AMBULATORY ANESTHESIA

SECTION EDITOR PAUL F. WHITE

SOCIETY

FOR

AMBULATORY ANESTHESIA

Sublingual Piroxicam for Postoperative Analgesia: Preoperative Versus Postoperative Administration: A Randomized, Double-Blind Study
Hans-F. Gramke, MD, Jurgen J. J. Petry, MD, Marcel E. Durieux, MD, PhD, Jean-P. Mustaki, MD, Marcel Vercauteren, MD, PhD, Georges Verheecke, MD, and Marco A. E. Marcus, MD, PhD
University Hospital Maastricht, Department of Anesthesiology and Pain Treatment, Maastricht, The Netherlands; Hopital de Zone Morges, Service d Anesthesiologie, Morges, Switzerland; Universitair Ziekenhuis Antwerpen, Dienst Anesthesiologie, Edegem, Belgium; Klinieken Noord Antwerpen, Dienst Anesthesiologie, Brasschaat, Belgium

Nonsteroidal antiinflammatory drugs have been used to obtain preemptive analgesia. We investigated, in this randomized, double-blind study, whether sublingual (s.l.) piroxicam given before was more effective than that given after surgery. Fifty-two patients scheduled for laparoscopic bilateral inguinal hernia repair under general anesthesia were enrolled. Group PRE (25 patients) received 40 mg of piroxicam s.l. 2 h before surgery and a placebo 10 min after surgery. Group POST (27 patients) were treated with a placebo 2 h before surgery and received 40 mg of piroxicam s.l. 10 min after surgery. After an initial dose of 100 mg tramadol IV,

patient-controlled analgesia with tramadol was started and recorded. Visual analog scores were assessed in the recovery and at 6, 20, and 30 h postoperatively. Significantly lower visual analog scores were found in group PRE at 6 and 20 h. Significantly smaller cumulative tramadol consumption was observed after 30 h in group PRE. In summary, our findings suggest that preoperative s.l. piroxicam is more effective than the postoperative administration. Because of the low pain scores in both groups, the clinical relevance of these findings is not clear from this study. (Anesth Analg 2006;102:7558)

he concept of preemptive analgesia consists of an antinociceptive treatment that prevents central neural sensitization that amplifies postoperative pain. This implies that the treatment has been established when the noxious stimulus starts. The use of nonsteroidal antiinflammatory drugs (NSAIDs) for preemptive analgesia is controversial. Although OHanlon et al. (1) described clinically relevant effects of preemptive NSAIDs, many other studies that have used NSAIDs for preemptive treatment of postoperative pain have yielded negative results (2 4). One possible explanation for the negative findings may be the route and the timing of administration. In the studies cited, the compounds were administered by oral, rectal, IM, and IV routes. Some of these may

Accepted for publication October 20, 2005. Address correspondence and reprint requests to M.A.E. Marcus, MD, PhD, University Hospital Maastricht, Department of Anesthesiology, PO Box 5800, 6202 AZ Maastricht, The Netherlands. Address e-mail to mmar@sane.azm.nl. DOI: 10.1213/01.ane.0000197611.89464.98
2006 by the International Anesthesia Research Society 0003-2999/06

require a prolonged time before being active. Piroxicam, a NSAID, exists in various forms of administration, including a fast dissolving dosage form (FDDF) for sublingual (s. l.) use. When FDDF tablets of piroxicam are administered s.l. they disintegrate instantly, releasing the drug into the saliva. One advantage of the s.l. route is that the patient can take the medication preoperatively or postoperatively without water. Therefore the time of administration can be chosen as the theoretically best time for maximal effect irrespective of logistics (for instance, presence of IV cannula) and preoperative fasting. Piroxicam has a long plasma half-life of about 50 h and reaches its peak serum levels after 3 to 5 h after oral administration (5,6). The pharmacokinetic profile of piroxicam FDDF is similar to that of the oral route as a result of ingestion of the saliva (7). However, an earlier increase in serum levels during the first 30 min after administration has been described with the FDDF and piroxicam absorption by oral mucosa in rats (8). In the present study, we investigated the use of s.l. administered piroxicam for preemptive analgesia in a prospective, randomized, double-blind study design:
Anesth Analg 2006;102:7558

755

756

AMBULATORY ANESTHESIA GRAMKE ET AL. PREEMPTIVE ANALGESIA WITH SUBLINGUAL PIROXICAM

ANESTH ANALG 2006;102:7558

patients received piroxicam either preoperatively (with postoperative placebo) or postoperatively (with preoperative placebo). We hypothesized that preoperative administration of piroxicam s.l. would reduce postoperative pain and opioid requirements more than would postoperative piroxicam.

Table 1. Demographic Data Group PRE Male/female Age (yr) Weight (kg) 21/4 56 15 75 11 Group POST 23/4 54 18 73 12

Values are mean standard deviation or n.

Methods
After approval by the local IRB and after obtaining written informed patient consent, 52 patients (ASA physical status III) were enrolled in this prospective, randomized, double-blind study. Patients were eligible for the study if scheduled for laparoscopic treatment of bilateral inguinal hernia under general anesthesia. A hospitalization of at least 48 h had to be anticipated for inclusion. Excluded were patients 18 yr or older than 65 yr of age and those with a known hypersensivity to NSAIDs, a history of gastric ulcer, renal insufficiency, migraine, or postoperative nausea and vomiting. When patients met the inclusion criteria, they were enrolled in the study. Randomization was performed by sealed numbered envelopes. One group of patients (group PRE) received piroxicam 40 mg s.l. 2 h before operation and 10 min postoperatively received a placebo. The other group (group POST) received a placebo 2 h preoperatively and 10 min postoperatively received piroxicam 40 mg s.l. General anesthesia was induced by IV administration of 2 mg/kg propofol, 1.5 g/kg fentanyl, and 0.5 mg/kg atracurium. Patients were tracheally intubated and mechanically ventilated. Anesthesia was maintained with continuous infusion of propofol (6-10 mg kg1 h1), additional boluses of 50 g fentanyl if necessary and additional neuromuscular blocking drug as required until 20 min before the anticipated end of the intervention. Residual neuromuscular blocking drug was antagonized at the conclusion of surgery with neostigmine (0.05 mg/kg) and 1 mg of atropine. Wounds were not infiltrated with local anesthetic. After surgery when the second s.l. dose of study medication was administered an initial dose of 100 mg of tramadol was given IV over a period of 30 min to all patients. Afterwards postoperative analgesia was supplied by IV tramadol, using patient-controlled analgesia (PCA). Settings were as follows: bolus dose 20 mg, lockout period 20 min, 4-h maximum dose of 200 mg. No basal infusion rate was used. Nausea and vomiting were recorded and treated with 4 mg ondansetron IV every 6 h if needed. The morning after the operation all patients received 40 mg piroxicam s.l. Intensity of postoperative pain was measured using a 10-cm visual analog scale

(VAS), anchored at no pain and worst pain I can imagine. All VAS scoring was done with the patient at rest. In the recovery room, and at 6, 20, and 30 h postoperatively, a VAS score was recorded. VAS scoring in the recovery room was done when patients had sufficiently recovered, the initial dose of tramadol had already been administered at that time. The scoring was performed by a blinded investigator or blinded nursing staff trained in scoring postoperative pain with VAS. The cumulative PCA tramadol use was recorded. The primary outcome of this study was the VAS score 6 h after the operation. Assuming a standard deviation of 2.3, 21 patients per group were required to detect a difference between the preoperative and postoperative administration group of 2 points with a power of 80% and a significance level 0.05 (2-sided). The data are presented as median and interquartile range. The Mann-Whitney U-test was used to compare the VAS scores and the cumulative use of tramadol between group PRE and group POST. A P value 0.05 was considered statistically significant.

Results
Fifty-two patients were enrolled in the study. There were 25 patients in group PRE (piroxicam 2 h preoperatively, placebo 10 min postoperatively) and 27 patients in group POST (placebo 2 h preoperatively, piroxicam 10 min postoperatively). The demographic data for the two groups were similar (Table 1). All interventions lasted 1.5 h ( 10 min) and were conducted by the same surgeon. VAS pain data are presented in Table 2. There were no differences between the groups, except at 6 h (P 0.029) and 20 h (P 0.030) postoperatively, when the VAS score was significantly higher in group POST than in group PRE. Further, Table 2 shows the cumulative consumption of tramadol by PCA excluding the initial dose. At 30 h postoperatively the cumulative dose of tramadol was significantly larger in group POST (P 0.045). The incidence of nausea was similar in both groups (4 patients in group PRE and 3 patients in group POST).

Discussion
Our results indicate that administration of piroxicam s.l. 2 hours before surgery reduces pain scores at 6

ANESTH ANALG 2006;102:7558

AMBULATORY ANESTHESIA GRAMKE ET AL. PREEMPTIVE ANALGESIA WITH SUBLINGUAL PIROXICAM

757

Table 2. VAS Scores and Cumulative Tramadol Consumption at Different Time Points Recovery VAS score (cm) Group PRE (n25) VAS score (cm) Group POST (n27) Cumulative tramadol group PRE (n25) Cumulative tramadol Group POST (n27) 1.0 (02.5) 1.0 (1.02.5) NA NA 6h 1.5 (0.43.2) 3.5* (1.24.2) 40 mg (2590 mg) 60 mg (40100 mg) 20 h 1.1 (0.22.4) 2.0* (1.03.0) 120 mg (40185 mg) 140 mg (80240 mg) 30 h 0.2 (01.4) 1.9 (02.3) 135 mg (60200 mg) 200 mg* (115320 mg)

Data are presented as median and interquartile range. VAS visual analogue scale; Cumulative tramadol cumulative tramadol consumption by patient controlled analgesia (excluding the 100-mg initial loading dose); NA not applicable. *P 0.05.

hours and at 20 hours after surgery compared with piroxicam given postoperatively. In addition, less cumulative opioid consumption by PCA was found in the group receiving piroxicam s.l. preoperatively at 30 hours after surgery. These results support the hypothesis of a preemptive action of preoperatively delivered piroxicam s.l. The concept of preemptive analgesia stresses the prevention of central sensitization processes (9). In this context, the long half-life of piroxicam appears beneficial. The preoperative administration of piroxicam provides antinociceptive treatment before the noxious stimulus, and it is still effective in the postoperative period, when the inflammatory response at the site of surgery can generate noxious inputs. We chose to combine piroxicam with tramadol, an analgesic with both opioid and non-opioid sites of action. It affects not only the receptor but also norepinephrine and serotonin pathways. There is evidence that tramadol combined with piroxicam provides better analgesia than each of the products used alone (10). When tramadol is used for PCA an IV initial loading dose is usually used before starting PCA (11). In this study we administered an IV initial dose of 100 mg of tramadol after surgery and before the first VAS score was recorded. Early administration of a loading dose of tramadol (even before emergence from general anesthesia) has been described in the literature (12). This technique does not conflict with a preemptive analgesic approach. However, as a result of this study design, we recorded very low VAS scores in both groups immediately postoperatively. Furthermore, very low VAS scores were found at 30 hours, which may reflect the administration of piroxicam to all patients on the day after the operation. We found statistically significant differences between groups, but the clinical relevance of these differences is questionable. Both groups had only mild postoperative pain, as described by the excellent VAS scores. Most striking are the extremely low VAS scores measured in the recovery in both groups. Considering the time of administration of piroxicam and the initial dose of tramadol and the pharmacokinetic profile of both

drugs, we hypothesize that these very low initial VAS scores can be attributed to the initial loading dose of tramadol. A smaller initial dose of tramadol would have probably interfered less with the evaluation of the analgesic effects of piroxicam. OHanlon et al. (1) reported the effectiveness of tenoxicam as a preemptive analgesic, when used as part of a multimodal postoperative analgesia regimen. Indeed, the combination of a procedure of short duration in which tissue damage is transmitted mainly by somatic afferent fibers, analgesia with preoperative tenoxicam IV and postsurgical infiltration using longacting local anesthetics, as well as assessment of pain and analgesic consumption in the early postoperative period, were mentioned in the accompanying editorial as essential factors in the observed effect (13). However, our study suggests that effective preemptive analgesia can also be obtained under different conditions. In our study, the wound was not infiltrated with local anesthetics, and the procedure was of a medium duration. We cannot predict if wound infiltration would have masked the preemptive effect of piroxicam or might have made the analgesic regimen even more effective. There are only a few reports regarding postoperative analgesia with sublingual administration of NSAIDs (5,7). Compared with IV administration, the FDDF s.l. has some advantages. The tablet melts in contact with saliva and no water is necessary. The analgesic action of piroxicam reaches its full efficacy 1 hour after administration and can last up to 24 hours. The time to onset of antiinflammatory action is 2 hours (14). Thus, piroxicam administered 2 hours before the procedure will exert an antiinflammatory effect at the time of surgical trauma, which is likely to be part of its analgesic effect. This effect will, of course, be absent when the compound is given postoperatively. The long half-life of piroxicam (50 hours) results in relatively stable plasma concentrations throughout the day at once-daily doses. In conclusion, our findings suggest that s.l. piroxicam FDDF given 2 hours preoperatively statistically

758

AMBULATORY ANESTHESIA GRAMKE ET AL. PREEMPTIVE ANALGESIA WITH SUBLINGUAL PIROXICAM

ANESTH ANALG 2006;102:7558

improves postoperative analgesia after bilateral inguinal hernia repair compared with s.l. piroxicam FDDF administered postoperatively. Considering the low pain scores in both groups and only minor improvements in absolute VAS scores, the clinical importance of these findings is not clear. Nevertheless, s.l. drug administration may be a useful approach for providing postoperative analgesia.

References
1. OHanlon DM, Thambipillai T, Colbert ST. Timing of preemptive tenoxicam is important for postoperative analgesia. Can J Anaesth 2001;48:1626. 2. Campbell WI, Kendrick R, Patterson C. Intravenous diclofenac sodium: does its administration before operation suppress postoperative pain? Anaesthesia 1990;45:7636. 3. Flath RK, Hicks ML, Dionne RA, Pelleu GB Jr. Pain suppression after pulpectomy with preoperative flurbiprofen. J Endod 1987; 13:33947. 4. Crocker S, Paech M. Preoperative rectal indomethacin for analgesia after laparoscopic sterilisation. Anaesth Intensive Care 1992;20:33740. 5. Pookarnjanamorakot C, Laohacharoensombat W, Jaovisidha S. The clinical efficacy of piroxicam fast-dissolving dosage form for postoperative pain control after simple spine surgery: a double-blinded randomized study. Spine 2002;27:44751.

6. Stoelting RK Nonsteroidal antiinflammatory drugs. In: Stoelting RK, ed. Pharmacology & physiology in anesthetic practice, 3rd ed. Philadelphia: Lippincott-Raven, 1999:24758. 7. OHanlon JJ, Muldoon T, Lowry D, McCleane G. Improved postoperative analgesia with preoperative piroxicam. Can J Anaesth 1996;43:1025. 8. Diez-Ortego I, Cruz M, Largo R. Studies of piroxicam absorption by oral mucosa. Arzneim-Forsch/Drug Res 2002;52: 3857. 9. Ong CKS, Lirk P, Seymour RA, Jenkins BJ. The efficacy of preemptive analgesia for acute postoperative pain management: a meta-analysis. Anesth Analg 2005;100:75773. 10. Lauretti GR, Mattos AL, Lima IC. Tramadol and betacyclodextrin piroxicam: effective multimodal balanced analgesia for the intra- and postoperative period. Reg Anesth 1997;22: 243248. 11. Karamanhog lu B, Turan A, Memis D. Preoperative oral rofecoxib reduces postoperative pain and tramadol consumption in patients after abdominal hysterectomy. Anesth Analg 2004;98: 103943. 12. Pang WW, Hurng-Shengh W, Chien-Chiung T. Tramadol 2.5 mg/kg appears to be the optimal intraoperative loading dose before patient-controlled analgesia. Can J Anaesth 2003;50: 4851. 13. Katz J. Pre-emptive analgesia: importance of timing. Can J Anaesth 2001;48:10514. 14. Mather LE. Do the pharmacodynamics of the nonsteroidal antiinflammatory drugs suggest a role in the management of postoperative pain. Drugs 1992;44 Suppl 5:112.