Classification and pathology of primary progressive aphasia
Jennifer M. Harris, BSc Claire Gall, MB, ChB Jennifer C. Thompson, PhD Anna M.T. Richardson, MB, ChB David Neary, MD Daniel du Plessis, MB, ChB Piyali Pal, MD David M.A. Mann, PhD Julie S. Snowden, PhD Matthew Jones, MD
Correspondence to Dr. Jones: [email protected]
ABSTRACT
Objective: We aimed to determine the extent to which patients with progressive language impairment conform to 2011 primary progressive aphasia (PPA) classification and to examine clinicopathologic correlations within PPA variants.
Methods: Sixty-two consecutive patients with pathologically confirmed dementia who presented clinically with aphasia were identified. Patients with insufficient clinical information were excluded. PPA classifications were applied to anonymized clinical data taken from patients’ initial assessment by raters who were blinded to clinical and pathologic diagnosis.
Results: The final cohort comprised 52 patients, 30 of whom met basic PPA criteria. Twenty-five patients met one of the 3 PPA classifications (13 logopenic, 8 nonfluent/agrammatic, and 4 semantic). Five patients did not meet the criteria for any of the PPA variants. All patients who met semantic variant PPA and 75% of patients who met nonfluent/agrammatic variant PPA classifications had frontotemporal lobar degeneration spectrum pathology. Pathologies were heterogeneous in patients who met logopenic variant PPA criteria (46% Alzheimer disease [AD], 8% AD mixed with dementia with Lewy bodies, 23% frontotemporal lobar degeneration, and 23% other). Conclusion: The 2011 PPA recommendations classify a large proportion of patients who meet basic PPA criteria. However, some patients had aphasic syndromes that could not be classified, suggesting that the 2011 recommendations do not cover the full range of PPA variants. Classification of semantic variant PPA provides a good prediction of underlying pathology. Classification of logopenic variant does not successfully differentiate PPA due to AD from PPA due to other pathologies. Neurology® 2013;81:1832–1839 GLOSSARY AD 5 Alzheimer disease; AOS 5 apraxia of speech; CBD 5 corticobasal degeneration; C9orf72 5 chromosome 9 open reading frame 72; FTLD 5 frontotemporal lobar degeneration; GRN 5 progranulin; lvPPA 5 logopenic variant of primary progressive aphasia; MAPT 5 microtubule-associated protein tau; nfvPPA 5 nonfluent/agrammatic variant of primary progressive aphasia; PPA 5 primary progressive aphasia; svPPA 5 semantic variant of primary progressive aphasia; TDP-43 5 transactive response DNA-binding protein 43.
Primary progressive aphasia (PPA) is clinically heterogeneous. Recommendations proposed in 2011 to standardize classification1 identify 3 variants: nonfluent/agrammatic (nfvPPA), semantic (svPPA), and logopenic (lvPPA). The 2011 classification system makes no a priori assumptions about the underlying pathologic substrate. In this respect, it differs from diagnostic criteria for other neurodegenerative disorders2–4 and from the 1998 frontotemporal lobar degeneration (FTLD) criteria, whereby classifications of progressive nonfluent aphasia and semantic dementia were intended to predict FTLD pathology.5 Significant pathologic heterogeneity has been reported in PPA.6 However, clinicopathologic studies were conducted before the publication of the 2011 recommendations, so it is possible that the heterogeneity reflects variable methods of patient clinical subtype classification. Moreover, some patients previously classified as “nonfluent” might more accurately conform to the profile of lvPPA rather than nfvPPA. There is a need for systematic examination of the Supplemental data at www.neurology.org From the Manchester Academic Health Sciences Centre (J.M.H., C.G., J.C.T., A.M.T.R., D.N., D.d.P., P.P., D.M.A.M., J.S.S., M.J.), Cerebral Function Unit, Greater Manchester Neuroscience Centre, Salford Royal NHS Foundation Trust, Salford; and Institute of Brain, Behaviour and Mental Health (J.M.H., D.N., D.M.A.M., J.S.S., M.J.), University of Manchester, UK. Go to Neurology.org for full disclosures. Funding information and disclosures deemed relevant by the authors, if any, are provided at the end of the article. 1832
© 2013 American Academy of Neurology
ª"NFSJDBO"DBEFNZPG/FVSPMPHZ6OBVUIPSJ[FESFQSPEVDUJPOPGUIJTBSUJDMFJTQSPIJCJUFE
relationship between clinical variants of PPA as specified in the 2011 recommendations and their pathologic substrate. The 3 PPA variants do not appear to account for the full range of PPA syndromes,7,8 and patients may fulfill criteria for more than one variant.8 Whether these “unclassifiable” forms of PPA share a common clinical syndrome and underlying pathology remains to be determined. In this study, we aimed to establish, in a pathologically confirmed cohort, the extent to which patients presenting with language impairment fulfilled 2011 PPA classification criteria and to identify reasons for nonfulfillment. We evaluated clinicopathologic relationships to determine the degree to which clinical phenotypes predict underlying pathology. METHODS Sixty-two patients who presented clinically with progressive language impairment were identified from a larger cohort of patients who had been assessed in a specialist unit for early-onset dementia (the Cerebral Function Unit, Greater Manchester Neuroscience Centre) and subsequently undergone postmortem neuropathologic brain tissue examination. All brains were obtained with ethical approval and informed written consent. As part of their initial diagnostic assessment, all patients
Figure 1
had undergone neuropsychological examination using the Manchester Neuropsychological Profile,9–11 which includes systematic assessment of language abilities and documentation of language errors (see appendix e-1 on the Neurology® Web site at www.neurology.org). Additional published language tests had been administered, although these were variable across patients (appendix e-1). Copies of clinic letters and neuropsychological reports had all references to presumptive clinical diagnoses and personal identifiable information expunged. One of 2 raters, who were blinded to clinical and pathologic diagnoses and who had not contributed to the patients’ clinical care, rated the neuropsychological and clinical data against the 2011 PPA classification recommendations.1 Data from the patients’ initial diagnostic assessment were used. It was determined whether they met overall inclusion and exclusion criteria for PPA, i.e., basic PPA criteria, and specific criteria for one of the variants. Missing data were managed in accord with a previous study.12 Twenty-four patients’ case notes were assessed by both raters to enable evaluation of interrater reliability.
Data analysis. Data analysis was performed with SPSS statistics version 20 (IBM Corp., Armonk, NY). A significance level of p , 0.05 was used. Comparisons of demographic data in different pathologic groups were conducted using t tests, x2 test of associations, or Fisher exact test, as appropriate. Interrater reliability was evaluated using k statistics. RESULTS Patient cohort, demographics, and genetics.
The selection of cases is shown in figure 1. Ten patients were excluded from the original 62-patient cohort because of insufficient clinical data to rate against the
Case selection
AD 5 Alzheimer disease; FTLD 5 frontotemporal lobar degeneration; PPA 5 primary progressive aphasia. Other refers to pathology other than FTLD or AD. FTLD pathologic classification according to Mackenzie et al.13,14 Neurology 81
November 19, 2013
1833
ª"NFSJDBO"DBEFNZPG/FVSPMPHZ6OBVUIPSJ[FESFQSPEVDUJPOPGUIJTBSUJDMFJTQSPIJCJUFE
Table 1
Demographics of FTLD and non-FTLD cases Pathologic finding All (n 5 52)
FTLD (n 5 29)
Non-FTLD (n 5 22)
Test statistic (df) p
Range of first visit dates
3/1/1983 to 9/7/2010 3/1/1983 to 1/19/2010 6/23/1983 to 9/7/2010
Age at onset, y, mean (SD)
60.8 (8.1)
61.0 (9.0)
60.7 (7.0)
t 5 0.118 (49)
0.91
Age at death, y, mean (SD) (1 missing)
69.8 (6.9)
69.7 (6.5)
70.0 (7.8)
t 5 20.160 (48)
0.87
3.2 (2.1)
3.4 (2.4)
2.9 (1.7)
t 5 0.782 (49)
0.44
Family history in first-degree relative, % (n) (1 missing)
35.3 (18)
44.8 (13)
23.8 (5)
x 5 2.335 (1)
0.13
Sex, % male (n)
55.8 (29)
55.2 (16)
59.1 (13)
x2 5 0.078 (1)
1.00
Duration of illness at presentation, y, mean (SD)
2
Abbreviation: FTLD 5 frontotemporal lobar degeneration. One patient had mixed FTLD and non-FTLD pathology and is therefore only included in the “all” category. Sex and family history were compared with x2 test of associations and other demographic data were compared using an independent sample t test.
criteria. The final study cohort of 52 patients (29 male and 23 female) consisted predominantly of early-onset cases (table 1); 35.3% of patients had a family history of a related disorder in at least one first-degree relative. Twenty-nine had FTLD, 22 non-FTLD pathologies, and 1 patient had mixed FTLD and non-FTLD pathology. Progranulin (GRN), microtubule-associated protein tau (MAPT), and chromosome 9 open reading frame 72 (C9orf72) genetic results were available for 21 of 29 FTLD cases (72.4%). From the FTLD cases, none had a C9orf72 expansion (3 missing), 5 had GRN mutations (4 missing including 2 transactive response DNA-binding protein 43 [TDP-43] type A), and 3 had MAPT mutations. APOE status was available for 38 cases (missing: 9 FTLD, 4 non-FTLD, and 1 mixed FTLD and non-
Table 2
PPA subtype Nonfluent
Pathologic relationships in PPA variants Total no. of patients 8
No. of patients in each pathologic group 6 FTLD (2 TDP-43 type A—1 GRN mutation, 1 Pick disease, 3 corticobasal degeneration) 1 CJD 1 mixed DLB, AD, and vasculitis
Semantic Logopenic
4 13
4 FTLD (TDP-43 type C) 3 FTLD (TDP-43 type A—1 GRN mutation, other 2 genetic information not available) 7 AD (including 1 mixed AD and DLB) 1 DLB 1 CJD 1 cerebrovascular disease
Unclassified
5
3 FTLD (TDP-43 type A—2 GRN mutation) 2 AD (including 1 mixed AD and cerebrovascular disease)
Abbreviations: AD 5 Alzheimer disease; CJD 5 Creutzfeldt-Jakob disease; DLB 5 dementia with Lewy bodies; FTLD 5 frontotemporal lobar degeneration; GRN 5 progranulin; PPA 5 primary progressive aphasia; TDP-43 5 transactive response DNA-binding protein 43. FTLD pathologic classification according to Mackenzie et al13,14; TDP-43 subtyping according to Mackenzie et al.15 1834
Neurology 81
FTLD). The APOE status of FTLD cases are as follows: 2 e2e3, 10 e3e3, 7 e3e4, and 1 e4e4. The APOE status of non-FTLD cases are as follows: 2 e2e3, 9 e3e3, 1 e2e4, 5 e3e4, and 1 e4e4. Interrater reliability. There was 100.0% agreement between raters for meeting basic PPA criteria (k 5 1.000, p , 0.001). The percentage agreement for each PPA variant was as follows: nfvPPA 83.3% (k 5 0.429, p 5 0.007), svPPA 100.0% (k 5 1.000, p , 0.001), and lvPPA 95.8% (k 5 0.917, p , 0.001). The lower level of agreement for nfvPPA was attributable to inconsistencies between raters in evaluating the core feature, “Effortful, halting speech with inconsistent speech sound errors and distortions (apraxia of speech)” (table e-1). Accordingly, if this feature was rated as present or there was uncertainty about its presence, it was checked by 2 experienced clinicians (M.J. and J.T.) and a consensus was reached. Patients who met basic PPA clinical criteria. Thirty patients met basic criteria for PPA. Eight fulfilled criteria for nfvPPA, 4 for svPPA, and 13 for lvPPA. Five patients did not meet criteria for a specific PPA variant, henceforth referred to as PPA-unclassified. All patients with svPPA and the majority of patients with nfvPPA had FTLD pathology13,14 (table 2). Conversely, most cases of lvPPA had non-FTLD pathology, with Alzheimer disease (AD) pathology accounting for the largest proportion of cases. Four patients who met basic PPA criteria had genetic mutations (table 2). nfvPPA: Agrammatism vs apraxia of speech. nfvPPA can be subdivided because only 1 of 2 core features is required to reach a diagnosis: agrammatism or apraxia of speech (AOS). Of the 8 patients with nfvPPA, 2 had prominent agrammatism without AOS. Both had TDP-43 type A pathology.15 One of these patients had a GRN mutation. Neither patient had motor signs at presentation. Five had prominent AOS, 3 of whom had tau pathology (2 corticobasal degeneration [CBD], 1 Pick).
November 19, 2013
ª"NFSJDBO"DBEFNZPG/FVSPMPHZ6OBVUIPSJ[FESFQSPEVDUJPOPGUIJTBSUJDMFJTQSPIJCJUFE
One patient with CBD pathology had markedly dysprosodic speech with evidence of AOS without generalized buccofacial or limb apraxia. The other patient with CBD pathology presented to the clinic with a 5-year history, beginning with circumscribed difficulties with word-finding for the first 3 years, followed by emerging difficulties using her right arm and leg. The patient with Pick disease pathology had profound buccofacial and AOS without limb apraxia. Three years after symptom onset, mild problems using her limbs emerged with crude and impersistent actions. Of the remaining 2 patients, one had mixed pathology of AD, dementia with Lewy bodies, and vasculitis, and one had Creutzfeldt-Jakob disease (described in detail elsewhere16). The case with mixed pathology presented with an atypical pattern of deficits. His speech was nonfluent with elongation of syllables and loss of prosody, which was interpreted as AOS. There was evidence of impairments in syntactic comprehension, repetition, confrontation naming, spelling, reading, and calculation. He also showed some difficulties with visual construction and mild executive impairments. There was no generalized buccofacial or limb apraxia. He was followed up for 2 years, during which time he became parkinsonian and mute with some behavioral disinhibition. One patient with both AOS and agrammatism had CBD pathology; he also displayed limb apraxia at presentation. Unclassified PPA. Two patients with PPA-unclassified presented with profound, circumscribed anomia, characterized by impaired phonologic access. They developed limb apraxia over the course of the illness. Both patients had core features of lvPPA (one “impaired single-word retrieval in spontaneous speech and naming,” the second “impaired repetition of sentences and phrases”) and svPPA (one “impaired confrontation naming,” the second both core features). The second patient who had both core features of svPPA had no supportive features. Both patients had the supportive features of nfvPPA “impaired comprehension of syntactically complex sentences” and “spared object knowledge,” but neither patient had core features of nfvPPA at presentation. At follow-up, the first patient met criteria for both nfvPPA and lvPPA, whereas the second patient did not meet criteria for a PPA subtype. Both of these anomic patients had FTLD TDP-43 type A pathology and mutations in the GRN gene. One has been described in detail elsewhere.17,18 A third PPA-unclassified patient presented with fluent anomic speech with semantic errors and impaired syntactic understanding. At his initial visit, he showed impaired confrontation naming and impaired single-word retrieval, therefore having one core feature of both svPPA and lvPPA. However, he
did not have any other core features of svPPA or lvPPA. Motor and syntactic elements of his speech were spared, thus he exhibited neither of the core features of nfvPPA. He had FTLD TDP-43 type A pathology. Another unclassified patient (age at onset 72 years) presented with anomia with moderate semantic and phonologic impairment. In addition, he developed some behavioral features of behavioral variant FTD. He had both core features of svPPA, but he did not exhibit impaired object knowledge and his repetition was not spared. Thereby, he lacked sufficient supportive features to fulfill the criteria. During follow-up, he met criteria for svPPA. He had AD pathology, with additional vascular pathology in the temporal and parietal cortex. The final unclassified patient presented with a reported 5-year history of language disorder. He showed all supportive features of lvPPA, but despite the long duration of illness, he exhibited preserved sentence repetition and confrontation naming so did not fulfill lvPPA criteria. He subsequently fulfilled the criteria at follow-up 9 years after symptom onset. He showed AD pathology. Patients presenting with aphasia who did not meet PPA inclusion/exclusion criteria. Twenty-two patients pre-
sented with a prominent language deficit but did not meet basic PPA criteria, largely because of the presence of an exclusion criterion. Interestingly, many of these patients could be ascribed a PPA subtype, based on language features alone. Table 3 shows 2011 PPA subtype classification of these patients, the basis for exclusion, underlying pathology, and genetic mutations where present. A significant association was found between the presence of prominent initial behavioral disturbance and FTLD pathology (p 5 0.03). Conversely, the presence of prominent initial episodic memory, visual memory, and visuoperceptual impairments was associated with AD pathology (p 5 0.001). Frequency of individual features of PPA clinical classification across patient groups. Both the core fea-
tures and several of the supporting features of lvPPA were present in the majority of the 52 patients presenting clinically with aphasia across pathologic groups (figure 2). In contrast, both of the core features of nfvPPA and impaired single-word comprehension, one of the core features of svPPA, were present in very few cases. The majority of patients who had either of the core features of nfvPPA had FTLD pathology. DISCUSSION This study found that 83.3% of patients who met basic criteria for PPA could be ascribed a PPA subtype using the 2011 classification Neurology 81
November 19, 2013
1835
ª"NFSJDBO"DBEFNZPG/FVSPMPHZ6OBVUIPSJ[FESFQSPEVDUJPOPGUIJTBSUJDMFJTQSPIJCJUFE
Table 3
Relationships among exclusion criteria, pathologic diagnosis, and 2011 PPA classification Pathologic diagnosis (and genetic mutation where applicable)
2011 PPA classification
No. of patients
AD
lvPPA
5
AD
Unclassified
2
DLB (1AD)
Unclassified
1
Pick disease
nfvPPA
1
Pick disease
svPPA
1
FTDP-17T; MAPT mutation
svPPA
2
Mixed
svPPA
1
TDP type C
svPPA
1
TDP type C
svPPA and lvPPA
2
Prominent initial motor disorder
TDP type B
nfvPPA
2
Prominent initial motor and behavioral disturbance
FTDP-17T; MAPT mutation
svPPA
1
TDP type A; GRN mutation
lvPPA
1
CBD
nfvPPA
1
Pick disease
Unclassified
1
Exclusion criterion Prominent initial episodic memory, visual memory, and visuoperceptual impairments
Prominent initial behavioral disturbance
None (but do not meet all inclusion criteria)
Abbreviations: AD 5 Alzheimer disease; CBD 5 corticobasal degeneration; DLB 5 dementia with Lewy bodies; FTDP-17T 5 frontotemporal dementia with parkinsonism-17 with tau inclusions; GRN 5 progranulin; lvPPA 5 logopenic variant of PPA; MAPT 5 microtubule-associated protein tau; nfvPPA 5 nonfluent/agrammatic variant of PPA; PPA 5 primary progressive aphasia; svPPA 5 semantic variant of PPA; TDP 5 transactive response DNA-binding protein.
recommendations. The clinicopathologic associations seen in the PPA subtypes are discussed below. nfvPPA and svPPA were most frequently associated with FTLD pathologies, specifically FTLD TDP-43 type A and tau pathologies in nfvPPA and TDP-43 type C in svPPA. AD was the most common pathology in lvPPA. These findings are in keeping with previous biomarker19,20 and clinicopathologic studies.11,21,22 Within the nfvPPA group, we found patients with both prominent AOS and prominent agrammatism. The 2 with prominent agrammatism had TDP-43 type A pathology. The patients with prominent AOS were more varied. Two were highly atypical cases clinically: one had Creutzfeldt-Jakob disease and in life had been diagnosed as such because of rapid clinical deterioration, and the other had a pathologically mixed disorder of AD, dementia with Lewy bodies, and vasculitis as described in the results section. Three other patients with speech apraxia had tau pathologies. This finding supports the argument that speech apraxia may be a predictor of FTLD-tau pathology.23–25 Moreover, these cases highlight the fact that nfvPPA is itself not homogeneous, and encompasses etiologically significant subvariants. lvPPA showed the poorest pathologic specificity, being linked not only to AD but also to a variety of other pathologies. This may reflect lack of specificity and discriminating value within the diagnostic features. It has been pointed out that 3 of the 4 1836
Neurology 81
supporting features of lvPPA correspond to the sparing of features characteristic of nfvPPA and svPPA,8 rather than the presence of distinct features. The same factor analytic study8 showed that “sentence repetition,” a core feature of lvPPA, clustered with the features of nfvPPA. In keeping with lack of specificity, most of the features of lvPPA were reported in the majority of patients with PPA (figure 2). Two patients, who met criteria respectively for nfvPPA and lvPPA, had a pathologic diagnosis of Creutzfeldt-Jakob disease, highlighting the fact that Creutzfeldt-Jakob disease may present as a circumscribed aphasia. Those patients had a rapid onset and progression of symptoms and would not have met PPA diagnostic criteria, had the diagnostic rule of “2-years of progressive, circumscribed aphasia,” originally suggested by Weintraub and Mesulam,26 been applied. The limitation of the 2-year rule is that it precludes early diagnosis. Nevertheless, if future PPA classifications hope to improve predictability of pathology, then rate of symptom progression requires consideration. Five patients with PPA did not meet criteria for any subtype and were termed PPA-unclassified. It has been speculated that patients with mild deficits may have insufficient features to be diagnosed with a PPA variant.1 In line with this view, one patient with AD pathology fulfilled criteria for lvPPA only on follow-up. The other 4 patients had marked
November 19, 2013
ª"NFSJDBO"DBEFNZPG/FVSPMPHZ6OBVUIPSJ[FESFQSPEVDUJPOPGUIJTBSUJDMFJTQSPIJCJUFE
Figure 2
Frequency of features from PPA classification recommendations and pathologic relationships
Features on the y-axis in bold refer to core criteria; those in standard text are supporting criteria. FTLD pathologic classification according to Mackenzie et al.13,14 AD 5 Alzheimer disease; FTLD 5 frontotemporal lobar degeneration; lvPPA 5 logopenic variant of PPA; nfvPPA 5 nonfluent/agrammatic variant of PPA; PPA 5 primary progressive aphasia; svPPA 5 semantic variant of PPA; TDP 5 transactive response DNA-binding protein. Other refers to pathology other than FTLD or AD.
language impairment at presentation, so their nonclassification cannot be attributed to stage of illness. In each, the prevailing symptom was anomia. An anomic PPA subgroup has been proposed to account for such cases.7 It is worthy of comment, however, that anomia per se appears poorly predictive of underlying pathology. In the present series, 3 anomic patients had FTLD TDP-43 type A pathology, in 2 cases associated with GRN gene mutations, whereas the other had AD pathology. At follow-up, there was overlap between syndromes, with one patient meeting criteria for both nfvPPA and lvPPA. The PPAunclassified patients reinforce the view that more than 3 PPA syndromes exist. In particular, as noted previously,27 patients with GRN mutations may exhibit a PPA syndrome that does not conform to current classifications. Unsurprisingly, the presence of “prominent initial episodic memory, visual memory, and visuoperceptual impairments” predicted AD pathology. Importantly, those patients, with classic AD features, also typically
exhibited language features consistent with the lvPPA phenotype. This finding contrasts with reports28 that language characteristics in early-stage AD differ from those of lvPPA. The disparity may relate to differences in patient population: in the latter study, patients were older and presented a classic profile dominated by memory loss. By contrast, early-onset AD is more likely to present as a constellation of deficits in language, working and episodic memory, and perceptuospatial function.10 It is possible that the clinical similarities between lvPPA and AD are limited to early-onset AD. Most study patients who conformed to svPPA or nfvPPA language variants but did not fulfill basic criteria for PPA because of early behavioral disturbance had FTLD pathology. Behavioral changes in progressive nonfluent aphasia and semantic dementia are well documented, so overlapping PPA/behavioral syndromes may therefore be expected.29–32 Indeed, disinhibition and preference for sweet food have been found to be useful in distinguishing FTLD-PPA from Neurology 81
November 19, 2013
1837
ª"NFSJDBO"DBEFNZPG/FVSPMPHZ6OBVUIPSJ[FESFQSPEVDUJPOPGUIJTBSUJDMFJTQSPIJCJUFE
AD-PPA.33 Arguably, behavioral disturbances may cause more distress to, and difficulties for, carers than language symptoms, so may be reported as the “main cause of functional impairment” even when language disorder is the most striking feature on neuropsychological testing. Our data indicate that the presence of behavioral disorders in aphasic patients is highly suggestive of underlying FTLD pathology. Moreover, they suggest a continuum between the behavioral disorder of FTD and the semantic and nonfluent aphasias. There is a need for PPA classifications to recognize these mixed syndromes. This study has limitations; principally, although by necessity, it was conducted through retrospective blinded case note review. The 2011 PPA recommendations are intended for prospective use. Retrospective evaluation is dependent on the descriptions of linguistic and other cognitive features recorded by clinicians at the time. However, all PPA patients in our study had undergone detailed neuropsychological assessment, including systematic assessment of expressive and receptive language skills. If there were insufficient clinical data, then patients were excluded from further analysis (n 5 10). This study only considered the clinical aspects of the classification system. We have not included imaging data principally because there were some cases for which only basic CT scans were available or no scan data were available at all because of the long time scale over which patients were seen. Another limitation is that vascular and psychiatric disorders were underrepresented in the sample, and these disorders may mimic PPA. It would be of value to replicate the study in an independent pathologic cohort that includes a larger number of patients with vascular and psychiatric disorders. We were able to classify the majority of patients using the 2011 PPA recommendations, attesting to their clinical utility. However, the existence of “PPA-unclassified” patients, even within a relatively small sample of 30 patients, suggests that the tripartite classification of PPA is insufficient. The svPPA classification was strongly associated with TDP-43 type C pathology, whereas the lvPPA classification was associated with diverse pathologies. The nfvPPA subtype was rarely caused by Alzheimer pathology, but a broad range of other pathologies were seen. The findings suggest that although the current classification recommendations can be used to predict pathology in svPPA, they have limited utility in predicting specific pathology in nfvPPA and lvPPA. Finally, the separation of FTLD criteria from the 2011 PPA classification raises the problem of overlapping or mixed syndromes. Our findings suggest that the behavioral and visual memory/visual perceptual features that are currently exclusions in the PPA 1838
Neurology 81
classification recommendations could be used to improve prediction of the underlying pathologic diagnosis. AUTHOR CONTRIBUTIONS Jennifer Harris: acquisition of data (rating and coding data), interpretation of data and statistical analysis, drafting/revising the manuscript. Dr. Claire Gall: acquisition of data (rating and coding data), drafting/ revising the manuscript. Dr. Jennifer Thompson: study conception and design, acquisition of data (redacting clinical data), revising the manuscript. Dr. Anna Richardson: acquisition of data (original clinical data), revising the manuscript. Prof. David Neary: acquisition of data (design of clinical history-taking), revising the manuscript. Dr. Daniel du Plessis, Dr. Piyali Pal, and Prof. David Mann: acquisition and interpretation of neuropathologic data, revising the manuscript. Prof. Julie Snowden: study conception and design, analysis and interpretation of data, drafting/revising the manuscript. Dr. Matthew Jones: study conception and design, analysis and interpretation of data, drafting/revising the manuscript, study supervision and coordination.
ACKNOWLEDGMENT The authors thank all the neurologists and neuropsychologists who have worked at the Cerebral Function Unit and contributed to the original clinical evaluation of patients included in this study, and the clinical colleagues from the North West Region of UK who referred patients. The authors thank Steve Chew-Graham for his role as coordinator of the Manchester brain donation scheme, which is supported financially by the Alzheimer’s Research Trust and Alzheimer’s Society through the Brains for Dementia Research Initiative.
STUDY FUNDING No targeted funding reported.
DISCLOSURE J. Harris, C. Gall, J. Thompson, and A. Richardson report no disclosures. D. Neary receives royalties for Neuroanatomy (Crossman A, Neary D. Elsevier Churchill Livingstone, 1st edition 1995, 2nd edition 2000, 3rd edition 2005, 4th edition 2010). D. du Plessis and P. Pal report no disclosures. D. Mann receives Medical Research Council and Wellcome Trust support through the Neurodegeneration Programme. The Manchester brain donation scheme and the work of the Manchester Brain Bank is supported by Alzheimer’s Research UK and the Alzheimer’s Society through the Brains for Dementia Research Initiative. J. Snowden has received travel expenses from the World Federation of Neurology for teaching in Kolkata India, 2012. She is a member of the Editorial Advisory Board of 1) Neuropsychologia, 2006 to present; 2) Journal of the International Neuropsychological Society, 2011 to present; 3) Amyotrophic Lateral Sclerosis and Frontotemporal Degeneration, 2012 to present; and 4) Neuropsychology, 2013 to present. M. Jones reports no disclosures. Go to Neurology.org for full disclosures.
Received May 1, 2013. Accepted in final form August 19, 2013. REFERENCES 1. Gorno-Tempini ML, Hillis AE, Weintraub S, et al. Classification of primary progressive aphasia and its variants. Neurology 2011;76:1006–1014. 2. Rascovsky K, Hodges JR, Knopman D, et al. Sensitivity of revised diagnostic criteria for the behavioural variant of frontotemporal dementia. Brain 2011;134:2456–2477. 3. McKhann G, Knopman DS, Chertkow H, et al. The diagnosis of dementia due to Alzheimer’s disease: recommendations from the National Institute on Aging–Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 2011;7:263–269.
November 19, 2013
ª"NFSJDBO"DBEFNZPG/FVSPMPHZ6OBVUIPSJ[FESFQSPEVDUJPOPGUIJTBSUJDMFJTQSPIJCJUFE
4.
5.
6. 7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17. 18.
McKeith IG, Dickson DW, Lowe J, et al. Diagnosis and management of dementia with Lewy bodies: third report of the DLB Consortium. Neurology 2005;65:1863–1872. Neary D, Snowden J, Gustafson L, et al. Frontotemporal lobar degeneration: a consensus on clinical diagnostic criteria. Neurology 1998;51:1546–1554. Grossman M. Primary progressive aphasia: clinicopathological correlations. Nat Rev Neurol 2010;6:88–97. Mesulam M, Wieneke C, Thompson C, Rogalski E, Weintraub S. Quantitative classification of primary progressive aphasia at early and mild impairment stages. Brain 2012;135:1537–1553. Sajjadi SA, Patterson K, Arnold RJ, Watson PC, Nestor PJ. Primary progressive aphasia: a tale of two syndromes and the rest. Neurology 2012;78:1670–1677. Thompson JC, Stopford CL, Snowden J, Neary D. Qualitative neuropsychological performance characteristics in frontotemporal dementia and Alzheimer’s disease. J Neurol Neurosurg Psychiatry 2005;76:920–927. Snowden J, Stopford CL, Julien C, et al. Cognitive phenotypes in Alzheimer’s disease and genetic risk. Cortex 2007;43:835–845. Snowden J, Thompson JC, Stopford CL, et al. The clinical diagnosis of early-onset dementias: diagnostic accuracy and clinicopathological relationships. Brain 2011; 134:2478–2492. Harris J, Gall C, Thompson J, et al. Sensitivity and specificity of FTDC criteria for behavioral variant frontotemporal dementia. Neurology 2013;80:1881–1887. Mackenzie IR, Neumann M, Bigio EH, et al. Nomenclature for neuropathologic subtypes of frontotemporal lobar degeneration: consensus recommendations. Acta Neuropathol 2009;117:15–18. Mackenzie IR, Neumann M, Bigio EH, et al. Nomenclature and nosology for neuropathologic subtypes of frontotemporal lobar degeneration: an update. Acta Neuropathol 2010;119:1–4. Mackenzie IR, Neumann M, Baborie A, et al. A harmonized classification system for FTLD-TDP pathology. Acta Neuropathol 2011;122:111–113. Kobylecki C, Thompson JC, Jones M, et al. Sporadic Creutzfeldt-Jakob disease presenting as progressive nonfluent aphasia with speech apraxia. Alzheimer Dis Assoc Disord Epub 2012 Jun 22. Snowden J, Neary D. Progressive anomia with preserved oral spelling and automatic speech. Neurocase 2003;9:27–43. Snowden JS, Pickering-Brown SM, Du Plessis D, et al. Progressive anomia revisited: focal degeneration associated with progranulin gene mutation. Neurocase 2007; 13:366–377.
19.
20.
21.
22. 23.
24.
25.
26. 27.
28.
29.
30.
31.
32.
33.
Rabinovici GD, Jagust WJ, Furst AJ, et al. Abeta amyloid and glucose metabolism in three variants of primary progressive aphasia. Ann Neurol 2008;64:388–401. Leyton CE, Villemagne VL, Savage S, et al. Subtypes of progressive aphasia: application of the International Consensus Criteria and validation using beta-amyloid imaging. Brain 2011;134:3030–3043. Snowden J, Neary D, Mann DM. Frontotemporal lobar degeneration: clinical and pathological relationships. Acta Neuropathol 2007;114:31–38. Mesulam M. Primary progressive aphasia pathology. Ann Neurol 2008;63:124–125. Josephs KA, Boeve BF, Duffy JR, et al. Atypical progressive supranuclear palsy underlying progressive apraxia of speech and nonfluent aphasia. Neurocase 2005;11:283–296. Josephs KA, Duffy JR, Strand EA, et al. Characterizing a neurodegenerative syndrome: primary progressive apraxia of speech. Brain 2012;135:1522–1536. Josephs KA, Duffy JR, Strand EA, et al. Clinicopathological and imaging correlates of progressive aphasia and apraxia of speech. Brain 2006;129:1385–1398. Mesulam M, Weintraub S. Spectrum of primary progressive aphasia. Baillieres Clin Neurol 1992;1:583–609. Beck J, Rohrer J, Campbell T, et al. A distinct clinical, neuropsychological and radiological phenotype is associated with progranulin gene mutations in a large UK series. Brain 2008;131:706–720. Ahmed S, de Jager CA, Haigh AM, Garrard P. Logopenic aphasia in Alzheimer’s disease: clinical variant or clinical feature? J Neurol Neurosurg Psychiatry 2012;83: 1056–1062. Rosen HJ, Allison SC, Ogar JM, et al. Behavioral features in semantic dementia vs other forms of progressive aphasias. Neurology 2006;67:1752–1756. Bozeat S, Gregory CA, Ralph MA, Hodges JR. Which neuropsychiatric and behavioural features distinguish frontal and temporal variants of frontotemporal dementia from Alzheimer’s disease? J Neurol Neurosurg Psychiatry 2000; 69:178–186. Snowden JS, Bathgate D, Varma A, Blackshaw A, Gibbons ZC, Neary D. Distinct behavioural profiles in frontotemporal dementia and semantic dementia. J Neurol Neurosurg Psychiatry 2001;70:323–332. Liu W, Miller BL, Kramer JH, et al. Behavioral disorders in the frontal and temporal variants of frontotemporal dementia. Neurology 2004;62:742–748. Xiong L, Xuereb JH, Spillantini MG, Patterson K, Hodges JR, Nestor PJ. Clinical comparison of progressive aphasia associated with Alzheimer versus FTD-spectrum pathology. J Neurol Neurosurg Psychiatry 2011;82:254–260.
Neurology 81
November 19, 2013
1839
ª"NFSJDBO"DBEFNZPG/FVSPMPHZ6OBVUIPSJ[FESFQSPEVDUJPOPGUIJTBSUJDMFJTQSPIJCJUFE
Jennifer M. Harris, BSc Claire Gall, MB, ChB Jennifer C. Thompson, PhD Anna M.T. Richardson, MB, ChB David Neary, MD Daniel du Plessis, MB, ChB Piyali Pal, MD David M.A. Mann, PhD Julie S. Snowden, PhD Matthew Jones, MD
Correspondence to Dr. Jones: [email protected]
ABSTRACT
Objective: We aimed to determine the extent to which patients with progressive language impairment conform to 2011 primary progressive aphasia (PPA) classification and to examine clinicopathologic correlations within PPA variants.
Methods: Sixty-two consecutive patients with pathologically confirmed dementia who presented clinically with aphasia were identified. Patients with insufficient clinical information were excluded. PPA classifications were applied to anonymized clinical data taken from patients’ initial assessment by raters who were blinded to clinical and pathologic diagnosis.
Results: The final cohort comprised 52 patients, 30 of whom met basic PPA criteria. Twenty-five patients met one of the 3 PPA classifications (13 logopenic, 8 nonfluent/agrammatic, and 4 semantic). Five patients did not meet the criteria for any of the PPA variants. All patients who met semantic variant PPA and 75% of patients who met nonfluent/agrammatic variant PPA classifications had frontotemporal lobar degeneration spectrum pathology. Pathologies were heterogeneous in patients who met logopenic variant PPA criteria (46% Alzheimer disease [AD], 8% AD mixed with dementia with Lewy bodies, 23% frontotemporal lobar degeneration, and 23% other). Conclusion: The 2011 PPA recommendations classify a large proportion of patients who meet basic PPA criteria. However, some patients had aphasic syndromes that could not be classified, suggesting that the 2011 recommendations do not cover the full range of PPA variants. Classification of semantic variant PPA provides a good prediction of underlying pathology. Classification of logopenic variant does not successfully differentiate PPA due to AD from PPA due to other pathologies. Neurology® 2013;81:1832–1839 GLOSSARY AD 5 Alzheimer disease; AOS 5 apraxia of speech; CBD 5 corticobasal degeneration; C9orf72 5 chromosome 9 open reading frame 72; FTLD 5 frontotemporal lobar degeneration; GRN 5 progranulin; lvPPA 5 logopenic variant of primary progressive aphasia; MAPT 5 microtubule-associated protein tau; nfvPPA 5 nonfluent/agrammatic variant of primary progressive aphasia; PPA 5 primary progressive aphasia; svPPA 5 semantic variant of primary progressive aphasia; TDP-43 5 transactive response DNA-binding protein 43.
Primary progressive aphasia (PPA) is clinically heterogeneous. Recommendations proposed in 2011 to standardize classification1 identify 3 variants: nonfluent/agrammatic (nfvPPA), semantic (svPPA), and logopenic (lvPPA). The 2011 classification system makes no a priori assumptions about the underlying pathologic substrate. In this respect, it differs from diagnostic criteria for other neurodegenerative disorders2–4 and from the 1998 frontotemporal lobar degeneration (FTLD) criteria, whereby classifications of progressive nonfluent aphasia and semantic dementia were intended to predict FTLD pathology.5 Significant pathologic heterogeneity has been reported in PPA.6 However, clinicopathologic studies were conducted before the publication of the 2011 recommendations, so it is possible that the heterogeneity reflects variable methods of patient clinical subtype classification. Moreover, some patients previously classified as “nonfluent” might more accurately conform to the profile of lvPPA rather than nfvPPA. There is a need for systematic examination of the Supplemental data at www.neurology.org From the Manchester Academic Health Sciences Centre (J.M.H., C.G., J.C.T., A.M.T.R., D.N., D.d.P., P.P., D.M.A.M., J.S.S., M.J.), Cerebral Function Unit, Greater Manchester Neuroscience Centre, Salford Royal NHS Foundation Trust, Salford; and Institute of Brain, Behaviour and Mental Health (J.M.H., D.N., D.M.A.M., J.S.S., M.J.), University of Manchester, UK. Go to Neurology.org for full disclosures. Funding information and disclosures deemed relevant by the authors, if any, are provided at the end of the article. 1832
© 2013 American Academy of Neurology
ª"NFSJDBO"DBEFNZPG/FVSPMPHZ6OBVUIPSJ[FESFQSPEVDUJPOPGUIJTBSUJDMFJTQSPIJCJUFE
relationship between clinical variants of PPA as specified in the 2011 recommendations and their pathologic substrate. The 3 PPA variants do not appear to account for the full range of PPA syndromes,7,8 and patients may fulfill criteria for more than one variant.8 Whether these “unclassifiable” forms of PPA share a common clinical syndrome and underlying pathology remains to be determined. In this study, we aimed to establish, in a pathologically confirmed cohort, the extent to which patients presenting with language impairment fulfilled 2011 PPA classification criteria and to identify reasons for nonfulfillment. We evaluated clinicopathologic relationships to determine the degree to which clinical phenotypes predict underlying pathology. METHODS Sixty-two patients who presented clinically with progressive language impairment were identified from a larger cohort of patients who had been assessed in a specialist unit for early-onset dementia (the Cerebral Function Unit, Greater Manchester Neuroscience Centre) and subsequently undergone postmortem neuropathologic brain tissue examination. All brains were obtained with ethical approval and informed written consent. As part of their initial diagnostic assessment, all patients
Figure 1
had undergone neuropsychological examination using the Manchester Neuropsychological Profile,9–11 which includes systematic assessment of language abilities and documentation of language errors (see appendix e-1 on the Neurology® Web site at www.neurology.org). Additional published language tests had been administered, although these were variable across patients (appendix e-1). Copies of clinic letters and neuropsychological reports had all references to presumptive clinical diagnoses and personal identifiable information expunged. One of 2 raters, who were blinded to clinical and pathologic diagnoses and who had not contributed to the patients’ clinical care, rated the neuropsychological and clinical data against the 2011 PPA classification recommendations.1 Data from the patients’ initial diagnostic assessment were used. It was determined whether they met overall inclusion and exclusion criteria for PPA, i.e., basic PPA criteria, and specific criteria for one of the variants. Missing data were managed in accord with a previous study.12 Twenty-four patients’ case notes were assessed by both raters to enable evaluation of interrater reliability.
Data analysis. Data analysis was performed with SPSS statistics version 20 (IBM Corp., Armonk, NY). A significance level of p , 0.05 was used. Comparisons of demographic data in different pathologic groups were conducted using t tests, x2 test of associations, or Fisher exact test, as appropriate. Interrater reliability was evaluated using k statistics. RESULTS Patient cohort, demographics, and genetics.
The selection of cases is shown in figure 1. Ten patients were excluded from the original 62-patient cohort because of insufficient clinical data to rate against the
Case selection
AD 5 Alzheimer disease; FTLD 5 frontotemporal lobar degeneration; PPA 5 primary progressive aphasia. Other refers to pathology other than FTLD or AD. FTLD pathologic classification according to Mackenzie et al.13,14 Neurology 81
November 19, 2013
1833
ª"NFSJDBO"DBEFNZPG/FVSPMPHZ6OBVUIPSJ[FESFQSPEVDUJPOPGUIJTBSUJDMFJTQSPIJCJUFE
Table 1
Demographics of FTLD and non-FTLD cases Pathologic finding All (n 5 52)
FTLD (n 5 29)
Non-FTLD (n 5 22)
Test statistic (df) p
Range of first visit dates
3/1/1983 to 9/7/2010 3/1/1983 to 1/19/2010 6/23/1983 to 9/7/2010
Age at onset, y, mean (SD)
60.8 (8.1)
61.0 (9.0)
60.7 (7.0)
t 5 0.118 (49)
0.91
Age at death, y, mean (SD) (1 missing)
69.8 (6.9)
69.7 (6.5)
70.0 (7.8)
t 5 20.160 (48)
0.87
3.2 (2.1)
3.4 (2.4)
2.9 (1.7)
t 5 0.782 (49)
0.44
Family history in first-degree relative, % (n) (1 missing)
35.3 (18)
44.8 (13)
23.8 (5)
x 5 2.335 (1)
0.13
Sex, % male (n)
55.8 (29)
55.2 (16)
59.1 (13)
x2 5 0.078 (1)
1.00
Duration of illness at presentation, y, mean (SD)
2
Abbreviation: FTLD 5 frontotemporal lobar degeneration. One patient had mixed FTLD and non-FTLD pathology and is therefore only included in the “all” category. Sex and family history were compared with x2 test of associations and other demographic data were compared using an independent sample t test.
criteria. The final study cohort of 52 patients (29 male and 23 female) consisted predominantly of early-onset cases (table 1); 35.3% of patients had a family history of a related disorder in at least one first-degree relative. Twenty-nine had FTLD, 22 non-FTLD pathologies, and 1 patient had mixed FTLD and non-FTLD pathology. Progranulin (GRN), microtubule-associated protein tau (MAPT), and chromosome 9 open reading frame 72 (C9orf72) genetic results were available for 21 of 29 FTLD cases (72.4%). From the FTLD cases, none had a C9orf72 expansion (3 missing), 5 had GRN mutations (4 missing including 2 transactive response DNA-binding protein 43 [TDP-43] type A), and 3 had MAPT mutations. APOE status was available for 38 cases (missing: 9 FTLD, 4 non-FTLD, and 1 mixed FTLD and non-
Table 2
PPA subtype Nonfluent
Pathologic relationships in PPA variants Total no. of patients 8
No. of patients in each pathologic group 6 FTLD (2 TDP-43 type A—1 GRN mutation, 1 Pick disease, 3 corticobasal degeneration) 1 CJD 1 mixed DLB, AD, and vasculitis
Semantic Logopenic
4 13
4 FTLD (TDP-43 type C) 3 FTLD (TDP-43 type A—1 GRN mutation, other 2 genetic information not available) 7 AD (including 1 mixed AD and DLB) 1 DLB 1 CJD 1 cerebrovascular disease
Unclassified
5
3 FTLD (TDP-43 type A—2 GRN mutation) 2 AD (including 1 mixed AD and cerebrovascular disease)
Abbreviations: AD 5 Alzheimer disease; CJD 5 Creutzfeldt-Jakob disease; DLB 5 dementia with Lewy bodies; FTLD 5 frontotemporal lobar degeneration; GRN 5 progranulin; PPA 5 primary progressive aphasia; TDP-43 5 transactive response DNA-binding protein 43. FTLD pathologic classification according to Mackenzie et al13,14; TDP-43 subtyping according to Mackenzie et al.15 1834
Neurology 81
FTLD). The APOE status of FTLD cases are as follows: 2 e2e3, 10 e3e3, 7 e3e4, and 1 e4e4. The APOE status of non-FTLD cases are as follows: 2 e2e3, 9 e3e3, 1 e2e4, 5 e3e4, and 1 e4e4. Interrater reliability. There was 100.0% agreement between raters for meeting basic PPA criteria (k 5 1.000, p , 0.001). The percentage agreement for each PPA variant was as follows: nfvPPA 83.3% (k 5 0.429, p 5 0.007), svPPA 100.0% (k 5 1.000, p , 0.001), and lvPPA 95.8% (k 5 0.917, p , 0.001). The lower level of agreement for nfvPPA was attributable to inconsistencies between raters in evaluating the core feature, “Effortful, halting speech with inconsistent speech sound errors and distortions (apraxia of speech)” (table e-1). Accordingly, if this feature was rated as present or there was uncertainty about its presence, it was checked by 2 experienced clinicians (M.J. and J.T.) and a consensus was reached. Patients who met basic PPA clinical criteria. Thirty patients met basic criteria for PPA. Eight fulfilled criteria for nfvPPA, 4 for svPPA, and 13 for lvPPA. Five patients did not meet criteria for a specific PPA variant, henceforth referred to as PPA-unclassified. All patients with svPPA and the majority of patients with nfvPPA had FTLD pathology13,14 (table 2). Conversely, most cases of lvPPA had non-FTLD pathology, with Alzheimer disease (AD) pathology accounting for the largest proportion of cases. Four patients who met basic PPA criteria had genetic mutations (table 2). nfvPPA: Agrammatism vs apraxia of speech. nfvPPA can be subdivided because only 1 of 2 core features is required to reach a diagnosis: agrammatism or apraxia of speech (AOS). Of the 8 patients with nfvPPA, 2 had prominent agrammatism without AOS. Both had TDP-43 type A pathology.15 One of these patients had a GRN mutation. Neither patient had motor signs at presentation. Five had prominent AOS, 3 of whom had tau pathology (2 corticobasal degeneration [CBD], 1 Pick).
November 19, 2013
ª"NFSJDBO"DBEFNZPG/FVSPMPHZ6OBVUIPSJ[FESFQSPEVDUJPOPGUIJTBSUJDMFJTQSPIJCJUFE
One patient with CBD pathology had markedly dysprosodic speech with evidence of AOS without generalized buccofacial or limb apraxia. The other patient with CBD pathology presented to the clinic with a 5-year history, beginning with circumscribed difficulties with word-finding for the first 3 years, followed by emerging difficulties using her right arm and leg. The patient with Pick disease pathology had profound buccofacial and AOS without limb apraxia. Three years after symptom onset, mild problems using her limbs emerged with crude and impersistent actions. Of the remaining 2 patients, one had mixed pathology of AD, dementia with Lewy bodies, and vasculitis, and one had Creutzfeldt-Jakob disease (described in detail elsewhere16). The case with mixed pathology presented with an atypical pattern of deficits. His speech was nonfluent with elongation of syllables and loss of prosody, which was interpreted as AOS. There was evidence of impairments in syntactic comprehension, repetition, confrontation naming, spelling, reading, and calculation. He also showed some difficulties with visual construction and mild executive impairments. There was no generalized buccofacial or limb apraxia. He was followed up for 2 years, during which time he became parkinsonian and mute with some behavioral disinhibition. One patient with both AOS and agrammatism had CBD pathology; he also displayed limb apraxia at presentation. Unclassified PPA. Two patients with PPA-unclassified presented with profound, circumscribed anomia, characterized by impaired phonologic access. They developed limb apraxia over the course of the illness. Both patients had core features of lvPPA (one “impaired single-word retrieval in spontaneous speech and naming,” the second “impaired repetition of sentences and phrases”) and svPPA (one “impaired confrontation naming,” the second both core features). The second patient who had both core features of svPPA had no supportive features. Both patients had the supportive features of nfvPPA “impaired comprehension of syntactically complex sentences” and “spared object knowledge,” but neither patient had core features of nfvPPA at presentation. At follow-up, the first patient met criteria for both nfvPPA and lvPPA, whereas the second patient did not meet criteria for a PPA subtype. Both of these anomic patients had FTLD TDP-43 type A pathology and mutations in the GRN gene. One has been described in detail elsewhere.17,18 A third PPA-unclassified patient presented with fluent anomic speech with semantic errors and impaired syntactic understanding. At his initial visit, he showed impaired confrontation naming and impaired single-word retrieval, therefore having one core feature of both svPPA and lvPPA. However, he
did not have any other core features of svPPA or lvPPA. Motor and syntactic elements of his speech were spared, thus he exhibited neither of the core features of nfvPPA. He had FTLD TDP-43 type A pathology. Another unclassified patient (age at onset 72 years) presented with anomia with moderate semantic and phonologic impairment. In addition, he developed some behavioral features of behavioral variant FTD. He had both core features of svPPA, but he did not exhibit impaired object knowledge and his repetition was not spared. Thereby, he lacked sufficient supportive features to fulfill the criteria. During follow-up, he met criteria for svPPA. He had AD pathology, with additional vascular pathology in the temporal and parietal cortex. The final unclassified patient presented with a reported 5-year history of language disorder. He showed all supportive features of lvPPA, but despite the long duration of illness, he exhibited preserved sentence repetition and confrontation naming so did not fulfill lvPPA criteria. He subsequently fulfilled the criteria at follow-up 9 years after symptom onset. He showed AD pathology. Patients presenting with aphasia who did not meet PPA inclusion/exclusion criteria. Twenty-two patients pre-
sented with a prominent language deficit but did not meet basic PPA criteria, largely because of the presence of an exclusion criterion. Interestingly, many of these patients could be ascribed a PPA subtype, based on language features alone. Table 3 shows 2011 PPA subtype classification of these patients, the basis for exclusion, underlying pathology, and genetic mutations where present. A significant association was found between the presence of prominent initial behavioral disturbance and FTLD pathology (p 5 0.03). Conversely, the presence of prominent initial episodic memory, visual memory, and visuoperceptual impairments was associated with AD pathology (p 5 0.001). Frequency of individual features of PPA clinical classification across patient groups. Both the core fea-
tures and several of the supporting features of lvPPA were present in the majority of the 52 patients presenting clinically with aphasia across pathologic groups (figure 2). In contrast, both of the core features of nfvPPA and impaired single-word comprehension, one of the core features of svPPA, were present in very few cases. The majority of patients who had either of the core features of nfvPPA had FTLD pathology. DISCUSSION This study found that 83.3% of patients who met basic criteria for PPA could be ascribed a PPA subtype using the 2011 classification Neurology 81
November 19, 2013
1835
ª"NFSJDBO"DBEFNZPG/FVSPMPHZ6OBVUIPSJ[FESFQSPEVDUJPOPGUIJTBSUJDMFJTQSPIJCJUFE
Table 3
Relationships among exclusion criteria, pathologic diagnosis, and 2011 PPA classification Pathologic diagnosis (and genetic mutation where applicable)
2011 PPA classification
No. of patients
AD
lvPPA
5
AD
Unclassified
2
DLB (1AD)
Unclassified
1
Pick disease
nfvPPA
1
Pick disease
svPPA
1
FTDP-17T; MAPT mutation
svPPA
2
Mixed
svPPA
1
TDP type C
svPPA
1
TDP type C
svPPA and lvPPA
2
Prominent initial motor disorder
TDP type B
nfvPPA
2
Prominent initial motor and behavioral disturbance
FTDP-17T; MAPT mutation
svPPA
1
TDP type A; GRN mutation
lvPPA
1
CBD
nfvPPA
1
Pick disease
Unclassified
1
Exclusion criterion Prominent initial episodic memory, visual memory, and visuoperceptual impairments
Prominent initial behavioral disturbance
None (but do not meet all inclusion criteria)
Abbreviations: AD 5 Alzheimer disease; CBD 5 corticobasal degeneration; DLB 5 dementia with Lewy bodies; FTDP-17T 5 frontotemporal dementia with parkinsonism-17 with tau inclusions; GRN 5 progranulin; lvPPA 5 logopenic variant of PPA; MAPT 5 microtubule-associated protein tau; nfvPPA 5 nonfluent/agrammatic variant of PPA; PPA 5 primary progressive aphasia; svPPA 5 semantic variant of PPA; TDP 5 transactive response DNA-binding protein.
recommendations. The clinicopathologic associations seen in the PPA subtypes are discussed below. nfvPPA and svPPA were most frequently associated with FTLD pathologies, specifically FTLD TDP-43 type A and tau pathologies in nfvPPA and TDP-43 type C in svPPA. AD was the most common pathology in lvPPA. These findings are in keeping with previous biomarker19,20 and clinicopathologic studies.11,21,22 Within the nfvPPA group, we found patients with both prominent AOS and prominent agrammatism. The 2 with prominent agrammatism had TDP-43 type A pathology. The patients with prominent AOS were more varied. Two were highly atypical cases clinically: one had Creutzfeldt-Jakob disease and in life had been diagnosed as such because of rapid clinical deterioration, and the other had a pathologically mixed disorder of AD, dementia with Lewy bodies, and vasculitis as described in the results section. Three other patients with speech apraxia had tau pathologies. This finding supports the argument that speech apraxia may be a predictor of FTLD-tau pathology.23–25 Moreover, these cases highlight the fact that nfvPPA is itself not homogeneous, and encompasses etiologically significant subvariants. lvPPA showed the poorest pathologic specificity, being linked not only to AD but also to a variety of other pathologies. This may reflect lack of specificity and discriminating value within the diagnostic features. It has been pointed out that 3 of the 4 1836
Neurology 81
supporting features of lvPPA correspond to the sparing of features characteristic of nfvPPA and svPPA,8 rather than the presence of distinct features. The same factor analytic study8 showed that “sentence repetition,” a core feature of lvPPA, clustered with the features of nfvPPA. In keeping with lack of specificity, most of the features of lvPPA were reported in the majority of patients with PPA (figure 2). Two patients, who met criteria respectively for nfvPPA and lvPPA, had a pathologic diagnosis of Creutzfeldt-Jakob disease, highlighting the fact that Creutzfeldt-Jakob disease may present as a circumscribed aphasia. Those patients had a rapid onset and progression of symptoms and would not have met PPA diagnostic criteria, had the diagnostic rule of “2-years of progressive, circumscribed aphasia,” originally suggested by Weintraub and Mesulam,26 been applied. The limitation of the 2-year rule is that it precludes early diagnosis. Nevertheless, if future PPA classifications hope to improve predictability of pathology, then rate of symptom progression requires consideration. Five patients with PPA did not meet criteria for any subtype and were termed PPA-unclassified. It has been speculated that patients with mild deficits may have insufficient features to be diagnosed with a PPA variant.1 In line with this view, one patient with AD pathology fulfilled criteria for lvPPA only on follow-up. The other 4 patients had marked
November 19, 2013
ª"NFSJDBO"DBEFNZPG/FVSPMPHZ6OBVUIPSJ[FESFQSPEVDUJPOPGUIJTBSUJDMFJTQSPIJCJUFE
Figure 2
Frequency of features from PPA classification recommendations and pathologic relationships
Features on the y-axis in bold refer to core criteria; those in standard text are supporting criteria. FTLD pathologic classification according to Mackenzie et al.13,14 AD 5 Alzheimer disease; FTLD 5 frontotemporal lobar degeneration; lvPPA 5 logopenic variant of PPA; nfvPPA 5 nonfluent/agrammatic variant of PPA; PPA 5 primary progressive aphasia; svPPA 5 semantic variant of PPA; TDP 5 transactive response DNA-binding protein. Other refers to pathology other than FTLD or AD.
language impairment at presentation, so their nonclassification cannot be attributed to stage of illness. In each, the prevailing symptom was anomia. An anomic PPA subgroup has been proposed to account for such cases.7 It is worthy of comment, however, that anomia per se appears poorly predictive of underlying pathology. In the present series, 3 anomic patients had FTLD TDP-43 type A pathology, in 2 cases associated with GRN gene mutations, whereas the other had AD pathology. At follow-up, there was overlap between syndromes, with one patient meeting criteria for both nfvPPA and lvPPA. The PPAunclassified patients reinforce the view that more than 3 PPA syndromes exist. In particular, as noted previously,27 patients with GRN mutations may exhibit a PPA syndrome that does not conform to current classifications. Unsurprisingly, the presence of “prominent initial episodic memory, visual memory, and visuoperceptual impairments” predicted AD pathology. Importantly, those patients, with classic AD features, also typically
exhibited language features consistent with the lvPPA phenotype. This finding contrasts with reports28 that language characteristics in early-stage AD differ from those of lvPPA. The disparity may relate to differences in patient population: in the latter study, patients were older and presented a classic profile dominated by memory loss. By contrast, early-onset AD is more likely to present as a constellation of deficits in language, working and episodic memory, and perceptuospatial function.10 It is possible that the clinical similarities between lvPPA and AD are limited to early-onset AD. Most study patients who conformed to svPPA or nfvPPA language variants but did not fulfill basic criteria for PPA because of early behavioral disturbance had FTLD pathology. Behavioral changes in progressive nonfluent aphasia and semantic dementia are well documented, so overlapping PPA/behavioral syndromes may therefore be expected.29–32 Indeed, disinhibition and preference for sweet food have been found to be useful in distinguishing FTLD-PPA from Neurology 81
November 19, 2013
1837
ª"NFSJDBO"DBEFNZPG/FVSPMPHZ6OBVUIPSJ[FESFQSPEVDUJPOPGUIJTBSUJDMFJTQSPIJCJUFE
AD-PPA.33 Arguably, behavioral disturbances may cause more distress to, and difficulties for, carers than language symptoms, so may be reported as the “main cause of functional impairment” even when language disorder is the most striking feature on neuropsychological testing. Our data indicate that the presence of behavioral disorders in aphasic patients is highly suggestive of underlying FTLD pathology. Moreover, they suggest a continuum between the behavioral disorder of FTD and the semantic and nonfluent aphasias. There is a need for PPA classifications to recognize these mixed syndromes. This study has limitations; principally, although by necessity, it was conducted through retrospective blinded case note review. The 2011 PPA recommendations are intended for prospective use. Retrospective evaluation is dependent on the descriptions of linguistic and other cognitive features recorded by clinicians at the time. However, all PPA patients in our study had undergone detailed neuropsychological assessment, including systematic assessment of expressive and receptive language skills. If there were insufficient clinical data, then patients were excluded from further analysis (n 5 10). This study only considered the clinical aspects of the classification system. We have not included imaging data principally because there were some cases for which only basic CT scans were available or no scan data were available at all because of the long time scale over which patients were seen. Another limitation is that vascular and psychiatric disorders were underrepresented in the sample, and these disorders may mimic PPA. It would be of value to replicate the study in an independent pathologic cohort that includes a larger number of patients with vascular and psychiatric disorders. We were able to classify the majority of patients using the 2011 PPA recommendations, attesting to their clinical utility. However, the existence of “PPA-unclassified” patients, even within a relatively small sample of 30 patients, suggests that the tripartite classification of PPA is insufficient. The svPPA classification was strongly associated with TDP-43 type C pathology, whereas the lvPPA classification was associated with diverse pathologies. The nfvPPA subtype was rarely caused by Alzheimer pathology, but a broad range of other pathologies were seen. The findings suggest that although the current classification recommendations can be used to predict pathology in svPPA, they have limited utility in predicting specific pathology in nfvPPA and lvPPA. Finally, the separation of FTLD criteria from the 2011 PPA classification raises the problem of overlapping or mixed syndromes. Our findings suggest that the behavioral and visual memory/visual perceptual features that are currently exclusions in the PPA 1838
Neurology 81
classification recommendations could be used to improve prediction of the underlying pathologic diagnosis. AUTHOR CONTRIBUTIONS Jennifer Harris: acquisition of data (rating and coding data), interpretation of data and statistical analysis, drafting/revising the manuscript. Dr. Claire Gall: acquisition of data (rating and coding data), drafting/ revising the manuscript. Dr. Jennifer Thompson: study conception and design, acquisition of data (redacting clinical data), revising the manuscript. Dr. Anna Richardson: acquisition of data (original clinical data), revising the manuscript. Prof. David Neary: acquisition of data (design of clinical history-taking), revising the manuscript. Dr. Daniel du Plessis, Dr. Piyali Pal, and Prof. David Mann: acquisition and interpretation of neuropathologic data, revising the manuscript. Prof. Julie Snowden: study conception and design, analysis and interpretation of data, drafting/revising the manuscript. Dr. Matthew Jones: study conception and design, analysis and interpretation of data, drafting/revising the manuscript, study supervision and coordination.
ACKNOWLEDGMENT The authors thank all the neurologists and neuropsychologists who have worked at the Cerebral Function Unit and contributed to the original clinical evaluation of patients included in this study, and the clinical colleagues from the North West Region of UK who referred patients. The authors thank Steve Chew-Graham for his role as coordinator of the Manchester brain donation scheme, which is supported financially by the Alzheimer’s Research Trust and Alzheimer’s Society through the Brains for Dementia Research Initiative.
STUDY FUNDING No targeted funding reported.
DISCLOSURE J. Harris, C. Gall, J. Thompson, and A. Richardson report no disclosures. D. Neary receives royalties for Neuroanatomy (Crossman A, Neary D. Elsevier Churchill Livingstone, 1st edition 1995, 2nd edition 2000, 3rd edition 2005, 4th edition 2010). D. du Plessis and P. Pal report no disclosures. D. Mann receives Medical Research Council and Wellcome Trust support through the Neurodegeneration Programme. The Manchester brain donation scheme and the work of the Manchester Brain Bank is supported by Alzheimer’s Research UK and the Alzheimer’s Society through the Brains for Dementia Research Initiative. J. Snowden has received travel expenses from the World Federation of Neurology for teaching in Kolkata India, 2012. She is a member of the Editorial Advisory Board of 1) Neuropsychologia, 2006 to present; 2) Journal of the International Neuropsychological Society, 2011 to present; 3) Amyotrophic Lateral Sclerosis and Frontotemporal Degeneration, 2012 to present; and 4) Neuropsychology, 2013 to present. M. Jones reports no disclosures. Go to Neurology.org for full disclosures.
Received May 1, 2013. Accepted in final form August 19, 2013. REFERENCES 1. Gorno-Tempini ML, Hillis AE, Weintraub S, et al. Classification of primary progressive aphasia and its variants. Neurology 2011;76:1006–1014. 2. Rascovsky K, Hodges JR, Knopman D, et al. Sensitivity of revised diagnostic criteria for the behavioural variant of frontotemporal dementia. Brain 2011;134:2456–2477. 3. McKhann G, Knopman DS, Chertkow H, et al. The diagnosis of dementia due to Alzheimer’s disease: recommendations from the National Institute on Aging–Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 2011;7:263–269.
November 19, 2013
ª"NFSJDBO"DBEFNZPG/FVSPMPHZ6OBVUIPSJ[FESFQSPEVDUJPOPGUIJTBSUJDMFJTQSPIJCJUFE
4.
5.
6. 7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17. 18.
McKeith IG, Dickson DW, Lowe J, et al. Diagnosis and management of dementia with Lewy bodies: third report of the DLB Consortium. Neurology 2005;65:1863–1872. Neary D, Snowden J, Gustafson L, et al. Frontotemporal lobar degeneration: a consensus on clinical diagnostic criteria. Neurology 1998;51:1546–1554. Grossman M. Primary progressive aphasia: clinicopathological correlations. Nat Rev Neurol 2010;6:88–97. Mesulam M, Wieneke C, Thompson C, Rogalski E, Weintraub S. Quantitative classification of primary progressive aphasia at early and mild impairment stages. Brain 2012;135:1537–1553. Sajjadi SA, Patterson K, Arnold RJ, Watson PC, Nestor PJ. Primary progressive aphasia: a tale of two syndromes and the rest. Neurology 2012;78:1670–1677. Thompson JC, Stopford CL, Snowden J, Neary D. Qualitative neuropsychological performance characteristics in frontotemporal dementia and Alzheimer’s disease. J Neurol Neurosurg Psychiatry 2005;76:920–927. Snowden J, Stopford CL, Julien C, et al. Cognitive phenotypes in Alzheimer’s disease and genetic risk. Cortex 2007;43:835–845. Snowden J, Thompson JC, Stopford CL, et al. The clinical diagnosis of early-onset dementias: diagnostic accuracy and clinicopathological relationships. Brain 2011; 134:2478–2492. Harris J, Gall C, Thompson J, et al. Sensitivity and specificity of FTDC criteria for behavioral variant frontotemporal dementia. Neurology 2013;80:1881–1887. Mackenzie IR, Neumann M, Bigio EH, et al. Nomenclature for neuropathologic subtypes of frontotemporal lobar degeneration: consensus recommendations. Acta Neuropathol 2009;117:15–18. Mackenzie IR, Neumann M, Bigio EH, et al. Nomenclature and nosology for neuropathologic subtypes of frontotemporal lobar degeneration: an update. Acta Neuropathol 2010;119:1–4. Mackenzie IR, Neumann M, Baborie A, et al. A harmonized classification system for FTLD-TDP pathology. Acta Neuropathol 2011;122:111–113. Kobylecki C, Thompson JC, Jones M, et al. Sporadic Creutzfeldt-Jakob disease presenting as progressive nonfluent aphasia with speech apraxia. Alzheimer Dis Assoc Disord Epub 2012 Jun 22. Snowden J, Neary D. Progressive anomia with preserved oral spelling and automatic speech. Neurocase 2003;9:27–43. Snowden JS, Pickering-Brown SM, Du Plessis D, et al. Progressive anomia revisited: focal degeneration associated with progranulin gene mutation. Neurocase 2007; 13:366–377.
19.
20.
21.
22. 23.
24.
25.
26. 27.
28.
29.
30.
31.
32.
33.
Rabinovici GD, Jagust WJ, Furst AJ, et al. Abeta amyloid and glucose metabolism in three variants of primary progressive aphasia. Ann Neurol 2008;64:388–401. Leyton CE, Villemagne VL, Savage S, et al. Subtypes of progressive aphasia: application of the International Consensus Criteria and validation using beta-amyloid imaging. Brain 2011;134:3030–3043. Snowden J, Neary D, Mann DM. Frontotemporal lobar degeneration: clinical and pathological relationships. Acta Neuropathol 2007;114:31–38. Mesulam M. Primary progressive aphasia pathology. Ann Neurol 2008;63:124–125. Josephs KA, Boeve BF, Duffy JR, et al. Atypical progressive supranuclear palsy underlying progressive apraxia of speech and nonfluent aphasia. Neurocase 2005;11:283–296. Josephs KA, Duffy JR, Strand EA, et al. Characterizing a neurodegenerative syndrome: primary progressive apraxia of speech. Brain 2012;135:1522–1536. Josephs KA, Duffy JR, Strand EA, et al. Clinicopathological and imaging correlates of progressive aphasia and apraxia of speech. Brain 2006;129:1385–1398. Mesulam M, Weintraub S. Spectrum of primary progressive aphasia. Baillieres Clin Neurol 1992;1:583–609. Beck J, Rohrer J, Campbell T, et al. A distinct clinical, neuropsychological and radiological phenotype is associated with progranulin gene mutations in a large UK series. Brain 2008;131:706–720. Ahmed S, de Jager CA, Haigh AM, Garrard P. Logopenic aphasia in Alzheimer’s disease: clinical variant or clinical feature? J Neurol Neurosurg Psychiatry 2012;83: 1056–1062. Rosen HJ, Allison SC, Ogar JM, et al. Behavioral features in semantic dementia vs other forms of progressive aphasias. Neurology 2006;67:1752–1756. Bozeat S, Gregory CA, Ralph MA, Hodges JR. Which neuropsychiatric and behavioural features distinguish frontal and temporal variants of frontotemporal dementia from Alzheimer’s disease? J Neurol Neurosurg Psychiatry 2000; 69:178–186. Snowden JS, Bathgate D, Varma A, Blackshaw A, Gibbons ZC, Neary D. Distinct behavioural profiles in frontotemporal dementia and semantic dementia. J Neurol Neurosurg Psychiatry 2001;70:323–332. Liu W, Miller BL, Kramer JH, et al. Behavioral disorders in the frontal and temporal variants of frontotemporal dementia. Neurology 2004;62:742–748. Xiong L, Xuereb JH, Spillantini MG, Patterson K, Hodges JR, Nestor PJ. Clinical comparison of progressive aphasia associated with Alzheimer versus FTD-spectrum pathology. J Neurol Neurosurg Psychiatry 2011;82:254–260.
Neurology 81
November 19, 2013
1839
ª"NFSJDBO"DBEFNZPG/FVSPMPHZ6OBVUIPSJ[FESFQSPEVDUJPOPGUIJTBSUJDMFJTQSPIJCJUFE
