Article Text

Download PDFPDF

Assessing the efficacy of peripherally acting µ-opioid receptor antagonists (PAMORAs) in the treatment of opioid-induced constipation: authors reply
  1. Nicholas E Burr1,
  2. Alexander C Ford2
  1. 1 Leeds Gastroenterology Institute, St. James’s University Hospital, Leeds, UK
  2. 2 Leeds Institute of Biomedical and Clinical Sciences, University of Leeds, Leeds, UK
  1. Correspondence to Professor Alexander C Ford, Leeds Institute of Biomedical and Clinical Sciences, University of Leeds, Leeds LS9 7TF, UK; alexf12399{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

We thank Chedid and Camilleri for their carefully composed letter,1 and are delighted to have provoked further debate about the relative efficacy of drugs acting on µ-opioid receptors in the treatment of opioid-induced constipation (OIC) with our network meta-analysis.2 In fact, we had also written to another journal concerning a recent network meta-analysis examining this issue,3 which had reported erroneously that subcutaneous methylnaltrexone appeared to perform better than other medications for OIC.4 The authors confirmed in their reply to our letter that they had mistakenly included redundant data from duplicate publications of the same two randomised-controlled trials (RCTs) of methylnaltrexone in this meta-analysis,5 likely accounting for their findings.4

In response to the specific issues raised by Chedid and Camilleri,1 first as we reported in the supplementary materials section of our manuscript there were no significant differences in rates of reversal of analgesia with any of the therapies studied, although only 15 of the eligible and included RCTs reported these data. Second, our primary analysis only included RCTs that used a fixed 2:1 dose ratio of oxycodone/naloxone,6 7 hence if their suggestion were followed this result would be unchanged. Third, the end points used in the included RCTs were indeed heterogeneous. As a result, we performed analyses according to multiple separate end points, which we reported in both the main article and the supplementary materials, in order to assess the robustness of our findings. In some of these analyses naldemidine was ranked first, but naloxegol was not ranked first in any of these analyses, hence our statement in the abstract of the article that naloxone and naldemidine appeared to be the most efficacious treatments for OIC. Chedid and Camilleri correctly point out that we excluded an RCT by Webster et al,8 but only after contacting the authors to obtain dichotomous data, which we were informed were not available. Finally, they discuss oral morphine equivalents (OMEs). It is true that the participants in the two RCTs reported in a single paper by Chey et al were receiving a higher OME than in other RCTs,9 although the absolute magnitude of this difference, relative to most of the other studies of µ-opioid receptor antagonists, is small. However, accepting that the participants in the trial by Simpson et al were receiving substantially lower OMEs than those in Chey et al,6 9 excluding Simpson et al from our primary analysis did not improve the ranking of naloxegol to any great extent. Naldemidine was ranked first (P-score = 0.82), followed by alvimopan (P-score = 0.79) and naloxone (P-score = 0.78), with naloxegol seventh (P-score = 0.36).

In an ideal world, and where an effective therapy for a condition already exists, drug companies would undertake head-to-head studies of novel agents rather than conducting placebo-controlled trials. However, these trials would be expensive to conduct, as it is likely that they would need huge numbers of patients in order to demonstrate superiority of one drug over another. Network meta-analysis can circumvent this problem to some extent, although unfortunately RCTs do not always use an identical design, recruit homogeneous groups of patients, or measure identical end points. Despite this, we believe the results of our network meta-analysis are still useful for healthcare providers and patients.



  • Contributors ACF and NEB drafted the letter and approved the final draft of the manuscript.

  • Competing interests None declared.

  • Patient consent Not required.

  • Provenance and peer review Not commissioned; internally peer reviewed.

Linked Articles