Clinical characteristics and outcomes of acute ischemic stroke in patients with COVID-19: A systematic review and meta-analysis of global data

Objective. There is increased concern regarding acute ischemic stroke (AIS) in patients with coronavirus disease 2019 (COVID-19). The aim of this study was to depict the manifestations and outcomes of COVID-19-associated AIS. Methods. We systematically searched for eligible studies describing AIS in patients with COVID-19 using PubMed, Embase, and Web of Science up to November 29, 2021. We complied with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines and used the Newcastle–Ottawa Scale to assess data quality. The data were pooled using fixedand random-effects models. Results. Thirty-eight eligible studies involving 76,894 participants were included in this meta-analysis. Compared with AIS patients who did not have COVID-19, patients with COVID-19 were more likely to have anterior circulation stroke (odds ratio [OR]: 2.29, 95% confidence interval [CI]: 1.03 – 5.10; I2:  37%), particularly involving the internal carotid artery (OR: 1.85, 95% CI: 1.19 – 2.88; I2: 0); more severe neurological deficit (National Institutes of Health Stroke Scale [NIHSS]) (weighted mean difference [WMD]: 3.21, 95% CI: 2.13 – 4.29; I2: 64%); higher proportion of cryptogenic stroke (OR: 1.83, 95% CI: 1.24 – 2.70; I2: 62%), large vessel occlusion (OR: 1.68, 95% CI: 1.10 – 2.57; I2: 75%), and multi-territory involvement (OR: 2.64, 95% CI: 1.62 – 4.29; I2: 0%); higher C-reactive protein levels (WMD: 55.90, 95% CI: 33.32 – 78.49; I2: 67%), and D-dimer levels (standardized mean difference: 0.81, 95% CI: 0.52 – 1.10; I2: 59%). The proportion of poor outcomes were higher among patients with COVID-19, including increased risk of in-hospital death (OR: 3.70, 95% CI: 2.73 – 5.02; I2: 64%) and lower possibility of favorable discharge (OR: 0.49, 95% CI: 0.39 – 0.61; I2: 0). However, COVID-19 did not increase the risk of hemorrhagic transformation (OR: 1.34, 95% CI: 0.91 – 1.98; I2: 39%) and symptomatic intracerebral hemorrhage (OR: 1.46, 95% CI: 0.81 – 2.62; I2: 0). Conclusion. AIS patients with COVID-19 seem to display a pattern of large vessel occlusion and multi-territory infarcts. These patients have high inflammatory marker *Corresponding authors: Qingxiu Zhang (zhangqingxiu@163.com) Liqun Zhang (liqun.zhang@stgeorges.nhs.uk) Citation: Yao Z, Huang L, Cheng Y, et al. 2022, Clinical characteristics and outcomes of acute ischemic stroke in patients with COVID-19: A systematic review and meta-analysis of global data. Adv Neuro, 1(1): 28. https://doi.org/10.36922/an.v1i1.28 Received: December 8, 2021 Accepted: March 2, 2022 Published Online: March 24, 2022 Copyright: © 2022 Author(s). This is an Open Access article distributed under the terms of the Creative Commons Attribution License, permitting distribution, and reproduction in any medium, provided the original work is properly cited. Publisher’s Note: AccScience Publishing remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Advanced Neurology AIS in patients with COVID-19 Volume 1 Issue 1 (2022) 2 https://doi.org/10.36922/an.v1i1.28


Introduction
The outbreak of coronavirus disease 2019 (COVID- 19), caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has rapidly spread worldwide since December 2019, with more than 258 million confirmed cases and 5.17 million deaths as of November 24, 2021 [1] . Even though the infection mainly results in respiratory symptoms, an increasing number of cases in cerebrovascular disease, particularly acute ischemic stroke (AIS), have been confirmed [2] . The incidence of AIS varied from 1% to 3%, and reached up to 6% in seriously ill patients [3,4] . Emerging data suggest that stroke in the context of COVID-19 may be associated with increased mortality and disability and presents with unique manifestations [5,6] . Although most of these studies have limited sample sizes or are restricted to particular geographic regions, thus showing considerable heterogeneity among studies, these individual studies provide valuable data on patients with AIS who have COVID-19. Therefore, the meta-analysis that can break the regional limitations and collect outcomes and characteristics of patients with COVID-19 in the real world can provide new insights.
The previous meta-analyses have mainly focused on stroke risk factors and outcomes [7,8] . With emerging evidence, we performed an updated systemic review and meta-analysis to illustrate the specific clinical features, laboratory findings, neuroimaging, revascularization treatment, as well as short-term outcomes of patients with COVID-19 who have AIS, to assist with better identification and management of these patients.

Methods
This study was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines [9] .

Search strategy
We conducted a systematic search of PubMed, Embase, and Web of Science databases from their inception to November 29, 2021, with no language restrictions. The search algorithm was modified by an information specialist; details are available in Table S1. References cited in retrieved articles as well as any review articles were also reviewed to identify additional studies.

Inclusion and exclusion criteria
Two investigators independently screened the identified articles and selected studies using pre-specified criteria, with disagreements resolved through consensus. Studies were deemed eligible if they (1) were observational studies with information on clinical features and outcomes of new-onset ischemic stroke in patients with COVID-19; (2) included at least 20 patients with AIS over 18 years of age; (3) SARS-CoV-2 infection was confirmed with a positive polymerase chain reaction test or International Classification of Diseases, Tenth Revision (ICD-10) codes [10] ; and (4) the diagnosis of stroke was based on neuroimaging and clinical symptoms. We excluded comments, editorials, letters, reviews, case reports, small case series (<20 cases), animal studies, and duplicate publications involving the same patient cohorts.

Data extraction
Data extraction was conducted independently by two investigators using a pre-designed form. For each eligible article, we extracted the first author, publication year, study design, geographic region, recruitment period, clinical definition of COVID-19 used in the study, sample size, age, gender, clinical manifestations, laboratory findings, neuroimaging findings, and short-term (in hospital or on discharge) outcomes of AIS.
Stroke etiology was classified according to the Trial of ORG 10172 in Acute Stroke Treatment (TOAST) criteria [11] . Stroke severity was measured using National Institutes of Health Stroke Scale (NIHSS). Functional independence (i.e., favorable functional outcome) was defined as modified Rankin scale score 0-2.

Quality appraisal
We used the Newcastle-Ottawa Scale (NOS) to assess the methodologic quality of the selected studies [12] . Specifically, the NOS scale evaluates quality in three aspects, including selection of study groups, comparability of the study groups, and assessment of exposure or outcome of interest. A total of seven out of nine points is considered a low risk of bias, a score of 4 -6 points a moderate risk, and a score <4 points a high risk of bias. studies were included with 0.5 continuity correction [13] . We converted median and range in the reported data into mean and standard deviation [14] .
We used Q and I 2 statistics to assess statistical heterogeneity among studies [15] . Heterogeneity was considered to be present with an I 2 value >40%. We used a fixed-effects    model with the Mantel-Haenszel method to combine data when results were homogeneous. Otherwise, a random-effects model with the DerSimonian and Laird method was applied. In this study, we calculated the odds ratio (OR), standardized mean difference (SMD), or weighted mean difference (WMD) with the associated 95% confidence interval (CI) using both the fixed-and random-effects models, and compared them to assess potential heterogeneity.
Sensitivity analyses were performed by omitting one study at a time to investigate the robustness of the pooled results. When ten or more studies were proven eligible for meta-analysis, publication bias was assessed using funnel plots and Begg's test. All reported P-values are two-tailed, and a P < 0.05 was defined as statistically significant. Statistical analyses were conducted using RevMan version 5.0 (The Cochrane Collaboration, 2020) and Stata version 16.0 (StataCorp LLC, College Station, TX, USA).

Results
Literature searches of the three databases yielded a total of 7961 potentially relevant references. After removing duplicates and screening titles and abstract, 772 full-text articles were retrieved. Of these, 732 were removed as no data of interest were provided. Finally, 38 observational studies involving 76,894 individuals met the eligibility criteria and were included in the meta-analysis. The detailed process of study identification and selection is presented in Figure 1.

Study quality and publication bias
The methodological quality of each included study after critical appraisal using the NOS is summarized in Table S2.
Most included studies were assessed as low risk of bias (n = 31, 81.6%) whereas the remaining studies were assessed as moderate risk of bias (n = 7, 18.4%). There was no evidence of publication bias in the meta-analyses (Figures 2 and 3).

Etiology of AIS in patients with COVID-19
Regarding the etiology of stroke in patients with COVID-19 according to the TOAST criteria, cryptogenic stroke was the most common type ( In comparison to patients who did not have COVID-19, those with COVID-19 were more likely to develop cryptogenic stroke (OR: 1.83, 95% CI: 1.24 -2.70; I 2 : 62%; 11 studies); no differences were observed for other stroke subtypes (Table 3).

Imaging findings and stroke severity in patients with COVID-19
Patients with COVID-19 showed a higher proportion of large vessel occlusion (LVO)  Table 3).

Inflammation and coagulopathy in patients with AIS and COVID-19
As opposed to AIS patients without COVID-19, those with COVID-19 had higher levels of C-reactive protein

Outcomes of patients with COVID-19 in developing AIS
In patients with AIS and COVID- 19 In contrast to patients who did not have COVID-19 infection, COVID-19 status was associated with high in-hospital mortality (OR: 3.70, 95% CI: 2.73 -5.02; I 2 : 64%; 19 studies; Figure 4) and lower possibility of favorable   Table 3). We then omitted one study at a time to assess whether the statistical significance had changed. No study substantially influenced the results of the summary estimates (Figures 6 and 7).

Discussion
In this systematic review and meta-analysis investigating the clinical characteristic and outcomes of stroke in patients with COVID-19, we found that in comparison with patients not infected with COVID-19, those with COVID-19 were more likely to develop cryptogenic large vessel stroke that involved multiple territories, present with more severe stroke syndromes, have higher CRP and D-dimer levels, and have prolonged APTT/PT. Furthermore, COVID-19 was associated with an increased risk of in-hospital mortality and lower rates of functional independence on discharge in ischemic stroke patients, especially after reperfusion treatment.
In this meta-analysis, we found that patients with AIS and COVID-19 tended to have multi-territory infarcts with LVO. In this study, we first observed abnormalities in several coagulation and inflammatory markers in patients with COVID-19. Compared with non-COVID-19 patients, AIS patients with COVID-19 had higher or longer D-dimer, PT, and APTT levels. These results suggested that AIS in patients with COVID-19 may be a manifestation of SARS-CoV-2-related coagulation disorders [52] . Furthermore, recent clinical research reported that elevated CRP levels were closely related to increased stroke severity, hemorrhagic transformation, and in-hospital mortality [53,54] , which suggested that CRP might not only be a biomarker of inflammation but also acts as a direct participant in the pathological process of ischemic stroke [55] . Accumulated studies have shown that several potential mechanisms with COVID-19 are involved in the occurrence of AIS, mainly inducing thrombo-inflammation or immune thrombosis. Viral translation through angiotensin-converting enzyme 2 receptors expressed in vessel walls may contribute to endothelial dysfunction and thrombosis. Thrombo-inflammation is secondary to activation of immune cells involved in the defense against the virus and amplification of the cytokine system and complement cascade, resulting in activation of downstream pro-coagulant pathways [56] . COVID-19 infection may also induce cardiac arrhythmias resulting in embolic infarcts [57] . Further studies are needed to explore the potential underlying mechanisms are needed.
In this study, we also found that D-dimer levels were high in most patients with COVID-19, surpassing the threshold that has been identified as a predictor of in-hospital death [58] . This highlights the need to closely monitor patients with high levels of CRP and D-dimer for potential stroke, although the prevalence is relatively low. Consistent with previous studies [59,60] , our metaanalysis showed that patients with AIS and COVID-19 were more likely to have LVO and multi-territory infarcts; rapid patient evaluation is crucial for effective reperfusion treatment.
It is suggested that D-dimer and CRP levels may be associated with the severity of AIS in patients with COVID-19 [61,62] . Hence, for patients with hypercoagulable states, proper use of antithrombotic agents or antithrombotic therapy could be effective [63,64] . Tracking these biological markers will allow for early identification and even prediction of disease progression. Intensive studies on these markers may provide the basis for development of therapeutic and preventive strategies against COVID-19-related stroke.
Our meta-analysis demonstrated poor prognosis and high mortality in patients with COVID-19. The previous meta-analyses have reported a high mortality rate of 29.2% amongst patients with COVID-19 [65] . Our study reinforced this finding and found that patients with AIS and COVID-19 infection had a nearly 4-fold higher risk of mortality compared with their counterparts who did not have COVID-19 infection, as well as highly unfavorable outcome at discharge, even for younger patients. We noted that patients with AIS and COVID-19 who received intravenous thrombolysis treatment had higher rates of hemorrhagic transformation, which may be related to the deranged coagulation status in these patients. We found that patients with COVID-19 who developed AIS and who received reperfusion treatment tended to have poor outcomes and high mortality. Several aspects related to COVID-19 infection may explain our observation, including respiratory distress, multiorgan failure [21] , a high proportion of LVO and multi-territory involvement.  [20] , which may also contribute to the poor prognosis in these patients.
Only a few studies have reported 90-day functional outcome, making it difficult to draw conclusions in this meta-analysis. Martí-Fàbregas et al. concluded that 90-day functional outcome was comparable in patients with and without COVID-19 [35] , whereas a recent study demonstrated that 3-month outcome tended to be worse in patients with COVID-19 [66] . Additional well-designed studies are warranted to investigate functional outcomes beyond 3 months and factors contributing to long-term outcomes.
Our study has notable strengths. The large sample size and worldwide geographic coverage means that the findings of this meta-analysis have good generalizability. To minimize risk of bias, we restricted our meta-analysis to cohort, casecontrol, and cross-sectional studies with a large sample size and with low to moderate risk of bias based on strict quality assessment criteria. Our comparative data on patients with COVID-19 and AIS as well as patients without COVID-19 allow for clearer inferences regarding the impact of COVID-19 on the manifestations and outcomes of patients with AIS. Furthermore, we summarized data based on patients with AIS who received acute revascularization treatment and laboratory data in the context of COVID-19, which have not been explored in previous reviews.
Several potential limitations should also be noted. First, potential confounding variables may lead to an overestimation of association because we used unadjusted estimates for the meta-analysis. Second, we cannot fully exclude the possibility that there may be overlapping of some patients across the included studies. Third, the small number of events may reduce the reliability of some estimates. Fourth, the studies included in this analysis demonstrated significant methodological heterogeneity, although we tried to mitigate this using random-effects models; this should be considered when interpreting our results. Finally, because the follow-up duration for most included studies was short, long-term functional outcomes remain to be determined.

Conclusions
This systematic review and meta-analysis showed that patients with AIS who had COVID-19 infection tended to have cryptogenic LVO and multi-territory infarcts with high CRP and D-dimer levels. These patients had more severe stroke syndromes, worse functional outcome, and a higher in-hospital mortality rate, with or without reperfusion treatment. These findings provide evidence that vigilance regarding stroke is needed in patients with severe COVID-19 infection as well as a need for antithrombotic treatment. Further studies are required to elucidate the precise pathophysiological mechanism of cerebrovascular disease in patients with COVID-19 and best management.

Funding
This research was supported by the National Natural Science Foundation of China (No. 82071304 and 81671149 to Qingxiu Zhang).