Posterior Circulation Infarcts Simulating Anterior Circulation Stroke
Perspective of the Acute Phase
Background and Purpose Ischemic stroke patients whose initial clinical presentation suggests an involvement of the anterior circulation (AC) are sometimes found to have a posterior circulation (PC) infarct, a fact that may generate erroneous decisions in clinical management. We investigated the prevalence of this misdiagnosis in the first few hours after stroke onset.
Methods We performed a cohort study of 158 patients hospitalized within 5 hours of onset of a presumed AC ischemic stroke, as diagnosed on clinical grounds.
Results Final CT or pathology diagnosis was AC infarct in 128 patients (81%), a repeatedly negative CT in 14 (9%), PC infarct (5 pons, 1 midbrain and cerebellum, 6 supratentorial territory of the posterior cerebral artery) in 12 (8%), and other or undiagnosed lesions in 4 (3%). AC and PC stroke patients did not differ in terms of age, vascular risk factors, and initial severity, but the latter were more frequently men (83% versus 53%; P=.04), were hospitalized later (mean±SD, 168±86 versus 109±55 minutes; P=.001), and presented a pure motor hemiparesis or a sensorimotor stroke (50% versus 33%) more often than their counterparts. At baseline CT, PC stroke patients never exhibited an early parenchymal hypodensity in the carotid territory or a hyperdense middle cerebral artery, which were instead found in 59% (P=.0003) and 31% (P=.02) of AC stroke patients, respectively. Early neurological deterioration, 1 month case-fatality rate, and disablement in survivors were comparable in the two groups.
Conclusions Shortly after onset the clinical discrimination between AC and nontypical PC infarcts is not reliable, which explains the frequent occurrence of this misdiagnosis. Emergency CT scan helps in the differential diagnosis only when it demonstrates an early focal hypodensity within the carotid territory.
Typical infarcts in the posterior circulation (PC) are recognized by the various combinations of long-tract deficits associated with contralateral cranial nerve palsies, ataxia, gaze palsies, vertigo, bilateral motor or sensory deficit, and hemianopia.1 However, it is well known that PC infarcts may simulate an involvement of the anterior circulation (AC), a misdiagnosis that may lead to erroneous decisions in clinical management.2 Previous studies have investigated the location of PC strokes mimicking AC strokes and have emphasized the fact that only CT or MRI permits a correct diagnosis.2 3 4 However, the frequency of this misdiagnosis in the acute phase has never been assessed.
In the present study we investigated a cohort selected from a continuous series of patients admitted to a stroke unit with a clinical presentation suggestive of an ischemic stroke in the AC within the first 5 hours of onset, which roughly corresponds to the crucial time frame for most diagnostic and therapeutic procedures. Our aim was to assess how frequently a PC infarct mimics an infarct in the AC. We also tried to determine whether, despite a similar clinical presentation, patients with a PC infarct differed from those with a true AC infarct in terms of vascular risk factors, initial clinical characteristics and CT findings, and final outcome.
Subjects and Methods
The data for the present study were derived from the registry of our stroke unit, an eight-bed ward serving the Emergency Department of the hospital of the University “La Sapienza” of Rome. We reviewed our hospitalization log for the period from March 1987 to March 1989: in this period, 245 patients with cerebral infarction were admitted to our stroke unit. We retrieved all the records pertaining to patients admitted within 5 hours of onset of a neurological deficit suggesting an AC ischemic stroke.
The criteria used for bedside diagnosis of an AC or PC infarct were those of Bamford et al,5 with slight modifications. In particular, any clinical presentation characterized by hemiparesis alone or in combination with aphasia, neglect, hemianopia, or ipsilateral sensory deficit was considered to be indicative of a probable AC infarct. Excluded from the study were patients whose initial signs and symptoms suggested an involvement of the PC, ie, crossed syndrome, bilateral motor or sensory deficit, disorder of eye movement other than conjugate deviation toward the nonhemiparetic side, and cerebellar dysfunction.
The standardized management protocol on admission included a plain CT scan and a clinical examination, including registration of vital signs and blood pressure, electrocardiogram, and blood chemicals. During the study period, the first version of the Canadian Neurological Scale6 was used to quantify the neurological deficit. This scale evaluates level of consciousness, orientation, speech, and facial and limb strength with a global score ranging from 1.5 (maximal deficit) to 10 (no deficit). Past medical history was investigated to assess stroke risk factors such as hypertension, diabetes, atrial fibrillation, other potentially embolic cardiopathies, previous transient ischemic attacks, and cigarette smoking. Initial CT scans, performed in all cases not later than 6 hours after onset, were reviewed for the presence of early focal hypodensity involving the lentiform nucleus and/or the cortex within the carotid distribution, initial mass effect, or hyperdensity of the middle cerebral artery (MCA).7 The location of the infarcted area was determined on the control CT, which was systematically performed during the first week after stroke or at autopsy in cases of death before repeated CT scan. Only lesions congruous with the clinical presentation were taken into account. Parenchymal distribution of AC and PC was defined according to the criteria of Damasio.8
Patients were followed up for 1 month after stroke. During this period we evaluated the case-fatality rate and the functional outcome in survivors. Disablement was defined as a Barthel Index score9 lower than 60.
Univariate analysis was performed with the use of the t test and Fisher's exact test to compare demographic data, baseline clinical findings, vascular risk factors for stroke, findings on initial CT, and outcome of AC and PC stroke patients. Logistic regression analysis was then applied to the aforementioned variables to determine factors that discriminated between PC and AC infarcts. The analysis was performed with the use of the 1990 release of BMDP statistical software.10
There were 158 patients whose clinical diagnosis on admission was of presumed cerebral infarction in the AC. All were hospitalized and evaluated within 5 hours from onset. Seven patients died before a follow-up CT could be performed. In these cases, diagnosis was made at autopsy. The final diagnosis was of ischemic stroke in all patients but one (1%), who had a brain metastasis. Second CT or autopsy showed the qualifying infarct to be localized in the AC in 128 patients (81%) (124 territorial, 4 lacunar) and in the PC in 12 (8%). Two patients (1%) had two infarcts: one in the basal ganglia and the other in the pons, both of which could explain the clinical presentation. One patient (1%) had a final diagnosis on the clinical record of ischemic stroke, but his second CT scan could not be retrieved. Finally, 14 patients (9%) had a repeatedly negative CT. We did not observe any border zone infarctions between the MCA and the anterior cerebral artery or between the MCA and the posterior cerebral artery.
Patients with proven AC infarct (n=128) were compared with patients with proven PC infarct (n=12) in terms of demographic characteristics, vascular risk factors, and initial clinical findings (Table⇓). PC infarct patients were more frequently men, were admitted to the hospital approximately 1 hour later than other patients, and exhibited a nonsignificant trend toward a milder neurological deficit. The frequency of vascular risk factors was similar in the two groups, as were serum glucose levels and blood pressure at entry. In terms of the clinical presentation at entry, PC infarct patients had a pure motor hemiparesis or a sensorimotor stroke more often than their counterparts and never had a total hemispheric syndrome. The number of presumed right or left hemisphere ischemic strokes was similar in the two groups. Three PC patients were aphasic (two motor, one sensory). Repeated CT of the two patients with motor aphasia revealed an infarct in the lateral thalamus, while the patient with sensory aphasia had a temporo-occipital infarct. The remaining patients had infarcts in the pons (n=5), the occipital lobe and thalamus (n=2), the midbrain and cerebellum (n=1), and the occipital lobe alone (n=1). The clinical records were reexamined to verify whether any new signs and symptoms of typical PC infarcts had emerged or had been detected after admission; this was found to be the case in only one of 12 patients (8%), who had a midbrain and cerebellar infarction and exhibited a vertical gaze palsy that had probably been overlooked in the acute phase.
Among patients with PC infarct, the initial CT scan never showed early focal hypodensity or mass effect in the carotid distribution or a hyperdense MCA (Table⇑). These features were instead found in 59%, 31%, and 15% of patients with AC infarct, respectively. At a multivariate analysis including clinical and CT findings, focal parenchymal hypodensity within the carotid territory was the only factor able to discriminate between PC and AC infarcts. The corresponding odds ratio was undefined because no patients with PC infarct had an initial focal hypodensity on CT. The latter finding predicted the development of an AC infarct in 100% of cases (95% confidence interval, 97% to 100%), whereas the negative predictive value was as low as 15% (95% confidence interval, 11% to 19%).
During hospitalization, osmotics and intravenous or subcutaneous heparin were administered with similar frequency to PC and AC patients. As shown in the Table⇑, the prevalence of neurological deterioration within the first 4 days (defined as death, coma, or decrease of ≥1 point on the Canadian Neurological Scale in relation to baseline values), 1-month case-fatality rate, and disablement in survivors were similar in the two groups.
In our consecutive series of patients observed within 5 hours of the onset of a cerebral infarct whose clinical presentation did not suggest a PC stroke, we determined that 8% actually had an infarct in the brain stem/cerebellum or in the supratentorial territory of the posterior cerebral artery. An additional nine patients (6%) had a permanently negative CT and a pure motor hemiparesis, two findings that, when associated, suggest a diagnosis of pontine infarction.2 This means that approximately 10% to 20% of patients with a very early diagnosis of presumed AC stroke may have a PC infarct, a finding that casts doubt on the accuracy of a bedside diagnosis of infarct localization based on the arterial territory.
Theoretically, the relatively rapid examination (lasting 10 to 20 minutes) or the poor collaboration of acute stroke patients may mean that signs typical of PC involvement were missed in the emergency setting. This, however, occurred in only one of our patients, whose vertical gaze paresis was overlooked at the initial neurological examination and was disclosed by a more accurate clinical examination performed the day after stroke. Our results are therefore in agreement with those of previous studies which found that it is not always possible to distinguish AC from PC involvement based on the clinical presentation at entry.2 3 4 11 12 Early CT signs of infarct in the AC had an excellent positive predictive value for a final AC infarct but a largely unsatisfactory negative predictive value. Therefore, an unremarkable emergency CT did not help in the discrimination between AC and nontypical PC strokes.
The risk factor profile was comparable between our AC and PC patients, suggesting a similar etiology. The overall clinical course and final outcome were also comparable. However, in the acute phase, clinicians should bear in mind that a unilateral weakness may be indicative of an occlusion of the basilar artery,13 a life-threatening condition that requires prompt, aggressive treatment to restore perfusion.14 Another scenario in which territorial misdiagnosis generates therapeutic errors is typified by the case reported by Nighoghossian et al.2 This patient presented with a right pure hemiparesis, had a normal CT, and underwent endarterectomy for a left carotid stenosis disclosed by a selective carotid angiography. However, months later he developed a contralateral pure motor hemiparesis, and on that occasion MRI was still unremarkable at the supratentorial level but showed bilateral paramedian pontine infarcts. From cases like this it might be inferred that MRI is a more accurate tool for a differential diagnosis of territorial involvement, especially if the lesion is located in the brain stem. However, the sensitivity of MRI in the very acute phase of ischemia is not as high as that in the subacute chronic phase,15 16 and its real advantage over CT in this setting has yet to be established.17
The bedside differential diagnosis between AC and nontypical PC infarct in the first few hours after stroke is unreliable, as is that of ischemic versus hemorrhagic stroke18 and territorial versus lacunar infarct19 ; it can, in fact, be made only with a certain amount of confidence on the basis of the higher incidence of AC infarction. Diagnoses such as hemispheric stroke, AC stroke, or MCA stroke formulated in the first few hours after onset are probably not valid eligibility criteria for clinical trials because there is a sizable number of cases in which they are likely to be mistaken.
This study was supported by the Italian National Research Council (grant CNR 93.00465.PF40). Cesare Fieschi provided invaluable advice during the preparation of the manuscript. Lewis Baker reviewed English style and grammar.
Reprint requests to Corrado Argentino, MD, I Clinica Neurologica, Dipartimento di Scienze Neurologiche, Viale dell'Università 30, 00185 Roma, Italy.
- Received January 4, 1996.
- Revision received May 6, 1996.
- Accepted May 9, 1996.
- Copyright © 1996 by American Heart Association
Caplan LR. Stroke: A Clinical Approach. 2nd ed. Boston, Mass: Butterworth-Heinemann; 1995.
Nighoghossian N, Ryvlin P, Trouillas P, Laharotte JC, Froment JC. Pontine versus capsular pure motor hemiparesis. Neurology. 1993;43:2197-2201.
Chambers BR, Broder RJ, Donnan GA. Proximal posterior cerebral artery occlusion simulating middle cerebral artery occlusion. Neurology. 1991;41:385-390.
Hommel M, Besson G, Pollack P, Kahane MP, LeBas JF, Perret J. Hemiplegia in posterior cerebral artery occlusion. Neurology. 1990;40:1496-1499.
Côté R, Hachinski VC, Shurvell BL, Norris IW, Wolfson C. The Canadian Neurological Scale: a preliminary study in acute stroke. Stroke. 1986;17:731-737.
Mahoney FJ, Barthel DW. Functional evaluation: the Barthel Index. Md Med J. 1965;21:61-65.
Dixon WJ, ed. BMDP Statistical Software, 1990 Release. Berkeley, Calif: University of California Press; 1990.
Bogousslavsky J, Maeder P, Regli M, Meuli R. Pure midbrain infarction: clinical syndromes, MRI and etiologic pattern. Neurology. 1994;44:2032-2040.
Hacke W, Zeumer H, Ferbert A, Brueckman H, del Zoppo G. Intra-arterial thrombolytic therapy improves outcome in patients with acute vertebrobasilar occlusive disease. Stroke. 1988;19:1216-1222.
Yuh WT, Crain MR, Loes DJ, Greene GM, Ryals TJ, Sato Y. MR imaging of cerebral ischemia: findings in the first 24 hours. AJNR Am J Neuroradiol. 1991;12:621-629.
Alberts MJ, Faulstich ME, Gray L. Stroke with negative brain magnetic resonance imaging. Stroke. 1992;23:663-667.
Hawkins GC, Bonita R, Broad JB, Anderson NE. Inadequacy of clinical scoring system to differentiate stroke subtypes in population-based studies. Stroke. 1995;26:1338-1342.
Toni D, Del Duca R, Fiorelli M, Sacchetti ML, Bastianello S, Giubilei F, Martinazzo C, Argentino C. Pure motor hemiparesis and sensorimotor stroke: accuracy of the very early clinical diagnosis of lacunar strokes. Stroke. 1994;25:92-96.