`RESEARCH
`
`
`
`APPLICATION NUMBER:
`
`22-272Orig1s014
`
`
`OTHER REVIEW(S)
`
`
`
`Labeling Review for Regulatory Action
`
`electronic st u u
`
`Sharon Hertz, M.D.
`From
`
`Subject
`Deputy Division Director Smnmary Review
`NDA#/ Su. nlement #
`022272/S-014
`
`
`
`A licant Name
`
`Date of Submission
`
`PDUFA Goal Date
`
`Proprietary Name /
`Established (USAN) Name
`Dosa_e Forms / Stren_ h
`
`Proposed Indication(s)
`
`Purdue Pharma
`
`Se tember 12, 2012
`
`A .ril 14, 2012
`
`OxyContin (Oxycodone Hydrochloride Controlled-
`Release) Tablets
`Tablet
`
`For the management of moderate to severe pain when a
`continuous, around-the-clock opioid analgesic is
`needed for an extended 1 -riod of time.
`
`Introduction
`
`This supplemental application proposes the addition of labeling language describing the results
`of pre- and post-marketing data from in vitro and in vivo abuse potential studies to the DRUG
`ABUSE AND DEPENDENCE section,
`"9‘"
`
`The material intended to support of the
`proposed labeling changes were submitted in the NDA and to IND 029038. A review of the
`studies intended to support the labeling changes was conducted by Dr. Pamela Horn as
`appended to the review by Dr. Bob Rappaport, dated February 10, 2013. Dr. Horn reviewed
`the following material:
`
`IND 29038
`
`0 Abuse liability and pharmacokinetic studies
`0 css reviews of studies CTR-1016, 1018, 1019, 1021, and 1022, DARRTS,
`9/21/12
`
`0 Statistical review of study CTR-1018, DARRTS, 8/20/12
`0 Clinical Pharmacology review of studies CTR-1016, 1018, and 1021,
`DARRTS, 9/20/ 12
`
`0 Study reports CTR-1016, 1018, 1019, 1021, and 1022, submitted to IND
`9/17/10
`
`0 Pharmacy instruction manuals CTR-1018 and 1021
`
`NDA 22272:
`
`0
`
`In vitro studies: CSS review of the following documents, DARRTS, 9/4/09
`0 Comprehensive In Vitro Testing of the Controlled-Release Properties ofNew
`OCR Tablets After Physical and Chemical Manipulation — Summary Report
`
`NDA 22-272 Dep Dir labeling memo OxyContin SOl4.doc
`Page 1 of 11
`
`Reference ID: 3294221
`
`
`
`
`
`o Comprehensive in Vitro Testing of the Controlled-Release Properties of New
`OCR Tablets After Physical and Chemical Manipulation – Complete Dataset
`Appendix
`o Evaluation of the Resistance to Physical and Chemical Manipulation of
`Oxycodone HCl (10, 15, 20, 30, 40, 60 and 80 mg) TR Tablets Compared to
`Currently Marketed OxyContin (10, 15, 20, 30, 40, 60 and 80 mg)
`o Protocol for Creating Particle Size Fractions by Crushing and Milling Extended
`Release Oxycodone HCl Tablets - OxTR In Vitro Testing Plan - Experiments
`2a, b and 5a, b, c
`o Protocol TTP-PMP-M0043.00 - "Simple Extraction: pH-Dependent API
`Release Study for Extended Release Oxycodone HCl Tablets"
`o Protocol for Smoking (Inhalation) Testing of Physically Manipulated Extended
`Release Tablets Containing Oxycodone HCl
`o Validation Protocol for Simple Extraction Testing of Physically Manipulated
`Extended Release Tablets Containing Oxycodone HCl
`
`
`
`The applicant proposed the changes to Section 9 Drug Abuse and Dependence as shown in the
`following table. The labeling language is in the left column and the support for the language is
`in the right column.
`
`
`NDA 22-272 Dep Dir labeling memo OxyContin S014.doc
`
`Page 2 of 11
`
`Reference ID: 3294221
`
`(b) (4)
`
`
`
`
`
`NDA 22-272 Dep Dir labeling memo OxyContin S014.doc
`
`Page 3 of 11
`
`Reference ID: 3294221
`
`(b) (4)
`
`
`
`
`
`
`
`
`
`In deciding what data to include in the labeling, consideration was given to the intended target
`audience. In this case, in addition to the prescriber, there will be many stakeholders interested
`in understanding the potential effects of the abuse-deterrent changes to the formulation of
`OxyContin on the different routes of abuse. Therefore, the study results have been provided in
`table format with median and mean values, as a figure representing the continuous responder
`function for drug liking, and as text describing key cutoff points (e.g. no reduction, some
`reduction, 30% reduction, and 50% reduction in drug liking) in the responder analysis.
`
`The following table represents the final agreed-upon language for the package insert.
`
`NDA 22-272 Dep Dir labeling memo OxyContin S014.doc
`
`Page 4 of 11
`
`Reference ID: 3294221
`
`(b) (4)
`
`
`
`
`
`
`Risks Specific to Abuse of OxyContin
`
`OxyContin is for oral use only. Abuse of OxyContin poses a risk of overdose and death. Abuse
`may occur by taking intact tablets in quantities greater than prescribed or without legitimate
`purpose, by crushing and chewing or snorting the crushed formulation, or by injecting a
`solution made from the crushed formulation. The risk of overdose or death is increased with
`concurrent use of OxyContin with alcohol and other central nervous system depressants.
`Taking cut, broken, chewed, crushed, or dissolved OxyContin enhances drug release and
`increases the risk of overdose and death.
`
`With parenteral abuse, the inactive ingredients in OxyContin can result in death, local tissue
`necrosis, infection, pulmonary granulomas, and increased risk of endocarditis and valvular
`heart injury. Parenteral drug abuse is commonly associated with transmission of infectious
`diseases, such as hepatitis and HIV.
`
`
`Abuse Deterrence Studies
`
`
`
`OxyContin is formulated with inactive ingredients intended to make the tablet more difficult to
`manipulate for misuse and abuse. For the purposes of describing the results of studies of the
`abuse-deterrent characteristics of OxyContin resulting from a change in formulation, in this
`section, the original formulation of OxyContin, which is no longer marketed, will be referred to
`as “original OxyContin” and the reformulated, currently marketed product will be referred to as
`OxyContin.
`In Vitro Testing
`In vitro physical and chemical tablet manipulation studies were performed to evaluate the
`success of different extraction methods in defeating the extended-release formulation. Results
`support that, relative to original OxyContin, there is an increase in the ability of OxyContin to
`resist crushing, breaking, and dissolution using a variety of tools and solvents. The results of
`
`This first paragraph is not truly
`specific to OxyContin but is
`important to underscore in the
`context of discussing possible
`abuse.
`
`
`
`
`
`The inactive ingredients in this
`formulation, if injected, would risk
`the types of injury listed. The risk
`of endocarditis, valvular injury and
`infection is general
`
`This is a new title for this section,
`added to direct the reader to the
`specific studies of abuse-deterrent
`characteristics.
`The first sentence represents a
`general statement that is included
`because later data support the
`relevance.
`
`This is a restatement of the proposed
`first paragraph from the sponsor
`based on the following material:
`
`March 30, 2009 Complete Response:
`Comprehensive in vitro Testing –
`
`NDA 22-272 Dep Dir labeling memo OxyContin S014.doc
`
`
`
`Page 5 of 11
`
`Reference ID: 3294221
`
`
`
`
`
`these studies also support this finding for OxyContin relative to an immediate-release
`oxycodone. When subjected to an aqueous environment, OxyContin gradually forms a viscous
`hydrogel (i.e., a gelatinous mass) that resists passage through a needle.
`
`Summary Report: [3.2.P.2
`Pharmaceutical Development,
`Sequence 0020]
`Experiment 1: Mechanical
`Fractionation of Tablets
`Experiment 2: Extraction in Water
`Experiment 3: Dissolution in
`Ethanol
`Experiment 4: Extraction in
`Advanced Solvents
`Experiment 5: Syringability,
`Injectability and Vaporization for
`Inhalation
`
`Clinical Study
`In a randomized, double-blind, placebo-controlled, 5-period crossover pharmacodynamic
`study, 30 recreational opioid users with a history of intranasal drug abuse received intranasally
`administered active and placebo drug treatments. The five treatment arms were finely crushed
`OxyContin 30 mg tablets, coarsely crushed OxyContin 30 mg tablets, finely crushed original
`OxyContin 30 mg tablets, powdered oxycodone HCl 30 mg, and placebo. Data for finely
`crushed OxyContin, finely crushed original OxyContin, and powdered oxycodone HCl are
`described below. Drug-liking was measured on a bipolar drug liking scale of 0 to 100 where 50
`represents a neutral response of neither liking nor disliking, 0 represents maximum disliking
`and 100 represents maximum liking. Response to whether the subject would take the study
`drug again was also measured on a bipolar scale of 0 to 100 where 50 represents a neutral
`response, 0 represents the strongest negative response (‘definitely would not take drug again’)
`
`This section comes from the
`following study:
`September 16, 2010 Information
`Amendment [Sequence 0025, Serial
`Number 0559 to oxycodone IND
`029038]: OTR 1018 entitled : “A
`Single-Center, Double-Blind Study
`in Recreational Opioid Users to
`Evaluate the Abuse Potential,
`Pharmacokinetics, and Safety of
`Crushed and Intranasally
`Administered Oxycodone HCl
`Tamper Resistant Tablets”.
`
`NDA 22-272 Dep Dir labeling memo OxyContin S014.doc
`
`
`
`Page 6 of 11
`
`Reference ID: 3294221
`
`(b) (4)
`
`
`
`
`
`and 100 represents the strongest positive response (‘definitely would to take drug again’).
`
`Twenty-seven of the subjects completed the study. Incomplete dosing due to granules falling
`from the subjects’ nostrils occurred in
` (n=10) of subjects with finely crushed OxyContin,
`compared with
`(n=2) of subjects with finely crushed original OxyContin and no subjects
`with powdered oxycodone HCl.
`
`The intranasal administration of finely crushed OxyContin was associated with a numerically
`lower mean and median drug liking score and a lower mean and median score for take drug
`again, compared to finely crushed original OxyContin or powdered oxycodone HCl as
`summarized in Table 2.
`
`Table 2. Summary of Maximum Drug Liking (Emax) Data Following Intranasal
`Administration
`
`
`OxyContin (finely
`crushed)
`
`Oxycodone HCl
`(powdered)
`
`VAS Scale
`(100 mm)*
`
`
`
`Original
`OxyContin (finely
`crushed)
`89.3 (3.1)
`94.0 (2.7)
`80.4 (3.9)
`Mean (SE)
`Drug Liking
`100 (50-100)
`100 (51-100)
`88 (36-100)
`Median (Range)
`
`86.6 (4.4)
`89.6 (3.9)
`64.0 (7.1)
`Take Drug Again Mean (SE)
`100 (0-100)
`100 (20-100)
`78 (0-100)
`
`Median (Range)
`* Bipolar scales (0 = maximum negative response, 50 = neutral response, 100 = maximum positive response)
`
`Figure 1 demonstrates a comparison of drug liking for finely crushed OxyContin compared to
`powdered oxycodone HCl in subjects who received both treatments. The Y-axis represents the
`percent of subjects attaining a percent reduction in drug liking for OxyContin vs. oxycodone
`HCl powder greater than or equal to the value on the X-axis. Approximately 44% (n = 12) had
`no reduction in liking with OxyContin relative to oxycodone HCl. Approximately 56% (n =
`15) of subjects had some reduction in drug liking with OxyContin relative to oxycodone HCl.
`Thirty-three percent (n = 9) of subjects had a reduction of at least 30% in drug liking with
`
`
`DAAAP and CSS reviewed the
`study protocol and results and
`concluded that the study was
`properly designed and conducted.
`The data describe responses that
`support the conclusions described
`below.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`The figure, a continuous responder
`function, represents an additional
`analysis by FDA. This figure
`permits the reader to see that
`relative to plain oxycodone powder,
`what percent of subjects
`experienced a reduction in drug
`liking over the full range from no
`reduction to total reduction. Any
`
`NDA 22-272 Dep Dir labeling memo OxyContin S014.doc
`
`
`
`Page 7 of 11
`
`Reference ID: 3294221
`
`(b) (4)
`
`(b) (4)
`
`
`
`
`
`OxyContin compared to oxycodone HCl, and approximately 22% (n= 6) of subjects had a
`reduction of at least 50% in drug liking with OxyContin compared to oxycodone HCl.
`
`Figure 1: Percent Reduction Profiles for Emax of Drug Liking VAS for OxyContin vs.
`oxycodone HCl, N=27 Following Intranasal Administration
`
`
`(Note to sponsor: change the title of x-axis to read “Percent Reduction in Drug Liking for
`Finely Crushed OxyContin vs. Powdered Oxycodone HCl Following Intranasal
`Administration”)
`
`The results of a similar analysis of drug liking for finely crushed OxyContin relative to finely
`crushed original OxyContin, were comparable to the results of finely crushed OxyContin
`relative to powdered oxycodone HCl. Approximately 43% (n = 12) of subjects had no
`
`future product can conduct the same
`study with a comparison to
`powdered oxycodone and provide
`the same analysis.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`The sponsor felt that the reader
`would want to know how the new
`and old formulations of OxyContin
`would compare for drug liking.
`Rather than include another figure, a
`description of the results was
`included. The continuous responder
`
`NDA 22-272 Dep Dir labeling memo OxyContin S014.doc
`
`
`
`Page 8 of 11
`
`Reference ID: 3294221
`
`
`
`
`
`reduction in liking with OxyContin relative to original OxyContin. Approximately 57% (n =
`16) of subjects had some reduction in drug liking, 36% (n = 10) of subjects had a reduction of
`at least 30% in drug liking, and approximately 29% (n= 8) of subjects had a reduction of at
`least 50% in drug liking with OxyContin compared to original OxyContin.
`
`fimction is not a common analysis
`and the natural tendency for the
`reader is to compare to figures,
`when, in this case, that is not the
`correct approach to understand these
`analyses.
`
`Summarv
`
`The in vitro data demonstrate that OxyContin has physicochemical properties expected to make
`abuse via injection difficult. The data from the clinical study, along with support from the in
`vitro data, also indicate that OxyContin has physicochemical properties that are expected to
`reduce abuse via the intranasal route. However, abuse of OxyContin by these routes as well as
`the oral route is still possible.
`
`Additional data, including epidemiological data, when available, may provide further
`information on the impact of the current formulation of OxyContin on the abuse liability of the
`drug. Accordingly, this section may be updated in the future as appropriate.
`
`These summary statements are
`derived from suggested language in
`the draft Guidance for Industry:
`Abuse-Deterrent Opioids —
`Evaluation and Labeling. The in
`vitro data are sufficient to show that
`
`the new formulation makes injection
`difficult.
`
`Taken together, the clinical liking
`data and the in vitro data support
`that the new formulation is likely to
`result in a meaningful reduction in
`
`intranasal abuse.
`
`
`
`
`
`OxyContin contains oxycodone, an opioid agonist and Schedule II controlled substance with an
`abuse liability similar to other opioid agonists, legal and illicit, including fentanyl,
`hydromorphone, methadone, morphine, and oxymorphone. OxyContin can be abused and is
`subject to misuse, addiction, and criminal diversion [See Warnings and Precautions (5.1) and
`Drug Abuse and Dependence (9.1)].
`
`
`
`Restatement of basic risk
`information to provide context for
`the study findings.
`
`NDA 22-272 Dep Dir labeling memo OxyContin S014.doc
`
`
`
`Page 10 of 11
`
`Reference ID: 3294221
`
`(b) (4)
`
`
`
`
`
`Summary
`The addition of in vitro and in vivo study results describing the abuse-deterrent characteristics
`of the new formulation of OxyContin represents an important advance in labeling products
`intended to reduce the abuse of opioid analgesics. The development of formulations intended
`to deter abuse is one step toward creating safer opioid analgesics. As noted in the draft
`Guidance for Industry: Abuse-Deterrent Opioids — Evaluation and Labeling:
`
`
`The science of abuse deterrence is relatively new, and both the formulation
`technologies and the analytical, clinical, and statistical methods for evaluating those
`technologies are rapidly evolving. Therefore, FDA will take a flexible, adaptive
`approach to the evaluation and labeling of potentially abuse-deterrent products.
`
`
`The data and interpretation of the data from the in vitro and in vivo studies added to the
`OxyContin package insert describe the properties of the new formulation while acknowledging
`that, as opioid analgesics must be able to deliver the opioid to patients for the management of
`pain, the extent to which an abuse-deterrent product is able to reduce abuse will never be
`absolute. Also, abuse-deterrence properties must always be evaluated relative to a comparator
`that can provide context to the findings.
`
`This action is consistent with the labeling standards set forth in the memo by Dr. Douglas
`Throckmorton dated April 15, 2013.
`
`
`
`
`
`
`
`
`NDA 22-272 Dep Dir labeling memo OxyContin S014.doc
`
`Page 11 of 11
`
`Reference ID: 3294221
`
`
`
`---------------------------------------------------------------------------------------------------------
`This is a representation of an electronic record that was signed
`electronically and this page is the manifestation of the electronic
`signature.
`---------------------------------------------------------------------------------------------------------
`/s/
`----------------------------------------------------
`
`SHARON H HERTZ
`04/16/2013
`
`Reference ID: 3294221
`
`
`
`
`
`
`DEPARTMENT OF HEALTH & HUMAN SERVICES
`
`
`
`
`Public Health Service
`Food and Drug Administration
`Rockville, MD 20857
`
`
`
`
`
`
`
`
`
`To:
`
`From:
`
`Date:
`
`Re:
`
`
`
`
`
`
`
`
`
`Memorandum
`
`Douglas Throckmorton, MD
`
`Gerald Dal Pan, MD, MHS
`
`April 12, 2013
`
`Office-level Memorandum on Purdue’s Reformulated OxyContin
`
`_____________________________________________________________________________
`
`To assess the impact of Purdue Pharma’s (Purdue) reformulation of OxyContin on OxyContin
`abuse, several investigations have addressed changes in the prevalence of OxyContin abuse and
`the routes by which OxyContin is abused. Staff in OSE’s Division of Epidemiology II have
`reviewed reports of Purdue’s investigations and two investigations conducted by Research
`Triangle Institute.1,2,3,4 This memo focuses on five investigations that both are reasonably well
`designed and contain sufficient data for review.5 While some investigations suggest a decline in
`some aspects of OxyContin abuse, particularly via non-oral routes, these findings are not supported
`by other data. Most of the investigations report either incomplete data or no data at all. No robust
`data on overdose or death are available. Taken as a whole, these investigations suggest, but do not
`confirm, that the reformulation of OxyContin has resulted in a decline in non-oral abuse.
`Furthermore, the data available at this time cannot support a robust conclusion that the
`reformulation of OxyContin is responsible for an overall decline in OxyContin abuse.
`
`
`
`
`1 Trinidad, J. & Dormitzer, C. (2012). Review of "A Summary of the Findings of the Post-Marketing Epidemiology
`Study Program to Detect Changes in Patterns of Abuse and Misuse and their Consequences: Addiction, Overdose and
`Death (as of October 15, 2011)". FDA/CDER/OSE. Submitted to DARRTS on Apr 27, 2012, for NDA #022272.
`2 Trinidad, J., Dormitzer, C., & Kornegay, C. (2013). Review of "Report on the Findings as of May 2012: Post-
`Marketing Epidemiology Study Program to Assess the Effects of Reformulated OxyContin on Patterns of Abuse and
`Misuse and their Consequences (Addiction, Overdose and Death), Patient Adverse Events, and Unintentional
`Exposures". FDA/CDER/OSE. Submitted to DARRTS on Apr 12, 2013, for NDA #022272 and IND #029038.
`3 Trinidad, J. (2013). Addendum to Review of "Report on the Findings as of May 2012: Post-Marketing Epidemiology
`Study Program to Assess the Effects of Reformulated OxyContin on Patterns of Abuse and Misuse and their
`Consequences (Addiction, Overdose and Death), Patient Adverse Events, and Unintentional Exposures" – Client
`Treatment Study, National Survey on Drug Use and Health, National Poison Data System, and ARGUS.
`FDA/CDER/OSE. Submitted to DARRTS on Apr 12, 2013, for NDA #022272 and IND #029038.
`4 Trinidad, J. (2013). Discussion on Postmarketing Data for Reformulated OxyContin®. FDA/CDER/OSE. Submitted
`to DARRTS on Apr 12, 2013, for NDA #022272 and IND #029038
`
`5 The five investigations are NAVIPPRO, the Client Treatment Study, the National Survey of Drug Use and Health
`(NSDUH), the RADARS System Poison Control Center Program, and the National Poison Data System (NPDS).
`
`
`
`
`
`Reference ID: 3292602
`
`1
`
`
`
`
`
`Changes in Overall Abuse
`
`Using data from standardized self-administered questionnaires completed by clients entering
`substance abuse treatment centers, the NAVIPPRO investigation examined the prevalence of
`OxyContin abuse in the 30 days prior to admission. Among persons reporting prescription opioid
`abuse, the prevalence of reformulated OxyContin abuse after its introduction to the market (12.1%)
`was 49% lower than the prevalence of original OxyContin abuse before the introduction of the
`reformulated version
`. Similarly, among all clients, prevalence of OxyContin abuse
`(regardless of formulation) statistically significantly declined from
` before, to
` after, the
`introduction of the reformulation. This finding, however, does not portray a complete picture of
`the extent of abuse of extended-release oxycodone, as abuse of original OxyContin persisted seven
`quarter-years after the introduction of reformulated OxyContin. In addition, some of the decline in
`the prevalence of abuse of extended-release oxycodone can be attributed to the cessation of the
`distribution of generic formulations before the introduction of reformulated OxyContin; however,
`abuse of generic formulations of extended-release oxycodone persisted after the introduction of
`reformulated OxyContin, though not to the same extent as abuse of original OxyContin. Thus,
`further study of the prevalence of abuse of reformulated OxyContin compared to the prevalence of
`abuse of original OxyContin must continue in order to determine the significance of these findings.
`At this time, the NAVIPPRO findings do not provide a quantitative estimate of the extent of
`reduction that is attributable to the reformulation.
`
`Similar to the NAVIPPRO investigation, the Client Treatment Study (CTS) investigation, which
`was independently conducted by Research Triangle Institute on behalf of a commercial sponsor
`other than Purdue, assessed trends in the prevalence of OxyContin abuse by examining the
`
`prevalence of past 30-day abuse among clients entering a random sample of approximately
`publicly funded substance abuse treatment centers. This investigation found that, on average, 2.6%
`of clients admitted to substance abuse treatment centers one year prior to the introduction of
`reformulated OxyContin reported past 30-day abuse of OxyContin, while 2.9% reported past 30-
`day OxyContin abuse one year afterwards. There is quarter-to-quarter variability in these data, and
`there was a trend toward increasing OxyContin abuse just prior to the introduction of reformulated
`OxyContin. While the CTS investigation does not document a decrease in OxyContin abuse, the
`significance of and reason for the observed increase is not clear. It is also not clear why two
`different samples of addiction treatment centers (i.e., NAVIPPRO and CTS) report very different
`findings on the prevalence of reported OxyContin abuse. These differences need to be explored
`further before conclusions can be drawn as to which estimates best reflect a true population-based
`estimate.
`
`The National Survey of Drug Use and Health (NSDUH) investigation, which surveys a
`representative sample of the non-institutionalized general U.S. population, found no significant
`change in the prevalence of non-medical OxyContin use, measured as the percent of respondents
`reporting non-medical OxyContin use in the 30 days prior to interview.
`
`Data from the RADARS System Poison Control Center Program and the National Poison Data
`System were used to assess the extent of OxyContin abuse by measuring the number of poison
`control calls related to intentional abuse of OxyContin. The RADARS data, which are based on a
`non-random sample of poison control centers in the U.S., documented that, after the introduction
`of reformulated OxyContin, there was a 32% relative drop in the number of calls related to
`OxyContin abuse per unique recipients of OxyContin prescriptions. This decline occurred abruptly
`after the introduction of reformulated OxyContin, a pattern that is inconsistent with the gradual
`
`2
`
`
`
`Reference ID: 3292602
`
`(b) (4)
`
`(b) (4)
`
`(b) (4)
`
`(b) (4)
`
`
`
`
`
`decline in dispensed prescriptions for OxyContin regardless of formulation. A similar finding was
`not noted for other prescription opioids. Data from the National Poison Data System, which
`includes all poison control centers in the United States, found a 30% relative reduction in calls
`related to OxyContin abuse after the introduction of reformulated OxyContin, and a 19% relative
`reduction when the calls were adjusted for the number of prescriptions dispensed for OxyContin.
`This investigation also found a
` relative reduction in unintentional general exposures to
`OxyContin, a measure that should not be impacted by the reformulation of OxyContin. Like the
`NAVIPPRO findings, the findings from poison control center data do not provide a robust
`quantifiable estimate of the extent of reduction in OxyContin abuse that can be attributed to the
`reformulation.
`
`Preliminary data from other Purdue-sponsored investigations reported reductions in OxyContin-
`related drug diversion cases, lower average street prices for reformulated OxyContin compared to
`original OxyContin, less "doctor-shopping" for reformulated OxyContin, a decline in internet chat
`room activity on routes of administration and extraction of oxycodone after the reformulation of
`OxyContin (but no change in abuse-endorsing or abuse-discouraging messages), and lower abuse
`of reformulated OxyContin among persons who had previously abused original OxyContin in a
`single county in Kentucky. In an analysis of its spontaneous adverse event reporting database,
`Purdue reported a decrease in the number of reports of drug abuse, overdose and medication errors
`with reformulated OxyContin compared to original OxyContin. In an analysis of OxyContin
`prescribing, Purdue found a modest decrease in the overall number of prescriptions for OxyContin,
`with more marked reductions in prescriptions to cash-paying patients, prescriptions for the higher
`strengths (40mg and 80 mg) of OxyContin, and prescriptions written by physicians whose
`prescribing practices are suggestive of abuse and diversion. While these preliminary findings are
`consistent with a decrease in drug abuse with reformulated OxyContin compared to original
`OxyContin, most of these investigations have yet to be validated, and the relationship between the
`assessed measures and the actual prevalence of drug abuse in the US population is not yet known.
`
`In sum, the above investigations of the overall prevalence of OxyContin abuse used a variety of
`approaches and outcome measures, yet at this time these investigations cannot support a robust
`conclusion that the reformulation of OxyContin is responsible for a decline in OxyContin abuse,
`and cannot provide a valid numerical estimate of a reduction in abuse that could be attributed to
`the reformulation. Furthermore, the currently available data measure these trends for about 18
`months after the introduction of reformulated OxyContin, and thus cannot address whether the
`observed trends will be sustained.
`
`Changes in Route-specific Abuse
`
`To date, data on changes in route-specific abuse after the introduction of reformulated OxyContin
`are limited to NAVIPPRO and the Client Treatment Study. The NAVIPPRO investigation, the
`Purdue-sponsored investigation that best addresses routes of abuse, suggests that non-oral abuse
`occurred less with reformulated OxyContin than with original OxyContin. In the two years after
`the introduction of reformulated OxyContin, approximately 40% of persons who abused
`reformulated OxyContin abused it non-orally, while approximately 70% of persons who abused
`original OxyContin abused it non-orally during the same time period. The Client Treatment Study
`investigation found that, among clients who reported OxyContin abuse, the percent of clients
`usually abusing it non-orally increased from 44% in the year before the introduction of
`reformulated OxyContin, to 48% in the year afterwards.
`
`3
`
`
`
`Reference ID: 3292602
`
`(b) (4)
`
`
`
`
`
`Because the reformulation of OxyContin was designed to prevent specifically non-oral abuse of
`OxyContin, it is important to examine the similarities and differences between these two
`investigations.
`
`First, both investigations use data collected upon intake into substance abuse treatment centers, so
`that the persons reporting prescription drug abuse will be those who have had sufficient negative
`consequences of substance abuse that they have sought treatment, or have been forced to receive
`treatment, in a substance abuse treatment center. Each of the substance abuse treatment centers in
`the CTS study receives at least some federal funding, while those in the NAVIPPRO investigation
`receive funding from public, private, or both public and private funding streams. There are not
`sufficient data to describe more detailed differences between the population in the NAVIPPRO
`investigation and that in the CTS investigation. The important point is that both investigations
`focus on populations that are in substance abuse treatment centers, and have had clear problems
`with substance abuse.
`
`Second, the standardized self-administered questionnaires used in the NAVIPPRO investigation
`attempted to differentiate between original OxyContin, reformulated OxyContin, and other
`formulations of extended-release oxycodone. In the CTS investigation, clients in substance abuse
`treatment centers were asked to identify substances that they used from a list. ‘OxyContin’ appears
`on this list, but is not further sub-identified as original OxyContin or reformulated OxyContin. It is
`also not clear how clients were to report use of generic extended-release oxycodone. Because data
`from NAVIPPRO show persistence of abuse of original OxyContin for several quarter years after
`the introduction of reformulated OxyContin, it is important that analyses of route of administration
`differentiate amongst and assess specific formulations and account for all availabe formulations. In
`this regard, analyses of route of abuse that differentiate amongst specific formulations are more
`informative than analyses of route of administration that do not make such a differentiation.
`
`Third, in the NAVIPPRO investigation, the self-administered questionnaire asked clients to report
`on each route of administration, while in the CTS investigation, clients were asked to list only the
`‘usual’ route of administration.
`
`Fourth, in neither of these studies are data collected on the euphoric effects of using OxyContin via
`a non-oral route, or the clinical consequences of abusing OxyContin through non-oral routes.
`
`Fifth, the timeframe after introduction of reformulated OxyContin differs between the two studies.
`Purdue’s descriptive statistical analysis of oral and non-oral routes of administration in
`NAVIPPRO covers the period June 1, 2009 through August 8, 2010 as the “before” period and
`August 9, 2010 through May 31, 2012 as the “after” period. Before the introduction of
`reformulated OxyContin,
` of those who reported abuse of original OxyContin did so via a
`non-oral route. After the introduction of reformulated OxyContin,
` of those who reported
`abuse of original OxyContin did so via a non-oral route, while
` of those who reported
`abusing reformulated OxyContin did so via a non-oral route. See Table 1 and Figure 1 below,
`which is derived from the review of Purdue’s study reports, and was generated by the Division of
`Epidemiology II.
`
`
`
`Reference ID: 3292602
`
`4
`
`(b) (4)
`
`(b) (4)
`
`(b) (4)
`
`
`
`
`
`Table 1: Percent of adults† who, upon stating that they abused extended-release (ER) oxycodone, reported
`non-oral abuse of ER oxycodone, before (June 1, 2009, to August 8, 2010) and after (August 9, 2010, to March
`31, 2012) marketing started for reformulated OxyContin®, by ER oxycodone category
`
`Data source: Purdue’s response to FDA’s Information Request, dated Dec 20, 2012
`†Clients admitted into substance abuse treatment more than once are counted multiple times. The data in the tables do
`not adjust for potential clustering of abuse patterns for clients admitted into subst