1312 CLINICAL NOTES Global Aphasia Due to Thalamic Hemorrhage: Report and Review of the Literature A Case Rajeswari Kumar, MD, Ajay K. Masih, Julie Pardo, MS/CCC-SLP ABSTRACT. Kumar R, Masih AK, Pardo J. Global aphasia due to thalamic hemorrhage: a case report and review of the literature. Arch Phys Med Rehabil 1996; 77: 13 12-5. A 55-year-old patient was admitted with weakness in the right extremities and with significant language deficits. Both computed tomography (CT) and positron-emission tomography (PET) using 2-fluoro-2-deoxy-D-glucose were done within 1 month and repeated at 10 months after onset. Language evaluation was done using parts of the Western Aphasia Battery (WAB) and the Boston Diagnostic Aphasia Examination (BDAE) at 1 month after admission, at 3 months, and at 10 months. The initial CT scan revealed hemorrhage in the left thalamus with edema surrounding the hemorrhage. A followup CT scan after 10 months showed an old hemorrhage in the left thalamus with no new lesions. The initial PET scan revealed hypometabolism in the left thalamus, and a repeat PET scan at 10 months showed reduced uptake in the left frontal, left parietal, and left temporal cortex. An initial language evaluation showed impaired auditory and reading comprehension, poor verbal expression, impaired repetition, and difficulty with naming and with sentence completion. The patient was diagnosed with global aphasia. Follow-up language evaluation at 3 months and 10 months showed only minimal improvement in his communication. Global aphasia due to lesions in the thalamic region is rare and the prognosis is poor. 0 1996 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation DEFICITS that result from lesions in the cerebral cortex are well known. Recent reports have suggested Lthat ANGUAGE speech defects can occur in patients with lesions in the subcortical structures in the dominant hemisphere, including the thalamus, the basal ganglia, the internal capsule, and the paraventricular white matter (PVWM).‘.“ Positron-emission tomography (PET) has shown a close association between these subcortical structures and the language areas in the cortex.’ Global aphasia caused by subcortical lesion is rare. We report here a patient who developed global aphasia after a hemorrhagic stroke involving the left thalamus. CASE REPORT The patient was a 55-year-old right-handed man who was brought to the emergency room by a friend who had found the From the Department of Physical Medicine and Rehabilitation (Dr. Kumar) and the Deoartment of Audioloev and Sueech Patholozv (Ms. Pardo), Veterans Affairs Me&al Center, West Los Angeles, CA; the Eeiartment ofMedicine, UCLA School of Medicine (Dr. &mar); and the Loyola University Chicago S&itch School of Medicine (Mr. Masih), Maywood, IL. Submitted for publication March 21, 1996. Accepted May 13, 1996. No commercial party having a direct financial interest in the results of the research suooortine this article has or will confer a benefit won the authors or won anv orsanizazon with which the authors are associated. ’ Reprint re&ests to Rajeswari Kumar, MD, Department of Physical Medicine and Rehabilitation (117), Wilshire and Sawtelle Boulevard, Los Angeles, CA 90073. 0 1996 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation 0003-9993/96/7712-3956$3.00/O Arch Phys Med Rehabil Vol77, December 1996 patient unresponsive. His blood pressure was 220040mmHg. He was able to follow one-step commands, but his responses were inconsistent. Cranial nerve examination findings were significant for a right facial droop and a right hemianopsia. Motor examination revealed flaccid right upper and lower extremities, while motor strength on the left side was normal. Initial computed tomography (CT) showed a left thalamic hemorrhage with surrounding edema that had penetrated into the lateral ventricle. The patient was given medication to control blood pressure and was also started on Dilantin” and on Decadron.b A repeat CT scan obtained 3 days after admission showed a large hemorrhage involving the left thalamus (fig 1) which had entered the lateral ventricle. The edema surrounding the bleed was also impinging on the left internal capsule. The blood additionally entered into the subarachnoid space. There were no cortical lesions. One month after admission, PET showed hypometabolism in the thalamus and caudate nucleus. The study was limited because of movement artifact created by the patient as a result of agitation. Forty minutes after the intravenous administration of lOm/Ci of 2-fluoro-2-deoxy-D-glucose, imaging was initiated using the Siemens ECAT-953 tomographic scanner.’ During the 40-minute uptake period prior to scanning, the patient was resting quietly in a darkened room, exposed to ambient noise, with his eyes open and ears unplugged. He was in a fasting state and was imaged with head restraints. Multiple transaxial images of the head were obtained at 4.0~mm intervals from the vertex to the base of the brain. Identical regions were drawn on the processed image of each side of the brain, and relative metabolism of glucose was compared between the two sides. The relative metabolism of glucose was calculated averaging the values of the three best images of a particular area of interest. The relative metabolism of glucose was 3.30 @i/mL in the right thalamus and 1.69 &!i/mL in the left thalamus. There was also a relative hypometabolism in the left caudate nucleus, left parietal lobe, and left frontal lobe. The metabolism in the subfrontal and occipital region, however, was relatively spared (fig 2 and table 1). The initial speech evaluation and re-evaluations were conducted in multiple sessions. Assessment during these evaluations was often difficult because of the patient’s poor attention span and because of his difficulty in focusing on a task. The initial evaluation after 1 week found that expressive output was severely impaired, as was object naming and reading of short sentences. He did not have dysarthria. The Western Aphasia Battery (WAB) and the Boston Diagnostic Aphasia Examination (BDAE) were administered at 3 weeks. He had difficulty with confrontational naming and in reading short sentences. The errors were due to perseveration and paraphasia. Additionally, his repetition was impaired and he had poor expression. His auditory comprehension was reduced for both short and long phrases. Writing could not be tested because of his right-sided hemiparesis. He responded poorly with gestures. A communication board was attempted but he could not be oriented to the pictures (table 2). The aphasia quotient derived from the results of the WAB showed significant GLOBAL APHASIA DUE TO THALAMIC HEMORRHAGE, Table Area Studied 1313 Kumar 1: Glucose Uptake in the Brain Using 2-Fluoro-2-Deoxy-D-Glucose in PET Right-Sided Glucose Uptake, &i/mL (% SD) Left-Sided Glucose Uptake, &i/mL (% SD) 3.309 (1.6) 2.847 (3.3) 2.719 (2.4) 9.78 (9.0) 3.16 (7.5) 3.64 (7.1) 2.71 (4.7) 1.690 (1.9) 2.405 (1.5) 2.072 (2.7) 1.01 (10.9) 2.79 (9.6) 3.52 (6.9) 2.72 (5.0) Thalamus Frontal Parietal Temporal Caudate Occipital Subfrontal DISCUSSION The classical models of aphasia do not explain completely all of the aphasia syndromes. With the advent of modern neuroimaging techniques, the presence of aphasia has been attributed not only to lesions of the cerebral cortex but also to lesions within the subcortical region. Initial reports had suggested that lesions of the thalamus and of the basal ganglia could also lead to aphasia. More recent reports have suggested that aphasias can also result from lesions in the internal capsule and in the PVWM. In the literature subcortical aphasiashave been defined more clearly by their specific neuroanatomic locations rather than by specific language profiles. Consequently, this review presents subcortical aphasias from a neuroanatomic approach. Fig 1. CT showing hemorrhage into the left thalamus. deficits. The patient was considered a “1” on the WAB, which means that all communication is through fragmented expression and by reliance on inference, questioning, and guessing by the listener. The range of information that can be exchanged is limited, and the listener carried the burden of communication. Only three subsets from the BDAE were administered; therefore, the rating scale of the speech characteristics could not be reported. The patient was diagnosed as having global aphasia. Re-evaluation at 3 months and 10 months showed minimal improvement. He continued to have significant language impairment (table 2). Thalamus Thalamic aphasia results from specific lesions of the left thalamus. Thalamic infarcts are a disease of the posterior cerebral circulation and, thus, can be differentiated from infarcts of other subcortical white matter regions that are supplied by the middle cerebral artery.’ Thalamic aphasia is characterized by fluent speech with paraphasias and perseveration, by lack of spontaneous speech, by less impaired comprehension, and by relatively preserved repetition.2X6S7 A model by Ojemann’ hypothesized that the thalamus may assist in language decoding and semantic accuracy through feedback mechanisms involving thalamocortical pathways. Based on this model, thalamic lesions would produce verbal paraphasias. Table 2: Parts of the WAB Task Fig 2. PET with the left thalamus, 2-fluoro-2-deoxy-D-glucose showing basal ganglia, and frontotemporal reduced region. uptake in Auditory/reading comprehension Personally relevant yes/ no questions (WAB) Complex yes/no questions (WAB) Complex ideational material (BDAE) Body part identification WAB) Sequential commands (BDAE) Single word reading IWAB) Reading commands (WAB) Spoken/written expression Repetition (WAB) Object naming (WAB) Sentence completion (WAB) Responsive naming (WAB) Word fluencv (WABI Picture description (BDAE) Writing and the BDAE Adminstered 3 Weeks 3 Months IO Months 13114 12114 IO/l4 416 5/6 316 O/4 l/4 014 5115 15/18 7116 1 Part 1-2 Part 1-2 Part 0 516 216 0 2 Part 2 Part I-2 Syllables Unintelligible 1-9 Syllables 15120 I-7 Syllables 11120 215 415 315 115 0 4/5 0 Provided 8 elements Unable 415 0 Provided 10 elements Unable Unable Unable Arch Phys Med Rehabil Vol77, December 1996 1314 GLOBAL APHASIA DUE TO THALAMIC In Crosson’s model of aphasias6 the thalamus played roles in cortical arousal and activation and in preverbal semantic monitoring. Crosson also proposed that this semantic monitoring occurred by a feedback mechanism via a bidirectional cortico-thalamo-cortical pathway. This pathway sent semantic information from the anterior cortex to the ventral anterior thalamus, from the ventral anterior thalamus to the pulvinar, and from the pulvinar to the temperoparietal cortex. Semantic refinements that are processed in the temperoparietal cortex were then sent back to the anterior cortex by this same pathway for execution of speech.6 Based on radiological evidence, paraphasic speech with decreased speech output is considered to result from lesions to the ventrolateral and anterolateral thalamic nuclei, whereas neologistic speech with increased speech output is considered to result from lesions involving the pulvinar or posterolateral nuclei.g Lazzarino et al7 suggested that the dominant paramedian thalamic nuclei, and in particular the dorso-median thalamic nucleus, had some role in language disturbances due to their connections with both Broca’s and Wernicke’s area. The degree of involvement accounted for the different aphasic profiles.7 Nonfluent thalamic aphasias rarely occur. Megans et al” reported an isolated thalamic abscess that resulted in nonfluent aphasia with paraphasia. While comprehension was well preserved, there was also impairment of object naming, repetition, and reading. Initial single-photon emission computed tomography (SPECT) showed a large area of hypoperfusion of the left hemispheric cortex, suggesting secondary cortical dysfunction. Thus, cortical diaschisis which is defined as a transient depression of function in a region remote, but connected to the lesion, might have a role in aphasia patients with subcortical 1esions.l’ Basal Ganglia, Internal Capsule Basal ganglia lesions have also been reported to cause aphasias. In contrast to thalamic lesions, basal ganglia lesions do not produce a single aphasia profile. They are characterized by fluent or nonfluent output, intact or impaired repetition, and varying levels of performance in other language measurements.’ The basal ganglia collectively refer to the caudate nucleus, putamen, and globus pallidus. The anterior limb of the internal capsule lies between the caudate nucleus and the lenticular nucleus, while the posterior limb of the internal capsule lies between the lenticular nucleus and the thalamus. The basal ganglia and the internal capsule are considered together because vascular lesions frequently affect both structures. Crossor? suggested that the basal ganglia could affect language through their influence on the tonic excitatory neural activity of cortical language areas, or may function as a motor release system that allows formulated language for motor programming to be released at the proper time. He believed that the globus pallidus had an inhibitory influence on the anterior cortical language centers indirectly through the ventral anterior thalamic nucleus. Likewise, the globus pallidus is under the influence of the caudate nucleus, which is itself normally inhibited by temperoparietal mechanisms. Caudate stimulation by these temperoparietal mechanisms, therefore, results in inhibition of the globus pallidus and in subsequent excitation of anterior cortical mechanisms, resulting in language output. With language output, temperoparietal inhibition of caudate mechanisms is reestablished through frontocaudate pathways.6 Robin and Schienberg’ proposed that the basal ganglia could affect language through their effect on attention and perceptual integration or by causing a depression of cortical metabolic function. Mendez at al” concluded in their study of caudate Arch Phys Med Rehabil Vol77, December 1996 HEMORRHAGE, Kumar nuclei lesions that basal ganglia have a primary role in behavioral and cognitive functions, in processing information related to frontal regions, and probably in conceptually integrating material. The caudate lesions resembled prefrontal lesions by their impairment in the initiation and inhibition of spontaneous behavior. l1 D’Esposito and Alexander4 reported on cases of left putaminal hemorrhage (subcortical but not thalamic). For patients with persistent aphasia, these lesions resulted in nonfluency, impaired comprehension, and, perhaps, impaired repetition.4 Lesions involving the anterior limb of the internal capsule cause reduced speech output, which is thought to be due to interruption in the frontocaudate pathways6 Mega and Alexandelj looked only at striatocapsular subcortical lesions. These nonthalamic lesions produced problems with generative language, abnormalities of verbal fluency, deficient sentence generation with increased latencies, and perseverations. There was normal phonemic paraphasias, word repetition, oral reading, and word comprehension. The investigators believed that damage to the frontal-striatal system was responsible for the subcortical aphasia profiles. They specifically noted that lesions of the dorsolateral frontal circuit with lesions of the dorsolateral caudate or anterior PVWM resulted in language deficits. Lesions of the anterior cingulate and supplementary motor area circuit with the caudate resulted in impaired initiation. Putamenal lesions resulted in speech latency deficits3 Alexander et al5 also noted word-finding deficits with extensive injury to the striatum, posterior-medial portion of the anterior limb of the internal capsule, and the anterior-superior PVWM. Weiller et al,‘* however, concluded from their study of striatocapsular infarcts that aphasia or neglect was caused by selective neuronal loss of the cerebral cortex as a result of a prolonged middle cerebral artery occlusion and insufficient collateral blood flow. They found that the different aphasic syndromes did not differ in the amount of involvement of the putamen, pallidum, head of the caudate, or deep white matter. The difference was mainly due to hypometabolism and subsequent cortical neuronal loss.‘* Metter et al’ likewise demonstrated the effect of subcortical lesions on the frontal lobe. By measuring resting glucose metabolism with a PET scanner they showed that subcortical damage involving the internal capsule, corpus striatum, globus pallidus, and thalamus affected the frontal lobe. Lesions in the posterior limb of the internal capsule had the strongest effect. This phenomenon can be explained by diaschisis.’ Role of White Matter Naese8 originally described three syndromes that suggested white matter pathways are the critical structures in language disorders. The first was an anterior subcortical aphasia created by a lesion of the caudate nucleus, anterior limb of the internal capsule, and anterior putamen resulting in nonfluent speech and lesser impairment of comprehension and repetition (Broca’stype aphasia). The second involved posterior lesions, including lesions to the temporal isthmus, resulting in a Wemicke’s-type aphasia. The third involved both sites with a global aphasia syndrome. Naeser and Palumbo’3 reported on patients with subcortical aphasia as having lesions of the putamen and internal capsule with white matter extension (anterior, posterior, and superior). The anterior extension involved the anterior limb of the internal capsule and white matter near the frontal horn (deep to Broca’s area). The posterior extension involved white matter in the temporal isthmus (deep to Wemicke’s area). The investigators believed that the temporal isthmus was responsible for language GLOBAL APHASIA DUE TO THALAMIC comprehension deficits, as a result of damage to the auditory pathways that run from the medial geniculate body to Heschl’s gyms. They noted that the middle third of the PVWM contained both the body of the caudate and the motor-sensory projections of the mouth that are thought to be involved in speech production.5.‘3 The superior extension involved white matter near the body of the lateral ventricle. Global aphasia patients had larger subcortical lesions than nonaphasic patients. These patients also had hypoperfusion to the frontotemporal cortex. Thus, Naeser and Palumbo concluded that the white matter pathways that are essential for speech and comprehension are affected.13 D’Esposito and Alexander,” who investigated putaminal hemorrhage, and Weiller et al,‘* who studied striatocapsular infarcts, likewise observed that damage to the deep frontal and anterior PVWM resulted in nonfluency, while damage to the deep temporal white matter (posterior PVWM) was associated with impaired word comprehension. Damage to both the anterior and posterior PVWM resulted in a global aphasia. Naeser et alI4 found that combined deep subcortical lesions involving the medial subcollosal fasciculus and the middle third of the PVWM (containing the body of the caudate nucleus) prevented spontaneous speech, because of the loss of pathways for speech initiation, motor execution, or sensory feedback. Okuda et al” suggested that white matter lesions were associated with cortical hypoperfusion (measured with a SPECT scan) of the cortical perisylvian language areas, including Broca’s and Wernicke’s areas, which were critical in the development of global subcortical aphasia. Prognosis Prognosis for subcortical aphasia is reported to be good. Aphasia due to thalamic lesions tend to have a rapid and favorable recovery. 9,‘6 Weiller” reported that 64% of patients with striatocapsular infarcts had, at 1 year, complete resolution or only minor problems with language. Naeser and Palumbo’3 reported that patients with subcortical temporal lobe lesions, including lesions to the temporal isthmus, had a better recovery of auditory comprehension at 1 to 2 years poststroke when compared to patients with lesions involving greater than half of Wemicke’s cortical area. They also reported, however, that lesions in the subcortical white matter areas can severely limit recovery of spontaneous speech. Combined white matter lesions of the frontal area (medial subcollosal fasciculus area) and the middle third of the PVWM were particularly important in speech recovery. Conclusion Our patient showed anatomic damage to the left thalamus. PET scans at 1 month also showed relative hypometabolism in the region of the caudate, and in the frontal, temporal, and parietal cortex. Although the literature reports that thalamic aphasia is characterized by fluent language with paraphasias and relatively spared comprehension, our patient demonstrated a global aphasia. Global subcortical aphasia is reported to be rare. The secondary cortical dysfunction may be due to cortical diaschisis, as suggested by previous investigators.‘*‘0,‘2 Because of the lack of speech recovery over time, the patient’s global aphasia was more likely the result of white matter HEMORRHAGE, 1315 Kumar lesions in the frontal region, caudate region, and the temporal region, as suggested by the relative hypometabolism in these areas. According to the increasing evidence in the literature, damage to white matter in the frontal region and the caudate region would severely limit recovery of speech.13 Damage to white matter in temporal region would result in comprehension deficits. The prognosis for recovery of speech in global subcortical aphasia is reported to be poor. Acknowledgment: The authors extend their thanks to Charles Brown MD, Department of Nuclear Medicine and Earnest Sponvilli, MD, Department of PM&R at the VA Medical Center at West Los Angeles, CA, for his assistance in reviewing the PET scan. References 1. Metter EJ, Riege WH, Hanson WR, Jackson CA, Kempler D, van Lancker D. Subcortical structures in aphasia: an analysis based on (F-l 8)-fluordeoxyglucose, positive emission tomography, and computed tomography. Arch Neurol 1988;45: 1229.34. 2. Ron DA, Schienberg S. Subcortical lesions and aphasia. J Speech Hear Disord 1990;55:90-100. 3. Mega MS: Alexander MP. Subcortical aphasia: the core profile of capsulostriatal infarction. Neurology 1994;44:1824-9. 4. D’Esposito M, Alexander MP. Subcortical aphasia: distinct profiles following left putaminal hemorrhage. Neurology 1995;45:38-41. 5. Alexander MP, Naeser MA, Palumbo CL. Correlations of subcortcal CT lesion sites and aphasia profiles. Brain 1987; 110:961-91. 6. Crosson B. Subcortical functions in language: a working model. Brain Lang 1985;25:257-92. 7. Lazzarino LG, Nicolai A, Biasizzo E. Language disturbances from mesencephalo-thalamic infarcts: identification of thalamic nuclei by CT-reconstructions. Neuroradiology 1991;33:300-4. 8. Ojemann GA. Brain organization for language from the perspective of electrical stimulation mapping. Behav Brain Sci 1983;2:189-230. 9. Graff-Radford NR, Damasio AR. Disturbances of speech and language associated with thalamic dysfunction, Semin Neurol 1984;4: 162-8. 10. Megans J, van Loon J, Goffin J, Gybels J. Subcortical aphasia from thalamic abscess. J Neurol Neurosurg Psychiatry 1992; 55:3 19-21. 11. Mendez MF, Adams NL, Lewandowski KS. Neurobehavioral changes associated with caudate lesions. Neurology 1989;39:34954. 12. Weiller C, Willmes K, Reiche W, Thron A, Isensee C, Buell U, et al. The case of aphasia or neglect after striatocapsular infarction. Brain 1993; 116:1509-25. 13. Naeser MA, Palumbo CL. Neuroimaging and language recovery in stroke. J Clin Neurophysiol 1994; 11: 150-74. 14. Naeser MA, Palumbo CL, Helm-Estabrooks N, Stiassny-Eder D, Albert ML. Severe nonfluency in aphasia: role of the medial subcallosal fasciculus and other white matter pathways in recovery of spontaneous speech. Brain 1989; 112:1-38. 15. Okuda B, Tanaka H, Tachibana H, Kawabata K, Sugita M. Cerebral blood flow in subcortical global aphasia: perisylvian cortical hypoperfusion as a crucial role. Stroke 1994;25:1495-9. 16. Ozeren A, Sarica Y, Efe R. Thalamic aphasia syndrome. Acta Neurol Belg 1994;94:205-8. Suppliers a. Parke-Davis, Division of Warner-Lambert Road, Morris Plains, NJ 07950. b. Merck and Company, Inc., PO Box 4, West c. C.T.I. PET System, 810 Innovation Drive, 2562. Arch Phys Med Rehabil Company, 201 Tabor Point, PA 19486.0004. Knoxville, TN 37932. Vol77, December 1996