Stroke Literature Synopses: Clinical Science
Approximately one third of strokes are cryptogenic. Two recent articles address factors that contribute to our inability to determine stroke pathogenesis: inadequate diagnostic evaluations and unrecognized stroke risk factors.
Sipola et al (Computed tomography and echocardiography together reveal more high-risk findings than echocardiography alone in the diagnostics of stroke etiology. Cerebrovasc Dis. 2013;35:521–530) performed a prospective, single-center observational study assessing the accuracy of cardiac, aortic, and cervicocranial computed tomography angiography (CACC-CT) in addition to transthoracic and transesophageal echocardiography (TTE/TEE) for individuals with suspected cardiogenic ischemic stroke or transient ischemic attack (n=140; mean age, 60±10 years; 68% males). Individuals were included if they were <50 years of age and had a cryptogenic stroke or transient ischemic attack or if they had a stroke or transient ischemic attack with a suspected cardiogenic source (ie, history of previous heart disease, simultaneous or sequential strokes in different arterial territories, hemorrhagic transformation, simultaneous emboli to other organs, decreased consciousness at stroke onset, isolated aphasia, or visual field defect). Individuals with atrial fibrillation were excluded. High-risk cardiac findings included signs of previous transmural myocardial infarction, left ventricular ejection fraction <30%, intracardiac thrombus or tumor, infiltrative tumor into the pulmonary veins, valvular fibroelastoma or vegetation, mitral stenosis, and ≥4-mm-thick aortic atheromatous plaques.
Nine high-risk cardiac findings were found by TTE/TEE compared with 20 with CACC-CT (sensitivity, 41% versus 68%; P=0.052). The combination of CACC-CT and TTE/TEE improved sensitivity to 91%, with 98% specificity and 97% diagnostic accuracy. CACC-CT was more sensitive than TTE/TEE in detecting previous myocardial infarction, aortic plaque, tumors, and aortic thrombi; whereas TTE/TEE was more sensitive in detecting cardiac thrombi and fibroelastomas. CACC-CT is not useful for assessing ejection fraction or patent foramen ovale.
This study, which suggests that the combination of CACC-CT and TTE/TEE increases the yield of detecting high-risk cardioembolic sources compared with either evaluation alone, builds on previous literature that compared accuracy of one evaluation versus the other. The study has several limitations, including small sample size, homogenous population (all white, mostly male), variable experience of cardiologists performing echocardiograms, variations in imaging techniques (eg, 16-slice CT earlier in study versus 64-slice CT later), delay from symptom onset to diagnostic imaging (≈6 days), lack of established diagnostic criteria for CACC-CT imaging findings, and omission of possible evaluation-related complications. In addition, the authors’ criteria for suspected cardioembolism included conditions that may have been attributable to intracranial atherosclerosis (aphasia, visual, artery-to-artery embolism, or lacunar syndrome. Participants did not undergo brain MRI or intracranial arterial imaging, further compounding the problem. The study had a high proportion of overweight or obese (75%) individuals, limiting accuracy of echocardiography. In addition, all participants received both TTE and TEE; in practice, TEE is not typically performed if a cardiac source is detected on TTE. Finally, it remains unclear as to what extent the additional findings changed clinical management because some of the findings (eg, low ejection fraction, previous myocardial infarction, and aortic arch atherosclerosis) are not clear indications for anticoagulation. The additional information gleaned from CACC-CT needs to be weighed against the risks of the test (including allergic reaction to contrast, acute kidney injury, and radiation exposure). Future studies are needed in more diverse stroke or transient ischemic attack populations that have undergone thorough diagnostic evaluations, including brain MRI, intracranial and extracranial arterial imaging, laboratory evaluation, and cardiac rhythm monitoring, to determine whether the addition of CACC-CT adds sensitivity and changes management in individuals who have had cryptogenic or suspected cardioembolic stroke.
Previously unrecognized stroke risk factors may partially account for the high proportion of cryptogenic strokes. Cohort studies have revealed a higher risk of stroke in individuals with traumatic brain injury (TBI) compared with the general population; however, it is unclear whether stroke is a short-term consequence of TBI (eg, subarachnoid hemorrhage, dissection) or whether TBI confers a long-term risk of stroke. In a retrospective analysis of all emergency department visits and inpatient discharges for trauma in California from 2005 to 2009 (n=1 173 353), Burke et al (Traumatic brain injury may be an independent risk factor for stroke. Neurology. 2013;81:1–7) investigated the association between TBI and subsequent ischemic stroke. Individuals with TBI (n=436 630) and trauma patients without TBI were followed for a median duration of 28 months (interquartile range, 14–44). To exclude strokes that were direct sequelae of TBI, they excluded intracerebral hemorrhages, subarachnoid hemorrhages, and arterial dissections. Given the high incidence of seizures after TBI, and the potential for misdiagnosing a seizure as a stroke, the authors adjusted for emergency department visits and hospital admissions for epilepsy or convulsions. During the period of observation, 1% developed ischemic stroke (1.1% in TBI group and 0.9% in non-TBI group). After adjustment for covariates, individuals with TBI were more likely to be hospitalized for ischemic stroke than trauma patients without TBI (hazard ratio, 1.31; 95% confidence interval, 1.25–1.36). All TBI subtypes (skull fracture, concussion, cerebral laceration, other intracranial injury, and unspecified) had a similar magnitude of association with ischemic stroke. When stroke hospitalization within 7, 30, 60, or 365 days of trauma were excluded from the outcome measure, the ischemic stroke–TBI association only modestly decreased. The stroke–TBI association was more pronounced among individuals <50 years of age (odds ratio, 1.56; 95% confidence interval, 1.32–1.85) compared with those ≥50 years of age (odds ratio, 1.22; 95% confidence interval, 1.16–1.28).
The finding that TBI is a robust, independent long-term risk factor for ischemic stroke is noteworthy and novel; however, as the authors note, the absolute risk for stroke in this population remains small. In addition, because this study relies on retrospective analyses of administrative records, there are several limitations, including the inability to make causal associations, the possibility of unmeasured confounding, potential errors in diagnostic coding, and inability to ascertain pertinent clinical data. Further studies will be necessary to determine whether a similar association exists in other populations, to further delineate the pathophysiological mechanisms underlying the association between TBI and stroke, and to determine whether therapeutic interventions can be applied after TBI to reduce the risk of subsequent stroke.
- Received July 12, 2013.
- Accepted July 12, 2013.
- © 2013 American Heart Association, Inc.