30 minutes after emergence from anesthesia in the present study, a significantly higher number of patients in the ketamine group presented shivering compared with that in the control group from 30 minutes to 60 minutes after anesthesia. The pharmacokinetics of ketamine suggest that the blood concentration of ketamine in the present study may have declined gradually during the first 30 minutes after the end of ketamine administration and reached an undetectable level 30 minutes after the end of ketamine administration.⁷ The inconsistent results between the present study and the study demonstrating the efficacy of ketamine in preventing remifentanil-induced shivering³ may be due to the difference in blood concentration of ketamine during the observation period. The blood concentration of ketamine in the present study might have been around 40 ng/mL at the end of administration, and declined to an undetectable level within 30 minutes after the end of anesthesia, but the observation was performed until 120 minutes after administration, whereas, in Nakasuji's study, the blood concentration of ketamine might have remained at around 50–100 ng/mL which could possess an analgesic effect, during the 30 minutes of observation.

The core temperature had been maintained in all patients in the present study. Although the comparable core temperatures between the ketamine group and the control group, and between shivering patients and non-shivering patients suggested the possibility that remifentanil-induced shivering is non-thermogenic, the vasoconstriction threshold seemed to be higher in shivering patients than in non-shivering patients of the control group, whereas it was lower in shivering patients than in non-shivering patients of the ketamine group. In the present study, the vasoconstriction threshold was estimated at the end of anesthesia, i.e., just after the termination of ketamine and remifentanil administration, and in the presence of a blood concentration of ketamine of around 30 ng/mL. The presence of a difference in vasoconstriction threshold at the end of anesthesia in shivering patients between the ketamine and control groups suggests that the residual effect of ketamine may modulate the vasoconstriction threshold which is altered by remifentanil infusion during anesthesia in shivering patients.

This study is the first trial aiming at preventing the development of remifentanil-induced shivering by very low-dose ketamine infusion. The result was surprisingly against expectation. It is clear that, in animal studies, NMDA receptor antagonists prevent the development or expression of withdrawal symptoms from opioid.⁸ However, in clinical studies, whether NMDA receptor antagonists could prevent the development of withdrawal symptoms from opioid is inconclusive.9, 10 Angst et al¹⁰ demonstrated in volunteers that ketamine co-administered with remifentanil prevented the development of opioid-induced hyperalgesia, which was evaluated in the presence of an analgesic dose of ketamine. In the present study, it is clear that the residual analgesic or sub-analgesic effect of ketamine was diminished at the time wherein many patients in the ketamine group developed shivering (30–120 minutes after the end of administration). The dose of ketamine during anesthesia might have been too low to prevent the development of hyperalgesia, or to treat hyperalgesia in patients with neuropathic pain.¹¹However, in the present study, the values of RE and SE tended to be lower in the ketamine group. This might be due to the values of RE and SE that had been kept in a wide range (from 30 to 60), suggesting that the dose of ketamine in the present study might have enhanced remifentanil-induced analgesia.⁶

The reason why a larger number of patients in the ketamine group shivered than in the control group, especially from 30 minutes to 60 minutes after emergence from anesthesia in the present study, cannot be explained only by the dosage of ketamine. In a study on the development of opioid withdrawal, electrophysiologic recording of NMDA current in rat dorsal horn neurons suggested that NMDA current increases in the presence of a moderate dose of remifentanil and during the wash out period.¹² However, in the presence of a high dose of remifentanil, NMDA current is suppressed and a high voltage of NMDA currents is evoked during the washout period. This suppression of NMDA current followed by a high voltage of NMDA current in the washout period is suspected to be the result of a rebound of NMDA current.¹² In the present study, a very low dose of ketamine might have suppressed NMDA receptor activity during administration and might have triggered rebound and enhancement of NMDA receptor activity. In a clinical study, a high incidence of remifentanil-induced shivering in the patients who received a high dose of remifentanil was reported.¹³ Although the reason for the high incidence of shivering in the ketamine group is not clear, we believe that it may be related to suppression of NMDA activity followed by rebound of NMDA activity. We have administered epidural ropivacaine intraoperatively and postoperatively. There was no difference in the dose of ropivacaine between the ketamine and control groups. Although epidural anesthesia lowers the shivering threshold almost 0.6 °C by an apparent temperature increase,¹⁴ active warming mimics the reduction in shivering threshold.¹⁵ In the present study, the effect of epidural anesthesia on the incidence of shivering may have been minimal.

Very low-dose ketamine infusion (blood concentration of 40 ng/mL) should not be administered for prevention of remifentanil-induced shivering.

In conclusion, intraoperative infusion of very low-dose ketamine during remifentanil-based anesthesia may increase the incidence of postoperative shivering.

Acknowledgments

We appreciate Dr. Daniel I. Sessler, Outcome Research Institute, Cleveland Clinic, OH, USA, for reviewing this manuscript.