Training-Induced Brain Remapping in Chronic Aphasia: A Pilot Study Paolo Vitali, PhD, Jubin Abutalebi, MD, Marco Tettamanti, PhD, Massimo Danna, MSc, Ana-Inés Ansaldo, PhD, Daniela Perani, MD, Yves Joanette, PhD, and Stefano F. Cappa, MD Background. The neural correlates of training-induced improvements of cognitive functions after brain damage remain still scarcely understood. In the specific case of aphasia, although several investigations have addressed the issue of the neural substrates of functional recovery, only a few studies have attempted to assess the impact of language training on the damaged brain. Aims. The main goal of this study was to examine the neurobiological correlates of improved picture-naming performance in 2 aphasic patients who received intensive and specific training for a chronic and severe phonological anomia. Methods. In both participants, picture-naming performance was assessed before and after phonological cueing training. Training-induced changes in patients’ performance were correlated to brain activity patterns as revealed by pre- and posttraining event-related functional magnetic resonance imaging scanning. Results. Training-induced improvement was observed concurrently with changes in the brain activation patterns. Better performance was observed in the patient with the smaller lesion, partially sparing Broca’s area, who showed a left perilesional reactivation. Conversely, the patient with complete destruction of Broca’s area showed a posttraining activation in the right mirror frontal region. Conclusions. The results show that, even in the chronic stage, phonological strategies may improve impaired naming and induce cerebral reorganization. Key Words: Aphasia recovery—Speech rehabilitation—fMRI— Phonological training—Brain reorganization—Anomia. ur knowledge of how the brain recovers after an acute stroke, and how different types of intervention can affect functional recovery, remains limited. Spontaneous recovery after brain damage has generally been ascribed to the functional reactivation O From the Vita-Salute University and San Raffaele Scientific Institute, Milano, Italy (PV, JA, MT, MD, DP, SFC); Université de Montréal and Institut Universitaire de Gériatrie de Montréal, Québec, Canada (PV, A-IA, YJ); Interdisciplinary Center for Cognitive Studies, University of Potsdam, Germany (JA); Institute of Bioimaging and Molecular Physiology-CNR, Milano, Italy (DP, SFC). Address correspondence to Paolo Vitali, Centre de Recherche - Institut Universitaire de Gériatrie de Montréal, 4565, ch. Reine-Marie, Montréal, Québec, Canada H3W 1W5. E-mail: paolo.vitali@umontreal.ca. Vitali P, Abutalebi J, Tettamanti M, Danna M, Ansaldo A-I, Perani D, Joanette Y, Cappa SF. Training-induced brain remapping in chronic aphasia: a pilot study. Neurorehabil Neural Repair 2007;21:152–160. DOI: 10.1177/1545968306294735 152 and reorganization of the brain. With regard to language disorders following left hemispheric damage, several studies have demonstrated the fundamental role of perilesional areas in efficient and long-term recovery.1-5 Other investigations have also highlighted the capacity of the right hemisphere (RH) to take over impaired linguistic abilities, especially auditory comprehension, after damage to the dominant left hemisphere (LH).6-10 There is increasing evidence, in particular from the field of recovery of motor functions, that specific training can modulate the cortical reorganization resulting in an improved clinical outcome.11 However, the neural correlates of training-induced improvements of cognitive functions after brain damage remain still scarcely understood. In the specific case of aphasia, although several investigations have addressed the issue of the neural substrates of functional recovery,12-14 only a few studies have attempted to assess the impact of language training on the damaged brain.15-20 An important role of the RH is suggested by the study of Musso16 who investigated with positron emission tomography (PET) the neural correlates of intensive verbal comprehension training in a group of aphasics. Posttraining comprehension performances were positively correlated with the pattern of regional cerebral blood flow in the right homologues of Wernicke’s area and of Broca’s area. Moreover, Blasi17 found that the learning of a stem-completion task was associated with specific response decrements in the right frontal and occipital cortex, rather than in the left-sided network engaged by normal subjects. Training-induced effects were prevalent in the RH also in a recent study of “intentional” therapy.20 Other investigations, however, suggest that the engagement of spared left hemispheric regions is also crucial. An influential PET study by Belin15 involved patients with chronic nonfluent aphasia, who had shown considerable improvement in repeating words after the introduction of additional rehabilitation training with melodic intonation therapy (MIT). Improvement in word repetition with MIT intonation was concomitant with an RH deactivation and a significant increase of left frontal areas. The authors argued that the right-sided activations mediating defective word repetition with natural intonation might reflect a “maladaptive” functional reorganization, owing Copyright © 2007 The American Society of Neurorehabilitation Downloaded from nnr.sagepub.com at CARLETON UNIV on July 12, 2015 Training-Induced Brain Remapping in Chronic Aphasia Table 1. Demographic and Clinical Information on the Anomic Patients Patient Patient 1 Patient 2 Age (Years) & Handedness 24 right 52 right Lesion Type Lesion Site Closed head injury Left prefrontal cortex Left fronto-parieto-occipital white matter Territory of left middle cerebral artery, involving fronto-temporo-parietal areas and extending to the temporo-parieto-occipital junction Ischemic stroke to the presence of the left lesion itself, whereas actual recovery mediated by MIT training might be associated with the reactivation of left-hemispheric undamaged structures. Support for this hypothesis comes also from a recent study of the effects of repetitive transcranial stimulation to the RH.21 Interference with right-sided Broca’s area resulted in improved picture naming in 3 patients with chronic aphasia. Léger18 compared the results of functional magnetic resonance imaging (fMRI) during a naming task pre- and postrehabilitation in a patient with prominent phonological errors in speech production. The main difference between the 2 studies was a reactivation of perilesional left hemispheric areas, in particular, the Broca’s region and the supramarginal gyrus, in the posttreatment study. None of these studies dealt with a direct correlation between brain activity and single behavioral responses, and thus the evaluation of the impact of language rehabilitation on recovery was only partial. Our study represents the first use of a single-case, event-related (er-) fMRI approach to examine the functional correlates of training-induced improvement in overt picture naming in aphasic patients. The event-related paradigm is the only type of fMRI design that allows one to relate behavioral performance (ie, hit or defective overt retrieval) to the pattern of brain activation. Impaired word retrieval (anomia) is one of the most prominent and widespread aphasic deficits and is often the focus of rehabilitation. Most often, the therapy consists of providing phonological cues to facilitate word retrieval. Phonological cueing has been suggested to be effective especially in patients with defective retrieval of the word form (ie, phonological anomia).22 We, thus, assessed picture-naming performance before and after phonological cueing training in 2 participants with chronic aphasia who presented severe phonological anomia. Training-induced changes in patients’ performance were correlated to brain activity patterns as revealed by pre- and posttraining er-fMRI scanning. Given the heterogeneity between aphasic individuals with anomia, emphasis was placed on singlesubject analysis of the results because training-elicited mechanisms are not necessarily uniform among anomic patients.23 The advantage of a multiple single-case Months Postonset 12 48 longitudinal design is that individual confounds can be kept under control because each participant becomes his or her own control. In this respect, our hypothesis was that traininginduced improvement following a specific approach would be associated with activations in the areas surrounding the lesion or in the RH areas homologous to the damaged LH regions.12 Given that there is a general agreement concerning the crucial role of Broca’s area in lexical retrieval,24 the effects of the training on the activation of Broca’s area were specifically examined. METHODS Participants Two chronic aphasic individuals (patient 1 and patient 2) with extensive LH lesions participated in the study. Their demographic and clinical information is summarized in Table 1. The 2 participants gave their written informed consent before entering the study. This research was approved by the local ethics committee. Patient 1 Patient 1, a 24-year-old right-handed man, suffered a closed head injury (CHI) 1 year before being enrolled in this study, resulting in right hemiparesis and aphasia. One year after the CHI, and after several months of global and unspecific language therapy, the patient’s speech was still nonfluent and effortful, characterized by speech apraxia, agrammatism, and telegraphic language. Extensive linguistic assessment conducted with a standardized aphasia protocol25 at the time of the present study revealed a severe impairment of word repetition with relatively preserved comprehension abilities. The most prominent language impairment was a severe naming deficit, characterized by the difficulty in retrieving word forms. Patient 2 Patient 2, a 53-year-old right-handed man, suffered from an extensive LH stroke, which caused a right-sided Neurorehabilitation and Neural Repair 21(2);from2007 Downloaded nnr.sagepub.com at CARLETON UNIV on July 12, 2015 153 Vitali et al hemiplegia and global aphasia, 4 years before being enrolled in the study. Patient 2 never received structured speech therapy during the 4 years that preceded the present study, even if some improvement in his language abilities, especially verbal comprehension, occurred. When patient 2 was enrolled in the present study, he still presented with a right-sided hemiparesis and a language deficit characterized by nonfluent speech, agrammatism, good word repetition, and preserved verbal comprehension, as revealed by an extensive linguistic assessment conducted with a standardized aphasia protocol.25 Once again, the main language impairment was a severe anomia due to the difficulty in retriving word forms. Experimental Procedure The experimental pictures were selected on an individual basis from a series of standardized pictures of concrete objects that the patients were asked to name when enrolled in the present study.26 A number of stimuli that patients could not name were selected from this data set. Half of the selected pictures were used as training material for anomia training (training set), whereas the other half were employed as an untrained control condition (control set). Items were matched for word length, number of syllables, and frequency in Italian.27 Moreover, we selected a set of pictures that the patients could name either spontaneously or after being phonologically cued with the first syllable of the target word (spontaneous training). Training consisted of repeatedly cueing patients with the initial syllable of the target word, and subsequently adding the following syllables, until the correct answer was produced. Phonological training started after acquisition of a first pretraining er-fMRI session and was administered daily in 1-hour sessions to both patients until the participants achieved a score of 50% on picture naming with the training set. Specifically, this required 8 weeks of training for patient 1 and 4 weeks of training for patient 2, after which the second, posttraining er-fMRI was performed. During er-fMRI acquisitions, all the pictures from the 3 experimental sets (spontaneous, training, control) were visually presented for 4500 ms and the patients were asked to name them aloud. Recordings of the oral responses made by participants permitted us to monitor their performances during scanning. Overt naming attempts were recorded using a nonmagnetic microphone placed within the scanner head coil about 1 cm in front of the participant’s mouth, connected via an optical cable to a laptop computer situated outside the scanner room. Vocal responses were recorded in .wav format during each fMRI run and were scored off-line. 154 Er-fMRI Acquisition and Data Analysis Functional images (fifteen 6-mm-thick slice volumes, FOV = 280 × 280 mm, matrix 64 × 64) were acquired with a 1.5-Tesla General Electric Signa Horizon System (GE Healthcare, USA) using a gradient echo echoplanar pulse sequence (TE = 60 ms, TR = 2000 ms). A standard T1-weighted sequence (1 × 1 × 1 mm3 voxel size, matrix 256 × 256) was also used to acquire high-resolution anatomical images of patients’ brains. fMRI data were analyzed using a conventional statistical parametric mapping analysis (SPM-99, Wellcome Department of Imaging Neuroscience, London, UK).28 Functional images were realigned and subsequently spatially normalized with respect to the MNI brain template,29 by using the lesion-masking MATLAB routine devised by Brett and colleagues.30 Finally, normalized images were spatially smoothed with an 8-mm full width at half-maximum isotropic Gaussian kernel. For each patient, a single statistical analysis combining the data from the first and second er-fMRI sessions was performed. Effects of interest (successful naming, sequences of phonemic approximations,31 and phonological/semantic errors) were independently modeled in the statistical design along with effects of no interest (anomias, that is, no overt response to a given item). These effects of no interest were modeled to covary out confounding effects and were not considered in the analysis. In the pretraining er-fMRI session, simple main effects of errors (that is, sequences of phonemic approximations, phonological and/or semantic errors) and successful naming were calculated. In the posttraining er-fMRI session, we assessed the anatomic-functional correlates of the training-induced improvement by comparing brain activations associated with successfully named items of the training set and brain activations associated with successfully named items of the spontaneous set. RESULTS Task Performance The aphasic patients’ performance before and after training is summarized in Table 2. Table 3 reports the percentage of correct responses before and after training on the training, control, and spontaneous sets. In the case of patient 1, a dramatic decrease in the number of anomias following training was observed during the second fMRI scanning session, concurrently Neurorehabilitation and Neural Repair 21(2); 2007 Downloaded from nnr.sagepub.com at CARLETON UNIV on July 12, 2015 Training-Induced Brain Remapping in Chronic Aphasia Table 2. Patients’ Pre- and Posttraining Behavioral Data Patient Naming on Snodgrass and Vanderwart Test (%) Anomias 1st fMRI (%) Anomias 2nd fMRI (%) Correct Naming 1st fMRI (%) Correct Naming 2nd fMRI (%) Errors 1st fMRI (%) Errors 2nd fMRI (%) Patient 1 Patient 2 0.38 45.52 91.4 12.0 50.0 42.3 1.7 47.8 31.0 48.9 6.9 40.2 19.0 8.8 Table 3. Patients’ Pre- and Posttraining Performance on the 3 Experimental Sets Correct 1st er-fMRI (%) Correct 2nd er-fMRI (%) Delta % Patient T C S T C S T C S Patient 1 0.0 0.0 5.5 75.0 0.0 16.7 +0.0 Patient 2 23.3 40.0 78.1 60.0 26.7 59.4 +75.0 χ2 = 30** +36.7 χ2 = 8* +11.2 n.s. –18.7 n.s. –13.3 n.s. T = training picture set; C = control picture set; S = spontaneous picture set; Delta % = difference in performance (%) between the 1st and 2nd er-fMRI sessions; n.s. = nonsignificant. *Significant at P < .01. **Significant at P < .001. with a behavioral improvement on picture naming. In patient 2, phonological training resulted in a significant reduction in errors during the second fMRI scanning session, whereas an increase in the number of anomias was observed. On the other hand, in the latter patient, correct naming performance was stable as assessed behaviorally. However, after intensive training, both patients showed a significant naming improvement on the training picture set, which generalized neither to the control set nor to the spontaneous set. Functional Magnetic Resonance Imaging Results Brain Activity Associated With Errors The neural activity corresponding to errors in lexical retrieval was assessed only during the first er-fMRI session. The goal was to describe the neural substrate related to the production of errors observed before training. Patient 1’s sequences of phonemic approximations and semantic/phonological errors were associated with activity in the left inferior frontal gyrus (IFG) (see Table 4 and Figure 1A). The activation of the IFG during the production of semantic/phonological errors and phonemic approximations was also observed in patient 2, but in this case in the RH alone and together with the right precentral gyrus (PreCG), the right middle frontal gyrus (MFG), and the left hippocampus. Patient 2 also presented bilateral activations in the cingulate cortex (see Table 4 and Figure 2A). Brain Activity Associated With Correct Pretraining Responses Due to the insufficient number of successful naming events (n = 1), spontaneous naming by patient 1 was not considered. Picture naming by patient 2 was associated with activations in the right PreCG, the right IFG, the right MFG, and the right insula. The left precuneus was also activated (see Table 5 and Figure 2B). Brain Activity Associated With Training-Induced Performance Improvement Both patients activated after training the supramarginal gyrus (SMG) in the left inferior parietal lobule (IPL), and the pars triangularis of the IFG (patient 1 in the LH, but patient 2 in the RH). Furthermore, similar activations were observed in the precuneus (on the left for patient 1, and on the right for patient 2) and in the right posterior cingulate cortex. Patient 1 also showed an activation of the anterior portion of the left insula, the left middle temporal gyrus, the left temporo-occipital junction, and the right hippocampal gyrus. On the other hand, patient 2 showed an activation of the left MFG and the left superior frontal gyrus (see Table 6 and Figures 1B and 2C). DISCUSSION The main goal of this study was to examine, in 2 cases of severe and chronic aphasia, the neurobiological Neurorehabilitation and Neural Repair 21(2);from2007 Downloaded nnr.sagepub.com at CARLETON UNIV on July 12, 2015 155 Vitali et al Table 4. Stereotactic Coordinates and Z Scores for Brain Areas Associated With Pretraining Errors in Patient 1 and Patient 2 Brain Regions (Brodmann Area) Patient 1 Left inferior frontal gyrus (BA 47) Left inferior frontal gyrus (BA 45) Patient 2 Left anterior cingulate cortex (BA 32) Left hippocampus Right precentral gyrus (BA 6) Right inferior frontal gyrus (BA 47) Right middle frontal gyrus (BA 11) Right cingulate cortex (BA 24) X Y Z Z Statistic –56 –62 +16 +20 –12 +4 3.26 2.93 –12 –20 +38 +50 +44 +20 +0 +12 –22 +2 +26 +40 +48 +34 +40 –16 +28 –16 –12 –16 –8 3.77 3.87 3.49 3.58 3.36 3.25 3.93 BA = Brodmann area. Table 5. Stereotactic Coordinates and Z Scores for Brain Areas Associated With Correct Pretraining Responses (ie, spontaneous naming) by Patient 2 Brain Regions (Brodmann Area) X Y Z Z Statistic Patient 2 Left precuneus (BA 7) Right precentral gyrus (BA 6) Right inferior frontal gyrus (BA 47) Right middle frontal gyrus (BA 10) Right insula –14 +60 +54 +26 +46 –60 +2 +22 +50 –4 44 +24 –12 0 +8 3.76 3.79 3.57 4.13 3.72 BA = Brodmann area. correlates of improved picture-naming performance following intensive phonological training. We studied 2 aphasic patients characterized by different behavioral and neuropathological profiles. Despite the variation, both patients benefited from the phonological training. The functional results indicated that 2 different processes contributed to the naming improvement, namely, the restitution of function in crucial areas surrounding the lesion, and the contralateral compensation in homologous regions, suggesting that specific anomia training can lead to cerebral reorganization in chronic and severe aphasia. As expected, both patients improved naming on the training set, without generalization to the untreated items. This is in agreement with several investigations reporting that treatments for word-finding impairments produce clear effects on the treated items, whereas generalization to untreated items is often absent, or—when occurring— less robust than the item-specific effects.32 However, the cases were not comparable with respect to the overall performances in the task or to the brain activity patterns. Given that the patients were different in age, etiology of brain damage, and lesion size, these differences might explain the different courses and the different activation patterns. Each patient is thus discussed separately. 156 Patient 1 Naming improvement after training was greater in the younger patient (patient 1) with the smaller lesion due to CHI, partially sparing Broca’s area. In the case of CHI lesions, spontaneous recovery effects may extend until 1 year postonset, and the overall prognosis is related to age.23 These favorable prognostic factors might have contributed to naming recovery in patient 1. Nevertheless, our findings showed that phonological training was specifically effective, because significant posttraining improvement in naming was noted on the trained picture set only. This result underlines that, in some cases, phonological strategies could be useful to improve impaired picture naming even in the chronic stage. From the standpoint of brain activity, this patient showed a left perilesional reactivation following the training, involving the pars triangularis of Broca’s area. The role of Broca’s area in phonological processing is well-known.33 The preservation of Broca’s area in patient 1 may have enhanced the impact of phonological training. The activation of the left IFG during error production paralleled the left IFG activation during training-induced correct naming and may be similarly related to demands for lexical selection and retrieval.34 Neurorehabilitation and Neural Repair 21(2); 2007 Downloaded from nnr.sagepub.com at CARLETON UNIV on July 12, 2015 Training-Induced Brain Remapping in Chronic Aphasia Table 6. Training-Induced Patterns of Brain Activity in Patient 1 and Patient 2 Brain Regions (Brodmann Area) Patient 1 Left inferior frontal gyrus (BA 45/46) Left inferior frontal gyrus (BA 45) Left insula Left middle temporal gyrus (BA 21) Left temporo-occipital junction (BA 39/19) Left inferior parietal lobule (BA 40) Left precuneus (BA 7) Right posterior cingulate cortex (BA 30) Right hippocampal gyrus Patient 2 Left middle frontal gyrus (BA 9) Left middle frontal gyrus (BA 10) Left superior frontal gyrus (BA 9) Left superior frontal cortex (BA 10) Left inferior parietal lobule (BA 40) Right inferior frontal gyrus (BA 45) Right precuneus (BA 7) Right posterior cingulate cortex (BA 31) X Y Z Z Statistic –54 –42 –34 –34 –58 –46 –42 –36 –40 –56 –46 –18 +10 +12 +26 +34 +22 +4 +4 –68 –36 –80 –78 –76 –48 –40 –64 –46 –40 –46 +4 +12 0 +4 +4 0 16 16 12 +36 +36 44 +12 0 –16 2.88 2.78 3.53 3.09 3.31 2.71 2.81 2.73 2.44 2.74 2.54 3.42 3.77 2.81 2.68 –44 –34 –6 –18 –56 –56 +40 +2 +12 +18 +16 +40 +50 +52 +50 –46 –46 +24 –60 –66 –42 –44 +32 –4 +36 –8 +28 +32 +16 +52 +60 + 44 +32 3.02 2.86 2.65 2.93 3.14 2.65 3.56 3.53 2.53 3.25 2.90 BA = Brodmann area. Patient 1 also displayed training-induced modifications in the SMG portion of the left IPL. This area is another crucial component of the phonological loop of verbal working memory, in particular for phonological storage.35 It is thus likely that this patient adopted phonological compensatory strategies for successful lexical retrieval (see Léger18 for a similar interpretation). Additionally, activation of the right hippocampal region was found, which may reflect the memory consolidation processes for lexical items.36 Therefore, the present results are in agreement with the hypothesis that the word-learning system as a whole might play a role in anomia rehabilitation.37 Patient 2 Naming performance during the posttraining fMRI session was more complex for patient 2. The improvement was also significant on the training picture set, but it was less prominent than in patient 1 and accompanied by an increase in anomic events along with a dramatic reduction in errors. A number of factors can be evoked to explain the less prominent improvement after training in patient 2. He was older than patient 1 and had a larger lesion of ischemic origin. Furthermore, his training started 3 years later and was shorter. Additionally, patient 2 was already much better at picture naming than patient 1 during the pretraining session, and, thus, the impact of training might have been reduced by a ceiling effect. In patient 2, Broca’s area was completely destroyed by the ischemic lesion. Pre- and posttraining RH activations suggest a permanent functional impairment of the LH. It is possible, thus, that the damage to Broca’s area may have reduced the impact of phonological training; this approach seems more suitable for tapping into the LH’s phonological processing abilities. We observed an activation of the “mirror” homologous right Broca’s area after training, which may indicate compensatory activation of RH areas.2,6 Therefore, it may be that in patient 2 the right homologous Broca’s area mediated successful, though nonoptimal, lexical retrieval. This finding supports the hypothesis that after LH damage, preexisting language areas in the RH are reactivated and may play an important role in aphasia Neurorehabilitation and Neural Repair 21(2);from2007 Downloaded nnr.sagepub.com at CARLETON UNIV on July 12, 2015 157 Vitali et al Figure 1. Patient 1’s cortical patterns of regional activity during picture naming. The statistical images are thresholded at P level = .01 (uncorrected) to verify our a priori hypothesis regarding the involvement of Broca’s area and its right homologue in lexical retrieval and in training-induced improvement. Results are superimposed onto transversal slices of the patient’s T1*weighted normalized brain image. A) Cerebral activation associated with failures in lexical retrieval during the pretraining fMRI session. B) Cerebral activation underlying training-induced improvement in naming performance. recovery.3,6,8 One alternative explanation, partially applying to the brain changes observed in patient 1 too, is that the increased IFG activation after training, particularly because it was associated with increased anomias in patient 2, could also reflect increased articulation or increased effort, similar to the increased activity in the right homologous Broca’s area (in Broca’s proper for patient 1 during semantic/phonological errors and phonemic approximations. This interpretation is not fully supported by behavioral observations in the 2 patients, whose oral production at picture naming, after training, was clearly faster and less effortful on the trained picture set. It is also possible that the posttraining activation of the right IFG in patient 2 was linked not only to a lesser degree of training-induced improvement but also to a change in retrieval strategy, which might account for the lower number of errors and the larger number of anomias in this participant. Thus, the training approach may have favored a change from a lexical-semantic to a phonological strategy during picture naming. The latter strategy forces the patient to search for and utter only the correct item and to discard alternatives (ie, errors). Given the reduced phonological competence of the RH, patient 2 was not always successful in accessing the target, and thus an anomic event occurred. Patient 2 also offers interesting insights into the specific patterns of activation associated with errors in naming. Indeed, the event-related methodology employed in this study detected the pattern of brain activity related to erroneous naming. The strong bilateral activation of the anterior cingulate cortex might reflect cognitive control and attentional processing. One of the most prominent roles of the anterior cingulate is detecting and signaling the occurrence of conflicts in 158 Figure 2. Patient 2’s cortical patterns of regional activity during picture naming. The statistical images are thresholded at P level = .001 (uncorrected) for (A) and (B), and at P level = .01 for (C) to verify our a priori hypothesis regarding the involvement of Broca’s area and its right homologue in lexical retrieval and in training-induced improvement. Results are superimposed onto transversal slices of the patient’s T1*weighted normalized brain image. A) Cerebral activation associated with failures in lexical retrieval during the pretraining fMRI session. B) Cerebral activation associated with the pretraining spontaneous naming. C) Cerebral activation underlying traininginduced improvement in naming performance. information processing and, more specifically, to develop an error-likelihood signal.38 Spontaneous naming was associated with an almost exclusively right-sided brain activation, engaging the RH homologue regions of the LH regions involved in word-retrieval tasks in the healthy population. The MFG is involved in a multitude of cognitive functions, including action and speech initiation and executive and word-retrieval processes. The activation of the right PreCG and the right insula may reflect speech production and articulatory planning, respectively, that is, those articulatory movements associated with the production of word forms.33 The study has a number of limitations that must be considered in the interpretation of the results. In particular, because repeated fMRI studies were not done in the absence of treatment, it is unclear whether the functional changes were due to the treatment itself or attributable to repeating the study after practice of the task.39 Moreover, recent studies of motor and speech recovery indicate that some of the activations (particularly in the hemisphere contralateral to the lesion) observed in poststroke recovery may not reflect activity that is important to the task, but rather “maladaptive” activation unrelated to functional performance. In fact, inhibition of right hemisphere areas with repetitive transcranial magnetic stimulation can result in task improvement.21 It must be, however, underlined that our findings in patient 2 seem not to support this alternative view, as the enhanced Neurorehabilitation and Neural Repair 21(2); 2007 Downloaded from nnr.sagepub.com at CARLETON UNIV on July 12, 2015 Training-Induced Brain Remapping in Chronic Aphasia engagement of right-sided brain regions observed in this patient after training paralleled the behavioral improvement. In addition, our results are in strong agreement with previously reported results regarding recovery mechanisms in aphasic patients1,6,8,16,18 and in line with longitudinal studies indicating that the damaged brain can show plastic changes even several years postonset.40 On a general note, we think that multiple singlesubject er-fMRI study is a promising method to investigate the recovery process taking place in individual subjects, which should be tested in patients using other training approaches. ACKNOWLEDGMENTS This work was partially supported by a Human Frontier Science Program Grant (RG148/2000) to Stefano Cappa; by the 5th FW EEC project “Connectivity in language, Rehabilitation in stroke in Europe” (LSDE) (Daniela Perani and Paolo Vitali); by the FIRB project “Cerebral plasticity” to Daniela Perani (RBNE01AZ92); by the Italy/Quebec Scientific and Technological Cooperation (Yves Joanette, Stefano Cappa, and Paolo Vitali); and by a Health and Stroke Foundation of Canada grant to Yves Joanette. The sponsors had no role in study design or analysis. We wish to thank Dr Lorella Algeri for the evaluation and referral of patient 1, and Dr Zofia Laubitz for revising the manuscript for fluency in English. The authors declare having no competing interest. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. Heiss WD, Kessler J, Thiel A, Ghaemi M, Karbe H. Differential capacity of left and right hemispheric areas for compensation of poststroke aphasia. Ann Neurol. 1999;45:430-438. Karbe H, Thiel A, Weber-Luxenburger G, Herholz K, Kessler J, Heiss WD. Brain plasticity in poststroke aphasia: what is the contribution of the right hemisphere? Brain Lang. 1998;64:215-230. Cao Y, Vikingstad EM, George KP, Johnson AF, Welch KM. Cortical language activation in stroke patients recovering from aphasia with functional MRI. Stroke. 1999;30:2331-2340. Warburton E, Price CJ, Swinburn K, Wise RJ. Mechanisms of recovery from aphasia: evidence from positron emission tomography studies. J Neurol Neurosurg Psychiatry. 1999;66:155-161. Calvert GA, Brammer MJ, Morris RG, Williams SC, King N, Matthews PM. Using fMRI to study recovery from acquired dysphasia. Brain Lang. 2000;71:391-399. Weiller C, Isensee C, Rijntjes M, et al. Recovery from Wernicke’s aphasia: a positron emission tomographic study. Ann Neurol. 1995;37:723-732. Cappa SF, Perani D, Grassi F, et al. A PET follow-up study of recovery after stroke in acute aphasics. Brain Lang. 1997;56:55-67. Thulborn KR, Carpenter PA, Just MA. Plasticity of languagerelated brain function during recovery from stroke. Stroke. 1999;30:749-754. Rosen HJ, Petersen SE, Linenweber MR, et al. Neural correlates of recovery from aphasia after damage to left inferior frontal cortex. Neurology. 2000;55:1883-1894. 10. Price CJ, Crinion J. The latest on functional imaging studies of aphasic stroke. Curr Opin Neurol. 2005;18:429-434. 11. Taub E, Uswatte G, Elbert T. New treatments in neurorehabilitation founded on basic research. Nat Rev Neurosci. 2002;3:228-236. 12. Cappa SF. Recovery from aphasia: why and how? Brain Lang. 2000;71:39-41. 13. Thompson CK. The neurobiology of language recovery in aphasia. Brain Lang. 2000;71:245-248. 14. Heiss WD, Thiel A, Kessler J, Herholz K. Disturbance and recovery of language function: correlates in PET activation studies. Neuroimage. 2003;20(Suppl 1):S42-S49. 15. Belin P, Van Eeckhout P, Zilbovicius M, et al. Recovery from nonfluent aphasia after melodic intonation therapy: a PET study. Neurology. 1996;47:1504-1511. 16. Musso M, Weiller C, Kiebel S, Muller SP, Bulau P, Rijntjes M. Training-induced brain plasticity in aphasia. Brain. 1999;122(Pt 9):1781-1790. 17. Blasi V, Young AC, Tansy AP, Petersen SE, Snyder AZ, Corbetta M. Word retrieval learning modulates right frontal cortex in patients with left frontal damage. Neuron. 2002;36:159-170. 18. Léger A, Demonet JF, Ruff S, et al. Neural substrates of spoken language rehabilitation in an aphasic patient: an fMRI study. Neuroimage. 2002;17:174-183. 19. Peck KK, Moore AB, Crosson BA, et al. Functional magnetic resonance imaging before and after aphasia therapy: shifts in hemodynamic time to peak during an overt language task. Stroke. 2004;35:554-559. 20. Crosson B, Moore AB, Gopinath K, et al. Role of the right and left hemispheres in recovery of function during treatment of intention in aphasia. J Cogn Neurosci. 2005;17:392-406. 21. Naeser MA, Martin PI, Nicholas M, et al. Improved picture naming in chronic aphasia after TMS to part of right Broca’s area: an open-protocol study. Brain Lang. 2005;93:95-105. 22. Le Dorze G, Pitts C. A case study of the effects of different techniques for the treatment of anomia. Neuropsychol Rehabil. 1995;5:51-65. 23. Basso A. Prognostic factors in aphasia. Aphasiology. 1992;6:337-348. 24. Cappa SF, Perani D. Broca’s area and lexical-semantic processing. In: Grodzinsky Y, Amunts K, eds. Broca’s Region. New York: Oxford University Press; 2006. 25. Miceli G, Laudanna A, Burani C, Capasso R. BADA: batteria per l’analisi dei deficit afasici. Roma: CEPSAG, Università Cattolica del Sacro Cuore; 1994. 26. Snodgrass JG, Vanderwart M. A standardized set of 260 pictures: norms for name agreement, image agreement, familiarity, and visual complexity. J Exp Psychol [Hum Learn]. 1980;6:174-215. 27. De Mauro T, Mancini F, Vedovelli M, Voghera M. Lessico di frequenza dell’italiano parlato. Roma: Etaslibri; 1993. 28. Friston KJ, Holmes AP, Worsley KJ, Poline JB, Frith CD, Frackowiak RSJ. Statistical parametric maps in functional imaging: a general linear approach. Hum Brain Mapp. 1995;2:189-210. 29. Friston KJ, Ashburner J, Poline J-B, Frith CD, Frackowiak RSJ. Spatial registration and normalization of images. Hum Brain Mapp. 1995;2:165-189. 30. Brett M, Leff AP, Rorden C, Ashburner J. Spatial normalization of brain images with focal lesions using cost function masking. Neuroimage. 2001;14:486-500. 31. Joanette Y, Keller E, Lecours AR. Sequences of phonemic approximations in aphasia. Brain Lang. 1980;11:30-44. 32. Nickels LA, Best W. Therapy for naming deficits (part II): specifics, surprises and suggestions. Aphasiology. 1996;10:109-136. 33. Demonet JF, Thierry G, Cardebat D. Renewal of the neurophysiology of language: functional neuroimaging. Physiol Rev. 2005;85:49-95. 34. Thompson-Schill SL, D’Esposito M, Aguirre GK, Farah MJ. Role of left inferior prefrontal cortex in retrieval of semantic knowledge: a reevaluation. Proc Natl Acad Sci U S A. 1997;94:14792-14797. Neurorehabilitation and Neural Repair 21(2);from2007 Downloaded nnr.sagepub.com at CARLETON UNIV on July 12, 2015 159 Vitali et al 35. Paulesu E, Frith CD, Frackowiak RS. The neural correlates of the verbal component of working memory. Nature. 1993;362: 342-345. 36. Raboyeau G, Marie N, Balduyck S, Gros H, Demonet JF, Cardebat D. Lexical learning of the English language: a PET study in healthy French subjects. Neuroimage. 2004;22:1808-1818. 37. Cornelissen K, Laine M, Tarkiainen A, Jarvensivu T, Martin N, Salmelin R. Adult brain plasticity elicited by anomia treatment. J Cogn Neurosci. 2003;15:444-461. 160 38. Carter CS, Braver TS, Barch DM, Botvinick MM, Noll D, Cohen JD. Anterior cingulate cortex, error detection, and the online monitoring of performance. Science. 1998;280:747-749. 39. Cardebat D, Demonet JF, De Boissezon X, et al. Behavioral and neurofunctional changes over time in healthy and aphasic subjects: a PET Language Activation Study. Stroke. 2003;34:2900-2906. 40. Ansaldo AI, Arguin M, Lecours AR. Recovery from aphasia: a longitudinal study on language recovery, lateralization patterns, and attentional resources. J Clin Exp Neuropsychol. 2004;26:621-627. Neurorehabilitation and Neural Repair 21(2); 2007 Downloaded from nnr.sagepub.com at CARLETON UNIV on July 12, 2015