| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
*Department of Anesthesiology and
Division of Pain Therapy, University Hospital of Bern, Switzerland;
Department of Anesthesia and Intensive Care, Kantonsspital St. Gallen, Switzerland; and
Center for Sensory-Motor Interaction, University of Aalborg, Aalborg, Denmark
Address correspondence to Martin Luginbühl, MD, DEAA, Department of Anesthesiology, University Hospital, CH-3010 Bern, Switzerland. Address e-mail to martin.luginbuehl{at}dkf2.unibe.ch Reprints will not be available from the authors.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionConclusionReferences |
|---|
IMPLICATIONS: Coadministration of ketamine and morphine for pain relief is still controversial. Our experimental pain study with volunteers showed that ketamine enhances opioid analgesia without increasing sedation and reduces respiratory depression. Opioid-induced hyperalgesia and tolerance were not affected by ketamine and depended on the type of nociceptive stimulus. This may explain the conflicting results on opioid tolerance in previous studies.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
In this randomized, double-blinded, placebo-controlled crossover study in volunteers, we tested the following hypothesis: opioid-induced hyperalgesia and acute opioid tolerance, as well as their prevention by ketamine, occur with some, but not all, types of painful stimuli. A secondary hypothesis was that ketamine enhances opioid analgesia and modifies its side effect profile. These hypotheses were tested by using a multimodal, multistructure experimental pain approach (12). Remifentanil was selected as the opioid because it has been reported to rapidly induce tolerance (5,8) and because of its rapid pharmacokinetics, which make it extremely suitable for volunteer studies.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
Each volunteer was tested in four study sessions, at least 4 days apart, and each session was performed at the same time of day. Four different drug regimens—one per study session—were applied in a randomized, double-blinded, crossover fashion: placebo (normal saline), remifentanil only, and remifentanil plus ketamine (50 ng/mL) and remifentanil plus ketamine (100 ng/mL), respectively (Fig. 1). Randomization was performed by drawing lots from sealed envelopes.
|
After a trial series to familiarize the volunteers with all experimental pain tests, 6 test series were performed in each study session (Fig. 1). The baseline series was performed without drugs. Series 1 (pure ketamine or placebo) was performed 15 min after the start of a target-controlled infusion of either ketamine (target plasma concentration, 50 or 100 ng/mL) or placebo (normal saline). Series 2 was performed 10 min after a second target-controlled infusion with either remifentanil (target plasma concentration, 1 ng/mL) or placebo (normal saline) was started, while the first infusion was kept constant. The subsequent 3 test series were performed at target plasma remifentanil concentrations of 2, 1, and 0 ng/mL with the ketamine infusion unchanged and with 10 min of equilibration after each change of the remifentanil concentration. After completion of the session, the volunteers were discharged according to the criteria for ambulatory anesthesia used in our department.
Remifentanil (Ultiva®; Glaxo-Smith-Kline Switzerland, Münchenbuchsee, Switzerland) and racemic ketamine (Ketalar®; Pfizer Pharmaceuticals, Zurich, Switzerland) were given as target-controlled IV infusions with use of two Harvard 22 infusion pumps (Harvard Apparatus, Kent, UK) driven by the Stanpump program (S. Schafer, Palo Alto, CA) with the pharmacokinetic variable set of Minto et al. (13) for remifentanil and Domino et al. (14) for ketamine. The syringes were filled with either the drugs or normal saline by a study nurse otherwise not involved in the study. The selected remifentanil target concentrations corresponded to those suggested for conscious sedation.1 The ketamine target plasma concentration was kept constant at 50 or 100 ng/mL. These concentrations are less than the 350 ng/mL used in a previous study, in which central side effects were occasionally observed (15). Four venous blood samples were obtained for determination of ketamine plasma concentration—at baseline and at 3, 10, and 60 min or at 3, 20, and 120 min after the start of the ketamine infusion, with random selection of the sampling schedule. Three further blood samples were taken immediately before and 10 and 20 min after discontinuation of the ketamine infusion, so that seven samples were drawn from each subject. The blood samples were centrifuged at 3500 rpm for 30 min and the plasma frozen at -18°C for later analysis. The ketamine plasma concentrations were determined by high-performance liquid chromatography with a detection limit of 10 ng/mL and a coefficient of variation of 2.1% at a concentration of 106 ng/mL in an external laboratory (Biochem A GmbH, Riegel, Germany).
Remifentanil, unlike ketamine, has rapid pharmacokinetics and equilibration between plasma and effect site (16), which allows rapid achievement of a steady-state (17). Moreover, the reported inaccuracy of the pharmacokinetic variable set used (16) was in the range of 20% and was thus rather small. We therefore considered the predicted plasma remifentanil concentrations more reliable than the predicted ketamine concentrations and did not measure remifentanil plasma concentrations.
The same investigator (AG) performed all experimental pain tests in all subjects. The investigator and the volunteers were blinded to the type of drugs administered. The pain detection threshold after single and repeated IM electrical stimulation and repeated transcutaneous electrical stimulation and the pain tolerance threshold to pressure stimulation were determined in each test series in a randomized order. The order was randomly assigned by drawing lots at the first study session of each subject and was kept constant for the subsequent sessions.
For single IM electrical stimulation, two needle electrodes (Dantec, Skovlunde, Denmark) with 3-mm insulated tips were inserted 2 cm into the anterior tibial muscle, 15 cm distal to the lower border of the patella, 2 and 2.5 cm lateral to the lateral edge of the tibia (18). A 25-ms train of 5 1-ms square-wave impulses (perceived as a single stimulus) was delivered from a computer-controlled constant current stimulator (Noxitest Inc., Aalborg, Denmark). The current intensity was increased in steps of 0.1 mA from 0.2 to 2.0 mA, in steps of 0.5 mA from 2.0 to 5.0 mA, and in steps of 1.0 mA from 5.0 mA. After each stimulation, the subject was asked if he or she perceived the stimulus as painful. The stimulus intensity evoking a pain sensation (pain detection threshold) was recorded.
Repeated stimulation at constant intensity may evoke an increase in perception during the repeated stimulation, so that the last stimuli are perceived as painful (19). This temporal summation of nociceptive stimulation results from a short-lasting hyperexcitability of the spinal cord induced by the repeated stimulation (19). Five equal electrical stimuli—each of them similar to the aforementioned single IM stimulus—were delivered to the same IM electrodes at a frequency of 2 Hz (20). The repeated stimulation pain threshold was defined as the current intensity (mA) evoking an increasing sensation from the first to the last of the five equal stimuli, with the last one to three stimuli being perceived as painful (summation pain threshold). The summation threshold was identified by a stepwise increase of the current intensity, as described for single electrical stimulation.
For repeated transcutaneous electrical stimulation, two bipolar surface Ag/AgCl electrodes (Dantec) were placed on the skin below the lateral malleolus 2 cm apart, thus stimulating the sural nerve (20). Repeated electrical stimulation was performed as described for IM stimulation to determine the summation pain threshold.
For all the electrical stimulations, a maximal current of 80 mA was set as a cutoff limit to avoid tissue damage. The mean of 3 threshold determinations was used for data analysis.
For pressure pain tolerance threshold, pressure stimulation was applied to the center of the pulpa of the second and third toes. An electronic pressure algometer (Somedic AB, Stockholm, Sweden) (21) with a surface area of 64 mm2 was used. The pressure was increased at 30 kPa/s to a maximum pressure of 1500 kPa. The pain tolerance threshold was defined as the pressure (kPa) at which the subject felt that the pain was intolerable. The mean of the two determinations from the second and third toes was considered for data analysis.
Sedation was assessed with a 10-cm visual analog scale (0 = fully fit, 10 = hardly able to keep the eyes open) (8). The reaction time was determined by delivering a 1000-Hz tone from a computer at randomized intervals of 3 to 8 s. The time from the onset of the tone until the volunteer pressed a button was defined as reaction time. The mean of five consecutive measurements was recorded.
The occurrence of dreams or hallucinations, pruritus, and nausea was additionally recorded after each study session. To assess the respiratory and cardiovascular function, oxygen saturation (SpO2), end-expiratory CO2 recorded from a nasal probe, heart rate (electrocardiogram), and noninvasive blood pressure were recorded after equilibration but before noxious stimulation. Nasal breathing of the subjects was ensured during the recording of end-expiratory CO2.
All the above side-effect variables were recorded immediately before each test series, after equilibration at the new target drug concentrations. The statistical analysis was performed with SigmaStat 2.03 software (Jandel Scientific Corp., San Rafael, CA), with a P value <0.05 considered significant.
The main hypothesis was that remifentanil induced hyperalgesia and tolerance. Remifentanil-induced hyperalgesia was defined as a decrease in pain thresholds measured before the start and after the discontinuation of the remifentanil infusion (S1 and S5, respectively; see Fig. 1). Remifentanil tolerance was defined as a decrease in pain thresholds from the first to the second measurement performed at 1 ng/mL (S2 and S4, respectively; see Fig. 1). The pain thresholds at the two time points were compared with a paired Student’s t-test or a Wilcoxon’s rank test with the Bonferroni correction for multiple testing.
Regarding the secondary hypothesis, about the analgesic effect and side effects of remifentanil and ketamine, the pain thresholds and the side effects (sedation score, reaction time, SpO2, end-expiratory CO2, heart rate, and blood pressure) recorded in corresponding test series of the different sessions were compared with repeated-measurements analysis of variance on ranks (Friedman’s test), with the Student-Newman-Keuls test for multiple comparisons.
To detect a dose-response relation of remifentanil, the measurements from S3 (2 ng/mL of remifentanil) were compared with the measurements of S2 and S4 with repeated-measurements analysis of variance on ranks. Because the absolute values of the pain thresholds are of little clinical importance, they are presented as percentages of baseline to facilitate comparison between the different pain tests. Conversely, the absolute values of side effects (sedation, CO2, and SpO2) are clinically important and are presented as raw data.
A multiple pairwise comparison of the incidence of dreams, hallucinations, nausea, vomiting, and pruritus in the different treatment groups was performed with McNemar’s test for repeated measurements. The measured plasma ketamine concentrations at the different sampling points are presented as mean and SD in Figure 1.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
The pressure pain tolerance significantly decreased after discontinuation of remifentanil compared with the measurement before starting the remifentanil infusion. This was observed in all sessions except placebo, indicating an induction of hyperalgesia that is not prevented by ketamine (Fig. 2A). The electrical pain tests, in contrast, did not reveal any remifentanil-induced hyperalgesia. The pain thresholds to electrical stimulations even increased in some sessions (Fig. 2B–D). This increase may reflect either the increasing ketamine plasma concentrations (Fig. 1) or a certain decrease of pain sensitivity in the electrical pain tests.
|
Remifentanil increased all the pain thresholds, as well as the sedation score, in a dose-dependent manner (Fig. 2A–E). The reaction time, however, was not affected by remifentanil (data not shown).
A pure ketamine effect was expected in the second test series, i.e., before the start of the remifentanil infusion. Ketamine alone at a target plasma concentration of 100 ng/mL increased the pain threshold to repeated IM but not transcutaneous stimulation and also had a significant sedative effect (at both target concentrations). In combination with remifentanil, 100 ng/mL of ketamine increased the IM pain threshold (single and repeated stimulation) at both remifentanil concentrations. It did not further increase the sedative effect of remifentanil, however. Ketamine did not affect the repeated transcutaneous stimulation pain threshold, the pressure pain tolerance threshold, or the reaction time (data not shown).
Dreams, although pleasant, occurred in 5 of 14 subjects during ketamine infusion at a plasma target concentration of 100 ng/mL, in 1 subject during 50 ng/mL of ketamine, and in another during pure remifentanil infusion (pairwise comparison with McNemar’s test; not significant). One subject reported pleasant hallucinations under the pure remifentanil infusion and under remifentanil combined with 50 ng/mL of ketamine. Another subject had unpleasant hallucinations at the larger ketamine concentration and did not complete this session. Nausea occurred in one subject during the pure remifentanil session and in three subjects during the larger-dose ketamine session (not significant). The incidence of pruritus was 64% during remifentanil alone, 79% during remifentanil combined with the smaller dose, 50% during remifentanil combined with the larger dose of ketamine, and 0% during placebo administration (the ketamine effect was not significant).
There was no significant change in SpO2 throughout the study. The subjects received oxygen by a nasal cannula, however. The smallest SpO2 recorded was 90% at the largest remifentanil target concentration without ketamine in one subject (data not shown).
Median (interquartile range) end-expiratory CO2 during remifentanil at a 2 ng/mL target concentration with and without 50 ng/mL and 100 ng/mL of ketamine was 43.5 mm Hg (42–47 mm Hg), 42 mm Hg (38–43 mm Hg), and 37.5 mm Hg (35–47 mm Hg), respectively (P < 0.05). In the corresponding placebo series, it was 38 mm Hg (35–41 mm Hg) (P < 0.05 compared with pure remifentanil and remifentanil plus 50 ng/mL of ketamine). In the other series, the end-expiratory CO2 was similar in all study sessions.
The largest end-expiratory CO2 concentration recorded was 51 mm Hg in the pure remifentanil session. Heart rate and blood pressure did not change throughout the study in any subject. The pharmacokinetic variables by Domino et al. (14) used to drive the computer-controlled infusion pump did not allow achieving steady-state concentrations during test series 1 and 2 of the study (Fig. 1).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
In the electrical pain tests, in contrast, the single- and repeated-stimulation pain thresholds did not decrease but, rather, increased. The increase was statistically significant mostly in sessions in which ketamine was administered. This increase may thus reflect the increasing ketamine concentration during the study. We did not find any hyperalgesia with the electrical pain tests. Koppert et al. (22) found a decrease of pinprick hyperalgesia and touch-evoked allodynia during alfentanil administration. This was assessed in the skin area around an intracutaneous electrical stimulation needle, whereas we measured the decrease of pain threshold of the electrical stimulation itself.
We measured the pain tolerance threshold in the pressure pain model and the pain detection threshold in the electrical pain models. Therefore, either the type of nociceptive stimulation (pressure versus electrical) or the different pain intensity (pain detection versus pain tolerance threshold) may explain the different findings between the pressure pain and the electrical pain paradigms.
The decrease in the forth series (1 ng/mL of remifentanil) compared with the second series with the same remifentanil concentration could also be interpreted partially as hyperalgesia induced by the increased opioid exposure (2 ng/mL) in the third series. In this case, a mixed effect of tolerance and hyperalgesia would account for the decrease in pressure pain tolerance. The pathogenesis of tolerance and hyperalgesia may be similar (23), and opioid doses are often varied in clinical practice. Therefore, the distinction between tolerance and hyperalgesia during opioid administration is probably of little clinical relevance.
Remifentanil alone increased all pain detection and tolerance thresholds and the sedation score. Pain induced by repeated transcutaneous and IM electrical stimulation evokes central temporal summation (19). The well known involvement of NMDA receptors in central temporal summation (24) explains why adding small concentrations of the NMDA receptor antagonist ketamine increased the analgesic effect of remifentanil in the repeated IM stimulation test, inducing temporal summation. The IM stimulation model detected the effect of ketamine alone and of ketamine added to remifentanil, whereas the transcutaneous stimulation model did not. Therefore, muscle pain models may be more sensitive than cutaneous pain tests for detecting the effects of ketamine. The pressure pain tolerance test did not detect the increase in analgesic effect achieved by adding ketamine to remifentanil. The reason may be that the pressure pain test did not evoke the same summation mechanisms in the dorsal horn neurons.
The increase of pain threshold to IM single and repeated electrical stimulation detected in the last test series (Fig. 2, B and C) mainly reflects the increasing ketamine concentration, because the simulated plasma remifentanil concentration in the last test series was in the range of 0.2 ng/mL, which did not have any detectable analgesic effect. Despite the sedative side effect of remifentanil and ketamine, the reaction time did not change throughout the study. This implies that the reaction time is not a reliable measure for sedation, probably because of the arousal effect of the test itself. The fact that reaction time remained unchanged demonstrates that during the pain testing the volunteers were adequately alert to perform the pain tests. The sedation score, however, was always assessed before testing in the nonstimulated subjects.
Ketamine antagonized the respiratory depressant effect of remifentanil, which is consistent with the respiratory-stimulating effect of ketamine reported in previous studies (25,26). This is a clinically relevant finding that may provide an additional reason for combining opioids with ketamine in pain management.
Coadministration of ketamine and morphine for clinical pain relief is still controversial (27,28). Our results provide experimental support for the advantages of combining NMDA receptor antagonists and opioids in treating human pain. Analgesia is enhanced, spinal cord hyperexcitability is more profoundly attenuated, respiratory depression is reduced, and opioid-induced sedation is not further increased. However, we found no effect of ketamine on opioid-induced hyperalgesia. Different effects of remifentanil on different types of clinical pain could partly explain the controversies over whether opioids induce hyperalgesia and tolerance in clinical conditions.
|
Conclusion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  B. Singler, A. Troster, N. Manering, J. Schuttler, and W. Koppert Modulation of Remifentanil-Induced Postinfusion Hyperalgesia by Propofol Anesth. Analg., June 1, 2007; 104(6): 1397 - 1403. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. R. Absalom, M. Lee, D. K. Menon, S. R. Sharar, T. De Smet, J. Halliday, M. Ogden, P. Corlett, G. D. Honey, and P. C. Fletcher Predictive performance of the Domino, Hijazi, and Clements models during low-dose target-controlled ketamine infusions in healthy volunteers Br. J. Anaesth., May 1, 2007; 98(5): 615 - 623. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. C. Colpaert, K. Deseure, L. Stinus, and H. Adriaensen High-Efficacy 5-Hydroxytryptamine 1A Receptor Activation Counteracts Opioid Hyperallodynia and Affective Conditioning J. Pharmacol. Exp. Ther., February 1, 2006; 316(2): 892 - 899. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Hans, P.-Y. Dewandre, J. F. Brichant, and V. Bonhomme Comparative effects of ketamine on Bispectral Index and spectral entropy of the electroencephalogram under sevoflurane anaesthesia Br. J. Anaesth., March 1, 2005; 94(3): 336 - 340. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Schulte, A. Sollevi, and M. Segerdahl The Synergistic Effect of Combined Treatment with Systemic Ketamine and Morphine on Experimentally Induced Windup-Like Pain in Humans Anesth. Analg., June 1, 2004; 98(6): 1574 - 1580. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
= ketamine target concentration 100 ng/mL). The test series are labeled with BL (baseline), S1 (pure ketamine 50 or 100 ng/mL or placebo), S2 (remifentanil 1 ng/mL or placebo plus ketamine or placebo), S3 (remifentanil 2 ng/mL or placebo plus ketamine or placebo), S4 (remifentanil 1 ng/mL or placebo plus ketamine or placebo), and S5 (10 min after discontinuation of remifentanil, with ketamine or placebo maintained).
= remifentanil, 
= remifentanil + ketamine (50 ng/mL), 
= remifentanil + ketamine (100 ng/mL).
