The ideal drug therapy for postoperative

Transcription

1 Emerging pharmacologic options for treating postoperative ileus The ideal drug therapy for postoperative ileus (POI), a common surgical complication, would have several characteristics. It would selectively antagonize the inhibitory effects of all the potential factors implicated in the pathophysiology of POI (neurogenic, inflammatory, hormonal, and pharmacologic mediators) in the gastrointestinal (GI) tract. Currently, the most promising target is inhibition of the effects of endogenous and exogenous opioids in the GI tract. Mu-opioid receptors in the GI tract mediate reduced motility in surgical patients receiving postoperative opioid analgesic therapy. 1 However, the ideal agent should not penetrate the blood brain barrier, where antagonism of central opioid receptors would interfere with analgesia, and should also be free of agonist activity at µ-opioid receptors in GI smooth muscle. In addition, the ideal drug therapy for POI would be available in oral and injectable dosage forms for flexibility of administration. Preferably, to minimize potential systemic adverse effects, oral dosage forms would not be systemically absorbed. Furthermore, the Michael D. Kraft Purpose. Characteristics of the ideal drug therapy for postoperative ileus (POI); the pharmacology, efficacy, and safety of currently available nonselective opioid antagonists and the new peripherally selective opioid antagonists methylnaltrexone and alvimopan for the treatment of POI; and formulary considerations associated with the introduction of these new POI drug therapies are discussed. Summary. The ideal drug therapy for treating POI would selectively antagonize the inhibitory effects on the gastrointestinal (GI) tract of all of the potential factors implicated in the pathophysiology of POI (neurogenic, inflammatory, hormonal, and pharmacologic mediators). The most promising target to date is inhibition of the adverse GI effects of endogenous and exogenous opioids. Selective inhibition of the µ-opioid receptors in the GI tract, without reversing centrally mediated opioid-induced analgesia, may be beneficial in reducing POI. The nonselective opioid antagonists naloxone and nalmefene have not been studied for POI, and they cross the blood brain barrier. Therefore, they are not appropriate for preventing or treating POI. The peripherally selective opioid antagonist methylnaltrexone shortens the duration of POI and the hospital length of stay (LOS). Alvimopan, a more extensively studied peripherally selective opioid antagonist, has been shown to reduce the duration of POI, frequency of postoperative nausea and vomiting, and hospital LOS. Both methylnaltrexone and alvimopan also appear effective for treating opioidinduced constipation. Preliminary results of a long-term study of alvimopan safety have revealed some potential concerns, and the significance of the adverse effects must be understood before the most appropriate role of alvimopan in patient care can be determined. Restricting the prescribing of new POI drug therapies to certain types of patients, surgeries, and prescribers; incorporating these therapies into preoperative and postoperative policies, procedures, and protocols; and the potential cost savings from reducing hospital LOS are among the considerations in adding these agents to health-system formularies. Conclusion. Peripherally selective opioid receptor antagonists are promising new drug therapies that can reduce the clinical and economic burden of POI. Index terms: Alvimopan; Blood brain barrier; Drugs, body distribution; Economics; Formularies; Hospitals; Ileus; Mechanism of action; Methylnaltrexone; Nalmefene; Naloxone; Opiate antagonists; Opiates; Pharmacokinetics; Postoperative complications; Postoperative nausea and vomiting; Protocols; Toxicity Am J Health-Syst Pharm. 2007; 64 (Suppl 13):S13 20 Michael D. Kraft, Pharm.D., is Clinical Assistant Professor, University of Michigan, College of Pharmacy, and Clinical Coordinator and Clinical Pharmacist, University of Michigan Health System, 1500 East Medical Center Drive, Ann Arbor, MI (mdkraft@umich.edu). Based on the proceedings of a symposium held December 5, 2006, during the ASHP Midyear Clinical Meeting in Anaheim, CA, and supported by an educational grant from Adolor Corporation and GlaxoSmithKline. Dr. Kraft received an honorarium for his participation in the symposium and for the preparation of this article. Dr. Kraft reports that he has served as a consultant and speaker for Adolor Corporation and GlaxoSmithKline. Copyright 2007, American Society of Health-System Pharmacists, Inc. All rights reserved /07/1002-0S13$6.00. DOI /ajhp Am J Health-Syst Pharm Vol 64 Oct 15, 2007 Suppl 13 S13

2 ideal POI treatment would be easy to administer, well tolerated, and inexpensive, and it would not interact with other drugs or disease states. Nonselective opioid antagonists Naloxone and nalmefene are nonselective tertiary opioid antagonists that are currently available. Neither drug is selective for µ-opioid receptors in the GI tract (both drugs cross the blood brain barrier and act at central receptors as well as peripheral receptors). 2 Neither drug is approved by the Food and Drug Administration (FDA) for the treatment of POI. Naloxone can antagonize the effects of opioids in the GI tract; however, it has never been evaluated in a prospective, randomized, controlled trial for the prevention or reduction of POI. In addition, naloxone is not selective for GI opioid receptors and can cross the blood brain barrier and lead to reversal of anesthesia. 2,3 Furthermore, the naloxone dose required to reverse opioid effects is patientspecific and highly variable, and it may cause withdrawal symptoms in opioid-dependent patients. 3,4 Therefore, naloxone should not be used for preventing or treating POI. Nalmefene hydrochloride is a nonselective opioid-receptor antagonist, and like naloxone, it has never been evaluated in a prospective, randomized, controlled trial for the prevention or reduction of POI. Nalmefene glucuronide, an inactive metabolite of nalmefene hydrochloride, has demonstrated evidence of selectivity for GI opioid receptors in animal studies. Because of this, nalmefene glucuronide was evaluated for the reduction of opioid-induced constipation in humans. 5 Ascending oral doses were used in five opioiddependent male volunteers who were maintained on methadone therapy. 5 Nalmefene glucuronide was continued until laxation (i.e., a bowel movement) or withdrawal symptoms occurred. In all five subjects, nalmefene glucuronide precipitated withdrawal symptoms an average of nine hours after its initiation. The investigators surmised that the withdrawal effects were the result of biotransformation of the glucuronide to the parent drug in the colon, which was then systemically absorbed and crossed the blood brain barrier. Because of the lack of prospective data and the potential to reverse analgesia, neither nalmefene hydrochloride nor nalmefene glucuronide should be used for preventing or treating POI. Selective opioid antagonists Two novel quaternary opioid antagonists with selectivity for peripheral (i.e., GI) µ-opioid receptors, methylnaltrexone (Figure 1) and alvimopan (Figure 2), have been developed that might prove useful Figure 1. Chemical Structures of Naltrexone and N-methylnaltrexone Naltrexone Figure 2. Chemical Structure of Alvimopan for the treatment of POI. Neither drug is currently approved by FDA. Methylnaltrexone Methylnaltrexone is a quaternary µ-opioid receptor antagonist that was formed by adding a methyl group to naltrexone. Naltrexone is a nonselective opioid antagonist that crosses the blood brain barrier and can reverse opioid-induced analgesia and cause opioid withdrawal symptoms in opioid-dependent patients. 6 The addition of the methyl group results in the formation of a protonated quaternary ammonium group, thereby increasing the polarity of the molecule and conferring low lipophilicity. Methylnaltrexone does not penetrate the blood brain barrier. 7 The drug is not demethylated to an appreciable extent in humans. 1,8 N-methylnaltrexone + CH 3 S14 Am J Health-Syst Pharm Vol 64 Oct 15, 2007 Suppl 13

3 Oral absorption of methylnaltrexone is limited. 6,9 The elimination half-life is about two to three hours after i.v. administration, 7,10 and approximately 40 50% of an i.v. dose is recovered in the urine (i.e., the kidneys play an important role in elimination of the drug) However, after oral administration, less than 1% of the unchanged parent drug is found in the urine, which is likely due to the limited oral absorption. 9 Steady state is achieved rapidly with repeated i.v. doses (after two to three doses), and accumulation of the drug and alteration in pharmacokinetics do not occur with repeated administration. 10 Methylnaltrexone has been shown to reverse opioid-induced slowing of GI transit time. 9,12 14 It does not antagonize the analgesic effect of opioid analgesics or precipitate withdrawal in opioid-dependent patients. 12,14 The drug is well tolerated, with mild to moderate abdominal cramping but no serious adverse effects. 10,11 Two methylnaltrexone dosage forms have been explored in clinical trials. An injectable dosage form has been evaluated in two Phase III studies of patients with opioid-induced bowel dysfunction and a Phase II study of patients at high risk of developing POI. 15,16 OBD is a spectrum of adverse GI effects associated with opioid use, including constipation, abdominal pain and distention, bloating, gastroesophageal reflux, hard dry stools, staining, and incomplete evacuation. 2 Two Phase I studies of oral uncoated and coated methylnaltrexone dosage forms have been completed in healthy volunteers. 9,13 In a Phase III, randomized, doubleblind, placebo-controlled study, 154 patients with opioid-induced constipation (i.e., no bowel movement for 48 hours) and advanced medical illness were randomly assigned to receive a single 0.15-mg/kg or 0.3-mg/kg dose of methylnaltrexone or placebo subcutaneously (s.c.). 15 Laxation within 4 hours was the primary efficacy endpoint, and laxation within 24 hours was a secondary endpoint. Both methylnaltrexone doses were significantly more effective than placebo in producing laxation within 4 hours and 24 hours after administration (p < for both doses and laxation within 4 hours, p = for 0.15 mg/kg and laxation within 24 hours, and p = for 0.30 mg/ kg and laxation within 24 hours). 15 The percentage of patients with laxation within 4 hours after drug administration was 58% with methylnaltrexone 0.3 mg/kg (the higher dose), 62% with methylnaltrexone 0.15 mg/kg (the lower dose), and 13% with placebo. The percentage of patients with laxation within 24 hours after administration was 64% with the higher methylnaltrexone dose, 68% with the lower methylnaltrexone dose, and 33% with placebo. The median time to laxation was significantly shorter with both methylnaltrexone doses (45 minutes with 0.3 mg/kg and 70 minutes with 0.15 mg/kg) compared with placebo (>24 hours, p < for both doses). The effect of the drug appeared to be dose related. The most common adverse effects from methylnaltrexone were abdominal cramps and flatulence, with rates of 30% to 40% and 15% to 20%, respectively. 15 Opioid withdrawal was not reported. The impact of methylnaltrexone on the duration of POI was evaluated in a Phase II, randomized, doubleblind study of 65 patients who underwent segmental colectomies. 16 Patients were randomly assigned to receive methylnaltrexone 0.3 mg/kg or placebo i.v. every six hours beginning within 90 minutes after surgery and continuing for up to seven days or until no more than 24 hours had elapsed after GI recovery, which was defined as the ability to tolerate solid food and having the first bowel movement. Compared with placebo, methylnaltrexone significantly reduced the average time until the patient could tolerate full liquids (70 hours versus 100 hours with placebo, p = 0.05) and the time to the first bowel movement (97 hours versus 120 hours with placebo, p = 0.01). 16 Methylnaltrexone also shortened the average time until the patient could tolerate solid foods (100 hours versus 125 hours with placebo) and the time to GI recovery by about one full day (124 hours versus 151 hours), although these differences were not significant. Patients treated with methylnaltrexone were eligible for hospital discharge significantly sooner (after 119 hours on average) than patients treated with placebo (149 hours, p = 0.03). The mean time to actual discharge was about one full day shorter in the methylnaltrexone group (140) hours than in the placebo group (165 hours), a difference that was not significant. Methylnaltrexone was well tolerated, with no serious adverse effects. 16 There were no differences in pain scores between the two groups. Four methylnaltrexone-treated patients and nine placebo-treated patients withdrew from the study. The most commonly reported adverse effects were fever and nausea, and the frequency of these effects was similar in the two groups. The methylnaltrexone-associated shortening of the duration of POI and hospital length of stay (LOS) by about one day in this study is clinically significant and could translate into substantial cost savings. 16 Two Phase III, randomized, double-blind, placebo-controlled, parallel-group studies of i.v. methylnaltrexone in patients with POI are under way. 17,18 One of these studies will enroll 495 adults who have undergone segmental colectomy by open laparotomy. 17 The primary endpoint is the time between the end of surgery and the first bowel movement. August 2007 was the anticipated completion date Am J Health-Syst Pharm Vol 64 Oct 15, 2007 Suppl 13 S15

4 of the study. In July 2006, methylnaltrexone received fast-track designation from FDA, which expedites the regulatory review process. 19 Alvimopan Alvimopan (formerly known as ADL ) is a quaternary µ-opioid receptor antagonist with a pharmacologic profile that differs from that of methylnaltrexone. Alvimopan has considerably greater binding affinity for µ-opioid receptors than for d- and k-opioid receptors. 20 Methylnaltrexone binds preferentially to µ-opioid receptors, but it is not highly selective for µ-opioid receptors compared with d- and k-opioid receptors. 21 Alvimopan has higher binding affinity for µ-opioid receptors than does methylnaltrexone. 21 The clinical significance of these differences remains to be determined. Alvimopan has a moderately large molecular weight and low lipophilicity, so it does not cross the blood brain barrier. 20 The drug has limited oral bioavailability (about 6% in humans) and a short half-life of about hours after oral administration of a single dose. 22 The major route of excretion is fecal; the kidneys play a negligible role in elimination of the drug after oral dosing. 23 Alvimopan reverses opioidinduced slowing of GI transit time and constipation without antagonizing the analgesic effect of opioid analgesics or precipitating withdrawal in opioid-dependent patients. 22,24 26 It has been studied in patients with opioid-induced bowel dysfunction (OBD) and patients at high risk of developing POI Opioid-induced bowel dysfunction. The use of alvimopan for the treatment of OBD was explored in several Phase II and Phase III studies. One Phase III randomized, placebo-controlled study involved 168 patients with OBD (defined as less than three bowel movements per week) who had been receiving opioid analgesic therapy (the equivalent of at least 10 mg/day of oral morphine) for at least one month for nonmalignant pain or opioid dependence. 26 Patients were randomly assigned to receive a single daily dose of alvimopan 0.5 or 1 mg or placebo for 21 days. The average percentage of patients having at least one bowel movement within eight hours after study drug administration each day during the 21-day treatment period was significantly higher with alvimopan 1 mg (54%) and 0.5 mg (43%) than placebo (29%, p < 0.001). The effect of alvimopan appeared to be dose related. The median time to the first bowel movement was significantly shorter after the first dose of alvimopan 1 mg compared with placebo (3 hours versus 21 hours, respectively, p < 0.001). The median time to first bowel movement was also shorter after the first dose of alvimopan 0.5 mg (7 hours), but this was not statistically significant. Alvimopan was well tolerated; 18 of 54 (33%) of patients in the placebo group, 22 of 58 (38%) of patients in the 0.5-mg group, and 27 of 56 (48%) of patients in the 1-mg group experienced at least one of the most common adverse effects. 26 The most common adverse effects were abdominal cramping, nausea, vomiting, diarrhea, and flatulence. The authors described most of these effects as mild to moderate in severity and were transient, resolving within one week. Alvimopan did not antagonize opioid-induced analgesia. Postoperative ileus. The efficacy and safety of alvimopan for the treatment of POI have been evaluated in several Phase III, randomized, double-blind, placebo-controlled trials involving patients who underwent major abdominal surgery for bowel resection or hysterectomy In three of these studies, alvimopan 6 mg and 12 mg or placebo was given orally at least two hours before surgery and twice daily after surgery for up to seven days or until hospital discharge The time to GI recovery, a composite measure of the time to toleration of solid food or passage of flatus or stool (whichever occurred later), was used as the primary efficacy endpoint, and this endpoint was known as GI-3. Because the time to passage of flatus is subjective and reporting may be unreliable, a composite measure of the time to toleration of solid food or passage of stool was used as a secondary efficacy endpoint (GI-2). The time to readiness for hospital discharge, time to writing of the discharge order, visual analogue scale pain scores, daily postoperative opioid analgesic consumption, frequency of postoperative nasogastric (NG) tube reinsertion because of failure to resume normal GI function, and safety also were evaluated. Adult men and women (at least 18 years old) who were scheduled to undergo major abdominal surgery as described above were enrolled in the studies Patients who had recently received opioid analgesics (within the previous one to four weeks, depending on the study protocol), patients currently using nonsteroidal antiinflammatory drugs or epidural opioid analgesics or local anesthetics, and patients with severe concomitant diseases were excluded. All patients received standardized postoperative care, with opioid analgesics administered by patient-controlled analgesia, removal of the NG tube at the end of surgery, oral liquids offered and ambulation encouraged on postoperative day 1, and solid food offered on postoperative day 2. In one of the Phase III POI studies involving 451 patients who underwent partial colectomies or simple or radical hysterectomies, alvimopan 6 mg significantly reduced the time to endpoint GI-3 from an average of hours with placebo to 86.2 hours with alvimopan 6 mg, a difference of about 14 hours (hazard ratio [HR], 1.45; p = 0.003). 27 Compared with placebo, alvimopan 6 mg also significantly reduced the time to endpoint GI-3 among a subset S16 Am J Health-Syst Pharm Vol 64 Oct 15, 2007 Suppl 13

5 of patients undergoing partial colectomy or radical hysterectomy by 16.9 hours (HR, 1.51; p = 0.004) and the time to endpoint GI-2 in all patients by 15 hours (HR, 1.46; p = 0.007). The time to reach these outcomes in the alvimopan 12-mg group tended to be lower than in the placebo group but did not reach statistical significance. The time to writing of the discharge order was significantly shorter in the alvimopan 6-mg group (108 hours) compared with the placebo group (122 hours, p = 0.033), while the time in the alvimopan 12-mg group (115 hours) was shorter than in the placebo group but did not reach statistical significance. The time to readiness for hospital discharge was significantly shorter with both alvimopan 6 mg (97 hours, p < 0.001) and 12 mg (98 hours, p = 0.004) than with placebo (112 hours). Opioid analgesic consumption was similar in the alvimopan 6-mg and placebo groups and was significantly higher in the alvimopan 12-mg group compared with placebo (p = 0.045). Pain control measured with visual analogue scales was similar in all three groups. The authors described that alvimopan was well tolerated, with similar rates of adverse effects in the placebo and alvimopan groups. 27 The most common adverse effects were nausea and vomiting. The frequency of these effects was slightly lower in the alvimopan 6-mg group and 53% lower in the alvimopan 12-mg group compared with the placebo group (suggesting that the reduction in nausea and vomiting from alvimopan was dose dependent). In the alvimopan 12-mg group, the rates of vomiting (15%) and constipation (6.8%) were significantly lower than in the placebo group (32% and 14.4%, respectively). The percentage of patients withdrawing from the study because of adverse effects was similar with placebo (15%) and alvimopan 12 mg (16%) and lower with alvimopan 6 mg (7%). However, the percentage of patients withdrawing from the study for any reason was higher in the alvimopan 12-mg group (27%) than in the placebo group (21%) and the alvimopan 6-mg group (16%). The investigators surmised that the unexpected efficacy results in the alvimopan 12-mg group (i.e., the apparent lack of dose-related efficacy) might be attributed to the high rate of study withdrawal in that group. A previously published Phase III POI study of alvimopan 6 mg, 12 mg, or placebo evaluated 469 patients undergoing large or small bowel resection or radical hysterectomy. 29 There were no significant differences among the groups with respect to the mean age, mean or median duration of surgery, or intraoperative opioid analgesic use. The average daily postoperative opioid analgesic consumption was significantly higher (p values not reported) in the alvimopan 6-mg group (33.6 mg) than in the alvimopan 12-mg (27.1 mg) and placebo groups (27.0 mg), but the investigators did not consider these differences clinically significant. In this study, alvimopan had doserelated effects on the time to both endpoint GI-3 and endpoint GI-2 that were significantly different from placebo. 29 Endpoint GI-3 occurred after an average of 120 hours in the placebo group compared with 105 hours in the alvimopan 6-mg group (p < 0.05 versus placebo) and 98 hours in the alvimopan 12-mg group (p < versus placebo). Similarly, endpoint GI-2 occurred after an average of 133 hours, 113 hours, and 105 hours with placebo, alvimopan 6 mg, and alvimopan 12 mg, respectively (p = for alvimopan 6 mg versus placebo and p < for alvimopan 12 mg versus placebo). The time to writing of the hospital discharge order was 146 hours with placebo, 133 hours with alvimopan 6 mg, and 126 hours with alvimopan 12 mg. The difference between the 12-mg dose and placebo was significant (p = 0.003), but also represents a potential reduction in length of hospital stay of nearly one day, which is also clinically important. Alvimopan was generally well tolerated in this study. 29 The frequency of adverse effects was generally similar among the placebo, 6-mg, and 12-mg groups, although the authors did not describe the severity of any adverse effects. Nausea and vomiting were the most commonly reported adverse effects, and alvimopan appeared to reduce the frequency of nausea in a dose-dependent manner (64% in the placebo group, 61% in the alvimopan 6-mg group, and 55% in the alvimopan 12-mg group). This antiemetic effect of alvimopan may be an added benefit in surgical patients who receive opioid analgesics. The frequency of postoperative NG tube reinsertion was 14.8% in the placebo group, 8.4% in the alvimopan 6-mg group, and 4.8% in the alvimopan 12-mg group. 29 The difference between placebo and the 12-mg dose was significant (p = 0.004). The frequency of POI as a treatmentrelated adverse effect (i.e., prolonged POI) was 15.8% in the placebo group compared with 8.3% in the alvimopan 6-mg group (p = 0.043) and 6.3% in the alvimopan 12-mg group (p = 0.005). The percentage of patients who were rehospitalized for any reason within 10 days after discharge was lower in alvimopan-treated patients (approximately 4%) than in placebotreated patients (approximately 8%). 29 These findings suggest that alvimopan use may allow safe early discharge of patients undergoing major abdominal surgery. An additional Phase IIIb study of alvimopan was completed, and the results were recently reported. 31 This was a randomized, double-blind, placebo-controlled study of 654 patients, but there were several differences between this study protocol and the previous Phase III studies. This study included only patients Am J Health-Syst Pharm Vol 64 Oct 15, 2007 Suppl 13 S17

6 undergoing laparotomy for small or large bowel resection (i.e., patients undergoing hysterectomy were not among the study population); it only evaluated the alvimopan 12-mg dose compared with placebo (i.e., no 6-mg group); and the preoperative alvimopan dose was given minutes before surgery (rather than at least two hours before surgery), and dosing was continued postoperatively for up to seven days or until discharge, as was done in previous studies. The results of this study were consistent with previous Phase III alvimopan studies. There were no significant differences between the two groups in preoperative, intraoperative, or postoperative opioid analgesic consumption. The mean time to achieve the primary endpoint of GI-2 was significantly shorter in the alvimopan group (92 hours) compared with the placebo group (112 hours, p < 0.001). The frequency of POI-related postoperative morbidity (defined as postoperative NG tube reinsertion or POI resulting in prolonged hospital stay or readmission) was significantly lower in the alvimopan group (7%) than the placebo group (14%, p = 0.003). The average hospital stay was significantly shorter in the alvimopan group (5.2 days) than in the placebo group (6.2 days). This finding was significant (p < 0.001), and a one-day reduction in average LOS also represents a clinically significant outcome. If hospital LOS after laparotomy and bowel resection could be reduced, this could result in significant cost savings to the health care system and improved patient satisfaction. Alvimopan was well tolerated; the most commonly reported adverse effects were nausea, vomiting, and abdominal distention. 31 The frequency of nausea and vomiting in the alvimopan group (58% and 14%, respectively) were significantly lower than in the placebo group (66% and 25%, respectively; p = for nausea and p < for vomiting). There was no significant difference between alvimopan and placebo in the frequency of abdominal distention (18% and 20%, respectively). The authors did not describe the severity of the adverse effects. The promising results of this and other Phase III studies suggested that FDA would likely approve the 12-mg dose of alvimopan to reduce POI in patients undergoing laparotomy and bowel resection. However, the FDA issued an approvable letter in November 2006 calling for additional safety data and a risk management plan because of preliminary findings from an ongoing safety study of alvimopan (study 014). 32 This study was a Phase III, blinded, placebo-controlled, long-term (12-month) safety evaluation of alvimopan 0.5 mg twice daily for the treatment of OBD in patients with chronic noncancer pain. A sixmonth interim analysis revealed an increase in the frequency of serious cardiovascular events in alvimopantreated patients compared with placebo-treated patients that was not significant but nevertheless was concerning. The affected patients were at high risk for cardiovascular disease, and the risk of serious cardiovascular events did not appear to be related to the duration of alvimopan use. These cardiovascular events were consistent with epidemiological expectations for the patient population, and at that time the combined results of all patients in the population with chronic disease who received alvimopan did not suggest that they were at higher risk for serious cardiovascular events. This study was recently completed, and the results were reported in a press release. 33 A total of 805 patients were enrolled and randomized in a 2:1manner (i.e., 538 patients in the alvimopan group and 267 in the placebo group). The proportion of patients who experienced serious adverse events was similar between the alvimopan and placebo groups (13% and 11%, respectively), but there was a numerical imbalance in serious cardiovascular events (myocardial infarction [MI] and all cardiovascular events) and neoplasm cases in the alvimopan group. 33 In study 014, 2.6% of patients in the alvimopan group experienced serious adverse events compared with 1.12% in the placebo group. The cardiovascular events occurred in patients with high risk for cardiovascular disease. Seven patients in the alvimopan group had MIs; no patients in the placebo group had an MI. Five of the MIs occurred at two investigational sites and did not appear to be related to the duration of alvimopan treatment. Most of the MIs occurred during the first 12 weeks of treatment. A higher number of neoplasms was also observed in the alvimopan group (15, 2.8%), with four considered serious, compared with the placebo group (2, 0.7%). In addition, the frequency of fractures was higher in patients receiving alvimopan, but this is in contrast to previous studies, which demonstrated a similar or lower rate. Currently, the manufacturers of alvimopan (Adolor and GlaxoSmithKline) have suspended all studies of the drug until the data can be further analyzed and the significance of study results is better understood. The companies complete response to the FDA approvable letter with respect to the use of alvimopan to reduce POI was targeted for the third quarter of Formulary considerations The potential approval by the FDA of methylnaltrexone and alvimopan will prompt pharmacy and therapeutics committees to evaluate these agents for inclusion in healthsystem formularies and consider establishing criteria for the safe and cost-effective use of the drugs. Criteria for patient selection might restrict use of the drugs to patients undergoing bowel resection or possibly other abdominal surgeries associated with a higher risk of POI S18 Am J Health-Syst Pharm Vol 64 Oct 15, 2007 Suppl 13

7 (e.g., total abdominal hysterectomy). These agents may also be approved for the treatment of OBD in patients receiving chronic opioid therapy, but additional data are needed to determine optimal criteria and uses. There are several unanswered questions with respect to the use of methylnaltrexone and alvimopan for reducing POI. It is not clear whether these agents would be beneficial in patients undergoing laparoscopy, a procedure associated with less bowel manipulation, less pain, lower opioid analgesic requirements, and a shorter duration of POI and hospital stay compared with laparotomy In addition, it is not currently known if preoperative dosing of these medications would be required for maximal benefit. All of the studies of alvimopan for reducing POI required preoperative dosing. The ongoing studies with methylnaltrexone will initiate the medication after surgery. The most important issue is the concern about safety with alvimopan. The complete study results need to be analyzed, published, and better understood before the most appropriate use of alvimopan can be determined. If cardiovascular safety remains a concern after the results of study 014 become available, restrictions could be placed on the use of alvimopan in patients with cardiovascular disease. In addition, there is no evidence to date demonstrating precipitation of opioid withdrawal symptoms in opioiddependent patients by either methylnaltrexone or alvimopan. However, the studies for reduction of POI have excluded opioid-dependent patients; the use of these agents for the reduction of POI might be restricted in such patients until additional safety data are available. Institutional preoperative and postoperative policies, procedures, and protocols will need to be modified to ensure the optimal use of a new drug therapy for reducing POI. The timing of drug administration with respect to surgery, route of administration, location of administration in the hospital, and patient convenience should be considerations in modifying policies, procedures, and protocols. The approval of methylnaltrexone and alvimopan would likely have a financial impact on health systems. The acquisiton cost of the drugs is unknown. Restricting use of these drugs to certain types of surgical procedures (e.g., open laparatomy with bowel resection, hysterectomy), prescribers (e.g., GI or gynecologic surgeons), or both could limit the impact on the pharmacy budget of adding either drug to the formulary and could encourage appropriate use for optimal patient outcomes. Substantial cost savings could be realized from the reductions in hospital LOS observed in some clinical trials of methylnaltrexone and alvimopan. 16,31 Whether the shortening of the time to readiness for hospital discharge and the time to writing of the hospital discharge order observed in some alvimopan clinical studies translate into safe reductions in actual LOS remains to be confirmed. 27,29 Conclusion Methylnaltrexone and alvimopan are promising new agents that could become an important component of perioperative care for the treatment of POI after bowel resection and, potentially, other major abdominal surgeries. These therapies may also reduce the economic burden of POI on health systems. References 1. Holzer P. Opioids and opioid receptors in the enteric nervous system: from a problem in opioid analgesia to a possible new prokinetic therapy in humans. Neurosci Lett. 2004; 361: Kurz A, Sessler DI. Opioid-induced bowel dysfunction: pathophysiology and potential new therapies. Drugs. 2003; 63: Liu M, Wittbrodt E. Low-dose oral naloxone reverses opioid-induced constipation and analgesia. J Pain Symptom Manage. 2002; 23: McEvoy GK, ed. Naloxone hydrochloride. In: AHFS Drug Information Bethesda, MD: American Society of Health-System Pharmacists; 2007: Cheskin LJ, Chami TN, Johnson RE et al. Assessment of nalmefene glucuronide as a selective gut opioid antagonist. Drug Alcohol Depend. 1995; 39: Foss JF. A review of the potential role of methylnaltrexone in opioid bowel dysfunction. Am J Surg. 2001; 182(suppl):19S 26S. 7. Foss JF, O Connor MF, Yuan CS et al. Safety and tolerance of methylnaltrexone in healthy humans: a randomized, placebocontrolled, intravenous, ascending-dose, pharmacokinetic study. Clin Pharmacol. 1997; 37: Kotake AN, Kuwahara SK, Burton E et al. Variations in demethylation of N- methylnaltrexone in mice, rats, dogs, and humans. Xenobiotica. 1989; 19: Yuan CS, Foss JF, Osinski J et al. The safety and efficacy of oral methylnaltrexone in preventing morphine-induced delay in oral-cecal transit time. Clin Pharmacol Ther. 1997; 61: Yuan CS, Doshan H, Charney MR et al. Tolerability, gut effects, and pharmacokinetics of methylnaltrexone following repeated intravenous administration in humans. J Clin Pharmacol. 2005; 45: Yuan CS. Clinical status of methylnaltrexone, a new agent to prevent and manage opioidinduced side effects. J Support Oncol. 2004; 2: Yuan CS, Foss JF, O Connor M et al. Methylnaltrexone for reversal of constipation due to chronic methadone use. JAMA. 2000; 283: Yuan CS, Foss JF, O Connor M et al. Effects of enteric-coated methylnaltrexone in preventing opioid-induced delay in oral-cecal transit time. Clin Pharmacol Ther. 2000; 67: Yuan CS, Foss JF, O Connor M et al. Methylnaltrexone prevents morphineinduced delay in oral-cecal transit time without affecting analgesia: a double-blind randomized placebo-controlled trial. Clin Pharmacol Ther. 1996; 59: Thomas J, Lipman A, Slatkin N et al. A phase III double-blind placebo-controlled trial of methylnaltrexone (MNTX) for opioid-induced constipation (OIC) in advanced medical illness (AMI) (abstract 8003). Presented at the American Society of Clinical Oncologists Annual Meeting. Orlando, FL; 2005 May d60f5624ba07fd506fe310ee37a01d/ vgnextoid= 76f8201eb61a7010Vgn VCM100000ed730ad1RCRD&vmview= abst_detail_view&confid=34&abstractid= (accessed 2007 Mar 8). 16. Viscusi E, Rathmell J, Fichera A et al. A double-blind, randomized, placebocontrolled trial of methylnaltrexone (MNTX) for post-operative bowel dysfunction in segmental colectomy patients (abstract A893). Presented at the American Society of Anesthesiologists Annual Meeting. Atlanta, GA; 2005 Oct asaabstracts.com/strands/asaabstracts/ abstract.htm;jsessionid=115f34a273578f2 Am J Health-Syst Pharm Vol 64 Oct 15, 2007 Suppl 13 S19

8 AFCA EDFA68?year=2005&index= 11&absnum=1630 (accessed 2007 Mar 8). 17. Study evaluating IV methylnaltrexone for the treatment of post operative ileus. NCT ?order=2. (accessed 2007 Mar 8). 18. Study of intravenous (IV) methylnaltrexone bromide (MNTX) in the treatment of postoperative ileus (POI). (accessed 2007 Mar 8). 19. Wyeth and Progenics announce methylnaltrexone receives FDA fast track designation for treatment of postoperative ileus. view/7613 (accessed 2007 Mar 8). 20. Zimmerman DM, Gidda JS, Cantrell BE et al. Discovery of a potent, peripherally selective trans-3,4-dimethyl-4-(3-hydroxyphenyl) piperidine opioid antagonist for the treatment of gastrointestinal motility disorders. J Med Chem. 1994; 37: Beattie DT, Cheruvu M, Mai N et al. The in vitro pharmacology of the peripherally restricted opioid receptor antagonists, alvimopan, ADL and methylnaltrexone. Naunyn Schmiedebergs Arch Pharmacol. 2007; May; 375(3): Epub 2007 Mar 6. query.fcgi?orig_db=pubmed&db=pubmed& cmd=search&term=%22naunyn-schmied eberg%27s+archives+of+pharmacology% 22%5BJour%5D+AND+2007%5Bpdat% 5D+AND+Beattie%5Bauthor%5D (accessed 2007 Mar 19). 22. Schmidt WK. Alvimopan (ADL ) is a novel peripheral opioid antagonist. Am J Surg. 2001; 182(suppl 5A): 27S 38S. 23. Camilleri M. Alvimopan, a selective peripherally acting mu-opioid antagonist. Neurogastroenterol Motil. 2005; 17: Liu SS, Hodgson PS, Carpenter RL et al. ADL , a trans-3,4-dimethyl-4- (3-hydroxyphenyl) piperidine, prevents gastrointestinal effects of intravenous morphine without affecting analgesia. Clin Pharmacol Ther. 2001; 69: Taguchi A, Sharma N, Saleem RM et al. Selective postoperative inhibition of gastrointestinal opioid receptors. N Engl J Med. 2001; 345: Paulson DM, Kennedy DT, Donovick RA et al. Alvimopan: an oral, peripherally acting, µ-opioid receptor antagonist for the treatment of opioid-induced bowel dysfunction a 21-day treatment-randomized clinical trial. J Pain. 2005; 6: Delaney CP, Weese JL, Hyman NH et al. Phase III trial of alvimopan, a novel, peripherally acting, mu opioid antagonist, for postoperative ileus after major abdominal surgery. Dis Colon Rectum. 2005; 48: Viscusi ER, Goldstein S, Witkowski T et al. Alvimopan, a peripherally acting mu-opioid receptor antagonist, compared with placebo in postoperative ileus after major abdominal surgery. Surg Endosc. 2006; 20: Wolff BG, Michelassi F, Gerkin TM et al. Alvimopan, a novel, peripherally acting (opioid antagonist: results of a multicenter, randomized, double-blind, placebo-controlled, phase III trial of major abdominal surgery and postoperative ileus. Ann Surg. 2004; 240: Herzog TJ, Coleman RL, Guerrieri JP et al. A double-blind, randomized, placebocontrolled phase III study of the safety of alvimopan in patients who undergo simple total abdominal hysterectomy. Am J Obst Gyn. 2006; 195: Leslie J, Steinbrook R, Viscusi E et al. Alvimopan oral dosing minutes before and BID after bowel resection accelerates GI recovery (abstract A122). Presented at the American Society of Anesthesiologists Annual Meeting. Chicago, IL; 2006 Oct asaabstracts/searchresults.htm;jsessionid=1 15F34A273578F2AFCA EDFA68? base=1&index=1&display=10&highlight= true&highlightcolor=0&bold=true&italic= false (accessed 2007 Mar 9). 32. GlaxoSmithKline press release. FDA issues approvable letter for Entereg for POI. Id=4&pageId=402&newsid=923 (accessed 2007 Mar 9). 33. GlaxoSmithKline and Adolor Corporation press release. GSK and Adolor announce preliminary results from Phase 3 safety study of alvimopan (Entereg/ Entrareg ). phoenix.zhtml?c= &p= irolnewsarticle&t=regular&id=982960& (accessed 2007 Jul 16). 34. GlaxoSmithKline and Adolor Corporation press release. Complete response to the POI approvable letter now targeted for 3Q zhtml?c=120919&p=irol-newsarticle&t= Regular&id= (accessed 2007 Jul 16). 35. Delaney CP, Kiran RP, Senagore AJ et al. Casematched comparison of clinical and financial outcome after laparoscopic or open colorectal surgery. Ann Surg. 2003; 238: Weeks JC, Nelson H, Gelber S et al. Short term quality of life outcomes following laparoscopic-assisted colectomy vs open colectomy for colon cancer: a randomized trial. JAMA. 2002; 287: Lacy AM, Garcia-Valdedasas JC, Delgado S et al. Laparoscopy-assisted colectomy versus open colectomy for treatment of nonmetastatic colon cancer: a randomised trial. Lancet. 2002; 359: S20 Am J Health-Syst Pharm Vol 64 Oct 15, 2007 Suppl 13