European Journal of Neurology 1999, 6:353-356 CASE REPORT Linguistic impairment after right cerebellar stroke: a case report Marina Gasparini, Vittorio Di Piero, Olga Ciccarelli, Maria Maddalena Cacioppo, Patrizia Pantano and Gian Luigi Lenzi Department of Neurosciences, ‘La Sapienza’ University, Rome, lialy Correspondence to Dr Marina Gasparini, Department of Neurosciences, viale dell’Universita, 30 00185 Roma, Italy Tel: +-39 06 445 4907; fax: +39 06 445 7376 Received 26 June 1998; accepted 9 November 1998 Increasing evidence suggests the involvement of the cerebellum in non-motor functions. We describe a patient who presented speech deficits after a right cerebellar infarction without other cognitive impairments. 9mT¢-HMPAO single- photon emission computed tomography showed a marked hypoperfusion in the right cerebellar hemisphere, but did not show crossed cerebellar-cerebral diaschisis. The absence of neuroradiological evidence of structural supratentorial lesions supports the cerebellar role in modulating not only the motor system, but also higher order functions. Eur J Neurol 6:353- 356 © 1999 Lippincott Williams & Wilkins Keywords: cerebellar stroke, speech deficits, crossed cerebello-cerebral diaschisis INTRODUCTION Anatomical and behavioral evidence suggests that the cerebellum contributes not only to the motor processes of fluent speech, but also to cognitive processes of word expression (for a review, see Leiner et al., 1991). In normal subjects, neuroimaging studies have shown that the inferior lateral area in the right cerebellum is strongly activated during word-processing tasks, whereas in the motor task of speaking, activation occurs in the superior anterior lobe (Fox et al., 1985; Peterson et ai., 1989). It has been demonstrated that the lateral area can send its output to Broca’s area in the frontal lobe, via the cortico-ponto-cerebellum pathways (Schmahmann, 1991). In patients with agrammatic speech following right cerebellar lesions, recent reports have yielded clinical evidence that the cerebellum plays a modulating role in non-motor linguistic processes (Silveri et al., 1994; Marién et al., 1996) We describe the occurrence of linguistic impairment associated with a right cerebellar infarct. CASE REPORT A 51-year-old man was referred to our clinic 4 months 1351-5101 © 1999 Lippincott Williams & Wilkins after an episode of right limb ataxia and speech disturbances due to a cerebellar infarct in the superior lateral area of the right hemisphere. The patient was a right-handed bank clerk with 13 years of education. His past medical history was unremarkable, apart from hypertension in recent years. He did not suffer from psychiatric disturbances and had a normal psychophysiological development. At entry, the clinical examination showed a mild right cerebellar syndrome and linguistic abnormalities. He underwent magnetic resonance imaging that showed a right cerebellar lesion; no signal abnormalities or focal atrophies were observed at the supratentorial level (Fig. 1). A single-photon emission computed tomography study, performed at rest with °°"Tc-HMPAO, showed markedly reduced perfusion in the right cerebellar hemisphere, but did not show any supratentorial abnormalities of perfusion distribution. Neuropsychological examination Performance in tasks exploring attention, visuospatial skills and frontal lobe functions were normal. Memory functions were slightly reduced in immediate recall for European Journal of Neurology 1999, Vol6.No3 353 é GASPARINI ET AL. FRONT FIGURE 1. Magnetic resonance images: T2-weighted axial (a) and T1-weighted coronal (b) slices performed 4 months after onset of neurological symptoms show an ischemic lesion in the superior area of the right cerebellar hemisphere. verbal material, whereas both delayed recall and recognition were normal. Neurolinguistic findings Language functions were assessed using the NRDA- Language Examination (O.S., 1974). In both writing and reading, sporadic omissions, transpositions and morphologic errors were present. Nevertheless, apart from markedly reduced verbal fluency (15/47), the language examination was quantitatively normal. By contrast, spontaneous speech appeared to be somewhat agrammatic and dysprosodic, although not dysarthric, and was accompanied by manifest word- finding difficulties. A speech sample was collected for the quantitative analysis (Table 1). It consisted of 197 words, with 19.28% of verbal perseverations excluded from further quantitative analysis. The speech rate was reduced (approximately 65.6 words/min); word utter- TABLE 1. Short sample of spontaneous speech (patient's de- scription of shaving) | begin to ... to... to make the lather, to make the lather to, to pass the razor after. The razor has to move downwards, downwards ... without ... ehm ... taking care not to cut oneself and the s ... the second phase that, the one, the one ... that from the neck, the neck you need to be more careful because more delicate. 354 European Journal of Neurology 1999, Vol 6 No 3 ance was uninterrupted, but multiple, long pauses were present (18 pauses in eight sequences). Repetition of initial phonemes sometimes hampered verbal output, although no phonemic or semantic paraphasias were observed. Additional features are shown in Table 2. The mean length of utterance (MLU)-lexical largely prevailed over the MLU-morphologic. The former defined the mean number of words produced in syntactically correct strings, ignoring agreement errors; the latter defined the mean number of words produced in syntactically and morphologically correct strings. The low ratio of 0.57 (normal ratio about 1.15) indicated the patient’s difficulty in producing morpho- logically appropriate speech. A task exploring sentence comprehension was also administered. It consisted of 17 syntactically complex TABLE 2. Quantitative analysis of speech production Speech sample 197 words Verbal perseverations 38/197 (19.28%) Speech rate 65.66 words/min Content words 64/159 (40%) Free-standing grammatic morphemes 96/159 (60%) Mean length utterance (MLU) MLU-lexical 22.0 MLU-morphologic 12.71 Ratio (morphologic/lexical) 0.57 LINGUISTIC IMPAIRMENT AFTER STROKE sentences, characterized by (a) role reversal, (b) spatial and temporal relation, (c) comparison, (d) symmetry, and (e) double negation. The patient was asked to judge the sentence content, first under auditory and then under visual presentation. He made nine mistakes (53%) in each task, and was able to perform the task only after the sentences had been fragmented into simple phrases. DISCUSSION Previous reports of linguistic impairment following tight cerebellar lesions have stressed the agrammatic nature of speech. Silveri et al. (1994) described a patient with agrammatism limited to spontaneous speech without involvement of sentence comprehen- sion, but with a marked involvement of grammatic morpheme, in terms of both substitution and omission rate. A similar linguistic picture was present in the patient described by Marién er al. (1996), who also presented word-finding difficulties and fragmentation of speech sequences. Interestingly, both cases were characterized by a marked cerebello-cerebral diaschisis with decreased perfusion in the left cerebral hemisphere, thus suggest- ing a ‘frontal’ involvement in the speech deficit. One appealing hypothesis explaining agrammatic speech produced by a lesion outside cerebral structures (Nespoulous ef al., 1988) suggested that agrammatism does not depend on disorders affecting syntactic know- ledge, but rather on deficits lying outside the linguistic system, such as a delay in the processes underlying sentence construction. If the speed of the ‘online application’ of syntactic rules is reduced, the represen- tation of grammatic morphemes in the working mem- ory will decay and sentences will become disjointed. It is likely that this mechanism also underlies sentence comprehension, which requires the manipulation of simultaneous complex morphosyntactic information. It has also been argued that the verbatim representation of a sentence is held in the memory as the basis for syntactic re-interpretation (Gathercole and Baddeley, 1993). All these operations would require a temporal modulation provided by a functional system also including the controlateral cerebellum (Silveri et al., 1994). In particular, this area seems to be involved in the correct timing of morphosyntactic processes for sentence representation, as well as in the temporal computation of non-motor cognitive tasks (Keele and Ivry, 1991). Functional neuroimaging studies have shown that a lesion of right cerebellar structures might induce a deactivation of left cerebral structures, owing to the loss of inputs through cerebello-cerebral pathways. This disconnection may be the cause of the linguistic deficits described by Silveri and Marién. By contrast, in our case, agrammatism involved spontaneous speech as well as, although to a lesser extent, reading and writing, and was also accompanied by difficulties in both complex sentence comprehension and word finding. Moreover, free-standing grammatic morphemes prevailed over content words, which is at variance with what Silveri ef al. observed in their patient. Surprisingly, we did not observe any hypoperfusion (i.e. diaschisis) in the left cerebral hemisphere. This phenomenon is a frequent finding at rest during the early phase of stroke, whereas during later stages, it is reported to be detectable by activation paradigm (Di Piero et al., 1990). In our patient, the lack of functional neuroimaging findings makes it difficult to correlate the linguistic impairment with a cerebello-cerebral disconnection. The absence of structural supratentorial lesions sug- gests, rather, that the right cerebellar infarction—and, as a consequence, the slower timing in sentence representation—may be responsible for the language symptoms. Our findings, therefore, seem to support the cerebellar role in modulating cerebral functions outside the motor system. REFERENCES Di Piero V, Chollet F, Dolan R, Thomas D, Frackowiak R (1990). The functional nature of cerebellar diaschisis. Stroke 21:1365—-1369. Fox P, Raichle M, Thach W (1985). Functional mapping of the cerebellum with positron emission tomography. Proc Natl Acad Sci USA 82:7462~7466. Gathercole S, Baddeley A (1993). Working memory and language. Hove: Lawrence Erlbaum Associates. Keele S, Ivry R (1991). Does the cerebellum provide a common computation for diverse tasks? Ann NY Acad Sci 608:179-211. Leiner H, Leiner A, Dow R (1991). The human cerebro- cerebellar system: its computing, cognitive, and language skills. Behav Brain Res 44:113-128. Marién P, Saerens J, Nanhoe R, Moens E, Nagels G, Pickut B, et al. (1996). Cerebellar induced aphasia: case report of cerebellar induced prefrontal aphasic language phenomena supported by SPECT findings. J Neurol Sci 144:34—43. Nespoulous J, Dordain M, Perron C, er al. (1988). Agramma- tism in sentence production without comprehension defi- cits: reduced availability of syntactic structures and/or of grammatical morphemes? A case study. Brain Lang 33:273-295. OS. (1974). Language examination. Firenze: Special Organi- zation Neuropsychological Research Development Associa- tion. Peterson S, Fox P, Posner M, Mintun M, Raichle M (1989). Positron emission tomographic studies of the processing of European Journal of Neurology 1999, Vol6No3 355 356 single words. J Cognitive Neurosci 1:153—170. Schmahmann J (1991). An emerging concept. The cerebellar contribution to higher function. Arch Neurol 48:1178- 1187, European Journal of Neurology 1999, Vol 6 No 3 GASPARINI ET AL. Silveri MC, Leggio MG, Molinari M (1994). The cerebellum contributes to linguistic production: a case of agrammatic speech following a right cerebellar lesion. Neurology 44:2047—2050.