Postoperative ileus (POI) is a temporary

Transcription

1 Drug Selection Perspectives Pharmacologic Options to Prevent Postoperative Ileus Yu-Chen Yeh, Elissa V Klinger, and Prabashni Reddy Postoperative ileus (POI) is a temporary impairment of gastrointestinal tract motility after abdominal or other surgery and is characterized by a clustering of bowel distension, lack of bowel sounds, accumulation of gastrointestinal gas and fluid, and delayed passage of flatus and stool. 1,2 Reasons for developing POI appear to be multifactorial: electrical activity in the gastrointestinal tract is decreased by anesthesia and surgery, while local inflammation and hyperactivity of the sympathetic nervous system further contribute to the development of POI. 1,2 The incidence of POI has been reported to vary from 4% to 32%. 3,4 Risk factors include the type of surgery, with a higher incidence associated with abdominal and pelvic surgery, open laparatomy, longer surgery time, higher estimated blood loss, prolonged opioid use, and inhalational anesthesia, as well as preexisting factors such as gastrointestinal disease and physical inactivity. 5,6 Consequences of POI include patient discomfort, decreased patient satisfaction, and complications such as pain and nausea 6 ; little information on the economic impact of POI has been reported. 3,7 Numerous surgical, anesthetic, and postoperative strategies have been used to limit POI. These include laparoscopic surgery, epidural anesthesia, opioid-sparing regimens, early enteral feeding, and early ambulation. Often, these are combined into a fast-track protocol that has been collectively shown to further decrease POI and length of stay (LOS) Regarding pharmacologic options, multiple agents including albumin, 14 β-blockers, 15 erythromycin, gum chewing, and metoclopramide have been studied in this setting. Two new drugs, both peripheral mu-opioid receptor antagonists, methylnaltrexone and alvimopan, have recently been evaluated in POI Author information provided at the end of the text. OBJECTIVE: To summarize the evidence on pharmacologic options in preventing postoperative ileus (POI). DATA SOURCES: The Cochrane Database of Reviews and OVID databases and Food and Drug Administration (FDA) Web site were searched (1950 April 2009) using the term postoperative ileus. STUDY SELECTION AND DATA EXTRACTION: Meta-analyses and randomized controlled trials were included for review. The FDA Web site was searched for clinical reviews and label information for drugs indicated for the prevention of POI. DATA SYNTHESIS: Three meta-analyses, 2 on gum-chewing and 1 on alvimopan, and 18 clinical trials were identified. Only gum chewing and alvimopan were effective in preventing POI. Gum chewing reduced the time to first flatus and bowel movement (weighted mean difference 21h; p = and 33h; p = , respectively). In one meta-analysis, gum chewing significantly reduced length of stay (LOS) by 2.4 days (p < ) but this was not replicated in the second meta-analysis. Alvimopan shortened the time to reach a composite endpoint of solid food intake, plus/minus flatus, and bowel movement (93 vs 105 h; p < 0.001). A higher incidence of myocardial infarction was observed in a 12- month study of alvimopan for the treatment of opioid-induced bowel dysfunction, but not in studies in patients undergoing bowel resection. Alvimopan decreased the time to written hospital discharge order (hazard ratio 1.35; p < 0.01), while the significance of a reduction in LOS ( days) was not reported. CONCLUSIONS: Gum chewing and alvimopan are effective in preventing POI, but given safety concerns and higher cost with alvimopan, gum chewing may be preferred. KEY WORDS: alvimopan, gum-chewing, ileus, postoperative. Ann Pharmacother 2009;43: Published Online, 14 Jul 2009, theannals.com, DOI /aph.1M121 The purpose of this review is to summarize the evidence supporting pharmacologic options in preventing POI. Specifically, efficacy, safety, and cost data from meta-analyses and randomized controlled trials are summarized. Data Sources and Selection The Cochrane Database of Reviews (CDSR, DARE, CCTR, CLHTA, CLEED) and OVID MEDLINE databases from 1950 to April 2009 were searched using the search term postoperative ileus, limited to the English-language literature, humans, clinical trials, comment, comparative study, consensus development conference, controlled clinical trial, editorial, guideline, journal article, letter, metaanalysis, multicenter study, practice guideline, randomized 1474 The Annals of Pharmacotherapy 2009 September, Volume 43 theannals.com Downloaded from aop.sagepub.com by guest on October 11, 2013

2 controlled trial, and reviews. In addition, the Food and Drug Administration (FDA) Web site was searched for clinical reviews and label information and was further supplemented by a manual bibliographic search. Articles were considered for inclusion if they were meta-analyses or randomized controlled trials and focused on pharmacologic options to prevent POI. Articles were excluded if they were Phase 1 studies or healthy volunteer studies. In addition, studies were excluded if they described the use of pharmacologic options after the development of POI (ie, treatment of POI). A total of 361 abstracts were identified from the search strategy. Eighteen relevant articles were included in the review. Three meta-analyses/pooled analyses were identified (Table 1): 2 on gum-chewing 19,20 and 1 on alvimopan. 32 Although the gum-chewing meta-analyses pooled data from the same 5 clinical trials, as the conclusions differed slightly, both analyses are presented in this review. In addition, 18 clinical trials were identified: one each on albumin, 14 β- blockers, 15 gum-chewing, 21 bisacody l,33 escin, 34 and mosapride 35 ; 3 each on methylnaltrexone and erythromycin ; and 6 on metoclopramide (Table 2). If studies were included in the meta-analyses, these were not summarized individually, although a brief overview and critique are provided. For 2 of the methylnaltrexone studies, only limited information is available. With the exception of the gum-chewing studies, all studies involved laparatomy procedures; few indicated whether epidural analgesia, opioid-sparing regimens, or other measures were employed to decrease POI. The studies included in the gum chewing meta-analyses were conducted in the US (n = 2), Japan (n = 2), and the UK (n = 1), while the alvimopan trials included in the meta-analysis were conducted in the US and Europe. Among the other 18 identified randomized controlled trials, 7 were conducted in the US, while the remaining were conducted in Scandinavia (n = 8), China (n = 1), Japan (n = 1), and Switzerland (n = 1). Endpoints in the studies included time to first flatus, time to first bowel movement, time to solid food intake, and LOS. In the alvimopan studies, a composite endpoint was used, GI-3, the time to recovery of gastrointestinal function as defined by the later of the following 2 endpoints: time that patient first tolerates solid food or time that patient first passes flatus or bowel movement. 32 The data on the use of each of the agents are summarized below. Albumin The use of albumin has been evaluated in one study. 14 It has been hypothesized that POI might be shortened by restoration of normal albumin concentrations, as marked hypoalbuminuria along with ileus have been observed following surgery. While treatment was randomized, assignment was based on whether patients had even (received albumin) or odd (control) hospital record numbers. Moreover, the study was single-blind and did not include a power calculation in its design. In the trial, the majority of patients underwent abdominal aortic aneurysm repair (57% vs 56%) or aortofemoral bypass (30% vs 28%, respectively). The goal of therapy was to maintain an albumin level of greater than or equal to 3.5 g/dl. Use of albumin did not appear to reduce the time to first flatus, time to first oral intake, or LOS, although p values were not reported and the lack of variance data precluded an independent calculation of significance. In addition, the study did not evaluate the effect of albumin on time to first bowel movement, an endpoint regarded as a more objective assessment of POI as compared with time to flatus. 36 Alvimopan Alvimopan was approved by the FDA in May 2008 to accelerate the time to upper and lower gastrointestinal recovery following partial large- or small-bowel resection surgery with primary anastomoses. 37 The drug is a peripheral mu-receptor antagonist believed to locally antagonize the effects of narcotic analgesics that are given during the perioperative period. The meta-analysis pooled data from 5 trials and only from patients undergoing bowel surgery. 32 Four of these trials had enrolled patients undergoing either bowel surgery or abdominal hysterectomies and the manufacturer s initial New Drug Application submission was for a proposed indication in patients undergoing major abdominal or complex pelvic surgery. 36 However, in the second-cycle submission, both the manufacturer and FDA acknowledged that alvimopan was not efficacious in the abdominal hysterectomy subpopulation. To address deficiencies in the original NDA submission, a fifth trial was conducted exclusively in patients undergoing bowel resection. Thus, data from the bowel surgery population in 5 studies were extracted and pooled for analysis. 32 These were post hoc analyses and had not been prespecified. Alvimopan 12 mg was administered before surgery and then twice daily after surgery for a maximum of 7 days or 15 doses. Alvimopan significantly reduced the time to reach a composite endpoint, GI-3, by approximately half a day. In addition, alvimopan reduced the time to reach another composite endpoint, GI-2, which was the time to recovery of gastrointestinal function as defined by the later of the following 2 endpoints: time that patient first tolerates solid food or time that patient first passes bowel movement by some 17 hours. 32 In their review of the data, the FDA noted that GI-2 was the more relevant endpoint given that it excluded time to flatus, an endpoint that is likely to be unreliable, as patients may not be aware when it occurs. 36 Alvimopan significantly decreased the time to written hospital discharge order in a pooled analysis of US studies (p 0.010); data from other countries were excluded, given the likely difference in practice patterns. 32 LOS was theannals.com The Annals of Pharmacotherapy 2009 September, Volume

3 Y-C Yeh et al. Table 1. Meta-Analyses of Drugs to Prevent Postoperative Ileus Time to Reference Intervention Design Pts. Time to Flatus Bowel Movement Length of Stay (days) Conclusion De Castro (2008) 19 gum chewing 3 studies, n = 5 mean age ± SD,61.9 y De Castro De Castro De Castro favorable effect on time to flatus and Chan (2007) 20 times/day (n = 78) pts., n = % female WMD: 19.3 h WMD: 29.7 h WMD 1.3 bowel movement colorectal resection standard postoperative all randomized laparoscopic surgery (95% CI 8 to 30) (95% CI 13 to 46) (95% CI 0.6 to 3.2) similar results LOS significantly in Chan (n = 3) but care (n = 80) (1 study) Chan Chan when Japanese study excluded not in De Castro (n = 4) epidural analgesia (2 WMD: 20.8 h; WMD: 33.3 h; Chan latter may have conducted inappropriate unknown, 2 yes, p = p = WMD 2.4; p < analysis by combining median with 1 partial) readmission and reoperation rates mean data similar (p = 0.24 and p = 0.83, respectively) Senagore (2007) 32 MITT population studies, N = 5 age 61 ± 14 y GI-3 93 vs 105 h; HR = 1.32, p < in 4 North American studies with alvimopan 12 mg accelerates time to GI bowel resection alvimopan 12 mg bid, pts., N = 1877 female 49.5% GI h vs 118 h; HR = 1.44, similar discharge criteria, recovery as evidenced by GI-3, GI-2, post hoc analysis placebo bid starting all randomized laparotomy p < alvimopan accelerated time to and hospital discharge readiness on day of surgery until no epidurals, local effect on GI-3 and GI-2 consistent hospital discharge written order hospital discharge or anesthetics, or NSAIDs across age, sex, and race HR 1.35; p 0.01 maximum of 7 days used GI = gastrointestinal; GI-3 = time to recovery of GI function as defined by the latter of the following 2 endpoints: time that patient first tolerates solid food or time that patient first passes flatus or bowel movement; GI-2 = time to recovery of GI function as defined by the latter of the following 2 endpoints: time that patient first tolerates solid food or time that patient first passes bowel movement; LOS = length of stay; MITT = modified intent-to-treat; NSAIDs = nonsteroidal antiinflammatory drugs; WMD = weighted mean difference. shorter with alvimopan ( days); however, it was not clear whether this difference was statistically significant. 36 Assuming the duration of therapy is 4 days (based on the Senagore et al. 32 meta-analysis in which the GI-2 endpoint was reached 101 h after surgery) and the average wholesale cost per dose is $75, the cost per patient of treatment with alvimopan could be as high as $675. Alvimopan is contraindicated in patients taking opioids for more than 7 days immediately prior to alvimopan due to a risk of precipitating opioid withdrawal. A higher incidence of myocardial infarction was observed in a 12-month study of alvimopan 0.5 mg twice a day for the treatment of opioid-induced bowel dysfunction in opioid-experienced patients with chronic noncancer pain. 36 These events occurred primarily between 1 and 4 months after the start of treatment. Cardiovascular risk factors (eg, age, body mass index 30 kg/m 2, diabetes, hypertension, smoking) were similar between the 2 groups. This imbalance has not been observed in other studies of alvimopan, including studies in patients undergoing bowel resection, and a causal relationship with alvimopan has not been established. This risk will be further evaluated in a prospective trial in patients undergoing radical cystectomy. 36 As a result of these findings, the FDA mandated that the manufacturer develop a Risk Evaluation and Mitigation Strategy. Enrollment in the Entereg Access and Support Program (EASE) is required to purchase the drug. Hospitals that perform bowel resection are eligible and need to acknowledge that: 1. educational materials are received and distributed to healthcare professionals responsible for ordering, dispensing, or administering alvimopan; 2. systems, order sets, protocols, or other measures are in place to limit use to 15 doses; 3. the hospital will not dispense alvimopan for outpatient use; and 4. the hospital will not transfer the patient on alvimopan to a nonregistered hospital. The manufacturer will assess the effectiveness of the EASE program by healthcare professional surveys. β-blockers The effect of β-blockers on POI has been evaluated in one small study (N = 40) in patients undergoing colonic surgery. 15 As inhibition of intestinal motility is thought to be mediated in part by α- and β-adrenoceptors and propranolol had been found to enhance intraluminal pressure in healthy volunteers, it has been postulated that β-adrenoceptor antagonists might prevent POI following colonic surgery. 38 This study was conducted over 20 years ago when practice patterns were likely different and the study was further limited by lack of a power calculation. Patients were randomized to 1 of 4 groups: (1) propranolol 4 mg twice a day intravenously then 40 mg twice a day orally on passage of flatus (n = 10); 1476 The Annals of Pharmacotherapy 2009 September, Volume 43 theannals.com

4 Pharmacologic Options to Prevent Postoperative Ileus Table 2. Randomized Controlled Trials of Drugs to Prevent Postoperative Ileus Time to Time to Time to Flatus Bowel Movement Oral Intake Length of Stay Reference Population Regimen (mean ± SD) (mean ± SD) (mean ± SD) (mean ± SD) Adverse Events Conclusions Albumin Woods (1993) 14 age (mean ± SD), albumin to maintain 4.1 vs 4.2 days; p values NR 4.0 vs 3.8 days; 9.2 vs 8.4 days; bronchitis, 8% vs 0% albumin does not reduce the R, conducted in 68.1 ± 10.1 vs level 3.5 g/dl and variance NR p values and p values and UTI, 5% vs 3% time to first flatus, oral intake, US 70.2 ± 7.4 y (n = 37) variance NR variance NR respiratory insufficiency, or LOS female, 24% vs 22% control (n = 32) 5% vs 0% abdominal aortic lymphocele/leak, 0% vs 9% aneurysm, 57% vs atelectasis, 3% vs 9% 56% aorto-femoral bypass, 30% vs 28% β-blockers Hallerback age (mean ± SEM), propranolol 4 mg iv propranolol vs placebo propranolol 4 mg NR NR pulse rate 11 ± 2 vs 3 ± propranolol shortens time to (1987) ± 4 vs 53 ± 4 vs bid; 40 mg po bid 52 ± 5 vs 56 ± 5 h; vs propranolol 10 2 beats/min first bowel movement but not R, DB, PC, 56 ± 4 y on passage of p = NR mg vs placebo 82 systolic blood pressure 9 ± to first flatus conducted in female, 50% vs 70% flatus (n = 10) ± 11 vs 79 ± 8 vs 3 vs 0 ± 3 mm Hg Norway vs 68% placebo (n = 10) 110 ± 9 h; p < 0.01 elective colonic starting morning surgery after operation or propranolol 10 mg iv bid; 80 mg bid po on passage of first flatus (n = 10) placebo (n = 10) starting 30 min before operation Bisacodyl Zingg (2008) 33 age (mean ± SD), bisacodyl 10 mg po median (range), 2.0 (1 7) median (range), median (range), median (range), tendency for longer duration bisacodyl shortens time to first R, DB, PC, 67.9 ± 13.2 vs 66.4 starting 1 day vs 2.0 (1 7) days; p = (1 8) vs 4.0 solid food (4 92) vs of nasogastric tube bowel movement but not time conducted in ± 14.5 y; p = 0.48 before surgery for (1 8) days; p = (2 30) vs (6 74) days; reinsertion in the bisacodyl to first flatus or tolerance of Switzerland female, 49% vs 37%; 4 days (n = 83) (2 23) days; p = 0.77 group (mean ± SD), 2.8 ± solid food or LOS p = 0.12 placebo (n = 86) p = vs 1.8 ± 0.9 days; p = rectosigmoid resection, % vs 63% (p = 0.64) anterior resection 13% vs 8% (p = 0.33) DB = double-blind; LOS = length of stay; NR = not reported; PC = placebo-controlled; R = randomized; UTI = urinary tract infection. (continued on page 1478) theannals.com The Annals of Pharmacotherapy 2009 September, Volume

5 Y-C Yeh et al. Table 2. Randomized Controlled Trials of Drugs to Prevent Postoperative Ileus (continued) Time to Time to Time to Flatus Bowel Movement Oral Intake Length of Stay Reference Population Regimen (mean ± SD) (mean ± SD) (mean ± SD) (mean ± SD) Adverse Events Conclusions Erythromycin Lightfoot age, NR erythromycin iv 125 median (IQ range), 5 median (IQ range), median (IQ median (IQ no significant difference in erythromycin not effective in (2007) 16 female, 18% vs 0%; mg q8h (maximum (4 6) vs 5 (4 6) days; 6 (4 6) vs 5 (4 7) range), general range) 10 the incidence of allergic reducing POI R, DB, PC, p = doses) (n = 11) p = 0.35 days; p = 0.98 diet 9 (8 10) vs (9 10) vs 9 reactions, QTc interval conducted in US cystectomy with placebo (n = 11) 8 (7 13) days; (8 14) days; prolongation, nausea, urinary diversion started on POD1 p = 0.60 p = 0.63 emesis, abdominal pain secondary to bladder cancer or interstitial cystitis Smith (2000) 17 pts. who completed erythromycin 200 mg postoperative day (mean postoperative day postoperative postoperative rash, 2% vs 0% erythromycin time to first R, PC, conducted study iv q6h at 1000 on ± SE) 4.1 ± 1.3 vs 4.4 ± (mean ± SE) 5.2 ± day (mean ± day (mean ± arrhythmias, 0% vs 6%; flatus but not the time to first in US age (mean ± SD), 1st POD until 1.1; p = vs 5.4 ± 1.3; SE) solid food SE) 7.5 ± 2 vs p = 0.04 bowel movement, oral intake, 63.2 ± 12.6 vs 61.4 patient on solid p = ± 1.9 vs ± 2.8; p = or LOS; note, primary end- ± 11.7 y; p = 0.43 food or POD5 ± 1.8; p = point placement of nasofemale, 34% vs 39%; (n = 65), placebo gastric tube NS between p = 0.53 (n = 69) groups operation (p = 0.84) right colectomy, 34% vs 25% low anterior resection, 31% vs 31% anterior resection, 12% vs 19% Bonacini (1993) 18 median age (range); erythromycin iv ± 29.4 vs 53.9 ± 81.8 ± 32.9 vs ± 44.4 vs postoperative, rash, 7% vs 3% erythromycin does not reduce R, DB, PC, 42 (18 80) vs 40 mg q8h for 9 doses 27.7 h; p = NS ± 28.0 h; p = NS 71.7 ± 65.5 h; ± abdominal pain, 5% vs 6% time to flatus, bowel moveconducted in US (18 81) y (n = 41) p = NS vs ± burning venous access, 2% ment, oral intake, or LOS female, 76% vs 64% placebo (n = 36) h; p = NS vs 0% cholecystectomy, 41% vs 47% celiotomy, 44% vs 47% Escin Xie (2009) 34 age (mean ± SD), 59 escin 5 mg (n = 18), ± h; p = ± h; not assessed not assessed none escin 25 mg once daily R, DB, PC, ± 11 vs 57 ± 10 vs 15 mg (n = 18), vs ± h; p = vs shortens the time to first conducted in 64 ± 10 vs 57 ± 11 y mg (n = 18) iv p = vs ± ± h; bowel movement and first China female, 56% vs 72% infusion once daily h; p = vs p = vs flatus vs 44% vs 44% given 6 h after ± h ± h; colorectal surgery completion of the p = vs surgery until first ± h bowel movement or up to 7 days placebo (n = 18) DB = double-blind; IQ = interquartile; LOS = length of stay; NR = not reported; NS = not significant; PC = placebo-controlled; POD = postoperative day; POI = postoperative ileus; R = randomized. (continued on page 1479) 1478 The Annals of Pharmacotherapy 2009 September, Volume 43 theannals.com

6 Pharmacologic Options to Prevent Postoperative Ileus Table 2. Randomized Controlled Trials of Drugs to Prevent Postoperative Ileus (continued) Time to Time to Time to Flatus Bowel Movement Oral Intake Length of Stay Reference Population Regimen (mean ± SD) (mean ± SD) (mean ± SD) (mean ± SD) Adverse Events Conclusions Gum chewing McCormick age (mean ± SD), 60 1 stick gum chewed open colectomy 2.9 ± 1.2 open colectomy 3.6 NR open colectomy NR gum chewing hastens time to (2005) 21 ± 14 y for 15 min qid (n = vs 2.8 ± 1.6; p = 0.83 ± 1.1 vs 3.9 ± 1.4; 5.6 ± 1.4 vs 5.3 first bowel movement and R, conducted in open colectomy, 42% 53) laparoscopic 2.2 ± 0.9 vs p = 0.50 ± 1.2; p = 0.50 LOS in patients undergoing US laparoscopic colec- sips of clear liquids 2.4 ± 1.3; p = 0.31 laparoscopic 2.6 ± laparoscopic 4.0 laparoscopic colectomy tomy, 58% qid (n = 35) 1.0 vs 3.3 ± 1.3; ± 1.2 vs 5.3 ± p = ; p = Methylnaltrexone Viscusi (2005) 28 open segmental methylnaltrexone NR mean ± SEM 97 ± mean ± SEM full mean ± SEM withdrawals 12% vs 28% methylnaltrexone time to first R, DB, PC, colonic resection 0.3 mg/kg over 20 6 vs 120 ± 10 h; liquids 70 ± ± 6 vs 165 most common adverse bowel movement, full liquid country unknown min q6h post- p = 0.01 vs 100 ± 19 h; ± 16 h; p = NS events were fever and intake, and discharge eligibilisurgery for 24 h p = 0.05 discharge eligi- nausea ty; no effect on LOS or solid after tolerate solid solid food 100 ± bility 119 ± 7 food intake food or first bowel 11 vs 125 ± 17 h; vs 149 ± 17 h; abstract form only movement, up to 7 p = 0.12 p = 0.03 days (n = 33) placebo (n = 32) Metoclopramide Cheape (1991) 22 age, mean (range), metoclopramide 10 not assessed not assessed fluids 3.5 ± 1.7 not assessed not assessed metoclopramide does not R, SB, conducted 59.5 (14 89) y mg iv q8h starting vs 3.5 ± 2.0 reduce time to oral intake in US female, 60% vs 62% in recovery room (difference procedure, segmental until regular diet 95% CI 0.78 colectomy, 43% vs initiated to 0.78) 43% control group solid food 4.8 ± abdominal colectomy, 1.9 vs 5.0 ± % vs 13% (difference 95% CI to 0.56) Tollesson age (mean ± SEM), metoclopramide 20 p = NS p = NS NR NR none metoclopramide did not time (1991) ± 4 vs 49 ± 4 y mg iv tid after to first flatus or stool R, DB, PC, female, 50% vs 60% operation for 10 conducted in cholecystectomy doses (n = 10) Sweden placebo (n = 10) Jepsen (1986) 24 age, median (range), metoclopramide 10 NR median (range) net oral intake NR NR metoclopramide prolonged R, DB, PC, 54 (36 69) y mg iv after opera (1 80.9) h larger in placebo ileus and had no effect on the conducted in female, 53% tion, then 10 mg iv vs 44.7 ( ) group on post- time to first oral intake Denmark laparotomy for qid for 5 days h; p = operative day 3 aortoiliac vascular (n = 30) (0.99 vs 1.75 L); procedures placebo (n = 25) days 1, 2, 4, 5; p = NS DB = double-blind; LOS = length of stay; NR = not reported; NS = not significant; PC = placebo-controlled; R = randomized; SB = single-blind. (continued on page 1480) theannals.com The Annals of Pharmacotherapy 2009 September, Volume

7 Y-C Yeh et al. Table 2. Randomized Controlled Trials of Drugs to Prevent Postoperative Ileus (continued) Time to Time to Time to Flatus Bowel Movement Oral Intake Length of Stay Reference Population Regimen (mean ± SD) (mean ± SD) (mean ± SD) (mean ± SD) Adverse Events Conclusions Metoclopramide (continued) Davidson abdominal surgery/ metoclopramide 10 no significant difference; no significant 10 doses to NR any AE, 10.4% vs 9.6% metoclopramide had no effect (1979) 25 laparotomy mg im (n = 58) results not shown difference; results tolerance to sedation, 16% vs 11% on the time to first flatus or R, DB, PC, group A (no incision placebo (n = 57) not shown solid food restlessness, 5% vs 4% bowel movement but time conducted in US into stomach, before surgery and group A, 83% vs dystonia 0% vs 5% to return to solid food in duodenum, or tid following surgery 48%; p < 0.05 patients not undergoing GI intestines made) until solid food group B, 24% vs anastomoses (n = 58) tolerated and/or 36%; p = NS age, mean (range), defecation 48.8 (20 81) vs 52.1 im, then po when (21 84) y oral fluids taken female, 52% vs 55% group B (GI entry, anastomosis, or suture) (n = 57) age, mean (range), 52.8 (19 81) vs 49.7 (30 77) y female 45% vs 46% Breivik (1971) 26 age (mean ± SD), metoclopramide earlier in active group; intestinal peristalsis NR NR NR metoclopramide hastened time R, DB, PC, 42.5 ± 16.8 vs mg im p value NR (mean ± SEM a ) to bowel sounds and passage conducted in ± 13.3 y placebo, at end of bowel sounds end of ± 1.26 vs of flatus; also vomiting Norway female, 100% operation, 2000, POD1, 50% vs 33% 8.60 ± 0.99; cholecystectomy and 0700, and end of POD2, 52% vs 26% p < hours on p values NR POD1 Kivalo (1970) 27 age 30 y, 16% vs metoclopramide 10 NR no stools by POD4 NR NR none metoclopramide reduced post- PC, conducted in 24%; y, 66% mg im, 2 doses 32% vs 64%; operative nausea, vomiting, Finland vs 46%; >50 y, 8% (n = 90) p < and constipation vs 16% placebo (n = 86) gynecological surgery Mosapride Narita (2008) 35 mean age (range), mosapride 15 mg mean (range), 48.5 mean (range), 32.7 not assessed mean (range), none mosapride shortens the time to R, DB, PC, 64.2 (35 82) vs po 3 times daily ( ) vs 69.3 ( ) vs 6.7 (5 19) vs first bowel movement and conducted in 70.6 y (55 85); starting on the ( ) h; p = ( ) h; 8.4 (6 19) days; LOS following laparoscopic Japan p = 0.13 morning of POD1 p = 0.28 p = 0.04 colectomy female, 55% vs 55%; until hospital p > 0.99 discharge or up to hand-assisted laparo- 10 days (n = 20) scopic colectomy placebo (n = 20) AE = adverse events; DB = double-blind; GI = gastrointestinal; LOS = length of stay; NR = not reported; NS = not significant; PC = placebo-controlled; POD = postoperative day; R = randomized. a Intestinal paralysis score: bowel sounds present: 2000 hours day of operation 4, POD hours 3, POD hours 2, POD hours 1; flatus passed: 2000 hours day of operation, 84 POD hours, 73 POD hours, 6 POD The Annals of Pharmacotherapy 2009 September, Volume 43 theannals.com

8 Pharmacologic Options to Prevent Postoperative Ileus (2) placebo starting on the morning after the operation (n = 10); (3) propranolol 10 mg twice a day intravenously followed by 80 mg twice a day orally on passage of first flatus (n = 10); or (4) placebo (n = 10) starting 30 minutes before the operation. Propranolol shortened the mean time to first bowel movement (propranolol 4 mg vs propranolol 10 mg vs placebo (mean ± SEM, 82 ± 11 vs 79 ± 8 h vs 110 ± 9 h; each propranolol group vs placebo; p < 0.01), but not the time to the passage of first flatus. Both heart rate and systolic blood pressure were significantly lower in patients who received propranolol. These results are promising, but the small sample size suggests that further studies are warranted before the findings can be applied in the clinical setting. Bisacodyl Bisacodyl is a laxative that stimulates colonic peristalsis. The effect of bisacodyl in preventing POI has been evaluated in one recent randomized, double-blind, placebo-controlled trial involving 169 patients undergoing open or laparoscopic colorectal resection. 33 The sample size was calculated based on a 20% reduction in time to reach the GI- 3 endpoint, with a significance of 5% and 90% power. Bisacodyl 10 mg, given twice daily by mouth, starting a day before surgery and continued for 3 days after surgery, significantly reduced the time to reach the GI-3 endpoint (median 3.0 vs 3.7 days; p = 0.007) and time to first bowel movement (median 3.0 vs 4.0 days; p = 0.001) compared with placebo. However, no significant difference was observed in time to first flatus, tolerance of solid food, or length of stay. In addition, for unexplained reasons, patients in the bisacodyl group seemed to have a longer duration of nasogastric tube reinsertion (mean ± SEM 2.8 ± 1.3 days vs 1.8 ± 0.9 days; p = 0.055). While the results appear promising, the longer duration of nasogastric tube reinsertion in the treatment group suggests the need for further studies. Erythromycin Three studies using intravenous erythromycin in the prevention of POI were identified At low doses, erythromycin has prokinetic effects in the gut through binding to gastrointestinal tract motilin receptors. 39 Daily doses ranged from 375 to 800 mg and were administered in 3 4 divided doses. All studies were placebo controlled and 2 of these employed a double-blind design. Only one of the studies included a power calculation, but based this on a 2-day reduction in time to first oral intake with erythromycin. The studies were conducted in a range of surgeries including cystectomy, colectomy, and cholecystectomy. Time to first bowel movement, time to first solid food intake, and LOS were all similar among patients given erythromycin and placebo. Time to flatus was significantly shorter in one study by approximately 8 hours, but the primary endpoint, placement of a nasogastric tube, was similar between the groups. 17 Furthermore, patients given erythromycin were also at a higher risk of experiencing arrhythmias (0% vs 6%; p = 0.04). The development of antibiotic resistance was not assessed in these studies but is a concern when using erythromycin for noninfective indications. 40 Escin Escin is a natural mixture of triterpene saponins with antiinflammatory, vascular protection, and veinotonic effects; its effects in the prevention of POI were evaluated in one small, randomized, double-blind study in China. 34 Based on the power calculation, 72 patients were randomized to receive 5, 15, or 25 mg of escin or placebo once daily administered by intravenous infusion. Time to passage of gas and time to bowel movement were significantly reduced with escin 15 mg (mean ± SEM ± h; p = 0.036; ± h; p = 0.035) and 25 mg (65.50 ± h; p = 0.022; ± h; p = 0.019) compared with placebo (82.81 ± h; ± h). Given the small sample size in this pilot study, additional research is needed to confirm these findings. Gum chewing Gum chewing is thought to promote physiological stimulation of the cephalic-vagal axis, thereby increasing bowel motility and gastrointestinal stimulation without the complications associated with early water intake or postoperative feeding. 19,20 Both meta-analyses included the same 5 trials and, as the results differed slightly, findings from both are included here. All of the included trials were small, with only 2 including a power analysis. Of note, the majority of the studies involved patients undergoing colectomy via open laparotomy rather than laparoscopic surgery. In all studies, patients were instructed to chew sugarless gum 3 times a day starting on the first postoperative day until first passage of flatus or bowel movement. 20 In some studies, the duration of chewing was specified and varied from 5 to 45 minutes Blinding of therapy allocation to the patients and surgeons was challenging and was not implemented in most of the studies, except for one trial in which patients assigned to the placebo group wore a wrist bracelet and were unaware that the bracelet was used as a placebo. 42 In both meta-analyses, gum chewing reduced the time to first presence of bowel sounds, flatus, bowel movement, and oral intake. 19,20 An additional randomized controlled trial of gum chewing was identified in which the time to first bowel movement and LOS in patients undergoing laparoscopic colectomy were significantly reduced, although no significant effect was found among patients who underwent open colectomy. 21 theannals.com The Annals of Pharmacotherapy 2009 September, Volume

9 Y-C Yeh et al. Safety issues were not noted in the gum-chewing analyses. 19,20,36 The cost is approximately $3 per patient. 44 In one of the meta-analyses, gum chewing significantly reduced LOS (n = 3 studies) by 2.4 days, which may translate into considerable cost savings. 20 However, LOS from 3 countries, the US, Japan, and the UK, were pooled in these analyses. As practice patterns may differ among countries, this limits the applicability of the findings. Of note, this reduction in LOS was not replicated in the second metaanalysis (n = 4 studies). However, it included a study in which LOS data were reported as a median value, and it may be methodologically inappropriate to include means and medians in the analysis. 19 Metoclopramide Six studies with intravenous or intramuscular metoclopramide were identified Metoclopramide stimulates gastrointestinal motility through peripheral and central mechanisms and has been shown to increase gastric contractions and gastric emptying by stimulating smooth muscle contractions in the small intestine, thereby decreasing intestinal transit time. 45 Of note, studies of metoclopramide in POI were conducted during the 1970s through 1980s when practice patterns are likely to have differed from current practices. In addition, several of the studies did not assess the effect of metoclopramide on time to first bowel movement, an endpoint that is more objective than time to first flatus, while another study used a nonvalidated scoring system to assess efficacy. 26,36 In the various trials, surgery types included colectomy, cholecystectomy, and laparotomy, as well as aortoiliac procedures and gynecological surgery. Metoclopramide appeared to be ineffective in preventing POI as assessed by time to first flatus, first bowel movement, and oral intake, as well as LOS. Methylnaltrexone Methylnaltrexone, like alvimopan, is a peripheral mureceptor antagonist. It was approved for the treatment of opioid-induced constipation in patients with advanced illness who are receiving palliative care, when response to laxative therapy has not been sufficient. 46 It is contraindicated in patients who have known or suspected mechanical gastrointestinal obstruction. 46 While methylnaltrexone significantly improved POI in a Phase 2 study, 28 it failed to meet its primary or secondary endpoints in 2 Phase 3 trials; details of the latter trials are not available Mosapride Mosapride is a prokinetic agent that stimulates the serotonin 5-hydroxytryptan- 4 receptor, increases the release of acetylcholine at parasympathetic nerve endings, and promotes bowel motility. A small study evaluated the effect of mosapride in patients undergoing hand-assisted laparoscopic colectomy (N = 40). 35 Compared with placebo, mosapride 15 mg 3 times daily by mouth significantly reduced the time to first bowel movement (mean 48.5 vs 69.3 h; p = 0.015) and length of postoperative hospital stay (mean 6.7 vs 8.4 days; p = 0.040). It appeared to be a safe agent in this study. However, it is commercially available only in Japan. Discussion Based on our review of the literature, both gum chewing and alvimopan appear to be effective in the prevention of POI. 19,20,32 Gum chewing reduced the time to first flatus and bowel movement by approximately a day each, while alvimopan shortened the time to reach a composite endpoint of solid food intake, plus/minus flatus, and bowel movement by half a day. When considering whether to use these agents, clinicians should assess the clinical value of these data for their patients. β-blockers, escin, and bisacodyl showed promise in postoperative ileus in small studies, but additional data are needed before their use can be recommended. 15,33,34 In contrast, albumin, erythromycin, and metoclopramide have little evidence to support their use in this setting. 14,16-18,22-27,47 Moreover, many of these studies were conducted over years ago when operative techniques and patient management likely differed from current day practices, thus limiting the applicability of their results. 15,24-27 While total surgical time, estimated blood loss, and total opiate dose were found to be significantly associated with the risk of POI, 5 few studies included in this review reported results by risk factor. Given the randomization design of the studies, it is likely that the risk of POI at baseline was comparable among the treatment groups; indeed, a few studies specifically noted this. 14,16,18,20,21,23 The lack of benefit with methylnaltrexone, which, like alvimopan, is a mu-receptor antagonist, is somewhat surprising. One explanation could be an inadequate dose. Indeed, the alvimopan dose in the POI prophylaxis trials was more than 20-fold higher than the dose used in the opioid-induced constipation trials. 36 In contrast, the methylnaltrexone doses in POI prevention and opioid-induced constipation were 12 or 24 mg every 6 hours by intravenous infusion 28 and mg/kg subcutaneously, 48,49 respectively. With gum chewing and alvimopan, efficacy in preventing POI is only in the setting of colorectal resection. 19,20,32 It is noteworthy that alvimopan was not effective in patients undergoing gynecologic surgeries. 36 As the incidence of POI has been shown to be lower in patients undergoing gynecologic surgeries compared with colorectal resections, 50 the likelihood of detecting a benefit with any intervention in this setting is reduced. What, if any, differences exist between alvimopan and gum chewing? Several issues need to be considered: pa The Annals of Pharmacotherapy 2009 September, Volume 43 theannals.com

10 Pharmacologic Options to Prevent Postoperative Ileus tient population, safety, administration, implementation, and economics. Unlike alvimopan trials, gum-chewing studies have been conducted in a broader range of patients, including those receiving epidural analgesia and laparoscopic surgery, 19,20 whereas alvimopan studies excluded patients who received epidurals, local anesthetics, or nonsteroidal antiinflammatory drugs. 32 Adverse effects with gum were not noted in any of the trials In contrast, concerns about safety with alvimopan, specifically the risk of myocardial infarction, have led to the need for an FDA-mandated Risk Evaluation and Mitigation Strategy (REMS). 37 The administration requirements of gum chewing and alvimopan differ. With gum chewing, patients need to be able to follow instructions and have sufficient coordination to adhere to therapy; thus, some patients (eg, those who are intubated or who have neurologic impairment) will not be candidates for therapy. Alvimopan is available only in oral form and, therefore, patients who are not to receive oral intake may not be eligible for this drug. While data on delivery of alvimopan via nasogastric tubes are not available, survey data show that the majority of patients have nasogastric tubes removed on the first postoperative day and, therefore, need for administration of the drug in the manner is likely to be a rare event. 51 Both agents have special requirements for procurement. With chewing gum, a practical issue is which hospital department will stock the item, as this is not traditionally carried in hospitals. Pharmacy and dietary departments are 2 possible options for procuring and dispensing gum. With alvimopan, the REMS program does entail an administrative burden and the hospital needs to ensure that it has the support and systems in place to adhere to the program s requirements. Clinical decision support in institutions that have computerized physician order entry systems may facilitate this. On the economic front, based on acquisition cost, alvimopan is more expensive compared with chewing gum. 32,44 However, the effect on other resources needs to be considered. Alvimopan significantly decreased the time to a written hospital discharge order. 32 Unlike with reduction in LOS, it is unclear how to assess the economic benefit of this endpoint. In contrast, data suggest that gum chewing may significantly reduce LOS by a day, although this is based on pooled data from different countries where practice patterns likely differ. 20 Data in surgeries other than colorectal resection are needed before use can be recommended in other settings. It is, however, known that alvimopan is not effective in gynecologic surgery, likely due to a lower incidence of POI in these patients. 3,4,6 Also, there are no data showing the effect of alvimopan on POI in patients undergoing laparoscopic surgery. However, this is believed to be associated with a lower incidence of POI and, indeed, is one of the recommendations in fast-track protocols to reduce POI Studies that compare the efficacy of alvimopan with that of gum chewing as well as their use in combination would be valuable. With alvimopan, data on LOS would facilitate a full assessment of the economic impact of this agent. Given that the endpoint in many of these studies is time to first bowel movement, it is perhaps surprising that only limited data on laxatives as preventive interventions for POI are available. 33 While mechanical bowel preparation with solutions such as polyethylene glycol and mannitol has been associated with an increased risk of anastomotic leakage, studies with other types of laxatives are needed. 52 However, it is not clear whether laxatives have already been incorporated into POI trials for other interventions (eg, alvimopan and chewing gum). In addition to studies of the effects of various drugs in preventing POI, further work to understand how best to define POI would be valuable. This would facilitate the design of future trials and selection of endpoints, which was an issue with the alvimopan trials. Objective, clinically, and economically meaningful endpoints, such as time to first bowel movement and LOS, are among those that would be informative to practitioners. While intuitively, POI is a problem, limited data demonstrate the consequences of POI. Of particular interest would be the impact on LOS and costs. All available data are published in abstract form only 3 ; studies showing the economic burden of POI are lacking. Review of the literature suggests that a comprehensive approach to preventing POI is needed. There is a considerable volume of data to support surgical and anesthetic prophylactic options, and these should be undertaken when feasible in all patients undergoing surgery associated with risk of POI. Both chewing gum and alvimopan may be considered as additional options in patients undergoing colorectal resections. While alvimopan appears to be efficacious in preventing POI, questions about safety remain. Furthermore, using alvimopan will increase pharmacy expenditures. In contrast, chewing gum is not only efficacious but also may reduce costs associated with large- or small-bowel resections, given that it likely reduces LOS; a few patients, however, may not be candidates for chewing gum. Given the safety concerns, numerous contraindications, and higher cost with alvimopan, chewing gum may be preferred. Summary Review of pharmacologic options in the prevention of POI shows that gum chewing and alvimopan are efficacious in patients undergoing colorectal resections while albumin, β-blockers, erythromycin, metoclopramide, and methylnaltrexone lack data to support their use in any setting. Questions about the safety of alvimopan remain, given the increased risk of myocardial infarction shown in a long-term study among patients with chronic noncancer pain, while safety issues with gum chewing have not yet been found. Alvimopan will increase pharmacy expenditures; the cost of acquiring gum is substantially lower. Both agents have several practical issues around procurement, storage, and adminis- theannals.com The Annals of Pharmacotherapy 2009 September, Volume

11 Y-C Yeh et al. tration, and neither has compelling data demonstrating a reduction in nondrug costs, including LOS. Given the safety concerns, numerous contraindications, and higher cost with alvimopan, gum chewing may be the preferred option. Yu-Chen Yeh MS, Senior Pharmacist, Center for Drug Policy, Partner s Healthcare, Needham, MA Elissa V Klinger MS, Project Manager, Center for Drug Policy, Partner s Healthcare Prabashni Reddy PharmD, Director, Center for Drug Policy, Partners Healthcare Reprints: Dr. Reddy, Center for Drug Policy, Partners Healthcare,115 Fourth Ave., Needham, MA 02494, fax 781/ Financial disclosure: None reported. References 1. Holte K, Foss NB, Andersen J, et al. Liberal or restrictive fluid administration in fast-track colonic surgery: a randomized, double-blind study. Br J Anaesth 2007;99: Maron DJ, Fry RD. New therapies in the treatment of postoperative ileus after gastrointestinal surgery. Am J Ther 2008;15: Saunders WB, Bowers B, Moss B, Bell TJ, Wang PF. Recorded rate and economic burden associated with postoperative ileus (abstract 30346). Presented at: American Society of Health-System Pharmacists Midyear Clinical Meeting, Orlando, FL, December 7, Schmitt SL, Cohen SM, Wexner SD, Nogueras JJ, Jagelman DG. Does laparoscopic-assisted ileal pouch anal anastomosis reduce length of hospitalization. In J Colorectal Dis 1994;9: Artinyan A, Nunoo-Mensah JW, Balasubramaniam S, et al. Prolonged postoperative ileus definition, risk factors, and predictors after surgery. World J Surg 2008;32: Senagore AJ. Pathogenesis and clinical and economic consequences of postoperative ileus. Am J Health Syst Pharm 2007;64(suppl):S Iyer S, Saunders W. Burden of postoperative ileus (POI) in colectomy surgery patients in the United States. Presented at: International Society of Pharmacoeconomics and Outcomes Research Annual International Meeting, Arlington, VA, May Charoenkwan K, Phillipson G, Vtyavanich T. Early versus delayed (traditional) oral fluids and food for reducing complications after major gynecological surgery. Cochrane Database Syst Rev 2007:CD Liu SS, Wu CL. Effect of postoperative analgesia on major postoperative complications: a systematic update of the evidence. Anesth Analg 2007; 104: Maffezzini M, Gerbi G, Campodonico F, Parodi D. Multimodal perioperative plan for radical cystectomy and intestinal diversion. I. Effect on recovery of intestinal function and occurrence of complications. Urology 2007;69: Schlachta CM, Burpee SE, Fernandez C, Chan B, Mamazza J, Poulin EC. Optimizing recovery after laparoscopic colon surgery (ORAL-CS). Surg Endosc 2007;21: Basse L, Jakobsen DH, Billesbølle, Werner M, Kehlet H. A clinical pathway to accelerate recovery after colonic resection. Ann Surg 2000;232: Jørgensen H, Wetterslev J, Møiniche S, Dhal JB. Epidural local anaesthetics versus opioid-based analgesic regimens for postoperative gastrointestinal paralysis, PONV, and pain after abdominal surgery. Cochrane Database Syst Rev 2001:CD Woods MS, Kelley H. Oncotic pressure, albumin and ileus: the effect of albumin replacement on postoperative ileus. Am Surg 1993;59: Hallerback B, Carlsen E, Carlsson K, et al. β-adrenoreceptor blockade in the treatment of postoperative adynamic ileus. Scand J Gastroenterol 1987;22: Lightfoot AJ, Eno M, Kreder KJ, O Donnell MA, Rao SSC, Williams RD. Treatment of postoperative ileus after bowel surgery with low-dose intravenous erythromycin. Urology 2007;69: Smith AJ, Nissan A, Lanouette NM, et al. Prokinetic effect of erythromycin after colorectal surgery: randomized, placebo-controlled, double-blind study. Dis Colon Rectum 2000;43: Bonacini M, Quiason S, Reynolds M, Gaddis M, Pemberton B, Smith O. Effect of erythromycin on postoperative ileus. Am J Gastroenterol 1993; 88: de Castro SM, van den Esshert JW, van Heek NT, et al. A systematic review of the efficacy of gum chewing for the amelioration of postoperative ileus. Dig Surg 2008;25: Chan MK, Law WL. Use of chewing gum in reducing postoperative ileus after elective colorectal resection: a systematic review. Dis Colon Rect 2007;50: McCormick JT, Garvin R, Caushaj P, et al. The effects of gum-chewing on bowel function and hospital stay after laparoscopic vs open colectomy: a multi-institution prospective randomized trial. J Am Coll Surg 2005; 201(suppl):S Cheape JD, Wexner SD, James K, Jagleman DG. Does metoclopramide reduce the length of ileus after colorectal surgery? A prospective randomized trial. Dis Colon Rectum 1991;34: Tollesson PO, Cassuto J, Faxén A, Rimbäck G, Mattsson E, Rosén S. Lack of effect of metoclopramide on colonic motility after cholecystectomy. Eur J Surg 1991;157: Jepsen S, Klærke, Nielsen PH, Simonsen O. Negative effect of metoclopramide in postoperative adynamic ileus. A prospective, randomized, double-blind study. Br J Surg 1986;73: Davidson ED, Hersh T, Brinner RA, Barnett SM, Boyle LP. The effects of metoclopramide on postoperative ileus: a randomized, double-blind study. Ann Surg 1979;190: Breivik H, Lind B. Anti-emetic and propulsive peristaltic properties of metoclopramide. Br J Anaesth 1971;43: Kivalo I, Miettinen K. The effects of metoclopramide on postoperative nausea and bowel function. Ann Chirug Gynaecol Fenniae 1970;59: Viscusi E, Rathmell J, Fichera A, et al. A double-blind, randomized, placebo-controlled trial of methylnaltrexone (MNTX) for post-operative bowel dysfunction in segmental colectomy patients (abstract 893). Presented at: American Society of Anesthesiologists Annual Meeting, Orlando, FL, October 24, Progenics Pharmaceuticals, Inc. and Wyeth Pharmaceuticals, Inc. Progenics and Wyeth initiate global phase 3 clinical program of intravenous methylnaltrexone to treat post-operative ileus. September 20, Press release. wyeth_html/home/news/pressreleases/2006/ html (accessed 2009 Mar 20). 30. Progenics Pharmaceuticals, Inc. and Wyeth Pharmaceuticals, Inc. Wyeth and Progenics provide update on phase 3 clinical trial of intravenous methylnaltrexone for postoperative ileus. March 12, Press Release. news/pressreleases/2008/ html (accessed 2009 Mar 20). 31. Progenics Pharmaceuticals, Inc. and Wyeth Pharmaceuticals, Inc. Wyeth and Progenics announce preliminary clinical trial results for Relistor oral and intravenous formulations. May 22, Press Release. www. wyeth.com/news/archive?nav=display&navto=/wyeth_html/home/ news/pressreleases/2008/ html (accessed 2009 Mar 20). 32. Senagore AJ, Bauer JJ, Du W, Techner L. Alvimopan accelerates gastrointestinal recovery after bowel resection regardless of age, gender, race, or concomitant medication use. Surgery 2007;142: Zingg U, Miskovic D, Pasternak I, Meyer P, Hamel CT, Metzger U. Effect of bisacodyl on postoperative bowel motility in elective colorectal surgery: a prospective, randomized trial. Int J Colorectal Dis 2008;23: Xie Q, Zong X, Ge B, et al. Pilot postoperative ileus study of escin in cancer patients after colorectal surgery. World J Surg 2009;33: Narita K, Tsunoda A, Takenaka K, Watanabe M, Nakao K, Kusano M. Effect of mosapride on recovery of intestinal motility after hand-assisted laparoscopic colectomy for carcinoma. Dis Colon Rectum 2008;51: NDA review for regulatory action. May, cder/foi/nda/2008/021775_entereg_capsules_medr.pdf (accessed 2009 Apr 30) The Annals of Pharmacotherapy 2009 September, Volume 43 theannals.com