Figure. Diagnostic tests in young adults with ischemic stroke. The lighter bars show the percentage of patients who underwent a certain test and the black bars show the percentage of tests showing a positive result relative to stroke etiology (ie, diagnostic yield). CSF indicates cerebrospinal fluid; CT, computed tomography; MRI, magnetic resonance imaging. See “Methods” section for individual tests included under each panel.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Ji R, Schwamm LH, Pervez MA, Singhal AB. Ischemic Stroke and Transient Ischemic Attack in Young Adults: Risk Factors, Diagnostic Yield, Neuroimaging, and Thrombolysis. JAMA Neurol. 2013;70(1):51–57. doi:10.1001/jamaneurol.2013.575
Author Affiliations: Department of Neurology, Massachusetts General Hospital, Boston, Massachusetts (Drs Ji, Schwamm, Pervez, and Singhal); and Department of Neurology, Capital Medical University, Beijing, China (Dr Ji).
Background Approximately 10% to 14% of ischemic strokes occur in young adults.
Objective To investigate the yield of diagnostic tests, neuroimaging findings, and treatment of ischemic strokes in young adults.
Design We retrospectively reviewed data from our Get with the Guidelines–Stroke database from 2005 through 2010.
Setting University hospital tertiary stroke center.
Patients A total of 215 consecutive inpatients aged 18 to 45 years with ischemic stroke/transient ischemic attack. The mean (SD) age was 37.5 (7) years; 51% were male.
Results There were high incidence rates of hypertension (20%), diabetes mellitus (11%), dyslipidemia (38%), and smoking (34%). Relevant abnormalities were shown on cerebral angiography in 136 of 203 patients, on cardiac ultrasonography in 100 of 195, on Holter monitoring in 2 of 192; and on hypercoagulable panel in 30 of 189 patients. Multiple infarcts were observed in 31% and were more prevalent in individuals younger than age 35 years. Relevant arterial lesions were frequently detected in the middle cerebral artery (23%), internal carotid artery (13%), and vertebrobasilar arteries (13%). Cardioembolic stroke occurred in 47% (including 17% with isolated patent foramen ovale), and 11% had undetermined stroke etiology. The median National Institutes of Health Stroke Scale score was 3 (interquartile range, 0-9) and 81% had good outcome at hospital discharge. Of the 29 patients receiving thrombolysis (median National Institutes of Health Stroke Scale score, 14; interquartile range, 9-17), 55% had good outcome at hospital discharge and none developed symptomatic brain hemorrhage.
Conclusions This study shows the contemporary profile of ischemic stroke in young adults admitted to a tertiary stroke center. Stroke etiology can be determined in nearly 90% of patients with modern diagnostic tests. The causes are heterogeneous; however, young adults have a high rate of traditional vascular risk factors. Thrombolysis appears safe and short-term outcomes are favorable.
Quiz Ref IDApproximately 10% to 14% of ischemic strokes occur in adults ages 18 to 45 years.1-7 Stroke etiology in this population differs by geographic region and has greater heterogeneity than in older individuals.5,8,9 Even within this population, the etiologic spectrum varies according to study type (hospital based1-3 vs epidemiologic10), the extent and availability of diagnostic evaluation, or the criteria used to ascribe etiology. To our knowledge, data concerning ischemic stroke in young adults in the United States is limited to a few studies published more than a decade ago.2-5 Advances in stroke diagnostic tests, particularly neuroimaging, refinement in classification schema,11 and the advent of thrombolysis and stroke unit care, may have significantly impacted the profile of ischemic stroke in young adults. We studied the diagnosis, management, and outcome of ischemic stroke in young adults admitted to our comprehensive stroke center. A major objective was to determine the yield of diagnostic tests and characterize brain and vascular imaging findings, which, to our knowledge, has not been the focus of prior studies.
This retrospective study was approved by our hospital's human research committee. We reviewed our American Heart Association Get with the Guidelines–Stroke database12 from 2005 through 2010 (n = 2643 cases) to identify 215 consecutive patients ages 18 to 45 years admitted with ischemic stroke or transient ischemic attack (TIA). Ischemic stroke was defined as a sudden focal neurologic deficit with imaging-confirmed infarct, and TIA as a transient focal neurologic deficit without stroke on brain imaging. We excluded patients with stroke from cerebral sinovenous thrombosis, subarachnoid hemorrhage, or interventions.
Medical records and brain scans were reviewed by a stroke neurologist (R.J.), with supervision (A.B.S.). We extracted data on patient demographics, relevant medical history (Table 1), National Institutes of Health Stroke Scale (NIHSS) score at admission (obtained prospectively as per our clinical stroke protocol), diagnostic test results, in-hospital treatment, and the modified Rankin Scale (mRS) score at hospital discharge. A favorable outcome was defined as a mRS score of 0 to 2 at hospital discharge.
The results of relevant blood tests were categorized as normal or abnormal based on our laboratory's reference ranges (Table 2). Cerebrospinal fluid examination results were positive if the white blood cell count was 5/μL or greater (to convert to ×109 per liter, multiply by 0.001) or the protein level was greater than 0.06 g/dL (to convert to grams per liter, multiply by 10.0). Transthoracic and transesophageal echocardiography reports were reviewed for cardioembolic sources, as described previously.11 Electrocardiography and Holter monitoring were considered positive if they showed atrial fibrillation/flutter. Lower-extremity Doppler ultrasonography and pelvic magnetic resonance venography were considered positive if deep vein thrombosis was documented.
We reviewed diffusion-weighted imaging and fluid-attenuated inversion recovery sequences for lesion topography. If magnetic resonance imaging was not performed, computed tomography was reviewed. The presence, number, location, laterality, and arterial territory of ischemic lesions were noted. Transfemoral, computed tomographic, and magnetic resonance angiography were reviewed for the presence of occlusion or stenosis in the culprit artery.
Stroke subtyping11 was based on information available at hospital discharge. Special consideration was given to patent foramen ovale (PFO) because of its controversial role in stroke and high detection rate in young stroke patients. Accordingly, PFO-associated strokes were subclassified as PFO with atrial septal aneurysm (ASA), PFO Plus (ie, PFO and predisposing risk for venous clot formation [eg, recent airline travel]), and isolated PFO.
SPSS version 16.0 for windows was used for statistical analysis. χ2, Fisher exact, and t tests were used as appropriate. P < .05 was considered statistically significant.
Quiz Ref IDOf 215 patients, 203 (94%) had ischemic stroke and 12 (6%) had TIA. The mean (SD) age was 37.5 (7) years, with a nearly equal distribution of men and women in the full cohort and in the 2 age groups. Most patients were white, as expected for our referral base in the Northeast United States. Quiz Ref IDHypertension (20%), diabetes mellitus (11%), dyslipidemia (38%), and smoking (34%) were common (Table 1). There were high frequency rates of prior headache (14%), illicit drug use (12%), and oral contraceptive use (9%). Hypertension and diabetes mellitus were more common in the 36-year-old to 45-year-old age group, and illicit drug use more common in the 18-year-old to 35-year-old age group. Men had a higher frequency of dyslipidemia, while women had a higher frequency of prior headaches. Three patients had postpartum stroke, and 2 had cancer-associated strokes.
Most of the patients underwent extensive blood tests, Holter monitoring, and echocardiography (Table 2, Figure). Cerebrospinal fluid examination and toxicology screening were performed when clinically indicated. The diagnostic yield was relatively low for Holter monitoring (1%), toxicology screening (5%), and vasculitis panel (<5%), as well as relatively high for echocardiography (51%) and angiography (64%). Lipid panel result abnormalities were significantly more common in men. Cerebrospinal fluid examination results were abnormal in 5 of 34 patients (15%).
Hypercoagulable panel screening results were positive in 16% of patients, including 3% with hereditary prothrombin gene or factor V Leiden mutations and 13% with low protein C, protein S, or antithrombin III levels. The most common abnormality was a low protein S level, mostly in women (18% vs 5%; P = .004), including 6 women taking oral contraceptives. Eight patients had persistently low levels on follow-up testing.
Contrast transthoracic echocardiography was performed in 91% of patients and transesophageal echocardiography in 40%. A positive transthoracic echocardiography or transesophageal echocardiography result was documented in 100 of 195 patients (51%). Ninety-six of 100 patients with a positive echocardiography result had a PFO, including 13 with ASA and 83 without ASA. Patent foramen ovale was implicated as a stroke etiology in 76 patients and was considered incidental in cases with alternate causes such as carotid dissection.
Computed tomography was performed in 97% of patients and magnetic resonance imaging in 98%; 203 patients had brain infarctions. Single infarcts were observed in 69% (Table 3). Multiple infarcts were more common in the 18-year-old to 35-year-old age group (41% vs 25%; P = .03). Supratentorial infarcts were more common in women (89% vs 67%; P < .01). Patients with supratentorial infarcts had significantly higher NIHSS scores at admission compared with those with infratentorial infarcts (median NIHSS score, 7 vs 3, respectively; P < .001). Of the affected arterial territories, 56% were middle cerebral, 3% were anterior cerebral, 7% were combined middle and anterior cerebral, 8% were posterior cerebral, 21% were vertebrobasilar, and 5% were multiple.
Transfemoral, computed tomographic, or magnetic resonance cerebral angiography was performed in 99% of patients and 20% additionally underwent vascular ultrasonography studies. Embolic occlusion, severe stenosis (>70% by visual estimate), or underlying arteriopathies such as premature atherosclerosis (Table 4) or cerebral artery dissection or Moyamoya disease (Table 5) were observed in 64% of patients, including 42% with angiographic abnormalities in the anterior circulation arteries, 14% in the posterior circulation arteries, and 8% with diffuse abnormalities from conditions such as reversible cerebral vasoconstriction syndrome. Proximal middle cerebral artery lesions were most common (23%), followed by lesions of the internal carotid artery and the vertebrobasilar system.
The most common stroke subtype was cardioembolic (47%): 5% were PFO with ASA, 14% were PFO Plus, 17% were isolated PFO, and 11% had established cardioembolic sources (Table 4 and Table 5). Large-artery atherosclerosis and small-vessel disease were relatively uncommon, and they occurred exclusively in patients older than age 36 years. More than one-third of patients proved to have other well-defined causes. Multiple etiologies were found in 4 patients (2%). The etiology remained undetermined in only 9%, including 18 patients with cryptogenic stroke despite a thorough workup and 1 patient who died before testing was completed.
Four patients underwent hemicraniectomy for malignant brain edema. At the time of hospital discharge, 87% of patients were taking antiplatelet agents or anticoagulants (5% were on both), and 32% were taking statins. The median NIHSS score at admission was 3 (interquartile range, 0-9). The mean (SD) length of stay was 6.2 (6) days (range, 1-54 days). More than 80% of patients had good clinical outcome (mRS score, 0-2). Poor outcome (mRS score, 5-6) was documented in 29 patients, including 7 who died (Table 4).
Thrombolysis was administered to 29 patients (13.5%), including 19 who received intravenous tissue plasminogen activator therapy alone, 6 who received intra-arterial thrombolysis, and 4 who received combined intravenous/intra-arterial thrombolysis. Their median NIHSS score at admission was 14 (interquartile range, 9-17) and mean (SD) length of stay was 7.1(5) days (range, 1-20 days). Angiographic occlusion was documented in 93%. Sixteen patients (55%) had good outcome, and none developed symptomatic intracerebral hemorrhage.
Our study shows the contemporary profile of ischemic stroke in young adults in a tertiary referral center in the Northeast United States. The age and sex distribution in our cohort is similar to that of other US-based studies with a similar age range.2,3 However, studies with a higher age cutoff have found a higher proportion of males.1Quiz Ref IDSimilar to recent studies,1,7 there were high incidence rates (compared with young individuals without stroke13) of modifiable stroke risk factors such as dyslipidemia, hypertension, diabetes mellitus, and smoking, which contribute to higher rates of recurrent vascular events.14 Despite the relatively high incidences of modifiable risk factors, stroke etiology was attributed to large-artery atherosclerosis and small-vessel disease in less than 10%, suggesting that vascular risk factors increase the susceptibility to stroke from other causes. These data re-emphasize the need to implement evidence-based stroke prevention targeting traditional risk factors in young adults.1,7
A novel feature is that 99% of patients were adjudicated as having a complete diagnostic workup. An extensive battery of tests was routinely performed, which might explain why we found a relatively low rate of undetermined stroke etiology compared with other studies.1,3,8 The use of a validated algorithm11,15 to classify etiology into standard TOAST (Trial of Org 10172 in Acute Stroke Treatment) subtypes lends confidence to our results. Patent foramen ovale–associated strokes were considered cardioembolic, as per TOAST criteria.16 Unlike prior publications,17 we further classified PFO-associated strokes based on the presumed risk level (ie, PFO with ASA, PFO Plus, and isolated PFO). This subclassification was justified by the high detection rate of PFO in our study, the possible significance of transient hypercoagulable panel abnormalities in patients with PFO, and existing uncertainties about PFO treatment,17 as well as the relationship between PFO and stroke. This level of detail will allow reclassification of isolated PFO or PFO Plus from cardioembolic to undetermined, if the stroke risk attributable to PFO changes in the future.
Quiz Ref IDYoung adults with stroke typically undergo a wide spectrum of diagnostic tests, but there is little knowledge about the tests' yield or cost-effectiveness. We found a low yield for inpatient Holter monitoring, serum/urine toxicology, and vasculitis panel tests. Hypercoagulable panel results at admission were positive in 30 patients, 17 of whom had transient hypercoagulable states; the rest had persistent abnormalities or inherited mutations. Conceivably, transient reductions in the levels of these clotting cascade proteins elevate stroke risk, particularly in patients with PFO, which further justifies the PFO Plus category used in our study. We note that the association between thrombophilia and stroke remains controversial, and the consequences of treating young adults with anticoagulants based on such results needs to be carefully considered.18 Cardiac ultrasonography was positive in 100 patients (51%), including 96 with a PFO. The true diagnostic yield of cardiac ultrasonography is probably much lower because PFO was deemed incidental in patients with alternate etiologies and because the significance of an isolated PFO remains uncertain.19 Our data reflect the diagnostic yield of tests in a real world academic stroke setting. Further studies are warranted to determine cost-effectiveness and design appropriate stroke diagnostic algorithms in young adults.
The detailed analysis of brain and vascular imaging provides insights into the pathophysiology of stroke in this population. Single infarctions were more common in the 36-year-old to 45-year-old age group, consistent with the higher incidence of small-vessel disease and atherosclerosis in this age group. Both age groups showed a similar distribution of supratentorial vs infratentorial infarcts. In contrast, Putaala et al1 found no significant difference in the number of single infarcts between age groups, and older individuals had a higher incidence of anterior-circulation lesions. These differences may be explained by the distinct populations and different age groups being studied. We found a high yield of vascular imaging (substantially higher than in the general stroke population20,21 or thrombolysed patients22), and arteriopathies are known to be the most common cause of stroke in young individuals.23 These data justify routine cerebrovascular imaging in young adults with stroke-like symptoms.
In our study, nearly 15% of patients required constant nursing care or died before hospital discharge. Young stroke survivors are known to develop substantial emotional and socioeconomic issues,24 and more than 10% develop recurrent vascular events within 5 years.14 From the perspective of secondary prevention, it is reassuring that 92% of our patients received antiplatelet or anticoagulant therapy, or both, at hospital discharge. However, the ideal duration, safety, and efficacy of these medications in young stroke patients is not clear. Cholesterol-lowering medications were administered to only 32% of patients, possibly reflecting uncertainties in exposing young adults with nonatherosclerotic stroke to long-term statin therapy. A relatively large number received thrombolysis, and no patient developed hemorrhagic complications; however, to our knowledge, there is little data concerning the safety and efficacy of thrombolysis in young adults or those with relatively uncommon stroke etiologies. Such knowledge gaps need to be addressed in prospective studies, and treatment and prevention guidelines that specifically address stroke in young adults need to be developed.
In conclusion, this study provides contemporary information on risk factors, diagnostic tests, imaging, thrombolysis, and secondary prevention of ischemic stroke in young adults. The clinical setting in a tertiary stroke center limits the generalizability of our results. Nevertheless, our data may prove useful in developing cost-effective diagnostic strategies, understanding the pathophysiology, and refining the management of ischemic stroke in young adults.
Correspondence: Aneesh B. Singhal, MD, ACC-729C, Department of Neurology, Massachusetts General Hospital, Boston, MA 02114 (firstname.lastname@example.org).
Accepted for Publication: May 16, 2012.
Published Online: October 29, 2012. doi:10.1001/jamaneurol.2013.575
Author Contributions:Study concept and design: Singhal. Acquisition of data: Ji, Schwamm, Pervez, and Singhal. Analysis and interpretation of data: Ji, Schwamm, Pervez, and Singhal. Drafting of the manuscript: Ji, Pervez, and Singhal. Critical revision of the manuscript for important intellectual content: Ji, Schwamm, Pervez, and Singhal. Statistical analysis: Ji and Singhal. Administrative, technical, and material support: Schwamm. Study supervision: Schwamm, Pervez, and Singhal.
Conflict of Interest Disclosures: Dr Schwamm has served as a consultant for the Massachusetts Department of Health and Lundbeck International Steering Committee, and as an unpaid chair for the American Heart Association's Get with the Guidelines–Stroke program.
Funding/Support: This study was funded by grants R01NS051412 and R21NS077442 from the National Institutes of Health/National Institute of Neurological Disorders and Stroke; grant 30700240 from the National Nature Science Foundation of China; and grant 2008B30 from the Nova Program, Beijing Science and Technology Commission.