Introduction

Mechanical thrombectomy (MT) of occluded middle cerebral artery (MCA) or distal internal carotid artery (ICA) is a standard treatment for acute ischemic stroke within the first six hours from stroke onset [1]. However, recently published results of the randomized trials DAWN (DWI or CTP Assessment with Clinical Mismatch in the Triage of Wake-UP and Late Presenting Strokes Undergoing Neurointervention with Trevo) and DEFUSE-3 (The Endovascular Therapy Following Imaging Evaluation for Ischemic Stroke) showed that time window for MT may be extended beyond 6 hours in patients carefully selected on basis of the brain imaging findings and clinical deficit [2, 3]. In case of basilar artery occlusion (BAO), the evidence from randomized trials is still missing; however, MT may be considered regarding a high risk of fatal outcome.

Anticoagulation therapy (AT) is generally used as a prevention of systemic or cerebral embolisation; however, the patients on AT may suffer an embolic event including acute ischemic stroke (AIS) [4,5,6,7,8]. In case of an occlusion of large cerebral vessel, MT may be performed also in the patients using AT, but it might be associated with a higher risk of symptomatic intracerebral hemorrhage or other bleeding events and poorer outcome due to affected coagulation. Published results from the positive randomized trials and following the Highly Effective Reperfusion evaluated in Multiple Endovascular Stroke Trials (HERMES) meta-analysis did not provide reliable data concerning the safety and efficacy of MT in these patients [9,10,11,12,13,14]. Moreover, anticoagulation agents including a “new” direct oral anticoagulants (DOAC) have different safety profile, and the number of treated patients has an increasing trend worldwide, especially due to the DOAC in a few lasts years.

Thus we aimed to compare safety and efficacy of MT performed in AIS patients using AT to those without AT.

Methods

All consecutive non-selected patients, who were treated with MT between years 2009 and 2016 in both centers, were enrolled in this retrospective bi-center study. In all patients, the symptomatic occlusion of cerebral artery was detected on computed tomographic angiography (CTA) or magnetic resonance angiography (MRA). MT was performed in case of occlusion of middle cerebral artery (MCA; M1 or M2 segment), or distal part of internal carotid artery (ICA) or in case of occlusion of basilar artery (BA). In case of arterial occlusion in anterior circulation, MT was initiated within first 6 h after stroke onset. In case of BA occlusion, MT was performed within first 24 h after stroke onset. The size of hypodense area corresponding to acute infarction larger than 2/3 of MCA territory present on admission CT in the patients with IS in the anterior circulation was only imaging exclusion criterion for MT. Prior severe disability and known fatal malignity were the clinical exclusion criteria. Patients with “wake-up” strokes in the anterior circulation were treated with MT if no hypodense area corresponding to acute infarction was present on admission CT or no hypersignal area corresponding to acute infarction was present on admission MRI sequence fluid attenuation inversion recovery (FLAIR). Admission clinical status was evaluated using the National Institutes of Health Stroke scale (NIHSS) by a certified neurologist. In patients without AT, intravenous thrombolysis (IVT) was administered within first 4.5 h from stroke onset prior MT. In patients with BAO, IVT was performed within first 24 h from stroke onset. No reversal agent was used for a normalization of coagulation parameters before the initiation of MT.

Mechanical thrombectomy was performed using stent retrievers (Solitaire®, Catch Device® and Trevo®). Achieved recanalization status was assessed according to the Thrombolysis in Cerebral Infarction Scale (TICI) on the final angiogram [15].

In all patients, the occurrence of intracerebral hemorrhage (ICH) was assessed on the control CT or MRI after 24 h. Symptomatic ICH was defined as a local remote parenchymal hematoma (type 2) or subarachnoid hemorrhage associated with at least four-point increase in NIHSS score or leading to death [16].

In the patients using AT, the anticoagulation was restarted after the control brain CT/MRI excluding ICH. In the patients without AT, antiplatelet therapy was initiated after the control brain imaging excluding ICH.

Neurological deficit was evaluated using the NIHSS after 24 h and clinical outcome after 3 months using the modified Rankin scale (mRS). A score 0–2 points was considered a good outcome. The mRS scoring was performed by an experienced certified neurologist and mostly during scheduled outpatient visits. In some patients, the scoring was performed using phone call with patient’s relatives or caregivers.

The study protocol was in compliance with the Declaration of Helsinki (1975) and was approved by the ethical committee of our hospitals.

Statistical analysis

SPSS software (version 22.0; SPSS, Chicago, Illinois) was used for the statistical analysis. Fisher’s exact test and Kruskal–Wallis test were used for nonparametric variables. Data normality was tested using Shapiro–Wilk test. Multivariate logistic regression analysis was used to identify possible predictors of ICH, SICH, and good clinical outcome. All tests used an α-level of 0.05 for significance.

Results

In total, 703 patients were treated with MT using stent retrievers in both centers during the investigated period (years 2009–2016). Demographic and baseline characteristics of these patients are shown in Table 1. Eighty-eight (12.5%) patients (46% of males, mean age 75.5 ± 11.8 years,) used AT in the time of AIS and had a median baseline NIHSS score of 17 points. Patients on AT were significantly older and had more frequently atrial fibrillation (AF), ischemic cardiopathy (ICD) and hypertension without medication (Table 1). IVT was performed before MT in 501 (81.5%) patients without AT and 114 (18.5%) patients without AT were treated with MT alone.

Table 1 Patients’ demographic and baseline clinical characteristics
Full size table

The recanalization (TICI 2b and 3) was reached in 80% of AT patients and complete recanalization (TICI 3) in 65% of patients (p—1.000) (Table 2). Good 90-day clinical outcome (mRS 0–2) was present in 36% of AT patients and 35% of patients died within 3 months. ICH after MT was observed in 34% (p = 0.38) and SICH in 9% of patients on AT (p = 0.136, Table 2). In the group of patients without AT, no difference was found in the rate of ICH and SICH according to prior IVT; 29.1% of ICH and 5.0% of SICH in patients treated with prior IVT versus 25.4% of ICH and 5.3% of SICH in patients treated with MT alone.

Table 2 Results
Full size table

Patients on AT had significantly poorer 90-day clinical outcome after MT than patients without AT (p = 0.03, Table 2, Fig. 1). Patients on AT with poor outcome after MT had more frequently AF (69 vs. 93%, p—0.005), higher admission NIHSS (13.5 vs. 17, p—0.004), and higher rate of SICH (0 vs. 14.5%, p—0.047) (Table 3).

Fig. 1
figure 1

Distribution of mRS in patients on AT and without AT

Full size image
Table 3 Comparison of patients on AT with good and poor 3-month clinical outcome after MT
Full size table

Fifty (57%) of anticoagulated patients were on warfarin, 15 (17%) were on DOAC, and remaining 22 (26%) were on the therapeutic dose of low molecular weight heparin (LMWH). No significant difference was found in the recanalization rate, 90-day clinical outcome, and SICH occurrence among the individual types of AT (Table 4). In pooled analysis, no difference in the clinical outcome and rate of SICH was found between patients on warfarin with admission International normalized ratio (INR) < 1.7 and ≥ 1.7 (Table 5).

Table 4 Comparison of results according to individual types of AT
Full size table
Table 5 Mechanical thrombectomy in patients on warfarin
Full size table

Multivariate logistic regression analysis did not show the age, hypertension, diabetes mellitus (DM), ICD, admission NIHSS, recanalization, and complete recanalization as predictors of SICH in patients on AT (Table 6).

Table 6 Multivariate logistic regression analysis for prediction of SICH after MT in AT patients
Full size table

Discussion

Ischemic stroke is not rare in the patients using anticoagulation therapy, especially in the presence of AF. Large randomized clinical studies showed a risk of AIS about 1–1.4%/year on warfarin and about 0.9 to 1.34/year on DOAC [4,5,6]. Li et al. [8] reported risk of AIS between 2.6–5.0%/year in the patients using warfarin, which depended on time in the therapeutic range. Sjögren et al. [7] recently reported the risk of AIS 1.03%/year in patients on warfarin and 1.04%/year in patients on DOAC.

The results of our study showed that MT might be safe and relatively effective also in patients using AT. The higher rate of poor 90-day clinical outcome and higher 3-month mortality (35%) in the presented study may be related to higher rate of AF, higher admission NIHSS and higher rate of SICH after MT in these patients (Table 3). Presence of AF is generally associated with a poor outcome and with a higher mortality after IS [17].

Up to date, only very limited data are available in terms of MT performed in AIS patients treated with anticoagulants; however, MT is being performed in some patients on AT in experienced centers. Benavente et al. [18] analyzed retrospectively 30 patients on warfarin treated with MT. Good outcome was present in 46.7% of patients with the rate of 3-month mortality 6.7% in the analyzed group and SICH was detected in 16.7% of cases [18].

Patients using AT have generally a higher risk of ICH, and thus the occurrence of SICH after MT in our AT patients (9%) exceeded reported numbers from previous positive randomized trials (4.4%) [14], but the difference in the SICH rate between patients on AT and without AT was found not significant in our cohort (Table 2). A trend of low SICH rate was observed in subgroup of patients on DOAC (Table 4), which may correspond to a known better safety profile of DOAC with lower incidence of ICH compared to warfarin [19]. Moreover, a specific antidotum for immediate reversal of anticoagulant effect of dabigatran was approved recently [20] and may be considered to use before MT in patients with high risk of bleeding complications [21].

Interestingly, no difference was found in the rate of ICH, SICH, and mortality between subgroups of patients using warfarin according to the admission INR (Table 5), even though Benavente et al. [18] reported significantly higher mortality in the patients with admission INR ≥ 1.7. Our results showed no difference in the recanalization rate between both groups. Moreover surprisingly, none of the analyzed baseline, clinical, and recanalization parameters was found as a predictor of SICH after MT including the value of INR (Table 6).

Our study has some limitations. Retrospective bi-center study was conducted with a relatively small sample size; however, the risk of ischemic stroke is generally low in patients on AT, and only very limited data about MT in these patients are available up to now. We present retrospective patients’ cohort for 7 years, but all patients were treated with stent retrievers and were enrolled for the treatment according to the same criteria during whole analyzed period (see “Methods” section).

In conclusion, MT performed in patients using AT was not associated with a significantly higher rate of SICH. Patients on AT with poor outcome after MT had more frequently AF, higher admission NIHSS, and higher rate of SICH. Further large multicenter prospective studies are strongly needed, because of an increasing number of patients on AT worldwide.