754 Journal of Pain and Symptom Management Vol. 42 No. 5 November 2011

Transcription

1 754 Journal of Pain and Symptom Management Vol. 42 No. 5 November 2011 Brief Report Characterization of Abdominal Pain During Methylnaltrexone Treatment of Opioid-Induced Constipation in Advanced Illness: A Post Hoc Analysis of Two Clinical Trials Neal E. Slatkin, MD, Richard Lynn, MD, Chinyu Su, MD, Wenjin Wang, PhD, and Robert J. Israel, MD California Cancer Specialists Medical Group (N.E.S.), Monrovia, California; Wyeth Pharmaceuticals, Pfizer Inc. (R.L., C.S., W.W.), Collegeville, Pennsylvania; and Progenics Pharmaceuticals (R.J.I.), Tarrytown, New York, USA Abstract Context. Methylnaltrexone is a selective peripherally acting mu-opioid receptor antagonist that decreases the constipating effects of opioids without affecting centrally mediated analgesia. In two double-blind, placebo-controlled, Phase III studies of methylnaltrexone for opioid-induced constipation in patients with advanced illness, abdominal pain was the most common adverse event (AE) reported. Objectives. This analysis sought to further characterize the Medical Dictionary for Regulatory Activities-defined abdominal pain AEs experienced in these studies. Methods. A post hoc analysis of verbatim descriptions was used to further assess AEs characterized as abdominal pain in both trials. Descriptive summary statistics were used to assess severity of abdominal pain, effect of abdominal pain on global pain scores, and other characteristics. Logistic regression analysis was used to determine the association of baseline characteristics with abdominal pain. Results. Most verbatim descriptions of abdominal pain referred to abdominal cramps or cramping. Abdominal pain AEs were mostly mild to moderate in severity and did not affect patients global evaluation of pain. The incidence of abdominal pain AEs in methylnaltrexone-treated patients was greatest after the first dose and decreased with subsequent doses. No association between abdominal pain AEs and most baseline patient characteristics was noted. Conclusion. Abdominal pain AEs in methylnaltrexone-treated patients in clinical trials are usually described as cramps or cramping, are mostly mild to moderate in severity, and decrease in incidence with subsequent Address correspondence to: Neal E. Slatkin, MD, Palliative Care Center, Silicon Valley, 4850 Union Avenue, San Jose, CA 95124, USA. pallcaredr@ yahoo.com Ó 2011 U.S. Cancer Pain Relief Committee Published by Elsevier Inc. All rights reserved. Accepted for publication: February 6, /$ - see front matter doi: /j.jpainsymman

2 Vol. 42 No. 5 November 2011 Abdominal Pain During Methylnaltrexone Treatment 755 dosing. J Pain Symptom Manage 2011;42:754e760. Ó 2011 U.S. Cancer Pain Relief Committee. Published by Elsevier Inc. All rights reserved. Key Words Methylnaltrexone, Relistor Ò, constipation, abdominal pain, opioids, mu-opioid receptor antagonist, opioid-induced constipation Introduction Opioids are frequently used to treat pain syndromes related to serious medical illness, but their use is associated with various adverse events (AEs), the most common of which is constipation. 1 The prevalence of opioid-induced constipation (OIC) in patients with advanced illness is estimated to be as high as 90%. 2,3 Severe OIC may result in opioid dose reduction or limit the further upward titration of opioids if needed for pain control and thus compromise analgesia. 1,4,5 Efforts to reverse this side effect have often involved the use of laxatives and stool softeners; however, these agents can be unpredictable in terms of the induction and timing of laxation and are often ineffective. 1,3,6 Methylnaltrexone (Relistor Ò ; Pfizer Inc., Philadelphia, PA, and Progenics Pharmaceuticals, Inc., Tarrytown, NY) is a selective peripherally acting mu-opioid receptor antagonist approved for the treatment of OIC in patients with advanced illness who are receiving palliative care and whose response to laxative therapy has been insufficient. 7,8 It is currently the only medication approved by the U.S. Food and Drug Administration and other regulatory agencies specifically for use in palliative care patients and the only drug approved for the treatment of OIC in any patient population. Methylnaltrexone decreases the constipating effects of opioids without affecting centrally mediated analgesia. 9 The effectiveness of subcutaneous methylnaltrexone in inducing laxation in advanced illness patients with OIC has been demonstrated in two double-blind, placebo-controlled studies. 10,11 In both studies, a significantly greater proportion of OIC patients receiving methylnaltrexone experienced laxation within four hours of the first dose (primary outcome measure) compared with those who received placebo. In both studies, abdominal pain was the most frequently recorded AE, based on the Medical Dictionary for Regulatory Activities (MedDRA) categorization of preferred terms. To further characterize these abdominal pain AEs, a post hoc analysis was performed on data from the double-blind phase of each study. To assess changes in abdominal pain over time, open-label phase study data also were included for this evaluation. This analysis aimed to provide clinically useful information for clinicians who prescribe methylnaltrexone for their advanced illness patients with OIC. Methods The design of each study is illustrated in Fig. 1. Both studies included a double-blind, placebo-controlled phase followed by an optional open-label extension phase. 10,11 Detailed methods for both studies have been previously reported. 10,11 This post hoc analysis identified incidents of AEs coded as abdominal pain as categorized by the MedDRA version 6.0 preferred terms. In patients with these abdominal pain (preferred term) AEs, their individual case report forms were reviewed for verbatim descriptions of the AEs. The incidence of abdominal pain across all double-blind and open-label phases was summarized. Abdominal pain severity was rated as mild, moderate, or severe and was captured in the patients study diaries at four hours after dosing. Patients receiving methylnaltrexone were under direct observation for at least four hours after at least the first dose of treatment; AEs were elicited and graded for severity, if present, including that of abdominal pain. The summaries for the double-blind phase included all patients who received at least one dose of study drug (methylnaltrexone or placebo) in the double-blind phase. The summaries across all phases included all patients who received at least one dose of methylnaltrexone (doubleblind or open-label) in the two studies. Each

3 756 Slatkin et al. Vol. 42 No. 5 November 2011 Study 1 Double-Blind Day 1 R (x 1 dose) a n Methylnaltrexone SC d 0.15 mg/kg Eligible o Methylnaltrexone SC Patients m 0.30 mg/kg i z Placebo e OL Study 2 QOD Dosing Days Eligible Patients R a n d o m i z e Methylnaltrexone SC (0.15 mg/kg) Placebo Methylnaltrexone SC (0.15 mg/kg) or Methylnaltrexone SC (0.30 mg/kg)* Placebo Placebo* OL Fig. 1. Study designs of two double-blind, placebo-controlled, Phase III trials in patients with advanced illness and opioid-induced constipation. *Dose escalation was allowed at the discretion of the primary investigator in patients with fewer than three rescue-free bowel movements by Day 8. OL ¼ open label; QOD ¼ every other day; SC ¼ subcutaneous. patient was counted once for the incidence of abdominal pain by the MedDRA preferred term. For the incidence of these events by verbatim terms, each unique AE description (verbatim term) that subsequently coded to the preferred term of abdominal pain was counted once. Descriptive summary statistics were used for the post hoc analysis. Logistic regression was used to examine the potential effects of baseline patient characteristics on abdominal pain. The initial logistic model included age, gender, race, cancer status, elapsed time from last bowel movement to first study dose, baseline World Health Organization status, baseline opioid use, baseline constipation distress scores, baseline pain scores, and treatment group. The final model was obtained by using the stepwise regression method. Results Characterization of Abdominal Pain According to Verbatim Descriptions The patients enrolled in these studies had advanced illness, that is, the illness was considered incurable. Most patients had either an actively progressive solid tumor or hematologic malignancy (Table 1) leading to their advanced clinical state; pain-related manifestations of the underlying illness constituted the primary reason for the use of opioid analgesics. During the double-blind period in Studies 1 and 2, 47 of the 165 patients receiving methylnaltrexone (28.5%) compared with 12 of 123 patients receiving placebo (9.8%) reported abdominal pain (preferred term) as an AE. Of the 165 patients in the two studies who received methylnaltrexone, 40 (24.2%) described their experience of abdominal pain (preferred term) as either abdominal cramping or cramps when verbatim terms were assessed (Table 2). Other verbatim reports included abdominal pain in eight (4.8%) patients and abdominal pressure in one (0.6%), with two of these nine patients using one of these descriptors and that of abdominal cramping or cramps. Thus, 40 of 47 methylnaltrexonetreated patients (85.1%) with abdominal pain (preferred term) AEs described this experience as cramping or cramps. Abdominal Pain Severity Abdominal pain (preferred term) AEs were most often mild to moderate in intensity (Table 3). Severe abdominal pain was reported by three (1.8%) methylnaltrexone-treated patients and by three (2.4%) placebo patients. During the double-blind phase of the studies, two patients (1.2%) in the methylnaltrexonetreated group discontinued because of abdominal pain (neither of whom reported severe

4 Vol. 42 No. 5 November 2011 Abdominal Pain During Methylnaltrexone Treatment 757 Characteristic Table 1 Combined Baseline Patient Characteristics From Studies 1 and 2 a (n ¼ 52) (n ¼ 47) Study 1 Study 2 MNTX, 0.30 mg/kg (n ¼ 55) (n ¼ 71) (n ¼ 63) Primary diagnosis, n (%) Cancer 43 (83) 37 (79) 45 (82) 41 (58) 37 (59) Cardiovascular 2 (4) 4 (9) 2 (4) 7 (10) 8 (13) Other 7 (14) 6 (13) 8 (15) 23 (32) 18 (29) WHO performance status, n (%) (2) (4) 2 (4) 1 (2) 6 (8) 3 (5) 2 17 (33) 13 (28) 15 (27) 16 (23) 14 (22) 3 21 (40) 19 (40) 30 (55) 36 (51) 28 (44) 4 12 (23) 12 (26) 9 (16) 13 (18) 18 (29) Constipation distress, n (%) None 4 (8) 4 (9) 4 (7) 8 (11) 7 (11) A little bit 9 (18) 7 (15) 5 (9) 6 (8) 6 (10) Somewhat 10 (20) 9 (20) 13 (24) 11 (16) 9 (14) Quite a bit 18 (37) 14 (30) 21 (39) 18 (25) 16 (25) Very much 8 (16) 12 (26) 11 (20) 27 (38) 22 (35) Not reported 3 (6) 1 (2) 1 (2) 1 (1) 3 (5) ¼ placebo; MTNX ¼ methylnaltrexone; WHO = World Health Organization. a Patients receiving at least one dose of study drug in the double-blind studies. pain), whereas no patients in the placebo group discontinued for this reason. Relationship Between Abdominal Pain and Current and Worst Pain Scores During the trials, patient assessment of global pain was obtained on a Likert scale from 0 to 10, with higher scores indicating more severe pain. In those patients treated with methylnaltrexone, there was no difference in current pain perception (measured at four hours after the first study dose) regardless of whether abdominal pain was present with the first dose, with the average current pain score being 2.8 for both methylnaltrexone-treated patients with and without abdominal pain. In the placebo group, current pain scores with the first dose were slightly higher (score ¼ 3.3) in those who did not report abdominal pain with the first dose compared with placebo patients who did have abdominal pain (score ¼ 2.4). Worst pain scores over the past 24 hours (measured at 24 hours after the first study dose) were similar in those experiencing abdominal pain independent of the treatment received (5.6 in both methylnaltrexone-treated and placebo patients). In those who did not experience abdominal pain, the worst pain scores were 4.7 for the methylnaltrexone group and 4.8 for the placebo group. Thus, the results Table 2 Combined Abdominal Pain Adverse Events Reported in Studies 1 and 2 During the Double-Blind Phase a Adverse Event (n ¼ 110), n (%) MNTX, 0.3 mg/kg (n ¼ 55), n (%) Total (n ¼ 165), n (%) (n ¼ 123), n (%) Preferred term Abdominal pain 26 (23.6) 21 (38.2) 47 (28.5) 12 (9.8) Verbatim category b Abdominal pain 2 (1.8) 6 (10.9) 8 (4.8) 4 (3.3) Abdominal cramping 24 (21.8) 16 (29.1) 40 (24.2) 8 (6.5) Other c 1 (0.9) 0 (0) 1 (0.6) 0 (0) MTNX ¼ methylnaltrexone; ¼ placebo. a Each patient was counted once for incidence of abdominal pain in preferred term. For the adverse event (AE) verbatim description, the same patient could be counted under different verbatim categories if more than one abdominal pain AE (preferred term) was reported. b An AE was categorized as abdominal pain if the verbatim term related to the abdomen contained the word pain. An AE was categorized as abdominal cramping if the verbatim term related to the abdomen contained the word cramp(s) or cramping, and did not contain the word pain. All other AEs were categorized as other. c One patient had an AE defined by the verbatim term of abdominal pressure.

5 758 Slatkin et al. Vol. 42 No. 5 November 2011 Table 3 Severity of Abdominal Pain (Preferred Term) Adverse Events Occurring During the Double-Blind Phase Study 1 Study 2 Combined Severity, n (%) (n ¼ 47) MNTX, 0.30 mg/kg (n ¼ 55) (n ¼ 52) (n ¼ 63) (n ¼ 71) All MNTX (n ¼ 163) All (n ¼ 123) Mild 9 (19.1) 12 (21.8) 1 (1.9) 5 (7.9) 4 (5.6) 26 (16.0) 5 (4.1) Moderate 4 (8.5) 7 (12.7) 1 (1.9) 7 (11.1) 3 (4.2) 18 (11.0) 4 (3.3) Severe 1 (2.1) 2 (3.6) 1 (1.9) 0 2 (2.8) 3 (1.8) 3 (2.4) Discontinuations because of abdominal pain AEs, n (%) (3.2) 0 2 (1.2) 0 MTNX ¼ methylnaltrexone; ¼ placebo; AEs ¼ adverse events. indicate that neither the presence of abdominal pain nor the treatment administered influenced overall current and worst pain scores. 10,11 Relationship Between Abdominal Pain and Laxation Response The percentage of patients who experienced a rescue-free bowel movement within four hours after the first double-blind dose, segregated by treatment and the presence of an abdominal pain AE with the first dose, is shown in Table 4. Among patients reporting abdominal pain, 80% of those receiving methylnaltrexone vs. 33% of those on placebo had a rescue-free bowel movement within four hours after the first dose. In contrast, in patients without abdominal pain after the first dose, 47.2% of methylnaltrexone-treated patients and 13.7% of placebo patients reported rescue-free laxation within four hours. Association Between Abdominal Pain and Baseline Characteristics We further examined the potential effect of baseline patient characteristics on abdominal pain AEs across the double-blind phase using a logistic regression model. The baseline characteristics, age, gender, race, cancer status, elapsed time from last bowel movement to first study dose, baseline opioid use, constipation distress scores, or pain scores were not retained in the final model using a stepwise method with entry level 0.10 and stay level Only baseline WHO status (1 vs. 4; odds ratio [OR] 3.92; 95% confidence interval [CI] 1.05e14.64; P ¼ 0.042) and treatment (methylnaltrexone vs. placebo; OR 4.36; 95% CI 2.09e9.13; P < ) were retained in the final model and were potential predictors for the occurrence of abdominal pain across the double-blind phase (Table 5). However, because of the small number of patients with WHO Status 1 in these trials, and with no significant differences observed between WHO Status 2 vs. Status 4 (OR 0.44; P ¼ 0.104) and between WHO Status 3 vs. Status 4 (OR 1.11; P ¼ 0.782), whether the baseline WHO status is a true predictor of the occurrence of abdominal pain is unknown. Incidence of Abdominal Pain Over Time To ascertain the proportion of patients experiencing abdominal pain with repeated doses, data from all double-blind and openlabel phases of the two Phase III studies were examined for rates of abdominal pain AEs for each dose number. For the first dose of methylnaltrexone, the incidence of abdominal pain was 23%; for the second dose, the Table 4 Patients With Rescue-Free Bowel Movement Within Four Hours After the First Dose of Treatment During the Double-Blind Phase Rescue-Free Bowel Movement, n/n (%) Treatment Abdominal Pain Yes No Methylnaltrexone Yes 32/40 (80.0) 8/40 (20.0) No 59/125 (47.2) 66/125 (52.8) Placebo Yes 2/6 (33.3) 4/6 (66.7) No 16/117 (13.7) 101/117 (86.3)

6 Vol. 42 No. 5 November 2011 Abdominal Pain During Methylnaltrexone Treatment 759 Table 5 Relationship of Abdominal Pain to Patient Characteristics During the Double-Blind Phase (Final Model) a Factor OR Estimate (95% CI) P-value Baseline WHO status 1 b vs (1.05e14.64) vs (0.17e1.18) vs (0.52e2.39) Treatment Methylnaltrexone vs. placebo 4.36 (2.09e9.13) < a Final model analysis of abdominal pain across the double-blind phase. This final model was determined by using a stepwise method with a entry level = 0.10 and retain a level ¼ This analysis included double-blind phases from both Study 1 and Study 2. b One patient (from Study 1) with a baseline WHO status of 0 was combined with those whose WHO status was 1 for this analysis. incidence decreased to 13%, and by the fifth dose, the incidence was less than 10% (Fig. 2). Discussion This post hoc analysis of two double-blind, placebo-controlled trials of subcutaneous methylnaltrexone in advanced illness patients with OIC sought to characterize the most commonly reported AE associated with methylnaltrexone in these two studies, that being abdominal pain. While treatment regimens varied between the two studies and one included a greater percentage of patients with cancer owing to the patient mix at the respective research sites (w80% vs. w60%, respectively), the two are more similar than different. Both studies used virtually the same entry (inclusion and exclusion) criteria, enrolled patients who were receiving opioids for the treatment of pain associated with their advanced illness or its associated comorbidities, and were under the care of palliative care practitioners. The primary focus of this analysis was related to the first dose of methylnaltrexone received by patients. Hence, the differing research designs, with respect to the number and sequencing of doses administered, would not be expected to impact the primary endpoint. Results from this analysis indicate that abdominal pain (preferred term) AEs found in association with methylnaltrexone were most commonly described verbatim as abdominal cramps or cramping. These abdominal pain AE reports did not impact the patients global evaluation of pain. AEs were mostly mild to moderate in intensity, decreased in incidence after the first dose, and led to treatment discontinuation in only two (1.2%) patients treated with methylnaltrexone, neither of whom reported severe pain. 30 Patients With Abdominal Pain (Preferred Term) Adverse Events, % Methylnaltrexone Dose Number a Fig. 2. Incidence of abdominal pain AEs in methylnaltrexone-treated patients across double-blind and open-label phases of Studies 1 and 2. An abdominal pain AE for any given dose occurred after that dose and before the next dose, unless the given dose was the last dose received. In that case, the abdominal pain AE occurred between dose administration and 30 days thereafter. a Represents the dose number of methylnaltrexone a patient received. For example, the first open-label dose of methylnaltrexone for a patient who previously received double-blind placebo would be considered Dose 1, whereas the first open-label dose of methylnaltrexone for a patient who previously received seven double-blind methylnaltrexone doses would be considered Dose 8.

7 760 Slatkin et al. Vol. 42 No. 5 November 2011 Among patients who received methylnaltrexone and reported an AE categorized as abdominal pain, 80% were responders (i.e., had a rescue-free bowel movement within four hours) vs. only 20% of nonresponders reporting such pain. One possible explanation is that the occurrence of abdominal pain was at least in part related to the experience of having a bowel movement, as a higher incidence of abdominal pain also occurred in placebo patients who achieved laxation within four hours. The fact that a rescue-free bowel movement occurred four times more often in those having abdominal pain compared with those who did not strongly suggests that the process of having a laxation, and not visceral withdrawal related to methylnaltrexone, best explains its occurrence. To the extent that abdominal pain may have been related to methylnaltrexone treatment independent of the process of laxation, the occurrence of this AE declined with repeated dosing despite no significant decline in the response rate. 11 Conclusion While a prospective investigation characterizing abdominal pain associated with the use of methylnaltrexone would be preferable, this post hoc analysis suggests that abdominal pain AEs in this patient population are generally described verbatim as cramps or cramping, are mostly mild to moderate in severity, and decrease in incidence with subsequent dosing. Disclosures and Acknowledgments This research was sponsored by Progenics Pharmaceuticals. Subanalyses were performed by Wyeth Research, which was acquired by Pfizer Inc. in October Editorial support was provided by Peloton Advantage, LLC, and was funded by Pfizer Inc. Progenics Pharmaceuticals has a proprietary commercial interest in methylnaltrexone. Robert J. Israel is an employee of and stockholder in Progenics Pharmaceuticals. At the time of this study, Richard Lynn, Chinyu Su, and Wenjin Wang were employees of and stockholders in Wyeth. Neal E. Slatkin has served as a consultant to Wyeth, Pfizer Inc., and Salix. He was also a principal investigator on both of the clinical trials from which this subanalysis was conducted. No author received an honorarium or other form of financial support related to the development of this manuscript. The authors thank John H. Simmons, MD, of Peloton Advantage for assistance with manuscript preparation.they would also like to acknowledge Michelle Rhiner, NP, and Chantal Kerr, RN, who assisted with study recruitment and data collection. References 1. Bell TJ, Panchal SJ, Miaskowski C, et al. The prevalence, severity, and impact of opioid-induced bowel dysfunction: results of a US and European Patient Survey (PROBE 1). Pain Med 2009;10:35e National Comprehensive Cancer Network (NCCN). NCCN clinical practice guidelines in oncology: Palliative care, V Fort Washington, PA: NCCN, Emanuel EJ, Hauser J, Emanuel LL. Palliative and end-of-life care. In: Fauci AS, Brauwald E, Kasper DL, et al, eds. Harrison s principles of internal medicine, 17th ed. New York: McGraw-Hill, 2008: 66e McNicol E, Horowicz-Mehler N, Fisk RA, et al. Management of opioid side effects in cancerrelated and chronic noncancer pain: a systematic review. J Pain 2003;4:231e Cook SF, Lanza L, Zhou X, et al. Gastrointestinal side effects in chronic opioid users: results from a population-based survey. Aliment Pharmacol Ther 2008;27:1224e Thomas J. Cancer-related constipation. Curr Oncol Rep 2007;9:278e Relistor Ò [package insert]. Philadelphia, PA, and Tarrytown, NY: Wyeth Pharmaceuticals Inc. and Progenics Pharmaceuticals, Yuan CS. Methylnaltrexone mechanisms of action and effects on opioid bowel dysfunction and other opioid adverse effects. Ann Pharmacother 2007;41:984e Yuan CS, Doshan H, Charney MR, et al. Tolerability, gut effects, and pharmacokinetics of methylnaltrexone following repeated intravenous administration in humans. J Clin Pharmacol 2005;45:538e Slatkin N, Thomas J, Lipman AG, et al. Methylnaltrexone for treatment of opioid-induced constipation in advanced illness patients. J Support Oncol 2009;7:39e Thomas J, Karver S, Cooney GA, et al. Methylnaltrexone for opioid-induced constipation in advanced illness. N Engl J Med 2008;328:2332e2343.