More than one million cesarean deliveries are performed each year in the United States and the number is increasing as a result of changing obstetric management (1). Many cesarean deliveries are performed under regional anesthesia, with neuraxial opioids and supplemental oral or IV opioids given for postoperative pain relief. However, despite this approach, cesarean deliveries result in moderate-to-severe postoperative pain that is often incompletely relieved by pain management protocols (2).
The addition of nonsteroidal antiinflammatory drugs (NSAIDs) to a post-cesarean analgesic regimen has been shown to improve post-cesarean pain and reduce opioid requirements (3–7). Potential maternal side effects (e.g., antiplatelet and gastrointestinal) and effects on the breast-feeding infant have raised concerns about their use in the post-cesarean delivery setting (8,9). Cyclooxygenase-2 specific inhibitors (COX-2 inhibitors) have been shown to be effective postoperative analgesics, decreasing pain scores and analgesic consumption after general, orthopedic, and dental procedures (10–13). Studies comparing COX-2 inhibitors to NSAIDs have demonstrated similar analgesic efficacy and opioid-sparing after surgery in a non-obstetric setting (14,15). COX-2 inhibitors are thus a potentially attractive alternative for post-cesarean use because of their minimal platelet inhibition compared with NSAIDs (10,16). No previous study has evaluated the analgesic efficacy of COX-2 inhibitors in this setting.
The objective of this randomized, double-blind, placebo-controlled study was to determine the postoperative analgesic efficacy of a selective COX-2 inhibitor, valdecoxib, in women undergoing cesarean delivery under spinal anesthesia.
After IRB approval and informed consent, 48 healthy ASA physical status I–II women undergoing elective cesarean delivery were enrolled in this randomized placebo-controlled study. All parturients aged 18–45 yr with a singleton pregnancy and gestation more than 37 wk undergoing elective cesarean delivery without postpartum tubal ligation were considered suitable for enrollment. Exclusion criteria included any severe uncontrolled medical condition (including pregnancy-induced hypertension, renal disorders, coagulation disorder, significant cardiovascular disease, and peptic ulcer disease); use of narcotic analgesics, anticonvulsants, corticosteroids, antidepressants, anxiolytics, or NSAIDS < 24 h before cesarean delivery; body mass index >40; prior myomectomy or vertical uterine scar; a history of alcohol, analgesic, or narcotic abuse; any significant antepartum, intrapartum, or postpartum hemorrhage; inadequate intraoperative neuraxial anesthesia requiring conversion to general anesthesia; previous breast surgery or other breast problems resulting in inability to breast-feed; allergies to sulfa, opioid, or NSAID medications, and NSAID-sensitive asthmatics.
All patients received spinal anesthesia with 12 mg hyperbaric bupivacaine, 10 μg fentanyl, and 100 μg intrathecal morphine. Patients were randomized into 2 groups to receive either PO valdecoxib 20 mg or placebo twice daily for the 72-h study period. Group assignments were determined by a computer-generated number sequence and were contained in opaque envelopes to insure blinding of the investigators. No block randomization was used in the study. The first dose of study drug was given 90 min after the spinal block, with subsequent doses given every 12 h “around-the clock” for a total of 6 doses. Treatments were administered via capsules filled with either valdecoxib 20 mg or lactose (placebo), thereby ensuring double-blinding. The first dose was given by the study investigator and this time was considered Time 0. Subsequent doses were given by postpartum nurses as part of the postoperative medication orders. Patients were evaluated by a blinded observer at 2, 4, 8, 12, 18, 24, 36, 48, 60, and 72 h.
Supplemental analgesia was administered according to a strict postoperative pain management protocol. Two oral opioids routinely used in our institution were permitted, hydrocodone 5 mg/acetaminophen 500 mg (Vicodin®; Abbott Laboratories, North Chicago, IL) and oxycodone 5 mg/acetaminophen 325 mg (Percocet®; Endo Pharmaceuticals, Chadds Ford, PA), with IV morphine used for severe or resistant pain. No NSAIDs or other oral opioid analgesics were permitted during the study period. Postoperative pain management was standardized as follows:
- Mild pain (verbal pain scores [0 = no pain and 10 = worse pain imaginable] VPS <3 were offered 1 tablet (hydrocodone 5 mg or oxycodone 5 mg) every 4 h as needed.
- Moderate pain (VPS 3–7) were offered 2 tablets (hydrocodone 10 mg or oxycodone 10 mg) every 4 h as needed.
- Severe pain (VPS >7) or moderate pain not responsive to oral analgesics (as above) were offered IV morphine 2 to 4 mg every 10 min as necessary until comfortable (up to 10 mg/h).
Maximum doses allowed were oxycodone 60 mg or hydrocodone 60 mg or 4 g acetaminophen (in combination with the oral opioid) in a 24-h period.
The primary end-points of the study were pain intensity rating and postoperative analgesic medication consumption. Pain intensity was measured using VPS (0–10 with 0 = no pain and 10 = the worst pain you can imagine) and a categorical rating scale (none, mild, moderate, or severe) at rest and at movement (defined as sitting up 90°). Pain scores were recorded at rest and at movement at 2, 4, 8, 12, 18, 24, 27, 36, 48, 51, 60, and 72 h after cesarean delivery. The total amount of supplemental opioid analgesic medication used in the 72-h study period was documented. Postoperative analgesics (hydrocodone and oxycodone) were converted to morphine mg-equivalents for analysis. Standardized opioid conversion values adapted from Goodman and Gillman’s text (17) were used to determine the mg-morphine equivalents, with oral hydrocodone 7.5 mg and oral oxycodone 7.5 mg being considered equivalent to 10 mg IV morphine.
Secondary end-points included time to first analgesic request (minutes after spinal block) and the proportion of patients requiring no postoperative IV medication. Patient satisfaction and functional activity scores were also recorded every 24 h after cesarean delivery. Patient satisfaction was assessed on Days 1, 2, and 3 using a categorical rating scale (0 = extremely unsatisfied, 1 = unsatisfied, 2 = mildly satisfied, 3 = satisfied, 4 = extremely satisfied). Functional activity was assessed using a scale response (1 = can sit with assistance, 2 = sit without assistance, 3 = stand with assistance, 4 = stand without assistance, 5 = walk with assistance, 6 = walk without assistance). Hospital stay and time to hospital discharge were also recorded.
Potential maternal adverse events including pruritus, nausea and vomiting, sedation, postpartum vaginal bleeding, and hypertension were assessed daily at 27, 51, and 72 h after cesarean delivery. Any fetal adverse events (e.g., feeding difficulties, bleeding) were also recorded. All adverse events were reported and the outcome, severity, and possible relationship to study drug were documented. Nausea was assessed using a numerical rating scale (0–10 with 0 = no nausea and 10 = the worst nausea imaginable); vomiting was assessed every 24 h as yes/no and number of episodes per 24-h period; pruritus was assessed using a numerical rating scale (0–10 with 0 = no itching and 10 = the worst itching imaginable); and sedation was assessed using a 5-point ordinal scale: 0 = alert, 1 = occasionally drowsy, 2 = frequently drowsy, 3 = sleepy but easy to arouse, 4 = somnolent and difficult to arouse.
Vaginal blood loss was defined as 0 = none, 1 = mild, 2 = moderate, or 3 = severe. Hypertension was defined as >25% increase in systolic blood pressure from baseline. Nausea and vomiting were treated with granisetron 1 mg and repeated every 15 min as necessary up to a maximum of 3 mg, and pruritus was treated with small doses of naloxone 40 μg repeated every 5 min to a maximum of 400 μg as necessary. Early breast-feeding success was measured by time to first breast feeding and the number of breast feedings per day. Late breast-feeding success was assessed approximately 1 mo after discharge with a telephone follow-up questionnaire to determine continuing breast-feeding success. Parturients were asked, “What is your current method of feeding your baby?” [1 = exclusively breast milk-fed (vitamins and minerals not included); 2 = fully breast milk-fed (less than daily use of any other liquids); 3 = partially breast milk-fed (daily use of formula); 4 = no longer breastfeeding].
An a priori sample size calculation revealed that 37 patients per study group were required to detect a 30% reduction in incisional pain scores on movement at 36 h after cesarean delivery (power = 0.8; P = 0.05) (4). However, because of concerns about the safety of COX-2 inhibitors that became apparent from published reports during this study, a decision was made to terminate the study early until the safety concerns were better understood.
Analyses were done on an intent-to-treat basis using all parturients who enrolled into the study. Descriptive statistics were used to summarize demographic, outcome, and side effect data. Outcome measures of interest between the two groups were compared using Student’s t-test and the Wilcoxon two-sample test where appropriate. Normal distribution was determined using QQ plots and the Kolmogorov-Smirnov test. Associations among discrete variables were investigated using Pearson’s χ2 and Fisher’s exact test where appropriate. Longitudinal data analysis was performed using random coefficients mixed models regression (SAS PROC MIXED) to model baseline outcome measure values and their change over time during the follow-up period. Post hoc power analysis was based on two-sample Student’s t-test of means. Analyses were performed with SPSS 11.0 statistical package (SPSS, Chicago, IL), Microsoft Excel and SAS 9.1 statistical package (SAS, Cary, NC) with P < 0.05 considered statistically significant.
Of the 48 patients enrolled, 47 completed the entire protocol. The one patient (in the valdecoxib group) who did not complete the protocol was withdrawn by her obstetrician on postoperative day 1 because of her obstetrician’s desire to administer nonselective NSAIDs. The intention-to-treat population consisted of all the patients enrolled and data were analyzed in all 48 patients. Demographic data were similar between the groups (Table 1, P = not significant).
We did not observe any significant differences in our primary outcomes between the groups. Pain at rest and with movement were similar between the groups (Figs. 1 and 2). Pain scores decreased over time during the 72-h study period but we detected no differences in pain between the 2 groups (P = 0.37 and P = 0.42 for pain at rest and movement respectively). There were no differences between the groups in IV morphine requirements, time to first analgesic request, or total analgesic use in morphine equivalents (Table 2).
Adverse events and possible side effects, such as vaginal bleeding, hypertension, sedation, pruritus, and nausea, were not significantly different between the study groups (Table 3). Overall, most patients were satisfied or extremely satisfied, and we detected no differences in the level of satisfaction between the study groups on Days 1, 2, or 3 after cesarean delivery. Other secondary outcome measures, such as breast-feeding success and functional activity, were also not significantly different between groups (data not shown).
Post hoc power analysis indicates that we had 80% and 90% power to detect a clinically significant difference of 30% difference in pain scores and analgesic consumption, respectively (Table 4).
Under the conditions of our study, we were unable to demonstrate significant analgesic efficacy or an opioid-sparing effect with the addition of valdecoxib after cesarean delivery under spinal anesthesia with intrathecal opioids. In contrast, previous studies using NSAIDs in this setting showed significant post-cesarean opioid-sparing and reductions in pain scores (3–7).
Intrathecal morphine administered at the time of spinal anesthesia for the control of post-cesarean delivery pain is usually quite effective for the first 24 hours (18). In this setting, peak postsurgical pain scores tend to shift from the first to the second and third postoperative days as the morphine effects wear off (2,4). Previous studies have shown that the addition of an NSAID to the post-cesarean delivery pain management regimen may not provide additional analgesia during the first 24 hours but does so at 36 hours (4,19). Despite our 72-hour study period, we were unable to demonstrate reduced pain scores at any time with valdecoxib administration.
Because of concerns about the safety of COX-2 inhibitors that became apparent from published reports during this study, we did not feel it was ethical to continue until further safety evaluations were undertaken. However, based on our data, we do not believe we missed any significant difference in analgesic consumption and pain scores between the study groups, and we believe the study had adequate power (Table 4) to have detected a clinically significant analgesic benefit from valdecoxib.
The relatively low pain scores in both study groups make detection of a treatment effect difficult and may provide a possible explanation for this lack of analgesic effect. A possible explanation for the lack of treatment effect might relate to the dose of the medication studied. However, we consider this unlikely, as we used a standard postoperative dose (40 mg daily), in keeping with other postoperative studies, and valdecoxib reaches an analgesic efficacy plateau at this dose in the dental surgery model (20). It may be relevant that changes in estrogen and progesterone levels during pregnancy increase COX-2 expression (21). Although the significance of such increased COX-2 expression is unknown (21), it may necessitate higher dose requirements of COX-2 inhibitors.
The lack of analgesic efficacy cannot be explained by the duration of treatment because steady-state concentrations develop after three doses of valdecoxib. Administration of COX-2 inhibitors for 3 days postoperatively should have provided an adequate duration to have demonstrated a significant treatment effect. Another possible explanation for valdecoxib’s lack of analgesic efficacy compared with NSAIDs in this setting may lie in the COX-2 inhibitors’ selectivity. NSAIDs with both COX-1 and COX-2 effects are particularly effective in relieving pain of a cramping nature after cesarean delivery (3). It could be that COX-2 inhibitors are not as effective at treating this component of the pain compared with NSAIDs.
Although there have been no previous studies designed to evaluate the analgesic efficacy of COX-2 inhibitors after cesarean delivery, Lee et al. (22) similarly found a lack of apparent analgesia when evaluating the effect of celecoxib on intrathecal morphine-induced pruritus after cesarean delivery. They found overall low pain scores but no difference between placebo and celecoxib in analgesia or pruritus (22). However, their findings must be viewed with caution, as pain was a secondary end-point. Also, patients were only evaluated for 24 hours after cesarean delivery when the analgesic effects of intrathecal morphine administered in the spinal anesthetic might have masked any analgesic effect of celecoxib (22). Considering our findings with valdecoxib, together with a lack of analgesic effect with celecoxib reported by Lee et al. (22), we speculate that the COX-2 inhibitors may be less effective in this post-cesarean pain model as compared with NSAIDs, which have been demonstrated to have a significant analgesic effect after cesarean delivery (3–7).
In the current study, we observed low pain scores throughout the study period. This is likely a consequence of the efficacy of intrathecal morphine and our post-cesarean pain management protocol, which allowed patients generous amounts of opioid analgesics to be administered as necessary. This probably contributed to the relatively large amount of analgesic morphine-equivalents consumed after cesarean delivery. If patients titrated their pain effectively with available analgesics, we would have expected less analgesic consumption in the valdecoxib study group if the valdecoxib had added a significant opioid-sparing effect.
Since finishing the study and analyzing the data, Pfizer, Inc. suspended sales of valdecoxib (Bextra) in the United States on April 7, 2005 because of an unfavorable overall benefit versus risk profile compared with NSAIDs. This announcement followed the September 30, 2004 recall of rofecoxib (Vioxx) by Merck because of potential myocardial and stroke risk. Whether the cardiovascular risks are a specific drug effect or a COX-2 inhibitor class effect is yet to be determined. Post-cesarean delivery patients are not at high risk for cardiovascular disease, and the 3-day postoperative use is unlikely to cause the problems associated with chronic administration. We believe the findings from this study are still important, as other COX-2 inhibitors (e.g., celecoxib) remain on the market, and this is the first study to evaluate the efficacy of this class of analgesics in the post-cesarean model. The reduced potential risk of hemorrhage in the mother and the breast-feeding infant with COX-2 inhibitors compared with NSAIDs is theoretically advantageous in this setting (9). In addition, COX-2 inhibitors should be safe during breast-feeding, as there is minimal drug transfer to the infant. Hale et al. (23) found the mean celecoxib concentration in breast milk was 66 μg/L, representing an absolute infant dose of only 9.8 μg·kg−1·d−1 or a relative infant dose of 0.3%, which is unlikely to cause untoward effects in breast-fed infants (23). These values are similar to those after NSAID administration and are much lower than those associated with most opioid analgesics used post-cesarean delivery (24).
We were unable to detect any differences in adverse effects or in early and late breast-feeding success. Because we found no significant differences in pain or analgesic consumption, it is not surprising that we found no differences in opioid-related side effects such as pruritus, sedation, and nausea or vomiting. In our study, valdecoxib was found to be safe, with no maternal or infant adverse events reported. However, this study was not powered to determine significant differences in many of these secondary end-points. A study of postoperative patients after coronary artery bypass grafting found an increase in thrombotic risk in patients receiving valdecoxib (25). Although this thrombotic risk was not confirmed in a meta-analysis of approximately 8000 patients with osteoarthritis and rheumatoid arthritis treated with valdecoxib in randomized clinical trials, it is warranted at this stage to avoid COX-2 inhibitors and NSAIDs in high-risk patients e.g., patients undergoing coronary artery bypass grafting (26). Although we experienced no thrombotic events and arterial blood pressures were similar between the valdecoxib and placebo groups, postpartum post-cesarean patients may be at higher risk for thrombotic complications. We elected to study valdecoxib because we felt that both this drug and its IV prodrug preparation, parecoxib, would be useful in the perioperative setting (27). In light of the side effects and thrombotic cardiovascular risks that became apparent after study completion, (26,28) COX-2 inhibitors are probably best avoided in this potentially prothrombotic population. Whether this holds true for all COX-2 inhibitors is unclear at this time but, given the efficacy and safety of NSAIDs in these patients, we do not currently see a place for COX-2 inhibitors in the treatment of post-cesarean pain.
This is the first study to evaluate the analgesic efficacy of a COX-2 inhibitor in the post-cesarean delivery setting. Although Pfizer, Inc. suspended the sales of valdecoxib, we feel the findings from this study are still clinically relevant. First, other COX-2 inhibitors like celecoxib are still available and other COX-2 inhibitors may be introduced in the future. Second, valdecoxib’s IV prodrug preparation, parecoxib is a potentially useful perioperative analgesic adjunct, and third, valdecoxib may be reintroduced in the future for specific indications, e.g., short-term postoperative pain management.
In conclusion, under the conditions of this study, we were unable to demonstrate significant analgesic efficacy of valdecoxib compared with placebo. Patients did not experience any adverse events related to short-term postoperative administration of valdecoxib. In light of the lack of significant analgesic efficacy observed and the potential thrombotic cardiovascular risks associated with valdecoxib, its routine use in the management of pain after cesarean delivery is not supported at this time. Whether these findings can be generalized to all COX-2 inhibitors is unclear, but based on findings from our study and data currently available, we recommend NSAIDs over COX-2 inhibitors for the treatment of post-cesarean pain.
1. Martin JA, Kochanek KD, Strobino DM, et al. Annual summary of vital statistics–2003. Pediatrics 2005;115:619–34.
2. Cohen SE, Subak LL, Brose WG, Halpern J. Analgesia after cesarean delivery: patient evaluations and costs of five opioid techniques. Reg Anesth 1991;16:141–9.
3. Huang Y, Tsai S, Huang C, et al. Intravenous tenoxicam reduces uterine cramps after cesarean delivery. Can J Anesth 2002;49:384–7.
4. Angle P, Halpern SH, Leighton BL, et al. A randomized controlled trial examining the effect of naproxen on analgesia during the second day after cesarean delivery. Anesth Analg 2002;95:741–5.
5. Pavy T, Paech MJ, Evans SF. The effect of intravenous ketorolac on opioid requirement and pain after cesarean delivery. Anesth Analg 2001;92:1010–4.
6. Cardoso M, Carvalho JCA, Amaro AR, et al. Small doses of intrathecal morphine combined with systemic diclofenac for postoperative pain control after cesarean delivery. Anesth Analg 1998;86:538–41.
7. Lim NL, Lo WK, Chong JL, Pan AX. Single dose diclofenac suppository reduces post-cesarean PCEA requirements. Can J Anaesth 2001;48:383–6.
8. American Academy of Pediatrics Committee on Drugs: The transfer of drugs and other chemicals into human milk. Pediatrics 1994;93:137–50.
9. Greer I, Johnston J, Tulloch I, Walker JJ. Effect of maternal ketorolac administration on platelet function in the newborn. Eur J Obstet Gynecol Reprod Biol 1988;29:257–60.
10. Fenton C, Keating GM, Wagstaff AJ. Valdecoxib: a review of its use in the management of osteoarthritis, rheumatoid arthritis, dysmenorrhoea and acute pain. Drugs 2004;64:1231–61.
11. Sinatra R. Role of COX-2 inhibitors in the evolution of acute pain management. J Pain Symptom Manage 2002;24:S18–27.
12. Desjardins PJ, Shu VS, Recker DP, et al. A single preoperative oral dose of valdecoxib, a new cyclooxygenase-2 specific inhibitor, relieves post-oral surgery or bunionectomy pain. Anesthesiology 2002;97:565–73.
13. Gan TJ, Joshi GP, Zhao SZ, et al. Presurgical intravenous parecoxib sodium and follow-up oral valdecoxib for pain management after laparoscopic cholecystectomy surgery reduces opioid requirements and opioid-related adverse effects. Acta Anaesthesiol Scand 2004;48:1194–207.
14. Romsing J, Moiniche S. A systematic review of COX-2 inhibitors compared with traditional NSAIDs, or different COX-2 inhibitors for post-operative pain. Acta Anaesthesiol Scand 2004;48:525–46.
15. Mehlisch DR, Desjardins PJ, Daniels S, Hubbard RC. Single doses of parecoxib sodium intravenously are as effective as ketorolac in reducing pain after oral surgery. J Oral Maxillofac Surg 2003;61:1030–7.
16. Leese PT, Talwalker S, Kent JD, Recker DP. Valdecoxib does not impair platelet function. Am J Emerg Med 2002;20:275–81.
17. Reisine T, Pasternak G. Opioid analgesics and antagonists. In: Hardman JG, ed. Goodman and Gillman’s the pharmacologic basis of therapeutics. New York: McGraw-Hill, 1996:535.
18. Palmer C, Emerson S, Volgoropolous D, Alves D. Dose-response relationship of intrathecal morphine for postcesarean analgesia. Anesthesiology 1999;90:437–44.
19. Cohen S, Desai JB, Ratner EF, et al. Ketorolac and spinal morphine for postcesarean analgesia. Int J Obstet Anesth 1996;5:14–8.
20. Desjardins PJ, Grossman EH, Kuss ME, et al. The injectable cyclooxygenase-2-specific inhibitor parecoxib sodium has analgesic efficacy when administered preoperatively. Anesth Analg 2001;93:721–7.
21. Chan VS. A mechanistic perspective on the specificity and extent of COX-2 inhibition in pregnancy. Drug Saf 2004;27:421–6.
22. Lee L, Irwin MG, Lim J, Wong CK. The effect of celecoxib on intrathecal morphine-induced pruritus in patients undergoing caesarean section. Anaesthesia 2004;59:876–80.
23. Hale TW, McDonald R, Boger J. Transfer of celecoxib into human milk. J Hum Lact 2004;20:397–403.
24. Hale TW. Anesthetic medications in breastfeeding mothers. J Hum Lact 1999;15:185–94.
25. Nussmeier NA, Whelton AA, Brown MT, et al. Complications of the COX-2 inhibitors parecoxib and valdecoxib after cardiac surgery. N Engl J Med 2005;352:1081–91.
26. Jones SF, Power I. Postoperative NSAIDs and COX-2 inhibitors: cardiovascular risks and benefits. Br J Anaesth 2005;95:281–4.
27. White PF. Changing role of COX-2 inhibitors in the perioperative period: is parecoxib really the answer? Anesth Analg 2005;100:1306–8.
28. Ray WA, MacDonald TM, Solomon DH, et al. COX-2 selective non-steroidal anti-inflammatory drugs and cardiovascular disease. Pharmacoepidemiol Drug Saf 2003;12:67–70.