Is Isolated Aphasia a Typical Presentation of Presumed Cardioembolic Transient Ischemic Attack or Stroke?Fennis Th.F.M.a · Compter A.b · van den Broek M.W.C.a · Koudstaal P.J.d · Algra A.c · Koehler P.J.e
aDepartment of Neurology, Ziekenhuis St. Jansdal, Harderwijk, bUMC Utrecht Stroke Center, Department of Neurology and Neurosurgery and cJulius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht, dDepartment of Neurology, Erasmus Medical Center, Rotterdam, and eDepartment of Neurology, Atrium Medical Center, Heerlen, the Netherlands Corresponding Author
Background: Previous studies have suggested that patients with a transient ischemic attack (TIA) or minor ischemic stroke and isolated aphasia should be carefully screened for a potential cardiac source of embolism. Most of these publications, however, were case reports or small-series. The purpose of this study was to assess the relationship between isolated aphasia and atrial fibrillation (AF) as the cause of presumed cardioembolic TIA or stroke within the setting of 2 large multicenter trials. Methods: The frequency of isolated aphasia was compared between patients with a TIA or minor ischemic stroke either with AF [European Atrial Fibrillation Trial (EAFT), n = 1,001] or without AF [Dutch TIA Trial (DTT), n = 3,150]. We analyzed data with univariable and multivariable logistic regression. Isolated aphasia was defined as aphasia without dysarthria, visual-field defects or motor or sensory deficits of the arm, leg or face. Because dysarthria can be difficult to detect in aphasic patients, a second analysis was done without excluding dysarthric patients. In a third analysis, we excluded patients with a symptomatic lacunar infarct from the DTT, as these patients were overrepresented due to the exclusion of patients with AF. Subgroup analysis was performed for patients presenting with TIA and minor stroke. Results: Of 4,151 patients, 210 (5.1%) had isolated aphasia, 109 from the EAFT and 101 from the DTT, crude odds ratio (OR) 3.69, 95% confidence interval (CI) 2.79-4.89. Patients with isolated aphasia were older (mean age 70.3 vs. 66.8 years, p < 0.01), more often female (OR 1.87, 95% CI 1.41-2.46), and more often had diabetes (OR 1.73, 95% CI 1.16-2.59) and hypercholesterolemia (OR 1.83, 95% CI 1.11-3.03) than those without aphasia. After simultaneous adjustment for age, sex, diabetes and hypercholesterolemia, patients with isolated aphasia still had AF more often than patients without isolated aphasia (adjusted OR 2.94, 95% CI 2.16-4.01). Both after inclusion of patients with dysarthria in the group of patients with isolated aphasia and after exclusion of patients with a symptomatic lacunar infarct, essentially the results remained the same. Patients presenting with isolated aphasia due to a TIA tended to have AF more often than patients with a minor ischemic stroke. Conclusions: Isolated aphasia is an independent sign of AF in patients with a TIA or minor ischemic stroke. Careful cardiac screening seems warranted in patients with isolated aphasia, as secondary prevention is different in patients with a cardiac source of embolism.
© 2013 S. Karger AG, Basel
Acute isolated aphasia without paresis or other neurological symptoms is a well-known clinical stroke syndrome. Previous studies have suggested that patients with isolated global aphasia should be carefully screened for a potential cardiac source of embolism [1,2,3]. Most of these publications, however, were case reports or small-series. Moreover, the definition of isolated aphasia varied. If isolated aphasia would indeed be related to cardioembolic stroke, this would have important clinical implications. In patients presenting with persistent atrial fibrillation (AF), a cardiac embolic source is evident, but in those with paroxysmal AF or other subclinical cardiac disorders, further investigations are needed to detect a cardioembolic source. Once found, secondary prevention in these patients will be different from in those in whom a vascular cause of the stroke is most probable . The purpose of this study was to investigate the relationship between isolated aphasia and AF as presumed cardioembolic stroke within the setting of 2 large multicenter trials.
The frequency of isolated aphasia was compared between patients from 2 randomized controlled trials which included patients with a TIA or nondisabling ischemic stroke (grade 3 or less on the modified Rankin scale) in the previous 3 months. One trial included only patients with (nonrheumatic) AF [European Atrial Fibrillation Trial (EAFT)]  and the other only patients without AF [Dutch TIA trial (DTT)] [6,7]. In the EAFT trial, 1,007 patients were recruited. Five of them had been enrolled inappropriately and the notification form for 1 was missing; consequently, the data of 1,001 patients were available for our analysis. In the DTT, 3,150 patients participated in the study. In both studies, the clinical symptoms of the qualifying event were recorded in detail. Isolated aphasia was defined as aphasia without dysarthria, visual-field defects or motor or sensory deficits of the arm, leg or face. Because dysarthria can be difficult to detect in aphasic patients, a second analysis was made without excluding dysarthric patients. Furthermore, in the DTT, patients with AF were excluded, leading to an overrepresentation of symptomatic lacunar infarcts . We therefore did a third analysis excluding patients with symptomatic lacunar infarcts from the DTT. Logistic regression was used to calculate odds ratios (ORs) and corresponding 95% confidence intervals (CIs). We adjusted ORs for differences between patients from the 2 trials. In addition, we did a subgroup analysis comparing patients presenting with a TIA and minor ischemic stroke. Both studies were approved by the institutional review board and were thus performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki. All patients provided informed consent prior to their inclusion. When the patient was unable to understand the study procedures due to the TIA or stroke, consent was obtained from a proxy.
In this study, we analyzed 4,151 patients; 3,150 (75.9%) without AF and 1,001 (24.1%) with AF (table 1). The male:female ratio was 1:0.6. Isolated aphasia was present in 210 (5.1%) patients. Patients with isolated aphasia more often had AF, diabetes and hypercholesterolemia than patients with other stroke types (table 2). In addition, patients with isolated aphasia were older (mean age 70.3 vs. 66.8 years, p < 0.01) and were more often female (OR 1.87, 95% CI 1.41-2.46). After simultaneous adjustment for age, sex, diabetes and hypercholesterolemia, patients with isolated aphasia still had AF more often than patients without isolated aphasia (table 2). Inclusion of dysarthric patients in the group of patients with isolated aphasia essentially gave the same results, with minimal differences in risk estimates and overlapping 95% CIs. After exclusion of patients with symptomatic lacunar infarcts in the DTT, the results remained essentially the same. The relationship between isolated aphasia and AF tended to be stronger amongst patients with a TIA (adjusted OR 4.33, 95% CI 2.44-7.70) than patients with a minor ischemic stroke (adjusted OR 2.55, 95% CI 1.77-3.67).
|Table 1. Baseline characteristics|
|Table 2. Relationship of isolated aphasia with several baseline characteristics, in particular, atrial fibrillation|
Our data show that in patients with a TIA or minor ischemic stroke who present with isolated aphasia, AF is more common than in TIA or stroke patients with other symptoms or a combination of symptoms. This confirms the hypothesis that cardioembolic sources are more likely to be found in stroke patients with isolated aphasia than in those with other symptoms.
Before our results can be accepted, some issues with respect to the differences between the 2 randomized clinical trials used in our study need to be addressed. Aphasia and other signs of cortical infarction are more frequent than lacunar stroke in patients with AF [9,10,11]. In the DTT, patients with a cardiac source of embolism were excluded, which led to an overrepresentation of patients with a lacunar infarct . This implies that in the non-AF group, aphasia was a priori less frequent. In a subgroup analysis, we adjusted for the difference between the 2 trials by excluding patients from the DTT who fulfilled the Trial of Org 10172 in Acute Stroke Treatment (TOAST) criteria for symptomatic lacunar infarcts. This did not influence our results much. However, classification of stroke subtype using clinical and imaging data remains inferior to pathological confirmation.
Second, both trials included patients with grade 3 or less on the modified Rankin scale. The majority of patients presenting with isolated aphasia are independent in daily living; however, patients with severe global aphasia are not. It is unlikely that these patients were excluded more often in 1 trial than in the other. This assumption is justified, as the catchment areas of the trials were largely similar. Even if exclusion of isolated aphasia would have occurred more often in 1 of the trials, it would have been more likely to happen in the AF group, which would mean that our results are an underestimation of the relationship of isolated aphasia with AF. Third, the mean age of AF patients was higher than non-AF patients, reflecting the fact that the frequency of AF increases with age. This probably also explains why women were overrepresented in the AF group, as their life expectancy is greater. The relationship between isolated aphasia and AF remained after adjustment for age and sex. Interestingly, this relation tended to be stronger for patients presenting with a TIA. This could partly be explained by the tendency of TIAs to be more frequent in the left hemisphere than ischemic strokes .
Global aphasia typically results from large perisylvian lesions, therefore adjacent motor areas are usually affected [2,10]. Consequently, isolated global aphasia without hemiparesis is a striking and relatively rare phenomenon. The hypothesis that isolated aphasia is of cardioembolic origin was first introduced 1982 when 3 patients were described with global aphasia without hemiparesis . These patients had two anatomical lesions in both the anterior (Broca) and posterior (Wernicke) language areas without affecting the motor cortex in between. Because separate branches of the middle cerebral artery supply these two language areas, an embolic source of stroke was suggested. Subsequent reports described a number of cases suggesting other anatomic possibilities, some with only one lesion [13,14,15,16]. The majority of cases seem to be of cardioembolic origin; in a combined analysis of 22 patients with isolated aphasia, 13 patients met the TOAST criteria for cardioembolic stroke [2,17]. Some of these cases, however, presented with a right-sided hemiparesis that improved quickly.
In conclusion, our data suggest that isolated aphasia is an independent sign of a potential cardiac source of embolism. This finding may have implications for TIA or ischemic stroke patients who present with isolated aphasia and in whom there is no obvious cardiac source of emboli, e.g. in patients with paroxysmal AF. In such patients, a careful cardiac screening, including prolonged rhythm monitoring, seems warranted.
We gratefully acknowledge the contribution of the EAFT and DTT study groups.
UMC Utrecht Stroke Center, Department of Neurology and Neurosurgery
G 03.232, University Medical Center Utrecht, PO Box 85500
NL-3508 GA Utrecht (The Netherlands)
Received: October 12, 2012
Accepted: January 23, 2013
Published online: April 23, 2013
Number of Print Pages : 4
Number of Figures : 0, Number of Tables : 2, Number of References : 17
Vol. 35, No. 4, Year 2013 (Cover Date: May 2013)
Journal Editor: Hennerici M.G. (Mannheim)
ISSN: 1015-9770 (Print), eISSN: 1421-9786 (Online)
For additional information: http://www.karger.com/CED