Share this article on:

Small-Dose S(+)-Ketamine Reduces Postoperative Pain When Applied with Ropivacaine in Epidural Anesthesia for Total Knee Arthroplasty

Himmelseher, Sabine MD; Ziegler-Pithamitsis, Doris MD; Argiriadou, Helena MD; Martin, Jan MD; Jelen-Esselborn, Sabine MD; Kochs, Eberhard MD

doi: 10.1097/00000539-200105000-00040
Regional Anesthesia And Pain Medicine: Research Report

Reduction of nociceptive input through blockade of N-methyl-d-aspartate (NMDA) receptors has been reported. We compared the effects of epidural S(+)-ketamine versus placebo on postoperative pain in a randomized, double-blinded study in 37 patients undergoing unilateral knee arthroplasty. After lumbar epidural anesthesia with ropivacaine (10 mg/mL, 10–20 mL), 19 patients received 0.9% epidural saline, and 18 patients received 0.25 mg/kg epidural S(+)-ketamine 10 min before surgical incision. After surgery, patient-controlled epidural analgesia with ropivacaine was provided. During the first 8 h after surgery, visual analog scale pain rating was similar between groups. Twenty-four and 48 h after surgery, patients anesthetized with ropivacaine had higher visual analog scale ratings at rest and during movement (P < 0.05) than patients anesthetized with S(+)-ketamine and ropivacaine. Forty-eight hours after surgery, patients anesthetized with ropivacaine also consumed more ropivacaine (558 ± 210 mg) (P < 0.01) than those anesthetized with S(+)-ketamine and ropivacaine (319 ± 204 mg). Adverse events were similar between groups. Patients who received S(+)-ketamine and ropivacaine rated the quality of their pain therapy better than those who received ropivacaine alone (P < 0.05). We conclude that the combination of S(+)-ketamine and ropivacaine in epidural anesthesia increases postoperative pain relief when compared with ropivacaine.

Klinik für Anaesthesiologie, Technische Universität München, München, Germany

January 9, 2001.

Implications: Epidural S(+)-ketamine applied with ropivacaine before surgery is a rational approach to decrease injury-induced pain sensitization. Epidural blockade with an N-methyl-d-aspartate receptor antagonist and a local anesthetic may provide better analgesia in the postoperative period than a local anesthetic alone.

Address correspondence and reprint requests to Sabine Himmelseher, Klinik für Anaesthesiologie, Klinikum rechts der Isar, Technische Universität München, Ismaningerstr. 22, D-81675 München, Germany. Address e-mail to

The efficacy of postoperative pain therapy after total knee replacement surgery is a major factor in functional outcome (1). Efforts to improve postoperative pain relief have concentrated on pain preventive regimens directed at decreasing noxious input and sensitization in pain pathways (2,3). Because nociceptive transmission and synaptic plasticity are critically related to activation of excitatory glutamate receptors of the N-methyl-d-aspartate (NMDA) subtype (2,3), therapy that aims at minimizing changes in the NMDA receptor system may prevent pain.

Racemic ketamine blocks NMDA receptor activation in a noncompetitive manner (4,5). In pain-preventive regimens, subanesthetic-dose IV ketamine reduced postoperative pain (6,7), and in combination with epidural opioids or local anesthetics, epidural ketamine increased postoperative analgesia (8–10). S(+)-ketamine, the left-handed optical isomer of racemic ketamine, has a fourfold higher affinity for NMDA receptors than right-handed R(−)-ketamine (5,11). Investigative trials indicate that this difference results in a clinical analgesic potency of S(+)-ketamine that is two times higher in comparison with racemic ketamine (12,13). S(+)-ketamine is now available in some European countries, and a preservative-free formulation that may be administered into the epidural space has been introduced.

Because epidural anesthesia and postoperative epidural analgesia have been associated with accelerated rehabilitation after knee arthroplasty (1), we compared the efficacy of postoperative patient-controlled epidural analgesia (PCEA) by using ropivacaine after epidural anesthesia with ropivacaine compared with that with S(+)-ketamine and ropivacaine.

Back to Top | Article Outline


After ethical committee approval and informed consent, 42 patients undergoing elective unilateral total knee arthroplasty under epidural anesthesia were enrolled in this randomized, prospective, double-blinded study. Exclusion criteria included contraindications to epidural anesthesia or the anesthetics used; age <18 yr or >80 yr; weight <45 kg or >145 kg; evidence of severe cardiovascular, renal, hematologic, or hepatic disease; ASA physical status >III; preexisting neurological or psychiatric illnesses; chronic pain syndromes; alcohol or drug abuse; and difficulties in communication or expected inability to understand patient-controlled analgesia.

On the preoperative visit, patients were made familiar with the visual analog scale (VAS), which ranged from 0 = no pain to 10 = worst pain imaginable. Use of the PCEA device (Graseby Medical Ltd., Watford, UK) was explained.

On the day of surgery, patients were premedicated with oral midazolam (3.75–7.5 mg) 1 h before surgery. In the operation room, an IV infusion was established, and prehydration with at least 500 mL lactated Ringer’s solution was provided. An epidural catheter was preferentially inserted at the L3-4 interspace, but variations in the site of insertion were allowed from T12 through L4. The epidural space was identified by loss of resistance. The catheter was advanced 3–5 cm, and a 4-mL test dose of ropivacaine (10 mg/mL) (Astra Chemicals, Wedel, Germany) eliminated subarachnoid insertion. Five minutes after a negative response, 10–15 mL epidural ropivacaine was titrated in aliquots of 5 mL to a sensory block of at least dermatome T12. If adequate sensory block was not achieved within 45 min, an additional 5-mL dose of ropivacaine was injected. A limit of a maximum dose of 250 mg ropivacaine within a 3-h interval was set. For patient comfort, light sedation with IV propofol or midazolam was provided. Sensory block was confirmed before surgery was started.

According to a computer-generated table of random numbers, patients were allocated to receive either 0.25 mg/kg epidural S(+)-ketamine (Parke-Davis Pharmaceuticals, Freiburg, Germany) or an equivalent volume of epidural 0.9% saline 10 min before surgical incision. The S(+)-ketamine dose chosen was based on open-label pilot studies in the same setting. We studied the effects of 0.125, 0.25, 0.5, and 0.75 mg/kg epidural S(+)-ketamine, and we observed increased postoperative pain relief and a 15%–40% decrease in ropivacaine demand after 0.25 mg/kg S(+)-ketamine. The smaller doses assessed resulted in negligible effects, whereas the larger doses studied caused increased patient sedation. Therefore, 0.25 mg/kg S(+)-ketamine was used for the investigation. Only commercially available preservative-free S(+)-ketamine was applied. Patients and investigator anesthetists were blinded as to patient group allocation.

On completion of surgery, analgesic therapy was commenced. The PCEA device was programmed to administer a bolus dose of 18.75 mg (5 mL) of ropivacaine (3.75 mg/mL) with a 10-min lockout and a maximum dose limit of 200 mg of ropivacaine over 4 h. Postoperative evaluations were performed at 2-h intervals for the first 8 h after surgery and then at 24 and 48 h after surgery. VAS scores at rest and during movement and PCEA consumption of ropivacaine were recorded. The time until first analgesic medication request was noted. Sensation to touch (with the scale 0 = none, 1 = pinprick, 2 = touch, to 3 = correct identification of a number written on the skin) was determined on the skin surrounding the operated area and the corresponding region on the contralateral extremity. Motor block was assessed bilaterally according to a modified Bromage scale (0 = no motor block, 1 = ability to move knee and ankle, 2 = ability to move ankle, 3 = inability to move ankle).

A standardized plan for postoperative rescue analgesia was prepared. If VAS scores were higher than 4 cm, nonsteroidal antiinflammatory drugs (50–100 mg diclofenac or 0.5–1 g paracetamol administered per rectum or orally) were first used, followed by the nonopioid analgesic metamizol (1–2 g IV). In case of pain persistence after 45 min, subcutaneous injections of piritramide (1.5 up to 7.5 mg) were provided. If the VAS scores were higher than 5 cm, piritramide was administered IV. If analgesia was not achieved with these treatments, the PCEA was discontinued, and 5–10 mL (18.75–37.5 mg) boluses of ropivacaine could be epidurally applied at the discretion of the anesthetist. If, after 45 min, pain control was considered inadequate (VAS scores higher than 4 cm), a continuous infusion of 0.2% epidural ropivacaine was provided. Finally, an IV patient-controlled analgesia device with piritramide could be set up.

Adverse events were noted. Special attention was paid to all side effects that could be associated with ketamine (5–10). Patients were questioned about unusual dreams, hallucinations, or other psychological effects possibly related to ketamine. Noninvasive blood pressure and heart rate were monitored, and the state of sedation was assessed (4-point score: 0 = no sedation, 1 = mild sedation, 2 = moderate sedation, 3 = severe sedation) at 30-min intervals. Patients’ ratings of the quality of their overall pain therapy were recorded at 48 h with a 4-point scale of 1 = poor, 2 = fair, 3 = good, or 4 = excellent.

Calculations from the pilot studies showed that 15 patients per group would allow detection of a difference of 20% in postoperative ropivacaine consumption with an overall α error at the 0.05 level and an 80% power. All P values were two sided. Where appropriate, results are presented as count and were statistically compared between groups by contingency table analysis. Continuous variables are reported as mean ± sd or median and range (25th–75th percentile) when data were not normally distributed. Discrete results are expressed as median and range (25th–75th percentiles). Nonparametric data were analyzed with the Kruskal-Wallis test combined with Dunn’s test for intergroup comparison. Parametric data were compared between groups by analysis of variance and post hoc testing with Dunnett’s test. Analysis was performed with the Statistical Package for the Social Sciences (SPSS, Inc., Chicago, IL) and StatView 4.5 (Abacus Concepts, Berkeley, CA).

Back to Top | Article Outline


Of the 42 enrolled patients, one patient had to be withdrawn because of failure to place the epidural catheter and one because of high dermatome block above T4 before epidural injection of the study medication. One patient who developed anaphylactic shock after infusion of an antibiotic in the preoperative period was excluded. Premature study discontinuation occurred in one patient anesthetized with S(+)-ketamine and ropivacaine who developed postoperative fever for which the epidural catheter was removed, and in one patient anesthetized with ropivacaine who had early postoperative catheter displacement. The data of 19 patients who received ropivacaine and of 18 patients anesthetized with S(+)-ketamine and ropivacaine were eligible for analysis. Patient characteristics and preoperative and intraoperative data did not differ between groups (Table 1).

Table 1

Table 1

In the postoperative period, time to first analgesic request was comparable between patients anesthetized with ropivacaine (170 min; range, 120–195 min) or S(+)-ketamine and ropivacaine (125 min; range, 90–240 min). The degree of sensory and motor block of the right and left leg was similar between groups (data not shown). VAS scores at rest and during movement differed at 24 and 48 h after surgery, with higher scores in patients who received ropivacaine (P < 0.05) (Fig. 1). Patients anesthetized with ropivacaine requested more ropivacaine via PCEA 48 h after surgery than patients anesthetized with S(+)-ketamine and ropivacaine (P < 0.01) (Fig. 2). Consumption of nonsteroidal antiinflammatory drugs and metamizol did not differ between groups (Table 2). Three patients of the Ropivacaine group needed IV piritramide, and two patients of the S(+)-Ketamine/Ropivacaine group received subcutaneous piritramide (Table 2). One patient anesthetized with ropivacaine required a continuous epidural infusion of 0.2% ropivacaine at the beginning of the second 24-h postoperative period, and another one at the end of the 48-h study. One patient who received ropivacaine felt insufficient pain relief despite an apparently adequate epidural block during the first postoperative night, requiring an IV PCA device with piritramide.

Figure 1

Figure 1

Figure 2

Figure 2

Table 2

Table 2

The incidences of postoperative adverse events were equivalent between the groups (Table 2). Postoperative hypotension (<90 mm Hg) was managed with IV fluid therapy; two patients in the Ropivacaine and two patients in the S(+)-Ketamine/Ropivacaine group suffered from wound bleeding requiring transfusion of packed red cells and fresh frozen plasma (Table 2). Patients anesthetized with S(+)-ketamine/ropivacaine rated the quality of their pain therapy better than patients anesthetized with ropivacaine (P < 0.05). No patient of the S(+)-Ketamine/Ropivacaine group indicated psychological events. Four patients who received ropivacaine after surgery reported dysphoric feelings of anxiety or nervousness.

Back to Top | Article Outline


This first clinical study on use of S(+)-ketamine together with ropivacaine in epidural anesthesia showed enhanced pain relief 24 and 48 hours after knee arthroplasty and decreased analgesic demand 48 hours after surgery. There was no increase in adverse events after 0.25 mg/kg epidural S(+)-ketamine.

Total knee arthroplasty is associated with a frequent incidence of intense postoperative pain, especially during early attempts of mobilization (1). Our analysis focused on the first 48 hours after surgery, during which many patients experience a peak pain level. We chose to compare postoperative analgesia with epidural ropivacaine after epidural anesthesia with S(+)-ketamine and ropivacaine versus ropivacaine. Epidural anesthesia with local anesthetics is often used for knee arthroplasty, and studies with epidural ropivacaine indicate safe and effective analgesia with minimal side effects after major orthopedic surgery (14).

Concern has been expressed about neurotoxicity after neuraxial use of ketamine, because spinal myelopathy has been reported with intrathecal injection of large doses (15). However, in animals and humans, there is no evidence for neurological injury after repeated intrathecal injection of preservative-free ketamine (16,17). Little is known about the pharmacokinetic distribution of epidural ketamine. In a dog model, racemic ketamine was rapidly absorbed from the epidural space into the cerebrospinal fluid and plasma with a longer plasma half-life than if applied IV (18). In humans, a single five-milligram epidural injection gained rapid access to the systemic circulation with 80% bioavailability (19). One possible criticism of this study is therefore the lack of blood measurements of S(+)-ketamine and its metabolites. Thus we cannot comment on systemic analgesic effects possibly induced by epidural S(+)-ketamine. A limitation of our investigation may be the lack of a patient group receiving an equipotent dose of racemic ketamine relative to S(+)-ketamine. Ethical considerations dominated our decision not to investigate such a group, because epidural small-dose racemic ketamine combined with bupivacaine failed to improve postoperative pain relief after knee arthroplasty with epidural anesthesia in a study by Weir and Fee (20).

The postoperative VAS scores obtained after ropivacaine in this investigation compare with VAS scores for unilateral knee arthroplasty with epidural anesthesia and postoperative epidural analgesia that have been reported in the literature (1,9,14). The few trials available that studied preoperative use of epidural racemic ketamine, however, cannot be directly compared with our study. Choe et al. (8) found that patients receiving anesthetic-dose ketamine and morphine before the incision versus at the end of gastrectomy during general anesthesia had longer lasting postoperative pain control and reduced analgesic demand after 48 hours. Wong et al. (9) observed decreased opioid consumption and lower VAS pain scores over three days when patients received preoperative small-dose ketamine, morphine, and lidocaine compared with use at 30 minutes after the start of knee arthroplasty with epidural anesthesia or with general anesthesia and postincisional ketamine. Abdel-Ghaffar et al. (10) reported a prolonged time interval to first postoperative analgesic request and a reduced 24-hour analgesic demand when preoperative small-dose ketamine was used for hysterectomy with general anesthesia.

Activation of NMDA receptors within the spinal cord is thought to be central to the genesis of acute and chronic pain (2,3). Surgically induced tissue-damaging injury sustains altered processing of afferent noxious input and pain hypersensitivity acting in part through NMDA receptor activation (2,3). The most persuasive mechanism that underlies the analgesic action of subanesthetic-dose ketamine is related to NMDA-receptor blockade (4,5). Ketamine may inhibit a facilitated state of excitability by suppressing a progressive increase in nociceptive responses caused by wind-up phenomenon, long-term potentiation, and summation pain (2,12,13). Once pain sensitization from incisional and inflammatory injury is established, ketamine reduces NMDA receptor-related altered pain transmission and thus hyperalgesia (12,13,21). This explains in part why a study comparing preoperative epidural anesthetic-dose ketamine with the same dose applied before wound closure failed to show a difference in analgesia after abdominal surgery with general anesthesia (22). One further explanation for the increased pain relief in our study is the potential complementary antinociceptive action of the two drug classes used. Local anesthetics do not block all pain-sensitizing mechanisms (23,24), and intrathecal ketamine does not completely abolish hyperalgesia when used alone (21).

The difference in pain between groups did not emerge until 24 hours after surgery. This period may reflect the interval required for the sensitizing effects of surgery to become established, or it may reflect the overwhelming contribution of peripheral inputs from the operated knee, which obscured the spinal component of nociception in the earlier postsurgical period. Sustained injury causes several periods of enhanced spinal activity (2,3,21). In particular, there appears to be a crucial time interval during the first minutes to hours after surgery, and there is a second phase of inflammatory injury that depends on the first period (2,3,21). Clearly, the delayed onset of pain relief after S(+)-ketamine is not likely to be the result of direct analgesic effects, because it greatly exceeds the pharmacologic duration of a single epidural drug injection (19). Nevertheless, because we did not compare preincisional S(+)-ketamine with the administration of postincisional S(+)-ketamine, we cannot decisively state that preemptive analgesia was proved. We can only speculate that S(+)-ketamine blocked pain sensitization during the entire observation period but that its antinociceptive effects were on the one hand clouded by the effects of ropivacaine during the first postoperative hours, and on the other hand did not become clinically manifest before patient mobilization was started. In addition, there is evidence that the binding of noncompetitive NMDA receptor antagonists occurs relatively slowly, and this may contribute to the delayed analgesic effects we found (25). Conceptually, repeated use of S(+)-ketamine could enhance the pain reduction observed. This should be addressed in the future.

Use of S(+)-ketamine in our awake patients was not associated with an increase in adverse events or the occurrence of unpleasant psychological effects. This may be because the patients were premedicated with a benzodiazepine or because a small drug dose was applied.

The patients treated with S(+)-ketamine rated the quality of their pain therapy better. Patient perception of wellness should influence the choice of analgesics for postoperative pain management. In our opinion, increased patient satisfaction justifies the adjunctive use of small-dose S(+)-ketamine, given the ease of implementation and small cost of this new regimen.

In summary, we suggest that the use of a small dose of epidural S(+)-ketamine with a local anesthetic before surgical injury may decrease injury-induced pain sensitization and therefore provide better analgesia in the postoperative period than that achieved with a dose of local anesthetic alone in the epidural space.

We gratefully acknowledge the support of Ms. Ingrid Kammergruber from our departmental pain service. We wish to thank the nurses on the orthopedic wards for their help in the care of patients. Our thanks are also expressed to Dr. Sieglinde Reil-Danner, at the time the study was performed at Astra Inc., Wedel, Germany, for providing ropivacaine.

Back to Top | Article Outline


1. Capdevila X, Barthelet Y, Biboulet P, et al. Effects of perioperative analgesic technique on the surgical outcome and duration of rehabilitation after major knee surgery. Anesthesiology 1999; 91: 8–15.
2. Woolf C, Chong MS. Preemptive analgesia: treating postoperative pain by preventing the establishment of central sensitization. Anesth Analg 1993; 77: 362–79.
3. Yaksh TL, Hua XY, Kalcheva I, et al. The spinal biology in humans and animals of pain states generated by persistent afferent input. Proc Natl Acad Sci U S A 1999; 96: 7680–6.
4. Orser BA, Pennefather PS, MacDonald JF. Multiple mechanisms of ketamine blockade of N-methyl- d -aspartate receptors. Anesthesiology 1997; 86: 903–17.
5. Øye I, Paulsen O, Maurset A. Effects of ketamine on sensory perception: evidence for a role of N-methyl- d -aspartate receptors. J Pharmacol Exp Ther 1992; 260: 1209–13.
6. Aida S, Yamakura T, Baba H, et al. Preemptive analgesia by intravenous low-dose ketamine and epidural morphine in gastrectomy: a randomized double-blind study. Anesthesiology 2000; 92: 1624–30.
7. Suzuki M, Tsueda K, Lansing PS, et al. Small-dose ketamine enhances morphine-induced analgesia after outpatient surgery. Anesth Analg 1999; 89: 98–103.
8. Choe H, Choi YS, Kim YH, et al. Epidural morphine plus ketamine for upper abdominal surgery: improved analgesia from preincisional versus postincisional administration. Anesth Analg 1997; 84: 560–3.
9. Wong CS, Lu CC, Cherng CH, Ho ST. Pre-emptive analgesia with ketamine, morphine and epidural lidocaine prior to total knee replacement. Can J Anaesth 1997; 44: 31–7.
10. Abdel-Ghaffar ME, Abdulatif M, Al-Ghamdi A, et al. Epidural ketamine reduces postoperative epidural PCA consumption of fentanyl/bupivacaine. Can J Anaesth 1998; 45: 103–9.
11. Zeilhofer HU, Swandulla D, Geisslinger G, Brune K. Differential effects of ketamine enantiomers on NMDA receptor currents in cultured neurons. Eur J Pharmacol 1992; 213: 155–8.
12. Mathisen LC, Skjelbred P, Skoglund LA, Øye I. Effect of ketamine, an NMDA receptor inhibitor, in acute and chronic orofacial pain. Pain 1995; 61: 215–20.
13. Arendt-Nielsen L, Nielsen J, Petersen-Felix S, et al. Effect of racemic mixture and the (S+)-isomer of ketamine on temporal and spatial summation of pain. Br J Anaesth 1996; 77: 625–31.
14. Muldoon T, Milligan K, Quinn P, et al. Comparison between extradural infusion of ropivacaine or bupivacaine for the prevention of postoperative pain after total knee arthroplasty. Br J Anaesth 1998; 80: 680–1.
15. Karpinsky N, Dunn J, Hansen L, Masliah E. Subpial vacuolar myelopathy after intrathecal ketamine: report of a case. Pain 1997; 73: 103–5.
16. Errando CL, Sifre C, Moliner S, et al. Subarachnoid ketamine in swine: pathological findings after repeated doses—acute toxicity study. Reg Anesth Pain Med 1999; 24: 146–52.
17. 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.
18. Pedraz JL, Calvo MB, Gascon AR, et al. Pharmacokinetics and distribution of ketamine after extradural administration in dogs. Br J Anaesth 1991; 67: 310–6.
19. Pedraz JL, Lanao JM, Calvo MB, et al. Pharmacokinetic and clinical evaluation of ketamine administered by IV and epidural routes. Int J Clin Pharmacol Ther Toxicol 1987; 25: 77–80.
20. Weir PS, Fee JPH. Double-blind comparison of extradural block with three bupivacaine-ketamine mixtures in knee arthroplasty. Br J Anaesth 1998; 80: 299–301.
21. Hatrick CT, Wise JJ, Patterson JS. Preemptive intrathecal ketamine delays mechanical hyperalgesia in the neuropathic rat. Anesth Analg 1997; 86: 557–60.
22. Kucuk N, Kizilkaya M, Tokdemir M. Preoperative epidural ketamine does not have a postoperative opioid sparing effect. Anesth Analg 1998; 87: 103–6.
23. Gottrup H, Hansen PO, Arendt-Nielsen L, Jensen TS. Differential effects of systemically administered ketamine and lidocaine on dynamic and static hyperalgesia induced by intradermal capsaicin in humans. Br J Anaesth 2000; 84: 155–62.
24. Brennum J, Nielsen PT, Horn A, et al. Quantitative sensory examination in human epidural anaesthesia and analgesia in man: dose-response effect of bupivacaine. Pain 1994; 56: 315–26.
25. Bonhaus DW, McNamara JO. N-methyl- d -aspartate receptor regulation of uncompetitive antagonist binding in rat membranes: kinetic analysis. Mol Pharmacol 1988; 34: 250–6.
© 2001 International Anesthesia Research Society