Human Abuse Liability & Abuse-Deterrent Formulations Category 4 Can we prove that ADF opioids really are? March 7, 2016 Richard C. Dart, MD, PhD Director, Rocky Mountain Poison and Drug Center Executive Director, RADARS System Professor, University of Colorado School of Medicine
Outline FDA Guidance on Abuse Deterrent Formulations A rigorous examination of the data currently available What s the problem? Some unique observations Low volume drugs are the real challenge
Abuse-Deterrent Opioids - Evaluation and Labeling Guidance for Industry Categories 1, 2, 3 have become routine. The goal of postmarket studies, Category 4, is to determine whether the marketing of a product with abuse-deterrent properties results in meaningful reductions in abuse, misuse, and related adverse clinical outcomes, including addiction, overdose, and death in the post-approval setting. Formal Studies Supportive Information Add to the totality of evidence to support ADF claim
National Academy of Medicine Session 4 Opioid Analgesics with Abuse- Deterrent Properties: Current Data and Future Opportunities Richard C. Dart, MD, PhD Director, Rocky Mountain Poison and Drug Center Executive Director, RADARS System Professor, University of Colorado
Opioids with Abuse Deterrent Properties: Current Data and Future Opportunities Oxycodone ER fulfills the Hill causation criteria Plausibility Temporality Effect Size Consistency Specificity Biologic gradient Coherence Experiment Analogy Alternative explanations International Society of Pharmacoepidemiology, August 26, 2015
Hill Criteria: Plausibility A plausible mechanism between cause and effect is helpful (but understanding of the mechanism is limited by current knowledge) Hill AB. Proc Royal Soc Med 1965;58:295-300.
Scientific Basis of Abuse-Deterrent Opioids Prescription drug abuse is like other drug abuse, except with an additional route of abuse: Oral = intact + chewed or crushed Intranasal Intravenous Importance of manipulating drug Crucial transition Changes perception of heroin use 1 Risk of acute (overdose, death) and chronic events (addiction, infections, death) higher after intranasal or IV abuse than oral abuse 1. Vosburg. J Child Adol Subst Abuse 2016.
Biological Plausibility Person in Pain Filling the Balloon Susceptible Person Intact Chewed Crushed Outcomes Addiction Recreational Abuser Overdose Abuse of Other Drugs Death Dart RC, Iwanicki JL. Can J Diag 2015;32:10.
Intervening in Prescription Drug Abuse Person in Pain Emptying the Balloon Guidelines Susceptible Person P D M P Intact A D F Chewed A D F Crushed Outcomes Addiction Recreational Abuser A D F Overdose Death A D F Dart RC, Iwanicki JL. Can J Diag 2015;32:10.
What if we had the perfect ADF? Two phases of abuse Obtaining the drug Friends and family Doctor shopping Pill mills Dealers Increased Intensity of Abuse Abusing the drug Neophyte swallow Experienced swallow
Hill Criteria: Temporality and Effect Size Effect has to occur after the cause (including a delay, if expected) Minimal delay expected for oxycodone ER All drug shipped after August 9, 2010 was reformulated version Pharmacy turnover of opioids is rapid A small association does not exclude a causal effect, though the larger the association, the more likely that it is causal. Hill AB. Proc Royal Soc Med 1965;58:295-300.
Oxycodone ER Prescriptions Dispensed Decreased Promptly After Reformulation 30% % Change in Prescriptions Dispensed 20% 10% 0% -10% -20% -30% Other Opioids Oxycodone ER -40% Reformulation of Oxycodone ER IMS, 2015-50% 2011 2012 2013 2014 2015 2016 Other Opioids = Oral dosage forms of opioid analgesics: hydrocodone, hydromorphone, morphine, oxymorphone, tramadol, tapentadol, and IR oxycodone
RADARS System - Mosaic Surveillance of Prescription Drug Abuse - 2015 Acute Health Events 49 Poison centers 46 states 512,609 cases 552,732 opioid mentions Drug Transactions Criminal Justice 260 agencies; 49 states 188,635 cases with 194,999 opioid mentions Entering Treatment Opioid Tx Program 72 programs; 42 states 43,861 cases with 201,099 opioid mentions Entering Treatment SKIP 135 practices, 47 states 12,328 cases with 80,617 opioid mentions General Population NMURx 30,0000 survey biannually All age groups Illicit Market Price StreetRx.com Users/Buyers, 50 states 25,250 price entries for an opioid Web Monitoring > 150 million sites monitored 13
Effect Size & Temporality: Oxycodone ER Abuse and Diversion, Population Adjusted, 2010-2016
Abuse Deterrent ER Oxycodone Reduces IV Abuse in Australia 4500 Introduction of Reformulated ER oxycodone 4000 3500 3000 Number of IV Cases 2500 2000 1500 1000 500 Non-abuse-deterrent oxycodone Reformulated ER oxycodone 0 Degenhardt et al., Drug Alc Depend 2015; 151: 56-57.
Hill Criteria: Consistency Consistent findings observed by different persons in different places with different samples strengthens the likelihood of an effect
Consistency: Oxycodone ER has Lower Rates Across Many Data Sources after Rx Adjustment Outcome Misuse Abuse Source RADARS (Poison Centers) RADARS (Poison Centers) NPDS (Poison Centers) NAVIPPRO (Treatment Centers) RADARS SKIP (Treatment Centers) Pre vs. Post % Change [95% CI] Since Reformulation ER Oxycodone Other Opioids Opioid Use Disorder Overdose Diversion Doctor Shopping RADARS OTP (Treatment Centers) Database of Opioid Users (Marketscan) Database of Opioid Users (Marketscan) RADARS (Drug Diversion) IMS Prescription Data Data adjusted for prescription volume Coplan et al. Clin Pharmacol Ther. 2016. -100% -50% 0% 50% 100% Decrease Increase
Additional Studies Epidemiology Cassidy 2014 Cicero 2015 Gomes 2012 Havens 2014 Hwang 2015 Larochelle 2015 McNaughton 2014 Sankey 2016 Hill Experimental (Abuse Liability) Darwish 2017 Gudin 2015 Hale 2015, 2016 Harris 2014 Setnick 2013 Stauffer 2009 Taber 2016 Webster 2011, 2012 Wen 2015 18
National Survey of Drug Use and Health, OxyContin Nonmedical Use 2 Introduction of Reformulated Oxycodone ER 1.6 # Cases of Past Year Nonmedical OxyContin Use (in millions) 1.2 0.8 0.4 0 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
20094 20102 20104 20112 20114 20122 20124 20132 20134 20142 Rate per 100,000 population Poison Centers: Response to Reformulation of Oxycodone ER and Oxymorphone ER Oxycodone ER Oxymorphone ER Other Opioids Reformulation of: OxyC ER OxyM ER Reformulation of: OxyC ER OxyM ER Reformulation of: OxyC ER OxyM ER 0.08 0.035 0.7 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0.03 0.025 0.02 0.015 0.01 0.005 0.6 0.5 0.4 0.3 0.2 0.1 0 0 0
Hill Criteria: Specificity The more specific an association between a factor and an effect is, the bigger the probability of a causal relationship 1 Results specific to oxycodone ER compared to other analgesic opioids FDA Concerns 2 1. Hill AB. Proc Royal Soc Med 1965;58:295-300. 2. International Society of Pharmacoepidemiology, August 26, 2015
Hill Criteria: Specificity Michna 2014 22
For data streams that collect information on a variety of events (e.g., abuse, adverse reactions), changes should be limited to abuse-related categories Abuse Misuse Suicide Unintentional Unintentional therapeutic error therapeutic error Unintentional general exposure Unintentional general exposure -100% -50% 0% 50% 100% Relative Change in Population Rate
For evaluation of ADFs, changes should be seen only in the routes of abuse expected to be affected Inhale or Inject Oral -100% -50% 0% 50% 100% Relative Change in Population Rate All routes of abuse should be low with an effective ADF
Rate Percent of States Confounders? RADARS PC 0.8 WA Rx Guidelines Reformulation Natl Drug Take Back FL TIRF REMS ER/LA REMS HC-APAP Tramadol 100% 0.7 90% 0.6 80% 0.5 Ox ER 70% 0.4 0.3 Other Opioids 60% 50% 0.2 40% 0.1 PDMPs 30% 0 20% All Other Opioids (per 100,000 population) Oxycodone ER (per 10,000 population) Operational PDMPS
26
The Problem FDA has legal regulatory requirements that they must fulfill. These are somewhat open to interpretation. FDA doesn t want to get burned again. Opinion: Unlikely that epidemiological data alone will satisfy FDA. 27
Solutions Prospective study mentioned by FDA repeatedly. But impossible to do? Improved epidemiological studies Can t meet FDA standard Will always have substantial issues with confounding and bias. 28
Experiment What if we had site that provided consistent stream of cloned substance abusers who could be randomly assigned a drug to abuse? Randomized New ADF drug IR version Non-ADF ER formulation of same API Patients are followed over time for abuse behaviors 29
Problems IRB/Ethics, is it ethical to let a person continue abuse once it s discovered? Most abuse is not by the patient If study works on one group or a few groups of patients, how do we know they are valid externally? If the drug is used throughout the US, will it still be abuse deterrent? Use epidemiological data to buttress and expand the data? 30
Results Might Look Like Havens, 2014 31
Problems for a Low Volume Category Killer Baseline for comparison is not high Abusers haven t been exposed to your product. Do they not like it, or just not know about it? Shortest possible time because sponsor wants to achieve market share Lack of familiarity may be difficult to identify product Limitations of subanalyses (e.g. route of abuse) because of small number of abuse cases and outcomes. 32
Low Volume Solutions Emphasizes need for a prospective study Epidemiological data Diffuse Enriched Prolonged 33
Conclusions and Implications Specificity, consistency and effect size indicate that abuse deterrent opioids are likely to be effective in reducing abuse and its outcomes Two drugs: Similar effects for crush-resistant oxymorphone ER Widespread use - prices, reduce the crucial transition from intact swallowing to crushing Education, training, and other interventions needed as well How can we do any type of study in abuse subjects?
35