Etiologic Subtypes of Ischemic Stroke in SARS-COV-2 Virus patients

Objective: To describe the ischemic stroke etiopathogenesis related to COVID-19 in a cohort of NYC hospitals. Background: Extra-pulmonary involvement of COVID-19 has been reported in the hepatic, renal and hematological systems. Most neurological manifestations are non-focal but few have reported the characteristics of ischemic strokes or investigated its pathophysiology. Methods: Over the last 6 weeks, data from four centers in New York City were collected to review the possible ischemic stroke types seen in COVID-19 positive patients. Their presentation, demographics, other related vascular risk factors, associated laboratory and coagulation markers, as well as imaging and outcomes were collected. Results: In our study, age range of patients was 25-75 with no significant male preponderance. 70% presented for acute hospitalization due the stroke. About a fifth did not have common risk factors for ischemic stroke like diabetes and hypertension. None had history of atrial fibrillation or smoking. 50% had poor outcome with four ending in mortality and one in a critical condition due ARDS. All had high Neutrophil/Lymphocyte ratio except one who demonstrated some neurological recovery. In 70% of our cases, D-dimer levels were collected, and all showed mild to severe elevation. None of the emergent large vessel occlusion (LVO) cases had known cardiac risk factors but two out of five were found to have cardiac abnormalities during the course of their hospitalization. All LVOs had hypercoagulable lab markers especially elevated D-dimer and/or Fibrinogen. The LVO patients were younger and sicker with a median age of 46 and mean NIHSS of 24 as opposed to non-LVOs with a median age of 62 and mean NIHSS of 6 respectively. Conclusion: COVID-19 related ischemic events can be small vessel, branch emboli or large vessel occlusions. The latter is often associated with either a hypercoagulable state or cardio-embolism. Patient outcomes were worse when multi-organ or pulmonary system failure prevailed. Keywords: COVID-19, Acute Ischemic strokes, Emergent Large Vessel Occlusion, Mechanical Thrombectomy


Introduction
The novel Coronavirus outbreak came to the fore in December 2019. Several features have been common manifestations of this disease: a primarily lower respiratory tract illness, a severe form that is more common in people with underlying diseases and higher mortality/case fatality in older populations. (1)(2)(3)(4) A particularly virulent form of immunological response, the cytokine storm syndrome especially seems to affect these vulnerable subgroups. (4)(5)(6) This syndrome is also often associated with extra-pulmonary complications of COVID-19. (5,7) In the last three months, multiple case reports have highlighted the extra-pulmonary complications of the disease. (7)(8)(9)(10) A prominent subgroup of these include thromboembolic complications. (9)(10)(11). They have also been often associated with lab markers suggesting an underlying inflammatory and hypercoagulable condition. (9,11,12) To this end, the specter of a COVID-19-specific coagulopathy has been raised, which is both inflammatory and consumptive, different from traditional disseminated intravascular coagulopathy (9).
Most neurological complications have focused on non-focal presentations like headaches, encephalopathy and skeletal muscle injury. (8,13) While some authors have mentioned the possibility of a form of meningoencephalitis, viral causation of neurological diseases have been mentioned less often. (14,15) Reports of patients presenting with focal signs secondary to ischemic strokes have been scant. (8,16) All rights reserved. No reuse allowed without permission.
was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The objective of this study was to prospectively study COVID-19 presenting as focal neurological disease caused by arterial thromboembolism to the intracranial circulation. We also wanted to better describe the different ischemic stroke subtypes seen in patients with SARS-COV2 infection especially with the view to assess its unique features seen in the context of COVID-19.

Methods
Retrospective case review protocol approval was obtained from the Institutional Review Boards at NYU School of Medicine, Brookdale Hospital University Medical Center, Jamaica Medical Center and Richmond University Medical Centers respectively. A retrospective analysis of the prospectively maintained stroke databases at these institutions between the period of March 15 to April 15, 2020 was performed. The period coincided with widespread testing in the hospitals where these patients were admitted. Patients who tested positive with SARS-COV-2 PCR were selected (Cobas SARS-COV-2 real time RT-PCR under EUA performed & Abbott Real-time SARS-CoV-2 PCR assay using M2000 platform). Only patients with ischemic strokes were selected. We excluded other neurovascular events like intracerebral hemorrhage, subarachnoid hemorrhage or venous thrombosis for the purpose of this analysis. The results from four NYC hospitals were pooled including two in Brooklyn, one in Queens and one in Staten Island.
Patients with both TIA and ischemic strokes were selected for this purpose. The clinical definition of an acute onset of a focal deficit to characterize an ischemic event was used. All patients underwent rapid imaging with CT and CT angiogram to confirm presence of a All rights reserved. No reuse allowed without permission.
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 (which this version posted May 8, 2020. Descriptive statistics were utilized as appropriate.

Results
Ten patients were included in this case series from four New York City hospitals: Richmond University Medical Center-Staten Island (1 patient), NYU Langone Health-Brooklyn (2 patients), Jamaica Hospital Medical Center (2 patients) and Brookdale Hospital University Medical Center (5 patients).
All rights reserved. No reuse allowed without permission.
was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Age of patients ranged between 27 and 75 years. About 80% of the patients had pre-existing conditions, the most common being diabetes (70%) and hypertension (60%). While some patients had comorbidities, no patient had preexisting atrial fibrillation, coronary disease, or cerebrovascular disease. One of the patients developed atrial fibrillation during their course in hospital. The majority of patients encountered (80%) presented to the emergency room (ER) with chief complaints of neurological deficits rather than respiratory symptoms (Table 1). Non-large vessel occlusions comprised lacunar infarcts, branch emboli or transient ischemic event (Table   2). Of note, only two patients (20%) presented within the 4.5 hour window for intravenous thrombolysis. Six (50%) presented more than 12 hours from last known normal and 1 presented more than 24 hours from last known normal. Nine patients (90%) were racial minorities and more than half (60%) were African-American.
The median age of large vessel stroke patients was 46 years, the youngest being in the twenties.
Median NIHSS was 24. While 80% of the patients had pre-existing conditions, the most common being diabetes (70%) and hypertension (60%), no patient had known coronary disease, cerebrovascular disease, or atrial fibrillation. One of the patients developed atrial fibrillation during their hospital course. Three patients reported no prior viral symptoms at the time of presentation. The most common viral illness complaints were cough and fever. Time interval from onset of viral illness to neurological complications was 1-21 days. Two patients who presented with LVO had no prior complaints of viral illness but went on to develop a severe course of the disease.
All rights reserved. No reuse allowed without permission.
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 (which this version posted May 8, 2020. Nevertheless, all but one demonstrated infarct on follow-up imaging. Most of the patients did poorly with mortality seen in three out of the five and one still critically ill. None were independent at the time of discharge. The mortality in two cases was secondary to pulmonary complications while one was due to infarct progression. The median age was 46 and NIHSS was 24. They also had a more severe form of viral illness compared with non-LVO patients. (Table 1)   [TABLE 1] Five patients presented with non-LVO syndromes. All had vascular risk factors of either diabetes or hypertension, but neither had atrial fibrillation, coronary or cerebrovascular disease. Their median age was higher (62) and NIHSS (6) lower than the LVO group. None received IV thrombolysis due to late presentation. Two had lacunar infarcts while two had infarcts related to distal embolization. One was deemed to be a transient ischemic attack (TIA). Their neurological All rights reserved. No reuse allowed without permission.
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 (which this version posted May 8, 2020. . https://doi.org/10.1101/2020.05.03.20077206 doi: medRxiv preprint and overall outcomes were better with only one mortality which was related to pulmonary progression. The other three were independent at the time of discharge. (Table 2) [  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 (which this version posted May 8, 2020.  17) The more severe form can also affect the heart, liver, kidneys and be associated with sepsis or septic shock. (7,10) Several case reports have also reported the possibility of a hypercoagulable condition caused by the virus. (9,11,18,19) This has specifically been seen in cases with higher incidence of VTE or PE. Although initially suspected to be due to poor prophylaxis, recent reports suggest a systemic hypercoagulable condition at play. (18,19)  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 (which this version posted May 8, 2020. In our series, we found that the demographic and disease features of COVID-19-related indices had a variable distribution. Our youngest patient was 27 while our oldest was 75. While the non-LVO group was older, the median age of both groups is lower than would be expected for the general population. In the case of LVO, a median age of 48 is expressly unusual. Only 40% of our cases were male which is different from the typical male preponderance in other studies. Ddimer was available in 70% of the cases and levels varied from 391 to 32000. Almost 80% had pre-existing conditions like diabetes and hypertension though none of them had history of atrial fibrillation or were chronic smokers. We observed that only six patients (60%) sought emergent All rights reserved. No reuse allowed without permission.
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 (which this version posted May 8, 2020. . https://doi.org/10.1101/2020.05.03.20077206 doi: medRxiv preprint medical care due to stroke symptoms, which result in a delay in care. Only three patients had a severe course of the pulmonary disease prior to the neurological event requiring them to be hospitalized. One was asymptomatic for three weeks. Five patients presented with emergent large vessel occlusion of which three underwent thrombectomy. One post-thrombectomy patient died from pulmonary progression. Most LVO cases did poorly with three ending in mortality while another one in a critical condition due to their lungs. All of these had high neutrophil/lymphocyte ratio except the one who had some neurological recovery. The LVO cases were typically younger, had a worse neurological presentation, more severe form of viral disease and higher levels of hypercoagulability markers than the non-LVO patients. In terms of our stroke etiological distributions, we encountered the gamut of small vessel strokes, branch or multifocal emboli, TIAs and large vessel occlusions. Most of these patients had vascular risk factors but nearly a fifth did not, and none had known coronary or cerebrovascular disease. Thus, it is possible that COVID-19 related thromboembolism may follow multiple separate pathogenetic pathways. In patients with pre-existing vascular risk factors, it may predispose to the ischemic event earlier than if purely driven by those underlying conditions. This could explain why some of our younger patients in the 40-50 age range presented with stroke. In these situations, it may follow the paradigm seen in other hypercoagulable conditions where two or more factors act synergistically to increase risk of stroke. (12) It can also be likened to the double-hit theory of many cancers. (24) It is also possible that in some cases this hypercoagulable condition by itself is enough to cause an embolic stroke. This may be correlated with elevated lab markers like D-Dimer or NLR ratio or presence of antiphospholipid antibodies. (8,23) A third pathway maybe related to its effect on the heart when it causes cardiomyopathy or myocarditis which in turn may predispose to cardio-embolism. (16,25) This All rights reserved. No reuse allowed without permission.
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 (which this version posted May 8, 2020. . https://doi.org/10.1101/2020.05.03.20077206 doi: medRxiv preprint mechanism may be mediated by ACE2 targeting which could affect the vascular endothelium or the heart directly. Finally, since the incidence of VTE is higher in these cases, the possibility of paradoxical emboli cannot be ignored. (12) To our knowledge, this is the first study that attempts to define the ischemic stroke etiologic subtype as well as pathophysiology, direct or indirect, in patients with SARS-COV-2 Virus.

Limitations
Our study has several limitations. This is a short analysis of neurovascular cases presenting to the hospital. As the number of cases in New York City and our hospitals continue to accumulate, we cannot determine the true rate of neurovascular events in COVID patients. Due to the stress on the resources created by COVID-19 some of the patients did not receive advanced imaging such as MRI to confirm whether a silent infarct may have occurred. Additionally, many cases had other risk factors and the embolic events may have been related to the underlying risk factor rather than the viral infection, or simply exacerbated by it. Finally, the type of strokes caused by the disease fit into several etiologic pathway rendering it difficult to delineate an appropriate treatment or prevention strategy.

Conclusion
Stroke in the setting of COVID-19 has an unusual presentation including atypical demographics and delayed time windows. Ischemic events can be small vessel, branch emboli or large vessel occlusions. The latter is often associated with either a hypercoagulable state or cardio-embolism.
Ischemic stroke can be a presenting symptom of COVID-19 and may not always be associated with severe disease markers including in the young, minorities and healthcare workers. All rights reserved. No reuse allowed without permission.
was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.