**3.3 Results**

### **3.3.1 Participant characteristics**

The 49 ALS patients and the 25 HCs did not differ in age or education level (Table 1). There were significantly more female HCs than female ALS patients (p=0.006). Gender correlated significantly with Troyer's average number of switches for phonemic fluency; thus, gender was entered as a covariate in this analysis. There were no significant demographic differences between the three groups of ALS patients coded for degree of cognitive dysfunction (Table 1), including site of onset (p = 0.36) and total ALS-FRS scores (p *=* 0.34).

*Phonemic fluency.* ALS patients generated fewer numbers of clusters than did HCs (p = 0.04; Figure 1). ALS patients also generated fewer switches between clusters; however, once gender was entered as a covariate in the analysis, this difference was no longer significant (p =0.14). The number of clusters differed significantly between the ALS groups, with the ALSintact group scoring higher than the ALS-mild and the ALS-FTD groups (p = 0.004; Table 2). The number of switches also differed between the ALS cognitive groups, with the ALS-

Verbal fluency tests were administered to ALS patients and HC in the following manner: Patients were asked to generate a list of words that began with a specific letter (F, A, & S was used for phonemic fluency) or category (Animals was used for semantic fluency) in a 1 minute period. Prior to administering the test, patients were told that proper nouns and root words with different suffices were not allowed. Words generated (including repetitions and rule breaks) were recorded verbatim. The total number of words generated, excluding repetitions and rule breaks, was the standard measure of analysis and Troyer's (Troyer et al,

*Phonemic fluency.* Clusters are scored for groups of phonemic words, or words that are similar based on phonemic rules for each letter. Troyer and colleagues (1997) defined parameters for scoring clusters, including: (a) words beginning with the same first two letters, (b) rhyming words, and (c) words that are the same syllabic length and differ only by a vowel sound. For example, *follow, fog, fond, foster, forget* is a cluster of four because the words all begin with the same phoneme, and the cluster size begins with the second word of a cluster. Each word that is not classified in a group of related phonemic words is scored as a cluster of zero. The *number of clusters* is defined as the total sum of individual clusters, including clusters of zero. The *cluster value* is defined as the sum number of consecutive related words excluding the first word of each grouping, or the sum of the values assigned to the clusters. *Switches* are defined as any break between clusters,

*Semantic fluency.* In the semantic fluency task, clusters are composed of words that are semantically related. Troyer's method defines the categories for finding semantic clusters in "Animals" as living environments, zoological categories, and human use, with each supraordinate category containing specific exemplars. The *cluster size* begins with the second word of a cluster. For example the group, *cow, horse, chicken,* and *rooster,* is scored as a cluster of three because they are all farm animals. Number of clusters, cluster value, and switches

The 49 ALS patients and the 25 HCs did not differ in age or education level (Table 1). There were significantly more female HCs than female ALS patients (p=0.006). Gender correlated significantly with Troyer's average number of switches for phonemic fluency; thus, gender was entered as a covariate in this analysis. There were no significant demographic differences between the three groups of ALS patients coded for degree of cognitive dysfunction (Table 1), including site of onset (p = 0.36) and total ALS-FRS scores

*Phonemic fluency.* ALS patients generated fewer numbers of clusters than did HCs (p = 0.04; Figure 1). ALS patients also generated fewer switches between clusters; however, once gender was entered as a covariate in the analysis, this difference was no longer significant (p =0.14). The number of clusters differed significantly between the ALS groups, with the ALSintact group scoring higher than the ALS-mild and the ALS-FTD groups (p = 0.004; Table 2). The number of switches also differed between the ALS cognitive groups, with the ALS-

1997) scoring methods for clustering and switching were utilized.

**3.2.2 Fluency scoring methods** 

**3.2.3 Verbal fluency components** 

including clusters of zero.

**3.3 Results** 

(p *=* 0.34).

were calculated as discussed above.

**3.3.1 Participant characteristics** 

intact patients switching more often than both the ALS-mild and the ALS-FTD groups (p =0 .004; Table 2). The cluster value scores did not differ significantly between the ALS groups (p=0 .13).


Fig. 1. Total Phonemic and Semantic Fluency Scores for ALS and HC groups

*Semantic fluency.* The total group of ALS patients generated fewer numbers of clusters (p = 0.01) and made significantly fewer switches between clusters (p =0.03) than did the HCs. The total groups did not differ significantly on the number of words within semantic clusters (cluster value=0.15). The ALS-FTD patients generated a smaller cluster value score than did ALS-intact and ALS-mild groups (p=0.03). The number of clusters and number of switches demonstrated trends toward significant differences between the groups (p=0.07, p= 0.06, respectively; Table 2).

Overview of Cognitive Function in ALS, with Special

criteria for Frontotemporal Lobar Degeneration.

patients (Chang et al., 2005).

observations in ALS patients could suggest temporal pathology.

Attention to the Temporal Lobe: Semantic Fluency and Rating the Approachability of Faces 713

changed in the last decade with several neuropsychological and functional imaging studies confirming common involvement of cortex outside the motor strip. The concept of primarily frontal lobe dysfunction in motor neuron disease was introduced by Montgomery and Erickson (1987) as well as Iwasaki et al. (1990). Several studies have since confirmed the association between FTLD, executive dysfunction, and ALS (eg. Massman et al., 1996; Strong et al., 2009). The largest study examining cognitive function in ALS to date found that 51% of patients had varying degrees of executive dysfunction (Ringholz et al., 2005; n = 279). These numbers confirmed an earlier study by Lomen-Hoerth and colleagues (2003) who had found evidence for frontal executive deficits in half of their patients, many of whom met

Whereas frontal pathology has become the focus of cognitive investigation in ALS patients, the integrity of temporal structures (apart from the hippocampal formation) in ALS has not received much attention. Temporal pathology is a hallmark of FTLD, and several behavioral

Several imaging studies and neuropathological investigations suggest that involvement of the temporal cortex, as well as the amygdala and other limbic structures in the disease process is very likely. Kew et al. (1993) showed reduced blood flow (rCBF) in the anterior cingulate cortex, the medial prefrontal cortex (Brodmann area 9 and 10), parahippocampal gyri and the anterior thalamic nuclear complex. Abrahams and colleagues (1995) observed decreased activity across a wide area of the frontal lobes, which also included the insular cortex and thalamic nuclear complex. In a small sample of clinically non-demented patients, there was a decrease in cerebral blood flow of the frontal and temporal lobes, despite normal MR imaging (Kokubo et al., 2003). Recently, a morphometric study of gray matter volume on MR scans revealed significant differences between patients with ALS and normal controls, predominantly in fronto-temporal areas, regardless of cognitive status; that is, the differences in gray matter volume between the ALS group as a whole and the control group were much more extensive than differences between cognitively normal and demented ALS

There have been a limited number of neuropathological studies looking at frontal and temporal pathology in ALS. Wilson et al. (2001) found changes that where overall more pronounced for cognitively affected patients (ubiquitin positive, alpha-synuclein-negative, and tau-negative neuronal inclusions), most pronounced in the cingulate cortex. Cognitive impairment was uniformly associated with superficial linear spongiosis, a pathologic feature common to several forms of frontotemporal dementia. Wilson and colleagues did not study temporal structures in more detail. In a group of ALS patients with cognitive impairment, and decreased frontal blood flow on SPECT, neuropathologic examination

showed spongy degeneration and neuronal loss in the frontal lobe (Abe et al., 1997).

involvement, especially impairments in semantic processing and amygdala function.

Some studies have specifically evaluated the limbic system in sporadic ALS. In the ALS-Parkinson-dementia complex of Guam, tau and alpha-synuclein aggregates are a common

Pathologically, there is also a special tie between semantic dementia, or temporal variant FTD, and ALS. Both FTD-MND (or FTD-ALS) and semantic dementia are characterized by ubiquinated inclusions (FTD-U); the clinical spectrum of patients seen with this histopathological finding varies from ALS, ALS with FTD and semantic dementia without ALS (Davies & Xuereb, 2007). This suggests that patients with ALS caused by this histopathological subtype would be expected to have overt or subtle features of temporal


### **3.4 Discussion**

These results support the findings that ALS patients demonstrate cognitive impairment localizing to both the frontal and temporal lobes, highlighting the frontotemporal neurocognitive phenotype of this disease (Lepow et al., 2010). ALS patients exhibited decreased phonemic and semantic fluency performances as compared to healthy nonneurologically impaired controls. Furthermore, in comparison to ALS patients whose cognition was intact, the subset of ALS patients with mild cognitive dysfunction or ALS-FTD demonstrated performance declines on standard measures of verbal fluency and the component processes of these measures. The component processes of verbal fluency provide a unique opportunity to further evaluate the ALS frontotemporal neurocognitive phenotype from slightly different perspective (Troyer et al., 1997). For phonemic fluency, the intact ALS sample generated fewer clusters and more switches than the ALS-mild and ALS-FTD patients, suggesting temporal involvement in ALS patients, with increasing frontal lobe involvement in patients with greater cognitive dysfunction. For semantic fluency, similar results were obtained with a greater emphasis on declines in clustering or increased temporal lobe dysfunction. These results suggest that verbal fluency measures identify frontal and temporal lobe involvement in the cognitive decline associated with ALS, particularly when the component processes are evaluated.

As a group, the ALS patients demonstrated temporal lobe involvement as compared to individuals without ALS. However, when the ALS patients were stratified based on their level of cognitive dysfunction, the influence of the frontal lobe involvement became more pertinent to their ability to perform this task. In conclusion, the differences in phonemic and semantic fluency scores between ALS patients and HCs suggest temporal lobe involvement in ALS patients with increasing frontal lobe involvement across the neurocognitive spectrum of the disease. A frontotemporal neurocognitive phenotype is revealed in ALS patients who demonstrate cognitive changes.

### **4. General discussion**

Up until the late 1980s, the prevalent view in the neurological literature was that ALS was a pure motor neuron disease only infrequently affecting cognitive function. This view has

**ALS mild**  Mean (SD)

30.1 (9.76) 1 18.3 (5.72) 1

7.15 (2.79)1 2.72 (1.62) 6.15 (2.81) 1

8.85 (3.83) 8.54 (3.26) 7.46 (3.20)

These results support the findings that ALS patients demonstrate cognitive impairment localizing to both the frontal and temporal lobes, highlighting the frontotemporal neurocognitive phenotype of this disease (Lepow et al., 2010). ALS patients exhibited decreased phonemic and semantic fluency performances as compared to healthy nonneurologically impaired controls. Furthermore, in comparison to ALS patients whose cognition was intact, the subset of ALS patients with mild cognitive dysfunction or ALS-FTD demonstrated performance declines on standard measures of verbal fluency and the component processes of these measures. The component processes of verbal fluency provide a unique opportunity to further evaluate the ALS frontotemporal neurocognitive phenotype from slightly different perspective (Troyer et al., 1997). For phonemic fluency, the intact ALS sample generated fewer clusters and more switches than the ALS-mild and ALS-FTD patients, suggesting temporal involvement in ALS patients, with increasing frontal lobe involvement in patients with greater cognitive dysfunction. For semantic fluency, similar results were obtained with a greater emphasis on declines in clustering or increased temporal lobe dysfunction. These results suggest that verbal fluency measures identify frontal and temporal lobe involvement in the cognitive decline associated with ALS, particularly

As a group, the ALS patients demonstrated temporal lobe involvement as compared to individuals without ALS. However, when the ALS patients were stratified based on their level of cognitive dysfunction, the influence of the frontal lobe involvement became more pertinent to their ability to perform this task. In conclusion, the differences in phonemic and semantic fluency scores between ALS patients and HCs suggest temporal lobe involvement in ALS patients with increasing frontal lobe involvement across the neurocognitive spectrum of the disease. A frontotemporal neurocognitive phenotype is revealed in ALS patients

Up until the late 1980s, the prevalent view in the neurological literature was that ALS was a pure motor neuron disease only infrequently affecting cognitive function. This view has

**ALS FTD** 

23.3 (6.54) 2 12.9 (4.05) 2

6.43 (2.17)2 1.67 (1.12) 5.43 (2.17) 2

6.43 (3.41)1 7.29 (1.98) 6.29 (1.98)

Mean (SD) **p-value** 

**0.009 0.02**

**0.004**  0.13 **0.004** 

**0.03**  0.07 0.06

**Table 2 ALS intact** 

when the component processes are evaluated.

who demonstrate cognitive changes.

**4. General discussion** 

**Total** 

**Troyer** 

**Phonemic Semantic** 

Switches

**3.4 Discussion** 

**Phonemic Fluency**  Number of Clusters Cluster Value Switches **Semantic Fluency**  Cluster Value Number of clusters

Mean (SD)

37.2 (10.9) 1,2 19.0 (5.29) 1,2

9.84 (2.34)1,2 3.20 (1.77) 8.84 (2.34) 1,2

10.94 (3.78)1 10.59 (3.73) 9.59 (3.73)

changed in the last decade with several neuropsychological and functional imaging studies confirming common involvement of cortex outside the motor strip. The concept of primarily frontal lobe dysfunction in motor neuron disease was introduced by Montgomery and Erickson (1987) as well as Iwasaki et al. (1990). Several studies have since confirmed the association between FTLD, executive dysfunction, and ALS (eg. Massman et al., 1996; Strong et al., 2009). The largest study examining cognitive function in ALS to date found that 51% of patients had varying degrees of executive dysfunction (Ringholz et al., 2005; n = 279). These numbers confirmed an earlier study by Lomen-Hoerth and colleagues (2003) who had found evidence for frontal executive deficits in half of their patients, many of whom met criteria for Frontotemporal Lobar Degeneration.

Whereas frontal pathology has become the focus of cognitive investigation in ALS patients, the integrity of temporal structures (apart from the hippocampal formation) in ALS has not received much attention. Temporal pathology is a hallmark of FTLD, and several behavioral observations in ALS patients could suggest temporal pathology.

Several imaging studies and neuropathological investigations suggest that involvement of the temporal cortex, as well as the amygdala and other limbic structures in the disease process is very likely. Kew et al. (1993) showed reduced blood flow (rCBF) in the anterior cingulate cortex, the medial prefrontal cortex (Brodmann area 9 and 10), parahippocampal gyri and the anterior thalamic nuclear complex. Abrahams and colleagues (1995) observed decreased activity across a wide area of the frontal lobes, which also included the insular cortex and thalamic nuclear complex. In a small sample of clinically non-demented patients, there was a decrease in cerebral blood flow of the frontal and temporal lobes, despite normal MR imaging (Kokubo et al., 2003). Recently, a morphometric study of gray matter volume on MR scans revealed significant differences between patients with ALS and normal controls, predominantly in fronto-temporal areas, regardless of cognitive status; that is, the differences in gray matter volume between the ALS group as a whole and the control group were much more extensive than differences between cognitively normal and demented ALS patients (Chang et al., 2005).

There have been a limited number of neuropathological studies looking at frontal and temporal pathology in ALS. Wilson et al. (2001) found changes that where overall more pronounced for cognitively affected patients (ubiquitin positive, alpha-synuclein-negative, and tau-negative neuronal inclusions), most pronounced in the cingulate cortex. Cognitive impairment was uniformly associated with superficial linear spongiosis, a pathologic feature common to several forms of frontotemporal dementia. Wilson and colleagues did not study temporal structures in more detail. In a group of ALS patients with cognitive impairment, and decreased frontal blood flow on SPECT, neuropathologic examination showed spongy degeneration and neuronal loss in the frontal lobe (Abe et al., 1997).

Pathologically, there is also a special tie between semantic dementia, or temporal variant FTD, and ALS. Both FTD-MND (or FTD-ALS) and semantic dementia are characterized by ubiquinated inclusions (FTD-U); the clinical spectrum of patients seen with this histopathological finding varies from ALS, ALS with FTD and semantic dementia without ALS (Davies & Xuereb, 2007). This suggests that patients with ALS caused by this histopathological subtype would be expected to have overt or subtle features of temporal involvement, especially impairments in semantic processing and amygdala function.

Some studies have specifically evaluated the limbic system in sporadic ALS. In the ALS-Parkinson-dementia complex of Guam, tau and alpha-synuclein aggregates are a common

Overview of Cognitive Function in ALS, with Special

*Neuroscience* 4: 165 – 178

*NY Acad. Sci.* 985: 326 – 340

1998 Jun 4;393(6684):470-4.

6;433(7021):68-72.

*Neurology, 62,* 556-562.

pp. 161 - 207

*Research. 12,* 189-198.

and ALS/FTLD*. Neurology* 2005;65:75-8

Attention to the Temporal Lobe: Semantic Fluency and Rating the Approachability of Faces 715

Adolphs R (2003) Cognitive Neuroscience of Human Social Behavior. *Nature Reviews* 

Adolphs R (2003) Is the human amygdala specialized for processing social information? *Ann* 

Adolphs R, Baron-Cohen S & Tranel D (2002) Impaired Recognition of social emotions following amygdala damage. *Journal of Cognitive Neuroscience* 14: 1264 - 1274 Adolphs R, Tranel D & Damasio AR (1998) The human amygdala in social judgment. *Nature.*

Adolphs R, Gosselin F, Buchanan TW, Tranel D, Schyns P & Damasio AR (2005). A

Baldo J., Schwartz S., Wilkins D., & Dronkers N. (2006). Role of frontal versus temporal

Bim, R. M., Kenworthy, L., Case, L., Caravella, R., Jones, T. B., Bandettini, P. A., et al. (2010).

Blair KS, Geraci M, Krelitz K, OteroM, Ernst M, Leibenluft E, Blair RJ & Pine DS (2011) The

Canning, S. J., Leach, L., Struss, D., Ngo, L., & Black, S. E. (2004). Diagnostic utility of

Chang JL, Lomen-Hoerth C, Murphy J, Henry RG, Kramer JH, Miller BL, Gorno-Tempini

Davies RR & Xuereb JH (2007) The histopathology of frontotemporal dementia *in* Hodges JR

Fagundo, A. B., Lopez, S., Romero, M., Guarch, J., Marcos, T., &. Salamero, M. (2008).

Guilford's structure of intellect theory. *Cogn Behav Neurol*; 24(1): 26-34. Folstein M. F., Folstein S. E., & McHugh P. R. (1975). Mini-mental state: A practical method

subgroups. *Clin Neurol Neurosurg*. 1997 May;99(2):79-86.

Frank B, Haas J, Heinze HJ, Stark E & Munte TF (1997) Relation of neuropsychological and

Freitas-Ferrari MC, Hallak JE, Trzesniak C, Filho AS. Machado-de-Sousa JP, Chagas MH,

*Journal of the International Neuropsychological Society, 12,* 896-900.

*Neurolmage, 49,* 1099-1107. doi:10.1016/j.neuroimage.2009.07.036

processing. *American Journal of Psychiatry* Jun 1 Epub ahead of print

mechanism for impaired fear recognition after amygdala damage. *Nature.* 2005 Jan

cortex in verbal fluency as revealed by voxel-based lesion symptom mapping.

Neural systems supporting phonemic search guided by letter and semantic category cues; A self-paced overt response fMRI study of verbal fluency.

pathology of social phobia is independent of developmental changes in faces

abbreviated fluency measures in Alzheimer disease and vascular dementia.

ML. (2005) A voxel-based morphometry study of patterns of brain atrophy in ALS

(ed.) *Frontotemporal Dementia Syndromes*, Cambridge University Press, Cambridge

Clustering and switching in semantic fluency: Predictors of the development of Alzheimer's disease. *International Journal of Geriatric Psychiatry, 23,* 1007-1013. Flaherty-Craig, C., Brothers, A., yang, C., Svoboda, R., Simmons, Z. (2011). Decreases in

problem solving and anosognosia in amyotrophic lateral sclerosis: Applications of

for grading the cognitive state of patients for the clinician. *Journal of Psychiatric* 

magnetic resonance findings in amyotrophic lateral sclerosis: evidence for

Nardi AE, Crippa JA (2010) Neuroimaging in social anxiety disorder: a systematic review of the literature*. Prog Neuropsychopharmacol Biol Psychiatry* 34(4):565-80

finding in the amygdala (e.g., Yamazaki et al., 2000). Case series of sporadic ALS patients with and without dementia have demonstrated ubiquitinated intraneuronal inclusions and spongiform changes in the neostriatum, the amygdala and the parahippocampal gyrus, as well as the temporal pole, anterior cingulate, orbitofrontal cortex and insula (Kawashima et al., 2001; Kato et al., 1994, also Tsuchiya et al., 2002).

We assume that the cognitive findings of decreased semantic fluency and abnormal approachability are the clinical correlate of the changes seen neuroradiologically and neuropathologically and suggest that many ALS patients may, in fact, have both clinically relevant amygdala dysfunction and difficulties with semantic processing.

Performance on the Approachability Paradigm was not related to frontal dysfunction. A similar lack of correlation between amygdala dysfunction and frontal cognitive changes was reported by Zimmerman and colleagues (2007). In their study, among the 8 patients with emotional perceptual impairment, one-half did not have depressive, or memory or cognitive symptoms on screening, whereas the remainder showed dementia symptoms alone or together with depressive symptoms. This finding is important in two ways: First, it suggests the response pattern seen in ALS patients in both studies was in fact due to amygdala involvement as hypothesized and was less likely to be the result of frontal dysfunction

Second, FTLD is known to have several subtypes with variable sites of onset, all of which can be seen in conjunction with ALS. Thus, it is not surprising to find that amygdala dysfunction and frontal dysfunction are not associated. It may be that there are groups of ALS patients that have predominantly temporal dysfunction at onset, while other groups have predominantly frontal onset. This also suggests that more ALS patients have clinical involvement outside the motor strip than the rough estimate of 50% percent from prior studies, which mainly concentrated on frontal cognitive dysfunction.
