Labeling for Abuse-Deterrent Drugs
- Posted by: Ken Phelps
- Published on: June 10, 2010
This past Tuesday (6/8/2010), we participated in a DIA-sponsored webinar entitled Understanding the Development and Label Allowances for 505(b)(2) Abuse-Deterrent Products. Joining me, Ruth Stevens our CSO and Cindy Phurrough, our Clinical Operations head, was Dr. Lynn Webster, Chief Medical Officer of Lifetree Clinical Research and Pain Clinic. Ruth reviewed the regulatory background, the types of studies that need to be done and what sections of the labeling can include information about product characteristics that might thwart abuse of the drug. Cindy reviewed practical aspects of the clinical studies, while Dr. Webster presented real-world experiences conducting liking studies. I’d like to use this post to highlight portions of this webinar.
From a technological standpoint, it is understood that the goal is to make changes to the API or formulation of a drug that has abuse potential in such a way that it dissuades the abuser from wanting to use the drug and choose something else. As Dr. Webster indicated, the real-world problem is how can we test the product in a controlled setting? We can’t, so the FDA will not allow labeling to state that the technology is abuse-deterrent until post-marketing studies. So, what can we put on the labeling in order to have a commercially meaningful difference upon which to market our product?
It turns out that you can get information about the in vitro and in vivo performance of your technology into a couple of sections of the labeling. The FDA will judge whether your technology provides an incremental difference from an existing, non-deterrent product — Ruth provided labeling language from examples of approved products.
Since abusers will try to get immediate highs, they will often attempt to crush the tablets to snort it or extract the actives to inject it. A product designed to be abuse-deterrent does not have to address both of these possibilities in order to get FDA approval (the question of market acceptance is another matter). Thus, a product like we reviewed recently, Acurox®, had three technologies:
- Addition of Niacin to induce an unpleasant flushing if the dose of drug product administered exceeds the labeled dose.
- Sodium laurel sulfate (SLS), intended to irritate the nasal mucosa if the drug product is pulverized and snorted and
- Polyethylene oxide, a substance that polymerizes upon wetting, forming a gel that is intended to be too viscous to inject through a needle and thus prevent misuse by injection. This excipient also has the potential to make the extraction of oxycodone from the matrix more difficult
In the webinar, Ruth reviewed the three types of studies that can be conducted to yield information that may be included on labeling at approval:
- In Vitro Data
Data from studies designed to evaluate the product’s resistance to attempts to defeat the abuse-deterrent properties. These studies should be based on information from abusers, and must be scientifically rigorous and blinded.
- Pharmacokinetic Data
Data from studies that evaluate the effects of different methods of physical manipulation identified in the in vitro studies on the pharmacokinetic profile.
- Clinical Data — Human Liking Studies
Data from studies of experienced drug abusers to evaluate the likability and euphorigenic effects of manipulated and intact product compared to the product that is not tamper resistant.
Dr. Webster’s presentation focused on the conduct of the clinical liking studies. Dr. Webster has had many years of experience designing, conducting and evaluating these kinds of studies. A major point he made was: we can design studies to measure differences between test and reference products, BUT, which measures are important in the REAL world and how much difference is enough to deter abuse? His conclusion: We don’t know. This is why FDA won’t allow the words ‘abuse-deterrent’ on labeling without actual experience studies post-approval. He used the following slide to make his point — though there is a 30% difference in likability, will that deter abuse in the real world? We know that a 30% difference in clinical efficacy between test and placebo is clinically significant, what should we use for abuse-deterrent standards?
Both Camargo and Lifetree are engaged by clients to consider these issues and we discuss them with FDA every month. We can say that the standards are evolving.