| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the *University of Maryland, School of Dentistry, Baltimore, Maryland; Practice of Endodontics, Boston, Massachusetts; National Institute of Nursing Research; Department of Nursing, Clinical Research Center; ||National Institute of Dental and Craniofacial Research, National Institutes of Health, Bethesda, Maryland; and ¶Department of Pharmacology, Faculty of Medicine, Ain Shams University, Cairo Egypt.
Address correspondence and reprint requests to Dr. Sharon M. Gordon, University of MD, School of Dentistry, Baltimore, MD. Address e-mail to SGordon{at}umaryland.edu.
Abstract
BACKGROUND: In addition to blocking nociceptive input from surgical sites, long-acting local anesthetics might directly modulate inflammation. In the present study, we describe the proinflammatory effects of bupivacaine on local prostaglandin E2 (PGE2) production and cyclooxygenase (COX) gene expression that increases postoperative pain in human subjects.
METHODS: Subjects (n = 114) undergoing extraction of impacted third molars received either 2% lidocaine or 0.5% bupivacaine before surgery and either rofecoxib 50 mg or placebo orally 90 min before surgery and for the following 48 h. Oral mucosal biopsies were taken before surgery and 48 h after surgery. After extraction, a microdialysis probe was placed at the surgical site for PGE2 and thromboxane B2 (TXB2) measurements.
RESULTS: The bupivacaine/rofecoxib group reported significantly less pain, as assessed by a visual analog scale, compared with the other three treatment groups over the first 4 h. However, the bupivacaine/placebo group reported significantly more pain at 24 h and PGE2 levels during the first 4 h were significantly higher than the other three treatment groups. Moreover, bupivacaine significantly increased COX-2 gene expression at 48 h as compared with the lidocaine/placebo group. Thromboxane levels were not significantly affected by any of the treatments, indicating that the effects seen were attributable to inhibition of COX-2, but not COX-1.
CONCLUSIONS: These results suggest that bupivacaine stimulates COX-2 gene expression after tissue injury, which is associated with higher PGE2 production and pain after the local anesthetic effect dissipates.
Tissue injury associated with surgical trauma directly and indirectly leads to the activation of nociceptors,1 with increased expression of proinflammatory cytokines and induction of cyclooxygenase-2 (COX-2),2 leading to peripheral and central sensitization with subsequent hyperalgesia.3 Prostaglandin E2 (PGE2) is an abundant eicosanoid released after surgical trauma and has been associated with inflammation and pain.4 We have previously shown that increases in PGE2 at the surgical site are associated with the onset of pain and inflammation in the third molar extraction model.5–7 Buvanendran et al.8 have shown an increase of the PGE2 concentration in cerebrospinal fluid (CSF) after hip arthroplasty in humans. The increase in PGE2 levels is associated with up-regulation of COX-2 both at the injury site9,10 and in the spinal cord.9,11,12 This widespread induction of COX-2 indicates the need to intervene at central as well as peripheral targets to suppress synthesis of proinflammatory prostanoids contributing to pain and inflammation.
Long-acting local anesthetics (LAs) such as bupivacaine and ropivacaine are used to provide prolonged perioperative pain relief and to attenuate the development of postoperative sensitization that manifests as hyperalgesia after the anesthetic effect has dissipated, thereby reducing postoperative pain and morbidity and accelerating recovery.13–16 Attenuation of postoperative pain beyond the duration of the commonly used 2% lidocaine with 1:100,000 epinephrine has been demonstrated for both etidocaine17 and bupivacaine13,16,18 and is additive with the effects of a nonsteroidal antiinflammatory drug (NSAID) administered preemptively before a minor surgical procedure.19 Treatment aimed at reducing sensory inflow into the central nervous system (CNS), such as regional local anesthesia during surgery, may be enhanced by COX-2 inhibition.
Although LAs are used for their conduction-blocking activity, they have actions on other cellular targets that could modulate inflammation.20 There are accumulating data suggesting an antiinflammatory action of LAs. For example, prolonged sciatic nerve block by bupivacaine limits edema and related pain after carrageenan-induced inflammation in the rat.21 Moreover, bupivacaine regulates the systemic inflammatory response elicited by carrageenan in an ex vivo murine model of local inflammation.22 Intratracheally instilled ropivacaine, in both low and clinically relevant doses, shows strong antiinflammatory effects on neutrophils, endothelial and epithelial lung cells in vitro and in vivo. Pham-Marcou et al.23 report that prolonged nerve block by 70% ethanol in mice decreased not only the local inflammatory reaction observed after carrageenan injection, but also its systemic consequences.
The present investigation studied the effect of two commonly used LAs (lidocaine and bupivacaine) on local COX-2 gene expression, PGE2 production and postoperative pain, and their interactions with a selective COX-2 inhibitor (rofecoxib) in the oral surgery model.
METHODS
The study design was a double-blind clinical trial with randomized parallel groups of subjects undergoing the surgical removal of impacted third molars, a clinical model of tissue injury. Subjects were outpatients referred for the extraction of impacted third molars. The surgical and experimental procedures were explained verbally and in writing, and written informed consent was obtained in accordance with, and with permission from, the National Institute of Dental and Craniofacial Research IRB. Subjects were men or women between the ages of 16 to 35 yr in general good health (ASA status I or II) with the clinical indication for third molar removal. Surgical criteria included the presence of two partial or fully impacted mandibular third molars in order to evaluate subjects experiencing similar pain levels.
Exclusion criteria included allergy to aspirin, NSAIDs, sulfites, or amide anesthetics; a history of peptic ulcers or gastrointestinal bleeding; and chronic use of medications confounding the assessment of the inflammatory response or analgesia, e.g., NSAIDs, COX-2 inhibitors, antihistamines, steroids, and antidepressants. Patients were also excluded for clinical signs suggestive of infection, inflammation, or preexisting pain or if unusual surgical difficulty occurred during the actual surgery. Pregnant or lactating females were ineligible.
Subjects were randomly assigned to either rofecoxib 50 mg or placebo orally, and local anesthesia 2% lidocaine with 1:200,000 epinephrine (Astra-Zeneca Pharmaceuticals, Wilmington, DE) or 0.5% bupivacaine with 1:200,000 epinephrine (Abbott Laboratories, Abbot Park, IL) in a factorial design. The first dose of medication (rofecoxib or placebo) was administered in the clinic 90 min before the surgery. All subjects were titrated immediately before surgery with IV administered midazolam, followed by LA administration. Subjects were tested for subjective signs of lower lip anesthesia 5 min postinjection and reinjected, if necessary, to achieve adequate local anesthesia. After demonstration of satisfactory local anesthesia, a 3-mm diameter tissue biopsy was taken from the area adjacent to one of the mandibular third molars. Two mandibular third molars were extracted and a surgical difficulty score was assigned for each tooth.
After extraction, a microdialysis probe (CMA/20 Microdialysis Probe; CMA/Microdialysis, North Chelmsford, MA) was placed along the buccal aspect of the mandible, beneath the mucogingival flap elevated for the surgical procedure. The probe fiber consists of a 10-mm flexible, nonmetallic, semipermeable dialysis membrane with a molecular cutoff ranging from 3000 to 20,000 Da. The probes were secured to an adjacent tooth with silk suture and the flap closed in the usual fashion using 3-0 chromic gut suture. Sterile lactated Ringer’s solution was pumped at 10 µL/min and samples collected at 20-min intervals after the completion of surgery, before pain onset. Subjects remained under observation for the first 4 h after surgery to evaluate pain and adverse events and collect samples by microdialysis. During the immediate postoperative period, subjects were allowed rescue medication after 1 h of evaluation with 975 mg of acetaminophen. Up to 60 mg of codeine could be requested as a second rescue drug if pain was unrelieved by acetaminophen. Patients rated their pain intensity every 20 min for the first 4 postoperative h and at 24 and 48 h using 100-mm visual analog scale. The subsequent doses of the study drug (rofecoxib or placebo) were self-administered by the patient at home after discharge.
At the conclusion of the observation period, the microdialysis probes were removed and the patients were discharged in the care of a responsible adult with two separate bottles of rescue pain medications: acetaminophen and codeine, with instructions for administration and how to record drug intake and pain in a diary with instructions to first take 975 mg of acetaminophen for pain unrelieved by the study medication and to use codeine 30 to 60 mg if pain was still unrelieved an hour after taking the acetaminophen. Patients recorded pain intensity at 24 and 48 h and were telephoned at 24 h to remind them to complete the pain questionnaires and to take the pain medication. At 48 h, patients returned for a second tissue biopsy along with completed pain and analgesic diaries.
Microdialysis samples were placed on dry ice after the collection period, and stored at –70°C until assayed by enzyme immunoassay using commercially available enzyme immunoassay kits for immunoreactive PGE2 or immunoreactive thromboxane B2 (TXB2) (Cayman Chemical Company, Ann Arbor, MI) following the manufacturer’s recommended methods.
Oral mucosal biopsies (n = 60) were used to detect gene expression using ABI Prism 7900 HT Sequence Detection System (Applied Biosystems, Foster City, CA). All reagents were purchased from Applied Biosystems and 2 µg of DNase-treated RNA was used to synthesize cDNA using random primers from the High-Capacity cDNA Archive Kit according to the manufacturer’s instruction. Polymerase chain reaction was performed with cDNA template using the PCR Master Mix with AmpErase UNG. Sequence–specific primers and TaqMan MGB probes were purchased from Assays-on-Demand Gene expression product. Quantification of gene expression was performed in a 20-µL reaction (384-well plate) and each sample was run in triplicate. The housekeeping gene 18S rRNA was used as endogenous control and negative controls were processed under the same conditions without a cDNA template. Data acquisition was conducted using User Bulletin #2 software (v1.6, Applied Biosystems). The threshold cycle (Ct) of 18 rRNA was used to normalize target gene expression (
Ct) to correct for experimental variations. The relative change in gene expression (Ct) was used for comparison of the gene expression in postsurgery tissue versus that in presurgery tissue using paired t-test. Two-sided 
level was set P < 0.05. Data were analyzed using SPSS v. 13.0 (SPSS, Chicago, IL). Unless otherwise specified, statistical significance was evaluated by means of one-way ANOVA followed by Newman–Keuls post hoc test with significance level of P < 0.05.
RESULTS
Of 136 patients enrolled, 22 failed to complete the data collection period, leaving 114 subjects remaining in the study to completion. Reasons for elimination included: failing to complete pain questionnaires at later time points, requiring a NSAID rescue analgesic (ketorolac tromethamine) to manage pain during the immediate postoperative period, or taking other analgesics aside from what was provided for rescue. The demographic and surgical variables of the subjects are summarized in Table 1, demonstrating that the four groups were similar in age, height, weight, surgical difficulty and the amount of LA and midazolam received.
|
Comparisons of pain intensity during the immediate postoperative period (Fig. 1) illustrate a gradual increase in pain intensity as the effect of the LA dissipated. The bupivacaine/rofecoxib group reported significantly less pain (P < 0.05) over the 4-h observation period in comparison to the other three treatment groups that did not differ significantly from each other. However, the pain suppression for the bupivacaine/rofecoxib group did not differ from that of the lidocaine/rofecoxib group at 24 or 48 h (Fig. 2). Moreover, there was significantly more pain reported by the bupivacaine/placebo group compared with the other three groups at the 24 h time point (P < 0.05). Although not statistically significant at the 48 h time point, the bupivacaine/placebo group reported higher pain intensity. As expected, acetaminophen intake was significantly lower in both rofecoxib groups, representing a difference in pain experienced between groups, but there was no significant difference in codeine intake among the four groups (Table 2).
|
|
|
PGE2 concentrations collected by microdialysis from the surgical site over the first 4 h postoperatively showed a gradual decrease over the first 90 min followed by increasing levels over the remainder of the observation period for the groups receiving placebo (Fig. 3a), with the bupivacaine/placebo group exhibiting a significant increase in PGE2 over the 4-h observation period (P < 0.05) as shown in Figure 3b. TXB2 concentrations, an indicator of COX-1 activity, were variable over time, but not different between groups during the 4-h observation period (Fig. 4).
|
|
Changes in gene expression of COX1/COX2 after bupivacaine treatment are shown in (Fig. 5). Compared with presurgical levels, at 48 h COX-1 gene expression was increased twofold (2.15 ± 0.44, P < 0.05, paired t-test) as was COX-2 (3.21 ± 0.79, P < 0.05). Bupivacaine induced a significant increase in COX-2 gene expression compared with presurgical tissue levels and to that of the lidocaine group (P < 0.05, Student’s t-test).
|
DISCUSSION
Bupivacaine up-regulated COX-2 gene expression and increased PGE2 production at the surgical site, accompanied by higher pain intensity at time points after the effects of the LA dissipated (24 and 48 h). This effect was ameliorated by rofecoxib for both LAs at later time points. The higher pain intensity in the bupivacaine group compared with the lidocaine group at 24 and 48 h contrasts with the results of our previous studies evaluating preemptive analgesic strategies,13,16 where administration of bupivacaine in the perioperative period resulted in less pain at 48 h that was attributed to suppression of postoperative nociceptive input. However, both previous studies were performed under general anesthesia using propofol and nitrous oxide. Many studies indicate that nitrous oxide inhibits excitatory N-methyl-d-aspartate glutamate transmission.24,25N-methyl-d-aspartate receptors are intimately involved in the development and maintenance of central sensitization.26,27 It is also suggested that a significant portion of the hypnotic action of propofol is mediated by enhancing the 
-aminobutyric acid (GABA)-induced chloride current through binding to the β-subunit of the GABAA receptor. Suppression of central inhibitory mechanisms mediated via GABAA receptors is also involved in central sensitization.28 Therefore, in the presence of nitrous oxide and propofol, the development of central sensitization would be greatly impaired, and this may have contributed to the results obtained in those studies. In addition, spinal propofol is reported to decrease PGE2 concentrations in the CSF of rats undergoing surgery or sham operation.12 The suppression of central inflammatory responses by general anesthesia, as described earlier8 is an explanation for failure of thoracoabdominal surgery to increase CSF levels of interleukin-1β, tumor necrosis factor-, or interleukin-6 in humans.29
The actions of LAs on inflammation are complex and reflect interplay of mechanisms at many levels. Although LAs block the conduction of nociceptive input, the induction of cytokines is not necessarily inhibited. For example, bupivacaine and tetrodotoxin did not diminish cytokine induction in carrageenan-induced hindpaw inflammation in rats,30 and sciatic nerve blockade in animals with hindpaw inflammation did not prevent increases in central COX-2 or PGE2 levels as effectively as administration of a cytokine inhibitor.9 Intraplantar infiltration with lidocaine and bupivacaine before carageenan transiently limited signs of inflammatory pain, but did not prevent it.31 Similarly, procaine failed to inhibit PGE2 release in mice spinal cord after formalin injection (Hamza M, unpublished observations). In agreement with the findings of the present study, Kroin et al.12 found an increase in pain behavior and secretion of PGE2 in the CSF of rats in response to spinal bupivacaine that was reduced by oral coadministration of rofecoxib. The 50-mg dose of rofecoxib has been shown (along with other coxibs) to rapidly penetrate the CNS as measured by CSF levels in healthy volunteers32,33 and to be 15% the level of plasma concentrations at the same dose.33 Hence, coxibs may act, in part, at the level of the human CNS.
Local tissue irritation resulting in an inflammatory infiltrate, and consequently pain, during or after anesthesia has been observed for some LAs.34–36 Histological studies indicate that bupivacaine produces the most and lidocaine the least inflammation in rat subcutaneous tissue.37 Pain at the injection site is also reportedly higher for bupivacaine. Morris et al.38 compared pain of injection from five different LAs and found that bupivacaine induced more pain than lidocaine. Hence, local tissue irritation resulting in cytokine release might explain the higher pain intensity we observed in the bupivacaine only group at 24 and 48 h.
It is important to note that the pH of the bupivacaine solution administered in the present study was lower (3.7) than that of lidocaine solution (4.2) and may have caused direct irritation at the site of administration. Buffering of LAs reduces the pain of injection and may attenuate tissue irritation.39 In the present study, it was not possible to include an unanesthetized control group. Thus, it is not clear if the lower PGE2 production and COX-2 expression in the lidocaine group was a proinflammatory effect for bupivacaine or an antiinflammatory effect of lidocaine. Intrathecal lidocaine has been reported to inhibit spinal prostaglandin production,40 and to inhibit the synthesis of leukotriene B4, PGE1, PGE2, and TXB2 when applied topically in combination with prilocaine for burn injury in the rat.41 In contrast, bupivacaine reportedly upregulates CSF PGE2 after spinal anesthesia.12 Lidocaine is an inhibitor of phospholipase A2 (PLA2) activity,42,43 which is a rate-limiting enzyme for the release of arachidonic acid from membrane phospholipids in response to inflammation.9 Therefore, inhibiting PLA2 may have contributed to lower PGE2 production in the lidocaine-treated group. An interesting finding in the present study is the up-regulation of COX-1 expression at 48 h, although TXB2 did not differ in the immediate postoperative period. In mice with collagen-induced arthritis, COX-1 expression was increased in both inflamed skin and spinal cord.44 COX-1 is also suggested to play an important role in spinal cord pain processing and sensitization in postoperative pain,45,46 and specific COX-1 inhibitors may be useful to treat postoperative pain.45 Furthermore, it was reported that COX-1 immunoreactivity is increased in the spinal cord after nerve injury.47
In the present study, the combination of rofecoxib with bupivacaine suppressed the onset and intensity of postoperative pain in the immediate postoperative period. Bupivacaine alone elicited more pain at later time points, significantly more at 24 h which was ameliorated by rofecoxib. Thus, the addition of rofecoxib offset the proinflammatory effect from bupivacaine at later time points. Rofecoxib was used in this study as a selective inhibitor of COX-2. Although rofecoxib is no longer on the market, based on the literature it is likely that COX-2 inhibition through other COX-2 inhibitors, or through dual-COX-1/2 activity would result in similar effects.
In summary, the results of the present study suggest that in the absence of general anesthesia, bupivacaine injected locally induces a proinflammatory effect, manifesting as up-regulation of COX-2 expression and PGE2 production, and increased pain intensity after the cessation of the anesthetic effect. These findings may have relevance to the clinical practice of using the long-acting LA bupivacaine to suppress the development of sensitization if our results are validated by further clinical studies.
ACKNOWLEDGMENTS
The authors thank Drs. George Grimes and Judy Starling of the NIH Pharmaceutical Development Service for contribution to the testing and blinding of the investigational agents used.
Footnotes
Accepted for publication September 10, 2007.
Supported by the Division of Intramural Research, NIH.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
