Scolaris Content Display Scolaris Content Display

Cochrane Database of Systematic Reviews Protocol - Intervention

Topiramate versus carbamazepine monotherapy for partial onset seizures or generalized onset tonic‐clonic seizures with or without other generalized seizure types

This is not the most recent version

Collapse all Expand all

Abstract

This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:

To compare the efficacy and tolerability of topiramate versus carbamazepine when used as monotherapy in patients with GTCS or partial onset seizures with or without other generalized seizure types.

Background

Description of the condition

Epilepsy is a chronic neurological condition characterised by paroxysm of seizures due to abnormal electrical discharge from central nervous system (CNS) neurons. The estimated incidence of epilepsy is approximately 50 per 100,000 and prevalence five to 10 per 1000 in the developed world (Sander 1996; IMF Link 2009). Between 2% and 3% of the population will be given a diagnosis of epilepsy at some time in their lives. In patients diagnosed with epilepsy, 20%‐30% achieve spontaneous remission without any treatment. Antiepileptic drug treatment achieves remission in 20%‐30% of patients. Unfortunately about 30%‐40% continue to have seizures despite therapy (Kwan 2004).

Description of the intervention

Antiepileptic drugs (AEDs) are the mainstay of treatment in patients with epilepsy. Studies have shown that in the majority of patients, monotherapy is equally or more effective, better tolerated and has fewer adverse effects than polytherapy. However, up to 30% will fail to respond to monotherapy with standard AEDs (Cockerell 1995; Hauser 1993), often requiring the use of newer antiepileptic drugs (AEDs). Adverse effects to standard AEDs can be severe enough to warrant withdrawal of therapy. This has generated a renewed interest in the development of newer AEDs over the past 15‐20 years.

How the intervention might work

Newer AEDs are believed to be equally effective and have better tolerability and side effect profiles. Topiramate, a newer AED, has been found useful when used as an add‐on to existing AEDs in patients with refractory epilepsy (Jette 2008). "If treatment fails to achieve seizure freedom for 12 months or more they are refractory (Kwan 2000)". Topiramate has also been used as monotherapy in patients with newly diagnosed epilepsy (Glauser 2007; Privitera 2003). Carbamazepine is a standard AED which has been in use as monotherapy and also as an add on in patients with generalized tonic‐clonic seizure (GTCS) or partial onset seizures with or without secondary generalization (Marson 2007).

Why it is important to do this review

We plan to undertake a systematic review comparing the efficacy, tolerability and side effects of topiramate used as monotherapy with carbamazepine monotherapy in order to better inform clinicians and patients.

Objectives

To compare the efficacy and tolerability of topiramate versus carbamazepine when used as monotherapy in patients with GTCS or partial onset seizures with or without other generalized seizure types.

Methods

Criteria for considering studies for this review

Types of studies

  • Randomized controlled trials or quasi randomized trials

  • Trials may be single blinded, double blinded or unblinded

  • Trials must include a comparison of topiramate and carbamazepine used as monotherap.

Types of participants

  • Adults or children of all age groups

  • No gender preference

  • New onset or chronic generalized tonic clonic or partial onset seizures, with or without other generalized seizure types

Types of interventions

One group of patients receives topiramate as monotherapy, while the other group receives carbamazepine as monotherapy. We will include all designs, such as initial monotherapy in newly diagnosed epilepsy; alternate monotherapy in persons who do not respond to initial therapy with another agent; or withdrawal to monotherapy designs.

Types of outcome measures

Primary outcomes

Time from randomization to withdrawal of allocated treatment (retention time): the allocated treatment may have been withdrawn due to poor seizure control, side effects or the use of additional add‐on treatment. This is a combined outcome measure which encompasses efficacy as well as tolerability (Commission 1998). Wherever possible we will try to stratify the withdrawals as whether due to poor seizure control, due to side effects or due to other reasons.

Secondary outcomes

  • Number or percentage seizure free after one year

  • Time to six‐, 12‐ or 24‐month remission from seizures

  • Time to first seizure post‐randomization

  • Severe adverse effects encountered during therapy

  • Quality of life, if assessed by validated scale.

Search methods for identification of studies

Electronic searches

We will search the following databases.

1. Cochrane Epilepsy Group Specialized Register.

2. CENTRAL (The Cochrane Central Register of Controlled Trials, The Cochrane Library).

3. MEDLINE. The proposed search strategy for MEDLINE is set out in Appendix 1. This strategy will be modified for use with the other databases.

We will not impose any language restrictions.

Searching other resources

(1) References from published studies

We will review the reference lists of retrieved studies to search for additional reports of relevant studies.

(2) Other sources

We will contact the manufacturers and original investigators, colleagues and experts in the field to identify any trials (published, unpublished or ongoing) not revealed by our searches.

Data collection and analysis

Selection of studies

Two review authors (VS and LNR) will independently assess the titles and abstracts of publications identified by the above search strategy for inclusion; we will select the full text for further assessment if the abstract suggests relevance. We will resolve disagreements, if any, by discussion and, if inconclusive, by referring to the third author (SR).

Data extraction and management

Two review authors (VS and LNR) will independently extract data from the selected studies on a data extraction form.
We will pilot test and modify the data extraction form if necessary. We will discuss disagreements, and solve them by consensus or referral to third author (SR).
We will extract the following data.

I. Methodological data

(1) Method of randomization
(2) Method of concealment of randomization
(3) Stratification factors
(4) Methods of blinding
(5) Description of withdrawals and dropouts
(6) Duration of baseline period
(7) Duration of treatment period
(8) Dose(s) of topiramate and carbamazepine tested
(9) Outcome assessment methods

II. Participant

(1) Number (total per group)
(2) Age and sex distribution
(3) Age at onset of epilepsy
(4) Seizure type and epilepsy syndrome
(5) Loss to follow up

III. Type of intervention: drug, dosage, type of preparation and route of administration
IV. Duration of follow up
V. Outcome measures: as listed above

Assessment of risk of bias in included studies

We will evaluate the methodological quality of the studies and the likelihood of bias according to the guidelines in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2009 ). We will exclude studies with high risk of bias or stratify them separately during analysis.

Measures of treatment effect

(1) Hazard ratio estimates (with confidence intervals);
(2) proportion of people achieving each outcome at certain time points at six, 12 and 24 months (with confidence intervals).

Unit of analysis issues

For time to event outcomes (e.g. time to seizure recurrence), we plan to use aggregate data methods (Parmar 1998; Williamson 2002) in the first instance and results will be presented as hazards ratios with 95% confidence intervals. If, after undertaking this analysis, it seems appropriate, we will attempt a further analysis using individual patient data. For time to seizure recurrence we also plan to provide estimates of the risk difference (RD with 95% confidence intervals (CI)) at six,12 and 24 months, from which the reciprocal will be taken as the estimate of the "number needed to harm".

Dealing with missing data

We will contact authors of identified studies to obtain any missing data. If studies are only in abstract form or published without any clear description of the methodology employed, we will contact the original trialists for clarification.

Assessment of heterogeneity

We will assess clinical and methodological diversity by comparing the distribution of important patient factors between trials (age, seizure type, certainty of diagnosis of epilepsy, duration of epilepsy, whether the drug was used in subjects who have not used AEDs before or among those who have failed one or more AEDs, and trial factors (randomization concealment, blinding, losses to follow up). Statistical heterogeneity will be assessed by using a I2 test where a value greater than 50% would indicate substantial heterogeneity (Higgins 2003).

Assessment of reporting biases

All authors may not have reported all outcomes as planned for this analysis. We will contact the original investigators for any outcome not reported in the publication. We will also try to ascertain the reason for withdrawals from study, if not mentioned in the publication.

Data synthesis

Provided no significant heterogeneity is present, we will synthesize data using a fixed‐effect model. Where there is significant heterogeneity, we will explore the causes for heterogeneity and the possibility of employing random‐effects models. For binary outcomes we will use risk ratio (RR) and 95% confidence intervals (CIs). For continuous data we will calculate mean difference (MD) with 95% CI. Primary analysis will be intention to treat, in which all participants are included in the treatment groups to which they were allocated, irrespective of the treatment or policy they actually received.

Subgroup analysis and investigation of heterogeneity

We will examine whether carbamazepine is inferior or superior to topiramate, or of equivalent efficacy. We will infer non‐inferiority if the efficacy does not differ by more than 15% (when the sample size is adequately powered).

Where possible we will undertake a subgroup analysis according to seizure type (partial onset versus generalized onset) and age groups (children less than 17 years old versus adults).

We plan to examine the dose‐response relationships using logistic regression, in the framework of generalized linear models (McCullagh 1989).

We will describe data on neurological examinations, CT scans, MRI scans, EEG results and effects on cognition, neuropsychological tests and quality of life.

Sensitivity analysis

We plan to do a sensitivity analysis to the test the robustness of the meta‐analysis. We plan to analyze the results by excluding and including low‐quality trials; by including or excluding results of drug‐naive patients with those who have failed one or more drugs.