Is methylnaltrexone effective in the treatment of constipation in hospice patients taking opioids?

Methylnaltrexone (Relistor) is effective for the treatment of opioid-induced constipation in hospice patients. However, long-term safety is not known, so it should not be widely used for nonterminally ill patients until longer studies have been performed. (LOE = 1b)

Thomas J, Karver S, Cooney GA, et al. Methylnaltrexone for opioid-induced constipation in advanced illness. N Engl J Med 2008; 358: 2332-2343.  [PMID:18509120]

Randomized controlled trial (double-blinded)

Industry

Concealed

Nursing home/extended care facility

Opiates attach to mu-opiod receptors in the gut and cause constipation. Naltrexone is an opioid antagonist, but it causes central nervous system side effects. Methylnaltrexone is an opioid antagonist that crosses the blood-brain barrier to a much smaller degree. In this study, 133 hospice patients taking long-term moderate-dose to high-dose opioids were randomized to receive methylnaltrexone (42 patients received 0.15 mg/kg, 20 received 0.30 mg/kg) or matching placebo. All participants had a terminal illness but were expected to live at least 1 month, and all had been taking a stable regimen of laxatives and opioids for at least 3 days prior to enrollment. Patients were followed up for 2 weeks, then came the 3-month open label phase of the study. Groups were balanced at the start of the study except that the mean and median doses of opiods (150 mg vs 100 mg per day of oral morphine equivalent) were higher in the intervention group; analysis was by intention to treat. The mean age of participants was 71 years, 44% were male, and 58% had cancer as their primary diagnosis. The intervention was more effective than placebo across several outcomes, including percentage with a bowel movement within 4 hours of the initial dose (48% vs 15%; P < .001; number needed to treat [NNT] = 4), percentage with a bowel movement within 4 hours after at least 2 of the first 4 doses (52% vs 8%; P < .001; NNT = 3), and symptoms scores. Pain (17% vs 13%), flatulence (13% vs 7%), dizziness (8% vs 3%), nausea (11% vs 7%), and fever (8% vs 3%) occurred more often in the intervention group. There were no significant changes in the pain scores or withdrawal symptoms between the 2 groups.