ESUS indicates embolic stroke of undetermined source; LAD, large artery disease; and SVD, small vessel disease.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Perera KS, Ng KKH, Nayar S, et al. Association Between Low-Dose Rivaroxaban With or Without Aspirin and Ischemic Stroke Subtypes: A Secondary Analysis of the COMPASS Trial. JAMA Neurol. 2020;77(1):43–48. doi:10.1001/jamaneurol.2019.2984
What is the association between low-dose rivaroxaban plus aspirin and ischemic stroke subtypes?
In this secondary analysis of a multinational randomized clinical trial of 27 395 participants with systemic atherosclerotic disease who received low-dose rivaroxaban plus aspirin, rivaroxaban alone, or aspirin alone, 291 participants experienced an ischemic stroke. Treatment with low-dose rivaroxaban plus aspirin was associated with a significant reduction in cardioembolic stroke and embolic stroke of undetermined source compared with treatment with rivaroxaban alone or aspirin alone; this reduction was not seen in participants with other ischemic stroke subtypes.
For patients with systemic atherosclerotic disease, treatment with low-dose rivaroxaban plus aspirin was associated with a significant reduction in cardioembolic stroke and embolic stroke of undetermined source.
The COMPASS (Cardiovascular Outcomes for People Using Anticoagulation Strategies) randomized clinical trial was stopped early owing to the efficacy of low-dose rivaroxaban plus aspirin in preventing major cardiovascular events. The main reason for early trial termination was the effect of combination therapy on reducing ischemic strokes.
To analyze the association between low-dose rivaroxaban with or without aspirin and different ischemic stroke subtypes.
Design, Setting, and Participants
This is a secondary analysis of a multicenter, double-blind, randomized, placebo-controlled study that was performed in 33 countries from March 12, 2013, to May 10, 2016. Patients with stable atherosclerotic vascular disease were eligible, and a total of 27 395 participants were randomized and followed up to February 6, 2017. All first ischemic strokes and uncertain strokes that occurred by this date were adjudicated using TOAST (Trial of Org 10172 in Acute Stroke Treatment) criteria. The analysis of ischemic stroke subtypes was evaluated using an intention-to-treat principle. Statistical analysis was performed from March 12, 2013, to February 6, 2017.
Participants received rivaroxaban (2.5 mg twice a day) plus aspirin (100 mg once a day), rivaroxaban (5 mg twice a day), or aspirin (100 mg once a day).
Main Outcomes and Measures
Risk of ischemic stroke subtypes during follow-up.
A total of 291 patients (66 women; mean [SD] age, 69.4 [8.5] years; 43 [14.8%] had a previous nonlacunar stroke) experienced an ischemic stroke. During the study, 49 patients (16.8%) received a diagnosis of atrial fibrillation. Applying TOAST criteria, 59 strokes (20.3%) were cardioembolic, 54 strokes (18.6%) were secondary to greater than 50% stenosis of the ipsilateral internal carotid artery, 42 strokes (14.4%) had a negative evaluation that met criteria for embolic stroke of undetermined source, and 21 strokes (7.2%) were secondary to small vessel disease. There were significantly fewer cardioembolic strokes (hazard ratio [HR], 0.40 [95% CI, 0.20-0.78]; P = .005) and embolic strokes of undetermined source (HR, 0.30 [95% CI, 0.12-0.74]; P = .006) in the combination therapy group compared with the aspirin-only group. A trend for reduction in strokes secondary to small vessel disease (HR, 0.36 [95% CI, 0.12-1.14]; P = .07) was not statistically significant. No significant difference was observed between the 2 groups in strokes secondary to greater than 50% carotid artery stenosis (HR, 0.85 [95% CI, 0.45-1.60]; P = .61). Rivaroxaban, 5 mg, twice daily showed a trend for reducing cardioembolic strokes compared with aspirin (HR, 0.57 [95% CI, 0.31-1.03]; P = .06) but was not associated with reducing other stroke subtypes.
Conclusions and Relevance
For patients with systemic atherosclerosis, low-dose rivaroxaban plus aspirin was associated with large, significant reductions in cardioembolic strokes and embolic strokes of undetermined source. However, these results of exploratory analysis need to be independently confirmed before influencing clinical practice.
ClinicalTrials.gov identifier: NCT01776424
Patients with systemic atherosclerotic disease have an increased risk of atherothrombotic events in multiple vascular territories (ie, cerebral, coronary, and peripheral).1,2 Systemic atherosclerotic disease, specifically carotid artery and aortic arch atherosclerotic plaques, can cause strokes secondary to arteriogenic emboli.3-5 Rivaroxaban, an oral factor Xa inhibitor, is effective for preventing ischemic strokes secondary to atrial fibrillation (AF).6 Currently antiplatelet therapy remains the mainstay for secondary prevention of most non–AF-associated ischemic stroke subtypes7,8 except for internal carotid artery stenosis of greater than 50%, for which surgical intervention is established to be superior to medical therapy.7-9
The Cardiovascular Outcomes for People Using Anticoagulation Strategies (COMPASS) trial tested rivaroxaban, 2.5 mg, twice daily plus aspirin, 100 mg, vs rivaroxaban, 5 mg, twice daily vs aspirin, 100 mg, in patients with stable atherosclerotic disease.10 This trial was stopped early owing to the efficacy of low-dose rivaroxaban plus aspirin in preventing the primary outcome of major cardiovascular events.10 The major driver of early termination was the effect of combination therapy on reducing ischemic strokes compared with aspirin (hazard ratio [HR], 0.51 [95% CI, 0.38-0.68]; P < .001). In the Rivaroxaban in Patients with Heart Failure, Sinus Rhythm, and Coronary Disease (COMMANDER HF) randomized trial that enrolled patients with heart failure, even though low-dose rivaroxaban plus antiplatelet therapy was neutral for the primary composite outcome, it demonstrated a 34% relative risk reduction of ischemic stroke when compared with antiplatelet therapy (HR, 0.66 [95% CI, 0.47-0.95]).11 Both trials demonstrated large, consistent reductions in ischemic strokes with the combination of low-dose rivaroxaban plus aspirin.
We sought to assess the effect of low-dose rivaroxaban plus aspirin and of rivaroxaban alone compared with aspirin on different subtypes of ischemic stroke and, in particular, to explore whether combination therapy might be more effective in the prevention of ischemic stroke subtypes other than strokes secondary to symptomatic carotid stenosis and known AF.
The design and main results of the COMPASS trial have been previously published (trial protocol in the Supplement).12 The COMPASS trial was a multicenter, double-blind, randomized placebo-controlled trial performed from March 12, 2013, to May 10, 2016, for which patients were recruited at 602 hospitals, clinics, or community practices from 33 countries across 6 continents.10 In this double-blind trial, 27 395 participants with stable atherosclerotic vascular disease were randomly assigned to receive rivaroxaban, 2.5 mg, twice daily plus aspirin, 100 mg, once daily, rivaroxaban, 5 mg, twice daily, or aspirin, 100 mg, once daily. Patients with medical conditions that required anticoagulation, including AF, were excluded. Patients were also excluded if they had experienced an ischemic stroke within 1 month, prior hemorrhagic stroke, or symptomatic lacunar stroke. The independent data and safety monitoring board recommended early termination of the study, owing to meeting prespecified efficacy stopping rules in the low-dose rivaroxaban plus aspirin vs aspirin comparison. All first ischemic strokes (those that were confirmed as ischemic by results of neuroimaging or autopsy) and uncertain strokes (these were presumed likely to be ischemic based on clinical features) that occurred by this date were adjudicated as follows, and the total sample was included in this analysis. The protocol was approved by health authorities and institutional review boards in all participating countries and written informed consent was obtained from all participants. Ethics approval was not required for this secondary analysis (the data were deidentified).
For the present secondary analysis, 5 stroke experts (K.S.P., K.K.H.N., S.N., L.C., and R.G.H.) who were unaware of treatment allocation adjudicated all ischemic and uncertain strokes to determine cause and subtype of the stroke. Information from all case report forms and source documents were reviewed; additional source documents were requested from the clinical sites (discharge summaries, neuroimaging reports, and/or an investigator narrative). Trial of Org 10172 in Acute Stroke Treatment (TOAST) criteria13 were used to classify the cause of the ischemic stroke. The TOAST criteria classifies strokes as secondary to cardioembolism, in the presence of intracranial arterial occlusions presumably due to an embolus arising in the heart from a high-risk source of embolism, (eg, AF, left ventricular thrombus, or metallic prosthetic valve) in the absence of greater than 50% stenosis of the ipsilateral internal carotid artery.13 Strokes are classified as secondary to large artery atherosclerotic disease, presence of more than 50% stenosis of the ipsilateral internal carotid artery with no evidence of cardioembolic source on results of either echocardiogram or cardiac monitoring. Subcortical strokes less than 1.5 cm that were associated with classic lacunar syndromes are classified as secondary to small vessel disease. According to TOAST classification, strokes of undetermined source include the following 3 subtypes: strokes with 2 or more causes identified, strokes with negative evaluation, and strokes that had an incomplete evaluation. Strokes are classified with negative evaluation when they are cortically based with no ipsilateral carotid stenosis greater than 50% and no cardioembolic source identified by results of echocardiogram and cardiac monitoring. Hence, these strokes met the criteria for embolic stroke of undetermined source (ESUS).14
Statistical analysis was performed from March 12, 2013, to February 6, 2017. All P values were from 2-sided tests and results were deemed statistically significant at P < .05. Summary statistics for baseline characteristics were presented as means and SDs for continuous variables, with comparison between groups performed using 2-sample t test and analysis of variance. Categorical variables were summarized as counts and proportions, with groups compared using the χ2 test. The analysis of ischemic stroke subtypes was evaluated using an intention-to-treat principle and the comparisons between each of the rivaroxaban-based groups and the aspirin control group were performed using stratified log-rank tests with 5% type I error. Hazard ratios and corresponding 95% CIs were obtained from stratified Cox proportional hazards regression models. At the time of termination, follow-up was 99% complete. All statistical analyses were conducted using SAS, version 9.4 software (SAS Institute Inc).
A total of 291 patients experienced strokes (272 ischemic and 19 uncertain) during the mean follow-up period of 23 months. Baseline characteristics of the patients who experienced strokes are presented in Table 1. There were no statistically significant differences between baseline characteristics of the participants with ischemic or uncertain strokes in the 3 treatment groups. Likewise, there were no baseline differences reported among the total sample, as previously reported.10 The mean (SD) age of the participants who experienced an ischemic or uncertain stroke was 69.4 (8.5) years and 66 were women. A total of 259 participants (89.0%) had coronary artery disease, 247 (84.9%) had hypertension, and 149 (51.2%) had type 2 diabetes. Of the 291 participants, 43 (14.8%) had a previous nonlacunar stroke and 31 (10.7%) had a history of asymptomatic carotid disease (≥50% stenosis) or had undergone carotid revascularization. A total of 49 participants (16.8%) who had ischemic or uncertain strokes received a diagnosis of AF during the trial prior to or after their strokes: 10 patients in the combination therapy group, 13 patients in the rivaroxaban, 5 mg group, and 26 in the aspirin group.
Applying TOAST criteria, of the 291 ischemic or uncertain strokes that occurred during the study, 59 (20.3%) were cardioembolic, 54 (18.6%) were secondary to carotid stenosis greater than 50%, 21 (7.2%) were lacunar strokes secondary to small vessel disease, 42 (14.4%) had a negative evaluation and met the criteria for ESUS, 108 (37.1%) had an incomplete evaluation, and 5 (1.7%) had 2 or more potential causes (Figure 1).
Strokes secondary to cardiac embolism were significantly less frequent in the low-dose rivaroxaban plus aspirin group compared with the aspirin alone group (12 of 9152 [0.1%] vs 30 of 9126 [0.3%]; HR, 0.40 [95% CI, 0.20-0.78]; P = .005) (Table 2 and Figure 2). Strokes with a negative evaluation (ie, ESUS) also were significantly less frequent in the low-dose rivaroxaban plus aspirin group compared with the aspirin alone group (6 of 9152 [0.07%] vs 20 of 9126 [0.2%]; HR, 0.30 [95% CI, 0.12-0.74]; P = .006). Even though there was a trend toward benefit for lacunar strokes in the combination therapy group vs the aspirin-only group, it did not reach statistical significance (4 of 9152 [0.04%] vs 11 of 9126 [0.1%]; HR, 0.36 [95% CI, 0.12-1.14]; P = .07). No significant difference was observed in the combination therapy group vs the aspirin-only group in stroke due to internal carotid artery stenosis greater than 50% (18 of 9152 [0.2%] vs 21 of 9126 [0.2%]; HR, 0.85 [95% CI, 0.45-1.60]; P = .61).
When comparing the rivaroxaban, 5 mg, twice daily group with the aspirin-only group, there were fewer cardioembolic strokes in the rivaroxaban, 5 mg, twice daily group, but this finding was not statistically significant (17 of 9117 [0.2%] vs 30 of 9126 [0.3%]; HR, 0.57 [95% CI, 0.31-1.03]; P = .06). There were no significant reductions seen in other stroke subtypes between these 2 groups (Table 2).
In patients with stable systemic atherosclerotic disease, the combination of low-dose rivaroxaban (2.5 mg twice daily) plus aspirin was significantly superior to aspirin for prevention of ischemic strokes. When considering ischemic stroke subtypes, combination therapy had a significant effect on cardioembolic stroke (relative risk reduction, 60%) and on ESUS (relative risk reduction, 70%). Rivaroxaban alone (5 mg twice daily) was associated with a trend toward reduction in cardioembolic stroke with no difference in other stroke subtypes, and was associated with an increased risk of bleeding,10 making it a less attractive treatment option for stroke prevention.
The cause of ischemic stroke is heterogeneous and most strokes are embolic in origin, arising from either cardiac or arterial sources.14 In patients with ESUS, nonstenotic (<50% stenosis) carotid plaques and aortic arch atheroma play a significant role as the underlying source of emboli.5 Mechanistically, emboli arising from nonstenotic plaques in arteries are likely to be composed of both red and white thrombi, and while the red thrombus is likely to respond to anticoagulation, the white thrombus component may respond better to antiplatelet agents.14 Factor Xa is also known to interact with protease-activated receptors 1 and 2. Activation of protease-activated receptor 1 and 2 receptors in the arterial wall has the potential to increase endothelial dysfunction and inflammation.15 Hence, inhibition of these receptors by rivaroxaban combined with the effect on thrombin generation and the anticyclooxygenase effect of aspirin may better reduce arteriogenic embolism in individuals with systemic atherosclerosis with nonstenotic plaques. However, this effect was not seen in ischemic strokes attributed to carotid artery stenosis greater than 50%. This finding could be a result of the small number of strokes in this analysis. However, another possibility is that, in the presence of arterial stenosis greater than 50%, there is likely interplay between hypoperfusion and emboli formation from a ruptured plaque in the stroke mechanism in which poor collateralization and marginal blood flow reduces the washout and clearance of emboli.16 Therefore, prevention of emboli without treating the hypoperfusion may be unlikely to be sufficient to reduce the occurrence of stroke; hence, the previously shown superiority of surgical interventions vs medical management with antithrombotic therapy.
In AF, rivaroxaban, 20 mg (reduced to 15 mg in renal failure), is proven to be effective for stroke prevention.6 Studies have shown that rivaroxaban at doses of 5 to 80 mg result in dose-dependent inhibition of factor Xa activity from 20% to 61%17,18 and that there are persistent anticoagulant effects sufficient to prevent clot formation, secondary to ongoing thrombin inhibition even at trough plasma levels.17-19 The association of low-dose rivaroxaban plus aspirin and rivaroxaban alone with cardioembolic strokes (some of which were associated with incident AF that occurred during the COMPASS trial) raises the possibility that even at low doses rivaroxaban may have sufficient anticoagulant effect to prevent generation of AF-associated thrombi and other cardiac sources in some patients. Whether rivaroxaban, 2.5 mg, twice daily combined with aspirin can provide a similar degree of stroke prevention as rivaroxaban, 20 mg, in patients with AF is unknown. Further studies to test this hypothesis are warranted.
In this population with systemic atherosclerosis, hypertension (84.9%), and diabetes (51.2%), there was only a small fraction of lacunar strokes. Lacunar strokes are strokes caused from intrinsic disease of the small penetrating arteries.20 Hypertension, diabetes, and other as yet undetermined genetic factors promote thickening of the media of penetrating arteries.20 Even though most of the participants had hypertension and diabetes, these risk factors were well controlled, as evidenced by the mean blood pressures on recruitment and follow-up, which could explain the low incidence of lacunar strokes seen in this group. Accurate subtyping of stroke may not be possible in all cases, even with advanced neuroimaging and vascular imaging techniques and some strokes thought to be lacunar may have cardioembolic or atheroembolic causes.21-23 The trend toward benefit for lacunar stroke with combination therapy could be secondary to an accumulative effect of combination therapy on various different causes.
Our analysis of the secondary outcome of ischemic strokes is hypothesis generating and has its own limitations. Since the overall COMPASS trial was stopped before its planned number of total events, 95% CIs are less precise than those originally planned for the primary outcomes, secondary outcomes, and subgroup analyses. Our analysis was not prespecified; hence, most data necessary for appropriate classification of strokes were not collected initially and since a stroke workup was not mandated by the study, some participants did not receive a complete stroke workup. These 2 factors contributed to the high number of strokes with incomplete evaluation. To minimize the effect of these factors on stroke subtyping, the adjudicators reviewed all available source documents and, when additional documents were necessary for appropriate classification, sites were requested to provide them. Although the number of events could be viewed as modest, the study patient population was stable at the time of randomization and the nearly 2-year follow-up, coupled with careful stroke subtype classification, is a strength of our analysis.
Low-dose rivaroxaban plus aspirin was associated with a reduced risk of all ischemic strokes and with large, significant reductions in cardioembolic strokes and ESUS among patients with systemic atherosclerotic disease; however, these observations need to be independently confirmed before influencing clinical practice. In the wake of 2 recent trials with neutral results assessing anticoagulant treatment for secondary stroke prevention in patients with ESUS,24,25 these findings make a compelling case for testing this combination in patients with ESUS, as well as in those with cardioembolic strokes.
Accepted for Publication: June 6, 2019.
Corresponding Author: Kanjana S. Perera, MBBS, Population Health Research Institute, McMaster University, DBCVSRI C4-106, 237 Barton St East, Hamilton, ON L8L 2X2, Canada (firstname.lastname@example.org).
Published Online: September 16, 2019. doi:10.1001/jamaneurol.2019.2984
Author Contributions: Dr Perera and Ms Dyal had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Perera, Ng, Catanese, Connolly, Yusuf, Eikelboom, Hart.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Perera, Dyal, Hart.
Critical revision of the manuscript for important intellectual content: Perera, Ng, Nayar, Catanese, Sharma, Connolly, Yusuf, Bosch, Eikelboom, Hart.
Statistical analysis: Ng, Dyal.
Obtained funding: Connolly, Yusuf, Bosch, Eikelboom.
Administrative, technical, or material support: Catanese, Connolly, Bosch.
Supervision: Yusuf, Hart.
Conflict of Interest Disclosures: Dr Perera reported receiving grants from Bayer AG during the conduct of the study and outside the submitted work. Dr Sharma reported receiving grants and personal fees from Bayer during the conduct of the study; and receiving grants from Bristol-Myers Squibb and honararium and salary support for conducting research from Bayer, Daiichi Sankyo, and Boehringer Ingelheim outside the submitted work. Dr Connolly reported receiving honararium and salary support for conducting research from Bayer during the conduct of the study; and receiving honararium and salary support for conducting research from Bristol-Myers Squibb, Portola, Janssen, and Daiichi Sankyo outside the submitted work. Dr Yusuf reported receiving grants and honararium and salary support for conducting research from Bayer during the conduct of the study and outside the submitted work. Dr Bosch reported receiving honararium and salary support for conducting research from Bayer during the conduct of the study. Dr Eikelboom reported receiving grants and honararium and salary support for conducting research from Bayer, Boehringer Ingelheim, Daiichi Sankyo, Bristol-Myers Squibb/Pfizer, and Janssen during the conduct of the study and outside the submitted work. Dr Hart reported receiving honararium and salary support for conducting research from Bayer AG during the conduct of the study. No other disclosures were reported.
Funding/Support: The study was supported by Bayer AG.
Role of Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit manuscript for publication.