Quantitative Template for Subtyping Primary Progressive Aphasia

ORIGINAL CONTRIBUTION Quantitative Template for Subtyping Primary Progressive Aphasia

Marsel Mesulam, MD; Christina Wieneke, BA; Emily Rogalski, PhD; Derin Cobia, PhD; Cynthia Thompson, PhD; Sandra Weintraub, PhD

Background: The syndrome of primary progressive Fourth Edition) classified all 16 patients in concor- aphasia (PPA) is diagnosed when a gradual failure of word dance with a clinical diagnosis that had been made be- usage or comprehension emerges as the principal fea- fore the administration of quantitative tests. All 3 PPA ture of a neurodegenerative disease. subtypes had distinctly asymmetrical atrophy of the left perisylvian language network. Each subtype also had dis- Objective: To provide a quantitative algorithm for clas- tinctive peak atrophy sites: PPA-G in the inferior frontal sifying PPA into agrammatic (PPA-G), semantic (PPA-S), gyrus (Broca area), PPA-S in the anterior temporal lobe, and logopenic (PPA-L) variants, each of which is known and PPA-L in Brodmann area 37. to have a different probability of association with Alzhei- mer disease vs frontotemporal lobar degeneration. Conclusions: Once an accurate root diagnosis of PPA is made, subtyping can be quantitatively guided using a Design: Prospective study. 2-dimensional template based on orthogonal tasks of Setting: University medical center. grammatical competence and word comprehension. Al- though the choice of tasks and the precise cutoff levels Patients: Sixteen consecutively enrolled patients with may need to be adjusted to fit linguistic and educational PPA who underwent neuropsychological testing and mag- backgrounds, these 16 patients demonstrate the feasi- netic resonance imaging recruited nationally in the United bility of using a simple algorithm for clinicoanatomical States as part of a longitudinal study. classification in PPA. Prospective studies will show whether this subtyping can improve clinical prediction Results: A 2-dimensional template that reflects perfor- of the underlying neuropathologic condition. mance on tests of syntax (Northwestern Anagram Test) and lexical semantics (Peabody Picture Vocabulary Test— Arch Neurol. 2009;66(12):1545-1551

HE CLASSIFICATION OF PRI- is particularly important for early-onset de- mary progressive aphasia mentias, in which the concordance be- (PPA) into subtypes has tween clinical predictions and postmor- acquired new relevance in tem confirmation can be quite low. light of postmortem series Although numerous studies1,3-5 have de- andT in vivo amyloid imaging showing that scribed clinical and neuropsychological individual variants have different likeli- characteristics of PPA subtypes, few have hoods of being caused by Alzheimer dis- included an unselected prospective co- ease (AD) vs frontotemporal lobar de- hort investigated using a unified battery generation (FTLD). The most frequent of easily administered tests specifically cho- associations have been reported between sen to probe the defining features of the the agrammatic variant (PPA-G) and FTLD subtypes. with tauopathy (FTLD-T), the semantic This study describes an empirically es- variant (PPA-S) and FTLD with ubiquitin/ tablished 2-dimensional quantitative tem- TAR-DNA binding protein 43 proteinopa- plate derived from performance on tests of thy (FTLD-TDP), and the logopenic vari- syntax and lexical semantics that success- ant (PPA-L) and AD.1-3 fully classified 16 consecutively investi- In the absence of definitive in vivo bio- gated patients with PPA. The biological va- markers for these diseases, the reliable clas- lidity of the resultant classification was Author Affiliations: Cognitive sification of PPA assumes considerable rel- supported by the presence of distinctive ana- Neurology and Alzheimer’s evance for increasing the accuracy with tomical patterns of peak cortical atrophy in Disease Center, Northwestern which the nature of the underlying patho- each variant. Whether this classification also University, Chicago, Illinois. logic abnormality can be predicted. This corresponds to differential neuropatho-

(REPRINTED) ARCH NEUROL / VOL 66 (NO. 12), DEC 2009 WWW.ARCHNEUROL.COM 1545

WAB % Correct Patient No./ Aphasia Word-Word % Correct % Correct % Correct % BDAE WAB Sex/Age at Symptom Quotient Association PPVT-4 NAT BNT Fluency Repetition Testing, y Education, y Handednessb Duration, y (100) (16) (36) (10) (60) (7) (100) Agrammatic PPA Variant 1/M/62 20 ϩ80 5.0 82.3 100.0 100.0 50 98.3 42.9 80 2/M/59 12 ϩ95 3.0 79.9 100.0 94.4 40 81.7 57.1 78 3/F/61 18 ϩ100 5.0 75.3 100.0 100.0 50 88.3 42.9 88 4/F/56 14 ϩ100 2.0 65.4 100.0 83.3 10 81.7 14.3 42 Semantic PPA Variant 5/F/53 16 ϩ100 3.0 65.9 50.0 22.2 60 6.7 100.0 74 6/M/63 18 ϩ100 5.5 88.2 93.8 47.2 100 23.3 100.0 97 7/F/54 12 ϩ100 3.0 83.2 75.0 27.8 100 10.0 100.0 91 8/F/64 16 ϩ100 7.5 80.6 62.5 38.9 100 8.3 100.0 88 9/F/56 18 ϩ100 2.5 75.5 75.0 38.9 100 5.0 100.0 85 Logopenic PPA Variant 10/M/69 15 ϩ100 2.5 92.0 100.0 97.2 90 98.3 71.4 90 11/M/63 18 ϩ100 2.5 97.1 100.0 100.0 100 96.7 100.0 89 12/M/58 16 ϩ100 2.0 86.9 100.0 97.2 90 90.0 100.0 90 13/F/65 13 ϩ100 5.0 78.6 100.0 83.3 70 83.3 71.4 88 14/F/75 16 ϩ100 2.5 97.2 100.0 97.2 100 88.3 100.0 90 15/M/64 18 ϩ90 2.0 93.2 100.0 100.0 70 98.3 85.7 89 16/M/48 16 ϩ80 6.0 83.2 100.0 100.0 100 98.3 71.4 90

Abbreviations: BDAE, Boston Diagnostic Aphasia Examination; BNT, Boston Naming Test; NAT, Northwestern Anagram Test; PPA, primary progressive aphasia; PPVT-4, Peabody Picture Vocabulary Test—Fourth Edition; WAB, Western Aphasia Battery. a Numbers in parentheses indicate the highest possible raw scores. b A“ϩ” score in the Edinburgh inventory indicates the degree of right-handedness.

logic processes remains to be determined by prospective SINGLE-WORD COMPREHENSION studies. Word comprehension (lexical semantics) is commonly tested by asking the patient to match a word to a picture. The auditory METHODS word comprehension subtest of the WAB was too easy. We, therefore, opted to use the PPVT-412 and selected a subset of 36 mod- Recruitment occurred in the context of a National Institutes erately difficult items (items 157-192). Each item requires the of Health–funded project that brought patients from through- patient to match a word representing an object, action, or at- out the United States to Northwestern University for a 3-day tribute to 1 of 4 picture choices. Because performance on the intensive research program. All of the patients who fulfilled the PPVT-4 could potentially be confounded by problems of pic- criteria for PPA, who could complete the 5 key diagnostic tests, ture recognition, its face validity as a measure of word compre- and who had a magnetic resonance image suitable for quanti- hension was further established by comparing scores with those tative morphometric analysis were included. Only images ob- on a word-word association task in which patients decided which tained within a few days of neuropsychological testing were used. of 2 pairs contained semantically matching words (eg, horse- The root diagnosis of PPA was made on the basis of a progres- saddle vs horse-slippers). Only patients with PPA-S with the low- sive language disturbance (ie, aphasia) that is initially the most est PPVT-4 scores showed less than 100% performance on the salient feature of the clinical picture (ie, primary) and that is caused word-word association task (Table 1). However, the impair- by neurodegeneration (ie, is progressive).6-8 The presence of an ment on this task was milder than that on the PPVT-4. In the aphasia was established by aphasia quotients derived from admin- future, a more difficult form of the word-word association task istration of the Western Aphasia Battery (WAB).9 All of the pa- could be substituted for the PPVT-4 to eliminate potential in- tients were right-handed, as determined using the Edinburgh in- terference from picture-recognition deficits. ventory10; 8 were men and 8 were women (Table 1). Duration of disease at the time of testing varied from 2.0 to 7.5 years. The progressive nature of the deficits and the fact that the language SYNTAX IN THE CONSTRUCTION disorder was the chief problem during the initial few years of the OF SENTENCES disease were documented by the history obtained from the patient, from medical records, and from at least 1 additional informant who Syntax, a major component of grammar, regulates the proper or- lived in the same household.11 All of the patients had received a dering of words into sentences. Its assessment is challenging. The descriptive diagnosis of PPA-G, PPA-L, or PPA-S based on an ini- WAB, for example, has no subtest for assessing syntax. In tradi- tial office evaluation (Table 2), and before administration of the tional aphasiology, fluency and phrase length have been used as quantitative tests, by clinicians with extensive experience with this surrogates for grammatical competence. However, it becomes dif- disease (M.M. and S.W.). Five language tests provided the basis ficult to decide whether apparent agrammatism in a dysfluent pa- for the quantitative classification: the Peabody Picture Vocabu- tient represents an economy of expression, consequences of dys- lary Test—Fourth Edition (PPVT-4), the Northwestern Anagram arthria, or a true insensitivity to rules of syntax. To circumvent Test (NAT), the Boston Naming Test, the Boston Diagnostic Apha- these problems, we designed the NAT.13 During administration sia Examination, and the WAB repetition subtest. of the NAT, the patient is asked to order single words, each printed

(REPRINTED) ARCH NEUROL / VOL 66 (NO. 12), DEC 2009 WWW.ARCHNEUROL.COM 1546

PPA Variant Descriptive Criteria Quantitative Criteria Root diagnosis Presence of aphasia determined by abnormality in word finding, word comprehension, word order, or other aspects of grammar. This can be determined by clinical examination or abnormal scores on batteries such as the WAB or the BDAE. Isolated impairments of speech alone (eg, dysarthria, apraxia of speech, and aphemia) do not qualify. The disorder is progressive. The aphasia is primary in the sense that it emerges as the most salient feature and becomes the chief impediment to customary daily living activities during the initial stages (1-2 y) of the disease. Neurodiagnostics do not reveal a cause for the aphasia other than neurodegeneration. Agrammatic The central feature is an abnormality in syntax (word order) or some other aspect of grammar in spoken or NAT score is Ͻ60% and written language in the presence of relatively preserved single-word comprehension. Fluency is usually PPVT-4 score is impaired, and speech is usually effortful and hesitant. Ն60% Semantic The central feature is an abnormality in single-word comprehension in the presence of relatively preserved PPVT-4 score is Ͻ60% grammar and fluency. Output is circumlocutory, occasionally uninformative, and frequently paraphasic. and NAT score is Naming is severely impaired. Ն60% Logopenic The central features are intermittent word-finding hesitations and phonemic paraphasias. Naming is PPVT-4 and NAT scores impaired but not as severely as in PPA-S and improves on phonemic cueing. Repetition may be impaired. are both Ն60% Fluent output in casual conversation can alternate with dysfluent speech, which emerges when the patient needs to convey precise information and cannot use circumlocution. Spelling can be impaired. Mixed Combination of agrammatism with comprehension deficit, usually accompanied by poor fluency and PPVT-4 and NAT scores frequent paraphasias are both Ͻ60%

Abbreviations: BDAE, Boston Diagnostic Aphasia Examination; NAT, Northwestern Anagram Test; PPA, primary progressive aphasia; PPVT-4, Peabody Picture Vocabulary Test—Fourth Edition; WAB, Western Aphasia Battery.

on a separate card, to be syntactically consistent with an action .nmr.mgh.harvard.edu) as previously described.11 Thickness depicted in a target picture.13 Printed words and arrows label each maps from the PPA group were statistically contrasted against actor and action in the picture to minimize the impact of single- those from 17 right-handed healthy volunteers (9 men and 8 word comprehension deficits on performance. Correct perfor- women; mean age, 64.4 years; mean educational level, 16.29 mance, therefore, specifically reflects the ability to order words years). There were no statistically significant differences in age into a sentence that has a syntactic structure consistent with the or educational level between groups. Differences in thickness depicted action. This test correlates with other tests of grammati- between groups were calculated by conducting a general lin- cal sentence production but not with tests of naming, single- ear model on every vertex along the cortical surface. False dis- word comprehension, or motor speech production. For PPA sub- covery rate methods were applied to adjust for multiple com- typing, we chose a subset of 10 items from the NAT that contain parisons.16 A significance threshold of PϽ.01 was used to detect object-extracted wh-questions and subject-extracted wh- areas of peak cortical thinning (ie, atrophy) in patients with questions. These items of intermediate difficulty could be per- PPA compared with controls. Because of the small sample size, formed even by patients with PPA-S with prominent word com- direct comparisons of subgroups was not performed. prehension deficits. The NAT inclusive of the 10-item subset (designated as object-extracted wh-questions and subject- extracted wh-questions) can be downloaded at http://www.soc RESULTS .northwestern.edu/NorthwesternAnagramTest/. Performances on the 5 language tests described in the NAMING, FLUENCY, AND REPETITION “Methods” section were expressed as a percentage of the highest possible scores for that test (Figure 1) and then The Boston Naming Test was used to assess the confrontation 14 were placed on a 2-dimensional map where the x and y axes naming of objects. It is a 60-item standardized test in which reflect the percentage scores on tests of grammaticality (mea- items are administered in order of decreasing frequency of oc- currence in the language. sured using the NAT) and word comprehension (mea- There are several measures of fluency, and the one selected sured using the PPVT-4) (Figure 2). The 60% range of for this study was “phrase length,” defined as the longest string performance on each axis, chosen empirically to fit the di- of words produced without pause in a speech sample. Recorded agnoses we had given during the initial office examina- samples while describing the “picnic” picture from the WAB were tion, divided the map into 4 quadrants (Figure 2). Accord- transcribed and rated by 2 raters (C.W. and S.W.) on a 7-point ing to the resultant map, the subtype is (1) PPA-S if the scale for phrase length taken from the Boston Diagnostic Apha- PPVT-4 score is less than 60% and the NAT score is 60% 15 sia Examination. Although the Boston Diagnostic Aphasia Ex- or greater, (2) PPA-G if the NAT score is less than 60% and amination also has a picture description task, the WAB picture the PPVT-4 score is 60% or greater, (3) PPA-L if the PPVT-4 has more actors and actions and provides more varied opportu- and NAT scores are both 60% or greater, and (4) mixed nities for speech production. Repetition was measured using the corresponding subtest from the WAB, which samples repeti- PPA if the PPVT-4 and NAT scores are both less than 60%. tion of single words, phrases, and sentences. Of the 16 patients, 4 were in the PPA-G group (patients 1-4), 5 were in the PPA-S group (patients 5-9), and IMAGING 7 were in the PPA-L group (patients 10-16). All of the subtypes displayed asymmetrically greater atrophy in the Images were acquired and reconstructed using the FreeSurfer left hemisphere (Figure 3). Peak atrophy in PPA-G in- image analysis suite (version 4.1.0; available at http://surfer cluded the inferior frontal gyrus (IFG, Broca area) and

(REPRINTED) ARCH NEUROL / VOL 66 (NO. 12), DEC 2009 WWW.ARCHNEUROL.COM 1547

Performance, % 30 20 10 0

P5 P6 P7 P8 PPA-S PPA-S PPA-S PPA-S 100 90 80 70 60 50 40

Performance, % 30 20 10 0

P9 P10 P11 P12 PPA-S PPA-L PPA-L PPA-L 100 90 80 70 60 50 40

Performance, % 30 20 10 0

P13 P14 P15 P16 PPA-L PPA-L PPA-L PPA-L 100 90 80 70 60 50 40

Performance, % 30 20 10 0 PPVT-4 NAT BNT F R PPVT-4 NAT BNT F R PPVT-4 NAT BNT F R PPVT-4 NAT BNT F R Test Test Test Test

Figure 1. Performance of the 16 patients on the 5 language tests. The height of the bars represents performance as a percentage, where 100% reflects a perfect score for that task. The horizontal lines show the 60% cutoff level for subtyping. BNT indicates Boston Naming Test; F, fluency as rated using the Boston Diagnostic Aphasia Examination; NAT, Northwestern Anagram Test; P1-P16, patients 1 to 16; PPA, primary progressive aphasia (PPA-G, agrammatic variant; PPA-S, semantic variant; and PPA-L, logopenic variant); PPVT-4, Peabody Picture Vocabulary Test—Fourth Edition; and R, repetition subtest of the Western Aphasia Battery.

the temporoparietal junction (TPJ). Additional atrophy terior temporal lobe, including the superior, middle, and was seen in the premotor and dorsolateral prefrontal cor- inferior temporal gyri and the fusiform gyrus. The PPA-L tices. The PPA-S group showed atrophy mostly in the an- group had peak atrophy in the TPJ and the posterior parts

(REPRINTED) ARCH NEUROL / VOL 66 (NO. 12), DEC 2009 WWW.ARCHNEUROL.COM 1548

©2009 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/04/2021 of the inferior temporal gyrus (Brodmann area 37). The IFG atrophy was prominent only in PPA-G, Brodmann P1 P16 P3 P15 P10 P11 area 37 atrophy only in PPA-L, and anterior temporal at- 100 P12 P2 rophy only in PPA-S. The leftward asymmetry was most P14 prominent in PPA-L. P4 P13 80

COMMENT (PPA-G) (PPA-L) 60 There is no one-to-one correspondence between anatomical components of the left perisylvian language network and P6 40 P9 specific language functions. In general, however, the fron- P8 Word Comprehension, % Word tal components are more closely related to fluency and gram- P7 mar, whereas the posterior and temporal components are P5 more closely related to lexical semantics and object 20 naming.17-19 Damage to different sectors of the language network can differentially hinder speech fluency, grammati- (PPA-M) (PPA-S) cal competence, word comprehension, word finding, spell- 20 40 60 80 100 ing,reading,andobjectnaming.Classicalaphasiology,based Grammaticality of Sentence Construction, % predominantly on the investigation of patients with focal cerebrovascular disease, delineated Broca, Wernicke, con- Figure 2. A 2-dimensional template based on single-word comprehension duction, and transcortical aphasias as prototypical mani- (Peabody Picture Vocabulary Test—Fourth Edition) and grammatical 20 structure of sentences (Northwestern Anagram Test). The 60% performance festations of damage to different parts of the network. level divides the template into 4 quadrants, 1 for each primary progressive The left perisylvian language network can also be- aphasia (PPA) subtype (PPA-G, agrammatic variant; PPA-S, semantic come the preferential target of degenerative disease. The variant; PPA-M, mixed variant; and PPA-L, logopenic variant). The values on resultant syndrome, a progressive and initially isolated lan- the x- and y-axes reflect the performance percentages shown in Figure 1. P1-P16 indicate patients 1 to 16. guage impairment, is known as PPA. As in the case of aphasias caused by cerebrovascular accidents, the aphasia in PPA can display numerous patterns. However, the clinicoanatomical correlations established in acute cerebrovascular lesions are not necessarily generalizable to those encountered in PPA.21 The differences probably reflect the IFG slow destruction of tissue by neurodegenerative disease, residual survival of neurons even in the most atrophic areas, and compensatory reorganizations of synaptic circuitry. PPA-G Recent developments showing that individual aphasic

patterns are differentially associated with the neuropatho- PM TPJ logic features of AD, FTLD-TDP, and FTLD-T have re- DF kindled the need to establish reliable subtyping of PPA. A IFG 37 widespread practice has been to use the progressive non- STG MTG fluent aphasia (PNFA) and semantic dementia (SD) syn- ITG dromes described by Neary et al4 as the two major vari- PPA-S ants of progressive aphasia. However, the PNFA designation, based on the core feature of “nonfluent spontaneous speech with at least one of the following: agrammatism, phonemic paraphasias, anomia,” seems, in retrospect, to have been too broad. The introduction of a logopenic PPA variant by 37 5 Gorno-Tempini et al has led to the division of PNFA into PPA-L agrammatic (PPA-G) and logopenic (PPA-L) subtypes of PPA. Patients with PPA-L may be dysfluent because of word- Figure 3. Distribution of cortical thinning. Red shading indicates a finding hesitations but do not show major impairments con- significance level of PϽ.01; yellow shading, PϽ.001. DF indicates dorsolateral prefrontal cortex; IFG, inferior frontal gyrus; ITG, inferior structing grammatical sentences. The heuristic value of this temporal gyrus; MTG, middle temporal gyrus; PM, premotor cortex; subdivision of PNFA was demonstrated by postmortem in- STG, superior temporal gyrus; TPJ, temporoparietal junction; and vestigations showing that PPA-L has a high association (60% 37, area 37 of Brodmann. in a recent postmortem series) with AD abnormalities, whereas the PPA-G variant has a high (80% in the same fore, subsume patients who are equally aphasic and agnosic series) association with FTLD-T.2 and who would, therefore, not fulfill the PPA criteria in Use of the SD nomenclature raises analogous con- Table 2. Moreover, patients with PPA and poor compre- cerns of heterogeneity. Its 2 core clinical features, which hension may not qualify for the diagnosis of SD in the ab- must both be present, are “loss of word meaning” and “per- sence of at least some perceptual disorder. We addressed ceptual disorder” characterized by prosopagnosia, asso- this question in a recent study11 in which we character- ciative agnosia, or both.4 The SD designation could, there- ized PPA-S as a syndrome where word comprehension defi-

(REPRINTED) ARCH NEUROL / VOL 66 (NO. 12), DEC 2009 WWW.ARCHNEUROL.COM 1549

©2009 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/04/2021 cit is the only obligatory core feature and the major cause All 3 subtypes had asymmetrical left hemispheric of disability. The reliable and reproducible diagnosis of atrophy that involved the perisylvian and additional PPA-S is of considerable practical importance because this temporal components of the language network. Each subtype has a high likelihood of being associated with the group also had unique anatomical signatures of peak neuropathologic features of FTLD-TDP.3,8 atrophy sites in the language network. These anatomical Previous subtyping approaches, such as the one by patterns agree with those described by Gorno-Tempini Neary et al,4 have generally relied on lists of features but and colleagues5 and, therefore, confirm the biological have rarely specified quantitative boundaries or specific validity of the subtyping method described herein. instruments. One of the several challenges has been the The distinctive atrophy patterns were concordant with implicit use of the term fluency, which can be impaired the clinical profiles. The areas of peak atrophy in PPA-S, by damage outside the language network, as a surrogate the subtype characterized by word comprehension defi- for grammatical competence in sentence construction, cits, overlapped parts of the language network known to which is a core function of the language network. This mediate word comprehension.23,24 The IFG was severely is one reason why so many patients with effortful speech atrophied only in PPA-G, a relationship that is consis- have been described as having PNFA, sometimes with- tent with the role of this area in syntax, fluency, and other out full documentation of a language impairment. In the aspects of grammatical competence.25 The atrophy in present study, we used a newly developed and easily ad- PPA-G also extended into other areas of the premotor ministered instrument, the NAT, to directly assess the and dorsolateral prefrontal cortices, a distribution that production of syntactically correct sentences. The test- may reflect the close relationship of this variant with cor- ing method minimizes the effect of poor single-word com- ticobasal degeneration.26 prehension and working memory deficits on perfor- In PPA-L, the major atrophy was in the posterior parts mance and dissociates low fluency from grammatical of the language network, including the TPJ and Brod- competence in constructing sentences. mann area 37. In light of new functional imaging data, it We chose the PPVT-4 for single-word comprehen- seems as if the TPJ, partially overlapping the Wernicke area, sion. Scores on the PPVT-4 had no significant correlation may not be critical for decoding the meaning of words de- with NAT scores and, therefore, assessed an orthogonal noting concrete objects and that this aspect of language aspect of language function. The face validity of the PPVT-4 may more closely depend on more anterior parts of the as a test of word comprehension was shown by its high lateral temporal lobe.23 The TPJ, especially the posterior concordance with the purely verbal paired word associa- part of the superior temporal gyrus, may play a particu- tion test administered to the same patients. We selected a larly important role in phonologic encoding,27 and its at- subset of items with difficulty levels that are likely to avoid rophy in PPA-L may underlie the frequent phonemic para- floor or ceiling effects. However, the cutoff level chosen phasias described in this variant.22 Brodmann area 37 has for this group of patients may need to be altered for popu- been linked to modality-independent lexical access,18 an lations with different educational levels. In fact, we have affiliation that is consistent with the word-finding impair- since found that it may be preferable to use only 24 of the ment characteristic of PPA-L. items (items 157-180) to decrease the rate of false- Delineation of PPA-L based on the preservation of positive results in the identification of the PPA-S variant. grammar and semantics may raise the concern that it The 2-dimensional mapping, based on PPVT-4 and NAT may merely reflect a less severe form of PPA rather than scores, with cutoff levels at 60%, allowed us to subtype a separate variant. However, note that the impaired patients in a manner that fit the descriptive clinical diag- word finding in PPA-L, often accompanied by addi- nosis made before the availability of PPVT-4 and NAT tional errors in spelling and calculation, can cause as scores. The other language tests shown in Figure 1 pro- much functional disability as arises in the other PPA vided supplementary but less specific information. Se- variants. As the disease progresses, patients with PPA-L vere impairments in naming on the Boston Naming Test may become more and more nonfluent because of fre- were seen only in PPA-S. However, a low Boston Naming quent word-finding hesitations. In our experience, Test score is unlikely to be specific to PPA-S because low however, such progression rarely, if ever, leads to emer- scores could also reflect impairments in lexical retrieval gence of the prominent impairments of sentence con- even when comprehension is intact. Fluency was lowest struction or semantics characteristic of PPA-G and in PPA-G, but 3 patients with PPA-L also had distinctly PPA-S. The PPA-L variant, therefore, has a trajectory of abnormal fluency scores (patients 10, 13, and 16), even progression that usually continues to distinguish it in the absence of motor or apraxic speech impairments. from the other PPA subtypes. Repetition abnormalities have been reported to consti- The most critical step in the process of subtyping is the tute a distinguishing feature of PPA-L.22 This was not the accurate root diagnosis of PPA and its delineation from pa- case in the present patients, probably because the WAB tients whose main problem lies in the areas of visual ag- repetition subtest is too easy for patients with relatively nosia, motor speech impairment, or amotivational states. mild impairment. Naming was preserved in some pa- Equally important is the need to eliminate patients whose tients with PPA-L and was only mildly impaired in oth- progressive aphasia emerges on a background of equally ers, leaving word-finding hesitations as the major area of severe amnesia, agnosia, or apathy. Once the root diagno- impairment in the spoken language of these patients. In sis of PPA has been made, the subsequent clinicoanatomi- clinical practice, we see patients with PPA-L and promi- cal subtyping can be achieved on the basis of 2 easily ad- nent retrieval-based object-naming deficits, although such ministered tests of syntax and semantics (Table 2). The patients were not represented in the present sample. literature2,3 indicates that PPA-G, PPA-S, and PPA-L have

(REPRINTED) ARCH NEUROL / VOL 66 (NO. 12), DEC 2009 WWW.ARCHNEUROL.COM 1550

©2009 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/04/2021 different probabilities of being linked to AD, FTLD-T, and ministrative, technical, and material support: Mesulam. FTLD-TDP. Future postmortem studies will show whether Study supervision: Mesulam and Weintraub. the subtyping algorithm described herein and validated on Financial Disclosure: None reported. a relatively small sample of 16 patients will lead to similar Funding/Support: This study was supported by grant relationships in additional samples and reliably improve DC008552 from the National Institute on Deafness and prediction of the underlying neuropathologic condition. Other Communication Disorders and by grant AG13854 All PPA subtypes share the common denominator of se- (Alzheimer’s Disease Center) from the National Insti- lective atrophy in the language network. The present sub- tute on Aging. typing approach is based on the nature of the most impaired language function at the early to middle stages of disease severity. This does not mean that other language REFERENCES functions in a subtype are intact. For example, patients with 1. Rabinovici GD, Jagust WJ, Furst AJ, et al. Aß amyloid and glucose metabolism in PPA-S may have a substantial proportion of their naming three variants of primary progressive aphasia. Ann Neurol. 2008;64(4):388-401. errors caused by lexical retrieval rather than by word com- 2. Mesulam M, Wicklund A, Johnson N, et al. Alzheimer and frontotemporal pa- prehension impairments, and many patients with PPA-L thology in subsets of primary progressive aphasia. Ann Neurol. 2008;63(6): 709-719. and PPA-G may show impairments in semantic prim- 3. Knibb JA, Xuereb JH, Patterson K, Hodges JR. Clinical and pathological charac- ing.11,28,29 It is, therefore, important to keep in mind that terization of progressive aphasia. Ann Neurol. 2006;59(1):156-165. 4. Neary D, Snowden JS, Gustafson L, et al. Frontotemporal lobar degeneration: a although subtypes are defined by the nature of the most consensus on clinical diagnostic criteria. Neurology. 1998;51(6):1546-1554. severe impairment, intersubtype boundaries become fuzzy 5. Gorno-Tempini ML, Dronkers NF, Rankin KP, et al. Cognition and anatomy in three when components of language function other than those variants of primary progressive aphasia. Ann Neurol. 2004;55(3):335-346. 6. Mesulam M-M, Weintraub S. Primary progressive aphasia and kindred disorders. of peak impairment are considered. As the disease In: Duyckaerts C, Litvan I, eds. Handbook of Clinical Neurology. New York, NY: progresses, testing may become increasingly difficult, and Elsevier; 2008:573-587. subtypes may no longer be identifiable. Also note that 7. Mesulam MM. Primary progressive aphasia: a language-based dementia. N Engl J Med. 2003;349(16):1535-1542. “grammar” and “word comprehension” are exceedingly 8. Rogalski EJ, Mesulam MM. Clinical trajectories and biological features of pri- complex constructs and that the NAT and the PPVT-4 cap- mary progressive aphasia (PPA). Curr Alzheimer Res. 2009;6(4):331-336. 9. Kertesz A. Western Aphasia Battery. San Antonio, TX: Psychological Corp; 1982. ture only a fragment of the corresponding processes. Nev- 10. Oldfield RC. The assessment and analysis of handedness: the Edinburgh inventory. ertheless, our goal was to provide a conceptual frame- Neuropsychologia. 1971;9(1):97-113. work for mapping subtypes according to performance along 11. Mesulam M, Rogalski E, Wieneke C, et al. Neurology of anomia in the semantic variant of primary progressive aphasia. Brain. 2009;132(pt 9):2553-2565. these 2 orthogonal subdomains of language, with the 2 tests 12. Dunn LM, Dunn DM. Peabody Picture Vocabulary Test. 4th ed. Toronto, On- serving as reliable (albeit partial) markers of impairment. tario: Pearson Canada Assessment Inc; 2006. Methods of classification tend to evolve, and the pres- 13. Weintraub S, Mesulam MM, Wieneke C, Rademaker A, Rogalski EJ, Thompson CK. The Northwestern Anagram Test: measuring sentence production in pri- ent approach will almost certainly be improved in the fu- mary progressive aphasia [published online ahead of print August 21, 2009]. ture. Other tests of grammar and semantics may prove to Am J Alzheimers Dis Other Demen. 2009;24(5):408-416. 14. Kaplan E, Goodglass H, Weintraub S. The Boston Naming Test. Philadelphia, PA: be more useful, and the cutoff level of performance will need Lea & Febiger; 1983. to be adjusted to accommodate different linguistic and edu- 15. Goodglass H, Kaplan E, Barresi B. Boston Diagnostic Aphasia Examination. 3rd cational backgrounds. The goal of the present study was ed. Austin, TX: Pro-Ed; 2001. 16. Genovese CR, Lazar NA, Nichols TE. Thresholding of statistical maps in functional to demonstrate the feasibility of a simple 2-dimensional tem- neuroimaging using the false discovery rate. Neuroimage. 2002;15(4):870-878. plate for mapping the major subtypes of PPA. Eventually, 17. Price CJ. The anatomy of language: contributions from functional neuroimaging. biomarkers will emerge and clinical subtyping will no longer J Anat. 2000;197(pt 3):335-359. 18. DeLeon J, Gottesman RF, Kleinman JT, et al. Neural regions essential for dis- serve the purpose of predicting the underlying neuropatho- tinct cognitive processes underlying picture naming. Brain. 2007;130(pt 5): logic condition. Even then, however, subtyping will help 1408-1422. 19. Mesulam M-M. Large-scale neurocognitive networks and distributed process- explore the molecular mechanisms that make individual ing for attention, language, and memory. Ann Neurol. 1990;28(5):597-613. sectors of the language network the selective targets of dif- 20. Benson F. Aphasia and related disorders: a clinical approach. In: Mesulam M-M, ferent neuropathologic diseases. ed. Principles of Behavioral Neurology. Philadelphia, PA: FA Davis; 1985:193- 238. 21. Mesulam M-M, Weintraub S. Spectrum of primary progressive aphasia. In: Rossor Accepted for Publication: June 24, 2009. MN, ed. Unusual Dementias. London, England: Baillière Tindall; 1992:583-609. Correspondence: Marek-Marsel Mesulam, MD, Cogni- 22. Gorno-Tempini ML, Brambati SM, Ginex V, et al. The logopenic/phonological variant of primary progressive aphasia. Neurology. 2008;71(16):1227-1234. tive Neurology and Alzheimer’s Disease Center, North- 23. Gitelman DR, Nobre AC, Sonty S, Parrish TB, Mesulam M-M. Language network western University, 320 E Superior St, Searle 11-453, specializations: an analysis with parallel task design and functional magnetic resonance imaging. Neuroimage. 2005;26(4):975-985. Chicago, IL 60611 ([email protected]). 24. Spitsyna G, Warren JE, Scott SK, Turkheimer FE, Wise RJS. Converging language Author Contributions: All authors had full access to all streams in the human temporal lobe. J Neurosci. 2006;26(28):7328-7336. the data in the study and take responsibility for the in- 25. Indefrey P, Hellwig F, Herzog H, Seitz RJ, Hagoort P. Neural responses to the production and comprehension of syntax in identical utterances. Brain Lang. 2004; tegrity of the data and the accuracy of the data analysis. 89(2):312-319. Study concept and design: Mesulam, Rogalski, Thomp- 26. Kertesz A, Martinez-Lage P, Davidson W, Munoz DG. The corticobasal degen- son, and Weintraub. Acquisition of data: Mesulam, eration syndrome overlaps progressive aphasia and frontotemporal dementia. Neurology. 2000;55(9):1368-1375. Wieneke, and Thompson. Analysis and interpretation of 27. Graves WW, Grabowski TJ, Mehta S, Gupta P. The left posterior superior tem- data: Mesulam, Wieneke, Rogalski, Cobia, Thompson, poral gyrus participates specifically in accessing lexical phonology. J Cogn Neurosci. 2008;20(9):1698-1710. and Weintraub. Drafting of the manuscript: Mesulam and 28. Rogalski E, Rademaker A, Mesulam M, Weintraub S. Covert processing of words Rogalski. Critical revision of the manuscript for important and pictures in nonsemantic variants of primary progressive aphasia. Alzheimer intellectual content: Mesulam, Wieneke, Rogalski, Dis Assoc Disord. 2008;22(4):343-351. 29. Vandenberghe RR, Vandenbulcke M, Weintraub S, et al. Paradoxical features of Cobia, Thompson, and Weintraub. Statistical analysis: word finding difficulty in primary progressive aphasia. Ann Neurol. 2005;57 Rogalski and Thompson. Obtained funding: Mesulam. Ad- (2):204-209.

(REPRINTED) ARCH NEUROL / VOL 66 (NO. 12), DEC 2009 WWW.ARCHNEUROL.COM 1551