Lobar Cerebral Hemorrhages: Acute Clinical Syndromes in 26 Cases Allan H. Ropper, MD, and Kenneth R. Davis, MD The acute syndromes and CT findings are described in 26 cases of spontaneous cerebral hemorrhage. Occipital hemorrhage (11 cases) caused severe pain around the ipsilateral eye and dense hemianopia. Left temporal hemorrhage (7 cases) began with mild pain in or just anterior to the ear, fluent dysphasia with poor auditory comprehension but relatively good repetition, and a visual deficit subtending less than a hemianopia. Frontal hemorrhage (4 cases) caused a distinctive syndrome beginning with severe contralateral arm weakness, minimal leg and face weakness, and frontal headache. Parietal hemorrhage (3 cases) began with anterior temporal (“temple”) headache and hemisensory deficit, sometimes involving the trunk to the midline. One patient had a right temporal hemorrhage. Spontaneous lobar hemorrhage and branch artery embolism in the same region produce similar clinical syndromes. Headache is a first and prominent symptom. A rapid but not instantaneous onset over several minutes, when combined with one of the typical syndromes, suggests lobar hemorrhage rather than other types of stroke. Ancillary investigations (including CT scanning, angiography in 11 patients, and autopsy in 4) disclosed 2 patients with bleeding diatheses due to warfarin, 2 with arteriovenous malformations, and l with metastatic tumor. Only 8 of the 26 patients had chronic hypertension (blood pressure greater than 130,435 mm Hg), suggesting that hypertension is not an etiological factor in most lobar hemorrhages. Ropper AH, Davis KR: Lobar cerebral hemorrhages: acute clinical syndromes in 26 cases. Ann Neurol 8:141-147, 1980 Bleeding located outside the basal ganglia and thalamus accounts for a minority of reported cerebral hemorrhages [ l , 2, 4 , 5 , 9, 11, 141. Because the symptoms and signs resemble those of emboiization of the same lobe [8, 121, and the mortality rate is lower than with other hemorrhages [4, 5, 91, lobar hemorrhages have not been well characterized clinically or pathologically. With the advent of computerized tomography (CT scanning), we were able to study the clinical and radiological characteristics in 26 cases of acute lobar hemorrhage. Four syndromes could be delineated. Materials and Methods Twenty-six patients with hemispheric cerebral hemorrhage located outside the basal ganglia and thalamus were studied. Sixteen patients were examined by one of the authors (A. H. R.). Ten additional cases were obtained by retrospective review of CT scans and charts of the Massachusetts General Hospital during the period July, 1975, through July, 1979. Data from the examined patients did not differ from those obtained retrospectively. Cases were excluded if: (1) the hemorrhage appeared to originate in the region of the basal ganglia or thalamus on CT scan; (2) the initial CT scan demonstrated an arteriovenous malformation (AVM) or tumor as the cause of From the Neurological-Neurosurgical ICU, Department of Neurology, and the Department of Radiology, Massachusetts General Hospital, Boston, MA. hemorrhage; (3) the patient had a known AVM; or (4) the history did not explicitly exclude recent head trauma or there was some sign of trauma such as scalp abrasion or hematoma. CT scans were performed in all patients within the first 48 hours after hemorrhage and again with intravenous infusion of contrast material several days later. A computed tomographic EMI-1005 cranial scanner with a 160 X 160 matrix was used with a scanning time of 1 minute per pair of 8 mm thick sections. Angiograms were performed in 11 patients; 2 patients had a second angiogram. The average age of the 26 patients was 65 years. The youngest was 18 and the oldest 85. Seven were younger than 60, 7 were in their sixties, 8 in their seventies, and 4 in their eighties. Thirteen were male and 13 female. There were 5 deaths, 3 of which were directly related to the acute hemorrhage. Autopsy material was available in 4 cases. Results Potential predisposing conditions for hemorrhage were present in 5 patients (Table). Two were taking warfarin at the time of their hemorrhage, 1 for a prosthetic aortic valve and 1 after a myocardial infarction. Prothrombin times were, respectively, 27 with a control of 12 seconds and 27 with a control of 14 seconds. All patients but the 2 o n warfarin had normal prothrombin times, partial thromboplastin Received Oct 11, 1979, and in revised form Dec 10. Accepted for publication Dec 16, 1979. Address reprint requests to Allan H . Ropper, MD, Department of Neurology, Massachusetts General Hospital, Boston, MA 02 7 14. 0364-5134/80/080141-07$01.25 @ 1979 by Allan H. Ropper 141 Underlying and Associated Conditions in 26 Sgbcortical Hemorrhages No. of PredisDosinn Condition Patients Definite factors Anticoagulation AVM Metastatic tumor Possible factors Neurosyphilis Postpartum state Polymyalgia rheumatica Blood pressure Hypertension prior to hemorrhage Normotension prior to hemorrhage (130/85 mm Hg or less) Definite (documented by previous hospitalization or doctors’ records) Probable (BP checked within 1 year, patient not told of hypertension) Prior BP unknown 2 2 1 1 1 1 8 14 3 1 AVM = arteriovenous malformation, BP = blood pressure. times, and platelet counts on admission. In 1 patient each, AVMs were found on a second angiogram one month after the hemorrhage and on a second CT scan one year after the hemorrhage. A metastatic fibrosarcoma was demonstrated by biopsy in 1 patient after signs of a miss lesion persisted. Blood prqssure (BPI history was available from dependable sources in 25 cases. Only 8 patients (31%) had a history of previous hypertension (BP > 130/85 mm Hg) (see the Table). Twelve patients, including the 8 with chronic hypertension, were hypertensive acutely with their hemorrhages. The highest acute BP was 260/160 mm Hg, with several at 200/100. Characteristics of Hemorrhages Hemorrhages were located in the occipital lobe in 11 patients, the temporal lobe in 8, the frontal lobe in 4 , and the parietal lobe in 3. Fifteen hemorrhages were in the left hemisphere and 11 in the right. The mean diameter of the clot at its greatest extent in the horizontal plane on CT scan sections (15 degrees cranial-caudal to the canthomental line) was 2 to 3 cm in 8 patients, 3 to 4 cm in 8 patients, and 4 to 6 cm in 10 patients. Measurements in the horizontal plane, rather than height, were considered of greatest value because of the propensity of such hemorrhages to enlarge mostly along the x and y coordinates of the CT section. Initial angiograms in 11 patients demonstrated only an avascular mass. 142 Annals of Neurology Vol 8 N o 2 August 1980 General Clinical Features The initial symptom was headache alone or predominating over other symptoms in 12 patients (severe in 8), neurological symptoms referable to the affected lobe in 10, and simultaneous headache and neurological symptoms in 3 . One patient was in coma. The headache preceded other symptoms by 30 minutes to 12 hours in 6 patients, and neurological symptoms antedated headache by several hours in 4 patients. In 7 patients headache was minimal or absent. Eleven patients vomited within the first hours of illness. Only 5 patients had a stiff neck on admission. No subhyaloid hemorrhages were seen. Excluding the patient with coma, 7 patienrs were drowsy on admission, 5 requiring auditory stimulation to arouse them and 2, physical stimulation. In 20 of the 26 patients, the onset of headache or neurological symptoms was sudden so that patients or observers were aware of a problem evolving over less than 5 or 10 minutes. However, the complete deficit did not arise in a lightning-like way. New symptoms frequently appeared during the first minutes to half-hour. In 9 patients additional symptoms appeared during the 30 minutes to 12 hours after the onset of headache. In 12 patients no new symptoms were reported after about 30 minutes, though an existing neurological deficit sometimes worsened. N o patient developed epilepsy with the onset of hemorrhage. Symptoms were first reported upon awakening or during preceding sleep in 9 patients, 30 minutes to 2 hours after awakening in 7, in the evening in 7, and at various other times in 3. None of the patients reported waking during the night with headaches. In 7 of the 11 patients who had lumbar punctures on admission, blood was found in the cerebrospinal fluid (CSF). Two of the 7 had temporal lobe hemorrhages, with clot reaching the cortical surface on CT. All the patients with occipital hemorrhages who had lumbar punctures were found to have blood in the CSF, though it was not always evident on the CT scan that the hemorrhage reached the subarachnoid space. Thirteen patients had electroencephalograms within 4 days of admission, and all were focally abnormal on the side of the hemorrhage. Six showed continuous focal slowing overlying the hemorrhage, and 1 had occasional focal spikes. Acute Clinical Syndromes In these 11 patients, headache was severe and was located in and around the ipsilateral eye in 7. Nine patients were explicitly aware of visual difficulty at the onset. Four described a hemianopia, 3 complained of “bumping into things,” and 2 noted only “blurred vision.” OCCIPITAL HEMORRHAGE. F i g I . C T scans without contrast infusiorr from 6 patients with occipital lobe hemorrhages. The clots appear t o extend to the co&cal szlrface. Allscans in this figure m d i n Figures 2 through 4 were performed within 24 hours of the onset of symptonzs. All 11 patients with occipital hemorrhages had hemianopias. Visual deficits were dense, complete, and congruous in 9 and spared a portion of the superior quadrant in 2. There was no obvious paresis, but 2 patients had minimal pronation drift of the arm and 3 had a Babinski sign on the side of the hemianopia. Three patients extinguished double simultaneous stimulation on the arm contralateral to the hemorrhage, and 2 had marked dysgraphia and dyslexia. There were no migranous visual phenomena, quadrantanopias, delirium, or fluctuation in deficits except in 1 patient, who developed a worsening hemiparesis and coma while taking warfarin. CT scans showed circular or ellipitical hemorrhages centered in the occipital white matter (Fig 1). TEMPORAL HEMORRHAGE. Only 1 patient had a right temporal hemorrhage, and arrived at the hospital comatose, providing insufficient information to describe an acute clinical syndrome in that lobe. Among the 7 patients with left temporal lobe hemorrhage, a family member or a friend initially noted a speech problem or confusion with difficulty in comprehension. Two patients complained to the family of trouble reading several hours before a severe headache. Five patients had mild to moderate left-sided headache centered around the ear o r just anterior to it. One patient vomited and then noted mild rightsided weakness. The initial examination in all 7 patients demonstrated dysphasia with similar characteristics. Speech was fluent, with poor comprehension and many paraphasias. Repetition was relatively good, though not perfect, in 5 patients. One patient was able to answer yeslno questions correctly, another had a paucity of speech in the first hours but was not mute, and 2 were initially logorrheic. Several patients were thought by other examiners to be “confused,” though they were neither drowsy nor inattentive. One patient was aphasic, delirious, and extremely agitated. Only 1 had right-sided weakness with a Babinski sign; the other 6 had no weakness. Two patients had minimal but definite sensory deficits with diminished reaction to pinprick on the limbs. Six had visual field deficits: 3 incongruous right hemianopias and 3 right inferior quadrantic defects. The CT scans showed rounded hemorrhages cen- Ropper and Davis: Lobar Cerebral Hemorrhage 143 tered in the temporal lobe white matter, reaching the surface in 2 cases (Fig 2). Several extended slightly into the parietal lobe. There was no ventricular o r subarachnoid blood. These 4 patients noted limb paresis initially, with headache occurring later in 2. Contralateral arm weakness was the most prominent symptom and caused patients to seek attention. The headache was moderately severe, bifrontal, and located predominantly on the side of the hemorrhage. All 4 patients had severe contralateral arm weakness when first examined. Two had an arm monoplegia. There was only mild leg weakness and minimal facial weakness. All the patients could walk. A mild gaze preference away from the weak arm was noted in 2 patients, and 3 patients were sleepy. No sensory deficits, change in affect, or language difficulty was found. (One frontal hemorrhage was on the left.) The CT scans showed rounded or oval hemorrhages centered in frontal lobe white matter (Fig 3). One extended slightly posteriorly to the anterior parietal lobe. FRONTAL HEMORRHAGE. PARIETAL HEMORRHAGE. All 3 patients had an acute severe headache in the anterior temporal (“temple”) region at the onset. Two were drowsy, 144 Annals of Neurology Vol 8 No 2 August 1980 Fig 2. CT scans, without contrast infusion, from 4 patients with left iemporal lobe hemorrhages. pig 3. CT scans, without contmst znfuston, from 3 pcrtients with frontal lobe hemorrhages. Fig 4. C T scans, without contrast infusion, from 3 patients with parietal lobe hemorrhages. The scan on the lefr is from pathological Case 2. and 1 complained of his hand “not feeling as if it were mine.” There was a severe contralateral hemisensory deficit involving the entire trunk to the midline in 2 patients and a moderate sensory deficit of the face, arm, and leg in 1 patient. All had a mild contralateral hemiparesis, with the arm slightly weaker than the leg and the face minimally affected. There was incomplete contralateral hemianopia in 2 patients and extinction in the contralateral visual field in 1. O n e patient had anosognosia of the left hand (right hemisphere hemorrhage), and 1 a mild fluent dysphasia with poor comprehension (left hemisphere hemorrhage). The CT scans are shown in Figure 4. Pa thological Pea tures A 64-year-old normotensive man died of a pulmonary embolus 55 days after a right occipital hemorrhage. The cortical surface appeared normal, but on horizontal sectioning there was a 4.5 x 3 cm clot in the center of the right occipital lobe. The epicenter of the clot was 2.5 cm from the cortical surface; its lateral border was just beneath the lateral occipital cortical surface (Fig 5A). No vascular CASE 1. malformations and no hypertensive vascular or amyloid changes were found with extensive sampling of the brain surrounding the hemorrhage. Heart weight and left ventricular wall thickness were normal. The position of the clot postmortem correlated well with the CT scan appearance two months before (Fig 5B). CASE 2. A 70-year-old normotensive woman suffered a left parietal hemorrhage and died of a pulmonary embolus 15 days later. On coronal sectioning of the hemispheres, a hemorrhage 6 cm in greatest diameter was found centered in the left posterior parietal lobe and reaching into the posterior frontal and anterior occipital lobes. No abnormal vessels or hypertensive vascular or amyloid changes were seen on extensive microscopic examination of the regions adjacent to the hemorrhage. Heart weight and size were normal. There was good correlation between the postmortem clot and the CT scan appearance 15 days earlier. CASE 3. An 84-year-old hypertensive woman with a right temporal hemorrhage and 100,000 red blood cells in the CSF on admission died 3 days later. Coronal sections of the hemispheres showed an elliptical hemorrhage in the midportion of the right temporal lobe with a largest vertical diameter of 4 cm. No abnormal vessels were seen on gross or microscopic examination of the regions adjacent to the hemorrhage. There was good correlation between the appearance of the clot and the CT scan. A 62-year-old normotensive man with an occipital hemorrhage and 373 red cells in the CSF worsened after CASE 4 . Ropper and Davis: Lobar Cerebral Hemorrhage 145 his steroid dose was reduced. Brain biopsy revealed a fibrosarcoma. At autopsy there was a large clot within a metastatic fibrosarcoma in the left occipital lobe. Fig 5 . (Case 1) (A) Horizontalsection of brain through midportion of lateral ventricles shows a 4.5 x 3 cm clot in the right occipital lobe. (B) Unenhanced CT scan from the same patient perfomzed 6 dayJ after the hemorrhage, 49 days prior to death. 146 Annals of Neurology Vol 8 No 2 August 1980 Discussion With improved control of hypertension, the incidence of putaminal and thalamic hemorrhage has decreased [6, 101. The relative proportion of spontaneous lobar hemorrhages has risen correspondingly at our institution in the last several years. Lobar subcortical hemorrhages have accounted for 10 to 32% of nontraumatic cerebral hemorrhages in several large series [1, 2 , 7 , 9 , 11, 15, 161. However, because of their relatively low mortality rate, they may be underrepresented in autopsy series [ 4 , 5, 91. Some lobar hemorrhages are due to identifiable lesions such as aneurysms [9, 111, arteriovenous or angiomatous malformations [l, 9, 11, 141, and neoplasms [9, 111. Others occur with bleeding diatheses, especially warfarin anticoagulation [2]. The proportion of lobar hematomas originating from vascular malformations varies widely in published series [7, 91. Telangiectasias in white matter are found incidentally at autopsy, but evidence of old hemorrhage is rare, suggesting that the lesions are unlikely to cause serious hemorrhage [ 131. Several authors have suggested that many subcortical hemorrhages are due to microscopic venous or angiomatous malformations that become hidden or obliterated with bleeding and therefore are not found at autopsy [ 1, 11, 141. This proposal is inherently difficult to substantiate. Four to 11% of lobar hemorrhages are reportedly due to unsuspected larger AVMs [7, 9, 11, 151. Angiograms or contrast CT scans performed soon after hemorrhage occasionally overlook AVMs that are compressed by adjacent clot. Two patients in our series had AVMs demonstrated only on radiographic studies repeated one and twelve months after their hemorrhages. Delaying angiography for several weeks deserves consideration in patients with lobar hemorrhage. Based on the findings from angiograms obtained at various times after stroke in 11 of our patients, extensive pathological studies in 3, and contrast CT scans in all 26, it is unlikely that undiagnosed AVMs accounted for many hemorrhages in our series [7]. Microaneurysms caused by hypertension, identical to those which occur in lenticulostriate vessels, are occasionally found in subcortical arteries. It is possible that hemorrhages located outside the basal ganglia and thalamus are related to these lesions [ 3 ] . H ~ data~ suggest~that chronic ~ hypertension ~ ~ precedes only 31% of lobar hemorrhages. Amyloid vascular changes have also been reported in association with cerebral hemorrhages. We were unable to find hypertensive small artery damage or amyloid , change in the cerebral hemispheres of 3 autopsied cases. The pathogenesis of the majority of lobar hemorrhages therefore remains unknown. The most distinctive acute syndrome occurred with lobar hemorrhage in the frontal lobe. There was severe contralateral arm weakness, with only mild face and leg weakness. Because of the isolated motor deficit in 3 of the 4 patients, the hemorrhage was initially thought to be an embolus to a rolandic branch artery. Similarities between the clinical syndromes of lobar hemorrhage and cerebral embolism raise the possibility that some hemorrhages begin as embolic strokes which become hemorrhagic infarctions. Though lobar hemorrhages are centered in regions corresponding approximately to areas infarcted by branch artery emboli, the clots often span two vascular territories. Early headache, a rapid but not instantaneous onset, and homogeneous appearance of the clot on CT scans suggest that acute vascular occlusion is not the inciting event in lobar hemorrhage. None of the patients with hemorrhages had extensive deficits similar to those caused by proximal middle cerebral artery thrombi o r putaminal hemorrhages. The dense hemiparesis, gaze preference, and drowsiness seen with putaminal hemorrhages did not occur simultaneously in lobar hemorrhages. Several general features of lobar hemorrhage are helpful in the differential diagnosis. Headache as a first and prominent symptom occurred in 12 patients (46%),less commonly than in embolic stroke. Absence of headache does not exclude hemorrhage, however. A combination of one of the four lobar neurological deficits described, coupled with headache, drowsiness or vomiting, new symptoms during the 12 hours after onset of stroke, and rapid but not instantaneous onset, should strongly suggest lobar hemorrhage rather than embolus. The occurrence of these hemorrhages means that most patients with stroke syndromes suggestive of partial lobar deficits will require CT scans for fully accurate diagnosis. Only 3 of our 26 patients died as a result of hemorrhage, suggesting that subcortical lesions are more benign than those centered in the ganglionic-thalamic region. This may be due to the smaller mean size of most lobar hemorrhages. Their peripheral location may also make brainstem compression less likely. Most of our patients had good resolution of neurological deficits, suggesting that routine surgical evacuation of the clot is inadvisable. The authors thank Dr C. Miller Fisher for his helpful advice. References 1. Adams R, Vander Eecken HM: Vascular disease of the brain. Annu Rev Med 4:213-220, 1953 2. Barron KD, Fergusson G: Intracranial hemorrhage as a complication of anticoagulant therapy. Neurology (Minneap) 9:447-455, 1959 3. Cole FM, Yates PO: Intracerebral microaneurysms and small cerebrovascular lesions. Brain 90:759-768, 1967 4. Fisher CM: The pathology and pathogenesis of intracerebral hemorrhage. In Fields WS (ed): Pathogenesis and Treatment of Cerebrovascular Disease. Springfield, IL, Thomas, 1961, pp 1-17 5. Freytag E: Fatal hypertensive intracerebral hematomas: a survey of the pathological anatomy of 393 cases. J Neurol Neurosurg Psychiatry 3 1:616-620, 1968 6. Furlan A, Whisnant JP, Elveback L: The decreasing incidence of primary intracerebral hemorrhage: a population study. Ann Neurol 5~367-373, 1979 7. Hayward RD, OReilly GVA: Intracerebral hemorrhage. Accuracy of computerized transverse axial tomography in predicting underlying aetiology. Lancet 1:l-4, 1976 8. Johansson SH: Hypertensive and normotensive intracerebral hemorrhage. Acta Psychiat Scand 36:suppl 150:90-94, 1961 9. McCormick WF, Rosenfield DB: Massive brain hemorrhage: a review of 144 cases and an examination of their causes. Stroke 4:949-954, 1973 10. Moser M, Goldman A: Hypertensive Vascular Disease. Philadelphia, Lippincott, 1967 11. Russell DB: The pathology of spontaneous intracranial hemorrhage. Proc R SOCMed 47:689-693, 1954 12. Schaafsma S: On the differential diagnosis between cerebral hemorrhage and infarction. J Neurol Sci 7233-94, 1968 13. Stebhens WE: Pathology of the Cerebral Blood Vessels. St. Louis, Mosby, 1972, pp 474-475 4. Vander Ark G, Edgar A, Kahn E: Spontaneous intracerebral hematoma. J Neurosurg 28:252-256, 1968 5 . Weisberg L: Computerized tomography in intracranial hemorrhage. Arch Neurol 36442-446, 1979 6. Zimmerman HM: Cerebral apoplexy: mechanisms and differential diagnosis. NY State J Med 49:2153-2158, 1949 Ropper and Davis: Lobar Cerebral Hemorrhage 147