Brain (1985), 108,485-516 NONHAEMORRHAGIC THALAMIC INFARCTION CLINICAL, NEUROPSYCHOLOGICAL AND ELECTROPHYSIOLOGICAL FINDINGS IN FOUR ANATOMICAL GROUPS DEFINED BY COMPUTERIZED TOMOGRAPHY P. J. ESLINGER1 and A. R. DAMASIO1 {From the ^Department of Neurology (lDivision of Behavioral Neurology and the 2 Division of Electrophysiology), University of Iowa College of Medicine, Iowa City, Iowa 52242, USA) SUMMARY Twenty-five patients with nonhaemorrhagic infarcts of the thalamus were studied clinically and by neuropsychological testing, computerized tomography and somatosensory evoked response (SER) recordings. Our aim was to determine whether the findings in these different tests would form distinct symptom clusters associated with different anatomical territories of the thalamus. Infarction conforming to the tuberothalamic arterial territory caused a facial paresis for emotional movements, severe neuropsychological deficits and a delay of the SER after P14. Infarction conforming to the interpeduncular profundus arterial territory caused a supranuclear vertical gaze paresis, severe neuropsychological deficits and a delay in the P60 component of the SER. Infarction conforming to the anterior choroidal territory caused a hemiparesis, moderate neuropsychological deficits and varied sensory evoked responses. Patients with infarctions conforming to the entire geniculothalamic territory had sensory loss in multiple modalities, minimal neuropsychological deficits and absence of sensory evoked responses after PI4. A lacune in this territory caused pure hemisensory loss involving part of the body for the modalities of pain and light touch but not proprioception or vibration. Neuropsychological deficits were uncommon and N32 and N60 were delayed in the SER. INTRODUCTION We have recently studied 25 patients with nonhaemorrhagic thalamic infarctions by computerized tomography (CT), and by clinical, neuropsychological and electrophysiological procedures. Our aim was to determine if, when we grouped the patients according to location of lesion as determined by CT, consistent clusters of clinical, neuropsychological and SERs would emerge in relation to the different sites of damage. Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 by N. R. GRAFF-RADFORD, 1 H. DAMASIO,1 T. YAMADA,2 486 N. R. GRAFF-RADFORD AND OTHERS METHODS Computerized Tomography We prepared templates of 4 mm brain slices through the thalamus and surrounding areas. The templates were based on our own human brain sections and on the Roberts and Hanaway Atlas (1970). Abnormal images detected on the CT scan were plotted on the appropriate templates by one of the investigators blind to clinical, neuropsychological and SER data. A composite of the vascular territories on 4 mm thalamic templates is seen infig.1. FIG. 1. Templates (4 mm) through the thalamus depicting the following four arterial territories. Vertical hatching = geniculothalamic territory; horizontal hatching = tuberothalamic territory; unbroken diagonal hatching = interpeduncular profundus territory; broken diagonal hatching = anterior choroidal territory. Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 Clinical and Neuropsychological Studies Twenty-five patients were examined neurologically, of which 23 underwent the following battery of neuropsychological tests: Temporal Orientation (Benton et al., 1983); WAIS-R (Wechsler, 1981); Multilingual Aphasia Examination (Benton and Hamsher, 1978); Wechsler Memory Scale (Wechsler, 1945) (prorated without the visual reproduction subtest); Benton Visual Retention Test—Administration A, Form C (Benton, 1974); Rey-Osterrieth Complex Figure (Rey, 1941; Osterrieth, 1945) (copy and delayed recall); Rey Auditory Verbal Learning (Rey, 1964); Facial Recognition (Benton and Van Allen, 1968); and Judgment of Line Orientation (Benton et al., 1978). Estimation of IQ, language, short-term memory and visual perception was based on the results of these tests. NONHAEMORRHAGIC THALAMIC INFARCTS 487 Somatosensory Evoked Responses RESULTS Anatomical Findings On the basis of CT data the patients were divided into four groups. The anatomical findings for each group are given in the following order: number of cases, site of lesion and probable arterial territory. For each group a CT scan and a diagram depicting the usual territory of the artery probably involved in the infarcts are shown. Posterolateral thalamic infarcts (n = 9). The lesions were situated posterolaterally in the thalamus in the territory of the geniculothalamic artery. Fig. 2A shows a CT scan and fig. 2B the arterial territory for this group. In 4 of the 9 patients there was Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 The electrophysiological studies were performed by another investigator, also blind to CT, clinical and neuropsychologicalfindings:silver/silver chloride disc electrodes were attached according to the 10-20 International System at F 3 , F 4 , C3, C4, P3, P4, O^ O2, F 7 , F 8 , T3, T4, T5, T6, and between the sixth and seventh cervical spines with an impedance of less than 5 KQ. Linked ears served as the reference. The median nerve was stimulated at the wrist with a pulse duration of 0.1 ms at a rate of 2.5/s. The intensity was adjusted to elicit a modest twitch of abductor pollicis brevis. The input was amplified by a 16-channel Grass EEG Recorder, Model 5P11. The frequency response was 3 to 1000 Hz (3 dB down). The output was summated by a 21 MX Hewlett Packard computer. The analysis time was 95 ms with a delay of 5 ms and the digitizer sampling time was 0.2 ms. There were four tests, each consisting of 1500 stimulations of the median nerve, left, right, and bilateral twice. The computer was programmed to edit out responses with seven successively overloaded points, so that potentials with unrealistically large potentials contaminated by ECG or muscle artefacts were deleted. The results were plotted on an x-y plotter. Yamada et al. (1984A) have recently reported that the latency of thefirstand second negative waves is different in frontal, central and parietal regions. In the frontal region these latencies are 17 and 29 ms, respectively, in the central region, 19 and 32 ms, and in the parietal region, 20 and 34 ms. They found that unilateral stimulation is reliable for assessment of early components (N17, N19, N20) but not later components (N32, N34, N60) because of intertrial variability of the latter in normals. However, bilateral stimulation normally causes symmetrical late responses and thus can be used to assess the later components. Bilateral stimulation was not helpful in assessing P14 and frontally recorded N17 and N29, since they are distributed bilaterally, even with unilateral stimulation. It would be expected that with bilateral median nerve stimulation, SERs recorded over each hemisphere should be the amalgam of both sets of stimulation. In practice, however, Yamada and coworkers (1983, 1984a, b) have not found this to be entirely true. With bilateral stimulation, the SER recorded over each hemisphere is predominantly from the contralateral hand, as if with bilateral stimulation there is an impediment to the ipsilateral components. However, we only used bilateral stimulation in conjunction with unilateral stimulation. The former is especially helpful in confirming abnormalities of the later SERs (N32, N34 and N60) which, as mentioned earlier, have significant intertrial variability with unilateral stimulation but not with bilateral stimulation. Measurements in milliseconds were made to P14 and the subsequent negative waves at each of the following areas: N17, N29 at frontal, N19, N32 and N60 at central and N20, N34 and N60 at parietal electrodes. We compared latencies of the left and right sides as well as latencies established for normals in our laboratory (normal + 3 SD). The details of normal features and abnormal criteria of topographic SEP recordings are described elsewhere (Yamada et al., 1984*). N. R. GRAFF-RADFORD AND OTHERS 488 B FIG. 2A. CT scan from Case 8 showing infarction in the geniculothalamic arterial territory. Note the concomitant posterior cerebral artery infarction, B, territory of the geniculothalamic artery. This is based upon the Schlesinger Atlas (1976), fig. 27.4 and 27.5. A = anterior nucleus; AC = anterior commissure; AV = anteroventral nucleus; B = basilar artery; C = carotid artery; CM = centrum medianum; Co = colliculi; DM = dorsomedial nucleus; DP = deep interpeduncular profundus artery; F = fornix; GPL = globus pallidus (lateral); GPM = globus pallidus (medial); LG = lateral geniculate; LP = lateroposterior nucleus; MG = medial geniculate; MAM = mammillary bodies; MT = midbrain tegmentum; MTT = mammillothalamic tract; OT = optic tract; PCe = posterior cerebral artery; PCo = posterior communicating artery; PL = pulvinar (lateral); PM = pulvinar (medial); Pu = pulvinar; R = reticular nucleus; RN = red nucleus; ST = subthalamic nucleus; Tt = tuberothalamic artery; VL = ventrolateral nucleus; VPL = ventral posterolateral nucleus; VPM = ventral posteromedial nucleus; ZI = zona inserta. (Abbreviations for figs. 2B, 4B, 5B and 6B.) Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 PCe NONHAEMORRHAGIC THALAMIC INFARCTS also infarction in the posterior cerebral artery territory. Lacunes may also occur in this territory. CT scanning is often negative, but with 4 mm cuts on a high resolution scanner we were able to demonstrate lacunes in 4 cases (fig. 3). Anterolateral thalamic infarction (n = 5). These are located in the territory of the tuberothalamic artery (also termed the polar artery by Percheron, the anterior internal optic artery by Duret and the premamillary pedicle by Foix and Hillemand, as quoted by Castaigne et ai, 1981). Two of our patients developed their infarct postoperatively after clipping of giant posterior communicating artery aneurysms. An example of a CT scan for this group is shown infig.4A and the arterial territory infig.4B. Medial thalamic infarction (n = 3). These are in the territory of the deep interpeduncular profunda arteries (also termed the paramedian thalamic arteries by Percheron, the posterior internal optic arteries by Duret and the thalamoperforating pedicle by Foix and Hillemand, as quoted by Castaigne et al., 1981). An illustrative CT scan is shown in fig. 5A and the usual arterial territory in fig. 5B. Note that one interpeduncular profundus artery may supply the medial thalamus bilaterally and thus lead to bilateral thalamic infarction. We have seen two such examples. Lateral thalamus and posterior internal capsule (n = 8). These patients had lesions in the posterior capsule which extended into the thalamus (fig. 6A). The lesion may involve part or the whole anterior choroidal territory (fig. 6B). Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 FIG. 3. CT scan of Case 8 showing a lacune in the geniculothalamic territory seen only in one cut. N. R. GRAFF-RADFORD AND OTHERS 490 FIG. 4A. CT scan of Case 14 showing infarction in the tuberothalamic arterial territory, B, territory of the tuberothalamic artery. This is based on fig. 28.4 of the Schlesinger Atlas (1976). Abbreviations are as forfig.2B. Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 A NONHAEMORRHAG1C THALAMIC INFARCTS 491 FIG. 5A. CT scan of Case 17 showing infarction in the interpeduncular profundus arterial territory. Note the bilateral lesions. This scan is contrast enhanced, B, territory of the interpeduncular profundus artery. The dotted lines indicate other areas which may also be supplied by this artery. This is based on fig. 28.6 and 28.7 of the Schlesinger Atlas (1976). Abbreviations are as forfig.2B. Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 A 492 N. R. GRAFF-RADFORD AND OTHERS Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 PCe FIG. 6A. CT scan of Case 23 showing infarction in the anterior choroidal arterial territory, B, the territory of the anterior choroidal artery. This is based onfig.27.2 of the Schlesinger Atlas (1976). Abbreviations are as forfig.2B. NONHAEMORRHAGIC THALAMIC INFARCTS 493 Clinical and Neuropsychological Findings A 58-year-old right-handed woman suffering from myotonic dystrophy noted the sudden onset of left-sided numbness on June 16,1982. She was admitted to hospital and on examination found to have cataracts, percussion myotonia and proximal weakness. These were long-standing deficits. Her mental status was normal, as were her cranial nerves. Her tendon reflexes were diminished but symmetrical; plantar reflexes were flexor. She had loss of sensation on her left side for pinprick, light touch, proprioception, vibration, two-point discrimination, graphaesthesia and stereognosis. No hyperaesthesia or dysaesthesia was noted. Her left hand showed pseudoathetoid movements with her eyes closed. Neuropsychological testing demonstrated difficulties with short-term visual memory for designs, constructional copy and recall of the Rey-Osterrieth complex figure, and controlled oral word association. The remainder of the examination was normal, including visual perceptual, intellectual, linguistic and short-term verbal memory abilities. There was no clinically significant amnesia or aphasia. The 5 patients with partial lesions in the geniculothalamic territory had either decreased pinprick or light touch appreciation, without loss for proprioception or vibration, involving only part of the contralateral body, for example for the face only in Case 9, the leg only in Case 6 and the face and arm only in Case 8. None had hemiparesis but 3 had visual field defects. Clinical assessment of mental status and neuropsychological testing revealed no remarkable deficits except for the patients with posterior cerebral artery infarction who had visual memory and visual perceptual deficits. Case 8 is an illustrative example of this subgroup. A 65-year-old right-handed clinical psychologist suddenly developed right-sided paraesthesiae. On examination he had right hemisensory loss for pinprick involving the face and hand but not the leg. Testing for vibration sense, proprioception, graphaesthesia and stereognosis in the right hand was normal. No motor deficits were noted. On clinical testing his mental status was normal. Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 The clinical and neuropsychological features of these cases are summarized in Tables 1 and 2; further clinical data and the actual neuropsychological test scores are given in the Appendix. Additional findings, divided by group, are as follows. Posterolateral thalamic infarction. Cases 1, 2, 7 and 9 had concomitant posterior cerebral artery infarction. The patients have been divided into two subgroups, first those with probable complete geniculothalamic infarction (the first 4 patients in Tables) and, secondly, those with partial geniculothalamic territory infarction (Cases 5 to 9 in Tables). Sensory loss for several modalities was present in all 4 patients in the first subgroup and is the clinical hallmark of this category. However, 1 also had dysaesthesia, 2 had hemipareses and 2 had visual field defects. In addition to thalamic damage, Cases 1 and 2 had sustained associated infarctions in the territory of the posterior cerebral artery. Our behavioural assessment of Case 1 revealed a severe left-sided hemispatial neglect and poor attention span leading to overall poor neuropsychological performances. Case 2 was not formally examined with neuropsychological tests. The remaining 2 patients (Cases 3 and 4) had no concomitant posterior cerebral infarction on CT scan, no visual field defects and only mild neuropsychological impairment. Case 4 illustrates this subgroup. Case Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 TABLE I CLINICAL FINDINGS Side of infarction Muscle strength Plantar reflexes Pinprick Light touch Propnoceptton Vibration Dysaesthesta Visual field defect R R R R R L L Decreased Decreased Normal Normal Normal Normal Normal Normal Normal Extensor Extensor Flexor Flexor Flexor Flexor Flexor Flexor Flexor Decreased Decreased Decreased Decreased Decreased* Decreased Normal Decreased Decreased** Decreased Decreased Decreased Decreased Decreased* Decreased Normal Decreased Decreased •* Decreased Decreased Decreased Decreased Decreased* Normal Normal Normal Normal Decreased Decreased Decreased Decreased Normal Normal Normal Normal Normal No Yes No No Yes Yes No No Yes Yes + Yes+ No Yes No Yes Yes^ No Yes+ Anterolaleral thalamic in fa rction R 67/M 10 R II 75/M 66/F R 12 R 67/F 13 60/F R 14 L L L R R Decreased Normal Normal Decreased Decreased Extensor Flexor Flexor Extensor Extensor Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Decreased Normal Normal Normal Normal Normal Normal Normal Normal Normal No No No Yes No Yes No No No No Medial thalamic infarction 69/F R 15 R 74/F 16 R 81/M 17 Bilateral R Bilateral Normal Decreased Normal Extensor Flexor Extensor Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal No No No No No No Lateral thalamic/posterior limb of internal capsule infarction 58/M R R Decreased 18 R R Decreased 63/F 19 R Decreased 72/M L 20 R L Decreased 47/M 21 67/M R L Decreased 22 R L Decreased 58/M 23 62/F R L Decreased 24 R L Decreased 67/M 25 Extensor Extensor Extensor Extensor Extensor Extensor Extensor Extensor Normal Normal Normal Decreased Decreased Normal Decreased Decreased Normal Normal Normal Normal Normal Decreased Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal No No No No No No No No No No Yes No No No No No No. AgejSex Handedness Posterolateral thalamic infiirction 1 48/F R R 78/F 2 64/F 3 L 4 58/F R R 66/M 5 19/M R 6 R 7 68/M R 65/M 8 R 9 60/M R R z o > T) rn ;> o o rj h Associated posterior cerebral infarct; * foot only; *• lower face only. > Z a o m 7° Case No Age/Sex Years of education Posterolateral thalamic infarctioni 48/F 10 1 2 78/F — 3 64/F 12 58/F 4 12 66/M 16 5 19/M 13 6 69/M 16 7 65/M 8 16 60/M 11 9 Anterolateral thahimic infarction 67/M 12 10 11 75/M 16 12 66/F 7 67/F 13 8 60/F 8 14 Medial thalamic infarction 69/F 16 15 16 74/F 12 81/M 12 17 TESTING Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 T A B L E 2. R E S U L T S O F N E U R O P S Y C H O L O G I C A L Side of lesion Verbal IQ Performance IQ Language Verbal memory Visual memory Visual Perception R R R R R R R L 57* — 109 99 95* 98 87* — 88 Normal — Normal Normal Normal Normal Normal — Normal Normal — Normal Normal Normal Normal Normal — Impaired _ — Normal Impaired Normal Normal Impaired _ Impaired Impaired — Normal Normal Normal Normal Impaired _ Impaired Left neglect, associated posterior cerebral infarct Associated posterior cerebral infarct L 78* — 116 106 117 103 120 — 89 L L L R R 67* 69* — 100 106 71* 75* — 76* 76* Impaired Impaired Impaired Normal Normal Impaired Impaired Impaired Normal Normal Impaired Impaired Impaired Impaired Impaired Impaired Impaired Impaired Impaired Disorientation Disorientation Disorientation Disorientation, dysarthria R+L 99* 78* 81* 75* 61* 65* Normal Normal Normal Impaired Impaired Impaired Impaired Impaired Impaired Impaired Impaired Impaired Disorientation Left neglect, disorientation Normal Normal Impaired"*""*" Normal Impaired** Impaired** Normal Normal Normal Normal Impaired Impaired Impaired Impaired Impaired Impaired Impaired Impaired Impaired Impaired Impaired Impaired Impaired — Impaired Impaired Normal Normal Normal Normal Normal Normal R R+ L Lateral Ihulumic/posterior limb of internal capsule infarction 87* 95* 58/M 12 18 R 63/F 80* 12 106 19 R 81 87 8 20 72/M L 47/M 12 112 HI 21 L 8 114 67/M 104 22 L 12 58/M 93* 88* 23 L 84 87 24 62/F 9 L 92 87 67/M 12 25 L — Other Associated posterior cerebral infarct z o z X > m 2 o 70 Associated posterior cerebral infarct o o H X z Dysarthria Dysarthria. Slow reading of paragraphs Dysarthria, occasional paraphasia Dysarthria Dysarthria * Below estimated IQ based on educational level and occupational background; — not available; *"+ mild difficulties with some tasks in Multilingual Aphasia Battery but not conforming to aphasic syndrome. •n > o H 496 N. R. GRAFF-RADFORD AND OTHERS Anterolateral thalamic infarction. All 5 patients had normal sensory findings save for transient proprioceptive loss in Case 8; 3 had hemipareses and 2 visual field defects. One with a visual field defect had also had posterior cerebral artery infarction. This group had severe and wide-ranging neuropsychological deficits in language, temporal orientation, intellect, memory, and visual perception and constructional praxis when the left side was affected. With right anterolateral thalamic infarction, nonverbal intellect, visuoperceptual performance, constructional praxis and visual memory for designs were compromized, with preservation of verbal functions, such as verbal intellect, speech, language processing and shortterm verbal memory. All patients with left-sided lesions were disorientated; those with right-sided lesions were not. Case 10 illustrates this group. Medial thalamic infarction. The 3 patients in this group had no sensory deficit. One had a hemiparesis, none had visualfielddefects. In the initial stages, speech and language were intact but intellect, memory, visual perception, constructional praxis and temporal orientation were impaired. One case (No. 16) had a visual neglect syndrome. All 3 had supranuclear vertical gaze paresis. Case 15 is the example from this group. A 69-year-old right-handed retired school teacher suddenly developed confusion, dysarthria and lethargy. Initially she had a waxing and waning level of attention. Examination showed a supranuclear vertical gaze paresis but normal sensory and motor function. Neuropsychological testing six weeks later revealed a euphoric woman who was disorientated for time and place and strikingly amnesic. Her memory was poor for both anterograde and retrograde material. Speech, naming, repetition and writing were normal and there was no evidence of aphasia. Verbal intellect was well preserved while nonverbal intellect was clearly impaired. Visual perceptive discrimination of unfamiliar faces, visuospatial judgement, constructional praxis and right-left orientation on confrontation were defective. On follow-up examination 18 months later she had improved in many respects but her amnesia remained. Intellectual, perceptual and constructional performances were within or very close to the normal range, but memory had improved little. Lateral thalamusjposterior limb of internal capsule. In this group, all 8 patients had a hemiparesis, 4 had diminished pinprick, 1 decreased light touch, but none had impaired proprioception or vibration sense. Only one had a visual field defect. Speech and language were usually spared but controlled oral word association and Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 A 67-year-old right-handed man suddenly developed right-sided weakness, difficulty in speaking and confusion. Examination revealed normal sensation for pinprick, light touch and vibration sense, but mild proprioceptive loss in the right hand. The latter had resolved by six weeks. There was a right facial paresis for emotional movements and a right hemiparesis. Neuropsychological testing six weeks after the event revealed pervasive impairments in speech and language characteristic of thalamic aphasia (Graff-Radford and Damasio, 1984), verbal and nonverbal intelligence, anterograde as well as retrograde memory for verbal and nonverbal material, temporal orientation, visual perception and constructional praxis. The patient continued to improve and was retested two years later. At this time speech and language testing were improved but the patient was still mildly aphasic; intellect, visual perception, praxis, orientation and memory had also improved but were still in the borderline to low average range. NONHAEMORRHAGIC THALAMIC INFARCTS 497 articulation were affected in about half of the cases. Right-sided lesions could disrupt nonverbal intellect, visual perceptual or spatial discrimination and shortterm visual memory for designs. In contrast, the left-sided lesions rarely affected IQ levels, visual perception and constructional praxis. Short-term verbal and visual memory was compromised in about half of the cases; however, these deficits were not associated with any clinically significant amnesia. Case 23 illustrates this group. Somatosensory Evoked Responses Major findings are tabulated in the Appendix. Posterolateral thalamic infarction. This group will again be subdivided into those patients with infarction involving the complete geniculothalamic territory and those with involvement of only part of the territory. Complete geniculothalamic territory. Unilateral median nerve stimulation of the affected arm resulted in no response after PI4 in both hemispheres. Unilateral median nerve stimulation of the nonaffected arm or bilateral median nerve stimulation caused a normal response in the unaffected hemisphere and an N17 wave over the affected hemisphere (fig. 7A, B, C). In a further patient with a clinically identical picture to this group, CT scan was negative but SERs were of the same pattern as this group. Partial geniculothalamic territory. We have 5 examples in this group. In 2, N17, N19 and N20 were normal but N29, N32 and N34 and N60 waves were delayed (fig. 8A, B). One patient with the clinical findings of a thalamic lacune (unilateral loss of pinprick sensation) but negative CT scan had an identical SER to this group. One patient had a normal SER. In 2 the SER could not be interpreted because 1 had a severe peripheral neuropathy (Case 7) and the other could not tolerate adequate stimuli (Case 4). Anterolateral thalamic infarction. In this group there were 4 patients on whom SERs were performed. In 3 there was a delay of N17, N19 and N20 over the frontal, central and parietal areas, respectively. Two of the cases had a delay of N29, N32 and N34; in all, N60 was delayed in the central region but the N60 response varied in each case over the parietal area (fig. 9A, B). In 1 case there was only a delay of N60. Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 On July 4,1982 a 58-year-old right-handed man suddenly developed a right hemiparesis and slurred speech. On examination, the weakness was greater in his face and arm than his leg. Sensory examination for pain, light touch, proprioception and vibration was intact bilaterally. Neuropsychological examination revealed an alert and well-orientated man without aphasia. Speech was fluent but dysarthric. Naming, repetition, spelling and aural comprehension were normal, but controlled oral word association was impaired and reading comprehension of paragraphs was below expectations. Verbal and nonverbal intellect appeared to have declined modestly, as did short-term memory for verbal material and geometric designs. Facial discrimination, judgement of line orientation and constructional praxis were normal. Neuropsychological examination one year after onset showed improved intellectual performance, controlled oral word association, reading and shortterm memory. Mild residual deficits in Rey's verbal memory test and in visual memory for designs were observed but he was not clinically amnesic. Some improvement in the strength of the right side was apparent. An arteriogram showed an absent left anterior choroidal artery. 498 N. R. GRAFF-RADFORD AND OTHERS Medial thalamic infarction. In the 3 patients in this group the most consistent delay noted was in the N60 (fig. 10A, B). Lateral thalamic Iposterior limb of internal capsule infarction. Of the 8 patients in this group, 7 had SERs and all had a normal N17 and a delayed N60. However, waves N19, N20, N29, N32, and N34 were variable, making this the most heterogenous group in relation to the SER findings. An example is shown in fig. 1 1A, B. DISCUSSION 100 ms FIG. 7A. Left unilateral stimulation in Case 1. Note the complete absence of evoked responses after brainstem PI 4. B,rightunilateral stimulation in Case 1. Note the normal contralateral responses over the left hemisphere and some ipsilateral responses over the right hemisphere, c, two trials of bilateral stimulation. They appear identical to those from right unilateral stimulation. Using only bilateral stimulation would have been deceptive in this case because of the ipsihemisphere response fromright-sidedstimulation. Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 Although vascular territories are somewhat variable from subject to subject and the extent of infarction within a given territory is also variable, the CT findings in our patients can be grouped according to four fairly distinct loci. Similarly, the findings noted above indicate that the clinical, neuropsychological and electrophysiological profiles associated with these anatomical groups are relatively distinct. 100 ms C, t 100 ms Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 T,.' 499 INFARCTS NONHAEMORRHAGIC THALAMIC 500 N. R. GRAFF-RADFORD AND OTHERS Right Left Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 O, lOOms + 5 100 ms 100 ms FIG. 8A. Right and left unilateral median nerve stimulation in Case 8. The responses to left unilateral stimulation are normal but there is a definite delay in N29 (black star) and a possible delay in N32 and N60 (hollow stars) with right-sided stimulation. B, two trials of bilateral stimulation in Case 8 which confirm the left hemisphere delay in N32 and N60. Note that the configuration of N32 and N60 at C 3 are almost identical to those with unilateral stimulation. NONHAEMORRHAGIC THALAMIC INFARCTS Right 501 Left 100 ms 100 ms 100 ms FIG. 9A. Right and left unilateral median nerve stimulation in Case 10. Note the normal responses over the right hemisphere and the definite delay in N17, N19 and N20 (solid black stars) and probable delay in N29, N32, N34 and N60 (hollow stars) on the left, B, two trials of bilateral stimulation in Case 10 which confirm both the earlier and later wave delays seen with unilateral stimulation. Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 5 UV 5 uV 502 N. R. GRAFF-RADFORD AND OTHERS Left 5/iV + 5MV 100 ms 100 ms 100 ms FIG. 10A. Right and left unilateral stimulation in Case 15 showing a possible delay at N34 at P4 and N60 at C, (open stars). N60 at P 4 may also be delayed, B, two trials of bilateral stimulation in Case 15 which confirm the delays in N34 at P 4 and N60 at C4 suspected with unilateral stimulation. However, the possible delay of N60 at P4 could not be confirmed. Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 O NONHAEMORRHAGIC THALAMIC INFARCTS Right 503 Left 100 ms 100 ms FIG. 1 1A. Right and left unilateral stimulation in Case 18 showing normal evoked responses over the left hemisphere and a delay in N19, N20 and N29 (black star), and a possible delay in N32 and N60. In contrast to Case 10 (fig. 9A, B) note that N17 is not delayed, B, two trials of bilateral stimulation which confirm the delays at N19, N20, N32 and N60. Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 100 ms 504 N. R. G R A F F - R A D F O R D AND OTHERS Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 Anatomical Findings The geniculothalamic artery is a branch of the posterior cerebral artery and is often occluded with posterior cerebral occlusion (Goto et al., 1979). We have seen 11 patients whom we believe have had infarctions in this distribution; 9 had CT confirmation while 2 had negative CTs. The latter can be explained by the timing of CTs which were performed at a time interval after the infarction when lesions may not be evident because of isodensity of the region of infarction. This point has been demonstrated with sequential CTs (Goto et al., 1979). The SERs and clinical findings are so similar to our other CT-confirmed cases that we think it is likely that there are similar lesions in the thalamus. There are other CT reports of lesions in this distribution with similar clinical findings (Goto et al., 1979; Beasley et al., 1980). There may or may not be concomitant posterior cerebral infarctions, depending upon the collateral circulation from the middle and anterior cerebral arteries (Goto et al., 1979). The geniculothalamic artery commonly supplies the following thalamic nuclei: pulvinar, ventroposterolateral, ventroposteromedial, part of the centrum medianum and the medial geniculate (fig. 1B) and has also been reported to supply part of the following nuclei: dorsomedial, posterolateral, reticular, parafascicular and lateral geniculate (Schlesinger, 1976). An important point is that this artery supplies the primary sensory nuclei. Thalamic lacunes with a 'pure hemisensory loss' are often not seen on CT (Fisher, 1982). Pathologically they are situated in the primary sensory nuclei of the thalamus. We have demonstrated 5 such cases on CT and the location is the posterolateral thalamus but, as expected, the lesions are not as extensive as in complete geniculothalamic occlusion. The tuberothalamic artery is a branch of the posterior communicating artery. We have now seen 5 patients with thalamic lesions in this distribution. Two of our cases suffered thalamic infarction after clipping of giant posterior communicating artery aneurysms. The CT appearance was typical of this group. This supports the idea that occlusion of the tuberothalamic artery, a branch of the posterior communicating artery, may cause infarction in the anterolateral quadrant of the thalamus. Archer et al. (1981), using CT with sagittal reconstruction and the Schaltenbrand and Wahren Atlas (1977), listed the nuclei destroyed in a case with an anterolateral thalamic infarction. There was an excellent overlap between Schlesinger's (1976) definition of the nuclei that should be supplied by the tuberothalamic artery and the findings of Archer et al. as to the actual nuclei involved in the thalamic infarction they studied. Our 5 patients with tuberothalamic infarctions had CT images similar to those in the case of Archer et al. (1981). Further confirmation that we are dealing with tuberothalamic artery territory comes from the CT examination in patients who have undergone stereotactic ventrolateral thalamotomy (Murayama et al., 1982). The ventrolateral thalamic nucleus (VL) is in the tuberothalamic arterial territory. Stereotactic lesions in VL have a similar (but less extensive) CT appearance to our tuberothalamic infarction group. NONHAEMORRHAGIC THALAMIC INFARCTS 505 Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 The deep interpeduncular profunda artery is a branch of the basilar root of the posterior cerebral artery. One artery may supply the medial thalamus bilaterally or unilaterally. Two of our patients in this group had bilateral medial thalamic infarctions. We have studied 1 patient with deep interpeduncular profunda artery occlusion by CT and subsequently at postmortem (Graff-Radford et al., 1985). The CT showed the typical medial thalamic low densities. There have been other case reports with lesions in this location (Mills and Swanson, 1978; Smith and Laguna, 1981) and the CT findings are very similar to our cases. Stereotactic thalamotomies for pain centred on the centrum medianum (a nucleus invariably supplied by this artery) caused a similar CT appearance to infarction in this arterial territory (Colombo et al., 1981). As the tuberothalamic artery may be small or even absent in certain patients (Lazorthes et al., 1976), the interpeduncular profunda may supply the dorsomedial, ventrolateral and anteroventral thalamic nuclei and, in addition, has been described as supplying the dorsolateral, posterolateral, ventroposteromedial and ventroposterolateral thalamic nuclei (fig. 5B). In the majority of cases, part of the dorsomedial, centrum medianum, and parafascicular nuclei are its main territory of supply (Schlesinger, 1976; Castaigne et al., 1981). The posterior limb of the internal capsule is irrigated by the anterior choroidal artery (Lazorthes et al., 1976). Takahashi et al. (1980) have demonstrated the CT findings of arteriographically proven anterior choroidal artery lesions. They wrote that the most vulnerable areas are the globus pallidus and the posterior limb of the internal capsule. Based on the CT analysis of our patients, we concur. The anterior choroidal artery territory, as depicted on CT templates, corresponds to the area of infarctions seen in Cases 18 to 25 (Damasio, 1983). We have examined a 48-year-old woman who had her right anterior choroidal artery sacrificed during the clipping of a giant internal carotid/posterior communicating aneurysm. The resulting subcortical infarction involved the medial globus pallidus, posterior internal capsule and lateral thalamus. The location of the lesion in this verified case of anterior choroidal artery interruption helps to validate the interpretation we offer for other patients in the group who had infarctions in the same region. The anterior choroidal artery supplies part of the amygdala, medial temporal lobe, globus pallidus, posterior limb of the internal capsule and lateral thalamus (Carpenter et al., 1954; Schlesinger, 1976). The thalamic nuclei that have been reported to receive blood supply from the anterior choroidal artery are the lateral geniculate, reticular, pulvinar, and ventroposterolateral (Schlesinger, 1976) (fig. 6B). Only 2 patients in this group had arteriography. One of them had an absent anterior choroidal artery and in the other the artery could not be traced beyond the plexal point. At this time we believe many of these lesions are due to occlusions of branches of the anterior choroidal artery. Some patients have complete anterior choroidal artery occlusion. 506 N. R. G R A F F - R A D F O R D AND OTHERS Clinical and Neuropsychological Findings Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 As noted, these different anatomical loci were associated with identifiable clinical and neuropsychological profiles. Geniculothalamic infarction may cause the Dejerine-Roussy (1906) syndrome. Two of our patients with lesions in this territory had this syndrome and a third without CT confirmation also developed it. But the clinical hallmark of a lesion involving the complete geniculothalamic distribution is loss of sensation in all primary modalities. At present we do not know why some with lesions in this territory lead to dysaesthesia and some do not. There do not appear to be major cognitive deficits, aphasia, apraxia or amnesia associated with this type of lesion, although a few abnormalities can be found on more demanding neuropsychological tests such as copying of a complex figure and controlled oral word association. Concomitant posterior cerebral occlusion can be associated with a hemineglect syndrome and more extensive neuropsychological deficits. A geniculothalamic lacune in the primary sensory nuclei causes the 'pure hemisensory loss'. We have found little or no neuropsychological deficit to accompany such lesion unless there is concomitant posterior cerebral artery occlusion. We have described elsewhere the clinical presentation and neuropsychological profiles of the anterolateral thalamic group (Graff-Radford et al., 1984). We now add another patient. These patients may have a hemiparesis and an emotional facial paresis but they do not usually have sensory deficits. In those with left-sided lesions, severe language deficits with aphasia are present (Graff-Radford and Damasio, 1984), along with marked decline in intellectual performance, visual perception, constructional praxis, memory and temporal orientation. Those with lesions on the right side have preserved verbal intellect, verbal memory, and speech and language. However, nonverbal intellect, visual perceptual discrimination, spatial judgement, visual memory for designs and constructional praxis are disturbed. Over a prolonged period the patients improve significantly but the pattern of deficits remains (Graff-Radford et al., 1984). We believe that impairment in memory performance exhibited by this group may in part be secondary to their difficulties with language and visuospatial processing. The medial thalamic group have no sensory or visual field deficit but may have a motor deficit. The latter could result from occlusion of a branch of the interpeduncular profundus artery to the cerebral peduncles. They also have vertical gaze paresis, probably related to infarction of the rostral part of the medial longitudinal fasciculus (Biittner-Ennever et al., 1982). In the early stages of damage, intellect, memory, visuospatial processing and orientation are all affected in this group but no speech disturbance or aphasia is evident in our 3 cases. When these patients are followed clinically and with serial neuropsychological testing, amnesia becomes their main cause of disability as intellect, visual perception and spatial processing improve. The severity of the initial neuropsychological deficits may in part be secondary to attentional difficulties. The patients can be drowsy in the beginning NONHAEMORRHAGIC THALAMIC INFARCTS 507 Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 with waxing and waning levels of attention. In the chronic phase, these patients could be described as having an amnestic syndrome. For two years after her stroke we have followed Case 15 who had bilateral medial thalamic lesions, larger on the right than the left. She has continued to exhibit anterograde memory impairment, most clearly demonstrated by her impaired performances on the Rey Auditory Verbal Learning Test and recall of the ReyOsterrieth Figure. Delayed recall (30 min) of the paragraph stories of the Wechsler Memory Scale was zero. These deficits were of clinical significance. Retrograde memory was also impaired. She scored 12 out of 48 on the Famous Faces portion of the Boston Remote Memory Test. Phonemic and semantic cues were of little help, but providing the first name to her led to improved recall. No temporal gradient of amnesia was evident. We consider that this patient's amnesia was similar to Case A.V. (Butters, 1984) who had a thalamic tumour and both anterograde and retrograde amnesia components, the latter without temporal gradient. However, our Case 15 is different from Case H.M. (Scoville and Milner, 1957; Penfield and Milner, 1958) in that H.M.'s anterograde deficit was more severe and there was no retrograde component before the age of 16. In D.R.B. (Damasio etal., 1985a), who suffered severe bilateral temporal lobe damage and basal forebrain damage due to herpes encephalitis, both the anterograde and retrograde components of memory are more impaired than in our Case 15. In cases of amnesia related to basal forebrain lesions (Damasio et ah, 19856), both the anterograde and retrograde amnesia is usually less severe. N.A. (Squire and Moore, 1979), who had a stab wound through the nostrils up to the left medial thalamus, had less retrograde amnesia than this patient. Our patient exhibited some similarity to patients with Korsakoff's syndrome in that there were both anterograde and retrograde components to the amnesia. However, the lack of temporal gradient in the retrograde component of the amnesia is dissimilar to the Korsakoff's patient reported by some (Albert et al., 1979), but not other researchers (Sanders and Warrington, 1971). Our evaluation of this patient's amnesia is based on the above-mentioned tests. A more extensive battery of tests may help further to characterize the amnesia. All the patients in the lateral thalamus/posterior limb of internal capsule group had a hemiparesis. Some of these patients would previously have been classified as 'pure motor hemiparesis'. On detailed testing, however, a few had minor sensory deficits; one had a transient visual field deficit—the anterior choroidal does supply the optic tract and the lateral geniculate body partially. Most had some neuropsychological deficit. Although the left-sided lesions do not cause an aphasia, dysarthria is common and there can be mild language processing difficulties such as in controlled oral word association or reading comprehension of paragraphs. Assessment of short-term verbal and visual memory may reveal mild deficits in some cases, but these do not appear to be related to clinically significant amnesia. Patients usually have difficulty with short-term verbal memory tasks such as the Rey Auditory Verbal Learning and the Wechsler Memory Scale. It is possible that the 508 N. R. GRAFF-RADFORD AND OTHERS Electrophysiological Findings We found that a loss of both proprioception and pain sensation from thalamic infarction was indicative of a complete geniculothalamic infarction. This corresponded to an absence of SERs of all waves after PI 4 (i.e. Nl 7, Nl 9, N20, N29, N32, N34 and N60) when the affected hand was stimulated alone. In contrast, a lacune in the same arterial territory caused impaired pain sensation but preserved proprioception. This corresponded to an intact N17, N19 and N20, but a delayed or absent N32. It is tempting to conclude that N19 reflects the medial lemniscus (proprioception) pathway and N32 the spinothalamic (pain) pathways, but with so few cases this conclusion is not well founded. A more general observation is that a partial lesion in the primary sensory nuclei (VPL and VPM) affects pain sensation and may correlate with N32 delay, whereas a complete lesion affects all modalities of sensation and correlates with a unilateral absence of all evoked responses after PI 4. Waves N17, N19 and N20 probably reflect sensory pathways (Ervin and Mark, 1964; Celesia, 1979). It is reasonable to ask why these waves are delayed in tuberothalamic infarction when the primary sensory nuclei are preserved and sensation is intact. We believe this may reflect a disturbance of the gating mechanism from destruction of part of the reticular nucleus of the thalamus, a nucleus irrigated in part by the tuberothalamic artery (Schlesinger, 1976). Yingling and Skinner (1977) have shown in the cat that stimulating the inferior thalamic peduncle to the frontal lobe or the reticular nucleus situated anterior to the ventroanterior thalamic nucleus diminishes sensory evoked potentials in visual, auditory and tactile modalities by interfering with the gating mechanism. Thus damage in this area may be the reason for the abnormal SERs. The lateral thalamic/posterior capsule group may also show a delay of N19, N20, N32 and N34 with minimal sensory deficit. Again, it is possible that the thalamic reticular nucleus is involved, this time adjacent to VPL and VPM, supporting the idea that SER delays can result from disturbed gating secondary to thalamic reticular nucleus destruction. Skinner and Yingling Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 mild language deficits exhibited by these patients may affect their verbal memory performance or vice versa. In addition, short-term visual memory testing (Benton Visual Retention Test and Recall of the Rey-Osterrieth Complex Figure) often falls in the defective range. This cannot easily be explained by a visuoperceptive disturbance because the patients usually perform well on tests of visual perception. The 2 cases with right-sided posterior limb/lateral thalamic lesions showed consistently defective visual perception, short-term visual memory for designs (perhaps secondary to the visuoperceptive disturbance) and nonverbal intellect performances. In contrast, language, short-term verbal memory and verbal intellectual abilities were normal. It must be remembered that the anterior choroidal artery supplies part of the medial temporal lobe. Thus the mild memory deficits, even if they are not secondary to verbal and visual perceptual abnormalities, are not necessarily related to the thalamic component of the infarction. NONHAEMORRHAGIC THALAMIC INFARCTS 509 Conclusions The clinical, neuropsychological and physiological features that characterize each anatomically defined group are as follows. 1. Patients with complete geniculothalamic infarction usually have multiple Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 (1977) also showed that, in the cat, stimulation of the reticular nucleus adjacent to VPL causes a delay in SERs in the tactile modality only. Nl 7 is recorded bifrontally with unilateral median nerve stimulation, suggesting a central origin of this wave. Others have also reported this wave (Abbruzzese et al., 1978; Desmedt and Cheron, 1981; Eisen, 1982; Mauguiere et al., 1983). It is the earliest wave to be affected in the patients with geniculothalamic infarction, possibly indicating that it originates in the primary sensory nuclei (VPL and VPM). This is supported by direct recording in the primary sensory thalamic nuclei which shows a wave of a similar latency, 17.5 ms (Fukushima et al., 1976) and 17 ms (Ervin and Mark, 1964). Simultaneous recordings in the thalamus and over the postcentral gyrus show that the cortically recorded wave is 2 ms later than that in the thalamus (Celesia, 1979). The origin of N19 is a source of controversy. Some attribute it to the thalamus (Goldie et al., 1981; Chiappa and Ropper, 1982) and others to thalamocortical pathways or cortex (Allison et al., 1980; Mauguiere et al., 1982). If N17 is thalamic, which seems possible from the above data, then N19 is likely to be generated more distally in the thalamocortical pathway or in the cortex itself. We believe that it is important to record in frontal, central and parietal scalp regions with unipolar electrodes. As can be appreciated from our data, we have noted dissociations between waves recorded at these areas with the thalamic infarctions. For example, we have seen a normal Nl 7 response in the frontal region but delayed N19 and N20 in the central and parietal areas (Case 18) or a delay in N32 at the central area but a normal N29 and N34 in the frontal and parietal regions (Case 20). Because of these dissociations and because the latencies recorded in the different scalp areas are consistently different, we believe that the origin of the waves recorded at various scalp areas is probably different. Most importantly, from a clinical standpoint, by recording in frontal, central and parietal areas, and by obtaining the patterns of SER abnormalities caused by different thalamic lesions, we are able to use the SER as a diagnostic aid in documenting and localizing lesions. Thus, in 2 cases, patients had normal CTs but distinctive SER patterns. In 1 case, the patient had a thalamic pain syndrome and absent SERs after PI4. This corresponded to the posterolateral thalamic infarction group. As mentioned earlier, Goto et al. (1979) have shown with serial CTs in posterolateral thalamic infarction that the infarction may 'apparently disappear' from the CT scan, which may explain our not seeing this lesion on CT. In the other case, a unilateral loss of response to pinprick corresponded to a contralateral delay in N32 and N34. Fisher (1978) has also reported that patients with pure hemisensory loss may have normal CT scans. In these 2 cases the somatosensory evoked responses were clinically useful in lesion localization. 510 N. R. GRAFF-RADFORD AND OTHERS Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 modality sensory loss, may have associated dysaesthesia, hemiparesis, hemianopia and choreiform movements. Neuropsychological evaluation reveals few abnormalities unless there is concomitant posterior cerebral infarction. With median nerve stimulation there is a complete absence of evoked responses after the brain-stem wave (P14) in the hemisphere ipsilateral to the lesion. A partial geniculothalamic lesion causes contralateral sensory loss involving part of the body for the modalities of pain and light touch but not proprioception or vibration. Occasionally there may be dysaesthesia but there are no motor signs. There may be a visual field defect if there is concomitant posterior cerebral infarction. Neuropsychologically there are few deficits unless there is posterior cerebral infarction. The most common somatosensory evoked response pattern is delay of the N32, N34 and N60 waves with normal N17, N19 and N20 responses. 2. Tuberothalamic infarction is often associated with facial paresis for emotional movement. Sometimes there is also a hemiparesis and visualfielddefect. Only rarely is there sensory loss. This group is severely impaired on neuropsychological evaluation with intellect, speech, language, all spheres of memory, visual perception and orientation all being affected with left-sided lesions. In contrast, right-sided lesions spare verbal intellect, speech, language and short-term verbal memory. The intellectual, perceptual, constructional and language defects improve, but a similar pattern of impairment remains. These patients often show a delay in N17, N19 and N20 as well as the later waves. Such abnormalities may also improve. 3. Interpeduncular profundus infarction often causes drowsiness in the early stages. Commonly there is supranuclear vertical gaze paresis and, on occasion, hemiparesis. These patients initially have significant cognitive impairment, especially affecting attention span, orientation, intellect, memory and visual perception, but language is preserved. Improvement is the rule, although the residual amnesic syndrome with anterograde and retrograde components may persist. Following median nerve stimulation there is often a delay in the N60 with preservation of the earlier waves. 4. Anterior choroidal infarction almost invariably causes a contralateral hemiparesis. Sometimes there is sensory loss for pinprick or light touch, but not proprioception or vibration. Dysarthria is common. Occasionally there is a visual field defect. Neuropsychological testing reveals that about half of the cases with leftsided lesions may exhibit mild deficiencies in memory and controlled oral word association, but aphasia, apraxia and amnesia are not evident. Right-sided lesions, on the other hand, most often affect only aspects of visual perception and visual memory for designs. Intellect, verbal memory and language are usually well preserved. Sensory evoked responses are variable. N60 is commonly delayed, however. Our findings indicate that when patients with nonhaemorrhagic thalamic infarctions are grouped according to the location of their lesions, as depicted by CT, the profiles of clinical, neuropsychological and somatosensory evoked responses reveal fairly consistent and identifiable symptom clusters. The recognition of these NONHAEMORRHAGIC THALAMIC INFARCTS 511 clusters may assist neurologists in the assessment of such patients, and may contribute to the understanding of these complex subcortical structures. But our findings must be tempered by the knowledge that arterial territories are subject to some variation. The complete validation of our results must await further postmortem evaluations. ACKNOWLEDGEMENTS Supported by Program Project Grant POI NS 19632-01 from NINCDS. We acknowledge Steven Cornell for help with obtaining CT images. We thank Frank Sindelar for his drawings of thalamic arterial territories and Jean Hulme for assistance in the preparation of this paper. ABBRUZZESE M, FA VALE E, LEANDRI M, RATTO S (1978) New subcortical components of the cerebral somatosensory evoked potential in man. Ada Neurologica Scandinavica, 58, 325-332. ALBERT M S, BUTTERS N, LEVIN J (1979) Temporal gradients in the retrograde amnesia of patients with alcoholic Korsakoff's disease. Archives of Neurology, Chicago, 36, 211-216. ALLISON T, GOFF W R, WILLIAMSON P D, VANGILDER J C (1980) On the neural origin of early com- ponents of the human somatosensory evoked potential. In: Clinical Uses of Cerebral, Brainstem and Spinal Somatosensory Evoked Potentials. Progress in Clinical Neurophysiology, Volume 7. Edited by J. E. Desmedt. Basel: Karger, pp. 51-68. ARCHER C R, ILINSKY I A, GOLDFADER P R, SMITH K R (1981) Case report. Aphasia in thalamic stroke: CT stereotactic localization. Journal of Computer Assisted Tomography, 5, 427-432. BEASLEY B A, DAVENPORT R W, NAKAGAWA H (1980) Computerized tomography localization of the thalamic syndrome—a case report. Computerized Tomography, 4, 243-245. BENTON A L (1974) Revised Visual Retention Test. 4th Edition. New York: Psychological Corporation. BENTON A L, HAMSHER K (1978) Multilingual Aphasia Examination. Iowa City: Department of Neurology, University of Iowa Hospitals. BENTON A L, HAMSHER K, VARNEY N, SPREEN O (1983) Contributions to Neuropsychological Assess- ment: A Clinical Manual. New York: Oxford University Press. BENTON A L, VAN ALLEN M W (1968) Impairment in facial recognition in patients with cerebral disease. Cortex, 4, 344-358. BENTON A L, VARNEY N R, HAMSHER K (1978) Visuospatial judgment. Archives of Neurology, Chicago, 35, 364-367. BUTTERS N (1984) Alcoholic Korsakoff's syndrome: an update. Seminars in Neurology, 4, 226-244. BUTTNER-ENNEVER J A, BUTTNER U, COHEN B, BAUMGARTNER G (1982) Vertical gaze paralysis and the rostral interstitial nucleus of the medial longitudinal fasciculus. Brain, 105, 125-149. CARPENTER M B, NOBACK C R, MOSS M L (1954) The anterior choroidal artery: its origins, course, distribution, and variations. Archives of Neurology and Psychiatry, Chicago, 71, 714-722. CASTAIGNE P, LHERMITTE F, BUGE A, ESCOUROLLE R, HAUW J J, LYON-CAEN D (1981) Paramedian thalamic and midbrain infarcts: clinical and neuropathological study. Annals of Neurology, 10, 127-148. CELESIA G G (1979) Somatosensory evoked potentials recorded directly from human thalamus and SmI cortical area. Archives of Neurology, Chicago, 36, 399-405. CHIAPPA K H, ROPPER, A H (1982) Evoked potentials in clinical medicine. New England Journal of Medicine, 306, 1205-1211. COLOMBO F, DETTORI P, PINNA V, BENEDETTI A (1981) Stereotactic thalamic lesions studied by CT scanner. Ada Neurochirurgica, 57, 205-212. Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 REFERENCES 512 N. R. GRAFF-RADFORD AND OTHERS DAMASIO A R, ESLINGER P J, DAMASIO H, VAN HOESEN G W, CORNELL S (1985a) Multimodal amnestic syndrome following bilateral temporal and basal forebrain damage. The case of patient D.R.B. Archives of Neurology, Chicago, 42, 252-254. DAMASIO A R, GRAFF-RADFORD N R, ESLINGER P J, DAMASIO H, KASSELL N (1985ft) Amnesia GOLDIE W D, CHIAPPA K. H, YOUNG R R, BROOKS E B (1981) Brainstem auditory and short-latency somatosensory evoked responses in brain death. Neurology, New York, 31, 248-256. GOTO K, TAGAWA K, UEMURA K, ISHII K, TAKAHASHI S (1979) Posterior cerebral artery occlusion: clinical, computed tomographic, and angiographic correlation. Radiology, 132, 357-368. GRAFF-RADFORD N R, DAMASIO A R (1984) Disturbances of speech and language associated with thalamic dysfunction. Seminars in Neurology, 4, 162-168. GRAFF-RADFORD N R, ESLINGER P J, DAMASIO A R, YAMADA T (1984) Nonhemorrhagic infarction of the thalamus: behavioral, anatomic, and physiologic correlates. Neurology, Cleveland, 34, 14-23. GRAFF-RADFORD N R, SCHELPER R L, ILINSKY I A, DAMASIO H (1985) CT and postmortem study of a non-hemorrhagic thalamic infarction. Archives of Neurology, Chicago. In press. LAZORTHES G, GOUAZE A, SALAMON C (1976) Vascularisation et Circulation de I'Encephale. Pans: Masson. MAUGUIERE F, BRUNON A M, ECHALLIER J F , COURJON J (1982) Early somatosensory evoked potentials in thalamo-cortical lesions of the lemniscal pathways in humans. Advances in Neurology, 32,321 -338. MAUGUIERE F, DESMEDT J E, COURJON J (1983) Neural generators of N18 and P14 far-field somatosensory evoked potentials studied in patients with lesion of thalamus or thalamo-cortical radiations. Electroencephalography and Clinical Neurophysiology, 56, 283-292. MILLS R P, SWANSON P D (1978) Vertical oculomotor apraxia and memory loss. Annals of Neurology, 4, 149-153. MURAYAMA Y, SOGABE K, MATSUMOTO K (1982) CT appearance of thalamic lesions in stereotactic surgery. Applied Neurophysiology, 45, 399-403. OSTERRIETH P A (1945) Le test de copie d'une figure complexe. Archives de Psychologie, 30, 205-353. PENFIELD W, MILNER B (1958) Memory deficit produced by bilateral lesions in the hippocampal zone. Archives of Neurology and Psychiatry, Chicago, 79, 475-497. REY A (1941) L'examen psychologique dans les cas d'encephalopathie traumatique (les problemes). Archives de Psychologie, 28, 286-340. REY A (1964) L'examen Clinique en Psychologie. Second Edition. Paris: Presses Universitaires de France. Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 following basal forebrain lesions. Archives of Neurology, Chicago, 42, 263-271. DAMASIO H (1983) A computed tomographic guide to the identification of cerebral vascular territories. Archives of Neurology, Chicago, 40, 138-142. DEJERINE J, ROUSSY G (1906) Le syndrome thalamique. Revue Neurologique, 14, 521-532. DESMEDT J E, CHERON G (1981) Non-cephalic reference recording of early somatosensory potentials to finger stimulation in adult or aging normal man: differentiation of widespread N18 and contralateral N20 from the prerolandic P22 and N30 components. Electroencephalography and Clinical Neurophysiology, 52, 553-570. EISEN A (1982) The somatosensory evoked potential. Canadian Journal of Neurological Sciences, 9, 65-77. ERVIN F R, MARK V H (1964) Studies of the human thalamus: IV. Evoked responses. Annals of the New York Academy of Sciences, 112, 81-92. FISHER C M (1978) Thalamic pure sensory stroke: a pathologic study. Neurology, New York, 28, 1141-1144. FISHER C M (1982) Lacunar strokes and infarcts: a review. Neurology, New York, 32, 871-876. FUKUSHIMA T, MAYANAGI Y, BOUCHARD G (1976) Thalamic evoked potentials to somatosensory stimulation in man. Electroencephalography and Clinical Neurophysiology, 40, 481-490. NONHAEMORRHAGIC THALAMIC INFARCTS 513 YAMADA T, KIMURA J, WILKINSON J T, KAYAMORI R (1983) Short- and long-latency median somato- sensory evoked potentials: findings in patients with localized neurological lesions. Archives of Neurology, Chicago, 40, 215-220. YAMADA T, DICKINS Q S, MACHIDA M, OISHI M, KIMURA J (1984a) SEPs to bilateral median nerve stimulation in man: method and clinical application. In: Evoked Potentials. Frontiers of Clinical Neuroscience. Edited by R. Q. Cracco and I. Bodis-Wollner. Baltimore: Williams and Wilkins. YAMADA T, KAYAMORI R, KIMURA J, BECK D O (19846) Topography of somatosensory evoked potentials after stimulation of the median nerve. Electroencephalography and Clinical Neurophysiology, 59, 29-43. YINGLING C D, SKINNER J E (1977) Gating of thalamic input to cerebral cortex by nucleus reticularis thalami. In: Attention, Voluntary Contraction and Event-Related Cerebral Potentials. Progress in Clinical Neurophysiology, Volume 1. Edited by J. E. Desmedt. Basel: Karger, pp. 70-96. {Received July 12, 1984. Revised October 11, 1984. Accepted November 6, 1984) Appendix on pp. 514-516 Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 ROBERTS M, HANAWAY J (1970) Atlas of the Human Brain in Section. Philadelphia: Lea and Febiger. SANDERS H I, WARRINGTON E K (1971) Memory for remote events in amnesic patients. Brain, 94, 661-668. SCHALTENBRAND G, WAHREN W (1977) Atlas for Stereotaxy of the Human Brain. Stuttgart: Thieme. SCHLESINGER B (1976) The Upper Brainstem in the Human. Berlin: Springer-Verlag. SCOVILLE W B, MILNER B (1957) Loss of recent memory after bilateral hippocampal lesions. Journal of Neurology, Neurosurgery and Psychiatry, 20, 11-21. SKINNER J E, YINGLING C D (1977) Central gating mechanisms that regulate event-related potentials and behavior: a neural model for attention. In: Attention, Voluntary Contraction and EventRelated Cerebral Potentials. Progress in Clinical Neurophysiology, Volume 1. Edited by J. E. Desmedt. Basel: Karger, pp. 30-69. SMITH M S, LAGUNA J F (1981) Upward gaze paralysis following unilateral pretectal infarction: computerized tomography correlation. Archives of Neurology, Chicago, 38, 127-129. SQUIRE L R, MOORE R Y (1979) Dorsal thalamic lesion in a noted case of human memory dysfunction. Annals of Neurology, 6, 503-506. TAKAHASHI S, KAWATA Y, UEMURA K (1980) CT findings on anterior choroidal artery occlusion. Rinsho Hoshasen, 25, 575-581. WECHSLER D A (1945) A standardized memory scale for clinical use. Journal of Psychology, 19, 87-95. WECHSLER D A (1981) Wechsler Adult Intelligence Scale (revised). New York: Psychological Corporation. N. R. GRAFF-RADFORD AND OTHERS 514 APPENDIX NEUROPSYCHOLOGICAL DATA Case No Age/sex Years educ Side of lesion Date of stroke -113 Verbal IQ 145 Perf. IQ 145 -8* 78' 57' —1 0 116 106 109 99 0 0 0 0 117 103 103 120 95* 98 84* 87* 05 02.84 21.06 84 -1 0 89 88 13 04.82 29.08 82 22.05.84 02 06 82 31.08 82 18.07 83 25.04 78 02 09 82 16.08 83 09.09 83 24 10.83 01.1183 -104' -81* -26' -61* -25* -1 -36' -1 0 0 67* 86* 71* 84* 69' 75* 94 95 100 102 106 76* 104 76* 0 97 78* 28.06 82 17 02 84 19.10.82 09 02 84 28 06.84 -102* -79* -102* -1 -1 99. 117 78' 81* 93 75' 93 61* 65" IT 18.1181 10.06 83 15 04 83 14.09 83 12.05 80 19.08 82 28 08 81 03.12.82 15.07 82 13 06.83 05 08 83 09.04 84 02.02 82 06 07 82 0 -2 -8* 0 0 0 0 0 0 0 1 _2 -2 0 95 106 81 84 111 104 104 106 93* 110 84 87 80* 87 81 112 103 114 114 88* 109 87 92 91 87 82 Dale tested Maximum scores Posterolateral thalamic infarction 1 48/F 10 — 78/F 2 3 64/F 12 4 58/F 12 66/M 5 16 R+ R+ R R R 21 10.82 27 10.82 03 03 83 18 06 83 06.06.80 19/M 13 R 2103 83 7 8 9 59/M 65/M 60/M 16 16 II R+ L L* 03 01.80 09 03 83 23 10 83 Anterolateral thalamic infarction 10 67/M 12 L 08 04.82 11 75/M 12 L 03 03.82 12 13 66/F 67/F 7 8 L* R 03 04 78 2608.82 14 60/F 8 R 19 08 83 Medial thalamic infarction 69/F 16 15 BLT 21 06.82 16 17 R BLT 0605 82 3101.82 74/F 81/M 12 12 Lateral thalamic/postenor limb of internal capsule infarction 18 58/M 12 R 1311 81 63/F R 12 01 06.83 19 8 20 1204 83 72/M L 47/M 21 22 67/M 12 8 L L 12 04.80 24 08 81 23 58/M n L 01 07.82 24 62/F 9 L 02 08.83 25 67/M + 12 L 3108.81 MAE naming MAE repetition MAE Token Test BDAE Reading SandP 60 14 44 10 12 60 56 12 13 44 44 50 14 8 18* 36* 44. 32* 16* 26* 22* 58 13 13 15 13 10 12 15 15 0' 30* 42 23' 15* 40 5* 6* 8 0 7* 7* 43 10 58 16 43 10 10 50 11 40 9 46 11 11 38' 9 58 50 50 56 10 25* 41 9 II 9 56 12 60 12 41 41 60 14 41 54 58 11 12 44 = Concomitant posterior cerebral artery cerebral infarct, * scores judged to be abnormal based on available norms as well as each patient's personal history, educational level, vocational background Abbreviations: Temp. Orient = Temporal Orientation Test, Verbal IQ = Wechslcr Verbal Intelligence Quotient; Perf. IQ = Wechsler Performance Intelligence Quotient; MAE = Multilingual Aphasia Examination, BDAE Reading S and P = Boston Diagnostic Aphasia Examination, reading of sentences and paragraphs; COW A = Controlled Oral Word Association of Multilingual Aphasia Examination; WMS = Wechsler Memory Scale, Ment. Cont.= Menial Control subtest of Wechsler Memory Scale (WMS), Parag. = Logical Memory (WMS); Paired Assoc. = Paired Associate Learning (WMS), Rey AVLT = Rey Auditory Verbal Learning Test (1st trial, 5th trial, recognition trial and delayed recall trial); BVRT (Coir Err.) = Benton Visual Retention Test (correct errors); R-O Complex Figure = Rey-Osterrieth Complex Figure; BRMT Famous Faces = Bcnton Remote Memory Test of Famous Faces; FRT = Facial Recognition Test, JLO = Judgment of Line Orientation; 3D Blocks = Three Dimensional Block Construction. 7 10 Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 6 28 10.82 _ 08.03 83 28 06.83 13.06 80 25.01.84 30 03.83 12 04 83 16.02 84 Temp Orient. 515 NONHAEMORRHAGIC THALAMIC INFARCTS N E U R O P S Y C H O L O G I C A L DATA (com.) ReyAVLT Trials (max. 15) MAE COW A WMS 58 143 37 22' 122 143 Mail Com. Parag Paired Assoc Con. Err 10 26+ 5 9* 8' 5 2 4 9* R-0 Complex Fig. BRMT Famous Drawing Memory Faces 36 36 21' •4* 120 99 94 143 13 4 7.6' 6.2' 11.2 114 11 9 10 7 15.7 14.4 12.2 130 15 5 32 101 116 132 10.4 5 4' 12.9 9.5 10 2 3' 10 9* 0 36 34 57* 76' 4 6' 79 5.8' 12.4 2' 4 4 14* 11' 7 15* 64' 5.8* 4.9' 1* 3 23' 12 3 6 4 2' 4 13' 6 10* 10* 7 16* 34 191 35 111' 1* 4 20* 11* 21' 36 1' 5' 2* 3 17' 12 32 3' 5 3 3 8 6 4 5 2' 4* 3* 3* 15* 10' 17' 13 2 9' 7 8 13* 8' 12* 11* 10' 10* 25 10* 7* 17' 1* 38 17' 46 9.9 19.1 11 3 16.0 99 12.0 10 3 6.2* 80 7.7 5.8* 4.6' 7.8 129 11 2 19.0 13 7 106 10 5 104 43 124 9.2 99 10.1 12.6 14 7 17* 41 14* 16* 24* 62' 101 62' 79' 89 10 1 14 8 76 82 11.8 5.2' 10.9 6 7' 6 7* 8.5 6 5' 9.2 5.4* 7.4' 94 46 15' 30 143 73 74 95 101 93 103 77' 108 95 92 93 94 11 8 64 7.8 85 14 2 90 10 1 12.8 147 76 62 12.3 9.0 14.9 69 84 8.8 9.7 11.3 10.2 61' 11 1 8.4 74 8.5 91 80 14.1 6.8 67 5 4' 76 104 12.2 7.2* 11 1 88 10 2 8.0 11.0 37 26 32 18' 29 33 41 22' 23' 2' 0* 5 4" 6* 0' 0* 6' 0' 4 7 12 11 15 10 13 5 9 11 7 0* 5 0* 4" 6' 5* 3' 11 1* 0* 2' 1* 7 3 3 11 4* 5* 14 5* 5' 10 0' 2* 4' 10 12 4' 4 4' 3' 6 4 9 3* 8 6 7 11 4* 10 12 13 5 1' 5* 4 7 3 5' 11 5 36 36 32 7' 7" 36 8' 3' 30 48 42 47 47 49 49 39' 25 26 25 25 27 20 28 50 29 29 7* 38* 39' 27' 49 43 10' 15* 14 29 16* 8 27 31 * 37' 8* 12* 19* 22' 33 191 37 31 7* 11' 13' 30 22 27 29 30' 51 25' 36' 36* 17' 25 29 6 44 2* 25 37' 24 4* 5* 29 29 5' 16* 43 35' 42 28* 35 24* 35 22}' 33 54 24 3' 29' 26' JLO 5* 19 49 43 41 45 47 44 21 40 23 43 39' 24 28 29 29 30 29 29 3 29 Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 29 36 34 17' 8 4' 4' 7 8 7 5 13.2 14 7 48 ID Blocks FRT N. R. GRAFF-RADFORD AND OTHERS 516 SOMATOSENSORY EVOKED RESPONSES Date of Stroke Date of SER Dy Mo Yr Dy Mo Yr NI7(F) N19(C) N20(P) N29(F) N32(C) N34(P) N60(C) N60(f Posterolateral thalamic infarction 1 04.11 82 21.10.82 03.11 82 27.10.82 2 09 03.83 03 03.83 3 24.06.83 4 18 06.83 01.01 84 5+ 06.06.80 6 30.03 83 21.03.83 04.02 84 03.01.80 7 09 09.83 09.03 83 8 08.02.84 9 23 10 83 ABS ABS ABS ABS NML NML » NML * ABS ABS ABS ABS NML NML * NML , ABS ABS ABS ABS NML NML * NML * ABS ABS ABS ABS NML DL * DL * ABS ABS ABS ABS NML DL * DL * ABS ABS ABS ABS NML DL * DL * ABS ABS ABS ABS NML DL * DL * ABS ABS ABS ABS NML DL Anterolateral thalamic infarction 10 08.04.82 26.05 82 11 21.04.82 03 03 82 12 03 04.78 13 26 08.82 03.09 82 14 19.08.83 30.08 83 DL DL NA DL NML DL DL NA DL NML DL DL NA DL NML DL DL NA NML NML DL DL NA NML NML DL DL NA NML NML DL DL NA DL DL ABS NML NA * DL Medial thalamic infarction 21 06 82 30.06.82 15 16 06.05.82 11 10 82 09.02.84 17 31.01.82 NML NML NML NML NML NML NML NML NML NML NML NML NML NML NML DL NML NML NML DL DL DL DL DL DL * NML NA NML * NML NML DL ABS NML NA NML DL NML NML DL * DL NA NML DL NML NML DL DL DL DL NA DL DL DL DL DL DL DL NA DL DL DL DL No. ABS = absent; DL = delay; NA = not available, NML = normal, * = unmterpretable, decreased sensation. NML NA NML DL NML NML = normal sensation arm, only leg has Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016 Lateral thalamic/postenor limb of internal capsule infarctions 18 11.11.81 DL 13 11 81 NML 19 01.06.83 07.06 83 NML * 14 04 83 20 12.04.83 NML NML 21 NA NA 12.04 80 22 03.12.82 NML 24 08 81 NML 16.07.82 23 01 07.82 NML * 24 02.08.83 08.08 83 NML NML 06 07 82 NML NML 25 31.08.81 DL *