Original Paper Received: March 16, 2017 Accepted: May 1, 2017 Published online: May 25, 2017 Eur Neurol 2017;78:15–21 DOI: 10.1159/000477167 Transcortical Sensory Aphasia after Left Frontal Lobe Infarction: Loss of Functional Connectivity Miseon Kwon a Woo Hyun Shim b Sang-Joon Kim b Jong S. Kim a Departments of a Neurology and b Radiology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, South Korea Abstract Background: The underlying mechanism of transcortical sensory aphasia (TSA) caused by lesions occurring in the left frontal lobe remains unclear. We attempted to investigate the mechanism with the use of functional MRI (fMRI). Methods: We studied 2 patients with TSA after a left frontal infarction identified by diffusion-weighted MRI. As control subjects, a patient with transcortical motor aphasia and a healthy normal adult were chosen. The Korean version of Western Aphasia Battery was performed initially and at 3 months post stroke. We performed fMRI using verb generation and sentence completion tasks. Resting-state fMRI (rs-fMRI) was also obtained for network-level analysis initially and at 3 months post stroke. Results: The results of diffusion- and perfusion-weighted MRI revealed no diffusion– perfusion mismatch. Initial fMRI in patients with TSA showed no reversed inter-/intrahemispheric activation patterns. rsfMRI showed significantly decreased resting-state functional connectivity in the language network in patients with TSA compared with the control subjects. Follow-up rs-fMRI studies showed improvement in functional connectivity along with the recovery of patients’ language function. Conclusion: Our data showed that the auditory comprehension deficits in patients with frontal lobe infarcts is attributed to © 2017 S. Karger AG, Basel E-Mail karger@karger.com www.karger.com/ene difficulty accessing the posterior language area due to functional disconnection between language centers in the acute stage of stroke. © 2017 S. Karger AG, Basel Introduction Transcortical sensory aphasia (TSA), characterized by impairment of auditory comprehension with preserved speech fluency and repetition, is usually attributed to temporal or parietal lesions [1]. Rarely, patients with left frontal lobe lesions presenting with TSA have been reported [2–6], but the underlying mechanisms remain unclear. We studied 2 patients with TSA due to left frontal lobe lesions using multimodal MRIs to clarify the mechanism. The results were compared with those of a patient with transcortical motor aphasia (TMA) caused by left frontal infarction and of a healthy, normal subject. Materials and Methods Patients We studied 2 patients (TSA1 and TSA2) who presented with TSA due to acute ischemic stroke and were admitted to the Asan Medical Center, Seoul, South Korea. TSA1 was a right-handed, 65-year-old man with 12 years of education. He was a current cigarette smoker and had hyperlipidemia. On the day of stroke Jong S. Kim, MD, PhD Department of Neurology Asan Medical Center, University of Ulsan College of Medicine 388-1, Poongnap-dong, Songpa-gu, Seoul 138-736 (South Korea) E-Mail jongskim @ amc.seoul.kr Downloaded by: University of Florida, Gainesville and Jacksonville 128.227.24.141 - 10/20/2017 12:20:47 PM Keywords Functional MRI · Frontal lobe · Transcortical aphasia TSA1 Fig. 1. Diffusion (first 3 figures in upper rows) and perfusion (first 3 figures in lower rows) MRIs showing an acute infarction in the left frontal areas without diffusion–perfusion mismatches. TSA2 had a small additional infarct in the left parietal area. Fluid attenuated inversion recovery images (last images in upper rows) showed moderate periventricular and white matter ischemic changes in TSA2 patient. MR angiography showed left carotid artery stenosis (arrows) in TSA1 and TMA, and occlusion of superior branch of the middle cerebral artery in TSA2 patients. TSA1, patient 1 with transcortical sensory aphasia; TSA2, patient 2 with transcortical sensory aphasia; TMA, patient with transcortical motor aphasia. (Figure 1 continued on next page.) onset, his wife found that he produced inappropriate speech such as “I cannot turn on the TV” while he was watching TV, and kept saying things out of context. On neurological examination, he did not have any neurologic deficits other than prominent aphasia. Diffusion-weighted imaging (DWI) MRI per- formed on the day of symptom onset showed a left frontal infarct (Fig. 1, TSA1). MR angiography (MRA) and carotid duplex scan showed moderate (55%) stenosis of the left carotid artery, and artery-to-artery embolism was considered a presumed stroke mechanism. 16 Eur Neurol 2017;78:15–21 DOI: 10.1159/000477167 Kwon/Shim/Kim/Kim Downloaded by: University of Florida, Gainesville and Jacksonville 128.227.24.141 - 10/20/2017 12:20:47 PM TSA2 TMA 1 Table 1. Results of language assessment in patients with aphasia and controls K-WAB Spontaneous speech (20) Fluency (10) Information contents (10) Auditory comprehension (10) Yes/no question (60) Word recognition (60) Sequential commands (80) Repetition (10) Naming (10) Object naming (60) Word fluency (20) Sentence completion (10) Responsive naming (10) AQ total (100) Subjects (sex/age) initial/follow-up test TSA1 (M/65) TSA2 (F/75) TMA (F/74) HN (M/67) 15/18 8/9 7/9 6.85/9.80 39/60 50/59 48/77 9.0/9.9 6.2/9.3 47/60 3/13 6/10 6/10 74.1/94.0 9/15 7/8 2/7 2.90/5.55 42/51 7/41 9/19 8.3/7.7 0.9/5.4 7/42 0/0 2/8 0/4 42.2/67.3 10/– 5/– 5/– 8.60/– 51/– 57/– 64/– 8.1 7.9 56/– 3/– 10/– 10/– 69.2/– 19/– 10/– 9/– 9.2/– 60/– 60/– 64/– 10/– 9.6/– 60/– 16/– 10/– 10/– 95.6/– TSA2 was a 75-year-old woman with 12 years of education, and had had hypertension, diabetes mellitus, and atrial fibrillation. She was reported to have irrelevant speech during a conversation with her son. On admission, she did not have any neurologic deficits except the prominent aphasia. DWI performed on the day of symptom onset showed a left frontal infarct (Fig. 1, TSA2). A small parietal infarct was also observed. MRA showed occlusion of superior branch of the middle cerebral artery, and there was no significant stenosis in the carotid artery. Cardiogenic embolism was considered a presumed mechanism of stroke. As controls, a patient with TMA and a healthy normal (HN) adult were recruited (Table 1). TMA was a 74-year-old, hyper- fMRI in Transcortical Sensory Aphasia Eur Neurol 2017;78:15–21 DOI: 10.1159/000477167 17 Downloaded by: University of Florida, Gainesville and Jacksonville 128.227.24.141 - 10/20/2017 12:20:47 PM K-WAB, Korean version of Western Aphasia Battery; AQ, aphasia quotient; TSA1, patient 1 with transcortical sensory aphasia; TSA2, patient 2 with transcortical sensory aphasia; TMA, patient with transcortical motor aphasia; HN, healthy normal adult; NA, not applicable. Language Tests Korean version of the Western Aphasia Battery (K-WAB) was performed 2–5 days after the stroke onset, and also at 3 months post stroke. Functional MRI Studies We performed functional MRI (fMRI) using verb generation and sentence completion tasks to assess the activation pattern during language processing for HN and the patients with aphasia within 1–4 days after the initial K-WAB test. Resting-state fMRI (rs-fMRI) was also obtained for network-level analysis. For TSA patients, rs-fMRI was repeated on the day of follow-up aphasia testing at 3 months post stroke. MRI data were acquired on a 3T system (Achieva; Philips Medical Systems, The Netherlands). A high-resolution 3D-T1 image was obtained with the following parameters: TR/TE, 9.9/4.6 ms; flip angle, 8°; field of view, 224 mm; matrix, 448 × 448; and slice thickness, 1 mm with no gap. The stimulus-induced fMRI and rs-fMRI were measured by blood oxygen level-dependent contrast image with following parameters: TR/TE, 3,000/35 ms; flip angle, 90°; field of view, 224 mm; matrix, 112 × 112; slice thickness, 3 mm with no gap; number of dynamics, 100 (task-based), 150 (restingstate). Parameters used for DWI were the following: TR/TE, 9,600/70 ms; flip angle, 90°; field of view, 224 mm, matrix 224 × 224; 32 diffusion directions with b = 1,000 and one with b = 0; slice thickness, 2 mm with no gap. For fMRI tests, words were presented visually (using E-Prime software) by back-projection to a screen mounted inside the bore and shown through a mirror mounted on the head coil. A stimulation paradigm consisting of 5 periods of 30 s “stimulation on” separated by 30 s “stimulation off” was repeated twice for sentence reading comprehension and for the verb generation task. The fMRI data were processed using FSL 5.0.6 (fMRI Expert Analysis Tool, www.fmrib.ox.ac.uk/fsl) and registered to a standard space (MNI-152 space). For the rs-fMRI data, several pre-processing steps were taken to remove the time course’s noise in each voxel: (1) motion correction, (2) de-spiking, (3) removal of linear trends (up to 2nd order), and (4) bandpass frequency filtering (0.01– 0.1 Hz). We also applied spatial smoothing with a 5 mm FWHM Gaussian kernel before connectivity measurements in a standard space (MNI-152 space). 18 Eur Neurol 2017;78:15–21 DOI: 10.1159/000477167 For each subject, the seed-driven rs-fMRI analysis was performed using AFNI (http://afni.nimh.nih.gov/afni/). We defined the seed-ROI based on the activation region when applying stimulus-induced fMRI, which was Broca’s (BA 44) area in the left hemisphere. The average time series of seed-ROI were computed and used to produce Pearson’s correlation coefficients maps. Diffusion Tensor Imaging Tractography The Diffusion Toolkit (http://www.trackvis.org/) was used for constructing fractional anisotropy and deterministic diffusion tensor imaging (DTI) tractography by a continuous tracking algorithm (50° angular threshold) after a applying DWI mask. Trackvis (http://www.trackvis.org/) was used for isolation of the arcuate fasciculus (AF) and fiber-track analysis. The ROIs for the AF track were drawn manually over the color fractional anisotropy maps. Results Language Tests and MRI Findings The K-WAB tests demonstrated mild-to-moderate (TSA1; aphasia quotient [AQ], 74.1/100) and moderateto-severe (TSA2; AQ, 42.2/100) aphasia (Table 1). Their spontaneous speech was fluent and grammatically correct with no dysarthria or apraxia of speech. TSA2 often showed perseveration/paraphasic errors, which made the speech content irrelevant. Their repetition ability was generally preserved. Auditory comprehension ability of TSA1 was impaired not only for grammatically complex sentences but also for simple words. TSA2 showed severe deficits even at the word level. The TMA patient showed mild-to-moderate (AQ 69.2/100) aphasia, and the AQ score of HN was within normal range (95.6). Follow-Up Language Tests Follow up K-WAB tests at 3 months post stroke showed that the AQ score of TSA1 increased to the normal range (AQ 94.0/100). However, he sometimes used vocal rehearsal in auditory comprehension tests, a strategy to compensate for incomplete recovery. The AQ score of TSA2 also increased, but she still had mild-to-moderate TSA (AQ 67.3/100). fMRI Results The results of single-subject analysis using nonnormalized data in the acute stage of stroke showed brain activation in TSA1 and TSA2 in perilesional areas of the frontal lobe and temporal regions during the sentence completion task (Fig. 2a). During the verb generation task (Fig. 2b), the activated areas were similar to those for the sentence completion tasks except that temKwon/Shim/Kim/Kim Downloaded by: University of Florida, Gainesville and Jacksonville 128.227.24.141 - 10/20/2017 12:20:47 PM tensive woman who developed speech disturbances and right hemiparesis. On neurological examination, he had prominent aphasia and right hemiparesis. DWI performed on the day of symptom onset showed a left frontal infarct (Fig. 1, TMA). MRA showed severe stenosis of the left carotid artery, and artery-to-artery embolism was considered a presumed cause of stroke. In all these 3 patients, there was no mismatch between DWI and perfusion-weighted MRI. Fluid attenuated inversion recovery images showed no significant white matter ischemic lesions except for TSA2 who showed moderate periventricular and white matter ischemic changes (Fig. 1). HN was a 67-year-old man without any neurological disturbances who was recruited from subjects who visited Asan Medical Center Health Promotion Center for the purpose of a routine health check-up. This study was approved by the Institutional Review Board of Asan Medical Center, and all the patients and control subjects provided written informed consent. R L 6.5 R TSA1 L TSA1 0 Threshold = 4 7 0 Threshold = 4 4 TSA2 TSA2 0 Threshold = 4 8 0 Threshold = 2.5 8 TMA TMA 0 Threshold = 5 11 0 Threshold = 5 7 HN HN 0 Threshold = 7 a R 6.5 L 0 Threshold = 4 b 0.8 R TSA1 L 0.8 TSA1 0 Threshold = 0.5 0.8 0 Threshold = 0.5 0.8 TSA2 TSA2 0 Threshold = 0.5 0.8 d 0 Threshold = 0.5 TMA 0 Threshold = 0.5 0.8 HN 0 Threshold = 0.5 Fig. 2. The results of the initial fMRI induced by sentence completion (a), and verb generation task (b) of the subjects. The white and black arrows indicate respectively Broca’s and Wernicke’s areas activated in each subject. In the initial (c) and follow-up rs-fMRI study (d), the white and black arrows indicate respectively the anterior and posterior language areas connected from the seed ROI (Broca’s regions activated on stimulus-induced fMRI area). The right column of (c) and (d) shows functional connectivity locations in sagittal view (with left 40–50 mm cutout option) over 3D volume rendered T1 images. TSA1, patient 1 with transcortical sensory aphasia; TSA2, patient 2 with transcortical sensory aphasia; TMA, patient with transcortical motor aphasia; HN, healthy normal adult. poral activation was decreased in TSA1. The overall signal was low in the verb generation task. TMA had stronger frontal activation and HN showed broad activation in the frontal and temporo-occipital areas compared with TSA patients during both tasks. DTI Tractography The results of tractography showing the AF in the left hemisphere of the patients with TSA in the acute stage demonstrated that the structural connectivity of language areas through the AF were preserved in these patients (Fig. 3). fMRI in Transcortical Sensory Aphasia Eur Neurol 2017;78:15–21 DOI: 10.1159/000477167 19 Downloaded by: University of Florida, Gainesville and Jacksonville 128.227.24.141 - 10/20/2017 12:20:47 PM c rs-fMRI Results The results of seed-based analysis of the rs-fMRI performed in the acute stage showed significantly decreased resting-state functional connectivity in the language network compared with the control subjects (Fig. 2c). There were weak correlations in the left perilesional and temporo-parietal areas in TSA patients. By contrast, TMA had strong correlations between the frontal and temporal language areas with some scattered regions in the frontal and temporal lobe of both hemispheres. A similar pattern was observed in HN. On the follow-up study, there were significant changes of the connectivity patterns in both TSA patients (Fig. 2d); there were stronger connections between the seed area and fronto-temporal language area and some additional connections in the frontal and parietal lobes. TSA1 TSA2 Discussion 20 Eur Neurol 2017;78:15–21 DOI: 10.1159/000477167 Fig. 3. The diffusion tensor tractographies of arcuate fasciculus in patients with TSA. TSA1, patient 1 with transcortical sensory aphasia; TSA2, patient 2 with transcortical sensory aphasia. hemisphere exceeding that of normal subjects during the language tasks (Fig. 2a, b), suggesting that this is an unlikely hypothesis. In our patients, the result of seed-based analysis of rsfMRI performed in the acute stage showed significantly decreased resting-state functional connectivity in the language network even though there was relatively preserved structural connectivity. Weaker correlations were observed between the seed area (Broca’s) and temporal regions compared with those seen in the control subjects (Fig. 2c). Follow-up tests revealed improved connectivity, with strong correlations in the superior and middle temporal gyri as well as the prefrontal and parietal lobes (Fig. 2d). These results together with relatively preserved structural connectivity shown in DTI tractography (Fig. 3) suggest that functional rather than structural disconnections between the language centers explain patients’ symptoms [11]. The preserved speech fluency in our TSA1 and TSA2 patients may be explained by the relatively preserved precentral gyrus and superior frontal lobe as compared to TSM patients (Fig. 1), that were reported Kwon/Shim/Kim/Kim Downloaded by: University of Florida, Gainesville and Jacksonville 128.227.24.141 - 10/20/2017 12:20:47 PM In patients with TSA with frontal lobe lesions, one of putative mechanisms of impaired auditory comprehension is the concomitant presence of hypoperfusion in the posterior language area [5, 7]. However, our patients did not show diffusion–perfusion mismatches (Fig. 1), and did not support this hypothesis. Although TSA2 patient had an additional parietal infarct (Fig. 1), the lesion was small and located far from language centers. Thus, it is highly unlikely that the lesion was causally associated with patient’s severe aphasia. It is also unlikely that the moderate white matter ischemic changes in this patient produced the patient’s aphasia, although these lesions may be attributed to the relatively slow recovery of language function as compared to TSA1 patient. Another hypothesis is an involvement of a part of the frontal lobe that plays a role in auditory comprehension of grammatically complex sentences [8, 9]. However, our patients demonstrated impaired auditory comprehension even at the single word level, illustrating that this hypothesis is also unlikely. Finally, anomalous inter- or intrahemispheric representation of language centers has been suggested as the mechanism, that is, Broca’s area representing receptive language and Wernicke’s area expressive language [6, 10]. However, although our fMRI study results showed individual differences, there was no reversed intrahemispheric activation pattern or dissociative activation in the right to be responsible for spontaneous speech production [2, 3, 12]. One notable limitation in this study was the small number of patients, which was attributed to the rareness of this condition. Therefore, our results may not be applicable for all such patients, and do not rule out the possibility of other mechanisms. Nevertheless, our results demonstrated that functional disconnections between language centers in the acute stage of the stroke explain the occurrence of TSA in patients with frontal lobe infarction. Acknowledgments This work was supported by Basic Science Research Program through the Ministry of Education of the Republic of Korea and National Research Foundation (NRF-2015R1D1A1A09060547) and by a grant from the Ministry for Health, Welfare and Family Affairs, Republic of Korea (HI14C1985). Disclosure Statement The authors report no conflicts of interest. References fMRI in Transcortical Sensory Aphasia 5 Sethi NK, Burke L, Torgovnick J, Arsura E: Transcortical sensory aphasia as a result of left frontal cortical-subcortical infarction. A case report. Eur Neurol 2007;57:52–53. 6 Kim EJ, Suh MK, Lee BH, et al: Transcortical sensory aphasia following a left frontal lobe infarction probably due to anomalously represented language areas. J Clin Neurosci 2009; 16:1482–1485. 7 Turken AU, Dronkers NF: The neural architecture of the language comprehension network: converging evidence from lesion and connectivity analyses. Front Syst Neurosci 2011;5:1. 8 Caplan D, Hildebrandt N, Makris N: Location of lesions in stroke patients with deficits in syntactic processing in sentence comprehension. Brain 1996; 119(pt 3):933– 949. Eur Neurol 2017;78:15–21 DOI: 10.1159/000477167 9 Stromswold K, Caplan D, Alpert N, Rauch S: Localization of syntactic comprehension by positron emission tomography. Brain Lang 1996;52:452–473. 10 Berthier ML: Unexpected brain-language relationships in aphasia: evidence from transcortical sensory aphasia associated with frontal lobe lesions. Aphasiology 2001;15:99–130. 11 Lim JS, Kang DW: Stroke connectome and its implications for cognitive and behavioral sequela of stroke. J Stroke 2015; 17: 256–267. 12 Enokido H, Torii H, Ainoda N, Hanyu T, Omori S: [Spontaneous speech disturbances in so called transcortical motor aphasia – comparison of 3 cases with the different lesion sites]. No To Shinkei 1984;36:895–902. 21 Downloaded by: University of Florida, Gainesville and Jacksonville 128.227.24.141 - 10/20/2017 12:20:47 PM 1 Boatman D, Gordon B, Hart J, Selnes O, Miglioretti D, Lenz F: Transcortical sensory aphasia: revisited and revised. Brain 2000; 123(pt 8):1634–1642. 2 Otsuki M, Soma Y, Koyama A, Yoshimura N, Furukawa H, Tsuji S: Transcortical sensory aphasia following left frontal infarction. J Neurol 1998;245:69–76. 3 Maeshima S, Kuwata T, Masuo O, et al: Transcortical sensory aphasia due to a left frontal subcortical haemorrhage. Brain Inj 1999;13:927–933. 4 Maeshima S, Osawa A, Nakayama Y, Miki J: Transcortical sensory aphasia following infarction in the left frontal lobe. Eur Neurol 2004;52:125–128.