Memory response and neuroimaging correlates of a novel cognitive rehabilitation program for memory problems in epilepsy: A pilot study

Abstract

Background:

Memory deficits are very common in epilepsy, but no standard of care exists to effectively manage them.

Objective:

We assessed effectiveness of cognitive rehabilitation (CR) on memory and neural plasticity in people with epilepsy (PWE) reporting memory impairments.

Methods:

Nine PWE completed 6 weekly sessions adapted from 2 generic CR programs enriched with information regarding epilepsy. Participants completed neuropsychological, mood, and quality of life (QOLIE-31) measures prior and after completion of CR; 5/9 participants also completed pre- and post-CR fMRI while performing a verbal paired associates learning task. FMRI data were analyzed using group spatial independent components analysis methods; paired t-tests compared spatial activations for pre-/post-CR.

Results:

Improvements were seen in immediate recall in Rey Auditory Verbal Learning Task, QOLIE-31, and read word recognition in paired associates task (all p’s≤0.05). FMRI changes comparing pre-to-post CR were noted through increased activation in the left inferior frontal gyrus (IFG) and anterior cingulate and decreased activation in the left superior temporal gyrus; also noted were decreased activations in the default mode network (DMN), right cingulate, right middle temporal gyrus, right supramarginal gyrus, and increased DMN activation in the left cuneus.

Conclusions:

This study demonstrates feasibility of conducting CR program in PWE with fMRI as a mechanistic biomarker. Improvements in cognition and cortical plasticity await confirmation in larger samples.

Keywords

Memory epilepsy memory impairment rehabilitation intervention fMRI paired associate learning task

1 Introduction

Memory impairment is a common co-morbidity in people with epilepsy (PWE), with close to half reporting subjective cognitive deficits (Helmstaedter & Kurthen, 2001). The etiology of memory impairment in epilepsy is likely multifactorial, including the etiology of the seizures themselves, network disruption due to seizures, side effects of anti-seizure drugs (ASDs), and other co-morbidities that can affect learning and memory (i.e. mood disturbances) (Black et al., 2010; Helmstaedter, Elger, & Lendt, 1994; Helmstaedter & Kurthen, 2001; Helmstaedter, Sonntag-Dillender, Hoppe, & Elger, 2004; Kent et al., 2006; Mitchell, Zhou, Chavez, & Guzman, 1992). Cognitive deficits in PWE contribute to poor health-related quality of life (QOL), employment difficulties, and scholastic underperformance (Baker, Jacoby, Buck, Stalgis, & Monnet, 1997; Baker, Nashef, & van Hout, 1997; Leidy, Elixhauser, Vickrey, Means, & Willian, 2001; Markand, Salanova, Whelihan, & Emsley, 2000; Smith, Baker, Davies, Dewey, & Chadwick, 1993). While several modalities including medications, psychotherapy, and behavioral interventions have been proposed as potential remedies for memory impairment associated with epilepsy, the efficacy of such interventions is not well established and there are no specific interventions available that would be considered standard of care (Caller et al., 2016; Del Felice et al., 2017; Fisher, Bortz, Blum, Duncan, & Burke, 2001; Radford, Lah, Thayer, & Miller, 2011).

Cognitive rehabilitation (CR) is a type of behavioral intervention that incorporates varied content and methods of delivery. It has been used to address cognitive impairment associated with several neurological conditions, including traumatic brain injury (TBI) where CR is considered standard of care in treating TBI-associated memory deficits. Various methods of CR have been studied in TBI, and many have demonstrated modest memory improvement (Fetta, Starkweather, & Gill, 2017). In particular, the “Cognitive Symptom Management and Rehabilitation Training (CogSMART)” program demonstrated sustained benefit up to one year after completion of treatment (Twamley, Jak, Delis, Bondi, & Lohr, 2014; Twamley et al., 2015). CR has also been studied in other neurological conditions such as Parkinson’s Disease (PD), multiple sclerosis (MS), Alzheimer’s Disease, and mild cognitive impairment with some improvements in various memory domains. However, improvements vary by condition and specific type of CR investigated (Brueggen et al., 2017; Choi & Twamley, 2013; Hampstead, Stringer, Stilla, Giddens, & Sathian, 2012; Leung et al., 2015; Mitolo, Venneri, Wilkinson, & Sharrack, 2015; Nauta et al., 2017). CR has been studied in PWE including one study of an in-person CR program that demonstrated improvements in verbal memory that were, in part, sustained for 12 weeks after completion; this program utilized a broad memory skills manual, “Making the Most of Your Memory”. (Radford et al., 2011; Radford, Say, Thayer, & Miller, 2010) Another phone-based CR program demonstrated improvements in QOL and attention (Caller et al., 2016). Additionally, CR specifically has been studied in rehabilitation settings after temporal lobe epilepsy surgery with positive results (Farina, Raglio, & Giovagnoli, 2015; Geraldi, Escorsi-Rosset, Thompson, Silva, & Sakamoto, 2017; Mosca et al., 2014).

The improvements in memory with CR are likely explained by neural plasticity (Eliassen, Holland, & Szaflarski, 2008). In PWE, it is recognized that fMRI activations seen with memory processes are overall less robust and of smaller volume than in healthy controls. Furthermore, longer duration of epilepsy leads to further reduction in these activations (Kent et al., 2006; Vannest et al., 2015). One study has demonstrated that PWE who have memory performance similar to healthy controls appear to have compensatory fMRI activations in the right insula, left cuneus, and bilateral anterior cingulate (Eliassen et al., 2008). However, to our knowledge, fMRI changes in response to CR in PWE who have not undergone epilepsy surgery have not been investigated to date.

The goals of this study were to determine feasibility, to generate preliminary efficacy data, and to provide proof of concept to incorporating fMRI to studying memory plasticity in response to a novel, epilepsy-specific CR program in PWE. We hypothesized that there would be significant improvements in verbal memory, mood, and QOL in PWE who complete a 6-week, in person, small group based CR program. Further, we also hypothesized that CR would be associated with change in activation patterns when measured with fMRI, especially in attention and verbal memory circuits.

2 Methods

2.1 Participants

PWE with reported subjective memory complaints were recruited from the University of Alabama at Birmingham. Inclusion criteria included: 1) diagnosis of epilepsy, 2) no other neurological conditions affecting the central nervous system such as a neurodegenerative disorder or medical problems that may affect ability to participate in the study (i.e. severe depression), 3) normal previous neuroimaging (mesial temporal sclerosis allowed), 4) taking no more than 4 concurrent ASDs, 5) completed at least high school education or equivalent, and 6) age 18–65. Exclusion criteria included 1) uncertain diagnosis of epilepsy (i.e. suspected non-epileptic events), 2) abnormal neuroimaging showing focal lesions, 3) known IQ <70 or inability to complete high-school education, 4) severe psychiatric disturbance, 5) any contraindication to MRI at 3T (i.e. metal in body; these participants received the CR program and neuropsychological testing only), and 6) taking ≥4 ASDs or combination therapy of topiramate and zonisamide, which have both been shown to have significantly detrimental effects on cognition and negative effects on fMRI signals (Dupont & Stefan, 2012; Fritz et al., 2005; Marmarou & Pellock, 2005; Szaflarski & Allendorfer, 2012).

2.2 Cognitive rehabilitation program

The overall goal was to create a CR program that built upon other effective memory training programs while tailoring the treatment to PWE through adding disease-specific information important to PWE with cognitive complaints. We based our intervention on a CR program that has been shown to have positive effects one year after treatment in TBI patients called “Cognitive Symptom Management and Rehabilitation Training (CogSMART)” (Twamley et al., 2014; Twamley et al., 2015). We supplemented this with a broad memory skills training program that has been used in PWE with some promising results, “Making the Most of Your Memory: an Everyday Memory Skills Program” (Radford et al., 2011; Radford et al., 2010). Our program was an in-person and small-group format with no more than 5 participants per group. Sessions were administered by trained session facilitators (2 psychology graduate students) with appropriate background and knowledge base; these students worked closely with a supervising neuropsychologist and epileptologist who developed the program and advised on the correct administration of the sessions. The program consisted of 6 sessions, with each session lasting approximately 1.5–2 hours. Sessions were structured and primarily didactic (via PowerPoint presentation), but included opportunity for open discussion and sharing applications of the content. All sessions had a similar overall structure: 1) educational component on memory and/or epilepsy, 2) education on an internal and external memory strategy with active practice in-session, and 3) education on a lifestyle issue pertinent to epilepsy and cognition (e.g., sleep). More detailed description of each of the CR sessions is presented in Table 1.

Table 1
Content of 6 sessions of the cognitive rehabilitation program.

Session Content

Session 1 - Introduction to 3 stages of memory (encoding, storage, and retrieval)

- Learning methods to improve attention and concentration, along with mindfulness, and acronyms for attention strategies

Session 2 - Brief introduction to brain anatomy/function.

- Using calendars to organize and plan tasks, using automatic places (i.e. key bowl)

- The link between nutrition and brain health

Session 3 - Participants learn how seizures negatively impact their memory

- Importance of repetition and rehearsal in memory emphasized

- Neurological benefits of exercise discussed

Session 4 - Participants learn how stress negatively affects mood and memory

- Remembering context and self-prompting as a means to improve retrieval

- Progressive muscle relaxation for stress reduction

Session 5 - Participants learn how sleep impacts health, cognition, and memory

- Visual imagery and linking items using stories as a means to elaborate encoding of items

- Benefits and limitations of using post-it notes to remember “to-do” items.

Session 6 - Review of previously discussed internal and external memory strategies with hands-on practice

- Lifestyle issues revisited

- Participants encouraged to set realistic expectations of themselves and seek support

Session	Content
Session 1	- Introduction to 3 stages of memory (encoding, storage, and retrieval)
	- Learning methods to improve attention and concentration, along with mindfulness, and acronyms for attention strategies
Session 2	- Brief introduction to brain anatomy/function.
	- Using calendars to organize and plan tasks, using automatic places (i.e. key bowl)
	- The link between nutrition and brain health
Session 3	- Participants learn how seizures negatively impact their memory
	- Importance of repetition and rehearsal in memory emphasized
	- Neurological benefits of exercise discussed
Session 4	- Participants learn how stress negatively affects mood and memory
	- Remembering context and self-prompting as a means to improve retrieval
	- Progressive muscle relaxation for stress reduction
Session 5	- Participants learn how sleep impacts health, cognition, and memory
	- Visual imagery and linking items using stories as a means to elaborate encoding of items
	- Benefits and limitations of using post-it notes to remember “to-do” items.
Session 6	- Review of previously discussed internal and external memory strategies with hands-on practice
	- Lifestyle issues revisited
	- Participants encouraged to set realistic expectations of themselves and seek support

2.3 Neuropsychological measures

After signing an IRB-approved informed consent, all participants completed neuropsychological testing to assess baseline cognitive status. Participants performed the same neuropsychological testing battery after completion of the CR program (typically one week after completion of the program), utilizing alternate forms, when available. The Rey Auditory Verbal Learning Test (RAVLT), a validated measure of verbal learning and memory, is administered by reading aloud 15 unrelated words to the participant over five consecutive trials, with an immediate recall trial after each presentation of the word list (learning trials). After the fifth trial, an interference trial of 15 new words is presented, and the participant is asked to recall this new list. The participant is then asked to recall as many words as possible from the original list. Delayed free recall and yes/no recognition trials are administered following a 20 minute delay (Lezak, 1976; Rosenberg, Ryan, & Prifitera, 1984). Immediate and delayed recall scores are generated. The Brief Visuospatial Memory Test-Revised (BVMT-R) is a validated measure of learning and memory that requires participants to reproduce as many figures as possible from an array of 6 figures presented for 10 second trials. There are 3 learning trials. A delayed recall trial is administered after a 25-minute delay. Drawings are judged on accuracy and position; a yes/no recognition trial is administered. (Benedict, Schretlen, Groniger, Dobraski, & Shpritz, 1996) Finally, the Digit Span subtest of the Wechsler Adult Intelligence Scale, Fourth Edition (WAIS-IV) is used to assess working memory (Egeland, 2015).

2.4 Behavioral measures

In order to assess mood and quality of life changes associated with our CR program, participants completed 3 behavioral measures at the time of enrollment (prior to initiation of CR program) and after completion of the CR program. The Quality of Life in Epilepsy-31 (QOLIE-31) is an abbreviated 31-item self-report questionnaire determined to have similar reliability to the original 89-question version (Cramer et al., 1998; Devinsky et al., 1995), The QOLIE-31 is comprised of 7 subscales including seizure worry, overall quality of life, emotional well-being, energy/fatigue, medication effects, work-driving-social limitations, and perceived cognitive function; additionally, a total score can also be calculated, which is what was used for this analysis. The Profile of Mood States (POMS) is a 65-item validated self-report questionnaire that has been widely used to assess mood, and has been used as a mood measure in many epilepsy clinical trials (McNair, Lorr, & Droppleman, 1971). The POMS consists of total mood disturbance (TMD, higher score indicating greater mood disturbance) based on 6 subscales: tension or anxiety, depression or dejection, anger or hostility, vigor or activity, fatigue or inertia, and confusion or bewilderment. High correlation between the depression/dejection subscale of POMS and Beck Depression Inventory has been reported (Griffith et al., 2005). For this study, we used TMD for analyses. Finally, the Epilepsy Self-Efficacy Scale (ESES) is a validated, 33-item self-report questionnaire that measures aspects of self-management in epilepsy. Questions include patients’ perceptions on how they feel they are able to handle taking medications/identifying side effects, lifestyle issues such as socialization, diet, and exercise, and comfort levels of living with seizures and epilepsy (Dilorio, Faherty, & Manteuffel, 1992).

2.5 Functional MRI

2.5.1 Image acquisition

Participants with no contraindication to MRI underwent scanning on a 3T Siemens Prisma fMRI at the University of Alabama at Birmingham’s Civitan International Neuroimaging Laboratory before and after completion of the CR program. High resolution T1 weighted anatomical images (parameters: TR/TE 2300/2.98 ms, FOV 25.6×25.6×19.2 cm, matrix 256×256, flip angle 9°, 1 mm slices, 192 sagittal slices) were acquired on the scanner. T2* images (parameters: TR/TE 2000/35.0 ms, FOV 24.0×24.0 cm, flip angle 90°, 4 mm slices, 30 axial slices) were collected using a clustered-sparse temporal image acquisition, with hemodynamics unrelated to sounds of hardware (HUSH) paradigm (Allendorfer, Kissela, Holland, & Szaflarski, 2012; Schmithorst & Holland, 2004).

2.5.2 FMRI and verbal paired associates learning task

During the fMRI scans, participants completed a verbal learning task intended to activate brain regions involved in verbal memory learning and attention (Mottaghy et al., 1999; Schefft et al., 2008; Vannest et al., 2012; Vannest et al., 2015). We programmed and presented the task with DirectRT software (Version 2008, Empirisoft, www.empirisoft.com). Participants were presented with a “ready” screen for the first 4 seconds. This was followed by 60 trials lasting 12 seconds each: a word pair was presented during the first 6 seconds of each trial (along with MRI silence to allow for participant response), followed by 6 seconds of fMRI data acquisition (3 image volumes collected) and an on-screen instruction for participants to “STOP” and remain silent during this time. Audio responses were recorded during the periods of MRI silence and performance during scanning was monitored. Of note, the task was not dependent on the verbal response of the participant; the task continued uninterrupted regardless of a response by the participant (Vannest et al., 2012; Vannest et al., 2015). A total of 60 related word pairs were presented to participants over 2 conditions (30 words per condition). In the “read” condition, participants were presented a pair of words and instructed to read the second word aloud (i.e. “bed-pillow”). In the “generate” condition participants were presented the first word and the first letter of the second word followed by asterisks indicating the number of missing letters, and the participant was instructed to generate the second word aloud (i.e. “spider-w**). All linguistic relationships were equally represented (i.e. associates, category members, synonyms, antonyms, and rhymes) (Siegel, Allendorfer, Lindsell, Vannest, & Szaflarski, 2012). The HUSH paradigm has been further described in detail previously (Allendorfer et al., 2012).

Immediately after the scan was complete, a post-test was performed to assess recognition of the second word in each of the 60 word pairs that were presented in the scanner. The post-test was a 3-item forced choice format. The second word of the previously presented word pairs in the scanner and two foils were presented on a computer screen, and participants chose which of the 3 words they recognized from the scanner task by pressing a designated key on the computer.

2.5.3 FMRI processing

Imaging data were pre-processed with Analysis of Functional NeuroImages software (AFNI) via a standard pre-processing pipeline (Cox, 1996). First, the functional images were split into 3 parts (first, second, and third volumes for each stimulus) to account for the signal intensity changes in the blood-oxygenation level dependent (BOLD) response over time (Schmithorst & Holland, 2004). The functional images were then motion-corrected in AFNI and then registered to the anatomical images (MNI) using FMRI software library (FSL) (Jenkinson, Bannister, Brady, & Smith, 2002), and smoothed to an effective smoothness of a Gaussian FWHM of 6 mm using AFNI’s 3dBlurToFWHM.

After pre-processing, group independent components analysis (ICA) was carried out using the Group ICA fMRI Toolbox (GIFT) in MATLAB, which yielded 41 independent components for each of the pre- and post-CR scans. The number of components chosen was based on a similar study with the same in-scanner verbal paired associates learning task (Vannest et al., 2015). Correlations with task time series were run in GIFT, and noise-related correlations were discarded. Given the small sample size, components correlated with the task-time series and meeting an arbitrary minimum correlation threshold of |r| = 0.15 were retained. If a component met threshold for task-relatedness across more than one of the three image volumes, the most highly correlated volume was selected for further analysis. Individual components meeting the above thresholds were visually examined in pre- and post- conditions for spatial sameness and grouped together in preparation for analysis.

2.6 Analysis

A participant was considered adherent to the CR program if they attended at least 5/6 of the CR sessions. All neuropsychological testing data and behavioral measures were scored according to the respective assessments’ scoring manuals; in scanner paired associates task were scored based on percent accuracy on the post-test. Means were calculated for pre- and post-CR for each assessment, and paired samples t-tests were performed for analysis. Because of a small sample, effect sizes (Cohen’s d) were also calculated to facilitate power analyses for replication studies.

After fMRI data were pre-processed and correlations were run in GIFT, paired sample t-tests were then performed (3dttest++) and results visualized in AFNI in order to compare spatial activations pre- and post-CR, with voxelwise uncorrected p-value set at ≤0.01 and minimum cluster threshold set at 20 voxels. The data were not corrected for multiple comparisons given the preliminary nature and the small sample size in this study.

3 Results

Of 12 participants who completed consent and enrolled in the study, 9 participants were considered adherent to the CR program by attending at least 5/6 sessions; five of these participants had no contraindication to fMRI and completed pre- and post-CR fMRI. The most common reason for drop out was due to lack of reliable transportation to study sessions. Participant demographics are listed in Table 2. From pre- to post-CR, there were statistically significant improvements in the RAVLT Immediate Recall raw composite score (p = 0.03, d = 0.389), QOLIE-31 total score (p = 0.008; d = 0.937), BVMT-R Delayed Recall raw score (p = 0.02, d = 0.937), and in the fMRI verbal paired associates learning task read condition (p = 0.03, d = 1.567). There were non-significant improvements seen in POMS total mood disturbance (lower score indicating improved TMD), WAIS-IV Digit Span, BVMT-R T-score for delayed recall, RAVLT Delayed Recall scores (raw and standard), and verbal paired associates learning task generate condition. Of importance, mean baseline (pre-CR) performance in the generate condition was non-significantly better than the read condition (77.8% vs. 69.4%, p = 0.15); this is in line with previous studies utilizing this task; it is hypothesized that people have improved recall when made to actively generate the second word in the word pair (Vannest et al., 2012). Given the small sample size, these results were not corrected for learning effects. There were no differences seen from pre- to post-CR in any of the QOLIE-31 and POMS subscales. Further, there was no significant worsening in any scores from pre- to post-CR. Data and results of analyses for all assessments are presented in Table 3.

Table 2
Demographics of study participants

Demographic N = 9

Gender 4 Male, 5 Female

Mean age 40.8±12.3 years (range 23–61 years)

Years of education 15±2.8 years (range 12–21 years)

Handedness 8 right, 1 left

Average Number of Anti-Seizure Drugs at Enrollment 2.2 (range 1–4)

Type of Epilepsy Idiopathic generalized epilepsy: 2

Left temporal lobe epilepsy: 1

Right temporal lobe epilepsy: 1

Bilateral temporal lobe epilepsy: 3

Right frontal lobe epilepsy: 1

Unclear seizure onset localization: 1

Demographic	N = 9
Gender	4 Male, 5 Female
Mean age	40.8±12.3 years (range 23–61 years)
Years of education	15±2.8 years (range 12–21 years)
Handedness	8 right, 1 left
Average Number of Anti-Seizure Drugs at Enrollment	2.2 (range 1–4)
Type of Epilepsy	Idiopathic generalized epilepsy: 2
	Left temporal lobe epilepsy: 1
	Right temporal lobe epilepsy: 1
	Bilateral temporal lobe epilepsy: 3
	Right frontal lobe epilepsy: 1
	Unclear seizure onset localization: 1

Table 3

Results of Pre- and Post-CR neuropsychological testing, behavioral measures, and fMRI-associated memory task. Data presented as means±standard deviation. Analysis performed with paired t-tests, with bold p-values statistically significant

	Pre-CR	Post-CR	p-value	Effect size (d)
RAVLT Immediate Recall total Word List Raw Score (Trials I-V) (n = 9)	47.6±9.1 words	51.1±8.9 words	0.030	0.389
RAVLT Immediate Recall Trials I-V Standard Score (n = 9)	92.6±16.8	99.2±16.3	0.032	0.399
RAVLT Short Delay Recall Raw Score (n = 9)	8.6±4.2	8.7±4.5	0.910	0.023
RAVLT Long Delay Recall Raw Score (n = 9)	7.4±4.9	7.6±4.4	0.880	0.043
BVMT-R Raw Score Delayed Recall (n = 9)	7.9±3.6	9.6±2.5	0.020	0.549
BVMT-R T-score Delayed Recall (n = 9)	41.9±16.5	49.3±14.1	0.058	0.482
WAIS-IV Digit Span (n = 9)	14.1±2.6	15.2±3.6	0.214	0.350
Pairs Encoding-Read (n = 5) (percent remembered)	69.4±6.5	82.8±10.2	0.030	1.567
Pairs Encoding-Generate (n = 5) (percent remembered)	77.8±12	78.3±14.5	0.950	0.038
POMS Total Mood Disturbance (n = 9)	62±31.6	46.7±13.9	0.300	0.627
QOLIE-31 total score (n = 9)	51.4±11.1	60.2±7.3	0.008	0.937
ESES (n = 9)	254.1±19.8	250.6±18.7	0.514	0.182

RAVLT: Rey Auditory Verbal Learning Test, BVMT-R: Brief Visual Memory Test-Revised, WAIS-IV: Wechsler Adult Intelligence Scale-Fourth Edition, POMS: Profile of Mood States, QOLIE: Quality of Life in Epilepsy, ESES: Epilepsy Self-Efficacy Scale.

Overall, fMRI spatial patterns associated with the verbal paired associates learning task were similar to those previously published (Vannest et al., 2012; Vannest et al., 2015). When comparing pre- to post-CR scans, in the generate condition we noted increased spatial activation in the left inferior frontal gyrus (IFG) and bilateral anterior cingulate gyrus, and decreased activation in the left superior temporal gyrus and left cerebellar tonsil. In the read condition, we noted increased activation in the left cuneus and left IFG and changes in the default mode network via decreased activation of the right cingulate gyrus, right middle temporal gyrus, right supramarginal gyrus, and increased activation in the left cuneus. Imaging results are presented in Fig. 1.

Fig.1

Changes in functional MRI activations seen in response to cognitive rehabilitation using verbal paired associates learning task. Top row demonstrates increases and decreases in blood-oxygenation level dependent (BOLD) contrast in response to active learning condition (generate); bottom row demonstrates similar findings in response to passive learning condition (read).

4 Discussion

Our preliminary data from a small number of participants indicate promising improvements in response to our epilepsy-specific CR program in aspects of the RAVLT, verbal paired associates learning task read condition, and BVMT-R. Additionally, we noted significant increases in the QOLIE-31 total score after completion of the CR program; this increase also neared clinical significance on average (mean difference of 8.8 points), as a clinically significant change has previously been determined to be 11.8 points (Wiebe, Matijevic, Eliasziw, & Derry, 2002). Although changes in these scores were not significant, we also noted moderate effect sizes in the BVMT-R T-score, WAIS-IV digit span, and POMS TMD. Given that memory problems impact health-related QOL, our improvements in the QOLIE-31 may be explained in part by the improvements seen in memory; however, a bi-directional relationship also cannot be excluded (Baker, Jacoby, et al., 1997; Baker, Nashef, et al., 1997; Leidy et al., 2001).

Findings from this preliminary CR study are promising in view of the findings of other behavioral interventions for the treatment of memory deficits in PWE who have not undergone epilepsy surgery published to date (Caller et al., 2016; Radford et al., 2011; Radford et al., 2010). One program was structured similarly to ours and utilized a generic memory skills manual (“Making the Most of Your Memory”). There were improvements seen immediately after completion and in part sustained at 3 months after completion of the program in RAVLT delayed recall scores. (Radford et al., 2011) A program that was home/phone based (The HOme Based Self-management and Cognitive Training CHanges lives; HOBSCOTCH) demonstrated a mean improvement of 7 points on the QOLIE-31 and improvements in attention measures that were significantly better than that of controls (Caller et al., 2016). However, both of these investigations had larger sample sizes and were compared to a control group, which provide stronger support for their hypotheses.

In the imaging portion of our study, we identified changes in neural circuits that support verbal memory and language in response to our CR program. Notably, in both of the generate and read conditions of the verbal memory fMRI encoding task, when comparing post- to pre- CR scans, we noted increased activation in the left IFG, while in the generate condition alone we observed decreased activation in the left superior temporal gyrus. The left IFG is involved in expressive language function, while the left STG is involved in receptive language. (Kim, Karunanayaka, Privitera, Holland, & Szaflarski, 2011) Pre-post CR increase in left IFG activation could be reflective of increased verbal processing that may contribute to the improvement in post-scan memory. Decreased left STG activation for the generate condition may be indicative of less reliance of receptive language centers in verbal memory processing. In the generate condition, we also noted increased activation of the anterior cingulate gyrus; these findings suggest possible increase in areas important in attention, executive function, and working memory (Cersosimo & Benarroch, 2013). Decreased activation was observed in the left cerebellar tonsil in the generate condition after CR completion, which is in part responsible for motor speech coordination/articulation (Starowicz-Filip et al., 2017). One possible explanation for this finding is higher efficiency of motor speech and articulation centers when producing speech as a response to CR. Finally, in the read condition we noted decreased activation in portions the default mode network (DMN) post-CR. The DMN is a network involved in functions of the resting brain state, with greater activity generally seen when an individual is focused internally rather than with attention demanding tasks; decreased activation could be indicative of improved attention in response to CR (Cechetto & Weishaupt, 2017, Benarroch, 2012). It is important to note, however, that given the small sample size, our fMRI findings might not be stable over time or may differ from these findings in our preliminary study. Additionally, in this preliminary study we only investigated fMRI changes with one recognition verbal memory task, which is only one aspect of memory.

To our knowledge, ours is the first study on fMRI changes in response to CR in PWE who have not received epilepsy surgery; however, CR has been studied in PWE who have undergone temporal lobe epilepsy surgery. In one study, during an in-scanner word generation task, comparing pre- and post-CR activations in patients post-left temporal lobe resection, there was increased activation seen in the anterior prefrontal cortex, pars opercularis, inferior prefrontal gyrus, cingulate cortex, and superior temporal gyrus (Geraldi et al., 2017). While some of the findings are in line with ours (such as increased activation in the inferior frontal gyrus and anterior cingulate), it is of interesting contrast that this study demonstrated increased STG activation where we demonstrated decreased STG activation after completion of the CR program. These discrepancies are likely related to sample sizes and types of fMRI probes. Further, a case report of a patient who underwent temporal lobectomy for drug resistant epilepsy and underwent a tailored CR program focused on visual imagery demonstrated a shift of activation from the anterior to posterior network comparing pre- to post-CR fMRI. The authors posited that this could be due to less requirement of attentional and executive process and more involvement of medial and posterior regions after rehabilitation which allowed the patient to efficiently use the visual imagery skills taught in the program (Mosca et al., 2014). These findings do differ from ours as we demonstrated increased activation in attentional networks; however, our program focuses on a wider range of strategies, on which the participants may have relied to improve their memory.

The neuroimaging effects of CR have been studied in other neurological conditions as well, such as TBI, PD, and MS (Campbell, Langdon, Cercignani, & Rashid, 2016; Diez-Cirarda et al., 2017; Galetto & Sacco, 2017; Kim et al., 2009; Laatsch & Krisky, 2006; Parisi et al., 2014). In one TBI study, increased activations in response to CR were demonstrated in the cerebellum, bilateral posterior parietal, left occipital, anterior cingulate, and precuneus (Galetto & Sacco, 2017; Kim et al., 2009). These findings were in part thought to be due to improved attention and increased neural efficiency resulting from CR. The findings in this study contrast ours somewhat as we noted decreased activation in the cerebellar tonsil after completion of CR. Therefore, the clinical significance of these differences is unclear and awaits further investigation. Another TBI study demonstrated a significant reduction in the pattern of fMRI activations post-CR during a reading task compared to baseline that appeared overall similar to activations present in healthy controls; this suggested that CR could make neural circuitry/activations during this task more “normal-appearing” (Laatsch & Krisky, 2006). While our analysis did not directly compare post-CR scans to healthy controls, the mechanisms of higher performance in memory in PWE are somewhat different, with evidence in other studies indicating that instead of looking more “normal”, PWE develop compensatory activations in order to perform at the level of healthy controls (Eliassen et al., 2008). In PD, CR induced increased activations in the left middle temporal and inferior frontal lobes, with positive correlations seen between brain connectivity and activation and the cognitive performance post-CR (Diez-Cirarda et al., 2017). In MS, increased activations in bilateral prefrontal cortex and right temporo-parietal regions were observed in one study while another study demonstrated increased anterior cingulate activation and associated networks important for executive function and attention after intervention (Campbell et al., 2016; Parisi et al., 2014). This study also noted the ability of CR to have compensatory-appearing action on connectivity that could be resilient to MS pathology. This latter study had comparable changes in activations to what we demonstrated in our small sample of PWE.

There are several limitations to this study. First, the sample size is too small to reliably detect the full extent of differences between pre- and post-CR, and the data are not compared to a control group. Effect sizes were calculated in this preliminary analysis with the expectation of guiding future statistical analysis. Whenever repeated neuropsychological testing is performed, there is unavoidable potential of learning effect; however, alternate forms of the tests were used in the pre- and post-CR testing sessions. Additionally, given small sample size, the type and laterality of epilepsy, seizure frequency, duration of epilepsy, and education level of the participants were not accounted for. These questions are important to address as type and laterality of epilepsy can influence patterns of memory deficit while duration of epilepsy, seizure frequency, and level of education can affect severity of memory deficits (Andrewes, Puce, & Bladin, 1990; Black et al., 2010; Helmstaedter, Kurthen, Lux, Reuber, & Elger, 2003; Mameniskiene, Rimsiene, & Puronaite, 2016). Mood disturbance is also a negative factor for memory in epilepsy, thus it is unclear if the memory improvements seen in our small sample are dependent on the improvements seen in mood (Paradiso, Hermann, Blumer, Davies, & Robinson, 2001). Finally, an in-person method of intervention delivery could limit program adherence; this is a particular concern in PWE as those with uncontrolled seizures cannot drive and may not have reliable transportation. However, it was felt at this preliminary stage that in-person sessions were optimal to maximize understanding of concepts and real-time feedback with the hands-on practice provided for the memory enhancing tools taught in the sessions. Finally, it is unclear how much the increased social interaction due to the in-person visits contributed to QOL and mood changes seen after the CR program was completed. There is an overlap in memory and mood circuits; thus, this could be considered a study confounder. However, we did not see significant improvement in mood as measured via the POMS TMD, so it is likely the fMRI changes seen comparing pre- to post-CR are more likely explained by improvement in memory.

Future directions for our CR program will need to address the current limitations. More participants will need to undergo the CR program and will need to be compared to a control group with repetitive behavioral and fMRI measures in order to increase statistical power and validity of results. Given behavioral interventions such as ours are somewhat difficult to establish a true “placebo” group, a possible solution to this would be to implement a “wait-list control” study design where participants undergo repeat baseline testing without intervention and thus serve as their own controls. A single type of epilepsy (i.e. temporal lobe epilepsy) should either be studied alone or should be an important analysis variable along with epilepsy laterality, seizure frequency, duration of epilepsy, and level of education. In order to determine if positive effects of the CR program on memory and behavioral measures are sustained, follow-up testing should be performed to determine durability of response. While for our preliminary study, the CR program was administered by 2 trained facilitators, for a larger study, fidelity of the program must be ensured; this would be accomplished by video recording all sessions, reviewing sessions by program developers, and these reviewers providing regular feedback to session facilitators. Additionally, other delivery methods of the intervention should be considered and tested, especially telemedicine, in order to allow for maximal adherence to the program and to make the CR program more widely available. Incorporation of strategies that have been successful in other forms of rehabilitation should be considered, such as a “transfer package” or intervention-based follow up after the program is complete in order to facilitate transfer of the skills learned in the program into daily life (Morris et al., 2019). Finally, with a larger sample, learning effects from repeated testing can be controlled for in power calculations.

In regards to the neuroimaging portion of our study, increasing sample size and comparing treatment vs. a control group will provide clarity to our results and serve as a mechanistic biomarker for CR’s effects on neural plasticity. While only spatial differences pre- vs. post-CR were compared in this preliminary analysis (given small sample size), in the future it will be helpful if BOLD signal intensity is also analyzed to determine any significant changes to this in response to CR. As mentioned previously, studying one particular type of epilepsy (e.g., TLE) and also controlling for laterality will improve the quality of the imaging data. Further, other domains and aspects of memory (i.e. working memory, visual memory, etc.) will need to be investigated in the scanner, as with this preliminary study we only investigated fMRI response to a verbal recognition task. In such study, neuropsychological test scores could be correlated with imaging findings to determine if imaging changes are associated with or facilitate function improvement. Another future direction would be to determine if post-CR fMRI activations in PWE demonstrate compensatory activations present in response to CR that differ from healthy controls. Imaging could also be helpful if there are activations present pre-CR that could predict response to the intervention, which could tailor patient selection to receive the intervention.

In summary, we have developed and tested a disease-specific CR program for epilepsy in a small number of participants with promising results in improvement in verbal memory, visual memory, and QOL after completion of the program. Further, this is the first study investigating changes in neural plasticity in response to CR in PWE who have not undergone epilepsy surgery via fMRI. Our results in this small sample support performing a larger, controlled study investigating the effects of this program in PWE.

Footnotes

Acknowledgments

The authors thank Amber Fahey, PhD, for help with study design and data collection.

References

Allendorfer, J.B., Kissela, B.M., Holland, S.K., & Szaflarski, J.P. (2012). Different patterns of language activation in post-stroke aphasia are detected by overt and covert versions of the verb generation fMRI task. Medical Science Monitor 18(3), Cr135–137.

Andrewes, D.G., Puce, A., & Bladin, P.F. (1990). Post-ictal recognition memory predicts laterality of temporal lobe seizure focus: Comparison with post-operative data. Neuropsychologia 28(9), 957–967.

Baker, G.A., Jacoby, A., Buck, D., Stalgis, C., & Monnet, D. (1997). Quality of life of people with epilepsy: A European study. Epilepsia 38(3), 353–362.

Baker, G.A., Nashef, L., van Hout

B.A.

(1997). Current issues in the management of epilepsy: The impact of frequent seizures on cost of illness, quality of life, and mortality. Epilepsia 38(Supplement 1), S1–8.

Benarroch, E. (2012). Chapter - Autonomic Control. In Robertson

(Ed.,) Primer on the Autonomic Nervous System, 3rd Edition (pp. 9–12): Elsevier.

Benedict, R., Schretlen, D., Groniger, L., Dobraski, M., & Shpritz, B. (1996). Revision of the Brief Visuospatial Memory Test: Studies of Normal Performance, Reliability, and Validity. Psychological Assessment 8(2), 145–153.

Black, L.C., Schefft, B.K., Howe, S.R., Szaflarski, J.P., Yeh, H.S., & Privitera, M.D. (2010). The effect of seizures on working memory and executive functioning performance. Epilepsy and Behavior 17(3), 412–419.

Brueggen, K., Kasper, E., Ochmann, S., Pfaff, H., Webel, S., Schneider, W., & Teipel, S. (2017). Cognitive Rehabilitation in Alzheimer’s Disease: A Controlled Intervention Trial. Journal of Alzheimers Disease 57(4), 1315–1324.

Caller, T.A., Ferguson, R.J., Roth, R.M., Secore, K.L., Alexandre, F.P., Zhao, W., ... & Jobst, B.C. (2016). A cognitive behavioral intervention (HOBSCOTCH) improves quality of life and attention in epilepsy. Epilepsy and Behavior 57(Pt A), 111–117.

10.

Campbell, J., Langdon, D., Cercignani, M., & Rashid, W. (2016). A Randomised Controlled Trial of Efficacy of Cognitive Rehabilitation in Multiple Sclerosis: A Cognitive, Behavioural, and MRI Study. Neural Plasticity 2016, 4292585.

11.

Cechetto, D.F., & Weishaupt, N. (2017). Introduction, In Cechetto

D. F. .

& Weishaupt

N. .

(Eds.)The Cerebral Cortex in Neurodegenerative and Neuropsychiatric Disorders (pp. xv–xix). San Diego: Academic Press.

12.

Cersosimo, M.G., & Benarroch, E.E. (2013). Chapter 5 - Central control of autonomic function and involvement in neurodegenerative disorders. In Buijs

R.M.

& Swaab

D.F.

(Eds.), Handbook of Clinical Neurology (Vol. 117, pp. 45–57): Elsevier.

13.

Choi, J., & Twamley, E.W. (2013). Cognitive rehabilitation therapies for Alzheimer’s disease: A review of methods to improve treatment engagement and self-efficacy. Neuropsychology Review 23(1), 48–62.

14.

Cox, R.W. (1996). AFNI: Software for analysis and visualization of functional magnetic resonance neuroimages. Computers and Biomedical Research 29(3), 162–173.

15.

Cramer, J.A., Perrine, K., Devinsky, O., Bryant-Comstock, L., Meador, K., & Hermann, B. (1998). Development and cross-cultural translations of a 31-item quality of life in epilepsy inventory. Epilepsia 39(1), 81–88.

16.

Del Felice, A., Alderighi, M., Martinato, M., Grisafi, D., Bosco, A., Thompson, P.J., ... & Masiero

(2017). Memory Rehabilitation Strategies in Nonsurgical Temporal Lobe Epilepsy: A Review. American Journal of Physical Medicine & Rehabilitation 96(7), 506–514.

17.

Devinsky, O., Vickrey, B.G., Cramer, J., Perrine, K., Hermann, B., Meador, K., & Hays, R.D. (1995). Development of the quality of life in epilepsy inventory. Epilepsia 36(11), 1089–1104.

18.

Diez-Cirarda, M., Ojeda, N., Pena, J., Cabrera-Zubizarreta, A., Lucas-Jimenez, O., Gomez-Esteban, J.C., .... & Ibarretxe-Bilbao

(2017). Increased brain connectivity and activation after cognitive rehabilitation in Parkinson’s disease: A randomized controlled trial. Brain Imaging and Behavior 11(6), 1640–1651.

19.

Dilorio, C., Faherty, B., & Manteuffel, B. (1992). The development and testing of an instrument to measure self-efficacy in individuals with epilepsy. The Journal of Neuroscience Nursing 24(1), 9–13.

20.

Dupont, S., & Stefan, H. (2012). Zonisamide in clinical practice. Acta Neurologica Scandinavica, Supplementum 29–35.

21.

Egeland, J. (2015). Measuring Working Memory With Digit Span and the Letter-Number Sequencing Subtests From the WAIS-IV: Too Low Manipulation Load and Risk for Underestimating Modality Effects. Applied Neuropsychology Adult 22(6), 445–451.

22.

Eliassen, J.C., Holland, S.K., & Szaflarski, J.P. (2008). Compensatory brain activation for recognition memory in patients with medication-resistant epilepsy. Epilepsy and Behavior 13(3), 463–469.

23.

Farina, E., Raglio, A., & Giovagnoli, A.R. (2015). Cognitive rehabilitation in epilepsy: An evidence-based review. Epilepsy Research 109, 210–218.

24.

Fetta, J., Starkweather, A., & Gill, J.M. (2017). Computer-Based Cognitive Rehabilitation Interventions for Traumatic Brain Injury: A Critical Review of the Literature. The Journal of Neuroscience Nursing 49(4), 235–240.

25.

Fisher, R.S., Bortz, J.J., Blum, D.E., Duncan, B., & Burke, H. (2001). A pilot study of donepezil for memory problems in epilepsy. Epilepsy and Behavior 2(4), 330–334.

26.

Fritz, N., Glogau, S., Hoffmann, J., Rademacher, M., Elger, C.E., & Helmstaedter, C. (2005). Efficacy and cognitive side effects of tiagabine and topiramate in patients with epilepsy. Epilepsy and Behavior 6(3), 373–381.

27.

Galetto, V., & Sacco, K. (2017). Neuroplastic Changes Induced by Cognitive Rehabilitation in Traumatic Brain Injury: A Review. Neurorehabilitation and Neural Repair 31(9), 800–813.

28.

Geraldi, C.V., Escorsi-Rosset, S., Thompson, P., Silva, A. C. G., & Sakamoto, A.C. (2017). Potential role of a cognitive rehabilitation program following left temporal lobe epilepsy surgery. Arquivos de Neuro-Psiquiatria 75(6), 359–365.

29.

Griffith, N.M., Szaflarski, J.P., Szaflarski, M., Kent, G.P., Schefft, B.K., Howe, S.R., & Privitera, M.D. (2005). Measuring depressive symptoms among treatment-resistant seizure disorder patients: POMS Depression scale as an alternative to the BDI-II. Epilepsy and Behavior 7(2), 266–272.

30.

Hampstead, B.M., Stringer, A.Y., Stilla, R.F., Giddens, M., & Sathian, K. (2012). Mnemonic strategy training partially restores hippocampal activity in patients with mild cognitive impairment. Hippocampus 22(8), 1652–1658.

31.

Helmstaedter, C., Elger, C.E., & Lendt, M. (1994). Postictal courses of cognitive deficits in focal epilepsies. Epilepsia 35(5), 1073–1078.

32.

Helmstaedter, C., & Kurthen, M. (2001). Memory and epilepsy: Characteristics, course, and influence of drugs and surgery. Current Opinion in Neurology 14(2), 211–216.

33.

Helmstaedter, C., Kurthen, M., Lux, S., Reuber, M., & Elger, C.E. (2003). Chronic epilepsy and cognition: A longitudinal study in temporal lobe epilepsy. Annals of Neurology 54(4), 425–432.

34.

Helmstaedter, C., Sonntag-Dillender, M., Hoppe, C., & Elger, C.E. (2004). Depressed mood and memory impairment in temporal lobe epilepsy as a function of focus lateralization and localization. Epilepsy and Behavior 5(5), 696–701.

35.

Jenkinson, M., Bannister, P., Brady, M., & Smith, S. (2002). Improved optimization for the robust and accurate linear registration and motion correction of brain images. Neuroimage 17(2), 825–841.

36.

Kent, G.P., Schefft, B.K., Howe, S.R., Szaflarski, J.P., Yeh, H.S., & Privitera, M.D. (2006). The effects of duration of intractable epilepsy on memory function. Epilepsy and Behavior 9(3), 469–477.

37.

Kim, K.K., Karunanayaka, P., Privitera, M.D., Holland, S.K., & Szaflarski, J.P. (2011). Semantic association investigated with functional MRI and independent component analysis. Epilepsy and Behavior 20(4), 613–622.

38.

Kim, Y.H., Yoo, W.K., Ko, M.H., Park, C.H., Kim, S.T., & Na, D.L. (2009). Plasticity of the attentional network after brain injury and cognitive rehabilitation. Neurorehabilitation and Neural Repair 23(5), 468–477.

39.

Laatsch, L., & Krisky, C. (2006). Changes in fMRI activation following rehabilitation of reading and visual processing deficits in subjects with traumatic brain injury. Brain Injury 20(13-14), 1367–1375.

40.

Leidy, N.K., Elixhauser, A., Vickrey, B., Means, E., & Willian, M.K. (2001). Seizure frequency and the health-related quality of life of adults with epilepsy. Neurology 57(11 Supplement 4), 69–73.

41.

Leung, I.H., Walton, C.C., Hallock, H., Lewis, S.J., Valenzuela, M., & Lampit, A. (2015). Cognitive training in Parkinson disease: A systematic review and meta-analysis. Neurology 85(21), 1843–1851.

42.

Lezak, M. (1976), Neuropsychological assessment. New York: Oxford University Press.

43.

Mameniskiene, R., Rimsiene, J., & Puronaite, R. (2016). Cognitive changes in people with temporal lobe epilepsy over a 13-year period. Epilepsy and Behavior 63, 89–97.

44.

Markand, O.N., Salanova, V., Whelihan, E., & Emsley, C.L. (2000). Health-related quality of life outcome in medically refractory epilepsy treated with anterior temporal lobectomy. Epilepsia 41(6), 749–759.

45.

Marmarou, A., & Pellock, J.M. (2005). Zonisamide: Physician and patient experiences. Epilepsy Research 64(1-2), 63–69.

46.

McNair, D.M., Lorr, M., & Droppleman, L. (1971), Profile of Mood States. San Diego, CA: Education and Industrial Testing Service.

47.

Mitchell, W.G., Zhou, Y., Chavez, J.M., & Guzman, B.L. (1992). Reaction time, attention, and impulsivity in epilepsy. Pediatric Neurology 8(1), 19–24.

48.

Mitolo, M., Venneri, A., Wilkinson, I.D., & Sharrack, B. (2015). Cognitive rehabilitation in multiple sclerosis: A systematic review. Journal of the Neurological Sciences 354(1-2), 1–9.

49.

Morris, D.M., Taub, E., Mark, V.W., Liu, W., Brenner, L., Pickett, T., .... & Uswatte

(2019). Protocol for a Randomized Controlled Trial of CI Therapy for Rehabilitation of Upper Extremity Motor Deficit: The Bringing Rehabilitation to American Veterans Everywhere Project. The Journal of Head Trauma Rehabilitation. Doi: 10.1097/HTR.0000000000000460

50.

Mosca, C., Zoubrinetzy, R., Baciu, M., Aguilar, L., Minotti, L., Kahane, P., & Perrone-Bertolotti, M. (2014). Rehabilitation of verbal memory by means of preserved nonverbal memory abilities after epilepsy surgery. Epilepsy and Behavior Case Reports 2, 167–173.

51.

Mottaghy, F.M., Shah, N.J., Krause, B.J., Schmidt, D., Halsband, U., Jancke, L., & Muller-Gartner, H.W. (1999). Neuronal correlates of encoding and retrieval in episodic memory during a paired-word association learning task: A functional magnetic resonance imaging study. Experimental Brain Research 128(3), 332–342.

52.

Nauta, I.M., Speckens, A.E.M., Kessels, R.P.C., Geurts, J.J.G., de Groot

, Uitdehaag, B.M.J., ... & de Jong, B.A. (2017). Cognitive rehabilitation and mindfulness in multiple sclerosis (REMIND-MS): A study protocol for a randomised controlled trial. BMC Neurology 17(1), 201.

53.

Paradiso, S., Hermann, B.P., Blumer, D., Davies, K., & Robinson, R.G. (2001). Impact of depressed mood on neuropsychological status in temporal lobe epilepsy. Journal of Neurology, Neurosurgery and Psychiatry 70(2), 180–185.

54.

Parisi, L., Rocca, M.A., Valsasina, P., Panicari, L., Mattioli, F., & Filippi, M. (2014). Cognitive rehabilitation correlates with the functional connectivity of the anterior cingulate cortex in patients with multiple sclerosis. Brain Imaging and Behavior 8(3), 387–393.

55.

Radford, K., Lah, S., Thayer, Z., & Miller, L.A. (2011). Effective group-based memory training for patients with epilepsy. Epilepsy and Behavior 22(2), 272–278.

56.

Radford, K., Say, M., Thayer, Z., & Miller, L. (2010), Making the Most of Your Memory: An everyday memory skills program. Sydney: ASSBI Resources.

57.

Rosenberg, S.J., Ryan, J.J., & Prifitera, A. (1984). Rey Auditory-Verbal Learning Test performance of patients with and without memory impairment. Journal of Clinical Psychology 40(3), 785–787.

58.

Schefft, B.K., Dulay, M.F., Fargo, J.D., Szaflarski, J.P., Yeh, H.S., & Privitera, M.D. (2008). The use of self-generation procedures facilitates verbal memory in individuals with seizure disorders. Epilepsy and Behavior 13(1), 162–168.

59.

Schmithorst, V.J., & Holland, S.K. (2004). Event-related fMRI technique for auditory processing with hemodynamics unrelated to acoustic gradient noise. Magnetic Resonance in Medicine 51(2), 399–402.

60.

Siegel, M., Allendorfer, J.B., Lindsell, C.J., Vannest, J., & Szaflarski, J.P. (2012). The effects of linguistic relationships among paired associates on verbal self-generation and recognition memory. Brain and Behavior 2(6), 789–795.

61.

Smith, D., Baker, G., Davies, G., Dewey, M., & Chadwick, D.W. (1993). Outcomes of add-on treatment with lamotrigine in partial epilepsy. Epilepsia 34(2), 312–322.

62.

Starowicz-Filip, A., Chrobak, A.A., Moskala, M., Krzyzewski, R.M., Kwinta, B., Kwiatkowski, S., ... & Przewoznik, D. (2017). The role of the cerebellum in the regulation of language functions. Psychiatria Polska 51(4), 661–671.

63.

Szaflarski, J.P., & Allendorfer, J.B. (2012). Topiramate and its effect on fMRI of language in patients with right or left temporal lobe epilepsy. Epilepsy and Behavior 24(1), 74–80.

64.

Twamley, E.W., Jak, A.J., Delis, D.C., Bondi, M.W., & Lohr, J.B. (2014). Cognitive Symptom Management and Rehabilitation Therapy (CogSMART) for veterans with traumatic brain injury: Pilot randomized controlled trial. Journal of Rehabilitation Research and Development 51(1), 59–70.

65.

Twamley, E.W., Thomas, K.R., Gregory, A.M., Jak, A.J., Bondi, M.W., Delis, D.C., & Lohr, J.B. (2015). CogSMART Compensatory Cognitive Training for Traumatic Brain Injury: Effects Over 1 Year. Journal of Head Trauma Rehabilitation 30(6), 391–401.

66.

Vannest, J., Eaton, K.P., Henkel, D., Siegel, M., Tsevat, R.K., Allendorfer, J.B., ... & Szaflarski

J.P.

(2012). Cortical correlates of self-generation in verbal paired associate learning. Brain Research 1437, 104–114.

67.

Vannest, J., Maloney, T., Kay, B., Siegel, M., Allendorfer, J.B., Banks, C., ... & Szaflarski, J.P. (2015). Age related-changes in the neural basis of self-generation in verbal paired associate learning. Neuroimage Clinical 7, 537–546.

68.

Wiebe, S., Matijevic, S., Eliasziw, M., & Derry, P.A. (2002). Clinically important change in quality of life in epilepsy. Journal of Neurology, Neurosurgery and Psychiatry 73(2), 116–120.