A rare, debilitating condition in pregnancy that causes nausea and vomiting so severe that some women end up terminating their pregnancies can be effectively treated with the seizure drug gabapentin.
That is the finding of a new study of 21 women with the condition led by the University at Buffalo and published on Oct. 29 in the American Journal of Obstetrics & Gynecology Maternal Fetal Medicine.
It is the first double-blind, randomized controlled trial that demonstrates an effective therapy in treating outpatients with hyperemesis gravidarum (HG), the condition that afflicted Princess Kate Middleton during her pregnancies.
A study participant tells her story in this video.
Led by Thomas Guttuso, Jr., MD, professor of neurology in the Jacobs School of Medicine and Biomedical Sciences at UB, and a physician with UBMD Neurology, the research was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development of the National Institutes of Health.
“Gabapentin is the first therapy shown to effectively reduce nausea and vomiting, and improve oral nutrition in patients with hyperemesis gravidarum,” Guttuso said.
The double-blind study was conducted on 21 women who didn’t respond to standard treatments and who required intravenous hydration.
Patients were randomized to receive either oral gabapentin (12 patients) or a standard-of-care treatment, oral ondansetron or oral metoclopramide (nine patients) for seven days.
Each patient kept track of symptoms using a validated home diary used in HG to record nausea, vomiting, retching and oral nutrition.
“The women taking gabapentin experienced a 52% greater reduction in nausea, vomiting and retching total scores than women taking standard-of-care treatment, which is quite significant,” noted Guttuso.
“Even more striking is the fact that gabapentin therapy provided a 96% increase in oral nutrition scores and a 254% increase in global satisfaction (a measure of how satisfied patients were with overall outcome) of treatment compared to standard-of-care therapy,” he added.
“It’s devastating”
“Nausea and vomiting in pregnancy is very common,” Guttuso explained, “but hyperemesis gravidarum is rare and on the very far end of the severity spectrum. It’s devastating.
The nausea and vomiting are just relentless all day and often all night, too, so these women can’t sleep. They can’t keep anything down. It wreaks havoc on virtually every aspect of a woman’s life.”
Patients become dehydrated and often have to be admitted to the hospital, he explained, where their condition improves with intravenous fluids. “But after discharge, at least a third end up being readmitted because none of the treatments we have right now are effective,” he said.
“This is a really tough disorder to manage,” agreed Vanessa Barnabei, MD, Ph.D., professor and former chair of the Department of Obstetrics and Gynecology at the Jacobs School, who helped recruit patients but who was not involved in the study.
“These women want to be pregnant, they want to do what’s best for the pregnancy, and they are really miserable,” she said. “Many of them come in and they haven’t had anything to eat or drink for two or three days or more because every time they try to eat or drink something they just bring it right back up.”
Of course, the inability to eat or drink can negatively impact the baby’s development. Guttuso noted that women with HG who gain less than 7 kg (about 15 pounds) during pregnancy are significantly more likely to have an infant delivered prematurely, of low birth weight and with a low Apgar score, which is used at birth to assess the health of newborns.
Guttuso said surveys show as many as 15% of women with HG have such severe symptoms that they end up terminating their pregnancies. One woman in the UB study experienced vomiting so violent that it ripped a hole in her esophagus; she had considered terminating her pregnancy but once she was on gabapentin, she was able to carry the baby to term. She also went on to have another child when she was again treated with gabapentin.
“Not only did gabapentin decrease nausea and vomiting in these women, but it significantly improved their oral nutrition as well,” Guttuso said. “This is very important because it’s the women who lose the most weight and have the worst nutrition while they’re pregnant who have more risks both to themselves and to the fetus,” he said.
“A therapy that can improve oral nutrition and decrease nausea and vomiting certainly could be of great value to women with hyperemesis gravidarum,” he continued.
Accidental discovery
Guttuso’s interest in gabapentin stretches back decades, the result of an accidental discovery he made as a neurology resident.
After Guttuso prescribed it to a breast cancer patient for her hot flashes, she informed him that it also appeared to fully resolve her refractory nausea and vomiting that had been induced by chemotherapy.
That led Guttuso to see if gabapentin might be able to treat HG. In 2010, he collaborated with other Buffalo physicians to conduct a pilot study with seven women with HG.
That study found that the women had an average 80% reduction in nausea and 94% reduction in vomiting. They returned to nearly normal levels of eating and drinking.
The next step is to enroll more women in a larger study. Guttuso noted that a major drawback of the current study was its very small size, driven, in part, by the reluctance of women with HG to stop taking whatever medication they had been prescribed, even if it wasn’t very effective.
“With a different study design and with more study sites, we will be able to enroll many more women,” said Guttuso. “If this subsequent larger study also shows positive results, this will reassure the medical community that this is a true effect that’s consistent and reliable.”
He is pursuing funding from public and private sources to support such a study.
Pregabalin and gabapentin are approved pharmacotherapies for the treatment of some epileptic and pain disorders, and pregabalin also for generalized anxiety disorder (Bocksbader et al., 2010; Calandre et al., 2016). Both pharmaceuticals are very closely related regarding their pharmacology (Bockbader et al., 2010; Calandre et al., 2016). Therefore, gabapentin and pregabalin can be placed in their own group of gabapentinoids (Rogawski and Bazil, 2008).
They are 3-substituted derivatives of the neurotransmitter γ-aminobutyric acid (GABA) and known inhibitors of α2d-subunit- containing voltage-dependent calcium channels (VGCC), more precisely the α2d type 1 and 2 proteins of the P/Q type of VGCCs (Tran-Van-Minh and Dolphin, 2010; Mico´ and Prieto, 2012).
By this action, they inhibit the trafficking of the α2d subunit complex to the plasma membrane and reduce the synaptic vesicle exocytosis (Tran-Van-Minh and Dolphin, 2010; Mico´ and Prieto, 2012). These VGCCs are located predominantly in presynaptic membranes and it was demonstrated that gabapentionids restrain stimulus-dependent synaptic transmitter release, mainly the excitatory transmitter glutamate and norepinephrine, but not dopamine (Dooley et al., 2000; Bockbader et al., 2010; Rogawski and Bazil, 2008).
Thereby, gabapentinoids may act against aberrant neuronal “overexcitation” and, likely, also against sensitization (Eroglu et al., 2009; Mico´ and Prieto, 2012). Additionally, therapeutic doses of gabapentinoids are dose-dependently associated with a modest increase of the extracellular GABA-concentration in brain tissue (Peng et al., 2008.; Bockbader et al., 2010; Cai et al., 2012; Calandre et al., 2016) and, thus, have weak GABAmimetic features that most likely drive the relaxation and euphoria experienced especially in the beginning of the drug therapy and during an overdose.
There is a substantial tolerance against this euphoric high which is typical for addictive GABAmimetics, e.g. benzodiazepines or propofol (Bonnet, 2011; Korpi et al., 2015). Pharmacokinetically, the gabapentinoids are nearly “ideal” pharmaceuticals with good tolerability (Zaccara et al., 2017), a low interaction potential (with the exception of combining with clozapine, opioids or sedatives (Englisch et al., 2012; Calandre et al., 2016;
Schjerning et al., 2016a; Quintero, 2017.; Abrahamsson, et al. 2017)), no metabolism and no protein binding (Bockbader et al., 2010; Calandre et al., 2016). However, they need dose reduction alongside increasing renal insufficiency (Verma et al., 1999; Calandre et al,, 2016).
Within the last decade, both, gabapentin and pregabalin, have become blockbuster prescription drugs with myriads of prescriptions worldwide (Kapril et al., 2014; Calandre et al.; 2016; Chiappini and Schifano, 2016; Kwork et al., 2017). Of note, a good portion of these drugs were prescribed “off-label” against anxiety, non- neuropathic pain, mood instability, insomnia, neurasthenia, somatoform disorders, and withdrawal symptoms from recreational drugs (Prescrire Int, 2012; Calandre et al., 2016; Freynhagen, et al., 2016; Kwok et al., 2017).
Since gabapentin and pregabalin became also easily obtainable over the internet and were sold on black markets, gabapentinoids have been assumed to possess considerable abuse liability (Schifano et al., 2011; Prescrire Int, 2012; Kapril et al., 2014).
This corresponds to pharmacoepidemiologic analyses of prescription data and a mounting number of records pointing to an abuse of gabapentinoids that have been spontaneously reported to pharmacovigilance databases, mainly in Scandinavia, the UK, and Germany (Chalabianloo and Schjøtt, 2009; Schwan et al., 2010; Gahr et al., 2013a; Bodén et al., 2014; Asomaning et al., 2016; Schjerning et al., 2016b).
Notably, the vast majority of the registered patients were currently or previously dependent on other substances, too, mostly opiates or sedatives (Chalabianloo and Schjøtt, 2009; Schwan et al., 2010; Prescrire Int, 2012; Gahr et al., 2013a; Bodén et al., 2014; Asomaning et al., 2016; Schjerning et al., 2016b).
This was supported by the latest analysis of the EudraVigilance database which included 11940 misuse reports of gabapentin (N=4301 records corresponding to 410 patients) and pregabalin (N=7639 records, 1315 patients) to the European Medicine Agency from Europa, East Asia, North and South America in the period 2004-2015 (Chiappini and Schifano, 2016). For both gabapentinoids, there was as considerable increase of those reports over time with a peak in 2013 (pregabalin, N=2154 records) and 2014 (gabapentin,
N=1001 records) (Chiappini and Schifano, 2016). These pharmacovigilance data were warning although, for reasons of methodology, remaining less specific towards the addiction risks of gabapentinoids, because it is cannot be excluded that they are simply innocent bystanders of other more powerful substance use disorders (SUD).
We attempt to estimate the addiction risks of gabapentinoids in several steps. Firstly, we conducted a review about animal and human studies focusing on rewarding properties (Panlilio and Goldberg, 2007) of gabapentinoids.
Secondly, we evaluated clinical studies and case reports having been related to gabapentin or pregabalin misuse according to fulfilled ICD-10-criteria of dependence (Dilling and Freyberger, 2006), information about the magnitude and durability of self-administrations (Panlilio and Goldberg, 2007) including relapses, and treatment-seeking behavior of affected patients.
Thirdly, we reviewed the overdose safety of gabapentin and pregabalin to assess the benefits and inconvenience to the consumer. Next, we discussed their addiction risks basing upon these findings and on a popular and useful explanation of how addiction can occur, i.e., the Incentive Sensitization Theory of Addiction (Berridge and Robinson, 2016). Finally, we provide an approach to compare the addiction risks of gabapentinoids with those of common substances of abuse.
Discussion
Pregabalin appeared to be somewhat more addictive than gabapentin, taking into consideration that the pregabalin use was more frequently associated with behavioral ICD-10-dependence symptoms, switches from prescription to self-administration and self-administrations themselves (Table 2). However, this review did not find convincing evidence of a substantial genuine addictive power of gabapentinoids in general which is primarily suggested from their limited rewarding properties and the very few cases with behavioral dependence symptoms of gabapentinoids without a prior substance abuse history (Yargic and Ozdemiroglu, 2011; Ashwini et al., 2015; Halaby et al., 2015; Driot et al., 2016; Table 2). In support, there were only 3 cases reports that mentioned relapses (Victorri-Vigneau et al., 2007; Grosshans et al., 2010; Yazdi et al., 2015; Table 2) and we could not found any publication reporting people who sought treatment for the use of gabapentinoids which appeared to be used mostly together with other substances of abuse (Table 1). Also, we could not
find any note on social hazards being attributed to the use of gabapentinoids in isolation. At this juncture, it should be mentioned that drugs of abuse were used rather infrequently in isolation from other substance dependences, which may point to contextual, social, and individual factors that maintain the substance use. Recently, some additional concerns about a substantial addictive power of gabapentinoids have arisen. Thus, a distribution analysis of the French Pharmacovigilance database found no significant association between the exposure to pregabalin and drug abuse or dependence, nevertheless, taking into account the limited specificity of a spontaneous reporting pharmacovigilance system (Bossard et al., 2016). Recently, Mutschler et al (2015) found no evidence of a non-medical pregabalin use in an opioid substitution program.
Impact on the mesolimbic reward system
Drugs of abuse are characterized by an increase of the dopamine activity in the mesolimbic reward system (Karoly et al., 2015; Volkow and Morales; 2015). Their repeated administrations drive neuroplastic changes in glutamatergic inputs to the striatum and midbrain dopamine neurons, which is associated with an increase of the motivational salience of drug cues, a reduction of the sensitivity to non-drug rewards, a weakening of self-control behavior; and an affection of the individual stress reactivity (Spiga et al., 2014; Karoly et al., 2015; Volkow and Morales, 2015; Fosnocht and Brand 2016; Scofield et al., 2016).
As yet, there is no evidence for gabapentinoids to increase the extracellular dopamine activity in the mesolimbic reward system. The only microdialysis study to this subject found that gabapentin (25-200 mg/kg, intraperitoneal, rats) produced a modest increase (approximately 50%) in extracellular nucleus accumbens GABA levels but failed to alter either the basal or the cocaine-enhanced dopamine activity in this key region of the reward system (Peng et al., 2008). This might restrict the ability of gabapentinoids to develop a substantial addictive power. Neuroimaging studies on the human reward system (Ernst and Luciana, 2015) under the influence of gabapentinoids are warranted.
Prevalence rates and motives of non-medical gabapentinoid use
As yet, epidemiologic surveys have not measured the prevalence rates of gabapentinoid abuse and dependence in the general population. These rates can be roughly estimated from an elderly German hospital population (life-prevalence of dependence: 0.25%, Cossmann et al., 2016) and younger British internet-population (life-prevalence of misuse: 0.5-1.1%, Kapril et al., 2014).
Opioid using (self-administrating) patients and patients in opioid substitution programs are at particular risk for gabapentinoid abuse and dependence with up-to- 6-month prevalence rates of up to 26% (Grosshans et al., 2013; Baird et al., 2013; McNamar et al., 2015; Wilens et al., 2015; Snellgrove, 2016; Smith et al., 2015; Bastiaens et al., 2016; Table 1).
There is robust evidence that opioid users including multiple drug users selected gabapentinoids mainly due to their special features to boost an euphoric high and reduce withdrawal symptoms while producing only few adverse effects (Schwan et al., 2010; Schifano et al., 2011; Grosshans et al,. 2013; Baird et al., 2013; Wilens et al., 2015; Smith et al.,2015; Bastiaens et al., 2016; Snellgrove, 2016). These cohorts clearly preferred pregabalin allowing a more rapid and stronger euphoric high than being possible with gabapentin (Tables 1 and 2). Most likely, this results from pharmacological differences of gabapentin and pregabalin (Bockbader et al., 2010; Calandre et al., 2016). Firstly, pregabalin is absorbed more rapidly (reached its maximal blood level within 1.5 hours after oral intake) and possesses a greater bioavailability (gabapentin: 33%–66%, pregabalin:
>90%). Secondly, while pregabalin is absorbed dose-independently gabapentin`s plasma concentrations have been found to have a non-linear relationship with increasing oral doses because its absorption underlies a saturation process in the gastrointestinal mucosa (Bockbader et al., 2010; Calandre et al., 2016). Thirdly, pregabalin has a stronger inhibitory action on the α2δ-subunits containing VCGG compared to gabapentin (Bockbader et al., 2010; Calandre et al., 2016). Especially the faster onset of a euphoric high and the linear relationship between blood concentrations and oral intake are supposed to be the reasons why pregabalin is
preferably self-administered by patients with experiences in substance abuse, such as opioid addicts. At this juncture, pregabalin would be associated more closely with the hazards of this population, such as dependence and overdose death (Hser et al, 2017), than gabapentin (Table 1-3).
Durability of gabapentinoid self-administrations
Regarding the self-administration for longer than one year, we did not find any case report about gabapentin and three case reports mentioning pregabalin self- administrations over a period of 2 and 4 years (Westin and Strom, 2010; Aldemir et al., 2013; Ashwini et al., 2015) (Table 2).
In a cohort of patients admitted to detoxify from other drugs, Snellgrove (2016) found that 13% of those patients, who had experience with non-medical pregabalin use had a period with pregabalin- dependence no longer than 2 years. As yet, only one prospective study on the longitudinal course of gabapentinoid has been published and this study found a considerable increase of gabapentin misuse in a special cohort of American prescription opioid dependents within 5 years (Smith et al., 2015) which support the view that opioid dependents are at particular risk for a non-medical co-use of gabapentinoids (Table 1).
Using the Danish nationwide Prescription Registry, Schjerning et al (2016b) found evidence for a long-term use of pregabalin when they analyzed pregabalin dispensing to patients in the period from 2004 to 2013. Of 42350 pregabalin-recipients, there were 2765 (6.5%) and 137 (0.3%) persons who had received prescriptions ≥ 600 mg/d and ≥ 1200 mg/d for longer than 12 months, respectively (Schjerning et al., 2016b).
This is interesting because those persons being dispensed pregabalin at higher than the maximum allowed dose most likely had other SUDs, too (Boden et al., 2014). Furthermore, it is not excluded that a good portion of the registered pregabalin dispensing is not used by the patients themselves and were diverted to others (e.g. family members, acquaintances) or to the black markets (Schifano et al., 2011; Kapil et al., 2014).
Safety of gabapentinoids
Unlike other substance of abuse, such as cannabis which euphoric high is scarcely influenced by tolerance due to repeated dosing (Wu and French, 2000), gabapentinoids are characterized by a rapid tolerance towards their desirable euphorization (Schifano et al., 2011; Calandre et al., 2016). This is suggested to drive a considerable overdosing (Schifano et al., 2011; Calandre et al., 2016). It is controversial how toxic overdoses of gabapentinoids can be (Table 3).
Overdosing of gabapentinoids in isolation appeared to be less toxic taking into account the cases reported to have ingested gabapentinoid amounts up to 25-fold their maximum therapeutic doses without developing serious sequelae (Fischer et al., 1994; Klein- Schwartz et al., 2003; FDA, 2004; Schauer et al., 2015). Otherwise, an intake of pregabalin in amounts 14 times higher than its maximum therapeutic dose was associated with coma and the necessity of mechanical ventilation (Wood et al., 2010). Gabapentinoid overdoses together with other substances are clearly more toxic (Table 3).
Thus, in the last years, there were increasing fatalities associated with pregabalin overdoses mostly in opioid using populations and infrequently also without opioids but in mixture with overdosed benzodiazepines or antidepressants (Table 3). Parenteral applications of gabapentin and pregabalin have been described (Schifano et al., 2011, Snellgrove, 2016) but seemed to be rare and, in the case of pregabalin, would not be associated with a stronger euphoria than achievable with an oral intake due to its high bioavailability of >90% (Calandre et al., 2016).
In comparison with propofol and benzodiazepines, gabapentinoids are certainly the safest GABAmimetics with a higher therapeutic index and wider dose margin between pleasure (euphoria/relaxation) and coma or death by overdosing (Bockbader et al., 2010; Bonnet, 2011; Calandre et al., 2016; Smith et al., 2016; Mersfelder and Nichols, 2016; Scherning et al., 2016a; Evoy et al., 2017; Quintero et al 2017).
Regarding the FTI as a new measure of harm, gabapentin and pregabalin are located clearly below that of opioids and most antidepressants and antipsychotics according to an analysis of the Finish national database of medicolegal autopsies and the Finish consumption figure of a particular substance (Ojanperä et al., 2016). Of note, the FTI of clonazepam was calculated to be lower than that of the
gabapentinoids (Ojanperä et al., 2016) which should await reproduction in alternative forensic samples. To this aspect, it should be considered that analyzing large registers is well suited to uncover trends but is less helpful to verify causal relationships. To illustrate this point, we can take a look at the deaths being reported in relation to drug poisoning in England and Wales basing upon a register from 2011 to 2015 (Office of National Statistics 2015).
This register reports increasing deaths related to gabapentinoids over the years but need to be interpreted with caution according to the limitations shown in the column of Table 3 which are all disclosed in detail on the official website (Office of National Statistics 2015). This quality of not being specific may explain the listed deaths related to cannabis and, most likely, a good portion of those being associated with gabapentinoids, too (Office of National Statistics 2015).
The increasing deaths related to gabapentin and pregabalin (Table 3) might have mainly reflected their increasing use resulting from their “bystanding” of another waxing potential lethal substance abuse, mostly that of opioids (Hser et al., 2017). However, considerable overdoses of gabapentinoids can be associated with respiratory depression (Spiller et al., 2002) and cardiac conduction disturbances (Rasima and Burkhart, 2006) which can make the difference whether a mixture with other potentially more life-threatening drugs is lethal or not.
“Anti-adverse selection” process or Pandora´s box in at-risk-populations?
One might speculate even on an “anti-adverse selection” process in the population of substance dependents, if assuming that overdosing gabapentinoids is far less life- threatening than overdosing other widely misused GABAmimetics, such as benzodiazepines, which are involved in 30% to 50% of overdose deaths related to opioid analgesics (Park et al., 2015). This means that gabapentinoids might have the potential to replace more toxic benzodiazepines on the prescriptions and black markets and thereby, would help to decrease hazards in the population of opioid and multidrug users: “more safe drugs for the riskiest patients” – however, on costs of “ground noises” of this population, such as inclusion of gabapentinoids in drug taking behavior and extreme overdosing – but even with fewer fatal consequences than with
benzodiazepines or other more toxic drugs. A similar process might have been happened between the late 1950ies and the late 1980ies when barbiturates had been largely replaced by benzodiazepines which appeared to have a markedly higher therapeutic index compared to barbiturates (Morgan, 1990; Coupey, 1997).
A presumptive signal for an anti-adverse selection might have come from the annual reports of the American Association of Poison Control Centers (Mowry et al., 2013, 2014, 2015, 2016). Therein, the total amount of registered fatality records had been fallen continuously from 2012 to 2015 whilst the number of gabapentinoid-related fatalities in drug mixtures became increased and this appeared to be much stronger attributed to gabapentin than to pregabalin (Table 3), although gabapentin is the lower toxic agent compared to pregabalin (Calandre et al., 2016).
A further argument could come from a 5 year cross sectional times series analysis in Canada which described a nearly 20-fold increase in the rates of pregabalin use from 2006 to 2014 and a simultaneous trend to a decrease of the co-use of other prescription drugs including benzodiazepines (Kwok et al., 2017). Further pharmacoepidemiological research is needed to verify this hypothesis as well as the opposite view that gabapentinoids possess features of Pandora´s box in risk populations, such as patients with substance abuse history (Prescrire Int, 2012; Häkkinen et al., 2014; Chiappini and Schifano, 2016; Eastwood and Davison, 2016; Elliott et al., 2017).
Comparing the addictive risks of gabapentinoids with those of traditional substances of abuse
Unlike traditional psychoactive drugs, there is less evidence for gabapentinoids to be misused in a long-term manner and to be associated with tenacious relapses and social hazards (as outlined above). This would support the hypothesis that gabapentinoids can induce a “liking” (euphoric/relaxing high) due to their GABAmimetic action but no or only minimal “wanting” (Berridge and Robinson, 2016) which corresponds with no or only minimal rewarding properties of gabapentinoids in animal experiments (c.f. 3.1). Non-treatment seeking cocaine abusers did not alter their choice to self-administer cocaine while on gabapentin maintenance (up to 3200
mg/day) (Hart et al., 2007) which also did not emphasize a robust addictive power (“wanting”) of gabapentinoids. However, the inhibition of presynaptic α2δ-subunit- containing VGCCs could even hinder the transfer (sensitization?) from “liking” to “wanting” (Figure 2) through “anti-sensitizing” (Dooley et al., 2000; Eroglu et al., 2009) and thereby, “anti-wanting” actions (Shibasaki et al., 2009; Kurokawa et al, 2011) (Figure 2). Of note, there is first evidence for an up-regulation of α2/δ subunits in the rodent limbic brain via repeated dopamine surges, e.g. mediated by methamphetamine administrations (Kurokawa et al., 2010, 2011).
Traditional drugs of abuse are characterized by a dominant “wanting” according to Berridge and Robinson´s “Incentive Sensitization Theory of Addiction” (Berridge and Robinson, 2016) (Figure 2). There are several further non-regulated medications being deemed to have less or no relevant addictive power (“wanting”), which were observed to be abused preferentially by patients with a history of another substance abuse disorder. These include some antidepressants (tranylcypromine, bupropion, tianeptine) (Haddad, 1999; Bernard et al 2011, Costa et al 2015), antipsychotics (quetiapine) (Montebello and Brett, 2015; Reeves and Burke, 2014; Reeves and Ladner, 2014) or pain relievers (flupirtine) (Gahr et al., 2013c).
On this background, we assume that gabapentin and pregabalin belong to those substances, which themselves have no relevant addictive power suis generis (“wanting”), but could become addictive in patients with prior substance abuse experiences These patients look for various options broadly being able to induce euphoria and to improve their affected stress reactivity that occur following substance addiction (Karoly et al., 2015; Volkow and Morales, 2015; Fosnocht and Brand, 2016).
Unlike for nicotine, alcohol or opiates, there is yet no evidence for gabapentinoids to facilitate the misuse of other drugs in patients without a history of SUDs. Although gabapentinoids are widely distributed, we have found only 4 cases supporting a “wanting” of gabapentinoids in patients without an abuse history and these 4 cases referred to pregabalin and not to gabapentin (Yargic and Ozdemiroglu, 2011; Ashwini et al., 2015; Halaby et al., 2015; Driot et al., 2016; Table 1). This is still
below the frequency of cases having reported extremely rare clinical phenomena, such as the toxicologically relevant gut fermentation syndrome (“endogeneous auto- brewery”) (Cordell and McCarthy, 2013; Welch et al., 2016). Thus, we assume that gabapentinoids are usually not related to the development of behavioral dependence (“wanting”) in patients without prior experiences with traditional drugs of abuse. Unlike gabapentin, dispensing of pregabalin is regulated by law in two countries (USA, Norway) and actually listed in schedules which control psychoactive drugs with the lowest risk of abuse (Calandre et al 2016; Westin and Strøm, 2010). In Germany, prescription of gabapentinoids is hitherto not regulated by his narcotic law, just as in most other countries of the world.
Considering the safety concerns and addictive power of gabapentin and pregabalin as outlined above, we propose Table 4 to compare the addictive risks of gabapentinoid use with those of traditional substances of abuse.
Table 1: Clinical and epidemiological studies about gabapentinoid-abuse or – dependence
| Study | Population | Methods | Results | Addiction history and Psychiatric comorbidity |
| Internet and hospital population | ||||
| Kapil et al., 2014 England and Wales | 16-59-year-old internet users, N=1500, 50.9% females | online survey about the use of use of recreational substances including gabapentin and pregabalin | life-time prevalence of misuse: baclofen 1.3% (N=19), gabapentin 1.1% (N=17), pregabalin 0.5% (N=8) misuse of these GABAergic drugs more than once weekly in 13.1% (N=5) | unknown |
| obtained from various sources, e.g. family or acquaintances (57.8%, N= 22) or from the Internet (47.3%, N=18) or abroad (7.8%, N=3) or legitimated prescriptions 13.1%, N=5) | ||||
| Cossman n et al., 2016 Germany | ≥ 65-year-old inpatients of a German general hospital in a metropolitan area, “Ruhrgebiet” 2013 (N=400) | structured DSM- IV-based face- to-face interview (SKID-I) focusing on the use of psychoactive substances including gabapentin and pregabalin | 12-months prevalence 0% and life-time prevalence 0,25 % for gabapentinoid-abuse and dependence, N=1 with transient dependence on gabapentin, no case of pregabalin abuse or dependence | the male gabapentin- dependent was prior and currently dependent on an opioid pain reliever |
| Addicted populations or recreational drug users | ||||
| Schifano et al., 2011 Europe | anonymous recreational drug users | evaluation of 108 internet web-sites with reference to gabapentinoid and clonazepam use experiences, 8 European languages | Pregabalin was attributed to be an „ideal psychotropic drug for recreational purposes.” incentives: to achieve alcohol/GHB/benzodiazepine- like effects mixed with euphoria; entactogenic feelings and DXM-like disassociation, and to cope with opiate/opioid withdrawal, similar effects with gabapentin but pregabalin | unknown |
| seemed to be preferred (“far less dosage to get the same recreational high”) routes of misuse: mostly oral, but intravenous, rectal (‘plugging’), and ‘parachuting’ (emptying the content of the capsule into a pouch) were reported, too usual pregabalin dosages: 600 to 5000 mg/d, rapid tolerance | ||||
| Piralishvili et al., 2013 Georgia | adults in opioid substitution programs, Tbilisi, N=506 (two females) | self-report questionnaire of psychoactive substances | non-medical use: 8.17% pregabalin | all dependent on opioids at least |
| Grosshan s et al., 2013 Germany | N=124 (34 female), 37.1±8.1 years old (range 20– 55) in opioid substitution program vs control group: N=111 patients treated for non- opiate addiction, mostly alcohol or cannabis, Mannheim | pregabalin urine screen | non-medical pregabalin use: 12.1% of the patients substituted with opioids. In the control group were 2.7% positive (all had been prescribed pregabalin for medical purposes) | all had another substance use disorder at least |
| Baird et al., 2013 Scotland | adults dependent on substances attended to 6 detoxification clinics in the Lothian region, N=129 | anonymous self- report questionnaire about psychoactive substances performed in 6 substance misuse clinics | non-medical use: N=25 (19%) gabapentin, N=4 (3%) pregabalin, N=19 (15%) methadone, N=61 (47%), buprenorphine N=2 (2%), cannabis N=55 (43%), heroin N=7 (5%). All patients admitting to using non- prescribed gabapentinoids (29/129 = 22%) were in opioid substitution programs incentives: to get “high”, to potentiate methadone effects | all dependent on opioids at least |
| McNamar a et al., 2015 Ireland | adults being in 6 different opioid substitution programs, N=440, 21 to 61 | drug urine screen including pregabalin in the National Drug Treatment Centre’s (NDTC) | 31 patients (7%) were found to use pregabalin for non-medical purposes, significantly more females (59%) | all dependent on opioids at least |
| years old, 34% females | Drug Analysis Laboratory | |||
| Smith et al., 2015 USA | adults using non-medical prescription opioids, Appalachian Kentucky, N=503, 78% females | prospective study, self- report questionnaire of psychoactive substances | 6-months prevalence of non- medical gabapentin-use: 15%; – which represented a 165% increase in use compared to reports from one-year prior, and a 2950% increase since 2008 within this cohort. incentive: to get high gabapentin use in 25 of the past 30 days was associated with abusing immediate- release oxycodone, buprenorphine, benzodiazepines, females, and reporting chronic medical conditions The two major sources of gabapentin were physicians (52%) and drug dealers (36%) | all dependent on opioid pain relievers |
| Wilens et al., 2015 USA | adult opioid dependent inpatients detoxifying voluntarily at the Massachusetts General Hospital, Boston, N=162, 49% female, 33 ±10 years old | self-report questionnaire of psychoactive substances | 28% reported using higher amounts of each medication than prescribed. Of opioid patients, 10% self-reported misusing clonidine, 22% gabapentin, 7% pregabalin, 25% clonazepam, and 11% amphetamine | all opioid dependent, at least anxiety 72%, ADHD 32%, depression 64%, PTSD 38%, bipolar disorder 28% |
| Bastiaens et al., 2016 USA | former inmates, living in a correctional community center, Pittsburgh, N= 250 adults, 37.2±12.1 years old, 36% females, 72% white | written questionnaire about the non- medical use in the past: opiates, gabapentin, buproprion, quetiapine, and fluoxetine | 16% (N=41) reported having misused gabapentin in the past (quetiapine 21.8%, bupropion 6%, fluoxetine 0%). Of patients with an opioid use disorder (N=145), 26 % endorsed gabapentin abuse while 4 % of patients without an opioid use disorder (N=105) endorsed the non-medical use of gabapentin (p<0.0001) 26% of opiate dependent patients reported illegally obtaining, overusing, or malingering problems to obtain gabapentin only one of 30 patients (3 %) | all patients had further substance use disorders, 72 % two substance use disorders, at least depressive disorders (25 %), ADHD (25 %), adjustment disorders (22 %), anxiety disorders (18 %), PTSD (11 %), bipolar disorder (8 %), schizophrenia (3 %) |
| with a cocaine use disorder and not comorbid with an opioid use disorder reported abusing gabapentin. | ||||
| Snellgrov e, 2015 Germany. | adults voluntarily treated in a ward for qualified detoxification, Ulm (N=253), 33,1±8,8 years old, 21% female, 100% white | structured DSM- IV-based face-to face interview about pregabalin use, visual analogue scales of pregabalin effects, pregabalin urine screens | pregabalin use in the last 30 days: 26% (N=65), lifetime: 56% (N=142) 24-months-prevalence: 7% und point prevalence: 3% for pregabalin-dependence, every 10th patient had a false positive memory about a recent pregabalin use obtained from family or acquaintances for free (44% of all pregabalin users), drug dealers (41%), physicians (30%) minimum and maximum dose = 100 and 4500 mg/d in the last 30 days, respectively, median: 600 mg/d incentives: to potentiate opioid high, to dampen anxiety or withdrawal symptoms 9% of patients with pregabalin experience used pregabalin over 12 months at ≥ 25 days/month | all dependent on other substances, mostly opioids or benzodiazepines, cannabis use = negative predictor for pregabalin use co-use. with opioids (39%), sedatives (20%), alcohol (19%) |
| Alblooshi et al., 2016 United Arab Emirates | male adults with substance use disorders from seven emirates/princip alities admitted to the National Rehabilitation Centre, Abu Dhabi 2015; N=250, mean age: 29.6 years, minimum: 18 years, maximum: 62 years | structured face- to-face interview about substance use | 84.4% were polysubstance abusers, mostly opioids and alcohol. Of whom, over 60% used prescription drugs non- medically, such as meprobamate, prociclidine, and pregabalin. Pregabalin was used by 68% in this group which were mainly below 30 years old. Pregabalin (commonly 7 to 14 capsules á 75 or 150 mg/day) was used in 41% of these cases together with other prescription drugs and in 27% not together with other prescription drugs | all patients with another substance use disorder at least |
| Mutschler et al., 2016 | adult patients in a methadone treatment | self-report measure focusing on | no case with pregaballin use could be identified by self- report and hair analysis | all dependent on opioids at least |
| Switzerla nd | program, Zurich (N=109) | pregabalin use, hair toxicology analysis | ||
| Heikman et al., 2016 Finland | adult patients in a methadone substitution program, Helsinki (N=(82) | retrospective study of 200 urine samples collected consecutively between 10/2013 and 4/2014 drug urine screen including pregabalin and gabapentin | gabapentin and pregabalin in 1 (0.5%) and 8 (4%) of the samples; the gabapentin- positive sample showed also benzodiazepines, amphetamine and cannabis; no information about concurrent drugs in the pregabalin-positive samples, no information about potential prescriptions | all had another substance use disorder at least |
Table 2: Differentiating case reports related to abuse and dependence of gabapentiniods according to the dependence-criteria of ICD-10a, addiction history, self-administration, long-term use, relapses
| Case reports | Cravin g (BD) | Loss of contro l (BD) | Narrowe d behavior b (BD) | Toleranc e (PD) | Withdraw al symptoms (PD) | Harmfu l use (PD) | Number of fulfilled depende nce criteriaa | History of other substance abuse or dependence apart from nicotine | Information about gender, comorbidity and patterns of consumption or relapses |
| Gabapentin (N=36) | |||||||||
| Markowitz et al., 1997 USA | U | U | U | None | None | unkno wn | 0 | Cocaine | 41-year-old woman posttraumatic stress disorder 400-1600 mg/d for three months to reduce cocaine craving, self- administration |
| Corá- Locatelli et al., 1998 USA | U | U | U | U | Yes | U | 1 | Negative | five patients with obsessive- compulsive disorder and mood disorder receiving 900- 3600mg/d for months |
| Rosebush et al., 1999 Canada | U | U | U | U | Yes | U | 1 | Negative | 48-year-old man bipolar disorder prescription of 500mg/d for 4 weeks |
| Norton, 2001 USA | U | U | U | U | Yes | U | 1 | Negative | 29-year-old man bipolar disorder prescription of 4800mg/d over 6 weeks |
| U | U | U | U | Yes | U | 1 | Negative | 36-year-old man bipolar disorder, chronic back pain prescription of 3600mg/d for 2 months | |
| U | U | U | U | Yes | U | 1 | Negative | 28-year-old man migraine prescription of 2400mg/d for 6 months | |
| Barrueto et al., 2002 USA | U | U | U | Yes | Yes | Yes | 3 | Negative | 34-year.old male chronic back pain 8000 mg/d for 9 month to reduce pain, self- administration |
| Drabkin and Calhoun, 2003 Canada | U | U | U | Yes | U | U | 1 | Negative | 35-year-old woman bipolar disorder prescription of 3600mg/d for 2 months |
| Reccoppa et al., 2004 | U | U | U | U | U | U | 0 | Cocaine | five 29 to 45- year-old male inmates snorting the |
| USA | powder of opened gabapentin 300 and 400mg capsules to get high bipolar disorder, anxiety disorder, antisocial personality disorder, impulse control disorder, neuropathic pain self- administration | ||||||||
| Tran et al., 2005 USA | U | U | U | U | Yes | U | 1 | Negative | 81-year-old woman schizoaffectiv e disorder, Parkinson´s disease, cerebrovascul ar disease prescription of 800mg/d over 5 years |
| Victorri- Vigneau et al., 2007 France | U | U | Yes | Yes | Yes | U | 3 | Alcohol | 67-year-old woman mood disorder up to 7200 mg/d, self- administration , relapse |
| Pittenger et al., 2007 USA | U | U | U | U | Yes | U | 1 | Alcohol, cocaine, opioids | 33-year-old man 3600 mg/d gabapentin over a few weeks, self- administration |
| U | U | U | Yes | Yes | Yes | 3 | Alcohol | 63-year-old man chronic back pain up to 4900 mg/d over |
| several months, self- administration | |||||||||
| Kruszewski et al., 2009 USA | Yes | U | U | Yes | Yes | Yes | 4 | Alcohol | 33-year-old male physician depression, anxiety high self- administered doses over several weeks |
| Hellwig et al., 2010 USA | U | U | U | U | Yes | U | 1 | Alcohol | 53-year-old woman liver cirrhosis 700 mg/d, prescribed over a few weeks |
| Finch et al., 2010 USA | U | U | U | U | Yes | U | 1 | U | 41-year-old man with an orthotopic liver transplant neuropathic pain prescription of gabapentin for several weeks |
| See et al., 2011 USA | U | U | U | U | Yes | U | 1 | Negative | 76-year-old woman depression, neuropathic pain prescription of 3600mg/d over 1 month |
| Di Fabio et al., 2013 Italy | U | U | U | U | Yes | U | 1 | Negative | 76-year-old woman bipolar disorder, neuropathic pain prescription of 900mg/d over two years |
| Mah and Hart, 2013 Canada | U | U | U | U | Yes | U | 1 | Negative | 75-year-old woman recurrent depression, |
| fibromyalgia, postherpetic neuralgia long-term prescription of 1800mg/d | |||||||||
| Reeves and Burke, 2014 USA | U | U | U | U | U | Yes | 1 | Cannabis, cocaine | 42-year-old man Taking up to 5 tablets of gabapentin 300 mg with 3 to 4 tablets of quetiapine 200 mg simultaneousl y to produce a sensation of sedation and euphoria, self- administration |
| Reeves and Ladner, 2014 USA | U | U | U | U | U | U | 0 | Opioid- maintenance program | 38-year-old man taking buprenorphin e/naloxone simultaneousl y with up to 1000 mg of quetiapine or with up to 2400 mg of gabapentin. to get relaxed and euphoric which was experienced as not as intense as during opiates, self- administration |
| Satish et al., 2015 India | Yes | Yes | yes | yes | yes | U | 5 | opioids | 26-year-old man ADHD substitution of a 3-years- lasting propoxyyphen e dependence with gabapentin (up to 12 g/d alongside 2 years), self- administration gabapentin high was different from the opioid |
| high | |||||||||
| Bonnet and Scherbaum, 2016 Germany | None | None | None | Yes | Yes | None | 2 | Negative | six 35-76- year-old patients, among them 4 females anxious depression or generalized anxiety disorder prescription of gabapentin (1200-3200 mg/d) over several weeks |
| Pregabalin (N=19) | |||||||||
| Oaklander and Buchbinder, 2010 USA | U | U | U | U | Yes | U | 1 | Negative | 80-year-old woman postherpetic neuralgia prescription of 375 mg/d for 49 weeks |
| Grosshans et al., 2010 Germany | Yes | U | Yes | Yes | Yes | U | 4 | Alcohol, cannabis, heroin | 47-year-old man self- administration of up to 7500mg/d to get high, relapse serum level 29 mg/l (therapeutic range: (2–8 mg/L) |
| Filipetto et al., 2010 USA | U | U | Yes | Yes | Yes | U | 3 | Opioids | 35-year-old woman neuropathic pain, anxious depression self- administration of 88500mg over a 28-day period |
| Westin and Strøm, 2010 Norway | U | Yes | U | Yes | Yes | U | 3 | Negative | 30-year-old woman generalized anxiety disorder, insomnia |
| self- administration to get euphoric – up to 1800mg/d for two years after a prescription of 600mg/d | |||||||||
| Yargic and Ozdemirogl u, 2011 Turkey | U | None | Yes | Yes | Yes | U | 3 | Benzodiaze pines, ketamine | 37-year-old man bipolar disorder, grand mal epilepsy self- administration of up to 3000 mg/d to get euphoric, over a period of 2 month, afterwards the patient learned to take therapeutic doses |
| Karosin et al., 2012 Austria | U | U | U | U | Yes | U | 1 | Negative | 28-year-old woman poststroke neuropathic pain prescription of 150 mg/d for 3 weeks |
| Skopp und Zimmer, 2012 Germany | U | Yes | Yes | Yes | U | U | 3 | Opioid maintenance program, cocaine, benzodiazep ines | adult man taken into police custody due to maniform irritation and aggression self- administration serum level: 25mg/l pregabalin – in combination with diazepam and methadone |
| Carrus and Schifano, 2012 South- | U | U | U | Yes | Yes | U | 2 | Benzodiaze pines, ecstasy, cocaine, cannabis | 32-year-old man antisocial personality disorder, |
| Europe | neuropathic pain self- administration of up to 4500mg/d over a period of 4 weeks to get energetic, empathic and relaxed – after an initial prescription of 600mg/d | ||||||||
| Yes | U | U | U | Yes | Yes | 3 | Ecstasy, alcohol, cannabis | 33-year-old man generalized anxiety disorder, bipolar disorder self- administration of up to 1500 mg/d within 4 weeks to get euphoric and relaxed “as with cannabis” – after an initial prescription of 300 mg/d | |
| Aldemir et al 2013 Turkey | Yes | U | Yes | Yes | Yes | Yes | 5 | Alcohol, cannabis, ecstasy, cocaine | 34-year old woman anxiety self- administration of up to 16000mg to feel more energetic and self-confident, visual hallucinations, for 3 years |
| Papazisis et al., 2013 Greece | U | U | Yes | Yes | U | U | 2 | Cannabis, alcohol | 19-year-old man generalized anxiety disorder self- administration of up to 1800 mg to get high, longer than 3 months after a |
| prescription of 600 mg/d, no motivation to detoxify | |||||||||
| Gahr et al., 2013b Germany | U | U | Yes | U | yes | U | 2 | Alcohol | 38-year-old woman bipolar disorder, borderline personality disorder, unspecific anxiety self- administration of 800 mg/d to get euphoric, for 4 months – after an initial prescription of therapeutic doses |
| Barrett et al., 2015 USA | U | Yes | U | U | Yes | Yes | 3 | Opioid painkillers | 61-year-old man neuropathic pain self- administration of more than the prescribed 300mg/d for 8 months |
| Ashwini et al., 2015 India | U | U | Yes | Yes | Yes | Yes | 4 | Negative | 30-year-old man neuropathic pain self- administration of up to 3000 mg/d, for 4 years, brought by his mother to admission due to suicidal ideations |
| Yazdi et al., 2015 Austria | U | Yes | Yes | Yes | Yes | Yes | 5 | Benzodiaze pines, alcohol, tramadol | 29-year-old man major depression, generalized anxiety disorder self- administration of up to 3000 |
| mg/d in combination with benzodiazepi ne or tramadol or quetiapine overdoses,, three admissions due deep sedation/mixe d intoxication, two relapses within a year | |||||||||
| Gahr et al., 2015 Germany | Yes | U | Yes | Yes | U | U | 3 | Benzodiaze pines | 33-year-old woman anxiety, borderline personality disorder self- administration of up to 3000 mg/d to get euphoric and relaxed, for 6 month |
| Halaby et et 2015 Lebanon | Yes (subsid ed within 6 weeks of abstine nce) | U | U | Yes | Yes | Yes | 4 | Negative | 26-year-old woman bipolar disorder, generalized anxiety disorder self- administration to calm negative feelings – up to 2400mg/d within 4 months |
| Nordgaart and Jürgens 2015 Denmark | U | U | U | Yes | Yes | U | 2 | Alcohol, opioid maintenance program | 38-year-old man generalized anxiety disorder self- administration of up to 8400 mg/d, bought 21 times |
| Driot et al 2016 France | Yes | U | Yes | Yes | Yes | U | 4 | Negative | young woman anxiety, depression, anorexia, |
| personality disorder prescription of < 300 mg/d – synergistic stimulating effects together with secondary tobacco use |
Notes: aDilling and Freyberger, 2004,b e.g. gabapentinoid seeking behavior.
Abbreviations: PD=physical dependence, BD=behavioral dependence, U=unknown; abuse = harmful use, dependence = ≥ 3 criteria within a year (Dilling and Fryberger, 2004)
Fulfilled BD-criteria in bold letters, negative addiction history is accentuated by shadowed boxes, relapses are highlighted in bold letters in the column on the right edge of the table
Table 3: Overdoses and fatalities associated with gabapentinoids
| Study | Population | Methods | Results | Addiction history and Psychiatric comorbidity | Limitations |
| Non-fatal overdosing | |||||
| Fischer et al., 1994 USA | 16-year-old male swallowed father´s gabapentin, approximately 50 g | case study | sedation, stable vital signs, plasma level 62 mg/L | cocaine | no estimation of period between ingestion and determination of blood levels |
| Fernandez et al., 1996a USA | 31-year-old man, self-poisoning with large amounts of valproic acid and gabapentin | case study | coma, shock, plasma concentrations: valproic acid 1306.9 mg/L (peak), gabapentin 60 mg/L, resolved during aggressive supportive care and concurrent hemoperfusion and hemodialysis to enhance elimination of valproic, negative urine drug screen | epilepsy | |
| Fernandez et al., 1996b | 32-year-old man, self-poisoning with approximately | case study | dizziness, somnolence, nystagmus, slurred speech, | unknown |
| USA | 91 g gabapentin and 54g valproic acid, beer and whiskey, he denied other coingestions | stable vital signs, normal respiratory rate of 24/minute, negative urine drug screen; plasma concentrations valproic acid 139.9 mg/L, gabapentin 44.5 mg/L | |||
| Stopforth, 1997 South Africa | 17-year-old girl swallowed approximately 40 gabapentin capsules á 300mg (12 g) together with 20 lamotrigine tablets á 100 mg | Case study | sedation, drowsiness, ataxia and lethargy, low serum potassium levels | epilepsy | no blood levels |
| Verma et al., 1999 USA | 30-year-old- female with epilepsy and chronic renal failure was treated with valproic acid 1250 mg/d which was augmented with 1800 mg/d gabapentin | case study | mild resting tremor, slight difficulty with naming objects and difficulty with performing three step commands, gabapentin plasma level 85 mg/L, valproate may have contributed to the symptoms | lupus erythematode s, at the age of 24 years left middle cerebral artery stroke which resulted in a right hemiparesis and expressive aphasia, chronic renal failure requiring hemodialysis three times per week, generalized tonic clonic seizures at the age of 27 years | |
| Klein- Schwartz et al., 2003 USA | pure gabapentin intoxications 20 patients ranging from 12 months to 83 years old, estimated doses of gabapentin ingestion | case series, 2-year multicenter prospective observational study of all gabapentin exposures reported to three poison | ten cases involved children and adolescents, 9 of the 20 cases were symptomatic with drowsiness, dizziness, nausea and vomiting, | unknown; of gabapentin exposures, 65% were acute-on- chronic, indicating that most cases involved the patient´s own | no information about blood levels, no conclusive information about addiction history and comorbidity |
| between 50 mg (child) and 35 g (48 years old) | centers from 4/1/1998 to 4/1/2000 | tachycardia, hypotension, none of the patients were admitted to medical care | medication | ||
| FDA, 2004 USA | premarketing clinical trials | medical review about pregabalin | six patients with reported ingested doses from 1500 to 8000 mg showed typical adverse effects (confusion, somnolence, dizziness, ataxia, diplopia, blurred vision), no death – one patient with an reported intake of 15g which had “resulted in no consequences” | unknown | no conclusive information about mixtures, comorbidity or serum levels |
| Spiller et al., 2002 USA | 61-year-old female; self- poisoning with gabapentin and quetiapine | case study | massive gabapentin and presumptive quetiapine overdose with a recorded serum gabapentin concentration of 104.5 mg/L associated with coma, respiratory depression requiring mechanical ventilation, and hypotension | unknown | no information about addiction history |
| Rasimas and Burkhart 2006 USA | 44-year-old female multiple prescription drug overdose plus alcohol and cannabis | case study | lethargy, drowsiness, tachycardia, hypotension, ataxia, dizziness, never lost consciousness, QRS widening and QTc prolongation attributed to a reported self- induced overdose with gabapentin and | alcohol, cannabis | no information about the ingested gabapentin and nefazodone amounts |
| nefazodone, serum levels of both substances obtained 2 hours post-ingestion were both within the therapeutic range (gabapentin 7 mg/L) | |||||
| Braga and Chidley, 2007 UK | 29-year-old man, self-poisoning with 32 g lamotrigine and 11.5 g pregabalin | Case study | agitated, unresponsivenes s, hemibalistic movements, facial grimacing, tachycardia, urinary retention, stable vital signs, pregabalin´s plasma level 45 mg/L | epilepsy | No conclusive information about addiction history and psychiatric comorbidity |
| Wood et al., 2010 UK | 54-year-old male, self- poisoning with 8.4 g of pregabalin, denied ingestion of other drugs | case study | significant neurological depression and coma approximately 3 h post-ingestion, remained cardiovascularly stable, serum concentration of pregabalin 66.5 mg/L; supportive care including mechanical ventilation | HIV, peripheral neuropathy, diabestes mellitus | no conclusive information about addiction history |
| Zacny et al., 2012 USA | healthy volunteers between the age of 31 and 39 years | double-blind, randomized, crossover design; in separate sessions, participants were exposed to 75 and (supratherape utic at first dosing) 150 mg pregabalin, 10 mg oxycodone and 75 mg pregabalin & 10 mg oxycodon | respiration rate (placebo: 12.9±0.6): significantly reduced with pregabalin 75 mg (10.3±0.7) and 150 mg (10.6±0.8), oxycodone 10 mg (10.5±0.7), oxycodone 10 mg & pregabalin 75 mg (9.6±0.5) pregabalin did not impact on psychomotor performance or | 15 volunteers reported use of cannabis, and some subjects also the use of stimulants, club drugs, hallucinogens, opioids | no information about the results of urine drug screens, although in the methods section is mentioned that upon arrival, breath alcohol, urine toxicology, and pregnancy (for females) tests had |
| liking or drug taking desire experimental combining 150 mg pregabalin with oxycodone 10 mg was not further studied because two subjects reported excessive drowsiness | been given no information about serum levels of the tested drugs | ||||
| Miljevic et al., 2012 Serbien | 54-year-old man, self-poisoning with 4.2 g pregabalin together with 21 mg bromazepam and 125 mg clomipramine | case study | conscious, alert, stable condition of cardiovascular and respiratory systems serum level: 20.8 mg/L | generalized anxiety disorder | no information about addiction history |
| Schauer et al., 2013 USA | 59-year-old man, self-poisoning with approximately 90 g gabapentin | case study | nausea, mild sedation, stable vital signs, normal cQT- interval negative urine drug screen gabapentin plasma level 72.8 mg/L, approximately 3 hours after ingestion | unknown | no information about addiction history and comorbidity |
| Millar et al., 2013 Northern Ireland, UK | 10 young adults presented to a Belfast emergency department following recreational pregabalin abuse | observational retrospective study | reported dosages ranged from 500–1400 mg pregabalin. Six (60%) patients presented with seizures (5 of which were ‘first’ seizures). Two patients (20 %) required intubation and ventilation and were admitted to the Intensive Care Unit. | unknown | no information about addiction history, comorbidity and the contribution of concurrent drugs of abuse or pharmaceutic als |
| Koschny et al., 2014 | 21-year-old female with | case study | asystole, cardiopulmonary | no information | no conclusive information |
| Germany | polyintoxication (approximately 1.75 g carvediol, 300 mg amlodipine, 6g amitriptyline, 500 mg torasemide, 1.5 g ketoprofen, 28 g nicotinic acid, 16 g gabapentin), self-poisoning | resuscitation, coma gabapentin´s peripheral blood concentration: 126.8 mg/L to facilitate drug removal, therapeutic plasma exchange was performed, after extubation without neurologic sequelae | about addiction history and comorbidity | ||
| Kriikku et al 2014 Finland | drivers apprehended for driving under the influence of drugs in 2012, N=206 pregabalin positive samples | retrospective review | pregabalin serum concentration- range: 0.68- 111.6 mg/L, median: 6.2 mg/L, nearly 50% cases had serum concentrations above the recommended therapeutic range (2.7-8.5 mg/L); no further substance besides pregabalin in three cases (0.7 mg/L, 17.9 mg/L, 18.2 mg/L) | additional use: benzodiazepin es (91%), cannabis (54%), amphetamines (44%), opioids (40%), alcohol 20%, in 43.2% > 5 substances simultaneousl y | no information about comorbidity or prescribed medications |
| Wills et al., 2014 USA | poison center data evaluating clinical outcomes from newer anticonvulsant overdoses | retrospective study using the ToxicallTM database from 1/1/2002 to 31/12/2011 | 94 cases with gabapentin (maximal dose 96g, median 6g) and 18 cases with pregabalin (maximal dose 9 g, median 2,4 g), no severe outcome, no fatality | unknown | no information about blood levels, addiction history, comorbidity |
| Fatalities | |||||
| Moore et al., 2005 USA | 54-year-old female, self- poisoned with metaxalone and | case study | postmortem central (heart) blood concentrations: | no drugs of abuse were found |
| gabapentin | gabapentin 24 mg/L, metaxalone (21 mg/L) and therapeutic concentrations of acetaminophen and citalopram | ||||
| Button et al., 2010 England, UK | 49-year-old female, 50-year- old male, self- poisoning with pregabalin | case-study, unpreserved femoral blood | pregabalin serum concentrations: 25.3 mg/L, 180 mg/L | additional use: benzodiazepin es, Z-drugs, opioids; antidepressant | causality douptful due to a mixture of contributing substances and missing information about comorbidities |
| Launiainen et al., 2011 Finland | population- based sample of deceased young adults aged 15- 34 years, 75% male, November 2006 to October 2008, N=1623 | register study, review of postmortem toxicology, background information from case referrals was used to distinguish the abuse of medicines from their therapeutic use | postmortem analyses found pregabalin positive in 68 cases (42 with abuse); for comparison: cannabis was positive in 221 cases (221 with abuse), morphine 149 (64), codeine 148 (55), tramadol 154 (84); any drug 677 (509) | blood alcohol was positive in 52% of the cases | no conclusive information about mixtures, comorbidity or serum levels; cases of known opioid substitution treatment and cases classified as suicides with the drug in question were not included |
| Middleton, 2011 USA | 82-year-old female, self- poisoning with gabapentin | case study | postmortem peripheral blood concentration of gabapentin 88 mg/L, clonazepam concentrations within the therapeutic range,residual gastric gabapentin dose was estimated to be 2210 mg | depression, several previous episodes of suicidal ideation, obesity, cardiomegaly | no conclusive information of period between ingestion and determination of blood levels, autopsy was performed ca. 22 h after her discovery |
| Lottner- Nau et al., 2013 Germany | autopsies of drug-related deaths within 2 years (October 2010 – | register study, retrospective observational study | pregabalin was found in 43 (4.4%) cases. The concentration | additional illicit and licit drugs in each case, opioids (100%), | causality douptful due to a mixture of contributing |
| September 2012), Munich, N=982 | range in femoral or heart blood was between 0.04 mg/L and 22.8 mg/L, median 5.18 mg/L | benzodiazepin es (77%), neuroleptics (33%), alcohol (30%) | substances and missing information about comorbidities | ||
| Llauniainen et al., 2014 Finland | population- based sample of 57903 Finish autopsy cases from the period between 01/01/2000 and 31/12/2010. Among them, there were 135 (0.23%) and 380 (0.66%) autopsy cases in which gabapentin and pregabalin were involved | register studies | 8 of the 135 (6%) records were related to gabapentin to be the main cause of death and 12 of the 380 (3.2%) records were related to pregabalin in this sense – as taken from death certificates | unknown | no conclusive information about mixtures, comorbidity |
| Priez- Barallon et al., 2014 France | 18 cases of deaths in which pregabalin was involved | case series | no significant differences between central and peripheral blood pregabalin concentrations, concentrations in peripheral blood ranged between 0.4 and 206.7 mg/L, pregabalin was suggested to be a likely factor in the cause of death in 3 cases, which used also opioids | unknown | no conclusive information about addiction history and comorbidity |
| Häkkinen et al., 2014 Finland | all medico-legal death cases in Finland in which gabapentin (N=43) and pregabalin (N=316) was found in postmortem toxicology from 2010 to 2011 | register study, 2-year retrospective observational study | median femoral blood concentrations of pregabalin were 15 mg/L in the abuser group and 5.8 mg/L in the other cases. For gabapentin, those concentrations were 12 mg/L (abuser group) and 8.3 mg/L (non-abuser | positive addiction history and drug abuse in 48.1% of pregabalin and 18.6 % of gabapentin cases; 91.4 of pregabalin abusers and 87.5 of gabapentin abusers had concomitant |
| group) | opioid use, other psychoactive substances were found in the remaining cases | ||||
| “only” pregabalin case, 31-year- old male, no valid pregabalin prescription | case study | femoral blood concentrations: pregabalin 110 mg/L – and ethanol (0.24%); traces of quetiapine and levomepromazin e; benzodiazepines were within therapeutic ranges | unknown addiction history, coronary disease | ||
| “pregabalin and opioid” case, 26- year-old male, no valid pregabalin prescription | case study | femoral blood concentrations: pregabalin 48 mg/L, concentrations of buprenorphine and benzodiazepines were within the therapeutic ranges | abuser of buprenorphine and amphetamine | ||
| Cantrell et al., 2015 USA | 47-year-old female deceased after the ingestion of approximately 26 tablets of 600 mg (15.6 g) gabapentin | case study | postmortem peripheral(femor al) gabapentin blood concentration 37 mg/L, central (heart) blood 32 mg/L (therapeutic concentrations 2- 20 mg/L), liver 26 mg/kg, vitreous 32 mg/L, gastric contents 6 mg concurrent medication: ibuprofen in therapeutic doses | chronic pain, obesity, no history of alcohol or illicit drug use, no other drugs including prescribed hydrocodone were detected | |
| Ojanperä et al., 2016 Finland | 39 (0.2%) and 6 (0.03%) fatalities were related to pregabalin and | register study, postmortem toxicology cases, | Mean FTI: pregabalin 1.92, gabapentin 0.91 | unknown | no information about possible |
| gabapentin, respectively, out of 19670 drug- related deaths | database review using the fatal toxicity index (FTI), expressed as the number of deaths per million DDD (defined daily doses) in 3 years, 2005, 2009, 2013 | increasing trend in FTI over the years for pregabalin (0.54, 1.54, 2,44) superior mean FTIs: methadone 42.65, dextropropoxyph ene 31.84, levomepromazin e 21.92, doxepine 13.99, chlorprothixene 7.11, oxycodone 6.76, amitriptyline 6.54, trimipramine 6.32, tramadol 5.69, sulpiride 4.66, propranolol 3.83, quetiapine 2.51, trazodone 2.44 | mixtures and comorbidities | ||
| Eastwood and Davison, 2016 UK | 70 cases over a two year period, analyses was carried out if use was suspected, mostly due to information about known prescription of pregabalin | register study, postmortem toxicology analyses | whole blood concentration: 0.005 to 225 mg/L, median: 8 mg/L. Of the cases, 67% had pregabalin concentrations within the assumed therapeutic range (0.4- 17mg/L). All cases had concurrent use of other substances, mostly opioids and benzodiazepines , One potential case (19-year- old male) where pregabalin (76 mg/L) appeared to be the cause of death; but contribution of low levels of | 13% heroin, 28% morphine, methadone 19%, cocaine 20%, 55% diazepam, alcohol 24% | No conclusive information about comorbidity |
| diazepam and sertraline detected could not be excluded | |||||
| Chiappini and Schifano, 2016 | EudraVigilance database, misuse reports of gabapentin (N=410 patients) and pregabalin (N=1315 patients) | database review | gabapentin: 86 (21%) fatality reports, in 3 (3.5%) of these reports no other drug was reported pregabalin: 27 (2%) fatality reports, in 5 (18.5%) of these reports no other drug was reported | unknown | no conclusive information about possible mixtures, comorbidities, serum levels |
| Office for National Statistics, 2016 England and Wales, UK | death related to drug poisoning in England and Wales from 1993 onwards | register study | increasing deaths in which gabapenoids are involved; gabapentin 2011 (4), 2012 (8), 2013 (9), 2014 (26), 2015 (49) and pregabalin 2011 (4), 2012 (4), 2013 (33), 2014 (38), 2015 (90) – All drug poisoned deaths 2011 (2652), 2012 (2597), 2013 (2955), 2014 (3346), 2015 (3674) – any opioid 2011 (1439), 2012 (1290), 2013 (1592), 2014 (1786), 2015 (1989) – any benzodiazepine 2011 (293), 2012 (284), 2013 (342), 2014 (372), 2015 (366) – any new psychoactive substance 2011 | unknown | i)the data are based only on information reported on the coroner’s death certificate and may not include every substance involved in the death, ii) in around 1 in 8 drug poisoning deaths, only a general description is recorded on the coroner’s death certificate (such as drug overdose or multiple drug toxicity), iii) in an additional third of all drug poisoning deaths, the death certificate mentions more than 1 |
| (31), 2012 (55), 2013 (63), 2014 (82), 2015 (114) – any amphetamine 2011 (62), 2012 (97), 2013 (120), 2014 (151), 2015 (157) – cocaine 2011 (112), 2012 (139), 2013 (169), 2014 (247), 2015 (320) – cannabis 2011 (7), 2012 (14), 2013 (11), 2014 (28), 2015 (21) – z-drugs 2011 (71), 2012 (83), 2013 (86), 2014 (100), 2015 (87) antipsychotics 2011 (104), 2012 (102), 2013 (107), 2014 (126), 2015 (101) propranolol 2011 (32), 2012 (39), 2013 (46), 2014 (54), 2015 (55) – tricyclic antidepressants 2011 (200), 2012 (233), 2013 (235), 2014 (253), 2015 (215) SSRI 2011 (127), 2012 (158), 2013 (150), 2014 (159), 2015 (150) paracetamol 2011 (207), 2012 (182), 2013 (226), 2014 (200), 2015 (197) Over half of the reported deaths involved an opoid | specific drug (where more than 1 drug is mentioned, it is not possible to tell which was primarily responsible for the death) iv) where more than 1 drug is mentioned on a death certificate, the death may be counted in more than one substance category; v) approximatel y 30% of all drug-related poisoning deaths also contain a mention of alcohol or long-term alcohol abuse (for example, cirrhosis) in addition to a drug, vi) no conclusive information about addiction history and comorbidity | ||||
| Mowry et al., 2013 USA | 2576 nonpharmacolog ical and pharmacological | national register of prehospital, hospital and | 38 fatalities in which gabapentinoids are involved | unknown | blood levels were presented only in 10 |
| exposures reported to the national poison data system from 57 of the nation´s poison centers in the period of 01/01/2012 and 12/31/2012 (30nd Annual Report)* | autopsy records or indirect cases reported through other sources | (1.5% of the whole 2576 records). The 7 (0.27%) pregabalin- involved fatalities were always contaminated with other pharmaceuticals, mostly antidepressants (N=3), opioids (N=3), benzodiazepines (N=3) – one record with blood level (0.44 mg/L). The 31 (1.2%) gabapentin- involved fatalities were always contaminated with other substances, mostly antidepressants (N=15), benzodiazepines (N=18), opioids (N=17) – 9 records with blood levels (0.8 – 29 mg/L) | records, no information about individual comorbidity or addiction history, autopsy in 19 of the 38 records | ||
| Mowry et al., 2014 USA | 2113 nonpharmacolog ical and pharmacological exposures reported to the national poison data system from 57 of the nation´s poison centers in the period of 01/01/2013 and 12/31/2013 (31nd Annual Report)* | national register of prehospital, hospital and autopsy records or indirect cases reported through other sources | 41 fatalities in which gabapentinoids were involved (1.9% of the whole 2113 records). The 7 (0.33%) pregabalin- related fatalities were always contaminated with other pharmaceuticals, mostly antidepressants (N=3), opioids (N=2), benzodiazepines (N=3). The 34 (1.6%) | unknown | blood levels were presented only in 5 records, no individual information about comorbidity or addiction history; autopsy only in 14 of the 41 records. One fatality (2.4%) of the 41 gabapentinoi d records (or 0.05% of the whole 2113 |
| gabapentin- related fatalities were nearly fully (N=33) contaminated with other pharmaceuticals, mostly antidepressants (N=22), benzodiazepines (N=12), opioids (N=13) – 5 records with blood levels (1.1 – 34 mg/L) – one of the 34 records was related to gabapentin alone | fatality records) was attributed solely to a gabapentinoi d alone (gabapentin), however, this record provided no information about blood levels, autopsy results, comorbidity or addiction history | ||||
| Mowry et al., 2015 USA | 1408 nonpharmacolog ical and pharmacological exposures reported to the national poison data system from 56 of the nation´s poison centers in the period of 01/01/2014 and 12/31/2014 (32nd Annual Report)* | national register of adult prehospital, hospital and autopsy records or indirect cases reported through other sources | 41 fatalities in which gabapentinoids were involved (2.9% of the whole 1408 records). The 11 (0.8%) pregabalin- related fatalities were always contaminated with other pharmaceuticals, mostly antidepressants (N=4), opioids (N=4), benzodiazepines (N=4) – 2 records had blood levels (24 and 21.3 mg/L). The 30 (2.1%) gabapentin- related fatalities were nearly fully (N=29) contaminated with other substances, mostly antidepressants (N=14), benzodiazepines (N=10), opioids | unknown | blood levels were presented only in 9 records, no information about individual comorbidity or addiction history, autopsy only in 16 of the 41 gabapentinoi d records. One fatality (2.4%) of these 41 records (or 0.07% of the whole 1408 fatality records) was attributed solely to a gabapentinoi d alone (gabapentin), however, this record provided no information about blood levels, autopsy |
| (N=9), ethanol (N=6) – 7 records had blood levels (1.9 – 26 mg/L) – one of the 30 records was related to gabapentin alone | results, comorbidity or addiction history | ||||
| Mowry et al., 2016 USA | 1371 nonpharmacolog ical and pharmacological exposures reported to the national poison data system from 55 of the nation´s poison centers in the period of 01/01/215 and 12/31/2015 (33nd Annual Report)* | national register of adult prehospital, hospital and autopsy records or indirect cases reported through other sources | 59 fatalities in which gabapentinoids were involved (4.3% of the whole 1371 records). The 8 (0.6%) pregabalin- related fatalities were always contaminated with other pharmaceuticals, mostly opioids (N=5) and antidepressants (N=6). The 51 (3.7%) gabapentin- related fatalities were nearly fully (N=50) contaminated with multiple other pharmaceuticals, mostly opioids (N=22), benzodiazepines (N=16), antidepressants (N=21) – gabapentin blood levels were presented in in 7 records (11-35.9 mg/L) – one of the 51 records was related to gabapentin alone | unknown | blood levels were presented only in 7 records, no information about comorbidity or addiction history, autopsy in 20 of the 59 gabapentinoi d records. One (1.7%) of these 59 records (or 0.07% of the whole 1371 fatality records) was attributed solely to a gabapentinoi d alone (gabapentin), however, this record provided no information about blood levels, autopsy results, comorbidity or addiction history |
| Abrahamss on et al., 2017 Sweden | nation-wide register data including all individuals who were dispensed methadone or | retrospective register-based open cohort study | 356 patients died (7.9%); 193 deaths were caused by overdoses (54.2%); | all opioid dependent | observational study not allowing conclusions about causalities; |
| buprenorphine as opioid maintenance for opioid dependence between July, 2005 and December, 2012, N=4501 | Z-drugs and pregabalin prescriptions were associated with overdose- death while benzodiazepine prescriptions were not associated with overdose-death but with non- overdose death | no information about socio- economic situation and comorbidity | |||
| Elliott et al., 2017 UK | requested and routine diagnostic investigation, pregabalin analysis was made in cases where pregabalin was prescribed or suspected to have been abused | case series | pregabalin was detected in 93 postmortem cases with 71 drug-related deaths among them. Pregabalin was attributed to be the main cause of death in 9 adult cases (blood levels 28- 182 mg/L). Of the 9 cases, al had additional substance abuse, mostly opiates (N=6), benzodiazepines (N=5), antidepressants (N=4), pregabalin prescription in 5 from the 9 cases. | all currently abusing or being dependent on traditional substances of abuse, mostly opiates and benzodiazepin es | no autopsy in 4 out of the 9 cases |
*only the annual reports to the US-register from the years 2012 to 2015 were considered, Annual reports are available as of 1999 (http://www.aapcc.org/annual-reports/) [accessed on 04/04/17]
Table 4: Addictive risks of gabapentinoids and traditional substances of abuse: a comparative appraisala
| Characteristics/substanc es of abuse | Opioids | Alcoho l | Gabapenti n | Pregabalin | Benzodia- zepines | Cannabis |
| Self-administration behavior (animals) “Wanting”a | ***** | **** | none | * (only on “overdose ”) | *** | ** |
| Physical dependence (tolerance, withdrawal symptoms) | ***** | **** | *** | *** | **** | ** |
| Behavioral = psychological dependence (craving, loss of control, addictive behavior) “Wanting”a | ***** | ***** | (*) (only in patients with history of SUD) | * (especially in patients with history of SUD) | **** | *** |
| Severity of addictionb | ***** | **** | * | ** | **** | *** |
| Transitions from prescription to self- administration Wanting”a | ***** | n/a | * | ** | ** | (**)e |
| Relapsing behavior/durability “Wanting”a | ***** | ***** | * | ** | **** | **** |
| Voluntary treatment- seeking behavior “Wanting”a | ***** | ***** | none | none | *** | *** |
| Overdose toxicity | ***** | *** | * | ** | **** | * |
| Social hazards (independent on co-use of other substances of abuse) “Wanting”a | ***** | ***** | n/ac | n/ac | *** | *** |
| Rapid euphorization “Liking”a | ***** (especially intravenou s) | **** | ** (especially on overdose) | **** (especially on overdose) | **** (especially on overdose) | *** |
| Easy to obtain | **** | ***** | **** | **** | **** | **** |
| Legal control of prescription/dispensing | ***** (most countries) | (**)d | none | * (Norway, USA) | ** (e.g. flunitrazepa | ***** (most countries |
| m in Germany) | ) |
The addictive power is expressed in the shadowed boxes. Notes: aaccording to Berridge and Robinson, 2016; baccording to the mean number of fulfilled operationalized dependence-criteria (ICD- 10; DSM-IV), cno relevant information in the literature, dconsidering predominantly Muslim countries, laws about young people and drinking alcohol, estrong overlap between medicinal and recreational cannabis users (Pacula et al., 2016). Abbreviations: SUD = substance use disorder.
= no effects, * = very weak effects, ** = weak effects,
*** = moderate effects, **** = strong effects, ***** = very strong effects. The estimation of the addictive power toxicity and safety of the gabapentinoids is based upon the present review. The estimation of the addictive power and safety of traditional drugs of abuse is based upon comprehensive reviews (e.g. Morgan, 1990; Coupey, 1997; Karoly et al., 2015; Korpi et al., 2015; Volkow and Morales, 2015; Brett and Murnion, 2015; Weaver; 2015; Bluth and Pincus, 2016; Quednow and Herdener, 2016) and the authors’ expertise in the treatment of drug- and alcohol addiction (e.g. Bonnet et al., 1999; Bonnet and Gastpar, 1999; Bonnet, 2011; Bonnet et al., 2015; Scherbaum, 2016).
Figure 2: Gabapentinoids and reward: simplified hypothethic scheme of subcortical sensitizing (grey) and de-sensitizing (green) actions of traditional drugs of abuse and gabapentinoids, adapted to parts of the Incentive Sensitization Theory of Addiction (Berridge and Robinson, 2016) and to parts of the Glutamate Homeostasis Theory of Addiction (Spiga et al., 2014; Scofield et al., 2016). The sensitization of drug cues and associations of drug-use and environmental (external) cues were considered to be primarily driven by the repeated stimulation of non-NMDA glutamate receptors in key regions of the mesolimbic reward system, more precisely, by the repeated activation of the AMPA- and metabotropic glutamate receptors (Spiga et al., 2014; Volkow and Morales, 2015; Scofield et al., 2016). Unlike traditional substances of abuse, gabapentinoids are hypothesized to act also de-sensitizing (Dooley et al., 2000; Eroglu et al., 2009) via inhibiton of α2-subunit containing VGCC and, thereby, are not able to induce a sustaining “wanting” (addictive power).
More information: Thomas Guttuso et al. Effect of gabapentin on hyperemesis gravidarum: a double-blind, randomized controlled trial., American Journal of Obstetrics & Gynecology MFM (2020). DOI: 10.1016/j.ajogmf.2020.100273
