doi:10.1093/brain/awr216
Brain 2011: 134; 3030–3043
| 3030
BRAIN
A JOURNAL OF NEUROLOGY
Subtypes of progressive aphasia: application
of the international consensus criteria and
validation using b-amyloid imaging
Cristian E. Leyton,1,2 Victor L. Villemagne,3,4,5 Sharon Savage,1 Kerryn E. Pike,3,4,6
Kirrie J. Ballard,7 Olivier Piguet,1,2 James R. Burrell,1,2 Christopher C. Rowe3,5 and
John R. Hodges1,2
1
2
3
4
5
6
7
Neuroscience Research Australia, Randwick, NSW, 2031, Australia
The University of New South Wales, Sydney, NSW, 2052, Australia
Department of Nuclear Medicine and Centre for PET, Austin Health, Heidelberg, VIC, 3084, Australia
The Mental Health Research Institute, The University of Melbourne, Melbourne, VIC, 3052, Australia
Department of Medicine, Austin Health, The University of Melbourne, Melbourne, VIC, 3084, Australia
School of Psychological Science, La Trobe University, Bundoora, Melbourne, VIC, 3086, Australia
Faculty of Health Science, The University of Sydney, Sydney, NSW, 2141, Australia
Correspondence to: Prof. John R. Hodges,
Neuroscience Research Australia,
Barker Street, Randwick,
NSW, 2031, Sydney, Australia
E-mail: [email protected]
Primary progressive aphasia comprises a heterogeneous group of neurodegenerative conditions with diverse clinical profiles and
underlying pathological substrates. A major development has been the publication of the recent International Consensus Criteria
for the three major variants namely: semantic, non-fluent/agrammatic and logopenic. The logopenic variant is assumed to
represent an atypical presentation of Alzheimer pathology although evidence for this is, at present, limited. The semantic
and non-fluent/agrammatic variants are largely associated with frontotemporal lobar degeneration with TDP-43 and tau pathology, respectively. The applicability of the International Consensus Criteria to an unselected clinical sample is unknown and no
agreed clinical evaluation scale on which to derive the diagnosis exists. We assessed 47 consecutive cases of primary progressive aphasic seen over a 3-year period in a specialist centre, using a newly developed progressive aphasia language scale. A
subgroup of 30 cases underwent 11C-labelled Pittsburgh Compound B positron emission tomography imaging, a putative
biomarker of Alzheimer’s disease that detects b-amyloid accumulation, and they were compared with an age-matched group
(n = 10) with typical, predominately amnestic Alzheimer’s disease. The application of an algorithm based on four key speech and
language variables (motor speech disorders, agrammatism, single-word comprehension and sentence repetition) classified 45 of
47 (96%) of patients and showed high concordance with the gold standard expert clinical diagnosis based on the International
Consensus Criteria. The level of neocortical b-amyloid burden varied considerably across aphasic variants. Of 13 logopenic
patients, 12 (92%) had positive b-amyloid uptake. In contrast, one of nine (11%) semantic variant and two of eight (25%)
non-fluent/agrammatic cases were positive. The distribution of b-amyloid across cortical regions of interest was identical in
cases with the logopenic variant to that of patients with typical Alzheimer’s disease although the total load was lower in the
aphasic cases. Impairments of sentence repetition and sentence comprehension were positively correlated with neocortical
burden of b-amyloid, whereas impaired single-word comprehension showed a negative correlation. The International
Received May 30, 2011. Revised July 12, 2011. Accepted August 1, 2011
ß The Author (2011). Published by Oxford University Press on behalf of the Guarantors of Brain. All rights reserved.
For Permissions, please email: [email protected]
Validation of PPA classification
Brain 2011: 134; 3030–3043
| 3031
Consensus Criteria can be applied to the majority of cases with primary progressive aphasic using a simple speech and language
assessment scale based upon four key variables. b-amyloid imaging confirms the higher rate of Alzheimer pathology in the
logopenic variant and, in turn, the low rates in the other two variants. The study offers insight into the biological basis of
clinical manifestations of Alzheimer’s disease, which appear topographically independent of b-amyloid load.
Keywords: progressive aphasia; Alzheimer’s disease; semantic dementia; progressive non-fluent aphasia;
compound B
11
C-labelled Pittsburgh
Abbreviations: ACE-R = Addenbrooke’s Cognitive Examination revised; PALS = progressive aphasia language scale; PiB = [11C]
Pittsburgh Compound B; PPA = primary progressive aphasia
Introduction
Relatively isolated aphasia due to neurodegeneration was reported
more than a century ago (Pick et al., 1994, 1997), but interest
was rekindled following the seminal report of Mesulam (1982,
2001) who coined the term primary progressive aphasia (PPA)
to describe patients with relentless language impairment without
generalized dementia. Following numerous case reports, it became
clear that patients followed two clinical patterns: fluent and nonfluent variants (Tyrrell et al., 1990; Snowden et al., 1992; Kertesz
et al., 1994, 2003; Hodges and Patterson, 1996; Knibb et al.,
2006). Cases with fluent variant, better known as semantic dementia, present a coherent syndrome with severe anomia and
impaired single-word comprehension associated with gradual disintegration of knowledge about words, people and objects, but
with preservation of the phonological and syntactic aspects of
language (Warrington, 1975; Snowden et al., 1989; Hodges
et al., 1992, 2010). Cases with non-fluent/agrammatic variant,
for the sake of convenience hereafter called non-fluent variant,
in contrast, have disturbances of language production (Grossman
et al., 1996; Hodges and Patterson, 1996; Knibb et al., 2006; Ash
et al., 2009; Wilson et al., 2010) with variable features that include oversimplification of language production and syntactic
errors, effortful, halting speech, distorted articulation and prosodic
changes (Grossman et al., 1996; Knibb et al., 2006; Ogar et al.,
2007; Ash et al., 2009; Rohrer et al., 2010b; Wilson et al., 2010).
Efforts to subdivide patients with non-fluent variant were unsuccessful until the identification of a third clinical variant, known
as logopenic variant (Gorno-Tempini et al., 2004), characterized
by word-finding problems with frequent hesitations, marked
anomia and difficulty repeating sentences or strings of words,
but flawless comprehension of single words and, in contrast
to non-fluent variant, preservation of spoken syntax and motor
aspects of speech (Gorno-Tempini et al., 2004, 2008;
Josephs et al., 2008).
Each variant is associated with a characteristic pattern of brain
atrophy that reflects the cognitive system compromised. Patients
with semantic dementia show the most consistent pattern, characterized by bilateral, but typically asymmetrical, polar and inferior
temporal atrophy (Hodges et al., 1992; Rosen et al., 2002;
Schroeter et al., 2008; Rohrer et al., 2009b; Mion et al., 2010).
In contrast, patients with non-fluent variant exhibit a more variable pattern with left-sided perisylvian atrophy, involving the inferior frontal gyrus, anterior portion of the insula and the middle
frontal gyrus (Nestor et al., 2003; Gorno-Tempini et al., 2004;
Wilson et al., 2010). Data on the pattern of atrophy in logopenic
variant are relatively limited, but it appears to be associated with
left inferior parietal and superior temporal atrophy (Gorno-Tempini
et al., 2004, 2008; Rohrer et al., 2009a).
The underlying pathology of PPA variants is also diverse.
Considered as a whole, the majority of patients with PPA have pathology within the spectrum of frontotemporal lobar degeneration that
comprises forms with tau-positive inclusions (Pick’s Disease, corticobasal degeneration, familial MAPt gene mutations) and
tau-negative cases with deposition of TDP-43, including patients
with mutation of the progranulin gene (Kertesz et al., 1994;
Hodges and Patterson, 1996; Turner et al., 1996; Forman et al.,
2006; Knibb et al., 2006; Cairns et al., 2007; Snowden et al.,
2007; Mesulam et al., 2008; Hu et al., 2010). Importantly, however, a significant proportion of cases, up to a quarter in some
series, have Alzheimer pathology at post-mortem (Knibb et al.,
2006; Mesulam et al., 2008). Clinicopathological studies indicate
a degree of convergence between clinical syndromes and underlying pathology. Those with semantic dementia have predominantly frontotemporal lobar degeneration-TDP-43 and rarely
frontotemporal lobar degeneration with tau-positive inclusions or
Alzheimer pathology (Kertesz et al., 2005; Deramecourt et al.,
2010; Hodges et al., 2010). The situation with non-fluent variant
is less clear since most clinicopathological studies pre-date the delineation of logopenic variant (Kertesz et al., 2005; Josephs et al.,
2006; Knibb et al., 2006). Motor speech disorders have been strongly associated with frontotemporal lobar degeneration with taupositive inclusions (Josephs et al., 2006), but some patients with
non-fluent variant have frontotemporal lobar degeneration
TDP-43 (Knibb et al., 2006; Snowden et al., 2007). Limited
pathological studies of logopenic patients support the concept
that this variant represents an atypical presentation of
Alzheimer’s disease (Mesulam et al., 2008; Rohrer et al., 2009a).
The ability to detect pathological changes in vivo represents a
major advance in neurodegenerative diseases and is clearly relevant to PPA. The PET tracer [11C] Pittsburgh Compound B (PiB)
has high affinity for fibrilllar b-amyloid peptide, which is one of the
hallmarks of Alzheimer pathology (Klunk et al., 2004). Not only is
PiB-PET highly sensitive to b-amyloid deposit in vivo, but it also
shows a strong anatomical correspondence with the b-amyloid
burden in pathologically confirmed cases (Bacskai et al., 2007;
Ikonomovic et al., 2008). Studies using PiB-PET in patients with
a clinical diagnosis of frontotemporal dementia have shown that
3032
| Brain 2011: 134; 3030–3043
between a quarter and a third had PiB-positive scans and most
had a language presentation (Rabinovici et al., 2007, 2008; Engler
et al., 2008). Based on convergent clinical and neuropathological
data, it would be predicted that the likelihood of PiB retention
would vary considerably across PPA variants. To date, only one
study of 15 patients with PPA (non-fluent variant = 6; semantic
dementia = 5; logopenic = 4) undergoing a PiB-PET has been reported (Rabinovici et al., 2008). This study found six of their patients to have a PiB-positive (40%) scan, with the following clinical
distribution: all four logopenic cases and one from each of the
other two categories. These findings are in keeping with the assumption that logopenic variant represents an atypical presentation of Alzheimer’s disease, but the number in each subgroup was
very small and it should be noted that this study was completed
before the formalization of criteria for the PPA subtypes.
To encompass the rapid developments in the field, a broad
ranging International Consensus Group recently published recommendations for the diagnosis and classification of PPA
(Gorno-Tempini et al., 2011). To harmonize terminology, semantic
dementia is referred to as semantic variant of PPA. The guidelines
define each variant according to the presence or absence of core
language features, with recommendations for the type of assessment required to classify subjects (Table 1). The applicability of
these criteria in a large sample of unselected clinic attendees with
PPA is, however, yet to be determined. A key question is the
validity of the separation of non-fluent variant from logopenic
variant based on the assumption that patients with logopenic variant PPA have predominantly Alzheimer’s disease as underlying
pathology. The assessment of aphasia is complex and the proposed criteria take a multidimensional approach that may be difficult to apply especially by those not versed in the evaluation of
patients with aphasia. An easily applicable bedside clinical scale
that is capable of capturing and grading the key language features
essential for the classification of PPA is clearly needed.
C. E. Leyton et al.
Our primary aim was to validate the new International
Consensus Group criteria for PPA using PiB-PET imaging to
detect cases with putative Alzheimer pathology and, thereby, to
test the hypothesis that patients with logopenic variant, in contrast
to other subtypes, have predominantly Alzheimer pathology. A
second aim was to establish the proportion of patients correctly
classified by the new criteria. A third aim was to develop a robust
and easily applicable assessment scale to be applied systematically
to a consecutive series of PPA cases.
Materials and methods
Participants
Consecutive cases with PPA referred to Frontier’s specialist
Frontotemporal Dementia Clinic at Neuroscience Research Australia
in Sydney (www.ftdrg.org) between January 2008 and December
2010 were recruited. Participants were regarded as having PPA if
the most prominent clinical feature was slowly progressive aphasia
for at least 2 years, without other significant cognitive or behavioural
symptoms, and preservation of daily living activities except those attributable to the language impairment (Mesulam, 2001). We excluded:
(i) participants with limited English proficiency (proficiency was defined
as those who had English as a second language and had lived and
worked in an English speaking country for 410 years); (ii) participants
with advanced disease who were essentially mute or had
Addenbrooke’s Cognitive Examination revised (ACE-R) (Mioshi et al.,
2006) 535; (iii) participants with concomitant motoneuron disease,
significant extrapyramidal features, a past history of stroke, epilepsy,
alcoholism or significant traumatic brain injury; (iv) participants with
abnormalities on MRI brain scan, other than atrophy; or (v) participants without an adequate digital video record. As a result, from a
total of 70 cases with PPA, 47 participants were eligible for the study.
Healthy control subjects (n = 52) were selected from the volunteer
panel of Frontier and matched case by case according to level of
Table 1 Summary of international recommendations for the diagnosis and classification of PPA
Primary progressive aphasia
(i) Most prominent clinical feature is a difficulty with language (word-finding deficits, paraphasias, effortful speech, grammatical and/or
comprehension deficits).
(ii) Aphasia should be the most prominent deficit at symptom onset and for the initial phases of the disease (other prominent symptoms,
such as behavioural, memory or visuospatial impairments should not be present at the onset).
(iii) No other conditions that may be better account for the language deficits are present (i.e. non-degenerative or psychiatric
conditions).
Semantic variant
(i) Poor confrontation naming and impaired single-word comprehension, explained by dissolution of semantic knowledge.
(ii) No motor speech disorders or agrammatism.
Non-fluent variant
(i) Either agrammatism or motor speech disorders with effortful, halting speech, inconsistent sound errors and distortions.
(ii) Spared single-word comprehension and object knowledge.
Logopenic variant
(i) Impaired single-word retrieval in spontaneous speech (speech fluency interrupted by word-finding pauses) and confrontational
naming; and impaired repetition of sentences and phrases.
(ii) Spared single-word comprehension and absence of motor speech disorders.
Validation of PPA classification
Brain 2011: 134; 3030–3043
education (mean: 13.2 years) and age (mean: 68.5 years). All participants underwent a brain MRI and a routine core neuropsychological
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.
Language assessment
A comprehensive assessment of speech and language skills was
performed and video-recorded. The assessment comprised a
semi-structured interview followed by a standardized sequence of language tasks required to classify patients into one of the three variants
of PPA (Table 1) (Gorno-Tempini et al., 2011). The assessment began
with an introductory conversation of 10–15 min initiated by the interviewer who introduced a similar range of topics in each case, such as
participant’s past employment, family, leisure activities and recent holidays. The interviewer took a mainly passive role, encouraging the
participant to talk. Video records of each conversation provided a naturalistic speech sample, which was analysed to determine two key
features: motor speech disorders and agrammatism. Cardinal features
| 3033
of motor speech disorders included presence of distortions, effortful
speech or articulatory groping and loss of normal prosody. Distortions
were defined as a lack of accurate articulation, yielding sounds that
were not recognizable as spoken English. Changes in prosody were
defined as the incorrect placement of stress and natural intonation
upon syllables and words within sentences. Features regarded as
agrammatism included errors in word arrangement (syntax), inflection
(morphology) or omission of grammatical structures resulting in oversimplification of language (Table 2). The informal conversation was
followed by several language tasks. Naming was assessed by presentation of animal models and household objects ranging in familiarity
(e.g. ‘horse’, ‘elephant’, ‘crocodile’, ‘hippopotamus’; and ‘toothbrush’,
‘scissors’, ‘tongs’, ‘whisk’). Distortions of the target word due to articulation problems were not regarded as errors. Single-word comprehension was assessed in two ways: first, by asking the participant to
point to objects (e.g. ‘point to the crocodile’), and secondly, asking
to repeat and define multisyllabic words of increasing difficulty (e.g.
‘caterpillar’, ‘perimeter’, ‘catastrophe’, ‘chrysanthemum’, etc.). Singleword repetition was evaluated in the repeat and define task described
above, and sentence repetition was assessed by repeating plausible
Table 2 Criteria for scoring variables on the progressive aphasia language scale
Item
Spontaneous speech
Motor speech
disorders
Agrammatism
Tasks
Naming
Single-word
repetition
Assessed features
1 (questionable)
2 (definite)
3 (severe)
Rarely present
Consistent or clearly present
Frequently present
Consistent or clearly
present
Output speed is affected
Some words are
unintelligible
The message is
understandable
Almost always present
Output speed is deeply
reduced
Level of intelligibility
Subtle and inconsistently
present
Normal output speed
Almost all words are
intelligible
The message is
understandable
Syntax and morphology
Largely preserved
Frequent and evident errors
Grammatical complexity
Tendency to use simple
structures
At least one error in 5 min
of spontaneous speech
Consistent and clear
oversimplification
Word retrieval when an
object is presented
Repetition of words
Occasional mistakes on
unfamiliar items
Occasional or subtle errors,
often involving difficult
words
Occasional mistakes, usually
on unfamiliar words (e.g.
herbivorous, indigenous)
Distortions, groping or
laboured speech
Changes in stress, intonation or prosody
Single-word
comprehension
Identification of words
Sentence
repetition
Ability to repeat sentences or strings of
words
Ability to follow commands of increasing
complexity
Sentence
comprehension
Score
Consistent mistakes
Clear and consistent errors
Many words are
unintelligible. The
message is hard to
understand
Telegraphic language
Frequent mistakes, often
involving familiar items
Frequent errors
Consistent mistakes
Highly familiar items are
often preserved (e.g.
dog or elephant)
Frequent mistakes
Minimal or occasional
mistakes
Consistent mistakes
Minimal or occasional
mistakes
Increased latency in
replying, requests for
repeating instructions,
and self-corrections
should be considered
Consistent mistakes, often
involving complex
instructions
Unable to repeat correctly
phrases of more than two
words
Consistent mistakes
involving direct and
simple instructions
3034
| Brain 2011: 134; 3030–3043
sentences (e.g. ‘the Chinese fan contained a rare emerald’). Finally,
sentence comprehension was evaluated by asking the patient to follow
instructions with different levels of complexity. Simple instructions
included direct two-step commands (e.g. ‘touch the pen and
then the scissors’), whereas complex instructions included commands
with reversible elements (e.g. ‘pick up the pen after touching the
scissors’).
Development of the progressive
aphasia language scale
To develop the scale, a set of eight language features were operationally defined and graded, taking into account the core features
specified in the consensus guidelines (Gorno-Tempini et al., 2011).
According to the magnitude and consistency of impairments, each language feature was scored from absent (0), subtle or questionable (1),
mild but definitely present (2) to moderate or severe (3) (Table 2). A
pilot version of the scale was then applied to a set of video recordings
of patients with PPA (n = 12) who were not included in the main
study, by two independent raters who had not been involved in
the initial diagnosis of the cases. The total inter-rater reliability
(Spearman’s r) was 0.93, ranging for each item from 0.63 to 1.00.
The item with the lowest reliability, ‘word-finding problems’, was
excluded. Based on the results of the inter-rater reliability study and
a number of minor modifications to the scoring criteria, a final version
of the Progressive Aphasia Language Scale (PALS) with seven items
was used in this study (see instructions of the PALS in the online
Supplementary Material).
Pittsburgh Compound B-positron emission tomography scans
Participants were invited to undergo a PET scan at the Austin Hospital,
Melbourne. Of the 35 patients asked to undergo a PET scan, 30
agreed to participate. Each participant received 370 MBq 11C-PiB
intravenously over 1 min. A 30-min acquisition in 3D mode starting
40 min after injection of PiB was performed with a Phillips AllegroTM
PET camera. A transmission scan was performed for attenuation correction. PET images were reconstructed using a 3D RAMLA algorithm.
Images were processed using a preset template of narrow cortical regions of interest placed by an operator who was blind to the participant’s clinical status (V.L.V.). PET data were not corrected for partial
volume effects. Standardized uptake values for PiB were calculated for
all brain regions examined, and standardized uptake value ratios were
generated by dividing all regional standardized uptake values by the
cerebellar cortex standardized uptake values. Neocortical b-amyloid
burden was expressed as the average standardized uptake value
ratios of frontal, superior parietal, lateral temporal, lateral occipital,
and anterior and posterior cingulate regions. Given the bimodal distribution of PiB-standardized uptake value ratios that is observed in
healthy controls, a hierarchical cluster analysis was performed on all
elderly healthy-control participants at Austin Health, n = 118, age
73.2 7.4 years, Mini-Mental State Examination 29.2 1.0
(mean SD) that yielded a cut-off for ‘high’ or ‘low’ neocortical standardized uptake value ratios of 1.50, consistent with cut-off values
used in previous PiB-PET studies (Rowe et al., 2010; Villemagne
et al., 2011). To compare positive cases, a sample of patients with
typical early-stage Alzheimer’s disease, with a predominately amnestic
presentation, was identified from the Austin Hospital database and
matched according to sex and age (n = 10, age 65.0 8.7 years,
Mini-Mental State Examination 20.8 2.5).
C. E. Leyton et al.
Statistical analysis
Non-parametric statistical methods were used to have a consistent
approach to non-normal distributed data, such as neocortical PiBstandardized uptake value ratios. Kruskal–Wallis followed by pair-wise
comparisons with Mann–Whitney tests were used for the analysis between PPA variants. Spearman rank-order correlations (denoted as ‘r’)
were used to evaluate associations between paired numeric measures.
Chi-squared tests were used to analyse categorical differences, while
the Fisher’s exact test was used to analyse differences in the performance of PALS items. The level of significance was adjusted for multiple
comparisons using the Bonferroni corrections and was set at 0.05,
unless otherwise stated.
Statistical analyses were undertaken using PASW 17.0 (IBM Corp.).
Results
Demographic characteristics and general cognition
Demographic data on the 47 cases with PPA (14 semantic variant,
15 non-fluent variant and 18 logopenic variant clinically classified)
are shown in Table 3. The overall PPA group did not differ from
controls in age, education or sex distribution. Similarly, the subgroup of 30 who undertook PiB-PET scans did not differ from the
other PPA cases in general demographic features, total time of
disease duration or general cognition (ACE-R) (data not shown).
Between PPA variants, however, logopenic cases had the largest
proportion of females and non-fluent cases showed the shortest
disease duration. With respect to general cognition, although
non-fluent cases showed better performance, all PPA variants
had lower ACE-R and Mini-Mental State Examination scores
than the matched controls (Table 3). On the ACE-R, non-fluent
cases performed better than the other PPA variants on the language and memory sub-domains, whereas no intergroup differences were found in the remaining sub-domains. Notably,
performance was equivalent and well preserved on the visuospatial subscale.
Application of the progressive aphasia
language scale
The performance of clinically classified participants on the language variables assessed by the PALS is shown in Fig. 1. Motor
speech disorders were severe in the majority of participants with
non-fluent variant and largely absent in the other groups, with a
significant difference between non-fluent variant and other
groups. Although rarely severe, agrammatism was also seen in
most participants with non-fluent variant and displayed the same
pattern of intergroup differences as for motor speech disorders. All
patients with semantic variant had moderate or severe anomia and
this represented a significantly greater proportion than seen in the
other two variants. Although a sizeable proportion of participants
with logopenic variant also had severe anomia, this proportion did
not differ from the non-fluent group. Single-word repetition was
essentially normal in semantic variant while most participants with
Validation of PPA classification
Brain 2011: 134; 3030–3043
| 3035
Table 3 Demographic and clinical characteristics of participants
Demographic*
Sex, female (%)
Age, years
Education, years
Disease duration, years
General cognition*
ACE-R (/100)
Attention (/18)
Memory (/26)
Fluency (/14)
Language (/26)
Visuospatial (/16)
Mini-Mental State Examination (/30)
Controls
(n = 52)
PPA
(n = 47)
P-value
40
69 (52–81)
13 (8–20)
–
43
66 (48–84)
13 (7–22)
4 (0–11)
95
18
24
12
25
16
29
65 (41–95)
15 (11–18)
14 (1–26)
5 (0–12)
16 (7–26)
14 (7–16)
23 (16–29)
(88–100)
(16–18)
(20–26)
(8–14)
(22–26)
(12–16)
(27–30)
PPA variant
Semantic
variant
(n = 14)
Non-fluent
variant
(n = 15)
Logopenic
variant
(n = 18)
P-value
NS
NS
NS
–
14a
64 (53–73)
12 (9–19)
5 (2–11)c
33
67 (48–84)
12 (7–22)
3 (0–8)b
72b
67 (54–79)
13 (9–19)
4 (1–11)
50.01
NS
NS
50.05
50.001
50.001
50.001
50.001
50.001
50.001
50.001
59 (43–79)c
16 (13–18)
12 (7–19)c
5 (0–10)
12 (7–20)a,c
14 (9–16)
24 (19–29)
73 (41–95)b
16 (12–18)
18 (1–26)a,b
5 (0–12)
20 (13–26)a,b
14 (7–16)
24 (17–29)
63 (44–78)
15 (11–18)
13 (4–23)c
6 (2–10)
16 (8–21)b,c
14 (10–16)
22 (16–26)
50.05
NS
50.01
NS
50.001
NS
NS
*Except for Sex, all values are expressed as mean (range).
a Significant difference from logopenic variant cases.
b Significant difference from semantic variant cases.
c Significant difference from non-fluent variant cases.
NS = not significant.
non-fluent variant and a sizeable proportion of logopenic cases
produced repetition errors. The difference between non-fluent
variant and logopenic variant did not reach statistical significance.
Single-word comprehension was invariably poor in semantic participants with no definite impairment in the other two groups.
Sentence repetition deficits were universal in non-fluent and logopenic variants, showing a significant difference between these and
the semantic group. Those with logopenic variant were severely
impaired, typically omitted words and could only produce the first
two or three words, while patients with non-fluent variant showed
less marked deficits characterized by distortions and some omissions. Finally, no intergroup differences were found in the performance of sentence comprehension.
Classification of primary progressive
aphasia variants: comparison of the
clinical gold standard and the PALS
The gold standard clinical diagnosis was based on the consensus
of two experienced clinicians (J.R.H. and J.R.B. or K.J.B.), following the clinical assessment and using the results of neuropsychological and language tests. This gold standard was compared
with a diagnosis based purely on the results of a simple short
speech and language evaluation, the PALS, with the aim of being
able to replace a lengthy expert-dependent diagnosis with
one based on the PALS. We developed an algorithm capable of
classifying the participants using empirical data. Based on the results detailed above, three components of the PALS were
excluded. Naming was excluded as all three groups contained
members with moderate to severe impairment. Single-word repetition deficits, although present in all non-fluent cases, were also
common in a large proportion of logopenic cases. Sentence comprehension performance did not vary significantly across the variants. The four features retained were, therefore: motor speech
disorders, agrammatism, single-word comprehension and sentence repetition. For the purposes of classification, scores were
dichotomized as present (definite or severe) or absent (questionable or absent). Since several core features may be present simultaneously in one case, a hierarchical approach was adopted and
a number of models contrasted. To determine the classifier level,
a correlational analysis of the four items was carried out
(Supplementary Table 1). Single-word comprehension was the
only item that showed a significant negative correlation with
the other items and was therefore the first classifier. The
second classifying feature was the presence of either agrammatism or motor speech disorders that were strongly correlated
(Spearman’s r = 0.68; P 5 0.001) but could be present independently. The third feature, sentence repetition, showed the most
diverse pattern of correlations. By far the best algorithm using
the four PALS features is shown in Fig. 2. This four-feature solution consistently classified a high proportion of cases and had a
good concordance with the clinical diagnosis. In 44 of 47
cases (94%), there was concordance between the expert clinical diagnosis and the PALS-based algorithm. As shown in Table 4,
the semantic and logopenic groups had a 100 and 94% concordance, respectively, between approaches, whereas the non-fluent
group had the lowest level, with 87% concordance. The discrepant cases included one clinically diagnosed logopenic case
classified as non-fluent by the PALS-based algorithm and two
clinically diagnosed non-fluent cases who had mild or questionable deficits on PALS items and consequently could not be
classified.
3036
| Brain 2011: 134; 3030–3043
C. E. Leyton et al.
Figure 1 Score distribution of the PALS items across the clinically defined PPA variants. NFV = non-fluent variant; LV = logopenic variant;
SV = semantic variant.
Pittsburgh Compound B retention status
Thirty patients (nine semantic variant, eight non-fluent variant and
13 logopenic variant) underwent a PiB-PET scan. Overall, half of
them showed significant neocortical PiB retention (Fig. 3). Of the
15 with PiB retention, 12 presented with the logopenic variant of
PPA. As predicted, the semantic and non-fluent groups showed
the lowest proportion of positive cases: one of nine semantic
(11%) and two of eight non-fluent cases (25%). In contrast, 12
of 13 logopenic (92%) cases had positive scans. Accordingly, the
semantic
and
non-fluent
groups
had
a
neocortical
PiB-standardized uptake value ratios level [mean (range)] of 1.19
(1.03–1.79) and 1.28 (1.02–1.87), respectively, both significantly
lower than the logopenic group, 2.02 (1.41–2.49), Kruskal–Wallis
test, P 5 0.001. The only unexpected negative logopenic case had
four neocortical right-sided regions that exceeded the
PiB-standardized uptake value ratios cut-off of 1.50: orbito and
ventrolateral frontal, lateral temporal and occipital (Fig. 4). On the
Validation of PPA classification
Brain 2011: 134; 3030–3043
| 3037
Total PPA cases
(n =47)
Present
(n = 14)
Single Word
Comprehension
Impaired
Absent
(n =33)
Motor Speech
Disorder or Agrammatism
Absent
(n= 14)
Motor Speech
Disorder or Agrammatism
Present
(n= 0)
Absent
(n =19)
Present
(n = 14)
Sentence repetition
Impaired
Absent
(n =2)
14 SV cases
0 unclassifiable
cases
2 unclassifiable
cases
Present
(n =17)
17 LV cases
14 NFV cases
Figure 2 Algorithm for diagnosing PPA cases based on four key clinical features. ‘Present’ represents a rating of definite or severe on
PALS. NFV = non-fluent variant; LV = logopenic variant; SV = semantic variant.
PALS-based non-fluent cases, 11 of 12 PALS-based logopenic
cases and one of nine semantic cases had PiB-positive scans.
Table 4 Comparison of diagnostic approaches
Clinical
based
Algorithm based
Logopenic Unclassifiable Total
Semantic Nonvariant
fluent variant
variant
Semantic
14
variant
Non-fluent 0
variant
Logopenic
0
variant
Total
14
0
0
0
14
13
0
2
15
1
17
0
18
14
17
2
47
other hand, the unexpected positive semantic case and the two
non-fluent cases had mean neocortical PiB-standardized uptake
value ratios of 1.79, 1.87 and 1.61, respectively, which were
lower than the mean neocortical PiB-standardized uptake value
ratios of logopenic cases (Fig. 3). With regards to the
PALS-based classification, of the three discrepant cases, two
underwent PiB scans. One unclassifiable case had a negative
scan, but one logopenic case diagnosed by the algorithm as
non-fluent had a positive scan. As a result, three of eight
Comparison between logopenic variant
and typical Alzheimer’s disease cases
Although logopenic variant had significantly lower PiBstandardized uptake value ratio levels than Alzheimer’s disease
cases in all cortical regions, except the primary visual and the lateral occipital cortices (Table 5), no difference in the pattern of PiB
retention was found between both groups (Fig. 4).
PALS, global and regional neocortical
PiB-standardized uptake value
ratio correlations
Neocortical PiB-standardized uptake value ratios did not correlate
with age, disease evolution or educational level in any of the PPA
groups. Mini-Mental State Examination (r = 0.43, P = 0.017),
but not ACE-R, was significantly associated with neocortical
PiB-standardized uptake value ratios (Supplementary Table 2).
To explore potential correlations between language performances
and cortical b-amyloid accumulation, PALS items were entered in
a correlation matrix with both regional and global neocortical
3038
| Brain 2011: 134; 3030–3043
PiB-standardized uptake value ratio values. Three items correlated
significantly with global PiB-standardized uptake value ratios: sentence repetition and sentence comprehension showed positive correlations (r = 0.64, P 5 0.001 and r = 0.45, P = 0.03,
respectively), while single-word comprehension showed a negative
C. E. Leyton et al.
correlation (r = 0.47, P = 0.009) (Supplementary Table 3). The
only item that showed a regional correlation was single-word
repetition that was positively correlated with the left mesial temporal region (r = 0.38, P = 0.047). No other items correlated with
PiB retention region.
Discussion
We found that 45 of 47 consecutive cases with PPA could be
classified into one of the three variants recently defined by the
International Consensus Group (Gorno-Tempini et al., 2011) offering the first objective support for these criteria. Moreover, the
algorithm-based approach in which key neurolinguistic features
were systematically assessed using a simple standardized speech
and language assessment, the PALS, was highly consistent with
the gold standard expert clinical classification. Four variables
emerged as critical for accurate classification. PiB-PET imaging
validated the hypothesis that the logopenic variant of PPA represents an atypical presentation of Alzheimer’s disease, which has
clear implications for therapy.
Figure 3 Neocortical PiB-standardized uptake value ratios
(SUVR) for each PPA variant (n = 30) and the Alzheimer’s disease (AD) sample (n = 10). The red dashed line indicates the PiB
positivity’s threshold. Horizontal lines represent the mean SUVRs
for each group. *P 5 0.01; **P 5 0.001. NFV = non-fluent
variant; LV = logopenic variant; SV = semantic variant.
Application of the PALS and primary
progressive aphasia classification
Although each PPA variant is associated with a spectrum of
impaired or preserved linguistic abilities, which have been fully
documented in a large number of neuropsychological studies
Figure 4 Representative examples of PiB-PET scanning. (A) PiB-PET image of a negative scan. (B) Image of the logopenic case who had a
positive scan. Arrows indicate the cortical regions with a PiB-standardized uptake values ratios higher than 1.50. (C and D) Images show
positive scans of logopenic and typical Alzheimer’s disease cases, respectively. SUVR = standardized uptake value ratios.
Validation of PPA classification
Brain 2011: 134; 3030–3043
Table 5 Comparison of regional PiB-standardized uptake
value ratios between logopenic variant and Alzheimer’s
disease cases
Cortical region
Alzheimer’s
disease
(n = 10)
P-value
(1.24–2.39)
(1.20–2.37)
2.55 (1.73–3.13)
2.49 (1.66–3.17)
0.003
0.004
(1.55–2.72)
(1.38–2.76)
2.74 (2.18–3.38)
2.70 (2.22–3.45)
0.002
0.002
(1.62–2.64)
(1.49–2.78)
2.75 (2.12–3.46)
2.69 (2.12–3.59)
50.001
0.002
(1.50–2.46)
(1.57–2.45)
2.88 (2.29–3.76)
2.84 (2.25–3.77)
50.001
50.001
(1.48–2.60)
(1.43–2.58)
2.72 (2.30–3.26)
2.69 (2.21–3.30)
0.001
0.001
(1.45–2.98)
(1.33–3.07)
2.82 (2.27–3.63)
2.86 (2.45–3.52)
0.003
0.005
(1.51–2.43)
(1.33–2.45)
2.60 (2.14–3.27)
2.63 (2.16–3.39)
50.001
50.001
(1.15–1.55)
(1.13–1.67)
1.79 (1.50–2.00)
1.81 (1.63–2.03)
50.001
50.001
(1.38–2.29)
(1.10–2.18)
2.42 (1.88–2.95)
2.48 (2.02–3.18)
0.001
50.001
(1.41–2.08)
(1.44–2.28)
2.07 (1.74–2.98)
1.99 (1.63–2.96)
0.069
0.926
(1.28–2.19)
(1.24–2.21)
2.00 (1.44–2.95)
1.94 (1.27–3.00)
0.352
0.733
(1.49–2.47)
(1.34–2.51)
(1.41–2.49)
2.58 (2.05–3.26)
2.57 (2.12–3.31)
2.58 (2.08–3.29)
0.001
0.001
0.001
Logopenic
variant
(n = 13)
Frontal lobe
Dorsolateral
Right
1.92
Left
1.87
Ventrolateral
Right
2.10
Left
2.10
Orbitofrontal
Right
2.07
Left
2.08
Gyrus recto
Right
2.02
Left
2.05
Cingulate cortex
Anterior cingulate
Right
2.03
Left
2.05
Posterior cingulate
Right
2.25
Left
2.28
Temporal lobe
Lateral temporal
Right
2.11
Left
1.98
Mesial temporal
Right
1.41
Left
1.41
Parietal lobe
Superior parietal
Right
1.95
Left
1.87
Occipital lobe
Lateral occipital
Right
1.80
Left
1.91
Visual cortex
Right
1.81
Left
1.86
Neocortex
Right
2.03
Left
2.02
Whole neocortex 2.02
All values expressed as mean (range).
(Hodges and Patterson, 1996; Knibb et al., 2006; Mesulam et al.,
2009; Rohrer et al., 2010b; Sapolsky et al., 2010), we found that
the distinction can be made on relatively simple grounds by considering four language variables: single-word comprehension,
motor speech disorder, agrammatism and impaired sentence repetition. These findings have implications for the international consensus criteria for PPA variants. Participants with semantic variant
showed the most homogeneous pattern characterized by impaired
| 3039
word comprehension, but preservation of speech articulation,
grammatical language production and sentence repetition. This
is in contrast to non-fluent and logopenic participants who
demonstrated preservation of single-word comprehension.
Although the proposed criteria include non-verbal-based semantic
tasks, such as object knowledge and reading/spelling tasks, all
semantic participants, who had undergone a full neuropsychological assessment, could be correctly classified applying the
PALS-based algorithm. Prior work has established that even patients with very early stage semantic variant (semantic dementia)
show impairment on non-verbal-based tests of semantic knowledge (Adlam et al., 2006), but the demonstration of such impairment is not necessary for clinical classification and may be
redundant in the criteria.
The cases with non-fluent variant presented either with motor
speech disorders, expressive agrammatism or both. The distinction
between non-fluent and logopenic variants is, however, demanding not only because of the overlapping deficits in these variants,
but also because the core deficits of non-fluent variant may be
subtle and inconsistent. Motor speech disorders are caused by
disturbance in the planning, programming or execution of movements for speech and assessment usually requires a complete battery of speech motor tasks to elicit characteristic disturbances in
articulation and/or prosody (Duffy, 2005, pp. 307–25; Ogar et al.,
2007; Rohrer et al., 2010a). Nevertheless, our simple clinical scale
was largely concordant with the more thorough evaluation. The
two clinically diagnosed non-fluent cases who were misclassified
by the PALS had mild speech deficits apparent to an expert that
the scale could not capture.
Of the three PPA variants, logopenic variant was identified
most recently (Gorno-Tempini et al., 2004, 2008; Rogalski
et al., 2011) but has emerged as a readily recognized syndrome
if appropriate clinical assessments are performed. Such cases
may be thought initially to have the non-fluent variant because
of their frequent pauses that disrupt the flow of conversation and
the generation of phonological errors, but logopenic cases
lack motor speech disorders or agrammatism and, instead, show
a clear reduction in sentence repetition that is often striking with
omission of words (Gorno-Tempini et al., 2004, 2008). In our
group, a few logopenic participants showed questionable motor
speech disorders or agrammatism that can lead to confusion.
For instance, articulatory errors, which represent motor planning
or programming errors, can distort spatial or temporal aspects of
a target phoneme to such a degree that it sounds like a different
phoneme. In these cases, the articulatory errors can be
mistaken for phonemic paraphasias that represent linguistically
based errors in phonemic selection (Ziegler, 2008). Moreover,
logopenic cases often present false starts that resemble articulatory groping (Wilson et al., 2010). Even so, using strictly the
PALS, no single logopenic case had definite motor speech
disorders.
In contrast to semantic cases in which single-word repetition
was largely preserved, all non-fluent cases and a sizeable proportion of logopenic cases had single-word repetition defects. The
nature of these defects, however, differed in each case.
Whereas patients with non-fluent variant displayed errors of articulation and syllabic fragmentation, logopenic cases showed
3040
| Brain 2011: 134; 3030–3043
phonological errors that indicate a breakdown in phonological integration (Canter et al., 1985; Croot, 2002; Ogar et al., 2007).
Impaired sentence repetition was also present in both non-fluent
and logopenic cases, but in contrast to single-word repetition, it
was the only item that showed definite or severe impairment in all
clinically defined logopenic participants and occurred without
motor speech disorders or agrammatism. In addition, rather than
defining logopenic variant by the absence of motor speech disorders or agrammatism, impaired sentence repetition positively
classified cases with logopenic variant. In non-fluent cases, sentence repetition errors are mainly due to articulatory defects that
also affect single-word processing and are associated with atrophy
in the inferior gyrus and dorsolateral regions of the left frontal lobe
and the anterior insular region (Nestor et al., 2003; Josephs et al.,
2006; Ogar et al., 2007; Rohrer et al., 20010a). Sentence repetition errors in logopenic variant reflect a breakdown in the phonological processing that also affects the repetition of single words.
Yet, the gross impairment in sentence repetition is almost certainly
secondary to a reduction of verbal short-term memory resources.
These findings are in keeping with the proposed anatomical locus
of pathology in logopenic variant, the left temporal–parietal junction, area 21 (Gorno-Tempini et al., 2008; Mesulam et al., 2009;
Rohrer et al., 2009a), a region shown to be critically involved both
in phonological processing and verbal short-term memory (Hickok
and Poeppel, 2007). The relative importance of phonological and
verbal short-term memory deficits in logopenic variant requires
future exploration.
Anomia was present across PPA variants and, therefore, was
not helpful in the classification of patients, despite its inclusion
in the International Consensus Criteria. The naming performance was, however, quantitatively and qualitatively different
in each PPA variant. Those with semantic variant are typically
extremely anomic and show a marked familiarity effect. Their
errors are typically semantic coordinates, superordinate substitutions or ‘don’t know’ responses (Hodges et al., 1992; Rogers
et al., 2004). Cases with logopenic variant are also markedly
anomic and may produce phonological errors or ‘don’t know’ responses (Gorno-Tempini et al., 2008). On the other hand,
those with non-fluent variant, despite their profound disruption
of conversational speech, show the best performance and their
errors mirror spontaneous speech with articulatory distortions
(Rohrer et al., 2010b). The qualitative analysis of naming errors
may differentiate PPA subtypes (Rohrer et al., 2008), but a
fine-grained analysis is beyond the scope of routine outpatient
language assessment. Finally, sentence comprehension was also
disappointing in its classificatory ability. Patients with non-fluent
variant and logopenic variant were equally mildly impaired but
for arguably different reasons. Those with non-fluent variant
have a true syntactic deficit, which has been demonstrated using
tasks that do not put high demands on working memory load
(Weintraub et al., 2009). Patients with logopenic variant are
also impaired on clinical tests of sentence comprehension that require subjects to process complex sentences. Their deficits almost
certainly reflect impairment in attentional resources and phonological short-term memory (Amici et al., 2007; Gorno-Tempini
et al., 2008).
C. E. Leyton et al.
Pittsburgh Compound B retention status
To date, only one PiB imaging study of logopenic cases exists
(Rabinovici et al., 2008). This study involved four participants
with logopenic variant, all of whom showed abnormal PiB retention. Our series of cases represent the largest to date and confirm
the strong association with PiB retention, suggesting, therefore,
underlying Alzheimer pathology in logopenic variant. Twelve out
of 13 logopenic cases were positive for PiB, with an average standardized uptake value ratio significantly higher than the other
variants. The only ‘negative’ case had a neocortical
PiB-standardized uptake value ratio of 1.41 (Fig. 2), just under
the cut-off, but had several cortical regions exceeding the standardized uptake value ratio threshold for positivity. Given that
neocortical PiB-standardized uptake value ratios represents overall
b-amyloid deposition across the brain, these positive regions were
‘diluted’ by the others. Even so, a neocortical PiB-standardized
uptake value ratio of 1.41 already indicates a substantial
b-amyloid deposition in the brain. This degree of PiB retention
likely represents an early stage in the disease process, particularly
knowing that asymptomatic elderly individuals with positive scans
present high retention in these regions (Becker et al., 2011) and
the b-amyloid burden increases with time, especially in the earlier
stages of Alzheimer’s disease (Villemagne et al., 2011). Follow-up
studies are needed to confirm this hypothesis.
One semantic and two non-fluent cases had unexpectedly positive PiB-PET scans. The explanation and clinical relevance of these
findings are not clear although two main possibilities exist. First,
positive PiB-PET scans may be a coincidental finding, taking into
account that up to a third of the cognitively normal elderly population displays positive PiB-PET scans (Pike et al., 2007; Aizenstein
et al., 2008; Mosconi et al., 2010; Rowe et al., 2010). PiB positivity, however, may represent a preclinical phase of Alzheimer’s
disease since b-amyloid deposition is associated with brain atrophy
even in asymptomatic cases (Becker et al., 2011; Oh et al., 2011).
In the same way, several longitudinal follow-up studies have
demonstrated an increased risk of cognitive decline in
PiB-positive healthy volunteers (Scheinin et al., 2009; Resnick
et al., 2010; Villemagne et al., 2011). Furthermore, b-amyloid
deposition increases with age paralleling the rise in prevalence of
Alzheimer’s disease, but with the increase in PiB occurring 15
years before Alzheimer’s disease onset (Rowe et al., 2010). It is
possible, therefore, that these unexpected positive cases have coincidental mild Alzheimer pathology that is not the cause of their
progressive aphasia. A second, more likely, explanation is that
Alzheimer’s disease can manifest as a range of aphasic syndromes.
Although most strongly associated with logopenic variant, a minority of cases with semantic variant have advanced Alzheimer
pathology at post-mortem: in the most recent update of the
Cambridge series, 3 of 24 semantic cases coming to post-mortem
had Alzheimer pathology (Hodges et al., 2010). Similarly, cases
with well-documented non-fluent variant can also have Alzheimer
pathology (Greene et al., 1996; Harasty et al., 2001; Knibb et al.,
2006; Mesulam et al., 2008; Hodges et al., 2010).
The pattern of cortical b-amyloid deposition in patients with
logopenic variant was comparable with typical cases with
Alzheimer’s disease although logopenic variant had a lower
Validation of PPA classification
b-amyloid burden. This finding raises fundamental questions concerning the relation between b-amyloid deposition, neuronal dysfunction and clinical symptomatology. Multiple studies have
shown that this relation is not direct or robust, suggesting that
other, as yet unknown, factors are involved in the development of
neurodegeneration (Jack et al., 2009; Scheinin et al., 2009; Becker
et al., 2011; Chetelat et al., 2010). The assessment of other biomarkers in vivo, such as tau, might help elucidate this issue
(Fodero-Tavoletti et al., 2011).
Three PALS items correlated with the degree of PiB retention
and were, therefore, predictive of b-amyloid deposition. Sentence
repetition and sentence comprehension were positively correlated
with the b-amyloid cortical burden. Both tasks depend heavily on
a common resource: working memory. As discussed earlier, impairments in working memory represent a pivotal deficit in logopenic variant and have been also described in typical cases with
Alzheimer’s disease (Peters et al., 2009). Given the strong association between the syndrome of logopenic variant and PiB status, it
is not surprising that these two language hallmarks were also
related to PiB retention. On the other hand, single-word comprehension, which was almost exclusively present in semantic patients, drove the association in the opposite direction,
highlighting the importance of semantic impairment in the classification of PPA cases.
Limitations
Although we have demonstrated the validity of the proposed criteria, using a biomarker, certain caveats need to be considered.
The stringent selection of participants may have yielded mostly
mild or moderate cases who tended to display circumscribed language impairments, making their classification easier. The hierarchical approach, together with the quantification of
impairments, however, may have overcome this issue since a
few contrasting features could discriminate cases with mixed characteristics. Another issue is the application of our results in routine
clinical practice. Although the PALS-based algorithm was applied
in a research setting, the scale was simple, with clear instructions
and scoring. We demonstrated a reasonable inter-rater reliability
and feel that the scale should be usable and reliable by a range of
clinicians. Further work, including the provision of illustrative video
clips available online, is planned.
Finally, although PiB-PET scanning is a putative biomarker of
Alzheimer pathology, the diagnosis ultimately requires histopathological confirmation. The inclusion of other biomarkers, such as
structural imaging or CSF protein measures may further contribute
to the diagnosis in vivo. Even so, multiple clinicopathological series
have consistently characterized semantic variant as a unique clinicopathological entity (Hodges and Patterson, 2007). Our results
support this fact: semantic cases showed the highest level of concordance between diagnostic approaches and, as was predicted,
the lowest likelihood of a positive PiB scan. In contrast, the differentiation of logopenic variant from non-fluent variant has
proven to be more elusive, in part, due to the intrinsic difficulties
in assessing key features. As a result, PiB-PET scanning may be
most useful in non-semantic cases to differentiate logopenic (PiB
positive) from non-fluent cases (PiB negative).
Brain 2011: 134; 3030–3043
| 3041
Conclusion
The recommendations for the diagnosis and classification of PPA
of the International Consensus Group are easily applicable and
appear to show a high clinical–pathological correlation, particularly
in logopenic variant. Moreover, the proposed PALS-based algorithm offers a simple bedside assessment that can readily classify
PPA cases based just on four speech and language variables. Our
findings raise questions regarding the need to include some features, such as anomia, word repetition, sentence comprehension
and reading, since patients can be reliably classified using four key
features.
Acknowledgements
The authors thank the Australian Imaging, Biomarkers and
Lifestyle (AIBL) study for providing PiB image data on the patients
with Alzheimer’s disease.
Funding
National Health and Medical Research Council (NHMRC) of
Australia Project Grant (# 630489, in part); CONICYT scholarship,
Chilean Government (C.E.L.); Australian Research Council
Federation Fellowship (#FF0776229 to J.R.H.).
Supplementary material
Supplementary material is available at Brain online.
References
Adlam AL, Patterson K, Rogers TT, Nestor PJ, Salmond CH, AcostaCabronero J, et al. Semantic dementia and fluent primary progressive
aphasia: two sides of the same coin? Brain 2006; 129: 3066–80.
Aizenstein HJ, Nebes RD, Saxton JA, Price JC, Mathis CA, Tsopelas ND,
et al. Frequent amyloid deposition without significant cognitive impairment among the elderly. Arch Neurol 2008; 65: 1509–17.
Amici S, Brambati SM, Wilkins DP, Ogar J, Dronkers NL, Miller BL, et al.
Anatomical correlates of sentence comprehension and verbal working
memory in neurodegenerative disease. J Neurosci 2007; 27: 6282–90.
Ash S, Moore P, Vesely L, Gunawardena D, McMillan C, Anderson C,
et al. Non-Fluent speech in frontotemporal lobar degeneration.
J Neurolinguistics 2009; 22: 370–83.
Bacskai BJ, Frosch MP, Freeman SH, Raymond SB, Augustinack JC,
Johnson KA, et al. Molecular imaging with Pittsburgh Compound B
confirmed at autopsy: a case report. Arch Neurol 2007; 64: 431–4.
Becker JA, Hedden T, Carmasin J, Maye J, Rentz DM, Putcha D, et al.
Amyloid-beta associated cortical thinning in clinically normal elderly.
Ann Neurol 2011; 69: 1032–42.
Cairns NJ, Bigio EH, Mackenzie IR, Neumann M, Lee VM, Hatanpaa KJ,
et al. Neuropathologic diagnostic and nosologic criteria for frontotemporal lobar degeneration: consensus of the Consortium for
Frontotemporal Lobar Degeneration. Acta Neuropathol 2007; 114:
5–22.
Canter GJ, Trost JE, Burns MS. Contrasting speech patterns in
apraxia of speech and phonemic paraphasia. Brain Lang 1985; 24:
204–22.
3042
| Brain 2011: 134; 3030–3043
Chetelat G, Villemagne VL, Bourgeat P, Pike KE, Jones G, Ames D, et al.
Relationship between atrophy and beta-amyloid deposition in
Alzheimer disease. Ann Neurol 2010; 67: 317–24.
Croot K. Diagnosis of AOS: definition and criteria. Semin Speech Lang
2002; 23: 267–80.
Deramecourt V, Lebert F, Debachy B, Mackowiak-Cordoliani MA,
Bombois S, Kerdraon O, et al. Prediction of pathology in primary progressive language and speech disorders. Neurology 2010; 74: 42–9.
Duffy JR. Motor speech disorders: substrates, differential diagnosis, and
management. Elsevier Mosby: St Louis, MO/Great Britain; 2005.
Engler H, Santillo AF, Wang SX, Lindau M, Savitcheva I, Nordberg A,
et al. In vivo amyloid imaging with PET in frontotemporal dementia.
Eur J Nucl Med Mol Imaging 2008; 35: 100–6.
Fodero-Tavoletti MT, Okamura N, Furumoto S, Mulligan RS, Connor AR,
McLean CA, et al. 18F-THK523: a novel in vivo tau imaging ligand for
Alzheimer’s disease. Brain 2011; 134: 1089–100.
Forman MS, Farmer J, Johnson JK, Clark CM, Arnold SE, Coslett HB,
et al. Frontotemporal dementia: clinicopathological correlations. Ann
Neurol 2006; 59: 952–62.
Gorno-Tempini ML, Brambati SM, Ginex V, Ogar J, Dronkers NF,
Marcone A, et al. The logopenic/phonological variant of primary progressive aphasia. Neurology 2008; 71: 1227–34.
Gorno-Tempini ML, Dronkers NF, Rankin KP, Ogar JM, Phengrasamy L,
Rosen HJ, et al. Cognition and anatomy in three variants of primary
progressive aphasia. Ann Neurol 2004; 55: 335–46.
Gorno-Tempini ML, Hillis AE, Weintraub S, Kertesz A, Mendez M,
Cappa SF, et al. Classification of primary progressive aphasia and its
variants. Neurology 2011; 76: 1006–14.
Greene JD, Patterson K, Xuereb J, Hodges JR. Alzheimer disease and
nonfluent progressive aphasia. Arch Neurol 1996; 53: 1072–8.
Grossman M, Mickanin J, Onishi K, Hughes E, D’Esposito M, Ding X-S,
et al. Progressive nonfluent aphasia: language, cognitive, and PET
measures contrasted with probable Alzheimer’s disease. J Cogn
Neurosci 1996; 8: 135–54.
Harasty JA, Halliday GM, Xuereb J, Croot K, Bennett H, Hodges JR.
Cortical degeneration associated with phonologic and semantic language impairments in AD. Neurology 2001; 56: 944–50.
Hickok G, Poeppel D. The cortical organization of speech processing. Nat
Rev Neurosci 2007; 8: 393–402.
Hodges JR, Mitchell J, Dawson K, Spillantini MG, Xuereb JH,
McMonagle P, et al. Semantic dementia: demography, familial factors
and survival in a consecutive series of 100 cases. Brain 2010; 133:
300–6.
Hodges JR, Patterson K. Nonfluent progressive aphasia and semantic
dementia: a comparative neuropsychological study. J Int
Neuropsychol Soc 1996; 2: 511–24.
Hodges JR, Patterson K. Semantic dementia: a unique clinicopathological
syndrome. Lancet Neurol 2007; 6: 1004–14.
Hodges JR, Patterson K, Oxbury S, Funnell E. Semantic dementia.
Progressive fluent aphasia with temporal lobe atrophy. Brain 1992;
115 (Pt 6): 1783–806.
Hu WT, McMillan C, Libon D, Leight S, Forman M, Lee VM, et al.
Multimodal predictors for Alzheimer disease in nonfluent primary progressive aphasia. Neurology 2010; 75: 595–602.
Ikonomovic MD, Klunk WE, Abrahamson EE, Mathis CA, Price JC,
Tsopelas ND, et al. Post-mortem correlates of in vivo PiB-PET amyloid
imaging in a typical case of Alzheimer’s disease. Brain 2008; 131:
1630–45.
Jack CR Jr, Lowe VJ, Weigand SD, Wiste HJ, Senjem ML, Knopman DS,
et al. Serial PIB and MRI in normal, mild cognitive impairment and
Alzheimer’s disease: implications for sequence of pathological events in
Alzheimer’s disease. Brain 2009; 132: 1355–65.
Josephs KA, Duffy JR, Strand EA, Whitwell JL, Layton KF, Parisi JE, et al.
Clinicopathological and imaging correlates of progressive aphasia and
apraxia of speech. Brain 2006; 129: 1385–98.
Josephs KA, Whitwell JL, Duffy JR, Vanvoorst WA, Strand EA, Hu WT,
et al. Progressive aphasia secondary to Alzheimer disease vs frontotemporal lobar degeneration pathology. Neurology 2008; 70: 25–34.
C. E. Leyton et al.
Kertesz A, Davidson W, McCabe P, Takagi K, Munoz D. Primary progressive aphasia: diagnosis, varieties, evolution. J Int Neuropsychol Soc
2003; 9: 710–9.
Kertesz A, Hudson L, Mackenzie IR, Munoz DG. The pathology and
nosology of primary progressive aphasia. Neurology 1994; 44:
2065–72.
Kertesz A, McMonagle P, Blair M, Davidson W, Munoz DG. The evolution and pathology of frontotemporal dementia. Brain 2005; 128:
1996–2005.
Klunk WE, Engler H, Nordberg A, Wang Y, Blomqvist G, Holt DP, et al.
Imaging brain amyloid in Alzheimer’s disease with Pittsburgh
Compound-B. Ann Neurol 2004; 55: 306–19.
Knibb JA, Xuereb JH, Patterson K, Hodges JR. Clinical and pathological characterization of progressive aphasia. Ann Neurol 2006; 59:
156–65.
Mesulam M, Wicklund A, Johnson N, Rogalski E, Leger GC,
Rademaker A, et al. Alzheimer and frontotemporal pathology in subsets of primary progressive aphasia. Ann Neurol 2008; 63: 709–19.
Mesulam M, Wieneke C, Rogalski E, Cobia D, Thompson C,
Weintraub S. Quantitative template for subtyping primary progressive
aphasia. Arch Neurol 2009; 66: 1545–51.
Mesulam MM. Slowly progressive aphasia without generalized dementia.
Ann Neurol 1982; 11: 592–8.
Mesulam MM. Primary progressive aphasia. Ann Neurol 2001; 49:
425–32.
Mion M, Patterson K, Acosta-Cabronero J, Pengas G, IzquierdoGarcia D, Hong YT, et al. What the left and right anterior fusiform
gyri tell us about semantic memory. Brain 2010; 133: 3256–68.
Mioshi E, Dawson K, Mitchell J, Arnold R, Hodges JR. The
Addenbrooke’s Cognitive Examination Revised (ACE-R): a brief cognitive test battery for dementia screening. Int J Geriatr Psychiatry 2006;
21: 1078–85.
Mosconi L, Rinne JO, Tsui WH, Berti V, Li Y, Wang H, et al. Increased
fibrillar amyloid-b burden in normal individuals with a family history of late-onset Alzheimer’s. Proc Natl Acad Sci USA 2010; 107:
5949–54.
Nestor PJ, Graham NL, Fryer TD, Williams GB, Patterson K, Hodges JR.
Progressive non-fluent aphasia is associated with hypometabolism
centred on the left anterior insula. Brain 2003; 126: 2406–18.
Ogar JM, Dronkers NF, Brambati SM, Miller BL, Gorno-Tempini ML.
Progressive nonfluent aphasia and its characteristic motor speech deficits. Alzheimer Dis Assoc Disord 2007; 21: S23–30.
Oh H, Mormino EC, Madison C, Hayenga A, Smiljic A, Jagust WJ.
beta-Amyloid affects frontal and posterior brain networks in normal
aging. Neuroimage 2011; 54: 1887–95.
Peters F, Collette F, Degueldre C, Sterpenich V, Majerus S, Salmon E.
The neural correlates of verbal short-term memory in Alzheimer’s disease: an fMRI study. Brain 2009; 132: 1833–46.
Pick A, Girling DM, Berrios GE. On the relationship between senile cerebral atrophy and aphasia. Hist Psych 1994; 5: 542–7.
Pick A, Girling DM, Berrios GE. On the symptomatology of left-sided
temporal lobe atrophy. Hist Psych 1997; 8: 149–59.
Pike KE, Savage G, Villemagne VL, Ng S, Moss SA, Maruff P, et al.
Beta-amyloid imaging and memory in non-demented individuals: evidence for preclinical Alzheimer’s disease. Brain 2007; 130:
2837–44.
Rabinovici GD, Furst AJ, O’Neil JP, Racine CA, Mormino EC, Baker SL,
et al. 11C-PIB PET imaging in Alzheimer disease and frontotemporal
lobar degeneration. Neurology 2007; 68: 1205–12.
Rabinovici GD, Jagust WJ, Furst AJ, Ogar JM, Racine CA, Mormino EC,
et al. Abeta amyloid and glucose metabolism in three variants of primary progressive aphasia. Ann Neurol 2008; 64: 388–401.
Resnick SM, Sojkova J, Zhou Y, An Y, Ye W, Holt DP, et al. Longitudinal
cognitive decline is associated with fibrillar amyloid-beta measured by
[11C]PiB. Neurology 2010; 74: 807–15.
Rogalski E, Cobia D, Harrison TM, Wieneke C, Thompson CK,
Weintraub S, et al. Anatomy of language impairments in primary progressive aphasia. J Neurosci 2011; 31: 3344–50.
Validation of PPA classification
Rogers TT, Lambon Ralph MA, Garrard P, Bozeat S, McClelland JL,
Hodges JR, et al. Structure and deterioration of semantic memory: a
neuropsychological and computational investigation. Psychol Rev
2004; 111: 205–35.
Rohrer JD, Knight WD, Warren JE, Fox NC, Rossor MN, Warren JD.
Word-finding difficulty: a clinical analysis of the progressive aphasias.
Brain 2008; 131: 8–38.
Rohrer JD, Ridgway GR, Crutch SJ, Hailstone J, Goll JC, Clarkson MJ,
et al. Progressive logopenic/phonological aphasia: erosion of the language network. Neuroimage 2009a; 49: 984–93.
Rohrer JD, Rossor MN, Warren JD. Apraxia in progressive nonfluent
aphasia. J Neurol 2010a; 257: 569–74.
Rohrer JD, Rossor MN, Warren JD. Syndromes of nonfluent primary
progressive aphasia: a clinical and neurolinguistic analysis. Neurology
2010b; 75: 603–10.
Rohrer JD, Warren JD, Modat M, Ridgway GR, Douiri A, Rossor MN,
et al. Patterns of cortical thinning in the language variants of frontotemporal lobar degeneration. Neurology 2009b; 72: 1562–9.
Rosen HJ, Gorno-Tempini ML, Goldman WP, Perry RJ, Schuff N,
Weiner M, et al. Patterns of brain atrophy in frontotemporal dementia
and semantic dementia. Neurology 2002; 58: 198–208.
Rowe CC, Ellis KA, Rimajova M, Bourgeat P, Pike KE, Jones G, et al.
Amyloid imaging results from the Australian Imaging, Biomarkers and
Lifestyle (AIBL) study of aging. Neurobiol Aging 2010; 31: 1275–83.
Sapolsky D, Bakkour A, Negreira A, Nalipinski P, Weintraub S,
Mesulam MM, et al. Cortical neuroanatomic correlates of symptom
severity in primary progressive aphasia. Neurology 2010; 75: 358–66.
Scheinin NM, Aalto S, Koikkalainen J, Lotjonen J, Karrasch M,
Kemppainen N, et al. Follow-up of [11C]PIB uptake and brain
volume in patients with Alzheimer disease and controls. Neurology
2009; 73: 1186–92.
Brain 2011: 134; 3030–3043
| 3043
Schroeter ML, Raczka K, Neumann J, von Cramon DY. Neural networks
in frontotemporal dementia–a meta-analysis. Neurobiol Aging 2008;
29: 418–26.
Snowden J, Neary D, Mann D. Frontotemporal lobar degeneration: clinical and pathological relationships. Acta Neuropathol 2007; 114: 31–8.
Snowden JS, Goulding PJ, Neary D. Semantic dementia: a form of circumscribed cerebral atrophy. Behav Neurol 1989; 2: 167–182.
Snowden JS, Neary D, Mann DM, Goulding PJ, Testa HJ. Progressive
language disorder due to lobar atrophy. Ann Neurol 1992; 31:
174–83.
Turner RS, Kenyon LC, Trojanowski JQ, Gonatas N, Grossman M.
Clinical, neuroimaging, and pathologic features of progressive nonfluent aphasia. Ann Neurol 1996; 39: 166–73.
Tyrrell PJ, Warrington EK, Frackowiak RS, Rossor MN. Heterogeneity in
progressive aphasia due to focal cortical atrophy. A clinical and PET
study. Brain 1990; 113 (Pt 5): 1321–36.
Villemagne VL, Pike KE, Chetelat G, Ellis KA, Mulligan RS, Bourgeat P,
et al. Longitudinal assessment of Ab and cognition in aging and
Alzheimer disease. Ann Neurol 2011; 69: 181–92.
Warrington EK. The selective impairment of semantic memory. Q J Exp
Psychol 1975; 27: 635–57.
Weintraub S, Mesulam MM, Wieneke C, Rademaker A, Rogalski EJ,
Thompson CK. The northwestern anagram test: measuring sentence
production in primary progressive aphasia. Am J Alzheimers Dis Other
Demen 2009; 24: 408–16.
Wilson SM, Henry ML, Besbris M, Ogar JM, Dronkers NF, Jarrold W,
et al. Connected speech production in three variants of primary progressive aphasia. Brain 2010; 133: 2069–88.
Ziegler W. Apraxia of speech. Handb Clin Neurol 2008; 88: 269–85.