Scolaris Content Display Scolaris Content Display

Base de datos Cochrane de Revisiones Sistemáticas Protocolo - Intervención

Non‐pharmacological treatments for improving memory in people with epilepsy

Declaraciones de intereses de los autores

Versión publicada: 08 diciembre 2015 Historial de versiones

https://doi.org/10.1002/14651858.CD011945
Contraer todo Desplegar todo

Abstract

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

We aim to assess the effects of non‐pharmacological interventions for the treatment of memory problems in people with epilepsy. More specifically, we aim to evaluate whether people with epilepsy receiving one of these treatments:

  1. may improve their memory, as measured by subjective patient‐reports, objective memory assessments, or both;

  2. may experience associated adverse effects;

  3. may improve in other neuropsychological domains, as assessed by validated measures (memory problems can occur due to a large number of cognitive difficulties such as word finding or executive functioning problems);

  4. may improve their quality of life, as measured by validated scales (memory is often reported as an important influence of individuals' quality of life);

Background

Description of the condition

Epilepsy is a chronic neurological disorder affecting nearly 50 million people worldwide (WHO 2015), characterised by recurrent involuntary brain activity manifesting in seizures (Chang 2003). It is associated with several comorbidities including psychological and cognition difficulties (Bell 2011). Cognitive impairment is a major concern among people with epilepsy. A survey of over 1000 patients with epilepsy found that cognitive impairment was ranked as their greatest concern (Fisher 2000). Memory difficulties are the most commonly reported cognitive impairment in epilepsy (Hendriks 2002) with 20% to 50% of people with epilepsy complaining of memory impairments (Hendriks 2002). Memory problems are cited to have a marked adverse effect on quality of life (Hall 2009).

Several factors may affect memory in people with epilepsy. The use of antiepileptic drugs (AEDs) seizure foci, interictal epileptiform activity (specific electrical abnormalities occurring in the brain between seizures), and seizure activity may disrupt the encoding of new information into memory and the retrieval of previously stored information (Hall 2009). Cognitive impairments in epilepsy are addressed indirectly by effective seizure control, careful selection and adjustment of AED doses, restructuring of antiepileptic drug polytherapy regimens, and treatment of co‐morbid depression (Shulman 2002). However, this systematic review will focus instead upon the non‐pharmacological interventions for the direct treatment of memory problems associated with epilepsy.

Description of the intervention

Neurostimulation

Vagus nerve stimulation (VNS)

Vagus nerve stimulation is a device, similar to a pacemaker, which generates regular electrical stimulation to the brain via the vagus nerve. It is a Food and Drug Administration (FDA) approved add‐on therapy for reducing seizure frequency in medically refractory partial epilepsy. The efficacy of VNS in reducing and preventing medically refractory seizures is supported by a significant amount of clinical evidence. The use of VNS has been shown to reduce seizure frequency by 24.5% to 30.9% (Landy 1993; Ben‐Menachem 1994; Handforth 1998; Panebianco 2015). However, the effects of VNS on cognition are less established with some studies reporting memory enhancement (Cramer 2001; Ghacibeh 2006) and others describing no change or deterioration in memory (Helmstaedter 2001; Hoppe 2001; McGlone 2008). In addition, a small randomised clinical trial found that the administration of VNS following learning improved word‐recognition memory in people with epilepsy (Clark 1999).

Deep brain stimulation (DBS)

Deep brain stimulation is a device which generates regular stimulation to specific parts of the brain via implanted electrodes in the brain. It is an FDA approved treatment for Parkinson’s disease, dystonia and central tremor. Research has demonstrated the efficacy of DBS for reducing seizure frequency in epilepsy (Sprengers 2014). Several studies of DBS of the anterior nucleus of thalamus (ANT) reported 50% or greater reduction in mean seizure frequency in 16% to 100% of epilepsy patients (Hodaie 2002; Kerrigan 2004; Andrade 2006; Lim 2007; Osorio 2007). In addition, a multicenter randomised trial of bilateral stimulation of the ANT in refractory partial epilepsy found stimulated patients experienced a 40% reduction in seizure frequency compared with 14% of controls (Fisher 2010). Several recent studies have examined the effects of DBS on cognition in people with epilepsy; one study of bilateral ANT stimulation found that DBS was associated with significantly improved performance in verbal fluency and delayed verbal memory tasks (Oh 2012), and stimulation of the entorhinal region during learning significantly enhanced performance on a spatial learning task which required patients to navigate through a virtual environment (Suthana 2012).

Transcranial magnetic stimulation (TMS)

Transcranial magnetic stimulation is a non‐invasive procedure which uses electromagnetic induction to deliver a magnetic field across the scalp stimulating nerve cells in regions of the brain (Ališauskienė 2005). Repetitive transcranial magnetic stimulation (rTMS) has been studied extensively as a treatment for a variety of psychiatric and neurological conditions including depression, bipolar disorder, schizophrenia, post‐traumatic stress disorder, Parkinson’s disease, dystonia, and epilepsy (Rossi 2009). A number of small studies and case reports have demonstrated the efficacy of rTMS for reducing seizure frequency in refractory partial epilepsy (Tergau 1999; Menkes 2000; Joo 2007) and epilepsia partialis continua (Graff‐Guerrero 2004; Misawa 2005; Rotenberg 2008). Currently, seven randomised controlled trials have examined efficacy of TMS for seizure reduction in epilepsy with four studies reporting significant results (Tergau 2003; Fregni 2006; Wang 2008; Sun 2012 ) and three studies reporting insignificant results (Theodore 2002; Cantello 2007; Joo 2007). One clinical trial (Fregni 2006), and one case report (Cotelli 2012) of rTMS in refractory epilepsy patients with malformations of cortical development found that this improved some aspects of cognition functioning.

Dietary therapies

The ketogenic diet is a high‐fat, moderate‐protein, low‐carbohydrate diet used in the treatment of refractory epilepsy, primarily in children. The ketogenic diet is based on a 3:1 or 4:1 fat to protein and carbohydrate ratio imitating the biochemical effects of fasting (Neal 2010). A low‐carbohydrate, high‐fat diet causes the body to enter a ketotic state, which leads to a reduction in seizure frequency (Huffman 2006). Two systematic reviews of the ketogenic diet report about 16% of children achieve seizure freedom and 33% to 56% experience a greater than 50% reduction in seizure frequency (Lefevre 2000; Keene 2006). A randomised controlled trial found the ketogenic diet to significantly reduce seizure frequency in children, with 38% of those assigned to the diet having a greater than 50% reduction in seizures compared with 6% on the waiting list control group (Neal 2010). A recent Cochrane review (Levy 2012) identified four randomised control trials that assessed the efficacy of the ketogenic diet and concluded that there are short to medium benefits, but tolerability is low over a longer period of time (Bergqvist 2005; Seo 2007; Kassoff 2008; Neal 2008).

Psychological therapies

Cognitive rehabilitation is a neuropsychological treatment program focused on the training of attention and memory functions in people with cognitive deficits due to brain injury. Although training strategies vary for each program, compensation, rather than restoration, is the primary aim of cognitive rehabilitation (Ponds 2006). Compensatory strategies can be either external or internal in nature. External memory strategies include methods of storing information externally and reminding oneself to complete a certain activity, such as the use of an electronic agenda (Koorenhof 2012). Internal memory strategies focus on the elaboration and visualization of information to facilitate later retrieval (Ponds 2006). Research has demonstrated the effectiveness of cognitive rehabilitation programs for the treatment of epilepsy‐related memory problems (Bresson 2007; Helmstaedter 2008; Radford 2011; Koorenhof 2012). However, methodologies and program characteristics varied significantly from study to study. A randomised controlled trial of two cognitive rehabilitation programs, the retraining method and the compensation method, found both treatment strategies to be effective in improving self‐reported cognitive outcomes and quality of life in people with epilepsy (Engelberts 2002).

How the intervention might work

Neurostimulation

The mechanisms of action of vagus nerve stimulation and how it may affect memory are not completely understood. However, the presumed mechanism of action of VNS involves activation of upstream neural networks from the input of the vagus nerve into the medulla and deep brain structures (George 2000). Some functional imaging evidence has indicated that the thalamus and mesial temporal lobe, structures involved in aspects of memory, are activated during vagus nerve stimulation in patients with epilepsy and treatment resistant depression (Lomarev 2002; Henry 2004). Secondary benefits may occur through improved alertness and attentional capacities, as well as the prospect of a reduction in AEDs (Tatum 2001; Kostov 2009; Shahwan 2009).

The mechanisms of deep brain stimulation are also not fully understood. However, mechanisms of action for improvement in seizure control and potential benefits for cognitive functioning likely depend upon the region in the brain selected for neurostimulation, as well as neural network interrelationships with targets of stimulation. Deep brain stimulation therapy in epilepsy thus far has most frequently targeted the anterior nucleus of thalamus (ANT). Three neuroanatomic circuits which are implicated in seizure propagation, the corticothalamic, mammillothalamic, and Papez circuits, involve the ANT (Mirski 1986; Takebayashi 2007; Lega 2010). Thus, DBS may exert a positive, indirect effect on cognition by decreasing abnormal, epileptic activity within these pathways (Zhong 2011). In addition, recent research has found that electrical stimulation during preoperative intracranial recording of another neuroanatomical target, the entorhinal cortex, improved spatial memory in patients with epilepsy (Suthana 2012).

In transcranial magnetic stimulation, an alternating electric current passes through a coil of wire, creating a rapidly changing magnetic field which is concentrated to a specific area, depending on coil shape (Hallett 2000). When a TMS device is positioned, the magnetic field passes through the skull and causes neurons near that area to generate action potentials which may have excitatory or inhibitory effects (Hemond 2007). More specifically, low‐frequency rTMS has been shown to reduce the excitability of neurons in the motor cortex (Chen 1997; Muellbacher 2000; Modugno 2003), which are hyperexcitable in medically refractory epilepsy (Macdonell 2002; Tassinari 2003; Cantello 2007). Research has demonstrated that rTMS may improve cognitive performance by affecting electrical activity of the brain (Klimesch 1999; Jensen 2002; Klimesch 2003;Hamidi 2009).

Dietary therapies

The mechanisms of action of the ketogenic diet and seizure reduction are not completely understood. The ketogenic diet may reduce seizure frequency through a variety of metabolic changes, including ketosis (Huffman 2006). Acetone, which is elevated during ketosis (Seymour 1999; Likhodii 2002), has been shown to have anticonvulsant properties (Likhodii 2003). Some evidence suggests that certain ketone bodies may protect neurons in structures implicated in learning and memory (Massieu 2003). Only one prospective study, which included 65 children being treated with ketogenic diets, has reported benefits in cognitive functioning (Pulsifer 2001). However, many other studies suggest that this intervention may positively affect cognition through seizure reduction and neuroprotection.

Psychological therapies

Cognitive rehabilitation is different from other potential treatments for epilepsy‐associated memory problems as it does not target a neurochemical or structural problem within the brain, but rather improves memory by training patients in a variety of compensatory strategies. These behavioural strategies require practice and aim to reduce patients’ reliance on impaired cognitive faculties, thereby reducing anxiety when faced with a memory task (Engelberts 2002). Mnemonic strategies, such as name‐face associations and the method of loci, use visual elaboration to facilitate memory by providing patients with additional relational cues that may serve to enhance recall (Ponds 2006). In addition, the use of an electronic agenda or other device provides patients with a way to store information externally and prompt themselves to complete a particular task (Koorenhof 2012).

Why it is important to do this review

Cognitive impairments are common in people with epilepsy, especially complaints of memory problems. Currently, few therapies exist for the direct treatment of memory problems in people with epilepsy and few studies have examined the efficacy of specific treatments for memory impairment in epilepsy. A systematic review focused on the use of non‐pharmacologic interventions for the treatment of memory problems in people with epilepsy may serve to generate further interest in this area and stimulate future investigations, leading to the development of therapies for memory problems in epilepsy. The findings of this review and subsequent research will be used to inform clinical practice about the most effective non‐pharmacological treatments for people with epilepsy who are experiencing memory difficulties.

In this systematic review, we aim to assess and summarize the existing evidence of the efficacy and adverse effects of non‐pharmacological interventions; a related future Cochrane review will evaluate the existing evidence of the efficacy and adverse effects of pharmacological interventions for the treatment of memory impairment in people with epilepsy.

Objectives

We aim to assess the effects of non‐pharmacological interventions for the treatment of memory problems in people with epilepsy. More specifically, we aim to evaluate whether people with epilepsy receiving one of these treatments:

  1. may improve their memory, as measured by subjective patient‐reports, objective memory assessments, or both;

  2. may experience associated adverse effects;

  3. may improve in other neuropsychological domains, as assessed by validated measures (memory problems can occur due to a large number of cognitive difficulties such as word finding or executive functioning problems);

  4. may improve their quality of life, as measured by validated scales (memory is often reported as an important influence of individuals' quality of life);

Methods

Criteria for considering studies for this review

Types of studies

We will include studies which meet the following criteria:

  1. All randomised controlled trials; randomised, quasi‐randomised and cluster randomised. These will be studies in which people with epilepsy are randomised to either an intervention group, a control group or a second intervention group.

  2. Studies including a non‐pharmacological intervention such as neurostimulation, dietary or psychological interventions.

  3. Studies assessing participants' memory functioning using a validated objective memory tests.

Types of participants

The following participants will be eligible for the interventions:

  1. People with a diagnosis of epilepsy with or without subjective memory or cognitive complaints, older than 12 years of age. We will exclude studies specifically recruiting children under 12 years of age as this review is focused on adolescent and adult populations, and we wish to exclude studies exploring interventions for children with infantile spasms or West Syndrome.

Comorbid conditions can also have an adverse effect on memory. Therefore, studies reporting interventions for the following participants will be excluded:

  1. People with non‐epileptic attack disorder (NEAD) or comorbid epilepsy and NEAD.

  2. People with epilepsy and other neurological disorders (such as traumatic brain injuries, learning disability).

  3. People with epilepsy and significant psychiatric conditions.

Studies reporting results from a mixed sample of participants who meet both the inclusion and exclusion criteria may be excluded depending on what proportion of participants meet the exclusion criteria. Studies with more than 20% of participants who meet the exclusion criteria will not be considered for inclusion in this review.

Types of interventions

Non‐pharmacological treatments such as vagus nerve stimulation, deep brain stimulation, transcranial magnetic stimulation, the ketogenic diet, and cognitive rehabilitation compared to no treatment, usual care, or another treatment. Studies comparing a non‐pharmacological intervention to a pharmacological intervention will be included in this review.

Types of outcome measures

Studies including subjective and objective memory assessments will be included. However, as there is often a discrepancy between these types of measures, results from subject and objective memory assessments will not be combined and will be analysed separately.

Primary outcomes

  1. Memory improvement: proportion of participants who experience improvement in memory as measured by subjective patient‐reports and/or objective memory assessments. The memory assessments must be from a validated scale including but not limited to: Wechsler Memory Scale‐forth edition (WMS‐IV) (Wechsler 2009), Rivermead Behavioural Memory Test (RBMT) (Wilson 1991), California Verbal Learning Test (CVLT) (Delis 2000) or Prospective and Retrospective Memory Questionnaire (PRMQ) (Crawford 2003). Subjective and objective memory reports will not be combined.

Secondary outcomes

  1. Reduced symptoms of depressed mood or anxiety.

  2. Incidence of adverse or harmful effects:

    1. increased seizure frequency

    2. cognitive side effects

    3. other adverse events

  3. Improvements in other neuropsychological domains (such as language or executive functioning), assessed by validated measures.

  4. Improved quality of life, measured by validated scales.

Search methods for identification of studies

Electronic searches

We will search the following databases:

No language restrictions will be imposed.

Searching other resources

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

We will handsearch three relevant journals from the past five years including Epilepsia, Epilepsy and Behaviour and Epilepsy Research.

We will search the Epilepsia Journal supplements from the past five years for congress proceedings.

Data collection and analysis

Selection of studies

Two review authors (CJ, SMM) will independently review the titles and abstracts of the studies identified by the electronic searches and remove studies that do not meet the inclusion criteria. The same two authors will independently review the full‐text reports to determine eligibility. Any disagreements will be discussed, and if necessary, be resolved by the opinion of the third author (ESL). In the event of multiple reports deriving from one study, the reports will be linked together. A final list of studies to be included in the review will be achieved.

Data extraction and management

Two review authors (CJ, SMM) will independently extract data from each included study, and cross check their results.

The following data will be extracted:

Participants

  • Total sample size

  • Total number of participants allocated to each group

  • Setting

  • Inclusion criteria

  • Age

  • Gender

  • Country

Methods

  • Study design

  • Duration of study

  • Sequence generation and allocation concealment

  • Method of blinding

  • Any other concerns about bias

Outcomes

  • Name and definition of outcome

  • Units of measurement

Results

  • Number of participants allocated to each intervention and control group

  • Sample size for each outcome

  • Missing data

  • Summary data for intervention and control groups (for example, means and standard deviations for all outcomes) See Types of outcome measures

Assessment of risk of bias in included studies

Two review authors (CJ, SMM) will independently assess the risk of bias of each study according to the methods detailed in The Cochrane Handbook for Systematic Reviews of Interventions. We will assess the risk of bias as low, high or unclear risk of bias. Disagreements will be resolved by a third author (ESL)

We will evaluate the following characteristics:

  1. method of randomisation

  2. allocation concealment

  3. preservation of blinding

  4. missing data, selective reporting and bias in analyses (for example, intention to treat)

  5. reporting bias

  6. any other sources of bias such as carry‐over effect accounted for satisfactorily in cross‐over studies

Measures of treatment effect

Ideally, for continuous outcomes such as memory, depression, anxiety, neuropsychological and quality‐of‐life outcomes, we hope to present effect estimates as the mean difference (MD). However, due to the variation in possible outcome measures available to researchers, we expect to present the data as the standardised mean difference (SMD) and risk ratios (RR).

In the case of a study reporting dichotomous data, such as adverse effect, effect estimates will be presented in the form of risk ratios with 95% CI.

Unit of analysis issues

In the event we discover unit of analysis issues across the included studies (for example cross‐over studies, cluster randomised or repeated measures), we plan to:

1. determine whether the methods in such studies were conducted appropriately.

2. combine extracted effect sizes from such studies through a generic inverse‐variance meta‐analysis.

Dealing with missing data

In the event of missing data, we will contact study authors to determine whether the data is missing at random.

Assessment of heterogeneity

Two review authors (CJ, SMM) will independently assess the clinical and methodological heterogeneity visually. Statistical heterogeneity will be assessed using a Chi² test and an I2 statistic. A P value of 0.01 will be used as a cut off for statistical significance on the Chi2 test and a I2 statistic of over 50% will be judged as having significant statistical significance.

Assessment of reporting biases

We will request all protocols from authors of identified studies and investigate reporting bias using the ORBIT matrix system (Kirkham 2010).

To examine publication bias, we will request any unpublished data from authors of the identified studies. We will look for small study effects to establish the likelihood of publication bias. Funnel plots will be examined if there are 10 or more studies that can be combined, in accordance with Cochrane recommendations (Higgins 2011).

Data synthesis

If two or more studies are identified and there is no significant clinical or methodological heterogeneity, they will be combined in a meta‐analysis. If the studies have no significant statistical heterogeneity, we will use a fixed‐effect model for the meta‐analysis. However, if there is significant statistical heterogeneity, we will use a random‐effects model for the meta‐analysis.

For outcomes which report continuous data, we will present the results of the meta‐analysis as a mean difference (MD), however, if there is statistical heterogeneity we will present the results as standardised mean difference (SMD). For outcomes which report dichotomous data, we plan to present the results of the meta‐analysis as risk ratios (RR).

In the event we deem meta‐analysis inappropriate, for example, if there is evidence of clinical heterogeneity, we plan to use a narrative format and discuss all comparisons according to the findings presented within the studies.

We expect to make the following comparisons:

  • intervention group versus control group on the effectiveness of an intervention for improving memory functioning for people with epilepsy as measured by objective assessments;

  • intervention group versus control group on the effectiveness of an intervention for improving memory functioning for people with epilepsy as measured by self‐report assessments

  • intervention group versus control group on the effectiveness of an intervention for reducing adverse effects for people with epilepsy;

  • intervention group versus control group on the effectiveness of an intervention for reducing symptoms of anxiety and depression for people with epilepsy;

  • intervention group versus control group on the effectiveness of an intervention for improving other neuropsychological functioning (for example reducing language or executive functioning difficulties) as measured by both objective and self‐report assessments;

  • intervention group versus control group on the effectiveness of an intervention for improving quality of life in people with epilepsy.

If included studies assessed comparisons not listed above, we plan to carry out a post hoc analysis.

Where possible, we plan to stratify by control group, study design, participant characteristics and measures used to ensure appropriate combination of study data.

Subgroup analysis and investigation of heterogeneity

If there is a sufficient number of studies, we will stratify comparisons according to age, epilepsy type, duration of epilepsy and seizure frequency.

Sensitivity analysis

In the event we identify any inconsistencies or peculiarities, we will conduct a sensitivity analysis. A second meta‐analysis will be carried out using only studies which are considered to be at low risk of bias. The results from this second meta‐analysis will be compared to the reports from the initial analysis which included all studies. If the results are unchanged, the evidence will be considered robust.

Summarising and interpreting results

We will use the GRADE approach to interpret findings (Schunemann 2011). We will use GRADE Profiler Software (GRADEPro 2004), and import data from Review Manager 5 (RevMan 2014), to create 'Summary of Findings' tables for each comparison included in the review for the primary outcomes.

The 'Summary of findings' table for each comparison will include information on overall quality of the evidence from the trials and information of importance for healthcare decision making. The GRADE approach determines the quality of evidence on the basis of an evaluation of eight criteria (risk of bias inconsistency, indirectness, imprecision, publication bias, effect size, presence of plausible confounding that will change effect and dose‐response gradient). We will use these to guide our conclusions and recommendations.