Efficacy of Treatments for Opioid-Induced Constipation: Systematic Review and Meta-analysis

Accepted Manuscript Efficacy of Treatments for Opioid-induced Constipation: A Systematic Review and Meta-Analysis Judy Nee, Mohammed Zakari, Michael A. Sugarman, Julia Whelan, William Hirsch, Shahnaz Sultan, Sarah Ballou, Johanna Iturrino, Anthony Lembo

PII: DOI: Reference:

S1542-3565(18)30087-9 10.1016/j.cgh.2018.01.021 YJCGH 55657

To appear in: Clinical Gastroenterology and Hepatology Accepted Date: 11 January 2018 Please cite this article as: Nee J, Zakari M, Sugarman MA, Whelan J, Hirsch W, Sultan S, Ballou S, Iturrino J, Lembo A, Efficacy of Treatments for Opioid-induced Constipation: A Systematic Review and Meta-Analysis, Clinical Gastroenterology and Hepatology (2018), doi: 10.1016/j.cgh.2018.01.021. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Efficacy of Treatments for Opioid-induced Constipation: Systematic Review and Meta-Analysis

RI PT

Short Title: Treatment of Opioid Induced Constipation: A Systematic Review and MetaAnalysis

SC

Grant support: none

Abbreviations:

NNH – number needed to harm NNT – number needed to treat OIC – opioid induced constipation OTC – over the counter

M AN U

CI – confidence interval

RR – relative risk

TE D

RCT – randomized controlled trials

AC C

EP

Authors: Judy Nee1, Mohammed Zakari1, Michael A Sugarman2, Julia Whelan1, William Hirsch1, Shahnaz Sultan3 Sarah Ballou1, Johanna Iturrino1, Anthony Lembo1

1. Beth Israel Deaconess Medical Center, Boston, MA, USA 2. Bedford Veterans Administration Medical Center, Bedford, MA, USA 3. University of Florida, Gainesville, FL, USA

Address correspondence and reprint requests to: Judy Nee, 330 Brookline Avenue, Boston, MA 02115; email: [email protected]; phone: 617-667-2193

Author contributions:

ACCEPTED MANUSCRIPT

Judy Nee – acquisition of data; analysis and interpretation of data; drafting of the manuscript

RI PT

Mohammed Zakari – acquisition of data; analysis and interpretation of data

SC

Michael Sugarman - analysis and interpretation of data; drafting of the manuscript; statistical analysis

M AN U

Julia Whelan – study concept and design; acquisition of data

Shahnaz Sultan – analysis and interpretation of data

TE D

William Hirsch - acquisition of data; analysis and interpretation of data; drafting of the manuscript

EP

Sarah Ballou - drafting of the manuscript; critical revision of the manuscript for important intellectual content

AC C

Anthony Lembo – study concept and design; analysis and interpretation of data; drafting of the manuscript; critical revision of the manuscript for important intellectual content

Conflict of interest statement: The above authors have no relevant conflicts of interest to report.

Manuscript Number: CGH 17-01222

ACCEPTED MANUSCRIPT

Title: Efficacy of Treatments for Opioid-induced Constipation: A Systematic Review and Meta-Analysis

SC

RI PT

Abstract: Background & Aims: Opioid induced constipation (OIC) is a common problem in patients on chronic opioid therapy for cancer-related and non-cancer related pain. Approved treatments for OIC are methylnaltrexone, naloxone, naloxegol, alvimopan, naldemedine, and lubiprostone. Since a meta-analysis performed in 2014, 2 new agents have been approved by the Food and Drug Administration (FDA) for treatment of OIC (naloxegol and naldemedine).

TE D

M AN U

Methods: We conducted a search of the medical literature following the protocol outlined in the Cochrane Handbook for systematic review. We searched MEDLINE, EMBASE, and EMBASE Classic, Web of Science, and the Cochrane Central Register of Controlled Trials until March 2017 to identify randomized controlled trials of peripheral u-opioid receptor antagonists (methylnaltrexone, naloxone, naloxegol, alvimopan, axelopran, or naldemedine), lubiprostone, or prucalopride. Response to therapy was extracted in a dichotomous assessment as an overall response to therapy. The effect of pharmacological therapies was pooled and reported as a relative risk (RR) of failure to respond to the treatment drug, with 95% CIs.

AC C

EP

Results: We included 27 placebo-controlled trials in our meta-analysis (23 trials evaluated μ-opioid receptor antagonists, 3 trials evaluated lubiprostone, and 1 trial evaluated prucalopride). In these trials, 5390 patients received a drug and 3491 received a placebo. Overall, μ-opioid receptor antagonists, lubiprostone, and prucalopride were superior to placebo for the treatment of OIC, with a RR of failure to respond to therapy of 0.70 (95% CI, 0.64–0.75) and an overall number needed to treat of 5 (95% CI, 4–7). When restricted to only FDA-approved medications for OIC, the RR of failure to respond to therapy was the RR was 0.69 (95% CI, 0.62-0.77) with a number needed to treat of 5 (95% CI, 4–7). Sensitivity analyses and meta-regression performed to account for heterogeneity showed that treatment was more likely to be effective in study populations taking higher doses of opiates at baseline or refractory to laxatives. Study duration and prespecified primary outcome did not affect the RR of failure. Participants who received μopioid receptor antagonists were significantly more likely to have diarrhea, abdominal pain, nausea, or vomiting than patients who received placebo.

ACCEPTED MANUSCRIPT

Conclusion: In a systematic review and meta-analysis, we found μ-opioid receptor antagonists to be safe and effective for the treatment of OIC. Prescription-strength laxatives (prucalopride, lubiprostone) are slightly better than placebo in reducing OIC.

M AN U

INTRODUCTION

SC

RI PT

KEY WORDS: opioid-induced bowel dysfunction; comparison; phase 3 trial; pharmacologic

With the widespread usage of opioids for non-cancer pain, undesired side effects such as sedation, respiratory depression, and GI symptoms are common. In particular, opioid induced constipation (OIC) is a common problem in individuals using chronic

TE D

opioids. OIC has been defined by Rome IV criteria as an abnormal change from baseline in bowel habits or defecation patterns after initiating opioid therapy, characterized by any of the following: reduced frequency of spontaneous bowel movements (SBMs)(<3

EP

bowel movements/week), development or worsening of straining to pass bowel movement, a sense of incomplete evacuation, or harder stool consistency or a patient’s

AC C

perception of distress related to bowel habits1,2. The true impact of OIC is likely underestimated due to underdiagnosis and

undertreatment. Current estimates of the prevalence of OIC in trials range from 15 to 70%2,3 in patients on chronic opioids, depending on the patient population, number of subjects, type and amount of opioid, definition of OIC, and outcome measurement. OIC can negatively impact an individual’s quality of life and work productivity and is

ACCEPTED MANUSCRIPT

associated with significant direct and indirect costs4,4,5. In addition, approximately a third of patients will reduce or discontinue prescribed opioids in an attempt to improve

RI PT

bowel function, which can result in worsening of pain symptoms6. Several newer treatments are currently approved by the FDA for the treatment of OIC including the peripherally acting μ-opioid receptor antagonists methylnaltrexone,

SC

naloxegol, naloxone, and recently naldemedine. Additionally, the prescription strength laxative lubiprostone is also approved by the FDA for the treatment of OIC. A previous

M AN U

meta-analysis by Ford et al. in 20137 included 17 trials and concluded that μ-opioid receptor antagonists are safe and effective for the treatment of OIC. Since then, both naloxegol and naldemedine have been approved by the FDA after completing Phase III clinical trials. Therefore, our aim is to update the previous meta-analysis by including

TE D

additional studies published for the treatment of OIC. METHODS

Search strategy and inclusion criteria

EP

We conducted a search of the medical literature following the protocol outlined in the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA)

AC C

statement8 using MEDLINE (1950 to March 2017), EMBASE (Elsevier Science: 1975Present) and EMBASE Classic (1947 to March 2017), Web of Science (Institute for Scientific Information: 1900 to March 2017), and the Cochrane Central Register of Controlled Trials (Update Software: 1996 to March 2017) (supplementary table 3). Randomized controlled trials (RCTs) evaluating the effect of medical therapy for OIC were eligible for inclusion. Studies using methylnaltrexone, naloxone, naloxegol,

ACCEPTED MANUSCRIPT

alvimopan, prucalopride, lubiprostone, axelopran (TD-1211), and naldemedine were identified. For RCTs that included an open-label extension, only the initial RCT period of

RI PT

studies were included in analyses. There is no universally accepted definition of OIC across studies. Two definitions of OIC were accepted including (1) constipation associated with initiation of opioids and

SC

(2) less than three spontaneous bowel movements (SBMs) per week with one or more

of symptoms of constipation hard stools, sensation of incomplete evacuation, moderate

M AN U

to severe straining in 25% of bowel movements) after initiation of opioids. Studies that recruited patients with organic or chronic idiopathic constipation were not eligible. Trials using any dose of pharmacological therapy were considered, and agents could be compared with each other or with placebo.

TE D

Studies on OIC were identified with the terms: constipation or gastrointestinal transit, chronic constipation, or slow transit. This search was also combined with the

EP

studies identified with the following terms: opiate alkaloids or analgesics, opioid, or the following free text terms: opiate, opiate$, opioid, or opioid$. This was also combined

AC C

with: narcotic antagonists, naltrexone, naloxone, naloxegol, mu opioid receptors, serotonin receptor agonists, serotonin receptor agonists, guanylate cyclase-coupled, guanylate cyclase, or chloride channels as well as methylnatrexone, alvimopan, 5-HT4, prucalopride, lubiprostone, linaclotide, naldemedine, or TD-1211. The search was conducted by a medical librarian (JW) and confirmed independently by two investigators (JN, MZ). The EMBASE database, which produced the greatest number of

ACCEPTED MANUSCRIPT

results, was searched both with keywords and subject terms, and with the McMaster validated EBM Center Hedge for RCTS. Results of the searches were combined with

RI PT

“OR” and duplicates removed. Subject experts and the research team selected highly relevant papers and the bibliographies of these papers were searched for additional

studies. Articles were assessed independently by two investigators (JN, MZ) using pre-

SC

designed eligibility forms, according to the pre-defined eligibility criteria.

M AN U

Outcome Assessment

The primary efficacy outcomes were defined as the efficacy of the therapy to fail compared to placebo. If possible, the author’s pre-specified outcome measure was used. The most common primary outcome was ≥3 complete spontaneous bowel

TE D

movements (CSBMs) a week over the trial period. If the primary endpoint could not be extracted through this method and responses from the authors could not be obtained, attempts were made to extract other clinically meaningful outcomes (e.g.improvement

EP

in SBMs, etc). Secondary efficacy outcomes included overall adverse events as well as diarrhea and, abdominal pain, nausea and vomiting.

AC C

Data extraction

All data were extracted independently by two investigators onto a Microsoft

Excel spreadsheet. as “response” or “no response” to pharmacologic treatment. The following clinical data were also extracted for each trial: setting (primary, secondary, or tertiary care), number of study centers, study country or countries, number of female patients, dose and duration of therapy, concomitant medications allowed, OIC criteria

ACCEPTED MANUSCRIPT

used, primary outcome measure used to define response to therapy. Data were extracted as intention-to-treat analyses, with drop outs assumed to be treatment

RI PT

failures, wherever trial reporting allowed. Assessment of bias risk & GRADE

The Cochrane Collaboration’s risk of bias tool was used to assess bias at the

M AN U

with discussion with third investigator (AL).

SC

individual study level (JN, MZ) (Supplementary table 2). Disagreements were resolved

The overall quality or certainty of evidence was assessed using the Grading of Recommendation, Assessment, Development, and Evaluation (GRADE9) framework. The overall quality or certainty of evidence across all outcomes is classified as high,

TE D

moderate, low, or very low. A summary of findings table, generated from the GradePro

EP

GDT10 platform was presented for each intervention (Table 2).

Data synthesis and statistical analysis

AC C

Data were pooled using random effects models. The effects of each

pharmacologic agent in OIC were expressed as the relative risk (RR) of failure to respond to therapy compared to placebo, with 95% CI. For studies that had multiple treatment groups with the same drug (e.g., comparing different dose levels), the groups were combined for analyses. The rate of adverse events was also expressed as the RRs for developing an adverse event on drug compared to placebo. The number needed to

ACCEPTED MANUSCRIPT

treat (NNT) was defined as the number of patients who would need to receive active therapy, over and above the placebo therapy, for one to experience an improvement in

RI PT

symptoms. The number needed to harm (NNH) was defined as the number of patients who would need to receive active therapy, over and above placebo therapy, for one to experience an adverse event leading to withdrawal. The NNT and NNH were calculated

SC

from the reciprocal of the absolute risk difference for efficacy and adverse events, respectively.

M AN U

All analyses were conducted with the “metafor” package11 for R12. Forest plots were generated to examine the mean RR for each drug (with 95% CIs). Funnel plots were also created to assess for evidence of asymmetry, and therefore possible publication bias or other small study effects. Heterogeneity across studies and drugs

14,15

TE D

was measured using the Q statistic13, which follows a chi-square distribution, and I 2 . Possible sources of residual heterogeneity were explored in sensitivity analyses.

Specifically, subgroup analyses were used to evaluate differences of effect in studies

EP

with cancer vs. non-cancer related pain as well as laxative vs. non-laxative refractory

AC C

patient populations. Subgroup analysis was also performed to determine the influence of different primary outcome definitions and study duration. Meta-regression was also performed based on baseline opioid dosage and placebo-response rate.

RESULTS Overall efficacy of OIC Agents

ACCEPTED MANUSCRIPT

The search strategy generated 886 citations, 35 of which appeared to be pertinent to the systematic review and were retrieved for further evaluation (figure 1).

RI PT

Of these, there were 26 publications with a total of 28 placebo-controlled trials were identified in this meta-analysis; one small methylnaltrexone trial was removed due to

being a clear outlier16 (see funnel plot analysis below), resulting in a total sample of 27 trials. All of the studies were published in English. In total, 24 trials evaluated –μ-opioid

SC

receptor antagonists in OIC including naloxone (n = 517–21), alvimopan (n = 422–25), naloxegol

M AN U

(n = 326,27), methylnatrexone (n = 716,28–32), naldemedine (n =433–35), and axelopran (TD-1211) (n =136). Additionally, four trials examined drugs that were not μ-opioid receptor

antagonists in the treatment of OIC, including three trials with lubiprostone37–39 and one with prucalopride40. Characteristics of each trial can be seen in Table 1. Each study’s

TE D

bias risk was assessed (supplementary table 2). Overall quality of evidence was evaluated with the GRADE assessment (Table 2). A funnel plot with all studies revealed significant asymmetry (Egger’s test:

EP

Z = -5.06, p = < .001), with one clear outlier16, a small methylnaltrexone study that found

AC C

results strongly in favor of the drug and which was inconsistent with other methylnaltrexone studies. In this study, all eleven participants administered the drug had a favorable response, and zero out of eleven participants administered placebo had a response. After removing this trial, the funnel plot was still significantly asymmetric (Z = -4.79, p < .001), but with outliers with regard to both high and low efficacy (Figure 2) that could be partially due to the inclusion of multiple drug classes and the overall heterogeneity amongst the 27 remaining trials(Q(26) = 105.0, p < .001; I2 = 77.9%). A

ACCEPTED MANUSCRIPT

comparison between low-bias (RR = 0.67, 95% CI: 0.58-0.78) and other trials (RR = 0.71, 95% CI: 0.64-0.78) revealed no significant difference with regard to efficacy (Q(1) = 0.23,

RI PT

p = .63); thus, no trials were removed for reasons pertaining to suspected bias. In total, 5390 participants were randomized to receive drug treatment and 3491 received placebo. A total of 2784 (51.3%) of those who received the drug had a

SC

favorable response, compared to 1157 (33.1%) of individuals randomized to placebo, for an overall number needed to treat (NNT) of 5 (Figure 3). The overall RR for failure to

M AN U

respond to therapy was significantly lower for those who received drug rather than placebo treatment (RR = 0.70, 95% CI 0.64-0.75). When this analysis was restricted to solely drugs and dosage levels that have been FDA-approved for the treatment of OIC (i.e., lubiprostone, methylnaltrexone, naldemedine, naloxegol, and naloxone), the RR

TE D

was 0.69 (95% CI: 0.62-0.77) with an NNT of 5 (95% CI: 4-7). Individually, the NNTs for lubiprostone, naloxegol, methylnaltrexone, naloxone, and naldemedine were 15 (95%

respectively.

EP

CI: 9-51), 7 (95% CI: 4-26), 3.4 (95% CI: 3-6), 4 (95% CI: 4-6), and 5 (95% CI: 4-8),

AC C

Given the heterogeneity observed in the included studies, meta-regression was then performed for possible moderating factors. The mean dosage of opioids at baseline in each study (converted to morphine equivalents) was a significant predictor of trial outcome (Q(1) = 4.75, p = .029), with higher doses associated with a lower RR. For example, studies with baseline morphine equivalent doses of 100mg, 200mg, and 400mg had predicted RRs of 0.79, 0.69, and 0.54, respectively. Secondly, the placebo response rate was evaluated. A wide range of placebo responses were observed across

ACCEPTED MANUSCRIPT

trials, ranging between 0%17 and 56%22. However, the proportion of placebo responders was not significantly related to the overall RR (Q(1) = 2.13, p = .145); that is,

RI PT

studies with higher rates of placebo responders also tended to have higher rates of drug responders, and the difference between drug and placebo appeared to be independent from the proportion of placebo responders within each trial.

Sensitivity analysis was performed to determine if the primary outcome of

SC

interest accounted for heterogeneity. Six studies22,23,26,34,36,41 utilized a more stringent

M AN U

outcome criteria, defined as improvement in weekly CSBM and more than three bowel movements per week; however, these six studies did not have a significantly different RR (0.75, 95% CI: 0.68-0.83) compared to the other 21 studies (RR = 0.68, 95% CI: 0.610.75; Q(1) = 0.83, p = .364). Furthermore, sensitivity analysis in regards to laxative

TE D

status showed that laxative refractory populations had superior efficacy with treatment (RR of failure 0.57, 95% CI: 0.47-0.69) than studies with non-laxative refractory populations (RR of failure = 0.75, 95% CI: 0.69-0.82); Q(1) = 6.89, p = .009).

EP

With regard to study duration, five trials28,29,31–33 had a duration of two weeks or less, and had a significantly superior response to drug treatment (RR of failure = 0.67,

AC C

95% CI: 0.61-0.74) compared to the other 22 trials (RR of failure = 0.85, 95% = 0.82-0.88) (p<0.001). However, amongst these 22 trials, there was not a significant difference (Q(1) = 3.43, p = .064) between trials with a duration ranging between 3 and 11 weeks17,19,20,24,25,25,30,36,40,41 (k = 10, RR = 0.81, 95% CI: 0.77-0.86) and trials that were 12 weeks or longer18,21–23,26,27,34,37–39 (k = 12, RR = 0.87, 95% CI: 0.83-0.91). Thus, while the

ACCEPTED MANUSCRIPT

shortest trials appeared to have the greatest efficacy, there was not a clear systematic

RI PT

relationship between overall trial duration and RR.

Methylnaltrexone. In total, six RCTs studied the use of methylnaltrexone in OIC after removing the small trial by Yuan16 as described above. A total of 1,619 patients participated in these trials, including 1,004 patients who received dosages of

SC

methylnatrexone ranging from as low as 12mg28–30,32 to 450mg41, from 0.5 to 4 weeks in

M AN U

duration. Methylnaltrexone was significantly more efficacious than placebo (RR = 0.62 [95% CI 0.49-0.78], p < .001). Out of the 1,004 individuals who received methylnatrexone, 485 (48.3%) failed to respond compared to the 442 patients of 618 on placebo (71.5%). There was significant heterogeneity between studies (Q(5) = 28.18, p <

TE D

.001; I2 = 77.2). Three studies29,31,32 with patients with cancer-related pain had significantly better results in favor of the drug (RR = 0.51, 95% CI: 0.41-0.63) than studies28,30,41 with non-cancer related pain (RR = 0.75, 95% CI = 0.63-0.90; Q(1) = 7.44, p

EP

= .006). After accounting for this variable as a moderator, the residual heterogeneity in methylnaltrexone studies was no longer significant (Q(4) = 7.37, p = .118, I2 = 46.9).

AC C

Only one study41 examined the other FDA-approved dose level of 450mg, which had an RR of 0.79 (95% CI: 0.66-0.94). The overall quality of evidence was high.

Naloxone. Five total trials17–21 examined the use of naloxone in OIC. A total of 482 patients received naloxone for a duration of three to twelve weeks, and 212 (43.9%) failed to respond compared to 250 of 356 patients on placebo (70.2%) (RR = 0.63, p <

ACCEPTED MANUSCRIPT

.001). There was no significant heterogeneity across studies (Q(4) = 1.23, p = .873; I2 = 0.0%). Because we rated down for risk of bias, the overall quality of evidence was moderate.

RI PT

Alvimopan. Four trials examined the use of alvimopan compared to placebo22–25.

Overall, 433 of 1,060 (40.8%) patients receiving alvimopan failed to respond compared to 280 of 519 (53.9%) patients on placebo. Three trials evaluated dosages of 0.5mg and

SC

1mg daily22–24, but the fourth trial used both a 1mg or 2mg total daily dosage25.

M AN U

Alvimopan was significantly more efficacious than placebo with a RR of failure of 0.68 (p < .001), with some heterogeneity between trials (Q(3) = 6.86, p = .076, I2 = 56.3%). This heterogeneity was explained by one study25 that used a higher dosage compared to the other trials and also showed the lowest RR of failure in the treatment of OIC. Because

TE D

we rated down for inconsistency, the overall quality of evidence was moderate.

Naldemedine. In total, there were four Phase III RCTs of naldemedine33–35. Overall, 367

EP

of 823 (44.6%) patients on naldemedine failed to respond compared to 459 of 702 (65.4%) patients on placebo. The drug was significantly more efficacious than placebo

AC C

(RR = 0.65, p < .001), with significant heterogeneity across trials (Q(3) = 11.20, p = .011, I2 = 79.6%), mostly due to one trial33 that had more favorable results for the drug compared to the other trials. In this study, naldemedine was randomized amongst cancer-related pain patients with 28.8% not responding on naldemedine compared to 65% on placebo. Comparatively, the three large studies34,35 examined non-cancer patients with chronic pain with 339 of 726 (46.7%) failing to respond to naldemedine

ACCEPTED MANUSCRIPT

compared to 397 of 607 who received placebo (65.4%). The overall quality of evidence was moderate because we rated down for inconsistency.

RI PT

Naloxegol. Three trials26,27 examined naloxegol compared to placebo. In total 1,522 patients were randomized to placebo or naloxegol. Of these, 572 of 981 (58.3%)

patients who received naloxegol failed to respond, compared to 382 of 541 (70.6%) who

SC

received placebo; the overall RR for naloxegol compared to placebo was 0.77 (95% CI

0.61-0.97, p = .026). The two trials26 that compared the 12.5mg and 25mg dose levels

M AN U

(the two doses currently approved by the FDA) did not find a significant difference in efficacy (RR for a superior response for 25mg compared to 12mg: 1.11, 95% CI: 0.941.30; p = .208). There was significant heterogeneity amongst the three trials (Q(2) = 8.44, p = .015; I2 = 86.4%), primarily because one study27 found superior efficacy

TE D

compared to the other two trials26, which were both conducted by the same research group. Since we rated down for inconsistency, the overall quality of evidence was

AC C

EP

moderate.

Lubiprostone. Three trials37–39 examined lubiprostone compared to placebo. Of 647 patients on lubiprostone, 400 (61. 8%) failed to respond compared to 438 of 637 (68.8%) patients who received placebo. There was no significant heterogeneity amongst the three trials (Q(1) = 1.40, p = .50; I2 = 0.0%). These studies provide modest evidence supporting lubiprostone compared to placebo (RR = 0.90 [95% CI 0.83-0.97], p = .005). The overall quality of evidence was high.

ACCEPTED MANUSCRIPT

Prucalopride and Axelopran (TD-1211). Only one placebo-controlled double blind trial has examined prucalopride40, with modest results in favor of the drug (RR = 0.88, 95% CI

RI PT

0.68-0.98; p = .032). Likewise, only one trial36 has been conducted examining axelopran (TD-1211), providing evidence to support superior efficacy compared to placebo (RR =

SC

0.60, 95% CI 0.44-0.81; p = .001). The overall quality of evidence was high for

M AN U

prucalopride and moderate for axelopran (we rated down for imprecision).

Adverse Effects

Twenty-three placebo-controlled RCTs provided data regarding adverse events experienced within each group. In these trials, 58.2% of participants given a drug

TE D

experienced at least one new-onset adverse event, compared with 53.0% of those given placebo. This difference was significant (IRR = 1.10 [95% CI: 1.05-1.16], p < .001), with an overall number needed to harm (NNH) of 21 and no significant heterogeneity across

EP

trials (Q(22) = 22.26, p = .44; I2= 0.0%). Thus, the frequency of adverse effects (Table 2) was combined across all drug classes. Where available, data for the three most common

AC C

adverse effects of diarrhea (k =22), abdominal pain (k = 22), and nausea/vomiting (k = 19) were also extracted. Participants who received a study drug were significantly more likely to experience all three adverse events compared to placebo (ps < .009). The overall dropout rate due to adverse events (only 20 studies reported these data) 7.4% and 4.7% for drug and placebo, respectively (p = .002), resulting in an NNH of 36 (95% CI: 22-96).

ACCEPTED MANUSCRIPT

DISCUSSION

RI PT

The purpose of this systematic review and meta-analysis is to provide a synthesis of the growing literature on available OIC treatment. Our findings further

support the efficacy of μ-opioid receptor antagonists in the treatment of OIC. Despite

SC

variable medication dosages and significant heterogeneity across studies, all studies

showed superiority of these agents compared to placebo. With the addition of newer

M AN U

agents to treat OIC, this shows a slightly higher overall number needed to treat of 5 compared to the previous meta-analysis7. The systematic review and meta-analysis of naloxegol, methylnaltrexone, naloxone, and naldemedine provided moderate to high quality of evidence.

TE D

After pooling all treatment for OIC, adverse effects were significantly higher in those receiving active drug compared to placebo, with a NNH of 20.

Amongst the

specific adverse events, participants who received drug treatment were significantly

EP

more likely to experience diarrhea, abdominal pain, nausea/vomiting compared to placebo, although the frequencies for these events were still relatively low in drug-

AC C

treated groups (8.5%, 12.8%, and 11.5% for diarrhea, abdominal pain, and nausea/vomiting, respectively). This meta-analysis confirms the overall safety of these medications in over 5,000 patients receiving u-opioid antagonists while maintaining opioid analgesic affect. While a limitation of this study is the significant heterogeneity across 27 studies, we have shown through sensitivity analysis and meta-regression the potential factors

ACCEPTED MANUSCRIPT

contributing to this heterogeneity. This is likely due to the inclusion of multiple agents, varying baseline opioid usage, and different subject populations (cancer vs. non-cancer related pain). Baseline narcotic dosage ranged from morphine equivalents of 20mg to

RI PT

2000mg not mentioned in others29,31,32,38,39. Opioid types varied from oxycodone to morphine to methadone. Although the type of opioid should not influence the

SC

prevalence of OIC, the differences in frequency and dosage may account for differences in symptom severity. In one study, daily usage of opioids led to constipation in 81%

M AN U

compared to 46% taking opioids only up to three times a week4. We have shown that higher baseline morphine equivalent dosage was associated with proportionately superior efficacy. Likewise, populations which were refractory to laxatives also showed superior efficacy. Both of these findings suggest that the subset of patients on opioids

TE D

with more severe constipation respond more favorably to agents targeted mechanistically against µ-opioid receptors. On the other hand, OIC criteria and duration of studies did not influence outcome in the treatment of OIC.

EP

Three trials using lubiprostone and one using prucalopride demonstrated that both agents were efficacious in the treatment of OIC; however, both drugs were

AC C

comparatively less efficacious when compared to u-opioid receptor antagonists, supporting previous findings7. The NNT of lubiprostone was higher at 15 compared to other u-opioid receptor antagonists. Laxatives are often used as first-line therapy in OIC. The effect of OTC laxatives in the treatment of OIC, however, is difficult to assess due to variable methodology, missing data, and risk of bias42,43. While our analysis supports

ACCEPTED MANUSCRIPT

the usage of lubiprostone for OIC compared to placebo, prescription laxatives may not be superior to over-the-counter laxatives.

RI PT

Despite Food and Drug Administration (FDA) approval for the u-opioid antagonists naloxegol, methylnatrexone, fixed-dosage oxycodone/naloxone, and

naldemedine no formal guidelines exist for their usage in clinical practice; thus, making

SC

them underutilized. While some consensus44 is developing in regards to utilizing these agents in clinical practice for those patients with chronic pain on opiate treatment, this

M AN U

provides further evidence demonstrating the efficacy and safety of u opioid receptor

AC C

EP

TE D

antagonists and lubiprostone in the treatment of OIC.45

ACCEPTED MANUSCRIPT

References: 1. Gaertner, J. et al. Definitions and outcome measures of clinical trials regarding opioid-induced constipation: a systematic review. J. Clin. Gastroenterol. 2015; 49: 9–16.

review of efficacy and safety. Pain 2004; 112: 372–380.

RI PT

2. Kalso, E., Edwards, J. E., Moore, R. A. & McQuay, H. J. Opioids in chronic non-cancer pain: systematic

3. Moore, R. A. & McQuay, H. J. Prevalence of opioid adverse events in chronic non-malignant pain:

SC

systematic review of randomised trials of oral opioids. Arthritis Res. Ther. 2005; 7: R1046-1051. 4. Bell, T. J. et al. The prevalence, severity, and impact of opioid-induced bowel dysfunction: results of

M AN U

a US and European Patient Survey (PROBE 1). Pain Med. Malden Mass 2009; 10: 35–42. 5. Candrilli, S. D., Davis, K. L. & Iyer, S. Impact of constipation on opioid use patterns, health care resource utilization, and costs in cancer patients on opioid therapy. J. Pain Palliat. Care Pharmacother. 2009; 23: 231–241.

TE D

6. Kurz, A. & Sessler, D. I. Opioid-induced bowel dysfunction: pathophysiology and potential new therapies. Drugs 2003; 63: 649–671.

7. Ford, A. C., Brenner, D. M. & Schoenfeld, P. S. Efficacy of pharmacological therapies for the

EP

treatment of opioid-induced constipation: systematic review and meta-analysis. Am. J. Gastroenterol. 2013; 108: 1566–1574; quiz 1575.

AC C

8. Moher, D., Liberati, A., Tetzlaff, J. & Altman, D. G. Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. J. Clin. Epidemiol. 2009; 62: 1006–1012. 9. Guyatt, G. H. et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ 2008; 336: 924–926. 10. GRADEpro GDT: GRADEpro Guideline Development Tool [Software]. McMaster Univ. Dev. Evid. Prime Inc Available Gradeproorg. 2015. 11. Viechtbauer, W. Conducting Meta-Analyses in R with the metafor Package. J. Stat. Softw. 2010; 36.

ACCEPTED MANUSCRIPT

12. R Development Core Team. R: a language and environment for statistical computing. 2011. 13. Hedges, L. & Olkin, I. Statistical methods for meta-analysis. 1, Academic Press, 1985. 14. Higgins, J. & Thompson, S. Quantifying heterogeneity in a meta-analysis. Stat. Med. 2002; 21: 1539–

RI PT

1558.

15. Higgins, J. P. T., Thompson, S. G., Deeks, J. J. & Altman, D. G. Measuring inconsistency in metaanalyses. BMJ 2003; 327: 557–560.

randomized controlled trial. JAMA 2000; 283: 367–372.

SC

16. Yuan, C. S. et al. Methylnaltrexone for reversal of constipation due to chronic methadone use: a

M AN U

17. Liu, M. & Wittbrodt, E. Low-dose oral naloxone reverses opioid-induced constipation and analgesia. J. Pain Symptom Manage. 2002; 23: 48–53.

18. Löwenstein, O. et al. Combined prolonged-release oxycodone and naloxone improves bowel function in patients receiving opioids for moderate-to-severe non-malignant chronic pain: a

TE D

randomised controlled trial. Expert Opin. Pharmacother. 2009; 10: 531–543. 19. Meissner, W. et al. A randomised controlled trial with prolonged-release oral oxycodone and naloxone to prevent and reverse opioid-induced constipation. Eur. J. Pain Lond. Engl. 2009; 13: 56–

EP

64.

20. Sanders, M. et al. New Formulation of Sustained Release Naloxone Can Reverse Opioid Induced

AC C

Constipation Without Compromising the Desired Opioid Effects. Pain Med. Malden Mass 2015; 16: 1540–1550.

21. Simpson, K. et al. Fixed-ratio combination oxycodone/naloxone compared with oxycodone alone for the relief of opioid-induced constipation in moderate-to-severe noncancer pain. Curr. Med. Res. Opin. 2008; 24: 3503–3512.

ACCEPTED MANUSCRIPT

22. Irving, G. et al. A randomized, placebo-controlled phase 3 trial (Study SB-767905/013) of alvimopan for opioid-induced bowel dysfunction in patients with non-cancer pain. J. Pain Off. J. Am. Pain Soc. 2011; 12: 175–184.

RI PT

23. Jansen, J.-P. et al. A randomized, placebo-controlled phase 3 trial (Study SB-767905/012) of

alvimopan for opioid-induced bowel dysfunction in patients with non-cancer pain. J. Pain Off. J. Am. Pain Soc. 2011; 12: 185–193.

SC

24. Paulson, D. M. et al. Alvimopan: an oral, peripherally acting, mu-opioid receptor antagonist for the treatment of opioid-induced bowel dysfunction--a 21-day treatment-randomized clinical trial. J. Pain

M AN U

Off. J. Am. Pain Soc. 2005; 6: 184–192.

25. Webster, L. et al. Alvimopan, a peripherally acting mu-opioid receptor (PAM-OR) antagonist for the treatment of opioid-induced bowel dysfunction: results from a randomized, double-blind, placebocontrolled, dose-finding study in subjects taking opioids for chronic non-cancer pain. Pain 2008;

TE D

137: 428–440.

26. Chey, W. D. et al. Naloxegol for opioid-induced constipation in patients with noncancer pain. N. Engl. J. Med. 2014; 370: 2387–2396.

EP

27. Webster, L. et al. A phase 2, double-blind, randomized, placebo-controlled, dose-escalation study to evaluate the efficacy, safety, and tolerability of naloxegol in patients with opioid-induced

AC C

constipation. Pain 2013; 154: 1542–1550. 28. Anissian, L. et al. Subcutaneous methylnaltrexone for treatment of acute opioid-induced constipation: phase 2 study in rehabilitation after orthopedic surgery. J. Hosp. Med. 2012; 7: 67–72. 29. Bull, J. et al. Fixed-Dose Subcutaneous Methylnaltrexone in Patients with Advanced Illness and Opioid-Induced Constipation: Results of a Randomized, Placebo-Controlled Study and Open-Label Extension. J. Palliat. Med. 2015; 18: 593–600.

ACCEPTED MANUSCRIPT

30. Michna, E. et al. Subcutaneous methylnaltrexone for treatment of opioid-induced constipation in patients with chronic, nonmalignant pain: a randomized controlled study. J. Pain Off. J. Am. Pain Soc. 2011; 12: 554–562.

patients. J. Support. Oncol. 2009; 7: 39–46.

RI PT

31. Slatkin, N. et al. Methylnaltrexone for treatment of opioid-induced constipation in advanced illness

32. Thomas, J. et al. Methylnaltrexone for opioid-induced constipation in advanced illness. N. Engl. J.

SC

Med. 2008; 358: 2332–2343.

33. Murata, T. et al. Treatment of opioid-induced constipation with naldemedine in patients with

M AN U

cancer: onset of action in a randomized phase 3 trial. Ann. Oncol. 2016; 27: 1466P. 34. Hale, M., Wild, J., Reddy, J., Yamada, T. & Arjona Ferreira, J. C. 598 Efficacy and Safety of Naldemedine for the Treatment of Opioid-Induced Constipation in Subjects With Chronic NonCancer Pain Receiving Opioid Therapy: Results From Two Phase 3 Clinical Trials. Gastroenterology

TE D

2016; 150: S121-122.

35. Webster, L. R., Yamada, T., Ferreira, A. & Camilo, J. A Phase 2b, Randomized, Double-Blind PlaceboControlled Study to Evaluate the Efficacy and Safety of Naldemedine for the Treatment of Opioid-

EP

Induced Constipation in Patients with Chronic Noncancer Pain. Pain Med. doi:10.1093/pm/pnw325 36. Vickery, R. et al. TD-1211 demonstrates a durable increase in bowel movement frequency and

AC C

return toward normal bowel function in a 5-week PH2B opioid-induced constipation (OIC) study. J. Pain 2014; 15: S77.

37. Spierings, E. Opioid-induced Bowel Dysfunction Pivotal Assessment of Lubiprostone (OPAL). 2015. 38. Cryer, B., Katz, S., Vallejo, R., Popescu, A. & Ueno, R. A randomized study of lubiprostone for opioidinduced constipation in patients with chronic noncancer pain. Pain Med. Malden Mass 2014; 15: 1825–1834.

ACCEPTED MANUSCRIPT

39. Jamal, M. M., Adams, A. B., Jansen, J.-P. & Webster, L. R. A randomized, placebo-controlled trial of lubiprostone for opioid-induced constipation in chronic noncancer pain. Am. J. Gastroenterol. 2015; 110: 725–732.

RI PT

40. Sloots, C. E. J., Rykx, A., Cools, M., Kerstens, R. & De Pauw, M. Efficacy and safety of prucalopride in patients with chronic noncancer pain suffering from opioid-induced constipation. Dig. Dis. Sci. 2010; 55: 2912–2921.

SC

41. Rauck, R., Slatkin, N. E., Stambler, N., Harper, J. R. & Israel, R. J. Randomized, Double-Blind Trial of Oral Methylnaltrexone for the Treatment of Opioid-Induced Constipation in Patients with Chronic

M AN U

Noncancer Pain. Pain Pract. Off. J. World Inst. Pain 2016. doi:10.1111/papr.12535 42. Candy, B. et al. Laxatives for the management of constipation in people receiving palliative care. in Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd. 2015. doi:10.1002/14651858.CD003448.pub4

TE D

43. Candy, B., Jones, L., Goodman, M. L., Drake, R. & Tookman, A. Laxatives or methylnaltrexone for the management of constipation in palliative care patients. Cochrane Database Syst. Rev. CD003448. 2011. doi:10.1002/14651858.CD003448.pub3

EP

44. Müller-Lissner, S. et al. Opioid-Induced Constipation and Bowel Dysfunction: A Clinical Guideline. Pain Med. Malden Mass. 2016. doi:10.1093/pm/pnw255

AC C

45. Argoff, C. E. et al. Consensus Recommendations on Initiating Prescription Therapies for OpioidInduced Constipation. Pain Med. Malden Mass 2015; 16: 2324–2337.

ACCEPTED MANUSCRIPT Table 1. Characteristics of randomized controlled trials of OIC agents vs. placebo. This table has been adapted from Ford et al

• Spierings 2015

37

NALOXONE 19 Meissner 2009

Simpson 2008

21

Lowenstein 2009

17

Liu 2002

Sanders 2015

18

Drug, dosage, schedule, duration of therapy

Criteria to define response to therapy

US and Canada, 79 sites, secondary and tertiary care

Chronic nonmalignant pain, not refractory

418 (64%)

24mcg orally BID x 12 weeks

SBM within 24hrs of first dose

US and Europe, 103 sites, primary and secondary care

Chronic nonmalignant pain, laxative status unknown

Stable dose >1 month, and <3 SBMs per week, with ≥1: hard/very hard stools, sensation of incomplete evacuation, moderate to very severe straining on >25% of BMs Stable dose >1 month, and <3 SBMs per week, with ≥1: hard/very hard stools, sensation of incomplete evacuation, moderate to very severe straining on 25% of BMs

431 (63%)

24mcg orally BID x 12 weeks

≥1 SBM improvement over baseline frequency for all treatment weeks and >= 3 SBMs/wk in 9 of 12 tx weeks

84 US and 4 Canadian sites, setting not reported

Chronic nonmalignant pain, refractory laxative status

Stable dose of opioids (any dosage) during 30 day screening period AND diagnosis of OIC or chronic constipation exacerbated with initiation of opioids

435 (61.1%)

24mcg orally BID x 12 weeks

≥3 SBM per week in over 50% of weeks (patient must be on treatment greater than 8 weeks)

Germany, 28 sites, secondary and tertiary care

Chronic nonmalignant pain, not laxative refractory Chronic nonmalignant pain, not laxative refractory

Stable oxycodone dose (40, 60, or 80mg/day) with concomitant constipation

202 (62.9%)

Naloxone 10m, 20, or 40mg qday

No need for laxatives

Opioid dose equivalent of >=20 to <=50mg/day of oxycodone, and constipation caused or aggravated by an opioid

322 (60.9)

Oxycodone PR/naloxone PR orally in a 2:1 fixed dose ratio for 12 weeks Oxycodone PR/naloxone PR orally 2:1 fixed dose ratio for 12 weeks Naloxone 2mg or 4mg daily

>=3 CSBMs/wk after 4 weeks

Full opioid agonist (oral or transdermal) but transdermal

Change in quality of life (slightly to substantially improved)

4 European countries, multiple sites, primary and secondary care Multi-national, multiple sites, secondary care USA, 1 site, secondary care

20

Europe, multiple sites, setting not reported

Chronic nonmalignant pain, not laxative refractory Chronic nonmalignant pain, laxative refractory Chronic nonmalignant pain, laxative

SC

RI PT

Total pts (% female)

M AN U

39

Criteria for OIC

TE D

• Jamal 2015

Patient group

EP

LUBIPROSTONE 38 • Cryer 2014

Country, # Centers, Setting

AC C

Study

7

Stable oxycodone dose 60-80mg/day with <3 CSBMs per week caused or worsened by opioids

265 (68.3)

Onset of constipation corresponds to their use of opioids and were on a stable dose of opioids for control of their pain

9 (55%)

Stable dose of opioids for the 3 weeks prior to the study, and <3 CSBMs per week caused or worsened by opioids

40 (52%)

>=3 CSBMs/wks at 4 wks

Satisfaction with relief of constipation

ACCEPTED MANUSCRIPT refractory

Multi-national, 153 sites, secondary and tertiary care Multi-national (North America and Europe), 148 sites, secondary and tertiary care

Chronic nonmalignant pain, laxative status not reported Chronic nonmalignant pain, laxative status not reported

Europe and US, 257 sites, secondary and tertiary sites

Chronic nonmalignant pain, not laxative refractory

• Chey 2014 – 05*

Europe and US, 257 sites, secondary and tertiary sites

Chronic nonmalignant pain, not laxative refractory

• Webster 2013

4 sites (Germany, US, Romania, Canada), setting not reported

Chronic nonmalignant pain, not laxative refractory

Jansen 201123

NALOXOGEL 26 • Chey 2014 – 04*

26

27

Stable for one month on Morphine 30mg to 1000mg daily AND <3 BMs/week associated with ≥1 of: hard/lumpy stools, straining, sensation of obstruction, or incomplete evacuation in ≥ 25% of BMs Stable for one month on Morphine 30mg to 1000mg daily AND <3 BMs/week associated with ≥1 of: hard/lumpy stools, straining, sensation of obstruction, or incomplete evacuation in ≥ 25% of BMs Stable oral opioid regimen of 30 to 1000 mg/day oral morphine equivalent doses for 2 weeks AND <3 BMs/week associated with ≥1 of: hard/lumpy stools, straining, sensation of

Almivopan 0.5mg or 1mg orally daily x 3 weeks

RI PT

Irving 201122

Stable opioid dose for ≥ 30mg morphine or equivalent and history of decreased BMs since starting opioids, with ≥1 of: hard stools, straining, or incomplete evacuation ≥ 25% of BMs Stable opioid dose for ≥ 30mg morphine or equivalent and history of decreased BMs since starting opioids, with ≥1 of: hard stools, straining Stable opioid dose for ≥ 30mg morphine or equivalent and history of decreased BMs since starting opioids, with ≥1 of: hard stools, straining

168 (58.3)

522 (63.8%)

≥ 1BM within 8hours of study drug each day during the treatment period

Almivopan 0.5mg BID, 1mg daily or 1mg BID orally x 6 weeks

≥3 CSBMs/week (different from Ford article)

485 (64.0%)

Almivopan daily or 0.5mg BID x 12 weeks

518 (63.0)

Almivopan daily or 0.5mg BID x 12 weeks

≥3 SBMs per week during therapy with an increase of ≥1 SBM per week from baseline ≥3 SBMs per week during therapy with an increase of ≥1 SBM per week from baseline

641 (61.3)

Naloxogel 12.5mg or 25 mg oral daily

≥3 SBMs/week and an increase from baseline of ≥ 1 SBM for ≥ 9 of 12 weeks and for ≥3 of the final 4 weeks

696 (63.3)

Naloxogel 12.5mg or 25 mg oral daily

≥3 SBMs/week and an increase from baseline of ≥ 1 SBM for ≥ 9 of 12 weeks and for ≥3 of the final 4 weeks

185(62. 2)

Naloxogel 5mg, 25mg, 50mg oral daily

Increase from baseline of ≥ 2 SBM per week across 28 day study

SC

9 countries, 113 sites, secondary and tertiary care

AC C

Webster 2008

Stable opioid dose for >=10mg morphine or equivalent, and <3 BMs/week associated with >= of: hard/lumpy stools, straining, sensation of obstruction, or incomplete evacuation

M AN U

25

Chronic nonmalignant pain or opioid dependent, not laxative refractory Chronic nonmalignant pain, not laxative refractory

TE D

USA, 22 sites, secondary and tertiary care

EP

ALMIVOPAN Paulson 200524

limited to less than 3 per cohort

ACCEPTED MANUSCRIPT obstruction, or incomplete evacuation

• Bull 2015*

29

Yuan 2000

16

USA, primary, secondary care

41

31

PRUCALOPRIDE 40 Sloots 2010

Country not reported, Multicenter, setting not reported USA, 17 sites, primary and secondary care

Chronic nonmalignant pain, laxative refractory

Multi-national, 60 sites, secondary and

Chronic nonmalignant pain, not laxative

Advanced illness (life expectancy 1 month or more), laxative refractory

RI PT

Advanced illness (life expectancy 1 month or more), laxative refractory Advanced illness (life expectancy 1 month or more), not laxative refractory Methadonemaintenance program, not laxative refactory

USA, 1 site, teriatry care

• Rauck 2016

Slatkin 2009

USA and Canada, 27 sites, primary and secondary care

Methlynatrexone 12mg SC daily 4-7 days

BM within 4 hours of first dose

460 (60.2)

Methylnatrexone 12mg SC daily or on alternate days for 4 weeks

≥3 rescue-free BMs per week

133 (56.7)

Methlnaltrexone 0.15mg/kg SC daily on alternate days for 2 weeks

≥3 rescue-free BMs per week

Stable dose of opioids (not ≥50% reduction in last 3 days) and laxatives (≥3 days) and <3 BMs in last week and no bowel movement in 24 hours or no bowel movement in 48 hours

230(48. 7)

Methylnaltrexone 8mg or 12mg every other day x 2 weeks

Rescue free BM within 4 hours after≥2 of the first 4 doses in the first week

Methadone-induced constipation with ≤1 BM in prior 3 days, or ≤2 BMs in previous week

22 (59.1)

Methylalnatrexone 0.015, 0.05, 0.1 then 0.2mg/kg intravenously every 1 minutes to a total of 0.365mg/kg, 2 days Methlynatrexone 150, 300, or 450mg orally daily x 4 weeks, then as required for 8 weeks Methylnatrexone 0.15 or 0.3mg/kg SC as a one-time dose

Defecation during or within 1 minute after cessation of the infusion

Prucalapride 2mg or 4mg orally daily for 4 weeks

Increase of ≥1 CSBM per week from baseline

SC

32

33 (66.7)

M AN U

Thomas 2008

Acutely constipated, received u-opioid anlagesics, and no BM x ≥48hours, difficulty in having an SBM, or inability to have SBM Opioid dose equivalent of > 50mg/day of morphine for ≥2 weeks, <3 rescue-free BMs per week associated with ≥1 of: hard or lumpy stools, straining, or incomplete evacuation Stable opioid regimen for ≥2 weeks, <3 BMs in previous week, and no clinically meaningful BM within 48hours before first study dose

TE D

30

Orthopedic procedure, laxative status not reported Chronic onmalignant pain, not laxative refractory

EP

Michna 2011

USA, 16 sites, secondary and tertiary care sites USA, multiple sites, setting not reported

Opioid dose equivalent to ≥50mg/day of morphine for ≥2 weeks, <3 rescue-free BMs per week associated with ≥1 of BSS 1 or 2, straining or incomplete evacuation on 25% of BMs Stable opioid regimen for ≥3 days, and no clinically meaningful BM within 48hours before first study dose

803(62. 8)

Constipation secondary to chronic daily opioid use

196 (61.2)

AC C

METHYLNALTREXONE 28 Anissian 2012

154 (45.5)

≥3 rescue-free BMs per week with an increase of ≥1 BM per week from baseline for 3 of first 4 weeks Rescue free BM within 24hours

ACCEPTED MANUSCRIPT

Country not reported, sites tertiary

Chronic malignancyrelated pain, laxative refractory Chronic nonmalignant pain, not laxative refractory

• Hale (COMPOSE I, 34 II), 2016*

Country not reported, sites not reported

• Webster 2017

69 sites US and Europe

Chronic nonmalignant pain, laxative refractory

Country not reported, sites not reported

Chronic nonmalignant pain

35

AXELOPRAN (TD1211) 36 • Vickery 2014*

Opioid induced constipation with <5 BMs in the prior 2 weeks of randomization

Reduced bowel movement (BM) frequency, development or worsening of straining while passing stool, sense of incomplete bowel evacuation, and hard stool formation following initiation of opioid therapy Stable dose of opioid therapy for ≥1month, <3 SBMs/week despite laxative use and one or more of the following symptoms in at least 25% of bowel movements: straining, feeling of incomplete evacuation, and/or hard/small stools, defined as Bristol Stool Scale (BSS) score lower than 3.

193 (38.3)

TE D

EP

≥3 SBMs per week with an increase of ≥1 BM per week from baseline

547 (60.1)

Naldemedine 0.2mg daily x 12 weeks

≥3 SBMs per week with an increase of ≥1 BM per week from baseline

244 (68.4%)

Naldemedine 0.1mg, 0.2mg, 0.4mg daily x 28 days

≥3 SBMs per week in the last 2 weeks of the treatment period and an increase of ≥1 BM per week from baseline in

217 (59.0)

TD-1211 5mg daily x 5 weeks

≥3 SBMs per week with an increase of ≥1 BM per week from baseline in 3 weeks of weeks 2 through 5

Opioid dose equivalent of > 30mg/day of morphine for ≥2 weeks, <5 rescue-free BMs during 2 week baseline associated with ≥1 of: straining or incomplete evacuation with 25% of BMs • Indicates studies that were not included in Ford et al’s 2013 meta-analysis

AC C

Naldemedine 0.2 mg daily x 2 weeks

RI PT

refractory

SC

tertiary care

M AN U

NALDEMEDINE 33 • Murata 2016*

ACCEPTED MANUSCRIPT

Methylnaltrexone compared to Placebo/Usual Care for OIC Quality of the evidence (GRADE)

1622 (6 RCTs)

⨁⨁⨁◯ HIGHa

M AN U

SC

Failure to respond

№ of participants (studies) Follow-up

Relative effect (95% CI)

RI PT

Outcomes

Cancer related pain: RR 0.51 (0.41-

Anticipated absolute effects Risk with Placebo/Usual Care

Risk difference with Methylnaltrexone

63 per 100

31 fewer per 100 (37 fewer to 23 fewer)

85 per 100

21 fewer per 100 (31 fewer to 8 fewer)

0.63) Non-cancer related pain: RR 0.75 (0.63-

0.90)

Outcomes

Failure to respond

AC C

Naloxone compared to Placebo/Usual care for OIC

EP

TE D

Explanations: a. The I2 was 77%, but we didn’t rate down for inconsistency since some of the heterogenity could be explained by different patient populations Three studies of methylnaltrexone in individuals with cancer-related pain had significantly better results (Q(1) = 7.44, p = .006) in favor of the drug than three studies in individuals with non-cancer related pain. Because the test of interaction was significant, we present the effects of methylnaltrexone in these two populations as separate effect estimates.

№ of participants (studies) Follow-up

Quality of Relative the evidence effect (GRADE) (95% CI)

838 (5 RCTs)

⨁⨁⨁◯ MODERATEa

Anticipated absolute effects Risk with Placebo/Usual care

RR 0.63 70 per 100 (0.56 to 0.71)

Risk difference with Naloxone 26 fewer per 100 (31 fewer to 20 fewer)

ACCEPTED MANUSCRIPT

Naloxone compared to Placebo/Usual care for OIC № of participants (studies) Follow-up

Quality of Relative the evidence effect (GRADE) (95% CI)

RI PT

Outcomes

Anticipated absolute effects Risk with Placebo/Usual care

Risk difference with Naloxone

Alvimopan compared to Placebo/Usual Care for OIC № of participants (studies) Follow-up

Quality of Relative the evidence effect (GRADE) (95% CI)

M AN U

Outcomes

SC

Explanations: a. We rated down for risk of bias because there were concerns about randomization, allocation concealment and blinding in some of the studies.

1579 (4 RCTs)

⨁⨁⨁◯ MODERATEa

Risk with Placebo/Usual Care

RR 0.68 54 per 100 (0.58 to 0.80)

TE D

Failure to respond

Anticipated absolute effects Risk difference with Alvimopan 17 fewer per 100 (23 fewer to 11 fewer)

AC C

EP

Explanations: a. We rated down for inconsistency (I2=56.3%). Some heterogeneity between trials may be due to different dosages: 0.5-1mg daily compared to 1-2mg daily.

Naldemedine compared to Placebo/Usual Care for OIC Bibliography: Outcomes

№ of participants (studies) Follow-up

Quality of Relative the evidence effect (GRADE) (95% CI)

Anticipated absolute effects Risk with Placebo/Usual Care

Risk difference with Naldemedine

ACCEPTED MANUSCRIPT

Naldemedine compared to Placebo/Usual Care for OIC Bibliography: Quality of Relative the evidence effect (GRADE) (95% CI)

1525 (4 RCTs)

⨁⨁⨁◯ MODERATEa

Anticipated absolute effects Risk with Placebo/Usual Care

RR 0.65 65 per 100 (0.52 to 0.82)

SC

Failure to respond

№ of participants (studies) Follow-up

RI PT

Outcomes

Risk difference with Naldemedine 23 fewer per 100 (31 fewer to 12 fewer)

M AN U

Explanations: a. We rated down for inconsistency which may be attributed to different patient populations. One study42 was in patients with cancer-related pain while the other three studies were in patients with non-cancer related pain43,44.

Naloxegol compared to Placebo/Usual Care for OIC

№ of participants (studies) Follow-up

Quality of Relative the evidence effect (GRADE) (95% CI)

1522 (3 RCTs)

⨁⨁⨁◯ MODERATEa

TE D

Outcomes

Risk with Placebo/Usual Care

RR 0.77 71 per 100 (0.61 to 0.97)

Risk difference with Naloxegol 16 fewer per 100 (28 fewer to 2 fewer)

AC C

EP

Failure to respond

Anticipated absolute effects

Explanations: a. There was significant heterogeneity across studies with I2= 86% therefore we rated down for inconsistency.

Lubiprostone compared to Placebo/Usual Care for OIC Outcomes

№ of participants (studies) Follow-up

Quality of the evidence (GRADE)

Relative effect (95% CI)

Anticipated absolute effects Risk with Placebo/Usual Care

Risk difference with Lubiprostone

ACCEPTED MANUSCRIPT

Quality of the evidence (GRADE)

Failure to respond

1284 (3 RCTs)

⨁⨁⨁◯ HIGH

Prucalopride compared to Placebo/Usual Care for OIC

№ of participants (studies) Follow-up

TE D

Outcomes

RR 0.90 (0.83 to 0.97)

Quality of Relative the evidence effect (GRADE) (95% CI)

Anticipated absolute effects Risk with Placebo/Usual Care

Risk difference with Lubiprostone

69 per 100

7 fewer per 100 (12 fewer to 2 fewer)

Anticipated absolute effects Risk with Placebo/Usual Care

196 (1 RCT)

⨁⨁⨁◯ HIGH

RR 0.88 77 per 100 (0.68 to 0.98)

№ of

Quality of

Relative

Risk difference with Prucalopride 9 fewer per 100 (25 fewer to 2 fewer)

AC C

EP

Failure to respond

Relative effect (95% CI)

SC

Outcomes

RI PT

№ of participants (studies) Follow-up

M AN U

Lubiprostone compared to Placebo/Usual Care for OIC

Axelopran compared to Placebo/Usual Care for OIC Outcomes

Anticipated absolute effects

ACCEPTED MANUSCRIPT

the evidence effect (GRADE) (95% CI)

201 (1 RCT)

⨁⨁⨁◯ MODERATEa

Risk with Placebo/Usual Care

RR 0.60 62 per 100 (0.44 to 0.81)

RI PT

Failure to respond

participants (studies) Follow-up

Risk difference with Axelopran 25 fewer per 100 (34 fewer to 12 fewer)

Explanations: a. We rated down for the following: unclear risk of bias, additionally, one study did not have optimal information size (few events).

M AN U

SC

GRADE Working Group grades of evidence High quality: We are very confident that the true effect lies close to that of the estimate of the effect Moderate quality: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low quality: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect Very low quality: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

AC C

EP

TE D

Table 2. GRADEpro quality of evidence summary

ACCEPTED MANUSCRIPT

Placebo

%

Total n

n with event

%

RR (log)

Adverse Event

k

Total n

n with event

Any event

23

5529

3219

58.2

3248

1722

53.0

1.10

Diarrhea

22

5399

458

8.5

3182

147

4.7

1.04

Abdominal pain

22

5377

689

12.8

3198

177

5.5

1.07

RI PT

Study drug

Nausea and/or vomiting

19

4705

542

11.5

2586

262

10.1

1.03

NNH

95% CI

1.05-1.16

20.6

14.3-36.8

1.03-1.05

26.3

20.2-37.7

1.04-1.09

16.4

12.3-16.4

1.01-1.05

45.6

25.0-256.4

SC

M AN U

TE D EP AC C

95% CI

Studies identified through literature search (n=886)

RI PT

ACCEPTED MANUSCRIPT

M AN U

Potential articles included in meta-analysis (n=619)

SC

Duplicates removed (n=267) (n=257)

TE D

Excluded after title/abstract review (n=584)

AC C

EP

Potential articles included for detailed assessment (n=35) Articles excluded (n=9) Not placebo controlled (n=4) Data not extractable (n=5)

Articles included in analyses (n=26)

Figure 1. PRISMA flow diagram of studies identified, excluded, and finally included for the metaanalysis

SC M AN U

0.30

AC C

0.60

EP

TE D

0.45

Standard Error

0.15

RI PT

0.00

ACCEPTED MANUSCRIPT

-1.5

-1

-0.5

0

0.5

1

Log Relative Risk

Figure 2. Funnel plot for all studies included in the analysis. Negative log relative risk values represent studies with superior efficacy for the drug compared to placebo.

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

Search Strategies for Meta-analysis on Drugs for Opiod Induced Constipation Strategies: PubMed

RI PT

(((("linaclotide"[Supplementary Concept] OR (MD-1100[All Fields] AND acetate[tw]) OR Linzess[tw] OR Linaclotide Acetate[tw]) OR ("Lubiprostone"[Mesh] OR Amitiza[tw]) OR ("naloxegol"[Supplementary Concept] OR Movantik[tw] OR NKTR-118[tw])) OR "micro opioid receptor antagonist"[tw])) AND ((("Constipation"[Mesh] OR constipation[tw] OR "Gastrointestinal transit"[Mesh] OR "gastrointestinal transit"[tw] OR "slow transit"[tw]) AND (randomized controlled trial[Publication Type] OR randomized[Title/Abstract] OR placebo[Title/Abstract])))

SC

EMBASE

M AN U

'linaclotide' OR 'linaclotide'/exp OR linaclotide OR 'linzess' OR 'linzess'/exp OR linzess OR 'lubiprostone' OR 'lubiprostone'/exp OR lubiprostone OR 'amitiza'/exp OR 'amitiza' OR 'naloxegol' OR 'naloxegol'/exp OR naloxegol OR 'movantik' OR 'movantik'/exp OR movantik OR 'nktr 118'/exp OR 'nktr 118' OR 'mu opiate receptor antagonist'/exp OR 'mu opiate receptor antagonist' OR 'alvimopan'/exp OR 'alvimopan' AND ('constipation' OR 'constipation'/exp OR constipation OR 'gastrointestinal transit'/exp OR 'gastrointestinal transit' OR 'slow transit' OR 'constipat*') AND ('opiate'/exp/dd_ae OR 'opiate' OR 'opiate induced' OR 'opioid induced' OR oic) AND random OR placebo OR 'double blind' Cochrane Library

Cochrane Trials

TE D

TX (( "Constipation" OR "Gastrointestinal transit" OR "gastrointestinal transit" OR "slow transit" OR bowel ) AND ( "linaclotide" OR “MD-1100 acetate” OR Linzess OR “Linaclotide Acetate” OR "Lubiprostone" OR Amitiza OR "naloxegol" OR Movantik OR NKTR-118 OR "micro opioid receptor antagonist" ))

AC C

EP

TX (( "Constipation" OR "Gastrointestinal transit" OR "gastrointestinal transit" OR "slow transit" OR bowel ) AND ( "linaclotide" OR “MD-1100 acetate” OR Linzess OR “Linaclotide Acetate” OR "Lubiprostone" OR Amitiza OR "naloxegol" OR Movantik OR NKTR-118 OR "micro opioid receptor antagonist"))

Web of Science

( linaclotide OR linzess OR lubiprostone OR amitiza OR naloxegol OR movantik OR nktr 118 OR alvimopan OR mu opiate receptor antagonist OR micro opioid receptor antagonist ) AND ( constipation OR constipat* OR gastrointestinal transit OR slow transit ) AND (opiate OR opioid induced OR opioid OR OIC)

ACCEPTED MANUSCRIPT

Blinding of Participants or Personnel

Blinding of Outcome Assessment

Incomplete Outcome Data

Selective Reporting / Reporting Bias

Other Bias

Yes No Unclear

Yes Yes Unclear

Yes Yes Unclear

Yes Yes Unclear

No No Unclear

No No No

None None None

Yes No No No Yes

Yes No No Yes Yes

Yes No No Yes Yes

Yes Yes No Yes Yes

No No No No No

No No Yes No No

None None None None None

Yes No No No

Yes Unclear Yes Yes

Yes Unclear Unclear Unclear

Yes Yes Yes Yes

No No No No

No No No No

None None None None

Yes Yes

Yes Yes

Yes Yes

Yes Yes

No No

No No

None None

Yes Yes Yes Yes Yes Yes Yes

No Yes Yes Yes Yes Yes Yes

No Yes Yes Yes Yes Yes Yes

Yes Yes Yes Yes Yes Yes Yes

No No No No No No No

No No No No No No No

None None None None None None None

No

No

No

Yes

No

No

None

TE D

M AN U

SC

Allocation Concealment

EP

LUBIPROSTONE 38 • Cryer 2014 39 • Jamal 2015 37 • Spierings NALOXONE 19 Meissner 2009 21 Simpson 2008 18 Lowenstein 2009 17 Liu 2002 20 Sanders ALMIVOPAN 24 Paulson 2005 25 Webster 2008 22 Irving 2011 23 Jansen 2011 NALOXOGEL • Chey 201426 Webster 201327 METHYLNALTREXONE 28 Anissian 2012 30 Michna 2011 32 Thomas 2008 29 • Bull 2015* 16 Yuan 2000 41 • Rauck 2016 31 Slatkin 2009 PRUCALOPRIDE 40 Sloots 2010

Random Sequence Generation

AC C

Study

RI PT

Supplementary 2. Table of Individual Bias Risk for Randomized Controlled Trials of Pharmacological Therapies Versus Placebo in OIC.

ACCEPTED MANUSCRIPT

Unclear Unclear Yes

Unclear Unclear Yes

Unclear Unclear Yes

Unclear

Unclear

Unclear

Unclear

No No No No

EP

TE D

M AN U

SC

Indicates studies that were not included in Ford et al’s 2013 meta-analysis

AC C

•

Unclear Unclear Yes

RI PT

NALDEMEDINE 33 • Murata 2016 34 • Hale 2016 35 • Webster 2017 AXELOPRAN (TD-1211) 36 • Vickery *

No No No

None None None

No

None

ACCEPTED MANUSCRIPT

PRISMA 2009 Checklist Section/topic

Reported on page #

# Checklist item

Title

RI PT

TITLE 1

Identify the report as a systematic review, meta-analysis, or both.

Title Page

2

Provide a structured summary including, as applicable: background; objectives; data sources; study eligibility criteria, participants, and interventions; study appraisal and synthesis methods; results; limitations; conclusions and implications of key findings; systematic review registration number.

Abstract

Rationale

3

Describe the rationale for the review in the context of what is already known.

Introduction

Objectives

4

Provide an explicit statement of questions being addressed with reference to participants, interventions, comparisons, outcomes, and study design (PICOS).

Introduction

Protocol and registration

5

Indicate if a review protocol exists, if and where it can be accessed (e.g., Web address), and, if available, provide registration information including registration number.

Search Strategy and Inclusion Criteria

Eligibility criteria

6

Specify study characteristics (e.g., PICOS, length of follow-up) and report characteristics (e.g., years considered, language, publication status) used as criteria for eligibility, giving rationale.

Information sources

7

Describe all information sources (e.g., databases with dates of coverage, contact with study authors to identify additional studies) in the search and date last searched.

Search Strategy and Inclusion Criteria

Search

8

Present full electronic search strategy for at least one database, including any limits used, such that it could be repeated.

Appendix S1 (need to include)

Study selection

9

State the process for selecting studies (i.e., screening, eligibility, included in systematic review, and, if applicable, included in the meta-analysis).

Search Strategy

Structured summary

SC

ABSTRACT

M AN U

INTRODUCTION

AC C

EP

TE D

METHODS

Search Strategy and Inclusion Criteria

ACCEPTED MANUSCRIPT

PRISMA 2009 Checklist and Inclusion Criteria 10

Describe method of data extraction from reports (e.g., piloted forms, independently, in duplicate) and any processes for obtaining and confirming data from investigators.

Data items

11

List and define all variables for which data were sought (e.g., PICOS, funding sources) and any assumptions and simplifications made.

Risk of bias in individual studies

12

Describe methods used for assessing risk of bias of individual studies (including specification of whether this was done at the study or outcome level), and how this information is to be used in any data synthesis.

Data Extraction and Risk of Bias

Summary measures

13

State the principal summary measures (e.g., risk ratio, difference in means).

Data Synthesis and Statistical Analysis

Synthesis of results

14

Describe the methods of handling data and combining results of studies, if done, including measures of consistency (e.g., I2) for each meta-analysis.

AC C

EP

TE D

M AN U

SC

RI PT

Data collection process

Section/topic

Search Strategy and Inclusion Criteria Data Extraction and Risk of Bias

Data Synthesis and Statistical Analysis

Page 1 of 2

# Checklist item

Reported on page #

Risk of bias across studies

15

Specify any assessment of risk of bias that may affect the cumulative evidence (e.g., publication bias, selective reporting within studies).

Data Synthesis and Risk of Bias

Additional analyses

16

Describe methods of additional analyses (e.g., sensitivity or subgroup analyses, meta-regression), if done, indicating which were pre-specified.

Data Synthesis and Statistical

ACCEPTED MANUSCRIPT

PRISMA 2009 Checklist Analysis

RI PT

RESULTS 17

Give numbers of studies screened, assessed for eligibility, and included in the review, with reasons for exclusions at each stage, ideally with a flow diagram.

Overview of Included Studies (Figure 1)

Study characteristics

18

For each study, present characteristics for which data were extracted (e.g., study size, PICOS, follow-up period) and provide the citations.

Throughout Results Section (Table 1)

Risk of bias within studies

19

Present data on risk of bias of each study and, if available, any outcome level assessment (see item 12).

Risk of Bias (supplementary 1, Table 2)

Results of individual studies

20

For all outcomes considered (benefits or harms), present, for each study: (a) simple summary data for each intervention group (b) effect estimates and confidence intervals, ideally with a forest plot.

Throughout Results Section

Synthesis of results

21

Present results of each meta-analysis done, including confidence intervals and measures of consistency.

Throughout Results Section

Risk of bias across studies

22

Present results of any assessment of risk of bias across studies (see Item 15).

Publication Bias

Additional analysis

23

Give results of additional analyses, if done (e.g., sensitivity or subgroup analyses, meta-regression [see Item 16]).

Throughout Results Section

Summary of evidence

24

Summarize the main findings including the strength of evidence for each main outcome; consider their relevance to key groups (e.g., healthcare providers, users, and policy makers).

Throughout Discussion Section

Limitations

25

Discuss limitations at study and outcome level (e.g., risk of bias), and at review-level (e.g., incomplete retrieval of identified research, reporting bias).

Throughout Discussion Section

M AN U

TE D

EP

AC C

DISCUSSION

SC

Study selection

ACCEPTED MANUSCRIPT

PRISMA 2009 Checklist 26

Provide a general interpretation of the results in the context of other evidence, and implications for future research.

27

Describe sources of funding for the systematic review and other support (e.g., supply of data); role of funders for the systematic review.

RI PT

Conclusions

FUNDING Funding

Throughout Discussion Section Funding Section

SC

From: Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med 6(6): e1000097. doi:10.1371/journal.pmed1000097

For more information, visit: www.prisma-statement.org.

AC C

EP

TE D

M AN U

Page 2 of 2