Abstract
Brivaracetam (BRV) is a novel antiepileptic drug recently licensed for the treatment of partial epilepsy in adults and adolescents over 16 years old. Like levetiracetam (LEV), it is a ligand of the synaptic vesicle protein SV2A. BRV has been shown in animal models and in studies using human brain slices to have a higher SV2A affinity and faster penetration into the brain. Its efficacy and safety have been shown in several randomized, controlled studies. The recommended initial dose is 50–100 mg, divided into two daily doses. Up-titration to a 200 mg daily dose is possible. Dizziness and somnolence are frequent side effects. There are some hints that BRV may be less frequently associated with behavioural adverse events than LEV. Long-term efficacy and safety and BRV use in special patient groups have to be assessed in the future.
Keywords
Introduction
About 15 antiepileptic drugs (AEDs) have been introduced since 1992, somewhat arbitrarily summarized as the ‘new’ or ‘newer’ AEDs as opposed to the older ones that had been introduced before. In general, the newer AEDs have advantages with regard to tolerability and drug–drug interactions. For instance, enzyme-inducing properties are expressed to a much lesser extent than in the older AEDs. Nevertheless, the newer AEDs may, of course, also have side effects, for instance psychiatric and cognitive ones. Aggressive behaviour and depression under levetiracetam (LEV) [Mula et al. 2003] and cognitive decline under topiramate have to be mentioned here [Brandt et al. 2015]. In addition, it seems that the number of patients with pharmacoresistant epilepsy has only moderately declined despite the availability of the newer AEDs [Schmidt et al. 2014].
Recently, Brivaracetam (Briviact®; BRV) has been approved in the European Union and the United States for the adjunctive treatment of partial seizures with or without secondary generalization in adults and adolescents with epilepsy from 16 years on [UCB, 2016]. The drug is available as film-coated tablets in dosages of 10, 25, 50, 75 and 100 mg, as an oral solution in a concentration of 10 mg/ml and as a solution for injection or infusion at the same concentration. A starting dose of 50 or 100 mg daily is recommended, and dose-escalation up to 200 mg/day is possible. The drug is administered divided in two daily doses. The tablets contain lactose. The dosages of the different formulations are equivalent [UCB, 2016].
The objectives of this paper are as follows: First, information on the mechanism of action of BRV and data from animal models will be presented to provide the reader with a framework of the development of the drug. Data on pharmacokinetics, efficacy, and tolerability will be reported. The focus of the paper is, actually, a clinical one: What is the current evidence with regard to efficacy and safety of BRV? Is there any difference to LEV? Does BRV show efficacy in patients who are currently or were previously on LEV? Do we have to expect behavioural side effects? What is known about pharmacokinetics, especially possible drug–drug interactions? Where is there still a lack of data? What can be done to fill these gaps?
Mechanism of action and preclinical data
BRV, as also LEV, is a ligand of the synaptic vesicle protein, SV2A. This is the most widely distributed isoform of SV2, an integral membrane protein present on all synaptic vesicles, and nearly ubiquitous in the central nervous system, as well as being present in endocrine cells [Lynch et al. 2004]. It modulates exocytosis of transmitter-containing vesicles [Kaminski et al. 2012]. Although the role of SV2A is not fully understood, it seems to have a role in epileptogenesis: SV2A-deficient mice show an increased vulnerability to epileptic seizures and also rapid development of kindling [Kaminski et al. 2012]. Acting on voltage-dependent calcium-channels or on AMPA and GABAA receptors does not seem to be a relevant mechanism of BRV [Niespodziany et al. 2015]. BRV showed higher affinity to SV2A in animal models and in studies with human brain slices [Matagne et al. 2008; Gillard et al. 2011]. The efficacy of the drug has been tested in several established animal models. It was more potent than LEV in audiogenic mice, Genetic Absence Epilepsy Rats of Strasbourg (GAERS), in amygdala- and cornea-kindling [Matagne et al. 2008] and in a rat model of post-hypoxic myoclonus [Tai et al. 2007]. Besides the suppression of seizures, BRV seemed to act as an antiepileptogenic in different kindling models [Dupuis et al. 2015]. BRV is more lipophilic than LEV [Klitgaard et al. 2016] and penetrates more rapidly into the brain in vitro, as measured by the use of Caco-2 cells, an in vitro model of the blood-brain-barrier [Garberg et al. 2005], and in a rodent model [Nicolas et al. 2015]. This may play a role in the treatment of status epilepticus or seizure clusters rather than in chronic epilepsy treatment. Various animal models of cognitive function did not show any evidence of negative effects of BRV on cognition [Detrait et al. 2010].
Pharmacokinetics and interactions
BRV has linear and predictable pharmacokinetics. Absorption after oral intake is fast and complete. Serum concentration peaks after a median of 0.5–2 h respectively, 3 h when BRV is administered with fatty food [Sargentini-Maier et al. 2007, 2008; Rolan et al. 2008]. Fatty food leads also to a decreased maximal plasma concentration but does not affect the area-under-the-curve (AUC). Steady-state is reached after 2 days. Distribution volume is about 0.5 l/kg which is slightly lower than the total body water [Stockis et al. 2013]. Less than 20% of the drug is bound to plasma proteins. This is a low value with regard to possible pharmacokinetic interactions [Sargentini-Maier et al. 2008]. Elimination half-life is about 9 h. The drug is extensively metabolized, primarily (60%) by hydrolysis, which is independent of the cytochrome P450 (CYP) system. A second metabolization pathway (30%) is via hydroxylation mediated by CYP2C19 [Stockis et al. 2014b] but not by CYP2C8 and CYP2C9 [Nicolas et al. 2012]. BRV has mainly three metabolites (acid, hydroxy, and hydroxyacid) all of which are not pharmacologically active. About 90% of the drug will be excreted by the kidneys within 72 h after intake, about 5–8% unchanged and >90% as pharmacologically inactive metabolites [Stockis et al. 2013]. Clearance was 30% lower in poor than in extensive CYP2C19 metabolizers in a study with healthy Japanese volunteers [Stockis et al. 2014b].
In patients with hepatic impairment, exposure to BRV increased by 50–60%, irrespective of severity of hepatic impairment. Plasma half-life of BRV increased as well (up to about 18 h) [Stockis et al. 2013]. Thus, hepatic impairment probably requires dose adjustment. In patients with renal impairment, the AUC of BRV was only slightly higher (mean ratio 1.21; 90% confidence interval 1.01–1.45) compared with healthy controls. Maximum serum concentration [Cmax] was not affected [Sargentini-Maier et al. 2012]. In contrast, exposure to the acid, hydroxy (BRV-OH), and hydroxyacid metabolites was markedly increased. However, nonclinical toxicology studies indicated that safety issues were not related to increased levels of metabolites. These observations suggest that dose adjustment of BRV should not be required at any stage of renal dysfunction [Sargentini-Maier et al. 2012].
There is no evidence of an inducing or inhibiting effect of BRV on CYP3A4 activity [Stockis et al. 2015b]. BRV at a dose of 200 mg twice daily doubles the serum concentration of carbamazepine-epoxide. It did not significantly alter the carbamazepine (CBZ) AUC. The influence of BRV on CBZ epoxide increases, in our opinion, the propensity to side effects under the combination of BRV and CBZ, as it is the case for the interaction between CBZ and valproic acid [Schoeman et al. 1984]. On the other hand, CBZ, phenytoin and phenobarbital reduce the BRV plasma concentration by 20–25% [Stockis et al. 2015a].
Rifampin, a potent CYP inducer, did not significantly affect the Cmax of BRV, but significantly decreased the AUC of BRV by 45%, mostly through induction of the CYP2C19 pathway. Adaptation of BRV dose should be considered in patients on treatment with rifampin [Stockis et al. 2016]. Gemfibrozil, a potent in vitro inhibitor of CYP2C9, did not influence the pharmacokinetics of BRV and its hydroxylation into BRV-OH [Nicolas et al. 2012].
BRV in a daily dose of 100 mg had no effect on the plasma concentrations of ethinylestradiol and levonorgestrel, which were taken as a combination oral contraceptive (OC), and the BRV concentration in serum in turn was not affected by the OC [Stockis et al. 2014a]. BRV 400 mg/day, a dose that is currently above the licensed dose range but might be reached in off-label use in the future, co-administered with a combination OC in healthy women, reduced the AUC for ethinylestradiol by 27% and that for levonorgestrel by 23% but did not affect the levels of endogenous hormones (i.e. did not result in ovulation) [Stockis et al. 2013].
There is only sparse information on BRV plasma concentrations in the literature. Data are available from the largest BRV trial [Klein et al. 2015]. Mean BRV plasma levels were between 2.02–2.06 µg/ml under BRV 200 mg/day and between 1.06–1.15 µg/ml under BRV 100 mg/day. Plasma and saliva levels of BRV are highly correlated according to data that have, however, been obtained only from healthy men [Rolan et al. 2008]. In these, mean BRV Cmax plasma values of 3.5 µg/ml, 7.7 µg/ml and 13.3 µg/ml were measured at steady state (day 14) after daily doses of 200 mg, 400 mg and 800 mg, respectively. BRV population pharmacokinetics and exposure-response modelling in adults with partial-onset seizures indicated that co-administration with CBZ, phenytoin, and phenobarbital/primidone decreased BRV exposure by 26%, 21%, and 19%, respectively, without significant effect on response [Schoemaker et al. 2016]. Further research on BRV pharmacokinetics including studies from routine therapeutic drug monitoring will be necessary.
Efficacy
A proof-of-concept study was performed in the photosensitivity paradigm [Kasteleijn-Nolst Trenite et al. 2007]. A good efficacy of BRV with regard to the photoparoxysmal response in patients with photosensitive epilepsy could be shown in this study. Of note, photosensitive epilepsies are generalized epilepsies in most cases but the usefulness of photosensitivity studies has been shown also in drugs primarily intended for the treatment of focal epilepsies [Kasteleijn-Nolst Trenite et al. 2015].
Gold-standard are randomized, double-blind, placebo-controlled multicentre trials (RCTs). BRV has been tested in several RCTs. In the two phase IIB and four phase III trials (see Table 1) a novel criterion was applied as the primary efficacy parameter: the percent reduction in baseline-adjusted partial-onset seizure frequency over placebo. This efficacy parameter was expected to have a greater power compared with median percent reductions of seizure frequency and 50%-responder rate (i.e. the rate of participants experiencing an at least 50% reduction of seizure frequency), criteria, that have been usually applied in previous AED trials [Vickers et al. 2001]. However, median percent reductions of seizure frequency from baseline and 50%-responder rate were additionally assessed in the trials mentioned above. Details of these studies are shown in Table 1. The findings can be summarized as follows: Although the primary outcome parameter was not always reached, significant improvements of important outcome parameters (50%-responder rate or median percent reduction from baseline) were found for BRV doses of 100 mg [Ryvlin et al. 2014] or 100 and 200 mg [Klein et al. 2015], in one trial also for 50 mg [Biton et al. 2014] and in another, that had, however, a short treatment period, also under 5 and 20 mg [French et al. 2010]. The innovative predefined efficacy parameter, the percent reduction in baseline-adjusted partial-onset seizure frequency over placebo, as mentioned above, was not reached in three studies. No significant superiority of BRV 50 mg over placebo was reached in one of the earlier studies [Ryvlin et al. 2014]. BRV 100 mg was, however, effective. Of note, the study did not meet its primary efficacy end point due to the predefined sequential testing strategy (50, 100, 20 mg/day). The primary efficacy criterion was also not reached in two subsequent trials [Van Paesschen et al. 2013; Kwan et al. 2014]. One of these two studies [Kwan et al. 2014] had assessment of tolerability as the primary objective. Finally, more patients (2414 randomized patients) were included in the BRV studies than in the pivotal trials of other AEDs. Most of the trials were on focal epilepsy but patients with generalized epilepsies could be included in one of the trials [Kwan et al. 2014]. In that trial, 49 [10.2%] patients had generalized epilepsy. No details on specific syndromes are given. A meta-analysis of the RCTs of BRV (except the last one which had not yet been published at that time) showed that BRV at daily doses >5 mg led to a significantly higher 50%-responder rate than placebo [Tian et al. 2015].
Summary of the pivotal studies.
In
AED, antiepileptic drug; BRV, brivaracetam; CI, confidence interval; LEV, levetiracetam; PBO, placebo; POS, partial-onset seizure.
Additional requirements may apply.
Austria, Belgium, Czech Republic, Germany, Hong Kong, India, Italy, Norway, Republic of South Africa, Russian Federation, Singapore, South Korea, Sweden, Taiwan, and Ukraine.
BRV was also tested in two randomized, double-blind, placebo-controlled trials in patients with Unverricht–Lundborg disease. The disease had been genetically ascertained. The effect of BRV on action myoclonus was not statistically significant in these trials [Kalviainen et al. 2016].
As BRV shows similarities with LEV, it is of special interest to know whether it is either advantageous to combine these two drugs or to try BRV in patients who have already had LEV before. With regard to the first question it could be shown that it does no harm to combine BRV and LEV but that, on the other hand, BRV does not add anything to the efficacy in these cases. Whilst concomitant LEV use was allowed in five of the six pivotal studies (see Table 1), it was excluded in the last, and biggest, one [Klein et al. 2015]. A total of 412 patients of this study had previously been treated with LEV, but that had been, according to the inclusion criteria, withdrawn at least 90 days before study entry. LEV had been withdrawn in about 2/3 of the patients due to lack of efficacy and in about 1/5 due to side effects. The efficacy of BRV could be demonstrated for all groups, though more clearly in the ‘LEV-naïve’ patients compared with those who had already had LEV, and among the latter more clearly in those who had LEV discontinued because of side effects, compared with those, who had discontinued it because of lack of efficacy. In detail: reduction in 28 day adjusted frequency of partial-onset seizures over placebo under 100 mg BRV was 15.8% (p = 0.024) in patients with previous exposure to LEV and 29.5% in LEV-naïve patients (p < 0.001), under 200 mg BRV 19.4% (p = 0.005) in patients with previous exposure to LEV and 27.1% in LEV-naïve patients (p < 0.001). The 50%-responder rate in patients with previous exposure to LEV was 16.8% under placebo, 28.7% under BRV 100 mg (p = 0.016) and 31.3% under BRV 200 mg (p = 0.007). It was 27.6% under placebo, 50.9% under BRV 100 mg (p < 0.001) and 45.2% under BRV 200 mg (p = 0.008) in LEV-naïve patients. With regard to patients who had previously been treated with LEV, reduction in 28 day adjusted frequency of partial-onset seizures over placebo under 100 mg BRV was 17.6% and 14.4% under 200 mg BRV in those patients in whom LEV had been withdrawn because of insufficient efficacy. Respective values were 12.5% under BRV 100 mg and 23.7% under BRV 200 mg in those in whom LEV had been withdrawn because of adverse events (AEs). No p-values are given.
Recently, a phase III study on intravenously (iv) administered BRV has been published [Klein et al. 2016]. Treatment was either initiated by oral medication and then switched to iv BRV or administered iv from the beginning. Intravenous BRV was given either as bolus injection over 2 min or as infusion over 15 min. The oral treatment was given in double-blind manner (placebo or BRV), the iv treatment was open-label. Over the course of the study, 29.5% of all patients experienced somnolence and 14.3% experienced dizziness. The figures were quite similar among all groups. The days under iv treatment were defined as the evaluation period. During this period, the incidence of dizziness was higher in the group that received iv BRV as an initial treatment as compared with the group that switched from oral BRV (11.8% versus 3.8%). Generally, iv BRV was well tolerated. Of note, this was a study on safety and tolerability, not on efficacy.
Safety and tolerability
Headache, fatigue and dizziness have been found as important side effects in several of the above mentioned studies [Biton et al. 2014; Ryvlin et al. 2014; Klein et al. 2015]. Headache was, however, also frequent in placebo groups. A recent meta-analysis found dizziness, somnolence, fatigue and irritability to occur significantly more frequent under BRV than under placebo [Lattanzi et al. 2016]. BRV in doses of 75 mg and 400 mg did not prolong the QT-interval in healthy volunteers [Rosillon et al. 2008]. Cognitive and psychiatric tolerability deserves special attention in new AEDs. A study with acute dosing of 2 × 10 mg BRV compared with moderate doses of LEV and lorazepam, and to placebo, did not show deterioration in a test battery of electrophysiological and neuropsychological measurements. This study probably does not have clinical significance because of the small BRV dose and also because of the lack of longterm data [Meador et al. 2011].
With regard to psychiatric tolerability, two studies are currently of importance: the last (and biggest) of the pivotal trials, the N01358 study [Klein et al. 2015], and a study with a switch from LEV to BRV in patients with nonpsychotic behavioural side-effects under LEV [Yates et al. 2015].
In the N01358 study [Klein et al. 2015], the frequency of psychiatric AEs was as follows: 10.3% under BRV 100 mg, 11.3% under BRV 200 mg and 7.7% under placebo. Specifically, anxiety (1.1% placebo, 2.2% under BRV, all doses), insomnia (1.1% versus 2.0%) and depression (0.4% versus 0.8%) were reported. According to a systematic review of the safety profile of LEV based on the data from placebo-controlled clinical trials in epilepsy, reports of behavioural problems in LEV and placebo groups were 13.5% versus 6.0% [French et al. 2001]. Behavioural problems were specified as agitation, hostility, anxiety, apathy, emotional lability, depersonalization or depression. The difference between verum and placebo is obviously lower for BRV than for LEV. These figures should, however, be interpreted with caution: First of all, direct comparison of data gained from different studies is not possible. Second, we do not know whether the psychiatric AEs registered in the BRV study represent the same entities as the behavioural problems in the LEV studies. The AEs in the LEV studies were classified using the Coding Symbols for Thesaurus of Adverse Reaction Terms (COSTART) [Food and Drug Administration, 1997], in the BRV study according to the Medical Dictionary for Regulatory Activities (MedDRA; version 15.0).
The course of nonpsychotic behavioural AEs after switching from LEV to BRV has been evaluated in a small study [Yates et al. 2015]. Patients who, among other inclusion criteria, experienced behavioural AEs within 16 weeks of LEV treatment initiation were enrolled in this open-label study. An immediate switch from LEV to BRV 200 mg/day (without titration) was performed. Because of recruitment difficulties, enrolment was closed at 29 patients. It can be speculated that the time frame of 16 weeks after initiation of LEV may have been too short to identify behavioural problems and refer the patients for study inclusion. Overall, 93.1% of the study subjects who switched from LEV to BRV had a clinically meaningful reduction in behavioural AEs, as determined by the investigator, and 62% experienced a complete remission at the end of the study. No patient reported a worsened intensity from baseline to the end of the treatment period. The study has some limitations, such as the small sample size, the open, unblinded character, a poorly differentiated terminology of nonpsychotic behavioural disorders and a possible regression to the mean as an explanation for the improvement.
Conclusion
BRV is a promising new AED for the treatment of partial epilepsy. Efficacy and tolerability have been proven in several RCTs. There are, however, open questions. As always when a new AED has just been introduced to the market, data on longterm efficacy and safety as a parameter of ‘real-life performance’ of a drug are sparse. The same applies to data on special patient groups, especially those with intellectual disability. With special regard to the similarity to LEV, it has to be assessed whether behavioural tolerability of BRV is really better than that of LEV. The role of therapeutic drug monitoring [Brandt et al. 2008, 2011] has also yet to be established.
Footnotes
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Conflict of interest statement
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Christian Brandt has received personal compensation from Otsuka, Eisai, Desitin, Pfizer, and UCB Pharma for serving on scientific advisory boards, for speaking activities, and for congress travel, and financial support for research activities from UCB Pharma and Otsuka. He was an investigator in several clinical trials with BRV and is a co-author of the N01358 BRV study.
Theodor May undertook industry-funded travel with support from UCB (Monheim, Germany) and Desitin (Hamburg, Germany), served on scientific advisory boards and received honoraria for speaking engagements from Eisai (Frankfurt, Germany), UCB and Desitin.
Christian Bien gave scientific advice to Eisai (Frankfurt, Germany) and UCB (Monheim, Germany), undertook industry-funded travel with support of Eisai (Frankfurt, Germany), UCB (Monheim, Germany), Desitin (Hamburg, Germany), and Grifols (Frankfurt, Germany), obtained honoraria for speaking engagements from Eisai (Frankfurt, Germany), UCB (Monheim, Germany), Desitin (Hamburg, Germany), diamed (Köln, Germany), Fresenius Medical Care (Bad Homburg, Germany), and Biogen (Ismaning, Germyn) received research support from diamed (Köln, Germany) and Fresenius Medical Care (Bad Homburg, Germany). He is a consultant to the Laboratory Krone, Bad Salzuflen, Germany, regarding neural antibodies.