key: cord-0794664-3g8tev0f authors: Sluis, W. M.; Linschoten, M.; Biesbroek, J. M.; den Hertog, H. M.; Ribbers, T.; Nieuwkamp, D.; van Houwelingen, R. C.; Dias, A.; van Uden, I. W.; Kerklaan, J. P.; Bienfait, H. P.; Vermeer, S. E.; de Jong, S. W.; Ali, M.; Wermer, M. J.; de Graaf, M. T.; Brouwers, P. J.; Asselbergs, F. W.; Kappelle, L. J.; van der Worp, H. B.; Algra, A. M.; consortium, CAPACITY-COVID collaborative title: Risk, clinical course and outcome of ischemic stroke in patients hospitalized with COVID-19: a multicenter cohort study date: 2021-03-12 journal: nan DOI: 10.1101/2021.03.11.21253189 sha: e72f6125267c1feff2f84d64b6be1a2ff9592516 doc_id: 794664 cord_uid: 3g8tev0f Background and purpose The frequency of ischemic stroke in patients with COVID-19 varies in the current literature, and risk factors are unknown. We assessed the incidence, risk factors, and outcomes of acute ischemic stroke in hospitalized patients with COVID-19. Methods We included patients with a laboratory confirmed SARS-CoV-2 infection admitted in 16 hospitals participating in the international CAPACITY-COVID registry between March 1st and August 1st, 2020. Patients were screened for the occurrence of acute ischemic stroke. We calculated the cumulative incidence of ischemic stroke and compared risk factors, cardiovascular complications, and in-hospital mortality in patients with and without ischemic stroke. Results We included 2147 patients with COVID-19, of whom 586 (27.3%) needed treatment at an intensive care unit (ICU). Thirty-eight patients (1.8%) had an ischemic stroke. Patients with stroke were older, but did not differ in sex or cardiovascular risk factors. Median time between onset of COVID-19 symptoms and diagnosis of stroke was two weeks. The incidence of ischemic stroke was higher among patients who were treated at an ICU (16/586; 2.7% versus 22/1561; 1.4%; p=0.039). Pulmonary embolism was more common in patients with (8/38; 21.1%) than in those without stroke (160/2109; 7.6%; adjusted RR: 2.08; 95%CI:1.52-2.84). Twenty-seven patients with ischemic stroke (71.1%) died during admission or were functional dependent at discharge and in-hospital mortality. Patients with ischemic stroke were at a higher risk of in-hospital mortality (adjusted RR 1.56; 95%CI:1.13-2.15) than patients without stroke. Conclusions In this multicenter cohort study, the cumulative incidence of acute ischemic stroke in hospitalized patients with COVID-19 was approximately 2%, with a higher risk in patients treated at an ICU. The majority of stroke patients had a poor outcome. The association between ischemic stroke and pulmonary embolism warrants further investigation. ischemic stroke were at a higher risk of in-hospital mortality (adjusted RR 1.56; 95%CI:1.13-2.15) than patients without stroke. In this multicenter cohort study, the cumulative incidence of acute ischemic stroke in hospitalized patients with COVID-19 was approximately 2%, with a higher risk in patients treated at an ICU. The majority of stroke patients had a poor outcome. The association between ischemic stroke and pulmonary embolism warrants further investigation. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted March 12, 2021. ; https://doi.org/10.1101/2021.03.11.21253189 doi: medRxiv preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted March 12, 2021. Coronavirus disease 2019 (COVID-19) is a pandemic that has affected millions of people worldwide. Although respiratory symptoms are the predominant feature of the disease, the clinical course of COVID-19 may be complicated by venous and arterial thromboembolic events. 1, 2 Pulmonary embolism accounts for the majority of these events, but other cardiovascular complications, including ischemic stroke, have also been reported. In contrast to some of the early reports suggesting an increased risk of acute ischemic stroke among patients hospitalized with COVID-19, results from later reports are less consistent. The occurrence of ischemic stroke varied greatly, ranging from 0.01% to 6.9%. [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] This may be explained in part by differences in study design, size, case-findings methods, and settings. Studies that reported clinical details have suggested an increased severity of stroke symptoms, a higher incidence of cryptogenic strokes, and a worse outcome, 3, 4, 18 including higher inhospital mortality rates, [3] [4] [5] in patients with COVID-19 than in those without. Nevertheless, large cohort studies reporting data on stroke severity, etiology, and outcome are still limited, probably due to the fact that ischemic stroke was often not reported as primary outcome and strokes were not assessed by neurologists. In addition, little data are available on the relationship between ischemic stroke and the occurrence of other cardiovascular complications in patients with COVID-19. To improve our understanding of the relation between COVID-19 and acute ischemic stroke, we assessed risk factors, time course, hospital setting, stroke severity, the relationship with other cardiovascular complications, and outcomes of ischemic stroke in patients hospitalized with COVID-19 during the first wave of the pandemic across 16 centers in the Netherlands. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted March 12, 2021. ; https://doi.org/10.1101/2021.03.11.21253189 doi: medRxiv preprint This study was conducted within the CAPACITY-COVID international patient registry (www.capacity-covid-eu; NCT04325412). Within this registry, data generated during routine clinical care was collected in a standardized manner. Details regarding CAPACITY-COVID have been outlined in detail elsewhere. 12 In short, CAPACITY-COVID extended the case report form (CRF) of the International Severe Acute Respiratory and emerging Infection Consortium (ISARIC) to collect in-depth information on cardiovascular history, the use of cardiovascular medication, and cardiac and thromboembolic events of patients hospitalized with COVID-19. STROCORONA was incorporated as a substudy within CAPACITY-COVID, to obtain additional detailed information on neurovascular history and the occurrence of acute ischemic stroke during hospitalisation, including data on vascular risk factors, etiology, severity, and outcome. Sixteen Dutch hospitals participated in STROCORONA. Ethical approval was obtained in all participating hospitals and the necessity of a consent procedure was determined conform local regulations. The majority of participating sites had an opt-out approach, were patients received written information about the study during hospitalisation. 19 The data used for this study, including de-identified individual participant data and a data dictionary defining each field or variable within the dataset, can be made available upon reasonable request to the data access committee of CAPACITY-COVID. These data will be made available following publication of this work. A data sharing agreement must be signed before any data are shared. We included adult patients with a laboratory-confirmed SARS-CoV-2 infection (determined by a positive polymerase chain reaction (PCR) test result from a nasopharyngeal swab) who All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted March 12, 2021. ; https://doi.org/10.1101/2021.03.11.21253189 doi: medRxiv preprint 8 were admitted to a hospital during the first wave of the pandemic in the Netherlands (from March 1 st to August 1 st , 2020). Patients who were suspected of COVID-19 but did not have a PCR-proven COVID-19 diagnosis were excluded. We retrieved data on demographics, comorbidities, pre-hospital medication, the need of mechanical ventilation, treatment at a high-dependency or intensive care unit (ICU) at any time during admission, in-hospital mortality, and the occurrence of other cardiac or thromboembolic complications: deep vein thrombosis (DVT), pulmonary embolism (PE), acute coronary syndrome (ACS), endocarditis, and new-onset atrial fibrillation (AF). Outcome definitions of in-hospital complications have been reported previously. 19 For STROCORONA, patient files of all eligible cases were systematically screened and scored by neurologists or other physicians with experience in stroke research to identify acute ischemic stroke during hospitalisation. In addition, data on prior neurovascular diseases, including transient ischemic attack (TIA), ischemic stroke, intracerebral hemorrhage, subarachnoid hemorrhage, or vascular dementia, were collected. 'Ischemic stroke' was defined as a sudden onset of focal neurological signs originating from the brain or retina that persists for more than 24 hours or until death, confirmed with neuroimaging demonstrating either infarction in the corresponding vascular territory or absence of another apparent cause. 20 We graded stroke severity at the time of diagnosis with the National Institutes of Health Stroke Scale (NIHSS; range 0 to 42, with higher scores indicating more severe neurological deficits) and collected data on acute stroke treatment (intravenous thrombolysis (IVT), endovascular treatment (EVT) and antithrombotic treatment), timing (median time between onset of COVID-19 symptoms and stroke diagnosis), and imaging findings (vascular territory, intracranial large vessel occlusion). We classified stroke etiology with the Trial of Org 10172 in Acute Stroke Treatment (TOAST) criteria and scored stroke outcome at discharge with the modified Rankin Scale (mRS). 21 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted March 12, 2021. ; https://doi.org/10.1101/2021.03.11.21253189 doi: medRxiv preprint Baseline characteristics are summarized with descriptive statistics as median (interquartile range; IQR), mean (standard deviation; SD), or frequencies (proportions) where appropriate. We did not impute missing data (details provided in Supplementary Material, please see https://www.ahajournals.org/journal/str). We calculated the cumulative incidence of ischemic stroke with corresponding 95% confidence intervals (95% CI) and stratified according to age and sex. Because three hospitals that participated in STROCORONA excluded patients that had no cardiovascular risk factors or were not consulted by a cardiologist during admission, we performed a sensitivity analysis excluding these three centers. We compared the occurrence of other cardiovascular complications and in-hospital mortality between patients with and without ischemic stroke with χ 2 or Student's t tests as appropriate and calculated risk ratios (RR) and 95% (CI) with Poisson regression. 22 We adjusted RRs for age, sex, and treatment on an ICU. For stroke outcome, we calculated the proportion of ischemic stroke patients with an unfavorable outcome (defined as death or dependency (mRS of ≥ 3)) at discharge. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted March 12, 2021. ; https://doi.org/10.1101/2021.03.11.21253189 doi: medRxiv preprint We included a total of 2147 patients in STROCORONA ( Figure 1 of the Supplementary Material, please see https://www.ahajournals.org/journal/str). Table 1 (Table 2 ). These patients were older than patients without ischemic stroke and had a lower BMI, but did not differ in terms of sex, cardiovascular comorbidities, and pre-hospital medication (Table 1 ; please see also Table 3 in the Supplementary Material, https://www.ahajournals.org/journal/str). Also after stratification by age we did not find differences in cardiovascular risk factors between patients with and without ischemic stroke (Supplementary Table 4 ). In a sensitivity analysis excluding three hospitals that excluded patients without cardiovascular risk factors or cardiologist consultation during admission, baseline characteristics and cumulative stroke incidence were similar (Supplementary Table 5 ). The median time between onset of COVID-19 symptoms and stroke diagnosis was 14 days (IQR: 9-25 days) for all patients, 23 days (IQR: 13-29) for patients who received ICU treatment and 10 days (IQR: 3-18) for patients treated on a general ward only (p=0.031; Figure 1 and Table 3 ). Stroke symptoms occurred as presenting complaint in 4 (10.5%) patients. The cumulative incidence of acute ischemic stroke was 2.7% in patients who were treated at an ICU (16/586; 95% CI: 1.7%-4.4%) and 1.4% in patients who received treatment on a general ward alone (22/1561; 95% CI: 0.9%-2.1%; p=0.039). Age-and sex stratified cumulative incidence are given in Table 2 and details about stroke severity, subtype, imaging, acute treatment and outcome in Table 3 . Stroke patients treated at All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted March 12, 2021. ; an ICU were younger than those treated at a general ward only (ICU: 63.4 years (SD: 15.2); general ward: 79.2 (SD: 8.1); p<0.001), frequently had other thromboembolic events (ICU: 8/16 (50%); general ward: 2/22 (9.1%); p=0.020), and had more severe strokes (ICU: median NIHSS 22.0; IQR: 3.8-30.0; general ward: 5.0; IQR: 2.8-17.5; p=0.050; Table 3 ). Eighteen patients (47.4%) had a stroke of undetermined etiology, however in 6 (33.3%) the diagnostic workup was incomplete due to early withdrawal of care. Differences between patients with cryptogenic stroke and those with a stroke of known etiology are summarized in the Supplement ( Table 6 ). The occurrence of other cardiovascular complications in patients with and without ischemic stroke is given in the Supplementary Material (Table 7 , please see https://www.ahajournals.org/journal/str). Pulmonary embolism was more common in patients with (8/38 (21.1%)) than in patients without ischemic stroke (160/2109 (7.6%); p=0.002), also after adjustment for age, sex, and treatment on an ICU (aRR 2.08; 95% CI: 1.52-2.84). The median time between onset of COVID-19 symptoms and occurrence of PE was 18 days (IQR: (Table 8) . All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted March 12, 2021. ; https://doi.org/10.1101/2021.03.11.21253189 doi: medRxiv preprint In our multicenter cohort of hospitalized patients with COVID-19, the overall cumulative incidence of acute ischemic stroke was 1.8%, with a higher rate of ischemic stroke in patients who needed treatment at an ICU (2.7%). Patients with acute ischemic stroke were older, but did not have more cardiovascular risk factors. In addition, patients with ischemic stroke were twice as likely to have pulmonary embolism and were at higher risk of in-hospital mortality. The overall incidence of ischemic stroke of 1.8% in hospitalized patients with COVID-19 is in line with previously published hospital-based COVID-19 cohorts, which reported cumulative incidences ranging between 1.0% and 2.4%. [3] [4] [5] Lower stroke rates have been found in studies that reported on a combination of hospitalized and non-hospitalized patients. 6, 23 Higher rates of up to 6.9% have been reported in populations consisting of ICU patients alone or in other selected patient populations. 5, 24, 25 In addition to differences in population and hospital setting, the variation in reported incidence of ischemic stroke may also be explained by other factors. First, most studies were performed in Asia and North America, with only a few European cohorts. [3] [4] [5] 7 Geographical variation may explain some of the heterogeneity, with a higher incidence reported in Asia than in North America and Europe. 3 In addition, regional differences in COVID-19 surges may have resulted in a higher threshold for seeking medical attention in some pandemic areas, especially for patients with mild symptoms of stroke. Second, in most studies ischemic stroke was recorded as one of various cardiovascular events, with case-ascertainment often not performed by neurologists or stroke physicians. [6] [7] [8] 26 This may have resulted in a systematic bias in the estimation of the cumulative stroke incidence among hospitalized patients with COVID-19. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted March 12, 2021. ; https://doi.org/10.1101/2021.03.11.21253189 doi: medRxiv preprint In contrast to some of the previous cohorts, 24, 27 our findings suggest that patients with COVID-19 who had acute ischemic stroke did not have more cardiovascular risk factors than patients who did not have a stroke. One explanation for this discrepancy may be that older patients with more vascular risk factors may not have been hospitalized or admitted to an ICU in the first place, because of treatment restrictions or patient preferences, which may have led to reduced survival rates in this group. 28 In addition, the greater severity of COVID-19 illness among hospitalized patients, especially those treated at an ICU, as well the increased risk of medical complications during hospitalization, such as venous thromboembolism due to immobilisation, may, at least partially, have contributed to the stroke risk in hospitalized COVID-19 patients without vascular risk factors. 29 To our knowledge, this is the first study to date to report on an association between pulmonary embolism and acute ischemic stroke in hospitalized patients with COVID-19 and to illustrate the time course of these cardiovascular complications. Although these findings need to be confirmed in additional studies, they may shed new light on the mechanisms by which COVID-19 is related to ischemic stroke, particularly in patients with no other vascular risk factors. Acute respiratory infections in general can act as a trigger for the short-term risk of ischemic stroke and myocardial infarction and are associated with a high risk of cardiovascular-related death, especially in older patients and in those with cardiovascular comorbidity. 30, 31 Two recent studies have compared the occurrence of acute ischemic stroke in hospitalized patients with COVID-19 versus those with influenza. One study found that patients with COVID-19 appeared to have an increased stroke risk (COVID-19: 1.6%; influenza: 0.2%), whereas the other study found the risk of ischemic stroke to be similar in patients with COVID-19 (1.2%) All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted March 12, 2021. ; https://doi.org/10.1101/2021.03.11.21253189 doi: medRxiv preprint 14 and influenza (1.2%). 9, 27 In SARS-CoV (SARS) and Middle-East Respiratory Syndrome (MERS), the occurrence of acute ischemic stroke has only been reported sporadically. 32 Pathophysiological mechanisms that could link COVID-19 to thromboembolic events include direct viral-induced endotheliitis, postinfectious immune-mediated responses, prothrombotic coagulopathy, and the occurrence of a hyperinflammatory state, with elevated D-dimer levels and antiphospholipid antibodies frequently found in patients with COVID-19 and thromboembolic complications. 2 Several studies have found that patients with COVID-19 who had acute ischemic stroke were more likely to die. [3] [4] [5] It remains unclear whether this association with an increased mortality is driven by disease severity and the prothrombotic state triggered by COVID-19. Alternatively, other confounding factors, such as impeded functional recovery due to fever and infection and withdrawal of care in patients with COVID-19 and ischemic stroke, may also play a role. 33, 34 Our study has some limitations. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted March 12, 2021. ; https://doi.org/10.1101/2021.03.11.21253189 doi: medRxiv preprint strokes with undetermined etiology in our study and others. Among patients with pulmonary embolism and ischemic stroke, the diagnostic work-up to rule out a patent foramen ovale was often not performed. In contrast, the relatively large proportion of patients with a cardioembolic atiology may reflect the accessibility of telemetry and ECG. Second, we were unable to adjust for changes in admission and treatment strategies that occurred during the first wave of the pandemic. This may have affected the rate of ischemic stroke and in-hospital mortality and may hamper the generalizability of our results to later waves and other settings. Due to the novelty of this pandemic, comparisons with hospital populations from previous years and across different waves of COVID-19 should be interpreted with caution. 7, 36 The main overall strength of the CAPACITY-COVID consortium is that it is a multinational and multi-disciplinary collaborative effort to systematically record all cardiovascular complications in patients with COVID-19 in a longitudinal fashion. By extending the consortium with STROCORONA, we were able not only to add cerebrovascular expertise and detailed acute ischemic stroke data, it also allowed us to link various cardiovascular complications in hospitalized patients with COVID-19. In conclusion, in our multicenter cohort the overall cumulative incidence of acute ischemic stroke in hospitalized patients with COVID-19 was approximately 2%, with a higher risk in patients treated at an ICU. Our finding that patients with COVID-19 and ischemic stroke were twice as likely to have pulmonary embolism than patients without stroke warrants further investigation, but may be relevant for our understanding of thrombotic mechanisms associated with COVID-19. Irrespective of potential pathophysiological relevance, our findings underscore the importance of appropriate antithrombotic strategies and increased awareness of stroke symptoms in hospitalized patients with COVID-19. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted March 12, 2021. ; https://doi.org/10.1101/2021.03.11.21253189 doi: medRxiv preprint Funding CAPACITY-COVID is funded by the Dutch and British Heart Foundation, the EuroQol Research Foundation, Novartis Global, Amgen Europe, Novo Nordisk Nederland, Servier Nederland and Daiichi Sankyo Nederland. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted March 12, 2021. ; https://doi.org/10.1101/2021.03.11.21253189 doi: medRxiv preprint All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Coagulation abnormalities and thrombosis in patients with COVID-19 Neurological associations of COVID-19 Stroke in covid-19: A systematic review and meta-analysis COVID-19 and stroke: Incidence and etiological description in a high-volume center Increased risk of acute stroke among patients with severe COVID-19: A multicenter study and meta-analysis Registry of arterial and venous thromboembolic complications in patients with COVID-19 Sars-cov-2 and stroke in a New York healthcare system Cerebrovascular events and outcomes in hospitalized patients with covid-19: The SVIN COVID-19 multinational registry Acute cerebrovascular events in hospitalized covid-19 patients COVID-19 related stroke in young individuals Cardiac complications in patients hospitalised with COVID-19 Proposed standardized neurological endpoints for cardiovascular clinical trials: An academic research consortium initiative Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. Toast. Trial of org 10172 in acute stroke treatment No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted March 12 Overestimation of risk ratios by odds ratios in trials and cohort studies: Alternatives to logistic regression Prevalence and outcomes of acute ischemic stroke among patients