Intracranial Hemorrhage, Outcome, and Mortality After Intra-Arterial Therapy for Acute Ischemic Stroke in Patients Under Oral Anticoagulants
Background and Purpose—Use of intravenous tissue-type plasminogen activator (IV tPA) for acute ischemic stroke is restricted to patients with an international normalized ratio (INR) less than 1.7. However, a recent study showed increased risk of symptomatic intracranial hemorrhage after IV tPA use in patients with oral anticoagulants (OAC) even with an INR less than 1.7. The present study assessed the risk of symptomatic intracranial hemorrhage, clinical outcome, and mortality after intra-arterial therapy (IAT) in patients with and without previous use of OAC.
Methods—Consecutive patients treated with IAT from December 1992 to October 2010 were included. Clinical outcome and mortality were assessed 90 days after stroke onset. Patients with and without previous use of OAC were compared.
Results—Overall, 714 patients were treated with IAT. Twenty-eight patients (3.9%) were under OAC at time of symptom onset. Median INR in the OAC group was 1.79 (interquartile range [IQR], 1.41–2.3) and 1.01 (IQR, 1.0–1.09; P<0.0001) in the group without OAC. Patients treated with OAC at admission underwent more often mechanical-only IAT than did patients without OAC (46.4% versus 12.8%; P<0.0001). Comparing patients with and without previous use of OAC, we did not find any statistical difference in the rate of symptomatic intracranial hemorrhage (7.1% versus 6.0%; P=0.80), unfavorable outcome (modified Rankin Scale score, 3–6; 67.9% versus 50.9%; P=0.11), and mortality (17.9% versus 21.6%; P=0.58).
Conclusions—Previous use of OAC did not significantly increase the risk of symptomatic intracranial hemorrhage after IAT or the risk of unfavorable outcome and mortality 90 days after IAT.
Symptomatic intracranial hemorrhage (sICH) is a feared complication of thrombolysis for acute ischemic stroke (AIS). The risk of sICH in patients with AIS treated with intravenous tissue-type plasminogen activator (IV tPA) ranges from 5% to 7%.1,2 Current guidelines restrict the use of IV tPA to patients presenting with an INR of less than 1.7.3 A recent article compared the risk of sICH after IV tPA in patients with and without oral anticoagulants (OAC) at time of stroke onset.4 Patients with OAC presented a 10-fold increased risk of sICH. Remarkably, the INR was less than 1.7 in all included patients, casting doubts on the safety of IV tPA in patients with OAC at time of stroke onset.
Trials on pharmacological intra-arterial therapy (IAT) excluded patients with admission INR greater than 1.7.5–8 Since the publication of these trials, several mechanical recanalization devices became available, making it possible to lower the dose or even obviate the use of fibrinolytic drugs. In the Mechanical Embolus Removal in Cerebral Ischemia (MERCI) and Multi MERCI trials, patients with abnormal hemostasis were included.9,10 Causes of hemostatic abnormalities were not limited to OAC, but included heparin use, disseminated intravascular coagulation, drug-induced thrombocytopenia, and malignancy. Compared with patients with normal hemostasis, patients with hemostatic abnormalities did not have higher sICH rates.11 To date, it is not known whether OAC itself influences the risk of sICH, clinical outcome, and mortality 90 days after IAT. As the sICH risk is increased after IAT, both pharmacological and mechanical, we aimed to investigate whether prior use of OAC additionally increased the risk of sICH, unfavorable clinical outcome and mortality 90 days after IAT for stroke.
Patients and Methods
The Institutional Review Board of the University of Bern gave its consent for the use of the database for retrospective clinical research purposes.
Patient Selection and Work-Up
We analyzed consecutive patients with AIS who underwent IAT in the stroke unit of the University Hospital of Bern, Switzerland from December 1992 to October 2010. Details of patient work-up have been described previously.12 Use of OAC at the time of symptom onset was assessed by asking patients or their next of kin. Stroke etiology was classified according to the criteria of the Trial of Org 10172 in Acute Stroke Treatment.13 We did not reverse OAC, eg, with fresh-frozen plasma or vitamin K, before or during IAT.
Angiography and Thrombolysis
Criteria for IAT were: clinical diagnosis of AIS; baseline National Institutes of Health Stroke Scale (NIHSS) score of at least 4 points or isolated hemianopia or aphasia; absence of hemorrhage on cranial computed tomography or magnetic resonance imaging; cerebral digital subtraction angiography showing vessel occlusion correlating with the neurological deficit; less than 6 hours from symptom onset to intra-arterial thrombolysis, less than 8 hours to mechanical thrombectomy in anterior circulation stroke, and less than 12 hours from stroke onset to IAT in vertebrobasilar artery occlusion; and no individual clinical or laboratory findings advising against thrombolysis. IAT was performed after 4-vessel angiography for assessment of vessel occlusion and collateral circulation. The IAT strategy employed was classified as: mechanical only (including clot retrieval, aspiration, percutaneous transluminal angioplasty and/or stenting); pharmacological only (intra-arterial fibrinolytic therapy with urokinase [urokinase HS Medac]); and combined, ie, both mechanical and pharmacological IAT. Thromboaspiration was performed with 7 French and 8 French guide catheters in the carotid and vertebral arteries, and with 5 French catheters in the intracranial arteries. Mechanical recanalization techniques included the Catch Thromboembolectomy System (Balt), the Phenox CRC (Phenox GmbH), Solitaire FR Revascularization Device (ev3, Inc), Merci Retrieval System (Concentric Medical, Inc) and Penumbra System (Penumbra, Inc). Urokinase was administered via a microcatheter placed in or directly in front of the occluding thrombus. The administered dose of urokinase was documented in international units (IU).
Vessel recanalization was quantified according to the Thrombolysis in Myocardial Infarction (TIMI) classification and was dichotomized into poor (TIMI grades 0 and 1) and good (TIMI grades 2 and 3).14 We did not use either anticoagulants with heparin during the first 24 hours or the combination of aspirin with clopidogrel for the first 48 hours.
Identification and Classification of Intracranial Hemorrhage
Magnetic resonance imaging or computed tomography was performed routinely within 24 hours after thrombolysis or if any clinical deterioration was detected. Intracranial hemorrhage (ICH) was classified as symptomatic if a parenchymal hematoma type 2 was accompanied by a 4-point increase in NIHSS score or led to death, according to the Safe Implementation of Thrombolysis in Stroke-Monitoring Study (SITS-MOST) criteria15; otherwise, ICH was defined as asymptomatic.
The modified Rankin Scale and mortality were assessed 90 days after stroke during an outpatient visit or through a structured phone interview performed by neurologists or trained study nurses. Outcome was dichotomized as favorable (0–2) and unfavorable (3–6).16
Patients were grouped based on previous use of OAC. For recanalization and outcome analysis, patients were additionally grouped based on admission INR (≥1.7 versus <1.7). Subgroup analysis for recanalization and outcome were performed as follows: (1) for patients treated with a pharmacological only or pharmacological/mechanical recanalization strategy, (2) and for patients treated with a pharmacological-only therapy. Categorical variables were summarized as counts (percentage) and continuous variables as medians and interquartile ranges (IQR), because they were not distributed normally. The Fisher exact test was performed for cross-tabulation. Mann-Whitney U test (Wilcoxon rank sum test) was used for noncategorical variables. Power calculations for the differences in the rates between the 2 patient groups were performed for the rates of sICH, unfavorable clinical outcome, and mortality 90 days after stroke. The α significance level was set at P<0.05. Data were analyzed using the Stata 11 (StataCorp 2009; Stata Statistical Software: Release 11).
Baseline Patient Characteristics
In total, 714 patients with AIS underwent IAT from December 1992 to October 2010 at our stroke unit. Median age of the cohort was 65.7 years (IQR, 55.5–74); median NIHSS on admission was 15 (IQR, 11–19), and median time from symptom onset to IAT was 275 minutes (IQR, 223–335). Twenty-eight patients (3.9%) were taking OAC at time of symptom onset. Table 1 summarizes baseline characteristics of the 2 patient groups.
Compared with patients without previous use of OAC, patients with OAC more often had a previous stroke (25.0% versus 6.4%; P<0.0001), atrial fibrillation (75.0% versus 28.0%; P<0.0001), and were less frequently on antiplatelet therapy (3.6% versus 9.2%; P<0.0001). In patients with previous use of OAC, stroke severity assessed by the NIHSS score tended to be higher compared with those without OAC (median NIHSS score, 17 versus 15; P=0.17). The INR in patients under OAC was significantly higher (1.79 versus 1.01; P<0.0001), as was activated partial thromboplastin time (35.9 s versus 28.0 s; P<0.0001).
Recanalization Therapy and Outcome
Recanalization strategies differed significantly between patients with and without previous use of OAC (Table 2). In particular, patients with OAC underwent more frequently a mechanical-only IAT without thrombolytic medication (46.4% versus 12.8%) than did patients without OAC. The median dose of urokinase administered was lower in patients with previous use of OAC (375 000 IU versus 1 000 000 IU; P<0.001). The partial or complete recanalization rate did not differ between the 2 patient groups (71.4% versus 70.0%; P=0.87), and the rate of sICH (7.1% versus 6.0%; P=0.80) or asymptomatic ICH (10.7% versus 17.4%; P=0.36) in the first 24 hours following IAT did not differ either. The detailed characteristics of the 28 patients with OAC are outlined in Supplemental Table I (http://stroke.ahajournals.org). Table 3 summarizes the characteristics of recanalization therapy and outcome stratified by admission INR (≥1.7 versus <1.7). The administered dose of urokinase differed significantly between the 2 groups (125 000 IU versus 1 000 000 IU; P=0.001). Rates of the outcome variables including sICH did not differ significantly (5.0% versus 6.5%; P=0.93).
Because the variables sICH, clinical outcome, and mortality did not significantly differ between patients with and without OAC, we performed a power analysis for all 3 variables. Our analysis had 80% power to detect a relative risk increase of at least 9.66 for sICH and 1.57 for unfavorable outcome in patients with previous use of OAC. For mortality, our analysis had 80% power to detect a relative risk increase of at least 3.08 in patients without OAC use.
In this retrospective study of consecutive patients with AIS treated with IAT, previous use of OAC did not significantly increase the risk of sICH. Moreover, OAC did not increase the risk of unfavorable outcome and mortality 90 days after stroke.
Our results differ from the recently reported 10-fold sICH risk increase in patients with previous use of OAC treated with IV tPA.4 At least 2 reasons could account for this difference. First, our stroke team individualized the IAT strategy (mechanical, pharmacological, or both) based on OAC status and coagulation parameters. In patients with previous use of OAC, the treating neuroradiologists adopted more frequently a mechanical-only strategy and used, if at all, lower doses of urokinase compared with in patients without OAC. This approach likely reduced the sustained fibrinolytic activity in patients with previous use of OAC, possibly lowering the risk of sICH. Supporting this, the only 2 patients with previous use of OAC who suffered a sICH were treated with relatively high doses of urokinase (900 000 IU and 1 000 000 IU). Moreover, they both suffered from an intracranial carotid bifurcation occlusion with involvement of A1 and M1 segments (carotid T occlusion) and from an extensive ischemic region, both associated with poor outcome despite IAT and increased sICH risk after recanalization.17
Second, large-vessel occlusions are more likely to be the cause of stroke in patients with previous use of OAC as opposed to in patients without OAC, in whom stroke caused by small-vessel disease is more frequently encountered. In the present study on IAT, large-vessel occlusions were encountered in all patients, both with and without OAC, as opposed to in the cited study on IV tPA, where large-vessel occlusions were more likely in patients with OAC, increasing sICH risk.4
A recently published posthoc analysis of the MERCI and Multi MERCI trials showed similar low sICH rates.11 The risk of sICH was not increased in patients with abnormal as compared with patients with normal hemostasis (8.6% versus 8.5%). Similar to our study, a dose-response for INR and sICH was not reported, further questioning the role of OAC in increasing sICH risk in patients treated with IAT. However, important differences to our study need to be pointed out. First, our work is based on everyday clinical practice, allowing neuroradiologists to individualize the IAT strategy, which was not limited to the MERCI Clot Retriever. As a new finding, we observed that patients with an admission INR ≥1.7 treated with pharmacological or pharmacological/mechanical IAT did not have an increased risk of sICH, unfavorable outcome, or death 90 days after IAT (Supplemental Table II; http://stroke.ahajournals.org) as compared with patients with admission INR <1.7. Second, we focused on the specific role of OAC on sICH risk following IAT rather than on abnormal hemostasis. In particular, we were able to include 13 patients of 28 patients (46.4%) with previous use of OAC, but with INR less than 1.7. This patient subgroup is relevant in everyday practice, as subtherapeutic OAC represents a serious risk factor for stroke, especially cardioembolic strokes. Accordingly, we found a cardioembolic etiology twice as often in patients with previous use of OAC compared with those without (89.3% versus 44.5%).
As the clinical impact of ICH represents a continuum reaching from no symptoms through clinical worsening to death, we will briefly discuss asymptomatic ICH rates. The asymptomatic ICH rate was higher in patients without OAC compared with those with OAC, yet without reaching statistical significance (17.4% versus 10.7%; P=0.36). The higher urokinase doses employed in patients without previous use of OAC might explain the higher ICH rate in this group.18 Conversely, once ICH is present, OAC might increase the impact of ICH from an asymptomatic to a symptomatic stage.
The differences in recanalization strategies according to previous use of OAC did not influence recanalization rates, clinical outcome, or mortality rates. In fact, the rates of unfavorable outcome and mortality were not significantly increased in patients with previous use of OAC treated by IAT. In contrast, in the posthoc analysis of the MERCI and Multi MERCI trials, patients with abnormal hemostasis had lower rates of good outcomes compared with those of patients with normal hemostasis (9.4% versus 35.3%; P=0.0002).11 Again, the definition itself of abnormal hemostasis may explain these discrepant results. The laboratory-based definition of abnormal hemostasis in the MERCI posthoc analysis led to inclusion of severely ill patients, including disseminated intravascular coagulation, sepsis, and cancer. As a consequence, the etiology of the abnormal hemostasis itself probably biased the clinical outcome beyond stroke and recanalization rate. In contrast, a patient selection bias in our study may also have influenced the clinical outcome toward a more positive trend. At our institution, the decision whether to perform IAT is discussed among neurologists and interventional neuroradiologists, especially in patients with previous use of OAC, where a selection bias in favor of patients in good general health condition may have taken place. Comparisons of clinical outcome and mortality rates with patients treated with IV tPA are not possible because the cited study on patients with previous use of OAC before tPA did not report clinical outcome data.4
Our study has several limitations. First, the retrospective design may not allow verifying whether a patient selection bias took place, thus influencing the rates of sICH, clinical outcome, and mortality. In turn, this limitation is inherent to the everyday practice design, as we cannot perform an intention-to-treat analysis based on the prespecified treatment, which may or may not be finally implemented. Moreover, the denominator of candidates for IAT from which this group was culled was not available. Second, missing statistical significance does not necessarily mean missing clinical effect, especially in the subgroup analyses. For example, regarding the sICH rate, we were able to conclude with 80% confidence that OAC on admission increases the risk of sICH by 9.66 or more. Third, coagulation parameters were not routinely documented on the day after IAT, so we were not able to evaluate the impact of INR evolution on sICH. Finally, some patients with previous use of OAC had an INR less than 1.7, suggesting improper OAC intake or dosage.
In summary, previous use of OAC did not increase the risk of sICH after IAT to the same extent as was reported after IV tPA. Moreover, OAC use did not increase the risk of unfavorable outcome and mortality 90 days after stroke. In selected patients with previous use of OAC, mechanical IAT with or without low-dose thrombolytic drugs may represent a therapeutic option.
The authors wish to thank Pietro Ballinari for his valuable statistical advice.
Continuing medical education (CME) credit is available for this article. Go to http://cme.ahajournals.org to take the quiz.
The online-only Data Supplement is available at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.111.615476/-/DC1.
- Received January 25, 2011.
- Accepted May 3, 2011.
- © 2011 American Heart Association, Inc.
- Adams HP Jr..,
- del Zoppo G,
- Alberts MJ,
- Bhatt DL,
- Brass L,
- Furlan A,
- et al
- del Zoppo GJ,
- Higashida RT,
- Furlan AJ,
- Pessin MS,
- Rowley HA,
- Gent M.
IMS Study Investigators Combined intravenous and intra-arterial recanalization for acute ischemic stroke: the Interventional Management of Stroke Study. Stroke. 2004;35:904–911.
IMS II Trial Investigators The Interventional Management of Stroke (IMS) II Study. Stroke. 2007;38:2127–2135.
- Smith WS,
- Sung G,
- Starkman S,
- Saver JL,
- Kidwell CS,
- Gobin YP,
- et al
- Smith WS,
- Sung G,
- Saver J,
- Budzik R,
- Duckwiler G,
- Liebeskind DS,
- et al
- Nogueira RG,
- Smith WS,
- MERCI and Multi MERCI Writing Committee
- Meier N,
- Nedeltchev K,
- Brekenfeld C,
- Galimanis A,
- Fischer U,
- Findling O,
- et al
- Adams HP Jr.,
- Bendixen BH,
- Kappelle LJ,
- Biller J,
- Love BB,
- Gordon DL,
- Marsh EE 3rd.
- van Swieten JC,
- Koudstaal PJ,
- Visser MC,
- Schouten HJ,
- van Gijn J
- Arnold M,
- Nedeltchev K,
- Mattle HP,
- Loher T,
- Stepper F,
- Schroth G,
- et al
- Brekenfeld C,
- Remonda L,
- Nedeltchev K,
- Arnold M,
- Mattle HP,
- Fischer U,
- et al