CONCLUSION. COVID-19 is associated with large vessel occlusion (LVO) strokes but not with small vessel occlusion (SVO) strokes.
Citation: American Journal of Roentgenology: 1-6. 10.2214/AJR.20.23847
Reprinted for educational purposes and social benefit, not for profit.
BACKGROUND. An increase in frequency of acute ischemic strokes has been observed among patients presenting with acute neurologic symptoms during the coronavirus disease (COVID-19) pandemic.
OBJECTIVE. The purpose of this study was to investigate the association between COVID-19 and stroke subtypes in patients presenting with acute neurologic symptoms.
METHODS. This retrospective case-control study included patients for whom a code for stroke was activated from March 16 to April 30, 2020, at any of six New York City hospitals that are part of a single health system. Demographic data (age, sex, and race or ethnicity), COVID-19 status, stroke-related risk factors, and clinical and imaging findings pertaining to stroke were collected. Univariate and multivariate analyses were conducted to evaluate the association between COVID-19 and stroke subtypes.
RESULTS. The study sample consisted of 329 patients for whom a code for stroke was activated (175 [53.2%] men, 154 [46.8%] women; mean age, 66.9 ± 14.9 [SD] years). Among the 329 patients, 35.3% (116) had acute ischemic stroke confirmed with imaging; 21.6% (71) had large vessel occlusion (LVO) stroke; and 14.6% (48) had small vessel occlusion (SVO) stroke. Among LVO strokes, the most common location was middle cerebral artery segments M1 and M2 (62.0% [44/71]). Multifocal LVOs were present in 9.9% (7/71) of LVO strokes. COVID-19 was present in 38.3% (126/329) of the patients. The 61.7% (203/329) of patients without COVID-19 formed the negative control group. Among individual stroke-related risk factors, only Hispanic ethnicity was significantly associated with COVID-19 (38.1% of patients with COVID-19 vs 20.7% of patients without COVID-19; p = 0.001). LVO was present in 31.7% of patients with COVID-19 compared with 15.3% of patients without COVID-19 (p = 0.001). SVO was present in 15.9% of patients with COVID-19 and 13.8% of patients without COVID-19 (p = 0.632). In multivariate analysis controlled for race and ethnicity, presence of COVID-19 had a significant independent association with LVO stroke (odds ratio, 2.4) compared with absence of COVID-19 (p = 0.011).
CONCLUSION. COVID-19 is associated with LVO strokes but not with SVO strokes.
CLINICAL IMPACT. Patients with COVID-19 presenting with acute neurologic symptoms warrant a lower threshold for suspicion of large vessel stroke, and prompt workup for large vessel stroke is recommended.
Coronavirus disease (COVID-19) is a clinical manifestation of infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) , which was first detected in Wuhan, China, in December 2019 and declared a pandemic on March 11, 2020 . COVID-19 most commonly presents with pneumonialike symptoms, including fever, dyspnea, cough, and severe hypoxia. However, the spectrum of clinical manifestations is broader and includes gastrointestinal disease, headache, altered mental status, anosmia, and confusion [3, 4].
Neurologic manifestations of COVID-19 have been described. In one study from China , neurologic symptoms were found in as many as 36% of patients admitted to the hospital with COVID-19. Common manifestations include altered mental status, dysarthria, aphasia, and diffuse corticospinal tract signs , many of which may be seen in patients with acute ischemic infarction. Study results have suggested that endothelial injury in the vessel walls is caused by SARS-CoV-2 binding to angiotensin-converting enzyme 2 (ACE-2) receptors. The result is activation of a cytokine cascade that aggravates the vessels and ultimately leads to vascular thrombosis [6–8]. The exact pathophysiologic mechanism of SARS-CoV-2 infection has not yet been identified in human cerebral vessels. ACE-2 receptors were initially reported in the heart, kidney, and testis , though they were later found to also be expressed ubiquitously throughout blood vessels and in the glial tissue of the brain . Vascular thrombosis and a hypercoagulable state have been widely reported in patients with COVID-19, resulting in deep vein thrombosis [11, 12], myocardial infarction [13, 14], and pulmonary embolisms [15–17]. Further case reports [18, 19] have described patients with COVID-19 presenting with large vessel strokes.
We have observed an increase in frequency of acute ischemic strokes in patients presenting with acute neurologic symptoms during the COVID-19 pandemic. To our knowledge, no study has assessed the association of COVID-19 with either large vessel or small vessel strokes. In this study, we investigated the association between COVID-19 and stroke subtypes in patients presenting with acute neurologic symptoms.
Study Design and Participants
This retrospective case-control study was compliant with HIPAA. Institutional review board approval was obtained on an expedited basis with a waiver of written informed consent. The study included patients from a single hospital system consisting of six hospitals across the New York City boroughs of Manhattan, Queens, and Brooklyn. From March 16 to April 30, 2020, the hospital system treated a total of 9814 patients with known COVID-19 status, 5862 of whom had positive test results for COVID-19 (Fig. 1). This included 1334 patients in the ICU, 1018 of whom had COVID-19. A stroke code was activated for 356 patients who had suspected acute stroke. This code is activated by any health care provider in the emergency department or inpatient areas to rule out an acute stroke when a patient has acute neurologic symptoms [20–22]. Activation of the code prompts an immediate evaluation by the stroke rapid response team and is followed by neuroimaging consisting of head CT and vascular imaging, such as CT angiography of the head and neck. Brain MRI is often performed on the same or the next day.
View larger version (126K)
Enrollment of patients was continuous throughout the study period. Patients were excluded if they died before neuroimaging (n = 7) or if they had no clinical data in the electronic medical record (EMR) (n = 20). This yielded a final sample of 329 patients. The distribution of patients across the three New York City boroughs was 266 patients from Manhattan, 41 patients from Queens, and 22 patients from Brooklyn.
Coronavirus disease (COVID-19) is associated with large vessel occlusion stroke but not small vessel occlusion stroke. After stratification for race and ethnicity, the risk of large vessel stroke among patients with COVID-19 was 2.4 times as high as that among patients without COVID-19.
Physicians should lower their threshold of suspicion for large vessel stroke in patients with COVID-19 who present with acute neurologic symptoms.
Major vascular risk factors related to stroke, such as hypertension, coronary artery disease, diabetes mellitus type 2, atrial fibrillation, congestive heart failure, dyslipidemia, smoking status, National Institutes of Health (NIH) Stroke Scale score, laboratory data, and body mass index, were collected from the EMR. Demographic information (age, sex, and race or ethnicity), COVID-19 status, and time to presentation were also collected from the EMR. COVID-19 was confirmed by in-house in vitro reverse transcription–polymerase chain reaction (RT-PCR) testing of nasopharyngeal swabs (cobas 6800/8800 systems, Roche) under emergency authorization from the U.S. Food and Drug Administration. Patients were considered to not have COVID-19 if the RT-PCR result was negative. However, because of the suboptimal sensitivity of the RT-PCR [23–25], if patients continued to experience typical pulmonary symptoms of COVID-19, such as cough and shortness of breath, and if chest CT or CT angiography showed typical features of COVID-19 (n = 6), these patients were considered to have COVID-19.
All neuroimaging of each patient was retrospectively reviewed for presence of stroke, stroke subtype (LVO or SVO), and location. Assessed imaging included head CT (n = 329), CT angiography (n = 327), MRI (n = 250), MR angiography (n = 78), and digital subtraction angiography (n = 59). This retrospective review was performed by a fellowship-trained neuroradiologist (J. S., with 1 year of experience) blinded to radiology reports and clinical history, including COVID-19 status. A second reader (S. K., third-year radiology resident) blinded to other clinical history independently recorded this information on the basis of a retrospective review of the prospective clinical radiology reports (which had been interpreted by five neuroradiologists) for each patient’s imaging. This second reader did not review the images when assessing the radiology reports. For patients with incongruency between the retrospective imaging review and the assessment of the prospective reports, the imaging was retrospectively reviewed by a third reader (P. B., fellowship-trained neuroradiologist with 4 years of experience) blinded to imaging reports and clinical data, whose assessment served as the final diagnosis.
Second-order vessel occlusions were counted as LVOs, as described by Rennert et al. . Thus, LVOs included the internal carotid artery, middle cerebral artery segments M1 and M2, anterior cerebral artery segments A1 and A2, posterior cerebral artery segments P1 and P2, basilar artery, anterior inferior cerebellar artery, vertebral artery, and posterior inferior cerebellar artery. SVO was defined as perforating vessel occlusion not appreciable to the reader on vascular imaging that resulted in acute lacunar infarction on MRI and CT combined with CT angiography, MR angiography, and digital subtraction angiography findings.
A univariate analysis was performed to evaluate for associations between stroke-related risk factors and COVID-19 status. The Pearson chi-square test was performed for categoric variables, and a t test was performed for continuous variables. To reduce confounding effects, stroke-related risk factors associated with COVID-19 status identified at univariate analysis were applied to a multivariate logistic regression analysis assessing the effect of COVID-19 status on stroke subtype outcome. Patients without COVID-19 formed a negative control group in these analyses. Analyses were conducted with the SPSS statistical package for Microsoft Windows (version 25, IBM).
Clinical Characteristics of Study Sample
The study sample consisted of 329 patients for whom a code for stroke was activated (175 [53.2%] men, 154 [46.8%] women; mean age, 66.9 ± 14.9 [SD] years). The mean time from symptom onset to hospital presentation was 11.2 ± 7.6 hours. The mean NIH Stroke Scale score was 9.5 ± 8.5. COVID-19 was present in 38.3% (126/329) of all of the patients for whom a code for stroke was activated. The other 61.7% (203/329) of patients formed the negative control group.
The first two readers provided congruent imaging assessments in 97.6% (321/329) of cases. The third reader interpreted the imaging in the other 2.4% (8/329) to reach a final assessment. Acute ischemic stroke was found on imaging in 35.3% (116/329) of the sample. LVOs were found in 21.6% (71/329) and SVOs in 14.6% (48/329) of the patients (Table 1). The most commonly observed LVO [62.0% (44/71)] was middle cerebral artery segment M1-M2 occlusion. The second most frequent occlusion in patients with LVO stroke [21.1% (15/71)] was in the internal carotid artery. Multifocal LVOs were observed in only 9.9% (7/71) of patients with LVO stroke.
Univariate Associations of Coronavirus Disease With Clinical Characteristics and Imaging Findings
At univariate analysis, the only individual stroke-related risk factor significantly associated with COVID-19 was race or ethnicity (Table 2). Hispanic patients accounted for 38.1% of those with compared with 20.7% of patients without COVID-19 (p = 0.001). LVO was present in 31.7% of patients with COVID-19 compared with 15.3% of patients without COVID-19 (p = 0.001). SVO was present in 15.9% of patients with COVID-19 and 13.8% of patients without COVID-19 (p = 0.632).
Independent Association of Coronavirus Disease and Large Vessel Occlusion Stroke in Multivariate Regression
After stratification for ethnicity and race in multivariate analysis, the presence of COVID-19 exhibited a significant independent association (p = 0.011) with LVO stroke (odds ratio, 2.4) compared with the absence of COVID-19 (Table 3). No other stroke-related risk factor was associated with LVO in the multivariate analysis.
To our knowledge, this is the first study to show an association between COVID-19 and large vessel strokes. In our study 62% of the large vessel strokes involved occlusion of the M1-M2 segments of the middle cerebral artery. Results of prior studies suggest that the pathologic mechanism involves direct binding to ACE-2 receptors, which are expressed throughout the endothelial cells of the vasculature, which leads to vascular thrombosis within the vessels through a cytokine storm [6–8, 27]. This process likely accounts for the increased risk of thrombosis—including deep vein thrombosis [11, 12], pulmonary embolisms [15–17], and strokes [18, 28]—in patients with COVID-19. Besides being cryptogenic and in situ occlusion from plaque rupture, LVOs can be cardioembolic and in rare cases can embolize from the venous system by bypassing through a right-to-left shunt, such as a patent foramen ovale. Although cardioembolic causes can be considered, multifocal occlusion was observed in only 9.9% of patients with LVOs in our study.
We conducted this retrospective study because of our anecdotal observation of an increased frequency of acute strokes in patients presenting with acute neurologic symptoms during the COVID-19 pandemic. We included stroke code patients presenting during the COVID-19 pandemic who underwent emergency stroke workups. Race or ethnicity was the only stroke-related risk factor associated with COVID-19 status; specifically, a significantly greater proportion of Hispanic patients had COVID-19. A prior report  also showed a disproportionate impact of COVID-19 on the Hispanic community in New York City.
Many of the common vascular comorbidities were not associated with LVOs in our study, likely owing to the small sample size. Erythrocyte sedimentation rate and C-reactive protein level are commonly elevated in a cytokine storm and are associated with severe COVID-19 . These inflammatory markers were elevated in patients with COVID-19 compared with negative control subjects, although the difference was not significantly different. Obesity has also been associated with severe forms of COVID-19, as found in a study showing that higher body mass index was associated with a significant increase in morbidity . However, we did not observe a significant association between body mass index and COVID-19 status.
The association between LVO strokes and COVID-19 has important clinical implications. Although the incidence of SARS-CoV-2 infection in New York City is declining, a large population of patients continue to present with COVID-19. Patients with active SARSCoV-2 infection should be monitored closely, especially if they present with acute neurologic symptoms, with a lower threshold of suspicion of LVOs. Health care providers in the emergency department and inpatient areas should be cognizant of this association and not delay activating a stroke code. This association may aid neurointerventionalists assessing the presence and location of an LVO if they are aware of this elevated risk in the COVID-19 population. Additionally, this association may imply that patients with LVO during the COVID-19 pandemic who have not undergone testing for SARS-CoV-2 infection or are waiting for results warrant higher suspicion and appropriate precautions.
There were several limitations to our study. First, we used the reference standard of RT-PCR from a nasal swab, the sensitivity and specificity of which for diagnosis of COVID-19 have not been definitively established. Although the test is being routinely used to assess for COVI-19 status, it was approved under emergency use authorization and has recently been found to have a sensitivity as low as 70% [23–25]. To address this problem, we assessed clinical manifestations and chest CT and CT angiography findings to identify likely false-negative diagnoses based on RT-PCR results, although only six patients with negative RTPCR results were classified as having COVID-19 on this basis. Second, because this was a retrospective study, we could not control for possible unknown confounding variables. We attempted to overcome confounding effects of common stroke-related risk factors by assessing the association between these factors and COVID-19 status and incorporating such risk factors into a multivariate analysis. Finally, we acknowledge that this study may show an association between COVID-19 and LVO strokes, but we did not evaluate causality.
To our knowledge, this is the first study to describe an association between COVID-19 and large vessel strokes. Patients with COVID-19 presenting with acute neurologic symptoms warrant a lower threshold of suspicion of large vessel stroke, and prompt workup for large vessel stroke is recommended. Future investigation may focus on the exact pathophysiologic mechanism of large vessel strokes in patients with COVID-19 and validate contributing risk factors through a larger study.
We thank Zahi A. Fayad and Valentin Faveau of the Biomedical Engineering and Imaging Institute, Icahn School of Medicine at Mount Sinai, for contributions in revision of this article. We also thank the Mount Sinai COVID Informatics Center.