Potentiation by ketamine of fentanyl antinociception. I. An experimental study in rats showing that ketamine administered by non-spinal routes targets spinal cord antinociceptive systems

R. Nadeson, A. Tucker, E. Bajunaki and C. S. Goodchild*

Monash University Department of Anaesthesia, Level 5, Block E, Monash Medical Centre, 246 Clayton Road, Clayton, Victoria, Australia 3168*Corresponding author

Accepted for publication: December 21, 2001


    Abstract
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Background. Ketamine has been found to exert antinociceptive effects in animals and to be analgesic at subanaesthetic doses in humans. This study was designed to investigate the involvement of spinal cord mechanisms in the potentiation of opioid analgesia by parenteral non-spinal administration of ketamine.

Methods. Thresholds for nociception were measured in an acute pain model in rats that allowed identification of antinociceptive effects due to drug action in the spinal cord. Dose–response curves for the antinociceptive effects of ketamine alone and ketamine in conjunction with the µ opioid fentanyl were constructed.

Results. Intraperitoneal ketamine up to 3.75 mg kg–1 caused no sedative or antinociceptive effects and intrathecal ketamine caused dose-dependent, spinally mediated antinociceptive effects. Injections of ketamine doses that caused no antinociceptive effects when given alone (intrathecal 25 µg and intraperitoneal 3.75 mg/kg) significantly increased spinally mediated antinociception produced by intrathecal fentanyl injections when assessed using noxious heat (tail-flick test) but not when assessed by noxious electrical current (electrical current threshold test).

Conclusions. We conclude that ketamine can potentiate the effects of fentanyl by an interaction at the level of the spinal cord even when ketamine is given via a non-spinal route of administration.

Br J Anaesth 2002; 88: 685–91

Keywords: analgesics opioid, fentanyl; anaesthetics i.v., ketamine; receptors, amino acid; spinal cord, sensory block


    Introduction
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
The involvement of spinal cord N-methyl-D-aspartate (NMDA) receptors in the generation and maintenance of acute and chronic pain states in animal models is well established.17 The dissociative anaesthetic drug ketamine has been shown to be an NMDA antagonist and, as such, to potentiate the antinociceptive effects of opioids assessed neurophysiologically in the spinal cord.8 9 This property seems to be due to the S isomer of ketamine.10

These observations in animal studies have prompted many clinical studies that have shown that racemic ketamine (mixture of R and S isomers) is useful in the management of postoperative pain in humans1117 and also cancer pain.18 In most of these studies ketamine was given by a non-spinal parenteral route (i.v.) and thus the site of action, brain or spinal cord, is uncertain. It is often assumed that the site of action in the production of pain relief is NMDA receptors in the spinal cord. However, ketamine does interact with NMDA receptors in the brain and thus it is possible that the observed potentiation of opioid analgesia is due to interaction of the drugs in the brain. No formal evidence has ever been provided in a human or animal behavioural experiment for interaction at the level of the spinal cord between non-spinally administered ketamine and an opioid to produce additive antinociceptive effects or potentiation of antinociceptive effects. Even when ketamine was given epidurally17 or when it was shown to prolong the analgesic actions of intrathecal neostigmine,19 the interaction between the drugs could have occurred at a site other than the spinal cord. Spinal cord NMDA receptors may not have been activated to produce these effects even though such interactions have been shown neurophysiologically at this level in the central nervous system (CNS).8 9

This study set out to investigate the interaction between a µ opioid agonist (fentanyl) and ketamine at the level of the spinal cord. The resulting antinociceptive effect was assessed in an acute pain model in rats that allowed identification of antinociceptive effects due to drug action in the spinal cord.20 Dose–response relationships for the antinociceptive effects of intrathecally and intraperitoneally administered ketamine alone and in conjunction with the µ opioid fentanyl given intrathecally were investigated.


    Methods
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Experiments were performed on male Wistar rats (weight 180 ± 10 g (mean ± SEM)). In all experiments, we followed international guidelines for the care of laboratory and experimental animals and for the investigation of pain in laboratory animals.21 This work was carried out with permission of the Monash University Standing Committee on Ethics in Animal Experimentation (SCEAE Project No. 93017). Dose–response relationships were investigated for ketamine given intraperitoneally and intrathecally alone and in combination with fentanyl administered intrathecally. Chronically implanted subarachnoid catheters were used to administer intrathecal injections. In all experiments in which nociceptive thresholds were measured, we tested tail-flick latency (TFL) and response to noxious electric current [electric current threshold (ECT) test]. The dose–response relationship for other CNS disturbances, such as sedation and motor incoordination for intraperitoneal ketamine, was also investigated using a rotarod task in order to define an antinociceptive dose that caused no other effects.

Sedative and motor effects of intraperitoneal ketamine
Rats were assessed for competency to perform the rotarod test as described previously.22 Each rat had no previous exposure to the rotarod test. Rats were placed on the rotarod accelerator treadmill (Ugo Basile 7650 accelerator rotarod, Comerio (VA), Italy) set at the minimal speed for two training sessions of 1–2 min at intervals of 30–60 min. After this conditioning period, the animals were placed on to the rotarod at a constant speed of 25 r.p.m. As the animal took grip of the drum, the accelerator mode was selected on the treadmill, i.e. the rotation rate of the drum was increased linearly at 20 r.p.m. every minute thereafter. The time was measured from the start of the acceleration period until the rat fell off the drum; this was the control (pretreatment) performance time for each rat. The maximum running time was 30 s. This test was performed on each rat four times with an interval of 30 min between runs. The mean performance time was calculated as an average of the last three control performance times. The rats were then injected intraperitoneally with ketamine (dose 0–10 mg kg–1 in 0.5 ml saline; n=11–18 for each dose). Each rat was then placed on the rotarod with the test parameters used in the control (pretreatment) period. The time the animal remained on the rotarod was measured every 5 min for 30 min. The shortest time during this period that the rat managed to balance on the rotarod was expressed as a percentage of that animal’s own mean control performance. The results from all 11–18 rats at each dose were combined to calculate the mean and SEM.

Catheter implantation
Lumbar intrathecal catheters (dead space volume 5–8 µl) were implanted under halothane anaesthesia and aseptic surgical conditions as described previously.23 Portex catheters were introduced into the lumbar intrathecal space by a lumbar laminectomy to lie next to the lumbar and sacral segments of the spinal cord responsible for innervation of the tail and hind legs. The catheters were tunnelled subcutaneously to an exit wound at the base of the neck. The exact volume of the catheter was measured before insertion, thus allowing accurate dosing of drugs injected through the catheter into the cerebrospinal fluid. Correct catheter placement was confirmed by injecting 2% lidocaine solution 10 µl down the catheter 10 min after recovery from general anaesthesia. The catheter was judged to be intrathecal if paralysis and dragging of the hind legs occurred within 30 s of this injection. All animals failing this test were rejected from further study, as were those with neurological deficit after surgery, as shown by dragging of one or both hind limbs when walking or tail paralysis or anaesthesia; the rejection rate for these reasons accounts for 5% of the rats in these studies. The lidocaine test was also performed after each experiment to confirm that all drugs injected down the intrathecal catheters were injected into the lumbar intrathecal space. Rats were rested for 24 h before further experiments were performed. Only one experiment per day, up to a maximum of five, was performed on each animal.

Nociceptive tests
The ECT and TFL tests were performed as described previously.23 Both of these tests were applied in sequence every 5 min. The electrical test paradigm allows demonstration of drug action at the level of the spinal cord after lumbar intrathecal injection, i.e. the measurement of an increase in nociceptive threshold in the tail with no change at the neck. Such a differential block proves that the effects of the drug were confined to the caudal segments of the spinal cord responsible for tail and hind limb innervation. A rise in TFL in the same experiment in which such a differential effect is demonstrated must also therefore be due to action of the drug on the caudal segments of spinal cord.

Electrical current threshold test
Each rat was placed in a darkened restrainer and electrical stimulating electrodes were applied to the skin of the tail and neck. Brief bursts of electrical current (0.5 s train; 2 ms pulse width; 50 Hz; 0–10 mA) were delivered in turn to each set of electrodes (tail first, 15 s before the neck) every 5 min. The minimum current necessary to evoke an obvious nociceptive response (a squeak or sharp withdrawal), defined by the ‘up–down’ method, was measured at each skin site every 5 min until three stable consecutive readings had been obtained. The intrathecal drug was then given and measurements continued at each skin site every 5 min for a further 45–50 min. In order to standardize results obtained at each skin site, the response to intrathecal drug was calculated as the ratio of the mean of the three stable readings taken before intrathecal drug injection, as described previously.23 These values for responses to each dose of drug were combined and expressed as mean (SEM) on agonist dose–response curves.

Tail-flick latency test
This was performed every 5 min immediately before the electrical measurement in the tail. The heat source from a tail flick unit (Ugo Basile 7360) was directed onto the blackened tip of the tail and the power adjusted so the rat flicked its tail away from the noxious stimulus in approximately 3 s before any drug. At this intensity, three stable control (pre-drug injection) readings were obtained and then the test drug was injected. Tail-flick latency measurement continued every 5 min thereafter for 30–45 min. The response to each dose of drug was calculated as a percentage of the maximum possible effect (MPE):


Cut-off time was 10 s. Values were combined for each drug dose and expressed as mean (SEM) for plotting dose–response curves for intrathecal ketamine alone. For fentanyl dose–response curves from experiments in which fentanyl was given alone and in combination with either intrathecal or intraperitoneal ketamine, individual values for the TFL and neck and tail ECT thresholds were plotted on scatter plots against the dose of intrathecal fentanyl. Logistic regression [y=a loge(x)+b] using sums of squares was performed on these data. The resulting lines, with 95% confidence limits, were plotted superimposed on the means (SEM) for the ketamine dose–response curves and on the scatter plots of TFL and ECT responses for intrathecal fentanyl.

Ketamine dose–response relationships
Eight rats were used in 24 experiments in which ECT and TFL responses were measured after intraperitoneal injections of ketamine (Parke Davis, Cawarra, NSW, Australia; dose range 2.5–7.5 mg kg–1; six observations at each dose). Values for antinociceptive responses in the TFL and the ECT (neck and tail) tests were combined for each dose and expressed as mean (SEM) and tabulated.

Ketamine (25, 35 and 50 µg dissolved in 5 µl saline) was injected intrathecally in 23 experiments in 12 rats. The antinociceptive effects assessed with ECT (tail and neck) and TFL were combined for each dose of ketamine to calculate the mean and SEM and plotted on dose–response curves.

Intrathecal fentanyl dose–response relationships
The antinociceptive effects in the ECT test (tail and neck) and TFL test in response to a range of doses of intrathecal fentanyl (0.05–1.0 µg) were measured for fentanyl alone (39 experiments in 28 rats) and fentanyl in combination with intrathecal ketamine 25 µg (40 experiments in 28 rats) and intraperitoneal ketamine 3.75 mg kg–1 (30 experiments in eight rats).

Values for ECT (neck and tail) and TFL for each dose of fentanyl were plotted on scatter plots and linear regressions were used to calculate dose–response curves.


    Results
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
The results of all experiments in this study that were performed on rats with intrathecal catheters had a positive result in the lidocaine test both after catheter implantation surgery and after each experiment. Thus, in all experiments in which drugs were injected via intrathecal catheters, the compounds were administered into the lumbosacral cerebrospinal fluid. As with previous experiments of this type,23 in all animals in which experiments were repeated daily, we compared the control (pre-drug) nociceptive thresholds obtained on each day with those obtained on other occasions in the same rat. Although there were day-to-day changes in nociceptive thresholds, indicating changes in resistance due to electrode placement, there were no consistent increases or decreases in nociceptive thresholds with either test that would indicate progressive neurological damage or cumulative drug effects from one day to the next. No rats in these experiments showed overt signs of neurological damage, such as anaesthesia of the tail or hind limbs, autotomy or paralysis.

Intraperitoneal ketamine
The sedative and motor coordination effects assessed with the rotarod test and the antinociceptive effects assessed with the TFL and ECT tests, with a range of doses of intraperitoneal ketamine, are shown in Table 1. The lower doses of ketamine (2.5 and 3.75 mg kg–1) caused no loss of rotarod performance. Doses of intraperitoneal ketamine up to 7.5 mg kg–1 caused no antinociceptive effects in either the TFL or the ECT test.


View this table:
[in this window]
[in a new window]
 
Table 1 Percentage depression in rotarod task performance compared with control, TFL [% maximal possible effect (MPE)] and ECT (tail and neck data are the ratio with respect to control) measurements in rats given a range of doses of ketamine intraperitoneally (0–10 mg kg–1)
 
Intrathecal ketamine
Figure 1 shows dose–response curves for ketamine administered intrathecally. Intrathecal ketamine caused dose-related antinociceptive effects in TFL and tail ECT with no significant changes in neck ECT at any of the doses tested (P>0.05, Student’s paired t-test). The 25-µg intrathecal dose of ketamine caused no significant antinociceptive effects in either test (P>0.05, Student’s paired t-test).



View larger version (14K):
[in this window]
[in a new window]
 
Fig 1 Dose–response curves for intrathecal ketamine (dose range 25–50 µg). Intrathecal ketamine caused dose-related, spinally mediated antinociception revealed by tail ECT and TFL with no change in neck ECT thresholds. Data points are means and bars show the SEM. Tail ECT, circles; neck ECT, squares; TFL, triangles.

 
Intrathecal fentanyl dose–response relationships
Fentanyl given intrathecally caused dose-related antinociceptive effects, as revealed by the TFL and ECT tests, to the same extent as published previously with this model [23]. The dose–response curves for intrathecal fentanyl given alone are shown in Figures 2 and 3. The curves for the antinociceptive effects of the combinations of ketamine with intrathecal fentanyl are also shown on these figures for the purposes of comparison.



View larger version (18K):
[in this window]
[in a new window]
 
Fig 2 Dose–response curves for intrathecal fentanyl (dose range 0.05–1.0 µg). (A) Curves for tail and neck ECT after intrathecal fentanyl given alone. (B) Curves for intrathecal fentanyl in combination with intrathecal ketamine (25 µg). (C) Curves for TFL effects of intrathecal fentanyl given alone and in combination with intrathecal ketamine (25 µg). The data were subjected to linear regression to produce lines bounded by 95% confidence limits.

 


View larger version (17K):
[in this window]
[in a new window]
 
Fig 3 Dose–response curves for intrathecal fentanyl (dose range 0.05– 1.0 µg). (A) Curves for neck and tail ECT measurements after intrathecal fentanyl given in combination with intrathecal ketamine (3.75 µg). (B) Curves for TFL effects of intrathecal fentanyl given alone (triangles) and in combination with intraperitoneal ketamine (3.75 µg) (circle with a dot in the middle). Results of individual experiments are shown as points on a scatter plot. The data were subjected to linear regression to produce lines bounded by 95% confidence limits.

 
Antinociceptive effects in the tail were seen as increases in TFL and ECT in the tail with no change in neck ECT values [Fig. 2A (ECT) and C (TFL)]. This indicates that the fentanyl injected via the intrathecal catheter was confined to the caudal segments of the spinal cord.

Figure 2B (ECT tail and neck) and 2C (TFL) show the dose–response relationships for intrathecal fentanyl when ketamine 25 µg was given intrathecally in combination with each dose of intrathecal fentanyl. The sloping parts of these curves have been subjected to linear regressions, which are shown as lines bounded by 95% confidence limits. It can be seen that the addition of ketamine 25 µg to the intrathecal injection markedly potentiated the spinally mediated antinociceptive effects of fentanyl in the TFL test. The estimated dose of fentanyl to cause 50% of the MPE [i.e. the median effective dose (ED50)] was 0.12 µg for fentanyl in combination with ketamine 25 µg. This was significantly less than the ED50 for fentanyl given alone (0.4 µg) and outside the 95% confidence limits for the dose–response curve for fentanyl given alone. By contrast, the addition of ketamine to the intrathecal injection of fentanyl did not alter the ECT dose–response curve compared with fentanyl given alone (compare Fig. 2A with B). In the experiments in which there was a rise in TFL, there was also a rise in tail ECT with no change in neck ECT. Thus the interaction between fentanyl and ketamine occurred in the caudal segments of the spinal cord responsible for tail innervation.

Intraperitoneal injection of a small dose of ketamine (3.75 mg kg–1) that had no effect when given alone also potentiated the TFL effects of intrathecal fentanyl (Fig. 3B). The ED50 for the fentanyl dose–response curve in combination with intraperitoneal ketamine was 0.13 µg, a value similar to that for the combination of intrathecal ketamine and fentanyl (compare Fig. 3B with Fig. 2C). The dose–response curves for the two combinations were coincident. As in the experiments with intrathecal ketamine, in the intraperitoneal ketamine experiments in which there was a rise in TFL after intrathecal fentanyl there was also a rise in tail ECT with no change in neck ECT (Fig. 3A). Thus the interaction between fentanyl, which was given intrathecally, and ketamine, given intraperitoneally, occurred in the caudal segments of the spinal cord responsible for tail innervation. Intraperitoneal ketamine did not alter the ECT effects of intrathecal fentanyl (compare Fig. 3A with Fig. 2A).


    Discussion
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Many clinical studies have shown that ketamine is useful at subanaesthetic doses in the management of postoperative pain in humans.1117 In most of these studies the ketamine was given i.v. or s.c. The site of action for this analgesic effect could therefore be in either the brain or the spinal cord. It is often assumed that the site of action in causing analgesia and subsequent reduction in requirements for other analgesics, such as opioids, is the NMDA receptors in the spinal cord.

The involvement of spinal cord NMDA receptors with the generation and maintenance of acute and chronic pain states in animal models is well established.17 Ketamine has been shown to be an NMDA antagonist. This action has been shown to potentiate the antinociceptive effects of opioids assessed neurophysiologically in the spinal cord.8 9 This property seems to be due to the S isomer of ketamine.10 Thus it seems logical to assume that ketamine causes antinociceptive and analgesic effects and is opioid-sparing by an action at spinal cord NMDA receptors. The study reported in this paper addressed the question of spinal cord involvement.

It is clear from experiments with intrathecal ketamine alone that this drug can exert a spinally mediated antinociceptive effect because it caused increases in TFL and ECT in the tail with no change in neck ECT measurements. Furthermore, intrathecal coadministration of ketamine and fentanyl showed potentiation of fentanyl TFL antinociception. This was due to an interaction of the two drugs in the spinal cord because tail ECT values also increased with no change at the neck. In these experiments, tail ECT values were the same for intrathecal fentanyl alone and coadministered with intrathecal ketamine; potentiation occurred only for TFL. In these experiments the observer was not blinded to the nature of the drug treatment and thus observer bias could have affected the result. This is unlikely as there is no interpretation required by the observer in measuring TFL. The apparatus makes this measurement automatically when the rat moves its tail away from the heat source. This selectivity for TFL interaction by ketamine is interesting and without explanation. This result suggests that different neurophysiological mechanisms, involving different neurotransmitters, are activated by the different noxious stimuli used in the TFL and ECT tests. The TFL activates C-polymodal nociceptive fibres, which are also thought to play a role in human acute pain states. This interaction of ketamine and fentanyl at the level of the spinal cord could therefore be argued to predict a similar potentiation in humans.

By contrast, intraperitoneal injections of ketamine did not cause any antinociceptive effects, assessed with ECT and TFL, even at doses that caused significant motor incoordination in the rotarod test. However, the coadministration experiments followed a similar course to those with intrathecal dosing. The dose of intraperitoneal ketamine was one that caused no defect in motor coordination and was not sufficient to cause changes in either nociceptive threshold when the drug was given alone. This dose had no effect on the degree of tail ECT antinociception caused by intrathecal fentanyl, as with intrathecal ketamine. However, the ECT measurements were useful in that they revealed that intrathecal fentanyl was confined in its actions to the caudal segments of the spinal cord. Under these conditions intraperitoneal ketamine potentiated the TFL effects of intrathecal fentanyl. It is theoretically possible that ketamine had some action at a supraspinal site even though the dose was reduced to below that which caused motor incoordination. However, this action would have had to activate a descending pathway to the spinal cord that then potentiated the spinal cord antinociceptive effects of intrathecal fentanyl. The final pathway for the interaction between the opioid and ketamine occurred at the level of the spinal cord. One can make these conclusions because the fentanyl was restricted in its actions to the caudal spinal cord, as shown by increases in tail ECT and not neck ECT.

The fact that the same TFL dose–response relationship for intrathecal fentanyl was obtained when it was combined with ketamine given intrathecally and intraperitoneally implies that the same mechanisms were activated in the potentiation of the TFL effects of intrathecal fentanyl by ketamine given by either route. Moreover, there was evidence of confinement of ketamine to the caudal segments of the spinal cord when it was given intrathecally (increase in tail ECT but not neck ECT). Thus the argument about the coincidence of dose–response curves for intrathecal fentanyl in combination with ketamine given by the two different routes also opposes a supraspinal action of intraperitoneally administered ketamine.

We conclude that the administration of ketamine by a non-spinal parenteral route potentiates the spinally mediated antinociception caused by the µ-selective opioid fentanyl by an action at the level of the spinal cord. Although ketamine is known to have actions other than as an antagonist at NMDA receptors, it is likely that ketamine targets these receptors in the spinal cord when it is injected by any non-spinal parenteral route. We suggest that this may be the mechanism by which ketamine causes pain relief in humans and decreased opioid consumption after surgery.


    References
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
1 Baranauskas G, Nistri A. Sensitization of pain pathways in the spinal cord: cellular mechanisms. Prog Neurobiol 1998; 54; 349–65[ISI][Medline]

2 Dickenson AH. Central acute pain mechanisms. Ann Med 1995; 27: 223–7[ISI][Medline]

3 Dickenson AH, Chapman V, Green GM. The pharmacology of excitatory and inhibitory amino acid-mediated events in the transmission and modulation of pain in the spinal cord. Gen Pharmacol 1997; 28: 633–8[Medline]

4 Fürst S. Transmitters involved in antinociception in the spinal cord. Brain Res Bull 1999; 48: 129–41[ISI][Medline]

5 Silva E, Cleland CL, Gebhart GF. Contributions of glutamate receptors to the maintenance of mustard oil-induced hyperalgesia in spinalized rats. Exp Brain Res 1997; 117: 379–88[ISI][Medline]

6 Yaksh TL, Hua XY, Kalcheva I, Nozaki-Taguchi N, Marsala M. The spinal biology in humans and animals of pain states generated by persistent small afferent input. Proc Natl Acad Sci USA 1999; 96: 7680–6[Abstract/Free Full Text]

7 Mao J, Price DD, Hayes RL, Lu J, Mayer DJ, Frenk H. Intrathecal treatment with dextrorphan or ketamine potently reduces pain-related behaviors in a rat model of peripheral mononeuropathy. Brain Res 1993; 605: 164–8[ISI][Medline]

8 Chapman V, Dickenson AH. The combination of NMDA antagonism and morphine produces profound antinociception in the rat dorsal horn. Brain Res 1992; 573: 321–3[ISI][Medline]

9 Mennerick S, Zorumski CF. Presynaptic influence on the time course of fast excitatory synaptic currents in cultured hippocampal cells. J Neurosci 1995; 15: 3178–92[Abstract]

10 Joó G, Horvath G, Klimscha W, et al. The effects of ketamine and its enantiomers on the morphine- or dexmedetomidine-induced antinociception after intrathecal administration in rats. Anesthesiology 2000; 93: 231–41[ISI][Medline]

11 Schmid RL, Sandler AN, Katz J. Use and efficacy of low-dose ketamine in the management of acute postoperative pain: a review of current techniques and outcomes. Pain 1999; 82: 111–25[ISI][Medline]

12 Sethna NF, Liu MW, Gracely R, Bennett GJ, Max MB. Analgesic and cognitive effects of intravenous ketamine–alfentanil combinations versus either drug alone after intradermal capsaicin in normal subjects. Anesth Analg 1998; 86: 1250–6[Abstract]

13 Adriaenssens G, Vermeyen KM, Hoffmann VL, Mertens E, Adriaensen HF. Postoperative analgesia with i.v. patient-controlled morphine: effect of adding ketamine. Br J Anaesth 1999; 83: 393–6[Abstract/Free Full Text]

14 Javery KB, Ussery TW, Steger HG, Colclough GW. Comparison of morphine and morphine with ketamine for postoperative analgesia. Can J Anaesth 1996; 43: 212–5[Abstract]

15 Menigaux C, Fletcher D, Dupont X, Guignard B, Guirimand F, Chauvin M. The benefits of intraoperative small-dose ketamine on postoperative pain after anterior cruciate ligament repair. Anesth Analg 2000; 90: 129–35[Abstract/Free Full Text]

16 Suzuki M, Tsueda K, Lansing PS, et al. Small-dose ketamine enhances morphine-induced analgesia after outpatient surgery. Anesth Analg 1999; 89:; 98–103[Abstract/Free Full Text]

17 Chia YY, Liu K, Liu YC, Chang HC, Wong CS. Adding ketamine in a multimodal patient-controlled epidural regimen reduces postoperative pain and analgesic consumption. Anesth Analg 1998; 86: 1245–9[Abstract]

18 Yang CY, Wong CS, Chang JY, Ho ST. Intrathecal ketamine reduces morphine requirements in patients with terminal cancer pain. Can J Anaesth 1996; 43: 379–83[Abstract]

19 Lauretti GR, Azevedo VMS. Intravenous ketamine or fentanyl prolongs postoperative analgesia after intrathecal neostigmine. Anesth Analg 1996; 83: 766–70[Abstract]

20 Nadeson R, Guo Z, Porter V, Gent JP, Goodchild CS. GABAA receptors and spinally-mediated antinociception in rats. J Pharmacol Exp Ther 1996; 278: 620–6[Abstract]

21 Zimmerman M. Ethical guidelines for investigation of experimental pain on conscious animals. Pain 1983; 16: 109–10[ISI][Medline]

22 Nadeson R, Goodchild CS. Antinociceptive properties of propofol: involvement of spinal cord gamma-aminobutyric acidA receptors. J Pharmacol Exp Ther 1997; 282: 1181–6[Abstract/Free Full Text]

23 Edwards M, Serrao JM, Gent JP, Goodchild CS. On the mechanism by which midazolam causes spinally mediated analgesia. Anesthesiology 1990; 73: 273–7[ISI][Medline]