ORIGINAL STUDY
Impaired Cognitive Flexibility in Amyotrophic Lateral Sclerosis Jessica Evans, BA,*w Christopher Olm, MSc,w Leo McCluskey, MD,w Lauren Elman, MD,w Ashley Boller, BA,w Eileen Moran, MSc,w Katya Rascovsky, PhD,w Teagan Bisbing, MSc,w Corey T. McMillan, PhD,w and Murray Grossman, MDw
Background and Objective: Up to half of patients with amyotrophic lateral sclerosis (ALS) may have cognitive difficulty, but most cognitive measures are confounded by a motor component. Studies relating impaired cognition in ALS to disease in gray matter and white matter are rare. Our objective was to assess executive function in patients with ALS using a simple, untimed measure with minimal motor demands, and to relate performance to structural disease. Methods: We gave the Visual-Verbal Test to 56 patients with ALS and 29 matched healthy controls. This brief, untimed measure of cognitive flexibility first assesses participants’ ability to identify a feature shared by 3 of 4 simple geometric designs. The participants’ cognitive flexibility is challenged when they are next asked to identify a different feature shared by another combination of 3 of the same 4 geometric designs. In a subset of 17 patients who underwent magnetic resonance imaging, regression analyses related test performance to gray matter atrophy and reduced white matter fractional anisotropy. Results: The patients with ALS showed significant impairment in cognitive flexibility (P < 0.01), with 48.2% making an error on the test. Regression analyses related impaired cognitive flexibility to gray matter atrophy in inferior frontal and insular regions, and to reduced fractional anisotropy in white matter projections in the inferior fronto-occipital and uncinate fasciculi and corpus callosum. Conclusions: Our patients with ALS had impaired cognitive flexibility on an untimed measure with minimal motor demands,
Received for publication January 25, 2014; accepted April 16, 2014. From the *Department of Psychology, Villanova University, Villanova, Pennsylvania; and wDepartment of Neurology and Penn Frontotemporal Degeneration Center, University of Pennsylvania, Philadelphia, Pennsylvania. Supported in part by the National Institutes of Health (AG032953, AG017586, AG038490, NS044266, NS053488, AG043503), the ALS Association, and the Wyncote Foundation. The authors declare no conflicts of interest. Reprints: Murray Grossman, MD, 2 Gibson, Hospital of the University of Pennsylvania, 3400 Spruce St, Philadelphia, Pennsylvania 191044283 (e-mail: [email protected]). Supplemental Digital Content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Website, www.cogbehavneurol.com. Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.
Cogn Behav Neurol
Volume 28, Number 1, March 2015
a finding related in part to a large-scale frontal network that is degraded in ALS. Key Words: amyotrophic lateral sclerosis, cognitive flexibility, prefrontal, white matter, anisotropy (Cogn Behav Neurol 2015;28:17–26)
ALS = amyotrophic lateral sclerosis. DWI = diffusion-weighted imaging. FA = fractional anisotropy. FTD = frontotemporal degeneration. MMSE = Mini-Mental State Examination. MRI = magnetic resonance imaging. SPM8 = Statistical Parametric Mapping Version 8. VVT = Visual-Verbal Test.
A
myotrophic lateral sclerosis (ALS) is a motor neuron disease that causes progressive degeneration of upper and lower motor neurons (Keirnan et al, 2011; Mitchell and Borasio, 2007). Pathologic studies have shown that ALS is a multisystem disorder associated overwhelmingly with transactive response DNA-binding protein of approximately 43 kD (TDP-43) pathology (Brettschneider et al, 2013; Geser et al, 2009). Histopathologic and genetic research has solidified the link between ALS and frontotemporal degeneration (FTD) by showing that TDP-43 pathology is also found in many patients with FTD, and repeat expansions in chromosome 9 open reading frame 72 (C9orf72) are found in families with both of these conditions (DeJesus-Hernandez et al, 2011; Irwin et al, 2013; Neumann et al, 2006; Renton et al, 2011). Further, a spectrum of cognitive changes paralleling the deficits in FTD has been found in up to half of patients with ALS (Murphy et al, 2007; Phukan et al, 2012). Several studies have shown that patients with ALS have limited executive control (Goldstein and Abrahams, 2013). Our present study focuses on these patients’ cognitive flexibility. This facet of executive function involves perceiving the sensorimotor and cognitive attributes of a pattern, exerting inhibitory control over this established pattern, and mentally switching to identify a new pattern. Perhaps the most common sign of a deficit in cognitive flexibility is difficulty with letter-guided naming fluency (Abrahams et al, 2000). However, executive deficits can be difficult to assess in patients with ALS because their www.cogbehavneurol.com |
17
Evans et al
motor disorder may compromise their ability to meet a test’s requirements for timed performance or implementation in a specific response modality, eg, speaking or clicking a computer key. The recognition of these confounds led to important innovations to help control for them. For example, Abrahams et al (2000) corrected the oral verbal fluency task for slowed motor articulation by giving patients the traditional timed oral fluency task and following it with a timed control condition in which they read aloud the word lists that they had generated during the oral fluency task. More recently, an untimed measure of cognitive flexibility in ALS was reported that minimizes dependence on timing and response modality: The DelisKaplan Executive Function System Sorting Test (Delis et al, 2001) uses recognition performance to document difficulty on a categorization task (Libon et al, 2012; Taylor et al, 2013). Unfortunately, this measure is complex and time-consuming to administer. A concise, straightforward task was still needed to assess cognitive flexibility in a clinical setting. In the present study, we examined cognitive flexibility in ALS using the Visual-Verbal Test (VVT), developed by Feldman and Drasgow in 1959. This task requires the patient first to identify an attribute (color, shape, size, orientation, “fill” lines) that is shared by 3 of 4 simple geometric designs, and then to identify a different feature shared by a different set of 3 of the same 4 designs (Supplemental Digital Content 1, http://links.lww.com/ CBN/A58). The VVT is well suited to measuring cognitive flexibility in ALS because it is brief, non-social, and untimed; the response modality is immaterial, minimizing any confound by a patient’s motor limitations. Prefrontal cortex is affected early in ALS (Brettschneider et al, 2013). Impaired cognitive flexibility on measures of patients’ executive control is thought to be caused in part by prefrontal disease (Abrahams et al, 1996, 2004; Kew et al, 1993; Libon et al, 2012), although scores may be confounded by the difficulty of most measures and the motor response requirements. Functional magnetic resonance imaging (MRI) studies in healthy adults have linked inferior frontal cortex (Goel and Vartanian, 2005) and insula (Chang et al, 2013) with cognitive flexibility and set switching. Disease in white matter tracts has also been implicated in impaired executive performance (Abrahams et al, 2005), although again the cognitive measures required motor skills. Impaired category naming fluency has been related to reduced fractional anisotropy (FA) in the cingulum (Sarro et al, 2011), presumably linking simple cognitive switching, supported by ventral frontal regions, with more dorsal regions that modulate the decision-making process on the task. In 2013, Pettit et al used a largely motor-free measure of dual tasking to relate executive function in patients with ALS to reduced FA in several white matter regions in the frontal lobe. In this study, we minimized motor confounds in evaluating the neuroanatomic basis for an executive
18 | www.cogbehavneurol.com
Cogn Behav Neurol
Volume 28, Number 1, March 2015
function deficit in ALS by relating VVT performance to MRI images of gray matter atrophy and reduced FA in white matter. We hypothesized that patients with ALS would be impaired on the VVT, and the impairment would be related to disease in a large-scale network involving inferior prefrontal and insular gray matter regions integrated by frontal white matter projections to form a prefrontal decision-making network.
METHODS Participants We recruited 56 patients with ALS from the outpatient Neurology Clinic at the University of Pennsylvania (Table 1). As described below, we determined that 47 of the patients had ALS alone and the other 9 had ALS combined with FTD. All of the patients with ALS alone and 6 of the 9 with ALS-FTD met the revised El Escorial criteria for ALS (Brooks et al, 2000); records were unavailable for the 3 other patients with ALS-FTD. ALS onset had been bulbar in 11 of the 56 patients, cervical in 20, and lumbosacral in 21; we could not determine the form of motor onset in 4 of the patients with ALS-FTD. We categorized our patients as having either ALS alone or ALS-FTD based on a physician’s (author L.M., L.E., or M.G.) evaluation using published diagnostic criteria (Rascovsky et al, 2011; Strong et al, 2009). All the patients in the ALS-FTD group had been diagnosed with concurrent behavioral variant FTD and motor neuron disease, and in all the FTD phenotype had appeared first. The ALS-only group exhibited only motor neuron disease; clinical evaluation showed that they had mild if any cognitive impairment. We tested the patients an average of about 6 months after they had received their diagnosis. We evaluated them with the Amyotrophic Lateral Sclerosis Functional Rating Scale–Revised (Cedarbaum et al, 1999). We excluded patients who had other potential causes for their cognitive difficulty, such as metabolic or psychiatric disorders. No patients were taking centrally acting medications that could compromise their cognitive function. We recruited 29 elderly controls from the community. The controls were matched with the ALS group for sex and education. The patients were significantly younger than these controls, although age did not correlate with VVT performance: r53 = 0.17; P > 0.3. We used a separate group of 36 healthy seniors as controls solely for the MRI portion of the study. These controls, described below under the Imaging Acquisition and Analysis section, did not take part in the cognitive portion of the study and we refer to them only in the context of the imaging. We gave the patients and cognitive controls the Mini-Mental State Examination (MMSE) (Folstein et al, 1975) to gauge their general cognitive function. The patients had significantly lower MMSE scores than the controls, even though we scored the test proportionately, giving allowances for items that the patients could not perform because of a motor deficit. Nevertheless, the Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
Cogn Behav Neurol
Volume 28, Number 1, March 2015
Impaired Cognitive Flexibility in ALS
TABLE 1. Demographic and Cognitive Performance Data for Patients with Amyotrophic Lateral Sclerosis (ALS), Patients with ALS and Frontotemporal Degeneration (FTD), and Controls ALS ALS-FTD ALS Total Controls (n = 47) (n = 9) (n = 56) (n = 29) Men/women Age (years) Education (years) Disease duration (months) ALS Functional Rating Scale1 Mini-Mental State Exam2 Visual-Verbal Test3 Initial categorization failures (number of errors) Cognitive flexibility (maximum = 10) Letter fluency4 Words/minute total Digit span5 Digits forward Digits backward
29/18 58.19 (13.04) 15.09 (2.80) 6.85 (8.66) 34.13 (8.48) (n = 47) 27.85 (3.23) (n = 47)
6/3 62.44 (9.81) 14.11 (3.41) 5.40 (6.62) 38.57 (3.05) (n = 5) 24.89 (6.33)* (n = 9)
35/21 58.88 (12.59)* 14.93 (2.89) 6.71 (8.44) 34.73 (8.10) (n = 52) 27.38 (4.00)* (n = 56)
13/16 67.76 (7.67) 16.07 (2.67)
0.04 (0.20)
0.22 (0.67)
0.07 (0.32)
0.00 (0.00)
8.55 (2.40)* (n = 47)
6.67 (4.03)* (n = 9)
8.25 (2.77)* (n = 56)
9.41 (1.15) (n = 29)
11.63 (5.31)* (n = 42)
6.81 (1.26)*w (n = 5)
10.91 (5.21)* (n = 47)
14.31 (5.20) (n = 29)
6.72 (1.24) 4.24 (1.763) (n = 40)
6.80 (0.84) 4.20 (1.30) (n = 5)
6.69 (1.10) 4.33 (1.49) (n = 45)
6.95 (1.17) 5.09 (1.74) (n = 29)
29.25 (1.11) (n = 29)
Data are shown as either numbers or mean (standard deviation). *ALS differs from controls, P < 0.05. wALS differs from ALS-FTD, P < 0.05. 1 Cedarbaum et al, 1999. 2Folstein et al, 1975. 3Feldman and Drasgow, 1959. 4Spreen and Strauss, 1991. 5 Wechsler, 1995.
patients’ average MMSE scores were within our adjusted normal range. The study protocol was approved by the University of Pennsylvania Institutional Review Board. All participants and caregivers gave written informed consent before taking part.
Cognitive Tests In addition to the MMSE, we tested cognitive function in the 56 patients and 29 controls with a brief battery that included letter-guided category naming fluency (F-A-S or just F) (Spreen and Strauss, 1991), to measure cognitive flexibility; digit span forward and backward (Wechsler, 1995), to measure attention and working memory; and the VVT, to measure cognitive flexibility (Table 1). During the study we gave the letter fluency task to 47 of the patients and the 29 controls. Early on, we asked 30 of the patients to say as many words starting with the letter F as they could within 1 minute (Spreen and Strauss, 1991). Then we asked them to repeat the task for the letter A and then S. Partway through the study, we changed the test procedure slightly. When we tested 17 more of the patients, we asked them to say only words starting with the letter F. To maximize the available data for analysis, we combined the numbers of F words from both sets of patients. Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
We gave the digit span task to 45 of the patients and the 29 controls. In the digit span forward, participants are asked to repeat a series of spoken number sequences, each of which is longer than the 1 before it. For the digit span backward, participants must repeat the ever-lengthening number sequences in the reverse order from which they are presented. Visual-Verbal Test We gave all participants the VVT. In this test, participants are shown a card that displays 4 simple geometric designs. These designs can be grouped into 2 sets of 3, based on color, size, shape, orientation, or pattern of lines filling the design. Participants are asked first to identify a set of 3 designs that share a feature. They may respond by speaking and/or pointing. On the same card, they are then asked to identify a second set of 3 designs that share a different feature. Supplemental Digital Content 1 (http://links. lww.com/CBN/A58) shows 2 sample cards from the VVT. The upper set of geometric designs can be grouped by color (ABC) and size (ABD). The lower set can be grouped by shape (ABC) and color (ACD). In this study, we gave participants 2 practice trials to ensure that they understood the task. The actual study consisted of 10 untimed trials. www.cogbehavneurol.com |
19
Evans et al
Grouping the first set of 3 geometric designs shows that participants can perceive the designs and group a subset of items based on a shared feature. For this part of the task, we score the number of times that a participant could not identify a shared feature across the designs. Thus, a score of zero indicates perfect performance. Disengaging from the first set of designs and identifying the set based on the second shared feature requires inhibitory control and set switching, which are components of cognitive flexibility. For this part of the task, we count the number of times that participants could switch from the first set to the second set. A score of 10 on this part of the task indicates that participants successfully identified all of the second sets and had intact cognitive flexibility. Although the designs may have differences in the saliency of particular features, the task does not become more difficult over the course of the 10 trials. We performed analysis of variance, t test, and Pearson correlation using Statistical Package for the Social Sciences Version 21 (IBM SPSS Statistics, International Business Machines, Armonk, New York).
Imaging Acquisition and Analysis A subset of 17 of the patients (11 with ALS alone and 6 with ALS-FTD) had high-resolution T1-weighted structural MRI scans to assess their gray matter atrophy, and diffusion-weighted imaging (DWI) to analyze their reduced white matter FA. We scanned only 17 patients because the others had trouble breathing when they lay prone at length in the magnet bore. The patients who had imaging matched the full cohort of 56 patients in sex, age, education, disease duration, and VVT performance (all P values >0.1). A Siemens 3.0 T Trio scanner (Siemens Medical Solutions, Erlangen, Germany) with an 8-channel head coil collected the images. We obtained T1-weighted 3dimensional spoiled gradient-echo images with repetition time = 1620 msec, echo time = 3 msec, flip angle = 15 degrees, matrix = 192 256, slice thickness = 1 mm, and in-plane resolution = 1.0 1.0 mm. For 15 of the 17 scanned patients, we performed a 30-directional DWI sequence using a single-shot, spinecho, diffusion-weighted echo planar imaging sequence (repetition time = 8100 msec, echo time = 83 msec, fat saturation, matrix size = 128 128, number of slices = 70, voxel size = 2 mm isotropic). We collected 30 volumes with diffusion weighting (b-value = 1000 seconds/ mm2) along 30 non-collinear directions, and we collected 1 volume per participant without diffusion weighting (b-value = 0 seconds/mm2). We collected additional b-value = 0 volumes for 8 participants. For the other 2 of the 17 scanned patients, we collected DWI images using a 12-directional sequence with a single-shot, spin-echo, diffusion-weighted echo planar imaging sequence (matrix size = 128 128, number of slices = 40, voxel size = 3 mm isotropic, repetition
20 | www.cogbehavneurol.com
Cogn Behav Neurol
Volume 28, Number 1, March 2015
time = 6500 msec, echo time = 99 msec), acquiring 12 non-collinear, non-coplanar, isotropic diffusion gradients. We also performed MRI scans on a different healthy control group of 36 seniors recruited from the community who were similar in age and other demographic characteristics to our 17 patients (all P values >0.5). Of these controls, 28 underwent 30-directional scans and 8 had 12-directional scans. As we described in an earlier study (Grossman et al, 2013), we normalized the images to a standard space and segmented them using the PipeDream interface (SourceForge, Slashdot Media, San Francisco, California) (http:// sourceforge.net/projects/neuropipedream/) to the Advanced Normalization Tools toolkit (Penn Image Computing & Science Lab, University of Pennsylvania, Philadelphia, Pennsylvania) (http://www.picsl.upenn.edu/ ANTS/) (Avants et al, 2008). A local 1-mm3 resolution T1 template consisted of 25 healthy seniors and 25 patients with FTD. We registered our current participants’ images to the local template, transformed into Montreal Neurological Institute space, and down-sampled to 2-mm3 resolution. We smoothed the images in Statistical Parametric Mapping Version 8 (SPM8) (Wellcome Trust Centre for Neuroimaging, London, United Kingdom; available online at http://www.fil.ion.ucl.ac.uk/spm/software/spm8) using a 5-mm full-width half-maximum isotropic Gaussian kernel to minimize individual gyral variations. To compare our patients’ gray matter atrophy to the controls’, we used a whole-brain voxel-wise analysis in SPM8 with height threshold = q < 0.05 (false discovery rate-corrected), minimum cluster size = 50 voxels, and peak voxel threshold = P < 0.001. We related VVT cognitive flexibility performance to gray matter atrophy using the multiple regressions module in SPM8, constrained to regions of gray matter atrophy. Clusters survived a height threshold = P < 0.005 (uncorrected), minimum cluster size = 40 voxels, and peak voxel threshold = P < 0.001. To preprocess the DWIs, we used PipeDream and Advanced Normalization Tools as described above. We removed motion and distortion artifacts using affine coregistration of each DWI to the unweighted (b-value = 0) image(s). We used a linear least-squares algorithm (Salvador et al, 2005) implemented in the Camino software toolkit (University College London Microstructure Imaging Group, London, United Kingdom) (http:// cmic.cs.ucl.ac.uk/camino/) (Cook et al, 2006) to compute diffusion tensors. We reoriented the tensors using the preservation of principal directions method (Alexander et al, 2001). We used the diffusion tensor image to compute FA for each participant, and registered each FA image to the T1 image. We used the symmetric diffeomorphic procedure in Advanced Normalization Tools to warp T1 images to the local template, and we warped each FA image to Montreal Neurological Institute space by applying the T1-to-template and template-to-Montreal Neurological Institute warps. We smoothed the FA images using a 4-mm full-width half-maximum isotropic Gaussian kernel. Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
Cogn Behav Neurol
Volume 28, Number 1, March 2015
The 2-sample t test module in SPM8 compared FA differences between the patients and controls, constrained to white matter by averaging all patient and control FA images and generating a mask consisting of voxels with FAZ0.25. We compared patients to controls at a false discovery rate-corrected height threshold = q < 0.01 with 200-voxel extent. Using deterministic tractography in Camino, we tracked white matter fibers in the healthy elderly template generated using the 30-directional diffusion tensor imaging sequence described above. We retained fiber tracts passing through voxels of reduced FA, as determined by reduced FA in our patients with ALS compared with our controls, to define the mask for regression analyses. We used the multiple regression module of SPM8 to relate VVT cognitive flexibility to white matter tracts with significantly reduced FA in our patients with ALS, using height threshold = P < 0.005, 50-voxel extent, and peak voxel threshold = P < 0.001.
RESULTS Behavioral Analyses We found impaired cognitive flexibility in our patients with ALS. The 1-way analysis of variance showed
Impaired Cognitive Flexibility in ALS
that the patients and controls did not differ in their ability to identify an initial set of shared-feature designs: F1,83 = 1.416, P = 0.237. However, the patients were significantly less able to identify a second set of sharedfeature designs, consistent with a deficit in cognitive flexibility: F1,83 = 4.674, P = 0.034. Each of the patient subgroups differed from the controls: ALS alone, t74 = 6.928, P = 0.04; ALS-FTD, t36 = 3.342, P = 0.002. We found at least 1 cognitive flexibility error in 27 (48.2%) of the 56 patients, including 9 (69.2%) of 13 patients with autopsy-confirmed ALS; 9 (31%) of the 29 controls had a similar error rate. Cognitive flexibility measured by the VVT correlated with performance on the MMSE: r54 = 0.660, P < 0.001; letter-guided category naming fluency: r45 = 0.350, P = 0.016; and digit span backward: r43 = 0.573, P < 0.001. We found no difference between the subgroups whose ALS onset had been bulbar, cervical, or lumbosacral (all P values > 0.3).
Image Analyses Figure 1 illustrates the MRI analyses in our 17 scanned patients. Figure 1A shows significant gray matter atrophy, primarily in a bilateral frontal distribution, extending into the anterior temporal and parietal
FIGURE 1. Gray matter atrophy and reduced white matter fractional anisotropy, with regression analyses, in 17 patients with amyotrophic lateral sclerosis. Panel A: The patients had significant gray matter atrophy (shown in red) relative to controls. Vertical white lines mark the locations of the slices shown in Panel B. Panel B: Coronal slices illustrate gray matter regressions relating the patients’ gray matter atrophy to impaired cognitive flexibility performance (purple) and reduced white matter fractional anisotropy (green); the regressions shown in blue relate reduced fractional anisotropy to impaired cognitive flexibility performance. The numbers indicate the coordinates of the slices in the Y axis. R = right. L = left. Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
www.cogbehavneurol.com |
21
Cogn Behav Neurol
Evans et al
lobes. Table 2 summarizes the peak foci of the gray matter atrophy, corresponding to areas of atrophy illustrated in Figure 1. Regression analysis, illustrated in Figure 1B and summarized in Table 2, shows that the patients’ cognitive flexibility deficits were related to areas of significant atrophy in bilateral inferior frontal and left insular regions. Figure 1B also illustrates areas of significantly reduced FA in several white matter tracts. Table 3 summarizes areas of reduced FA in the white matter, corresponding to areas visualized in Figure 1B. These areas are the corpus callosum, descending fibers from motor cortex, and projections involving inferior frontal regions: the uncinate fasciculus, inferior fronto-occipital fasciculus, inferior longitudinal fasciculus, and cingulum. Regression analysis, also summarized in Table 3 and illustrated in Figure 1B, shows that difficulty with cognitive
Volume 28, Number 1, March 2015
flexibility was related to reduced FA in frontal white matter, specifically the inferior fronto-occipital fasciculus and uncinate fasciculus, and in the corpus callosum.
DISCUSSION Our findings suggest that the VVT is sensitive to impairments in cognitive flexibility in patients with ALS. The VVT is a simple, brief, untimed measure to which participants can respond by either speaking or pointing. Therefore, this measure is well suited for clinical use in evaluating cognition in patients with ALS. Previous studies suggested that as many as 50% of patients with ALS have cognitive difficulty (Murphy et al, 2007; Phukan et al, 2012). We confirm these estimates with our current finding that 48% of patients with ALS had limited cognitive flexibility on the VVT. Our regression analyses related impaired performance on this
TABLE 2. Anatomic Locations of Significant Gray Matter Atrophy in Patients with Amyotrophic Lateral Sclerosis (ALS) Relative to Controls, and Regressions Relating Impaired Cognitive Flexibility to Atrophy Anatomic Locus (Brodmann Area)
MNI Coordinates x y z
z Score
Cluster Size (Voxels)
Gray matter atrophy R primary motor (4) R premotor (6) R inferior frontal (44) R middle frontal (10) R middle frontal (46) R cingulate (32) R medial frontal (9) R medial frontal (11) R orbital frontal (11) R insula L inferior frontal (47) L middle frontal (10) L middle frontal (9) L cingulate (24) L orbital frontal (11) L insula R superior temporal (22) R superior temporal (38) R middle temporal (21) R middle temporal (21) R inferior temporal (20) R fusiform (20) R lingual (19) L middle temporal (21) L middle temporal (21) L fusiform (20) L hippocampal (28) R postcentral (2)
44 34 42 28 38 12 20 12 24 40 40 34 36 12 28 42 50 48 50 52 52 40 22 54 50 44 20 62
8 2 8 50 34 34 28 34 32 0 24 46 16 38 36 2 42 14 2 14 6 18 44 12 2 18 10 18
36 50 30 6 22 24 36 20 14 4 2 6 30 12 14 6 6 20 18 12 32 28 10 14 32 30 16 28
3.86 4.25 5.74 4.56 3.97 5.2 4.74 4.4 4.38 5.11 3.71 3.89 3.49 4.67 3.68 3.63 3.49 3.34 4.08 3.86 4.52 4.77 4.35 4.14 3.63 4.35 3.74 3.68
109 73 266 227 141 247 165 147 145 710 135 73 83 546 106 90 50 73 65 115 201 127 76 299 73 94 62 68
Regressions R inferior frontal (47) L orbital frontal (11) L insula
32 32 36
28 34 20
0 12 4
3.47 3.26 3.64
53 57 42
MNI = Montreal Neurological Institute. R = right. L = left.
22 | www.cogbehavneurol.com
Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
Cogn Behav Neurol
Volume 28, Number 1, March 2015
Impaired Cognitive Flexibility in ALS
TABLE 3. Reductions in White Matter Fractional Anisotropy in Patients with Amyotrophic Lateral Sclerosis (ALS) Relative to Controls, and Regressions Relating Impaired Cognitive Flexibility to Reduced Fractional Anisotropy MNI Coordinates Anatomic Locus Cluster Size (White Matter Tract) x y z z Score (Voxels) Reduced fractional anisotropy R inferior frontal white matter R orbital frontal white matter R inferior fronto-occipital fasciculus R middle fronto-occipital fasciculus B corpus callosum (frontal and body) B fornix L inferior frontal gyrus white matter L inferior fronto-occipital fasciculus L inferior longitudinal fasciculus L inferior longitudinal fasciculus L cingulum L cerebral peduncle
15 10 26 31 0 2 35 44 41 38 21 18
8 23 28 11 16 6 12 35 21 5 26 13
13 12 12 8 29 17 23 2 15 22 17 16
5.23 5.2 4.34 3.84 5.55 5.59 4.28 4.52 4.97 4.28 4.69 6.02
951 418 248 410 11958 1114 619 630 2000 291 283 1446
Regressions R uncinate fasciculus R corpus callosum (frontal) R corpus callosum (frontal) R corpus callosum (frontal) L uncinate fasciculus L inferior frontal gyrus white matter L inferior fronto-occipital fasciculus L corpus callosum (frontal) L corpus callosum (frontal) L corpus callosum (frontal)
23 17 11 19 17 37 38 15 14 13
21 55 30 7 26 5 17 52 44 19
11 6 7 50 10 23 5 2 23 44
3.72 5.02 4.18 3.2 3.27 3.65 3.92 4.11 3.51 3.09
99 1273 5078 69 256 115 455 745 268 82
MNI = Montreal Neurological Institute. R = right. B = bilateral. L = left.
task to gray matter regions in prefrontal cortex (areas of known pathological burden in ALS) and to interconnecting frontal white matter projections of the uncinate and inferior fronto-occipital fasciculi and corpus callosum. To some extent, informal observation and a clinical interview can reveal impaired cognition in patients’ dayto-day performance. However, quantification of cognitive difficulty is essential to determining progression and prognosis, and may be helpful in evaluating response during a treatment trial. Quantifying impairment has proven difficult with most standard cognitive measures, such as the MMSE and Addenbrooke Cognitive Examination–Revised (Leyton et al, 2010), because many components of these tests require a motor response. Indeed, in an unpublished survey of our clinical cohort of 123 patients with ALS, we found that only 59% were able to complete all components of the MMSE, while 100% successfully completed the VVT. Measures of executive function such as letter-guided category naming fluency are sensitive to cognitive deficits in ALS (Goldstein and Abrahams, 2013). However, some patients with a bulbar motor disorder have slowed speech (Ash et al, 2014), which might confound their performance on a timed cognitive task requiring speech Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
production. Abrahams et al (2004) introduced an important innovation to letter-guided category naming fluency: a control task (timed reading of the words produced by the patient) that compensates in part for the motor deficit, thereby defining more clearly an impairment on the task’s cognitive component. Still, even this control does not fully eliminate the confounding motor deficit. Abrahams et al (2014) published norms for a written version of letter-guided category naming fluency, although this only shifts the motor confound to another response modality. The VVT also minimizes confounds related to other aspects of executive function. For example, identifying a deficit in cognitive flexibility is less confounded with the VVT than with the Wisconsin Card Sorting Task, a lengthy measure of cognitive flexibility that also requires significant working memory (Anderson et al, 1991), because patients taking the VVT are explicitly told that they need to make a mental shift. Patients with ALS have difficulty with lengthy and demanding measures of cognitive flexibility. Libon et al (2012) and Taylor et al (2013) asked patients with ALS to group 16 cards into sets according to a variety of shared features (Delis et al, 2001). Although this task was not timed, the patients showed significant impairment. One www.cogbehavneurol.com |
23
Evans et al
problem was that they had motor difficulty manipulating the cards into groups. To minimize the motor component of the task, the researchers gave the patients a recognition procedure, asking them to recognize the rule for grouping the cards. Still, their performance remained significantly impaired. We think that this task is too lengthy to be suited for administration in a busy clinic. In the present study, we assessed our patients’ cognitive flexibility with the VVT, which is a brief, untimed measure that does not depend on a specific response modality. Our patients had little difficulty perceiving their first pattern of shared features across geometric designs. This finding leads us to believe that visual perception and perceptual categorization of geometric features did not play a major role in the patients’ cognitive deficit. However, they were significantly impaired on the subsequent measure of cognitive flexibility, because they had difficulty identifying a second set of 3 stimuli that shared a different feature. Eslinger et al (2007) reported impaired VVT performance in patients with FTD. We found significant impairment even in our subgroup of patients with ALS who did not have concurrent FTD. Our patients’ VVT performance correlated with their performance on letterguided category naming fluency and digit span backward tasks, which are traditional measures of executive function and working memory. Widespread gray matter and white matter disease is now recognized in ALS (Agosta et al, 2007; Ellis et al, 2001; Verstraete et al, 2012). The affected areas extend beyond primary motor cortex to include inferior and dorsolateral prefrontal cortical regions that are important for executive function, as well as anterior temporal and parietal gray matter areas. This spread corresponds to the anatomic distribution of pathology in ALS (Geser et al, 2008). Likewise, patients can have substantial disease in the associated white matter corticospinal fibers and white matter tracts that have been implicated in the prefrontal projections that subserve large-scale cognitive networks thought to play a role in executive function (Mioshi et al, 2013). This distribution, too, corresponds to the white matter pathology in ALS (Douaud et al, 2011). In the present study, we found disease in gray matter regions and white matter tracts similar to those reported in previous studies. Moreover, we were able to identify a compromised network of gray matter regions and white matter tracts associated with impaired cognitive flexibility in ALS. Regression analysis implicated inferior frontal cortex in our patients’ impaired cognitive flexibility. Previous functional MRI work in healthy adults has demonstrated that inferior frontal cortex contributes to inhibitory control and set switching (Goel and Vartanian, 2005). Another functional MRI study in healthy adults showed a role for the insula in cognitive flexibility and set switching (Chang et al, 2013). Our regression analysis also implicated the insula in cognitive flexibility difficulty in ALS. Previous work has also associated dorsolateral prefrontal areas with impaired cognitive flexibility in patients with ALS (Abrahams et al, 1996, 2004). There are
24 | www.cogbehavneurol.com
Cogn Behav Neurol
Volume 28, Number 1, March 2015
several reasons why we may not have seen this association in the present study. One is that the prior work may have combined complex cognitive challenges with the demands of a motor component in patients who had a motor deficit. However, our own previous study (Libon et al, 2012) linked category recognition impairment to dorsolateral prefrontal gray matter atrophy on a task that made minimal motor demands. This finding raises the possibility that the relative cognitive demands of the task itself influence the anatomic distribution of prefrontal cortex required for cognitive flexibility: More demanding switching tasks might be associated with dorsolateral prefrontal cortex, and less demanding tasks with ventral frontal regions. Functional MRI studies in healthy adults have shown different patterns of prefrontal activation depending on the cognitive demands of the task (Badre and D’Esposito, 2007; Koechlin and Jubault, 2006). More work is needed to examine the role of varying cognitive task demands explicitly in the gray matter as a basis for impaired cognitive flexibility in ALS. White matter disease has been seen consistently enough in patients with ALS to warrant the proposal that imaging changes in white matter may serve as a biomarker of ALS (Foerster et al, 2013). We, too, observed significant white matter disease in the corpus callosum and white matter tracts in the frontal lobe. Rare studies have been performed that relate white matter disease directly to motor deficits in ALS (Verstraete et al, 2011). Previous studies related impaired letter-guided category naming fluency to reduced white matter volume in ALS (Abrahams et al, 2005) and to reduced FA in the cingulum (Sarro et al, 2011). More recent work has related FA in several white matter regions of interest to performance on a largely motor-free dual task, but not to measures of information processing speed (Pettit et al, 2013). In the present study, our regressions related white matter projections in the frontal lobe to impaired cognitive flexibility in ALS. Among the implicated tracts was the inferior fronto-occipital fasciculus, which links prefrontal with posterior temporal regions. Temporal regions presumably contributed to our participants’ perception of the visual stimuli in the VVT, while inferior frontal regions maintained control over processing of the perceived sets. We also found that the uncinate fasciculus plays a role in cognitive flexibility. This fasciculus links inferior frontal and anterior temporal regions that may be related to inhibitory control. Finally, our regression analysis showed a role for the anterior corpus callosum in cognitive flexibility. This region may be important in coordinating performance across the left and right frontal regions. Our work thus describes a potential connectome that might mediate cognitive flexibility deficits in ALS. Several caveats should be kept in mind when considering our findings. While we collected behavioral data on a large number of patients with ALS, we were able to obtain MRI data for only 17 (30%). Second, the materials and procedures that we used in this study did not permit us to determine the precise role of each frontal Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
Cogn Behav Neurol
Volume 28, Number 1, March 2015
region and white matter projection in cognitive flexibility. Finally, in Table 1 we report an unvalidated MMSE score that adjusts for items that some patients could not complete because of motor limitations; we include this measure to provide a rough clinical guide to the patients’ overall cognitive function. These caveats notwithstanding, our findings are consistent with the hypothesis that many patients with ALS exhibit a deficit in cognitive flexibility even when performing a simple, untimed task that minimizes motor confounds, and this deficit appears to be associated with disease in a large-scale neural network involving both gray matter and white matter regions in non-motor portions of the frontal lobe. REFERENCES Abrahams S, Goldstein LH, Kew JJM, et al. 1996. Frontal lobe dysfunction in amyotrophic lateral sclerosis: a PET study. Brain. 119:2105–2120. Abrahams S, Goldstein LH, Simmons A, et al. 2004. Word retrieval in amyotrophic lateral sclerosis: a functional magnetic resonance imaging study. Brain. 127:1507–1517. Abrahams S, Goldstein LH, Suckling J, et al. 2005. Frontotemporal white matter changes in amyotrophic lateral sclerosis. J Neurol. 252:321–331. Abrahams S, Leigh PN, Harvey A, et al. 2000. Verbal fluency and executive dysfunction in amyotrophic lateral sclerosis (ALS). Neuropsychologia. 38:734–747. Abrahams S, Newton J, Niven E, et al. 2014. Screening for cognition and behaviour changes in ALS. Amyotroph Lateral Scler Frontotemporal Degener. 15:9–14. Agosta F, Pagani E, Rocca MA, et al. 2007. Voxel-based morphometry study of brain volumetry and diffusivity in amyotrophic lateral sclerosis patients with mild disability. Hum Brain Mapp. 28:1430–1438. Alexander DC, Pierpaoli C, Basser PJ, et al. 2001. Spatial transformations of diffusion tensor magnetic resonance images. IEEE Trans Med Imaging. 20:1131–1139. Anderson S, Damasio H, Jones R, et al. 1991. Wisconsin Card Sorting Test performance as a measure of frontal lobe damage. J Clin Exp Neuropsychol. 13:909–922. Ash S, Olm C, McMillan CT, et al. 2014. Deficits in sentence expression in amyotrophic lateral sclerosis. Amyotroph Lateral Scler Frontotemporal Degener. 8:1–9. Avants BB, Epstein CL, Grossman M, et al. 2008. Symmetric diffeomorphic image registration with cross-correlation: evaluating automated labeling of elderly and neurodegenerative brain. Med Image Anal. 12:26–41. Badre D, D’Esposito M. 2007. Functional magnetic resonance imaging evidence for a hierarchical organization of the prefrontal cortex. J Cogn Neurosci. 19:2082–2099. Brettschneider J, Del Tredici K, Toledo JB, et al. 2013. Stages of pTDP43 pathology in amyotrophic lateral sclerosis. Ann Neurol. 74:20–38. Brooks BR, Miller RG, Swash M, et al. 2000. El Escorial revisited: revised criteria for the diagnosis of amyotrophic lateral sclerosis. Amyotroph Lateral Scler Other Motor Neuron Disord. 1:293–299. Cedarbaum JM, Stambler N, Malta E, et al. 1999. The ALSFRS-R: a revised ALS functional rating scale that incorporates assessments of respiratory function: BDNF ALS Study Group (Phase III). J Neurol Sci. 169: 13–21. Chang LJ, Yarkoni T, Khaw MW, et al. 2013. Decoding the role of the insula in human cognition: functional parcellation and large-scale reverse inference. Cereb Cortex. 23:739–749. Cook PA, Bai Y, Nadjati-Gilani S, et al. 2006. Camino: open-source diffusion-MRI reconstruction and processing. Proc Intl Soc Mag Reson Med. 14:2759. DeJesus-Hernandez M, Mackenzie IR, Boeve BF, et al. 2011. Expanded GGGGCC hexanucleotide repeat in noncoding region of C9ORF72 causes chromosome 9p-linked FTD and ALS. Neuron. 72: 245–256.
Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
Impaired Cognitive Flexibility in ALS
Delis DC, Kaplan E, Kramer JH. 2001. Delis-Kaplan Executive Function Battery. San Antonio, Texas: The Psychological Corporation. Douaud G, Filippini N, Knight S, et al. 2011. Integration of structural and functional magnetic resonance imaging in amyotrophic lateral sclerosis. Brain. 134:3467–3476. Ellis CM, Suckling J, Amaro E. 2001. Volumetric analysis reveals corticospinal tract degeneration and extramotor involvement in ALS. Neurology. 57:1571–1578. Eslinger PJ, Moore P, Troiani V, et al. 2007. Oops! Resolving social dilemmas in frontotemporal dementia. J Neurol Neurosurg Psychiatry. 78: 457–460. Feldman MJ, Drasgow J. 1959. The Visual-Verbal Test. Los Angeles, California: Western Psychological Services. Foerster BR, Dwamena BA, Petrou M, et al. 2013. Diagnostic accuracy of diffusion tensor imaging in amyotrophic lateral sclerosis: a systematic review and individual patient data meta-analysis. Acad Radiol. 20:1099–1106. Folstein MF, Folstein SE, McHugh PR. 1975. Mini-mental state: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 12:189–198. Geser F, Martinez-Lage M, Robinson J, et al. 2009. Clinical and pathological continuum of multisystem TDP-43 proteinopathies. Arch Neurol. 66:180–189. Geser F, Winton MJ, Kwong LK, et al. 2008. Pathological TDP-43 in parkinsonism-dementia complex and amyotrophic lateral sclerosis of Guam. Acta Neuropathol. 145:115–133. Goel V, Vartanian O. 2005. Dissociating the roles of right ventral lateral and dorsal lateral prefrontal cortex in generation and maintenance of hypotheses in set-shift problems. Cereb Cortex. 15: 1170–1177. Goldstein LH, Abrahams S. 2013. Changes in cognition and behaviour in amyotrophic lateral sclerosis: nature of impairment and implications for assessment. Lancet Neurol. 12:368–380. Grossman M, Peelle JE, Smith EE, et al. 2013. Category-specific semantic memory: converging evidence from bold fMRI and Alzheimer’s disease. Neuroimage. 68:263–274. Irwin DJ, McMillan CT, Brettschneider J, et al. 2013. Cognitive decline and reduced survival in C9orf72 expansion frontotemporal degeneration and amyotrophic lateral sclerosis. J Neurol Neurosurg Psychiatry. 84:163–169. Keirnan MC, Vucic S, Cheah BC, et al. 2011. Amyotrophic lateral sclerosis. Lancet. 377:942–955. Kew JJ, Goldstein LH, Leigh PN, et al. 1993. The relationship between abnormalities of cognitive function and cerebral activation in amyotrophic lateral sclerosis: a neuropsychological and positron emission tomography study. Brain. 116:1399–1423. Koechlin E, Jubault T. 2006. Broca’s area and the hierarchical organization of human behavior. Neuron. 50:963–974. Leyton CE, Hornberger M, Mioshi E, et al. 2010. Application of Addenbrooke’s Cognitive Examination to diagnosis and monitoring of progressive primary aphasia. Dement Geriatr Cogn Disord. 29:504–509. Libon DJ, McMillan C, Avants B, et al. 2012. Deficits in concept formation in amyotrophic lateral sclerosis. Neuropsychology. 26:422–429. Mioshi E, Lillo P, Yew B, et al. 2013. Cortical atrophy in ALS is critically associated with neuropsychiatric and cognitive changes. Neurology. 80:1117–1123. Mitchell JD, Borasio GD. 2007. Amyotrophic lateral sclerosis. Lancet. 369:2031–2041. Murphy JM, Henry RG, Langmore S, et al. 2007. Continuum of frontal lobe impairment in amyotrophic lateral sclerosis. Arch Neurol. 64:530–534. Neumann M, Sampathu DM, Kwong LK, et al. 2006. Ubiquinated TDP-43 in frontotemporal lobar degeneration and amyotrophic lateral sclereosis. Science. 314:130–133. Pettit LD, Bastin ME, Smith C, et al. 2013. Executive deficits, not processing speed relates to abnormalities in distinct prefrontal tracts in amyotrophic lateral sclerosis. Brain. 136:3290–304. Phukan J, Elamin M, Bede P, et al. 2012. The syndrome of cognitive impairment in amyotrophic lateral sclerosis: a populationbased study. J Neurol Neurosurg Psychiatry. 83:102–108.
www.cogbehavneurol.com |
25
Evans et al
Rascovsky K, Hodges JR, Knopman D, et al. 2011. Sensitivity of revised diagnostic criteria for the behavioural variant of frontotemporal dementia. Brain. 134:2456–2477. Renton AE, Majounie E, Waite A, et al. 2011. A hexanucleotide repeat expansion in C9ORF72 is the cause of chromosome 9p21-linked ALS-FTD. Neuron. 72:257–268. Salvador R, Pena A, Menon DK, et al. 2005. Formal characterization and extension of the linearized diffusion tensor model. Hum Brain Mapp. 24:144–155. Sarro L, Agosta F, Canu E, et al. 2011. Cognitive functions and white matter tract damage in amyotrophic lateral sclerosis: a diffusion tensor tractography study. AJNR Am J Neuroradiol. 32:1866–1872. Spreen O, Strauss EA. 1991. A Compendium of Neuropsychological Tests. New York: Oxford University Press.
26 | www.cogbehavneurol.com
Cogn Behav Neurol
Volume 28, Number 1, March 2015
Strong MJ, Grace GM, Freedman M, et al. 2009. Consensus criteria for the diagnosis of frontotemporal cognitive and behavioural syndromes in amyotrophic lateral sclerosis. Amyotroph Lateral Scler. 10:131–146. Taylor LJ, Brown RG, Tsermentseli S, et al. 2013. Is language impairment more common than executive dysfunction in amyotrophic lateral sclerosis? J Neurol Neurosurg Psychiatry. 84:494–498. Verstraete E, Veldink JH, Hendrikse J, et al. 2012. Structural MRI reveals cortical thinning in amyotrophic lateral sclerosis. J Neurol Neurosurg Psychiatry. 83:383–388. Verstraete E, Veldink JH, Mandl RC, et al. 2011. Impaired structural motor connectome in amyotrophic lateral sclerosis. PLoS One. 6:e24239. Wechsler D. 1995. Wechsler Adult Intelligence Scale. San Antonio: Psychological Press.
Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
Impaired Cognitive Flexibility in Amyotrophic Lateral Sclerosis Jessica Evans, BA,*w Christopher Olm, MSc,w Leo McCluskey, MD,w Lauren Elman, MD,w Ashley Boller, BA,w Eileen Moran, MSc,w Katya Rascovsky, PhD,w Teagan Bisbing, MSc,w Corey T. McMillan, PhD,w and Murray Grossman, MDw
Background and Objective: Up to half of patients with amyotrophic lateral sclerosis (ALS) may have cognitive difficulty, but most cognitive measures are confounded by a motor component. Studies relating impaired cognition in ALS to disease in gray matter and white matter are rare. Our objective was to assess executive function in patients with ALS using a simple, untimed measure with minimal motor demands, and to relate performance to structural disease. Methods: We gave the Visual-Verbal Test to 56 patients with ALS and 29 matched healthy controls. This brief, untimed measure of cognitive flexibility first assesses participants’ ability to identify a feature shared by 3 of 4 simple geometric designs. The participants’ cognitive flexibility is challenged when they are next asked to identify a different feature shared by another combination of 3 of the same 4 geometric designs. In a subset of 17 patients who underwent magnetic resonance imaging, regression analyses related test performance to gray matter atrophy and reduced white matter fractional anisotropy. Results: The patients with ALS showed significant impairment in cognitive flexibility (P < 0.01), with 48.2% making an error on the test. Regression analyses related impaired cognitive flexibility to gray matter atrophy in inferior frontal and insular regions, and to reduced fractional anisotropy in white matter projections in the inferior fronto-occipital and uncinate fasciculi and corpus callosum. Conclusions: Our patients with ALS had impaired cognitive flexibility on an untimed measure with minimal motor demands,
Received for publication January 25, 2014; accepted April 16, 2014. From the *Department of Psychology, Villanova University, Villanova, Pennsylvania; and wDepartment of Neurology and Penn Frontotemporal Degeneration Center, University of Pennsylvania, Philadelphia, Pennsylvania. Supported in part by the National Institutes of Health (AG032953, AG017586, AG038490, NS044266, NS053488, AG043503), the ALS Association, and the Wyncote Foundation. The authors declare no conflicts of interest. Reprints: Murray Grossman, MD, 2 Gibson, Hospital of the University of Pennsylvania, 3400 Spruce St, Philadelphia, Pennsylvania 191044283 (e-mail: [email protected]). Supplemental Digital Content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Website, www.cogbehavneurol.com. Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.
Cogn Behav Neurol
Volume 28, Number 1, March 2015
a finding related in part to a large-scale frontal network that is degraded in ALS. Key Words: amyotrophic lateral sclerosis, cognitive flexibility, prefrontal, white matter, anisotropy (Cogn Behav Neurol 2015;28:17–26)
ALS = amyotrophic lateral sclerosis. DWI = diffusion-weighted imaging. FA = fractional anisotropy. FTD = frontotemporal degeneration. MMSE = Mini-Mental State Examination. MRI = magnetic resonance imaging. SPM8 = Statistical Parametric Mapping Version 8. VVT = Visual-Verbal Test.
A
myotrophic lateral sclerosis (ALS) is a motor neuron disease that causes progressive degeneration of upper and lower motor neurons (Keirnan et al, 2011; Mitchell and Borasio, 2007). Pathologic studies have shown that ALS is a multisystem disorder associated overwhelmingly with transactive response DNA-binding protein of approximately 43 kD (TDP-43) pathology (Brettschneider et al, 2013; Geser et al, 2009). Histopathologic and genetic research has solidified the link between ALS and frontotemporal degeneration (FTD) by showing that TDP-43 pathology is also found in many patients with FTD, and repeat expansions in chromosome 9 open reading frame 72 (C9orf72) are found in families with both of these conditions (DeJesus-Hernandez et al, 2011; Irwin et al, 2013; Neumann et al, 2006; Renton et al, 2011). Further, a spectrum of cognitive changes paralleling the deficits in FTD has been found in up to half of patients with ALS (Murphy et al, 2007; Phukan et al, 2012). Several studies have shown that patients with ALS have limited executive control (Goldstein and Abrahams, 2013). Our present study focuses on these patients’ cognitive flexibility. This facet of executive function involves perceiving the sensorimotor and cognitive attributes of a pattern, exerting inhibitory control over this established pattern, and mentally switching to identify a new pattern. Perhaps the most common sign of a deficit in cognitive flexibility is difficulty with letter-guided naming fluency (Abrahams et al, 2000). However, executive deficits can be difficult to assess in patients with ALS because their www.cogbehavneurol.com |
17
Evans et al
motor disorder may compromise their ability to meet a test’s requirements for timed performance or implementation in a specific response modality, eg, speaking or clicking a computer key. The recognition of these confounds led to important innovations to help control for them. For example, Abrahams et al (2000) corrected the oral verbal fluency task for slowed motor articulation by giving patients the traditional timed oral fluency task and following it with a timed control condition in which they read aloud the word lists that they had generated during the oral fluency task. More recently, an untimed measure of cognitive flexibility in ALS was reported that minimizes dependence on timing and response modality: The DelisKaplan Executive Function System Sorting Test (Delis et al, 2001) uses recognition performance to document difficulty on a categorization task (Libon et al, 2012; Taylor et al, 2013). Unfortunately, this measure is complex and time-consuming to administer. A concise, straightforward task was still needed to assess cognitive flexibility in a clinical setting. In the present study, we examined cognitive flexibility in ALS using the Visual-Verbal Test (VVT), developed by Feldman and Drasgow in 1959. This task requires the patient first to identify an attribute (color, shape, size, orientation, “fill” lines) that is shared by 3 of 4 simple geometric designs, and then to identify a different feature shared by a different set of 3 of the same 4 designs (Supplemental Digital Content 1, http://links.lww.com/ CBN/A58). The VVT is well suited to measuring cognitive flexibility in ALS because it is brief, non-social, and untimed; the response modality is immaterial, minimizing any confound by a patient’s motor limitations. Prefrontal cortex is affected early in ALS (Brettschneider et al, 2013). Impaired cognitive flexibility on measures of patients’ executive control is thought to be caused in part by prefrontal disease (Abrahams et al, 1996, 2004; Kew et al, 1993; Libon et al, 2012), although scores may be confounded by the difficulty of most measures and the motor response requirements. Functional magnetic resonance imaging (MRI) studies in healthy adults have linked inferior frontal cortex (Goel and Vartanian, 2005) and insula (Chang et al, 2013) with cognitive flexibility and set switching. Disease in white matter tracts has also been implicated in impaired executive performance (Abrahams et al, 2005), although again the cognitive measures required motor skills. Impaired category naming fluency has been related to reduced fractional anisotropy (FA) in the cingulum (Sarro et al, 2011), presumably linking simple cognitive switching, supported by ventral frontal regions, with more dorsal regions that modulate the decision-making process on the task. In 2013, Pettit et al used a largely motor-free measure of dual tasking to relate executive function in patients with ALS to reduced FA in several white matter regions in the frontal lobe. In this study, we minimized motor confounds in evaluating the neuroanatomic basis for an executive
18 | www.cogbehavneurol.com
Cogn Behav Neurol
Volume 28, Number 1, March 2015
function deficit in ALS by relating VVT performance to MRI images of gray matter atrophy and reduced FA in white matter. We hypothesized that patients with ALS would be impaired on the VVT, and the impairment would be related to disease in a large-scale network involving inferior prefrontal and insular gray matter regions integrated by frontal white matter projections to form a prefrontal decision-making network.
METHODS Participants We recruited 56 patients with ALS from the outpatient Neurology Clinic at the University of Pennsylvania (Table 1). As described below, we determined that 47 of the patients had ALS alone and the other 9 had ALS combined with FTD. All of the patients with ALS alone and 6 of the 9 with ALS-FTD met the revised El Escorial criteria for ALS (Brooks et al, 2000); records were unavailable for the 3 other patients with ALS-FTD. ALS onset had been bulbar in 11 of the 56 patients, cervical in 20, and lumbosacral in 21; we could not determine the form of motor onset in 4 of the patients with ALS-FTD. We categorized our patients as having either ALS alone or ALS-FTD based on a physician’s (author L.M., L.E., or M.G.) evaluation using published diagnostic criteria (Rascovsky et al, 2011; Strong et al, 2009). All the patients in the ALS-FTD group had been diagnosed with concurrent behavioral variant FTD and motor neuron disease, and in all the FTD phenotype had appeared first. The ALS-only group exhibited only motor neuron disease; clinical evaluation showed that they had mild if any cognitive impairment. We tested the patients an average of about 6 months after they had received their diagnosis. We evaluated them with the Amyotrophic Lateral Sclerosis Functional Rating Scale–Revised (Cedarbaum et al, 1999). We excluded patients who had other potential causes for their cognitive difficulty, such as metabolic or psychiatric disorders. No patients were taking centrally acting medications that could compromise their cognitive function. We recruited 29 elderly controls from the community. The controls were matched with the ALS group for sex and education. The patients were significantly younger than these controls, although age did not correlate with VVT performance: r53 = 0.17; P > 0.3. We used a separate group of 36 healthy seniors as controls solely for the MRI portion of the study. These controls, described below under the Imaging Acquisition and Analysis section, did not take part in the cognitive portion of the study and we refer to them only in the context of the imaging. We gave the patients and cognitive controls the Mini-Mental State Examination (MMSE) (Folstein et al, 1975) to gauge their general cognitive function. The patients had significantly lower MMSE scores than the controls, even though we scored the test proportionately, giving allowances for items that the patients could not perform because of a motor deficit. Nevertheless, the Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
Cogn Behav Neurol
Volume 28, Number 1, March 2015
Impaired Cognitive Flexibility in ALS
TABLE 1. Demographic and Cognitive Performance Data for Patients with Amyotrophic Lateral Sclerosis (ALS), Patients with ALS and Frontotemporal Degeneration (FTD), and Controls ALS ALS-FTD ALS Total Controls (n = 47) (n = 9) (n = 56) (n = 29) Men/women Age (years) Education (years) Disease duration (months) ALS Functional Rating Scale1 Mini-Mental State Exam2 Visual-Verbal Test3 Initial categorization failures (number of errors) Cognitive flexibility (maximum = 10) Letter fluency4 Words/minute total Digit span5 Digits forward Digits backward
29/18 58.19 (13.04) 15.09 (2.80) 6.85 (8.66) 34.13 (8.48) (n = 47) 27.85 (3.23) (n = 47)
6/3 62.44 (9.81) 14.11 (3.41) 5.40 (6.62) 38.57 (3.05) (n = 5) 24.89 (6.33)* (n = 9)
35/21 58.88 (12.59)* 14.93 (2.89) 6.71 (8.44) 34.73 (8.10) (n = 52) 27.38 (4.00)* (n = 56)
13/16 67.76 (7.67) 16.07 (2.67)
0.04 (0.20)
0.22 (0.67)
0.07 (0.32)
0.00 (0.00)
8.55 (2.40)* (n = 47)
6.67 (4.03)* (n = 9)
8.25 (2.77)* (n = 56)
9.41 (1.15) (n = 29)
11.63 (5.31)* (n = 42)
6.81 (1.26)*w (n = 5)
10.91 (5.21)* (n = 47)
14.31 (5.20) (n = 29)
6.72 (1.24) 4.24 (1.763) (n = 40)
6.80 (0.84) 4.20 (1.30) (n = 5)
6.69 (1.10) 4.33 (1.49) (n = 45)
6.95 (1.17) 5.09 (1.74) (n = 29)
29.25 (1.11) (n = 29)
Data are shown as either numbers or mean (standard deviation). *ALS differs from controls, P < 0.05. wALS differs from ALS-FTD, P < 0.05. 1 Cedarbaum et al, 1999. 2Folstein et al, 1975. 3Feldman and Drasgow, 1959. 4Spreen and Strauss, 1991. 5 Wechsler, 1995.
patients’ average MMSE scores were within our adjusted normal range. The study protocol was approved by the University of Pennsylvania Institutional Review Board. All participants and caregivers gave written informed consent before taking part.
Cognitive Tests In addition to the MMSE, we tested cognitive function in the 56 patients and 29 controls with a brief battery that included letter-guided category naming fluency (F-A-S or just F) (Spreen and Strauss, 1991), to measure cognitive flexibility; digit span forward and backward (Wechsler, 1995), to measure attention and working memory; and the VVT, to measure cognitive flexibility (Table 1). During the study we gave the letter fluency task to 47 of the patients and the 29 controls. Early on, we asked 30 of the patients to say as many words starting with the letter F as they could within 1 minute (Spreen and Strauss, 1991). Then we asked them to repeat the task for the letter A and then S. Partway through the study, we changed the test procedure slightly. When we tested 17 more of the patients, we asked them to say only words starting with the letter F. To maximize the available data for analysis, we combined the numbers of F words from both sets of patients. Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
We gave the digit span task to 45 of the patients and the 29 controls. In the digit span forward, participants are asked to repeat a series of spoken number sequences, each of which is longer than the 1 before it. For the digit span backward, participants must repeat the ever-lengthening number sequences in the reverse order from which they are presented. Visual-Verbal Test We gave all participants the VVT. In this test, participants are shown a card that displays 4 simple geometric designs. These designs can be grouped into 2 sets of 3, based on color, size, shape, orientation, or pattern of lines filling the design. Participants are asked first to identify a set of 3 designs that share a feature. They may respond by speaking and/or pointing. On the same card, they are then asked to identify a second set of 3 designs that share a different feature. Supplemental Digital Content 1 (http://links. lww.com/CBN/A58) shows 2 sample cards from the VVT. The upper set of geometric designs can be grouped by color (ABC) and size (ABD). The lower set can be grouped by shape (ABC) and color (ACD). In this study, we gave participants 2 practice trials to ensure that they understood the task. The actual study consisted of 10 untimed trials. www.cogbehavneurol.com |
19
Evans et al
Grouping the first set of 3 geometric designs shows that participants can perceive the designs and group a subset of items based on a shared feature. For this part of the task, we score the number of times that a participant could not identify a shared feature across the designs. Thus, a score of zero indicates perfect performance. Disengaging from the first set of designs and identifying the set based on the second shared feature requires inhibitory control and set switching, which are components of cognitive flexibility. For this part of the task, we count the number of times that participants could switch from the first set to the second set. A score of 10 on this part of the task indicates that participants successfully identified all of the second sets and had intact cognitive flexibility. Although the designs may have differences in the saliency of particular features, the task does not become more difficult over the course of the 10 trials. We performed analysis of variance, t test, and Pearson correlation using Statistical Package for the Social Sciences Version 21 (IBM SPSS Statistics, International Business Machines, Armonk, New York).
Imaging Acquisition and Analysis A subset of 17 of the patients (11 with ALS alone and 6 with ALS-FTD) had high-resolution T1-weighted structural MRI scans to assess their gray matter atrophy, and diffusion-weighted imaging (DWI) to analyze their reduced white matter FA. We scanned only 17 patients because the others had trouble breathing when they lay prone at length in the magnet bore. The patients who had imaging matched the full cohort of 56 patients in sex, age, education, disease duration, and VVT performance (all P values >0.1). A Siemens 3.0 T Trio scanner (Siemens Medical Solutions, Erlangen, Germany) with an 8-channel head coil collected the images. We obtained T1-weighted 3dimensional spoiled gradient-echo images with repetition time = 1620 msec, echo time = 3 msec, flip angle = 15 degrees, matrix = 192 256, slice thickness = 1 mm, and in-plane resolution = 1.0 1.0 mm. For 15 of the 17 scanned patients, we performed a 30-directional DWI sequence using a single-shot, spinecho, diffusion-weighted echo planar imaging sequence (repetition time = 8100 msec, echo time = 83 msec, fat saturation, matrix size = 128 128, number of slices = 70, voxel size = 2 mm isotropic). We collected 30 volumes with diffusion weighting (b-value = 1000 seconds/ mm2) along 30 non-collinear directions, and we collected 1 volume per participant without diffusion weighting (b-value = 0 seconds/mm2). We collected additional b-value = 0 volumes for 8 participants. For the other 2 of the 17 scanned patients, we collected DWI images using a 12-directional sequence with a single-shot, spin-echo, diffusion-weighted echo planar imaging sequence (matrix size = 128 128, number of slices = 40, voxel size = 3 mm isotropic, repetition
20 | www.cogbehavneurol.com
Cogn Behav Neurol
Volume 28, Number 1, March 2015
time = 6500 msec, echo time = 99 msec), acquiring 12 non-collinear, non-coplanar, isotropic diffusion gradients. We also performed MRI scans on a different healthy control group of 36 seniors recruited from the community who were similar in age and other demographic characteristics to our 17 patients (all P values >0.5). Of these controls, 28 underwent 30-directional scans and 8 had 12-directional scans. As we described in an earlier study (Grossman et al, 2013), we normalized the images to a standard space and segmented them using the PipeDream interface (SourceForge, Slashdot Media, San Francisco, California) (http:// sourceforge.net/projects/neuropipedream/) to the Advanced Normalization Tools toolkit (Penn Image Computing & Science Lab, University of Pennsylvania, Philadelphia, Pennsylvania) (http://www.picsl.upenn.edu/ ANTS/) (Avants et al, 2008). A local 1-mm3 resolution T1 template consisted of 25 healthy seniors and 25 patients with FTD. We registered our current participants’ images to the local template, transformed into Montreal Neurological Institute space, and down-sampled to 2-mm3 resolution. We smoothed the images in Statistical Parametric Mapping Version 8 (SPM8) (Wellcome Trust Centre for Neuroimaging, London, United Kingdom; available online at http://www.fil.ion.ucl.ac.uk/spm/software/spm8) using a 5-mm full-width half-maximum isotropic Gaussian kernel to minimize individual gyral variations. To compare our patients’ gray matter atrophy to the controls’, we used a whole-brain voxel-wise analysis in SPM8 with height threshold = q < 0.05 (false discovery rate-corrected), minimum cluster size = 50 voxels, and peak voxel threshold = P < 0.001. We related VVT cognitive flexibility performance to gray matter atrophy using the multiple regressions module in SPM8, constrained to regions of gray matter atrophy. Clusters survived a height threshold = P < 0.005 (uncorrected), minimum cluster size = 40 voxels, and peak voxel threshold = P < 0.001. To preprocess the DWIs, we used PipeDream and Advanced Normalization Tools as described above. We removed motion and distortion artifacts using affine coregistration of each DWI to the unweighted (b-value = 0) image(s). We used a linear least-squares algorithm (Salvador et al, 2005) implemented in the Camino software toolkit (University College London Microstructure Imaging Group, London, United Kingdom) (http:// cmic.cs.ucl.ac.uk/camino/) (Cook et al, 2006) to compute diffusion tensors. We reoriented the tensors using the preservation of principal directions method (Alexander et al, 2001). We used the diffusion tensor image to compute FA for each participant, and registered each FA image to the T1 image. We used the symmetric diffeomorphic procedure in Advanced Normalization Tools to warp T1 images to the local template, and we warped each FA image to Montreal Neurological Institute space by applying the T1-to-template and template-to-Montreal Neurological Institute warps. We smoothed the FA images using a 4-mm full-width half-maximum isotropic Gaussian kernel. Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
Cogn Behav Neurol
Volume 28, Number 1, March 2015
The 2-sample t test module in SPM8 compared FA differences between the patients and controls, constrained to white matter by averaging all patient and control FA images and generating a mask consisting of voxels with FAZ0.25. We compared patients to controls at a false discovery rate-corrected height threshold = q < 0.01 with 200-voxel extent. Using deterministic tractography in Camino, we tracked white matter fibers in the healthy elderly template generated using the 30-directional diffusion tensor imaging sequence described above. We retained fiber tracts passing through voxels of reduced FA, as determined by reduced FA in our patients with ALS compared with our controls, to define the mask for regression analyses. We used the multiple regression module of SPM8 to relate VVT cognitive flexibility to white matter tracts with significantly reduced FA in our patients with ALS, using height threshold = P < 0.005, 50-voxel extent, and peak voxel threshold = P < 0.001.
RESULTS Behavioral Analyses We found impaired cognitive flexibility in our patients with ALS. The 1-way analysis of variance showed
Impaired Cognitive Flexibility in ALS
that the patients and controls did not differ in their ability to identify an initial set of shared-feature designs: F1,83 = 1.416, P = 0.237. However, the patients were significantly less able to identify a second set of sharedfeature designs, consistent with a deficit in cognitive flexibility: F1,83 = 4.674, P = 0.034. Each of the patient subgroups differed from the controls: ALS alone, t74 = 6.928, P = 0.04; ALS-FTD, t36 = 3.342, P = 0.002. We found at least 1 cognitive flexibility error in 27 (48.2%) of the 56 patients, including 9 (69.2%) of 13 patients with autopsy-confirmed ALS; 9 (31%) of the 29 controls had a similar error rate. Cognitive flexibility measured by the VVT correlated with performance on the MMSE: r54 = 0.660, P < 0.001; letter-guided category naming fluency: r45 = 0.350, P = 0.016; and digit span backward: r43 = 0.573, P < 0.001. We found no difference between the subgroups whose ALS onset had been bulbar, cervical, or lumbosacral (all P values > 0.3).
Image Analyses Figure 1 illustrates the MRI analyses in our 17 scanned patients. Figure 1A shows significant gray matter atrophy, primarily in a bilateral frontal distribution, extending into the anterior temporal and parietal
FIGURE 1. Gray matter atrophy and reduced white matter fractional anisotropy, with regression analyses, in 17 patients with amyotrophic lateral sclerosis. Panel A: The patients had significant gray matter atrophy (shown in red) relative to controls. Vertical white lines mark the locations of the slices shown in Panel B. Panel B: Coronal slices illustrate gray matter regressions relating the patients’ gray matter atrophy to impaired cognitive flexibility performance (purple) and reduced white matter fractional anisotropy (green); the regressions shown in blue relate reduced fractional anisotropy to impaired cognitive flexibility performance. The numbers indicate the coordinates of the slices in the Y axis. R = right. L = left. Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
www.cogbehavneurol.com |
21
Cogn Behav Neurol
Evans et al
lobes. Table 2 summarizes the peak foci of the gray matter atrophy, corresponding to areas of atrophy illustrated in Figure 1. Regression analysis, illustrated in Figure 1B and summarized in Table 2, shows that the patients’ cognitive flexibility deficits were related to areas of significant atrophy in bilateral inferior frontal and left insular regions. Figure 1B also illustrates areas of significantly reduced FA in several white matter tracts. Table 3 summarizes areas of reduced FA in the white matter, corresponding to areas visualized in Figure 1B. These areas are the corpus callosum, descending fibers from motor cortex, and projections involving inferior frontal regions: the uncinate fasciculus, inferior fronto-occipital fasciculus, inferior longitudinal fasciculus, and cingulum. Regression analysis, also summarized in Table 3 and illustrated in Figure 1B, shows that difficulty with cognitive
Volume 28, Number 1, March 2015
flexibility was related to reduced FA in frontal white matter, specifically the inferior fronto-occipital fasciculus and uncinate fasciculus, and in the corpus callosum.
DISCUSSION Our findings suggest that the VVT is sensitive to impairments in cognitive flexibility in patients with ALS. The VVT is a simple, brief, untimed measure to which participants can respond by either speaking or pointing. Therefore, this measure is well suited for clinical use in evaluating cognition in patients with ALS. Previous studies suggested that as many as 50% of patients with ALS have cognitive difficulty (Murphy et al, 2007; Phukan et al, 2012). We confirm these estimates with our current finding that 48% of patients with ALS had limited cognitive flexibility on the VVT. Our regression analyses related impaired performance on this
TABLE 2. Anatomic Locations of Significant Gray Matter Atrophy in Patients with Amyotrophic Lateral Sclerosis (ALS) Relative to Controls, and Regressions Relating Impaired Cognitive Flexibility to Atrophy Anatomic Locus (Brodmann Area)
MNI Coordinates x y z
z Score
Cluster Size (Voxels)
Gray matter atrophy R primary motor (4) R premotor (6) R inferior frontal (44) R middle frontal (10) R middle frontal (46) R cingulate (32) R medial frontal (9) R medial frontal (11) R orbital frontal (11) R insula L inferior frontal (47) L middle frontal (10) L middle frontal (9) L cingulate (24) L orbital frontal (11) L insula R superior temporal (22) R superior temporal (38) R middle temporal (21) R middle temporal (21) R inferior temporal (20) R fusiform (20) R lingual (19) L middle temporal (21) L middle temporal (21) L fusiform (20) L hippocampal (28) R postcentral (2)
44 34 42 28 38 12 20 12 24 40 40 34 36 12 28 42 50 48 50 52 52 40 22 54 50 44 20 62
8 2 8 50 34 34 28 34 32 0 24 46 16 38 36 2 42 14 2 14 6 18 44 12 2 18 10 18
36 50 30 6 22 24 36 20 14 4 2 6 30 12 14 6 6 20 18 12 32 28 10 14 32 30 16 28
3.86 4.25 5.74 4.56 3.97 5.2 4.74 4.4 4.38 5.11 3.71 3.89 3.49 4.67 3.68 3.63 3.49 3.34 4.08 3.86 4.52 4.77 4.35 4.14 3.63 4.35 3.74 3.68
109 73 266 227 141 247 165 147 145 710 135 73 83 546 106 90 50 73 65 115 201 127 76 299 73 94 62 68
Regressions R inferior frontal (47) L orbital frontal (11) L insula
32 32 36
28 34 20
0 12 4
3.47 3.26 3.64
53 57 42
MNI = Montreal Neurological Institute. R = right. L = left.
22 | www.cogbehavneurol.com
Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
Cogn Behav Neurol
Volume 28, Number 1, March 2015
Impaired Cognitive Flexibility in ALS
TABLE 3. Reductions in White Matter Fractional Anisotropy in Patients with Amyotrophic Lateral Sclerosis (ALS) Relative to Controls, and Regressions Relating Impaired Cognitive Flexibility to Reduced Fractional Anisotropy MNI Coordinates Anatomic Locus Cluster Size (White Matter Tract) x y z z Score (Voxels) Reduced fractional anisotropy R inferior frontal white matter R orbital frontal white matter R inferior fronto-occipital fasciculus R middle fronto-occipital fasciculus B corpus callosum (frontal and body) B fornix L inferior frontal gyrus white matter L inferior fronto-occipital fasciculus L inferior longitudinal fasciculus L inferior longitudinal fasciculus L cingulum L cerebral peduncle
15 10 26 31 0 2 35 44 41 38 21 18
8 23 28 11 16 6 12 35 21 5 26 13
13 12 12 8 29 17 23 2 15 22 17 16
5.23 5.2 4.34 3.84 5.55 5.59 4.28 4.52 4.97 4.28 4.69 6.02
951 418 248 410 11958 1114 619 630 2000 291 283 1446
Regressions R uncinate fasciculus R corpus callosum (frontal) R corpus callosum (frontal) R corpus callosum (frontal) L uncinate fasciculus L inferior frontal gyrus white matter L inferior fronto-occipital fasciculus L corpus callosum (frontal) L corpus callosum (frontal) L corpus callosum (frontal)
23 17 11 19 17 37 38 15 14 13
21 55 30 7 26 5 17 52 44 19
11 6 7 50 10 23 5 2 23 44
3.72 5.02 4.18 3.2 3.27 3.65 3.92 4.11 3.51 3.09
99 1273 5078 69 256 115 455 745 268 82
MNI = Montreal Neurological Institute. R = right. B = bilateral. L = left.
task to gray matter regions in prefrontal cortex (areas of known pathological burden in ALS) and to interconnecting frontal white matter projections of the uncinate and inferior fronto-occipital fasciculi and corpus callosum. To some extent, informal observation and a clinical interview can reveal impaired cognition in patients’ dayto-day performance. However, quantification of cognitive difficulty is essential to determining progression and prognosis, and may be helpful in evaluating response during a treatment trial. Quantifying impairment has proven difficult with most standard cognitive measures, such as the MMSE and Addenbrooke Cognitive Examination–Revised (Leyton et al, 2010), because many components of these tests require a motor response. Indeed, in an unpublished survey of our clinical cohort of 123 patients with ALS, we found that only 59% were able to complete all components of the MMSE, while 100% successfully completed the VVT. Measures of executive function such as letter-guided category naming fluency are sensitive to cognitive deficits in ALS (Goldstein and Abrahams, 2013). However, some patients with a bulbar motor disorder have slowed speech (Ash et al, 2014), which might confound their performance on a timed cognitive task requiring speech Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
production. Abrahams et al (2004) introduced an important innovation to letter-guided category naming fluency: a control task (timed reading of the words produced by the patient) that compensates in part for the motor deficit, thereby defining more clearly an impairment on the task’s cognitive component. Still, even this control does not fully eliminate the confounding motor deficit. Abrahams et al (2014) published norms for a written version of letter-guided category naming fluency, although this only shifts the motor confound to another response modality. The VVT also minimizes confounds related to other aspects of executive function. For example, identifying a deficit in cognitive flexibility is less confounded with the VVT than with the Wisconsin Card Sorting Task, a lengthy measure of cognitive flexibility that also requires significant working memory (Anderson et al, 1991), because patients taking the VVT are explicitly told that they need to make a mental shift. Patients with ALS have difficulty with lengthy and demanding measures of cognitive flexibility. Libon et al (2012) and Taylor et al (2013) asked patients with ALS to group 16 cards into sets according to a variety of shared features (Delis et al, 2001). Although this task was not timed, the patients showed significant impairment. One www.cogbehavneurol.com |
23
Evans et al
problem was that they had motor difficulty manipulating the cards into groups. To minimize the motor component of the task, the researchers gave the patients a recognition procedure, asking them to recognize the rule for grouping the cards. Still, their performance remained significantly impaired. We think that this task is too lengthy to be suited for administration in a busy clinic. In the present study, we assessed our patients’ cognitive flexibility with the VVT, which is a brief, untimed measure that does not depend on a specific response modality. Our patients had little difficulty perceiving their first pattern of shared features across geometric designs. This finding leads us to believe that visual perception and perceptual categorization of geometric features did not play a major role in the patients’ cognitive deficit. However, they were significantly impaired on the subsequent measure of cognitive flexibility, because they had difficulty identifying a second set of 3 stimuli that shared a different feature. Eslinger et al (2007) reported impaired VVT performance in patients with FTD. We found significant impairment even in our subgroup of patients with ALS who did not have concurrent FTD. Our patients’ VVT performance correlated with their performance on letterguided category naming fluency and digit span backward tasks, which are traditional measures of executive function and working memory. Widespread gray matter and white matter disease is now recognized in ALS (Agosta et al, 2007; Ellis et al, 2001; Verstraete et al, 2012). The affected areas extend beyond primary motor cortex to include inferior and dorsolateral prefrontal cortical regions that are important for executive function, as well as anterior temporal and parietal gray matter areas. This spread corresponds to the anatomic distribution of pathology in ALS (Geser et al, 2008). Likewise, patients can have substantial disease in the associated white matter corticospinal fibers and white matter tracts that have been implicated in the prefrontal projections that subserve large-scale cognitive networks thought to play a role in executive function (Mioshi et al, 2013). This distribution, too, corresponds to the white matter pathology in ALS (Douaud et al, 2011). In the present study, we found disease in gray matter regions and white matter tracts similar to those reported in previous studies. Moreover, we were able to identify a compromised network of gray matter regions and white matter tracts associated with impaired cognitive flexibility in ALS. Regression analysis implicated inferior frontal cortex in our patients’ impaired cognitive flexibility. Previous functional MRI work in healthy adults has demonstrated that inferior frontal cortex contributes to inhibitory control and set switching (Goel and Vartanian, 2005). Another functional MRI study in healthy adults showed a role for the insula in cognitive flexibility and set switching (Chang et al, 2013). Our regression analysis also implicated the insula in cognitive flexibility difficulty in ALS. Previous work has also associated dorsolateral prefrontal areas with impaired cognitive flexibility in patients with ALS (Abrahams et al, 1996, 2004). There are
24 | www.cogbehavneurol.com
Cogn Behav Neurol
Volume 28, Number 1, March 2015
several reasons why we may not have seen this association in the present study. One is that the prior work may have combined complex cognitive challenges with the demands of a motor component in patients who had a motor deficit. However, our own previous study (Libon et al, 2012) linked category recognition impairment to dorsolateral prefrontal gray matter atrophy on a task that made minimal motor demands. This finding raises the possibility that the relative cognitive demands of the task itself influence the anatomic distribution of prefrontal cortex required for cognitive flexibility: More demanding switching tasks might be associated with dorsolateral prefrontal cortex, and less demanding tasks with ventral frontal regions. Functional MRI studies in healthy adults have shown different patterns of prefrontal activation depending on the cognitive demands of the task (Badre and D’Esposito, 2007; Koechlin and Jubault, 2006). More work is needed to examine the role of varying cognitive task demands explicitly in the gray matter as a basis for impaired cognitive flexibility in ALS. White matter disease has been seen consistently enough in patients with ALS to warrant the proposal that imaging changes in white matter may serve as a biomarker of ALS (Foerster et al, 2013). We, too, observed significant white matter disease in the corpus callosum and white matter tracts in the frontal lobe. Rare studies have been performed that relate white matter disease directly to motor deficits in ALS (Verstraete et al, 2011). Previous studies related impaired letter-guided category naming fluency to reduced white matter volume in ALS (Abrahams et al, 2005) and to reduced FA in the cingulum (Sarro et al, 2011). More recent work has related FA in several white matter regions of interest to performance on a largely motor-free dual task, but not to measures of information processing speed (Pettit et al, 2013). In the present study, our regressions related white matter projections in the frontal lobe to impaired cognitive flexibility in ALS. Among the implicated tracts was the inferior fronto-occipital fasciculus, which links prefrontal with posterior temporal regions. Temporal regions presumably contributed to our participants’ perception of the visual stimuli in the VVT, while inferior frontal regions maintained control over processing of the perceived sets. We also found that the uncinate fasciculus plays a role in cognitive flexibility. This fasciculus links inferior frontal and anterior temporal regions that may be related to inhibitory control. Finally, our regression analysis showed a role for the anterior corpus callosum in cognitive flexibility. This region may be important in coordinating performance across the left and right frontal regions. Our work thus describes a potential connectome that might mediate cognitive flexibility deficits in ALS. Several caveats should be kept in mind when considering our findings. While we collected behavioral data on a large number of patients with ALS, we were able to obtain MRI data for only 17 (30%). Second, the materials and procedures that we used in this study did not permit us to determine the precise role of each frontal Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
Cogn Behav Neurol
Volume 28, Number 1, March 2015
region and white matter projection in cognitive flexibility. Finally, in Table 1 we report an unvalidated MMSE score that adjusts for items that some patients could not complete because of motor limitations; we include this measure to provide a rough clinical guide to the patients’ overall cognitive function. These caveats notwithstanding, our findings are consistent with the hypothesis that many patients with ALS exhibit a deficit in cognitive flexibility even when performing a simple, untimed task that minimizes motor confounds, and this deficit appears to be associated with disease in a large-scale neural network involving both gray matter and white matter regions in non-motor portions of the frontal lobe. REFERENCES Abrahams S, Goldstein LH, Kew JJM, et al. 1996. Frontal lobe dysfunction in amyotrophic lateral sclerosis: a PET study. Brain. 119:2105–2120. Abrahams S, Goldstein LH, Simmons A, et al. 2004. Word retrieval in amyotrophic lateral sclerosis: a functional magnetic resonance imaging study. Brain. 127:1507–1517. Abrahams S, Goldstein LH, Suckling J, et al. 2005. Frontotemporal white matter changes in amyotrophic lateral sclerosis. J Neurol. 252:321–331. Abrahams S, Leigh PN, Harvey A, et al. 2000. Verbal fluency and executive dysfunction in amyotrophic lateral sclerosis (ALS). Neuropsychologia. 38:734–747. Abrahams S, Newton J, Niven E, et al. 2014. Screening for cognition and behaviour changes in ALS. Amyotroph Lateral Scler Frontotemporal Degener. 15:9–14. Agosta F, Pagani E, Rocca MA, et al. 2007. Voxel-based morphometry study of brain volumetry and diffusivity in amyotrophic lateral sclerosis patients with mild disability. Hum Brain Mapp. 28:1430–1438. Alexander DC, Pierpaoli C, Basser PJ, et al. 2001. Spatial transformations of diffusion tensor magnetic resonance images. IEEE Trans Med Imaging. 20:1131–1139. Anderson S, Damasio H, Jones R, et al. 1991. Wisconsin Card Sorting Test performance as a measure of frontal lobe damage. J Clin Exp Neuropsychol. 13:909–922. Ash S, Olm C, McMillan CT, et al. 2014. Deficits in sentence expression in amyotrophic lateral sclerosis. Amyotroph Lateral Scler Frontotemporal Degener. 8:1–9. Avants BB, Epstein CL, Grossman M, et al. 2008. Symmetric diffeomorphic image registration with cross-correlation: evaluating automated labeling of elderly and neurodegenerative brain. Med Image Anal. 12:26–41. Badre D, D’Esposito M. 2007. Functional magnetic resonance imaging evidence for a hierarchical organization of the prefrontal cortex. J Cogn Neurosci. 19:2082–2099. Brettschneider J, Del Tredici K, Toledo JB, et al. 2013. Stages of pTDP43 pathology in amyotrophic lateral sclerosis. Ann Neurol. 74:20–38. Brooks BR, Miller RG, Swash M, et al. 2000. El Escorial revisited: revised criteria for the diagnosis of amyotrophic lateral sclerosis. Amyotroph Lateral Scler Other Motor Neuron Disord. 1:293–299. Cedarbaum JM, Stambler N, Malta E, et al. 1999. The ALSFRS-R: a revised ALS functional rating scale that incorporates assessments of respiratory function: BDNF ALS Study Group (Phase III). J Neurol Sci. 169: 13–21. Chang LJ, Yarkoni T, Khaw MW, et al. 2013. Decoding the role of the insula in human cognition: functional parcellation and large-scale reverse inference. Cereb Cortex. 23:739–749. Cook PA, Bai Y, Nadjati-Gilani S, et al. 2006. Camino: open-source diffusion-MRI reconstruction and processing. Proc Intl Soc Mag Reson Med. 14:2759. DeJesus-Hernandez M, Mackenzie IR, Boeve BF, et al. 2011. Expanded GGGGCC hexanucleotide repeat in noncoding region of C9ORF72 causes chromosome 9p-linked FTD and ALS. Neuron. 72: 245–256.
Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
Impaired Cognitive Flexibility in ALS
Delis DC, Kaplan E, Kramer JH. 2001. Delis-Kaplan Executive Function Battery. San Antonio, Texas: The Psychological Corporation. Douaud G, Filippini N, Knight S, et al. 2011. Integration of structural and functional magnetic resonance imaging in amyotrophic lateral sclerosis. Brain. 134:3467–3476. Ellis CM, Suckling J, Amaro E. 2001. Volumetric analysis reveals corticospinal tract degeneration and extramotor involvement in ALS. Neurology. 57:1571–1578. Eslinger PJ, Moore P, Troiani V, et al. 2007. Oops! Resolving social dilemmas in frontotemporal dementia. J Neurol Neurosurg Psychiatry. 78: 457–460. Feldman MJ, Drasgow J. 1959. The Visual-Verbal Test. Los Angeles, California: Western Psychological Services. Foerster BR, Dwamena BA, Petrou M, et al. 2013. Diagnostic accuracy of diffusion tensor imaging in amyotrophic lateral sclerosis: a systematic review and individual patient data meta-analysis. Acad Radiol. 20:1099–1106. Folstein MF, Folstein SE, McHugh PR. 1975. Mini-mental state: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 12:189–198. Geser F, Martinez-Lage M, Robinson J, et al. 2009. Clinical and pathological continuum of multisystem TDP-43 proteinopathies. Arch Neurol. 66:180–189. Geser F, Winton MJ, Kwong LK, et al. 2008. Pathological TDP-43 in parkinsonism-dementia complex and amyotrophic lateral sclerosis of Guam. Acta Neuropathol. 145:115–133. Goel V, Vartanian O. 2005. Dissociating the roles of right ventral lateral and dorsal lateral prefrontal cortex in generation and maintenance of hypotheses in set-shift problems. Cereb Cortex. 15: 1170–1177. Goldstein LH, Abrahams S. 2013. Changes in cognition and behaviour in amyotrophic lateral sclerosis: nature of impairment and implications for assessment. Lancet Neurol. 12:368–380. Grossman M, Peelle JE, Smith EE, et al. 2013. Category-specific semantic memory: converging evidence from bold fMRI and Alzheimer’s disease. Neuroimage. 68:263–274. Irwin DJ, McMillan CT, Brettschneider J, et al. 2013. Cognitive decline and reduced survival in C9orf72 expansion frontotemporal degeneration and amyotrophic lateral sclerosis. J Neurol Neurosurg Psychiatry. 84:163–169. Keirnan MC, Vucic S, Cheah BC, et al. 2011. Amyotrophic lateral sclerosis. Lancet. 377:942–955. Kew JJ, Goldstein LH, Leigh PN, et al. 1993. The relationship between abnormalities of cognitive function and cerebral activation in amyotrophic lateral sclerosis: a neuropsychological and positron emission tomography study. Brain. 116:1399–1423. Koechlin E, Jubault T. 2006. Broca’s area and the hierarchical organization of human behavior. Neuron. 50:963–974. Leyton CE, Hornberger M, Mioshi E, et al. 2010. Application of Addenbrooke’s Cognitive Examination to diagnosis and monitoring of progressive primary aphasia. Dement Geriatr Cogn Disord. 29:504–509. Libon DJ, McMillan C, Avants B, et al. 2012. Deficits in concept formation in amyotrophic lateral sclerosis. Neuropsychology. 26:422–429. Mioshi E, Lillo P, Yew B, et al. 2013. Cortical atrophy in ALS is critically associated with neuropsychiatric and cognitive changes. Neurology. 80:1117–1123. Mitchell JD, Borasio GD. 2007. Amyotrophic lateral sclerosis. Lancet. 369:2031–2041. Murphy JM, Henry RG, Langmore S, et al. 2007. Continuum of frontal lobe impairment in amyotrophic lateral sclerosis. Arch Neurol. 64:530–534. Neumann M, Sampathu DM, Kwong LK, et al. 2006. Ubiquinated TDP-43 in frontotemporal lobar degeneration and amyotrophic lateral sclereosis. Science. 314:130–133. Pettit LD, Bastin ME, Smith C, et al. 2013. Executive deficits, not processing speed relates to abnormalities in distinct prefrontal tracts in amyotrophic lateral sclerosis. Brain. 136:3290–304. Phukan J, Elamin M, Bede P, et al. 2012. The syndrome of cognitive impairment in amyotrophic lateral sclerosis: a populationbased study. J Neurol Neurosurg Psychiatry. 83:102–108.
www.cogbehavneurol.com |
25
Evans et al
Rascovsky K, Hodges JR, Knopman D, et al. 2011. Sensitivity of revised diagnostic criteria for the behavioural variant of frontotemporal dementia. Brain. 134:2456–2477. Renton AE, Majounie E, Waite A, et al. 2011. A hexanucleotide repeat expansion in C9ORF72 is the cause of chromosome 9p21-linked ALS-FTD. Neuron. 72:257–268. Salvador R, Pena A, Menon DK, et al. 2005. Formal characterization and extension of the linearized diffusion tensor model. Hum Brain Mapp. 24:144–155. Sarro L, Agosta F, Canu E, et al. 2011. Cognitive functions and white matter tract damage in amyotrophic lateral sclerosis: a diffusion tensor tractography study. AJNR Am J Neuroradiol. 32:1866–1872. Spreen O, Strauss EA. 1991. A Compendium of Neuropsychological Tests. New York: Oxford University Press.
26 | www.cogbehavneurol.com
Cogn Behav Neurol
Volume 28, Number 1, March 2015
Strong MJ, Grace GM, Freedman M, et al. 2009. Consensus criteria for the diagnosis of frontotemporal cognitive and behavioural syndromes in amyotrophic lateral sclerosis. Amyotroph Lateral Scler. 10:131–146. Taylor LJ, Brown RG, Tsermentseli S, et al. 2013. Is language impairment more common than executive dysfunction in amyotrophic lateral sclerosis? J Neurol Neurosurg Psychiatry. 84:494–498. Verstraete E, Veldink JH, Hendrikse J, et al. 2012. Structural MRI reveals cortical thinning in amyotrophic lateral sclerosis. J Neurol Neurosurg Psychiatry. 83:383–388. Verstraete E, Veldink JH, Mandl RC, et al. 2011. Impaired structural motor connectome in amyotrophic lateral sclerosis. PLoS One. 6:e24239. Wechsler D. 1995. Wechsler Adult Intelligence Scale. San Antonio: Psychological Press.
Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
