Research International Journal of Stroke 2021, Vol. 16(4) 437–447 Cerebrovascular events and outcomes in ! 2020 World Stroke Organization hospitalized patients with COVID-19: Article reuse guidelines: The SVIN COVID-19 Multinational sagepub.com/journals-permissions DOI: 10.1177/1747493020959216 Registry journals.sagepub.com/home/wso James E Siegler1,2 , Pere Cardona3, Juan F Arenillas4,5 , Blanca Talavera3, Ana N Guillen3, Alba Chavarrı́a-Miranda3 , Mercedes de Lera3, Priyank Khandelwal5, Ivo Bach6, Pratit Patel5, Amit Singla7, Manuel Requena8,9 , Marc Ribo8,9, Dinesh V Jillella10, Srikant Rangaraju10, Raul G Nogueira10,11, Diogo C Haussen10,11, Alejandro R Vazquez12, Xabier Urra12,13, Ángel Chamorro12,13 , Luis S Román14, Jesse M Thon1,2, Ryna Then1,2, Emma Sanborn1,2, Natalia P de la Ossa15 , Mònica Millàn15, Isaac N Ruiz15, Ossama Y Mansour16 , Mohammed Megahed17, Cristina Tiu18,19, Elena O Terecoasa18,19, Răzvan A Radu18, Thanh N Nguyen20,21,22, Gioacchino Curiale20, Artem Kaliaev21, Alexandra L Czap23, Jacob Sebaugh23, Alicia M Zha23, David S Liebeskind24, Santiago Ortega-Gutierrez25, Mudassir Farooqui25 , Ameer E Hassan26,27, Laurie Preston26, Mary S Patterson28, Saif Bushnaq28, Osama Zaidat28 and Tudor G Jovin1,2; on behalf of the SVIN COVID-19 Multinational Registry & Task Force 17 Department of Neurology, Stroke and Neurointervention division, 1 Cooper Neurologic Institute, Cooper University Hospital, Camden, NJ, Alexandria University, Alexandria, Egypt 18 USA Department of Critical Care Medicine, Alexandria University, 2 Cooper Medical School of Rowan University, Camden NJ, USA Alexandria, Egypt 19 3 Department of Neurology, Hospital Universitari, Bellvitge, Barcelona, Department of Neurology, University Emergency Hospital Bucharest, Spain Bucharest, Romania 20 4 Department of Neurology, Hospital Clı́nico Universitario, Valladolid, ‘‘Carol Davila’’ University of Medicine and Pharmacy, Bucharest, Spain Romania 21 5 Neurovascular Research Laboratory, Instituto de Biologı́a y Genética Department of Neurology, Boston Medical Center, Boston University Molecular, Universidad de Valladolid, Consejo Superior de School of Medicine, MA, USA 22 Investigaciones Cientı́ficas, Madrid, Spain Department of Radiology, Boston Medical Center, Boston University 6 Department of Neurology, Robert Wood Johnson University Hospital, School of Medicine, MA, USA 23 New Brunswick, NJ, USA Department of Neurosurgery, Boston Medical Center, Boston 7 New Jersey Medical School, Newark, NJ, USA University School of Medicine, MA, USA 24 8 Department of Neurosurgery, Robert Wood Johnson University Department of Neurology, McGovern Medical School, University of Hospital, New Brunswick, NJ, USA Texas Health Science Center, Houston, TX, USA 25 9 Stroke Unit, Department of Neurology, Vall d’Hebron Research Department of Neurology, Ronald Reagan UCLA Medical Center, Los Institute, Barcelona, Spain Angeles, CA, USA 26 10 Departament de Medicina, Universitat Autónoma de Barcelona, Department of Neurology, Neurosurgery and Radiology, University of Barcelona, Spain Iowa Hospitals and Clinics, Iowa City, IA, USA 27 11 Department of Neurology, Emory University School of Medicine, Department of Clinical Neuroscience Research, Valley Baptist Medical Atlanta, GA, USA Center, Harlingen, TX, USA 28 12 Department of Neurology, Grady Memorial Hospital, Atlanta, GA, USA Department of Neurology, University of Texas Rio Grande Valley, 13 Department of Neurology, Hospital Clı́nic, Barcelona, Spain Harlingen, TX, USA 29 14 Area of Neuroscience, Institut d’Investigacions Biomèdiques August Pi I Neuroscience Institute, Bon Secours Mercy Health St. Vincent Hospital, Sunyer (IDIBAPS), Barcelona, Spain Toledo, OH, USA 15 Department of Radiology, Hospital Clı́nic, Barcelona, Spain Corresponding author: 16 Stroke Unit, Neuroscience Department, Hospital Universitari Germans Tudor G Jovin, Cooper University Health Care Camden, NJ 08901, USA. Trias i Pujol, Carretera Canyet s/n, Badalona, Barcelona, Spain Email: jovin-tudor@cooperhealth.edu International Journal of Stroke, 16(4) 438 International Journal of Stroke 16(4) Abstract Background: Severe acute respiratory syndrome-coronavirus-2 (SARS-CoV-2) has been associated with a significant risk of thrombotic events in critically ill patients. Aim: To summarize the findings of a multinational observational cohort of patients with SARS-CoV-2 and cerebrovas- cular disease. Methods: Retrospective observational cohort of consecutive adults evaluated in the emergency department and/or admitted with coronavirus disease 2019 (COVID-19) across 31 hospitals in four countries (1 February 2020–16 June 2020). The primary outcome was the incidence rate of cerebrovascular events, inclusive of acute ischemic stroke, intracranial hemorrhages (ICH), and cortical vein and/or sinus thrombosis (CVST). Results: Of the 14,483 patients with laboratory-confirmed SARS-CoV-2, 172 were diagnosed with an acute cerebro- vascular event (1.13% of cohort; 1130/100,000 patients, 95%CI 970–1320/100,000), 68/171 (40.5%) were female and 96/172 (55.8%) were between the ages 60 and 79 years. Of these, 156 had acute ischemic stroke (1.08%; 1080/100,000 95%CI 920–1260/100,000), 28 ICH (0.19%; 190/100,000 95%CI 130–280/100,000), and 3 with CVST (0.02%; 20/100,000, 95%CI 4–60/100,000). The in-hospital mortality rate for SARS-CoV-2-associated stroke was 38.1% and for ICH 58.3%. After adjusting for clustering by site and age, baseline stroke severity, and all predictors of in-hospital mortality found in univariate regression (p < 0.1: male sex, tobacco use, arrival by emergency medical services, lower platelet and lympho- cyte counts, and intracranial occlusion), cryptogenic stroke mechanism (aOR 5.01, 95%CI 1.63–15.44, p < 0.01), older age (aOR 1.78, 95%CI 1.07–2.94, p ¼ 0.03), and lower lymphocyte count on admission (aOR 0.58, 95%CI 0.34–0.98, p ¼ 0.04) were the only independent predictors of mortality among patients with stroke and COVID-19. Conclusions: COVID-19 is associated with a small but significant risk of clinically relevant cerebrovascular events, particularly ischemic stroke. The mortality rate is high for COVID-19-associated cerebrovascular complications; there- fore, aggressive monitoring and early intervention should be pursued to mitigate poor outcomes. Keywords All cerebrovascular diseases/stroke, intracranial hemorrhage, cerebral venous thrombosis, COVID-19 Received: 22 July 2020; accepted: 20 August 2020 among consecutively evaluated patients with COVID- Introduction 19. Furthermore, given the high short-term mortality of Thrombotic complications of the novel human corona- hospitalized patients with COVID-19 (5–20%),8,9 we virus responsible for the systemic acute respiratory syn- sought to evaluate the short-term functional outcomes drome, SARS-CoV-2, have been well described.1,2 and survival of patients with COVID-19 and associated Early reports from Wuhan, China, indicated a 2–5% cerebrovascular disease. risk of acute ischemic stroke among hospitalized patients with the coronavirus disease 2019 (COVID- 19).3 Other series have reported that patients with Methods SARS-CoV-2 are at risk of embolic strokes with severe deficits and a high in-hospital mortality rate.4–6 Study design and participants This risk may be more than seven times greater for A retrospective observational registry involving pro- patients with SARS-CoV-2 than patients with influenza spectively gathered data from 17 healthcare networks according to one comparative analysis.7 Although ini- (31 unique hospitals) across four countries (USA, tial estimates indicated a 1 in 20 risk of stroke in Spain, Egypt, and Romania) was queried. patients with COVID-19, more recent data report a Participating centers were recruited based on affiliation rate of 1–2% among hospitalized patients.7 with the Society of Vascular and Interventional Neurology (SVIN) and contacts among site investiga- tors. All consecutive patients 18 years of age who Aims were evaluated in each center’s emergency department To better illustrate the relationship between SARS- (ED) and/or admitted between 1 February 2020 and 16 CoV-2 and associated cerebrovascular events (CVE), June 2020 were eligible for inclusion. To be included, we established a multinational registry to estimate the patients must have been diagnosed with SARS-CoV-2 global incidence rate of these neurologic complications using oropharyngeal polymerase chain reaction (PCR) International Journal of Stroke, 16(4) Siegler et al. 439 or IgG and/or IgM antibody presence in the sera using Coull method.14 After event rates were found to vary commercial assays. All imaging to confirm the diagno- according to site COVID-19 volume, Pearson’s correl- sis of acute ischemic stroke, intracranial hemorrhage, ation coefficient was calculated to estimate the associ- or cortical vein and/or dural sinus thrombosis (CVST) ation between the COVID-19 volume at individual sites was performed and interpreted at the discretion of the and (1) CVE incidence rate, (2) stroke incidence rate, treating physician. This study was approved under a and (3) in-hospital mortality. waiver of informed consent by the local institutional As part of the secondary objective, we evaluated the review board at each participating center, and it is clinical, laboratory, and radiographic factors that were reported in accordance with the STrengthening the associated with in-hospital mortality. In addition to Reporting of OBservational studies in Epidemiology using the aforementioned descriptive statistics, the statement. two-sample test of proportions was used to determine significant differences in-hospital mortality between each of the various stroke mechanisms. All relevant Data collection covariates were entered into univariate regression Patient demographic information, including age, sex, models estimating the odds of mortality among patients race, and ethnicity, as well as pertinent medical history, with ischemic stroke. Variables significant to p < 0.1 in National Institutes of Health Stroke Scale (NIHSS) these models were added to age and baseline NIHSS score, neuroimaging, treatment, laboratory testing (known significant predictors of mortality)15,16 in a (hematologic parameters, d-dimer), functional outcome multivariable logistic regression model, which was clus- at the time of discharge using the modified Rankin tered by site. A separate multivariable model was gen- Scale (mRS),10,11 and discharge disposition were cap- erated among patients with ischemic stroke due to an tured. De-identified data elements (including age, which intracranial occlusion estimating the odds of in-hospital was binned by decade for de-identification purposes) mortality, incorporating age, baseline NIHSS, and all were documented by local investigators on a HIPAA- variables significant to p < 0.1 in univariate regression. compliant, online platform, as previously described,12 All tests were performed at the two-sided level using and were made accessible to the data coordinating STATA 15.0 (College Station, TX), with p  0.05 con- center (Cooper University Hospital) for analysis. sidered statistically significant. As this was an explora- Intracranial occlusion was defined broadly as any intra- tory analysis, p-values are reported for convention and cranial arterial occlusion identified on vascular ima- should be interpreted with caution. No adjustments ging. Stroke etiology was classified according to the were made for multiple comparisons. Missing data Trial of Org 10172 in Acute Stroke Treatment as pre- were not imputed. Fully de-identified data will be viously described.13 Sites also reported their overall in- made available upon reasonable request of the corres- hospital mortality rate among all COVID-19 patients ponding author. throughout the study period. Data elements with a <50% completion rate (e.g. intubation and/or trache- Results ostomy, length of hospital stay) were not reported in order to limit the selection bias. Ninety-day outcomes Of the 14,483 patients with laboratory-confirmed (including mRS, living situation, and survival) were SARS-CoV-2, 172 were diagnosed with an acute CVE unavailable given the recency of the study period and (1.13% of COVID-19 cohort; 1130 per 100,000 the generally protracted hospitalizations for COVID-19 patients, 95%CI 970–1320 per 100,000). Among patients. patients with CVE, 68/171 (40.5%) were female, 94/ 138 (68.1%) with reported race were White, and 96/ 172 (55.8%) were between the ages 60 and 79 years. Statistical analysis Between sites, CVE rates ranged from 0.19 to 5.04% No sample-size calculations were performed prior to (Figure 1), with sites treating more COVID-19 patients data collection. Descriptive statistics were used to sum- reporting a lower rate of CVE (r ¼0.659, p ¼ 0.004) marize continuous and categorical variables. Normality and acute ischemic stroke (r ¼ 0.668, p ¼ 0.003). The of continuous data was assessed histographically and most common CVE was acute ischemic stroke, confirmed using the Shapiro–Wilk test. Continuous observed in 156 patients (1.08%; 1080 per 100,000, variables were reported as medians with interquartile 95%CI 920–1260 per 100,000), with 28 patients range. The primary outcome was the incidence rate of having radiographically confirmed primary ICH CVEs, which was summated by healthcare institutions (0.19%; 190 per 100,000, 95%CI 130–280 per 100,00) and across the entire cohort, with approximate bino- and 3 with CVST (0.02%; 20 per 100,000, 95%CI 4–60 mial 95% confidence intervals (95%CI) per 100,000 per 100,000). Fourteen patients had acute stroke and hospitalized COVID-19 patients using the Agresti– intracerebral hemorrhage (8.1% of cohort), with a International Journal of Stroke, 16(4) 440 International Journal of Stroke 16(4) similar distribution of age and comorbidities as the patients with stroke (median NIHSS 13 (IQR 5–21)) overall stroke population in this cohort (Table 1). or ICH (median NIHSS 22 (IQR 7–25)). The most Demographic data are summarized in Table 1. common mechanism of acute stroke among COVID- Deficits were generally moderate or severe among 19 patients was cryptogenic (42.6%) followed by cardi- oembolism (27.1%). An intracranial occlusion was identified in 53/107 patients with intracranial arterial imaging (49.5%; Table 2). Figure 1. Prevalence of CVEs by site. The median in-hospital COVID-19 mortality rate was 11.7% (IQR 8.6–18.4%), with no correlation between COVID-19 volume and in-hospital mortality among COVID-19 patients (r ¼ 0.002), or among patients with COVID-19 and CVE (r ¼ 0.01; Supplemental Figure 1). There was a non-significant variation in mortality across sites (p ¼ 0.09), but this did not vary significantly by site COVID-19 volume (Supplemental Figures 1 and 2). More than one-third of patients with acute stroke and COVID-19 died during hospitalization, while more than half of patients with ICH and COVID-19 died (Supplemental Table 1). The median time from stroke onset to death was four days (IQR 1–10) among patients with stroke, 33 of Incidence rate of any CVEs and stroke are plotted using vertical bars whom died within seven days (21.2% of all stroke (primary y-axis) while the total COVID-19 ED and/or hospitalization rate are patients). The median time from ICH onset to death plotted as dots (secondary y-axis). was 3.5 days (IQR 1–7), 11 of whom died within seven CVE: cerebrovascular event; COVID-19: coronavirus disease 2019. days (39.3% of all ICH patients). Although there was Figure 2. In-hospital mortality rate according to stroke etiology. The distribution of stroke etiologies is plotted using vertical bars (primary y-axis), and mortality rate according to stroke mechanism is plotted as squares with standard error bars (secondary y-axis). International Journal of Stroke, 16(4) Siegler et al. 441 Table 1. Demographic data All patients with suspected or Stroke þ radiographically Intracranial intracranial confirmed Acute ischemic hemorrhage hemorrhage CVEa (n ¼ 172) stroke (n ¼ 156) (n ¼ 28) (n ¼ 14) Age, no. (%) <30 0/172 (0%) 0/156 (0%) 0/28 (0%) 0/14 (0%) 30–39 5/172 (2.9%) 4/156 (2.6%) 1/28 (3.6%) 0/14 (0%) 40–49 19/172 (11.1%) 15/156 (9.6%) 6/28 (21.4%) 3/14 (21.4%) 50–59 24/172 (14.0%) 20/156 (12.8%) 4/28 (14.3%) 0/14 (0%) 60–69 50/172 (29.1%) 49/156 (31.4%) 4/28 (14.3%) 4/14 (28.6%) 70–79 46/172 (26.7%) 45/156 (28.9%) 5/28 (17.8%) 4/14 (28.6%) 80–89 23/172 (13.4%) 18/156 (11.5%) 8/28 (28.9%) 3/14 (21.4%) >89 5/172 (2.9%) 5/156 (3.2%) 0/28 (0%) 0/14 (0%) Sex, no. female (%) 68/171 (40.5%) 61/155 (39.4%) 10/28 (35.7%) 4/14 (28.6%) Race, no. (%) White 94/138 (68.1%) 84/125 (67.2%) 20/24 (83.3%) 10/12 (83.3%) Black 37/138 (26.8%) 30/125 (28.0%) 3/24 (12.5%) 2/12 (16.7%) Asian 3/138 (2.2%) 2/125 (1.6%) 1/24 (4.2%) 0/12 (0%) More than one race 2/138 (1.5%) 2/125 (1.6%) 0/24 (0%) 0/12 (0%) Other 2/138 (1.5%) 2/125 (1.6%) 0/24 (0%) 0/12 (0%) Hispanic ethnicity, no. (%) 62/154 (40.3%) 58/140 (41.4%) 10/27 (37.0%) 6/14 (42.9%) Method of arrival, no. (%) Emergency medical services 67/115 (58.3%) 59/105 (56.2%) 12/19 (63.2%) 5/10 (50.0%) Transfer from outside hospital 30/115 (26.1%) 28/105 (26.7%) 6/19 (31.6%) 4/10 (40.0%) Walk-in/private vehicle 13/115 (11.3%) 13/105 (12.4%) 1/19 (5.3%) 1/10 (10.0%) Other 5/113 (4.4%) 5/105 (4.8%) 0/19 (0%) 0/10 (0%) Diagnosis of COVID-19,b no. (%) Oropharyngeal PCR 169/172 (98.3%) 153/156 (98.1%) 26/28 (92.9%) 12/14 (85.7%) Serum IgM and/or IgG 12/172 (1.7%) 11/156 (7.1%) 2/28 (7.1%) 2/14 (14.3%) Known COVID-19 22/172 (12.8%) 19/156 (12.2%) 3/28 (10.7%) 0/14 (0%) exposure, no. (%) Medical history, no. (%) Hypertension 121/170 (71.2%) 111/154 (72.1%) 18/28 (64.3%) 9/14 (64.3%) (continued) International Journal of Stroke, 16(4) 442 International Journal of Stroke 16(4) Table 1. Continued All patients with suspected or Stroke þ radiographically Intracranial intracranial confirmed Acute ischemic hemorrhage hemorrhage CVEa (n ¼ 172) stroke (n ¼ 156) (n ¼ 28) (n ¼ 14) Diabetes mellitus 68/168 (40.5%) 65/152 (42.8%) 7/28 (25.0%) 5/14 (35.7%) Dyslipidemia 61/155 (39.4%) 58/142 (40.1%) 9/25 (36.0%) 6/13 (46.2%) Atrial fibrillation 30/163 (18.4%) 27/148 (18.2%) 5/22 (18.5%) 2/14 (14.3%) Congestive heart failure 30/170 (17.7%) 27/154 (17.5%) 6/22 (21.4%) 4/10 (40.0%) Active tobacco use 17/150 (10.8%) 15/145 (10.3%) 3/26 (11.5%) 1/14 (7.1%) Prior stroke 19/154 (12.3%) 15/140 (10.7%) 6/27 (22.2%) 2/14 (14.3%) Chronic renal insufficiency 18/160 (11.3%) 16/144 (11.1%) 5/28 (17.9%) 3/14 (21.4%) (stage III/IV or dialysis-dependent) Chronic obstructive pulmonary 12/148 (8.1%) 11/135 (8.2%) 2/26 (7.7%) 1/14 (7.1%) disease and/or asthma Cancer 9/147 (6.1%) 8/134 (6.0%) 4/26 (15.4%) 3/14 (21.4%) a CVE includes acute ischemic stroke, intracranial hemorrhage, or cortical vein and/or dural sinus thrombosis. b 9 patients with a CVE, including 8 patients with an acute stroke were diagnosed with COVID-19 using both serum antibodies and oropharyngeal PCR. CVE: cerebrovascular event; COVID-19: coronavirus disease 2019; PCR: polymerase chain reaction. no significant difference in mortality according to remained the only independent predictor of in-hospital stroke mechanism (p ¼ 0.22), there was a significantly mortality among patients with an intracranial occlusion higher mortality rate among patients with stroke plus (aOR 5.03, 95%CI 1.99–12.74, p < 0.01; Supplemental intracranial hemorrhage (OR 4.07, 95%CI 1.19–13.85, Table 3). p ¼ 0.03) and a non-significantly higher mortality for stroke of cryptogenic mechanism versus stroke due to Discussion ‘‘other’’ mechanisms (p ¼ 0.07), and cryptogenic strokes versus strokes due to cardioembolism In this observational cohort of more than 14,000 (p ¼ 0.098; Figure 2). Among all patients with a COVID-19 patients treated at 31 hospitals in four stroke, those more likely to die were male (OR 2.16, countries, approximately 1 in 100 patients were diag- 95%CI 1.04–4.48, p ¼ 0.04), had more severe deficits nosed with a clinically relevant or radiographically con- (b ¼ 0.017, 95%CI 0.009–0.026, p < 0.001), and a firmed cerebral infarction. The reported incidence rate cryptogenic stroke mechanism (OR 2.08, 95%CI of any CVE varied between 0.19 and over 5% in this 0.94–4.63, p ¼ 0.07). When added to age (p ¼ 0.03 in cohort, with a more than twofold higher rate of crypto- unadjusted regression) in the final multivariable genic ischemic stroke (42%) when compared to historic model after adjustment for all variables significant to cohorts (17%)17 and half of patients having an p < 0.1 in univariate regression (male sex, tobacco use, embolic occlusion. The high prevalence of cryptogenic arrival by emergency medical services, lower admission infarctions with an embolic radiographic appearance platelet and lymphocyte counts, and intracranial occlu- validates prior reports suggesting a prothrombotic sion)—and after clustering by site—cryptogenic stroke mechanism of stroke among patients with COVID- mechanism (aOR 5.01, 95%CI 1.63–15.44, p < 0.01), 19.7 Furthermore, the severity of disease in patients older age (aOR 1.78, 95%CI 1.07–2.94, p ¼ 0.03), and with a cryptogenic (SARS-CoV-2-associated) stroke is lower lymphocyte count on admission (aOR 0.58, reflected by the high mortality rate among cryptogenic 95%CI 0.34–0.98 p ¼ 0.04) remained the only inde- stroke patients with COVID-19 (39%). Importantly, pendent predictors of in-hospital mortality the early mortality risk resonates with other reports (Supplemental Table 2). Cryptogenic stroke also of increased mortality with SARS-CoV-2 among International Journal of Stroke, 16(4) Siegler et al. 443 Table 2. Clinical and radiographic features Acute ischemic stroke (n ¼ 156) Intracranial hemorrhage (n ¼ 28) NIHSS at baseline, median (IQR) 13 (5–21) (n ¼ 130) 22 (7–25) (n ¼ 16) Glasgow Coma Scale at baseline, median (IQR) n/a 6 (4–15) (n ¼ 15) Head CT performed, no. (%) 135/151 (89.4%) 28 (100%) First CT indicating acute stroke, no. (%) 76/132 (57.6%) n/a Intracranial hemorrhage on CT, no. (%) 14/132 (10.6%) 28 (100%) Any vascular imaging performed, no. (%) 105/156 (67.3%) 14/28 (50.0%) Intracranial occlusion identified, no. (%) 53/107 (49.5%) n/a MRI brain performed, no. (%) 55/151 (36.4%) 8/27 (29.6%) MRI evidence of acute stroke, no. (%) 45/55 (81.8%) 5/8 (62.5%) CT or MRI location of infarction, no. (%) n/a Cortical 81/101 (80.2%) n/a Subcortical supratentorial 54/101 (53.5%) n/a Infratentorial 12/101 (11.9%) n/a Stroke etiology, no. (%) Cardioembolism 35/129 (27.1%) n/a Large vessel atherosclerotic disease 15/129 (11.6%) n/a Small vessel disease 4/129 (3.1%) n/a Cryptogenic 55/129 (42.6%) n/a Other 11/129 (8.5%) n/a ICH compartment,a no. (%) Intraparenchymal n/a 20 (71.4%) Subarachnoid n/a 7 (25.0%) Subdural n/a 2 (7.1%) Intraventricular n/a 8 (28.6%) Epidural n/a 0 (0%) IPH location*, no. (%) Supratentorial n/a 13/18 (72.2%) Infratentorial n/a 7/18 (38.9%) IPH volume, median cc (IQR) n/a 20 (10–66.2) (n ¼ 11) ICH score, median (IQR) n/a 3 (1–3) (n ¼ 10) (continued) International Journal of Stroke, 16(4) 444 International Journal of Stroke 16(4) Table 2. Continued Acute ischemic stroke (n ¼ 156) Intracranial hemorrhage (n ¼ 28) IPH etiology, no. (%) Coagulopathy n/a 8/16 (50.0%) Hypertension n/a 5/16 (31.3%) Other n/a 3/16 (18.8%) a Some patients experienced intracranial hemorrhage in multiple compartments with intraparenchymal involvement of supra- and infratentorial regions. NIHSS: National Institutes of Health Stroke Scale; IQR: interquartile range; CT: computed tomography; MRI: magnetic resonance imaging; ICH: intracranial hemorrhage; IPH: intraparenchymal hemorrhage. patients with comorbid cerebrovascular disease,18 and 50% rate of intracranial occlusion—consistent although it vastly exceeds what has been historically with an embolic mechanism. It is worth noting that described in non-SARS-CoV2 patients (6% at one while strokes in this cohort were more severe than in year for patients with cryptogenic stroke vs. 21% for historic cohorts, we did not account for critical illness cardioembolism and large vessel disease).19 This sug- and/or sedation, which could have artificially aug- gests that COVID-19-associated cryptogenic stroke is mented the NIHSS in these patients. Regardless of con- a truly unique stroke mechanism with a high probabil- founding by sedation or critical illness, the observation ity of early mortality. Whether stroke is a driving factor that COVID-19 patients with cryptogenic strokes (pre- for early mortality in COVID-19, or a clinical bio- sumably related to the prothrombotic milieu triggered marker of disease severity, should be explored in by SARS-CoV-2) were significantly more likely to die future studies. during their hospitalization speaks to the gravity of An unrelated observation was that the hospital cerebrovascular complications. More importantly, it COVID-19 volume was strongly and inversely corre- should galvanize our efforts to closely monitor for lated with the reported CVE rates. Several explanations such complications and manage aggressively when an may account for this. First, sites with higher volumes of event is suspected. COVID-19 patients may have broader testing or admis- sion criteria for asymptomatic or mild patients, among whom a cerebrovascular complication may be less Limitations common. Second, sites with higher volumes of While this study is the largest observational cohort of COVID-19 patients may have experienced significant consecutive COVID-19 patients with cerebrovascular resource limitations and as such, may have missed complications to date, it is limited by its retrospective stroke diagnoses among critically ill hospitalized nature, inclusion of patients largely representing the patients with COVID-19. Third, sites responsible for United States and Spain, and unavailability of some rapid triage and management of a large number of outcome data (e.g. length of stay, 90-day functional COVID-19 patients may have sought to limit health- status). Because we prioritized data collection for hos- care provider exposure to patients or restrict certain pitalized patients with COVID-19 and CVE, we are diagnostic testing (e.g. neuroimaging with CT and/or unable to report any comparisons between patients MRI) to prevent delays in the care of other patients with CVE and those without CVE. The incidence of while these resources could be decontaminated. And CVE observed in this cohort is not reflective of the finally, sites with higher COVID-19 volumes may overall incidence of CVE among COVID-19 patients have implemented more aggressive antithrombotic because these data were limited to sick patients who treatment earlier in the course of hospitalization for presented to the ED and/or were hospitalized. In add- COVID-19. However, data regarding prophylactic ition, sites participating in this study were affiliated with treatment was not captured in this investigation. the SVIN, with large referral networks, which could Our study also validates recent data from a New have confounded the severity of cases and prevalence York health system that reported COVID-19 patients of intracranial occlusion. However, among patients with acute stroke had a higher probability of crypto- who arrived via transfer from outside institutions, genic embolic stroke mechanism with more severe def- there was no higher prevalence of an occlusion (6/36 icits.8 Most patients in our cohort had moderate or vs. 10/35, p ¼ 0.23) or greater severity of deficits severe deficits according to the NIHSS, with an 80% (median NIHSS 17 vs. 11, p ¼ 0.47) when compared prevalence of cortical involvement on neuroimaging to patients who arrived by other means (data not International Journal of Stroke, 16(4) Siegler et al. 445 otherwise shown). In a separately reported analysis high as 5%), and the risk of poor outcomes was signifi- involving 11 of the sites included here, the rate of intra- cant. This builds on existing literature suggesting the cranial occlusion among all-comers with stroke stroke risk to be 1%;7,8 however, the risk may be admitted during a similar study period was 19.6% higher in some populations or those with more frequent (183/933 patients, unpublished data), which is slightly neuroimaging and/or aggressive neurologic monitoring. lower than what has been reported in observational It is likely that this approximately 1% risk of clinically studies (30%).20 Therefore, we do not believe that significant stroke reflects a floor rate, and the true rate this COVID-19 population reflects institutional biases could be even higher. In our cohort of geographically which could lead to a higher rate of LVO patients, irre- distinct and medically unique patients, the observed spective of COVID-19 status. While we observed a high high mortality rate validates one recent meta-analysis rate of cryptogenic stroke in this COVID-19 popula- of 4448 patients with COVID-1921 using a larger multi- tion, the diagnosis of a cryptogenic mechanism (and center data set. Patients with stroke may have suffered minimum diagnostic testing for stroke evalu- from a more virulent strain of SARS-CoV-222 and/or ation—including long-term cardiac event monitoring, greater complications of COVID-19, which puts them intracranial vascular imaging, etc.) was left to the dis- at a higher risk of long-term disability and in-hospital cretion of the treating physician. It is possible that mortality. The severity of cerebrovascular disease patients with cryptogenic stroke had less comprehen- during the COVID-19 pandemic may be compounded sive diagnostic evaluations due to severity of their ill- by the lack of in-hospital family support with visitor ness and low probability of survival, which could have restrictions, patients being managed in a COVID-19 confounded these results. However, the time from unit rather than a dedicated stroke unit, and possibly stroke onset to death was a median of four days, indi- less contact with physical therapy and occupational cating that a timely evaluation could have been pursued services until patients are liberated from COVID-19 to rule out traditional mechanisms of stroke. Stroke precautions. However, it may also be that the throm- mechanisms in this COVID-19 population are being botic complications of the virus could drive the reported separately. Furthermore, these data reflect adverse outcomes. Additional studies are called upon the experience of healthcare centers within and asso- to firmly establish COVID-19 as a unique mechanism ciated with the SVIN network, with 62% of patients of stroke and to better characterize the long-term com- being treated outside the United States. Importantly, plications of COVID-19-associated cerebrovascular the primary objective of this study was to estimate the disease. incidence rate of CVE and outcomes at healthcare insti- tutions that could monitor for neurologic complica- tions of COVID-19. More than 80% of patients with Acknowledgements acute ischemic stroke and 100% of patients with ICH This work would not have been possible without the incred- and CVST had radiographically confirmed events. It is ible assistance from the following collaborators of the SVIN possible, if not probable, that the lack of serial neuroi- COVID-19 Multinational Registry and Task Force: Patricia maging of critically ill patients with COVID-19 may Feineigle (CUH), Mohamad Abdalkader (BMC), Sergio Amaro (HCB), Hugo Aparicio (BMC), Juan F Arenillas underestimate the event rates reported in this multicen- (Valladolid), Ivo Bach (RWJMS), Jordi Blasco (HCB), Saif ter study. The question of how best to prevent throm- Bushnaq (Mercy St. Vincent), Alejandro Bustamante botic complications in patients with COVID-19 also (Germans Trias), Ángel Chamorro (HCB), Judith Clark remains unanswered. The non-randomized allocation (BMC), Alexandra Czap (UTH), Natalia Perez de la Ossa of treatment in the form of antiplatelet therapy (single (Germans Trias), Mercedes de Lera Alfonso (Valladolid), or dual), anticoagulation, or both, in conjunction with Shashvat Desai (UPMC), Laura Dorado (Germans Trias), the small number of total stroke patients, precludes a Luis R Esparragoza (Germans Trias), Denise Evans meaningful comparison of treatment effects. The best (VBMC), Mudassir Farooqui (UI), Meritxell Gomis course of treatment can only be clarified by randomized (Germans Trias), Ameer E Hassan (VBMC), Tamer Hassan clinical trials evaluating immunotherapies and antith- (Alexandria), Diogo Haussen (Emory Grady), Maria rombotics or by pooling of these data with other obser- Hernández-Pérez (Germans Trias), Mark Heslin (CUH), Chris Higham (CUH), Ashutosh P Jadhav (UPMC), Dinesh vational studies. V Jillella (Emory Univ.), Tudor G Jovin (CUH), Artem Kaliaev (BMC), Priyank Khandelwal (RWJMS), Rakeshh Conclusions Khatri (Lutheran Health), Amy Krueger (Mercy St. Vincent), Carlos Laredo (HCB), David S Liebeskind The arterial thrombotic complications of COVID-19 (UCLA), Italo Linfante (Baptist), Laura Llull (HCB), are not to be understated. While the risk of a clinically Antonio López (HCB), Ossama Y Mansour (Alexandria), relevant stroke was low in this multicenter study, there Racheal McCoy (CUH), Mohammed Megahed was considerable variability in the incidence rate (as (Alexandria), Mònica Millàn (Germans Trias), Mahmoud H International Journal of Stroke, 16(4) 446 International Journal of Stroke 16(4) Mohammaden (Emory Grady), Leigh Moore (Mercy St. thrombotic complications in critically ill ICU patients Vincent), Fadi Nahab (Emory Grady), Thanh N Nguyen with COVID-19: an updated analysis. Thromb Res (BMC), Raul G Nogueira (Emory Grady), Isaac N Ruiz 2020; 191: 148–150. (Germans Trias), Vı́ctor Obach (HCB), Darko Q Orozco 2. Klok FA, Kruip M, van der Meer NJM, et al. Incidence (UI), Santiago Ortega-Gutierrez (UI), Mohamed Osman of thrombotic complications in critically ill ICU patients (Mercy St. Vincent), Pratit Patel (RWJMS), Mary S with COVID-19. Thromb Res 2020; 191: 145–147. Patterson (Mercy St. Vincent), Gonzalo Valle Peñacoba 3. Mao L, Wang M, Chen S, et al. Neurological manifest- (Valladolid), Leonardo Pisani (Emory Grady), Pere C ations of hospitalized patients with COVID-19 in Wuhan, Portela (Bellvitge), Laurie Preston (VBMC), Razvan A Radu China: a retrospective case series study. JAMA Neurol (Bucharest), Vivek Rai (OhioHealth Riverside), Anna Ramos- 2020; 77: 683–690. Pachón (Germans Trias), Ankit Rana (CUH), Srikant 4. Escalard S, Maı̈er B, Redjem H, et al. Treatment of acute Rangaraju (Emory Grady), Arturo Renú (HCB), Manuel ischemic stroke due to large vessel occlusion with Requena (Valld’Hebron), Marc Ribo (Valld’Hebron), Jose R COVID-19: experience from Paris. Stroke 2020; 51: Romero (BMC), Salvatore Rudilosso (HCB), Emma Sanborn 2540–2543. (CUH), Sunil Sheth (UTH), Julie G Shulman (BMC), James E 5. Oxley TJ, Mocco J, Majidi S, et al. Large-vessel stroke as Siegler (CUH), Amit Singla (RWJMS), Ainsley Smith (CUH), a presenting feature of Covid-19 in the young. N Engl J Amy Starosciak (Baptist), Lauren Thau (CUH), Ephrem Med 2020; 382: e60. Teklemariam (CUH), Elena O Terecoasa (Bucharest), Ryna 6. Sweid A, Hammoud B, Bekelis K, et al. Cerebral ischemic Then (CUH), Jesse M Thon (CUH), Cristina Tiu and hemorrhagic complications of coronavirus disease (Bucharest), Vlad E Tiu (Bucharest), Israr Ul Haq (Mercy 2019. Int J Stroke. Epub ahead of print 26 June 2020. St. Vincent), Xabier Urra (HCB), Martha Vargas (HCB), DOI: 10.1177/1747493020937189. Alejandro R Vazquez (HCB), Vı́ctor Vera (HCB), Osama 7. Yaghi S, Ishida K, Torres J, et al. SARS-CoV-2 and Zaidat (Mercy St. Vincent), Cynthia Zevallos (UI), and stroke in a New York Healthcare System. Stroke 2020; Alicia M Zha (UTH). 51: 2002–2011. 8. Bonow RO, Fonarow GC, O’Gara PT and Yancy CW. Declaration of conflicting interests Association of Coronavirus Disease 2019 (COVID-19) The author(s) declared the following potential conflicts of inter- with myocardial injury and mortality. JAMA Cardiol est with respect to the research, authorship, and/or publication 2020; 5: 751–753. of this article: RGN reports consulting fees for advisory roles 9. Richardson S, Hirsch JS, Narasimhan M, et al. with Anaconda, Biogen, Cerenovus, Genentech, Imperative Presenting characteristics, comorbidities, and out- Care, Medtronic, Phenox, Prolong Pharmaceuticals, Stryker comes among 5700 patients hospitalized with COVID- Neurovascular and stock options for advisory roles with 19 in the New York City Area. JAMA 2020; 323: Astrocyte, Brainomix, Cerebrotech, Ceretrieve, Corindus 2052–2059. Vascular Robotics, Vesalio, Viz-AI, and Perfuze. No other 10. Rankin J. Cerebral vascular accidents in patients over the authors report any competing financial interests. age of 60. II. Prognosis. Scott Med J 1957; 2: 200–215. 11. van Swieten JC, Koudstaal PJ, Visser MC, Schouten HJ Funding and van Gijn J. Interobserver agreement for the assess- ment of handicap in stroke patients. Stroke 1988; 19: The author(s) received no financial support for the research, 604–607. authorship, and/or publication of this article. 12. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N and Conde JG. Research electronic data capture ORCID iDs (REDCap) – a metadata-driven methodology and work- James E Siegler https://orcid.org/0000-0003-0287-3967 flow process for providing translational research inform- Juan F Arenillas https://orcid.org/0000-0001-7464-6101 atics support. J Biomed Inform 2009; 42: 377–381. Alba ChavarrÚa-Miranda https://orcid.org/0000-0002- 13. Adams HP Jr, Bendixen BH, Kappelle LJ, et al. 7368-4939 Classification of subtype of acute ischemic stroke. Manuel Requena https://orcid.org/0000-0002-5671-6484 Definitions for use in a multicenter clinical trial. Ángel Chamorro https://orcid.org/0000-0003-0493-5340 TOAST. Trial of Org 10172 in Acute Stroke Treatment. Natalia P de la Ossa https://orcid.org/0000-0002-6063-211X Stroke 1993; 24: 35–41. Ossama Y Mansour https://orcid.org/0000-0001-8814-6431 14. Agresti A and Coull BA. Approximate is better than Mudassir Farooqui https://orcid.org/0000-0003-3697-5697 ‘‘exact’’ for interval estimation of binomial proportions. Am Stat 1998; 52: 119–126. Supplemental material 15. Saposnik G, Guzik AK, Reeves M, et al. Stroke prognos- tication using age and NIH Stroke Scale: SPAN-100. Supplemental material for this article is available online. Neurology 2013; 80: 21–28. 16. Koennecke HC, Belz W, Berfelde D, et al. Factors References influencing in-hospital mortality and morbidity in 1. Klok FA, Kruip M, van der Meer NJM, et al. patients treated on a stroke unit. Neurology 2011; 77: Confirmation of the high cumulative incidence of 965–972. International Journal of Stroke, 16(4) Siegler et al. 447 17. Hart RG, Catanese L, Perera KS, et al. Embolic stroke of 20. Lakomkin N, Dhamoon M, Carroll K, et al. Prevalence undetermined source: a systematic review and clinical of large vessel occlusion in patients presenting with acute update. Stroke 2017; 48: 867–872. ischemic stroke: a 10-year systematic review of the litera- 18. Aggarwal G, Lippi G and Henry BM. Cerebrovascular ture. J Neurointerv Surg 2019; 11: 241–245. disease is associated with an increased disease severity in 21. Pranata R, Huang I, Lim MA, Wahjoepramono EJ and patients with Coronavirus Disease 2019 (COVID-19): a July J. Impact of cerebrovascular and cardiovascular dis- pooled analysis of published literature. Int J Stroke 2020; eases on mortality and severity of COVID-19-systematic 15: 385–389. review, meta-analysis, and meta-regression. J Stroke 19. Li L, Yiin GS, Geraghty OC, et al. Incidence, outcome, Cerebrovasc Dis 2020; 29: 104949. risk factors, and long-term prognosis of cryptogenic 22. Yao H, Lu X, Chen Q, et al. Patient-derived mutations transient ischaemic attack and ischaemic stroke: a impact pathogenicity of SARS-CoV-2. medRxiv. Epub population-based study. Lancet Neurol 2015; 14: ahead of print 23 April 2020. DOI: 10.1101/2020.04.14. 903–913. 20060160. International Journal of Stroke, 16(4)