- Email: [email protected]
Cyclooxygenase-2 inhibition potentiates morphine antinociception at the spinal level in a postoperative pain model
Original Articles
Cyclooxygenase-2 Inhibition Potentiates Morphine Antinociception at the Spinal Level in a Postoperative Pain Model Jeffrey S. Kroin, Ph.D., Asokumar Buvanendran, M.D., Robert J. McCarthy, Pharm.D., Hila Hemmati, M.D., and Kenneth J. Tuman, M.D. Background and Objectives: After peripheral inflammatory stimuli, spinal cord cyclooyxgenase–2 (COX-2) mRNA and protein levels increase, whereas COX-1 is unchanged. In animal models of inflammatory pain, intrathecal COX-2 selective inhibitors suppress hyperalgesia. However, the role of spinal COX-2 inhibition in postoperative pain is not well elucidated. This study investigates whether a water-soluble COX-2 selective inhibitor, L-745,337, can modify allodynic responses in a rat model of postoperative pain. Methods: Allodynia was induced in the left plantar hindpaw by surgical incision. Animals then received intrathecal (0-80 g) or subcutaneous (0-30 mg/kg) L-745,337 coadministered with intrathecal morphine (0-2 nmol). Reduction of mechanical allodynia (increased withdrawal threshold) was quantified with calibrated von Frey hairs. Results: L-745,337 alone, whether intrathecal or systemic, had no effect on withdrawal threshold. When intrathecal L-745,337 at doses of 40 to 80 g was combined with a subthreshold dose (0.5 nmol) of morphine, withdrawal thresholds were increased in a dose-dependent manner. Adding 80 g L-745,337 to 1 nmol morphine produced an antiallodynic effect greater than that of morphine at twice the dose. Subcutaneous L-745,337, up to 30 mg/kg combined with intrathecal morphine resulted in the same antiallodynic response as morphine alone. Conclusion: These results suggest a spinal interaction of COX-2 inhibition with opiate analgesia may allow a reduction of postoperative pain with lower doses of opiate. Reg Anesth Pain Med 2002;27:451-455. Key Words:
Cyclooxygenase, Intrathecal, Morphine, Postoperative pain, Rats.
A
t the spinal level, the 2 cyclooxygenase (COX) enzyme systems are expressed differently in response to peripheral inflammatory stimuli. After hindpaw injection of complete Freund’s adjuvant,1-3 COX-2 mRNA levels increase significantly in the lumbar spinal cord within a few hours, re-
From the Department of Anesthesiology, Rush Medical College at Rush-Presbyterian-St. Luke’s Medical Center, Chicago, IL. Accepted for publication May 15, 2002. Presented in part at the annual meeting of the International Anesthesiology Research Society, Fort Lauderdale, FL, March 17, 2001. Reprint requests: Jeffrey S. Kroin, Ph.D., Department of Anesthesiology, Rush Medical College, 1653 W. Congress Parkway, Chicago, IL 60612. E-mail: [email protected]. © 2002 by the American Society of Regional Anesthesia and Pain Medicine. 1098-7339/02/2705-0002$35.00/0 doi:10.1053/rapm.2002.35521
main elevated for 24 hours, and then normalize over the next few days. However, COX-1 mRNA levels remain unchanged.1,2 Similarly, peripheral injection of complete Freund’s adjuvant causes a transient large increase in COX-2 protein levels in the lumbar spinal cord,3,4 with no change in COX-1 protein levels.4 The injection of peripheral inflammatory agents also produces mechanical and thermal hyperalgesia in the hindpaws. Systemic administration of nonselective COX-1/COX-2 nonsteroidal anti-inflammatory drugs (NSAIDs), such as indomethecin, or COX-2 selective inhibitors, such as L-745,337, can block this hyperalgesia.5,6 Intrathecal COX-2 inhibitors also attenuate inflammation-induced thermal and mechanical hyperalgesia,3,6-8 but a COX-1 selective agent does not.6 All of the available evidence suggests a prominent role of COX-2 in the spinal response to peripheral inflammation.
Regional Anesthesia and Pain Medicine, Vol 27, No 5 (September–October), 2002: pp 451–455
451
452
Regional Anesthesia and Pain Medicine Vol. 27 No. 5 September–October 2002
The role of spinal COX-2 in the modulation of postoperative pain still needs to be defined. Using a rat foot-incision model,9 intrathecal COX-2 inhibitors administered alone had a minimal effect in attenuating mechanical hyperalgesia.10 However, there are clinical studies showing a morphine-sparing effect of NSAIDs in postoperative pain,11 and, more recently, COX-2 inhibitors administered before orthopedic surgery reduced postoperative opioid use.12 Therefore, the optimal strategy for using COX-2 inhibitors intrathecally to modulate postoperative pain may be in combination with opiates. The present study investigates whether a water soluble COX-2 inhibitor, L-745,337, administered intrathecally can potentiate morphine antinociception in an animal postoperative pain model.
Methods This study was approved by our Institutional Animal Care and Use Committee. Experiments were performed on male Sprague-Dawley rats (Sasco, Wilmington, MA) weighing 225 to 250 g. Under 2% halothane anesthesia, an intrathecal polyethylene catheter (0.61-mm outer diameter) was implanted through the cisterna magna and advanced 8.5 cm caudally with the tip at the lumbar enlargement.13,14 All skin margins were closed, leaving 3 cm of catheter above the skull for future injections. Any animal exhibiting neurological impairment after catheter placement was excluded from the study. Five days after intrathecal catheter implantation, an incision was made in the plantar surface of the left hindpaw, according to the method of Brennan et al.9 Briefly, under 2% halothane anesthesia, a 1-cm longitudinal incision was made with a number 11 scalpel blade through the skin and fascia starting from the edge of the heel and extending toward the toes. The plantaris muscle was also incised longitudinally, leaving intact the muscle origin and insertion. After applying gentle pressure for hemostasis, the skin was opposed with 2 mattress sutures of 5-0 nylon. On the next day, rats were placed over a metal grid, and mechanical allodynia was evaluated by using von Frey filaments (20-282 mN) applied to the hindpaw medial to the incision.9 Testing was performed in triplicate with 5 minutes separating each trial. Animals were then randomly assigned to receive an intrathecal bolus injection of morphine alone at 0, 0.5, 1.0, or 2.0 nmol; morphine 0.5 nmol plus L-745,337 at 0, 20, 40, 60, or 80 g; morphine 1.0 nmol plus L-745,337 80 g; or L-745,337 80 g alone. The volume of each intrathecal injection was 8 L, and an equal volume of saline was used to
flush the 7-L catheter dead space. Five other groups of animals received L-745,337 2 mg/kg (500 g) subcutaneously, followed 30 minutes later by intrathecal morphine at doses of 0, 0.5, or 1.0 nmol or L-745,337 30 mg/kg (7.5 mg) subcutaneously, followed 30 minutes later by intrathecal morphine at 0 or 0.5 nmol. Mechanical allodynia was again evaluated in triplicate at 55, 60, and 65 minutes after the last intrathecal injection. Rats were also evaluated for motor coordination on a rotating rod (3 minutes at 10 rpm) to test for motor side effects with any of the drug combinations. All personnel performing allodynia and coordination testing were blinded to the solutions administered. The withdrawal threshold was determined as the minimum force producing a response of the 3 trials. If there was no withdrawal response to the 282 mN filament, a cutoff value of 300 mN was recorded. Withdrawal thresholds pre- and postinjection were analyzed by the Wilcoxon signed rank test, whereas comparisons between different injection combinations were analyzed by using the Kruskal-Wallis H test followed by the Mann-Whitney U test, with P ⬍ .05 considered statistically significant. The specific drugs used in this study were morphine sulfate soluble 30 mg tablets (Eli Lilly, Indianapolis, IN) and L-745,337 (Merck Frosst Canada, Kirkland, Quebec). All drug dilutions were made with preservative-free 0.9% sodium chloride injection. The maximum solubility of L-745,337 at room temperature is 10 mg/mL.
Results All groups of animals exhibited reduced thresholds for foot withdrawal (median of 20 mN v 300 mN before surgery) at 24 hours after foot incision, with no difference between groups (Fig 1). An intrathecal injection of 80 g L-745,337 alone produced no change in withdrawal threshold. The 0.5nmol dose of intrathecal morphine alone also failed to change the withdrawal threshold. When L-745,337 was added to 0.5-nmol intrathecal morphine injection, there was no potentiation at the 20-g dose of the COX-2 inhibitor. However, combining L-745,337 at doses of 40, 60, and 80 g with 0.5 nmol morphine produced an increase in withdrawal threshold. The 40-g dose produced an effect greater than the 20-g dose but less than the 60- and 80-g doses. Compared with intrathecal morphine alone, the addition of 80 g L-745,337 consistently increased the withdrawal threshold (Fig 2). The withdrawal threshold of 1.0 nmol morphine with 80 g L-745,337 was greater than that of a 2.0-nmol morphine injection. The combination of 0.5 nmol mor-
Spinal COX-2 Inhibition Potentiates Morphine
Fig 1. Effect of different combinations of intrathecal morphine (M) and L-745,337 (L), either intrathecal (I.T.) or subcutaneous (S.C.), on withdrawal threshold as measured with von Frey filaments after plantar hindpaw incision. Subcutaneous injection performed 30 minutes before intrathecal injection. The results are expressed as medians (horizontal line) with first and third quartiles (boxes), and 10 and 90 percentiles (error bars). There were 10 to 15 animals per group. †Different from preinjection. *Different from postinjection 0.5 nmol morphine plus 20, 60, or 80 g L. ‡Different from postinjection 0.5 nmol morphine plus 20 or 40 g L; all P ⬍ .05.
phine with 80 g L-745,337 was not different than an injection of 1.0 to 2.0 nmol morphine alone. A subcutaneous injection of L-745,337 at 2 or 30 mg/kg did not change the withdrawal threshold nor did it potentiate intrathecal morphine at a dose of 0.5 nmol. Intrathecal morphine at 1.0 nmol did increase the withdrawal threshold (Fig 2), but the subcutaneous addition of 500 g L-745,337 did not potentiate this effect (not shown). No drug combination produced motor side effects.
•
Kroin et al.
453
ing water-insoluble compounds in solvents and injecting them intrathecally is that dosing may be inconsistent. This occurs because of precipitation of the compound as soon as the injectant mixes with cerebrospinal fluid. As a single agent, intrathecal L-745,337 had no effect in this postsurgical pain model, which is consistent with Yamamoto and Sakashita’s10 experience with another COX-2 inhibitor NS398. Similarly, COX-2 inhibitors administered intrathecally have no effect on thermal pain responses (hot plate or tail-flick test) in normal rats.3,17-19 This is different from inflammatory pain models in which intrathecal COX-2 inhibitors alone can block hyperalgesia.3,6-8 Although systemic L-734,337 at 30 mg/kg had no effect on postsurgical allodynia, oral administration effectively suppresses carrageenaninduced hindpaw hyperalgesia with an ID50 ⫽ 0.37 mg/kg.5 Intrathecal morphine alone is effective in reducing allodynia in the foot incision model.20 Combining morphine with the COX-2 inhibitor, L-745,337, allowed an equivalent antiallodynic effect at half the morphine dose (Fig 2). Therefore, intrathecal COX-2 inhibitors share the opioid-sparing properties usually associated with systemic NSAIDs. In this study, it is clear that the effect of intrathecal L-745,337 is occurring at the spinal level. Although 80 g of intrathecal L-745,337 potentiated intrathecal morphine, this did not occur with up to 30 mg/kg of this COX-2 inhibitor. Therefore, the COX-2 inhibitor is mediating morphine allodynia in the spinal cord and not by reuptake into the circulatory system followed by peripheral COX-2 inhibition. Clinical studies have shown that systemic COX
Discussion L-745,337 is a water-soluble COX-2 specific inhibitor that is 4 times more potent than indomethacin when given orally to reduce carrageenaninduced paw hyperalgesia.5 In several in vitro tests, L-745,337 was shown to have greater than a 100fold selectivity for COX-2 over COX-1.15 For intrathecal use, the water solubility of the compound gives it several advantages over other available COX-2 inhibitors. Typically, a vehicle containing dimethyl sulfoxide is used to solubilize the COX-2 inhibitor to create an injectable formulation. However, a vehicle containing as little as 10% dimethyl sulfoxide alone can affect allodynia,16 and although investigators typically run vehicle controls to compensate for such effects, it is not an ideal formulation. Another experimental problem with dissolv-
Fig 2. Effect of different doses of intrathecal morphine on withdrawal threshold after drug injection, with or without 80 g intrathecal L-745,337. †Different from morphine alone at the same dose. ‡Different from 2 nmol morphine. *Different from preinjection; all P ⬍ .05.
454
Regional Anesthesia and Pain Medicine Vol. 27 No. 5 September–October 2002
inhibitors when combined with epidural opioids improved postoperative pain relief21,22 or decreased postoperative epidural opioid usage.23 Besides potential species differences, the results of clinical studies may differ from our study because of a different route of drug administration (epidural versus intrathecal). Another important difference is that the clinical studies all involved cesarean delivery, with its inflammatory components, whereas the animal study uses an incisional wound pain model. In thermal withdrawal tests (tail-flick test), intrathecal COX-2 inhibitors or nonselective COX inhibitors did not potentiate intrathecal morphine when coadministered.18,19 Therefore, the efficacy of intrathecal COX-2 inhibition appears to be related to the degree of inflammation in the pain model: acute pain studies (no inflammation) show no analgesic effect of COX-2 inhibition alone and no morphine potentiation, postoperative pain models (some inflammatory component) show no analgesic effect of COX-2 inhibition alone but morphine potentiation, and inflammatory models show analgesic effect of COX-2 inhibition alone. The relative importance of COX-2 pathways in the induction and maintenance of postoperative pain is unknown. In preliminary experiments, L-745,337 80 g was injected intrathecally 10 minutes before the hindpaw incision, but there was no reduction in allodynia at 1, 2, or 24 hours afterwards.24 This is different from the hyperalgesia produced by the inflammatory agent carrageenan injected into the hindpaw, which can be blocked for about 3 hours if the COX-2 inhibitor SC58125 is given intrathecally 10 minutes before.8 An antagonist of the prostaglandin E2–type EP1 receptor injected into a hindpaw incision site can suppress allodynia consistent with a peripheral release of prostaglandins in the foot incision model.25 In addition, it has been recently reported that a similar prostaglandin antagonist given intrathecally attenuated allodynia in the same model.26 In inflammatory pain models, hyperalgesia can be similarly attenuated by intrathecal treatment with an inhibitor of central prostanoid production.3 Therefore, it is not evident why spinal L-745,337 alone does not reduce postoperative pain, although another COX-2 inhibitor, NS398, has also been shown ineffective intrathecally.10 The exact mechanism by which a COX-2 inhibitor can modulate -opioid activity in the spinal cord has not been determined. There is analgesic synergy between the NSAID ketorolac and morphine given intrathecally in the rat hindpaw formalin test.27 Interestingly in a neuropathic pain model, in which there is not a significant inflammatory component,
a COX-2 inhibitor added to spinal morphine potentiated its antiallodynic effect.28 Various models have been proposed of the spinal interaction of opiates and NSAIDs.27,29 In the dorsal horn, opioids act at the presynaptic terminal to inhibit release of excitatory neurotransmitters, and there are also prostanoid receptors in lamina 1 and 2, most likely on primary afferent terminals.30 However, it is also possible that the NSAID acts within the terminal to modulate arachidonic acid pathways involved in opiate activity.29 This occurs in midbrain periaqueductal grey neurons, although COX-1 inhibition rather than COX-2 appears to be more important in that region.31 Because preclinical toxicity studies have not been published for the intrathecal administration of any COX-2 inhibitor or any mixed NSAID, the safety of these compounds with or without opiates still must be determined. The water solubility of L-745,337 avoids the issue of vehicle toxicity, but the potential adverse effect of delivering relatively high doses of these enzyme inhibitors to the spinal subarachnoid space is relatively unknown. In conclusion, the intrathecal combination of a water-soluble COX-2 inhibitor with morphine decreases allodynia compared with morphine alone, whereas the systemic administration of COX-2 inhibitor was not effective. Our findings suggest the need for future studies to determine whether this opiate-sparing property may allow spinal morphine to be used postoperatively with fewer side effects.
References 1. Beiche F, Scheuerer S, Brune K, Geisslinger G, Goppelt-Struebe M. Up-regulation of cyclooxygenase-2 mRNA in the rat spinal cord following peripheral inflammation. FEBS Letters 1996;390:165-169. 2. Hay CH, Trevethick MA, Wheeldon A, Bowers JS, De Belleroche JS. The potential role of spinal cord cyclooxygenase-2 in the development of Freund’s complete adjuvant-induced changes in hyperalgesia and allodynia. Neuroscience 1997;78:843-850. 3. Samad TA, Moore KA, Sapirstein A, Billet S, Allchorne A, Poole S, Bonventure JV, Woolf CJ. Interleukin-1-mediated induction of COX-2 in the CNS contributes to inflammatory pain hypersensitivity. Nature 2001;410:471-475. 4. Gardiner NJ, Giblett S, Grubb BD. Cyclooxygenases in rat spinal cord: Selective induction of COX-2 during peripheral inflammation. Br J Pharm 1997;120: 71P. 5. Boyce S, Chan C-C, Gordon S, Li C-S, Rodger IW, Webb JK, Rupniak NMJ, Hill RG. L-745,337: A selective inhibitor of cyclooxygenase-2 elicits antinociception but not gastric ulceration in rats. Neuropharmacology 1994;33:1609-1611. 6. Yaksh, TL, Dirig DM, Conway CM, Svensson C, Luo
Spinal COX-2 Inhibition Potentiates Morphine
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
ZD, Isakson PC. The acute antihyperalgesic action of nonsteroidal, anti-inflammatory drugs and release of spinal prostaglandin E2 is mediated by the inhibition of constitutive spinal cyclooxygenase-2 (COX-2) but not COX-1. J Neurosci 2001;21:5847-5853. Yamamoto T, Nozaki-Taguchi N. Role of spinal cyclooxygenase (COX)-2 on thermal hyperalgesia evoked by carageenan injection in the rat. Neuroreport 1997; 8:2179-2182. Dirig DM, Isakson PC, Yaksh TL. Effect of COX-1 and COX-2 inhibition on induction and maintenance of carrageenan-evoked thermal hyperalgsia in rats. J Pharmacol Exp Ther 1998;285:1031-1038. Brennan TJ, Vandermeulen EP, Gebhart GF. Characterization of a rat model of incisional pain. Pain 1996; 64:493-501. Yamamoto T, Sakashita Y. The role of the spinal opioid receptor Like1 receptor, the NK-1 receptor, and cyclooxygenase-2 in maintaining postoperative pain in the rat. Anesth Analg 1999;89:1203-1208. Gillies GW, Kenny GN, Bullingham RE, McArdle CS. The morphine sparing effect of ketorolac tromethamine. A study of a new, parenteral non-steroidal anti-inflammatory agent after abdominal surgery. Anesthesia 1997;42:727-731. Reuben SS, Connelly NR. Postoperative analgesic effects of celecoxib or rofecoxib after spinal fusion surgery. Anesth Analg 2000;91:1221-1225. Stevens CW, Yaksh TL. Dynorphin A and related peptides administered intrathecally in the rat: a search for putative kappa opiate receptor activity. J Pharmacol Exp Ther 1986;238:833-838. Kroin JS, McCarthy RJ, Von Roenn N, Schwab B, Tuman KJ, Ivankovich AD. Magnesium sulfate potentiates morphine antinociception at the spinal level. Anesth Analg 2000;90:913-917. Warner TD, Giuliano F, Vojnovic I, Bukasa A, Mitchell JA, Vane JR. Nonsteroid drug selectivities for cyclo-oxygenase-1 rather than cyclo-oxygenase-2 are associated with human gastrointestinal toxicity: A full in vitro analysis. Proc Natl Acad Sci U S A 1999;96: 7563-7568. Zhao A, Chen SR, Eisenach JC, Busija DW, Pan HL. Spinal cyclooxygenase-2 is involved in development of allodynia after nerve injury in rats. Neuroscience 2000;97:743-748. Yamamoto T, Nozaki-Taguchi N. Analysis of the effects of cyclooxygenase (COX)-1 and COX-2 in spinal nociceptive transmission using indomethacin, a nonselective COX inhibitor, and NS-398, a COX-2 selective inhibitor. Brain Res 1996;739:104-110. Powell KJ, Hosokawa A, Bell A, Sutak M, Milne B, Quirion R, Jhamandas K. Comparative effects of cyclo-oxygenase and nitric oxide synthase inhibition on the development and reversal of spinal opioid tolerance. Br J Pharmacol 1999;127:631-644. Wong CS, Hsu MM, Chou R, Chou YY, Tung CS.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
•
Kroin et al.
455
Intrathecal cyclooxygenase inhibitor administration attenuates morphine antinociceptive tolerance in rats. Br J Anaesth 2000;85:747-751. Zahn PK, Gysbers D, Brennan TJ. Effect of systemic and intrathecal morphine in a rat model of postoperative pain. Anesthesiology 1997;86:1066-1077. Sun HL, Wu CC, Lin MS, Chang CF, Mok MS. Combination of low-dose epidural morphine and intramuscular diclofenac sodium in postcesarean analgesia. Anesth Analg 1992;75:64-68. Tzeng JI, Mok MS. Combination of intramuscular ketorolac and low dose epidural morphine for the relief of post-cesarean pain. Ann Acad Med Singapore 1994;23(suppl 6):10-13. Pavy TJG, Paech MJ, Evans SF. The effect of intravenous ketorolac on opioid rquirement and pain after cesarean delivery. Anesth Analg 2001;92:1010-1014. Buvanendran A, Kroin JS, Hemmati H, McCarthy RJ, Ivankovich AD. Co-administration of L-745,337, a cyclooxygenase-2 (COX-2) inhibitor, potentiates morphine antinociception at the spinal level in a model of postoperative pain. Anesth Analg 2001;92: S226. Omote K, Kawamata T, Nakayama Y, Kawamata M, Hazama, Namiki A. The effects of peripheral administration of a novel selective antagonist for prostaglandin E receptor subtype EP1, ONO-8711, in a rat model of postoperative pain. Anesth Analg 2001;92: 233-238. Yamamoto H, Omote K, Kawamata T, Nakayama Y, Namiki A. Antihyperalgesic effects of intrathecal antagonist for prostaglandin E receptor subtype EP1 in postoperative pain. Anesthesiology 2001;95:A866. Malmberg A, Yaksh TL. Pharmacology of the spinal action of ketorolac, morphine, ST-91, U50488H, and L-PIA on the formalin test and an isobolographic analysis of the NSAID interaction. Anesthesiology 1993;79:211-213. Lashbrook JM, Ossipov MH, Hunter JC, Raffa RB, Tallarida RJ, Porreca F. Synergistic antiallodynic effects of spinal morphine with ketorolac and selective COX1- and COX2-inhibitors in nerve-injured rats. Pain 1999;82:65-72. Christie MJ, Vaughan CW, Ingram SL. Opioids, NSAIDs and 5-lipoxygenase inhibitors act synergistically in brain via arachidonic acid metabolism. Inflamm Res 1999;48:1-4. Matsumura K, Wantanabe YU, Onoe H, Watanabe Y. Prostacyclin receptor in the brain and central terminals of the primary sensory neurons: An autogradiographic study using a stable prostacyclin analogue [3H]iloprost. Neuroscience 1995;65:493-503. Vaughan CW. Enhancement of opioid inhibition of GABAergic synaptic transmission by cyclo-oxygenase inhibitors in rat periaqueductal grey neurons. Br J Pharmacol 1998;123:1479-1481.
Cyclooxygenase-2 Inhibition Potentiates Morphine Antinociception at the Spinal Level in a Postoperative Pain Model Jeffrey S. Kroin, Ph.D., Asokumar Buvanendran, M.D., Robert J. McCarthy, Pharm.D., Hila Hemmati, M.D., and Kenneth J. Tuman, M.D. Background and Objectives: After peripheral inflammatory stimuli, spinal cord cyclooyxgenase–2 (COX-2) mRNA and protein levels increase, whereas COX-1 is unchanged. In animal models of inflammatory pain, intrathecal COX-2 selective inhibitors suppress hyperalgesia. However, the role of spinal COX-2 inhibition in postoperative pain is not well elucidated. This study investigates whether a water-soluble COX-2 selective inhibitor, L-745,337, can modify allodynic responses in a rat model of postoperative pain. Methods: Allodynia was induced in the left plantar hindpaw by surgical incision. Animals then received intrathecal (0-80 g) or subcutaneous (0-30 mg/kg) L-745,337 coadministered with intrathecal morphine (0-2 nmol). Reduction of mechanical allodynia (increased withdrawal threshold) was quantified with calibrated von Frey hairs. Results: L-745,337 alone, whether intrathecal or systemic, had no effect on withdrawal threshold. When intrathecal L-745,337 at doses of 40 to 80 g was combined with a subthreshold dose (0.5 nmol) of morphine, withdrawal thresholds were increased in a dose-dependent manner. Adding 80 g L-745,337 to 1 nmol morphine produced an antiallodynic effect greater than that of morphine at twice the dose. Subcutaneous L-745,337, up to 30 mg/kg combined with intrathecal morphine resulted in the same antiallodynic response as morphine alone. Conclusion: These results suggest a spinal interaction of COX-2 inhibition with opiate analgesia may allow a reduction of postoperative pain with lower doses of opiate. Reg Anesth Pain Med 2002;27:451-455. Key Words:
Cyclooxygenase, Intrathecal, Morphine, Postoperative pain, Rats.
A
t the spinal level, the 2 cyclooxygenase (COX) enzyme systems are expressed differently in response to peripheral inflammatory stimuli. After hindpaw injection of complete Freund’s adjuvant,1-3 COX-2 mRNA levels increase significantly in the lumbar spinal cord within a few hours, re-
From the Department of Anesthesiology, Rush Medical College at Rush-Presbyterian-St. Luke’s Medical Center, Chicago, IL. Accepted for publication May 15, 2002. Presented in part at the annual meeting of the International Anesthesiology Research Society, Fort Lauderdale, FL, March 17, 2001. Reprint requests: Jeffrey S. Kroin, Ph.D., Department of Anesthesiology, Rush Medical College, 1653 W. Congress Parkway, Chicago, IL 60612. E-mail: [email protected]. © 2002 by the American Society of Regional Anesthesia and Pain Medicine. 1098-7339/02/2705-0002$35.00/0 doi:10.1053/rapm.2002.35521
main elevated for 24 hours, and then normalize over the next few days. However, COX-1 mRNA levels remain unchanged.1,2 Similarly, peripheral injection of complete Freund’s adjuvant causes a transient large increase in COX-2 protein levels in the lumbar spinal cord,3,4 with no change in COX-1 protein levels.4 The injection of peripheral inflammatory agents also produces mechanical and thermal hyperalgesia in the hindpaws. Systemic administration of nonselective COX-1/COX-2 nonsteroidal anti-inflammatory drugs (NSAIDs), such as indomethecin, or COX-2 selective inhibitors, such as L-745,337, can block this hyperalgesia.5,6 Intrathecal COX-2 inhibitors also attenuate inflammation-induced thermal and mechanical hyperalgesia,3,6-8 but a COX-1 selective agent does not.6 All of the available evidence suggests a prominent role of COX-2 in the spinal response to peripheral inflammation.
Regional Anesthesia and Pain Medicine, Vol 27, No 5 (September–October), 2002: pp 451–455
451
452
Regional Anesthesia and Pain Medicine Vol. 27 No. 5 September–October 2002
The role of spinal COX-2 in the modulation of postoperative pain still needs to be defined. Using a rat foot-incision model,9 intrathecal COX-2 inhibitors administered alone had a minimal effect in attenuating mechanical hyperalgesia.10 However, there are clinical studies showing a morphine-sparing effect of NSAIDs in postoperative pain,11 and, more recently, COX-2 inhibitors administered before orthopedic surgery reduced postoperative opioid use.12 Therefore, the optimal strategy for using COX-2 inhibitors intrathecally to modulate postoperative pain may be in combination with opiates. The present study investigates whether a water soluble COX-2 inhibitor, L-745,337, administered intrathecally can potentiate morphine antinociception in an animal postoperative pain model.
Methods This study was approved by our Institutional Animal Care and Use Committee. Experiments were performed on male Sprague-Dawley rats (Sasco, Wilmington, MA) weighing 225 to 250 g. Under 2% halothane anesthesia, an intrathecal polyethylene catheter (0.61-mm outer diameter) was implanted through the cisterna magna and advanced 8.5 cm caudally with the tip at the lumbar enlargement.13,14 All skin margins were closed, leaving 3 cm of catheter above the skull for future injections. Any animal exhibiting neurological impairment after catheter placement was excluded from the study. Five days after intrathecal catheter implantation, an incision was made in the plantar surface of the left hindpaw, according to the method of Brennan et al.9 Briefly, under 2% halothane anesthesia, a 1-cm longitudinal incision was made with a number 11 scalpel blade through the skin and fascia starting from the edge of the heel and extending toward the toes. The plantaris muscle was also incised longitudinally, leaving intact the muscle origin and insertion. After applying gentle pressure for hemostasis, the skin was opposed with 2 mattress sutures of 5-0 nylon. On the next day, rats were placed over a metal grid, and mechanical allodynia was evaluated by using von Frey filaments (20-282 mN) applied to the hindpaw medial to the incision.9 Testing was performed in triplicate with 5 minutes separating each trial. Animals were then randomly assigned to receive an intrathecal bolus injection of morphine alone at 0, 0.5, 1.0, or 2.0 nmol; morphine 0.5 nmol plus L-745,337 at 0, 20, 40, 60, or 80 g; morphine 1.0 nmol plus L-745,337 80 g; or L-745,337 80 g alone. The volume of each intrathecal injection was 8 L, and an equal volume of saline was used to
flush the 7-L catheter dead space. Five other groups of animals received L-745,337 2 mg/kg (500 g) subcutaneously, followed 30 minutes later by intrathecal morphine at doses of 0, 0.5, or 1.0 nmol or L-745,337 30 mg/kg (7.5 mg) subcutaneously, followed 30 minutes later by intrathecal morphine at 0 or 0.5 nmol. Mechanical allodynia was again evaluated in triplicate at 55, 60, and 65 minutes after the last intrathecal injection. Rats were also evaluated for motor coordination on a rotating rod (3 minutes at 10 rpm) to test for motor side effects with any of the drug combinations. All personnel performing allodynia and coordination testing were blinded to the solutions administered. The withdrawal threshold was determined as the minimum force producing a response of the 3 trials. If there was no withdrawal response to the 282 mN filament, a cutoff value of 300 mN was recorded. Withdrawal thresholds pre- and postinjection were analyzed by the Wilcoxon signed rank test, whereas comparisons between different injection combinations were analyzed by using the Kruskal-Wallis H test followed by the Mann-Whitney U test, with P ⬍ .05 considered statistically significant. The specific drugs used in this study were morphine sulfate soluble 30 mg tablets (Eli Lilly, Indianapolis, IN) and L-745,337 (Merck Frosst Canada, Kirkland, Quebec). All drug dilutions were made with preservative-free 0.9% sodium chloride injection. The maximum solubility of L-745,337 at room temperature is 10 mg/mL.
Results All groups of animals exhibited reduced thresholds for foot withdrawal (median of 20 mN v 300 mN before surgery) at 24 hours after foot incision, with no difference between groups (Fig 1). An intrathecal injection of 80 g L-745,337 alone produced no change in withdrawal threshold. The 0.5nmol dose of intrathecal morphine alone also failed to change the withdrawal threshold. When L-745,337 was added to 0.5-nmol intrathecal morphine injection, there was no potentiation at the 20-g dose of the COX-2 inhibitor. However, combining L-745,337 at doses of 40, 60, and 80 g with 0.5 nmol morphine produced an increase in withdrawal threshold. The 40-g dose produced an effect greater than the 20-g dose but less than the 60- and 80-g doses. Compared with intrathecal morphine alone, the addition of 80 g L-745,337 consistently increased the withdrawal threshold (Fig 2). The withdrawal threshold of 1.0 nmol morphine with 80 g L-745,337 was greater than that of a 2.0-nmol morphine injection. The combination of 0.5 nmol mor-
Spinal COX-2 Inhibition Potentiates Morphine
Fig 1. Effect of different combinations of intrathecal morphine (M) and L-745,337 (L), either intrathecal (I.T.) or subcutaneous (S.C.), on withdrawal threshold as measured with von Frey filaments after plantar hindpaw incision. Subcutaneous injection performed 30 minutes before intrathecal injection. The results are expressed as medians (horizontal line) with first and third quartiles (boxes), and 10 and 90 percentiles (error bars). There were 10 to 15 animals per group. †Different from preinjection. *Different from postinjection 0.5 nmol morphine plus 20, 60, or 80 g L. ‡Different from postinjection 0.5 nmol morphine plus 20 or 40 g L; all P ⬍ .05.
phine with 80 g L-745,337 was not different than an injection of 1.0 to 2.0 nmol morphine alone. A subcutaneous injection of L-745,337 at 2 or 30 mg/kg did not change the withdrawal threshold nor did it potentiate intrathecal morphine at a dose of 0.5 nmol. Intrathecal morphine at 1.0 nmol did increase the withdrawal threshold (Fig 2), but the subcutaneous addition of 500 g L-745,337 did not potentiate this effect (not shown). No drug combination produced motor side effects.
•
Kroin et al.
453
ing water-insoluble compounds in solvents and injecting them intrathecally is that dosing may be inconsistent. This occurs because of precipitation of the compound as soon as the injectant mixes with cerebrospinal fluid. As a single agent, intrathecal L-745,337 had no effect in this postsurgical pain model, which is consistent with Yamamoto and Sakashita’s10 experience with another COX-2 inhibitor NS398. Similarly, COX-2 inhibitors administered intrathecally have no effect on thermal pain responses (hot plate or tail-flick test) in normal rats.3,17-19 This is different from inflammatory pain models in which intrathecal COX-2 inhibitors alone can block hyperalgesia.3,6-8 Although systemic L-734,337 at 30 mg/kg had no effect on postsurgical allodynia, oral administration effectively suppresses carrageenaninduced hindpaw hyperalgesia with an ID50 ⫽ 0.37 mg/kg.5 Intrathecal morphine alone is effective in reducing allodynia in the foot incision model.20 Combining morphine with the COX-2 inhibitor, L-745,337, allowed an equivalent antiallodynic effect at half the morphine dose (Fig 2). Therefore, intrathecal COX-2 inhibitors share the opioid-sparing properties usually associated with systemic NSAIDs. In this study, it is clear that the effect of intrathecal L-745,337 is occurring at the spinal level. Although 80 g of intrathecal L-745,337 potentiated intrathecal morphine, this did not occur with up to 30 mg/kg of this COX-2 inhibitor. Therefore, the COX-2 inhibitor is mediating morphine allodynia in the spinal cord and not by reuptake into the circulatory system followed by peripheral COX-2 inhibition. Clinical studies have shown that systemic COX
Discussion L-745,337 is a water-soluble COX-2 specific inhibitor that is 4 times more potent than indomethacin when given orally to reduce carrageenaninduced paw hyperalgesia.5 In several in vitro tests, L-745,337 was shown to have greater than a 100fold selectivity for COX-2 over COX-1.15 For intrathecal use, the water solubility of the compound gives it several advantages over other available COX-2 inhibitors. Typically, a vehicle containing dimethyl sulfoxide is used to solubilize the COX-2 inhibitor to create an injectable formulation. However, a vehicle containing as little as 10% dimethyl sulfoxide alone can affect allodynia,16 and although investigators typically run vehicle controls to compensate for such effects, it is not an ideal formulation. Another experimental problem with dissolv-
Fig 2. Effect of different doses of intrathecal morphine on withdrawal threshold after drug injection, with or without 80 g intrathecal L-745,337. †Different from morphine alone at the same dose. ‡Different from 2 nmol morphine. *Different from preinjection; all P ⬍ .05.
454
Regional Anesthesia and Pain Medicine Vol. 27 No. 5 September–October 2002
inhibitors when combined with epidural opioids improved postoperative pain relief21,22 or decreased postoperative epidural opioid usage.23 Besides potential species differences, the results of clinical studies may differ from our study because of a different route of drug administration (epidural versus intrathecal). Another important difference is that the clinical studies all involved cesarean delivery, with its inflammatory components, whereas the animal study uses an incisional wound pain model. In thermal withdrawal tests (tail-flick test), intrathecal COX-2 inhibitors or nonselective COX inhibitors did not potentiate intrathecal morphine when coadministered.18,19 Therefore, the efficacy of intrathecal COX-2 inhibition appears to be related to the degree of inflammation in the pain model: acute pain studies (no inflammation) show no analgesic effect of COX-2 inhibition alone and no morphine potentiation, postoperative pain models (some inflammatory component) show no analgesic effect of COX-2 inhibition alone but morphine potentiation, and inflammatory models show analgesic effect of COX-2 inhibition alone. The relative importance of COX-2 pathways in the induction and maintenance of postoperative pain is unknown. In preliminary experiments, L-745,337 80 g was injected intrathecally 10 minutes before the hindpaw incision, but there was no reduction in allodynia at 1, 2, or 24 hours afterwards.24 This is different from the hyperalgesia produced by the inflammatory agent carrageenan injected into the hindpaw, which can be blocked for about 3 hours if the COX-2 inhibitor SC58125 is given intrathecally 10 minutes before.8 An antagonist of the prostaglandin E2–type EP1 receptor injected into a hindpaw incision site can suppress allodynia consistent with a peripheral release of prostaglandins in the foot incision model.25 In addition, it has been recently reported that a similar prostaglandin antagonist given intrathecally attenuated allodynia in the same model.26 In inflammatory pain models, hyperalgesia can be similarly attenuated by intrathecal treatment with an inhibitor of central prostanoid production.3 Therefore, it is not evident why spinal L-745,337 alone does not reduce postoperative pain, although another COX-2 inhibitor, NS398, has also been shown ineffective intrathecally.10 The exact mechanism by which a COX-2 inhibitor can modulate -opioid activity in the spinal cord has not been determined. There is analgesic synergy between the NSAID ketorolac and morphine given intrathecally in the rat hindpaw formalin test.27 Interestingly in a neuropathic pain model, in which there is not a significant inflammatory component,
a COX-2 inhibitor added to spinal morphine potentiated its antiallodynic effect.28 Various models have been proposed of the spinal interaction of opiates and NSAIDs.27,29 In the dorsal horn, opioids act at the presynaptic terminal to inhibit release of excitatory neurotransmitters, and there are also prostanoid receptors in lamina 1 and 2, most likely on primary afferent terminals.30 However, it is also possible that the NSAID acts within the terminal to modulate arachidonic acid pathways involved in opiate activity.29 This occurs in midbrain periaqueductal grey neurons, although COX-1 inhibition rather than COX-2 appears to be more important in that region.31 Because preclinical toxicity studies have not been published for the intrathecal administration of any COX-2 inhibitor or any mixed NSAID, the safety of these compounds with or without opiates still must be determined. The water solubility of L-745,337 avoids the issue of vehicle toxicity, but the potential adverse effect of delivering relatively high doses of these enzyme inhibitors to the spinal subarachnoid space is relatively unknown. In conclusion, the intrathecal combination of a water-soluble COX-2 inhibitor with morphine decreases allodynia compared with morphine alone, whereas the systemic administration of COX-2 inhibitor was not effective. Our findings suggest the need for future studies to determine whether this opiate-sparing property may allow spinal morphine to be used postoperatively with fewer side effects.
References 1. Beiche F, Scheuerer S, Brune K, Geisslinger G, Goppelt-Struebe M. Up-regulation of cyclooxygenase-2 mRNA in the rat spinal cord following peripheral inflammation. FEBS Letters 1996;390:165-169. 2. Hay CH, Trevethick MA, Wheeldon A, Bowers JS, De Belleroche JS. The potential role of spinal cord cyclooxygenase-2 in the development of Freund’s complete adjuvant-induced changes in hyperalgesia and allodynia. Neuroscience 1997;78:843-850. 3. Samad TA, Moore KA, Sapirstein A, Billet S, Allchorne A, Poole S, Bonventure JV, Woolf CJ. Interleukin-1-mediated induction of COX-2 in the CNS contributes to inflammatory pain hypersensitivity. Nature 2001;410:471-475. 4. Gardiner NJ, Giblett S, Grubb BD. Cyclooxygenases in rat spinal cord: Selective induction of COX-2 during peripheral inflammation. Br J Pharm 1997;120: 71P. 5. Boyce S, Chan C-C, Gordon S, Li C-S, Rodger IW, Webb JK, Rupniak NMJ, Hill RG. L-745,337: A selective inhibitor of cyclooxygenase-2 elicits antinociception but not gastric ulceration in rats. Neuropharmacology 1994;33:1609-1611. 6. Yaksh, TL, Dirig DM, Conway CM, Svensson C, Luo
Spinal COX-2 Inhibition Potentiates Morphine
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
ZD, Isakson PC. The acute antihyperalgesic action of nonsteroidal, anti-inflammatory drugs and release of spinal prostaglandin E2 is mediated by the inhibition of constitutive spinal cyclooxygenase-2 (COX-2) but not COX-1. J Neurosci 2001;21:5847-5853. Yamamoto T, Nozaki-Taguchi N. Role of spinal cyclooxygenase (COX)-2 on thermal hyperalgesia evoked by carageenan injection in the rat. Neuroreport 1997; 8:2179-2182. Dirig DM, Isakson PC, Yaksh TL. Effect of COX-1 and COX-2 inhibition on induction and maintenance of carrageenan-evoked thermal hyperalgsia in rats. J Pharmacol Exp Ther 1998;285:1031-1038. Brennan TJ, Vandermeulen EP, Gebhart GF. Characterization of a rat model of incisional pain. Pain 1996; 64:493-501. Yamamoto T, Sakashita Y. The role of the spinal opioid receptor Like1 receptor, the NK-1 receptor, and cyclooxygenase-2 in maintaining postoperative pain in the rat. Anesth Analg 1999;89:1203-1208. Gillies GW, Kenny GN, Bullingham RE, McArdle CS. The morphine sparing effect of ketorolac tromethamine. A study of a new, parenteral non-steroidal anti-inflammatory agent after abdominal surgery. Anesthesia 1997;42:727-731. Reuben SS, Connelly NR. Postoperative analgesic effects of celecoxib or rofecoxib after spinal fusion surgery. Anesth Analg 2000;91:1221-1225. Stevens CW, Yaksh TL. Dynorphin A and related peptides administered intrathecally in the rat: a search for putative kappa opiate receptor activity. J Pharmacol Exp Ther 1986;238:833-838. Kroin JS, McCarthy RJ, Von Roenn N, Schwab B, Tuman KJ, Ivankovich AD. Magnesium sulfate potentiates morphine antinociception at the spinal level. Anesth Analg 2000;90:913-917. Warner TD, Giuliano F, Vojnovic I, Bukasa A, Mitchell JA, Vane JR. Nonsteroid drug selectivities for cyclo-oxygenase-1 rather than cyclo-oxygenase-2 are associated with human gastrointestinal toxicity: A full in vitro analysis. Proc Natl Acad Sci U S A 1999;96: 7563-7568. Zhao A, Chen SR, Eisenach JC, Busija DW, Pan HL. Spinal cyclooxygenase-2 is involved in development of allodynia after nerve injury in rats. Neuroscience 2000;97:743-748. Yamamoto T, Nozaki-Taguchi N. Analysis of the effects of cyclooxygenase (COX)-1 and COX-2 in spinal nociceptive transmission using indomethacin, a nonselective COX inhibitor, and NS-398, a COX-2 selective inhibitor. Brain Res 1996;739:104-110. Powell KJ, Hosokawa A, Bell A, Sutak M, Milne B, Quirion R, Jhamandas K. Comparative effects of cyclo-oxygenase and nitric oxide synthase inhibition on the development and reversal of spinal opioid tolerance. Br J Pharmacol 1999;127:631-644. Wong CS, Hsu MM, Chou R, Chou YY, Tung CS.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
•
Kroin et al.
455
Intrathecal cyclooxygenase inhibitor administration attenuates morphine antinociceptive tolerance in rats. Br J Anaesth 2000;85:747-751. Zahn PK, Gysbers D, Brennan TJ. Effect of systemic and intrathecal morphine in a rat model of postoperative pain. Anesthesiology 1997;86:1066-1077. Sun HL, Wu CC, Lin MS, Chang CF, Mok MS. Combination of low-dose epidural morphine and intramuscular diclofenac sodium in postcesarean analgesia. Anesth Analg 1992;75:64-68. Tzeng JI, Mok MS. Combination of intramuscular ketorolac and low dose epidural morphine for the relief of post-cesarean pain. Ann Acad Med Singapore 1994;23(suppl 6):10-13. Pavy TJG, Paech MJ, Evans SF. The effect of intravenous ketorolac on opioid rquirement and pain after cesarean delivery. Anesth Analg 2001;92:1010-1014. Buvanendran A, Kroin JS, Hemmati H, McCarthy RJ, Ivankovich AD. Co-administration of L-745,337, a cyclooxygenase-2 (COX-2) inhibitor, potentiates morphine antinociception at the spinal level in a model of postoperative pain. Anesth Analg 2001;92: S226. Omote K, Kawamata T, Nakayama Y, Kawamata M, Hazama, Namiki A. The effects of peripheral administration of a novel selective antagonist for prostaglandin E receptor subtype EP1, ONO-8711, in a rat model of postoperative pain. Anesth Analg 2001;92: 233-238. Yamamoto H, Omote K, Kawamata T, Nakayama Y, Namiki A. Antihyperalgesic effects of intrathecal antagonist for prostaglandin E receptor subtype EP1 in postoperative pain. Anesthesiology 2001;95:A866. Malmberg A, Yaksh TL. Pharmacology of the spinal action of ketorolac, morphine, ST-91, U50488H, and L-PIA on the formalin test and an isobolographic analysis of the NSAID interaction. Anesthesiology 1993;79:211-213. Lashbrook JM, Ossipov MH, Hunter JC, Raffa RB, Tallarida RJ, Porreca F. Synergistic antiallodynic effects of spinal morphine with ketorolac and selective COX1- and COX2-inhibitors in nerve-injured rats. Pain 1999;82:65-72. Christie MJ, Vaughan CW, Ingram SL. Opioids, NSAIDs and 5-lipoxygenase inhibitors act synergistically in brain via arachidonic acid metabolism. Inflamm Res 1999;48:1-4. Matsumura K, Wantanabe YU, Onoe H, Watanabe Y. Prostacyclin receptor in the brain and central terminals of the primary sensory neurons: An autogradiographic study using a stable prostacyclin analogue [3H]iloprost. Neuroscience 1995;65:493-503. Vaughan CW. Enhancement of opioid inhibition of GABAergic synaptic transmission by cyclo-oxygenase inhibitors in rat periaqueductal grey neurons. Br J Pharmacol 1998;123:1479-1481.