The consequences of drug misuse on post-marketing surveillance

KEYWORDS:

• Pharmacovigilance
• addiction
• drug misuse
• addictive liability
• novel psychoactive substances
• prescription drugs
• prescribing medications

Introduction

Over the past decade, the ‘traditional’ drug scenario has shown significant changes because of the emergence of a range of molecules, e.g. the novel psychoactive substances (NPS), which are either already existing or newly created molecules [1]. A range of prescribed medications are currently being used as NPS [1]. Overall, the misuse and diversion of medications is a significant and increasing public health concern [2], with 5.4% of British respondents aged 16–19 years old having abused a prescription drug in the past 12 months [3]. It is a matter of concern that, for a range of prescribing molecules (e.g. gabapentinoids), the formal pre-marketing processes had not been able to appropriately identify their potential for abuse, a potential which has however emerged overtime [4,5]. Similarly, drugs such as benzodiazepines and z-hypnotics were considered ‘safe’ for many years before their addictive liability levels were identified.

Hence, in this article, we aimed at commenting on the different factors relating to pre- and post-marketing prescription drugs’ abuse liability assessment; issues likely to be complicated by recent changes in drug scenarios.

Pre-marketing assessment of the abuse liability of a new drug

For a new drug to be approved and to be considered ‘safe’, a range of abuse liability assessments are carried out, which are based on a multi-faceted profile of the drug’s chemistry, pharmacology, clinical manifestations, similarity to other drugs in a class, and potential for public health risks [6]. Pre-marketing trials are designed to provide useful information on abuse liability levels [2,7–9]. However, these trials typically involve the administration of carefully controlled, daily limited, therapeutic dosages. A weakness of these trials is that subjects with a current/previous history of drug misuse/addiction are excluded [10]. As a consequence, the molecule’s potential for abuse will be fully appreciated only when the real world client population, involving vulnerable individuals, is exposed to it [5].

Current and optimal post-marketing pharmacovigilance approaches to detect signals of abuse liability of a product

Challenges in pharmacovigilance database analyses

Spontaneous pharmacovigilance reports may play a major role in the identification of drugs’ addictive liability levels. However, relevant data are often incomplete, and spontaneous reporters often use different terminologies and/or a range of terms in parallel [6]. Furthermore, the rate at which cases are reported is dependent on issues, such as: time since launch; existing regulatory activities; media attention; and literature reports [11]. A range of proxies of prescription drug misuse are typically being used across pharmacovigilance studies, including: number of prescribers, number of dispensing pharmacies, early refills, and volume of drugs dispensed [12]. Although monitoring trends in reporting patterns may improve signal detection, optimal approaches should consider a range of further methodological options [13], including the use of: (a) appropriate denominator data, e.g. considering the estimated number of people in the general population exposed to a given molecule; (b) calculation of the proportional reporting ratio (PRR). This is a measure of disproportionality of reporting, with a PRR greater than 1 suggesting that the adverse event is more commonly reported for individuals taking the drug of interest, relative to the comparison drug(s) [14]; (c) data mining algorithms, which are database screening tools used to distinguish proper ‘signals’, from ‘noise’ (e.g. data lacking of public health interest) [15]; and (d) Bayesian hierarchical models, helpful in analyzing a huge volume of adverse drug reaction (ADRs’) data, with outputs being used in signal validation [16].

Role of post-marketing, multiple sources/social media, surveys

As a single source may not be informative enough, some surveillance programs (e.g. the French CEIP Addiction vigilance network) consider simultaneously a range of indicators, focusing on: suspicious prescriptions; range of misused prescribing psychotropics; health insurance databases; and health professional sentinel networks [17]. At a wider level, the Europe-wide toxicovigilance system monitors almost 600 NPS, including listed prescription drugs, identifying signals of related serious harms and responding through risk communication/risk assessment [18].

Although further data could derive from law enforcement; drug abuse treatment programs; poison centers; and medical emergency departments [19], one could wonder about how effective these agencies are at early detecting the abuse potential of psychotropics. In fact, it is only when the molecule’s availability has reached a critical mass that the Police will be able to seize the molecule’s first batches; relating near misses will present to the emergency rooms; and clients will start asking treatment agencies for help. An increase in online trafficking/debate on a specific psychoactive drug typically precedes the occurrence of clinical incidents at the population level [5,20,21]. Using an integrated web mapping/social media system, phenomena, such as abuse of both synthetic cannabimimetics/‘spice’ and synthetic cathinones were all identified as emerging trends at a time when very limited or no scientific publications were available [20]. With a similar web/social media-based approach, an extensive description of gabapentinoid addictive liability was provided at a time when the potential of pregabalin abuse was not even mentioned in the prescribers’ aids [5].

The use of social media/web data for ADR monitoring, including assessment of search logs from healthcare professionals [22], is receiving increasing levels of attention [23]. The amount of user-generated data on social media and other Internet-based venues provides the possibility to systematically mine, aggregate, and analyze textual/unstructured data (e.g. ‘infodemiology’), with ‘infoveillance’ being used for surveillance purposes [24]. Similarly, the ‘netnography’ [25] systematic approach provides an innovative methodology to collect online data on NPS/prescription drugs’ abuse/misuse.

While considering the web-based approach, White et al. [26] analyzed search queries collected from 80 million consenting users and found that the ADR detection performance via search logs was comparable to the adverse event system detection of US FDA, even increasing accuracy levels by 19%. Finally, Sarker et al. [27] recently assessed Twitter users’ posts as a resource for the automatic monitoring of prescription medication abuse. They found that, with a calculated 82% accuracy overall, the percentage of tweets containing abuse signals were significantly higher for the three case medications (Adderall 23%; quetiapine: 5%, oxycodone: 12%) than the proportion for the control medication (metformin: 0.3%). There are different approaches to ADR detection/extraction from web/social media; these include: supervised classification techniques (to detect posts containing ADR mentions) and lexicon-based approaches (for extraction of ADR mentions from texts, such as Tweets) [28].

What happens when the abuse potential of a drug has been identified? A few case studies

If during the post-marketing phase, a potential for abuse is identified, a range of risk management programs (RMPs) [6] are increasingly being required to address the regulatory agencies’ concerns [29]. RMPs typically require the implementation of a ‘risk minimization action plan’, which may well include either enabling access to a drug only under highly restricted conditions or drug scheduling [30]. A few case studies, highlighting a range of different post-marketing situations, are illustrated below.

Drugs with a limited potential for abuse where post-marketing measures are currently not being considered: olanzapine; quetiapine; venlafaxine; and tropicamide

Olanzapine, a popular second-generation antipsychotic (SGA), is being anecdotally identified online as the ‘ideal trip terminator’ after a psychedelic drug binge. The molecule is self-prescribed, and for a few days only, at very large daily dosages (e.g. up to 50 mg) [31,32]. Another SGA, quetiapine (‘Q ball’), is anecdotally advised to help ‘come off the psychedelic trip’. Vulnerable subjects (e.g. adolescents and inmates) may be particularly at risk [33]. Methods of quetiapine misuse include inhaling/injecting crushed tablets [34]. Reasons for abuse of atypical antipsychotics may include the desire of ‘feeling mellow’ [35].

Similar issues have been raised with venlafaxine/‘baby ecstasy’ [36,37], a popular antidepressant whose misuse is possibly related to the increase it produces in dopamine neurotransmission mainly at the prefrontal cortex level [38].

Conversely, changes in routes of administration are behind abuse of tropicamide, an ophthalmic anticholinergic compound which, when injected, may be associated with hallucinations, dysphoria, psychomotor agitation, tachycardia and suicidal ideation [39].

The abuse potential of the above molecules is not mentioned in the medications’ package leaflet/patient information leaflet [40–43], and it has not been the subject of pharmacovigilance studies.

Drugs possessing a clear potential of abuse where post-marketing measures have been/are being considered: gabapentinoids; oxycodone; and hydromorphone

Gabapentinoids are approved in Europe for the treatment of epilepsy/partial seizures and neuropathic pain, with pregabalin being licensed for generalized anxiety disorder as well [5]. In parallel with increasing prescribing levels, a growing black market is currently being observed [44]. Overall, pregabalin is characterized by higher potency, quicker absorption rates, and greater bioavailability than gabapentin [5]. Typical abusers of these compounds are individuals with a history of recreational polydrug use. Euphoria, improved sociability, opiate-like sedation, dissociation, and psychedelic effects are associated with gabapentinoid high/very high dosage intake. Intravenous use, rectal ‘plugging’ and smoking of pregabalin have also been reported [45], with opiates/opioids being frequently misused in combination [46]. Pregabalin has been approved in Canada and the United States, since 2005, and EU approval to treat generalized anxiety disorder was received in 2006 [47]. Yet, a proper debate regarding abuse and dependence did not start in the medical literature before 2010 [48]. In the United Kingdom, gabapentinoids are currently the subject to formal attention/rescheduling to prevent diversion and abuse [44].

Oxycodone

Its non-medical use is a relatively recent concern, especially in the United States [6] and Canada [47]. Because of the controlled-release formulation, at the time of approval its abuse potential was considered by FDA to be similar to other opioid analgesics. Crushing the controlled-release capsule, followed by intravenous injection or snorting, has however led to high levels of abuse [6]. As a result of this, an abuse-deterrent formulation was introduced in 2010 [49], and labeling changes were implemented.

Hydromorphone

The FDA approved this molecule in 2004 for the management of moderate/severe pain [6]. Eventually, a sponsor’s study identified that, compared to taking hydromorphone alone, when co-administered with alcohol the blood hydromorphone concentrations were significantly higher [6]. Eventually, sales and marketing of the drug were suspended.

Conclusions

In this article, it is suggested that in Europe, much as is done in the United States, post-marketing surveillance for substance abuse is needed as a matter of routine to assess the abuse potential of newly released drugs with addictive liability levels (e.g. opiates/opioids; stimulants, etc.). Overall increasing levels of access to the web over the past 15 years or so may have contributed to the current scenario of prescribed drugs’ abuse. Indeed, social networks may play a role in prescription drugs’ aggressive marketing/distribution from rogue websites [50]. Furthermore, within the online drug fora communities, there are some educated/informed users who typically ‘test’ a range of psychotropics to achieve the state of consciousness they find most pleasurable, eventually making this information available to peers [51].

Guidelines have been drafted to carry out laboratory-based testing to improve post-marketing medications’ abuse predictive ability [2,7–9]. However, overall failures of pre-marketing clinical trials in detecting a range of psychotropics’ addictive potential [4] may have complicated the current scenario. Abuse liability-focused laboratory testing may need to consider interaction studies with alcohol and/or other drugs. Furthermore, tamper-resistant delivery systems need to be routinely considered [2,52].

Pharmaceutical manufacturers applying for approval of novel molecules with clearly identified levels of abuse (e.g. opiates/opioids, stimulants, etc.) could submit post-marketing plans for a strategy addressing education, prevention, detection and abuse/diversion issues. Post-authorization safety studies to monitor the effectiveness of any risk minimization measures would be desirable as well [6].

In the United Kingdom, many websites offering unauthorized medicines for sale have seen been shut down, but this has just shifted the sellers to servers in overseas countries, with a sort of ‘playing catch up’ scenario [53]. Since part of the problem is the demand, and not the supply, it may be unrealistic to design a reliable system to stop online medicine-related transactions [53].

Finally, while a continuum of related professional training is needed, it may be important to consider a strategy increasing clients’ access to treatment services [54] possibly enhancing links between community pharmacists (first professionals to identify a repeat supplies’ issue) and prescribers/clinicians.

Financial & competing interests disclosure

This paper was supported in part by grants of the European Commission (Drug Prevention and Information Programme 2014-16; contract no. JUST/2013/DPIP/AG/4823; EU-MADNESS project). Further financial support was provided by the EU Commission-targeted call on cross border law enforcement cooperation in the field of drug trafficking - DG Justice/DG Migrations and Home Affairs (JUST/2013/ISEC/DRUGS/AG/6429) Project EPS/NPS (Enhancing Police Skills concerning Novel Psychoactive Substances; NPS). F Schifano is an AVMD member; EMA Advisory board member. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.