,ingio V Pergamon N" I ,, 97 1 .3 1994 Elmrier Science Ltd Primes in Grenl Ndlaln W_8 79]? 94 c600 . 0 00 RIGHT TEMPORAL COMPENSATORY MECHANISMS IN A DEEP DYSPHASIC PATIENT : A CASE REPORT WITH ACTIVATION STUDY BY SPECT DOMINTQLE CARDEBAT, * J EAN-FR .A N(DIS DFMONFT, PTERRF CET SIS . MICHFI F fUFL, GFR .ARD VIALLARD and JEAN-PIERRE MARC-VLRGNES INSERM U 230 and Department of Neurology, CHI ; Purpan, Toulouse . France (Rereired 27 April 199.3 : accepted 17 June 1993) brain activation study using SPFCI' and "' Xe in a deep dysphasic patient with left tempera] lesion is presented . The activation paradigm consisted of a passive listening to foreign language as baseline . a phoneme monitoring condition and a semantic word monitoring condition . 'flue specific activation of the right middle temporal cortex observed in the semantic condition is congruent with the hypothesis of a compensatory role of the right hemisphere in processing concrete words . This case illustrates the interest of functional imaging for a better understanding of neural mechanisms of functional recovery after brain injury . Abstract-A DEEP DYSPHASIA is a rare syndrome [5 .7] occurring after a lesion in the left temporal parietal region of the brain . It is characterized by a nearly complete incapacity to repeat pronounceable non-words and abstract words and by a large proportion of verbal substitutions in the repetition of concrete words . Furthermore, deep dysphasic patients seem to partially understand the meaning of concrete words without handling their phonological form . The capacity of the right hemisphere for language, more especially for understanding concrete words, has been invoked to account for the observed dissociation between the phonological and semantic components of word processing [5] . This paper describes an activation study with SPECT and "' Xe in a deep dysphasic patient . The aim of the study is to demonstrate, with an imaging technique, the role of the right hemisphere in the comprehension of concrete words . CASE REPORT A 22-year-old right-handed male patient suffered from a left-sided ischernic cerebral vascular accident which was attributed to a probable complicated migraine . The initial language assessment demonstrated a fluent aphasia with severe auditory comprehension deficit suggesting a pure word deafness . M R 1 performed 2 months post-onset demonstrated a lesion involving the posterior half of the superior temporal gyros, the supra-marginal gyros *Address for correspondence and reprints : Dominique Cardebat. INSERM 1,230, Service de Neurologie . CHU Purpan .31059 Toulouse Cedes, France . 97 98 D . CARImBAT et ulL and the angular gyrus (see Fig . I I . An extensive language assessment was performed 3 months post-onset . Patient's performance fitted with the diagnosis of deep dysphasia (for details, sec Rcf.[]] ) . For instance, this patient repeated, without any delay, "cherry" instead of "raspberry'' or "ballet" instead of "theatre" . Auditory comprehension improved since normal performance were observed, for concrete words only, on semantic judgement tasks . However, performance was dramatically poor on a rhyming task . Thus . meaning of words was adequately processed whereas phonological analysis seemed impossible . METHODS Regional Cerebral Blood Flow Ir( BF) was assessed quantitatively f21 using SPELT ITOMOMATIC' 64. MFDIMA'I ](',('openhagenI,and intravenous injection of I" Xe I22'_o MBq) .at 6 months post -onset . Data were collected simultaneously 14 min) froln 3 transverse slices . 2-em-thick, centred at 1, 5 and 9 cm above the orbitotocalal line and during 3 cxpcrinsntal conditions performed on the sane day wilt a 40-min interval between the measurement- The sohlect's head was carefully positioned according to graduate external markers, and marks were also drawn on the akin in order to readjust the head's position front one experimental condition to the other. Changes in rC'RF' were evaluated in the two cerehellar hemispheres (OM + I em), in 14 Regions of Interest in the mid slice (Figs 2) and in the superior part of each hemisphere in the OM +9 em slice- Regions of interest (ROIs) were drawn and processed using art interactive software on a Macintosh microcomputer . 'I he first condition (hasline) consisted in listening to a 4-nun excerpt of scientific communication given in Italian la language totally unknown to the patient)- to the second condition (semantic condition I, the patient had to press down a button each time he was hearing the noun of bird in a 4-nin French connected speech . In the third condition i t honological condition), thepaucnt, again listeningto4-min French connected speech, had to pressdown a button each time he was hearing the phonettle ;e when pronounced at the end of a word . RESULTS Under the semantic condition, the patient detected 77% of the targets whereas lie completely failed in detecting the targets under the phonological condition . Table I gives the flow values (ml)100 g/min) in the Regions of Interest of the mid slice (OM + 5 cm) and under the three conditions . Low flow values were observed on the one hand in the left Posterior Temporal region which corresponded to the core of the lesion and on the other, to a lesser extent- in the left Middle Temporal (OM+5 cm) and the left Superior Frontal Parietal (OM 19 cm) regions in which remote hypopcrfusion effects probably took place . Compared to baseline. both the semantic and phonological tasks induced a global increase in CBF . Lateral asymmetry indexes (right-left difference in rCBF for homologous ROIs in each cerebral hemisphere) were used to normalized values to assess activation effects . The only outstanding change, readily observable on the Ilow maps (Fig . 3) . is a focal increase in the right middle temporal region under the semantic condition . This selective, localized activation resulted in a large increase ( + I2m1/100g;min) of the asymmetry index (right-left difference) in the middle temporal region in the semantic -baseline comparison (fable 2) . By contrast, for the phonological-baseline comparison, only a mild increase was found in this region which was comparable to that observed in all ROIs for each comparison . DISCUSSION In our patient . the selective activation of the right middle temporal region was associated with a good performance during the semantic condition as opposed to performance at random level and absence of marked selective r('BF increase during the phonological condition . SPEC! AND BRAIN ACTIVATION IN STROKE Fig . 1 . T2 weighted axial MR image obtained at 46 mm above the orbitomeatal line (that is 12 mm above and parallel to the bicommissural plane according to the stereotactic space from TALAIRACH and TouRNOUx [9] . This image demonstrates a hyperintensity corresponding to the sequclac of an infarction in the posterior parts of the superior and middle temporal gyri (Brodmann's areas 22 and 39) in the left hemisphere . 99 100 D . CARUEBAT et ,I . Fig . 3 . Cerebral blood flow maps at 50 mm above and parallel to orbitomcatal line . (A) Baseline ; (B I semantic condition ; and (C) phonological condition . Note the focal hyperflow (arrow) in the right middle temporal region under the semantic condition . 101 SPLCI AND BRAIN ACTIVATION IN SIRIIKL Fig .2 . Regions olliiteresl drawn on the (HE hap i1 50 miii ahose and parallel In orbitnme ;tIil line . I and 2 : left and right prefrontal region,,,3 and 4 :left and right inferior frontal regions : Sand 6 : cIt and right middle Iemporal regions : 7 and 8 : left and Iight posterior temporal regions : 9 and 10 : left and right rnedial occipital regions : I I and 12 : left and right striatum regions : 13 and 14 : left and right thalamus regions . I ]its result supports the assumption that, in our patient the right middle temporal region is involved in the comprehension of concrete words . Clinical studies on aphasic patients [6] as well as PET activation studies in normals [3] have demonstrated that auditory comprehension is mainly processed in the left temporal region (including Wernicke's area) . However . data from patients with isolated right hemisphere after commissurotomy [10] or data from aphasic patients with good recovery of comprehension [8] have shown that the right hemisphere can achieve the semantic processing of frequent concrete words . A compensatory mechanism can be proposed to account for the right middle temporal activation observed in our patient when he was performing the semantic task . Such an activation may reflect the latent semantic capacities of the right hemisphere when freed from the inhibitory effects of the left hemisphere, i e . the dominant hemisphere for language in right-handers . This hypothesis has been proposed in commissurolomized patients in whom lexical semantic processing was demonstrated in the right hemisphere [9] . 102 D . CARDI BAT et al . Table L Flow v,due, in the 18 Regions of Interest during the three experimental conditions Baseline Cerebellum I'refronlal Inform frontal Middle temporal Posterlortcmporil Medial occipital Striatum thalamus Superior frontal parietal Len Right Ldl Right 1<0 Right Lefl Righl Lca Right I ett Right Left Right Left Right Left Right 41 43 43 49 41 46 33 45 23 41 40 46 39 38 36 43 34 35 Semantic condition Phonological condition 5, 54 56 66 711 57 66 45 61 _8 51 5] 52 5(, 59 57 63 49 5'- 57 6, Sri 51 53 55 S7 60 63 47 49 Re,ultsarcexpressed in in! 100g:min . Table 2 . Conrpal'isous of lateral asymmetry indexes Iright left) on the three experimental conditions Lateral asymmetry indexes Baacline condition Cerebellum Prefonlal Inferior frontal Middle temporal Posterior temporal Medial oeeipilal Striatum Thalamus Superior frontal parietal 12 18 Senianlie condition Across-condition diBcrnee in lateral asymmetry indexes Phonological condition ' 4 9 16 '7 0 Semantic Baseline difference 1 - 5 4 7 1 Phonological Baseline difference 6 -1 I Results arc expressed in nil,'1OOgInin . Such a mechanism might also occur, after a first stage of transhenlispheric diaschisis, in patients sufermg a vascular lesion in the left cortical areas devoted to auditory comprehension . Moreover, latent language abilities have been also recently suggested by language activation studies with PL T in normal subjects [3 4] . To our knowledge, single case studies of aphasic patients with cerebral blood flow activation using SPFCT or PET have never been reported . Yet, these functional imaging techniques offer the unique opportunity w lend the pathophysiological hypotheses on the compensatory role to the non-dominant hemisphere in aphasic patients . SPF( T n9) BRAIN AC nbnnON IN STROKE °. 103 REFFRENC'FS L C,ARDLBAT . D_ PULL . M .- DiNioNCT . J . F . and NLsrorlots .) . L . Les dilidrentes "baize' de Ill repetition : Analyse (Tun cas do (IN sphasic profonde . Ret . tiPwopsrrhnl . 1, 215 232 . 1991 . _ . CLLSIS . P_ GOU>mA',. T ., I ILNRIKSLN . L, and Lsssi : . N . A . A method for calculating regional cerebral blood floss from emission cumpuled to Inogaphy of inert gas concentrations . J . Cnmpia( cosier . Trrrnri 1 , .5 . 641 645 . 1981 3 . DrsION .T, J . P . . CiOLLET . I . . Rvossv_ S . . CARDIBA . D . . NtsitUnt s . 1 . L . . NA isi . R . . Rnstit, A . and FR 4CRow1AA, R . S . 1 . The anatomy of phonological and semantic proeessme in normal euhiects . 1ham 115, 1753 176K, 199 1 , 4. Di%IONJT . 3 . F . . Pla(I :, C., Wist . RR and ER,u'anxans, R . S . J, Involsnncut of the right angular gyms in semantic processes and of the supra-marginal gnus in phonological pro es,ses : An ucus allon study iismc PF I in normal subjects . J. Cereh . Blnud Floir Veto/i . 13 SI, 547 . 1993 . ` . Dt linnu .LJ . R . and PON(LT, M . Deep dysphasia In a case of phonemic deafness : Role of the right hemisphere in audiirn} language comprehension, Vrmvpsrcholoyin 24, 769 71,9 1986 . 6 . KI If Tisz . A . Localization of lesions in NN ernickes aphasia . In Loci/_tiion iri NruropsIch rlagia, A Kran .sz IEditor) . pp. 'l1' 2 29 . Academic Press . Ness York, 1983 . 7 . MI( fit I . f-and NSRUS'rSKY, E . Deep dysphasia : An analog it deep dyslcxia in the auditors modality /tram Lamp,IS.'1' 2'1 .1983 . 8 . PArsv