Effects of Prophylactic Nalmefene on the Incidence of Morphine-related Side Effects in Patients Receiving Intravenous Patient-controlled Analgesia 

After we received institutional review board approval, we enrolled 120 consenting women (classified as American Society of Anesthesiologists physical status I, II, or III) who were scheduled for lower abdominal surgery. Patients with significant cardiovascular, pulmonary, renal, or hepatic dysfunction, and those with a history of allergy to opioids or their antagonists and a history of drug or alcohol dependence were excluded from the study. All patients were given instructions regarding the use of the PCA system.

After premedication with 2 mg intravenous midazolam, general anesthesia was induced with 3-5 mg/kg intravenous thiopental, and tracheal intubation was facilitated using 0.6 mg/kg intravenous rocuronium. Anesthesia was maintained with desflurane (3 or 4%) in a mixture of 60% nitrous oxide in oxygen and intermittent bolus doses of fentanyl or sufentanil. Heart rate, arterial blood pressure (noninvasive), pulse, oximetry, temperature, end-tidal carbon dioxide, and end-tidal desflurane concentrations were monitored. No prophylactic antiemetic agents were administered. At the end of the surgery, all patients received neostigmine, 50 [micro sign]g/kg, and 10 [micro sign]g/kg intravenous glycopyrrolate to antagonize neuromuscular blockade. When spontaneous breathing resumed, 1 or 2 mg intravenous morphine was administered, as necessary, to achieve a respiratory rate of 12-15 breaths/min. At the conclusion of the surgery, patients were assigned randomly, using a computer-generated random-number sequence, to receive saline (group 1), 15 [micro sign]g nalmefene (group 2), or 25 [micro sign]g nalmefene (group 3) intravenously. A blinded person who was not involved in data collection administered the study drug.

In the postanesthesia care unit (PACU), patients who reported pain received morphine (2-5 mg intravenously every 5 min) until they were pain free. After this “loading dose,” IV-PCA morphine was started. Patients could self-administer 3 mg morphine (the demand dose) every 10 min (the lockout interval). If analgesia was inadequate, the lockout time was decreased or a background infusion of 0.5-1 mg/h morphine was started, or both. The total dose of morphine used during the 24-h study was recorded.

The episodes of nausea or vomiting and the need for treatment were recorded by a blinded observer 30 min after the patients arrived in the PACU. The severity of pain was assessed, 30 min after arrival in the PACU, using a four-point categoric scale (0 = none, 1 = slightly uncomfortable, 2 = moderate but tolerable, 3 = severe and intolerable). Rescue medications for nausea, vomiting, or both were administered if nausea lasted for more than 15 min or if at any time the patients requested them. The rescue medications for nausea and vomiting consisted of 0.625-1 mg droperidol, 4 mg ondansetron, or 5-10 mg prochlorperazine given intravenously. Rescue medication for pruritus was 25-50 mg diphenhydramine given intravenously. The need for antiemetic and antipruritic medications during the 24-h study period was recorded. In addition, patient recollection of nausea, vomiting, or itching in the previous 24 h was noted. Furthermore, a 10-cm visual analog scale was used to assess retrospectively the overall severity of pain (0 and 10 labeled as “no pain” and “worst pain imaginable”) and the overall pain relief (0 and 10 labeled as “no pain relief” and “complete pain relief”) at 24 h.

An SAS computer package (SAS Institute, Cary, NC) was used for all statistical analyses. For the purpose of analysis, for patients receiving fentanyl, the total amount of fentanyl used was converted to analgesic equivalents of sufentanil (1 [micro sign]g sufentanil equivalent to 10 [micro sign]g fentanyl). Continuous variables (e.g., demographic data, duration of anesthesia, and intraoperative and postoperative opioid use) were analyzed using an analysis of variance test with Bonferroni correction for post hoc comparisons. Frequency distributions were determined for each of the nonparametric outcome measurements (e.g., nausea, vomiting, pruritus, pain) by group. Treatment groups were compared using chi-square contingency Table analysisand likelihood ratio chi squares. If the overall likelihood ratio chi-square was significant, the overall chi-square was partitioned to evaluate pairwise differences between groups. [7]A P value less than 0.05 was considered significant.

One hundred twenty patients participated in the study, with 40 patients enrolled in each of three groups. One patient in group 2 was excluded from the statistical analyses because of protocol violation. There were no significant differences among the groups with respect to age, weight, height, duration of anesthesia, and use of intraoperative opioids (Table 1).

There were no significant differences between the groups in the number of patients reporting nausea and vomiting and the need for antiemetic treatment in the PACU (Table 2). Although there was no difference in the overall incidence of pain among the groups, a significantly more patients receiving nalmefene 25 [micro sign]g reported severe pain in the PACU (Table 2).

The need for antiemetic and antipruritic medications during the 24-h study period was significantly less in the patients receiving nalmefene compared with those receiving placebo (Table 3). However, there was no difference between the nalmefene groups. Significantly more patients in the placebo group recalled being nauseated and remembered reporting itching compared with the patients in the nalmefene groups (Table 3). There was no difference in the number of patients who remembered vomiting during the 24-h study period (Table 3). Although the total consumption of morphine during the 24-h study period was similar in the three groups, the overall 24-h pain scores were significantly less in the patients receiving nalmefene (Table 3). In addition, the overall 24-h pain relief scores were significantly greater in patients receiving nalmefene compared with those receiving placebo.

The incidence of postoperative nausea and vomiting (PONV) is predictably high with the use of opioids. Interestingly, most patients consider PONV to be more distressing and debilitating than pain. [8]Indeed, PONV was cited by 71% of patients as the reason for the poor rating they gave their perioperative experiences. [9]In addition to reducing patient satisfaction, PONV may delay recovery and increase the duration of hospital stay. In a study evaluating morphine-related side effects in patients receiving placebo, 0.25 [micro sign]g [middle dot] kg^-1[middle dot] h^-1naloxone, or 1 [micro sign]g [middle dot] kg^-1[middle dot] h^-1naloxone in conjunction with IV-PCA, the incidences of nausea were 80, 45, and 35%, respectively. [2]Semple et al. [10]reported PONV in as many as 88% of patients receiving IV-PCA with morphine after abdominal hysterectomy. Similarly, Gan et al. [11]reported the incidence of PONV with IV-PCA with morphine as much as 80% without administration of prophylactic antiemetic agents.

In this study, the prophylactic administration of nalmefene significantly reduced the need for antiemetics during the 24-h study period. However, the absence of antiemetic effects of nalmefene in the PACU may be related to the timing of the administration of nalmefene. It is possible that if nalmefene had been administered 15-30 min before the end of surgery, the incidence of nausea in the PACU may have been reduced. Several studies have evaluated the antiemetic effectiveness of droperidol, promethazine, or metoclopramide added to IV-PCA morphine. [1]However, these drugs have a minimal effect on PONV. In addition, they may be associated with side effects. Increased sedation and extrapyramidal reactions have been reported as late as 24 h after small doses of droperidol were given. [12]Furthermore, the addition of antiemetic drugs to morphine or the use of a naloxone infusion increases the complexity of IV-PCA therapy.

The addition of droperidol to morphine PCA after a prophylactic dose at induction has been shown to reduce the degree of nausea and the incidence of requests for rescue antiemetic, but it also increases the degree of sedation. [13]In contrast, the addition of droperidol (5 mg) in PCA morphine (30 mg) after a single dose of droperidol (1.25 mg) at the end of surgery resulted in a greater degree of sedation but did not improve the antiemetic effect. [11]A recent study showed that the simultaneous titration of promethazine (0.625 mg/PCA dose, an average of 17.6 mg/24 h) and morphine decreased nausea associated with PCA therapy. [14]The addition of metoclopramide (0.5 mg/ml) to morphine reduced the incidence of severe PONV during the first 6 h, but the overall incidence of PONV was not affected. [15]Use of transdermal hyoscine with PCA morphine reduced the incidence of PONV only minimally (from 88% to 78%). [10] 

The incidence of pruritus in patients receiving IV-PCA morphine after abdominal hysterectomy was reported to be 55%, compared with 20-25% in patients receiving a naloxone infusion in conjunction with IV-PCA. [2]In a study of women undergoing cesarean section, the incidence of pruritus while they received IV-PCA was 60%; however, only 5% of these patients required treatment. [16]In a similar patient population, Harrison et al. [17]reported a 35% incidence of pruritus, with 11% of patients requiring treatment. In this study, nearly 23% of patients in the placebo group needed antipruritic treatment, and prophylactic administration of nalmefene significantly decreased the need for treatment to 5%.

The patients in the three groups self-administered similar doses of morphine during the 24-h study period. However, the patients receiving nalmefene characterized their pain retrospectively as less severe in the previous 24 h (i.e., lower pain scores and higher pain relief scores) than did those who received placebo. This supports the observations by previous investigators that the end point chosen by patients using IV-PCA opioids is not necessarily total pain relief, but rather a balance between analgesia and the side effect profile. [18]It is also suggested that, frequently, pain and itching are closely associated. Therefore, the higher incidence of pruritus in the placebo group may have resulted in the higher pain scores. [19] 

Several investigators reported an analgesic effect of low-dose naloxone. [2,20]It is possible that low doses of nalmefene may also have some analgesic effects. In contrast to the findings in this study, Gan et al. [2]observed no difference in the pain scores in patients receiving placebo or naloxone infusion; however, morphine use was significantly less in the patients receiving low-dose naloxone infusion compared with those receiving placebo. The 24 h assessment may have represented the severity of pain at that specific time rather than the overall degree of pain during the previous 24 h. However, this element of bias should have been similar in the three study groups. Nevertheless, more data collection time points would have better clarified the effects of nalmefene on the intensity of postoperative pain.

Until recently, naloxone was the only pure opioid antagonist available. Its onset of action after intravenous administration is rapid (1 or 2 min), but its duration of action is relatively short (30-60 min). Therefore, it must be administered as an infusion to achieve a prolonged effect. In a placebo-controlled study, Gal and DiFazio [4]reported that nalmefene could antagonize the analgesia and respiratory depression associated with opioid administration during a prolonged period of time. In addition, increasing the dose of nalmefene was effective in further prolonging its duration of action. Konieczko et al. [5]reported similar conclusions using 0.4 mg nalmefene given intravenously. These authors also reported that nalmefene was four times more potent than naloxone at the [micro sign]- and [Greek small letter kappa]-opioid receptors. [5]However, Glass et al. [6]reported that nalmefene is equipotent to naloxone but had a longer duration of action after a single intravenous dose. The doses of nalmefene (15 and 25 [micro sign]g) used in this study were chosen based on the findings of previous reports with naloxone (0.25 [micro sign]g [middle dot] kg^-1[middle dot] h^-1)[2]and assume that nalmefene is as potent as naloxone.

Although high doses of nalmefene have been used safely, [3,21]administration of 75 [micro sign]g nalmefene to reverse postoperative respiratory depression has been reported to cause acute pulmonary edema. [22]The other possible side effects associated with opioid antagonists include recurrence of postoperative pain, tachycardia, hypertension, dizziness, headache, nausea, and vomiting. Therefore, nalmefene should be used with caution in patients with significant cardiovascular dysfunction. The initial recommended dose to reverse postoperative opioid-related respiratory depression is 0.1 [micro sign]g/kg, with a maximum dose of 1 [micro sign]g/kg. [23]Use of higher doses is likely to increase the incidence and severity of side effects.

In conclusion, in patients receiving IV-PCA with morphine, prophylactic administration of nalmefene decreased the need for antiemetic and antipruritic medications without affecting the quality of analgesia.

The authors thank Joan Reisch, Ph.D., Director, Academic Computing Services, University of Texas Southwestern Medical Center at Dallas, for statistical analysis and interpretation.