Topiramate for alcoholism treatment: novel pharmacogenetic evidence for the journey to personalized medicine?

The small number of medications approved for the treatment of alcoholism is, by itself, an important justification for the significant efforts directed towards medication development. Perhaps an even more important issue is why a medication (approved or used off-label) does not always reproduce the same results across randomized clinical trials (RCTs). Indeed, even clinical trials that find effects report that there are both treatment responders and non-responders, and the ability to identify responders and non-responders could help explain variation across clinical trials. This variation forms the foundation of ‘personalized medicine’ approaches, and highlights the need to identify patient sub-types that respond better to a certain medication.

Although typologies ( Leggio et al., 2009 ) provide phenotypic traits that may be applied in RCTs, pharmacogenetics ( Heilig et al., 2011 ) represents an even more compelling approach, as it is more reproducible, biologically-based and mechanistic.

Although in its infancy, the alcoholism pharmacotherapy literature has already shown the potential value of pharmacogenetics as a way to develop personalized medicine. The functional OPRM1 polymorphism A118G plays an important role in the effects of naltrexone on alcohol consumption, with G-allele carriers responding more positively to treatment than A-allele homozygotes ( Oslin et al., 2003 ; Anton et al., 2008 ; Barr et al., 2010 ; Kranzler et al., 2013 ). A functional polymorphism in the promoter region of the serotonin transporter gene, SLC6A4 , known as 5-HTTLPR, results in a long (L) and a short (S) allele. Alcoholic patients who were L homozygotes had significant reductions in alcohol use when treated with ondansetron, as compared to S carriers ( Kenna et al., 2009b , 2014; Johnson et al., 2011 , 2013 ). It has also been shown that the tri-allelic 5-HTTLPR polymorphism moderated the effects of sertraline and age at onset of alcohol dependence on the frequency of drinking and heavy drinking ( Kranzler et al., 2011 ). Finally, clinical trials testing the role of acamprosate in alcoholism have generated conflicting results ( Kiefer and Mann, 2010 ), but preliminary findings suggested that an intronic SNP (rs13273672) in the gene for GATA-binding protein 4 ( GATA4 ) was associated with relapse, an effect that was mainly based on patients’ positive response to acamprosate ( Kiefer et al., 2011 ).

Topiramate is an anti-epileptic medication that is currently approved to treat epilepsy, prevent migraine, and (in combination with phentermine) promote weight loss. Topiramate increases GABA _A -facilitated neuronal activity and simultaneously antagonizes AMPA and kainate glutamate receptors ( Kenna et al., 2009a ). In the first RCT of topiramate, Johnson et al. (2003) demonstrated the efficacy of this drug in reducing alcohol consumption and promoting abstinence in 150 alcoholic patients randomized to placebo or topiramate 300 mg/d. A second, larger RCT ( n = 371) confirmed a robust effect of topiramate 300 mg/d as a treatment for alcoholism ( Johnson et al., 2007 ), while an additional small clinical trial ( n = 106) with alcoholic patients receiving a residential treatment program failed to show any differences between topiramate and placebo ( Likhitsathian et al., 2013 ).

The overall conclusion is that topiramate represents a promising medication for alcoholism ( Johnson and Ait-Daoud, 2010 ). However, no research has been conducted to identify possible endophenotypic characteristics of alcoholic patients who may or may not respond to topiramate. Additionally, a clinical concern with topiramate has been the potential for significant adverse effects, especially cognitive and neurological ones. For example, in one study, paresthesia and difficulties with concentration were significantly more frequent in the topiramate than in the placebo group ( Johnson et al., 2007 ). Of note, however, most adverse effects appear to be dose-related ( Kenna et al., 2009a ), suggesting that lower doses of topiramate may be safer. Interestingly, a human laboratory placebo-controlled pilot study comparing two doses of topiramate (200 and 300 mg/d) found that the effects of topiramate on alcohol use were evident by the third week of treatment (i.e. while the drug was still being titrated), suggesting that lower doses than 300 mg/d may be effective ( Miranda et al., 2008 ). This same study also found that topiramate, compared to placebo, reduced alcohol-related stimulation effects, which was suggested as a possible biobehavioral mechanism of action, but this was only observed for those receiving 200 mg/d of the medication ( Miranda et al., 2008 ). In summary, preliminary evidence suggested that investigating lower doses of topiramate, such as 200 mg/d, may be of scientific and clinical value.

A recently completed RCT by Kranzler et al., offers new and compelling evidence to support this view. This study, conducted in 138 heavy drinkers who were interested in reducing their alcohol consumption to safe levels, used the lower 200 mg/d dose of topiramate. The main results, published elsewhere ( Kranzler et al., 2014b ), indicate that even at this dosage, the medication was significantly better than placebo in reducing heavy drinking days and increasing abstinent days. Perhaps more interesting, however, was the secondary finding that the efficacy of topiramate, which exerts effects on glutamate activity through kainate receptors containing GluK1 or GluK2 subunits, was moderated by a single nucleotide polymorphism (rs2832407) in GRIK1 , the gene encoding the GluK1 receptor subunit. In the subsample of European Americans ( n = 122), topiramate was effective in reducing heavy drinking days only in rs2832407 C-allele homozygotes. In addition, a previous pharmacogenetic analysis of the human laboratory pilot study mentioned previously ( Miranda et al., 2008 ) showed that rs2832407 was associated with the severity of topiramate-induced side effects ( Ray et al., 2009 ). The RCT by Kranzler et al. (2014) did not find an effect of the SNP on adverse events, suggesting that the kainate receptor does not play a unique role in mediating topiramate-related adverse effects.

In the present issue of the International Journal of Neuropsychopharmacology , Kranzler et al. ( 2014a ) expand on their initial findings from the same RCT with respect to genetic moderation by GRIK1 . They report that rs2832407*C allele homozygotes treated with topiramate consumed significantly less alcohol, as measured by patients’ daily reports obtained using interactive voice response (IVR) technology, an innovative approach that takes advantages of technology-based assessment techniques (for a review, see ( Gurvich et al., 2013 )). In addition, rs2832407*C allele homozygotes reported larger decreases in both positive alcohol expectancies and the desire to drink. Thus, the results of the present study corroborate findings from the parent report concerning drinking behavior ( Kranzler et al., 2014b ), and add novel information in the form of medication × genotype effects on the motivation to drink.

The present findings are of clinical relevance as they suggest a path to consider topiramate within a personalized medicine framework. The authors provide a comprehensive discussion of possible directions that this research may lead in, including enrichment strategies for future RCTs (e.g. prospective genotyping), development of animal models via the manipulation of GluK1, and medications development focused on the GluK1-containing kainate receptor.

While it may be premature to envisage genotyping rs2832407 to inform the decision on whether to treat individuals with topiramate, an RCT with prospective genotyping for this SNP could not only validate these findings but also provide an approach that, in the future, might be applied to real-world clinical settings. Although speculative at this stage, the present findings may represent the avant-première of a much anticipated future where primary care physicians can take a buccal swab, analyze patients’ genetic make-up, and then decide which medication to prescribe to their alcoholic patients.

A few years ago, Kranzler et al. (2009) reported a significant association between rs2832407 and alcohol dependence. The more recent RCT provides additional evidence in support of the role of this SNP in alcohol use disorder by suggesting that the kainate receptor plays a key role in the mechanism by which topiramate affects not only drinking behavior but, perhaps at a more fundamental level, the motivation to drink. Despite these compelling pharmacogenetic data for topiramate, the need for functional validation remains. In this regard, genetic manipulations such as with GluK1 knockout (KO) mice may provide important information. Although one may dispute the clinical validity of KO models, a reverse translational approach could provide functional validation of the present clinical findings and generate useful information, even beyond the role of topiramate sensu stricto.

Last but not least, the present findings could also serve as the basis for the development of novel medications for alcoholism that specifically target the GluK1 receptor. Although selective GluK1 ligands have been synthesized (e.g. Venskutonyte et al. (2011) ) and tested in humans, including healthy controls ( Petersen et al., 2014 ) and patients with chronic pain conditions ( Chappell et al., 2014 ), it may be some time before these compounds can be tested in alcoholic patients. In actuality, the efficacy of topiramate may be due to its multiple pharmacological mechanisms of action, rather than those specific to the GluK1 receptor. Nonetheless, the development of a selective GluK1 ligand with potential effects on alcohol consumption remains an important avenue for clinical research.

In summary, although additional studies are needed, including replications and extensions to non-European-American populations, the Kranzler et al. findings are of scientific, pharmacological, pharmacogenetic, and clinical interest and represent an exciting platform for future research aimed at novel personalized medicine approaches.

Acknowledgments

This work was supported by the Division of Intramural Clinical and Biological Research of the National Institute on Alcohol Abuse and Alcoholism (NIAAA) and the Intramural Research Program of the National Institute on Drug Abuse (NIDA). The content of this review is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Statement of Interest

None.

References