Pain is commonly treated with opioids and non-opioid analgesics, e.g., nonselective nonsteroidal antiinflammatory drugs (NSAIDs). Opioids provide good analgesic effect, but their use is limited because of side effects and potential addiction (1). Although nonselective NSAIDs can provide analgesic efficacy in many clinical situations (2,3), their use is not appropriate in presurgical, perisurgical, or postsurgical settings, as they can increase the incidence of bleeding complications. Nonselective NSAIDs inhibit prostaglandin synthesis by inhibiting both forms of the cyclooxygenase (COX) enzyme. The inhibition of COX-2 is primarily responsible for the therapeutic effects of NSAIDs by diminishing inflammation, pain, and fever whereas inhibition of COX-1, which has an important role in gastric mucosal protection, can lead to detrimental gastrointestinal effects (4,5) and inhibition of platelet aggregation (6). COX-2 selective inhibitors, conversely, do not inhibit the COX-1 isoform in the clinical therapeutic dose range, thus not affecting platelet aggregation. COX-2 selective inhibitors, such as rofecoxib, celecoxib, and valdecoxib, have demonstrated efficacy and safety similar to that of nonselective NSAIDs in patients with osteoarthritis and rheumatoid arthritis (7–9). These drugs have also demonstrated efficacy in acute pain, including postorthopedic surgery pain (10).
Etoricoxib, a new COX-2 selective inhibitor, has demonstrated efficacy and safety in osteoarthritis, rheumatoid arthritis, acute gouty arthritis, dental surgery pain, and primary dysmenorrhea (11–15). The present study assessed the analgesic effect of etoricoxib for short-term use in patients suffering from pain after orthopedic (total knee or total hip) surgery.
This was a randomized, double-blind, placebo-controlled, parallel-group, two-part study conducted at eight clinical centers in the United States. Local IRBs reviewed and approved the protocol and informed consent. All patients gave written informed consent before any study procedure was performed. Men and women 18 yr of age or older scheduled to have either total hip or total knee replacement (or revisions) were eligible for participation. Women were postmenopausal, surgically sterilized, or agreed to practice birth control (abstinence, IM, implanted contraceptives, or double-barrier contraception). Patients were not eligible if they had uncontrolled hypertension; angina or congestive heart failure, recent history of myocardial infarction, coronary angioplasty, coronary arterial bypass, stroke or transient ischemic attack, uncontrolled diabetes mellitus or renal disease, clinical malabsorption, neoplastic disease, bleeding disorders, or any disease that the investigator believed would have confounded the study results or posed a risk to the patient. Other exclusion criteria were allergy to aspirin, hydrocodone bitartrate, acetaminophen, or NSAIDs and history of significant clinical or laboratory abnormalities that contraindicated the use of selective or nonselective NSAIDs. Furthermore, patients were not eligible if they had a recent history of narcotic or alcohol abuse or required treatment for depression with tricyclic antidepressants.
All patients underwent orthopedic surgery without restriction of the anesthesia used; however, treatments with hypnotics, sedatives, tranquilizers, or muscle relaxants were excluded within 6 h before and until 12 h after dosing. Generally, patients’ initial pain was controlled by epidural or patient-controlled administration pump of opioids with a minimum washout period of 30 min. If a patient had received oral opioids, the minimum washout period was 6 h. NSAIDs were excluded after surgery; however, if they were used, the patients were subjected to a washout period of 24 to 72 h, as determined by the half-life of the drug. After appropriate washout of analgesic medication, consenting patients who reported moderate or severe pain within 72 h after completion of surgery were eligible for the study. Patients were randomized according to a computer-generated allocation schedule to receive etoricoxib 120 mg, controlled-release naproxen sodium 1100 mg, or placebo (1:1:1) on day 1, followed by etoricoxib 120 mg and placebo (1:2) on days 2 to 7 (Fig. 1). The allocation was stratified within each study center by baseline pain intensity (moderate or severe) and type of surgery (knee or hip). The study medication was administered as 4 tablets: 2 tablets of etoricoxib 60 mg or matching placebo and 2 tablets of controlled-release naproxen sodium (NAPRELAN®; Wyeth, Collegeville, PA) 550 mg or matching placebo.
In Part I (day 1), two stopwatches were started at the time of dosing. Patients stopped the first stopwatch when they experienced perceptible pain relief and the second when they experienced meaningful pain relief (16). During the 24-h period after the first dose of study medication, patients rated their pain intensity (none, slight, moderate, or severe) and the pain relief (none, a little, some, a lot, or complete) at 14 prespecified time points (30, 60, 90, 120 min, 3, 4, 5, 6, 7, 8, 10, 12, 20, and 24 h). At 8 and 24 h postdose or at the time of rescue medication, patients recorded a Global Evaluation of Response to Therapy by responding to the following question using a scale of poor, fair, good, very good, and excellent: “How would you rate the study medication you received for pain?” Hydrocodone/acetaminophen 7.5/500 mg (LORTAB™ 7.5; UCB Pharma, Smyrna, GA) was available as supplemental rescue medication in case the study medication would not provide sufficient pain relief. However, patients were encouraged to refrain from using rescue medication during the first 90 min after dosing.
All patients were eligible to participate in Part II (days 2 to 7). Patients who received etoricoxib on day 1 continued to receive etoricoxib on days 2 to 7, and patients who received placebo or naproxen sodium on day 1 received placebo on days 2 to 7 (Fig. 1). Patients took study medication once daily in the morning and they were permitted to use hydrocodone/acetaminophen to treat intolerable pain; no other analgesics were allowed. Each evening at bedtime, patients completed a Global Evaluation of Response to Therapy (poor, fair, good, very good, and excellent) and a pain questionnaire rating their worst, least, and average pain since the last dose of study medication using a 0 (no pain) to 10 (worst imaginable pain) numeric rating scale (17).
The primary end-point was total pain relief over 8 h. Secondary and other end-points to determine single-dose efficacy included sum of pain intensity difference over 8 h, Patient’s Global Evaluation at 8 and 24 h, percentage of patients with and time to confirmed perceptible pain relief, peak pain intensity difference and peak pain relief, and percent of patients using and time to use of rescue medication. The amount of rescue medication (tablets of hydrocodone/acetaminophen) used on day 1 was determined in a post hoc analysis. The end-points in Part II (days 2-7) were rescue medication use (key end-point), Patient’s Global Evaluation, and pain intensity assessment (worst, average, least pain).
A physical examination, evaluation of vital signs (arterial blood pressure, heart rate, respiration rate, and oral temperature), and laboratory safety tests (complete blood count, serum chemistry panel, and urinalysis) were performed before randomization and approximately 14 days after the last dose of study drug. Clinical and laboratory adverse experiences were reported throughout the study and the investigators judged intensity and causality to study medication.
Total pain relief over 8 h and sum of pain intensity difference over 8 h were determined as the time-weighted average with the last value carried forward for patients who discontinued. The analysis of variance model was used to analyze total pain relief over 8 h, sum of pain intensity difference over 8 h, Patient’s Global Evaluation, peak pain intensity difference, peak pain relief, amount of rescue medication used, and the Pain Intensity Assessment Scores. Patient’s Global Evaluation, pain relief, and pain intensity scores were defined on categorical scales; however, for analytical purposes they were treated as continuous data. Pain relief and pain intensity values that were missing after a patient used supplemental analgesic medication were estimated by carrying forward the last available values before calculating total pain relief over 8 h and sum of pain intensity difference over 8 h. The analysis of variance model included factors of treatment, baseline pain intensity, study site, and surgical procedure (hip or knee). Treatment effects were assessed by evaluating within-treatment least-squares means and the between-treatment difference least-squares means (95% confidence intervals) obtained from the analysis of variance model. The Cox proportional hazards regression model assessed treatment effects on the time-to-event variables, and percent of patients who experienced an event was analyzed using a logistic regression model with factors similar to those in the analysis of variance model.
Part II included two placebo groups. The protocol prespecified that only the placebo/placebo group (i.e., patients who received placebo in both Part I and Part II) would be used for the determination of the analgesic effect as a result of a possible confounding effect of controlled-release naproxen sodium used in Part I.
A total of 412 patients were screened and 228 patients were randomized (Fig. 1). The 3 most common reasons for patients not being randomized were 1) withdrawal of consent (50 patients); 2) no reports of moderate or severe pain (41 patients); and 3) use of unallowed medications (24 patients). Of the 228 randomized patients, 210 patients (92.1%) completed Part I and 186 (88.6%) completed Part II. Specifically, 19 (25.3%) patients in the placebo/placebo group, 11 (13.8%) in the etoricoxib 120 mg/etoricoxib 120 mg group, and 12 (16.4%) in the naproxen sodium 1100 mg/placebo group discontinued. Patient baseline characteristics were similar across treatment groups, including the time between end of surgery and allocation (Table 1).
On day 1 (single-dose efficacy), etoricoxib provided significantly greater overall analgesic effect compared with placebo as assessed by mean pain relief scores over 24 h. The pain relief curve was similar to that of the patients taking controlled-release naproxen sodium (Fig. 2). The least squares mean total pain relief over 8 h score for etoricoxib was significantly (P < 0.001) more than that of placebo and similar to that of naproxen sodium; sum of pain intensity difference over 8 h scores showed similar results (Tables 2 and 3). Furthermore, at 8 h, 53% of etoricoxib patients, 60% of naproxen sodium patients, and 26% of placebo patients evaluated the study medication to provide “good, very good, or excellent” pain relief; results at 24 h were similar to those at 8 h (Fig. 3).
Onset of analgesia, as assessed by the median time to confirmed perceptible pain relief, was significantly more rapid in the etoricoxib group than in the placebo group (P < 0.05) and similar to that of naproxen sodium. Furthermore, a significantly larger percent of etoricoxib patients, compared with placebo patients, achieved onset as assessed by the 2-stopwatch method (51.3% for etoricoxib versus 32.0% for placebo; P < 0.05); a similarly large group of patients (50.7%) experienced onset in the naproxen sodium group (Table 2).
The peak effect was assessed by both peak pain intensity difference and peak pain relief scores; the respective least squares mean scores for the etoricoxib (1.2 and 2.0 U) and controlled-release naproxen sodium (1.3 and 2.4 U) were significantly larger (P < 0.050) than those for placebo (0.9 and 1.5 U).
The median time to the first dose of rescue medication was significantly (P < 0.001) longer for etoricoxib compared with placebo and similar to that of naproxen sodium (Table 2). A larger percentage of patients in the placebo group (98.7%) required rescue medication compared with etoricoxib (82.5%) and naproxen sodium (75.3%) during the first 24 h (Table 2). The pain relief and the pain intensity difference scores at 12 and 24 h are other measures of maintenance of effect. The pain relief scores at 12 h and 24 h were 1.1 and 1.0 U for etoricoxib and 0.4 and 0.3 U for placebo indicating that at each of these time points, etoricoxib provided significantly better analgesia compared with placebo (P ≪ 0.002). Patients on controlled-release naproxen sodium reported similar values as those on etoricoxib. Furthermore, a post hoc analysis showed that patients in the placebo group used significantly more hydrocodone/acetaminophen tablets than patients in the etoricoxib and the controlled-release naproxen sodium groups on day 1, 3.1 tablets versus 1.9 and 1.7 tablets, respectively (P < 0.001).
On days 2–7 (Part II), the patients in the etoricoxib group used significantly (P < 0.001) fewer hydrocodone/acetaminophen tablets compared with those in the placebo group (Table 3). In addition, the patients in the etoricoxib group reported significantly greater daily Global Evaluation scores than did those in the placebo group (P < 0.001; Table 3), and a larger percentage of patients receiving etoricoxib (approximately 74%) reported good, very good, and excellent responses than did those receiving placebo (approximately 50%). Furthermore, patients in the etoricoxib group experienced significantly less pain as assessed by both “worse” and “average” pain than patients in the placebo group; “least” pain was similar between the two groups (Table 2).
The incidence of clinical adverse experiences was generally similar across the three treatment groups: placebo/placebo (placebo in Parts I and II), etoricoxib/etoricoxib (etoricoxib 120 mg in Parts I and II), and naproxen sodium/placebo (controlled-release naproxen sodium 1100 mg in Part I and placebo in Part II) (Table 4). The three most common adverse experiences were constipation, nausea, and fever. Constipation and fever were more commonly reported in the two placebo groups than in the etoricoxib group, whereas nausea occurred at similar incidences across treatment groups (Table 4). Two (2.7%), three (3.8%) and one (1.4%) serious clinical adverse experiences were reported in the placebo/placebo, etoricoxib/etoricoxib, and naproxen sodium/placebo groups, respectively. Most of these adverse experiences were common postsurgery complications: joint infection, cellulitis, hematoma, wound dehiscence, and postoperative complication. In addition, there was one report of atrial fibrillation (etoricoxib), one of congestive heart failure (5 days after the last dose of etoricoxib), and one episode of upper gastrointestinal hemorrhage resulting from bleeding from a Dieulafoy’s lesion (8 days after the last dose of etoricoxib). The latter two events had been judged in a blinded fashion as possibly related to study drug by the investigators. Laboratory adverse experiences occurred in 10 (13.7%) patients taking placebo, 5 (6.3%) patients taking etoricoxib, and 6 (8.5%) patients taking controlled-release naproxen sodium/placebo.
Nonselective NSAIDs inhibit both the COX-1 and COX-2 enzymes, whereas the COX-2 inhibitors, in the clinical dose range, selectively inhibit the COX-2 enzyme. The inhibition of the COX-2 enzyme is essential in the treatment of chronic and acute pain (10). The role of COX-1 inhibition in the treatment of acute pain is yet to be fully understood, although some argue that it may play an important role in treatment of nociceptive pain (18). However, the role of the COX-1 enzyme in NSAID-induced gastrointestinal side effects and its role in platelet function are well known. Therefore, the use of nonselective NSAIDs is limited in the treatment of acute pain in a postsurgical setting, as they may result in postsurgical gastrointestinal or bleeding complications (6,19), This two-part study evaluated the analgesic effect of etoricoxib, a new COX-2 selective inhibitor, in a postsurgical setting. In other studies of acute pain e.g., acute pain associated with removal of impacted third molars, etoricoxib 120 mg provided rapid onset of analgesia as well as an analgesic effect that was maintained over 24 hours (14,20). These studies have shown that, in acute pain associated with minor surgery, a nonselective NSAID or a COX-2 selective inhibitor alone can provide substantial pain relief to most patients. However, after major surgery, such as postorthopedic surgery, multimodal analgesic therapy is generally required, e.g., opioids and other drugs (21).
The present study demonstrated that etoricoxib 120 mg is effective in providing analgesia over a 7-day period compared with placebo in patients who had undergone total knee or hip replacement surgery. Patients experiencing moderate to severe pain within 72 hours after surgery received superior overall analgesic effect from etoricoxib compared with placebo over the first 24-hour postdose time period. Etoricoxib had a similar overall analgesic effect as controlled-release naproxen sodium, a once-daily nonselective NSAID, which was included on day 1 to validate the study and to benchmark the treatment effect. The study also demonstrated that etoricoxib could provide effective pain control in acute pain that lasted for several days. This was done through the key objective of Part II (days 2-7), where patients receiving etoricoxib 120 mg used significantly less rescue medication (hydrocodone/acetaminophen), had superior Patient’s Global Evaluation scores, and suffered significantly less pain than the patients who received placebo (P < 0.05 for all). Etoricoxib 120 mg once daily reduced the hydrocodone/acetaminophen rescue medication use with approximately 35% compared with placebo, which is the range of reduction that has been reported for other selective and nonselective NSAIDs in the treatment of postorthopedic surgery pain (10,22,23). Patients who received etoricoxib also reported superior global evaluation scores and less pain as well as less hydrocodone/acetaminophen rescue medication usage; this suggests that etoricoxib had an opioid-sparing effect and that etoricoxib along with supplemental opioid medication in the form of hydrocodone/acetaminophen provided superior analgesic effect compared with placebo with supplemental opioid medication.
Other important variables of analgesia, e.g., onset and duration of the analgesic effect were determined on day 1. The onset was assessed by the 2-stopwatch method, which is a well-validated method for single-center studies; however, for multicenter, in-hospital studies it is often difficult to achieve the required rigorous standardization of this method. In the etoricoxib group, 51.3% of the patients experienced onset of analgesia, compared with 50.7% of naproxen sodium patients, and 30.2% in the placebo group; these patients experienced onset of analgesia within 1 hour. However, when determining the median time to onset for all patients, etoricoxib patients had an onset of 2.6 hours or earlier compared with 1.4 hours or earlier for naproxen sodium patients. Thus, the percentage of patients may better describe the phenomena of onset of analgesia than the actual time in this multicenter study where etoricoxib had a time to onset of 2.6 hours and naproxen had a time to onset of sodium 1.4 hours, whereas within the subgroup of patients reporting analgesic onset, the time to onset occurred within 1 hour.
Clinical settings in which multimodal treatment of pain is required are not optimal for determining the duration of a single therapy. However, it is still possible to assess the analgesic effect over 24 hours through evaluation of the pain and pain relief scores over the 24-hour time period. Patients treated with etoricoxib reported significantly greater pain relief scores and lower pain intensity scores at 10, 12, 20, and 24 hours than did those who received placebo (Fig. 2).
Etoricoxib 120 mg was generally safe and well tolerated in this study. The overall incidence of clinical and laboratory adverse experiences in the etoricoxib group was similar or less than those in the placebo groups. Specifically, patients in the placebo groups had a numerically more frequent incidence of constipation compared with patients taking etoricoxib (Table 3). In this study, patients on etoricoxib used 35% less opioid-containing rescue medication compared with patients on placebo; it is conceivable that this decrease in the use of rescue medication was responsible for the numeric decrease in the number of episodes of constipation on etoricoxib compared with placebo. Moreover, in this study, there was a numeric trend in favor of etoricoxib compared with placebo for reported episodes of fever and headache, which are other commonly observed adverse experiences in a postsurgical setting. Unlike nonselective NSAIDs, etoricoxib does not affect platelet function (6,19) making it a useful analgesic in the postsurgery setting. Nonselective NSAIDs and COX-2 selective inhibitors have been associated with fluid retention and hypertension in some patients (24). In this study, the incidence of lower extremity edema, commonly observed after orthopedic surgery, was similar among all treatment groups. Hypertension was observed in 2 patients; one in the etoricoxib/etoricoxib group and one in the naproxen sodium/placebo group. However, there were four reports of hypotension, all in the etoricoxib group. All events were transient and self-limited and none was determined to be related to the study drug.
This study was designed to assess the analgesic effect of etoricoxib in the treatment of acute pain over several days. The study was not designed or powered to assess specific safety variables, e.g., bone healing. Although it is known that the cyclooxygenases are involved in bone healing and bone modulation (25), no prospective, randomized clinical trials in humans have been performed to assess this effect. In retrospective studies and analyses of the nonselective NSAID ketorolac and the COX-2 selective inhibitors rofecoxib and celecoxib, only ketorolac, a potent inhibitor of the COX-1 enzyme, has been shown to have an effect on bone healing (26,27).
In summary, etoricoxib 120 mg administered once daily for 7 days provided analgesia in a postsurgery setting. On day 1 etoricoxib was superior to placebo and similar to controlled-release naproxen sodium, a once-daily nonselective NSAID, with respect to overall analgesic effect, onset of effect, peak effect, and time to the first dose of rescue medication in the treatment of postorthopedic surgery pain when dosed within 72 hours after surgery. Over the entire 7 days of treatment, etoricoxib significantly reduced opioid-containing rescue medication (hydrocodone/acetaminophen) use versus placebo while decreasing pain intensity. The study results suggest that, in an acute postsurgery setting, treatment with etoricoxib, when taken with supplemental opioid medication such as hydrocodone/acetaminophen, results in a superior analgesic effect compared with placebo taken with supplemental hydrocodone/acetaminophen over 7 days. The study also suggests that by reducing the use of opioid-containing rescue medication, opioid-associated adverse experiences may also be reduced.
The authors wish to thank Heather Coughlin and Megan McGratty for their assistance with study start-up and monitoring; Heather O’Kelly for help with data collecting tools; Tersit Biftu for providing reports; Staci Gola and Mindi Turpin at JBA Research, Marc Maslanka at South Eastern Center for Clinical Trials, and Trish Hardee at Coastal Medical Research for site coordination; Anish Mehta for assistance with the manuscript; and Dr. Jan Markind for her critical review of the manuscript.
The Etoricoxib Postorthopedic Study Group consisted of the following investigators and organizations: M.H. Bourne MD (Salt Lake Orthopedic Clinic, Salt Lake City, UT), M. Jove MD (Atlanta Knee and Sports Medicine, Decatur, GA), G. Kimball MD (Alta View Hospital, Sandy, UT), W.P. McLeod MD (Bert Fish Medical Center, New Smyrna, FL), A. Paterson MD (UT Valley Regional Medical Center, Provo, UT), G.L. Rasmussen MD (Orthopedic Specialty Hospital, Murray, UT), J.R. Rhodes MD (Halifax Medical Center, Daytona Beach, FL), S.M. Rhondeau MD (LDS Hospital, Salt Lake City, UT), JBA Research in Salt Lake City, UT and Costal Medical Research in South Daytona, FL.
1. Cherny NL. Opioid analgesics: comparative features and prescribing guidelines. Drugs 1996;51:713–37.
2. Rosenblum M, Weller RS, Conrad PL, et al. Ibuprofen provides longer lasting analgesia than fentanyl after laparoscopic surgery. Anesth Analg 1991;73:255–9.
3. Buggy DJ, Wall C, Carton EG. Preoperative or postoperative diclofenac for laparoscopic tubal ligation. Br J Anaesth 1994;73:767–70.
4. Vane JR, Botting RM. New insights into the mode of action of anti-inflammatory drugs. Inflamm Res 1995;44:1–10.
5. Donnelly MT, Hawkey CJ. COX-II inhibitors: a new generation of safer NSAIDs? Aliment Pharmacol Ther 1997;11:227–36.
6. Schafer AI. Effects of nonsteroidal antiinflammatory drugs on platelet function and systemic hemostasis. J Clin Pharmacol 1995;35:209–19.
7. Weaver AL. Rofecoxib: clinical pharmacology and clinical experience. Clin Ther 2001;23:1323–38.
8. Clemett D, Goa KL. Celecoxib: a review of its use in osteoarthritis, rheumatoid arthritis and acute pain. Drugs 2000;59:957–80.
9. Goldman M, Schutzer S. Valdecoxib: a COX-2 inhibitor for treatment of osteoarthritis, rheumatoid arthritis, and primary dysmenorrhea. Formulary 2002;37:68–77.
10. Katz WA. Cyclooxygenase-2-selective inhibitors in the management of acute and perioperative pain. Cleve Clin J Med 2002;69(Suppl. 1):SI65–75.
11. Gottesdiener K, Schnitzer T, Fisher C, et al. Results of a randomized, dose-ranging trial of etoricoxib in patients with osteoarthritis. Rheumatol (Oxford) 2002;41:1052–61.
12. Matsumoto AK, Melian A, Mandel DR, et al. A randomized, controlled, clinical trial of etoricoxib in the treatment of rheumatoid arthritis. J Rheumatol 2002;29:1623–30.
13. Schumacher Jr., HR Boice JA, Daikh DI, et al. Randomised double blind trial of etoricoxib and indomethacin in treatment of acute gouty arthritis. BMJ 2002;324:1488–92.
14. Malmstrom K, Kotey P, Coughlin H, Desjardin PJ. A randomized, double-blind, parallel-group study comparing the analgesic effect of etoricoxib to placebo, naproxen sodium, and acetaminophen with codeine using the dental impaction pain model. Clin J Pain 2004;20:147–55.
15. Malmstrom K, Kotey P, Cichanowitz N, et al. Analgesic efficacy of etoricoxib in primary dysmenorrhea: results of a randomized, controlled trial. Gynecol Obstet Invest 2003;56:65–9.
16. Desjardins PJ, Black PM, Balm TK, et al. Onset of analgesia: further validation of a new stopwatch method. Clin Pharmacol Ther 1996;59:130.
17. Cleeland CS, Ryan KM. Pain assessment: global use of the brief pain inventory. Ann Acad Med 1994;23:129-38.
18. Ochi T, Motoyama Y, Goto T. The analgesic effect profile of FR122047, a selective cyclooxygenase-1 inhibitor, in chemical nociceptive models. Eur J Pharmacol 2000;391:49–54.
19. Dallob A, Hawkey CJ, Greenberg H, et al. Characterization of etoricoxib, a novel, selective COX-2 inhibitor. J Clin Pharmacol 2003;43:573–85.
20. Sinatra R. Role of COX-2 inhibitors in the evolution of acute pain management. J Pain Symptom Manage 2002; 24:S18–27.
21. DeAndrade RJ, Maslanka M, Reines HD, et al. Ketoralac versus meperidine for pain relief after orthopaedic surgery. Clin Orthop 1996;325:302–12.
22. Joris J. Efficacy of nonsteroidal antiinflammatory drugs in postoperative pain. Acta Anaesthesiol Belg 1996;47:115–23.
23. Gerstenfeld LC, Einhorn TA. COX inhibitors and their effects on bone healing. Expert Opin Drug Saf 2004;3:131–6.
24. Glassman SD, Rose SM, Dimar JR, et al. The effect of postoperative nonsteroidal antiinflammatory drug administration on spinal fusion. Spine 1998;23:834–8.
25. Reuben SS. Effect of nonsteroidal anti-inflammatory drugs on osteogenesis and spinal fusion. Reg Anesth Pain Med 2001;26:590–1.
26. Brater DC. Effects of nonsteroidal anti-inflammatory drugs on renal function: focus of cyclooxygenase-2-selective inhibition. Am J Med 1999;107:65S–70S.
© 2005 International Anesthesia Research Society
27. Malmstrom K, Sapre A, Coughlin H, et al. Etoricoxib in acute pain associated with dental surgery: a randomized, double-blind, placebo- and active comparator-controlled dose-ranging study. Clin Ther 2004;26:667–79.