<<

J7ournal ofNeurology, Neurosurgery, and Psychiatry 1996;60:403-410 403

Progressive supranuclear palsy presenting with J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.60.4.403 on 1 April 1996. Downloaded from dynamic

Thomas Esmonde, Elaine Giles, John Xuereb, John Hodges

Abstract have been noted in some patients with PSP Background-Progressive supranuclear during the course of their illness,2-6 descrip- palsy (PSP) is an akinetic-rigid syndrome tions of a presentation with a disorder of spo- of unknown aetiology which usually pre- ken language production are virtually absent sents with a combination of unsteadiness, from the medical literature; Perkin et al8 bradykinesia, and disordered eye move- reported five patients with atypical presenta- ment. often becomes dysarthric tions, two of whom had severe dysphasia in but language disorders are not well recog- the context of mild global dementia, one made nised. naming errors and later developed unintelligi- Methods-Three patients with PSP ble speech, and the other was described as a (pathologically confirmed in two) are non-fluent dysphasic. The lack of awareness of reported in which the presenting symp- this aspect is illustrated by the fact that two toms were those of difficulty with lan- recent books devoted entirely to PSP,9 10 guage output. although mentioning mild word finding diffi- Results-Neuropsychological testing culty and reduction in speech output during showed considerable impairment on a the course of the disease, do not refer to this range of single word tasks which require presenting feature. We have seen three active initiation and search strategies patients with PSP in whom the initial presen- (letter and category fluency, sentence tation was one of a verbal adynamia, resem- completion), and on tests of narrative bling the phenomenon described by Luria," language production. By contrast, nam- also sometimes termed dynamic aphasia.'2 All ing from pictures and from verbal three underwent detailed neuropsychological descriptions, and word and sentence and linguistic assessment, and in two the diag- comprehension were largely intact. The nosis was confirmed pathologically. degree of semantic memory impairment was also slight. Conclusions-Relatively selective involve- Patient 1 ment of cognitive processes critical for A 62 year old gardener presented in 1990 to a planning and initiating language output psychiatrist after a two year history of a change may occur in some patients with PSP. in personality. Initially, he had been disinhib- http://jnnp.bmj.com/ This presentation resembles the phenome- ited and aggressive, but had subsequently non of "verbal adynamia" or "dynamic become anergic and apathetic. Over this aphasia" seen in patients with frontal lobe period it was noticed by the patient, and his damage. Although definite cortical family, that his expressive output in both changes were present at postmortem speech and writing had diminished greatly. examination, it is likely that the neu- Practical abilities around the home seemed

ropsychological deficits reflect functional preserved. Examination in 1991 showed him on September 24, 2021 by guest. Protected copyright. Department of , Royal frontal deafferentation secondary to to be very non-fluent with pronounced word Victoria Hospital, interruption of frontostriatal feedback finding difficulties. Speech was limited to sin- Belfast, UK loops. gle words or short phrases which were gram- T Esmonde matically correct without phonological or The University (3 Neurol Neurosurg Psychiatry 1996;60:403-4 10) semantic paraphasias. By contrast, confronta- Neurology Unit, Addenbrooke's tion naming, word and sentence repetition, Hospital, Cambridge, and comprehension of simple and complex UK Keywords: progressive supranuclear palsy; dynamic commands were strikingly normal. Physical E Giles aphasia J Hodges examination showed a mild degree of supranu- clear vertical gaze limitation and poor horizon- Department of Pathology, Progressive supranuclear palsy (PSP) as origi- tal saccadic movements. His gait was wide Addenbrooke's nally described by Richardson et al in the based and slightly unsteady, but Romberg's Hospital, Cambridge, 1960s' characteristically presents as a parkin- test was negative. Other extrapyramidal signs UK J Xuereb sonian syndrome with axial rigidity, bradyki- were absent. Correspondence to: nesia, postural instability, and paralysis of Brain CT showed a mild degree of frontal Dr J R Hodges, Department conjugate gaze for voluntarily guided eye atrophy. Examination of CSF was normal. of Neurology, Addenbrooke's Hospital, movements. Dysarthria and swallowing diffi- The diagnosis at this stage was unclear, but Hills Road, Cambridge, UK. culties occur as a result of involvement of the PSP or dementia of frontal lobe type were Received 18 May 1995 muscles of articulation by extrapyramidal considered the most likely possibilities. and in final revised form 8 December 1995 rigidity. Although dysphasia (of a generally Over the ensuing 12 months, anxiety and a Accepted 15 December 1995 unspecified type) and word finding difficulties preoccupation with death became prominent 404 Esmonde, Giles, Xuereb, Hodges

features. Axial rigidity became pronounced, died after a total duration of illness of five J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.60.4.403 on 1 April 1996. Downloaded from and there was a loss of all voluntary eye move- years. ments. He became increasingly immobile and often fell. Speech was restricted to single word POSTMORTEM EXAMINATION repetitive utterances before mutism super- Macroscopically, the brain showed mild gen- vened in the terminal stages of his illness. He eralised gyral atrophy and pallor of the sub- died four years after the onset of the first stantia nigra. Microscopical examination symptoms. showed a few neurofibrillary tangles in the frontal cortex and an occasional tangle in the POSTMORTEM BRAIN EXAMINATION temporal cortex. Senile plaques were absent. Macroscopical examination of the brain Nerve cell loss and gliosis were present in the showed a degree of cerebral gyral atrophy, subthalamic nucleus. Neurofibrillary tangles more pronounced in the lateral frontal regions, were numerous in the , and pallor of the and locus Meynert's nucleus, the thalamus, the hypo- coeruleus in the brain stem. Microscopically, thalamus, and the subthalamic nucleus. severe nerve cell loss and gliosis were evident Tangles were also prominent in the red in the subthalamic nucleus and the central nucleus, substantia nigra, the nucleus of mesencephalic grey. Less severe neuronal loss Darkshevich, central mesencephalic grey, was present in the brain stem, pigmented brain stem pigmented nuclei, reticular forma- nuclei, and the dentate nucleus. Surviving tion, pontine nuclei, accessory dorsal olivary nerve cells in these subcortical and brain stem nucleus, and the dentate nucleus. As in the areas contained tau positive neurofibrillary previous patient, the neuropathological find- tangles; tangles were also present in the globus ings were diagnostic of PSP. pallidus, the red nucleus, the brain stem reticu- lar formation, the pontine nuclei, and particu- larly in the nucleus of Darkshevich in the Patient 3 rostral midbrain. Neurofibrillary tangles were A right handed housewife and exwarden of a numerous in the frontal cortex but there were nursing home presented at the age of 65 with a very few in the temporal cortex; senile plaques one year history of word finding difficulty in were sparse. The type and distribution of speech and very impaired conversational abili- lesions confirmed the diagnosis of PSP. ties. Her ability to use her right hand in writ- ing had slowed up and there was a tendency to overbalance easily. Despite the reduced verbal Patient 2 output and clumsiness, she continued to com- A retired dispatch manager presented at the plete her daily crossword puzzle. age of 67 when his wife noticed that over the She had somewhat immobile facies, mild previous two years there had been a reduction axial rigidity, and a hesitant gait. Voluntary in his speech, which was now reduced to single upward saccades and pursuit movements were words or short phrases. His comprehension abnormal. Her speech was mildly dysarthric seemed intact and he was only mildly forget- and slow, with reduced phrase length. ful. On further enquiry, it became apparent Paraphasic errors were absent. Naming, word that more recently his personality had changed and sentence repetition, and comprehension http://jnnp.bmj.com/ to the extent that he had become rigid, irrita- were normal. ble, and had lost his sense of humour, but he Brain CT, chest radiograph, thyroid func- did not complain of any of these difficulties tion, syphilis serology, and CSF examination spontaneously. His general abilities around the were all normal. home seemed normal, and he had been able to She was treated with levodopa, lisuride, and cook and clean when left on his own while his amantadine, without noticeable response and

wife was in hospital for a cataract operation. A her condition deteriorated. Five years from the on September 24, 2021 by guest. Protected copyright. degree of postural instability was evident from onset of first symptoms, she was chairbound the fact that he had had several falls. with very dysarthric speech and minimal spon- He had a dyspraxic gait and postural insta- taneous utterances. Rigidity was pronounced. bility. Eye movements were abnormal with Occasional instances of inability to inhibit an reduced down gaze pursuit and very slow vol- initiated movement were seen, for example, untarily guided horizontal and vertical sac- repetitive hand rubbing. Voluntary conjugate cades. He had prominent pout and bilateral eye movements were severely restricted, espe- grasp reflexes. His speech was appropriate but cially in the vertical plane. very reduced in output, consisting of short, On the basis of the criteria used by Golbe et grammatically correct phrases. Paraphasic al,'3 she conformed to a clinical diagnosis of errors were absent. Naming, word and sen- PSP (supranuclear gaze palsy, onset after 40, tence repetition, and comprehension were progressive course, bradykinesia, rigidity normal. He was fully oriented and showed greater axially than in the limbs, dysarthria, good recall of novel verbal material (a fictional and the absence of ). name and address). Brain CT showed mild frontal lobe atrophy, and CSF examination was normal. Neuropsychological assessment Over the subsequent three years, his condi- A battery of neuropsychological tests was tion steadily deteriorated, and there was no given to each patient to assess a broad range of response to treatment with levodopa. He cognitive abilities including memory, visuo- became dysphagic, mute, and immobile and perceptual functions, expressive language, Progressive supranuclearpalsy presenting with dynamic aphasia 405

Table 1 General cognition and memory J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.60.4.403 on 1 April 1996. Downloaded from Duration Dementia Logical Logical Recognition Recognition ofdisease rating memory memory memory memoty Digits Digits Rey Rey (months) MMSE scale (immediate) (delayed) (words) (faces) forward backward copy recall Patient 1 36 22 90 5-75 2-25 33 28 6 4 2-5 1 41 17 NT 3 1-25 NT NT 5 3 NT NT Patient 2 24 24 124 2-5 1-5 26 28 6 4 11 6 30 22 121 4-75 1-25 32 46 6 2 14 7 40 18 101 3 2-25 NT 37 6 3 4-5 2 48 21 80 3-75 1 NT NT 6 2 NT NT Patient 3 12 28 138 12 10-5 44 43 5 4 33 8 18 29 134 12-25 11 49 38 5 5 35 13-5 28 29 126 12 9 41 35 5 3 34 14 Maximum score 30 144 24 24 50 50 36 36 Controls 29-2 140-5 23-2 17 47-3 43-7 6-8 4-8 34 15-3 (SD) (1-0) (2 4) (7 8) (6 8) (2-8) (3 8) (1-0) (1-2) (3 0) (7 4) NT = Not tested; MMSE = mini mental state examination.

comprehension, and semantic knowledge. consisting of a photograph of a common The patients' results are illustrated in the object together with a photograph of the same accompanying tables and are compared with a object taken from a different angle and a visu- group of 24 normal volunteer subjects (18 ally similar foil. Patients 2 and 3 performed women, six men) with mean age 69-7 (SD normally at presentation but the performance 7 8) years, and mean 10-8 (SD 2 3) years edu- of patient 2 subsequently declined on follow cation selected from the MRC Applied up. The ability to copy the Rey figure was sig- Psychology Unit's subject panel. nificantly impaired in two patients (1 and 2), but preserved in the other. Two patients (2 GENERAL COGNITION and 3) also experienced difficulty on the Assessment of global cognitive function was judgement of line orientation task, obtaining restricted in patient 1, and to a lesser extent in scores of 19/30 and 23/30 respectively.20 patient 2, by their severe communication diffi- culties. Patient 3 performed normally on the mini mental state examination'4 and the Expressive language abilities dementia rating scale,'5 whereas patient 2 SINGLE-WORD PRODUCTION TASKS: VERBAL showed a mild degree of impairment on both FLUENCY AND NAMING ABILITIES measures. All three showed a significant One of the most striking dissociations in all decline on these measures late in the course of three patients was that between confronta- the disease (table 1). tional naming and verbal fluency. Two com- monly used verbal fluency tasks-letter and MEMORY category fluency-were given to all three Immediate auditory-verbal short-term mem- patients. For letter fluency, they were asked to ory, as measured by digit span forwards, was generate as many words as possible beginning normal at presentation in all three patients, with the letters "F, A, and S" excluding proper but became impaired as their disease pro- nouns and the same word with different suf-

gressed (table 1). Tests of longer term verbal fixes (fix, fixed, fixing, etc). For category flu- http://jnnp.bmj.com/ and visual memory showed more pronounced ency tests, subjects were given one minute for impairment, based on both recall (logical each of four "living" categories (animals, memory)'6 and recognition paradigms,'7 in birds, water creatures, and breeds of dog) and two of the three patients. Interpretation of the four "manmade" categories (household logical memory results was, however, compli- objects, vehicles, musical instruments, and cated by the patients' spoken language deficits. types of boat). On both initial letter and Although ability to copy the Rey figure'8 was semantic category based tasks all three impaired, it is notable that the percentage patients showed very pronounced impairment on September 24, 2021 by guest. Protected copyright. recalled was similar to that of controls. with a progressive decline over time. The nor- mal relation of better performance on category PERCEPTION fluency was maintained in all three (table 2). Basic perceptual abilities were assessed with The drastically impaired verbal fluency con- an object matching task'9 in which the subject is trasted with their normal, or near normal, per- presented simultaneously with three pictures formance on tests of naming ability, using as

Table 2 Language: single-word production tasks Duration Categoryfluency Naming ofdisease Naming to (months) Letterfluency Living x 4 Manmade x 4 pictures description Patient 1 36 3 17 15 36 17 41 4 15 13 42 20 Patient 2 24 7 25 24 38 18 30 6 27 20 42 17 40 4 19 12 43 18 48 7 16 12 45 15 Patient 3 12 21 30 29 42 23 18 13 27 23 45 24 28 10 23 24 44 23 Maximum score 48 24 Controls (SD) 44-6 (10-2) 58-3 (12-3) 56-2 (8-4) 43-6 (2-3) 22-5 (1-4) NT = Not tested. 406 Esmonde, Giles, Xuereb, Hodges

stimuli either 48 line drawings of animals and you say?" J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.60.4.403 on 1 April 1996. Downloaded from objects or 24 verbal descriptions (for example, Patient 2: "That's not mine." "What do we call a small vehicle with runners, Patient 3: "Somebody left a glove . .. a sheep- used on snow or ice" answer: sledge); the skin glove . .. somebody has dropped it." verbal description tasks were taken from the Each of the 10 responses were scored for semantic battery which has previously been understandability (information content and described in detail.2' 22 Examination of the ability to convey the appropriate message) and errors made on picture naming showed that intelligibility (the clarity of speech in terms of most were either visual confusions (for exam- phonetic, articulatory and prosodic compo- ple, "a row of books" for accordion) or cir- nents) on a five point scale (total score for cumlocutions ("shellfish with claws" for understandability and intelligibility = 50) lobster). according to the criteria applied by Blomert et al.25 Controls obtained mean scores of 95-6 NARRATIVE AND DISCOURSE ABILITIES (SD 4-8)% and 91*6 (SD 11-5)% for under- As one of the principal complaints of the standability and intelligibility respectively. The patients' families was the patients' lack of con- content of the responses of patient 2 was versation, we applied some measures aimed at reduced, and although appropriate, lacked assessing narrative language skills. normal conversational richness with a resul- The cookie theft picture description task23 tant laconic and sometimes telegrammatic requires subjects to produce a verbal descrip- style; he scored 62% on both understandabil- tion of a drawing of a domestic scene. All three ity and intelligibility of the message. On a sec- patients showed considerable poverty of lan- ond administration of the test, eight months guage output when presented with this task later, perseveration was prominent, with a ten- and also tended to perseverate. Their dency to echo the words used by the examiner; responses were analysed according to the his scores at this stage had fallen to 40% and number of T units and morphemes they con- 20%. Patient 3 scored within the normal range tained. A T unit is an independent clause (a (86% for understandability and 100% for verb and a noun) with its dependent clauses.24 intelligibility). Patient 1 was not available for A morpheme is the smallest unit of meaning testing. (for example, "house" = 1 morpheme, "houses" = 2 morphemes, "give" = 1 mor- SENTENCE COMPLETION TASKS pheme, "gave" = 2 morphemes). Trans- Single-word completions criptions of the patients' descriptions are To evaluate word finding and production abil- reproduced in the appendix with the T unit ity in a task which requires a constrained, but and morpheme scores. untimed semantic search, we used a sentence Patients 1 and 3 showed a progressive completion test based on that used by Costello decline in the total number of morphemes and Warrington'2 in their examination of a produced, but in patient 3 the absolute num- patient with dynamic aphasia resulting from a ber of T units remained fairly constant over left frontal astrocytoma. In this task, the sub- this period resulting in a steady fall in the ratio jects were given the stem of a sentence and of morphemes to T units. Other features of the then asked to provide an appropriate single

patients' narratives are the occasional syntactic word to complete it. Twenty sentences http://jnnp.bmj.com/ errors with omission of appropriate function required a noun for completion, and 20 a verb. words (particularly striking in patient 2), the In a half of the noun and verb completion sen- pauses while searching for words resulting in tences, the possible choice was limited by the aborted phrases, and the tendency to persever- stem (closed sentences- for example, "Most ation. cats see very well at . . ", "He mailed the let- To assess narrative language skills in a more ter without a . . . "), and in the remainder, the naturalistic setting, we applied the choice was wide (open sentences-for example, Amsterdam-Nijmegen everyday language "The man with a moustache wore a. . on September 24, 2021 by guest. Protected copyright. test,25 which has been extensively used to "Clumsily the woman loosened her. . . monitor recovery in patients with aphasic The control subjects had no difficulty with this . In this task, the subject is asked to task, and produced an average of only 1 1 imagine themselves in 10 everyday situations, (2 75%) errors for the 40 sentences which which they then describe in appropriate con- consisted of violations of the task requirements versational speech to the examiner. For exam- by adding more than one word to complete ple: the sentence, or occasional failures to produce 1 Examiner: "You have just moved to my a response. Patient 2 made 12 errors (30%), street, and you would like to meet me. You tending to perseverate words from the sen- ring my doorbell and say?" tence stem, or the previous response. Patient 3 Patient 2: "Meet me. Meet me. Meet me." also made 12 (30%) errors, violating the rules Patient 3: "How about meeting for a cuppa?" of the task consistently by producing short 2 Examiner: "You see your neighbour walk- phrases (rather than words). Patient 1 was ing by. You want to ask him to visit you some- unavailable by the time we devised this task. time. What do you say?" Patient 2: "Visit me. Visit me . . . That's all" Phrase completions Patient 3: "Could you come to visit me some- In a variant of the above task, we asked sub- time?" jects to complete 40 sentences with a short 3 Examiner: "You are at the butcher's and phrase. Controls made an average of 0 6 this is lying on the floor (a glove). What do (1 5%) errors which were either violations or Progressive supranuclear palsy presenting with dynamic aphasia 407

non-responses. Patient 2 produced nine used in the naming test. Four of the questions J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.60.4.403 on 1 April 1996. Downloaded from bizarre responses (22 5%) which failed to take are designed to elicit a "yes" answer, and four a account of the prosodic and semantic informa- "no" response. Two patients (2 and 3) per- tion in the cue (for example, "The girls had formed normally on this task (table 3), and

tried hard but" . . . "freedom"; "The business although patient 1 (and subsequently patient had crumbled so"... "the boy thought"; 2) produced results outside the normal range,

"The husband disappeared last year and" .. . there was evidence that they continued to "a war"). Patient 3 was unable to formulate a retain a significant amount of semantic knowl- response in 12 out of the 40 (30%) sentences, edge (77% and 80% of maximum score but did not produce any erroneous responses. respectively on their worst performances). The definitions subtest was given to each patient at presentation. In this test, subjects Semantic and syntactic comprehension are asked to generate definitions of 12 of the Comprehension of single word meaning was items in the battery in response to the spoken largely intact in all three patients. On the word word, each within the space of one minute. picture matching test from the psycholinguis- Normal subjects produce a mean of 6-8 pieces tic assessments of language processing in of factual information with hardly any perse- aphasia,*26 in which each target item is dis- verative or intrusive errors.21 The information played together with a close and distant supplied by each of the three patients was cor- semantic foil, a visually related foil, and an rect, but there was a diminution in the amount unrelated foil, two of the patients achieved produced (mean of 2-6, 5-6, and 5X 1 in near perfect scores (table 3). Likewise, on the patients 1 to 3 respectively) with frequent per- word-picture matching subtest from the severation. For example: semantic battery2l 22 in which subjects are pre- Patient 1: sented with eight target line drawings that Sledge: "You pull it." belong to the same category on an A4 sheet Lamp: "Needs a bulb in. Put it on. Needs and asked to point to one named by the exam- electricity."

iner (for example, eight different animals: Accordion: "Buttons on it ... got buttons on deer, rabbit, horse, cow, squirrel, fox, cat, and it, squeeze-box, squeeze it. Got it in front of mouse), the same two patients (2 and 3) per- me." formed normally. The impaired performance Patient 2: of patient 1, and the slight subsequent decline Crocodile: "An animal, it's got skin, lovely in patient 2, may reflect their considerable skin, it's got four feet, it swims, it's got a tail. problems with visual scanning. Eats man. Water-it likes water." Conceptual or semantic knowledge was Fox: "Bushy tail. Has four legs. It stinks. A probed with other subtests of the semantic sharp nose. Has fur all over it." battery, which assesses knowledge about the Patient 3: same 48 items-half living and half man- Peacock: "A bird. Has a big tail which it fans made-across a range of tests.2' 22 In the sort- out when its mating. They're quite big." ing task, the subject has to sort picture cards Toaster: "Household item. Can be for two or into defined categories at three levels (level 1: four pieces of toast. Its either pop-up or manu- living v manmade; level 2: land animal v bird v ally operated." http://jnnp.bmj.com/ water creature; household item v vehicle v Comprehension of syntactic structures was musical instrument; level 3: native v foreign tested using the test for the reception of gram- animal, fierce v non-fierce animal, electrical v mar.27 In this test, the subject is asked to point non-electrical item, etc). All three performed to one of four pictures in response to phrases normally on this task at presentation, but in or sentences of increasing syntactic complexity two patients (1 and 2), performance declined in which the vocabulary remains very simple

as their disease progressed. (for example, "The boys pick the apples"; on September 24, 2021 by guest. Protected copyright. In the semantic feature questionnaire from "The boy is chased by the dog"; "The pencil is the same battery, the subject is asked four on the book that is yellow" etc). In total, there questions relating to physical characteristics are 80 arrays divided into 20 blocks, each of (for example, "Is an elephant grey?") and four which tests comprehension of a particular syn- relating to functional attributes (for example, tactic structure (for example, plurality, gender, "Does an elephant eat meat?") of 24 items locative prepositions, negatives, reversible

Table 3 Semantic and syntactic comprehension Duration PALPA Word- Sorting test Semantic ofdisease Word picture picture feature (months) matching matching (level 1) (level 2) (level 3) questions Case 1 36 27 43 48 41 64 157 41 NT 43 44 35 57 154 Case 2 24 39 48 48 43 68 162 30 39 48 48 38 67 159 40 38 46 46 38 63 147 48 34 42 44 34 54 NT Case 3 12 40 46 48 48 66 177 18 40 48 47 47 68 174 28 40 48 48 46 64 175 Maximum score 40 48 48 48 72 192 Controls (SD) 37-3 (1-3) 47-4 (1 1) 48 (0 2) 46-9 (0-9) 68-8 (2-2) 178 (5-1) NT = Not tested; PALPA = psycholinguistic assessments of language processing in aphasia.26 408 Esmonde, Giles, Xuereb, Hodges

active, reversible passive, postmodified sub- opportunity to vary the phrase, this difficulty J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.60.4.403 on 1 April 1996. Downloaded from ject, relative and embedded sentences). One persisted. patient (3) obtained a near perfect score but One of the other striking findings in our the other two showed evidence of mild impair- patients was the degree of impairment on both ment. letter and category based verbal fluency. Normal subjects perform better on category than on letter based tests.29'2 Although clearly Discussion very poor at both tasks, the "normal" relation The three patients all presented with a consid- between the two tasks was maintained in all erable reduction in verbal output resulting in three patients with PSP, a finding in keeping difficulties with propositional language. With with that of a larger group of patients with disease progression, their speech became radi- PSP.29 This contrasts with the pattern seen in cally impoverished with eventual . At Alzheimer's disease in which category fluency a stage when conversational language and ver- is consistently poorer than letter fluency.29-12 It bal fluency were strikingly impaired they has been argued that the deficit in letter flu- exhibited normal, or near normal, perfor- ency reflects the early breakdown of semantic mance on tests requiring naming from pictures memory found in Alzheimer's disease, whereas and from oral descriptions, as well as tests of the equal impairment in letter and category word and sentence comprehension. Repetition fluency found in PSP probably results from of words and sentences also seemed to be impaired initiation and retrieval processes, intact, although this was not tested formally. which are common to both tasks. Phonological speech errors never occurred. Against this background of grossly impaired This pattern is best described as verbal ady- performance on tests of letter and category namia or dynamic aphasia." 12 Although dys- verbal fluency, and on sentence completion, it phasia is classically regarding as a feature of is notable that all three patients performed cortical dementias (such as Alzheimer's dis- within normal limits on a conventional 48 item ease), rather than diseases which involve pre- picture naming test and on a test requiring dominantly subcortical structures, as noted subjects to name from verbal descriptions. On earlier dysphasia has been reported by several confrontation naming, any errors reflected investigators during the course of PSP,2-6 and perceptual difficulties or mild impairment in on occasions has been described as a promi- access to phonological word forms (circumlo- nent feature at presentation.7 The nature of cutions). Semantic memory, as measured by the language disturbance in these patients has, performance on tasks which require subjects however, not been analysed in any detail and to access the central representational knowl- to the best of our knowledge, PSP presenting edge concerning a range of living and man- with dynamic aphasia has not been recorded made items, was only mildly affected in our previously. It should be noted that subtle patients. They performed normally on two changes in personality were also apparent to tasks requiring the matching of spoken words the relatives of the first two patients. to pictures, although it should be pointed out The reduction in language output in these that such tasks are relatively insensitive to patients progressed in parallel with the time semantic memory impairment.22 On the more course of their disease, and was greatest on demanding picture sorting and semantic fea- http://jnnp.bmj.com/ tests requiring the subject to generate dis- ture questionnaire subtests of the semantic test course, such as the Amsterdam-Nijmegen battery, our patients showed a mild degree of everyday language test,25 or narrative as in the impairment in keeping with the results of a cookie theft picture.23 Verbal output on these larger scale group study of patients with PSP.33 tasks was impoverished; the patients were able The difficulties noted with language pro- to produce appropriate information but could duction cannot, therefore, be adequately to a often manage communicate with incom- explained by loss of semantic information; it on September 24, 2021 by guest. Protected copyright. plete sentences only. Although almost all ele- has been repeatedly shown that even a mild ments of the cookie theft picture were degree of semantic disruption causes substan- mentioned, the information produced was tial impairment on standard tests of confronta- insufficient to describe fully the situation pre- tional naming. A major phonological or sented. As can be seen from the performances articulatory deficit can also be excluded, and of patient 3 on serial administration of the pic- although the patients did show abnormalities ture, there was a progressive decline in her of syntactic structure on elicited speech pro- output, as evidenced by the progressive reduc- duction tasks, a primary disorder of syntax tion in morphemes/T unit.24 cannot explain their profound difficulties with Tasks of sentence completion also proved narrative speech production, letter and cate- difficult for the two patients tested: they pro- gory fluency, and sentence completion tasks. duced an abnormally large number of omis- The characteristics of the language abnor- sions and occasional bizarre responses. With mality in our cases most closely resembles that disease progression, there was a pronounced described by Luria" who delineated the fea- tendency to perseverate using responses from tures of "dynamic aphasia" in patients with previous questions, or to produce echolalic traumatic frontal lobe damage. Luria reported responses with repetition of the sentence stem. that such patients have no difficulty with nam- When the patients were required to produce ing or repetition, but find expression of inter- more than one word in response to the stem, nally generated thoughts and the production which might be expected to place less of a con- of simple verbal statements very difficult due straint on performance because there is more to a "disturbance of the linear scheme of a sen- Progressive supranuclear palsy presenting with dynamic aphasia 409

tence". In a more recent case study, Costello have, however, shown subtle but definite corti- J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.60.4.403 on 1 April 1996. Downloaded from and Warrington"2 described a patient with a cal changes in PSP37 consisting of neocortical large left frontal tumour who had intact naming neurofibrillary tangles of the globose type and reading skills, but an almost total lack of affecting the larger pyramidal and small neu- spontaneous speech. They argued that Luria's rons and most often involving the precentral concept of dynamic aphasia was inappropriate gyrus. A very recent study38 of 17 patients with in their case, and that instead it was a failure at pathologically confirmed Steele-Richardson- the "higher order" stage of initial formulation Olszewski syndrome (a designation preferred of a phrase or sentence. It would seem, there- by the authors since not all patients had fore, that in some patients with dynamic apha- supranuclear gaze palsies) again reported the sia there might be a breakdown at the highest presence of mild cortical pathology that did "pragmatic" level of language planning-that not seem to correlate with the degree of is to say in converting thoughts to speech dementia. It is interesting to note that all of acts-whereas in others the functional locus the patients with supranuclear gaze palsies might be at a lower level of word selection, became demented whereas a lesser proportion sentence planning, and implementation. of those without ocular-motor abnormalities Although slowness in speech production did so, suggesting that the cognitive deficits and dysarthria have been reported in PSP by are indeed related more to subcortical than some authors,34 35 more profound linguistic cortical involvement. defects have been recorded only rarely. For Both of our patients who underwent post- instance, the six patients studied by Podoll et mortem examination had a degree of gener- at'6 had abnormalities of speech production alised cortical atrophy and showed specific that were characterised by short simple sen- neurofibrillary tangles involving the prefrontal tences and a tendency to leave sentences cortex. Despite these cortical changes, the unfinished; but semantics, syntax, and gram- brunt of the neuronal loss and globose tangle mar were found to be normal. The difficulties formation in PSP was in subcortical struc- in language production were considered by the tures, such as the substantia nigra, subthala- authors to be non-specific and secondary to a mic nucleus, nucleus basalis of Meynert, combination of extrapyramidal features and globus pallidus, corpus striatum, , dysarthria. Maher and Lees' enumerated the and various nuclei. It seems likely, clinical features in 52 patients with PSP and therefore, that subcortical structures were the noted dysphasia in four, but the nature of the site of initial pathology and that the dynamic language disturbance was not further eluci- aphasia in our patients is likely to reflect the dated. In a paper discussing brain neurotrans- interruption of frontostrial loops rather than mitter levels in five patients with pathologically direct cortical pathology. Brain PET studies of established PSP, Kish et a14 described severe patients with PSP have shown low striatal and aphasia with paraphasic word substitutions frontal uptake of '8F-2-fluorodeoxyglucose, developing in one patient in the last year of her and reduced and caudate uptake of life. One of the patients reported in a PET 18F-6-fluorodopa6 39; these findings have been study by Leenders et al6 had evidence of dys- interpreted as supporting the notion that the phasia but no further details were provided. In frontal dysfunction in PSP is secondary to a detailed single-case study of language in pathology in the . http://jnnp.bmj.com/ PSP, Lebrun et al5 described a 55 year old In summary, we have presented three cases man whose symptoms included speech diffi- of PSP in whom a disturbance of language culties resembling stuttering with a degree of output was the presenting complaint. Further dysphonia and a tendency to perseverate on analysis of their difficulties has led us to the last word or syllable in a phrase. believe that there may be a failure of language Neuropsychological evaluation showed intact production which is not explainable by articu-

comprehension and syntax in spoken speech, latory or phonological dysfunction or by a fun- on September 24, 2021 by guest. Protected copyright. and choice of words was appropriate. More damental loss of semantic knowledge. The recently, Daniele et al7 reported a patient with occurrence of bizarre responses in some of the a clinical diagnosis of PSP who had presented tests may reflect a failure of executive func- with dysarthria and difficulty writing. More tions responsible for overall supervision of detailed assessment disclosed word finding appropriateness. This phenomenon is well difficulty with semantic errors and omissions recognised in other diseases with predominant on confrontational naming as well as other pathology in the frontal lobes. The language deficits affecting particularly verb, rather than profile seen most closely resembles dynamic noun, processing. aphasia, which has been shown to occur in It is possible that our patients represent a patients with frontal lobe disease and reflects a subgroup of PSP in whom language is particu- breakdown at a high level of language planning larly affected. Alternatively, deficits in lan- and initiation. Dynamic aphasia may be more guage production may be more common than common in PSP than has been realised previ- is realised but may have been overlooked ously. Many issues remain to be considered in because narrative abilities have not been inves- future studies, including the fundamental tigated. question of whether this dynamic aphasia rep- Tuming to the question of the neu- resents a breakdown in all aspects of higher roanatomical and pathological basis for the order thought, language, and action planning, dynamic aphasia, PSP is traditionally regarded or is restricted to the domain of expressive lan- as a prototypic form of subcortical dementia. guage. It also remains open to question More contemporary neuropathological studies whether the observed neuropsychological pro- 410 Esmonde, Giles, Xuereb, Hodges

file reflects functional frontal deafferentation 7 Daniele A, Giutolisi L, Silveri MC, Colosimo C, Gainotti J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.60.4.403 on 1 April 1996. Downloaded from secondary to the interruption of frontostriatal G. Evidence for a possible neuroanatomical basis for lex- ical processing of nouns and verbs. Neuropsychologia feedback loops, or direct frontal cortical 1994;32: 1325-41. involvement, although the balance of evidence 8 Perkin GD, Lees AJ, Stem GM, Kocen RS. Problems in the diagnosis of progressive supranuclear palsy (Steele- perhaps favours the first explanation. Richardson-Olszewski syndrome). Can 7 Neurol Sci 1978; 5: 167-73. 9 Litvan I, Agid Y. Progressive supranuclear palsy: clinical and Appendix research approaches. New York: Oxford University Press, PATIENT 1 1992. (1) 36 months duration of illness 10 Tolosa E, Duvoisin R, Cruz-Sanchez FF. Progressive "The little boy's nearly toppling supranuclear palsy: diagnosis, pathology, and therapy. Wien: over/the cupboard's Springer-Verlag, 1994. open the wrong way/washing up the wrong way/cup- 11 Luria AR. The working brain. Harmondsworth: Penguin board's open the wrong way/ the boy's stool. . the Books, 1973. 12 boy's stool . .. the boy's stool started to back up." Costello A de L, Warrington EK. Dynamic aphasia: the selective impairment of verbal planning. Cortex 1989; Total T units = 5. Total morphemes in T units = 36. 25:103-14. Morphemes/T unit = 7-2. 13 Golbe LI, Davis PH, Schenberg BS, Duvoisin RC. Prevalence and natural history of progressive supranu- (2) 41 months duration clear palsy. Neurology 1988;38:1031-4. of illness 14 Folstein MF, Folstein SE, McHugh PR. "Mini-mental "Everything I see going on in the picture... yeah state". A practical method for grading the mental state of ... yeah... (mumbles) hungry/washing up patients for the clinician. _7 Psychiat Res 1975;12:189-98. 15 Mattis S. Dementia rating scale. Windsor: NFER-Nelson, (mumbles)." 1992. = Total T units 0 (no descriptive information). 16 Wechsler DA. Wechsler memory scale-revised. San Antonio: Psychological Corporation, 1987. PATIENT 2 17 Warrington E. Recognition memory test. Windsor: NFER- (1) 40 months duration of illness Nelson, 1984. 18 Rey A. Le test de copie defigure complexe. Paris: Editions cen- "Washin up... overflowing... cookie jar... wash- tre de psychologie appliquee, 1959. ing/little girl putting her hand up, and window, and 19 Humphreys GW, Riddoch MJ.The fractionation of visual In: tree, and curtains/she's got shoes on... her agnosia. Humphreys GW, Riddoch MJ, eds. Visual legs object processing. London: Erlbaum, 1987. crossed... water overflowing/ got the taps on 20 Benton AL, Hamsher KdeS, Vamey NR, Spreen 0. ... that's wrong." J7udgement of line orientation. New York: Oxford Total T units = 7. Total morphemes in T units = 31. University Press, 1983. 21 Salmon Morphemes/T unit = 4-4. Hodges JR, DP, Butters N. Semantic memory impairment in Alzheimer's disease: failure of access or degraded knowledge? Neuropsychologia 1992;30:30 1-14. PATIENT 3 22 Hodges JR, Patterson K. Is semantic memory consistently (1) 12 months duration of impaired early in the course of Alzheimer's disease? illness Neuroanatomical and diagnostic implications. "Oh, I've seen this before ... um ... the kitchen sink is Neuropsychologia 1995;33:441-59. flooding and the boy is falling off the kitchen stool raid- 23 Goodglass H, Kaplan E. The assessment of aphasia and ing the cookie jar/er, the girl is holding out her hand for related disorders. London: Henry Kimpton Publishers, 1976. the cookies, but er the boy hasn't given any to . her/um 24 Reed V. An introduction to children with language disorders. . . the ... the wi ... the mother is washing up ... wip- 2nd ed. New York: Merrill, 1994. ing up ... um ... but she's got a flood." 25 Blomert L. The Amsterdam-Nijmegen everyday language test In: Total T units = 6. Total morphemes in T units = 61. (ANELT). Steinbuchel N, von Cramon DY, Poppel E, eds. Neuropsychological rehabilitation. Berlin: Morphemes/T unit = 10-2. Springer-Verlag, 1992. 26 Kay J, Lesser R, Coltheart M. Psycholinguistic assessments of (2) 18 months duration of illness language processing in aphasia. Hove:Lawrence Erlbaum "Well the sink is overflowing and the housewife Associates Ltd, 1992. doesn't 27 Bishop DVM. Test for the reception of grammar. London: seem to be taking much notice/ she's washing up, wiping Medical Research Council, 1989. up, and the boy is stealing from the cookie jar but he's 28 HodgesJR, Salmon DP, Butters N. Differential impair- ment of semantic and http://jnnp.bmj.com/ fallen over because he's on a stool/I can't think of any- episodic memory in Alzheimer's and Huntington's disease: a controlled prospective study. thing else." .7 Neurol Neurosurg Psychiatry 1990;53: 1089-95. Total T units= 5. Total morphemes = 50. 29 Rosser A, Hodges JR. Initial letter and semantic category Morphemes/T unit= 10-0. fluency in Alzheimer's disease, Huntington's disease, and progressive supranuclear palsy. .7 Neurol Neurosurg Psychiatry 1994;57: 1389-94. (3) 28 months duration of illness 30 Monsch AU, Bondi MW, Butters N, Salmon DP, Katzman "The boy.... umr.., is pinching the cookie jar... R, ThalLJ. Comparisons of verbal fluency tasks in the cookies from the jar and giving them to detection of dementia of the Alzheimer type. Arch Neurol the girl and he's 1992;49: 1253-8. fallen over .., with the stoolumr .. the mother is ... er 31 Butters N, Granholm E, Salmon DP, Grant I, Wolfe J. on September 24, 2021 by guest. Protected copyright. ... having a flood .., the er ... kitchen sink is over- Episodic and semantic memory: a comparison of amnesic flowing/the curtains aren't pulled very nicely." and demented patients. .7 Clin Exp Neuropsychol 1987;9: = 479-97. Total T units 5. Total morphemes = 39. 32 Martin A, Fedio P. Word production and comprehension Morphemes/T unit= 6-5. in Alzheimer's disease: the breakdown of semantic knowledge. Brain Lang 1983;19:124-41. 1 Richardson JC, Steele J, Olszewski J. Supranuclear oph- 33 van der Hurk PR, Hodges JR. Episodic and semantic mem- thalmoplegia, , nuchal dystonia and ory in Alzheimer's disease (DAT) and progressive dementia: a clinical report on eight cases of "heteroge- supranuclear palsy (PSP): a comparative study. _7 Clin neous system degeneration". Trans Amer Neurol Assoc Exp Neuropsychol 1995;17:459-71. 1 963;88:25-9. 34 de Bruin VMS, LeesAJ.The clinical features of 67 patients 2 Maher ER, Smith EM, LeesAJ. Cognitive deficits in with clinically definite Steele-Richardson-Olszewski syn- Steele-Richardson-Olszewski syndrome (progressive drome. Behav Neural 1992;5:229-32. supranuclear palsy). _7 Neurol Neurosurg Psychiatry 35 Grafman J, Litvan I, Gowez C, Chase TN. Frontal lobe 1 990;48: 1234-9. function in progressive supranuclear palsy. Arch Neural 3 Maher ER, LeesAJ. The clinical features and natural his- 1990;47:553-8. tory of the Steele-Richardson-Olszewski syndrome (pro- 36 Podoll K, Schwarz M, Noth J. Language functions in pro- gressive supranuclear palsy). Neurology 1986;36:1005-8. gressive supranuclear palsy. Brain 1991;114: 1457-72. 4 Kish SJ, Chang U, Mirchandani L, Shannak K, 37 Lantos PL. The neuropathology of progressive supranu- Hornykiewicz 0. Progressive supranuclear palsy: rela- clear palsy. In: Tolosa E, Duvoisin R, Cruz-Sanchez FF, tionship between extrapyramidal disturbances, dementia, eds. Progressive supranuclear palsy: diagnosis, pathology, and brain neurotransmitter markers. Ann Neurol 1985; and therapy. Wien: Springer-Verlag, 1994. 18:530-6. 38 Daniel SE,-de Bruin-VMS, Lees AJ. The clinical and 5 Lebrun Y, Devreux F, Rousseau JA. Language and speech in pathological spectrum of Steele-Richardson-Olszewski a patient with a clinical diagnosis of progressive supranu- syndrome (progressive supranuclear palsy). Brain 1995; clear palsy. Brain Lang 1986;27:247-56. 118:759-70. 6 Leenders KL, Frackowiak RSJ, Lees AJ. Steele- 39 Brooks DJ (1994). PET studies in progressive supranuclear Richardson-Olszewski syndrome. Brain energy metabo- palsy. In: Tolosa E, Duvoisin R, Cruz-Sanchez FF, eds. lism, blood flow, and flurodopa uptake measured by Progressive supranuclear palsy: diagnosis, pathology, and positron emission tomography. Brain 1988;111:615-30. therapy. Wien: Springer-Verlag, 1994.