Is the logopenic-variant of primary progressive aphasia a unitary disorder?

CORTEX1427_proof ■ 13 April 2015 ■ 1/12 c o r t e x x x x ( 2 0 1 5 ) 1 e1 2

Available online at www.sciencedirect.com 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64

ScienceDirect Journal homepage: www.elsevier.com/locate/cortex

Research report

Q4

Q3

Is the logopenic-variant of primary progressive aphasia a unitary disorder? Cristian E. Leyton a,b,c,*, John R. Hodges b,c,d, Olivier Piguet b,c,d, Catriona A. McLean e, Jillian J. Kril f and Kirrie J. Ballard a,b a

Faculty of Health Sciences, The University of Sydney, Lidcome, NSW, Australia Neuroscience Research Australia, Randwick, NSW, Australia c ARC Centre of Excellence in Cognition and its Disorders, Sydney, NSW, Australia d School of Medical Sciences, The University of New South Wales, Sydney, NSW, Australia e Department of Anatomical Pathology, Alfred Hospital, Melbourne, VIC, Australia f Department of Pathology, Sydney Medical School, The University of Sydney, Sydney, NSW, Australia b

article info

abstract

Article history:

Logopenic progressive aphasia is one of the clinical presentations of primary progressive

Received 11 November 2014

aphasia and formally defined by the co-occurrence of impaired naming and sentence

Reviewed 19 January 2015

repetition. Impaired naming is attributed to failure of lexical retrieval, which is a multi-

Revised 30 January 2015

staged process subserved by anatomically segregated brain regions. By dissecting the

Accepted 17 March 2015

neurocognitive processes involved in impaired naming, we aimed to disentangle the clinical

Action editor Peter Garrard

and neuroanatomical heterogeneity of this syndrome. Twenty-one individuals (66.7% fe-

Published online xxx

males, age range 53e83 years) who fulfilled diagnostic criteria for logopenic variant and had at least two clinical and language assessments, 1 year apart, were recruited and matched for

Keywords:

age, sex distribution and level of education with a healthy control sample (n ¼ 18). All

Primary progressive aphasia

participants underwent a structural brain scan at the first visit and surface-wise statistical

Logopenic variant of primary pro-

analysis using Freesurfer. Seventeen participants with logopenic variant underwent amy-

gressive aphasia

loid imaging, with 14 demonstrating high amyloid retention. Based on their performance on

Alzheimer's disease

single-word comprehension, repetition and confrontation naming, three subgroups of

Anomia

logopenic cases with distinctive linguistic profiles and distribution of atrophy were identified. The first subgroup (n ¼ 10) demonstrated pure anomia and left-sided atrophy in the posterior inferior parietal lobule and lateral temporal cortex. The second subgroup (n ¼ 6), presented additional mild deficits in single-word comprehension, and also exhibited bilateral thinning of the fusiform gyri. The third subgroup (n ¼ 5) showed additional impaired single-word repetition, and cortical thinning focused on the left superior temporal gyrus. The subgroups differed in the proportion of cases with high amyloid retention and in the rate of decline of naming performance over time, suggesting that neurodegeneration spreads differentially throughout regions subserving word processing. In line with previous reports, these results confirm the extensive damage to the language network and, in part, explain the clinical heterogeneity observed across logopenic cases. © 2015 Published by Elsevier Ltd.

* Corresponding author. Faculty of Health Sciences, The University of Sydney, 75 East St. Libcombe, NSW, 2141, Australia. E-mail addresses: [email protected] (C.E. Leyton), [email protected] (J.R. Hodges), [email protected] (O. Piguet), [email protected] (C.A. McLean), [email protected] (J.J. Kril), [email protected] (K.J. Ballard). http://dx.doi.org/10.1016/j.cortex.2015.03.011 0010-9452/© 2015 Published by Elsevier Ltd.

Please cite this article in press as: Leyton, C. E., et al., Is the logopenic-variant of primary progressive aphasia a unitary disorder?, Cortex (2015), http://dx.doi.org/10.1016/j.cortex.2015.03.011

65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129

CORTEX1427_proof ■ 13 April 2015 ■ 2/12

2 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

1.

c o r t e x x x x ( 2 0 1 5 ) 1 e1 2

Introduction

The linguistic profile of the logopenic variant of primary progressive aphasia (lv-PPA) is defined by the co-occurrence of anomia, word-finding difficulties and impaired sentence repetition (Gorno-Tempini et al., 2011). These deficits result from the breakdown of several cognitive processes, including verbal short-term memory (Gorno-Tempini et al., 2008), lexical retrieval (Leyton, Piguet, Savage, Burrell, & Hodges, 2012) and phonological processing (Bonner & Grossman, 2012; Brambati, Ogar, Neuhaus, Miller, & Gorno-Tempini, 2009; Goll et al., 2011; Leyton & Hodges, 2013). Reflecting the involvement of these cognitive processes, the distribution of cortical atrophy in lv-PPA comprises an extensive swathe of the left hemisphere which underpins most of the components of the language network (Leyton et al., 2012; Mesulam, Wieneke, Thompson, Rogalski, & Weintraub, 2012; Rohrer et al., 2010). Although atrophy of the left parietal-temporal junction represents the anatomical signature of lv-PPA, the extent of brain atrophy varies considerably from case to case, pointing to the presence of lv-PPA endophenotypes with slightly different clinical profiles, disease severity and decline over time (Leyton, Ballard, Piguet, & Hodges, 2014; Machulda et al., 2013). Accordingly, although the vast majority of cases with lv-PPA have Alzheimer's disease (Chare et al., 2014; Harris et al., 2013; Mesulam et al., 2014; Rohrer, Rossor, & Warren, 2012), this pathology demonstrates variable neuroanatomical extension which results in heterogeneous clinical presentations (Warren, Fletcher, & Golden, 2012). This clinical and anatomical heterogeneity opens the possibility that specific cognitive processes are predominantly damaged in some cases, but not in others. In this sense, the investigation of confrontation naming provides a suitable paradigm to explore the diversity of cognitive deficits in lvPPA, as this task relies on the integration of separate, albeit interactive, steps requiring multiple cognitive processes anatomically segregated (Damasio, Tranel, Grabowski, Adolphs, & Damasio, 2004; DeLeon et al., 2007). Models of lexical production specify a number of discrete stages (Dell & O'Seaghdha, 1992; Levelt, 2001), each of which can be separately damaged and result in impaired naming. In the first stage, the item to be named should be recognised, for which the integrity and access to semantic representations are crucial. Failure at this stage, as in the semantic variant of PPA, not only results in profound anomia, but also in failure on object recognition and word-comprehension tasks. In the intermediate stage, referred to as lexical retrieval, the semantic representation is linked to its arbitrary phonological word form. In other words, although the item can be recognised, the specific target word is not yet retrieved. At the final, or postlexical, stage the phonological information is temporarily stored in the phonological buffer in order to execute the motor plan of the intended utterance. Consequently, failure at this level can result in marked difficulties with repetition, particularly for multisyllabic words and long sentences (Leyton, Savage, et al., 2014). Given that diagnostic criteria for lv-PPA explicitly exclude significant impairments in single-word comprehension, object knowledge or motor aspects of speech, it can be presumed

that the main mechanism underlying anomia in lv-PPA is impaired retrieval of the phonological form. Nevertheless, given the broad extension of pathological changes in lv-PPA (Leyton et al., 2012; Rohrer et al., 2010; Teichmann et al., 2013) over most of the left-sided language regions involved in semantic and lexical processing (Indefrey & Levelt, 2004; Price, Devlin, Moore, Morton, & Laird, 2005), it is possible that other stages of naming processing are also compromised. On these grounds, the concurrent analysis of performance across a range of single-word tasks and structural neuroimaging measures can not only contribute to decipher the clinical and neuroanatomical heterogeneity of the logopenic syndrome, but also reveal sub-groups with distinctive neurobiological features and prognosis. Furthermore, information related to the preservation and involvement of various linguistic components can provide the rational basis for planning and implementing behavioural interventions (Best et al., 2013). The aim of this study was to ascertain the neurocognitive processes involved in impaired naming in lv-PPA by examining patterns of performance on single-word processing tasks and naming errors. Accordingly, we hypothesised that this set of tasks would allow the identification of coherent clinical subgroups with distinctive patterns of brain atrophy and neurobiological behaviour. As such, a second aim was to observe the clinical progression of lv-PPA over consecutive assessments and infer the proportion of cases with Alzheimer pathology in each group.

2.

Material and methods

2.1.

Participants

Twenty-one lv-PPA participants having at least two separate assessment sessions were recruited between 2007 and 2013 through the FRONTIER frontotemporal dementia clinical research group in Sydney Australia. Participants with limited English proficiency (high proficiency was assumed for those who had English as a second language but had lived and worked in an English speaking country for over 10 years) or with concomitant motor neuron disease, significant extrapyramidal features, past history of stroke, epilepsy, alcoholism, or significant traumatic brain injury were excluded from the study. All participants underwent a complete neurological evaluation, a routine neuropsychological assessment, and structural brain MRI. The clinical diagnosis of lv-PPA was retrospectively conducted at baseline assessment using a clinical protocol previously described (Leyton & Hodges, 2014; Leyton et al., 2011) and based on the current International Consensus recommendations (Gorno-Tempini et al., 2011). As such, PPA cases with anomia, word finding difficulties and impaired sentence repetition in absence of apraxia of speech, frank agrammatism or dissolution of semantic knowledge were included in the study. Cases with impaired naming and mild single-word comprehension in absence of other evidence of semantic involvement, but impaired sentence repetition were also classified as lv-PPA. This profile contrasted with that

Please cite this article in press as: Leyton, C. E., et al., Is the logopenic-variant of primary progressive aphasia a unitary disorder?, Cortex (2015), http://dx.doi.org/10.1016/j.cortex.2015.03.011

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

CORTEX1427_proof ■ 13 April 2015 ■ 3/12 c o r t e x x x x ( 2 0 1 5 ) 1 e1 2

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

of individuals diagnosed with the semantic variant of PPA. These cases not only experienced severe naming deficits, but also displayed marked single-word comprehension deficits accompanied by other evidence of semantic involvement, such as impaired object knowledge and surface dyslexia. Similarly, PPA cases with evidence of apraxia of speech, frank agrammatism or both deficits were diagnosed as non-fluent/ agrammatic variant of PPA (Gorno-Tempini et al., 2011). Seventeen of the 21 lv-PPA participants underwent a PiB-PET scan (Klunk et al., 2004) and those cases with a standardised uptake value ratio for neocortical PiB equal or higher than 1.5 were regarded as having high burden of amyloid brain denoting the presence of Alzheimer pathology (Rowe et al., 2010). Healthy control subjects with no history or clinical evidence of major neurological or psychiatric disorder (n ¼ 13) were recruited from the research group's volunteer panel and matched according to the number of years of formal education and age at the moment of assessment. The study received approval from the South Eastern Sydney and Illawarra Area Health Service and the University of New South Wales human ethics committees.

2.2.

Q1

Cognitive status and functional ability assessment

All participants were administered the Addenbrooke's Cognitive Examination revised (ACE-R) (Mioshi, Dawson, Mitchell, Arnold, & Hodges, 2006) to estimate their current global cognitive ability. This screening cognitive measure comprises items that evaluate attention and orientation, memory, verbal fluency, language, and visuospatial abilities. Additional neuropsychological testing included the auditory-verbal digit span task (forward and backward) from the WAIS-III (Wechsler, 1997), ReyeOsterrieth Complex Figure (Osterrieth, 1945; Rey, 1941), the Trail Making Test A and B (Reitan, 1955). A sentence repetition task selected from the Multilingual Aphasia Examination (Benton & Hamsher, 1989), in which participants have to repeat one-by-one 14 sentences of increasing length from 3 to 18 words, was also administrated. In addition, functional ability was determined using a carerbased questionnaire, the Frontotemporal Dementia Rating Scale (Mioshi, Hsieh, Savage, Hornberger, & Hodges, 2010).

2.3.

Single-word tasks

The Sydney Language Battery (Savage et al., 2013) consists of four single-word tasks using the same set of stimuli: visual confrontation naming, repetition of multisyllabic words, word comprehension, and semantic association. Visual confrontation naming, which is always administrated first, requires the participants to name 30 different colour photographs presented one at a time. These include 12 living and 18 nonliving semantic-category items, matched by word frequency and number of phonemes per word. The word repetition task requires the participant to listen and repeat the same 30 multisyllabic words one after the other. Any phonological error, pausing or re-starts were considered abnormal singleword repetition and coded as phonological or nonphonological errors, respectively. Single-word comprehension (word-picture matching), which is always administered

3

before the semantic association task (see below), is assessed by asking the participant to point to the picture that best matches the word spoken by the examiner in an array or photographs containing the target item and six semantically related or visually similar foils. Finally, the semantic association task (pictureepicture matching) requires the participant to select the picture most closely associated with the target picture from a set of four options. The four options are semantically related to each other, but only one option is semantically related to the target (e.g., strawberry: cream, butter, cheese, milk). While the same pictures of the target items are used in the naming and in semantic association subtests, alternative versions are used in the word comprehension task, so that participants cannot identify the item based on visual memory.

2.4.

Analysis of naming errors

For the naming task, any marked hesitancy, phonological or semantic substitutions as well as omissions or ‘don't know’ responses were scored as errors. The classification of errors was conducted by a rater (C.E.L.) and based on the procedures described in Hodges, Graham, and Patterson (1995). In case of uncertain classification, errors were reviewed and reached consensus by second raters (K.J.B. and J.R.H.) Phonological errors and marked fragmentation of the utterance were coded as “sub-lexical” naming errors. Co-ordinate substitutions (e.g., rhinoceros for hippopotamus) and superordinate substitutions (e.g., animal for giraffe) were considered “semantic” errors. Approximate responses, circumlocutions or coherent explanations related to the item to be named, but not part of the aforementioned semantic errors, were considered as “lexical retrieval” errors. Examples of these errors include expressions like ‘pyramids and all that’ referring to hieroglyphic; ‘prehistoric animal’ or ‘they don't exist any more’ to dinosaur; and ‘something for cooking, you can eat’ referring to potatoes. Abandoned testing, empty or unrelated responses (e.g., ‘I've seen it before’, ‘I know it’); and incomplete, incomprehensible, or ‘I don't know’ responses were considered as ‘non-responses’.

2.5.

Statistical analysis

Statistical analyses were undertaken using SPSS 20.0 (IBM Corporation).

2.5.1.

Baseline performances

Given the non-normal distribution of scores on several tasks in the healthy control group due to ceiling effects, nonparametric ManneWhitney U tests were used to compare performances of continuous variables between lv-PPA and healthy controls. Chi square test was used to estimate differences in the distribution of categorical variables.

2.5.2.

Cluster analysis

A two-step cluster analysis was conducted to identify lv-PPA sub-groups based on their performance on single-word tasks. To select the single-word tasks to be entered to the cluster analysis, a preliminary analysis of covariance was conducted using naming performance as the dependent

Please cite this article in press as: Leyton, C. E., et al., Is the logopenic-variant of primary progressive aphasia a unitary disorder?, Cortex (2015), http://dx.doi.org/10.1016/j.cortex.2015.03.011

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

CORTEX1427_proof ■ 13 April 2015 ■ 4/12

4 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

c o r t e x x x x ( 2 0 1 5 ) 1 e1 2

variable and all the other three single-word tasks as independent variables. This analysis revealed that single-word comprehension (F(1,17) ¼ 7.5, p ¼ .014) and single-word repetition (F(1,17) ¼ 4.7, p ¼ .045), but not the semantic association task (F(1,17) ¼ 2.0, p ¼ .17), were significantly associated with naming performance. As a result, naming, single-word repetition and single-word comprehension performances were entered to an exploratory hierarchical cluster method to determine a priori the number of partitions in the sample. In this analysis, squared Euclidean distances were used as a proximity measure, and the Ward's method used as a clustering algorithm. The number of clusters to be used was determined by visual inspection of the dendrogram and then used as input for the k-means analysis. Given the iterative nature of this clustering method, this outcome was matched with the initial clustering solution and used as confirmatory method (Lange, Iverson, Senior, & Chelune, 2002). One-way analysis of variance (ANOVA) tests were carried out to estimate differences across resultant clusters for the relevant continuous variables. In addition, three separate repeatedmeasures ANOVAs were conducted to analyse naming performances on specific semantic-category, to compare frequency of type of naming errors, and to compare type of errors in single-word repetition across clusters. A non-significant test of Mauchly was assumed as sphericity; in case of violation of this assumption, degrees of freedom were adjusted using the Greenhouse-Geisser method. Overall significant differences in the models were further explored using post hoc pair-wise comparisons with Bonferroni corrections. The level of statistical significant was set at .05 unless stated otherwise.

2.5.3.

Longitudinal changes

To examine the rate of decline over time in each resultant cluster, performances on total ACE-R and each of the four single-word tasks were selected as outcome variables. Linear mixed-effect models (Laird & Ware, 1982) were used to evaluate performances on the selected outcome variables over time. Fixed effects included the clustering group, followup time, and the interaction of both effects. Patient's individual variability at baseline was the only random effect included (where we used a random intercept model). The variability of any estimated parameters was determined by the fixed and random components in the model. A significant effect of follow-up time would indicate that the outcome variable changes linearly with time, while a significant interaction between clustering group and follow-up time would indicate that the rate of change of the outcome variable differed across clustering cohorts. Residual errors of the models were assumed to be normally distributed, as were the random intercepts for the subjects' baseline response. All participants were assumed to be independent.

2.6. Imaging acquisition and cortical thickness calculation Whole-brain T1-weighted images were acquired for all participants at the baseline assessment using a 3T Philips MRI scanner with standard quadrature head coil (eight channels). The 3D T1-weighted images were acquired as follows: coronal

orientation, matrix 256  256, 200 slices, 1 mm in-plane resolution, slice thickness 1 mm, echo time/repetition time ¼ 2.6/ 5.8 msec, flip angle a ¼ 19 . To estimate cortical thickness, T1 images were processed using Freesurfer version 5.1 (http://surfer.nmr.mgh.harvard. edu/) (Fischl & Dale, 2000), following the pipeline described elsewhere (Leyton, Ballard, et al., 2014). The identification of cortical areas was conducted using automatic parcellation of the cerebral cortex (Destrieux, Fischl, Dale, & Halgren, 2010). The calculated cortical thickness was smoothed with a 20 mm full-width at half height Gaussian kernel. This level of blurring kernel was chosen to reduce the impact of imperfect alignment between cortices and thereby improving the signal-tonoise ratio (Lerch & Evans, 2005). Imaging statistical analysis was performed vertex-byvertex using general linear models to examine differences between each resultant lv-PPA cluster and the healthy control sample. A false discovery rate (Genovese, Lazar, & Nichols, 2002) was set at .001 to adjust p values for multiple comparisons. p values for group comparison analyses were mapped onto the inflated cortical surface representation of an average brain. Effect sizes for each cluster were also calculated according to the formula by Cohen, in line with previous studies (Hilti et al., 2013).

3.

Results

3.1. Demographic, cognitive and single-word tasks at baseline The average age of the lv-PPA cohort at baseline assessment was 66.9 ± 7.6 years, with estimated symptom duration of 3.5 ± 2.2 years, and 13.2 ± 3.6 years of formal education. The control sample was matched in age (67.7 ± 4.4; t(37) ¼ .4, p ¼ .69), education (12.7 ± 2.4; t(37) ¼ .5, p ¼ .61) and sex distribution (both samples had 66.7% of females). The lv-PPA sample performed lower than the healthy control sample on all cognitive tests and on total scores for each of the four single-word tasks (Table 1). An analysis of naming performance for living versus nonliving items revealed that lv-PPA named living items (53.6%) better than nonliving items (44.2%) (two-tailed, paired t(20) ¼ 3.1, p ¼ .006). Naming error analysis revealed that almost half (48.2%) of all incorrect responses on picture naming were omissions or abandoned test, followed by semantic errors (25.9%), lexical retrieval breakdown (14.0%) and post-lexical impairments (11.9%).

3.2.

Cluster analysis

The resultant dendrogram demonstrated three clusters made up of 10, 6 and 5 cases (Fig. 1). The K-means clustering analysis demonstrated complete congruence with the initial 3-cluster solution. The analyses of variance revealed no differences in demographic features or length of symptoms across cluster (Table 2). Except for ACE-R performance, which was higher in Cluster 1, none of the other cognitive tasks were significantly different across clusters. All clusters had low performance on sentence repetition, but post hoc non-

Please cite this article in press as: Leyton, C. E., et al., Is the logopenic-variant of primary progressive aphasia a unitary disorder?, Cortex (2015), http://dx.doi.org/10.1016/j.cortex.2015.03.011

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

CORTEX1427_proof ■ 13 April 2015 ■ 5/12

5

c o r t e x x x x ( 2 0 1 5 ) 1 e1 2

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Table 1 e Neuropsychological performance in healthy controls and lv-PPA patients. lv-PPA (n ¼ 21)

Healthy controls (n ¼ 18)

ACE-R (/100) Digit span Forwards Digit span Backwards RCF Copy (/36) RCF Recall (/36) TMT A (sec) TMT B (sec) Single-word tasks Naming (/30) Comprehension (/30) Repetition (/30) Semantic association (/30)

ManneWhitney U

Mean

SD

Range

Mean

SD

Range

z

p

96.7 11.9 8.3 32.7 18.7 28.8 73.0

±2.5 ±2.0 ±2.4 ±3.1 ±6.7 ±9.0 ±23.2

[91e100] [8e14] [6e13] [25e36] [7e35] [13e50] [36e112]

65.6 6.3 3.2 24.4 6.6 73.0 260.9

±11.1 ±2.2 ±1.4 ±9.1 ±5.4 ±64.9 ±141.4

[44e91] [2e10] [0e6] [5.5e35] [0e18.5] [24e300] [77e450]

5.32 5.05 5.29 3.64 4.41 3.68 4.73

<.001 <.001 <.001 <.001 <.001 <.001 <.001

27.8 29.2 29.9 28.2

±1.9 ±1.3 ±0.2 ±1.4

[23e30] [25e30] [29e30] [25e30]

14.4 26 25.1 24.6

±7.2 ±2.6 ±5.8 ±3.6

[2e26] [21e30] [11e30] [16e30]

5.11 4.27 4.80 3.98

<.001 <.001 <.001 <.001

ACE-R: Addenbrooke's Cognitive Evaluation-Revised; SD: standard deviation; RCF: Rey Complex Figure; TMT: Trail Making Test.

corrected contrasts demonstrated that Cluster 3 had the worst performance. The analyses of single-word tasks demonstrated distinct profiles across clusters. Cluster 1, in addition to impaired sentence repetition, showed a reduced performance (<80% of maximal score) on naming. In contrast, Cluster 2 exhibited additional reduced performance on single-word comprehension and semantic association, whilst Cluster 3 showed impaired performance on naming and single-word repetition only. In all clusters, a specific semantic-category effect was present on naming (F(1,18) ¼ 8.4, p ¼ .01) reflecting, as indicated above, better performance on naming living items over non-living semantic-category items. No interaction between clusters and naming performance on semantic-category items was found (F(2,18) ¼ .6,

p ¼ .55). As shown in Fig. 2A, ‘non-responses’ was the most common type of error in the whole lv-PPA sample (F(1.8,31.7) ¼ 11.2, p < .001). The interaction between the type of naming error and clusters showed a significant effect (F(3.5,31.7) ¼ 3.9, p ¼ .013). Post hoc comparisons revealed that the ‘sub-lexical’ type of errors occurred more frequently in Cluster 3 than in the other clusters (p < .001, significance not shown in Fig. 2A), whereas within Cluster 2 non-responses occurred more frequently. No other differences were detected between or within clusters. The analysis of single-word repetition (Fig. 2B) showed that phonological errors were the most frequent type of error in the whole lv-PPA sample (F(1,18) ¼ 19.2, p < .001). The interaction effect between the type of error and clusters was significant (F(2,18) ¼ 8.5,

Fig. 1 e Dendrogram displays the arrangement obtained by the hierarchical cluster analysis (Ward's method) of participants with logopenic aphasia. Clusters are demarked with brackets and cases are identified with numbers placed at the terminal leaves of the dendrogram. The red vertical line indicates the cut-point that divides the sample into 3 clusters. The x-axis represents the clustering distance rescaled. Please cite this article in press as: Leyton, C. E., et al., Is the logopenic-variant of primary progressive aphasia a unitary disorder?, Cortex (2015), http://dx.doi.org/10.1016/j.cortex.2015.03.011

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

CORTEX1427_proof ■ 13 April 2015 ■ 6/12

6 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

c o r t e x x x x ( 2 0 1 5 ) 1 e1 2

Table 2 e Demographic, cognitive and single-word task performance across clusters. Cluster 1 Age, yrs Education, yrs Symptom duration, yrs Follow-up time, yrs Number of visits ACE-Ra/100 FRS Rasch score Digit span Forwards Digit span Backwards Sentence Repetition (/14) RCF Copy/36 RCF Recall/36 TMT A (sec) TMT B (sec) Single-word tasks Naminga (/30) Living (%) Nonliving (%) Comprehensionb (/30) Repetitionc (/30) Semantic association (/30)

Cluster 2

Cluster 3

65.4 13.8 3.2 1.6 2 73.3 1.3 7.0 3.8 5.1 26.0 8.5 79.6 241.7

± 7.9 ± 3.4 ± 1.5 ± 0.7 ±2 ± 7.3 ± 1.7 ± 2.3 ± 1.2 ± 2.4 ± 9.8 ± 6.3 ± 80.9 ± 150.6

67.8 14.4 2.7 1.9 3 58.5 1.7 6.7 3.0 5.0 21.7 4.4 88.3 315.7

± 8.8 ± 4.5 ± 2.0 ± 0.8 ±2 ± 3.6 ± 1.1 ± 0.8 ± 0.9 ± 1.0 ± 10.3 ± 5.0 ± 59.5 ± 156.2

68.8 ± 6.3 10.4 ± 1.5 5.1 ± 3.1 1.6 ± 0.6 2±1 58.8 ± 14.1 2.5 ± 2.2 4.4 ± 2.3 2.4 ± 1.8 2.4 ± 1.1 24.4 ± 7.0 5.5 ± 3.3 41.6 ± 15.2 233.4 ± 109.9

20.2 72.5 63.9 27.9 28.0 26.0

± 4.3 ± 17.6 ± 16.2 ± 1.2 ± 2.3 ± 2.2

8.7 37.5 21.2 23.3 28.0 22.5

± 4.9 ± 18.1 ± 15.9 ± 2.0 ± 1.8 ± 4.4

9.6 ± 4.7 35.0 ± 14.9 30.0 ± 17.3 25.6 ± 2.3 15.8 ± 3.0 24.2 ± 4.3

p .69 .14 .16 .73 .52d .004 .47 .07 .16 .044e .68 .33 .43 .55 <.001

<.001 <.001 .17

Scores are means ± SD, except for Number of visits, where it is mode ± range. ACE-R: Addenbrooke's Cognitive Evaluation-Revised; FRS: Frontotemporal Dementia Rating Scale; RCF: Rey Complex Figure; TMT: Trail Making Test. a Cluster 1 > other clusters (post hoc test with Bonferonni correction). b Cluster 2 < Cluster 1 (post hoc test with Bonferonni correction). c Cluster 3 < other clusters. d KruskaleWallis test. e Pair-wise non-corrected contrasts: Cluster 3 vs Cluster 1, p ¼ .019 and Cluster 3 vs Cluster 2, p ¼ .058.

p ¼ .003). Post hoc contrasts showed that phonological errors were more common in Cluster 3 than in other clusters (p < .001).

3.3. Clinical evolution and neurobiological characterisation Clinical assessments conducted over follow-up visits demonstrated relentless worsening of language function, accompanied by overall cognitive deterioration in almost all cases. Three cases developed non-cognitive motor deficits (Table 3). Two participants from Cluster 1 exhibited myoclonic jerks, while one case from Cluster 3 with a negative PiB-PET scan developed poor balance and marked parkinsonism, warranting the diagnosis of cortical basal syndrome. During the follow-up, two cases died and underwent

neuropathological examination. One of the participants (P1) who developed myoclonic jerks in Cluster 1 showed Alzheimer pathology combined with Lewy body disease, while the other case (P2), who had not undergone amyloid-b imaging in Cluster 2, demonstrated Alzheimer pathology (summary of pathology reports and plot of SUVR across clusters in Supplementary Material).

3.4.

Longitudinal assessment

A preliminary analysis of covariance between clusters revealed no effect of age, education or length of disease at presentation on baseline performances of the outcome variables (Supplementary Table 1). Performance of the lv-PPA cohort on all outcome measures declined over time; however, not all participants declined at the same pace. Participants

Fig. 2 e Distribution of type of naming (A) and repetition (B) errors across lv-PPA clusters. Asterisks indicate significant difference in respect to other types of errors in the cluster. Please cite this article in press as: Leyton, C. E., et al., Is the logopenic-variant of primary progressive aphasia a unitary disorder?, Cortex (2015), http://dx.doi.org/10.1016/j.cortex.2015.03.011

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

CORTEX1427_proof ■ 13 April 2015 ■ 7/12

7

c o r t e x x x x ( 2 0 1 5 ) 1 e1 2

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Table 3 e Neurobiological characterisation of clusters and relevant non-cognitive deficits occurring at follow-up. Cluster

n

PiB neg/PiB pos

Non-cognitive deficits

Pathological examination

Cluster 1 Cluster 2 Cluster 3 Total

10 6 5 21

1/9 1/3 1/2 3/14

2 Myoclonic jerks e 1 Parkinsonism 3

1 AD þ LB 1 AD 0 2

AD þ LB: Alzheimer pathology combined with Lewy body inclusions. More details in Supplementary Material.

in Cluster 1 tended to decline more rapidly than participants from other clusters in ACE-R and demonstrated a significantly more rapid decline in naming. (Table 4 and Fig. 3).

3.5.

Cortical thickness analyses

The combined lv-PPA sample showed the expected pattern of cortical changes compared with healthy controls, characterised by widespread cortical thinning spanning the left temporal lobe, posterior cingulate and the inferior parietal cortices (Supplementary Material Fig. 1 and Table 2). Analyses conducted on each cluster separately revealed overlapping, but distinct, patterns of cortical thinning across clusters (Fig. 4). Cluster 1 showed left-sided cortical thinning focused on the inferior parietal lobe, including the supramarginal and angular gyri, as well as the superior temporal gyrus and part of the inferior temporal gyrus. Cluster 2 demonstrated, in addition, more severe and extensive thinning in the left inferiormedial aspects of temporal lobe cortex, temporal pole and the right fusiform cortex. Cluster 3 showed the least overall cortical thinning involving mainly in the left superior temporal gyrus, including the planum temporale and inferior insula.

4.

Discussion

Our findings confirm clinical heterogeneity in lv-PPA as it is reflected in the varying neuroanatomical extent of neurodegeneration seen across clinical subgroups. Although all cases showed anomia, impaired sentence repetition, and atrophy involving the left posterior temporal lobe and inferior parietal area, three distinct profiles of performance on singleword tasks and extent of cortical thinning were identified. Cluster 1, which comprised the largest proportion of the lvPPA cohort, displayed the typical naming profile of logopenic syndromedcharacterised by pure anomia, circumscribed left temporo-parietal atrophy and, importantly, as their PiB-PET

positive scan suggested, Alzheimer pathology in almost all cases. Cluster 2 and Cluster 3 exhibited more global cognitive impairments and deficits in single-word repetition and comprehension, suggesting in addition impairments at the level of phonological output and semantic processing, respectively. The involvement across multiple language processes in lv-PPA has been suggested in previous studies. Deficits of phonological encoding is suggested by the abnormal repetition of multisyllabic words and non-words and by the presence of phonological errors in cases with lv-PPA (Leyton, Ballard, et al., 2014; Leyton, Savage, et al., 2014; Petroi, Duffy, Strand, & Josephs, 2014), while the semantic involvement is suggested by the progressive impairment in single-word comprehension (Leyton, Hsieh, Mioshi, & Hodges, 2013). Although those deficits might be explained by the inclusion of cases with more advanced level of disease, performance differences across clusters are unlikely to be exclusively attributed to a disease-severity effect, as the clusters demonstrated no differences in age, education, estimated length of disease or level of functional status and had equivalent period of follow-ups. Moreover, rather than a diffuse extension, each cluster demonstrated a discrete anatomical extension of atrophy that conforms with the distribution reported by Machulda et al. (2013). In their mild typical logopenic phenotype, characterised by atrophy involving left temporoparietal cortex, the distribution of atrophy is equivalent to our Cluster 1; their severe typical logopenic group with extensive bilateral atrophy is similar to our Cluster 2, while their mild atypical group with atrophy confined to the left superior temporal cortex is homologous to our Cluster 3. Differences in the proportion of PiB positive cases across the three clusters suggest that each cluster displays different likelihood of Alzheimer pathology. This presumption, however, needs to be taken with cautious. Indeed, the cut-off SUVR value of 1.5 separating negative from positive cases is arbitrary. As the SUVR plot shows (Supplementary Material Fig. 1), PiB negative cases Clusters 1 and 2 showed higher values than the negative case in Cluster 3. The former cases

Table 4 e Estimated changes of outcome measures over time across clusters. Outcome measure

ACE-R Confrontation Naming Single-word repetition Single-word comprehension Semantic association

Main effect follow-up time

Cluster group and follow-up time interaction

Estimated Annual decline Cluster 1

Cluster 2

Cluster 3

F

p

F

p

Coefficient

Coefficient

Coefficient

90.9 31.3 21.5 13.2 15.4

<.001 <.001 <.001 .001 <.001

3.1 10.5 0.3 1.7 0.5

.055 <.001 n.s. n.s. n.s.

17.8 5.0 2.2 2.9 2.3

13.3 2.3 2.3 1.2 2.8

9.3 0.1 3.2 1.2 4.3

Please cite this article in press as: Leyton, C. E., et al., Is the logopenic-variant of primary progressive aphasia a unitary disorder?, Cortex (2015), http://dx.doi.org/10.1016/j.cortex.2015.03.011

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

CORTEX1427_proof ■ 13 April 2015 ■ 8/12

8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

c o r t e x x x x ( 2 0 1 5 ) 1 e1 2

Fig. 3 e Modelled ACE-R and single-word task performances over time for each cluster.

displayed at baseline assessment and on subsequent assessments indistinguishable clinical features from their cohort, whereas the case with low PiB burden in Cluster 3 developed extrapyramidal motor deficits on subsequent assessments, warranting the alternative diagnosis of cortical basal syndrome. All participants demonstrated relentless decline over time in all tasks, reflecting the progressive nature of the neurodegenerative process. Nevertheless, the degree of decline varied across clusters: lv-PPA patients from Cluster 1 declined more rapidly on both measures of global cognition and naming than the other clusters. This differential rate of decline suggests not

only a more accelerated destruction of the language network in Cluster 1 but also a differential neuroanatomical spread of neurodegeneration across clusters. Although current evidence suggests a rather eccentric neuroanatomical spread, which is initiated in the left temporo-parietal junction and spread out to other language and non-language areas (Rohrer et al., 2013), the pattern of decline and distribution of brain atrophy observed across clusters suggests a diverging spread into cortical regions subserving semantics (Cluster 2) and phonological processing (Cluster 3). Yet, the low performance at baseline of Clusters 2 and 3 raises concerns that floor effects may have driven the longitudinal trends. This possibility,

Please cite this article in press as: Leyton, C. E., et al., Is the logopenic-variant of primary progressive aphasia a unitary disorder?, Cortex (2015), http://dx.doi.org/10.1016/j.cortex.2015.03.011

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

CORTEX1427_proof ■ 13 April 2015 ■ 9/12 c o r t e x x x x ( 2 0 1 5 ) 1 e1 2

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

9

Fig. 4 e Distribution of cortical thinning for each of the lv-PPA clusters. Level of significant set at false discovery rate <.001. Statistical information and effect sized are in Supplementary Material 2.

however, seems less likely if we consider that ACE-R has a broad scaling range which minimises ceiling and floor effects and is capable of tracking progression in PPA (Leyton, Hornberger, Mioshi, & Hodges, 2010). Moreover, the presence of phenotypes with dissimilar prognosis was demonstrated by Machulda et al. (2013) in their mild atypical logopenic group, which despite having a longer disease duration, showed atrophy confined to the superior temporal gyrus. The findings raise the question of what should be considered as the core criteria for lv-PPA and how to classify cases in Clusters 2 and 3. Application of the lv-PPA label is undoubtedly justified for Cluster 1 cases. Cases in Cluster 2, who showed additional single-word comprehension deficits, yet were not classified as semantic variant of PPA since their semantic impairment was mild without loss of object knowledge. Moreover these patients did not show the marked anterior temporal lobe atrophy characteristic of semantic variant (Davies et al., 2005) and had marked impairment of sentence repetition which is not a feature of semantic cases. It is well recognised that semantic deficits occur in Alzheimer's disease (Hodges & Patterson, 1995). It is of interest that patients in Cluster 2 had extensive cortical involvement in regions not ‘typically’ described in lv-PPA. The semantic deficits can be accounted for damage to the inferior and medial aspects of the temporal lobes, in line with reports that emphasise the role of fusiform gyrus in semantic processing (Mion et al., 2010). Such cases can either be considered within an extended lv-PPA phenotype or alternatively as mixed or unclassifiable cases. Conversely, cases in Cluster 3 had additional deficits in single-word repetition implying impairment in post-lexical processing and creating a potential overlap with nonfluent/ agrammatic PPA variant, although these patients did not show either apraxia of speech or agrammatism which are the

core deficits diagnostic of the nonfluent/agrammatic PPA syndrome (Gorno-Tempini et al., 2011). In this cluster, most errors were phonological errors, suggesting that, in addition to difficulties in retrieving the phonological form of words, these patients experience marked deficits at the level of phonological output processing (i.e., the stage prior to the motor plan execution, which also includes the capacity of buffering phonological representations). Accordingly, converging imaging and clinical evidence suggests that breakdown at the level of phonological integration stage of word production may be the main cause of impaired single-word repetition in lv-PPA (Leyton, Savage, et al., 2014). This contention is also supported by the existence of phonological errors, often observed in connected speech of lv-PPA cases (Leyton, Ballard, et al., 2014; Petroi et al., 2014), and impairment in verbal shortterm memory, which are the other cardinal cognitive processes involved in lv-PPA, and also explains the marked impairment in sentence repetition in this cluster (Leyton & Hodges, 2013). Accordingly, the anatomical correlate of these deficits is likely to be the left posterior superior temporal gyrus (Leyton, Ballard, et al., 2014), although it is possible that the conjoint involvement, either functional or neuroanatomical, of the connections to the articulatory areas may have also impacted the performance on single-word repetition. The pattern of performance on naming revealed a category-specific semantic impairment characterised by a relative advantage in naming natural items over manufactured or nonliving items, a pattern that contrasts to that described in typical Alzheimer's disease (Garrard et al., 2001; Grossman et al., 2013) and semantic variant PPA cases (Libon et al., 2013). Given the relevance of visuoperceptual features to the semantic representation of living items, loss of this category-specific semantic is associated with damage to

Please cite this article in press as: Leyton, C. E., et al., Is the logopenic-variant of primary progressive aphasia a unitary disorder?, Cortex (2015), http://dx.doi.org/10.1016/j.cortex.2015.03.011

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

CORTEX1427_proof ■ 13 April 2015 ■ 10/12

10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

c o r t e x x x x ( 2 0 1 5 ) 1 e1 2

ventral and inferolateral temporal cortices. Conversely, the semantic representation of nonliving items may rely upon perceptual and/or motor/action features, and engage, therefore, heteromodal cortices such as the left temporo-parietal junction (Humphreys & Forde, 2001; Martin, 2007; Patterson, Nestor, & Rogers, 2007). The contrasting pattern of categoryspecific semantic performance can be explained by the distribution of brain atrophy observed in lv-PPA compared with typical Alzheimer's disease or semantic variant PPA: while lvPPA cases have atrophy focused on the left temporal and parietal cortices, Alzheimer's disease and semantic variant PPA have maximal atrophy involving inferior and medial aspects of temporal lobes (Grossman et al., 2013; Libon et al., 2013). One potential weakness of the current study is the application of cluster analysis and the relatively small number of cases. While cluster analysis is not an inferential statistical method, this procedure allowed the systematic classification of lv-PPA cases and provided a rigorous framework to demonstrated our empirical observation that some lv-PPA cases seemed to be worse on repetition, whereas others were much more anomic. This other issue is that of comparing effects between unbalanced sub-groups, which can undermine the statistical power and limit the generalisation of results. The estimated sample size for each cluster, however, was comparable (see Supplementary Material 2) and, as was mentioned, the cluster partition, which is data driven, yielded coherent clinical and neuroanatomical phenotypes that were consistent with a previous report (Machulda et al., 2013). In addition, longitudinal analyses revealed differences in the pattern of deficits across clusters, which suggest differences in the extent of neurodegeneration. Undoubtedly, replication in larger samples is warranted to generalise our findings. Nevertheless, the heterogeneity of the clinical features displayed by lv-PPA cases raises the issue of diagnosis: Should lv-PPA be limited to cases falling into Cluster 1, or should it be applied broadly to encompass clusters 2 and/or 3?

5.

Conclusions

The analysis of single-word processing demonstrated an extensive involvement across several stages of word production in lv-PPA. This involvement varied across cases, suggesting that different neurobiological factors, some not yet fully understood, may play a role in the clinical diversity. Although in most cases the underlying pathology is presumably Alzheimer's disease, a few cases evolved with atypical features that prompted the consideration of an alternative diagnosis. Alternatively, the diverging neuroanatomical involvement may reflect the pathological spreading of same pathology detected at multiple stages. Further research is needed to replicate our findings and to better characterise the progression of this syndrome.

Funding This work was supported by a National Health and Medical Research Council (NHMRC) project grant (630489); an NHMRC

program grant (APP1037746); the Australian Research Council (ARC) Centre of Excellence in Cognition and its Disorders (CE110001021); a Sydney University Postdoctoral Fellowship (to C.E.L.); an ARC Future Fellowship (FT120100355 to K.J.B.); an ARC Federation Fellowship (FF0776229 to J.R.H.); and an NHMRC Career Development Fellowship (APP1022684 to O.P.). Q2

Acknowledgements The authors are grateful to the participants and their families for supporting our research. The authors also thank Prof Rowe and his team for conducting the PiB scanning. We would like to thank the participants in the FRONTIER brain donor program and Lauren Bartley for coordinating this research program. Human brain tissue was collected and characterised through the Australian Brain Bank Network, specifically by the Sydney Brain Bank (supported by Neuroscience Research Australia and the University of New South Wales), and by the Victorian Brain Bank Network (supported by Neurosciences Australia, the University of Melbourne, the Mental Health Research Institute of Victoria, the Alfred Hospital, and the Victorian Forensic Institute of Medicine).

Supplementary data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.cortex.2015.03.011.

references

Benton, A. L., & Hamsher, K. (1989). Multilingual aphasia examination. Manual of instructions (2nd. ed.). Iowa City: AJA Associates. Best, W., Greenwood, A., Grassly, J., Herbert, R., Hickin, J., & Howard, D. (2013). Aphasia rehabilitation: does generalisation from anomia therapy occur and is it predictable? A case series study. Cortex, 49(9), 2345e2357. http://dx.doi.org/10.1016/ j.cortex.2013.01.005. Bonner, M. F., & Grossman, M. (2012). Gray matter density of auditory association cortex relates to knowledge of sound concepts in primary progressive aphasia. Journal of Neuroscience, 32(23), 7986e7991. http://dx.doi.org/10.1523/ JNEUROSCI.6241-11.2012, 32/23/7986 [pii]. Brambati, S. M., Ogar, J., Neuhaus, J., Miller, B. L., & GornoTempini, M. L. (2009). Reading disorders in primary progressive aphasia: a behavioral and neuroimaging study. Neuropsychologia, 47(8e9), 1893e1900. http://dx.doi.org/ 10.1016/j.neuropsychologia.2009.02.033. S0028-3932(09)001110 [pii]. Chare, L., Hodges, J. R., Leyton, C. E., McGinley, C., Tan, R. H., Kril, J. J., et al. (2014). New criteria for frontotemporal dementia syndromes: clinical and pathological diagnostic implications. Journal of Neurology Neurosurgery and Psychiatry, 85(8), 865e870. http://dx.doi.org/10.1136/jnnp-2013-306948. Damasio, H., Tranel, D., Grabowski, T., Adolphs, R., & Damasio, A. (2004). Neural systems behind word and concept retrieval. Cognition, 92(1e2), 179e229. Davies, R. R., Hodges, J. R., Kril, J. J., Patterson, K., Halliday, G. M., & Xuereb, J. H. (2005). The pathological basis of semantic

Please cite this article in press as: Leyton, C. E., et al., Is the logopenic-variant of primary progressive aphasia a unitary disorder?, Cortex (2015), http://dx.doi.org/10.1016/j.cortex.2015.03.011

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

CORTEX1427_proof ■ 13 April 2015 ■ 11/12 c o r t e x x x x ( 2 0 1 5 ) 1 e1 2

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

dementia. Brain, 128(Pt 9), 1984e1995. http://dx.doi.org/ 10.1093/brain/awh582. awh582 [pii]. DeLeon, J., Gottesman, R. F., Kleinman, J. T., Newhart, M., Davis, C., Heidler-Gary, J., et al. (2007). Neural regions essential for distinct cognitive processes underlying picture naming. Brain, 130(Pt 5), 1408e1422. http://dx.doi.org/10.1093/brain/ awm011. awm011 [pii]. Dell, G. S., & O'Seaghdha, P. G. (1992). Stages of lexical access in language production. Cognition, 42(1e3), 287e314. Destrieux, C., Fischl, B., Dale, A., & Halgren, E. (2010). Automatic parcellation of human cortical gyri and sulci using standard anatomical nomenclature. NeuroImage, 53(1), 1e15. http:// dx.doi.org/10.1016/j.neuroimage.2010.06.010. S1053-8119(10) 00854-2 [pii]. Fischl, B., & Dale, A. M. (2000). Measuring the thickness of the human cerebral cortex from magnetic resonance images. Proceding of the National Academy of Sciences United States of America, 97(20), 11050e11055. http://dx.doi.org/10.1073/ pnas.200033797, 200033797 [pii]. Garrard, P., Lambon Ralph, M. A., Watson, P. C., Powis, J., Patterson, K., & Hodges, J. R. (2001). Longitudinal profiles of semantic impairment for living and nonliving concepts in dementia of Alzheimer's type. Journal of Cognitive Neuroscience, 13(7), 892e909. http://dx.doi.org/10.1162/089892901753165818. Genovese, C. R., Lazar, N. A., & Nichols, T. (2002). Thresholding of statistical maps in functional neuroimaging using the false discovery rate. NeuroImage, 15(4), 870e878. Goll, J. C., Kim, L. G., Hailstone, J. C., Lehmann, M., Buckley, A., Crutch, S. J., et al. (2011). Auditory object cognition in dementia. Neuropsychologia, 49(9), 2755e2765. http:// dx.doi.org/10.1016/j.neuropsychologia.2011.06.004. S00283932(11)00278-8 [pii]. Gorno-Tempini, M. L., Brambati, S. M., Ginex, V., Ogar, J., Dronkers, N. F., Marcone, A., et al. (2008). The logopenic/ phonological variant of primary progressive aphasia. Neurology, 71(16), 1227e1234. http://dx.doi.org/10.1212/ 01.wnl.0000320506.79811.da, 01.wnl.0000320506.79811.da [pii]. Gorno-Tempini, M. L., Hillis, A. E., Weintraub, S., Kertesz, A., Mendez, M., Cappa, S. F., et al. (2011). Classification of primary progressive aphasia and its variants. Neurology, 76(11), 1006e1014. http://dx.doi.org/10.1212/WNL.0b013e31821103e6. WNL.0b013e31821103e6 [pii]. Grossman, M., Peelle, J. E., Smith, E. E., McMillan, C. T., Cook, P., Powers, J., et al. (2013). Category-specific semantic memory: converging evidence from bold fMRI and Alzheimer's disease. NeuroImage, 68, 263e274. http://dx.doi.org/10.1016/ j.neuroimage.2012.11.057. Harris, J. M., Gall, C., Thompson, J. C., Richardson, A. M., Neary, D., du Plessis, D., et al. (2013). Classification and pathology of primary progressive aphasia. Neurology. http://dx.doi.org/ 10.1212/01.wnl.0000436070.28137.7b. € nggi, J., Vitacco, D. A., Kraemer, B., Palla, A., Hilti, L. M., Ha Luechinger, R., et al. (2013). The desire for healthy limb amputation: structural brain correlates and clinical features of xenomelia. Brain, 136(1), 318e329. http://dx.doi.org/10.1093/ brain/aws316. Hodges, J. R., Graham, N., & Patterson, K. (1995). Charting the progression in semantic dementia: Implications for the organisation of semantic memory. Memory, 3(3/4), 463e495. Hodges, J. R., & Patterson, K. (1995). Is semantic memory consistently impaired early in the course of Alzheimer's disease? Neuroanatomical and diagnostic implications. Neuropsychologia, 33(4), 441e459. doi: 0028-3932(94)00127-B [pii]. Humphreys, G. W., & Forde, E. M. (2001). Hierarchies, similarity, and interactivity in object recognition: “category-specific” neuropsychological deficits. Behavioral and Brain Sciences, 24(3), 453e476. discussion 476e509.

11

Indefrey, P., & Levelt, W. J. M. (2004). The spatial and temporal signatures of word production components. Cognition, 92(1e2), 101e144. Klunk, W. E., Engler, H., Nordberg, A., Wang, Y., Blomqvist, G., Holt, D. P., et al. (2004). Imaging brain amyloid in Alzheimer's disease with Pittsburgh Compound-B. Annals of Neurology, 55(3), 306e319. http://dx.doi.org/10.1002/ana.20009. Laird, N. M., & Ware, J. H. (1982). Random-effects models for longitudinal data. Biometrics, 38(4), 963e974. Lange, R. L., Iverson, G. L., Senior, G. J., & Chelune, G. J. (2002). A primer on cluster analysis applications to cognitive rehabilitation research. Journal of Cognitive Rehabilitation, 20(1), 16e33. Lerch, J. P., & Evans, A. C. (2005). Cortical thickness analysis examined through power analysis and a population simulation. NeuroImage, 24(1), 163e173. http://dx.doi.org/ 10.1016/j.neuroimage.2004.07.045. S1053-8119(04)00418-5 [pii]. Levelt, W. J. (2001). Spoken word production: a theory of lexical access. Proceedings of the National Academy of Sciences United States of America, 98(23), 13464e13471. http://dx.doi.org/ 10.1073/pnas.231459498, 98/23/13464 [pii]. Leyton, C. E., Ballard, K. J., Piguet, O., & Hodges, J. R. (2014). Phonologic errors as a clinical marker of the logopenic variant of PPA. Neurology, 82, 1636e1642. http://dx.doi.org/10.1212/ wnl.0000000000000387. Leyton, C. E., & Hodges, J. R. (2013). Towards a clearer definition of logopenic progressive aphasia. Current Neurology and Neuroscience Reports, 13(11), 396. http://dx.doi.org/10.1007/ s11910-013-0396-6. Leyton, C. E., & Hodges, J. R. (2014). Differential diagnosis of primary progressive aphasia variants using the international criteria. Aphasiology, 28(8e9), 909e921. http://dx.doi.org/ 10.1080/02687038.2013.869306. Leyton, C. E., Hornberger, M., Mioshi, E., & Hodges, J. R. (2010). Application of Addenbrooke's cognitive examination to diagnosis and monitoring of progressive primary aphasia. Dementia and Geriatric Cognitive Disorders, 29(6), 504e509. http:// dx.doi.org/10.1159/000313980, 000313980 [pii]. Leyton, C. E., Hsieh, S., Mioshi, E., & Hodges, J. R. (2013). Cognitive decline in logopenic aphasia: more than losing words. Neurology, 80(10), 897e903. http://dx.doi.org/10.1212/ WNL.0b013e318285c15b. WNL.0b013e318285c15b [pii]. Leyton, C. E., Piguet, O., Savage, S., Burrell, J., & Hodges, J. R. (2012). The neural basis of logopenic progressive aphasia. Journal of Alzheimer's Disease, 32(4), 1051e1059. http:// dx.doi.org/10.3233/JAD-2012-121042. QRNJ17317W21442W [pii]. Leyton, C. E., Savage, S., Irish, M., Schubert, S., Piguet, O., Ballard, K. J., et al. (2014). Verbal repetition in primary progressive aphasia and Alzheimer's disease. Journal of Alzheimer's Disease. http://dx.doi.org/10.3233/jad-132468. Leyton, C. E., Villemagne, V. L., Savage, S., Pike, K. E., Ballard, K. J., Piguet, O., et al. (2011). Subtypes of progressive aphasia: application of the international consensus criteria and validation using beta-amyloid imaging. Brain, 134(10), 3030e3043. http://dx.doi.org/10.1093/brain/awr216. Libon, D. J., Rascovsky, K., Powers, J., Irwin, D. J., Boller, A., Weinberg, D., et al. (2013). Comparative semantic profiles in semantic dementia and Alzheimer's disease. Brain, 136(8), 2497e2509. http://dx.doi.org/10.1093/brain/awt165. Machulda, M. M., Whitwell, J. L., Duffy, J. R., Strand, E. A., Dean, P. M., Senjem, M. L., et al. (2013). Identification of an atypical variant of logopenic progressive aphasia. Brain and Language. http://dx.doi.org/10.1016/j.bandl.2013.02.007. S0093934X(13)00050-3 [pii]. Martin, A. (2007). The representation of object concepts in the brain. Annual Review of Psychology, 58, 25e45. http://dx.doi.org/ 10.1146/annurev.psych.57.102904.190143.

Please cite this article in press as: Leyton, C. E., et al., Is the logopenic-variant of primary progressive aphasia a unitary disorder?, Cortex (2015), http://dx.doi.org/10.1016/j.cortex.2015.03.011

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

CORTEX1427_proof ■ 13 April 2015 ■ 12/12

12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

c o r t e x x x x ( 2 0 1 5 ) 1 e1 2

Mesulam, M. M., Weintraub, S., Rogalski, E. J., Wieneke, C., Geula, C., & Bigio, E. H. (2014). Asymmetry and heterogeneity of Alzheimer's and frontotemporal pathology in primary progressive aphasia. Brain, 137(Pt 4), 1176e1192. http:// dx.doi.org/10.1093/brain/awu024. Mesulam, M. M., Wieneke, C., Thompson, C., Rogalski, E., & Weintraub, S. (2012). Quantitative classification of primary progressive aphasia at early and mild impairment stages. Brain, 135(5), 1537e1553. http://dx.doi.org/10.1093/brain/aws080. Mion, M., Patterson, K., Acosta-Cabronero, J., Pengas, G., Izquierdo-Garcia, D., Hong, Y. T., et al. (2010). What the left and right anterior fusiform gyri tell us about semantic memory. Brain, 133(11), 3256e3268. http://dx.doi.org/10.1093/ brain/awq272. Mioshi, E., Dawson, K., Mitchell, J., Arnold, R., & Hodges, J. R. (2006). The Addenbrooke's cognitive examination revised (ACE-R): a brief cognitive test battery for dementia screening. International Journal of Geriatric Psychiatry, 21(11), 1078e1085. http://dx.doi.org/10.1002/gps.1610. Mioshi, E., Hsieh, S., Savage, S., Hornberger, M., & Hodges, J. R. (2010). Clinical staging and disease progression in frontotemporal dementia. Neurology, 74(20), 1591e1597. http:// dx.doi.org/10.1212/WNL.0b013e3181e04070, 74/20/1591 [pii]. Patterson, K., Nestor, P. J., & Rogers, T. T. (2007). Where do you know what you know? The representation of semantic knowledge in the human brain. Natural Review Neuroscience, 8(12), 976e987. http://dx.doi.org/10.1038/nrn2277. nrn2277 [pii]. Petroi, D., Duffy, J. R., Strand, E. A., & Josephs, K. A. (2014). Phonologic errors in the logopenic variant of primary progressive aphasia. Aphasiology, 1e21. http://dx.doi.org/ 10.1080/02687038.2014.910591. Price, C. J., Devlin, J. T., Moore, C. J., Morton, C., & Laird, A. R. (2005). Meta-analyses of object naming: effect of baseline. Human Brain Mapping, 25(1), 70e82. http://dx.doi.org/10.1002/ hbm.20132. Reitan, R. M. (1955). The relation of the trail making test to organic brain damage. Journal of Consulting Psychology, 19(5), 393e394.

Rohrer, J. D., Caso, F., Mahoney, C., Henry, M., Rosen, H. J., Rabinovici, G., et al. (2013). Patterns of longitudinal brain atrophy in the logopenic variant of primary progressive aphasia. Brain and Language, 127(2), 121e126. http://dx.doi.org/ 10.1016/j.bandl.2012.12.008. Rohrer, J. D., Ridgway, G. R., Crutch, S. J., Hailstone, J., Goll, J. C., Clarkson, M. J., et al. (2010). Progressive logopenic/ phonological aphasia: erosion of the language network. NeuroImage, 49(1), 984e993. http://dx.doi.org/10.1016/ j.neuroimage.2009.08.002. S1053-8119(09)00863-5 [pii]. Rohrer, J. D., Rossor, M. N., & Warren, J. D. (2012). Alzheimer's pathology in primary progressive aphasia. Neurobiology of Aging, 33(4), 744e752. http://dx.doi.org/10.1016/ j.neurobiolaging.2010.05.020. S0197-4580(10)00234-4 [pii]. Rowe, C. C., Ellis, K. A., Rimajova, M., Bourgeat, P., Pike, K. E., Jones, G., et al. (2010). Amyloid imaging results from the Australian imaging, biomarkers and lifestyle (AIBL) study of aging. Neurobiology of Aging, 31(8), 1275e1283. http:// dx.doi.org/10.1016/j.neurobiolaging.2010.04.007. S01974580(10)00164-8 [pii]. Savage, S., Hsieh, S., Leslie, F., Foxe, D., Piguet, O., & Hodges, J. R. (2013). Distinguishing subtypes in primary progressive aphasia: application of the sydney language battery. Dementia and Geriatric Cognitive Disorders, 35(3e4), 208e218. http:// dx.doi.org/10.1159/000346389, 000346389 [pii]. Teichmann, M., Kas, A., Boutet, C., Ferrieux, S., Nogues, M., Samri, D., et al. (2013). Deciphering logopenic primary progressive aphasia: a clinical, imaging and biomarker investigation. Brain, 136(Pt 11), 3474e3488. http://dx.doi.org/ 10.1093/brain/awt266. Warren, J. D., Fletcher, P. D., & Golden, H. L. (2012). The paradox of syndromic diversity in Alzheimer disease. Nature Review Neurology, 8(8), 451e464. http://dx.doi.org/10.1038/ nrneurol.2012.135. nrneurol.2012.135 [pii]. Wechsler, D. (1997). WAIS-III administration and scoring manual. San Antonio: The Psychological Corporation.

Please cite this article in press as: Leyton, C. E., et al., Is the logopenic-variant of primary progressive aphasia a unitary disorder?, Cortex (2015), http://dx.doi.org/10.1016/j.cortex.2015.03.011

38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74