A multimodal mapping study of conduction aphasia with impaired repetition and spared reading aloud

Neuropsychologia 70 (2015) 214–226

Contents lists available at ScienceDirect

Neuropsychologia journal homepage: www.elsevier.com/locate/neuropsychologia

A multimodal mapping study of conduction aphasia with impaired repetition and spared reading aloud Barbara Tomasino a,n, Dario Marin a,1, Marta Maieron b, Serena D'Agostini c, Irene Medeossi d, Franco Fabbro a,e, Miran Skrap c, Claudio Luzzatti f a

IRCCS “E. Medea”, San Vito al Tagliamento (PN), Italy Fisica Medica, A.O. Santa Maria della Misericordia, Udine, Italy c Unità Operativa di Neuroradiologia, A.O. Santa Maria della Misericordia, Udine, Italy d IMFR Gervasutta, Udine, Italy e Dipartimento di Scienze Umane, Università di Udine, Italy f Dipartimento di Psicologia, Università di Milano-Bicocca, Milano, Italy b

art ic l e i nf o

a b s t r a c t

Article history: Received 29 May 2014 Received in revised form 11 February 2015 Accepted 17 February 2015 Available online 18 February 2015

The present study explores the functional neuroanatomy of the phonological production system in an Italian aphasic patient (SP) who developed conduction aphasia of the reproduction type following brain surgery. SP presented with two peculiar features: (1) his lesion was localized in the superior temporal gyrus, just posterior to the primary auditory cortex and anterior/inferior to and neighboring the Sylvian parietal temporal (Spt) area, and (2) he presented with severely impaired repetition and spelling from dictation of words and pseudowords but spared reading-aloud of words and pseudowords. Structural, functional, fiber tracking and intraoperative findings were combined to analyze SP's pattern of performance within a widely used sensorimotor control scheme of speech production. We found a dissociation between an interrupted sector of the arcuate fasciculus terminating in STG, known to be involved in phonological processing, and a part of the arcuate fasciculus terminating in MTG, which is held to be involved in lexical-semantic processing. We argue that this phonological deficit should be interpreted as a disorder of the feedback system, in particular of the auditory and somatosensory target maps, which are assumed to be located along the Spt area. In patient SP, the spared part of the left arcuate fasciculus originating in MTG may support an unimpaired reading performance, while the damaged part of the left arcuate fasciculus originating in STG may be responsible for his impaired repetition and spelling from dictation. & 2015 Elsevier Ltd. All rights reserved.

Keywords: Superior temporal gyrus (STG) Magnetic resonance imaging (MRI) Conduction aphasia Phonological processing Reading and writing Sylvian parietal temporal (Spt) area

1. Introduction Conduction aphasia (CA) is a language deficit characterized by disordered repetition and good comprehension (e.g., Benson and Ardila, 1996). According to Wernicke's (1874) associationist model of language processing, CA is caused by a disconnection between the area processing recognition of sound images of words (the posterior superior part of the temporal lobe, also known as Wernicke's area) and the area where the motor images of words are represented (the posterior part of the left inferior frontal gyrus, also known as Broca's area), owing to a lesion of the left arcuate fasciculus (Wernicke, 1874; see De Bleser et al., 1993 for a review). The deficit is due to an impaired encoding of phonological word n Correspondence to: IRCCS “E. Medea”, Polo Regionale del Friuli Venezia Giulia, Via della Bontà, 7, 33078 San Vito al Tagliamento (PN), Italy. Fax: þ39 0434 842797. E-mail address: [email protected] (B. Tomasino). 1 Co-first author

http://dx.doi.org/10.1016/j.neuropsychologia.2015.02.023 0028-3932/& 2015 Elsevier Ltd. All rights reserved.

forms and their mapping to the associated sequential articulatory gestures. CA is characterized by phonemic paraphasias, phonemic neologisms, continuous repairs, and conduite d'approche behavior (repetitive self-correction attempts) (Anderson et al.,. 1999; Buchsbaum et al., 2011; Fridriksson et al., 2010; Goodglass, 1992; Hickok, 2000; Hillis, 2007). Besides the classical Wernicke's description of CA, a further condition of disproportionate repetition deficit has been reported by Shallice and Warrington (1977) who distinguished between Reproduction and Repetition subtypes. In the repetition subtype of CA (see also Caplan et al., 1986), patients do not present with phonemic paraphasias, phonemic neologisms, continuous repairs, and conduites d'approche but with a selective damage of phonological short-term memory. The functional damage causing reproduction AC, rather than as a mere disconnection between the phonological input and output lexicons, is usually localized at the level of the phonological output buffer (Baldo et al., 2008; Caramazza et al., 1986; Romani, 1992;

B. Tomasino et al. / Neuropsychologia 70 (2015) 214–226

Shallice et al., 2000) or at the level of the interface between the buffer and the phonological output lexicon. Activation in areas involving the left frontal inferior gyrus and including parts of the middle frontal and inferior precentral gyri (Chein and Fiez, 2001) might correspond to functional processing in the phonological output buffer (Jacquemot and Scott, 2013). In particular, Baddley (2003) localized the phonological output buffer in Broca's area/the premotor cortex (for a review see Jacquemot and Scott, 2013). However, this localization is not consistent with a frequently reported lesion in CA patients, namely damage to the left temporal/ inferior parietal area, although a lesion of the posterior temporal/ parietal area might impact on Broca's area/the premotor cortex since the arcuate fasciculus connects these areas. Furthermore, a functional damage to the phonological output buffer hypothesized by the cognitive information processing models (e.g. Patterson, 1996) implies a disproportionate phonological deficit not only in repetition but also during reading-aloud of words and pseudowords (a prediction that is actually implicit also for CA following Wernicke–Lichtheim's account, since reading-aloud, too, would require a functional association pathway running in the external capsule, such as the arcuate fasciculus). According to one of the recent dual-stream models of speech processing (Hickok and Poeppel 2007), CA is a deficit of sensorymotor integration for speech. In other words, a disproportionate phonological deficit in repetition can be explained as a damage to a system serving as interface between the auditory target and the corresponding motor speech output (Hickok, 2000; Hickok et al., 2000, 2003; Hickok and Poeppel, 2004, 2007). Thus, the deficits in repetition and speech production characterizing CA of the reproduction subtype reflect an impairment in the capacity of auditory representations of speech to constrain and guide the corresponding articulatory representations (Hickok, 2000; Hickok and Poeppel, 2004; Wise et al., 2001). Hickok and Poeppel argue that inputs to the auditory-phonological network define the auditory targets of speech acts. The predicted auditory consequences of a motor speech unit can be checked against the auditory target. If the anticipated speech sounds match the auditory targets, such representation remains active and a normal articulation process follows; if they mismatch, a correction signal is generated to activate the correct motor unit. A lesion to the Sylvian parietaltemporal (Spt) area would disrupt the ability to generate forward predictions in the auditory cortex and thus the ability to perform internal feedback monitoring. In the present study, we analyzed the phonological processing of SP, a conduction aphasic patient suffering a profound repetition deficit, fluent but paraphasic speech output, and preserved auditory comprehension. SP presented with two peculiar features: (i) impaired word and pseudoword repetition and spelling from dictation but spared word and pseudoword reading abilities, and (ii) his lesion was localized in the superior temporal gyrus, just posterior to the primary auditory cortex and anterior/inferior to the Spt area. If SP's deficits were of the “repetition subtype” (Shallice and Warrington, 1977), his pattern of performance (repetition deficit, spelling deficit, and spared reading aloud) could have simply been explained as due to a phonological short-term memory deficit. However, this view fails to account for SP's phonemic paraphasias, phonemic neologisms, continuous repairs and conduite d'approche behavior. In addition, the dissociation between spared reading and impaired repetition and writing to dictation is unusual (Ardila, 2010; Benson et al., 1973) showing that phonemic errors in readingaloud (Goodglass, 1992) as well as writing (from mild spelling difficulties to profound agraphia) (Bernal and Ardila, 2009) parallel the phonological production errors in CA (see also (Caramazza et al., 1981) for a similar observation). In fact, very few studies on CA investigated written expression in detail (Balasubramanian,

215

2005). None of the literature cases in which the patients' reading and spelling performance was reported showed a pattern of damage with impaired repetition and spelling from dictation but no reading-aloud deficit, as found for patient SP (see Fridriksson et al., 2010 for a review). In the present multi-modal mapping study we investigated the functional neuroanatomy of SP's performance. We combined structural and functional brain imaging techniques, intraoperative stimulation mapping as well as neuropsychological testing to better understand the neurological and functional substrates involved in CA with impaired phonological output in repetition, but spared phonological output in reading-aloud. The dissociation shown by SP cannot be explained by a standard information processing model of word and pseudoword processing (e.g. Patterson, 1986), because models postulate that the phonological deficit characterizing reproduction AC does necessarily involve both repetition and reading-aloud performance. As already reported, the same prediction is implicitly suggested by Wernicke and Lichtheim's account of CA, since reading-aloud requires the contribution of the arcuate fasciculus via the external capsule. We focused on the examination of the dorsal stream and, in particular, made reference to an anatomical dissociation reported by Glasser and Rillings (2008) between the part of the superior longitudinal fasciculus (SLF) terminating in the middle temporal gyrus, (MTG, which is the anatomical counterpart of the dorsal stream), and the part terminating in the superior temporal gyrus (STG) (see Catani et al., 2005 for a different reconstruction account of the SLF).

2. Materials and methods 2.1. Participants SP is a 42-year-old, right-handed (Oldfield, 1971), monolingual native speaker of Italian. He has an educational level of 13 years and works as a fireman. He was admitted to the Udine General Hospital one week before the study start for a seizure with loss of consciousness. A neurological examination did not reveal any focal deficits and any overt signs of language impairment. The patient had no family history of developmental language problems or learning disabilities nor had suffered neurological problems in the past. Conventional T2-weighted MR imaging revealed a low-grade glioma (approximately 6.4 cc in volume). According to the Anatomy toolbox (Eickhoff et al., 2005), the region of interest (ROI) drawn on the patient's lesion and normalized to the MNI template was localized in the left superior and middle parts of the temporal lobe and lays posterior to the primary auditory cortex (as evidenced in Fig. 1A and C). The patient's fMRI and DTI maps were compared with those obtained from a group of 18 monolingual Italian native speakers acting as controls (12F, 6M; mean age 47.7 77.6, age range 35–61; handedness: mean laterality index 94.0 79.8, range 66.7–100; education: mean 13 years 70). All participants had normal or corrected-to-normal vision and reported no history of neurological illness (except for the patient), psychiatric disease, or drug abuse. The participants' informed consent to participate in the study was obtained in line with the Declaration of Helsinki. The study was approved by the local Ethics Committee. The glioma was removed surgically about 20 days after the clinical diagnosis. Fig. 1B and D portrays the removal of the tumor from the left hemisphere.

216

B. Tomasino et al. / Neuropsychologia 70 (2015) 214–226

Fig. 1. Pre-surgical (A) and post-surgical (B) T2-weighted MR images of the patient's lesion (thickness of the axial slices: 5 mm). (C) The lesion is localized posterior to the primary auditory cortex pre- (on the left) and post-surgery (on the right). We considered only those voxels of the functional ROIs that were located within the cytoarchitectonically defined MPMs of the primary auditory cortex provided by the SPM Anatomy toolbox (Eickhoff et al., 2005) on the word listening fMRI-related maps. (D) Intraoperative picture showing the results of the cortical stimulation mapping (on the left) and the results of surgical resection (on the right). Rendered brain templates of the patient pre- (on the left) and post-surgery (on the right) are shown above the intraoperative pictures to indicate the head position during mapping. (E) Pre-operative (red) and post-operative (yellow) and 1-year follow-up (blue) reconstructions of the sectors of the arcuate fasciculus terminating in the MTG and number of fibres and mean FA for controls and patient SP. (F) Pre-operative (red) and post-operative (yellow) and 1-year follow-up (blue) reconstructions of the sectors of the arcuate fasciculus terminating in the STG and number of fibres and mean FA for controls and patient SP, indicating that the lesion affected this tract. (G) A possible explanation for SP's dissociation is that damage to the STG sector with preserved MTG sector of the arcuate fasciculus may cause a pattern of performance with impaired repetition and spelling from dictation but preserved reading-aloud. The model proposed by Glasser and Rilling (2008) is superimposed on the pre- and post-operative and the 1-year follow-up reconstructions of the sectors of the arcuate fasciculus terminating in the MTG and in the STG shown by patient SP. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

B. Tomasino et al. / Neuropsychologia 70 (2015) 214–226

2.2. Neuropsychological testing 2.2.1. Pre- and post-surgical and follow-up evaluations The patient's cognitive functions were tested by means of a neuropsychological test battery assessing general cognitive functioning [Raven's Colored Progressive Matrices (Basso et al., 1987) and the Mini Mental State Examination (MMSE) (Magni et al., 1996)]. Other tests included digit span [forward and backward (Orsini et al., 1987)], praxis [ideomotor (De Renzi et al., 1980) and oral apraxia testing (Spinnler and Tognoni, 1987)], in addition to a battery for the analysis of language disorders (Miceli et al., 1994) administered before surgery, one week post-surgery and one year post-surgery. 2.3. fMRI and DTI study 2.3.1. Paradigms and tasks Language tasks included word listening and reading aloud of sentences and pseudowords. In the baseline conditions (N ¼6), a fixation cross (15 s) was presented between blocks. In all tasks, each block (N ¼4, 15 s each) started with an instruction (3 s) and included seven trials (2430 ms each). Stimuli were selected from the Battery for the Analysis of Language Disorders (Miceli et al., 1994). In the sentence reading-aloud task, the patient was asked to “read aloud the following sentences as accurately and as quickly as possible”. In the pseudoword reading-aloud task, the patient was asked to “read aloud the following pseudowords as accurately and as quickly as possible”. The Presentations software (Version 9.9, Neurobehavioral Systems Inc., CA, USA) was used for stimulus presentation and synchronization with the MR scanner. Participants viewed the stimuli via VisuaStim Goggles (Resonance Technology). Prior to acquisitions, they practiced the tasks outside the scanner. 2.3.2. fMRI and DTI data acquisition Magnetic resonance imaging was performed 6–10 days prior to craniotomy. A Philips Achieva 3-T (Best, Netherlands) whole-body scanner was used to acquire DTI, anatomical and functional images, using a SENSE-Head-8 channel head coil and a custom-built head restrainer to minimize head movements. Diffusion tensor data were acquired using an axial diffusionweighted, single-shot, spin-echoplanar imaging sequence covering the whole brain (repetition time ¼8800 ms; echo time ¼74 ms, bandwidth ¼1287 Hz/pixel; flip angle ¼90°; field of view ¼224  224 cm2; 70 contiguous axial slices, 1.6 mm slice thickness; matrix size ¼224  224 voxels). Two b values were used; seven images at 0 s/mm2 (no diffusion weighting) and 64 non-coplanar images at 1000 s/mm 2 (diffusion-weighting b value) were acquired. The gradient directions were uniformly distributed on a sphere. Functional images were obtained using a single-shot gradient echo, echoplanar imaging (EPI) sequence. EPI volumes (N ¼ 54 for each fMRI task) contained 34 axial slices (TR ¼2500 ms, TE ¼ 35 ms, FOV ¼230 mm, matrix: 128  128; slice thickness of 3 mm with no gaps, 90° flip angle, voxel size: 1.79  1.79  3.3 mm3 ) and were preceded by four dummy images allowing the MR scanner to reach a steady state. High-resolution T2-weighted and post-gadolinium contrast T1weighted anatomical MR images were acquired for the intra-operative stereotactic surgical navigation system. To this purpose, a T1-weighted 3D magnetization-prepared, rapid acquisition gradient-echo fast field echo (T1W_3D_TFE SENSE) pulse sequence (TR ¼8.1007 ms, TE ¼3.707 ms, FOV ¼240.000 mm, 190 sagittal slices of 1 mm thickness, flip angle ¼8°, voxel size: 1  1  1) and a T3-weighted 3D magnetization-prepared, rapid acquisition gradient-echo fast field echo (T2W_3D_TFE SENSE) pulse sequence (TR ¼2500 ms, TE ¼368.328 ms, FOV ¼240.000 mm, 190 sagittal

217

slices of 1 mm thickness, flip angle ¼90°, voxel size: 1  1  1) were acquired. 2.3.3. fMRI data processing All calculations were performed on UNIX workstations (Ubuntu 8.04 LTS, i386, http://www.ubuntu.com/) using MATLAB r2007b (The Mathworks Inc., Natick, MA/USA) and SPM5 (Statistical Parametric Mapping software, SPM; Wellcome Department of Imaging Neuroscience, London, UK http://www.fil.ion.ucl.ac.uk/ spm). Dummy images were discarded prior to further image processing. Pre-processing included spatial realignment of the images to the reference volume of the time series, segmentation producing the parameter file used for normalization of EPI data to a standard EPI template of the Montreal Neurological Institute provided by SPM5, re-sampling to a voxel size of 2  2  2 mm3, and spatial smoothing with a 6-mm FWHM Gaussian kernel to meet the statistical requirements of the General Linear Model and to compensate for residual macro-anatomical variations across subjects. To delineate the language networks related to the word listening, sentence reading and pseudoword reading tasks, we modeled the alternating epochs by using a simple boxcar reference vector. A general linear model for block designs was applied to each voxel of the functional data by modeling the activation and the baseline conditions for each subject and their temporal derivatives by means of reference waveforms, which correspond to boxcar functions convolved with a hemodynamic response function (Friston et al., 1995a, 1995b). Furthermore, we included six additional regressors that modeled the head movement parameters obtained from the realignment procedure. Accordingly, a design matrix, which comprised contrast modeling alternating intervals of “activation” and “baseline”, was defined. At the single subject level, specific effects were assessed by applying appropriate linear contrasts to the parameter estimates of the baseline and experimental conditions resulting in t-statistics for each voxel. Low-frequency signal drifts were filtered using a cut-off period of 128 s. These t-statistics were then transformed into Z-statistics constituting statistical parametric maps (SPM{Z}) of differences between the experimental conditions and between the experimental conditions and the baseline. SPM{Z} statistics were interpreted in light of the theory of probabilistic behavior of Gaussian random fields (Friston, et al., 1995a, 1995b). For each task we calculated the following contrast images: the main effects of CONDITION (task–baseline_pre-surgery4 task–baseline_post-surgery, and task–baseline_post-surgery 4task–baseline_pre-surgery), using a threshold of p o.05, corrected for multiple comparisons at the cluster level (using FWE), with a height threshold at the voxel level of p o.001, uncorrected. The anatomical interpretation of the functional imaging results was performed using the SPM Anatomy toolbox (Eickhoff et al., 2005). 2.3.4. DTI data processing Three-dimensional tract reconstruction was performed using DTIStudio (version 3.0.3) software (Jiang et al., 2006). Deterministic tractography was performed in the left hemisphere of the patient and control subjects using the Fiber Assignment by Continuous Tracking (FACT) algorithm (Mori et al., 1999; Mori and van Zijl, 2002; Wakana et al., 2004) to track two segments of the arcuate fasciculus, one terminating in the posterior superior temporal gyrus (STG) and the other terminating in the middle temporal gyrus (MTG) (Glasser and Rilling, 2008). A fractional anisotropy (FA) threshold of 1.5 and a turning angle 445° were used as criteria to restrict the algorithm to yield biologically plausible results (Wakana et al., 2007). The tracts were reconstructed using a multi-ROI approach: the first ROI was placed on a coronal view at the middle of the

218

B. Tomasino et al. / Neuropsychologia 70 (2015) 214–226

posterior limb of the internal capsule on the intense triangleshape green structure which identifies the arcuate fasciculus and neighboring fibers (Makris et al., 2005). To divide the arcuate fasciculus into STG and MTG segments, a second ROI was placed on the tract terminations in the posterior STG to select the arcuate part of the tract, and a third ROI was placed around the terminations on the superior temporal sulcus in the MTG to resolve the phonological part of the tract. Potential contaminating fibers were removed. FA and number of fibers were evaluated. Furthermore, we performed a further fibre tracking study of the SLF using Catani et al. (2005) reconstruction method. 2.4. Intra-operative methods During surgery a navigation system was used (StealthStation TReon plus; Medtronic Sofamor-Danek) showing the T1-weighted and T2-weighted high-resolution 3D images loaded with fMRI and DTI. Awake surgery was performed. Cortical and subcortical functional mapping was performed by direct electrical stimulation following conventional methods (Berger et al., 1990; Berger and Ojemann, 1992; Szelenyi et al., 2010). Cortical stimulation mapping included counting, object and action naming, word, pseudoword and sentence repetition, as well as sensorimotor mapping. For the naming task, we used the same 28 black and white pictures used in the fMRI paradigm. The patient was asked to produce the correct name by saying “This is …” in Italian. The length of each stimulation block was set at 2.5 s on the stimulation software. The picture was presented for 2 s 500 ms after the stimulation starting point. Audio/video data were recorded by a video system. Cortical mapping was performed immediately after opening the dura in order to avoid brain shift. The real position of the probe on the brain surface indicating the DES positive site was acquired and sent to the neuronavigation system to be projected on T1 or T2 images and saved as a snapshot.

3. Results 3.1. Neuropsychological testing 3.1.1. Pre-surgical evaluations The patient's performance on the Raven's Colored Progressive Matrices (score: 33/36, cut-off 18: (Basso et al., 1987)), and the Mini Mental State Examination was normal (score 30/30, cut-off

25: (Magni et al., 1996)). SP reached a digit span (forward) of 5 (Orsini et al., 1987) (backward of 4: (Orsini, 2003)). He showed no ideomotor apraxia (72/72, cut-off score 52: (De Renzi et al., 1980)) nor signs of dysarthria or oral apraxia (20/20) (De Renzi et al., 1966). His spontaneous speech was fluent and grammatically and lexically correct. A mild phonemic paraphasia was observed, but SP always corrected himself. His verbal comprehension was preserved (Token test, 35/36, cut-off 29; De Renzi and Faglioni, 1978). His verbal fluency (PLF) was 45 (cut-off 20: (Novelli et al., 1996)). Phonological and lexical-semantic abilities were examined using the Battery for the analysis of language disorders (Miceli et al., 1994). Phonemic discrimination was unimpaired (58/60). SP's performance on word and pseudoword transcoding tasks was normal (see Table 1): word and pseudoword repetition (45/45 and 36/36 correct responses respectively), word and pseudoword reading aloud (90/92 and 45/45 correct responses respectively) and word and pseudoword writing to dictation (25/25 and 25/25 correct responses respectively) were well preserved. Oral picture naming was unimpaired. SP produced 30/30 correct responses to nouns and 27/28 correct responses to actions indicating well-preserved word-finding abilities. 3.1.2. Post-surgery evaluation SP's performance on the Raven's Colored Progressive Matrices was 33/36. SP showed no ideomotor apraxia (72/72) nor signs of either dysarthria or oral apraxia (20/20). SP had an altered digit span (forward) of 3 (backward of 2). The clinical assessment revealed a fluent spontaneous speech disorder characterized by phonemic paraphasia with frequent conduites d'approche, and good comprehension. The patient was aware of his difficulties. Auditory (98%) and visual noun and verb comprehension were spared (100% correct). SP's performance on the Token test was moderately impaired (19/36; cut-off 29). Phonemic discrimination was unimpaired (58/60). The patient was 60% and 64% correct at oral naming of nouns and verbs (errors were mostly phonological, as nfocchio in place of fiocco “bow” or nsiffischiare in place of fischiare “to whistle”. No semantic paraphasia or anomia emerged), 50% correct at written naming of nouns and 41% correct at written naming of verbs (his spelling errors were similar to those emerged during oral naming tasks: e.g. nteleffono in place of telefono “telephone” or nafanniare in place of annaffiare “to water”). He was very poor at repetition of both words (5/45) and pseudowords (18/ 36) (11% and 50% correct respectively), with a strong length effect (word repetition: he correctly repeated 18% of trisyllabic words

Table 1 Patient SP's neuropsychological performance before surgery, after surgery and after follow-up. Bold values indicate impaired performance. Raw Score Pre

Phonological discrimination Phonological transcoding (Pseudowords)

Phonological transcoding (Words)

Noun comprehension Verb comprehension Noun naming Verb naming

Raw score Post

Raw score Follow up

Pre- vs. post-surgery

Pre vs follow up

Post vs follow up

X2

P

X2

P

X2

P

Auditory Repetition

60/60 36/36

58/60 18/36

59/60 10/36

.51 24

.48 o.001

0 40.69

1 o .001

0 3.74

1 .053

Reading Writing Repetition

45/45 25/25 45/45

44/45 13/25 5/45

44/45 2/25 10/45

1.01 15.79 72

1 o.001 o.001

1.01 42.59 52.27

1 o .001 o .001

ns 11.52 2

ns o.001 .16

Reading Writing Auditory Visual Auditory Visual Oral Written Oral Written

90/92 25/25 – – – – 30/30 – 27/28 –

83/92 13/25 39/40 40/40 20/20 20/20 18/30 11/22 18/28 9/22

92/92 13/25 39/40 40/40 20/20 20/20 25/30 17/22 22/28 17/22

3.48 15.79 – – – – 15 – 9.16 –

.06 o.001 – – – – o.001 – .002 –

.51 15.79 – – – – 3.49 – 2.61 –

.48 o .001 – – – – .061 – .11 –

7.48 ns ns ns ns ns 2.95 3.53 1.4 6.02

.006 ns ns ns ns ns .085 .06 .24 .014

B. Tomasino et al. / Neuropsychologia 70 (2015) 214–226

and 0% of quadrisyllabic words; pseudoword repetition: he correctly repeated 87.5% of monosyllabic items, 50% of disyllabic items and 0% of trisyllabic items). Word and pseudoword reading was almost preserved (83/92 and 44/45 correct responses, i.e. 90% and 98% respectively). The few errors made during word reading were primarily phonological substitutions (as ndope in place of dopo “after”). Spelling from dictation of words and pseudowords was severely impaired (13/25 and 13/25 correct responses, i.e. 52% and 52%, respectively). 3.1.3. Follow-up at one year post-surgery The patient underwent speech therapy before the follow-up. He showed no ideomotor apraxia (72/72) and no signs of either dysarthria or oral apraxia (20/20). SP still presented with an altered digit span (forward) of 3 (backward of 2) and scored 23/36 on the Token test. His spontaneous speech was fluent with less severe phonemic paraphasia. His comprehension was adequate. Verbal comprehension and speech were slightly improved. Auditory (98%) and visual noun and verb comprehension (100% correct) were unimpaired. Naming had improved [83% and 78% correct responses at oral naming of nouns (25/30 correct) and verbs (22/28 correct)]. As in the previous assessment, his errors were all of phonological type (e.g. ncarmiriere in place of cameriere “waiter” or nversere in place of versare “to pour”). No semantic errors or anomias emerged. In both written naming of nouns and verbs he performed 17/22 correct, i.e. he was 77% correct (his writing errors demonstrated features similar to those emerging during the oral naming task, e.g. nbittoglia in place of bottiglia “bottle” or ntindere in place of tingere “to paint”). His repetition performance remained severely impaired. SP could correctly repeat 10/45 words and 10/ 36 pseudowords (22% and 28% correct, respectively): he correctly repeated 28% of trisyllabic words and 12% of quadrisyllabic words; in the pseudoword repetition task he correctly produced 56% of monosyllabic items, 10% of disyllabic items and 0% of trisyllabic items. Reading of words (92/92) and pseudowords (44/45) was preserved (100% and 98% of correct responses, respectively).

219

Writing from dictation of words (13/25) and pseudowords (2/25) remained profoundly impaired (52% and 8% of correct responses, respectively). In summary, the post-operative and the follow-up evaluations showed a pattern of impairment that is compatible with “conduction aphasia of the reproduction subtype” (Shallice and Warrington, 1970). Post-followup SP showed an improvement in oral speech output and naming but continued to display a severe impairment in repetition and spelling on dictation (words and pseudowords) while reading abilities were almost preserved (see Table 1). 3.2. fMRI results The neural network underlying the linguistic tasks was assessed by a whole-brain analysis (p o.05, FWE corrected for multiple comparisons at the cluster level, with a height threshold at the voxel level of p o.001). The network of areas found in the patient's pre-surgical fMRI mapping was comparable to that of control participants and consisted of several activation clusters (see the Supplementary Materials). 3.2.1. Main effect of condition Pre- versus post-surgery fMRI mapping (and vice versa). The network of areas differentially recruited by pre- vs. post-surgical fMRI mapping involved several clusters of activity (See Fig. 2). As regards word listening (vs. rest), we found differential activation in the left superior temporal gyrus, extending to the planum temporale (cluster 1a, see Fig. 2A), the left supramarginal gyrus (cluster 2, see Figs. 2A and 3A), the left middle temporal gyrus (cluster 3, see Figs. 2A and 3A), the right superior temporal gyrus, the precuneus bilaterally, and the right inferior frontal gyrus (see Fig. 2A and Table 2) in the pre- vs. post-surgery fMRI mapping comparison. The post- vs. pre-surgery fMRI mapping comparison showed a differential activation in the right Heschl's gyrus and the right middle temporal gyrus, in the superior occipital gyrus

Fig. 2. Network of areas differentially recruited in the pre- vs. post-surgical (and vice versa) fMRI mapping. The figure portrays the relative increase in neural activity associated with word listening (A), sentence reading (B), and pseudoword reading tasks (C) (performed at p o .05, corrected for multiple comparisons at the cluster level, family-wise error, with a height threshold at the voxel level of p o .001, uncorrected). Labels indicate that parts of cluster 1a were activated for reading sentences (1b) and reading pseudowords (1c), while clusters 2 and 3 were exclusively activated for word listening, and clusters 5a and 5b and 6a and 6b were exclusively activated in the postvs. pre-fMRI comparison for reading. Activations were superimposed on a rendered brain template provided by SPM5.

220

B. Tomasino et al. / Neuropsychologia 70 (2015) 214–226

Fig. 3. Relative increase in neural activity in the temporal cortex associated with word listening (A), sentence reading (B) and pseudoword reading (C) in the pre-surgical vs. the post-surgical (and vice versa). Activation areas are superimposed on axial, sagittal and coronal sections of the patient's pre-surgical T2 brain scan.

bilaterally, the left inferior frontal gyrus, and the left pre- and post-central gyri (see Figs. 2A and 3A and Table 2). As regards sentence reading (vs. rest), parts of cluster 1a found in the left superior temporal gyrus for word listening (see above) were differentially activated also in pre- vs. post- surgery fMRI comparison (cluster 1b, see Fig. 2B), in addition to a cluster in the left posterior occipito-temporal gyrus (cluster 4), the right middle and superior temporal gyrus, the inferior frontal gyrus, and the middle occipital gyrus bilaterally (see Figs. 2B and 3B and Table 3). In the post- vs. pre-surgery fMRI-mapping comparison activations were found in the left supramarginal gyrus (cluster 2a, see Figs. 2B-3B), the left inferior temporal gyrus (cluster 6a), the left inferior and superior parietal cortex, and the lingual gyrus bilaterally (see Table 2). As regards pseudoword reading (vs. rest) we found that parts of cluster 1 emerged in the left superior temporal gyrus for word listening and sentence reading (see above) were differentially activated (cluster 1c, see Figs. 2C–3C), in addition to the right inferior posterior temporal gyrus, the pre- and post-central gyrus bilaterally, the SMA bilaterally, the right inferior frontal gyrus, and the left superior frontal gyrus. In the post- vs. pre-surgery fMRImapping comparison, activations were found in the left supramarginal gyrus and superior temporal gyrus (cluster 2b, see Figs. 2C–3C), the left posterior inferior temporal gyrus (cluster 6b), the right middle and superior temporal gyri, the left inferior frontal gyrus bilaterally, the occipital cortex bilaterally, and the right inferior parietal cortex (see Table 4).

other in the middle temporal gyrus (MTG) (Glasser and Rilling, 2008). We found a dissociation between the sector of the arcuate fasciculus terminating in the STG and the sector terminating in the MTG. At the follow-up examination the sector terminating in the left STG differed significantly between SP and controls in terms of FA (Z¼  4.30, p o.05, with 0.39 70.09 FA for SP and 0.47 70.019 for healthy controls). All other parameters did not differ significantly between SP and healthy controls (see Table 5 and Fig. 1E–F). Furthermore, Table 6 and Fig. 4 show the number of fibres and the FA for the anterior, posterior and long segments constituting the arcuate fasciculus, according to Catani et al.'s (2005) method of reconstruction. The analysis indicates that the long sector of the superior longitudinal fasciculus (SLF) (see Catani et al., 2005) is not significantly different between SP and controls both in terms of number of fibres and FA at all three time points (pre-surgery, one week post-surgery and follow-up evaluation), with the exception of a significant difference in FA at the follow-up. Furthermore, we found that the anterior segment of SLF is less represented in terms of number of fibres in SP than in the control participants, independently of the examination time point (pre-surgery, oneweek post-surgery, and follow-up evaluation). Last, we found that at one week after surgery the posterior segment of the SLF was less represented, in terms of FA, in SP than in the control participants. However, this difference was no longer evident at the follow-up. 3.4. Intraoperative mapping

3.3. DTI results The different sectors of the arcuate fasciculus have all a frontal termination in the posterior part of the frontal lobe, but a different origin: one originates in the superior temporal gyrus (STG), the

During cortical stimulation of the functional tissue sorrounding the lesion, different patterns of direct electrical stimulation (DES) effects were found (Fig. 1D). In detail, stimulation of the inferior frontal gyrus (site 3) elicited speech arrest. The patient reported he

B. Tomasino et al. / Neuropsychologia 70 (2015) 214–226

Table 2 Brain regions showing significant increase in BOLD response for word listening (vs. fixation) in (i) pre- 4post-surgery fMRI mapping, and (ii) in post- 4pre-surgery fMRI mapping. Region

Side

MNI x

Word listening: pre- 4post-surgery Precuneus LH Precuneus RH Supramarginal gyrus LH Superior Temporal gyrus LH Heschls gyrus LH Temporal pole LH Middle Temporal gyrus LH Superior Temporal gyrus RH Inferior Frontal gyrus (pars RH triangularis) Inferior Frontal gyrus (pars RH orbitalis) Word listening: post- 4pre-surgery Superior Occipital gyrus RH Cuneus RH Calcarine gyrus RH Cuneus LH Superior Occipital gyrus LH Postcentral gyrus LH Middle Temporal gyrus RH Heschls gyrus RH Rolandic operculum RH Middle Frontal gyrus RH Inferior Frontal gyrus (pars LH triangularis) LH Inferior Frontal gyrus (pars opercularis) LH Inferior Frontal gyrus (pars triangularis) Inferior Frontal gyrus (pars LH orbitalis) Precentral gyrus LH Insula LH

y

Z z

Cluster size

4.69 4.46 5.98 Inf 7.02 5.72 4.99 4.74 4.23

62 97 76 856 – – 44 109 53

4

4.13

–

20

Table 3 Brain regions showing significant increase in BOLD response for sentence reading (vs. fixation) in (i) pre- 4post-surgery fMRI mapping, and (ii) in post- 4presurgery fMRI mapping. Region

Side MNI x

Vox

6  82 44 0  54 38  64  46 26  60  32 16  64  10 8  60 6 4  66  38  4 56  42 14 46 28 2 54

221

20  88 8  82 12  94  12  86  16  88  48  10 44 8 44  22 52  20 30 36  56 14

32 34 2 30 20 54  22 10 16 24 28

6.60 4.93 6.41 4.68 4.10 5.19 4.23 Inf 5.62 6.13 6.04

 52 6

18

4.72

 46 32

0

5.35 140

 52 26

6

5.20

 48 0  30 22

34 0

5.27 41 4.41 69

117 – 837 83 57 42 261 – 88 159

For each region of activation, the MNI space coordinates are provided with reference to the maximally activated voxel within an area of activation, as indicated by the highest Z-value (Po 0.05, corrected for multiple comparisons at the cluster level, height threshold P o.001, uncorrected). LH/RH ¼ left/right hemisphere.

could not talk (“Non riuscivo ad esprimere la parola. È la stessa sensazione che ho avuto durante le crisi”, “I could not speak. It's the same feeling I had during seizures”). Stimulation of the lower part of the precentral gyrus (site 4) evoked a contraction of the lips (“Ho avuto contrazione delle labbra, come se stesse per uscire la saliva”, “I felt a contraction of the lips as if saliva was about to drip from my mouth”). Stimulation of the lower part of the postcentral gyrus (site 2) caused paresthesia of the tongue and teeth, and stimulation of another site -slightly superior to the previous one in the lower part of the precentral gyrus (site 1)- evoked paresthesia of the right-side of the tongue. Stimulation of the middle temporal gyrus (site 5) caused anomia. SP hesitated and then produced the target word. In addition, his production was contaminated by phonemic paraphasias (e.g., nginocco in place of ginocchio, knee; nado in place of dado, dice). Stimulation of other two sites in the middle temporal gyrus (sites 6 and 7) caused anomia as well. SP showed difficulty to name objects or actions. (“Questo è un …, non ce la facevo a buttar fuori la parola”, This is a… I couldn't utter the word”). In addition, DES of the middle temporal gyrus (site 6) elicited phonemic paraphasias during word and sentence repetition (“la eeeh, la moto insegna la macchina” in place of la moto insegue la macchina, the bike follows the car; “il cavallo è ninsiguato dal cane” in place of il cavallo è inseguito dal cane, the horse is chased by the dog; “il ncavàno incine il cane” in place of il cavallo

Sentence reading: pre 4 post Middle Occipital gyrus Middle Occipital gyrus Angular gyrus Superior parietal lobule Postcentral gyrus Rolandic operculum Precentral gyrus Precentral gyrus Precentral gyrus Inferior Frontal gyrus (Pars orbitalis) Inferior Frontal gyrus (Pars opercularis) Middle Temporal gyrus Superior Temporal gyrus Middle Temporal gyrus Superior Temporal gyrus Middle Temporal gyrus Sentence reading: post 4 pre Lingual gyrus Lingual gyrus Calcarine gyrus Inferior Parietal lobule Superior Parietal lobule Supramarginal gyrus Middle cingulate cortex Inferior Temporal gyrus

Z y

z

Cluster size Vox

LH RH RH RH LH LH LH LH RH LH

 32  94 38  78 32  66 24  76  60  4  64  10  60 2  34  2 48 2  52 22

6 34 46 50 18 12 4 32 40 4

6.46 5.05 4.48 4.00 5.86 5.45 3.56 5.13 4.58 5.04

RH

44

16

32

4.66 49

LH LH LH RH RH

 66  68  50 64 68

 44  28  70  34  34

10 12 8 8 0

5.11 80 3.87 4.83 54 4.32 50 3.84

RH RH LH LH LH LH RH LH

18  74 20  50  24  70  34  70  26  78  48  36 10 10  44  64

8 8 8 46 46 24 40 6

Inf 5.14 5.74 5.65 5.44 Inf 4.62 4.60

47 156

77

148 47 80

679 88 88 108 63 39 48

For each region of activation, the MNI space coordinates are provided with reference to the maximally activated voxel within an area of activation, as indicated by the highest Z-value (Po .05, corrected for multiple comparisons at the cluster level, height threshold Po .001, uncorrected). LH/RH ¼left/right hemisphere.

insegue il cane, the horse is chasing the dog; “balata” in place ofbolata, pseudoword; “svolga” in place of svolge, (he) does). Stimulation of the temporo-parietal cortex (sites 8–11) evoked errors in phonological processing during the verbal auditory comprehension and repetition task. No positive sites were found during subcortical mapping. The results are consistent with of the view that in the case of negative mapping neuropsychological deficits may arise post-operatively.

4. Discussion In the present study we explored the pattern of impairment observed in an Italian conduction aphasic patient (of the reproduction type) who developed a fluent but paraphasic speech output with profound phonological deficits in repetition and preserved auditory comprehension after surgery for a tumor circumscribed to the left superior temporal gyrus. Patient SP had two peculiarities: first, his lesion was anterior/inferior to and neighboring the Sylvian parietal-temporal (Spt) area; second, in addition to his repetition impairment, he presented spared word and pseudoword reading aloud and impaired word and pseudoword spelling from dictation. The pattern of symptoms observed is inconsistent with the classical disconnection hypothesis between Broca's and Wernicke's area as well as with standard information processing models of word and pseudoword processing (e.g. Patterson, 1986), while it is compatible with an impaired sensorymotor integration for speech (Hickok and Poeppel, 2007). One might argue that a pattern of damage with impaired

222

B. Tomasino et al. / Neuropsychologia 70 (2015) 214–226

Table 4 Brain regions showing significant increase in BOLD response for pseudo-word reading (vs. fixation) in (i) pre- 4post-surgery fMRI mapping, and (ii) in post4pre-surgery fMRI mapping.

Table 5 Results from the DTI reconstruction of the parts of the arcuate fasciculus terminating in the MTG and in the STG (FA ¼ Fractional anisotropy). MTG

Region

Z

Side MNI x

Pseudo-word reading: pre 4post Superior Parietal lobule LH Superior Parietal lobule RH Angular gyrus RH Middle Occipital gyrus RH Middle Occipital gyrus RH Postcentral gyrus LH Rolandic Operculum RH Postcentral gyrus RH Postcentral gyrus LH Superior Frontal gyrus LH Postcentral gyrus LH Precentral gyrus LH Precentral gyrus RH Inferior Frontal gyrus (pars RH opercularis) Precentral gyrus RH Inferior Frontal gyrus (pars RH triangularis) Rolandic Operculum RH Superior Medial gyrus RH Supplementary Motor Area LH (SMA) Supplementary Motor Area RH (SMA) Middle Temporal gyrus LH Inferior Temporal gyrus RH Pseudo-word reading: post 4pre Superior Occipital gyrus RH Calcarine gyrus RH Calcarine gyrus LH Calcarine gyrus RH Inferior Parietal lobule RH Supramarginal gyrus LH Superior Temporal gyrus LH Middle Temporal gyrus LH Inferior Temporal gyrus LH Middle Temporal gyrus RH Superior Temporal gyrus RH Inferior Frontal gyrus (pars LH opercularis) Inferior Frontal gyrus (pars LH triangularis) Inferior Frontal gyrus (pars LH orbitalis) Inferior Frontal gyrus (pars LH opercularis) Inferior Frontal gyrus (pars RH opercularis)

y

 32  64 32  68 40  60 40  80 36  84  48  30 50 8 64 4  26  30  22  8  60  4  38  10 46 2 58 14

z

54 52 50 34 12 48 18 18 62 52 20 38 38 34

Vox

4.65 5.67 4.94 4.80 4.69 5.72 4.80 4.47 5.03 4.88 Inf 5.32 6.80 6.14

129 366

48 53 109 203

Number of fibers Pre-surgery 655 (Z¼0.29, p 4.05, n.s.) 409 One week (Z¼  0.69, postp 4.05, n.s.) surgery One year 396 follow up (Z¼  0.74, p 4.05, n.s.) Healthy 588 7 257.41 controls

FA

Number of fibers

FA

0.4787 0.12 (Z¼  0.7, p 4.05, n.s.) 0.459 7 0.12 (Z¼  1.4, p 4.05, n.s.) 0.45 7 0.69 (Z¼  1.7, p 4.05, n.s.) 0.4967 0.026

60 (Z¼  0.94, p 4.05, n.s.) –

0.44 70.12 (Z¼  1.38, p 4.05, n.s.) –

5 (Z¼  1.68, p 4.05, n.s.) 1307 74.084

0.39 70.093 (Z¼  4.3, po .05) 0.477 0.019

465 320

22 42

 28 64 14 12

4.62 72 4.23 197

60 0 4

8 64 10

10 20 50

4.23 4.19 47 6.68 65

4

12

46

3.79

 68 46

 36 8 5.25 64  72  4 4.28 70

24  92 16  90 8  80 28  64 60  48  50  34  58  32  56  64  60  58 68  40 68  34  60 10

28 4 4 6 44 24 22 2 6 4 12 6

4.81 4.81 4.57 4.44 5.28 6.67 4.33 5.63 3.55 4.76 4.48 5.87

 48 34

0

5.66 143

 44 24

 2 3.45

 58 10

22

4.70 116

50

0

4.68 57

16

STG

Cluster size

42 321 64 61 90 66 76 68

For each region of activation, the MNI space coordinates are provided with reference to the maximally activated voxel within an area of activation, as indicated by the highest Z-value (P o.05, corrected for multiple comparisons at the cluster level, height threshold Po .001, uncorrected). LH/RH ¼ left/right hemisphere.

repetition and spared reading-aloud could be due to a functional dissociation between visual input (preserved) and auditory input (damaged). We argue that, based on data reported in the literature concerning conduction aphasia and the functional models of phonological output, such dissociation is too a simplistic view compared with the state of the art on such issue. In point of fact, the very large majority of patients suffering from CA of the reproduction type show a phonological damage that actually involves repetition, reading aloud, naming and spontaneous speech homogeneously, i.e. a functional impairment at the phonological output buffer level, instead of the dissociation between Broca's and Wernicke's areas assumed by the Wernicke–Lichtheim's

Bold values indicate a significant difference between the patient's and controls' FA values for the tract terminating in the STG.

account. However, this functional damage is incompatible with the dissociation found in SP. The tractography reconstruction of SP's white matter tracts clearly showed (in the post-operative MRI) an interruption of the sector of the arcuate fasciculus terminating in STG. As a consequence, the phonological circuitry necessary for normal repetition was damaged. When the input is auditory, as it happens for word and pseudoword repetition, or for writing from dictation, SP's altered performance depends on an impaired sector of the arcuate fasciculus terminating in STG. By contrast, SP's reading was spared, thus suggesting that the preserved part of the arcuate fasciculus terminating in MTG still allows a link between the visual input and the speech output processing areas. SP's dissociation may account for the observed pattern of performance with impaired repetition and spelling from dictation but preserved reading aloud. Indeed, by reviewing data reported in literature we found that phonemic errors usually occur also in oral reading (Goodglass, 1992), as well as in writing (from mild spelling difficulties to profound agraphia) (Bernal and Ardila, 2009). Such errors parallel the phonological production errors in conduction aphasia (Goodglass, 1992). However, only very few studies of conduction aphasia specifically looked at writing disorders (Axer et al., 2001; Goodglass, 1992). SP produced a large number of phonemic paraphasias during repetition and phonological/spelling errors when writing words and pseudowords, which were however read aloud flawlessly. Furthermore, in the majority of the reviewed cases (Anderson et al., 1999; Axer et al., 2001; Balasubramanian, 2005; Baldo et al., 2008; Bartha and Benke, 2003; Benson et al., 1973; Buchsbaum et al., 2011; Caplan et al., 1986; Damasio and Damasio, 1980; Fridriksson et al., 2010; Gandour et al., 1991; Gandour, 2013; Mendez and Benson, 1985; MoritzGasser and Duffau, 2013; Quigg et al., 2006), reading impairments were not explicitly reported or, in a few cases, the patients suffered from the repetition subtype of conduction aphasia, which is a phonological short-term memory disorder and thus does not affect reading performance. None of the 53 cases of CA (reproduction subtype) we could trace in literature suffered from a phonological output deficit in repetition and no phonological impairment in reading-aloud. Thus, the dissociation emerging in SP between impaired word and pseudoword repetition (and writing) and spared readingaloud, lead us to address the role of the dorsal stream in phonological processing (e.g., Dick and Tremblay, 2012; Glasser and Rilling, 2008; Hickok, 2009; Hickok and Poeppel, 2007). The dorsal stream has been related to the functionality of the superior longitudinal fasciculus/arcuate fasciculus (SLF/AF) (Dick

B. Tomasino et al. / Neuropsychologia 70 (2015) 214–226

223

Table 6 Results from the DTI reconstruction of the arcuate fasciculus according to Catani et al.'s (2005) method (FA ¼Fractional anisotropy). Bold values indicate a significant difference between the patient's and controls' FA and number of fibres. ANTERIOR Number of fibers Pre-surgery

POSTERIOR FA

LONG

Number of fibers

FA

Number of fibers

449 0,45 927 (Z ¼  2,28, po .05) (Z¼  0,01, p 4.05, n. (Z¼  0,15, p4 .05, n.s.) s.) 574 0,41 919

0,4 1928 (Z ¼  1,50, p 4.05, n. (Z¼ 1,07, p 4.05, n.s.) s.) 0,31 1060

(Z ¼  2,05, po .05) (Z ¼  2,31, po 05)

(Z¼  0,17, p 4.05, n.s.)

(Z ¼  4,59, p o.05)

One year follow up

428 0,37 (Z ¼  2,32, po .05) (Z ¼  4,60, po .05)

925 (Z¼ -0,15, p4 .05, n. s.)

Healthy controls

1656,83 7 528,69

986,50 7 406,65

0,41 (Z ¼  1,15, p 4.05, n. s.) 0,447 0,03

One week postsurgery

0,45 70,02

and Tremblay, 2012; see also Saur et al., 2008). However, there is disagreement in the literature about the role and the origins, terminations, and extent of the fibre pathways forming the dorsal and the ventral streams (Dick and Tremblay, 2012). Our case study gave us the opportunity to relate neuropsychological data with neuroanatomy. We further addressed the role and the origins and

FA 0,487 (Z¼  0,40, p4 .05, n. s.) 0,46

(Z¼  0,86, p 4.05, n.s.) (Z¼  1,34, p 4.05, n. s.) 1266 0,44 (Z¼  0,40, p 4.05, n.s.) (Z¼  2,03, po .05) 1448,007 449,58

0,50 7 0,03

terminations of the dorsal stream by reconstructing the fibres presurgery, one week post-surgery, and at the follow-up evaluation. In particular we used two different approaches to reconstruct the superior longitudinal fasciculus/arcuate fasciculus (SLF/AF): the methods proposed by Catani et al. (2005) and by Glasser and Rilling (2008). Catani and colleagues (2005; Thiebaut de Schotten

Fig. 4. Pre-operative (red) and post-operative (yellow) and 1-year follow-up (blue) reconstructions of the anterior, posterior and long segment of the arcuate fasciculus. The figure also depicts the number of fibres and mean FA for controls and patient SP. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

224

B. Tomasino et al. / Neuropsychologia 70 (2015) 214–226

et al., 2011b) delineated the SLF/AF into three segments: a long direct temporo-frontal segment (phonetic) that is essentially the classical SLF/AF pathway, and an indirect semantic pathway (via 2 segments that connect the inferior parietal lobe to both the temporal and the frontal lobes). By contrast, Glasser et al. (2008) proposed to separate the arcuate fasciculus into two terminations: the STG pathway (phonologic processing) and the MTG pathway (lexical-semantic processing). The main difference between the two methods is that Catani et al. model combines the STG and MTG pathways of the arcuate fasciculus into a single segment, arguing that the entire pathway conveys phonetic information directly to the frontal lobe. Glasser et al. showed instead that the STG and MTG terminations of the arcuate fasciculus connect cortices with different functions and suggested therefore that they should be considered as two separate pathways. The tractography reconstruction of SP's white matter tracts showed that the long segment of the superior longitudinal fasciculus (SLF), corresponding to the sector of SLF related to the dorsal phonological route (Duffau et al., 2003; Makris et al., 2005), did not significantly differ from that of the control participants in terms of number of fibres and FA at all three time points, with the exception of a significant difference in FA at follow-up. However, this difference could not explain the pattern observed since SP's deficit emerged after surgery (and not only at follow-up). In addition, the lesion site would not explain why pseudoword reading was spared since, according to Catani's model (Catani et al., 2005), the SLF long segment is directly involved in phonological processing. Furthermore, we found that the anterior segment of the SLF was less represented –in terms of fibres– in SP than in the control participants independently of the examination time point. Thus, it could not explain SP's pattern of deficit since his pre-surgery performance was in the normal range. Last, we found that the posterior SLF segment was less represented in terms of FA in SP than in the control participants at one week after surgery. However, this difference was no longer evident at the follow-up evaluation and could not explain SP's pattern of performance, since his repetition performance was still impaired at follow-up. By contrast, the DTI tractography reconstruction of SP's fibers according to Glasser and Rilling (2008) clearly showed (in the post-operative MRI) an interruption of the sector of the arcuate fasciculus terminating in STG. As a consequence, the phonological circuitry necessary for normal repetition was damaged. When the input is auditory, as it happens for word and pseudoword repetition or writing from dictation, SP's altered performance depends on an impaired sector of the arcuate fasciculus terminating in STG. Furthermore, SP's reading performance was spared, thus suggesting that the preserved part of the arcuate fasciculus terminating in MTG still allows a link between visual input and the speech output processing areas. It is held that reading is supported by a lexical-semantic route, which promotes direct access from word shapes to meaning, and includes the left occipito-temporal region (Cohen et al., 2000), which is more activated post- than pre-surgery. The inferior fronto-occipital fasciculus (IFOF), connecting the visual occipital areas with the inferior and middle temporal gyrus and with the orbitofrontal cortex, has been proposed as the neuroanatomical correlate of the lexical reading route (Epelbaum et al., 2008; Vandermosten et al., 2012a, 2012b). The DTI results indicate that the patient's IFOF as well as the inferior longitudinal fasciculus (ILF) are preserved. SP made phonological errors in repetition and in writing (reading is undamaged), which indicated that the impaired processing route is rather the dorsal route. The left arcuate fasciculus subserves sublexical reading along the orthographic-to-phonological conversion route (Vandermosten et al., 2012a, 2012b). Our data suggest, therefore, that the part of the arcuate fasciculus terminating in STG is not relevant to warrant a preserved reading performance. The spared reading

performance of SP could be supported in principle via the ventral stream. This however could only explain the word reading performance but could not account for the fact that pseudoword reading was proficient. To sum up, only the model proposed by Glasser and Rilling (2008) as compared to Catani et al. (2005) could explain our pattern of dissociation. In SP, the spared part of the left arcuate fasciculus originating in MTG may support correct functioning of his sublexical route of reading, while the damaged part of the left arcuate fasciculus that originates in STG may be responsible for his impaired repetition and spelling from dictation. This interpretation has been visually rendered in Fig. 1G by applying the model proposed by Glasser and Rilling (2008) to the pattern of dissociation emerging in SP. Glasser and Rilling (2008) argued that their findings may be interpreted within the framework of Hickok and Poeppel's (2004) language production model. We integrated our fMRI and DTI results into the Directions into Velocities of Articulators (DIVA) model (Golfinopoulos et al., 2010), which is a neurocomputational approach to Hickok and Poeppel's model (2004). SP could correctly detect his own overt speech upon hearing, tried to repair his errors and developed a conduite d'approche behavior. The sensory response to self-generated speech, which is processed in the socalled auditory and somatosensory state maps (which may be localized along Heschl's gyrus/the planum temporale and the inferior postcentral gyrus (Golfinopoulos et al., 2010; Hickok et al., 2011)) according to the DIVA model was not affected by the lesion. It is held that the critical part of the STS that is involved in phonological processing is anterior to the anterolateral part of Heschl's gyrus and posterior to the posterior part of the Sylvian fissure (Hickok, 2009; Hickok and Poeppel, 2007). As revealed by the neuropsychological assessment (especially the phonological auditory discrimination task), SP's early phonological input processing was spared. SP's lesion is located in the posterior and superior part of the left temporal lobe. According to the DIVA model, in this sector of the temporal cortex, when producing a speech sound, the target maps can predict the sensory consequences of the sounds to be produced prior to the arrival of an actual sensory feedback and can use such estimates to provide rapid corrective feedback. We argue that the lesion impaired the ability to calculate the difference between the actual (SP correctly detects his own overt speech upon hearing and tries to correct errors) and expected sensory responses for the produced speech sounds (likely altered). With regard to both word listening and sentence and pseudowords reading, a cluster of activation was found in the Spt area (Fig. 2: clusters 1a, 1b and 1c, respectively), which was more activated pre- than post-surgery. There is a strong correspondence between the coordinates of this area (x¼  60, y¼  32 and z¼ 16) and the coordinates at which Golfinopoulos et al. (2010) localized the target and error maps (x ¼  64.6, y¼  33.2 and z ¼13.5, see Table 1). The decreased fMRI activation in the Spt area found during reading may not be crucial since SP's reading-aloud was preserved. However, SP's altered performance on word and pseudoword repetition or writing from dictation depends on a decreased Spt activation during the sensorimotor translation processing (and the processing of expectations). In addition to the Spt area, a greater pre- vs. post-surgery activation was observed in the left supramarginal gyrus (cluster 2, Fig. 2), where the somatosensory target maps are localized according the DIVA model. This pattern confirms the role of the Spt area and the supramarginal gyrus in the emergence of CA (Hickok, 2000, 2009; Hickok and Poeppel, 2004). In other words, phonological errors occur because sensory representations of speech are prevented from providing online guidance for speech sound sequencing (Hickok and Poeppel, 2007).

B. Tomasino et al. / Neuropsychologia 70 (2015) 214–226

fMRI studies in patients (especially neurosurgical patients) are rare and have the advantage of reporting, at variance with stroke patients' data, small and selective lesions. We are aware that single-case fMRI studies have the potential pitfall of leading to posthoc interpretation of results and raise a range of distinct problems. However, fMRI studies in patients may be relevant for testing cognitive models of mental processing (see Chatterjee, 2005; Price and Friston, 1999; Price et al., 2006) when performed taking in consideration some key issues in the experimental design and data analysis. In particular, a pre-requisite for interpreting fMRI data in patients is to study them “with tasks they can perform” (Price and Friston, 1999). As regards the present fMRI investigation, we used two fMRI tasks that SP could perform normally, namely a word listening task and two reading tasks (sentence and pseudoword reading). Despite the pitfall of single-case fMRI studies, we argue that such investigations, combined with neuropsychological data and structural imaging, are worth to be performed and challenge current models of language functioning. In the present study, for example, given the models proposed by Catani et al. (2005) and by Glasser and Rilling (2008), only the second model could explain the pattern of dissociation seen in SP. To conclude, this paper describes a dissociated pattern of performance in a patient with conduction aphasia showing impaired repetition and spelling from dictation but spared reading-aloud. SP's pattern of impairment suggests that different anatomical lesions and functional patterns of damage can be related to conduction aphasia. In other words, different combinations of symptoms (repetition, picture naming, reading, and/or spelling deficits) can be caused by a damage to different parts of the anatomical network, including white and grey matter, causing different patterns of functional damage. We successfully adopted a model that showed great explanatory power (Golfinopoulos et al., 2010; Hickok, 2009; Hickok and Poeppel, 2004, 2007), combined with a method of fiber reconstruction that dissociates the part of the arcuate fasciculus terminating in STG and MTG (Glasser and Rilling, 2008). We thus underline the importance of combining neuropsychological, functional and DTI data with a cognitive model in order to account for the dysfunctional pattern of neuropsychological patients.

Appendix A. Suplementary Information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.neuropsychologia. 2015.02.023.

References Anderson, J.M., Gilmore, R., Roper, S., Crosson, B., Bauer, R.M., Nadeau, S., et al., 1999. Conduction aphasia and the arcuate fasciculus: a reexamination of the Wernicke–Geschwind model. Brain Lang. 70, 1–12. Ardila, A., 2010. A review of conduction aphasia. Curr. Neurol. Neurosci. Rep. 10, 499–503. Axer, H., von Keyserlingk, A.G., Berks, G., von Keyserlingk, D.G., 2001. Supra- and infrasylvian conduction aphasia. Brain Lang. 76, 317–331. Baddley, A., 2003. Working memory: looking back and looking forward. Nat. Rev. Neurosci. 4, 829–839. Balasubramanian, V., 2005. Dysgraphia in two forms of conduction aphasia. Brain Cognit. 57, 8–15. Baldo, J.V., Klostermann, E.C., Dronkers, N.F., 2008. It's either a cook or a baker: patients with conduction aphasia get the gist but lose the trace. Brain Lang. 105, 134–140. Bartha, L., Benke, T., 2003. Acute conduction aphasia: an analysis of 20 cases. Brain Lang. 85, 93–108. Basso, A., Capitani, E., Laiacona, M., 1987. Raven's coloured progressive matrices: normative values on 305 adult normal controls. Funct. Neurol. 2, 189–194. Benson, D.F., Ardila, A., 1996. Aphasia: A Clinical Perspective. Oxford University Press, New York.

225

Benson, D.F., Sheremata, W.A., Bouchard, R., Segarra, J.M., Price, D., Geschwind, N., 1973. Conduction Aphasia. A clinicopathological study. Arch. Neurol. 28, 339–346. Berger, M.S., Ojemann, G.A., 1992. Intraoperative brain mapping techniques in neuro-oncology. Stereotact. Funct. Neurosurg. 58, 153–161. Berger, M.S., Ojemann, G.A., Lettich, E., 1990. Neurophysiological monitoring during astrocytoma surgery. Neurosurg. Clin. N. Am. 1, 65–80. Bernal, B., Ardila, A., 2009. The role of the arcuate fasciculus in conduction aphasia. Brain 132, 2309–2316. Buchsbaum, B.R., Baldo, J., Okada, K., Berman, K.F., Dronkers, N., D'Esposito, M., et al., 2011. Conduction aphasia, sensory-motor integration, and phonological short-term memory-an aggregate analysis of lesion and fMRI data. Brain Lang. 119, 119–128. Caplan, D., Vanier, M., Baker, C., 1986. A case study of reproduction conduction aphasia I: word production. Cognit. Neuropsychol. 3, 99–128. Caramazza, A., Basili, A.G., Koller, J., Berndt, R.S., 1981. An investigation of repetition and language processing in a case of conduction aphasia. Brain Lang. 14, 235–271. Caramazza, A., Miceli, G., Villa, G., 1986. The role of the output phonological buffer in reading, writing and repetition. Cognit. Neuropsychol. 3, 37–76. Chatterjee, A., 2005. A madness to the methods in cognitive neuroscience? J. Cognit. Neurosci. 17, 847–849. Chein, J.M., Fiez, J.A., 2001. Dissociation of verbal working memory system components using a delayed serial recall task. Cerebr. Cortex 11, 1003–1014. Cohen, L., Dehaene, S., Naccache, L., Lehericy, S., haene-Lambertz, G., Henaff, M.A., et al., 2000. The visual word form area: spatial and temporal characterization of an initial stage of reading in normal subjects and posterior split-brain patients. Brain 123, 291–307. Damasio, H., Damasio, A.R., 1980. The anatomical basis of conduction aphasia. Brain 103, 337–350. De Bleser, R., Cubelli, R., Luzzatti, C., 1993. Conduction aphasia, misrepresentations and word representations. Brain Lang. 45, 475–494. De Renzi, E., Faglioni, P., 1978. Normative data and screening power of a shortened version of the Token Test. Cortex 14, 41–49. De Renzi, E., Motti, F., Nichelli, P., 1980. Imitating gestures. A quantitative approach to ideomotor apraxia. Arch. Neurol. 37, 6–10. De Renzi, E., Piezcuro, A., Vignolo, L.A., 1966. Oral apraxia and aphasia. Cortex 2, 50–73. Eickhoff, S., Stephan, K.E., Mohlberg, H., Grefkes, C., Fink, G.R., Amunts, K., et al., 2005. A new SPM toolbox for combining probabilistic cytoarchitectonic maps and functional imaging data. Neuroimage 25, 1325–1335. Epelbaum, S., Pinel, P., Gaillard, R., Delmaire, C., Perrin, M., Dupont, S., et al., 2008. Pure alexia as a disconnection syndrome: new diffusion imaging evidence for an old concept. Cortex 44, 962–974. Fridriksson, J., Kjartansson, O., Morgan, P.S., Hjaltason, H., Magnusdottir, S., Bonilha, L., et al., 2010. Impaired speech repetition and left parietal lobe damage. J. Neurosci. 30, 11057–11061. Friston, K.J., Frith, C.D., Turner, R., Frackowiak, R.S.J., 1995a. Characterising evoked hemodynamics with fMRI. Neuroimage 2, 157–165. Friston, K.J., Holmes, A.P., Worsley, K.J., Poline, J.-B., Frith, C.D., Frackowiak, R.S.J., 1995b. Statistical parametric maps in functional imaging: a general linear approach. Hum. Brain Mapp. 2, 189–210. Gandour, J., Marshall, R.C., Young Kim, S., Neuburger, S., 1991. On the nature of conduction aphasia: a longitudinal case study. Aphasiology 5, 291–306. Gandour, J.T., 2013. A functional deficit in the sensorimotor interface component as revealed by oral reading in Thai conduction aphasia. J. Neurolinguist. 26, 337–347. Glasser, M.F., Rilling, J.K., 2008. DTI Tractography of the human Brain's Language Pathways. Cerebr. Cortex 18, 2471–2482. Golfinopoulos, E., Tourville, J.A., Guenther, F.H., 2010. The integration of large-scale neural network modeling and functional brain imaging in speech motor control. Neuroimage 52, 862–874. Goodglass, H., 1992. Diagnosis of Conduction Aphasia In: Kohn, S.E. (Ed.), Conduction Aphasia. Lawrence Erlbaum Associates, N.J.: Hillsdale, pp. 39–49. Hickok, G., 2009. The functional neuroanatomy of language. Physics of Life Reviews 6, 121–143. Hickok, G., 2000. Speech perception, conduction aphasia, and the functional neuroanatomy of language In: Grodzinsky, Y., Shapiro, L., winney, D. (Eds.), Language and the Brain. Academic Press, San Diego. Hickok, G., Buchsbaum, B., Humphries, C., Muftuler, T., 2003. Auditory-motor interaction revealed by fMRI: speech, music, and working memory in area Spt. Cognit. Neurosci. 15, 673–682. Hickok, G., Erhard, P., Kassubek, J., Helms-Tillery, A.K., Naeve-Velguth, S., Strupp, J.P., et al., 2000. A functional magnetic resonance imaging study of the role of left posterior superior temporal gyrus in speech production: implications for the explanation of conduction aphasia. Neurosci. Lett. 287, 156–160. Hickok, G., Houde, J., Rong, F., 2011. Sensorimotor integration in speech processing: computational basis and neural organization. Neuron 69, 407–422. Hickok, G., Poeppel, D., 2004. Dorsal and ventral streams: a framework for understanding aspects of the functional anatomy of language. Cognition 92, 67–99. Hickok, G., Poeppel, D., 2007. The cortical organization of speech processing. Nat. Revi. Neurosci. 8, 393–402. Hillis, A.E., 2007. Aphasia: progress in the last quarter of a century. Neurology 69, 200–213. Jacquemot, C., Scott, S.K., 2013. What is the relationship between phonological short-term memory and speech processing? Trends Cognit. Sci. 10, 480–486.

226

B. Tomasino et al. / Neuropsychologia 70 (2015) 214–226

Jiang, H., van Zijl, P.C., Kim, J., Pearlson, G.D., Mori, S., 2006. DtiStudio: resource program for diffusion tensor computation and fiber bundle tracking. Comput. Methods Progr. Biomed. 81, 106–116. Lichtheim, L., 1885. On Aphasia. Brain 7, 433–484. Magni, E., Binetti, G., Padovani, A., Cappa, S.F., Bianchetti, A., Trabucchi, M., 1996. The mini-mental state examination in alzheimer's disease and multi-infarct dementia. Int. Psychogeriatr. 8, 127–134. Mendez, M.F., Benson, D.F., 1985. Atypical conduction aphasia. Arch. Neurol. 42, 886–891. Makris, N., Kennedy, D.N., McInerney, S., Sorensen, A.G., Wang, R., Caviness, V.S., et al., 2005. Segmentation of subcomponents within the superior longitudinal fascicle in humans: a quantitative, in vivo, DT-MRI study. Cerebr. Cortex 15, 854–869. Miceli, G., Laudanna, A., Burani, C., Capasso, R., 1994. Batteria Per l'analisi Dei Deficit Sfasici. CEPSAG, Roma (B.A.D.A.). Mori, S., Crain, B.J., Chacko, V.P., van Zijl, P.C., 1999. Three-dimensional tracking of axonal projections in the brain by magnetic resonance imaging. Ann. Neurol. 45, 265–269. Mori, S., van Zijl, P.C., 2002. Fiber tracking: principles and strategies-a technical review. NMR in Biomed. 15, 468–480. Moritz-Gasser, S., Duffau, H., 2013. The anatomo-functional connectivity of word repetition: insights provided by awake brain tumor surgery. Front. Hum. Neurosci., 7. Novelli, G., Papagno, C., Capitani, E., Laiacona, M., Vallar, G., Cappa, S.F., 1996. Tre test clinici di ricerca e produzione lessicale. Taratura su soggetti normali. Arch. Psicol. Neurol. Psichiatr. 47, 477–505. Oldfield, R.C., 1971. The assessment and analysis of handedness: the Edinburgh Inventory. Neuropsychologia 9, 97–113. Orsini, A., 2003. La memoria diretta e la memoria inversa di cifre in soggetti dai 16 ai 64 anni. Boll. Psicol. Appl. 239, 73–77. Orsini, A., Grossi, D., Capitani, E., Laiacona, M., Papagno, C., Vallar, G., 1987. Verbal and spatial immediate memory span: normative data from 1355 adults and 1112 children. Ital. J. Neurol. Sci. 8, 539–548. Patterson, K., 1986. Lexical but not semantic spelling? Cognit. Neuropsychol. 3, 341–367. Price, C.J., Friston, K.J., 1999. Scanning patients with tasks they can perform. Hum. Brain Mapp. 8, 102–108. Price, C.J., Crinion, J., Friston, K.J., 2006. Design and analysis of fMRI studies with neurologically impaired patients. J. Magn. Reson. Imag. 23, 816–826. Quigg, M., Geldmacher, D.S., Elias, W.J., 2006. Conduction aphasia as a function of

the dominant posterior perisylvian cortex. J. Neurosurg. 104, 845–848. Romani, C., 1992. Are there distinct input and output buffers? Evidence from an aphasic patient with an impaired output buffer. Lang. Cognit. Process. 7, 131–162. Saur, D., Kreher, B.W., Schnell, S., Kummerer, D., Kellmeyer, P., Vry, M.S., et al., 2008. Ventral and dorsal pathways for language. Proc. Natl. Acad. Sci. USA 105, 18035–18040. Shallice, T., Rumiati, R.I., Zadini, A., 2000. The selective impairment of the phonological output buffer. Cognit. Neuropsychol. 17, 517–546. Shallice, T., Warrington, E.K., 1970. Independent functioning of verbal memory stores: a neuropsychological study. Q. J. Exp. Psychol. 22, 261–273. Shallice, T., Warrington, E.K., 1977. Auditory short term memory impairment and conduction aphasia. Brain Lang. 4, 479–491. Spinnler, M., Tognoni, G., 1987. Standardizzazione e taratura italiana di test neuropsicologici. Ital. J. Neurol. Sci. Suppl. 8, 1–120. Szelenyi, A., Bello, L., Duffau, H., Fava, E., Feigl, G.C., Galanda, M., et al., 2010. Intraoperative electrical stimulation in awake craniotomy: methodological aspects of current practice. Neurosurg. Focus 28, E7. Vandermosten, M., Boets, B., Poelmans, H., Sunaert, S., Wouters, J., Ghesquiere, P., 2012a. A tractography study in dyslexia: neuroanatomic correlates of orthographic, phonological and speech processing. Brain 135, 935–948. Vandermosten, M., Boets, B., Wouters, J., Ghesquiere, P., 2012b. A qualitative and quantitative review of diffusion tensor imaging studies in reading and dyslexia. Neurosci. Biobehav. Rev. 36, 1532–1552. Wakana, S., Caprihan, A., Panzenboeck, M.M., Fallon, J.H., Perry, M., Gollub, R.L., et al., 2007. Reproducibility of quantitative tractography methods applied to cerebral white matter. Neuroimage 36, 630–644. Wakana, S., Jiang, H., Nagae-Poetscher, L.M., van Zijl, P.C., Mori, S., 2004. Fiber tractbased atlas of human white matter anatomy. Radiology 230, 77–87. Wernicke C. 1874. Der Aphasische Symptomencomplex. Eine Psychologische Studie auf Anatomischer Basis. Breslau: Cohn & Weigert. (English Translation edited by R.S. Cohen and M.W. Wartofsky: The symptom complex of aphasia. A psychological study on anatomical basis. In: Boston Studies in the Phylosophy of Science; Proceedings of the Boston Colloquium for the Phylosophy of Science, 1966-1968, Volume 4. Dordrecht: Reidel; pp. 34-97). Wise, R.J., Scott, S.K., Blank, S.C., Mummery, C.J., Murphy, K., Warburton, E.A., 2001. Separate neural subsystems within ‘Wernicke's area’. Brain 124, 95.