Key Points
Question
What is the optimal antithrombotic strategy for cancer-associated cryptogenic stroke?
Findings
In a secondary analysis of 1015 participants with cryptogenic stroke and atrial cardiopathy in the Atrial Cardiopathy and Antithrombotic Drugs in Prevention After Cryptogenic Stroke (ARCADIA) randomized clinical trial, 137 participants had history of cancer. Among this subgroup, the risk of major ischemic or major hemorrhagic events was nominally but not significantly lower with apixaban compared with aspirin, but analyses were underpowered given small numbers.
Meaning
Among ARCADIA trial participants with history of cancer, the risk of major ischemic or hemorrhagic events did not differ significantly with apixaban vs aspirin, but a clinically important benefit of apixaban could not be ruled out.
Abstract
Importance
Approximately 10% to 15% of ischemic strokes are associated with cancer; cancer-associated stroke, particularly when cryptogenic, is associated with high rates of recurrent stroke and major bleeding. Limited data exist on the safety and efficacy of different antithrombotic strategies in patients with cancer and cryptogenic stroke.
Objective
To compare apixaban vs aspirin for the prevention of adverse clinical outcomes in patients with history of cancer and cryptogenic stroke.
Design, Setting, and Participants
Post hoc analysis of data from 1015 patients with a recent cryptogenic stroke and biomarker evidence of atrial cardiopathy in the Atrial Cardiopathy and Antithrombotic Drugs in Prevention After Cryptogenic Stroke (ARCADIA) trial, a multicenter, randomized, double-blind clinical trial conducted from 2018 to 2023 at 185 stroke centers in North America. Data analysis was performed from October 15, 2023, to May 23, 2024.
Exposures
Oral apixaban, 5 mg (or 2.5 mg if criteria met), twice daily vs oral aspirin, 81 mg, once daily. Subgroups of patients with and without cancer at baseline were examined.
Main Outcomes and Measures
The primary outcome for this post hoc analysis was a composite of major ischemic or major hemorrhagic events. Major ischemic events were recurrent ischemic stroke, myocardial infarction, systemic embolism, and symptomatic deep vein thrombosis or pulmonary embolism. Major hemorrhagic events included symptomatic intracranial hemorrhage and any major extracranial hemorrhage.
Results
Among 1015 participants (median [IQR] age, 68 [60-76] years; 551 [54.3%] female), 137 (13.5%) had a history of cancer. The median (IQR) follow-up was 1.5 (0.6-2.5) years for patients with history of cancer and 1.5 (0.6-3.0) years for those without history of cancer. Participants with history of cancer, compared with those without history of cancer, had a higher risk of major ischemic or major hemorrhagic events (hazard ratio [HR], 1.73; 95% CI, 1.10-2.71). Among those with history of cancer, 8 of 61 participants (13.1%) randomized to apixaban and 16 of 76 participants (21.1%) randomized to aspirin had a major ischemic or major hemorrhagic event; however, the risk was not significantly different between groups (HR, 0.61; 95% CI, 0.26-1.43). Comparing participants randomized to apixaban vs aspirin among those with cancer, events included recurrent stroke (5 [8.2%] vs 9 [11.8%]), major ischemic events (7 [11.5%] vs 14 [18.4%]), and major hemorrhagic events (1 [1.6%] vs 2 [2.6%]).
Conclusions and Relevance
Among participants in the ARCADIA trial with history of cancer, the risk of major ischemic and hemorrhagic events did not differ significantly with apixaban compared with aspirin.
Trial Registration
ClinicalTrials.gov Identifier: NCT03192215
This secondary analysis of the Atrial Cardiopathy and Antithrombotic Drugs in Prevention After Cryptogenic Stroke (ARCADIA) randomized clinical trial compares apixaban vs aspirin for the prevention of adverse clinical outcomes in patients with history of cancer and cryptogenic stroke.
Introduction
Approximately 10% to 15% of patients with ischemic stroke have a history of cancer, among whom about half have active cancer at the time of stroke.1,2 Recurrent thromboembolism is common in these patients.3,4 While the exact mechanism of stroke remains undetermined in approximately 50% of patients with active cancer, making the stroke cryptogenic, autopsy and laboratory studies suggest that cancer-mediated hypercoagulability may contribute.5 Cancer can produce a prothrombotic state through secreted procoagulant factors, endothelial dysfunction, or circulating microparticles.5,6 Based on these considerations, anticoagulation is often prescribed for patients with cancer and cryptogenic stroke. Further, these patients often have concomitant venous thromboembolism, and low-molecular-weight heparins and direct oral anticoagulants are proven to be effective at preventing and treating venous thromboembolism in patients with cancer.7,8,9
However, the benefit of anticoagulation in cancer-associated stroke without concomitant venous thromboembolism remains unproven. Conventional stroke risk factors, such as atherosclerosis of the aorta and great vessels, remain common in patients with cancer, and these factors are typically treated with antiplatelet agents.10,11 Additionally, classic hypercoagulable causes of stroke in cancer, like marantic endocarditis and disseminated intravascular coagulation, are rarely definitively diagnosed in patients with cancer before death, even with targeted investigation.3,12 Patients with cancer also face up to a 20% annual risk of major bleeding with anticoagulant therapy.13
Given these conflicting considerations, the optimal antithrombotic strategy for cancer-associated cryptogenic stroke is uncertain. The 2021 American Heart Association/American Stroke Association guidelines state, “Patients who had a stroke attributable to hypercoagulability from cancer may be at a particularly high risk of bleeding with anticoagulation. In these patients, the benefit of anticoagulants for secondary stroke prevention is not well established, and further research is needed.”14 To address this knowledge gap, we conducted a post hoc subgroup analysis of the Atrial Cardiopathy and Antithrombotic Drugs in Prevention After Cryptogenic Stroke (ARCADIA) randomized clinical trial to compare apixaban vs aspirin in patients with recent cryptogenic stroke and history of cancer.
Methods
Design
The ARCADIA trial was a randomized, double-blind, phase 3 clinical trial conducted from 2018 to 2023 at 185 North American sites in the National Institutes of Health (NIH) StrokeNet and the Canadian Stroke Consortium (the trial protocol is available in Supplement 1).15 The study was approved by the relevant institutional review boards. Patients or their surrogates provided written informed consent. Data analysis was performed from October 15, 2023, to May 23, 2024. This study followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline.
Population
The ARCADIA trial enrolled patients aged 45 years or older with cryptogenic stroke and biomarker evidence for atrial cardiopathy. Cryptogenic stroke was defined per the consensus criteria for embolic stroke of undetermined source (ESUS), which requires a standardized diagnostic evaluation to exclude well-established causes of stroke.16 Atrial cardiopathy, as reported previously, was defined according to specific blood, echocardiogram, and electrocardiogram biomarker thresholds associated with increased stroke risk.17
Key exclusion criteria included atrial fibrillation, a clear indication for treatment-dose anticoagulant therapy (eg, venous thromboembolism within 3 months), a clear indication for antiplatelet therapy (eg, recent coronary stent), history of spontaneous intracranial hemorrhage, serum creatinine concentration of 2.5 mg/dL or higher (to convert to micromoles per liter, multiply by 88.4), clinically significant bleeding diathesis, hemoglobin concentration lower than 9 g/dL (to convert to grams per liter, multiply by 10.0), platelet count less than 100 × 103/µL (to convert to ×109 per liter, multiply by 1.0), clinically significant gastrointestinal bleeding within 1 year, or a modified Rankin Scale (mRS) score of 5 (the protocol in Supplement 1 lists full eligibility criteria). Active cancer, regardless of patient age, was not considered a distinct stroke mechanism or a clear indication for anticoagulant therapy. Patients with cancer were eligible for enrollment unless they were participating in another drug or acute stroke intervention clinical trial or site investigators considered their participation unsafe.
A history of cancer was recorded for all participants at the time of enrollment. As cancer was not the focus of ARCADIA, data on the site, histology, stage, activity, and diagnosis date of recorded cancers were not systematically collected or available for analyses.
Intervention
Patients who met the biomarker criterion for atrial cardiopathy were randomized in a 1:1 ratio to active apixaban and placebo aspirin or to active aspirin and placebo apixaban. Randomization could occur as early as poststroke day 3 and as late as poststroke day 120. Among participants with an initial NIH Stroke Scale (NIHSS) score of 11 or higher, hemorrhagic transformation of the index stroke, or uncontrolled hypertension, randomization was delayed until at least poststroke day 14. Participants randomized to active apixaban received an oral apixaban dose of 5 mg twice daily, except for participants meeting 2 or more of the standard adjusted-dose criteria (age ≥80 years, weight ≤133 lbs [to convert to kilograms, multiply by 0.45], or creatinine concentration ≥1.5 mg/dL), who received a dose of 2.5 mg twice daily. Participants randomized to active aspirin received an oral aspirin dose of 81 mg once daily. Participants, medical staff, and study investigators were all blinded to study drugs.
Mandatory study visits were scheduled around the following times after randomization: 30 days, 90 days, 180 days, 270 days, 360 days, and then every 180 days thereafter.
Measurements
For this post hoc analysis, the primary study outcome was a composite of major ischemic or major hemorrhagic events. Major ischemic events were defined as recurrent ischemic stroke, myocardial infarction, systemic embolism, symptomatic deep vein thrombosis (DVT), or symptomatic pulmonary embolism. Major hemorrhagic events were defined as any symptomatic intracranial hemorrhage or any major extracranial hemorrhage per ARCADIA criteria: clinically overt bleeding associated with a drop in hemoglobin by 2 g/dL or more during a 24-hour period or transfusion of 2 or more units of blood, bleeding in a critical site besides the brain, or bleeding causing death. Asymptomatic hemorrhagic transformation of the index stroke was not considered a major hemorrhagic event. Major hemorrhagic events were included in the composite end point to assess the net clinical benefit of apixaban vs aspirin because patients with cancer and stroke have an increased risk of both ischemic and hemorrhagic events and any reduction in ischemic events with apixaban could be offset by increased hemorrhagic events.18,19,20 Nonstroke ischemic outcomes were included because they are common, disabling, and potentially fatal events in patients with cancer and stroke, making them clinically relevant.3,4 Further, if a patient with cancer develops venous thromboembolism, they require indefinite anticoagulant therapy, resolving the question about how they should be treated for secondary stroke prevention.
Secondary efficacy outcomes were (1) recurrent ischemic stroke, (2) ischemic or hemorrhagic stroke, (3) any major arterial ischemic event, (4) symptomatic venous thromboembolism, and (5) any major ischemic event. Secondary safety outcomes were (1) all-cause mortality, (2) symptomatic intracranial hemorrhage, and (3) any major hemorrhagic event.
Data for the following baseline self-reported characteristics were also collected: age; sex; race (Asian, Black or African American, White, or Other race [defined as Alaska Native or American Indian, Native Hawaiian or Other Pacific Islander], or >1; site investigators and coordinators were instructed to ask participants to report self-identified race and ethnicity, which were then categorized per NIH guidelines); ethnicity; history of prior stroke, heart failure, coronary disease, peripheral vascular disease, hypertension, diabetes, or tobacco use (dichotomized as never vs past or current use). Body mass index and the NIHSS and mRS scores at enrollment were recorded.
Statistical Analysis
The intention-to-treat dataset was used for this analysis. Participants were censored at the time of noninformative death, withdrawal, or loss to follow-up. Fisher exact test and the Wilcoxon rank sum test were used to compare categorical and continuous study covariates between treatment groups. Incidence rates of study outcomes were calculated per 100 person-years of follow-up. Cox regression was used to test the null hypothesis that the hazard ratio (HR) comparing apixaban vs aspirin for the primary study end point of major ischemic or major hemorrhagic events was 1. Cox models were performed for participants with and without history of cancer. A test of interaction was conducted for cancer and study treatment. Log-log plots and visual inspection of the Kaplan-Meier curves confirmed that the proportional hazards assumption was met. Secondary outcomes were also analyzed via Cox regression.
In a sensitivity analysis, primary outcome rates between treatment groups were compared adjusting for the competing risk of death using the Fine and Gray subdistribution hazards model approach.
Performed analyses were exploratory, based on the convenience of a clinical trial sample. There were no missing data in this analysis. A 2-sided P < .05 was considered statistically significant. The analyses were performed using Stata version 15 statistical software (StataCorp LLC).
Results
Participant Characteristics
Among 1015 participants randomized in ARCADIA (median [IQR] age, 68 [60-76] years; 551 [54.3%] female), 137 (13.5%) reported a history of cancer at the time of stroke (the flow diagram is shown in the eFigure in Supplement 2). Participants were enrolled a median (IQR) of 13 (3-68) days after their index stroke. Among the 137 participants with cancer, the median (IQR) age was 74 (68-80) years, 75 (54.7%) were women, and 119 (86.9%) were White (Table 1). Their median (IQR) baseline NIHSS score was 1 (0-2). Among the 878 participants without cancer, the median (IQR) age was 67 (59-75) years, 476 (54.2%) were women, and 641 (73.0%) were White. Their median (IQR) baseline NIHSS score was 1 (0-3). Compared with participants without cancer, participants with cancer were older and more often White, while medical comorbidities and stroke severity were similar between groups. Of the 137 participants with cancer, 76 were randomized to aspirin and 61 were randomized to apixaban. Treatment group characteristics were well balanced among patients with cancer (Table 2).
Table 1. Characteristics of Participants Stratified by History of Cancer at Enrollment.
| Characteristica | Cancer (n = 137) | No cancer (n = 878) |
|---|---|---|
| Demographic | ||
| Age, median (IQR), y | 74 (68-80) | 67 (59-75) |
| Sex, No. (%) | ||
| Female | 75 (54.7) | 476 (54.2) |
| Male | 62 (45.3) | 402 (45.8) |
| Race, No. (%)b | ||
| Black or African American | 15 (10.9) | 199 (22.7) |
| White | 119 (86.9) | 641 (73.0) |
| Otherc | 3 (2.2) | 38 (4.3) |
| Ethnicity, No. (%)b | ||
| Hispanic or Latino | 8 (5.8) | 74 (8.4) |
| Not Hispanic or Latino | 129 (94.2) | 804 (91.6) |
| Medical comorbidities, No. (%) | ||
| Coronary artery disease | 17 (12.4) | 87 (9.9) |
| Congestive heart failure | 9 (6.6) | 62 (7.1) |
| Diabetes | 33 (24.1) | 282 (32.1) |
| Hypertension | 103 (75.2) | 681 (77.6) |
| Prior stroke | 23 (16.8) | 174 (19.8) |
| Tobacco use | 58 (42.3) | 372 (42.4) |
| BMI, median (IQR) | 28 (25-32) | 29 (25-33) |
| Stroke severity or functional status score, median (IQR) | ||
| Baseline NIHSS | 1 (0-2) | 1 (0-3) |
| Baseline modified Rankin Scale | 1 (0-2) | 1 (0-2) |
Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); NIHSS, National Institutes of Health Stroke Scale.
There were significant between-group differences for age and race only.
Race and ethnicity were self-reported. Site investigators and coordinators were instructed to ask participants to report self-identified race and ethnicity, which were then categorized per National Institutes of Health guidelines.
Includes Alaska Native or American Indian, Native Hawaiian or Other Pacific Islander, or more than 1.
Table 2. Characteristics of Participants With History of Cancer at Enrollment Stratified by Treatment Group.
| Characteristica | Aspirin (n = 76) | Apixaban (n = 61) |
|---|---|---|
| Demographic | ||
| Age, median (IQR), y | 74 (69-80) | 73 (68-79) |
| Sex, No. (%) | ||
| Female | 38 (50.0) | 37 (60.7) |
| Male | 38 (50.0) | 24 (39.3) |
| Raceb | ||
| Black or African American | 9 (11.8) | 6 (9.8) |
| White | 65 (85.5) | 54 (88.5) |
| Otherc | 2 (2.6) | 1 (1.6) |
| Ethnicityb | ||
| Hispanic or Latino | 3 (3.9) | 5 (8.2) |
| Not Hispanic or Latino | 73 (96.1) | 56 (91.8) |
| Medical comorbidities, No. (%) | ||
| Coronary artery disease | 8 (10.5) | 9 (14.8) |
| Congestive heart failure | 6 (7.9) | 3 (4.9) |
| Diabetes mellitus | 17 (22.4) | 16 (26.2) |
| Hypertension | 56 (73.7) | 47 (77.0) |
| Prior stroke | 14 (18.4) | 9 (14.8) |
| Tobacco use | 30 (39.5) | 28 (45.9) |
| BMI, median (IQR) | 28 (25-32) | 28 (25-32) |
| Stroke severity or functional status score, median (IQR) | ||
| Baseline NIHSS | 1 (0-3) | 1 (0-2) |
| Baseline modified Rankin Scale | 1 (0-2) | 1 (0-2) |
Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); NIHSS, National Institutes of Health Stroke Scale.
No significant between-group differences existed for any studied participant characteristic.
Race and ethnicity were self-reported. Site investigators and coordinators were instructed to ask participants to report self-identified race and ethnicity, which were then categorized per National Institutes of Health guidelines.
Includes Alaska Native or American Indian, Native Hawaiian or Other Pacific Islander, or more than 1.
Primary Outcome
Median (IQR) follow-up time was 1.5 (0.6-2.5) years in participants with history of cancer and 1.5 (0.6-3.0) years in participants without history of cancer. The incidence rate for the primary composite outcome (major ischemic or hemorrhagic events) was 10.6 (95% CI, 7.1-15.8) per 100 person-years among participants with cancer vs 5.9 (95% CI, 4.8-7.3) per 100 person-years among participants without cancer (HR, 1.73; 95% CI, 1.10-2.71; P = .02). Adjusting for demographic characteristics and comorbidities, cancer remained associated with an increased risk for the primary composite outcome (adjusted HR, 1.73; 95% CI, 1.09-2.75; P = .02).
Among the cohort with cancer, 8 of 61 participants (13.1%) randomized to apixaban and 16 of 76 participants (21.1%) randomized to aspirin developed a major ischemic or hemorrhagic event (Table 3). The accompanying incidence rates were 7.8 (95% CI, 3.9-15.7) per 100 person-years for the apixaban group and 12.8 (95% CI, 7.8-20.9) per 100 person-years for the aspirin group. However, the risk was not significantly different between groups (HR, 0.61; 95% CI, 0.26-1.43; P = .26). At 1 year from randomization, the Kaplan-Meier rate of the primary composite outcome was 7.5% (95% CI, 2.9%-19.0%) among participants randomized to apixaban and 17.9% (95% CI, 10.5%-29.6%) among participants randomized to aspirin (Figure 1). Among participants without history of cancer, the HR for the primary outcome with apixaban vs aspirin was 0.97 (95% CI, 0.64-1.45; P = .87). The P value for an interaction between cancer and study treatment was .35.
Table 3. Outcomes Among Participants With History of Cancer at Enrollment Stratified by Treatment Group.
| Outcome | Aspirin (n = 76) | Apixaban (n = 61) | HR (95% CI) | ||
|---|---|---|---|---|---|
| No. (%) | Incidence rate, No./100 person-years (95% CI) | No. (%) | Incidence rate, No./100 person-years (95% CI) | ||
| Primary outcome | |||||
| Major ischemic or major hemorrhagic event | 16 (21.1) | 12.8 (7.8-20.9) | 8 (13.1) | 7.8 (3.9-15.7) | 0.61 (0.26-1.43) |
| Secondary efficacy outcome | |||||
| Recurrent ischemic stroke | 7 (9.2) | 5.3 (2.5-11.2) | 5 (8.2) | 4.7 (1.9-11.2) | 0.87 (0.28-2.76) |
| Ischemic or hemorrhagic stroke | 9 (11.8) | 6.8 (3.6-13.2) | 5 (8.2) | 4.7 (1.9-11.2) | 0.68 (0.23-2.03) |
| Major arterial ischemic event | 9 (11.8) | 7.0 (3.7-13.5) | 6 (9.8) | 5.6 (2.5-12.5) | 0.79 (0.28-2.23) |
| Symptomatic DVT or PE | 6 (7.9) | 4.7 (2.1-10.4) | 1 (1.6) | 0.9 (0.1-6.7) | 0.21 (0.02-1.71) |
| Major ischemic event | 14 (18.4) | 11.2 (6.6-18.9) | 7 (11.5) | 6.6 (3.1-13.9) | 0.59 (0.24-1.47) |
| Secondary safety outcome | |||||
| All-cause mortality | 4 (5.3) | 3.0 (1.1-8.1) | 3 (4.9) | 2.8 (0.9-8.7) | 0.94 (0.21-4.19) |
| Symptomatic ICH | 2 (2.6) | 1.5 (0.4-6.1) | 0 | NA | NA |
| Major hemorrhagic event | 2 (2.6) | 1.5 (0.4-6.1) | 1 (1.6) | 1.0 (0.1-6.9) | 0.61 (0.06-6.73) |
Abbreviations: DVT, deep vein thrombosis; HR, hazard ratio; ICH, intracranial hemorrhage; NA, not applicable; PE, pulmonary embolism.
Figure 1. Cumulative Rates of Major Ischemic or Major Hemorrhagic Events in Participants With Cancer Randomized to Apixaban vs Aspirin.
Kaplan-Meier curves estimate the cumulative incidence functions of major ischemic or major hemorrhagic events among participants with history of cancer on enrollment in the Atrial Cardiopathy and Antithrombotic Drugs in Prevention After Cryptogenic Stroke (ARCADIA) trial. The shaded areas represent the 95% CIs for the estimated rates. Hashmarks represent participant censoring for death, withdrawal, or loss to follow-up.
Secondary Outcomes
The incidence rates for secondary outcomes among participants with history of cancer are provided in Table 3. Rates of all efficacy outcomes were numerically, but not significantly, lower in the apixaban group. For the apixaban vs aspirin groups, events included recurrent stroke (5 [8.2%] vs 9 [11.8%], respectively) and major ischemic events (7 [11.5%] vs 14 [18.4%], respectively). The incidence rates for recurrent stroke were 4.7 (95% CI, 1.9-11.2) per 100 person-years for the apixaban group and 6.8 (95% CI, 3.6-13.2) per 100 person-years for the aspirin group (HR, 0.68; 95% CI, 0.23-2.03; P = .49). One year from randomization, the Kaplan-Meier rates for recurrent stroke were 5.8% (95% CI, 1.9%-17.0%) with apixaban and 9.1% (95% CI, 4.2%-19.2%) with aspirin (Figure 2).
Figure 2. Cumulative Rates of Recurrent Stroke in Participants With Cancer Randomized to Apixaban vs Aspirin.
Kaplan-Meier curves estimate the cumulative incidence functions of recurrent ischemic or hemorrhagic stroke among participants with history of cancer on enrollment in the Atrial Cardiopathy and Antithrombotic Drugs in Prevention After Cryptogenic Stroke (ARCADIA) trial. The shaded areas represent the 95% CIs for the estimated rates. Hashmarks represent participant censoring for death, withdrawal, or loss to follow-up.
Major hemorrhagic events among participants with history of cancer were rare (3 in total). This comprised 1 major extracranial hemorrhage (1.6%) in the apixaban group and 2 symptomatic intracranial hemorrhages (2.6%) in the aspirin group. One year from randomization, the all-cause mortality rate was 3.8% (95% CI, 1.0%-14.2%) with apixaban and 3.1% (95% CI, 0.8%-11.7%) with aspirin.
Sensitivity Analysis
When adjusting for the competing risk of death, the relative hazard for major ischemic or hemorrhagic events with apixaban vs aspirin was largely unchanged. The subdistribution HR was 0.61 (95% CI, 0.26-1.41; P = .25).
Discussion
In an exploratory analysis of the ARCADIA randomized clinical trial, among the subgroup with history of cancer who had a cryptogenic stroke, the risk of major ischemic or hemorrhagic events with apixaban compared with aspirin was nominally but not significantly lower, with an estimated HR of 0.61 (95% CI, 0.26-1.43) in favor of apixaban. This numerical difference favoring apixaban was driven by the difference in venous thromboembolism between groups, although recurrent stroke was also numerically fewer with apixaban than aspirin. Meanwhile, major hemorrhagic events were fairly rare with either antithrombotic drug. Given the small sample size of this patient subgroup, these analyses were underpowered, and a clinically important benefit of apixaban could not be ruled out. The results should be considered hypothesis generating and do not answer the question about whether anticoagulant therapy benefits patients with cryptogenic stroke and history of cancer.
The subgroup of ARCADIA participants with history of cancer had considerably higher incidence rates for major ischemic or hemorrhagic events than participants without history of cancer, highlighting that patients with cancer and cryptogenic stroke are an especially high-risk population. When a clear indication for anticoagulant therapy such as DVT is absent, the optimal antithrombotic treatment strategy for these patients is unknown. The most commonly used strategies are on-label antiplatelet therapy based on best-practice guidelines for the secondary prevention of cryptogenic stroke or off-label anticoagulant therapy based on pathophysiological considerations and the high risk for recurrent thromboembolic events in patients with cancer.5,14 Practice patterns often depend on institutional and geographic factors, and few high-quality data support one strategy over the other.21
In a study by Seok et al,22 among 29 patients in Korea with active cancer and acute ischemic stroke, anticoagulant use was associated with reductions in serial D-dimer concentration, a surrogate marker for recurrent stroke risk and hypercoagulability. In a larger follow-up study,23 patients with cancer and stroke whose D-dimer level decreased with anticoagulation to less than the cohort’s median baseline level had improved 1-year survival. These studies did not evaluate antiplatelet therapy, and it is uncertain whether aspirin or other antiplatelet drugs affect D-dimer levels in this population. Conversely, among 263 patients with active cancer and ischemic stroke at a US cancer center, there were no differences in the risks of recurrent stroke or other thromboembolism between patients treated with anticoagulant or antiplatelet therapy.3 Additionally, among 543 patients with history of cancer enrolled into the NAVIGATE ESUS randomized trial,24 which compared rivaroxaban, 15 mg daily, vs aspirin, 100 mg daily, in patients with ESUS, both recurrent ischemic stroke risk (annual rate, 7.7% vs 5.4%; HR, 1.43; 95% CI, 0.71-2.87) and major bleeding risk (annual rate, 2.9% vs 1.1%; HR, 2.57; 95% CI, 0.67-9.96) were nonsignificantly higher in patients receiving rivaroxaban. However, these findings should be interpreted cautiously as most cancers in the NAVIGATE ESUS trial were apparently inactive, with only 9% diagnosed in the prior year. Further, the nonstandard dose of rivaroxaban limits generalization to other anticoagulant therapies.
Two prospective trials have focused on cancer-associated stroke. The Trial of Enoxaparin vs Aspirin in Patients With Cancer and Stroke (TEACH) was a pilot randomized, open-label clinical trial, conducted at 3 centers in New York, of enoxaparin, 1 mg/kg twice daily, vs aspirin, 81 to 325 mg once daily, among 20 patients with active solid or hematological cancer and acute ischemic stroke.25 The trial’s primary aims were to evaluate the feasibility of each treatment and determine whether equipoise existed among physicians and patients. TEACH found that while both physicians and patients had equipoise regarding randomization between these 2 antithrombotic strategies, patients generally preferred oral medications over injections. The small number of patients prohibited any meaningful comparisons of clinical outcomes between groups. The Edoxaban for the Treatment of Coagulopathy in Patients With Active Cancer and Acute Ischemic Stroke (ENCHASE) study was a randomized, open-label clinical trial, conducted at 2 centers in Korea, of edoxaban, 60 mg once daily, vs enoxaparin, 1 mg/kg twice daily, among 40 patients with cancer and ESUS.26 Its main findings were that interval changes in serum D-dimer levels and transcranial Doppler microembolic signals at 7 and 90 days after index stroke were similar between treatment groups. ENCHASE did not include an antiplatelet group. In this context, our findings add meaningfully to the literature on cancer-associated cryptogenic stroke by further supporting the potential for improved outcomes with anticoagulant over antiplatelet therapy while also reinforcing the current existence of equipoise regarding these treatments.
This study’s strengths include the multicenter, randomized, double-blind design; the use of apixaban, arguably the most effective and safest oral anticoagulant, administered at a standard dose for stroke prevention; a patient population with diversity in race, ethnicity, and sex; and longer follow-up time and more granular characterization of events compared with the NAVIGATE ESUS analysis.24
Limitations
This study has limitations. First, data on cancer type, stage, treatments, and diagnosis date were not systematically collected in ARCADIA; therefore, we were unable to determine how these important cancer characteristics might affect the risk-benefit profile of apixaban vs aspirin. Similarly, it should be highlighted that this was an unselected cohort of patients with cancer, and we do not know which documented cancers were active during the follow-up period. The observed increased risk for major ischemic or hemorrhagic events in this subgroup is consistent with the known increased risk for thromboembolic and bleeding outcomes in patients with active cancer after stroke.4,27 In a post hoc analysis, we reviewed event descriptions, which included admission notes and discharge summaries for study participants with a history of cancer who developed a recurrent stroke outcome. Among these patients, 36% had clear documentation of active cancer. This proportion, as well as the overall proportion of patients with any reported cancer in ARCADIA (13.5%), is nearly identical to published, nationally representative US data from 2019,2 when ARCADIA was enrolling. Second, this exploratory subgroup analysis was underpowered and should be considered hypothesis generating only. Third, the ARCADIA population was unique in that randomized patients not only had cryptogenic stroke but also had biomarker evidence for atrial cardiopathy. While data do not exist to support such an association, atrial cardiopathy could have modified the treatment effects of apixaban vs aspirin in the subgroup of patients with cancer.
Fourth, the anticoagulant studied in ARCADIA was apixaban at a dosage of 5 mg (or 2.5 mg) twice daily, so our results may not be generalizable to other anticoagulant medications or doses. Future head-to-head trials with other anticoagulants are needed. Small biomarker-driven investigations have suggested that subcutaneous low-molecular-weight heparins may be more effective than other forms of anticoagulation at correcting coagulopathy in patients with cancer and cryptogenic stroke.28,29 The ENCHASE trial,26 however, demonstrated that edoxaban and enoxaparin had comparable effects on biomarkers of hypercoagulability and cerebral thromboembolism. Moreover, multiple randomized clinical trials have proven comparable efficacy and safety between factor Xa inhibitors and low-molecular-weight heparins for the treatment of cancer-associated venous thromboembolism,8,30 another common form of cancer-associated thrombosis. These findings have led to the preferential real-world use of oral factor Xa inhibitors because of their easier route of administration.31 Furthermore, this practice is supported by the findings in the TEACH trial,25 where injections were the leading reason for enrollment failure and study drug discontinuation, highlighting their poor long-term feasibility in cancer-associated stroke. Fifth, we lacked data on D-dimer, an important risk biomarker in cancer-related stroke.
Conclusions
Among participants in the ARCADIA trial with history of cancer, the risk of major ischemic and hemorrhagic events with apixaban compared with aspirin nominally but not significantly favored apixaban; however, analyses were underpowered given small numbers. A definitive trial of anticoagulant vs antiplatelet treatment in patients with active cancer and cryptogenic stroke is needed.
Trial Protocol
eFigure. ARCADIA Trial Cancer Subgroup Analysis Flow Diagram
Data Sharing Statement
References
- 1.Wilbers J, Sondag L, Mulder S, Siegerink B, van Dijk EJ; Dutch String-of-Pearls Stroke Study Group . Cancer prevalence higher in stroke patients than in the general population: the Dutch String-of-Pearls Institute (PSI) stroke study. Eur J Neurol. 2020;27(1):85-91. doi: 10.1111/ene.14037 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Otite FO, Somani S, Aneni E, et al. Trends in age and sex-specific prevalence of cancer and cancer subtypes in acute ischemic stroke from 2007-2019. J Stroke Cerebrovasc Dis. 2022;31(12):106818. doi: 10.1016/j.jstrokecerebrovasdis.2022.106818 [DOI] [PubMed] [Google Scholar]
- 3.Navi BB, Singer S, Merkler AE, et al. Recurrent thromboembolic events after ischemic stroke in patients with cancer. Neurology. 2014;83(1):26-33. doi: 10.1212/WNL.0000000000000539 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Navi BB, Zhang C, Sherman CP, et al. Ischemic stroke with cancer: hematologic and embolic biomarkers and clinical outcomes. J Thromb Haemost. 2022;20(9):2046-2057. doi: 10.1111/jth.15779 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Navi BB, Kasner SE, Elkind MSV, Cushman M, Bang OY, DeAngelis LM. Cancer and embolic stroke of undetermined source. Stroke. 2021;52(3):1121-1130. doi: 10.1161/STROKEAHA.120.032002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Navi BB, Sherman CP, Genova R, et al. Mechanisms of ischemic stroke in patients with cancer: a prospective study. Ann Neurol. 2021;90(1):159-169. doi: 10.1002/ana.26129 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Lee AY, Levine MN, Baker RI, et al. ; Randomized Comparison of Low-Molecular-Weight Heparin Versus Oral Anticoagulant Therapy for the Prevention of Recurrent Venous Thromboembolism in Patients With Cancer (CLOT) Investigators . Low-molecular-weight heparin versus a coumarin for the prevention of recurrent venous thromboembolism in patients with cancer. N Engl J Med. 2003;349(2):146-153. doi: 10.1056/NEJMoa025313 [DOI] [PubMed] [Google Scholar]
- 8.Agnelli G, Becattini C, Meyer G, et al. ; Caravaggio Investigators . Apixaban for the treatment of venous thromboembolism associated with cancer. N Engl J Med. 2020;382(17):1599-1607. doi: 10.1056/NEJMoa1915103 [DOI] [PubMed] [Google Scholar]
- 9.Carrier M, Abou-Nassar K, Mallick R, et al. ; AVERT Investigators . Apixaban to prevent venous thromboembolism in patients with cancer. N Engl J Med. 2019;380(8):711-719. doi: 10.1056/NEJMoa1814468 [DOI] [PubMed] [Google Scholar]
- 10.Plummer C, Henderson RD, O’Sullivan JD, Read SJ. Ischemic stroke and transient ischemic attack after head and neck radiotherapy: a review. Stroke. 2011;42(9):2410-2418. doi: 10.1161/STROKEAHA.111.615203 [DOI] [PubMed] [Google Scholar]
- 11.Kikuno M, Ueno Y, Takekawa H, et al. ; CHALLENGE ESUS/CS Collaborators . Distinction in prevalence of atherosclerotic embolic sources in cryptogenic stroke with cancer status. J Am Heart Assoc. 2021;10(21):e021375. doi: 10.1161/JAHA.120.021375 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Merkler AE, Navi BB, Singer S, et al. Diagnostic yield of echocardiography in cancer patients with ischemic stroke. J Neurooncol. 2015;123(1):115-121. doi: 10.1007/s11060-015-1768-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Kamphuisen PW, Beyer-Westendorf J. Bleeding complications during anticoagulant treatment in patients with cancer. Thromb Res. 2014;133(suppl 2):S49-S55. doi: 10.1016/S0049-3848(14)50009-6 [DOI] [PubMed] [Google Scholar]
- 14.Kleindorfer DO, Towfighi A, Chaturvedi S, et al. 2021 Guideline for the prevention of stroke in patients with stroke and transient ischemic attack: a guideline from the American Heart Association/American Stroke Association. Stroke. 2021;52(7):e364-e467. doi: 10.1161/STR.0000000000000375 [DOI] [PubMed] [Google Scholar]
- 15.Kamel H, Longstreth WT Jr, Tirschwell DL, et al. ; ARCADIA Investigators . Apixaban to prevent recurrence after cryptogenic stroke in patients with atrial cardiopathy: the ARCADIA randomized clinical trial. JAMA. 2024;331(7):573-581. doi: 10.1001/jama.2023.27188 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Hart RG, Diener HC, Coutts SB, et al. ; Cryptogenic Stroke/ESUS International Working Group . Embolic strokes of undetermined source: the case for a new clinical construct. Lancet Neurol. 2014;13(4):429-438. doi: 10.1016/S1474-4422(13)70310-7 [DOI] [PubMed] [Google Scholar]
- 17.Kamel H, Longstreth WT Jr, Tirschwell DL, et al. The atrial cardiopathy and antithrombotic drugs in prevention after cryptogenic stroke randomized trial: rationale and methods. Int J Stroke. 2019;14(2):207-214. doi: 10.1177/1747493018799981 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Navi BB, Howard G, Howard VJ, et al. The risk of arterial thromboembolic events after cancer diagnosis. Res Pract Thromb Haemost. 2019;3(4):639-651. doi: 10.1002/rth2.12223 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Navi BB, Reiner AS, Kamel H, et al. Association between incident cancer and subsequent stroke. Ann Neurol. 2015;77(2):291-300. doi: 10.1002/ana.24325 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Ajabnoor AM, Parisi R, Zghebi SS, et al. Common cancer types and risk of stroke and bleeding in patients with nonvalvular atrial fibrillation: a population-based study in England. J Am Heart Assoc. 2023;12(19):e029423. doi: 10.1161/JAHA.123.029423 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Lun R, Siegal D, Ramsay T, Dowlatshahi D. Cancer and stroke: what do we know and where do we go? Thromb Res. 2022;219:133-140. doi: 10.1016/j.thromres.2022.09.014 [DOI] [PubMed] [Google Scholar]
- 22.Seok JM, Kim SG, Kim JW, et al. Coagulopathy and embolic signal in cancer patients with ischemic stroke. Ann Neurol. 2010;68(2):213-219. doi: 10.1002/ana.22050 [DOI] [PubMed] [Google Scholar]
- 23.Lee MJ, Chung JW, Ahn MJ, et al. Hypercoagulability and mortality of patients with stroke and active cancer: the OASIS-CANCER study. J Stroke. 2017;19(1):77-87. doi: 10.5853/jos.2016.00570 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Martinez-Majander N, Ntaios G, Liu YY, et al. ; NAVIGATE ESUS Investigators . Rivaroxaban versus aspirin for secondary prevention of ischaemic stroke in patients with cancer: a subgroup analysis of the NAVIGATE ESUS randomized trial. Eur J Neurol. 2020;27(5):841-848. doi: 10.1111/ene.14172 [DOI] [PubMed] [Google Scholar]
- 25.Navi BB, Marshall RS, Bobrow D, et al. Enoxaparin vs aspirin in patients with cancer and ischemic stroke: the TEACH pilot randomized clinical trial. JAMA Neurol. 2018;75(3):379-381. doi: 10.1001/jamaneurol.2017.4211 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Chung JW, Hwang J, Kim HJ, et al. Edoxaban for the treatment of hypercoagulability and cerebral thromboembolism associated with cancer: a randomized clinical trial of biomarker targets. Int J Stroke. Published online March 21, 2024. doi: 10.1177/17474930241239266 [DOI] [PubMed] [Google Scholar]
- 27.Kiyuna F, Sato N, Matsuo R, et al. ; Fukuoka Stroke Registry Investigators . Association of embolic sources with cause-specific functional outcomes among adults with cryptogenic stroke. JAMA Netw Open. 2018;1(5):e182953. doi: 10.1001/jamanetworkopen.2018.2953 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Jang H, Lee JJ, Lee MJ, et al. Comparison of enoxaparin and warfarin for secondary prevention of cancer-associated stroke. J Oncol. 2015;2015:502089. doi: 10.1155/2015/502089 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Kim HJ, Chung JW, Bang OY, et al. The role of factor Xa-independent pathway and anticoagulant therapies in cancer-related stroke. J Clin Med. 2021;11(1):123. doi: 10.3390/jcm11010123 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Raskob GE, van Es N, Verhamme P, et al. ; Hokusai VTE Cancer Investigators . Edoxaban for the treatment of cancer-associated venous thromboembolism. N Engl J Med. 2018;378(7):615-624. doi: 10.1056/NEJMoa1711948 [DOI] [PubMed] [Google Scholar]
- 31.Riaz IB, Fuentes H, Deng Y, et al. Comparative effectiveness of anticoagulants in patients with cancer-associated thrombosis. JAMA Netw Open. 2023;6(7):e2325283. doi: 10.1001/jamanetworkopen.2023.25283 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Trial Protocol
eFigure. ARCADIA Trial Cancer Subgroup Analysis Flow Diagram
Data Sharing Statement