Recurrent pure sensory transient ischemic attacks: In vivo demonstration of associated thalamic infarction

J Stroke Cerebrovasc Dis

1994;4:174-178 © 1994 National Stroke Association

Recurrent Pure Sensory Transient Ischemic Attacks: In Vivo Demonstration of Associated Thalamic Infarction Ernesto Fernandez-Beer, M.D., Ijames T. Patrick, M.D., Ph.D., Jose Biller, M.D., and Jeffrey L. Saver, M.D.

We present a case study and literature review to clarify the pathoanatomic substrate of recurrent pure sensory transient ischemic attacks (TlAs). A 60-year-old woman had had over 50 transient episodes ofleft hemibody hypoesthesia over the previous 15 years. Interictal neurologic examination was normal. Magnetic resonance imaging (MRI) showed a small infarct in the ventroposterolateral (VPL) nucleus of the right thalamus. Symptoms resolved with warfarin therapy. Ischemia-related pure sensory symptoms present as one of two clinical syndromes: (a) sudden onset of persistent unilateral face, arm, and/or leg hypoesthesia/paresthesias, preceded by no or few TlAs, or (b) frequent espisodes of transient unilateral hypoesthesia/paresthesias not progressing to permanent deficit. The first is widely recognized; recent computed tomography/MRI series have reported 105 cases associated with thalamic, brainstern, or thalamocortical projection sensory pathway infarction. Although equally common, the syndrome of recurrent transient hemihypoesthesia is under-recognized and has a less welldescribed pathologic substrate. One previous case with postmortem documentation of a VPL thalamic infarction has been reported. We believe this is the first reported case of recurrent transient hemihypoesthesia associated with VPL thalamic infarction imaged in vivo. Key Words: Sensory transient ischemic attackThalamic infarction-Transient ischemic attack.

First reported in the late 19th century (1), the syndrome of pure sensory stroke was revived and analyzed in clinical and pathologic detail by C. M. Fisher almost three decades ago (2). He described the clinical features of 26 patients who acutely developed isolated hemisensory deficits. Pathologic study of a single case demonstrated a lacunar infarct affecting the ventrolateral thalamus. The nature of the underFrom the Department of Neurology, Stroke Program, Northwestern University Medical School, Chicago, IL,and the IDent Neurologic Institute, Millard Fillmore Hospital, Buffalo, NY, U.S.A. Address correspondence and reprint requests to Dr. J. L Saver at Department of Neurology, Northwestern University Medical School, 233 E.Erie Street, Suite 500, Chicago, IL 60611, U.S.A. 174 ] STROKE CEREBROVASCDIS, VOL. 4, NO.3, 1994

lying arterial disease was not investigated. A few years later, he reported two additional autopsy-confirmed cases of pure hemisensory loss associated with ventrolateral thalamic infarction (3). In these two cases, the responsible vasculopathy was determined to be lipohyalinotic occlusive disease of a small penetrating artery. In 1982, Fisher reported the clinical profile of 135 cases he had observed with some form of isolated, unilateral sensory disturbance (4). He concluded that cases due to ischemia appeared as one of two clinical syndromes: (a) sudden onset of persistent unilateral face, arm, leg, and/or trunk hypoesthesia or paresthesias, preceded by no 01: few transient ischemic attacks (TIAs), or (b) frequent episodes of transient, unilateral, hypoesthesia or paresthesias not progress-

PURE SENSORYTRANSIENTISCHEMIC AITACKS

ing to a permanent deficit. He emphasized the point that both presentations were almost equally frequent. Subsequent large-scale epidemiologic and neuroradiologic studies have confirmed and extended Fisher's observations regarding the first of these syndromes-pure sensory stroke (neurologic deficit lasting more than 24 h). Pure sensory stroke accounts for 1.2-1.8% of all strokes (5,6). Involvement of face, arm, and leg simultaneously is most common, but more restricted deficits can occur, especially face and arm or arm and leg (7,8). In recent computed tomographic (CT) and magnetic resonance imaging (MRI) series, 105 cases due to ischemia (6-10) have been reported. Correlative lesions were visualized by CT in 24% and by MRI in 89%. The most common infarct sites were thalamus (35%), sensory projection pathways (27%), and brainstem (6%). Rarely, the syndrome is produced by hemorrhage or nonvascular processes (10). In contrast, the second syndrome Fisher delineated-recurrent pure sensory TIAs not progressing to a fixed deficit-has not received major attention in subsequent studies. It is likely to be clinically underrecognized. Its pathologic substrate remains obscure. Fisher reported the only postmortem study to date, documenting a posterolateral thalamic infarct in a patient who, 5 years earlier, had suffered a flurry of spells of hemiparesthesias confined to a single 7-h period (4, Case 13). In patients with pure sensory TIAs that recur over weeks to years, no pathologic or neuroimaging evidence of structural disturbance has been previously documented. We present a case study of pure sensory spells associated with in vivo MRI evidence of an ischemic infarction in a location appropriate to explain the patient's symptoms and discuss the differential diagnosis and pathophysiology of recurrent pure sensory TIAs.

Case Report A 60-year-old, right-handed, Latin American woman presented to our emergency room complaining of repeated, transient, left hemicorporal, sensory disturbances. On the day of admission, she had already experienced five episodes. She described them as abrupt in onset, involving simultaneously the whole left side of her body, not beginning focally or exhibiting a march. There were no other associated symptoms. She characterized the paresthesias as tingling, numb, and heavy feelings not affecting motor function. Each episode lasted 3-5 min and had no identifiable precipitating factors. Once the paresthesias subsided, she returned to normal function.

She further related that similar episodes of leftsided paresthesias had been occurring unpredictably for the last 15 years averaging three to four per year; never had she had so many in one day. The episodes she was experiencing on this admission were described as more intense and longer lasting, prompting her to seek medical attention. Her past medical history was significant for gestational hyperglycemia, postpartum deep venous thrombosis with pulmonary embolism, treated with a 6-month course of oral anticoagulation, hypercholesterolemia, and obesity. There was no history of hypertension, cardiac problems, tobacco use, or alcohol consumption. Examination in the emergency room was remarkable for moderate obesity. Her blood pressure was 200/110 and regular heart rate at 78 beats/min. Auscultation of the heart, neck, and cranium was unrevealing. No neurologic abnormalities were found. While still in the emergency room, we witnessed an episode of transient paresthesias during which no objective sensory or motor deficit was detected. Subsequently, hyperventilation did not reproduce the patient's symptoms. Routine hematologic and biochemical tests of the peripheral blood were noncontributory. Prothrombin time, partial thromboplastin time, sedimentation rate, antithrombin Ill, and luetic serology were normal. Unenhanced cranial computed tomography (CT) obtained on admission showed no overt lesion. MRI of the brain obtained 2 days after admission revealed an area of abnormal increased signal intensity on T2weighted images consistent with ischemic change in the right thalamus. Template mapping localized the lesion to the ventroposterolateral nucleus in the distribution of the geniculothalamic artery (11,12) (Figs. 1 and 2). No contrast enhancement was seen after administration of gadolinium. Right carotid and vertebrobasilar angiography demonstrated no significant atherosclerotic disease intra- or extra cranially. Transesophageal echocardiography (TEE)showed an atrial septal aneurysm and a low-lying atrial septal defect with contrast demonstration of a right-to-left shunt. Waking electroencephalography and median nerve somatosensory evoked potentials were nor-

maL Hospital Course Heparin was initiated on admission. During her 10day stay in the hospital, she had two additional, brief (less than 1 min) episodes of left-sided paresthesias. On the fifth day, warfarin anticoagulation was begun

J STROKE CEREBROVASC

DIS, VOL. 4, NO.3, 1994

175

E. FERNANDEZ-BEERETAL

blood pressure has been normal without treatment on four consecutive follow-up visits.

Discussion

Figure 1. AxialTl-weighted MRldemonstrates high-intensity signal abnormalitq in the right uenirocaudal thalamus.

Figure 2. Template mapping (12) localizes the lesion to the ueniroposierolateral nucleus oftheright thalamus, in thedistribution of thegeniculothalamic artery. (Thalamic structures numbered on the right include: 9, inferior thalamic peduncle; 10, fascicular nucleus; 11, thalamic fasciculus; 12, mamillothalamic tract; 13, ventral-lateral nucleus; 14, medial nucleus; 17, central nucleus; 18, habenula; 19, pulvinar. Amongvascular territories depicted on left, dots indicate the geniculoihalamic territory.)

and has been maintained to the present at an international normalized ratio (INR) of 2.0-3.0. In the 9 months of follow-up, she had only one very brief (a few seconds) episode 5 months after discharge. Her 176

] STROKECEREBROVASCDIS, VOL.4, NO.3, 1994

The differential diagnosis of recurrent, transient, hemicorporal, pure sensory spells includes migraine auras (13), simple partial seizures (14), hyperventilation (15), and ischemia. The possibility that our patient's recurrent sensory symptoms reflected a migrainous phenomenon that eventually culminated in infarction is appealing and cannot be completely excluded. However, the simultaneous involvement of face, arm, and leg without a march, the brief duration of the episodes, the absence of associated headaches, and the absence of personal or family history of vascular headaches is strong clinical evidence against migraine. Simple partial seizures must be considered in any patient presenting with episodic hemisensory disturbances. They pose a particular diagnostic challenge, since routine scalp-recorded ictal EEG tracings detect ictal discharges in only 11% of patients with simple partial seizures. The sensitivity is increased to 89% if subdural grid recordings are utilized (16). However, in our patient, the lack of other associated symptoms during the ictus, their sporadic and nonprogressive course over many years without antiepileptic treatment, the absence of a rapid march, the association with a subcortical rather than cortical lesion, and to a lesser extent the normal interictal EEG make an epileptic etiology highly unlikely. Hyperventilation attacks are typically characterized by bilateral paresthesias of the upper limbs and, less often, the lower limbs and face, accompanied by lightheadedness, headache, blurring of vision, and/or loss of consciousness (15). Although somatic symptoms are most commonly bilateral, 10% exhibit unilateral paresthesias, predominantly left-sided as in our patient (17). However, the absence ofidentifiable triggering factors, the lack of lightheadedness or other typical accompanying symptoms, the sporadic occurrence, the associated structural brain lesion, the failure to reproduce the symptoms with hyperventilation, and the response to anticoagulant treatment practically exclude this possibility as an explanation for our patient's spells. An ischemic mechanism is the most likely explanation for this patient's intermittent pure sensory symptomatology. The lack ofa march to the spells, the improvement with anticoagulant therapy, and the association with an MRI-documented abnormality

PURE SENSORYTRANSIENTISCHEMIC ATTACKS

compromising the sensory nucleus of the thalamus are all consonant with an ischemic etiology. The paraclinical investigations performed disclosed the presence of a known cause of cardioembolism. It is well-recognized that small subcortical strokes may be caused by cardiac or artery-to-artery embolization, albeit rarely (18). However, the sporadic and stereotypical nature of the events over many years, and the normal carotid and vertebrobasilar angiogram argue strongly against cardioembolic or large-vessel atherothromboembolic disease. This leaves us with the hypothesis that our patient's spells were pure sensory TIAs secondary to intrinsic occlusive disease of a small, penetrating artery. In this patient, the most likely etiologies are micro atheroma or lipohyalinosis, despite the absence of a well-documented history of arterial hypertension. It is known that about 40% of patients with small subcortical strokes are normotensive (19). The syndrome of recurrent pure sensory TIAs not progressing to a lasting deficit, illustrated by our patient, is likely more common than is currently appreciated. Pure sensory TIAs account for approximately 20% of all lacunar TIAs (20). In Fisher's series (4), among 100 patients with hemisensory symptoms felt likely to be due to ischemia, almost half (42%) exhibited pure sensory TIAs without subsequent deficit lasting longer than 24 h. In contrast, as with other lacunar stroke syndromes (6,21), the large preponderance (about 80%) of pure sensory strokes occur without preceding TIAs (4,10). In the minority of cases in which pure sensory TIAs do herald subsequent persisting clinical deficits, the transient spells tend to occur only in the immediately preceding hours or days, rather than over weeks or months (4,10). Among patients who experience pure sensory TIAs alone, about half will have only one or two attacks, but half will have three or more and a quarter will have between 10 and 100 episodes (4). Our patient reported approximately 60 spells over 15 years. Most pure sensory TIAs are brief, with over 50% lasting less than 30 min (4). Accurate recognition of pure sensory TIAs is important to patient management. In a series of patients with "crescendo" lacunar TIAs, 4 of 9 exhibited pure sensory TIAs, and carotid angiography was negative in all (22). Patients with suspected pure sensory TIAs should perhaps not be subjected to conventional angiography with its attendant risks. Also, prolonged trials of anticonvulsant medication and prophylactic antimigraine agents should not be pursued in patients whose clinical symptomatology strongly suggests recurrent pure sensory TIAs, to avoid unnecessary

exposure to adverse effects of pharmacotherapy. Since some patients with pure sensory TIAs do progress to persistent deficit, antithrombotic therapy should be considered, recognizing that the benefit of anti platelet or anticoagulant agents in lacunar TIAs is unproven. The localization and pathogenesis of recurrent pure sensory TIAs is iII-defined. Most likely, as in pure sensory infarct, the most common site of ischemia is the thalamus. Fisher's patient exhibiting a posterolateral thalamic infarct remains the only pathologic case reported. Rothrock et aI. (11) briefly described one patient with TIAs occurring within a 24h period whose MR! showed a left anterolateral thalamic lacune. Our patient with recurrent TIAs over 15 years exhibited a right ventroposterolateral thalamic infarct on MR!. Doubtless, transient ischemia at other sites implicated in pure sensory stroke may also produce transient sensory symptoms, such as brainstem sensory pathways and thalamocortical projection fibers. Patients with recurring TIAs associated with pathologically or radiologically documented infarction may have partially impaired, but just adequately compensated, sensory processing especially vulnerable to alterations in collateral blood flow. Transient ischemia that produces neuronal dysfunction but not infarction, however, is presumably also a frequent substrate for pure sensory TIAs.

References 1. Bourneville A. De l'hernianesthesie liee a une lesion d'un hemisphere du cerveau. Prog Med 1883;1:2446. 2. Fisher CM. Pure sensory stroke involving face, arm, and leg. Neurology 1965;15:76-80. 3. Fisher CM. Thalamic pure sensory stroke: a pathologic study. Neurology 1978;28:1141-4. 4. Fisher CM. Pure sensory stroke and allied conditions. Stroke 1982;13:434-47. 5. Bamford J, Sandercock PAG, Jones L, et a1. The natural history of lacunar infarction: The Oxfordshire Community Stroke Project. Stroke 1987;18:545-51. 6. Charnorro A, Sacco RL, Mohr JP, et a1. Clinical-eomputed tomographic correlations of lacunar infarction in the Stroke Data Bank. Stroke 1991;22:175-81. 7. Hommel M, Besson G, Le Bas jF, et a1. Prospective study of lacunar infarction using magnetic resonance imaging. Stroke 1990;21:546-54. 8. Arboix A, Marti-Vilalta JL,Garcia [H. Clinical study of 227 patients with lacunar infarcts. Stroke 1990;21:8427. 9. Chimowitz MI, Furlan AJ, Silva CA, et a1. Etiology of motor or sensory stroke: a prospective study of the predictive value of clinical and radiological features. Ann Neuro11991;30:519-25. 10. Kim JS. Pure sensory stroke. Clinical-radiological correlates of 21 cases . Stroke 1992;23:983-7. ] STROKE CEREBROVASC DIS,VOL 4, NO.3, 1994

177

£ FERNANDEZ-BEERET AL

11. Rothrock JF, Lyden PD, Hesselink JR, et al. Brain magnetic resonance imaging in the evaluation of lacunar stroke. Stroke 1987;18:781-6 . 12. PuIlicino PM. Diagrams of perforating artery territories in axial, coronal and sagittal planes. In: PuIlicino PM, Caplan LR, Hommel M, eds. Cerebral small artery disease. New York: Raven Press, 1993:41-72. 13. Fernandez-Beer E, Saver JL,Biller J. Stroke symptoms associated with migraine headaches. Heart Dis Stroke 1993;2:69-74. 14. Devinsky 0, Kelley K, Porter RJ,et al. Clinical and electroencephalographic features of simple partial seizures. Neurology 1988;38:1347-52. 15. Perkin GD, Joseph R Neurological manifestations of the hyperventilation syndrome. J R Soc Med 1986;79: 448-50. 16. Devinsky 0, Sato S, Kufta CV, et al. Electroencephalographic studies of simple partial seizures with subdur-

178

J STROKECEREBROVASCDIS, VOL 4, NO.3,

1994

al electrode recordings. Neurology 1989;39:527-33. 17. Blau IN, Wiles CM, Solomon FS. Unilateral somatic symptoms due to hyperventilation. Br Med J 1983;286: 1108. 18. Lodder J, Bamford JM, Sandercock PAG, et al. Are hypertension or cardiac embolism likely causes of lacunar infarction? Stroke 1990;21:375-81. 19. Millikan C, Futrell N. The fallacy of the lacune hypothesis. Stroke 1990;21:1251-7. 20. Hankey GJ, Warlow CPo Lacunar transient ischaemic attacks: a clinically useful concept? Lancet 1991;337: 335-8. 21. Norrving B, Cronqvist S. Clinical and radiologic features of lacunar versus nonlacunar minor stroke. Stroke 1989;20:59-64. 22. RothrockJF, Lyden PD, YeeJ, et al. "Crescendo" transient ischemic attacks: clinical and angiographic correlations. Neurology 1988;38:198-201.