Introduction

Severe acute respiratory syndrome coronavirus 2 (SARS-Cov-2) is a new strain of the coronavirus family that emerged in 2019 and was first reported in Wuhan, China, but spread exponentially to become a worldwide pandemic in months. Infection with SARS-Cov-2 primarily presents as a severe acute respiratory illness in human subjects [1].

Understanding of the 2019-coronavirus disease (COVID-19) has evolved significantly over time. In addition to respiratory symptoms, COVID-19 may present with various gastrointestinal, renal, neurological and cardiovascular symptoms [2, 3]. Attention has been drawn recently to the thromboembolic sequelae of this disease, including ischemic strokes, due to the severe outcomes associated with these complication [4,5,6]. Despite a significant overall reduction in the number of stroke cases presenting to hospitals during the pandemic, studies have demonstrated a remarkably high incidence of acute cerebrovascular disease in those with COVID-19 [7,8,9].

Although the indications for endovascular thrombectomy (EVT) of COVID-19 related large vessel occlusions (LVOs) are the same as for the general population, a protected code stroke algorithm has been introduced to expedite management while also ensuring the safety of healthcare providers [10,11,12]. In this study, we employed meta-analytic techniques to pool data from the literature to explore the outcomes and potential complications of EVT among COVID-19 patients.

Methods

Institutional Case Series Study Design

Institutional review board approval was obtained for a retrospective study of COVID-19 positive patients with large vessel occlusions (LVO) who underwent endovascular intervention between March 1 and June 20. Informed consent was waived. We studied patients’ demographics, clinical presentations, laboratory findings, National Institutes of Health Stroke Scale (NIHSS), CT angiography (CTA) and CT perfusion (CTP) findings, mechanical thrombectomy procedure, immediate post-procedural outcome and neurological outcome at final follow-up whenever available.

Literature Search

We performed a literature research using the PubMed search engine on September 1st, 2020 using different combinations of the keywords COVID-19, SARS-CoV-2, stroke, large vessel occlusion and thrombectomy. Case series and treatment arms of observational studies were included. Studies that included COVID-19 negative patients and non-LVO pathologies were excluded. Patient’s demographics, comorbidities, anatomic thrombus location, NIHSS at presentation and at 24 h, angiographic outcome via the thrombolysis in cerebral infarction (TICI) scale, thrombectomy technical details, and mRS.

Results

Patient Demographics

In our series, we identified 3 patients (mean age 64.3 ± 15.1) with findings consistent with acute stroke on non-contrast head CT, followed by CTA head and CT perfusion studies (An example is demonstrated Fig. 1). All the 3 patients had a sizeable ischemic penumbra and were deemed candidates for EVT according to Endovascular Therapy Following Imaging Evaluation for Ischemic Stroke 3 (DEFUSE 3) criteria [13]. Two patients had anterior circulation LVO and one patient with posterior circulation LVO. Clinical presentation, treatment and outcome are detailed in (Table 1).

Fig. 1
figure 1

A 75-year-old female with multiple comorbidities; a CT perfusion demonstrates a core infarction of 35 ml and mismatch ration of 4.3. b CTA shows occlusion of the distal basilar artery (arrow). c Left vertebral angiogram again reveals basilar artery occlusion. d Post thrombectomy angiogram reveals TICI3 perfusion

Full size image
Table 1 Patient demographics, clinical presentations, laboratory work-up, endovascular and clinical outcomes
Full size table

A total of 635 abstracts were identified from the literature search, from which 35 were identified for detailed review. Eight studies, including our case series, consisting of 73 patients were included for final analysis (Table 2). The average age of the included patients was 62.2 years, 67.6% of whom were male. All but one study included demographic data regarding vascular comorbidities. Of these, hypertension was the most prevalent (59.7%, n = 43), followed by diabetes (30.6%, n = 22), atrial fibrillation (18.5%), and hyperlipidemia (15.4%). Six studies described the basic anatomic location of cerebral thrombus location, 65.7% (n = 48) of which involved the anterior circulation. Involvement of the M1 segment was most commonly reported (53.8%). Involvement of multiple vascular territories was noted in 34.6% of cases. Carotid involvement occurred in 19.2% of cases. The mean NIHSS at presentation was 20.4 (12–40).

Table 2 Pooled data: patient demographics, treatments, neurological outcomes and complications
Full size table

Intervention

In total, 47.9% (n = 35) of patients received tPA and 89% (n = 65) underwent endovascular thrombectomy. Four studies describing 42 patients provided technical EVT details. The most utilized methods were the ADAPT technique (40.5%, n = 17) and stent-aspiration combination (33.3%, n = 14). Suction aspiration only and stent retrieval were used as primary techniques in 16.7% (n = 7) and 9.5% patients (n = 4), respectively (Table 3).

Table 3 Summary of patient characteristics and outcomes
Full size table

Outcomes

TICI ≥ 2b was reported in 83.1% (n = 54) of cases. The mean NIHSS at 24 h after presentation (documented in three studies) was 20 (12–25), representing no significant difference compared to NIHSS at presentation. Six patients were reported to have early cerebral re-occlusion and cerebral hemorrhage occurred in three patients during the post-procedural period. In-hospital mortality was reported in 28.8% (n = 15) of patients (Table 3).

Discussion

The COVID-19 pandemic has demonstrated a direct and indirect impact on the occurrence and presentation of acute cerebrovascular disease. The known associated prothrombotic state leading to arterial and venous thrombosis has contributed to stroke being the most commonly reported neurological complication [4, 6]. Using meta-analytic methodology to analyze outcomes data following EVT, our study highlights the various challenges of treating COVID-19 patients presenting with stroke.

Although COVID-19 related LVO has been reported in young patients less than 50 years old, the mean age of this patients’ pool was 62.2 years and is comparable to the mean of 68 years in pre-pandemic reports [8, 14, 15, 16•, 17]. This review found the NIHSS at presentation among COVID-19 patients is typically moderate to severe with a mean of 20.4 (range 7–40), consistent with a recent systematic review that reported a mean of 19 [18•]. In the series of patients from our institution, the two patients with anterior circulation LVO had NIHSS of 12 and 17 comparable to large-scale EVT trials who report an NIHSS range of 14–20 (mean of 17) [17]. This contrasts to the NIHSS of 40 in the patient with basilar occlusion, which was reported in 12.5% of the overall included patient pool. Re-evaluation of the NIHSS at 24 h after presentation, which was reported in three of the eight studies, demonstrated no significant difference with the NIHSS at presentation, suggesting EVT has minimal effect in the acute phase [19, 20•].

A small but significant proportion of the patients who underwent EVT (8%) had early proximal arterial re-occlusion after a successful thrombectomy. Overall post procedure intracerebral hemorrhage was 4%, which is comparable to previous EVT trials [17]. Though successful revascularization (TICI ≥ 2b) was achieved in 83% of patients regardless of technique used, the in-hospital mortality rate was high at 28.8% compared to 15.3% among non-COVID-19 patients [17]. In our series of patients, persistent neurological deficits were noted in two patients and one patient ultimately died from stroke and COVID-19 related respiratory complications. Despite a high proportion of successful angiographic results, these relatively poor outcomes compared to non-COVID-19 stroke patients may be attributed to differences in clot composition, with an increased tendency for clot fragmentation and migration that results in microemboli as described by Wang et al. [16•]. The sequalae of such microemboli are further exacerbated by the severity of the underlying COVID-19 respiratory disease that in conjunction with stroke indicates a worse overall prognosis [6].

Recent literature has reported the incidence of stroke among COVID-19 patients to range between 1.1 and 5.7% [5, 6, 9, 21, 22]. However, as patients with multiple cardiovascular risk factors are significantly more likely to develop severe COVID-19 disease as well as acute cerebrovascular disease, the association between COVID-19 and stroke is complex [23, 24]. Future research that employ multi-institutional cohorts of larger patient populations will be required to more clearly elucidate the relationship between comorbidities and LVOs related to COVID-19.

Our results are subject to inherent limitations of the retrospective nature of case series and case reports, including significant heterogeneity of included studies. Furthermore, several of the studies included in this systematic review failed to report relevant variables such as NIHSS at 24 h and mRS, and most of these were reported on a mean basis making it difficult to evaluate the progression or outcome of each patient. Future studies with a larger cohort and more detailed long-term neurological follow-up are required.

Conclusion

This systematic literature analysis demonstrates overall poor outcomes and high mortality in COVID-19 patients presenting with stroke despite angiographically successful EVT of LVOs in the majority of patients. An unusual incidence of early intracerebral proximal arterial re-occlusion and propensity for clot fragmentation and microemboli may be significant contributing factors.