Brivaracetam: a Prospective and Multicentre Post-marketing Observational Study

Brivaracetam (BRV) is a new antiepileptic drug approved in March 2016 by Health Canada for the adjunctive treatment of focal epilepsy in adults. While randomized controlled trials represent the gold standard in measuring intervention efficacy, the generalizability of these findings to usual clinical practice remains uncertain. The primary objective of this study is to evaluate the effectiveness of BRV as an adjunctive treatment in epilepsy. The secondary objective of this study is to evaluate the tolerability of BRV as an adjunctive treatment in epilepsy. This is a prospective and multicentre post-marketing observational study. All consecutive adult patients (i.e. aged at least 18 years) in whom BRV is introduced in participating medical centres, ambulatory or hospitalized, will be approached to participate in the study. The investigators will exclude individuals with generalized epilepsy as those aged less than 18 years, in order to respect current Health Canada indications. The investigators will exclude individuals cognitively or physically unable to complete the study questionnaires. The investigators will collect data from participants during three clinical visits with their regular treating physician. These will be the baseline visit, the 3-month visit (three months following the initiation of BRV), and the 6-month visit. At each visit, the investigators will collect data on seizure type(s) and frequency. Study participants will also complete four questionnaires to measure irritability, anxiety, depression, and quality of life. There will be two primary study outcomes. These are: a) mean percent change in monthly seizure frequency; and b) proportion with at least a 50% decrease in seizure frequency. There will be several secondary study endpoints: a) mean change in irritability \[measured using the Brief Irritability Test (BITe)\]; b) mean change in generalised anxiety \[measured using the Generalized Anxiety Disorder - 7 (GAD-7) scale\]; c) mean change in depression \[measured using the Neurological Disorders Depression Inventory (NDDI-E) scale\]; d) mean change in quality of life \[measured using the 7-item Quality of Life Inventory in Epilepsy-10 (QOLIE-10) scale\]; e) the proportion of individuals that are seizure free, and f) change in distribution of seizure types (e.g. focal with motor seizures, generalized absence). The investigators will query for all adverse effects the participant may experience.

Study introduction/rationale: Brivaracetam (BRV) is a new antiepileptic drug (AED) approved in March 2016 by Health Canada for the adjunctive treatment of focal epilepsy in adults. Three phase III clinical trials (randomized, blinded, and placebo-controlled) have confirmed its efficacy with an acceptable adverse effects profile. While randomized controlled trials represent the gold standard in measuring intervention efficacy, the generalizability of these findings to usual clinical practice remains uncertain. These questions are in part because of certain aspects of the pivotal studies' design. These include: a) participation was limited to individuals with frequent focal seizures (at least eight seizures during an eight-week baseline period), with or without secondary generalization who were already treated with 1 or 2 AEDs; b) individuals aged less than 16 years and greater than 70 years (80 years in one study) were excluded, as well as individuals with cerebral neoplasms, psychogenic non-epileptic seizures, or status epilepticus within the preceding 12 months; d) BRV was introduced at fixed doses without a titration period; and e) the treatment period was limited to 12 The weeks. Inherent to any clinical trial, there remain questions as to what effect local clinical practice and patterns of reimbursement may have on intervention effectiveness (a measure of real-life efficacy that considers issues such as non-compliance). Clinical trials are also at high risk of the Hawthorne effect, where participants modify their behaviour (e.g. increased compliance) due to the knowledge that they are observed. Canada is one of the first countries where BRV was commercialized. This allows for an important opportunity for investigators to carry out a prospective observational study to establish BRV's effectiveness and tolerability in routine clinical practice. Primary study objective: The primary objective of this study is to evaluate the effectiveness of BRV as an adjunctive treatment in epilepsy. Secondary study objective(s): The secondary objective of this study is to evaluate the tolerability of BRV as an adjunctive treatment in epilepsy. Participant recruitment: This is a prospective and multicentre post-marketing observational study. As an observational study, the study investigators will not determine which individuals receive BRV nor how it is prescribed. The study investigators will not have any role in providing BRV to enrolled study participants. Data will be collected from participants during three clinical visits with their regular treating physician (MD). These will be the baseline visit, the 3-month visit (three months following the initiation of BRV), and the 6-month visit. The baseline visit will be the regularly scheduled clinical visit where the MD will decide to begin BRV in their patient. The patient will be asked by their MD if they are interested in participating in the proposed study. If the offer is accepted, the individual will be consented by the research assistant (RA). The baseline data questionnaire will be completed by the MD, with the input of the participant. The participant will then complete the four questionnaires to measure baseline irritability, anxiety, depression, and quality of life. The study participant will be given a seizure diary and asked to prospectively record their seizures. At the 3-month visit the MD will record the seizure frequency (preferably based on the seizure diary but retrospectively assessed if necessary) and any adverse effects reported by the participant. The participant will complete the 3-month irritability, anxiety, depression, and quality of life measures. The data collected at the 3-month visit will be re-collected at the 6-month visit. The study is planned for 10 recruitment sites across Canada. Dr Mark Keezer (Université de Montréal) is the study principal investigator. The Centre Hospitalier de l'Université de Montréal (CHUM) is the primary study site. Véronique Cloutier (CHUM) is the primary study coordinator. Study safety endpoints: This is an observational study. The investigators will include, however, a statement in the participant consent form stating "Brivaracetam is approved by Health Canada for the adjuvant treatment of focal epileptic seizures in adults. It is currently not approved for the treatment of generalized epilepsies nor epilepsy in children. There are no data on the safety of brivaracetam during pregnancy and during breast feeding." If during this study there is any information learned by the investigators that may be relevant to the health of the participant, the investigators will immediately communicate this to the study participant and their MD. Study specific assessments/instruments/scales: The baseline data questionnaire will collect the following participant data: age, gender, age of epilepsy onset, type and aetiology of epilepsy, seizure type(s) and average seizure frequency over the preceding one month, name of the AED(s) currently used and previously tried, and past medical history (including psychiatric). If levetiracetam was previously tried but since withdrawn, the reason for its discontinuation will be noted. When possible, the baseline mean number of seizures per month will be calculated from the seizure diary (if the participant completes at least 30 days of data). This will occur following their recruitment into the study but before the initiation of BRV, where the investigators expect there will be a delay. If at least 30 days of baseline seizure frequency are not recorded prospectively, the analyses will be conducted using the retrospectively collected frequency recorded in the baseline data questionnaire. The Brief Irritability Test (BITe), Generalized Anxiety Disorder - 7 (GAD-7) scale, Neurological Disorders Depression Inventory (NDDI-E) scale, and Quality of Life Inventory in Epilepsy-10 (QOLIE-10) scale, will be used to measure irritability, anxiety, depression, and quality of life, respectively. English- and French-language versions of these questionnaires have been validated, apart from the BITe which has only been validated in English. The investigators will follow standard cross-cultural methods to create a French-language version of the BITe. The original English version will be translated to French by a professional translator. A second translator, blinded to the original English-language instrument, will back-translate the new French-language instrument to English. One study investigator, fluent in both English and French, will then compare the documents to ensure that the different versions are concordant. Adverse effects will be recorded by the MD, spontaneously reported by the participant, or in response to a standardized question. Each participant will be asked the standardized question "Since the start of the brivaracetam, is there anything changed about your health?" For each reported adverse effect, the MD will judge whether it is mild (i.e. does not interfere with their daily lives), moderate (i.e. sometimes interferes with their daily lives), or severe (i.e. most of the time interferes with their daily lives). The MD will judge whether the adverse effect is unlikely, possibly, or probably related to BRV. This method of assessing for adverse effects is modeled on those used in phase IV trials of levetiracetam. Study sample size calculation: Eleven participants are sufficient to identify a mean percent seizure reduction of 20% \[standard deviation (SD): 15\], between the baseline versus the 3- and 6-month follow-ups, using ANCOVA for repeated measures. In addition, 76 participants are sufficient to identify a mean percent seizure difference of 5% between the 3- and 6-month follow-ups \[i.e. 25% (SD: 15) versus 20% (SD: 15)\], using paired student's T-test. These sample size calculations assume a power of 90%, no correlation between the baseline and follow-up parameters but an 80% correlation between 3- and 6-month follow-ups, and using a significance level of 0.05. Following the ten events per variable rule, 115 participants are sufficient to construct valid logistic regression models for the proportion of individuals with at least a 50% decrease in seizure frequency, assuming the inclusion of four independent variables \[follow-up timing (i.e. baseline, 3-month, or 6-month), age, sex, number of baseline AEDs\] and an expected prevalence of 40% of participants with at least a 50% decrease in seizure frequency. Of note, the ten events per variable rule is considered by some to be overly conservative. The expected effectiveness of BRV is based upon the pooled phase III trial data. Stratified analyses are planned and so to guarantee sufficient sample size our target is 200 participants. Study statistical methods: The effectiveness analyses will be limited to participants with at least three months of on-treatment follow-up data. The BRV tolerability analyses will be carried out in all participants that received at least one dose of BRV. Mean percent change in monthly (30 days) seizure frequency \[95% confidence interval (95% CI)\] will be compared between the baseline, 3- , and 6-month visits using ANCOVA for repeated measures. The same will be used to compare BITe, GAD-7, NDDI-E, and QOLIE-10 scores. For the analysis of these continuous outcomes, normality will be confirmed with visual inspection of histograms. If normality is not satisfied, the data will be transformed to meet normality, or the investigators will calculate median estimates and use non-parametric tests. The investigators will measure the association between the start of BRV and the average probability of participants experiencing at least a 50% decrease in seizure frequency (95% CI) at the baseline, 3-, and 6-month visits using conditional logistic regression. The same will be used to assess the association between BRV and the probability of participants being seizure-free (95% CI). The investigators will control for the following potentially confounding variables: sex, age, and the number of baseline AEDs. The investigators will stratify analyses by BRV starting dose (≥ 100 mg/day versus \< 100 mg/day), epilepsy type, the prior or concomitant use of levetiracetam (versus not), and the use of a retrospectively (versus prospectively) collected baseline seizure frequency data. Sensitivity analyses will include: the exclusion of any participant where changes to another AED occurred during the study period; the exclusion of any participant whose baseline mean seizure frequency was less that 4 seizures per month. The investigators will descriptively assess tolerability, as measured by the reported adverse effects. The mean duration of BRV use, the distribution of seizures types (baseline, versus 3 months and 6 months), and the reason (if any) for BRV discontinuation will also be presented descriptively. Statistical analyses will be carried out using STATA/SE, version 12.0 (StataCorp LP, College Station, Texas, USA). Statistical significance will be defined as a p-value \< 0.05.