CENTER FOR DRUG EVALUATION AND RESEARCH 21-610 21-611 MEDICAL REVIEW ?nd FDA CENTER FOR DRUG EVALUATION AND RESEARCH DIVISION OF ANESTHESIA, ANALGESIA, AND RHEUMATOLOGY PRODUCTS DEPUTY DIVISION DIRECTOR APPROVAL NIEMO DATE: SUBMISSION DATE: DRUG: TRADENAME: NDA: SPONSOR: DOSAGE FORM: DOSAGE INDICATIONS: June 22, 2006 December 22, 2005 Oxymorphone Hydrochloride Extended-Release Tablets Oxymorphone Hydrochloride Immediate-Release Tablets OPANA ER OPANA 21?610, 505 Response to Approvable Action 21-611 Endo Pharmaceuticals Oral 5, 10, 20, and 40 mg (Extended?Release) 5 and 10 mg (Immediate?Release) The management of moderate to severe pain requiring around-the- clock opioid therapy for an extended period of time (Extended- Release) The management of moderate to severe acute pain (Immediate- Release) RECOMMENDED ACTION: Approval of NDA 21-610 oxymorphone hydrochloride extended-release tablets and approval of NDA 21?611 oxymorphone hydrochloride immediate- release tablets. SUMMARY OF FINDINGS 1. There is evidence of ef?cacy of oxymorphone extended-release tablets dosed every 12 hours when titrated to ef?cacy and tolerability in chronic pain requiring around-the?clock opioids for an extended period of time. The safety ?nding of numerous reports of hypoxia requiring treatment with an opioid antagonist from a study of oxymorphone ER Deputy Director Memo 21-610 21 -61 in the postoperative period indicates that this formulation of oxymorphone is not appropriate for the postoperative setting. . There is evidence of ef?cacy for oxymorphone immediate-release tablets dosed as 10 mg and 20 mg every four to six hours in acute pain. A dose of 30 mg every six to eight hours in the postoperative setting in opioid naive patients did not provide any greater analgesia than the 20 mg close but did yield an unacceptably high rate of hypoxia requiring treatment with an opioid antagonist. This ?nding indicates that maximum initial dose in this setting should be 20 mg. . Elderly patients are most sUsceptible to hypoxia in the postoperative setting indicating the need for careful titration in this population. . Equianalgesic potency is unclear. An attempt to determine the relative analgesic potency of oxymorphone extended-release tablets compared to modi?ed-release oxycodone tablets and modi?ed-release morphine tablets was unsuccessful. Guidance for the conversion of opioid tolerant patients from oxycodone, morphine, fentanyl, and methadone to oxymorphone ER that will be included in the package insert has been informed by conversion ratios for each drug used during a clinical trial, but this cannot be considered adequate evidence of equianalgesic potency. Oxymorphone ER and oxymorphone IR are contraindicated in moderate and severe hepatic impairment due to increased bioavailability of 3.7-fold and 12.2-fold, respectively. . An in viva study of the effect of alcohol 20%, 4% and on the bioavailability of a single dose of 40 mg of OPANA ER in healthy, fasted volunteers found a highly variable effect on Cmax with concomitant administration of alcohol and OPANA ER. The change in Cmax ranged from a decrease of 50% to an increase of 270% across all conditions studied. As the in vitro study of oxymorphone ER with alcohol 20%, and 40%) did not demonstrate an increased rate of release in oxymorphone, the in viva effect may not be due to dose dumping but may instead be due to enhanced absorption of oxymorphone. Therefore alcohol should not be ingested in patients taking either oxymorphone ER or oxymorphone IR. . After oral dosing with a single dose of 40 mg in fasted subjects, the mean peak oxymorphone plasma level is 2.4 ng/mL and the median is 2 hours. The Applicant has recommended that oxymorphone ER and oxymorphone IR be taken on an empty stomach. A Risk Minimization Program has been prepared by the Applicant to address concerns associated with modified?release opioids. Deputy Director Memo 21-610 21-61 1 BACKGROUND Oxymorphone immediate-release tablets 2, 5, and 10 mg were ?rst approved by the FDA (NDA- 11-73 7) in 1959 and marketed under the trade name Numorphan. The Applicant reports that the oral tablets were removed from the market in 1979, for commercial reasons. Oxymorphone injectable, 1 mg/ml, for intramuscular and subcutaneous administration (NDA 11-707) and oxymorphone rectal suppository, 2 mg and 5 mg, (NDA 11?707 and NDA 11-738) were approved by the Agency in 1959. Both products are currently marketed in the US. Studies were performed under IND 56,919 (oxymorphone ER) and IND 58,602 (oxymorphone IR). This submission represents a complete response to an Approvable action dated October 15, 2003 for NDA 21-610 oxymorphone hydrochloride extended-release tablets (oxymorphone ERtablets and NDA 21?611 oxymorphone hydrochloride immediate-release (oxymorphone IR) 5 and 10 mg tablets. The original NDAs were submitted on December 19, 2002. An approvable action was issued because the two ef?cacy studies submitted to the initial application were not considered successful in providing support of ef?cacy (EN3202-015 and EN3202-025). One study of oxymorphone ER in the immediate postoperative period submitted in the original application (EN3202-012) did show evidence of analgesic ef?cacy but with an unacceptable rate of hypoxia. In this resubmission, the results of two new 12-week clinical trials (EN3202-031 and EN 3202- 032) have been submitted in support of ef?cacy for oxymorphone ER with support from an 18- day placebo?controlled ef?cacy study submitted in the initial application. The results of one new multiple-dose study in acute postoperative pain has been submitted in support of ef?cacy for oxymorphone IR. One multiple-dose study submitted with the initial application (EN3203-004) and one single?dose provided evidence of ef?cacy for oxymorphone IR. However, these studies failed to provide adequate support for the proposed dosing regimen of every six hours. Support for the safety side of the risk bene?t analysis is provided by the original safety database with patients who had been exposed to oxymorphone ER and oxymorphone IR plus additional data from the two new studies ef?cacy trials of oxymorphone ER with oxymorphone IR as rescue medication (EN3202-031, EN3202-032) and additional data from two open?label extension studies ongoing at the time of the safety update from the original submission 021 and EN3202-22) and two new open?label safety studies (EN3202-028 and EN3202-029). Additional safety information in the 188 includes data from two ef?cacy studies of the oxymorphone IR (EN3203-008 and EN3203-009). Two trade names have been proposed by the Applicant. During the ?rst review cycle, the Of?ce of Drug Safety had raised concerns about the name OPANA and confusion with tincture of opium, but given the differences in dosage form, this concern has been dismissed by the Of?ce of Drug Safety and the division. 3 Deputy Director Memo 7 21 -61 0 21-61 1 CMC Dr. ila Boal reviewed the CMC material submitted in support of this application and has recommended approval from the product quality standpoint. The primary means of controlling the drug release for oxymorphone ER is via the excipient a hydrophilic matrix and Xanthum gum in the formulation hydrates when exposed to water forming a polysaccharide gel network which enables additional water to diffuse into the granulation. Less soluble locust bean gum then begins to hydrate and form a cohesive gel matrix with the xanthum gum. Drug release the occurs primarily by diffusion . There were numerous CMC de?ciencies identi?ed in the original NDA submission. As the oxymorphone ER formulation contains a relatively small amount of drug substance, the potential for agglomeration or segregation within the bulk blended drug product required appropriate sampling for homogeneity testing . Particle size speci?cations based on i were established for the drug substance. A speci?cation for the particle size distribution of the particle size was set. The Applicant also provided r' a? I A 7 Acceptance criteria for degradation products were revised according to ICH Q3 recommendations. The stability data showed that the - 3 total degradation products remained within speci?cations beyond months supporting an expiration dating period of 36 months. The drug substance speci?cations have been revised according to ICH Q3A. The present as process impurities from the of the drug substance are controlled on an interim basis and will be tightened further based on action by the DMF vendor. Dr. Boal concludes that the Applicant has provided an adequate response to the de?ciencies identi?ed in the NDA action letter dated October, 15, 2003. PHARMACOLOGY AND TOXICOLOGY Dr. Mamata De performed the review of the nonclinical pharmacology and toxicology data. For this resubmission, carcinogenicity studies in the rat and mouse were performed and found to be negative for drug-related neoplasms. The mechanism of oxymorphone?induced positive ?ndings in the in viva micronucleus assay was studied by the applicant and adequate data was provided suggesting that the increase in the incidence of micronuclei formation is indirectly related to activation of opioid receptors. See Dr. De?s review for greater detail. The applicant elected to reduce the levels to below the ICH quali?cation threshold rather than qualify them at higher levels. The applicant provided in vitro genetic toxicology screens for f? . The ..were positive in the in vitro chromosome 4 Deputy Director Memo 21-610 21-61 1 aberration assay. The division has an agreement in place with the DMF holder for the reduction of the levels of these impurities over time. The was negative in both genetic toxicology studies and is considered quali?ed. CLINICAL PHARMACOLOGY AND BIOPHARMACEUTICS N0 de?ciencies were cited in the Clinical Pharmacology and Biopharmaceutics Review of the original NDA by Dr. David Lee. In the current submission, the applicant submitted the results of in vitro and in vivo alcohol oxymorphone ER interaction results, food effect study results, oxymorphone IR single- and multiple-dose information and results from a simulation of 4-hour and IR multiple dosing regimens based on a single?dose oxymorphone 6? hour pharmacokinetic information. The in vitro studies demonstrated that when the oxymorphone ER 40 mg formulation was mixed with 500 mL of 0. 1N HCI solutions containing ethanol 20%, and 40%) there was no increase in release of oxymorphone from the formulation. The release rates actually correlated inversely with the amount of ethanol. An in vivo study examined the effect of alcohol 20%, 4% and on the bioavailability of a single dose of 40 mg of oxymorphone-ER in healthy, fasted volunteers. The results showed that the oxymorphone mean AUC was 13% higher (not statistically signi?cant) after co- administration of 240 mL of 40% alcohol. The AUC was essentially unaffected in subjects following the co-administration of oxymorphone ER and ethanol (240 mL of 20% or 4% ethanol). There was a highly variable effect on Cmax with concomitant administration of alcohol and oxymorphone ER. The change in Cmax ranged from a decrease of 50% to an increase of 270% across all conditions studied. Following concomitant administration of 240 mL of 40% ethanol the Cmax increased on average by 70% and up to 270% in individual subjects. Following the concomitant administration of 240 mL of 20% ethanol, the Cmax increased on average by 31% and up to 260% in individual subjects. Following the concomitant administration of 240 mL of 4 ethanol, the Cmax increased 7% on average and by as much as 110% for individual subjects. After oral dosing with a single dose of 40 mg in fasted subjects, the mean peak oxymorphone plasma level is 2.4 ng/mL and the median Tmax is 2 hours. Following co-administration of oxymorphone ER and alcohol (240 mL of 40% ethanol) in fasted subjects, the mean peak oxymorphone level is 3.9 ng/mL and the median Tmax is 1.5 hours (range 0.75 6 hours). Alcohol interaction studies were not performed with oxymorphone IR. As the in vitro study of oxymorphone ER with alcohol failed to demonstrate an increased rate of release in oxymorphone, the in vivo effect may not be due to dose dumping and may be due to enhanced absorption of oxymorphone. Therefore alcohol should not be ingested in patients taking either oxymorphone ER or oxymorphone IR. The results of the food effect study submitted in this response is similar to the data submitted in the original applications. Two studies found that after the administration of oxymorphone ER, the Cmax was increased by approximately 50% in fed subjects compared to fasted subjects. The AUC was unchanged in one study and increased by approximately 18% in the other study in fed 5 Deputy Director Memo 21 -610 21-61 1 subjects. After oral dosing with a single dose of 40 mg, a peak oxymorphone plasma level of 2.8 ng/ml is achieved at 1.0 hour in fasted subjects and a peak of 4.25 ng/ml is achieved at 2.0 hours in fed subjects and that beyond the 12 hour time point, there is very little difference in the curves. The applicant recommends that oxymorphone ER be dosed at least one hour prior to or two hours after eating. For oxymorphone lR, food increased both oxymorphone-Cmax and AUC by 38%. Oxymorphone IR 5, 10, and 20 mg exhibited dose linearity after single and multiple dosing every six hours for four days. Simulations of four and eight?hour dosing regimen were created by the sponsor and found to be underestimates of the average steady?state concentrations by 14- 25%. The general pharmacokinetic parameters of oxymorphone ER are presented in the following table taken from the proposed package insert. Table 1: Pharmacokinetic parameters of oxymorphone ER after ?ve days of every 12 hour dosing. Regimen Dosage Cmax AUC y, (ng/mL) (ng'hr/mL) (hr) Single Dose 5 mg 027:0.13 4.54:2.04 11.30i10.81 10 m8 0.65i029 8.94i4.16 9.83i5.68 20 mg 1211-077 17.81i7.22 9.89i3.21 40 mg 2.59:1.65 37.90i16.20 935162.94 Multiple Dose 5 mg? 0.70i0.55 5.60:3.87 NA 10 m8 124:0.56 97713.52 NA 20 m8 2.54i1.35 19.28i8.32 NA 40 m8 4.47i1.91 36.98i13.53 NA CLINICAL STUDIES AND STATISTICAL ANALYSIS Dr. Christina Fang performed the clinical review and Dr. Dionne Price performed the statistical review. OXymorphone ER Overview of Original Application Oxymorphone ER In the initial application, the Applicant submitted the results of four placebo-controlled studies in support of ef?cacy, safety, and dosing recommendations for oxymorphone ER. The results of three open-label or active-controlled trials were submitted in support of safety and to establish relative potency with approved modi?ed?release opioids. The results from three open-label studies were submitted in support of safety. was a double?blind, parallel group, placebo-controlled, multiple?dose study of oxymorphone ER intended to evaluate the ef?cacy and safety of a 20 mg dose of oxymorphone ER, to characterize the analgesic onset and duration of oxymorphone ER effect in a single-dose paradigm, and to evaluate the opioid dose-sparing effects of oxymorphone ER after multiple dosing in patients with acute pain following orthopedic surgery. While there was evidence of 6 Deputy Director Memo 21 -610 21 ~61 1 ef?cacy, many patients failed to achieve meaningful pain relief during the ?rst 12 hours. Furthermore, there was a disproportionate number of subjects requiring treatment with naloxone raising concerns about the safety of oxymorphone ER in the postoperative setting. EN3202-015 was a 4?week, double-blind, parallel group, placebo- and active-controlled, multiple-dose study of oxymorphone ER intended to support the ef?cacy and safety of two doses of oxymorphone ER (20 and 40 mg) compared to placebo and OxyContin, in patients with osteoarthritis of the knee and/or hip. The primary ef?cacy endpoint, change from baseline in Arthritis PI by VAS, failed to provide evidence of ef?cacy for the two doses of oxymorphone ER studied when a conservative imputation method was used to impute missing data. EN3202-016 was a double-blind, placebo- and active-controlled, randomized withdrawal study of oxymorphone ER. After 7-10 days of double-blind titration with oxymorphone ER or OxyContin to reach a stable dose that provided satisfactory pain relief, patients were randomized to continue with current treatment or receive placebo for 18 days. Ef?cacy was demonstrated by the change from baseline to the 4-hour post-dose VAS PI on Day 18, which was statistically signi?cantly better for the oxymorphone ER group compared with placebo. EN3202-025 was a double-blind, parallel-group, placebo-controlled study of 10, 40, and 50 mg doses of oxymorphone ER The purpose of this 2-week, dose-ranging study was to identify the minimum effective dose and maximum tolerable dose in patients with moderate to severe pain due to osteoarthritis. The primary ef?cacy endpoint, change from baseline in Arthritis PI by VAS, failed to provide evidence of ef?cacy when an appropriate method of imputation was used, as for study 015. was an open-label, crossover study of oxymorphone ER, OxyContin, and MS Contin in patients with cancer pain to determine the relative potency of oxymorphone compared to oxycodone and morphine. The study design was unable to achieve the study?s objective due at least in part to the open-label design, allowance for dose adjustments and rescue medication, and limited dosing ?exibility of the oxymorphone ER (20 mg). was a randomized, double-blind, crossover study intended to demonstrate analgesic equivalence between oxymorphone ER and MS Contin and determine the equianalgesic dose ratio between these two products. A valid dose ratio could not be calculated as there was a statistically signi?cant sequence effect. was a randomized, double-blind, crossover study intended to demonstrate analgesic equivalence between oxymorphone ER and OxyContin and the equianalgesic dose ratio between these two products. An equianalgesic was not de?nitively established. Long-term safety data was obtained from the following three studies. was an open-label study enrolling patients who completed or discontinued early from studies and permitting continuation of treatment with study drug for up to two years. 7 Deputy Director Memo 21 -610 21 ?61 EN3202-021 was an open-label study enrolling patients who completed or discontinued early from studies EN3202- 016 and EN3202-019 permitting continuation of treatment with study from for up to one year. EN3202-022 was an open?label study enrolling patients who completed or discontinued early from study EN3202- 018 permitting continuation of treatment with study for up to one year. NEW EFFICACY STUDIES OXYMORPHONE ER - The current submission consists of new data from two completed adequate and well controlled ef?cacy studies, EN3202-031 (031), EN3202-032 (O32), safety data from two completed open? label safety studies, and EN3202-029, and safety data from two ongoing open-label studies, and EN3202-022. Study Study EN3202-031 was a multi-center, double?blind, placebo?controlled, randomized withdrawal design study with a 28-day open-label titration period followed by a 12-week double-blind treatment period. The study was intended to support ef?cacy and safety of oxymorphone ER in opioid?naive patients with chronic low back pain (LBP). It was also intended to de?ne a safe dose titration regimen and dose range. The study was to have enrolled male and non?pregnant female patients over 18 years of age with a history of moderate to severe chronic non-neuropathic LBP daily for at least several hours per day for a minimum of three months prior to screening. Patients were to have been on stable physical therapy, biofeedback therapy, acupuncture therapy, or herbal remedies) for back pain and were to have been opioid naive, de?ned as on no more than an equianalgesic equivalent of oxycodone of 5 mg/day over the 14 days prior to screening. Patients were to have an initial pain score of at least 50 mm on a lOO-mm VAS scale at baseline. Patients were to undergo open-label titration on oxymorphone ER beginning with 5 mg every 12 hours for two days and then titrating by increments of 5-10 mg every 12 hours every 3-7 days until stabilization, de?ned as a pain score of less than 40 mm on same dose for three of ?ve consecutive days. Patients who successfully reached a minimum dose of 10 mg every 12 hours were then to have been randomized into the double-blind treatment period, continuing on their stabilized dose or placebo. Rescue medication was not to have been permitted during the open- label titration. During the ?rst four days of the double-blind period, rescue was to have been permitted, oxymorphone IR 5 mg every 4 to 6 hours, but was to then have been limited to no more than two doses per day for the remainder of the 12-week double-blind period. The primary ef?cacy endpoint was to have been the change in average pain intensity from baseline to ?nal visit, by VAS over the prior 24 hours, Secondary endpoints were to have included the change from baseline to ?nal visit in patient?s global assessment of pain medication and physician?s global assessment of pain medication, an evaluation of compliance and study medication usage, and time to discontinuation due to lack of ef?cacy. Safety monitoring was to include adverse events, vital signs at most clinic visits, the Adjective Rating Scale for Withdrawal (ARS) and the Clinical Opiate Withdrawal Scale (COWS). Adverse events were to have been coded using the system. 8 Deputy Director Memo 21-610 21-61 1 Missing ef?cacy values were to have been imputed using screening PI (worst case) carried forward for dropouts due to an AB, baseline PI (best case) carried forward for dropouts due to opioid withdrawal in the placebo group, and last observation carried forward for dropouts due to all other reasons. The analysis population was to have been all patients who received at least one dose of study medication. The Applicant created a modi?ed intent-to?treat population based on the exclusion of 14 patients, one who failed to sign the HIPAA as required, and 13 patients who failed to reach the prespeci?ed minimum dose of oxymorphone ER 10 mg every 12 hours but were randomized into the second period of the study. Results A total of 326 patients entered open-label titration. One patient never received study drug. Of the remaining 325 patients, only 205 successfully completed the open?label titration period (including the 12 patients who did not reach the minimum oxycodone dose). The most common reason for discontinuation during the open-label titration was adverse events, accounting for 59 patients The next most common reason was failure to meet the titration-stabilization criteria for 23 patients consent was withdrawn by another eight patients, and ?investigator opinion accounted for another five patients. The reason for discontinuation was further investigated by Dr. Fang for those patients who discontinued due to withdrawn consent, Investigator Opinion, Other, and Sponsor Request. None of these patients? reason for withdrawal included either lack of ef?cacy or adverse events. Table 2. Patient Disposition During the Open-Label Titration Period: Number of Patients Patient Disposition Oxymorphone ER Entered Open-Label Titration Period 326 (100.0) All Treated (Open-Label Titration Period)a 325 (99.7) Not Treatedb 1 (0.3) Completed Open-Label Titration Period 205 (62.9) Discontinued in Open-Label Titration Period 120 (36.8) Adverse Event 59 (18.1) AE not due to opioid withdrawal 59 (18.1) Opioid withdrawal-AB 0 Patient did not meet Titration-Stabilization criteria 23 (7.1) Withdrew Consent -14 (4.3) Lost to Follow-up 8 (2.5) Investigator Opinion 5 (1.5) Protocol Violation 5 (1.5) >3 days of <80% compliance with study medication 1 (0.3) Other 4 (1.2) Lack of Efficacy 4 (i .2) Sponsor Request 2 (0.6) Randomized and Entered Double?Blind Treatment Period 205 (62.9) a All patients who received at least one dose of the Open-Label Titration medication. Patient 031-021 was not treated according to Drug Accountability data. Note: For Patient 036-014, the reason for discontinuation was coded to AE, instead of Protocoi Violation, to be consistent with the Adverse Event Case Report Form 9 Deputy Director Memo 21-610 21-61 1 Source: Applicant?s Table 3 Study Report en3202031.pdf During the double-blind period, there were further early discontinuations. Table 3 taken from Dr. Fang?s review shows that only 68% of patients on oxymorphone ER completed the double?blind period compared to 47% of the patients randomized to placebo. Given the limited rescue medication available, it is not surprising that the most common reason for dropout in the placebo patients was lack of ef?cacy It is surprising that lack of ef?cacy was also the most common reason for early discontinuation in the double-blind period for the oxyrnorphone ER group Adverse events was next most common for both treatment groups followed by withdrawn consent. Further evaluation of the CRFs of patients who withdrew consent by Dr. Fang did not ?nd these were due to adverse events. Table 3 (Dr. Fang?s Table 31-3) Patient disposition during the double-blind treatment period number of patients Patient Disposition Oxymorphone ER Placebo Overall Randomized and entered double-blind treatment period 105 (100.0) 100 (100.0) 205 (100.0) All treated patients dose of double-blind treatment) 105 (1000) 100 100.0) 205 (100.0) Completed Double-Blind Treatment Period 71 (67.6) 47 (47.0) 118 (57.6) Discontinued in Double-Blind Treatment Period 34 (32.4) 53 (53.0) 87 (42.4) Lack of Ef?cacy 12 (11.4) 35 (35.0) 47 (22.9) Adverse Event 9 (8.6) 8 (8.0) 17 (8.3) A8 not due to opioid withdrawal 8 (7.6) 6 (6.0) 14 (6.8) Opioid withdrawal-AB 1 (1.0) 2 (2.0) 3 (1.5) Withdrew Consent 7 (6.7) 4 (4.0) 11 (5.4) Protocol Violation 3 (2.9) 7 3 (3.0) 6 (2.9) Investigator Opinion 3 (2.9) 1 (1.0) 4 (2.0) Other 3 (2.9) 1 1.0) 4 (2.0) >3 days of <80% compliance with study medication 0 2 (2.0) 2 (1.0) Lost to Follow-up 1 (1.0) 1 (0.5) Sponsor Request 0 1 (1.0) 1 (0.5) Modified Intent-to-Treat 97 (92.4) 95 (95.0) 192 (93.7) From Applicant?s Table 4, Study Report en3202031.pdf The demographics and baseline characteristics were comparable for race, stabilized dose level and baseline pain ratings as noted in Table 10?2 in Dr. Fang?s review. The oxymorphone ER group had more patients over age 65 compared to the placebo group more female patients (56% vs. and were heavier on average (195 vs. 186 lbs). The heavier weight in the active group would, if anything, bias against ef?cacy for the active treatment. The larger number of older patients would potentially bias against safety for the active treatment. As noted by Dr. Fang, the protocol deviations did not appear to be serious enough to impact ef?cacy results from these patients. During titration, approximately 15% of patients received concomitant NSAIDs and 13% received an acetaminophen-containing product. During the double-blind period, 14% of patients received an acetaminophen-containing product. Seven percent were on antidepressants. Other products 10 Deputy Director Memo 21 -610 21-61 1 used by up to 13% of patients included lipid lowering drugs, proton pump inhibitors, and stool softeners. Overall, concomitant medications were fairly well balanced across treatment groups. Patients titrated to a wide range of oxymorphone doses. As shown in the table below, patients stabilized at total daily dose of 20 mg most commonly, followed mg. Two patients did titrate to 140 mg of oxymorphone daily, even though opioid nai've previously. Based on the overall distribution and mean total daily dose of oxymorphone, the patients randomized to oxymorphone ER and placebo were fairly evenly distributed. Table 4. Total Daily Dose of Oxymorphone at Which Patients Stabilized at Randomization, All Randomized Patien Double-blind Period ne ER Placebo Total Dai Dose 105 100 Mean SD 39.2 26.4 40.9 (25.3) Source Applicant?s Table, Appendix 16.2.2, Table 9, Study Report en3202031.pdf 8 2 3 3 1 2 Ef?cacy The analysis of the primary ef?cacy endpoint, change in average VAS pain intensity over the prior 24 hours from baseline to ?nal visit, showed a statistically signi?cantly smaller change for the oxymorphone ER-treated patients than placebo-treated patients using the MITT population. Based on the design of this study, an ideal outcome would have been no change for the oxymorphone ER group and worsening for the placebo group. Based on the analysis, a worsening of pain over the 12?week double?blind period would be re?ected as a positive score. As seen in the table below, both groups had positive scores but the change was smaller for the oxymorphone ER group than the placebo group. According to the Applicant, there was no substantial difference for the same analysis using the ITT population. The Applicant?s proposed method of imputing missing scores was appropriately conservative in that for patients who dropped out due to an adverse event, the worst case score was imputed. Therefore, if patients had an effect on pain intensity but could not tolerate study drug, a poor score was imputed. i 1 Deputy Director Memo 21 -61 0 21-61 1 Table 5. Mean Average Pain Intensity (VAS) and Mean Change from Baseline to Final Visit (Primary Analysis) Modi?ed Intent-to-Tre-at Population (Double-Blind Treatment Period) Oxymorphone ER Placebo Statistics? (N =95) Average Pain Intensity Baseline (Visit 5) Mean (STD) 18.5 (11.22) 19.3 (11.26) 'Final Visit Mean (STD) 29.9 (26.21) 46.2 (27.03) Change from Baseline to Final Visit Mean (STD) 11.4 (24.39) 26.9 (27.81) LSMean :t SE 10.6 2.50 27.7 d: 2.53 Treatment comparison vs. Placebo LSMean Difference -17.1 -- 95% CI (-24.21, ?10.04) -- P-value 0.0001 -- Modi?ed from Dr. Fang?s Table 31-4 The secondary outcome measure, mean average pain intensity change by visit, illustrates the study results in the following ?gure. It is notable that the pain scores worsened from the end of titration and the start of the double?blind period through about Day 4, then became fairly stable through the end of the study. Day 4 is when rescue medication was restricted. Figure 1. Mean Change in Baseline in Average Pain Intensity by Visit Figure 3. Mean Change from Baseline in Average Pain Intensity (V AS) by Visit Modi?ed Intent-to-Treat Population (Double-Blind Treatment PeriodMean Change from Baseline (VAS) H4 Day Data Source: amaendix 1.1. Average Pain Intensity VAS scores range from 0 mmqio pain to 100 mm=the worst pain imaginable. Note: Error bars represent standard errors -) Source: Applicant?s Figure 3, Study Report en3202031.pdf 12 Deputy Director Memo 21 ?6 I 0 21-61 1 Secondary endpoints, the change from baseline to ?nal visit in patient?s global assessment of pain medication and physician?s global assessment of pain medication, an evaluation of compliance and study medication usage, and time to discontinuation due to lack of ef?cacy are summarized in the following table taken from Dr. Fang?s review. The ?ndings are all supportive of oxymorphone ER as more effective than placebo. Table 6. Summary of Results for Secondary and Additional Efficacy Endpoints-MITT Oxymorphone Placebo value Study report ER reference Secondary endpoints Time to discontinuation due to lack of ef?cacy (see <0.0001 ?g 2, p70 ?gure 2 below) Number discontinued due to lack of ef?cacy 10/97 34/95 table 14.1, p907 Patient global at ?nal visit a <0'0001 table 13? p72 Proportion with good/very good/excellent 78/95 0322 A) 34/86 Physician global at ?nal visit 0 (0'0001 table 14? p73 Proportion with good/very good/excellent 80/96 (83'4 A) 32/87 Additional endpoints Average PI by visit (see ?gure 1 above) ?g 3, p76 Change in P1 with respect to stabilized dose level, 10.6 27.7 <0.0001 table 18.1, LSMean p913 Percent reduction in average PI (see ?gure 3 below) ?g 4, p81 Responders: 230% reduction in P1 79/97 47/91 <0.0001 ?g 4, p81 Time to discontinuation due to all reasons (see <0.0007 ?g 5, p82 ?gure 4 below) Proportion discontinued due to all reasons 32/97 52/95 table 15.1, p908 Daily rescue in ?rst four days of double-blind from 2.2 mg from 2.3 table 17, p85 treatment to 3.3 mg mg to 10.3 mg of days on rescue medication: Day 4 to ?nai visit 34.4 - 41.2% 55.1 65.3% table 33, p41 17 From Dr. Fang?s review, Table 10-8 The time to discontinuation due to lack of ef?cacy during the double?blind period shows a similar pattern as the average change in pain by visit. There is a ?urry of dropouts until approximately Day 10. Rescue medication became limited on Day 4. Appears This Way On Original 13 Deputy Director Memo 21 "610 21-61 Figure 2. Time to Discontinuation Due to Lack of Ef?cacy Figure 2. Time (Days) to Discontinuation Due to Lack of Ef?cacy During the Double-Blind Treatment Period?Modi?ed Intent-to-Treat Population (Double-Blind Treatment Period) 10': [rm-sea Oxymurphune ER Placebo 0.8-: 0.6-: 0.5-: quot. .. . 0.3% Proportion Discontinued Due Time to Dlsconn?nuation Due to LOE {Days} Data Source: .?xpeenrtita it? 1-2, Figure p?value fortrealment comparison between wrymorphone ER and placebo is 000i and based on log-rank test. Applicant?s Figure 2, Study Report en3202031.pdf Subgroup analyses were performed for the demographic and baseline characteristics. Younger patients randomized to oxymorphone ER had greater worsening of pain than patients over 65 years for placebo but there was little difference in the oxymorphone group. Patients with severe baseline pain did have a greater worsening over the course of the double-blind period, as did patients who stabilized on a lower dose. Neither ?nding is surprising. Table 10-7 in Dr. Fang?s review provides the details of these analyses. Summary Study 031, characterized by a titration to stabilized dose and randomization of responders to active treatment or placebo provides evidence of ef?cacy for oxymorphone ER in previously opioid nai've patients with chronic low back pain. The primary ef?cacy endpoint demonstrated a statistically significant difference in favor of active treatment with oxymorphone ER. The multiple secondary endpoints were supportive of this ?nding. Study EN3202-032 Study was to have been the same as Study EN3202-031 except that patients were to have been opioid tolerant de?ned as on a stable, ?xed dose of around-the?clock opioid for LBP for at least 2 weeks prior to screening with reasonable tolerance of the opioid and expected to need a total daily oxymorphone ER dose in the range of 20?220mg (equivalent to oral morphine: 60-660mg). See Dr. Fang?s review for the protocol review. 14 Deputy Director Memo 21-610 21 -6 1 There were several post hoc analyses performed by the Applicant. The following bullets are taken ?om Dr. Fang?s review. Mean change from baseline in average pain intensity (VAS) by visit using the imputation rules established for the primary analysis with an additional imputation rule that a previous post-baseline value was to be used to impute a missing ?intermediate? post-baseline value. a Time to discontinuation due to all reasons using the same method of analysis as for the endpoint of time to discontinuation due to lack of ef?cacy - Percentage of responders de?ned as at least 30% percent reduction in average pain intensity (VAS) from screening to ?nal visit, using a chi-square test of observed values (no data imputation) - The percent reduction at all levels (310%, 220%, 330% presented as a ?gure. Results A total of 251 patients entered open-label titration. One patient never received study drug. Of the remaining 250 patients, only 143 successfully completed the open-label titration period. The most common reason for discontinuation during the open-label titration was adverse events, accounting for 47 patients The next most common reason was failure to meet the titration-stabilization criteria for 17 patients consent was withdrawn by 15 patients, and 10 patients discontinued due to lack of ef?cacy. The reason for discontinuation was further investigated by Dr. Fang for those patients who discontinued due to withdrawn consent, Investigator Opinion, Other, and Sponsor Request. None of these patients? reason for withdrawal included either lack of ef?cacy or adverse events. Table-7. Patient Disposition During Open-Label Titration Period: Number of Patients Patient Disposition Oxymorphone ER Entered Open-Label Titration Period 251 (100.0) All Treated (Open-Label Titration Period)a 250 (99.6) Not Treatedb 1 (04) Completed Open-Label Titration Period 143 (57.0) Discontinued in Open-Label Titration Period 107 (42.6) Adverse Event 47 (18.7) AE not due to opioid withdrawal 47 (18.7) Opioid withdrawal-AB 0 Patient did not meet Titration-Stabilization criteria 17 Withdrew Consent 15 (6.0) Lack of Ef?cacy 10 (4.0) Lost to Follow?up 6 (2.4) Investigator Opinion 6 (2.4) Protocol Violation 4 (1.6) >3 days of <80% compliance with study medication 1 (0.4) Other 3 (1.2) Sponsor Request 2 (0.8) Randomized and Entered Double-Blind Treatment Period 143 (57.0) a All patients who received at least one dose of the Open-Label Titration medication. One patient was not treated according to Drug Accountability data. Source: Applicant?s Table 3 Study Report en3202032.pdf 15 Deputy Director Memo 2 -6l 0 21 ?61 i During the double-blind period, there were further early discontinuations. The following table shows that only 49% of patients on oxymorphone ER completed the double-blind period compared to 18% of the patients randomized to placebo. The most common reason for dropout in the placebo patients was lack of ef?cacy, 53%, in contrast to 11% of patients in the oxymorphone ER group. Adverse events were responsible for 11% of placebo patient discontinuations including five patients who experienced opioid withdrawal. The protocol called for a best case score to be imputed for any patients who discontinue early due to opioid withdrawal, and as these were only in the placebo group, there would be no inadvertent bias in favor of active treatment due to these withdrawals. Adverse events were also responsible for an additional 10% of oxymorphone ER?treated patients discontinuing study participation early. The failure of four patients to sign the HIPAA consent form resulted in the exclusion of these four patients from the ITT population and all ef?cacy analyses were conducted on the remaining 138 patients, although all patients were included in the safety analysis. Further evaluation of the CRFs of patients who withdrew consent by Dr. Fang did not ?nd these were due to adverse events. Table 8. Patient Disposition During the Double?Blind Treatment Period: Number of Patients Table 4. Patient Disposition During the Double-Blind Tre atment Period Number of Patients Patient Disposition Oxymorphone ER Placebo Overall Randomized and Entered Douhl e-Blind Treatment Period 70 (100.0) 73 (100.0) 143 (100.0) All Treated Patients (Double-Blind Treatment Period)? 70 (100.0) 72 (98.6) 142 (99.3) Not Treatedb 1 (1.4) 1 (0.7) Completed Double-Blind Treatment Period 49 (20. 0) 18 (24.7) - 62 (46.9) Discontinued in Double-Blind Treatment Peri ed? 21 (30. 0) 54 (7 4.0) 75 (52.4) Lack ofEf?cacy 8(1 1.4) 39 (53.4) 47 (32.9) Adverse Event 7(100) 8(110) 15(105) Opioid 0 5 (6.8) 5 (3.5) AEnot dueto opioid uimdiawal 2(100) 3(4.l) 10 (7.0) Investigator Opinion 2 (2. 9) 2 (2.7) 4 (2.8) Withdraw Consent (1.4) 2 (2.7) 3 (2.1) Protocol Violation 2 (2.9) 1(1.4) 3 (2.1) Used prohibited medication for more than 3 consecutive days 0 1 (1.4) 1 (0.7) Compliance with study medicationis less than 80% for more than 3 days 1 (1.4) 0 (0.7) Other 1 (1.4) 0 (0.7) LosttoFollow?up l(l.4) 1(l.4) 2 (1.4) Sponsor Request 0 1 (1.4) 1 (0.2) All Treated Patients (Double-Blind, Ef?cacy)d 69 (98. 6) 69 (94.5) 138 (96.5) Data Source: Appde 16. 2.2, able 2 ?All randomized patients who received at least one dose of the double~blind study medication bPatient 023-009 was randomized but not treated according to drug accotmta?oilinr data. ?7 Reasons for discontinuation are sorted in descending order of overall ?equmcy. dThe following patients were excluded from the All Treated Patients (Double-B lind, E?iracy) population due to not signing the HIPAA consent form 007-002 (Oxymorphone ER), 022-002 (Placebo), 022-004 (Placebo), 027-002 (Placebo). From Applicant?s Table 4, Study Report en3202032.pdf The demographics and baseline characteristics were comparable for age, stabilized dose level and baseline pain ratings as noted in Table 10?11 in Dr. Fang?s review. The oxymorphone ER group had more African American patients (15% vs. respectively), more female patients (57% vs. 33%, respectively) and were heavier on average (200 vs. 191 lbs, respectively). 16 Deputy Director Memo 21-610 21-611 The heavier weight in the active group would, if anything, bias against ef?cacy for the active treatment. In the Applicant?s review of blinded data, a major protocol violation was identi?ed in four patients who received randomized treatments without signing the HIPAA consent form. These patients were excluded from the ef?cacy analysis. During titration, approximately 43% of patients received concomitant opioid analgesics, and 21% received a benzodiazepine. Thirteen percent received NSAIDs including selective COX-2 inhibitors and 10% received an acetaminophen-containing product. During the double-blind period, 21% of patients received a benzodiazepine, 14% of patients received an acetaminophen containing product. Twenty-three percent were on antidepressants, although these were more common in patients randomized to oxymorphone ER. Beta blockers were similar across the oxymorphone ER and placebo groups, 16% and 15%, respectively, but ACE inhibitors were 14% and I the oxymorphone ER and placebo groups, respectively. Other products used by up to 16% of patients included lipid lowering drugs, proton pump inhibitors, and stool softeners. As patients were on relatively stable doses of concomitant medications, it is unlikely that any of these differences were able to impact the ef?cacy results. During titration, patients titrated to a wide range of oxymorphone doses. As shown in the table below, patients stabilized at total daily dose of 40 mg most commonly, followed mg. Fourteen patients stabilized on doses of 200 mg or higher of oxymorphone daily, including one patient titrated to 260 mg of oxymorphone per day. Based on the overall distribution and mean total daily dose of oxymorphone, the patients randomized to oxymorphone ER and placebo were fairly evenly distributed. Table 9. Stabilized Individual Dosing at Randomization of Double-Blind Treatment Oxymorphone ER (N 70) Placebo (N 72) Daily Dose (mg/day) 20 12 (17.1) 6 (8.3) 40 18 (25.7) 16 (22.2) 60 9 (12.9) 15 (20.8) 80 9 (12.9) 5 (6.9) 100 4 (5.7) 6 (8.3) 120 4 (5.7) 3 (4.2) 140 2 (2.9) 5 (6.9) 160 4(5.7) 8(11.1) 180 1(1.4) (1.4) 200 4 (5.7) 2 (2.8) 220 3 (4.3) 4 (5.6) 260 0 (1.4) Mean (SD) 80.9 (59.3) 93.3 (61.2) Source Applicant?s Table 9, Appendix 16.2.2, Study Report en3202032.pdf Ef?cacy The analysis of the primary ef?cacy endpoint, change in average VAS pain intensity over the prior 24 hours from baseline to ?nal visit, showed a statistically signi?cantly smaller change for the oxymorphone ER-treated patients than placebo-treated patients using the ITT population. Based on the design of this study, an ideal outcome would have been no change for the 17 Deputy Director Memo 21?610 21?611 oxymorphone ER group and worsening for the placebo group. Based on the analysis, a worsening of pain over the 12-week double-blind period would be re?ected as a positive score. As in Study 031, and as presented in the table below, both groups had positive scores but the change was smaller for the oxymorphone ER group than the placebo group. The Applicant?s proposed method of imputing missing scores was appropriately conservative in that for patients who dropped out due to an adverse event, a worst case score (screening) was imputed and for patients who dropped out due to opioid withdrawal, a best case score was imputed, baseline after titration. Therefore, if patients had an some benefit with pain intensity but could not tolerate study drug, a poor score was imputed. Similarly, a good score was imputed for those in the placebo group who dropped out due to opioid withdrawal, baseline after titration, again to avoid biasing in favor of study drug. Table 10. Mean Average PI (VAS) and Mean Change from Baseline to Final Visit Oxymorphone ER Placebo Statistics? m=69) Average Pain Intensity Baseline (Visit N363 Mean (STD) 23.9 (12.05) 22.2 (10.75) Final Visit Mean (STD) 31.3 (23.48) 54.5 (28.43) Change from Baseline to Final Visit Mean (STD) 7.9 (20.60) 32.4 (27.00) LSMean :t SE 8.7 2.95 31.6 2.93 Treatment comparison vs. Placebo LSMean Difference -23.0 -- P-value 0.0001 -- ?The Primary analysis used an ANCOVA model with treatment and center as effects, screening and baseline average pain intensity as covariates. The following imputation rules, for missing values, were used: Discontinued due to AE: Screening Discontinued due to Opioid withdrawal in placebo group: Baseline Discontinued for all other reasons: Last OCF;Patients who discontinued for all other reasons but without post-baseline pain score: Screening OCF. observation carried forward bOxymorphone ER patient 009-010 has a missing CRF/Visit Baseline value. Sourcc: Applicant?s Table 12, Study Report en3202032.pdf The mean average pain intensity change by visit is illustrated in the following ?gure. It is notable that the pain scores worsened from the end of titration through about Day 4, then became fairly stable through the end of the study. Day 4 is when rescue medication was restricted. Appears This Way On Original 18 Deputy Director Memo 21-610 21 -6 Figure 3. Mean Change from Baseline in Average Pain Intensity (VAS) by Visit All Treated Patients (Double-Blind, Ef?cacy) 13gu-ew Li" so: f, . 255 15: 10Day Mean Change from Baseline WAS) Pd 1 Data Source: ppmdix 2J3, fig-ewe 1 Note: Avaiage pain intensityr VAS scores range from 0 mm no pain to 100 mm the worst pain inngimble. Error bars represmt standard errors Source: Applicant?s Figure 3, Study Report en3202032.pdf Secondary endpoints, the change from baseline to ?nal visit in patient?s global assessment of pain medication and physician?s global assessment of pain medication, an evaluation of compliance and study medication usage, and time to discontinuation due to lack of ef?cacy are summarized in the following table taken from Dr. Fang?s review. The post hoc analyses, proportion of responders, proportion discontinued due to all reasons The ?ndings are all supportive of oxymorphone ER as more effective than placebo. Appears This Way On Original 19 Deputy Director Memo 2] ~610 21?6] Table 11. Summary of Results for Secondary and Additional Ef?cacy End aoints?MITT Oxymorphon I Placebo value Study report 6 ER reference Secondary endpoints Time to discontinuation due to lack of ef?cacy <0.0001 ?g 2, p71 (see ?gure 5 below) Number discontinued due to lack of 8/69, 37/69, table 14, ef?cacy p656 Patient global at ?nal visit 55/69 <0.0001 table 13, p73 Proportion with good/very good/excellent 22/67 . Physician global at ?nal visit 58/69 <0.0001 table 14, p74 Proportion with good/very good/excellent 18/65 Change in PQAS-20 baseline to ?nal visit 5.5 40.5 <0.0001 table 18, LSMean p667 Additional endpoints Average PI by visit (See ?gure 6 above) ?g 3, p76 Change in P1 with respect to stabilized dose 8.6 31.6 <0.0001 table 19, level, LSMean p668 Percent reduction in average PI (see ?gure 7 ?g 4, p81 below) Responders: 230% reduction in P1 55/69 23/66 <0.0001 Time to discontinuation due to all reaSOns (see <0.0001 ?g 5, p82 ?gure 8 below) Proportion discontinued due to all reasons 20/69 51/69 table 15, p657 Daily rescue in ?rst four days of double-blind from 5.6 mg from 11.0 table 17, p85 treatment to 6.5 mg mg to 15.6 mg of days on rescue medication: Day 4 to ?nal 61 .8-70.7% 62.2-66.7% table 34, visit p697 From Dr. Fang?s review, Table 10-8 The time to discontinuation due to lack of ef?cacy during the double-blind period shows a similar pattern to the average change in pain by visit. Again, there is a marked ?urry of dropouts between Days 4 and 8, but there are ongoing dropouts due to lack of ef?cacy through Day 22 for the oxymorphone ER patients and Day 36 for placebo patients. Rescue medication became limited on Day 4. Appears This Way On Original 20 Deputy Director Memo 21-610 21-6ll Figure 4. Time to Discontinuation Due to Lack of Ef?cacy Figure 2. Time (Days) to Discontinuation Due to Lack of Ef?cacy During the Double-Blind Treatment Period All Treated Patients (Double?Blind, Efficacy) 1'0? Oxymorphone ER Placebo I 0.9-: 0.3-: 0.5-: 0.5-: $3.3 - 0.4-: 0.3-: 0.2-: Proportion Discontinued Due to LOE 0.1 -f ?Flo-1Time to Discontinuation Due to LOE (Days) Data Source: appendix gm: Note: p-value for treatment comparison between oxymorphone ER and placebo is 0001 and has ed on log?rank test. Applicant?s Figure 2, Study Report en3202032.pdf Subgroup analyses were performed for the demographic and baseline characteristics. Younger patients randomized to oxymorphone ER had greater worsening of pain that patients over 65 years, but the converse is true for placebo. This is the reverse of what was found in Study 031. Patients with severe baseline pain did have a greater worsening over the course of the double-blind period, as did patients who stabilized on a lower dose. Neither ?nding is surprising. Table 10-16 in Dr. ang?s review provides the details of these analyses. Summary Study 032, characterized by a titration to stabilized dose, followed by randomization of responders to active treatment or placebo provides evidence of ef?cacy for oxymorphone ER in previously opioid-tolerant patients with chronic low back pain. The primary ef?cacy endpoint demonstrated a statistically signi?cant difference in favor of active treatment with oxymorphone ER. The multiple secondary endpoints were supportive of this ?nding. Overall Summary of Ef?cacy of Oxymorphone ER Studies 031 and 032 demonstrate evidence of ef?cacy of oxymorphone ER in managing chronic pain in patients who were opioid naive and opioid tolerant, respectively. The question arises of how worsening pain scores on active treatment can support a ?nding of ef?cacy. One explanation is the development of opioid tolerance, particularly in a formerly opioid naive population, This does not appear likely as most of the change occurred during the ?rst week and then pain scores were fairly constant over remaining 11 weeks of the double?blind period. Tolerance would more likely have appeared as a more gradual change over time. Another explanation is that the protocol imposed limitation on rescue medication contributed to 21 Deputy Director Memo 21 ?610 21?61 1 worsened pain. Chronic pain is known to ?uctuate and a modi?ed-release product with limited rescue may not be suf?cient to manage the ?uctuations. This hypothesis is supported by the time to discontinuation which shows a marked ?urry of dropouts between Days 4 and 8. Rescue medication became limited on Day 4. As this ?nding" likely re?ects constraints imposed by study participation and there is a consistently better response by patients in the active treatment group, this study can be considered support for a ?nding of ef?cacy. As with Study 031, in Study 032, the change in pain intensity lessened for both groups, more so for the placebo group. The pattern of change with respect to pain scores as well as time to discontinuation due to lack of ef?cacy similarly supports the likelihood that the primary reason for the worsening of pain scores early on was limitation of rescue. In this opioid-experienced population, it seems even less likely that this was due to tolerance. The time to discontinuation and the number of days during which pain intensity worsened both re?ect a change early in the double-blind period. Similarly, there is a consistently better response by patients in the active treatment group, and this study can also be considered support for a ?nding of ef?cacy. As an enrichment design in which only those who could be successfully titrated were randomized, one can argue that the results may not be generalizable to the wider chronic pain population. However, it is characteristic of chronic analgesic trials of opioids to have a large percentage of patients dropout due to adverse events and for ef?cacy to be demonstrated in only a percentage of patients enrolled. When the patients who dropout due to adverse events are among the randomized population, there is a large amount of missing data and therefore a need to impute data for the primary ef?cacy analysis. Imputation generally creates bias, and depending on the method chosen, can bias in favor or against the study drug. By enriching the randomized population for patients who can potentially be successfully treated with study drug, the problem of dropouts can be reduced, thereby reducing the amount of data that must be imputed. This allows the ef?cacy to be more clearly demonstrated. To permit an understanding of the generalizability of the results, the number of patients who fail to successfully titrate must be kept in consideration when reporting the study outcomes. OXYMORPHON IR Overview of Original Application Oxymorphone IR A Study EN3203-004 was a 48?hour, single and multi-dose, placebo- and active-controlled study in 300 patients with post-operative pain due to knee or hip total or partial arthroplasty involving osteotomy. The results of the primary ef?cacy endpoint, TOTPAR 0?8, along with secondary analyses of pain relief and change in pain intensity following the single dose period demonstrate that the oxymorphone IR 10, 20, and 30 mg doses were effective when compared with placebo. Effects of exclusion of patients based on the Sponsor?s de?nition of the evaluable population were explored in additional analyses which yielded the same results. An immediate-release formulation of oxycodone was not effective when compared with placebo. Additional analyses continued to support ?ndings of ef?cacy, but not consistently for the oxymorphone IR 10, 20, and 30 mg doses. The proportion of patients experiencing 50% pain relief and the median time to 50% pain relief were statistically signi?cantly greater for the oxymorphone 10 and 20 mg groups compared to placebo. Time to ?rst perceptible pain relief did not differ between the active treatment groups and placebo while time to meaningful pain relief was statistically 22 Deputy Director Memo 2] -6 0 2] ?6l 1 signi?cantly shorter for the three oxymorphone IR groups compared to placebo. The median time to re?medication was statistically signi?cantly longer for the oxymorphone IR 20 and 30 mg groups compared to placebo. Review of mean PR and PID scores at individual study time points failed to demonstrate any superiority of the oxymorphone IR 30 mg group over the oxymorphone IR 20 mg group while the data trended in favor of the oxymorphone IR 20 mg dose. The patient global evaluation of satisfaction with study medication was statistically signi?cantly better for the oxymorphone IR 10 and 20 mg groups compared to placebo, but not for the oxymorphone IR 30 mg group. The nature of the study design during the multi?dose period precluded drawing conclusions about ef?cacy during this portion of the study. During this time period, patients requiring remedication prior to three hours after the last dose or rescue medication were withdrawn from the study. The Sponsor attempted to determine the dosing interval of oxymorphone IR during this study period. The analysis of the dosing interval performed by the Sponsor suggests a dosing interval of 7 to 9 hours. However, this analysis fails to account for subjects being withdrawn from the study for requiring rescue medication within 3 hours of study medication dosing, and for subjects receiving rescue medication after 3 hours of study medication dosing. Further support that the dosing interval for oxymorphone IR is not 7 to 9 hours comes from the ?nding that more than half of the study patients on oxymorphone IR 10 mg withdrew from the study by Hour 4 and for patients on oxymorphone Hour 5. Study EN3203-005 was a single-dose, double-?blind, placebo-and active?control study of oxymorphone IR 10 mg, oxymorphone IR 20 mg, oxycodone 15 mg, and oxycodone 30 in 324 patients with postoperative pain due to orthopedic procedures involving osteotomy. The results support a ?nding of ef?cacy for oxymorphone IR 20 mg as well as the two oxycodone IR doses using the primary ef?cacy endpoint, TOTPAR 0-8, as well as nearly all of the secondary outcome measures. There was no ef?cacy demonstrated for the oxymorphonc IR 10 mg dose. Effects of an evaluable population excluding subjects requiring rescue medication within the ?rst hour were explored in alternate analyses of the primary outcome utilizing a more inclusive population. No differences in the outcomes were found. Patient global assessment of pain relief mirrored these ?ndings. For patients in the oxymorphone IR 20 mg group not requiring rescue medication in the ?rst hour, the median time to rescue was nearly 5 hours. NEW EFFICACY STUDIES Oxymorphone IR The current submission consists of new data from two completed adequate and well controlled ef?cacy studies,. Study Protocol Study EN3203-009 was a randomized, double-blind, placebo- and active?controlled, parallel, single- and multiple-dose (48 hours), study of oxymorphone IR in patients with moderate to severe pain following abdominal surgery. Enrollment was to be patients scheduled for surgery through an abdominal incision of at least 3 cm, anticipated hospitalization for at least 36 hours, need of oral opioid therapy for at least 48 hours, anticipated treatment with short-acting parenteral analgesia post-operatively. Patients 23 Deputy Director Memo 21-610 2 -6 were to be capable of undergomg a washout within 12 hours of the last dose of parenteral opioids (washout 245 minutes from IV analgesics and 24 hours from IM analgesics) and anticipated conversion to oral analgesics within 30 hours following surgery. Initial post-operative analgesia was to have been an opioid by IV PCA or non-PCA or IM, but not by epidural route. Following a washout from parenteral analgesia, patients with moderate to severe pain on a categorical scale and pain rated 250 mm on a 100-mm visual analog scale (VAS) were to be randomized to one of the four treatment groups to take study drug every four to six hours. Study treatment groups were oxymorphone IR 10 mg, oxymorphone IR 20 mg, oxycodone IR 15 mg and placebo. After the initial 6-hour period, dosing was to be every four to six hours. Patients able to complete the ?rst six-hour period and requiring rescue within four to six hours were to enter the multiple-dose period. Those requesting re-medication within the ?rst fdur hours were to be discontinued from the study. During the 48-hour multiple-dose period which was to follow the initial dose, patients were to complete a diary/electronic diary for dosing and pain assessments prior to each dose of the study medication. Those in need of re?medication sooner than every four hours between doses were to be discontinued from the study. The primary ef?cacy endpoint was to be the time to discontinuation due to all causes during the entire study. Secondary ef?cacy endpoints for the multiple?dose period were planned to include mean average pain intensity (PI) (the average pain during each dosing interval evaluated at the time prior to each additional dose) and mean current PI (which represented the end?of-dosing pain evaluated also at the time prior to each additional dose), and patient global and physician global evaluation of study medication (recorded at the end of 48 hours). Secondary ef?cacy endpoints for the single-dose period were to include sum of pain intensity differences (SPID), total pain relief (TOTPAR), hourly pain relief (PR) and hourly pain intensity difference (to be recorded at 15, 30, 45, and 60 minutes, 1.5 hours, 2 hours, and hourly thereafter through Hour 6 using both categorical and VAS scales), and time to the ?rst perceptible and meaningful pain relief (by double stopwatch) for the single?dose period. Safety and tolerability were planned to be evaluated by monitoring adverse events (ABS) throughout the study. Appears This WGV 0n Original 24 Deputy Director Memo 2 1-610 21 -6 1 The analysis population was to be the ITT population de?ned as all randomized patients who received at least one dose of double-blind study medication and completed one post-dose ef?cacy assessment. imputation of missing scores for the secondary analyses was to be based on the reason the data was missing. For patients who discontinued early or took the second dose of study medication prior to the 6-hour mark, baseline observation carried forward (BOCF) was to be used. For missing individual assessments, data was to be interpolated linearly. For PR, a score of zero was used in the BOCF method for discontinuation due to an AB. For patients who took a second dose, the last observation carried forward (LOCF) method was applied to PR and PI. For patients who took a second dose prior to completion of the initial six hours of assessments, the LOCF method was used to carry the last pain score collected in the Single-Dose Period to Hour 6. Post hoc changes to the analysis plan are described in Dr. Fang?s review. Results A total of 331 patients were enrolled and received at least one dose of study medication. One patient who received one dose of study drug was consented after surgery and so was excluded from all ef?cacy analyses. Two patients were excluded from the secondary ef?cacy analyses due to no post-dose data. Usually these patients would be included in the ITT population, but as they were included in the primary ef?cacy analysis, and there were only two patients, there is no need to reanalyze the secondary endpoints. Disposition is displayed in the following table. During the single-dose and multiple-dose periods, the placebo patients had the highest rate of discontinuation due to lack of ef?cacy. Discontinuations due to adverse events were highest for the oxymorphone IR 20 rug-group in the single-dose period, and for the oxymorphone 10 mg group during the multiple?dose period. Appears This Way 0n Original 25 Deputy Director Memo 21 -61 0 21?61 1 Table 12. Disposition Patient Disposition Oxymorphone Oxymorphone Oxycodone Placebo Total =82) Randomized 82 81 83 85 331 All Treated Patients 82 (100.00) 81 (100.00) 83 (100.00) 85 (100.00) 331 (100.00) Multiple-Dose Period (0-48 Hours) 5:336th treatment 31 (37.80) 32 (39.51) 34 (40.96) 15 17.65) 112 (33.84) Discontinued 51 (62.20) 49 60.49) 49 59.04) 70 (82.35) 219 66.16) Lack ofEf?cacy 34 41.46) 25 (30.86) 33 (39.76) 53 (62.35) 145 (43.81) Adverse Event 7 (8.54) 14 17.28) 11 13.25) 11 12.94) 43 12.99) Withdrew Consent 9 10.98) 7 8.64) 4 4.82) 5 5.88) 25 7.55) Protocol Violation 1 1.22) 1 1.23) 0 0.00) 1 1.18) 3 (0.91) Withdrew by Investigator 0.00) 1 1.23) 0 0.00) 0 0.00) 1 0.30) Other 0 0.00) 1 1.23) 1 1.20) 0 0.00) 2 0.60) Intent-to-Treat Patients 81 (98.78) 81 (100.00) 83 (100.00) 85 (100.00) 330 99.70) Single-Dose Period (0-6 Hours) $313,136th treatment 48 (58.54) 53 65.43) 49 (59.04) 45 52.94) 195 58.91) Discontinued 34 41.46) 28 (34.57) 34 (40.96) 40 (47.06) 136 41.09) Lack ofEf?cacy 29 35.37) 21 (25.93) 27 (32.53) 35 (41.18) 112 (33.84) Adverse Event 5 6.10) 4 (4.94) 4 (4.82) 4 (4.71) 17 5.14) Withdrew Consent 0 0.00) 1 1.23) 3 (3.61) 1 1.18) 5 1.51) Withdrew by Investigator 0 0.00) 1 1.23) 0 0.00) 0 0.00) 1 0.30) Other 0 0.00) 1 1.23) 0 0.00) 0 0.00) 1 (0.30) Intent-to-Treat Patients 80 97.56) 80 (98.77) 83 (100.00) 85 (100.00) 330 99.70) Source: Applicant?s Table 2, en3203009.pdf Baseline and demographic characteristics were well balanced across treatment groups. More than 98% of the patient population was female in this study. The mean age was 42 to 43 years across the treatment groups. Baseline pain was comparable by VAS score across treatment groups averaging 62 to 65 mm. Using the categorical scale, there were fewer severe ratings in the oxymorphone IR 20 mg group. See Dr. Fang?s review for further details. Concomitant medications were primarily stool softeners and laxatives, antibiotics, serotonin 5HT3 antagonists, and estrogens. There were differences in the frequency of use of these products across the treatment groups as demonstrated in the following table. It is unlikely that any of the imbalances noted would alter ef?cacy outcomes. 26 Appears This Way Deputy Director Memo On Original 21-610 21-611 Table 13. Concomitant Medications Table 5. Summary of Concomitant Medications Taken by at Least 5% of Patients All Treated Patients Oxymorphone IR. Oxymorphone IR Oxyeodcne Placebo ATC Class (94) Number of Patients with at Last One Concomitant 67 81.71) 68 83.95) 63 (75.90) 69 81.18) Medication So?eners,Emollients 13 (15.85) 14 (17.28) 11 (13.25) 12 (14.12) ContactLaxatives 9 (10.98) 16 (19.75) 10 (12.05) 10 (11.76) Cephalospcrins and Related Substances 15 (18.29) 8 9.88) 14 16.87) 6 7.06) Serotonin SHT3 Antagonists 10 (12.20) 14 (17.28) 9 (10.84) 8 9.41) An?noalkylEthers 11 (13.41) 12 14.81) 13 15.66) 4 4.71) Anilides 7.32) 10 12.35) 11 13.25) 3 3.53) Natural and Plain 5 6.10) 8 1; 9.88) 8 9.64) 7 8.24) Multivitamins, Other Combinations 9 10.98) 5 I: 6.17) 4 4.82) 7 8.24) OtherDrugs forFunmionaIBowelDisorders 8 9.76) 7 8.64) 5 6.82) 5 5.88) Selective SerotomnReuptake Inhibitors 6 7.32) 5 15.17) 8 9.64) 4 4.71) Thyroid Hormones 7 8.54) 3 3.70) '1 8.43) 7.06) H2-ReceptorAntagonists 4 4.88) 5 6.17) 4 4.82) 4 4.71) Source: Applicant?s Table 5, EN3203009.pdf Ef?cacy The primary ef?cacy endpoint was time to discontinuation due to all causes during the 48-hour treatment period. There were statistically signi?cant differences favoring all three active treatment groups compared to placebo. The median time to discontinuation was 17 hours and 55 minutes for the oxymorphone IR group, 20 hours and 15 minutes for the oxymorphone IR 20 mg group, 24 hours and 5 minutes for the oxycodone IR 15 mg group and four hours and 50 minutes for the placebo group. The Kaplan-Meier curves for this analysis clearly illustrate the differences between the active treatment groups and placebo. Appears This Way On Original 27 Deputy Director Memo 21-610 21-61 1 Figure 5. Time to Discontinuation Figure 2. Time (Hours) to DisContinuation Due to All Causes Intent-to-Treat Patients Multiple-Dose Period (0-48 Hours) ated Proportion of Patients Ox mo hone IR 10 or mitigate-20 m3 ?twee 91? 6 810121416182 222 262830 323435 38404244464 Time (H ours) to Discontinuation due to All Causes Source: Applicant?s Figure 2, EN3203009.pdf It is notable that while there were fewer dropouts in the active treatment groups, there was a substantial proportion of patients who discontinued early from all three active treatment groups. While patients undergoing surgery are frequently treated with oral opioid analgesics in the postoperative period, it is usually in the context of individualization of the dose, both amount and frequency of dosing. The lack of ?exibility in dosing in the context of a clinical trial of an opioid analgesic likely contributes substantially to early discontinuations and cannot be taken as evidence of lack of ef?cacy or tolerability. The presence of the oxycodone IR treatment group provides an opportunity to compare the oxymorphone IR doses to an approved product recognized as safe and effective. The behavior of patients in the oxycodone IR treatment group is very similar to the oxymorphone IR treatment groups. The secondary ef?cacy endpoints are described in detail in Dr. Fang?s review. Mean average PI over the 48-hour treatment period was statistically signi?cantly lower for the active treatment groups compared to placebo. The mean Pljust prior to the next dose was also statistically signi?cantly lower for the active treatment groups as compared to placebo. The patient global evaluation of study medication was statistically signi?cantly better for the two oxymorphone IR groups compared to placebo. The oxyeodone group trended better than placebo. The time speci?c analysis of PI difference and PR during the ?rst 6-hour dosing interval showed the longest periods of statistically signi?cant improvement compared to placebo for the oxymorphone IR 20 mg group. See Dr. Fang?s review for the details of these analyses. 28 Deputy Director Memo 221610 21-61 1 Multiple analyses of single-dose effects were performed and supported the ef?cacy of the active treatment groups although not all comparisons with placebo met statistical signi?cance. There were generally greater differences for the oxymorphone IR 20 mg group as compared to the oxymorphone 10 mg group. The median time to onset was 40 minutes for the oxymorphone IR 10 mg group, 34 minutes for the oxymorphone IR 20 mg group and 45 minutes for the oxycodone IR 15 mg group. The median time to remedication during the ?rst 6?hour dosing interval was four hours for all treatment groups. While there were no statistically signi?cant differences from placebo, this is a descriptive measure used to inform dosing interval and supports the currently proposed dosing interval of every four to six hours. Summary . Study 009 was a double-blind, placebo?controlled and active-controlled study of oxymorphone IR 10 mg and 20 mg every four to six hours in the postoperative setting. This study primarily enrolled women. There is evidence for ef?cacy of the two doses of oxymorphone studied compared to placebo, and overall similar results to the oxycodone IR 15 mg treatment group. There was evidence of a dose response across the two oxymorphone doses. Study 3203-008 Study 3203-008 was a randomized double-blind, eight?hour study of the analgesic ef?cacy and safety of oxymorphone IR 5 mg in patients following ambulatory arthroscopic knee surgery. Oxymorphone IR 5 mg was compared to placebo in patients once their postoperative pain reached mild to moderate on a categorical scale or between 30-70mm on a VAS. Patients could close as frequently as every hour for an eight-hour period. One hundred twenty two patients were enrolled. Mean pain intensity difference over the eight- hour treatment period did reveal a statistically signi?cant difference between treatment groups, but 93% of the oxymorphone IR 5 mg patients redosed at one hour, 78% redosed at 2 hours, and 65% redosed at 3 hours. In comparison, 81% of placebo patients redosed at one hour, 63% at 2 hours and 45% at 3 hours. The mean dose interval was 1.9 hours for the oxymorphone IR group and 1.6 hours for placebo. While there is evidence of analgesic ef?cacy for oxymorphone IR from this study, the results do not support 5 mg as an effective dose. Overall Summary of Ef?cacy of Oxymorphone IR Studies 004 and 009 provide evidence of ef?cacy for oxymorphone IR 10 and oxymorphone IR 120 in two postoperative pain populations. Study 004 explored patients undergoing knee or hip total or partial arthroplasty involving osteotomy. Study 009 studied patients undergoing non- laparoscopic abdominal surgery. The mean age of patients in Study 004 was approximately 62 years and included patients over 90 years of age. In Study 009 the mean age was approximately 42 years of age. Ef?cacy during the multiple-dose period of this Study 004 was dif?cult to fully interpret due to the availability of additional analgesic medication greater than three hours after dosing of study medication at the discretion of the Investigator. The lack of available rescue in 29 Deputy Director Memo 21-610 21-61 1 Study 009 provided results during the multiple-dose period that did not have this limitation and provide support for the ?nding of multiple?dose ef?cacy. In Study 004, there was no bene?t evident for the oxymorphone IR 30 mg dose over the 20 mg dose, but as will be discussed in the evaluation of safety, there were a clinically meaning?il increase in serious adverse events, particularly hypoxia for the 30 mg dose, indicating that it is not an appropriate initial dose of oxymorphone in the immediate postoperative period. The currently proposed dosing interval of every four to six hours is supported by the median time to remedication of four hours in Study 009 and the prior ?nding that more than half of the study patients on oxymorphone IR 10 mg withdrew from the study by Hour 4 and for patients on oxymorphone Hour 5. Study was a single-dose, double-blind, placebo-and active-control study of oxymorphone IR 10 mg, oxymorphone IR 20 mg, oxycodone 15 mg, and oxycodone 30 in 324 patients with postoperative pain due to orthopedic procedures involving osteotomy. The results support a ?nding of ef?cacy for oxymorphone IR 20 mg as well as the two oxycodone IR doses using the primary ef?cacy endpoint, TOTPAR 0-8, as well as nearly all of the secondary outcome measures. There was no ef?cacy demonstrated for the oxymorphone IR 10 mg dose. Effects of an evaluable population excluding subjects requiring rescue medication within the ?rst hour were explored in alternate analyses of the primary outcome utilizing a more inclusive population. No differences in the outcomes were found. The patient global assessment of pain relief mirrored these findings. For patients in the oxymorphone IR 20 mg group not requiring rescue medication in the ?rst hour, the median time to rescue was nearly 5 hours. INTEGRATED ANALYSIS OF SAFETY Extent of Exposure In the original application, there were 2542 unique patients in the oxymorphone clinical development program, 2108 of these patients in the Phase 2/3 program, and 1484 unique patients in oxymorphone ER Phase 2/3 clinical trials, 1089 of whom received oxymorphone ER and/or IR at some point. The remainder received oxycodone ER or IR, morphine ER, or placebo. There were 624 unique patients in oxymorphone IR Phase 2/3 clinical trials. Extent of Exposure- Oxymorphone ER In the current submission, a total of 2575 unique patients were enrolled in oxymorphone ER Phase 2/3 clinical trials, including 2011 who received oxymorphone ER. A total of 1828 subjects received oxymorphone ER as their last treatment in Phase 2/3 studies; 807 of these patients are from open?label Phase 2/3 studies. The following table describes the exposure to oxymorphone ER by dose and duration in placebo- and active-controlled clinical trials. Appears This Way 0? Original 30 Deputy Director Memo 21 ?6 0 21 ?61 1 Table 14 Extent of Exposure by Treatment Duration and Modal Daily Dose?Phase 2/3 ER Trials Oxymorphone ER dosage (mg/day) 510 >10-50 I >50-90 >90 I Total Duration Number of subjects on treatment 1 - 3 Days 65 (18.8) 238 (69) 19 (5.5) 23 (6.7) 345 (100) 4 - 10 Days 67 (10.9) 332 (54.2) 118 (19.3) 95 (15.5) 612 (100) 11 17 Days 14 (5.8) 195 (81.3) 20 (8.3) 11 (4.6) 240 (100) 18 - 24 Days 2 (3.3) 40' (65.6) 11 (18) 8 (13.1) 61 (100) 25 - 31 Days 2 (6.9) 16 (55.2) 7 (24.1) 4 (13.8) 29 (100) 3 Months 13 (3.8) 173 (50.3) 72 (20.9) 86 (25) 344 (100) 6 Months 16 (10.1) 54 (34) 29 (18.2) 60 (37.7) 159 (100) 9 Months 0 (0.0) 15 (24.2) 21 (33.9) 26 (41.9) 62 (100) 12 Months 0 (0.0) 45 (45) 25 (25) 30 (30) 100 (100) 12 Months 0 (0.0) 21 (51.2) 14 (34.1) 6 (14.6) 41 (100) 18 Months 0 (0.0) 7 (58.3) 4 (33.3) 1 (8.3) 12 (100) Total 179 (9) 1129 (56.6) 336 (16.9) 349 (17.5) 1993 (100) Patient Years 12.21 163.42 92.35 102.88 370.87 No dosing information is available for 18 patients and they are therefore excluded from this table. Total duration of exposure for all trials in which a patient participated. Patient years: the total number of days in a given modal dose group divided by 365.25. Note: The information in the table represented the exposure to oxymorphone over time based on the most frequent dose the patient took, not each patient?s total exposure. Based on Studies EN3202-012, 015, 016, 017, 018, 019, 025, 028, 029, 031, and 032. Source: Applicant Table 7, The following table presents the cumulative open-label exposure by study and duration on treatment. The overall exposure includes 433 patients with at least six months exposure and 267 patients with exposure of one year duration or longer. Table 15 Extent of Cumulative Exposure in Open-Label Phase 2/3 Oxymorphone ER Trials Months on treatment 3 Months I 6 Months 9 Months I 12 Months I 18 Months I 224 Months Study (duration) Number of subjects on treatment 020 (2-year) 197 (100) 104 (52.8) 94 (47.7) 86 (43.7) 42 (21.3) 26 (13.2) 021 (l-year) 239 (100) 160 (66.9) 126 (52.7) 111 (46.4) 3 (1.3) 0 (0.0) 022 (l?year) 24 (100) 9 (37.5) 7 (29.2) 7 (29.2) 7 (29.2) 1 (4.2) 028 (6?month) 124 (100) 63 (50.8) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 029 (1-year) 220 (100) 97 (44.1) 74 (33.6) 63 (28.6) 0 (0.0) 0 (0.0) Total 804 (100) 433 (53.9) 301 (37.4) 267 (33.2) 52 (6.5) 27 (3.4) N0 dosing information was available for three patients (two in Study 028, one in Study 029) and therefore, they were excluded from this table. Note: total exposure to oxymorphone was calculated on a per patient basis regardless of dose or changing doses for a patient within a study. Source: Applicant Table 6, 31 Deputy Director Memo 21-610 21-611 Extent of Exposure? Oxymorphone In total, 788 unique subjects received at least one dose of oxymorphone IR in clinical trials of oxymorphone IR, including 34 cancer patients who received oxymorphone IR during active? controlled trials and EN3202-019). A total of 554 patients received oxymorphone IR in Phase 2/3 IR trials. Table from Dr. Fang?s review provides a breakdown of patient numbers by study and treatment assignment. Table 16-. Summary of Exposure by Subset and Treatment Groups in All Oxymorphone IR Trials Treatment Oxymorphone IR Oxycodone IR Placebo Database 188 Update Overall ISS Update Overall [88 Update Overall All Trials 565 223 788 195 83 278 122 147 269 Phase Phase 2/3 368 223 591 195 83 278 123 147 270 Phase 2/3 IR post-op pain 334 223 557 195 EN3203-005 130 128 66 EN3202-008 60 0 62 EN3202-009 163 83 85 Phase 2/3 ER cancer pain 34 34 0 0 0 0 EN3202-018 18 EN3202-019 16 In Study 3203-004, 21 placebo patients were re-randomized to one of the three oxymorphone IR treatment groups (oxymorphone IR 10 mg: six patients; oxymorphone IR 20 mg: eight patients; oxymorphone IR 30 mg: seven patients); seven placebo patients were re?randomized to oxycodone IR. These 21 patients are presented in the table under both the active (oxymorphone or oxycodone) and the placebo treatment groups. Source: Dr. Fang?s Table 7- 10 (Table 50 on page 52 of the 188, Table 2 on page 42 of the report for Study 3203- 009, and page 68 of the report for Study 3203?008. The following table taken from Dr. ang?s review breaks down the exposure to oxymorphone IR by number of doses and duration of exposure. Table 17. Ex sure To 0 hone IR Number Of Doses And Duration Of sure. Dose Single >1 of dose/ patients Duration (hours) dose dose Mean Maximum Mean Maximum Stud 3203-008 5 doses 8 doses 7 hours 36 hours Study 3203-005 Stud 3203-004 3 doses 11 doses 17 hours 51 hours Stud 3203-009 5 doses 16 doses 21 hours 80 hours ubtbtal 70 Study 3203-005 3203?004 4 doses 13 doses 21 hours 57 hours 3203-009 5 doses 13 doses 23 hours 59 hours Subtotal 00 30 Stu 3203-004 3 doses 9 doses 16 hours 63 hours Total 7 Source: Page 58 of the report for Study 3203-005, Table 12 on page 50 of the report for Study 3203 ~008, Table 18 32 Deputy Director Memo 21-610 21-61 I on page 67 of the report for Study 3203-009, and Table 1 on page 5 of the submission dated June 13, 2006. Disposition In the Phase 2/3 oxymorphone ER trials, 64% of the 1828 patients discontinued early with adverse events being the most common cause Across the oxymorphone 1R trials, 42% of the 55 7 patients in oxymorphone IR Phase 2/3 trials discontinued early. Lack of ef?cacy was the most common reason for early discontinuation. This is comparable to the experience with the patients who received oxycodone IR in these trials: 34% of the 278 patients treated discontinued early, 22% due to lack of ef?cacy. SAFETY Deaths There were a total of 49 deaths in the oxymorphone development program. Thirty ?ve of these deaths were reviewed during the original application. Thirty four of these deaths occurred in patients with cancer pain. Dr. Fang has performed a detailed review of the deaths, reviewing each of the narratives. Thirteen of the deaths reported in this submission were reported from open-label studies, EN3202-021, 028, and 029. These were cancer patients who had been on oxymorphone treatment for a period ranging from two weeks to 10 weeks prior to the death. There does not appear to be a causative relationship between these deaths and study drug. The remaining death was a 73-year old woman with chronic low back pain and a history of hypertension, diabetes mellitus, hyperlipidemia, Parkinson?s disease, prior myocardial infarction, and breast cancer. The patient had several episodes of pneumonia and sinus infection after starting Study EN3202-032 while titrating onto oxymorphone ER with oxymorphone IR as rescue. The patient had previously been treated with OxyContin. It is unlikely that the patient?s pneumonia and subsequent death was related to use of study medication. Serious Adverse Events The nature of the serious adverse events during clinical trials with oxymorphone ER did not change appreciably from the original application. The overall incidence was comparable and the individual event frequencies did not change substantially. The four cases of drug interaction NOS were oxymorphone overdoses in the original application in the postoperative setting where PCA oxymorphone was also part of the acute postoperative treatment for pain. No further cases of overdose were reported. Appears This Way On Original 33 Deputy Director Memo 21-610 21 -61 1 Table 18. Number of Oxymerphone ER-treated Patients with Non-Fatal SAEs in Descending Frequency in Phase 2/3 ER Trials 120?Day Update Overall Safety Total patients treated with oxymorphone ER (N 1089) (N 972) 2011) preferred term Any adverse events 93 (8.5) 62(6.4) 150 (7.5) Chest pain NBC '7 (0.6) 4 (0.4) 11 (0.5) Pneumonia NOS 3 (0.3) 8 (0.8) 11 (0.5) Vomiting NOS 8 (0.7) 3 (0.3) 10 (0.5) Dehydration 5 (0.5) 4 (0.4) 9 (0.4) Nausea 6 (0.6) 4 (0.4) 9 (0.4) NOS 5 (0.5) 1 (0.1) 6 (0.3) Abdominal pain NOS 4 (0.4) 2 (0.2) 5 (0.2) Back pain 3 (0.3) 1 (0.1) 4 (0.2) Confusional state 0 (0.0) 4 (0.4) 4 (0.2) Drug interaction NOS 4 (0.4) 0 (0.0) 4 (0.2) Hypotension NOS 3 (0.3) 1 (0.1) 4 (0.2) Osteoarthritis aggravated 4 (0.4) 0 (0-0) 4 (0.2) Pulmonary embolism 2 (0.2) 2 (0.2) 4 (0.2) Pyrexia 2 (0.2) 2 (0.2) 4 (0.2) Urinary retention 3 (0.3) 1 (0.1) 4 (0.2) Atrial ?brillation 3 (0.3) 0 (0.0) 3 (0.1) Cholelithiasis 1 (0.1) 2 (0.2) 3 (0.1) Chronic obstructive airways disease exacerbated 2 (0.2) (0.1) 3 (0.1) Concomitant disease progression 2 (0.2) 1 (0.1) 3 (0.1) Depressed level of consciousness 3 (0.3) 0 (0.0) 3 (0.1) Malignant pleural effusion 1 (0.1) 2 (0.2) 3 (0.1) Pain exacerbated 2 0.2) 1 (0.1) 3 (0.1) Pain in limb 3 (0.3) 0 (0.0) 3 (0.1) Pain NOS 1 (0.1) 2 (0.2) 3 (0.1) Urinary tract infection NOS 3 (0.3) 0 (0.0) 3 (0.1) Venous thrombosis deep limb 3 (0.3) 0 (0.0) 3 (0.1) Chest pain aggravated 0 (0.0) 2 (0.2) 2 (0.1) Drug withdrawal 0 (0.0) 2 (0.2) 2 (0.1) Hypoxia 0 (0.0) 2 (0.2) 2 (0.1) Intractable pain 0 (0.0) 2 (0.2) 2 (0.1) Malignant neoplasm progression 0 (0.0) 2 (0.2) 2 (0.1) Mental status changes 0 (0.0) 2 (0.2) 2 (0.1) Obstructive uropathy 0 (0.0) 2 (0.2) 2 (0.1) Renal failure NOS 0 (0.0) 2 (0.2) 2 (0.1) Respiratory failure 0 (0.0) 2 (0.2) 2 (0.1) Syncope 0 (0.0) 2 (0.2) 2 (0.1) Subjects who either continued in one of the studies after the 120-Day Safety Update cut-off or subjects who experienced AEs not reported in the 120?Day Update (the AES had to be new or had to have increased in severity from AEs previously reported) or were from studies and EN3202-032. Overall column includes all serious AEs, some of which may not be classi?ed as serious before the 120-Day Safety cutoff date but later became serious. Source: Applicant Table 16 on page 171-179, 34 Deputy Director Memo 21 -610 21 ?61 The serious adverse events in the oxymorphone IR trials were also less frequent in the new data compared to the original safety database. Table 19. Number of Oxymorphone IR-Treated Patients with Non?Fatal SAEs in Descending Frequency in Phase 2/3 IR Trials 188 Update Overall Total patients treated with oxymorphone ER (N 334) (N 223) (N 557) preferred term Any adverse events 18 (5.4) 7 (3.1) 25 (4.5) Myocardial infarction 3 (0.9) (0.0) 3 (0.5) Venous thrombosis deep limb 3 (0.9) 0 (0.0) 3 (0.5) Corna NBC 2 (0.6) (0.0) 2 (0.4) Ileus 2 (0.6) 0 (0.0) 2 (0.4) Pneumonia NOS 1 (0.3) 1 (0.4) 2 (0.4) Pyrexia 1 (0.3) 1 (0.4) 2 (0.4) Adult respiratory distress 1 (0.3) 0 (0.0) (0.2) Cardiogenic shock I (0.3) 0 (0.0) 1 (0.2) Chest pain NBC 0 (0.0) (0.4) 1 (0.2) Con?ision 1 (0.3) 0 (0.0) 1 (0.2) Depressed level of consciousness 1 (0.3) (0.0) 1 (0.2) Disorientation 1 (0.3) 0 (0.0) (0.2) NOS 1 (0.3) 0 (0.0) (0.2) Hypotension NOS 1 (0.3) 0 (0.0) i (0.2) Hypoventilation (0.3) 0 (0.0) (0.2) Hypoxia 1 (0.3) 0 (0.0) 1 (0.2) Leukocytosis 0 (0.0) (0.4) 1 (0.2) Mental status changes 1 (0.3) 0 (0.0) 1 (0.2) Muscle contractions involuntary 1 (0.3) 0 (0.0) 1 (0.2) Nausea - 0 (0.0) (0.4) 1 (0.2) Pneumothorax NOS 1 (0.3) 0 (0.0) 1 (0.2) Postoperative ileus 0 (0.0) 1 (0.4) (0.2) Radius fracture 1 (0.3) 0 (0.0) 1 (0.2) Renal failure acute (0.3) 0 (0.0) 1 (0.2) Respiratory distress - (0.3) (0.0) (0.2) Small intestinal obstruction NOS 0 (0.0) 1 (0.4) 1 (0.2) Somnolence 1 (0.3) 0 (0.0) (0.2) Sweating increased 1 (0.3) 0 (0.0) 1 (0.2) Tachycardia NOS 0 (0.0) 1 (0.4) 1 (0.2) Tendon rupture 1 (0.3) 0 (0.0) (0.2) Tremor NBC 1 (0.3) (0.0) 1 (0.2) Uterine cancer NOS 0 (0.0) 1 (0.4) (0.2) Vaginal cellulitis 0 (0.0) 1 (0.4) (0.2) Vomiting NOS 0 (0.0) (0.4) (0.2) Wound infection NBC 1 (0.3) (0.0) 1 (0.2) New Subjects since the ISS. Source: Table 17 on pages 180-181 of the update safety report. Dr. Fang did a thorough review of cardiac serious adverse events and did not ?nd any potential causal relationship or safety signal of concern. Her thorough review of data found that 35 Deputy Director Memo 21?610 21 -61 these events were not apparently-drug related as most resolved while on oxymorphone or after rechallenge. Discontinuation Due to Adverse Events The frequency of discontinuation due to adverse events was lower in the new data than in the original application Dr. Fang notes that this may be due to the study design in which patients titrated onto study drug in Studies EN3202-031 and 032 which may have improved tolerability. Table 20. Number of Oxymorphone ER-treated Patients with AEs Causing Dropouts in Descending Frequency in Phase 2/3 ER trials 120-Day Safety Update Overall Total patients treated with oxymorphone ER (N 1089) (N 972) (N 201 l) preferred term Any adverse events 391 (35.9) 251 (25.8) 642 (31.9) Nausea 157 (14.4) 59 (6.1) 216 (10.7) Dizziness (exc vertigo) 93 (8.5) 32 (3.3) 124 (6.2) Vomiting NOS 85 (7.8) 23 (2.4) 108 (5.4) Somnolence 43 (3.9) 29 (3.0) 72 (3.6) Constipation 45 (4.1) 26 (2.7) 71 (3.5) Pruritus NOS 39 (3.6) -18 (1.9) 57 (2.8) Headache NOS 26 (2.4) 16 (1.6) 42 (2.1) Sweating increased 27 (2.5) 10 (1.0) 37 (1.8) Sedation 26 (2.4) 6 (0.6) 32 (1.6) Fatigue 16 (1.5) 12 (1.2) 28 (1.4) Insomnia NOS 10 (0.9) 9 (0.9) 19 (0.9) Abdominal pain NOS 10 (0.9) 8 (0.8) 18 (0 .9) Diarrhea NOS - 11 (1.0) 6 (0.6) 17 (0.8) Dry mouth 13 (1.2) 4 (0.4) 17 (0.8) Lethargy 8 (0.7) 8 (0.8) 16 (0.8) Concomitant disease progression (0.8) '2 (0.2) 14 (0.7) Confusion (1.2) 0 (0.0) 14 (0.7) Anxiety NBC 6 (0.6) 6 (0.6) 12 (0.6) NOS 6 (0.6) 5 (0.5) 11 (0.5) Nervousness 5 (0.5) 6 (0.6) 11 (0.5) Appetite decreased NOS 7 7 (0.6) 3 (0.3) 10 (0.5) Disorientation 7 (0.6) 3 (0.3) 10 (0.5) Subjects who either continued in one of the studies after the 120-Day Safety Update cut?off or subjects who experienced AEs not reported in the 120?Day Update (the AEs had to be new or had to have increased in severity from AEs previously reported) or were from studies EN3202-028, and EN3202-032. Overall column includes all AEs causing study drug discontinuation, some of which may not be included in the other two columns since these AEs existed before the 120?Day Safety Update, but the subject later discontinued without the AE worsening in severity. Source: Table 18 on page 184 to 196. The adverse events leading to study discontinuation were very similar in the new data as compared to the original submission. 36 Deputy Director Memo 21-610 21-61 1 Table 21. Number of Oxymorphone IR?Treated Patients with AEs Causing Dropouts in Descending Frequency in Phase 2/3 IR Trials 120-Day Safety Update Overall Total patients treated with oxymorphone IR (N 334) (N 223) (N 2 557) preferred term Any adverse events 34 (10.2) 21 (9.4) 55 (9.9) Nausea 6 (1.8) 8 (3.6) 14 (2.5) Vomiting NOS 6 (1.8) 6 (2.7) 12 (2.2) Sedation 4 (1.2) 2 (0.9) 6 (1.1) Somnolence 5 (1.5) 1 (0.4) 6 (1.1) Headache NOS 1 (0.3) 3 (1.3) 4 (0.7) Coma NBC 3 (0.9) 0 (0.0) 3 (0.5) Confusion 3 (0.9) 0 (0.0) 3 (0.5) Respiratory depression 3 (0.9) 0 (0.0) 3 (0.5) Abdominal pain NOS 1 (0.3) (0.4) 2 (0.4) Agitation I (0.3) 0 (0.0) 1 (0.2) Confusional state 0 (0.0) 1 (0.4) (0.2) Constipation (0.3) 0 (0.0) (0.2) Depressed level of consciousness 1 (0.3) 0 (0.0) 1 (0.2) Disorientation 1 (0.3) 0 (0.0) 1 (0.2) NOS 1 (0.3) 0 (0.0) 1 (0.2) Feeling abnormal 1 (0.3) 0 (0.0) 1 (0.2) Hallucination NOS 1 (0.3) 0 (0.0) 1 (0.2) Headache NOS aggravated 1 (0.3) 0 (0.0) 1 (0.2) Hypotension NOS 1 (0.3) 0 (0.0) 1 (0.2) Hypoventilation 1 (0.3) 0 (0.0) 1 (0.2) Hypoxia 1 (0.3) 0 (0.0) 1 (0.2) Ileus (0.3) 0 (0.0) 1 (0.2) Incision site complication 0 (0.0) 1 (0.4) 1' (0.2) Lethargy 1 (0.3) 0 (0.0) 1 (0.2) Mental status changes 1 (0.3) 0 (0.0) 1 (0.2) Migraine NOS 0 (0.0) 1 (0.4) 1 (0.2) Myocardial infarction . 1 (0.3) 0 (0.0) (0.2) Pruritus NOS 0 (0.0) 1 (0.4) 1 (0.2) Rash NOS 0 (0.0) 1 (0.4) 1 (0.2) ReSpiratory distress 1 (0.3) (0.0) 1 (0.2) Sweating increased 1 (0.3) 0 (0.0) 1 (0.2) Tachycardia NOS 0 (0.0) 1 (0.4) 1 (0.2) New Subjects since the 188. Source: Table 19 on pages 197 to 198 of the update safety report. Adverse Events The most common adverse events are for the most part consistent with known opioid adverse events and did not change appreciably in the current data when compared to the original database. The overall rate of adverse events, 85% in the oxymorphone ER group compared to 57% in the placebo group. Due to the availability of rescue medication in the clinical trials, the placebo patients also experienced some opioid-related adverse events however limits placed on rescue is several studies would limit the adverse events as well. 37 Deputy Director Memo Table 22. The Most Frequent AEs Reported in All (Controlled and Open-Label) Phase 2/3 ER Trials Treatment group ER Placebo Database 120-Day Update Overall 120-Day Update Overall Total patients (N 1089) (N 972) (N 2011) (N 350) (N 172) (N 522) preferred term Any adverse experience 976 (89.6) 759 (78.1) 1702 (84.6) 225 (64.3) 71 (41.3) 296 (56.7) Nausea 500 (45.9) 213 (21.9) 711 (35.4) 63 (18.0) 10 (5.8) 73 (14.0) Constipation 435 (39.9) 236 (24.3) 670 (33.3) 63 (18.0) 2 (1.2) 65 (12.5) Dizziness (exc vertigo) 280 (25.7) 111 (11.4) 388 (19.3) 35 (10.0) 3 (1.7) 38 (7.3) Vomiting NOS 256 (23.5) 91 (9.4) 346 (17.2) 23 (6.6) 2 (1.2) 25 (4.8) Pruritus NOS 264 (24.2) 75 (7.7) 339 (16.9) 42 (12.0) 1 (0.6) 43 (8.2) Somnolence 179 (16.4) 154 (15.8) 333 (16.6) 14 (4.0) 0 (0.0) 14 (2.7) Headache NOS 129 (11.8) 104 (10.7) 233 (11.6) 26 (7.4) 2 (1.2) 28 (5.4) Sweating increased 200 (18.4) 34 (3.5) 232 (11.5) 37 (10.6) 4 (2.3) 41 (7.9) Sedation 188 (17.3) 14 (1.4) 202 (10.0) 37 (10.6) 0 (0.0) 37 (7.1) Dry mouth 81 (7.4) 44 (4.5) 125 (6.2) 1 (0.3) 2 (1.2) 3 (0.6) Insomnia NBC 62 (5.7) 55 (5.7) 117 (5.8) 8 (2.3) 2 (1.2) 10 (1.9) Diarrhea NOS 68 (6.2) 47 (4.8) 115 (5.7) 18 (5.1) 8 (4.7) 26 (5.0) Fatigue 60 (5.5) 51 (5.2) 111(5.5) 4(l.1) 2(1.2) 6(1.1) Appetite decreased NOS 54 (5.0) 24 (2.5) 78 (3.9) 1 (0.3) 1 (0.6) 2 (0.4) Subjects who either continued in one of the studies after the 120-Day Safety Update cut-off or subjects who experienced AEs not reported in the 120?Day Update (the ABS had to be new or had to have increased in severity from AEs previously reported) or were from studies EN3202-028, EN3202-029, EN3202-031 and EN3202-032. Note: ABS included in this table occurred in 25% of subjects in either column. This table is sorted by Overall Total frequency in descending order. Source: Table 26 on pages 383 to 422. The common adverse events were also very similar but less frequent in the updated data compared to the original data. The comparison to oxycodonc IR is helpful in putting the adverse event pro?le of oxymorphone IR in context. There are no trends for more frequent events with oxymorphone IR as compared to oxycodone IR and for some 38 Deputy Director Memo Appears This Way On Original 2l-610 21-611 Table 23. The Most Frequent AEs Reported in All Phase 2/3 Oxymorphone IR Trials Oxymorphone IR Oxycodone IR Placebo Database 188 Update Overall ISS Update Overall ISS Update Overall #treated Any 114 (71.0) (50.2) (62.7) @546) (54.2) (61.5) (46.3) (3 8.8) (42.2) Nausea (16.5) (22.9) (19.0) (19.5) (27.7) (21.9) (6.5) (15.6) (11.5) Pyrexia (21.9) (2.7) (14.2) (15.9) (4.8) (12.6) (15.4) (2.0) (8.1) Somnolence (14.7) (1.3) (9.3) (13.8) (2.4) (10.4) (4.1) (0.7) (2.2) Vomiting NOS (7.8) (10.8) (9.0) (6.7) (9.6) (7.6) (4.1) (9.5) (7.0) Pruritus NOS (7.8) (8.1) (7.9) (6.2) (13.3) (8.3) (3.3) (4.1) (3.7) Headache NOS (3.0) (12.6) (6.8) (4. 1) (7.2) (5.0) (0.8) (7.5) (4.4) Dizziness (8.4) (3.6) (6.5) (5.1) (8.4) (6.1) (1.6) (2.7) (2.2) Constipation (5.1) (2.7) (4 .1) (7.2) (2.4) (5.8) (0.8) (1.4) (1.1) Confusion (4.5) (0.0) (2.7) (2.6) (0.0) (1.8) (1.6) (0.0) (0.7) Anemia NOS (3.9) (0.4) (2.5) (2.1) (0.0) (1.4) (3 .3) (0.0) (1.5) Dry mouth (2.4) (0.9) (1.8) (0.5) (0.0) (0.4) (0.0) (1.4) (0 .7) Tachycardia .1) (1.3) (1 .8) (0.5) (0.0) (0.4) (1.6) (0.0) (0.7) Hypoxia (2.4) (0.4) (1.6) (4.1) (1.2) 7 (3.2) (4.1) (0.7) (2.2) Note: The 28 patients who received both placebo and oxymorphone treatments were counted in both groups. Source: Table 27 on pages 423 to 435 of the update safety report. Laboratory Findings Dr. Fang reviewed the changes in white blood cell count noted in the prior review of safety. She found that there were no consistent ?ndings that could be attributed to study drug, but in fact many were due to laboratory sample mishandling. Several repeated tests were normal. Adverse Events of Interest Patients Requiring Opioid Antagonists - A total of 27 subjects in the original 188 received naioxone. Twenty-three of the 27 subjects requiring naloxone were enrolled in one of the three acute post-operative pain trials (EN3202- 012, EN3203-004, and EN 3203-005). The rates of naloxone use are detailed in Table 18. The incidence of use of naloxone in postsurgical patients receiving oxymorphone ER was 39 Deputy Director Memo 21-610 21 ?6 Table 18 Incidence of Naloxone Use in Acute Post-Operative Pain Trials Study Group EN3202-012 OXymorphone ER Placebo N365 4 0 EN3203 -004 and Oxymorphone IR Oxycodone IR Placebo EN3203-005 12 1 The four subjects in the oxymorphone ER group from the preceding table had an adverse event coded to the term ?drug interaction but these were actually overdoses. Each of these events occurred following use of a single dose of oxymorphone ER and several doses of PCA hydromorphone. One of these patient had received a 60 mg dose of oxymorphone ER as well. In the current update, there were four cases of hypoxia identi?ed in the oxymorphone ER patient groups. Two of these were serious. The narratives of these were reviewed with all of the serious adverse events. Patient 212?012 was a 44 year old man with metastatic lung cancer. He began treatment with oxymorphone in March of 2004 and had progressive including hypoxia on room air in . He was found to have a large pleural effusion. There is no evidence that the hypoxia was related to oxymorphone. Patient 127-004 was a 63 year old woman with neuropathic pain. She began oxymorphone in December of 2003. In the patient developed pneumonia with hypoxia that resolved with treatment. There is no evidence of a casual relationship between study medication and hypoxia in this patient. One case of hypoxia in study EN3202-032 was not considered serious, but the patient received naloxone in the setting of a tooth extraction at which time also receiving Vicodin. The dose of oxymorphone ER was not documented at that time. It is impossible to know whether there was a causal relationship between the oxymorphone and the event independent of effects of Vicodin. Drug-Drug Interactions No major clinically signi?cant drug-drug interactions were noted in the clinical trials. Withdrawal Effects In the initial application, there were six oxymorphone ER?treated subjects and two oxycodone ER-treated subjects with adverse events mapped to drug withdrawal Two of these events occurred in Study EN3202-016, in which subjects were re-randomized to placebo treatment after stabilization on double-blind treatment with oxymorphone ER or oxycodone ER. In the update, there were ?ve patients with of opioid withdrawal in Study EN3202- 032 after randomization to placebo. These were opioid tolerant patients and the available rescue medication was insuf?cient to avoid withdrawal There were also three patients in Study EN3202-03 1, two after randomization to placebo, one while on treatment with oxymorphone. 40 Deputy Director Memo 21 -6i 0 21 -61 As an opioid, the potential for opioid withdrawal is not surprising and patients who are opioid tolerant will need adequate taper if they are to be taken off oxymorphone. Care in conversion to other opioids will also need to be taken in order to avoid withdrawal as well as toxicity. Overdose As described in the original review of safety, there was (me accidental overdose by a patient who mistakenly took four tablets instead of one, for one dose. The subject contacted the study site and was monitored. No adverse events were noted. Several cases of patients requiring naloxone for oversedation or respiratory depression in the postoperative setting have been described. There were no reports of overdose in the updated data in the current submission. Drug Abuse and Abuse Liability Oxymorphone is a mu-agonist opioid analgesic and it?s abuse liability can be expected to be similar to morphine. Data from the clinical trials suggest that withdrawal will occur with abrupt discontinuation. There were two episodes of drug diversion at two clinical sites during the clinical development program as described in the original review. There were no new reports of diversion in the updated data. The Controlled Substances Staff provided a consultation. It appears that the recommendations were based on a review of the original NDA submission which identified problems with proposed dosing for the oxymorphone IR formulation and which found safety concerns with the oxymorphone ER product in the postoperative period. This review addresses those concerns. The CSS consult made the following recommendations: 1. Medication Guide CSS states that the potency and dosage for this product mandate a Med Guide for the ER formulation and a PPI for the IR formulatiOn. The relative potency of oxymorphone compared to other opioids is not well de?ned and may differ in the setting of opioid tolerance, but CSS notes relative potencies of approximately 0.333 for morphine and 1.33 for hydromorphone. The same source suggests a relative potency of 0.5 for oxycodone. The proposed for the oxymorphone ER formulation, 5, 10, 20,and 40 mg. However, oxymorphone ER and IR has not been found to have any greater evidence of toxicity in this safety database when compared to oxycodone ER and IR. The oral bioavailability of oxymorphone ER and IR is only 10%, in contrast to 60?80% for oxycodone ER which is available in dosage up to 80 mg (160 mg is approved, not currently marketed). The oral bioavailability for morphine ER is approximately 40 and these are available in dosage up to 100 mg. So while relative potency of the actual drug substance of oxymorphone may be greater than for morphine or oxycodone, the limited bioavailability substantially reduces the relative potency of the drug product. The only modi?ed-release oral opioid approved with a Medication Guide is Palladone. All of Palladone are indicated only for opioid tolerant patients due to the amount high potency hydromorphone. Oxymorphone ER has been safely titrated and used in opioid naive patients as demonstrated in Study EN3202-031. 41 Deputy Director Memo 21-610 21 -61 42 Therefore, I disagree that a Medication Guide is indicated for this oxymorphone ER. Similarly, reducing the dose of oxymorphone IR to no more than 20 mg in the postoperative period eliminated the serious adverse events and provided a safety pro?le comparable to oxycodone IR. Therefore I do not agree that a PPI is indicated for oxymorphone IR when it is not indicated for oxycodone IR. 2. The labeling should be revised to include more clearly the risks of prescribing the ER dosage form to non-opioid tolerant patients. I agree that this is extremely important. The following language is in the current package insert for oxymorphone ER. Initiation of Therapy Opioid?Naive Patients It is suggested that patients who are not opioid-experienced being initiated on chronic around-the-clock opioid therapy be started with OPANA ER 5 mg every 12 hours. Thereafter, it is recommended that the dose be individually titrated, preferably at increments of 5?10 mg every 12 hours every 3?7 days, to a level that provides adequate analgesia and minimizes side effects under the close supervision of the prescribing physician (see CLINICAL TRIALS: 12-Week Study in Opioid?Naive Patients with Low Back Pain) 3. The labeling should address the safety issues of high potency and high rate of adverse effects associate with oxymorphone. The adverse event pro?le for the two formulations of oxymorphone were comparable to the adverse events with oxycodone treatment arms in the same studies. 4. Revise the information provided to assist in conversion form other oral opioid to proposed products to a more user friendly, less confusion format with sources(s) for data cited. The current conversion table represents the conversion used in clinical trials and is intended to be a conversion guide and not an equianalgesic table. Based on the clinical trials, patients converting from other opioids did not need to titrate down after this conversion. To ensure greater safety across broad use, the package insert recommends calculating the conversion, halving the dose and then titrating as needed. This following is from the package insert: Conversion from Other Oral Opioids to OPANA ER For conversion from other opioids to OPANA ER, physicians and other health care professionals are advised to refer to published relative potency information, keeping in mind that conversion ratios are only approximate. In general, it is safest to start the OPANA therapy by administering half of the calculated total daily dose ER (see conversion ratio table below) in 2 divided doses, every 12 hours. The initial dose of OPANA ER can be gradually adjusted until adequate pain relief and acceptable side effects have been achieved. the following table provides approximate equivalent doses, which may be used as a Deputy Director Memo 21610 21 -61 1 guideline. In a Phase 3 clinical trial with an open-label titration period, patients were converted from their current opioid to OPANA ER using the following table as a guide. In general, patients were able to successfully titrate to a stabilized dose of OPANA ER within 4 weeks (see CLINICAL TRIALS: 12?Week Study in Opioid-Experienced Patients with Low Back). There is substantial patient variation in the relative potency of different opioid drugs and formulations. CONVERSION RATIOS TO OPAN A ER Approximate Equivalent Dose Oral Opioid Oral Conversion Ratio? Oxymorphone 10 mg 1 Hydrocodone 20 mg 0.5 Oxycodone 20 mg 0.5 Methadone 20 mg 0.5 Morphine 30 mg 0.333 aRatio for conversion of oral opioid dose to approximate oxymorphone equivalent dose. Select opioid and multiply the dose by the conversion ratio to calculate the approximate oral oxymorphone equivalent. 0 Sum the total daily dose for the opioid and multiply by the conversion ratio to calculate the oxymorphone total daily dose. 0 For patients on a regimen of mixed opioids, calculate the approximate oral oxymorphone dose for each opioid and sum the totals to estimate the total daily oxymorphone dose. 0 The dose of OPANA ER can be gradually adjusted, preferably at increments of 10 mg every 12 hours every 3?7 days, until adequate pain relief and acceptable side effects have been achieved (see lndividualization of Dose). 5. The sponsor should provide more details on the DrugLogic proportional analysis engine data and how it will be used to manage risk in the post-marketing surveillance plan for the proposed product. This comment has been forward to the Applicant. The risk minimization plan has been reviewed the Of?ce of Safety and Epidemiology and has been found to be acceptable. SPECIAL POPULATION There were too few non Caucasian subjects to adequately explore the effects of race and ethnicity on ef?cacy. Among oxymorphone ER-treated subjects in the Phase 2/3 ER trials, the incidence of dizziness, somnolence, and headache were higher in Caucasians compared to Blacks, but the relatively small proportion of Blacks makes the clinical signi?cance of this difference uncertain. There were too few subjects over the age of 65 years to provide a meaningful evaluation of the effects of age on ef?cacy in the only trial supporting the proposed indication, The frequency of some adverse events, such as somnolence and dizziness, may increase with increasing age in oxymorphone ER-treated subjects. 43 Deputy Director Memo 21-610 21-61 I OPANA ER should be used with caution in elderly patients. The plasma levels of oxymOrphone are about 40% higher in elderly (265 years of age) than in younger subjects (see CLINICAL PHARMACOLOGY). Elderly patients should initially receive smaller starting doses of oxymorphone and dose titration should proceed cautiously. Of the total number of subjects in clinical studies of oxymorphone ER, 27 percent were 65 and over, while 9 percent were 75 and over. No overall differences in effectiveness were observed between these subjects and younger subjects. There were several adverse events that were more frequently observed in subjects 65 and over compared to younger subjects. These adverse events included dizziness, somnolence, confusion, and nausea. Of the total number of subjects in clinical studies of oxymorphone IR, 31 percent were 65 and over, while 7 percent were 75 and over. No overall differences in effectiveness were observed between these subjects and younger subjects. There were several adverse events that were more frequently observed in subjects 65 and over compared to younger subjects. These adverse events included dizziness, somnolence, confusion, and nausea. In general, dose selection for elderly patients should be cautious, usually starting at the low end of the dosing range, re?ecting the greater frequency of decreased hepatic, renal or cardiac function, and of concomitant disease or other drug therapy. There were no consistent effects of gender on ef?cacy. Nausea, vomiting, and headache were notably more frequently in females 28.0%, and 14.6%, respectively), compared to males 13.9%, and respectively). This gender difference was not seen in placebo-treated subjects There will be a postmarketing commitment to ful?ll the requirements of PREA. DOSING The clinical pharmacokinetic and bioavailability studies were also conducted. A dosing interval of every 12 hours has been utilized in the Study EN3203-016 in which ef?cacy was demonstrated and is supported by the pharmacokinetic pro?le of oxymorphone extended?release tablets. Dose adjustments are called for in mild hepatic impairment, titration should begin low and proceed with close clinical monitoring. Oxymorphone is highly metabolized by the liver. Use of oxymorphone should be contraindicated in moderate and severe hepatic impairment. As oxymorphone plasma concentrations were relatively higher in the setting of renal impairment, dosing of oxymorphone should be started at low doses and titrated carefully in all categories of renal impairment under close clinical supervision. Patients over age of 65 exhibited higher plasma concentrations, AUC and Cmax. Therefore, dosing in patients over the age of 65 should begin with low starting doses and titrated carefully under close clinical supervision. 44 Deputy Director Memo 21-610 21 -61 1 The following is from the Dosage and Administration section from the package insert: Oxymorphone ER (OPANA ER) Initiation of Therapy Opioid-Naive Patients It is suggested that patients who are not opioid-experienced being initiated on chronic around- the?clock opioid therapy be started with OPANA ER 5 mg every 12 hours. Thereafter, it is recommended that the dose be individually titrated, preferably at increments of 5-1 0 mg every 12 hours every 3-7 days, to a level that provides adequate analgesia and minimizes side effects under the close supervision of the prescribing physician (see CLINICAL TRIALS: lZ-Week Study in Opioid-Naive Patients with Low Back Pain) Opioid-Experienced Patients Conversion from OPANA to OPANA ER Patients receiving OPANA (IR) may be converted to OPANA ER by administering half the patient's total daily oral OPANA (IR) dose as OPANA ER, every 12 hours. For example, a patient receiving 40 mg/day OPANA (IR) may require 20 mg OPANA ER every 12 hours. Conversion from Parenteral Oxymorphone to OPANA ER Given the absolute oral bioavailability of approximately 10%, patients receiving parenteral oxymorphone may be converted to OPANA ER by administering 10 times the patient's total daily parenteral oxymorphone dose as OPANA ER in two equally divided doses IV dose 10/2). For example, approximately 20 mg of OPANA ER, every 12 hours, may be required to provide pain relief equivalent to a total daily dose of 4 mg of parenteral oxymorphone. Due to patient variability with regards to opioid analgesic response, upon conversion patients should be closely monitored to ensure adequate analgesia and to minimize side effects. Conversion from Other Oral Opioids to OPANA ER For conversion from other opioids to OPANA ER, physicians and other health care professionals are advised to refer to published relative potency information, keeping in mind that conversion ratios are only approximate. In general, it is safest to start the OPANA therapy by administering half of the calculated total daily dose of OPANA ER (see conversion ratio table below) in 2 divided doses, every 12 hours. The initial dose of OPANA ER can be gradually adjusted until adequate pain relief and acceptable side effects have been achieved. the following table provides approximate equivalent doses, which may be used as a guideline. In a Phase 3 clinical trial with an open?label titration period, patients were converted from their current opioid to OPANA ER using the following table as a guide. In general, patients were able to successfully titrate to a stabilized dose of OPANA ER within 4 weeks (see CLINICAL TRIALS: 12-Week Study in Opioid-Experienced Patients with Low Back). There is substantial patient variation in the relative potency of different opioid drugs and formulations. Appears This Way On Original 45 Deputy Director Memo 21 -6i0 21 -61 1 CONVERSION RATIOS TO OPANA ER Approximate Equivalent Dose Oral Opioid Oral Conversion Ratio? Oxymorphone 10 mg 1 Hydrocodone 20 mg 0.5 Oxycodone 20 mg 0.5 Methadone 20 mg 0.5 Morphine 30 mg 0.333 ?Ratio for conversion of oral opioid dose to approximate oxymorphone equivalent dose. Select opioid and multiply the dose by the conversion ratio to calculate the approximate oral oxymorphone equivalent. 0 Sum the total daily dose for the opioid and multiply by the conversion ratio to calculate the oxymorphone total daily dose. 0 For patients on a regimen of mixed opioids, calculate the approximate oral oxymorphone dose for each opioid and sum the totals to estimate the total daily oxymorphone dose. 0 The dose of OPANA ER can be gradually adjusted, preferably at increments of 10 mg every 12 hours every 3-7 days, until adequate pain relief and acceptable side effects have been achieved (see Individualization of Dose). Oxymorphone IR (OPANA) Initiation of Therapy Opioid-Naive Patients Patients who have not been receiving opioid analgesics should be started on OPANA in a closing range of 10 to 20 mg every four to six hours depending on the initial pain intensity. If deemed necessary to initiate therapy at a lower dose, patients may be started with OPANA 5 mg. The dose should be titrated based upon the individual patient?s response to their initial dose of OPANA. This dose can then be adjusted to an acceptable level of analgesia taking into account the pain intensity and side effects experienced by the'patient. Initiation of therapy with doses higher than 20 mg is not recommended because of potential serious side effects (see CLINICAL TRIALS: Orthopedic Surgery). Conversion from Parenteral Oxymorphone to OPANA Given the absolute oral bioavailability of approximately 10%, patients receiving parenteral oxymorphone may be converted to OPANA by administering 10 times the patient?s total daily parenteral oxymorphone dose as OPANA, in four or six equally divided doses g. IV close 10/4). For example, approximately 10 mg of OPANA may be required to provide pain relief equivalent to a total daily IM dose of 4 mg oxymorphone. The dose can be titrated to optimal pain relief or combined with acetaminophen/NSAIDS for optimal pain relief. Due to patient variability with regard to opioid analgesic response, upon conversion patients should be closely monitored to ensure adequate analgesia and to minimize side effects. Conversion from Other Oral Opioids to OPANA 46 Deputy Director Memo 21-610 21 -61 1 For conversion from other opioids to OPANA, physicians and other healthcare professionals are advised to refer to published relative potency information, keeping in mind that conversion ratios are only approximate. In general, it is safest to start the OPANA therapy by administering half of the calculated total daily dose of OPANA in 4 to 6 equally divided doses, every 4?6 hours. The initial dose of OPANA can be gradually adjusted until adequate pain relief and acceptable side effects have been achieved. Additional comments Dr. Fang suggest that the Applicant enter into a postmarketing commitment to study the relative potency of oxymorphone and other opioids. As a safe conversion method from other opioids has been described, this can be left up to the Applicant?s discretion. Appears This Way On Original 47 Deputy Director Memo 21-610 21?61 1 This is a representation of an electronic record that was signed electronically and this page is the manifestation of the electronic signature. Sharon Hertz 6/22/2006 07:01:44 PM MEDICAL OFFICER 00?} I, CLINICAL REVIEW Application Type NDA Submission Number 21-610 Submission Code N000 Letter Date PDUF A Goal Date Reviewer Name Review Completion Date Established Name (Proposed) Trade Name Therapeutic Class Applicant Priority Designation Formulation Dosing Regimen Indication Intended Population 12-22-05, 3?28-06, 4-3-06, 5?23?06, 5-26-06, 6-2?06 June 22, 2006 Christina Fang, MD. June 22, 2006 Oxymorphone Extended?Release Tablets OPANATM ER Opioid analgesics Endo Pharmaceuticals Inc. 3s Extended?release tabletsIndividualized Moderate to severe chronic pain Adult patients requiring continuous, around-the-clock opioid therapy for an extended period of time Clinical Review of NDA 21 -61 0 N000 for oxymorphone extended release by Christina Fang 1 Table of Contents EXECUTIVE SUMMARY ..4 I .I RECOMMENDATION ON REGULATORY ACTION .. 4 1.2 RECOMMENDATION ON POSTMARKETING ACTIONS .. 4 1.2.1 Risk Management Activity .. ..4 1.2.2 Required Phase 4 Commitments ..4 1.2.3 Other Phase 4 Requests .. 4 L3 SUMMARY OF CLINICAL FINDINGS .. 4 1.3.1 Brief Overview ofClinical Program ..4 1.3.2 Ef?cacy .. 5 1.3.3 Safety .. 5 1.3.4 Dosing Regimen and Administration .. 6 1.3.5 Drug-Drug Interactions ..6 1.3.6 Special Populations ..6 INTRODUCTION AND BACKGROUND ..8 PRODUCT INFORMATION .. 8 2.2 CURRENTLY AVAILABLE TREATMENT FOR INDICATIONS .. 8 2.3 AVAILABILITY OF PROPOSED ACTIVE INGREDIENT IN THE UNITED STATES .. 8 2.4 IMPORTANT ISSUES WITH PHARMACOLOGICALLY RELATED PRODUCTS .. 8 2.5 PRESUBMISSION REGULATORY ACTIVITY .. 8 2.6 OTHER RELEVANT BACKGROUND INFORMATION .. 9 SIGNIFICANT FINDINGS FROM OTHER REVIEW DISCIPLINES ..9 3.1 CIVIC (AND PRODUCT MICROBIOLOGY, IF APPLICABLE) .. 9 3.2 ANIMAL .. 9 DATA SOURCES, REVIEW STRATEGY, AND DATA INTEGRITY ..9 4.1 SOURCES OF CLINICAL DATA .2 .. 9 4.2 TABLES OF CLINICAL STUDIES .. 10 4.3 REVIEW STRATEGY .. IO 4.4 DATA QUALITY AND INTEGRITY .. I I 4.5 COMPLIANCE WITH GOOD CLINICAL PRACTICES .. II 4.6 FINANCIAL DISCLOSURES ..1 1 CLINICAL PHARMACOLOGY .. 12 5.1 .. 12 5.2 PHARMACODYNAMICS .. 12 5.3 EXPOSURE-RESPONSE RELATIONSHIPS .. I2 INTEGRATED REVIEW OF EFFICACY .. 12 6.1 INDICATION .. 12 6.2 METHODS ..12 6.3 GENERAL DISCUSSION OF ENDPOINTS .. I2 6.4 STUDY DESIGN ..13 6.5 EFFICACY FINDINGS ..14 6.6 CLINICAL MICROBIOLOGY .. 17 6.7 EFFICACY CONCLUSIONS ..17 INTEGRATED REVIEW OF SAFETY .. 18 METHODS AND FINDINGS ..18 7.1.1 Deaths .. 18 7.1.2 Other Serious Adverse Events .. 18 7.1.3 Dropouts and Other Signi?cant Adverse Events .. 21 7.1.4 Other Search Strategies .. 22 7.1.5 Common Adverse Events ..22 7.1.6 Less Common Adverse Events ..25 7.1.7 Laboratory Findings ..25 7.1.8 Vital Signs ..26 2 Clinical Review of NBA 21-610 N000 for oxymorphone extended release by Christina Fang 7.1.9 Electrocardiograms (ECGs) .. 27 7.1.10 Immunogenicity .. 28 7.1.11 Human Carcinogenicity ..29 7.1.12 Special Safety Studies ..29 7.1.13 Withdrawal Phenomena and/0r Abuse Potential .. 29 7.1.14 Human Reproduction and Pregnancy Data ..29 7.1.15 Assessment of Effect on Growth .. 29 7.1.16 Overdose Experience ..29 7.1.17 Postmarketing Experience ..29 7.2 ADEQUACY OF PATIENT EXPOSURE AND SAFETY ASSESSMENTS .. 30 7.2.1 Description of Primary Clinical Data Sources (Populations Exposed and Extent of Exposure) Used to Evaluate Safety 30 7.2.2 Description of Secondary Clinical Data Sources Used to Evaluate Safety ..33 7.2.3 Adequacy of Overall Clinical Experience .. 33 7.2.4 Adequacy of Special Animal and/or In Vitro Testing .. 33 7.2.5 Adequacy of Routine Clinical Testing ..33 7.2.6 Adequacy ofMetabolic, Clearance, and Interaction Workup ..33 7.2.7 Adequacy for Potential Adverse Events for Any New Drug and Particularly for Drugs in the Class Represented by the New Drug; Recommendations for Further Study .. 33 7.2.8 Assessment onuaIity and Completeness of Data .. 34 7.2.9 Additional Submissions, Including Safety Update ..34 7.3 SUMMARY OF SELECTED DRUG-RELATED ADVERSE EVENTS, IMPORTANT LIMITATIONS OF DATA, AND CONCLUSIONS 34 7.4 GENERAL METHODOLOGY .. 34 7.4.1 Pooling Data across Studies to Estimate and Compare Incidence ..34 7.4.2 Explorations for Predictive Factors .. 35 7.4.3 Causality Determination .. 37 8 ADDITIONAL CLINICAL ISSUES ..38 8.1 DOSING REGIMEN AND ADMINISTRATION .. 38 8.2 DRUG-DRUG INTERACTIONS ..38 8.3 SPECIALPOPULATIONS .. 38 8.4 PEDIATRICS .. 38 8.5 ADVISORY COMMITTEE MEETING .. 38 8.6 LITERATURE REVIEW .. 38 8.7 POSTMARKETING RISK MANAGEMENT PLAN .. 38 8.8 OTHER RELEVANT MATERIALS .. 39 9 OVERALL ASSESSMENT ..39 CONCLUSIONS .. 39 9.2 RECOMMENDATION ON REGULATORY ACTION .. 39 9.3 RECOMMENDATION ON POSTMARKETING ACTIONS .: .. 39 9.3.1 Risk Management ..39 9.3.2 Required Phase 4 Commitments ..39 9.3.3 Other Phase 4 Requests ..39 9.4 LABELING REVIEW ..40 9.5 COMMENTS TO APPLICANT .. ..40 10 APPENDICES ..41 10.] REVIEW OF INDIVIDUAL STUDY REPORTS .. 4] 10.1.1 Study 031 ..41 10. .2 Study 032 .. 55 10.2 LABELING REVIEW .. 69 REFERENCES ..69 Clinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang 1 EXECUTIVE SUMMARY 1.1 Recommendation on Regulatory Action Oxymorphone ER (extended-release) is recommended for approval for the relief of moderate to severe chronic pain in patients requiring continuous, around?the?clock opioid. The recommendation for approval is based on the acceptable bene?t/risk ratio determined from the evaluation of ef?cacy and safety data submitted in 21?610. Oxymorphone ER studied at individualized dosage in highly selected responders was shown to be ef?cacious in treating chronic low back pain based on the replicable positive ?ndings from the studies of chronic low back pain. Oxymorphone ER has a similar safety pro?le as the other opioid drugs and is considered reasonably safe to be used with a careful initial dose titration to meet individual's acceptance of tolerance, for the treatment of chronic pain in both opioid naive and opioid experienced populations. 1.2 Recommendation on Postmarketing Actions None. 1.2.1 Risk Management Activity The Sponsor?s proposed post?marketing Risk Management Plan (RMP) for oxymorphone products is considered acceptable in general. The additional recommendations from the Of?ce of Drug Safety and the Controlled Substance Staff will be forwarded to the Sponsor. 1.2.2 Required Phase 4 Commitments None. 1.2.3 Other Phase 4 Requests There should be further studies of relative potency in comparison to the other commonly used opioids to well inform the labeling. 1.3 Summary of Clinical Findings 1.3.1 Brief Overview of Clinical Program The original submission 21?610 had twelve Phase 1 studies and ten Phase 2/3 studies, all reviewed in detail before an approvable regulatory action was granted by the Division in October 2003. There are four new Phase 2/3 studies in the current submission, including two ef?cacy studies of chronic low back pain (Studies 031 and 032) and two open-label safety studies of chronic pain (Studies 028 and 029). Clinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina'Fang 1.3.2 Ef?cacy The two studies (031 and 032) had a very similar design. They were studies of open-label, titration to effect followed by randomized withdrawal with a double?blind, placebo-controlled, parallel, multiple-dose design to evaluate chronic low back pain in opioid naive patients (Study 031) and opioid experienced patients (Study 032) conducted at multiple centers in the US The two studies had similar ?ndings. The ef?cacy was demonstrated by the primary outcome measure, the change in pain intensity from baseline to the ?nal visit, and supported by the protocol?de?ned secondary outcome measures, time to discontinuation due to lack of ef?cacy, and patient and physician global evaluation of medication. Treatment differences based on additional analyses were also shown in the following measurements: the change in P1 from baseline to the ?nal visit adjusted for stabilized dose level, time?speci?c measure of average PI by visit, percent reduction in average PI, proportion of responders (230% reduction in P1), percent discontinued due to lack of efficacy, time to discontinuation and proportion discontinued due to all reasons, and the use of rescue medication over the first four days and the percent of days used during the remainder of the 12?week double?blind treatment period (no difference in Study 032 for the laSt outcome in the list). The discontinuation from the open?label titration treatment was 37% due to all reasons and 26% due to AEs, lack of ef?cacy, and failure to meet titration?stabilization criteria in opioid naive patients in Study 031, and 43% due to all reasons and 31% due to AEs, lack of ef?cacy, and failure to meet titration? stabilization criteria in opioid experienced patients in Study 032. The stabilized dosage of oxymorphone ER was much higher in opioid?experienced patients than in opioid nai?ve patients and drug tolerance to the new Opioid treatment was suggested by ?ndings of worsening pain scores at the end of study in both trials. Dose response in ef?cacy was suggested by observation of larger treatment differences in the higher dosage groups in both studies. 1.3.3 Safety The length of exposure and the number of subjects exposed to long-term treatment appear adequate in the Overall database. The total exposure to oxymorphone ER treatment was reported in over 2000 patients and to placebo treatment in more than 500 patients in all the Phase 2/3 ER trials, including the studies reported in the original submission. A total of 433 patients had at least six months of exposure and 267 of them had at least 12 months of exposure in the ?ve open-label studies. The most frequent daily dose exposed was 10-50 mg/day followed by >90 mg/day and 50-90 mg/day There were 14 new reports of deaths (13 cancer deaths and one non-cancer death) in addition to 35 deaths reviewed with the original submission. The causes of deaths were most likely attributable to complications associated with the disease progression of end-stage cancer (most cases had only 0.5-2.5 months of oxymorphone treatment) in 13 cancer deaths and to community-acquired pneumonia in one non-cancer death based on the review of narratives. Serious AEs were reported in 150 of oxymorphone ER?treated subjects and the most common ones were chest pain, pneumonia, and vomiting in the Overall Safety database. The review of narratives ofthe six new cases of serious cardiac AEs identi?ed in the Update Safety database did not appear to suggest cardiac toxicity associated with oxymorphone treatment. Clinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang The Overall dropout rate was more than 60% in oxymorphone-treated subjects. The dropouts were mainly due to AEs, especiallysystems commonly associated with the use of opioid drugs. The most frequently occurring AEs in patients treated with oxymorphone ER were events expected in opioid users: nausea constipation dizziness (excluding vertigo, vomiting pruritus somnolence headache sweating increased and sedation The same set of was identi?ed as the most common, treatment-related AEs. Laboratory tests were not conducted in any of the new studies in the current submission. The review of laboratory findings suggested no safety signals for treatment-related decrease in WBC count and neutropenia (which were mostly due to laboratory sample mishandling) or LFT elevation. Vital signs were recorded in the two controlled studies (031 and 032) in the current submission. Other than a trend of small decrease in systolic and diastolic blood pressure in the first few weeks of the open-label treatment in Study 031, there were no remarkable trends based on the changes in group mean values. Based on the reanalysis of ECG data of the identi?ed cases, prolongation was mainly due to in treatment ?uctuation because most of the abnormalities were normalized upon rechallenge with the same or different oxymorphone formulations and the end?of?study measurements were normal in six of the seven cases. The ECG data from retrospective partial recollection of subjects treated in Studies 015, 020, and 025 provided no additional evidence for treatment-related prolongation. Eight cases of discontinuation due to opioid withdrawal were identi?ed in Studies 031 and 032 although the group mean scores of COWS and ARS at each scheduled visit during the double-blind treatment did not show the signs of opiate withdrawal in the two treatment groups. One case of on-study pregnancy was reported in Study 032 and ended as an elective abortion. There were no new reports of overdose in the current submission. There appeared to be an age-related increase in the incidence rates of dizziness and somnolence associated with oxymorphone treatment. The other observed higher incidence rates of some AEs in one subpopulation versus the other were probably due to normal variations. 1.3.4 Dosing Regimen and Administration The ef?cacy and safety ?ndings from the trials using titration?to?effect design support the proposed dosing regimen for opioid naive patients. The proposed conversion from other opioids in the dosing instruction for opioid experienced patients is not supported by data because the studies of relative potency between oxymorphone ER and other opioids had not provided useful information about the relative potency. 1.3.5 Drug-Drug Interactions There were no new studies of drug?drug interactions in the current submission. 1.3.6 Special Populations Elderly patients have been shown to have increased risks to oxymorphone?induced respiratory/CNS depression at higher starting doses in a post?operative setting based on the results of studies in the original submission. There was also an age?related increase in incidence rates of dizziness and somnolence. Because elderly patients are at higher risk to oxymorphone-induced drug toxicity due to higher levels of 6 Clinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang systemic exposure (about a 40% increase in total and maximum drug levels in comparison to younger subjects), oxymorphone treatment should be started at lower level with gradual titration under closer supervision. Abpeors This Way 0n Original Clinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang 2 INTRODUCTION AND BACKGROUND 2.1 Product Information The established name of the product is oxymorphone hydrochloride (HCL) and the proposed name is OPANATM ER. It is an extended?release formulation featured by the use of a hydrophilic matrix as a controlled-release drug delivery excipient. Oxymorphone HCL is a opioid? receptor agonist with the proposed mechanism of action at multiple CNS sites though interaction with opioid receptors. The proposed indication is for the relief of moderate to severe pain in patients requiring continuous, around-the-clock opioid therapy for an extended period of time, but not for mild pain, non persistent pain, or acute post-operative pain(as needed) analgesic. The proposed adult dosage for opioid naive patients is to initiate at 5 mg q12h followed by individualized titration at increments of 5-10 mg q12h every 3-7 days, to a level that provides adequate analgesia and minimizes side effects under the close supervision of the prescribing physician, and for opioid experienced patients, to convert from other opioids. 2.2 Currently Available Treatment for Indications The currently available treatments for the indication are mainly opioid analgesics, combination products containing an opioid as an active ingredient, and tramadol. 2.3 Availability of Proposed Active Ingredient in the United States The currently available drug products containing the active ingredient oxymorphone are Numorphan? injection 1 mg/ml and 1.5 mg/ml by subcutaneous and intramuscular administration (NBA 1 1-707) and Numorphan? oxymorphone rectal suppositories 5 mg (NBA 1 1?738). 2.4 Important Issues with Pharmacologically Related Products As an opioid agonist oxymorphone has similar pharmacological effects as the other drugs ofthe same class as described in the product labeling for opioid drugs. The major safety issues with the use of opioids are their potential for misuse, drug abuse and addiction and the potential for respiratory depression, eSpecially in the elderly or debilitated patients as well as in those suffering from conditions accompanied by hypoxia or hypercapnia when even moderate therapeutic doses may dangerously decrease pulmonary ventilation. The interaction ofthe opioid products with alcohol and drugs of abuse and with other CNS depressants may cause respiratory depression, hypotension, profound sedation, or coma. A common problem associated with the chronic use of opioid is the development of drug dependence and tolerance. 2.5 Presubmission Regulatory Activity The original submission 21-610 for oxymorphone ER dated December 19, 2002, was granted approvable by the Division on October 15, 2003. The Sponsor was requested to address a number of clinical deficiencies, which included providing replicable evidence for ef?cacy in a 12?week study of chronic pain and additional data to address safety concerns regarding liver function, WBC count, and QTC interval. In the subsequent interactions With the Sponsor as recorded in the meeting/teleconference minutes 8 Clinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang dated October 31, 2003, March 16, 2004, and May 25, 2004 the need for additional evidence to support ef?cacy was re?emphasized, the Sponsor's explanations for laboratory abnormality were accepted, and QTC abnormality was still a safety concern requiring further investigation. On July 12, 2004, a new protocol (Study 031) was submitted under Special Protocol Assessment (SPA) for studying multiple?dose effects of oxymorphone ER in patients with chronic low back pain. 2.6 Other Relevant Background Information There were ten phase 2/3 studies of oxymorphone ER submitted in the ?rst review cycle, three multiple? dose ef?cacy studies, one single?dose ef?cacy study, three studies of relative potency with respect to other opioid, and three open-label extension safety studies (refer to the original review for detail). The multiple-dose analgesic studies of osteoarthritis were placebo-controlled, parallel studies of 20 and 40 mg doses for four weeks (Study 015doses for two weeks (Study 025). Statistically signi?cant treatment differences in arthritic pain intensity (VAS) were not demonstrated in either study. The multiple-dose analgesic study of low back pain (Study 016) was designed in a similar way, but with shorter study duration in comparison to the two ef?cacy studies in the current submission. It had a design of titration?to?effect and randomized withdrawal with a placebo control and that the titration period was of no more than two weeks and was followed by an 18?day maintenance period, during which patients were given individualized dosing twice a day. Ef?cacy was supported by the results ofthe study. The single-dose analgesic study of the ER formulation (Study 012) raised the issues about safe use of the product in a post?operative setting. The three potency studies of a crossover design (Study 017, 018, and 019) did not result in providing useful information about the relative potency because ofthe design issues. Another NDA for the drug product containing the same active ingredient in different formulation submitted originally at about the same time (December 20, 2002) was NDA 21?61 1 for oxymorphone immediate- release (1R) formulation. The two NDAs have been resubmitted on the same date. The ef?cacy and safety of oxymorphone lR will be reviewed separately. 3 SIGNIFICANT FINDINGS FROM OTHER REVIEW DISCIPLINES 3.1 CMC (and Product Microbiology, if Applicable) Refer to the chemistry review. 3.2 Animal Pharmacology/Toxicology Refer to the pharmacology/toxicology review. 4 DATA SOURCES, REVIEW STRATEGY, AND DATA INTEGRITY 4.1 Sources of Clinical Data Ef?cacy data were from the two studies (Studies 03 and 032) of chronic low back pain. Safety data new to the current submission in the Update Safety database include data from the four newly completed studies: the two studies ofchronic low back pain (Studies 031 and 032) and the two open?label studies of chronic pain (Studies 028 and 029), as well as the newly collected data from the two open?label extension 9 Clinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang studies (Studies 021 and 022) already ongoing at the time of the 120-Day Safety data cutoff date. Overall safety data from all phase 2/3 studies are also used in the safety review. 4.2 Tables of Clinical Studies Table 4-1 Summary of Newly Completed Clinical Studies Used as Data Sources Protocol Study Type Study Design Dates of Dosage of Mean age/range Data of sites Study subj (range) relevance Gender (M, F) Race (W, NW) Dose ranging Open?label 11/24/04? Titration to 325 50.1 (18?85) Ef?cacy 31 sites ef?cacy study in titration followed 7/18/05 optimal doses of 160 M, 165 and safety opioid-naive by randomizedpatients with double-blind, then OM ER or chronic low back placebo-controlled Placebo pain I EN3202-032 Dose ranging Open-label 10/13/04? Titration to 250 49.1 (21-85) Ef?cacy 30 sites ef?cacy study in titration 8/19/05 optimal doses 1 18 M, 132 and safety patients with followed by OM ER 220mg, 219 W, 3! NW chronic low back randomized and then OM ER pain withdrawal, or Placebo double?blind, placebo-controlled Safety study in Multi-center, 6/1 1/03- OM ER 5mg tab 126 56.2 (19-84) Safety 29 Sites opioid-naive open-label 1/21/04 OM ER 10mg 55 M, 71 patients with tab 1 12 W, 14 NW chronic (non OM ER 20mg malignant) pain tab Safety study in Multi-center, 8/22/03? OM ER 5mg tab 221 56.8 (19-85) Safety 40 sites patients with open-label 3/5/05 OM ER 10mg 123 F, 98 cancer or tab 202 W, 19 NW neuropathic pain OM ER 20mg tab OM ER 40mg tab EN3202-021 Safety study in Multi-center, 3/14/01 - Completed 239 48.0 (24-81) Safety 44 sites adults with cancer open-label 7/2/03 studies 016 128 M, 1 i pain or chronic 019; 221 W, 18 NW lower back pain Dose titration in the first week EN3202-022 Safety study in Multi-center, open? 4/19/01? Completed study 24 57.0 (3671) Safety 25 sites cancer patients label 12/31/02 Di 8; Starting 19 F, 5 with moderate to dosage from 21 W, 3 NW severe pain previous controlled-study; Dose titration Source: Supplemental Table 1 on pages 61 to 69 ofthe updated safety report. 4.3 Review Strategy Each ef?cacy study is reviewed in detail in Section 10 and the results of the two ef?cacy studies are discussed together in Section 6. Review of safety is based on the pooled safety data from all phase 2/3 studies grouped into three categories: the Updated Safety database consisted of data from the six studies with new data in the current submission, the 120?Day Safety database consisted of safety data already reviewed in the ?rst review cycle, and the Overall Safety database combining the two. Clinical Review of NBA 21-610 N000 for oxymorphone extended release by Christina Fang 4.4 Data Quality and Integrity Two clinical sites, Site 13 (with enrollment of seven patients in Study 031 and 38 patients in Study 032) and Site 29 (with enrollment of 32 patients in Study 031 and 15 patients in Study 032), were selected for inspection. Inspection of site 13 (Investigator: Martin Hale, MD.) revealed under reporting of VAS scores at ?ve occasions in three subjects in the oxymorphone ER treatment groups as summarized in the table below. Table 4?2 Data Discrepancy at Site 13 of Study 032 Subject Visit Duration of double~ VAS reported by subject as VAS reported by subject blind treatment shown in source document as listed in CRF 2009 10 4 weeks 18 mm . 12 mm 2011 14 12 weeks 67 mm 57 mm 2034 8 2 weeks 15 mm 9 mm 2034 9 3 weeks 15 mm 9 mm 2034 14 12 weeks 15 mm 10 mm Two of these under reported scores for the 12?week assessment were included in the dataset for the primary analyses Inspection of Site 29 (Investigator: Gilbert Podolsky, MD.) revealed only one case of dosing error where the subject was randomized to 25 mg, instead of the correct dose of 30 mg, and received the dose for several days before the error was discovered. 4.5 Compliance with Good Clinical Practices The steps to ensure the accuracy and reliability of data included the selection of quali?ed Investigators and appropriate study centers, review of protocol procedures with the Investigators and associated personnel prior to the start of the study, and periodic monitoring visits by Sponsor personnel. Sponsor personnel reviewed CRFs for accuracy and completeness before, during, and after on-site monitoring visits; any discrepancies were resolved with the Investigator or designee, as appropriate. CRF data were entered into a clinical database by the electronic data capture system. After the resolution of data queries, the database was locked and the data transferred to the statistician. Selective audits of ?ve sites, Sites 12, 23, 29, 31, and 35, with a total enrollment of 106 patients in Study 031 and 62 patients in Study 032, were performed by Endo Pharmaceuticals Inc. to evaluate compliance with Good Clinical Practice guidelines according to the requirements of Endo?s Quality Assurance Audit Plan. The Sponsor provided the audit certi?cate in the submission. 4.6 Financial Disclosures The financial disclosure form signed by the Sponsor certi?ed that no financial arrangements had been made, where outcomes affects compensation, with any Principle Investigator or sub?investi gators involved in the clinical studies, and that these Investigators had no proprietary, significant equity interest, or any significant payments of other sorts as defined in 21 CFR 5420?). Clinical Review of NBA 21-610 N000 for oxymorphone extended release by Christina Fang 5 CLINICAL PHARMACOLOGY 5.1 Pharmacokinetics Refer to clinical pharmacology review. 5.2 Pharmacodynamics Refer to the clinical pharmacology review. 5.3 Exposure-Response Relationships Refer to the clinical pharmacology review. 6 INTEGRATED REVIEW OF EFFICACY 6.1 Indication The proposed indication for the oxymorphone extended-release formulation is for the relief of moderate to severe pain in patients requiring continuous, around-the-clock opioid therapy for an extended period of time (not intended for use as a pm analgesic). 6.2 Methods There are two controlled efficacy studies and EN3202-032 with similar designs. The results of each ofthese two studies are reviewed in detail in Section 10. The efficacy ?ndings from both studies are discussed together in this section as evidence to support efficacy claim. 6.3 General Discussion of Endpoints The efficacy endpoints are listed below: Primary ef?cacy endpoint: - Change in average pain intensity (VAS) (of past 24 hours) from baseline to final visit Secondary ef?cacy endpoints: - Change from baseline to final visit in patient?s global assessment of pain medication - Change from baseline to final visit in physician?s global assessment of pain medication - Change from baseline to final visit in Pain Quality Assessment Scale (Study 032 only) - Evaluation of compliance and study medication usage, time to discontinuation due to lack of efficacy Additional ef?cacy endpoints (in study report and not predefined in the protocol) were: - Average PI by visit Change in average Pl from baseline to final visit adding stabilized dose level as a factor - Percent reduction in average PI - Proportion of responders as defined by 230% reduction in Pl - Time to discontinuation due to all reasons - Proportion discontinued due to all reasons - Average daily dose of rescue medication during the first four days of double?blind treatment i2 Clinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang Days on rescue medication from Day 4 to ?nal visit In the ef?cacy studies of chronic pain indications the most commonly used primary ef?cacy endpoint is the change of pain intensity from baseline to ?nal clinic visit. The ef?cacy endpoints of secondary importance in chronic analgesic studies are functional assessment and patient global evaluation. However, the functional assessment was neither in the study plan nor in any previous communications with the Sponsor. The assessment of the use of rescue medication is also considered important in the ef?cacy evaluation of analgesics for chronic pain indications. The choices of the primary and secondary ef?cacy endpoints were agreed between the Division and the Sponsor through Special Protocol Assessments and meetings. In this reviewer's opinion, the addition of functional assessment of chronic analgesic effects of the study drug would make the set of endpoints more adequate in providing reasonable assessment of clinical benefit. 6.4 Study Design The two studies (031 and 032) had a very similar design. They were studies of open-label, titration to effect followed by randomized withdrawal with a double?blind, placebo-controlled, parallel, multiple-dose design to evaluate chronic low back pain in opioid naive patients (Study 031) and opioid experienced patients (Study 032) conducted at multiple centers in the US The ?ndings from the three chronic pain studies in the original submission have suggested a withdrawal design of having a better chance ofdemonstrating ef?cacy than a straight parallel design. The withdrawal design using an enriched population is one approach to address the problem with missing data due to high dropout rates commonly encountered in a study of a straight parallel design. .It deals with the problem upfront by eliminating the non?reSponders before randomization instead of waiting for them to dropout during the study. However, the interpretation of the results of a randomized withdrawal study should take the following factors into consideration. The study by design may overestimate bene?t by showing greater treatment differences in responders than what would be expected from a more general study population and underestimate toxicity by excluding those who can not tolerate the drug and thus have fewer reports of adverse events. The proportion of non-responders who dropped out during the open?label titration period should be counted in the report of outcome. There were no active controls included in the studies and thus, no intention to assess the relative potency to other approved modified?release opioid analgesics. Having an active control in a setting of individualized dosing might not provide accurate information on relative potency anyway in this reviewer's opinion. In both studies patients on the placebo arm had sudden switch from opioid treatment to placebo, which could potentially lead to opioid withdrawal and make the maintenance of blinding dif?cult. The effort in minimizing the potential bias was by allowing rescue medication, oxymorphone IR 5 mg q4-6h in the first four days. The dose restriction of rescue to :2 doses/day after the ?rst four days would help to reduce confounding effect of rescue. The choice ofthe study pepulation, patients with chronic non-neuropathic low back pain who were opioid nai?ve (Study 031) and opioid experienced (Study 032), was consistent to the proposed indication and considered representative samples ofthe target population. Chronic low back pain has been increasingly used as a reliable model in evaluation of chronic pain. The studies of oxymorphone in both opioid naive and opioid experienced patients could provide useful information on how to use the drug in these different populations and about whether there is a differential response to the study drug with respect to the past opioid experience. Clinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang The choice of individualized dosing regimen was based on the information obtained from previous controlled trials on ?xed dose levels (Studies 015 and 025) individualized dosing regimen (Study 016) in the original submission (refer to the?related reviews for detail), and preceding open-label studies of upward titration (Studies 028 and 029) in the current submission. The data obtained from the studies of individualize dosage provided essential information in formulating the dosing recommendation in the proposed labeling. The 12-week double-blind treatment duration is commonly required length of time to study durability effect of chronic analgesics. Controlled trials of longer duration (>12 weeks) might be able to provide more information on long-term ef?cacy since the development of drug tolerance is a potential problem with chronic use of opioids. 6.5 Ef?cacy Findings The results of the treatment comparison between oxymorphone ER and placebo are summarized for the two studies in the table below. The treatment differences between the oxymorphone ER treatment group and placebo were highly statistically signi?cant for the primary efficacy endpoint and all the secondary and additional endpoints tested. The effect size of the treatment difference was remarkable in almost all ef?cacy endpoints studied. a The mean increase in pain intensity from baseline to ?nal visit (primary endpoint) in the oxymorphone ER group was much smaller than the placebo group (an increase of 1 1 mm with oxymorphone ER treatment versus 28 mm with placebo treatment in Study 031 and 9 mm versus 32 mm in Study 032). The proportion of patients discontinued due to lack of ef?cacy in the oxymorphone ER group was much smaller than the placebo group (10% in the oxymorphone ER group versus 36% in the placebo group in Study 031 and 12% versus 54% in Study 032). - The proportion of patients discontinued due to all reasons in the oxymorphone ER group was also much smaller than the placebo group (33% in the oxymorphone ER group versus 55% in the placebo group in Study 031 and 29% versus 74% in Study 032). The proportion of patients who assessed. their pain medication as good, very good, or excellent at the ?nal visit was much higher in the oxymorphone ER group than placebo group (82% in the oxymorphone ER group versus 40% in the placebo group in Study 031 and 80% versus 33% in Study 032). - Physician's global evaluation of medication showed the same pattern that the proportion of evaluation as good, very good, or excellent for patients on oxymorphone ER was much higher than for patients on placebo (83% in the oxymorphone ER group versus 37% in the placebo group in Study 031 and 85% versus 28% in Study 032). - The change in the total Pain Quality Assessment scores (PQAS-20) from baseline to ?nal visit was much smaller in patients on oxymorphone ER than on placebo (5.5 in the oxymorphone ER group versus 40.5 in the placebo group in Study 032; the parameter was not studied in Study 031). The incorporation of patients? stabilized dose levels in the statistical model ofthe primary analyses did not change the magnitude of the treatment difference between the two study arms in either study. 14 Clinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang The percentage of responders de?ned by a 230% reduction in average pain intensity from screening to the ?nal visit was much higher in the oxymorphone ER group than the placebo group (81% in the oxymorphone ER group versus 52% in the placebo group in Study 031 and 80% versus 35% in Study 032). The increase in the amount of daily rescue usage from Day 1 to Day 4 (with no dose restriction in the ?rst four days) of the double-blind treatment was smaller in the oxymorphone ER group than the placebo group (up to 3.3 mg/day in the oxymorphone ER group versus 10.3 mg/day in the placebo group in Study 031 and up to 6.5 mg/day versus 15.6 mg/day in Study 032). The percentage of days over which the rescue was used during the time interval from Day 5 until the end of the double?blind treatment (rescue restricted to two doses/day) was smaller in the oxymorphone ER group than the placebo group (34 to 42% in the oxymorphone ER group versus 55 to 65% in the placebo group) in Study 031 and was similar for the 2 treatment arms in Study 032. The durability of treatment effects over the entire 12?week double-blind treatment period and associated effect size of treatment differences could also be appreciated visually in the graphs of the pain curves of visit-speci?c average pain intensity, as well as the survival curves for the time to discontinuation due to lack of ef?cacy and due to all reasons (refer to the review of individual studies). Table 6-1 Summary of the Ef?cacy Results of Study 031 and Study 032 Study 031 Study 032 Statistics Oxymorphone Placebo p-value Oxymorphone Placebo p?value ER ER Primary endpoint Baseline pain intensity (VAS) 18.5 (11.22) 19.3 (11.26) -- 23.9 (12.05) 22.2 (10.75) -- mean (STD) Final visit Pl (VAS) 29.9 (26.21) 46.2 (27.03) 31.3 (23.48) 54.5 (28.43) -- mean (STD) Change in P1, 10.6 :t 2.50 27.7 i 2.53 <0.0001 8.7 2.95 31.6 :t 2.93 <0.0001 Secondary endpoints Time to discontinuation due to lack <0.0001 <0.0001 of ef?cacy Number discontinued due to 10/97 34/95 8/69, 37/69 lack of ef?cacy Patient global at ?nal visit <0.0001 <0.0001 Proportion with Zgood evaluation 78/95 34/86 55/69 22/67 Physician global at ?nal visit <0.0001 <0.0001 Proportion with Zgood evaluation 80/96 32/87 58/69 18/65 Change in LSMean 5.5 40.5 <0.0001 Additional endpoints Change in P1 with respect to Close 10.6 27.7 <0.0001 8.6 31.6 <0.0001 level, LSMean Responders: 230% P1 decrease 79/97 (8 47/91 <0.0001 55/69 23/66 <0.0001 Time to discontinuation due to all <0.0007 <0.0001 reasons Proportion discontinued due to all 32/97 52/95 20/69 51/69 reasons - Daily rescue in ?rst 4 days of from 2.2 mg from 2.3 mg from 5.6 mg from 1 1.0 mg double-blind treatment to 3.3 mg to 10.3 mg to 6.5 mg to 15.6 mg of days on rescue medication: 34.4 - 41.2% 55.1 - 65.3% 61 .8?70.7% 62.2?66.7% Day 4 to ?nal visit The discontinuation from the open-label titration treatment was 37% due to all reasons and 26% due to ABS, lack of ef?cacy, and failure to meet titration-stabilization criteria in opioid na?r?ve patients in Study 15 Clinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang 031, and 43% due to all reasons and 31% due to ABS, lack of ef?cacy, and failure to meet titration? stabilization criteria in opioid experienced patients in Study 032. In terms of other factors that might have potential impact on the study results, the treatment groups in the two studies were basically balanced with regard to the demographic characteristics such as age, gender, race, and weight and with regard to the etiology of low back pain and the baseline pain intensity. There were no statistically significant interactions between the average pain intensity and demographic and screening disease characteristics. The distribution of individual closing levels at randomization was similar between the two treatment groups in Opioid naive patients in Study 031 and had more variations in opioid experienced patients in Study 032. The mean daily dose at stabilization was doubled in opioid experience patients (about 80 mg/day) in comparison to opioid naive patients (about 40 mg/day). There were differential dropout rates of about 30% in the oxymorphone ER treatment group (in both studies), and 53% (study 031) to 74% (study 032) in the placebo group, where most drOpouts in the placebo group were due to lack of ef?cacy as expected. The methods used in imputation of missing values were considered acceptable. However, the methods used in the sensitivity analyses were much less conservative than what was used in the original plan. The results of the two sensitivity analyses were consistent with that of the primary analyses. About 30% patients who received double?blind treatment had protocol deviations occurred at some point during the entire study (both open-label phase and double-blind phase), mostly due to off-schedule visit and missing visit. More cases of protocol deviation were reported in patients in the oxymorphone ER group than the placebo group. Thetypes of protocol deviation were not considered as having a noticeable impact on the results of primary and secondary efficacy evaluations. The potential impact could only have been on the change in P1 by visit (one of the additional ef?cacy parameters). The impact of off-schedule visit and missing visit had limited effect because of the relatively long time interval (one to two weeks) between the scheduled clinic visits (unlike hourly assessments) and the imputation rules for missing data speci?ed in the data analysis plan. In Study 031 a modi?ed intention-to?treat (MITT) population was used for ef?cacy analyses. The MITT population excluded 13 patients (eight on oxymorphone ER and ?ve on placebo), who were on oxymorphone ER 10 mg/day (5 mg BID) at the study entry, for failing to stabilize on a minimum of20 mg total daily dose. The results of analyses based on the entire ITT population were consistent to the results obtained from MITT-based analyses. In Study 032 the Sponsor excluded four patients (one on oxymorphone and three on placebo) from the ITT ef?cacy analyses because they did not sign the consent form to have their data eligible for ef?cacy analysis. The exclusion ofthis magnitude was considered unlikely'to change study results and ef?cacy conclusion if data would have been reanalyzed. . The possible explanations for the worsening ofpain from baseline to the ?nal visit while on oxymorphone treatment are tolerance to the opioid treatment in opioid have patients and tolerance to the new type of opioid treatment in opioid experienced patients. The results of subgroup analyses revealed a larger treatment difference in the subgroup stabilized at a higher daily dose than at a lower daily dose in both studies, suggesting possibly a dose response in ef?cacy. Gender difference was opposite in the two studies and thus non conclusive. Because ofthe relatively small sample size for the groups of elderly, non~Caucasian, and patients with severe pain at the screening, the treatment differences with respect to these variables could not be adequately evaluated. 16 Clinical Review of NBA 21-610 N000 for oxymorphone extended release by Christina Fang 6.6 Clinical Microbiology Not applicable. 6.7 Ef?cacy Conclusions Oxymorphone ER studied at individualized dosage in highly selected responders was shown to be ef?cacious in treating chronic low back pain based on the replicable positive ?ndings from the two studies. Appears This Way On Original Clinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang 7 INTEGRATED REVIEW OF SAFETY 7.1 Methods and Findings 7.1.1 Deaths A total of 49 deaths had been reported from all clinical trials in the oxymorphone development program. In addition to the 35 cases of deaths (34 cancer death and one non?cancer death) discussed in the safety review of the original submission, there were 14 cases of deaths reported in patients on oxymorphone treatment in the current submission. Thirteen of the 14 cases in the open-label safety studies (one case in Study 021, one case in Study 028, and 11 cases in Study 029) were cancer deaths most likely attributable to complications associated with the disease progression of end?stage cancer (most cases had only 0.5-2.5 months of oxymorphone treatment preceding deaths) based on the review of narratives. One death was reported from the study of low back pain (Study 032). The cause of death was unlikely to be related to the study drug based on the review of the narrative below. Patient 001-016 was a 73-year-old Caucasian female with chronic low back pain. Her medical history was notable for hypertension, Type II diabetes, diabetic neuropathy, hyperlipidemia, Parkinson?s disease, acid re?ux, cervical spondylosis, osteoarthritis, left shoulder and arm pain, myofascial pain, three myocardial infarctions (1991, 1993, 1995), quadruple bypass surgery (1995), stent placement mastectomy due to breast cancer, cholecystectomy, and cervical and lumbar laminectomy. She was treated with OxyContin 10 mg po tid and oxycodone 5 mg po qid from 2003 to April 3, 2005 for chronic low back pain. Her other concomitant medications included Neurontin, metoprolol, Lasix, Plavix, aspirin, Lantus insulin, l-lumalog, primidone, Aciphex, Phenergan, and Zyrtec. On April 4, 2005, she was enrolled in the study open-label titration period and began taking oxymorphone BR 10 mg po q12h and oxymorphone IR 5 mg PO (limited to 2 doses per day for breakthrough pain). Prior to enrollment in the study, the patient developed a sinus infection on April 1, 2005. She was admitted to the hospital on . and was treated for pneumonia. She stopped taking the study drug on the morning of April 5, 2005. The patient re?started the study drug on April '14, 2005. She was seen by her physician and she was again admitted to the hospital in the evening for pneumonia. The patient was administered cefepime 1000 mg 1V q24h from April 16 24, 2005. Oxymorphone 1R was discontinued on April 21, 2005 and Oxymorphone ER Was discontinued on April 23, 2005 due to missing study visits. The study investigator gave verbal orders to the attending physician to start the patient on OxyContin 10 mg tid and oxycodone 5 mg po qid for pain control. The patient was given OxyContin 10 mg PO QD on April 25 and April 26, 2005. The investigator assessed the event as unlikely related to the study drug but rather to a community?acquired infection. 7.1.2 Other Serious Adverse Events In the Overall database 150 ofoxymorphone ER-treated subjects and 11 of placebo subjects experienced one or more SAE during their participation in the studies. Incidence rates for the oxymorphone ER treatment-emergent, non-fatal SAES are presented by preferred term in the table below, which includes only the more frequently occurring SAEs reported in at least three subjects in the Overall database or at least two subjects in the Update database (safety data submitted only in the current submission). The most frequently occurring individual SAEs were chest pain (11 subjects), pneumonia (1 1 subjects), and vomiting (10 subjects). There were several SAEs reported by more subjects in the Update database in this submission than subjects in the l20?Day Safety database (safety data submitted during the first review cycle), including pneumonia in eight subjects, confusional state in four subjects, cholelithiasis, malignant pleural effusion, pain, chest pain aggravated, drug withdrawal hypoxia, intractable pain, malignant neoplasm progression, mental status changes, obstructive uropathy, renal failure, and respiratory failure, each in two subjects. 18 Clinical Review of NBA 21-610 N000 for oxymorphone extended release by Christina Fang Table 7?1 Number of 'Oxymorphone ER-treated Patients with Non?Fatal SAEs in Descending Frequency in Phase 2/3 ER Trials 120-Day Safety Update Overall Total patients treated with oxymorphone ER (N 1089) (N 972) (N 2011) preferred term Any adverse events 93 (8.5) 62(6.4) 150 (7.5) Chest pain NBC 7 (0.6) 4 (0.4) 11 (0.5) Pneumonia NOS 3 (0.3) 8 (0.8) 11 (0.5) Vomiting NOS 8 (0.7) 3 (0.3) 10 (0.5) Dehydration 5 (0.5) 4 (0.4) 9 (0.4) Nausea 6 (0.6) 4 (0.4) 9 (0.4) NOS 5 (0.5) (0.1) 6 (0.3) Abdominal pain NOS 2 (0.2) 5 (0.2) Back pain 3 (0.3) 1 (0.1) 4 (0.2) Confusional state 0 (0.0) 4 (0.4) 4 (0.2) Drug interaction NOS 4 (0.4) 0 (0.0) 4 (0.2) Hypotension NOS 3 (0.3) (0.1) 4 (0.2) Osteoarthritis aggravated 4 (0.4) 0 (0-0) 4 (0.2) Pulmonary embolism 2 (0.2) 2 (0.2) 4 (0.2) Pyrexia 2 (0.2) 2 (0.2) 4 (0.2) Urinary retention 3 (0.3) 1 (0.1) 4 (0.2) Atrial fibrillation 3 (0.3) 0 (0.0) 3 (0.1) Cholelithiasis (0.1) 2 (0.2) 3 (0.1) Chronic obstructive airways disease exacerbated 2 (0.2) (0.1) 3 (0.1) Concomitant disease progression 2 (0.2) 1 (0.1) 3 (0.1) Depressed level of consciousness 3 (0.3) (0.0) 3 (0.1) Malignant pleural effusion (0.1) 2 (0.2) 3 Pain exacerbated 2 0.2) I (0.1) 3 (0.1) Pain in limb 3 (0.3) 0 (0.0) 3 (0.1) Pain NOS 1 (0.1) 2 (0.2) 3 (0.1) Urinary tract infection NOS 3 (0.3) 0 (0.0) 3 (0.1) Venous thrombosis deep limb 3 (0.3) 0 (0.0) 3 Chest pain aggravated 0 (0.0) 2 (0.2) 2 (0.1) Drug withdrawal 0 (0.0) 2 (0.2) 2 (0.1) Hypoxia (0.0) 2 (0.2) 2 (0.1) Intractable pain 0 (0.0) 2 (0.2) 2 (0.1) Malignant neoplasm progression 0 (0.0) 2 (0.2) 2 (0.1) Mental status changes 0 (0.0) 2 (0.2) 2 (0.1) Obstructive uropathy 0 (0.0) 2 (0.2) 2 (0.1) Renal failure NOS ., 0 (0.0) 2 (0.2) 2 (0.1) Respiratory failure 0 (0.0) 2 (0.2) 2 (0.1) Syncope 0 (0.0) 2 (0.2) 2 (0.1) Subjects who either continued in one ol?the studies after the 120-Day Safety Update cut-offor subjects who experienced ABS not reported in the 120-Day Update (the AEs had to be new or had to have increased in severity from AEs previously reported) or were from studies EN3202-029, EN3202-031 and EN3202-032. Overall column includes all serious AEs, some of which may not be classified as serious before the l20-Day Safety cutoff date but later became serious. Source: Table 16 on pages 171?179. Serious cardiac AEs are selected to be reviewed in detail because ofthe safety issues regarding to the questionable prolongation discussed in the original NDA review. As shown in the table below, six cases of serious cardiac AEs were identi?ed in the Update safety database (one placebo patient in Study 031 and five patients treated with oxymorphone: one in Study 032 and four in Study 029). All six cases resolved with no sequela and were considered unlikely to be related to the study drug in the Investigators? opinion. Three of the six cases had continuous treatment while the ABS occurred, one had an interruption in treatment, and two were discontinued from the treatment due to the AEs. These events did not suggest a 19 Clinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang causal relationship by the study drug based on the review of narratives. As shown in the three examples provided below the events were more likely attributable to the concurrent medical conditions such as a positive medical history of cardiovascular disease. Table 7?2 Number of Oxymorphone ER?treated Patients with Non-Fatal Cardiac Serious AES in Phase 2/3 ER Trials lZO-Day Safety Update Overall Total patients treated with oxymorphone ER (N 1089) (N 972) (N 2011) preferred term (0.1) 0 (0.0) (0.0) Atrial ?brillation 3 (0.3) (0.1) 4 (0.2) Atrial ?utter 0 (0.0) (0.1) (0.0) Cardiac failure congestive 1 (0.1) (0.1) 2 (0.1) Cardio-respiratory arrest 1 (0.1) 0 (0.0) (0.0) Coronary artery occlusion (0.1) (0.0) 1 (0.0) Myocardial infarction 2 (0.2) 0 (0.0) 2 (0.1) Oedemalowerlimb (0.1) 0 (0.0) (0.0) Sinus tachycardia 0 (0.0) (0.1) 1 (0.0) Supraventricular tachycardia (0.1) 0 (0.0) 1 (0.0) Tachycardia NOS 0 (0.0) (0.1) 1 (0.0) Ventricular tachycardia 0 (0.0) (0.1) 1 (0.0) Source: Table 16 on pages 171479. Sample cases of serious cardiac AEs are described below. Atrial ?brillation--Subject EN3202-03 1?017-1006 was a 58?year-old male who received placebo during the study and experienced atrial ?brillation. He had a signi?cant cardiac history for coronary artery disease, prior MI, hypertension, atrial fibrillation, and by-pass surgery. Atrial ?utter?-Subject was an 83~year-old male who received oxymorphone ER 20 mg?40 mg/day for cancer pain from 1/29/04 until 8/5/04. The subject had a history of hypertension, coronary artery disease, and myocardial infarction. Concomitant medications included simvastatin, lisinopril, Aranesp, carboplatin and taxotere. The subject experienced pneumonia and atrial ?utter on a-I-Ijl?and was treated with diltiazem and amiodarone. He continued with study drug after the event without further episodes of atrial flutter. In the opinion ofthe investigator, the event of atrial flutter was related to subject?s pre-existing cardiac disease. The subject had progression of disease and was discharged from the study. Ventricular tachycardia-~Subject was an 82-year-old male who received oxymorphone ER 5 mg BID along with oxymorphone 1R 5 mg PRN for cancer pain from 1/21/04 until 6/28/04. The subject?s medical history included prostate cancer, myocardial infarction and hypertension. The subject experienced ventricular tachycardia and rectal hemorrhage on An ECG showed two types of ventricular tachycardia with different morphologies one of which may be an example ofa different circuit with some variation occurring spontaneously. It was doubted that it was supraventricular with aberrant conduction. 1n the opinion ofthe investigator, the events were unlikely to be related to study medication but more likely to be a consequence ofthe subjectis previous myocardial infarction and depressed ventricular systolic function. The subject continued with study drug post adverse events until 6/28/04 with no further episodes ofthe event. 20 Clinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang 7.1.3 Dropouts and Other Signi?cant Adverse Events 7.1.3.1 Overall pro?le of dropouts The disposition of subjects by treatment groups during the Phase 2/3 oxymorphone ER clinical trials is summarized in the table below. In the Overall database, 1828 subjects received oxymorphone ER and 522 subjects received placebo. The oxymorphone?treated subjects include 807 subjects from all the open- label studies of the entire database, 347 of whom were subjects from the open-label Studies 028 and 029 in the current submission. Discontinuation occurred in 64% subjects treated with oxymorphone ER and 45% of subjects treated with placebo (note: not all trials had a placebo arm). The most frequently reported reasons for discontinuation were A133 for the oxymorphone ER treatment group and lack of ef?cacy for the placebo group. The pattern of early discontinuation was similar between the Update database and the original l20?Day Safety database. The total rate of discontinuation was similar for the oxymorphone ER and placebo treatment groups in the Update database. Only ?ve of the 14 deaths in Update database were counted as the cause of discontinuation because most other cases had major events preceding death being listed as the reasons for dropouts. Table 7-3 Disposition of All Subjects by Treatment Groups in Phase 2/3 Oxymorphone ER Trials Oxymorphone ER Placebo Patient Status 120-Day Safety Update Overall l20-Day Safety Update Overall Treated 1078 (100.0) 750 (100.0) 1828 (100.0) 350 (100.0) 172 (100.0) 522 (100.0) Complete 407 (37.8) 240 (32.0) 667 (36.5) 221 (63.1) 65 (37.8) 286 (54.8) Ongoing 25 (2.3) (0.0) 0 (0.0) Discontinued 646 (59.9) 510 (68.0) 1 161 (63.5) 129 (36.9) 107 (62.2) 236 (45.2) Adverse events 394 (36.5) 242 (32.3) 639 (35.0) 20 (5.7) 16 (9.3) 36 (6.9) Death 0 (0.0) 5 (0.7) 5 (0.3) Lack of efficacy 84 (7.8) 51 (6.8) 135 (7.4) 94 (26.9) 74 (43.0) 168 (32.2) Lost to follow-up 18 (1.7) 23 (3.1) 42 (2.3) 4 (1.1) 2 (1.2) 6 (1.1) Other 150 (13.9) 189 (25.2) 340 (18.6) ll (3.1) 15 (8.7) 26 (5.0) Patients are categorized according to the last treatment received in the last trial in which they participated. The Ongoing patients at 120-day Safety Update either completed or discontinued study at this safety update. This category includes patients who discontinued the study due to Adverse Event reported on either the Study Termination or Adverse Event or Opioid CRF page. Patient was considered as discontinued Due to AE, although the Study Termination Reason was listed as Investigator Opinion. Source: Tables 1 and 2 on pages 16 and 17. 7.1.3.2 Adverse events associated with dropouts Adverse events associated with drOpouts of oxymorphone ER-treated subjects in the Phase 2/3 oxymorphone ER trials are presented by preferred term in the table below, which includes only the more frequently occurring AE-related discontinuation in at least 10 subjects. In the Overall database 32% of oxymorphone ER-treated subjects discontinued due to an AB. The most frequent AE?related discontinuation was due to AEs of CNS and GI systems and general A135. A135 resulting in the discontinuation of 1% or more of oxymorphone [ER-treated subjects included nausea dizziness (excluding vertigo, vomiting somnolence constipation pruritus (2.8), headache (2.1), sweating increased (1.8), sedation (1.6), and fatigue (1.4). The incidence rates for AEs leading to discontinuation were lower in the Update safety database than in the original 120-Day Safety database for the total AEs as well as the individual AEs. [Reviewer?s comments: A possible explanationfor the lower rate of discontinuation due to AEs in the Update database than the original [20-Day database is that all the subjects in the newly completed chronic 21 Clinical Review of NDA 21-610 N000 for oxymorphone extended reiease by Christina Fang studies underwent dose titration to a level of individually tolerated dose. Table 7-4 Number of Oxymorphone ER?treated Patients with AEs Causing Dropouts in Descending Frequency in Phase 2/3 ER trials 120-Day Safety Update Overall Total patients treated with oxymorphone ER (N 1089) (N 972) (N 2011) preferred term Any adverse events 391 (35.9) 251 (25.8) 642 (31.9) Nausea 157 (14.4) 59 (6.1) 216 (10.7) Dizziness (exc vertigo) 93 (8.5) 32 (3.3) 124 (6.2) Vomiting NOS 85 (7.8) 23 (2.4) 108 (5.4) Somnolence 43 (3.9) 29 (3.0) 72 (3.6) Constipation 45 (4.1) 26 (2.7) 71 (3.5) Pruritus NOS 39 (3.6) 18 (1.9) 57 (2.8) Headache NOS 26 (2.4) 16 (1.6) 42 (2.1) Sweating increased 27 (2.5) 10 (1.0) 37 (1.8) Sedation 26 (2.4) 32 (1.6) Fatigue l6(1.5) l2(l.2) 28(l.4) Insomnia NOS 10 (0.9) 9 (0.9) 19 (0.9) Abdominal pain NOS i0 (0.9) 8 (0.8) 18 (0 .9) Diarrhoea NOS 11 (1.0) 6 (0.6) 17 (0.8) Dry mouth 13 (1.2) 4 (0.4) 17 (0.8) Lethargy 8 (0.7) 8 (0.8) 16 (0.8) Concomitant disease progression (0.8) 2 (0.2) 14(0.7) Confusion (1.2) 0 (0.0) 14 (0.7) Anxiety NBC 6 (0.6) 6 (0.6) 12 (0.6) NOS 6 (0.6) 5 (0.5) 11 (0.5) Nervousness 5 (0.5) 6 (0.6) ll (0.5) Appetite decreased NOS 7 (0.6) 3 (0.3) 10 (0.5) Disorientation 7 (0.6) 3 (0.3) 10 (0.5) Subjects who either continued in one ofthe studies after the l20?Day Safety Update cut-off or subjects who experienced AES not reported in the 120-Day Update (the ABS had to be new or had to have increased in severity from AEs previously reported) or were from studies EN3202-029, EN3202-03I and EN3202-032. Overall column includes ail AEs causing study drug discontinuation, some of which may not be included in the other two columns since these ABS existed before the l20-Day Safety Update, but the subject later discontinued without the AE worsening in severity. Source: Table 18 on pages 184 to 196. 7.1.3.3 Other signi?cant adverse events Refer to the original NDA safety review. 7.1.4 Other Search Strategies None. 7.1.5 Common Adverse Events 7.1.5.1 Eliciting adverse events data in the development program In the four new chronic studies (028, 029, 031, and 032) adverse events were monitored and recorded throughout the trials. Safety data were collected at scheduled weekly visit during the titration period (all studies) and at visit during the maintenance period in Studies 028 and 029 and at Day 4 and Weeks during the double-blind treatment in Studies 031 and 032. In Studies 031 22 Ciinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang and032 vital signs were recorded during most of the clinic visits (screening and weekly visit for the open- label phase; baseline and visits at Day 4, Weeks for the double-blind phase) and withdrawal were assessed by using the Adjective Rating Scale for Withdrawal (ARS) and the Clinical Opiate Withdrawal Scale (COWS) at screening, baseline, and visits during the ?rst four weeks of double-blind treatment. Any AE that was ongoing at completion/termination of the study was followed until resolution or up to 30 days. 7.1.5.2 Appropriateness of adverse event-categorization and preferred terms The coding of ABS using preferred terms and categorization of ABS into system organ class by were appropriate. 7.1.5.3 Incidence of common adverse events In the Overall database AEs were reported in 1702 of oxymorphone ER-treated subjects and 296 of placebo subjects in the oxymorphone ER Phase 2/3 studies. The incidence of the most common AEs occurring in 25% subjects in all the Phase 2/3 studies (with or without the control groups) are presented by preferred term (in descending frequency in the Overall database) in the table below. The most? frequently occurring treatment~emergent ABS in patients treated with oxymorphone ER were nausea constipation dizziness (excluding vertigo, vomiting pruritus somnolence headache sweating increased (1 and sedation The incidence rates for the total ABS and the individual AEs were less in the Update safety database than in the original 120-Day Safety database in general. Table7-5 The Most Frequent AEs Reported in All (Controlled and Open?Label) Phase 2/3 ER Trials Treatment group Oxymorphone ER Placebo Database 120-Day Update Overall 120-Day Update Overall Total patients (N 1089) (N 972) (N 2011) (N 350) (N 172) (N 522) preferred term Any adverse experience 976 (89.6) 759 (78.1) 1702 (84.6) 225 (64.3) 71 (41.3) 296 (56.7) Nausea 500 (45.9) 213 (21.9) 71] (35.4) 63 (18.0) 10 (5.8) 73 (14.0) Constipation 435 (39.9) 236 (24.3) 670 (33.3) 63 (18.0) 2 (1.2) 65 (12.5) Dizziness (exc vertigo) 280 (25.7) 111 (11.4) 388 (19.3) 35 (10.0) 3 (1.7) 38 (7.3) Vomiting N08 256 (23.5) 91 (9.4) 346 (17.2) 23 (6.6) 2 (1.2) 25 (4.8) Pruritus N08 264 (24.2) 75 (7.7) 339 (16.9) 42 (12.0) 1 (0.6) 43 (8.2) Somnolence 179 (16.4) 54 (15.8) 333 (16.6) 14 (4.0) 0 (0.0) 14 (2.7) Headache NOS 129 (1 1.8) 104 (10.7) 233 (11.6) 26 (7.4) 2 (1.2) 28 (5.4) Sweating increased 200 (18.4) 34 (3.5) 232 (1 1.5) 37 (10.6) 4 (2.3) 41 (7.9) Sedation 188 (17.3) 14 (1.4) 202 (10.0) 37 (10.6) 0 (0.0) 37 (7.1) Dry mouth 81 (7.4) 44 (4.5) 125 (6.2) 1 (0.3) 2 (1.2) 3 (0.6) Insomnia NBC 62 (5.7) 55 (5.7) 117 (5.8) 8 (2.3) 2 (1.2) 10 (1.9) Diarrhoea NOS 68 (6.2) 47 (4.8) 115 (5.7) 18 (5.1) 8 (4.7) 26 (5.0) Fatigue 60 (5.5) 51 (5.2) 111(5.5) 4 (1.1) 2 (1.2) 6 (1.1) Appetite decreased N08 54 (5.0) 24 (2.5) 78 (3.9) 1 (0.3) i 1 (0.6) 2 (0.4) Subjects who either continued in one ofthe studies after the 120?Day Safety Update cut-off or subjects who experienced AEs not reported in the l20-Day Update (the AEs had to be new or had to have increased in severity from AEs previously reported) or were from studies EN3202-028, EN3202-029, and EN3202-032. Note: AEs included in this table occurred in 25% ofsubjects in either column. This table is sorted by Overall Total frequency in descending order. Source: Table 26 on pages 383 to 422. 23 Clinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang A summary of the most frequent AEs from the placebo-controlled Phase 2/3 oxymorphone ER clinical trials is presented in the table below, which includes ?ve controlled clinical trials (Studies 015, 016 025, 031, and 032) with 1259 patients on oxymorphone ER and 461 on placebo. There were more reports of the total ABS and the most common individual AEs in the oxymorphone ER group than the placebo group for most of the AEs listed in the table. 9 Table 7-6 The Most Frequent AEs Reported in Placebo-Controlled Phase 2/3 ER Trials Oxymorphone ER Placebo Total Total patients treated (N 1259) (N 461) (N 1720) preferred term Any adverse experience 1008 (80.1) 248 (53.8) 1256 (73.0) Nausea 417 (33.1) 61 (13.2) 478 (27.8) Constipation 347 (27.6) 61 (13.2) 408 (23.7) Dizziness (exc vertigo) 224 (17.8) 35 (7.6) 259 (15.1) Pruritus NOS 191 (15.2) 35 (7.6) 226 (13.1) Somnolence 216 (17.2) 10 (2.2) 226 (13.1) Vomiting NOS 196(15.6) 19(4.l) 215 (12.5) Headache NOS 154 (12.2) 26 (5.6) 180 (10.5) Sweating increased 108 (8.6) 40 (8.7) 148 (8.6) Sedation 74 (5.9) 35 (7.6) 109 (6.3) Dry mouth 80 (6.4) 3 (0.7) 83 (4.8) Diarrhoea N08 54 (4.3) 26 (5.6) 80 (4.7) Insomnia NBC 50 (4.0) 9 (2.0) 59 (3.4) Fatigue 49 (3.9) 6 (1.3) 55 (3.2) Abdominal pain NOS 32 (2.5) 7 (1.5) 39 (2.3) Appetite decreased NOS 36 (2.9) 2 (0.4) 38 (2.2) The Placebo Controlled Phase 2/3 ER Trials are: EN3202-015, EN3202-016, EN3202-025, EN3202-031 and EN3202-032. Note: This table is sorted by total frequency in descending order. Source: Table 28 on pages 436 to 448. 7.1.5.4 Common adverse event tables Refer to the tables in the preceding section. 7.1.5.5 Identifying common and drug?related adverse events The incidence rates of the most common drug-related (based on the investigators' opinion) AEs occurring in 25% subjects in the oxymorphone ER Phase 2/3 studies are presented by preferred term (in descending frequency in the Overall database) in the table below. in the Overall database, the most frequently occurring drug?related AEs in patients treated with oxymorphone ER were constipation nausea dizziness (excluding vertigo, pruritus somnolence vomiting sweating increased and sedation The incidence rates for the total ABS and for the most common drug-related individual AEs were much less in the placebo group. There were fewer reports of drug- related AEs for the total ABS and the most common individual AEs in the Update safety database than in the original l20?Day Safety database. Table 7-7 The Most Frequent Drug-Related AEs Reported in All Phase 2/3 ER Trials Treatment group Oxymorphone BR Placebo Datab ase [20?Day Update Overall l20?Day Update Overall Total patients (N 1089) (N 972) (N 2011) (N 350) (N 172) (N 522) preferred term Any adverse experience 870 (79.9) 595 (61.2) 1457 (72.5) 155 (44.3) 37 (21.5) 192 (36.8) Constipation 431 (39.6) 229 (23.6) 659 (32.8) 62 (17.7) 2 (1.2) 64 (12.3) Nausea 464 (42.6) 181 (18.6) 644 (32.0) 60 (17.1) 8 (4.7) 68 (13.0) 24 Clinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang Dizziness (exc vertigo) 270 (24.8) 96 (9.9) 363 (18.1) 34 (9.7) (0.6) 35 (6.7) Pruritus NOS 253 (23.2) 70 (7.2) 323 (16.1) 36 (10.3) 0 (0.0) 36 (6.9) Somnolence 171 (15.7) 150 (15.4) 321 (16.0) 12 (3.4) 0 (0.0) 12 (2.3) Vomiting NOS 230 (21.1) 57 (5.9) 287 (14.3) 19 (5.4) 1 (0.6) 20 (3.8) Sweating increased 193 (17.7) 23 (2.4) 214 (10.6) 36 (10.3) 2 (1.2) 38 (7.3) Sedation 187 (17.2) 14 (1.4) 201 (10.0) 36 (10.3) 0 (0.0) 36 (6.9) Headache NOS 83 (7.6) 63 (6.5) 146 (7.3) 14 (4.0) 1 (0.6) 15 (2.9) Dry mouth 73 (6.7) 40 (4.1) 113 (5.6) 1 (0.3) 2 (1.2) 3 (0.6) Source: Tables 20 and 21 on pages 35 and 36. 7.1.5.6 Additional analyses and explorations Refer to the original NDA safety review. 7.1.6 Less Common Adverse Events The incidence rates for most of the less common AES were either similar or lower in the Update Safety database than the 120?Day safety database. The following AEs had noticeably higher percentage of reports in the Update Safety database than the l20?Day safety database: lethargy, upper abdominal pain, anemia, anorexia, flushing, peripheral edema, pain in extremity, hypotension, malignant neoplasm progression, asthenia, confusion state, pharyngolaryngeal pain, rash, generalized pruritus, gastroesophageal reflux disease, contusion, and neutropenia. 7.1.7 Laboratory Findings There were no laboratory tests conducted in any of the new studies in the current submission. The safety concerns with the decreased WBC count, neutropenia, and LFT elevation had been addressed by the Sponsor in the submission (meeting package) dated February 17, 2004. Of the seven cases of low WBC and neutropenia, one had a less than 10% decrease and normal end?of?study values. The other six cases had abnormally low values due to laboratory sample mishandling. Three had repeated lab tests with the results within normal range and the other three were not available for laboratory retest. Of the five cases with abnormal LFT elevations, four had elevations at a post?operative setting and had concomitant medications and/or concurrent medical conditions known to increase liver enzymes. One had LFT and GGT elevationjudged by the Investigator as unlikely related to the study drug. The Sponsor's explanations were considered acceptable by the Division as documented in the meeting minutes for the meeting held on March 16, 2004. 7.1.7.1 Overview oflaboratory testing in the development program Refer to section 7.1.7 and the safety review ofthe original NDA. 7.1.7.2 Selection of studies and analyses for drug?control comparisons oflaboratory values Refer to section 7.1.7 and the safety review ofthe original NDA. 7.1.7.3 Standard analyses and explorations of laboratory data Refer to section 7.1.7 and the safety review of the original NDA. 7.1.7.4 Additional analyses and explorations Refer to section 7.1.7 and the safety review of the original 25 Clinical Review of NBA 21-610 N000 for oxymorphone extended release by Christina Fang 7.1.7.5 Special assessments Refer to section 7.1.7 and the safety review of the original NDA. 7.1.8 Vital Signs Vital signs were recorded in the Studies 031 and 032. The group mean changes in vital signs are summarized in the table below. Most of the changes in vital signs were ?uctuations around the screening values. There was a trend of small decrease in systolic blood pressure in the ?rst three weeks and small decrease in diastolic blood pressure in the ?rst two weeks of the open?label treatment period in Study 031. Table 7-8 Summary of Mean Change in Vital Signs Treatment period Open-label titration Double-blind treatment Vital signs (VS) Change in VS from screening Change in VS from screening Direction Range Direction Range Study 031 Treatment Systolic BP (mmHg) 1, -3.3 to ?6.7 Oxymorphone -2.6 to 1.5 Placebo -0.4 to -2.0 Diastolic BP (mmHg) 1 ?2.4 to -4.0 Oxymorphone -0.7 to 0.5 Placebo -l.0 to 0.7 Heart rate (beat/min) -0.1 to 1.4 Oxymorphone ?0.2 to 3.1 Placebo ?0.5 to 1.7 ReSpiration rate -0.1 to 0.0 Oxymorphone -0.7 to 0.3 (breath/min) . Placebo ~0.l to 0.4 Temperature (0F) to 0.2 Oxymorphone 0.0 to 0.1 Placebo 0.0 to 0.1 Study 032 Systolic BP (mmHg) 1 ?1.4 to 6.4 Oxymorphone ?2.6 to 2.1 Placebo ?2.1 to 7.1 Diastolic BP (mmHg) 1 ~09 to -2.5 Oxymorphone ~08 to 1.7 Placebo ?0.6 to 3.3 Heart rate (beat/min) 0.3 to 1.7 Oxymorphone -0.4 to 2.3 Placebo 2.7 to 4.8 Respiration rate -0.2 to 0.2 Oxymorphone -0.7 to 0.2 (breath/min) Placebo -0.7 to 0.1 Temperature (OF) 0.0 to 0.1 Oxymorphone -0.2 to 0.0 Placebo ~02 to 0.1 Source: Tables 43 and 44 on pages 794?81 8 of the report of Study 031 and Tables 44 and 45 on pages 740-764 ofthe report of Study 032. 7.1.8.1 Overview ofvital signs testing in the development program Refer to section 7.1.8 and the safety review ofthe original NDA. 7.1.8.2 Selection of studies and analyses for overall drug?control comparisons Refer to section 7.1.8 and the safety review of the original NDA. 7.1.8.3 Standard analyses and explorations of vital signs data Refer to section 7.1.8 and the safety review ofthe original NDA. 26 Clinical Review of NBA 21-610 N000 for oxymorphone extended release by Christina Fang 7.1.8.4 Additional analyses and explorations Refer to section 7.1.8 and the safety review of the original NDA. 7.1.9 Electrocardiograms (ECGS) There were no new ECG data in the current submission. According to the original NDA review there was a safety concern with abnormality identi?ed as interval 2430 msec (males) or 450 msec (females) or a change from pre-dose of 230 msec reported in 1 subjects from three Phase 1 clinical. trials, Studies 001, 002, and 003. Of the 11 subjects identi?ed four had abnormalities at baseline with no worsening on treatment. abnormalities for the other seven subjects are listed in the table below. The abnormalities identi?ed after treatment with oxymorphone 20 mg tablets in the ?rst treatment period in Subjects 1, 2, and 5 resolved upon rechallenge with the same formulation given in the subsequent period. The abnormality identi?ed following the treatment with oxymorphone 10 mg oral solution in the third treatment period in Subject 7 resolved upon rechallenge with the 20 mg tablet treatment given in the fourth period. For Subject 6, the increase in interval from 328 msec to 393 msec with 10 mg oral solution in the ?rst treatment period did not continue upon the rechallenge with the same formulation in the second treatment period, where interval was stabilized from a pre?dose value of 386 msec to a post-dose value of 385 msec. The length of the QT interval ?uctuated in an irregular pattern in the range of 367 msec and 473 msec during the three periods of treatment for Subject 3, for whom the abnormalities identi?ed were a 43 msec increase from a pre-dose value of 404 msec to a post-dose value of 447 msec. Similarly, the length of the interval ?uctuated in an irregular pattern in the range of 382 msec and 413 msec during the four periods of treatment for Subject 4, for whom the abnormalities identified were the 30 msec (borderline) increase from a pre-dose value of382 msec to a post?dose value of412 msec. Also, all ofthe values for the interval at the end of crossover treatments were <430 msec for six of the seven subjects. The data did not suggest a causal relationship between the oxymorphone treatment and prolongation of interval in this reviewer?s opinion. Table 7-9 Summary of Interval Data: and (msec) Subject ID Period Oxymorphone Treatment Predose Postdose Reason for Abnormality EN3202-002- 20 mg tablet 372 476 Postdose Z430 msec 001-001 2 20 mg tablet 439 347 3 10mg oral solution 387 364 2 EN3202-002- 20 mg tablet 358 49] Postdose 2430 msec and increase from 001?006 predose 230 msec 2 20 mg tablet 374 356 3 10 mg oral solution 388 336 3 EN3202-002- 10 mg oral solution 408 367 001-009 2 20 mg tablet 473 419 3 20 mg tablet 404 447 Postdose 2430 msec and increase from predose 230 msec 4 20 mg tablet 386 390 001-002 2 20 mg tablet 382 412 Increase from predose 230 msec 3 10mg oral solution 396 391 4 10 mg oral solution 396 413 5 EN3202-003- 20 mg tablet 421 433 Postdose 2430 msec 001'005 2 10 mg oral solution 470 442 3 20 mg tablet 392 378 4 10 mg oral solution 388 376 27 Clinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang 6 EN3202-003- 10 mg oral solution 328 393 Increase from predose 230 msec 001-012 2 10 mg oral solution 386 385 7 EN3202-003- I 10 mg oral solution 352 352 001-027 2 20 mg tablet 342 356 3 10 mg oral solution 374 426 Increase from predose 230 msec 4 20 mg tablet 362 358 Subject discontinued early. Source: Table 5 on page 3746 ofthe safety update report. The Sponsor performed a retrospective reanalysis of partially recollected data (not collected in the case report forms of the original submission) from Studies 015, 020, and 025 that provided the following ?ndings though the interpretation of data was greatly limited by incomplete data and poor data quality. - The incidences of on-study prolongation judged by a baseline value of 2430/450 msec or an increase from baseline by 230 msec, and on?study value of2500 msec were similar in the oxymorphone, oxycodone, and placebo treatment groups. - Nearly 50% of all subjects had on?study shortening of the interval (ranging from -1 to -270 msec), and about 20% who had baseline prolongation of 2 430/450 msec subsequently had on- study shortening in each treatment group. - Of the four subjects identi?ed as having an on?study prolongation of 2500 msec, one received placebo and three received oxymorphone treatment. Of the oxymorphone?treated patients Subject EN3202-025-01 1?010, had an increase from 392 msec at baseline to 505 msec at the end of the 2? week study with no cardiac AE. Subject had an increase from 438 msec at baseline to 518 msec after four weeks of treatments and recovery to baseline level during the 1-year extended treatment. Subject was hospitalized for an evaluation of myocardial ischemia in Study 017, who was subsequently enrolled in Study 020 and was found to have an increase from 425 msec to 501 msec after 17 days of extended treatment. The data again did not suggest an association between the oxymorphone treatment and prolongation of interval. 7.1.9.1 Overview of ECG testing in the development program, including briefreview of preclinical results Refer to section 7.1.9 and the safety review ofthe original NDA. 7.1.9.2 Selection of studies and analyses for overall drug-control comparisons Refer to section 7.1.9 and the safety review ofthe original NDA. 7.1.9.3 Standard analyses and explorations of ECG data Refer to section 7.1.9 and the safety review ofthe original NDA. 7.1.9.4 Additional analyses and explorations Refer to section 7.1.9 and the safety review ofthe original NDA. 7.1.10 Immunogenicity No data were available. 28 Clinicai Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang 7.1.11 Human Carcinogenicity There were no long-term exposure data for evaluation of human carcinogenicity. 7.1.12 Special Safety Studies Dose dumping of oxymorphone by co-administration with alcohol was suggested by in viva alcohol interaction study but not con?rmed by in vitro dissolution test (refer to the PK review for detail). 7.1.13 Withdrawal Phenomena and/or Abuse Potential The potential for drug abuse and dependence/withdrawal was evaluated in the safety review of the original submission and in the consultation from the Controlled Substances Staff. Two new studies (031 and 032) in the Update Safety database included two scales to measure withdrawal, the Clinical Opiate Withdrawal Scale (COWS) and the Adjective Rating Scale for Withdrawal (ARS). No signs of opiate withdrawal were suggested for the two treatments (oxymorphone and placebo) based on the mean COWS total score and the mean ARS total score at each visit during the double-blind treatment period for the all-treated-patient population. Similar results were seen for the subpopulation consisting of patients who discontinued for lack of ef?cacy, patients who discontinued due to all other reasons, and patients who completed the study. There were eight cases of discontinuation due to opioid withdrawal in the double-blind treatment period, including one oxymorphone-treated patient and two placebo patients in Study 031 and five placebo patients in Study 032. 7.1.14 Human Reproduction and Pregnancy Data One case of on?study pregnancy was reported in Study 032. Patient received 17 days of open-label oxymorphone ER treatment and then was discontinued from the trial due to a positive pregnancy test. Subsequent to her termination from the trial, the Principal investigator notified the Sponsor that the patient had an elective abortion. Refer to pharmacology/toxicology review for the studies of genotoxicity. 7.1.15 Assessment of Effect on Growth There were no pediatric studies. 7.1.16 Overdose Experience There were no reports of overdose in the Update safety database. Refer to the safety review of the original submission for the evaluation of overdose. 7.1.17 Postmarketing Experience The extended-release oral formulation has not been approved for marketing in the U.S. or abroad. Oxymorphone IR approved in 1959 was removed from the market for commercial reasonstablets were removed after seven years of marketing and the 10 mg tablet was removed after i 1 years of marketing. The search for the postmarketing reports on available formulations only revealed 37 unique cases in AERS database in patients treated with intravenous and suppository formulations, which were discussed in the safety review ofthe original submission. 29 Clinical Review of NDA 21-610 N000 for oxymorphone extended release by_Christina Fang 7.2 Adequacy of Patient Exposure and Safety Assessments 7.2.1 Description of Primary Clinical Data Sources (Populations Exposed and Extent of Exposure) Used to Evaluate Safety The primary clinical safety data source was the Update Safety database consisting of data from subjects in the two trials (Studies 021 and 022) ongoing and four trials (Studies 028, 029, 031, and 032) planned at the time of the 120-Day Safety data cutoff date. The safe data from the120?Day Safety database were described and analyzed in the ?rst cycle review of the original NDA submission. The Overall Safety database included the 120?Day Safety database and the Update Safety database. Safety data from the three safety populations (120?Day, Update, and Overall) will be listed side by side, wherever applicable for comparison purpose. 7.2.1 .1 Study type and design/patient enumeration The Update Safety database consisted of data from the following studies: two titration-to?effect, withdrawal design ef?cacy Studies (031 and 032), two open?label safety studies (028 and 029), and two open-label extension studies (021 and 022) which were ongoing at the data cutoff date of the 120-Day Safety Update. As summarized in the table below the data set included a total of 971 patients, 575 patients in the two 16-week ef?cacy studies, 126 patients in the six-month open-label Study 028, 221 patients in the 12?month open-label Study 029, 38 of the 239 patients treated in Study 021, and 12 of the 24 patients treated in Study 022. Table 7-10 Overview of Study Type and Design and Patient Enumeration Protocol Type Design Dates of Dosage of Demography Investigat0r(s Study subj Mean age (range) Gender (M, F) Race (W, NW) Dose ranging Open-label titration 1 1/24/04? Titration to 325 50.1 (18-85) '31 sites efficacy study in followed by 7/18/05 optimal doses of 160 M, 165 opioid-naive patients randomized, OM ER and then 289 W, 36 NW with chronic low double-blind, OM ER or back pain placebo~controlled Placebo EN3202-032 Dose ranging Open?label titration 10/13/04- Titration to 250 49.1 (21-85) 30 sites ef?cacy study in followed by 8/19/05 optimal doses 1 18 M, 132 patients with chronic randomized OM ER 220mg 219 W, 31 NW low back pain withdrawal, and then OM ER double?blind, or Placebo placebo?controlled EN3202-028 Safety study in Multi-center, 6/1 1/03- OM ER 5mg tab 126 56.2 (19?84) 29 sites opioid-naive patients open-label 1/21/04 OM ER 10mg tab 55 M, 71 with chronic pain OM ER 20mg tab 1 12 W, 14 NW EN3202-029 Safety study in Multi-center, 8/22/03~ OM ER 5mg tab 221 56.8 (19?85) 40 sites patients with cancer open-label 3/5/05 OM ER 10mg tab 123 F, 98 or neuropathic pain OM ER 20mg tab 202 W, 19 NW OM ER 40mg tab Safety study in Multi?center, 3/14/01? Completed studies 239 48.0 (24-81) 44 sites adults with cancer open-label 7/2/03 016 019; 128 M, 111 pain or chronic Dose titration in 221 W, 18 NW lower back pain the ?rst week Clinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang Safety study in Multi-center, Open- 4/19/01- Completed study 24 57.0 (36-71) 25 sites cancer patients label 12/31/02 018; 19 F, 5 with moderate to Starting dosage 21 W, 3 NW severe pain from previous controlled-study; Dose titration Source: Supplemental Table on pages 61 to 69 of the updated safety report. 7.2.1.2 Demographics In the Overall database the mean age of the subjects in the Phase 2/3 ER studies was about 55 years. over one quarter of these subjects were elderly (65 years of age or older) and close to 10% were 74 years of age or older. more than one half of the subjects were female and most (close to 90%) of the subjects were Caucasian. The oxymorphone ER and placebo groups had very similar distributions of demographic characteristics. Subjects in the Update database had younger mean age and smaller proportion of the elderly patients, and lighter weight than subjects in the 120-Day safety database. The differences were more noticeable in the placebo group. Also, more male patients received placebo treatment than female patients in the Update database. Table 7?11 Overall Demographics for Subjects in Phase 2/3 Oxymorphone ER Trials Treatment Oxymorphone ER Placebo atab as 120-Day Update Overall 120-Day Update Overall of patients (N 1089) (N 972) (N 2011) (N 350) (N 172) (N 522) Age (yrs) 1089 922 201 350 172 522 Mean 57.4 52.3 55.1 58.9 47.2 55.1 12.76 13.91 13.55 12.35 11.96 13.39 Minyears 731 (67.1) 744 (80.7) 1475 (73.3) 220 (62.9) 159 (92.4) 379 (72.6) 2 65 years 358 (32.9) 178 (19.3) 536 (26.7) 130 (37.1) 13 (7.6) 143 (27.4) 274 years 121 (11.1) 73 (7.9) 194 (9.6) 44 (12.6) 3 (1.7) 47 (9.0) Gender, 11 Female 619 (56.8) 491 (53.3) 1110 (55.2) 206 (58.9) 74 (43.0) 280 (53.6) Male 470 (43.2) 431 (46.7) 901 (44.8) 144 (41.1) 98 (57.0) 242 (46.4) Race, Asian 1 (0.1) 2 (0.2) 3 (0.1) Black 93 (8.5) 72 (7.8) 165 (8.2) 29 (8.3) 10 (5.8) 39 (7.5) Caucasian 968 (88.9) 822 (89.2) 1790 (89.0) 313 (89.4) 155 (90.1) 468 (89.7) Other 27 (2.5) 26 (2.8) 53 (2.6) 8 (2.3) 7 (4.1) 15 (2.9) Height (in) 1075 569 1644 350 172 522 Mean 66.7 67.3 66.9 66.6 67.6 66.9 4.29 4.38 4.33 3.81 4.67 4.14 Min, max 48, 78 52, 91 48,Weight (lbs) 1079 571 1650 350 172 522 Mean 198.6 193.4 196.8 202.4 188.2 197.7 51.64 45.42 49.62 50.14 42.54 48.19 Min, max 84, 427 95, 383 84, 427 110, 425 95, 334 95,425 Source: Tables 9 and 10 on pages 23 and 24 ofthe updated safety report. 7.2.1.3 Extent of exposure (dose/duration) The total exposure in the Phase trials consisted of about 2500 patients exposed to oxymorphone ER, or about 2000 to oxymorphone ER treatment and about 500 to placebo treatment categorized by their last 31 Clinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang dose received in the study, in the Overall database. The information on the cumulative exposure to oxymorphone ER in the ?ve open-label Phase 2/3 ER trials is presented in the table below. Overall, approximately 54% of the subjects (433/804) who participated in the open?label extension trials remained on study for six months, and 33% of the subjects (267/804) remained on study through Month 12. A total of 433 subjects had at least six months of exposure and 267 subjects had at least 12 months of exposure in these open?label studies. Table 7-12 Extent of Cumulative Exposure in Open-Label Phase 2/3 Oxymorphone ER Trials Months on treatment 3 Months I 6 Months I 9 Months I 12 Months I 18 Months I 224 Months Study (duration) Number of subjects on treatment 020 (2-year) 197 (100) 104 (52.8) 94 (47.7) 86 (43.7) 42 (21.3) 26 (13.2) 021 (l-year) 239 (100) 160 (66.9) 126 (52.7) 111 (46.4) 3 (1.3) 0 (0.0) 022 (l-year) 24 (100) 9 (37.5) 7 (29.2) 7 (29.2) 7 (29.2) 1 (4.2) 028 (6-month) 124 (100) 63 (50.8) 0 (0.0) (0.0) 0 (0.0) 0 (0.0) 029 (l-year) 220 (100) 97 (44.1) 74 (33.6) 63 (28.6) 0 (0.0) 0 (0.0) Total 804 (100) 433 (53.9) 301 (37.4) 267 (33.2) 52 (6.5) 27 (3.4) No dosing information was available for three patients (two in Study 028, one in Study 029) and therefore, they were excluded from this table. Note: total exposure to oxymorphone was calculated on a per patient basis regardless ol?dose or changing doses for a patient within a study. Source: Table 6 on page 20 ofthe updated safety report. The information on the extent of exposure to oxymorphone ER treatment by modal daily close category (the most frequent dose the patient took) and duration of treatment (of the most frequent dose) is summarized in the table below. The modal daily doses taken by subjects in the Phase 2/3 oxymorphone ER studies were in the following descending frequency: >10 to 50 mg/day in 57% of subjects 129), >90 mg/day in 18% of subjects >50 to 90 mg/day in 17% of subjects and 510 mg/day in 9% of subjects There were 371 patient?years of exposure to oxymorphone ER in the Phase 2/3 oxymorphone ER studies. The majority ofthese patient-years were in the >10 to 50 mg/day modal daily dose category (163.4 patient- years), the >90 1n g/day category (102.9 patient-years), and the >50 to 90 mg/day category (92.4 patient? years). Table 7-13 Extent of Exposure by Treatment Duration and Modal Daily Dose?Phase 2/3 ER Trials Oxymorphone ER dosage (mg/day) 310 I >10?50 >50?90 >90 1 Total Duration Number ofsubjects on treatment 1 - 3 Days 65 (18.8) 238 (69) 19 (5.5) 23 (6.7) 345 (100) 4 10 Days 67 (10.9) 332 (54.2) 118 (19.3) 95 (15.5) 612 (100) 11 - 17 Days 14 (5.8) 7 195 (81.3) 20 (8.3) 11 (4.6) 240 (100) 18-24 Days 263) 40(65.6) 11 (18) 8(13.1) 61(100) 25 31 Days 2 (6.9) 16 (55.2) 7 (24.1) 4 (13.8) 29 (100) 3 Months 13 (3.8) 173 (50.3) 72 (20.9) 86 (25) 344 (100) 6 Months 16(10.1) 54 (34) 29 (18.2) 60 (37.7) 159 (100) 9 Months 0 (0.0) 15 (24.2) 21 (33.9) 26 (41.9) 62 (100) 12 Months 0 (0.0) 45 (45) 25 (25) 30 (30) 100 (100) 12 Months 0 (0.0) 21 (51.2) 14 (34.1) 6 (14.6) 41 (100) 18 Months 0 (0.0) 7 (58.3) 4 (33.3) 1 (8.3) 12 (100) Total 179(9) 1129 (56.6) 336 (16.9) 349 (17.5) 1993 (100) Patient Years 12.21 163.42 92.35 102.88 370.87 No dosing information is available for two patients in Study 028, one patient in Study 029, nine patients in Study 031, and six patients in Study 032 and they are therefore excluded from this table. Total duration ofexposure for all trials in which a patient participated. 32 Clinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang Patient years: the total number of days in a given modal dosc group divided by 365.25. Note: The information in the table represented the exposure to oxymorphone over time based on the most frequent dose the patient took, not each patient?s total exposure. Source: Table 7 on page 21 of the updated safety report. 7.2.2 Description of Secondary Clinical Data Sources Used to Evaluate Safety Refer to the review of the original NDA. Other studies None. 7.2.2.2 Postmarketing experience None. 7.2.2.3 Literature Literature reports were reviewed in the safety review of the original submission. There were no new reports of ABS according to the Sponsor's most recent review of clinical literature. 7.2.3 Adequacy of Overall Clinical Experience The number of subjects exposed to long-term treatment and the length of exposure appear suf?cient for an adequate overall safety assessment. The level of exposure varied because of the changing dose with individual response and tolerance. The safety concerns with diagnostic laboratory test results and ECG findings from the original review were addressed in the meeting with the Division and in the current submission. The safety findings were in general expected of opioid treatment. The amount of controlled safety data was minimal, which makes the detection of new safety signals almost impossible. 7.2.4 Adequacy of Special Animal and/0r In Vitro Testing Non-clinical studies were considered adequate according to the pharmacology/toxicology review. 7.2.5. Adequacy ofRoutine Clinical Testing Refer to the safety review ofthe original NDA. 7.2.6 Adequacy of Metabolic, Clearance, and Interaction Workup Metabolic, clearance, and interaction workup were considered adequate according to the clinical pharmacology reviewer. 7.2.7 Adequacy of Evaluation for Potential Adverse Events for Any New Drug and Particularly for Drugs in the Class Represented by the New Drug; Recommendations for Further Study The safety database appears to have captured most of the expected opioid-related AEs. 33 Clinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang 7.2.8 Assessment of Quality and Completeness of Data DSI inspection discovered inadequate, inaccurate, and inconsistent data reporting in a few cases mainly involving efficacy data. The entire dataset for the treatment-emergent AEs classi?ed as "treatment~related? was not submitted until the information was requested by the reviewer. 7.2.9 Additional Submissions, Including Safety Update There have been no further safety updates. 7.3 Summary of Selected Drug?Related Adverse Events, Important Limitations of Data, and Conclusions The most frequently occurring AEs in patients treated with oxymorphone ER were nausea constipation dizziness (excluding vertigo, vomiting pruritus somnolence headache sweating increased and sedation in all Phase 2/3 studies (controlled and open-label studies) as well as in the placebo-controlled trials in the Overall database. The same set of was identified as the most common treatment?related ABS and most common ABS causing dropouts in the study. The incidence rates for these AES were much lower in the placebo group. All these are known AEs associated with the use of opioid drugs. The available data did not suggest treatment-related prolongation or cardiac toxicity based on the findings of back to normal range upon rechallenge and at the end of crossover treatment in the majority of cases reviewed and the review of serious cardiac AEs. There was no notable safety signal for treatment-related decrease in WBC and neutrophil count (other than problems with laboratory sample mishandling) or LFT elevation. The major limitation is the lack of controlled data, especially placebo-controlled long?term exposure from studies of a parallel design in the overall safety database. The data from open?label studies using historical controls had limited values in identifying new safety signals. in deciding the causality ofa particular AE there is a tendency to attribute the cause of AE to the treatment if it has a commonly recognized association with the drug class and attribute the cause of AE to the concomitant medications and/or concurrent illness if it is not a known effect of the drug class. Oxymorphone ER has a similar safety profile as the other opioid drugs and is considered reasonably safe to be used for the treatment of chronic pain with a careful initial dose titration to meet individual's acceptance of tolerance in a patient population not at high risks for the treatment. 7.4 General Methodology 7.4.1 Pooling Data across Studies to Estimate and Compare Incidence 7.4.1.1 Pooled data versus individual study data Ofthe six studies in the Update Safety database four studies (028, 029, 021, and 022) had no control groups for comparison and two studies (031 and 032) had a placebo-control group withdrawing from oxymorphone treatment in a study population highly selected for their response and tolerance to the initial oxymorphone treatment. The only noticeable treatment differences between the oxymorphone treatment 34 Clinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang group and placebo group in Study 031 (about 100 subjects per group) were more reports of nausea (12 versus nine), vomiting (eight versus one), constipation (sever versus one), dizziness (?ve versus three), headache (four versus two), and total number of ABS (58% versus 44%) in the active treatment arm than placebo arm. There were only one or two reports of any individual AB in patients receiving continued oxymorphone ER treatment during the double-blind treatment period in Study 032 (about 70 subjects per arm). Therefore, the individual study data had a very limited value for adequate evaluation of safety. 7.4.1.2 Combining data The safety data from all Phase 2/3 trials are combined and grouped by the 120?Day safety database already reviewed before, the Update safety database with the new information, and the Overall safety database combining the two. 7.4.2 Explorations for Predictive Factors 7.4.2.1 Explorations for dose dependency for adverse ?ndings Refer to the safety review of the original NDA. 7.4.2.2 Explorations for time dependency for adverse ?ndings Refer to the safety review of the original NDA. 7.4.2.3 Explorations for drug-demographic interactions Incidence rates for all treatment?emergent AEs occurring in 5% or more of oxymorphone ER?treated subjects in the Phase 2/3 oxymorphone ER trials are presented by age group in the table below. In the Overall database there were larger percentages of reports of nausea, dizziness (exc vertigo), vomiting, somnolence, dry mouth, fatigue, and appetite decreased in the elderly group than in the younger age group. There appeared to be an age?related increase in incidence rates of dizziness and somnolence. On the other hand, the percentages of reports of pruritus, headache, sweating increased, and sedation were lower in the elderly group than in the younger age group. The incidence rates in the Update safety database were much lower than what were reported in the 120?Day safety database for majority of the most commoniy occurring individual AES for all the age groups. Table 7-14 The Most Frequent AEs by Age Group and Preferred Term in Phase 2/3 ER Trials 120-Day Safety Total Update Total Overall Total 1) Age group Age subjects 731 358 121 789 183 74 1475 536 194 Any AB 649 (88.8) 327 (91.3) 110 (90.9) 598 (75.8) 161 (88.0) 65 (87.8) 1219 (82.6) 483 (90.1) 174 (89.7) Nausea 322 (44.0) 178 (49.7) 60 (49.6) 165 (20.9) 48 (26.2) 21 (28.4) 485 (32.9) 226 (42.2) 81 (41.8) Constipation 299 (40.9) 136 (38.0) 42 (34.7) 179 (22.7) 57 (31.1) 26 (35.1) 477 (32.3) 193 (36.0) 68 (35.1) Dizziness (exc' 177 (24.2) 103 (28.8) 42 (34.7) 78 9.9) 33 (18.0) 16 (21.6) 252 (17.1) 136 (25.4) 58 (29.9) vertigo) Vomiting NOS 163 (22.3) 93 (26.0) 32 (26.4) 69 8.7) 22 (12.0) 8 (10.8) 231 (15.7) 115 (21.5) 40 (20.6) Pruritus N08 204 (27.9) 60 (16.8) 22 (18.2) 60 7.6) 15 8.2) 6 8.1) 264 (17.9) 75 (14.0) 28 (14.4) Somnolence 106 (14.5) 73 (20.4) 35 (28.9) 121 (15.3) 33 (18.0) 13 (17.6) 227 (15.4) 106 (19.8) 48 (24.7) HeadacheNOS 100(13.7) 29(8.1) 8(6.6) 92(11.7) 12(6.6) 3(4.1) 192(13.0) 41(7.6) 11 (5.7) Sweating ?f 160 (21.9) 40 (11.2) 10 8.3) 25 3.2) 9 4.9) 3 4.1) 184 (12.5) 48 9.0) 13 (6.7) Sedation 153 (20.9) 35 9.8) 7 5.8) 12 1.5) 2 1.1) 1(i.4) 165 (11.2) 37 6.9) 8 4.1) 35 Clinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang I Dry mouth 46 6.3) 35 9.8) 12 9.9) 37 4.7) 7 3.8) 4 5.4) 83 (5.6) 42 7.8) 16 8.2) Insomnia NBC 43 5.9) 19 5.3) 10 8.3) 46 (5.8) 9 4.9) 5 6.8) 89 6.0) 28 5.2) 15 7.7) Diarrhoea NOS 50 6.8) 18 5.0) 6 5.0) 39 4.9) 8 4.4) 4 5.4) 89 6.0) 26 4.9) 10 5.2) Fatigue 39 5.3) 21 (5.9) 9 7.4) 37 4.7) 14 7.7) 8 (10.8) 76 5.2) 35 (6.5) 17 8.8) Appetite i NOS 29 4.0) 25 7.0) 9 7.4) 17 2.2) 7 3.8) 4 5.4) 46 3.1) 32 6.0) 13 (6.7) Pyrexia 37 5.1) 14 3.9) 4 3.3) 15 1.9) 5 2.7) 2 2.7) 51 (3.5) 19 3.5) 6 3.1) Age of subject from subject youngest recorded age. This table is sorted by Overall Total frequency in descending order. Subjects who either continued in one of the studies after the 120?Day Safety Update cut-off or subjects who experienced AEs not reported in the l20?Day Update (the AEs had to be new or had to have increased in severity from AEs previously reported) or were from studies EN3202-028, EN3202-029, EN3202-031 and EN3202-032. Source: Table 27 on page 45 of the updated safety report. Incidence rates for all treatment-emergent AEs occurring in 5% or more of oxymorphone ER?treated subjects in the Phase 2/3 oxymorphone ER trials are presented by gender in the table below. In the Overall database there were larger percentage of reports of nausea, vomiting, pruritus, headache, and dry mouth, in female than male patients. The pattern in the Update safety database was the same. The incidence rates in the Update safety database were much lower than what were reported in the 120?Day safety database for majority of the most commonly occurring individual AEs for both genders. Table 7-15 The Most Frequent AEs by Gender and Preferred Term in Phase 2/3 ER Trials 120-Day Safety Total Update Total Overall Total 1) Male Female Male Female Male Female Number of Subjects 470 619 453 519 901 1 1 10 Any Adverse Experience 404 86.0) 572 92.4) 344' 75.9) 415 80.0) 736 81.7) 966 87.0) Nausea - 181 (38.5) 319 (51.5) 67 14.8) 146 (28.1) 248 (27.5) 463 41.7) Constipation 197 (41.9) 238 (38.4) 112 24.7) 124 23.9) 309 (34.3) 361 (32.5) Dizziness (exc vertigo) 116 24.7) 164 (26.5) 51 11.3) 60 11.6) 166 18.4) 222 20.0) Vomiting NOS 78 16.6) 178 (28.8) 34 7.5) 57( 11.0) 112( 12.4) 234 21.1) Pruritus NOS 106 22.6) 158 25.5) 23 5.1) 52 10.0) 129 14.3) 210 18.9) Somnolence 69 14.7) 110( 17.8) 68 15.0) 86 16.6) 137 15.2) 196 17.7) Headache NOS 36 7.7) 93 15.0) 39 8.6) 65 12.5) 75 8.3) 158( 14.2) Sweating increased 87 18.5) 113 18.3) 19 (4.2) 15 2.9) 106 11.8) 126 11.4) Sedation 85 18.1) 103 16.6) 3 (0.7) 11 (2.1) 88 9.8) 114( 10.3) Dry mouth 24 5.1) 57 9.2) 16 3.5) 28 5.4) 40 4.4) i 85 7.7) Insomnia NBC 23 (4.9) 39 6.3) 30 6.6) 25 (4.8) 53 5.9) 64 (5.8) Diarrhoea NOS 24 5.1) 44 7.1) 21 (4.6) 26 5.0) 45 5.0) 70 6.3) Fatigue 28 6.0) 32 5.2) 20 4.4) 31 (6.0) 48 5.3) 63 (5.7) Appetite decreased NOS 22 4.7) 32 5.2) 13 (2.9) 11 2.1) 35 3.9) 43 3.9) Subjects who either continued in one of the studies after the l20-Day Safety Update cut-offor subjects who experienced AEs not reported in the 120-Day Update (the AEs had to be new or had to have increased in severity from ABS previously reported) or were from studies EN3202-028, EN3202-029, EN3202-031 and EN3202-032. Note: This table is sorted by Overall Total frequency in descending order. Source: Table 29 on page 49 of the updated safety report. In the Overall database only 165 ofthe 201 subjects were African American and 56 onOl 1 were classified as others. The safety data were mainly from the Caucasian population, which accounted for about 90% of the study population. Therefore, the discussion about drug?demographic interactions with respect to race is limited because of the dramatic imbalance in subpopulation size. Nevertheless, the incidence rates for all treatment-emergent AEs occurring in 5% or more of oxymorphone ER?treated subjects in the Phase 2/3 oxymorphone ER trials are presented by racial group in the table below. In the Overall database the incidence rates for dizziness (exc vertigo), somnolence, and sedation appeared to be higher in Caucasian study population than in African American population. The incidence rates in the Update safety database were much lower than what were reported in the l20-Day safety database for some of most commonly occurring individual AEs for all the racial groups. 36 Clinical Review of NDA 21-610 N000 for oxymorphone extended release by Christina Fang Table 7-16 The Most Frequent AEs by Racial Group and Preferred Term in Phase 2/3 ER Trials 120-Day Safety Total Update Total Overall Total 1) Caucasian Black Other Caucasian Black Other Caucasian Black Other ofsubjects 968 93 28 866 77 29 1790 165 56 Any AB 873 (90.2) 81 (87.1) 22 (78.6) 683 (78.9) 53 (68.8) 23 (79.3) 1526 (85.3) 131 (79.4) 45 (80.4) Nausea 447 (46.2) 42 (45.2) 11 (39.3) 191 (22.1) 18 (23.4) 4 (13.8) 637 (35.6) 59 (35.8) 15 (26.8) Constipation 397 (41.0) 27 (29.0) 11 (39.3) 221 (25.5) 11 (14.3) 4 (13.8) 617 (34.5) 38 (23.0) 15 (26.8) Dizziness (exc 253 (26.1) 18 (19.4) 9 (32.1) 105 (12.1) 3 3.9) 3 (10.3) 355 (19.8) 21 (12.7) 12 (21.4) Vertigo) Vomiting NOS 227 (23.5) 26 (28.0) 3 (10.7) 79 9.1) 9 (1 1.7) 3 (10.3) 306 (17.1) 34 (20.6) 6 (10.7) Pruritus NOS 237 (24.5) 22 (23.7) 5 (17.9) 64 7.4) 9 (1 1.7) 2 6.9) 301 (16.8) 31 (18.8) 7 (12.5) Somnolence 165 (17.0) 11 (11.8) 3 (10.7) 139 (16.1) 9 (1 1.7) 6 (20.7) 304 (17.0) 20 (12.1) 9 (16.1) HeadacheNOS 115(119) 8(8.6) 6(21.4) 96(11.1) 7(9.1) l(3.4) 211(11.8) 15(9.1) 7(12.5) Sweating 183 (18.9) 11 (11.8) 6 (21.4) 30 (3.5) 3 (3.9) 1 (3.4) 211 (11.8) 14 8.5) 7 (12.5) Sedation 176 (18.2) 9 9.7) 3 (10.7) 14 1.6) 0 0.0) 0 0.0) 190 (10.6) 9 5.5) 3 (5.4) Dry mouth 74 7.6) 6 6.5) 1(3.6) 39 (4.5) 4 5.2) 1 (3.4) 113 (6.3) 10 6.1) 2 3.6) Insomnia NBC 55 5.7) 4 4.3) 3 (10.7) 50 (5.8) 4 5.2) 1 (3.4) 105 (5.9) 8 (4.8) 4 7.1) Diarrhoea NOS 61 (6.3) 4 4.3) 3 (10.7) 42 (4.8) 2 2.6) 3 (10.3) 103 (5.8) 6 3.6) 6 (10.7) Fatigue 53 5.5) 6 6.5) 1 (3.6) 46 5.3) 5 6.5) 0 0.0) 99 5.5) 11 (6.7) I (1.8) Appetite 1 NOS 48 5.0) 4 4.3) 2 7.1) 22 2.5) 1 1.3) 1 3.4) 70 3.9) 5 3.0) 3 (5.4) Pyrexia 43 (4.4) 7 7.5) 1 (3.6) 19 2.2) 0 0.0) 1 3.4) 61 (3.4) 7 4.2) 2 3.6) Arthralgia 36 (3.7) 4 4.3) 2 7.1) 22 2.5) 1 (1.3) 0 0.0) 58 3.2) 5 3.0) 2 3.6) Abdominal pain 33 3.4) 4 4.3) l(3.6) 18 2.1) 5 6.5) 0 0.0) 51 (2.8) I 9 5.5) 1.8) NOS In?uenza 31 (3.2) 4 4.3) 2 7.1) 18 (2.1) 0 0.0) 0 0.0) 49 2.7) 4 2.4) 2 3.6) Dyspepsia 21 (2.2) 3 3.2) 2 7.1) 26 3.0) 1 1.3) 1 (3.4) 47(2.6) 4 2.4) 3 (5.4) Subjects who either continued in one ofthe studies after the 120-Day Safety Update cut-off or subjects who experienced A138 not reported in the 120-Day Update (the AEs had to be new or had to have increased in severity from AEs previously reported) or were from studies EN3202-028, EN3202-029, and EN3202-032. Note: This table is sorted by Overali Total frequency in descending order. Source: Table 31 on page 53 ofthe updated safety report. 7.4.2.4 Explorations for drug-disease interactions Refer to the safety review ofthe original NDA. 7.4.2.5 Explorations for drug~drug interactions Refer to the safety review ofthe original NDA. 7.4.3 Causality Determination Most of the commoniy occurring AEs were treatment-related AEs known to be associated with the use of opioid drugs. The data have limited values in identifying new safety signals due to the lack ofcontrolled data of long?term exposure. Appears This Way On Original Clinical Review of NBA 21-610 N000 for oxymorphone extended release by Christina Fang 8 ADDITIONAL CLINICAL ISSUES 8.1 Dosing Regimen and Administration The ef?cacy and safety ?ndings from the trials using titration to effect design support the proposed dosing regimen for opioid na'1've patients, to initiate at 5 mg ql2h followed by individualized titration at increments of 5?10 mg q12h every 3-7 days, to a level that provides adequate analgesia and minimizes side effects. The proposed conversion from other opioids in the dosing instruction for opioid experienced patients is not supported by data because the studies of relative potency between oxymorphone ER and other opioids had not provided useful information about the relative potency. 8.2 Drug-Drug Interactions Refer to the safety review of the original NDA. 8.3 Special Populations Elderly patients are at higher risk for oxymorphone treatment-related adverse events due to a higher level of systemic exposure (about a 40% increase in total and maximum drug levels in comparison to younger subjects). Elderly patients have been shown to have increased risks to oxymorphone-induced respiratory/CNS depression at higher starting doses in a post-operative setting based on the results of studies in the original submission. There was also an age?related increase in incidence rates of dizziness and somnolence. Therefore, there should be a lower starting doses, slower titration, and closer monitoring for the elderly patients. 8.4 Pediatrics p-u-n 8.5 Advisory Committee Meeting This 505(b)(2) application is not planned to be discussed at an Advisory Committee meeting. 8.6 Literature Review Literature reports were reviewed in the safety review ofthe original submission. There were no new reports of ABS according to the Sponsor's most recent review ofclinical literature. 8.7 Postmarketing Risk Management Plan The Sponsor?s proposed post?marketing Risk Management Plan for oxymorphone products has been reviewed in detail in the consultation from the Office of Drug Safety (ODS). There have been several 38 Clinical Review of NBA 21-610 N000 for oxymorphone extended release by Christina Fang communications with the Sponsor to address the concerns. To date most of the issues have been resolved. The additional recommendations from ODS include further education in patients and HCPs about the appropriate use of oxymorphone ER, the risk associated With inappropriate use, and the differences between the IR and ER formulations, strong warning against the use with alcohol and warning about starting at higher doses of oxymorphone ER in opioid naive patients in the labeling, inclusion of any pediatric (age 16 years or younger) use or medication error in the 15-day Alert Report, and a number of other comments about the Sponsor's pharmacovigilance and educational plans in RMP. The consultation from the Controlled Substance Staff provided additional recommendations regarding the details and the use of data analysis instrument in risk management, and a number of labeling revisions. 8.8 Other Relevant Materials The use of the proprietary name, OPANATM ER is considered acceptable by recommendations from the Division of Medication Errors and Technical Support. 9 OVERALL ASSESSMENT 9.1 Conclusions Oxymorphone ER studied at individualized dosage in highly selected responders was shown to be ef?cacious in treating chronic low back pain based on the replicable positive ?ndings from the studies of chronic low back pain. Oxymorphone ER has a similar safety pro?le as the other opioid drugs and is considered reasonably safe to be used with a careful initial dose titration to meet individual's acceptance of tolerance, for the treatment of chronic pain in both opioid naive and opioid experienced populations. 9.2 Recommendation on Regulatory Action Oxymorphone ER is recommended for market approval. 9.3 Recommendation on Postmarketing Actions None. 9.3.1 Risk Management Activity The Sponsor?s proposed post?marketing Risk Management Plan (RMP) for oxymorphone products is considered acceptable in general. The additional recommendations from the Office of Drug Safety and the Controlled Substance Staff will be forwarded to the Sponsor. 9.3.2 Required Phase 4 Commitments None. 9.3.3 Other Phase 4 Requests There should be further studies of relative potency in comparison to the other commonly used opioids to well inform the labeling. 39 Clinical Review 21?610 N000 for oxymorphone extended release by Christina Fang 9.4 Labeling Review The labeling will be reviewed separately. 9.5 Comments to Applicant Further studies of relative potency in comparison to the other commonly used opioids should be conducted. Risk Management Plan should incorporate all the recommendations from the Of?ce of Drug Safety and the Controlled Substance Staff. Appears This Way 0n Original 40 Clinical Review of NDA 21?610 N000 for oxymorphone extended release by Christina Fang 10 APPENDICES 10.1 Review of Individual Study Reports 10.2 Study 031 Protocol Study EN3202-031 was planned as a titration-to-effect, randomized withdrawal, double-blind, placebo? controlled, parallel, multiple-dose, dose range study of chronic low back pain (LBP) in opioid naive patients. The study was planned to have two periods: an open-label titration period and a double-blind, placebo?controlled treatment period. Eligible subjects were going to be patients with moderate to severe chronic (23 months duration), non? neuropathic low back pain (LBP), with suboptimal response to non?opioid analgesic regimens, and without prior use of around?the?clock opioid (opioid naive). Following screening assessments all eligible patients were planned to be started on open-label treatments with two days (Day 1 and 2) of oxymOrphone ER 5 mg q12 hours followed by upward titration at increments of 5 to] 0 mg q12 hours every three to seven days until reaching stabilization. To be eligible to enter the double-blind, placebo?controlled treatment, patients were going to be required to be stabilized within four weeks of open?label treatment. Stabilization was going to be defined as reaching a tolerated dose that would provide adequate pain relief, maintaining an average pain intensity score of :40 mm by visual analogue scale (VAS) for three of ?ve consecutive days immediately prior to randomization, and reaching a minimum dose of oxymorphone ER of 10 mg q12 hours (20 mg daily) prior to randomization. No rescue medication was going to be allowed during the titration period. Patients requiring rescue were to be discontinued. Anti-constipation medication was planned to be avaiiable throughout the study. Patients eligible for double-blind treatment were planned to be randomized to either continue on the stabilized dose of oxymorphone ER or replace oxymorphone ER treatment with placebo. During the 12? week double-blind treatment oxymorphone immediate release (IR) was going to be allowed as rescue pain medication for breakthrough pain and as an aid in tapering placebo patients to prevent opioid withdrawal. The use of rescue medication was planned as oxymorphone IR 5 mg q4-6h as needed during the first four days of double?blind treatment and at most twice daily thereafter. Patients who developed intolerance or inadequate pain control to their established dose of study drug were going to be terminated from the study. Patients were to be instructed to keep a daily diary record of the total oxymorphone ER (or placebo) dose, as well as any oxymorphone IR (rescue medication) dose. During the double-blind treatment period, patients were going to return to the site for safety and efficacy assessments at Day 4 and Weeks 1, 2, 3, 4, 6, 8, 10, and 12(i3 days). Efficacy was planned to be measured by the change in average pain intensity from baseline (last VAS pain score before randomization) to final study visit as the primary endpoint and by time to early discontinuation due to lack ofefficacy and patient/physician global satisfaction with study medication as the secondary endpoints. Safety and tolerability were planned to be evaluated by adverse events (AE5), vital signs, discontinuations due to drug-related AEs (tolerability), and investigator- and patient-rated signs and of opioid withdrawal. 4