[Skip to Navigation]
Sign In
Figure.  Trial Profile
Trial Profile
Table 1.  Classification of Subtypes of Recurrent Ischemic Stroke
Classification of Subtypes of Recurrent Ischemic Stroke
Table 2.  Participant Features at Baseline by Recurrent Ischemic Stroke Subtypea
Participant Features at Baseline by Recurrent Ischemic Stroke Subtypea
Table 3.  Baseline Participant Features and Functional Outcome Based on Identification of Atrial Fibrillation (AF) During Study Follow-Upa
Baseline Participant Features and Functional Outcome Based on Identification of Atrial Fibrillation (AF) During Study Follow-Upa
Table 4.  Location of Qualifying ESUS and Recurrent Ischemic Stroke
Location of Qualifying ESUS and Recurrent Ischemic Stroke
1.
Hart  RG, Pearce  LA, Aguilar  MI.  Meta-analysis: antithrombotic therapy to prevent stroke in patients who have nonvalvular atrial fibrillation.   Ann Intern Med. 2007;146(12):857-867. doi:10.7326/0003-4819-146-12-200706190-00007 PubMedGoogle ScholarCrossref
2.
Connolly  SJ, Eikelboom  J, Joyner  C,  et al; AVERROES Steering Committee and Investigators.  Apixaban in patients with atrial fibrillation.   N Engl J Med. 2011;364(9):806-817. doi:10.1056/NEJMoa1007432 PubMedGoogle ScholarCrossref
3.
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 PubMedGoogle ScholarCrossref
4.
Hart  RG, Catanese  L, Perera  KS, Ntaios  G, Connolly  SJ.  Embolic stroke of undetermined source: a systematic review and clinical update.   Stroke. 2017;48(4):867-872. doi:10.1161/STROKEAHA.116.016414 PubMedGoogle ScholarCrossref
5.
Ntaios  G, Vemmos  K, Lip  GY,  et al.  Risk stratification for recurrence and mortality in embolic stroke of undetermined source.   Stroke. 2016;47(9):2278-2285. doi:10.1161/STROKEAHA.116.013713 PubMedGoogle ScholarCrossref
6.
Hart  RG, Sharma  M, Mundl  H,  et al; NAVIGATE ESUS Investigators.  Rivaroxaban for stroke prevention after embolic stroke of undetermined source.   N Engl J Med. 2018;378(23):2191-2201. doi:10.1056/NEJMoa1802686PubMedGoogle ScholarCrossref
7.
Diener  HC, Sacco  RL, Easton  JD,  et al; RE-SPECT ESUS Steering Committee and Investigators.  Dabigatran for prevention of stroke after embolic stroke of undetermined source.   N Engl J Med. 2019;380(20):1906-1917. doi:10.1056/NEJMoa1813959PubMedGoogle ScholarCrossref
8.
Wachter  R, Freedman  B.  The role of atrial fibrillation in patients with an embolic stroke of unknown source (ESUS).   Thromb Haemost. 2017;117(10):1833-1835. doi:10.1160/TH17-08-0592 PubMedGoogle ScholarCrossref
9.
Poli  S, Bombach  P, Geisler  T.  Atrial fibrosis and its implications on a revised ESUS concept.   Neurology. 2019;93(4):141-142. doi:10.1212/WNL.0000000000007823 PubMedGoogle ScholarCrossref
10.
Kamel  H, Merkler  AE, Iadecola  C, Gupta  A, Navi  BB.  Tailoring the approach to embolic stroke of undetermined source: a review.   JAMA Neurol. 2019;76(7):855-861. doi:10.1001/jamaneurol.2019.0591 PubMedGoogle ScholarCrossref
11.
Tsivgoulis  G, Katsanos  AH, Köhrmann  M,  et al.  Embolic strokes of undetermined source: theoretical construct or useful clinical tool?   Ther Adv Neurol Disord. Published online May 24, 2019. doi:10.1177/1756286419851381PubMedGoogle Scholar
12.
Healey  JS, Gladstone  DJ, Swaminathan  B,  et al.  Recurrent stroke with rivaroxaban compared with aspirin according to predictors of atrial fibrillation: secondary analysis of the NAVIGATE ESUS randomized clinical trial.   JAMA Neurol. 2019;76(7):764-773. doi:10.1001/jamaneurol.2019.0617 PubMedGoogle ScholarCrossref
13.
Kasner  SE, Swaminathan  B, Lavados  P,  et al; NAVIGATE ESUS Investigators.  Rivaroxaban or aspirin for patent foramen ovale and embolic stroke of undetermined source: a prespecified subgroup analysis from the NAVIGATE ESUS trial.   Lancet Neurol. 2018;17(12):1053-1060. doi:10.1016/S1474-4422(18)30319-3 PubMedGoogle ScholarCrossref
14.
Hart  RG, Sharma  M, Mundl  H,  et al.  Rivaroxaban for secondary stroke prevention in patients with embolic strokes of undetermined source: design of the NAVIGATE ESUS randomized trial.   Eur Stroke J. 2016;1(3):146-154. doi:10.1177/2396987316663049 PubMedGoogle ScholarCrossref
15.
Hart  RG, Veltkamp  RC, Sheridan  P,  et al; NAVIGATE ESUS Investigators.  Predictors of recurrent ischemic stroke in patients with embolic strokes of undetermined source and effects of rivaroxaban versus aspirin according to risk status: the NAVIGATE ESUS Trial.   J Stroke Cerebrovasc Dis. 2019;28(8):2273-2279. doi:10.1016/j.jstrokecerebrovasdis.2019.05.014PubMedGoogle ScholarCrossref
16.
Ntaios  G, Papavasileiou  V, Milionis  H,  et al.  Embolic strokes of undetermined source in the Athens stroke registry: an outcome analysis.   Stroke. 2015;46(8):2087-2093. doi:10.1161/STROKEAHA.115.009334 PubMedGoogle ScholarCrossref
17.
Perera  KS, Vanassche  T, Bosch  J,  et al; ESUS Global Registry Investigators.  Embolic strokes of undetermined source: prevalence and patient features in the ESUS global registry.   Int J Stroke. 2016;11(5):526-533. doi:10.1177/1747493016641967 PubMedGoogle ScholarCrossref
18.
Portegies  ML, Selwaness  M, Hofman  A, Koudstaal  PJ, Vernooij  MW, Ikram  MA.  Left-sided strokes are more often recognized than right-sided strokes: the Rotterdam study.   Stroke. 2015;46(1):252-254. doi:10.1161/STROKEAHA.114.007385 PubMedGoogle ScholarCrossref
19.
Hedna  VS, Bodhit  AN, Ansari  S,  et al.  Hemispheric differences in ischemic stroke: is left-hemisphere stroke more common?   J Clin Neurol. 2013;9(2):97-102. doi:10.3988/jcn.2013.9.2.97 PubMedGoogle ScholarCrossref
20.
Rodríguez Hernández  SA, Kroon  AA, van Boxtel  MP,  et al.  Is there a side predilection for cerebrovascular disease?   Hypertension. 2003;42(1):56-60. doi:10.1161/01.HYP.0000077983.66161.6F PubMedGoogle ScholarCrossref
21.
Ntaios  G, Swaminathan  B, Berkowitz  SD,  et al; NAVIGATE ESUS Investigators.  Efficacy and safety of rivaroxaban versus aspirin in embolic stroke of undetermined source and carotid atherosclerosis.   Stroke. 2019;50(9):2477-2485. doi:10.1161/STROKEAHA.119.025168 PubMedGoogle ScholarCrossref
22.
Gupta  A, Gialdini  G, Lerario  MP,  et al.  Magnetic resonance angiography detection of abnormal carotid artery plaque in patients with cryptogenic stroke.   J Am Heart Assoc. 2015;4(6):e002012. doi:10.1161/JAHA.115.002012 PubMedGoogle Scholar
23.
Hyafil  F, Schindler  A, Sepp  D,  et al.  High-risk plaque features can be detected in non-stenotic carotid plaques of patients with ischaemic stroke classified as cryptogenic using combined (18)F-FDG PET/MR imaging.   Eur J Nucl Med Mol Imaging. 2016;43(2):270-279. doi:10.1007/s00259-015-3201-8 PubMedGoogle ScholarCrossref
24.
Bentzon  JF, Otsuka  F, Virmani  R, Falk  E.  Mechanisms of plaque formation and rupture.   Circ Res. 2014;114(12):1852-1866. doi:10.1161/CIRCRESAHA.114.302721 PubMedGoogle ScholarCrossref
25.
Simes  J, Robledo  KP, White  HD,  et al; LIPID Study Investigators.  D-dimer predicts long-term cause-specific mortality, cardiovascular events, and cancer in patients with stable coronary heart disease: LIPID study.   Circulation. 2018;138(7):712-723. doi:10.1161/CIRCULATIONAHA.117.029901 PubMedGoogle ScholarCrossref
26.
Goldberger  JJ, Arora  R, Green  D,  et al.  Evaluating the atrial myopathy underlying atrial fibrillation: identifying the arrhythmogenic and thrombogenic substrate.   Circulation. 2015;132(4):278-291. doi:10.1161/CIRCULATIONAHA.115.016795 PubMedGoogle ScholarCrossref
27.
Ntaios  G, Perlepe  K, Sirimarco  G,  et al.  Carotid plaques and detection of atrial fibrillation in embolic stroke of undetermined source.   Neurology. 2019;92(23):e2644-e2652. doi:10.1212/WNL.0000000000007611 PubMedGoogle ScholarCrossref
28.
Jaffre  A, Guidolin  B, Ruidavets  JB, Nasr  N, Larrue  V.  Non-obstructive carotid atherosclerosis and patent foramen ovale in young adults with cryptogenic stroke.   Eur J Neurol. 2017;24(5):663-666. doi:10.1111/ene.13275 PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

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.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Original Investigation
    July 6, 2020

    Characteristics of Recurrent Ischemic Stroke After Embolic Stroke of Undetermined Source: Secondary Analysis of a Randomized Clinical Trial

    Author Affiliations
    • 1Division of Brain Sciences, Imperial College London, London, United Kingdom
    • 2Department of Neurology, Alfried Krupp Krankenhaus, Essen, Germany
    • 3currently a biostatistics consultant, St Catharines, Ontario, Canada
    • 4Department of Clinical Therapeutics, National and Kapodistrian University of Athens, Athens, Greece
    • 5Population Health Research Institute, Department of Medicine, McMaster University, Hamilton, Ontario, Canada
    • 6Department of Neurology, University of Pennsylvania, Philadelphia
    • 7Department of Human Neurosciences, Sapienza University of Rome, Rome, Italy
    • 8Departamento de Neurología, Fleni, Buenos Aires, Argentina
    • 9Bayer AG, Wuppertal, Germany
    • 10Department of Clinical Neuroscience, Institute of Neurosciences and Physiology, Sahlgrenska Academy at University of Gothenburg, Gothenburg, Sweden
    • 11Department of Neurology, Sahlgrenska University Hospital, Gothenburg, Sweden
    • 12Faculty of Health and Medical Sciences, Medical School, University of Western Australia, Perth, Australia
    • 13Department of Clinical Sciences and Neurology, Lund University, Lund, Sweden
    • 14Department of Neurology, Skåne University Hospital, Lund, Sweden
    • 15Bayer, LLC, Whippany, New Jersey
    • 16Instituto Nacional de Neurologia y Neurocirugia Manual Velasco Suarez, Mexico City, Mexico
    • 17Department of Neurology, Faculty of Medicine, Selcuk University, Konya, Turkey
    • 18Institute of Neuroscience and Psychology, University of Glasgow, Queen Elizabeth University Hospital, Glasgow, United Kingdom
    • 19Department of Neuroscience, Hospital Clinic of Barcelona, Institute Reçerca Biomèdica August Pi i Sunyer, University of Barcelona, Barcelona, Spain
    • 20Department of Medicine, University of Alberta, Edmonton, Canada
    JAMA Neurol. 2020;77(10):1233-1240. doi:10.1001/jamaneurol.2020.1995
    Key Points

    Question  What are the characteristics and the etiology of recurrent strokes after embolic strokes of undetermined source?

    Findings  In this secondary analysis of a randomized clinical trial, recurrent ischemic stroke occurred in 309 of 7213 patients undergoing randomization. Among 270 classifiable recurrent strokes, 156 (58%) were embolic strokes of undetermined source, and 114 (42%) were not. Atrial fibrillation was found in 27 recurrent strokes (9%) and was associated with higher mortality and disability compared with other causes.

    Meaning  This trial analysis found that most stroke recurrences after embolic strokes of undetermined source were embolic and often of undetermined source; few were associated with atrial fibrillation, and these had worse outcomes.

    Abstract

    Importance  The concept of embolic stroke of undetermined source (ESUS) unifies a subgroup of cryptogenic strokes based on neuroimaging, a defined minimum set of diagnostic tests, and exclusion of certain causes. Despite an annual stroke recurrence rate of 5%, little is known about the etiology underlying recurrent stroke after ESUS.

    Objective  To identify the stroke subtype of recurrent ischemic strokes after ESUS, to explore the interaction with treatment assignment in each category, and to examine the consistency of cerebral location of qualifying ESUS and recurrent ischemic stroke.

    Design, Setting, and Participants  The NAVIGATE-ESUS trial was a randomized clinical trial conducted from December 23, 2014, to October 5, 2017. The trial compared the efficacy and safety of rivaroxaban and aspirin in patients with recent ESUS (n = 7213). Ischemic stroke was validated in 309 of the 7213 patients by adjudicators blinded to treatment assignment and classified by local investigators into the categories ESUS or non-ESUS (ie, cardioembolic, atherosclerotic, lacunar, other determined cause, or insufficient testing). Five patients with recurrent strokes that could not be defined as ischemic or hemorrhagic in absence of neuroimaging or autopsy were excluded. Data for this secondary post hoc analysis were analyzed from March to June 2019.

    Interventions  Patients were randomly assigned to receive rivaroxaban, 15 mg/d, or aspirin, 100 mg/d.

    Main Outcomes and Measures  Association of recurrent ESUS with stroke characteristics.

    Results  A total of 309 patients (205 men [66%]; mean [SD] age, 68 [10] years) had ischemic stroke identified during the median follow-up of 11 (interquartile range [IQR], 12) months (annualized rate, 4.6%). Diagnostic testing was insufficient for etiological classification in 39 patients (13%). Of 270 classifiable ischemic strokes, 156 (58%) were ESUS and 114 (42%) were non-ESUS (37 [32%] cardioembolic, 26 [23%] atherosclerotic, 35 [31%] lacunar, and 16 [14%] other determined cause). Atrial fibrillation was found in 27 patients (9%) with recurrent ischemic stroke and was associated with higher morbidity (median change in modified Rankin scale score 2 [IQR, 3] vs 0 (IQR, 1]) and mortality (15% vs 1%) than other causes. Risk of recurrence did not differ significantly by subtype between treatment groups. For both the qualifying and recurrent strokes, location of infarct was more often in the left (46% and 54%, respectively) than right hemisphere (40% and 37%, respectively) or brainstem or cerebellum (14% and 9%, respectively).

    Conclusions and Relevance  In this secondary analysis of randomized clinical trial data, most recurrent strokes after ESUS were embolic and of undetermined source. Recurrences associated with atrial fibrillation were a minority but were more often disabling and fatal. More extensive investigation to identify the embolic source is important toward an effective antithrombotic strategy.

    Trial Registration  ClinicalTrials.gov Identifier: NCT02313909

    Introduction

    Ischemic strokes are caused by a variety of mechanisms that are conventionally categorized into cardioembolic, extracranial or intracranial large artery atherosclerotic disease, lacunar (ie, small vessel disease), other defined entities, and cryptogenic origin. In most cases, secondary stroke prevention with antithrombotic drugs is performed with antiplatelet agents, but for high-risk cardioembolism, including atrial fibrillation, anticoagulants are preferred because of their superior efficacy compared with antiplatelets.1,2

    Cryptogenic stroke, representing about 20% of all strokes, has been an ill-defined category for decades. In 2014, Hart and coworkers3 proposed the concept of embolic stroke of undetermined source (ESUS). The ESUS concept unifies a large subgroup of cryptogenic strokes based on neuroimaging, a defined minimum set of diagnostic tests, and exclusion of specific causes. In several observational studies,4,5 ESUS carried a substantial annual stroke recurrence rate of 3% to 6% despite antithrombotic therapy. Beyond a clearer mechanistic characterization of strokes of unknown origin, a main purpose of the ESUS concept was to define a group of patients who might benefit from anticoagulants. However, 2 recently reported large randomized clinical trials, NAVIGATE-ESUS (New Approach Rivaroxaban Inhibition of Factor Xa in a Global Trial vs ASA to Prevent Embolism in Embolic Stroke of Undetermined Source)6 and RESPECT-ESUS (Randomized, Double-Blind, Evaluation in Secondary Stroke Prevention Comparing the Efficacy and Safety of the Oral Thrombin Inhibitor Dabigatran Etexilate vs Acetylsalicylic Acid in Patients with Embolic Stroke of Undetermined Source),7 did not find superior efficacy of direct oral anticoagulants over antiplatelets.

    Although these neutral results have led some researchers to suggest abandoning the ESUS concept,8,9 others have proposed identifying subgroups within the ESUS construct that are likely to benefit from either an anticoagulant or an antiplatelet preventive strategy.10,11 In any case, to develop a more tailored strategy for stroke prevention after ESUS, a better understanding of the characteristics and the causes of recurrent strokes after ESUS is needed. In the present exploratory analysis of the NAVIGATE-ESUS trial, we aim to describe the stroke subtype of recurrent ischemic strokes after ESUS, to explore the interaction with treatment assignment in each category, and examine the consistency of cerebral location of qualifying ESUS and recurrent ischemic stroke.

    Methods
    Patients and Procedures

    NAVIGATE-ESUS was an international, double-blinded, randomized phase 3 trial conducted at 459 centers in 31 countries and involving 7213 participants from December 23, 2014, to October 5, 2017 (trial protocol in Supplement 1). The study was approved by the relevant health authorities and the institutional review board at each trial site, and all patients provided written informed consent. This secondary post hoc analysis of the NAVIGATE-ESUS data followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline.

    The study rationale, design, participant features, and main results have been previously published.6,12-14 Briefly, patients with recent (7 days to 6 months) ischemic stroke visualized by neuroimaging were eligible if they met criteria for ESUS3 with minor modifications.14 Eligibility required that patients be 50 years or older. Participants who were aged 50 to 59 years were required to have 1 or more additional stroke risk factors consisting of stroke or transient ischemic attack before the qualifying stroke, diabetes, heart failure, hypertension, or tobacco smoking.14 Ischemic stroke was defined as a focal neurological deficit of sudden origin due to presumed arterial occlusion persisting for more than 24 hours and without evidence of primary hemorrhage on neuroimaging; if lasting fewer than 24 hours, neuroimaging evidence of brain infarction must have been present. Transthoracic echo without evidence of major cardioembolic source but not transesophageal echocardiography was required. Participants were randomly assigned to either rivaroxaban (15 mg once daily) or enteric-coated aspirin (100 mg once daily) to be taken with food. Median follow-up was 11 (interquartile range [IQR], 12) months when the trial was stopped at an interim analysis.6 Stroke outcome events observed during follow-up were verified centrally using a 2-tier process consisting of an algorithm linking reports by local physician investigators (overwhelmingly stroke neurologists) with criteria for stroke diagnosis, followed by conventional expert adjudication if all diagnostic criteria were not met. Expert adjudicators were blinded to assignment of treatment.

    For the present analyses concerning recurrent ischemic stroke during follow-up, patients with recurrent strokes that could not be defined as ischemic or hemorrhagic in the absence of relevant neuroimaging or autopsy (n = 5) were excluded. Recurrent ischemic strokes were classified by local investigators as ESUS, non-ESUS (specifically atherosclerotic, cardioembolic, lacunar, or of other defined etiology), or indeterminate (ie, mainly because insufficient diagnostic testing was performed). Based on overall clinical assessment, local investigators also determined arterial territory of the qualifying ESUS and the recurrent ischemic stroke, which were then categorized as single vs multiple territory, with single territory further categorized into left hemispheric, right hemispheric, or brainstem and/or cerebellar.

    Statistical Analysis

    Data were analyzed from March to June 2019. All analyses were performed on the intention-to-treat population. Patient characteristics were described using proportions for discrete variables and means with SDs or medians with IQRs for continuous variables. Characteristics were compared between groups using a χ2 test (or Fisher exact test if minimum expected cell count was <5) for categorical data and a unpaired t test (or Mann-Whitney test for nonparametric data) or analysis of variance (or Kruskal-Wallis for nonparametric data) for continuous variables. Time-to-event data were summarized by computing the annualized event rate (ie, the number of patients with an event divided by patient-years of exposure) with 95% CI computed assuming a Poisson distribution. Hazard ratios (HRs) and CIs from a Cox proportional hazards model were used to describe treatment effect within a group. There was no imputation of missing data. All tests were 2 sided, and statistical significance was accepted at the P < .05 level. No adjustments were made for multiple comparisons. Statistical analysis was performed using SPSS for Windows, version 24.0.0 (IBM Corp).

    Results

    The trial profile is presented in the Figure. A recurrent ischemic stroke was identified in 309 of the 7213 patients in NAVIGATE-ESUS trial (205 men [66%] and 104 women [34%]; mean [SD] age, 68 [10] years). Of these, stroke subtype was not classified for 39 patients (13%), mainly because the available diagnostic evaluation was incomplete according to the ESUS diagnostic criteria. Of the 270 classifiable recurrent ischemic strokes, 156 (58%) were reported as ESUS and 114 (42%) as non-ESUS (37 [32%] cardioembolic, 26 [23%] atherosclerotic, 35 [31%] lacunar, and 16 [14%] other determined cause). (Table 1). Annualized rates of recurrent ischemic stroke were 4.7% and 4.6% in those assigned rivaroxaban and aspirin, respectively.

    Patient characteristics at enrollment were similar among the different subtypes of recurrent ischemic stroke with the exception of current tobacco use, which ranged from 5 patients with a cardioembolic subtype (14%) to 18 patients with a lacunar subtype (51%) (Table 2). Recurrent ischemic strokes associated with cardioembolic subtype were more disabling as measured by the modified Rankin scale score (median change from baseline, 2 [IQR, 3] vs 0 [IQR, 1]) and were associated with a higher mortality rate than other stroke subtypes including ESUS (11% vs 1%) (Table 2).

    Because identification of atrial fibrillation based on patient characteristics would have a major effect on preventive antithrombotic therapy, we compared the features of participants with recurrent ischemic stroke who were diagnosed with atrial fibrillation during study follow-up (27 of 309 [9%]) vs those without AF (282 of 309 [91%]) (Table 3). Differences between these groups beyond the severity of the stroke were limited to a larger atrial diameter in patients with atrial fibrillation (mean [SD], 4.1 [0.7] vs 3.7 [0.8] cm) and proportionally fewer patients assigned to rivaroxaban among those with recurrent strokes who were diagnosed with atrial fibrillation after randomization (8 [30%] vs 148 [52%]).

    The effect of assigned treatment with rivaroxaban vs aspirin for prevention of recurrent ischemic stroke by stroke subtype is shown in eTable 1 in Supplement 2. Participants assigned to rivaroxaban vs aspirin tended to less frequently have a cardioembolic stroke (HR, 0.54; 95% CI, 0.28-1.1). In contrast, participants not assigned to aspirin tended to have higher risk for atherosclerotic (HR, 1.9; 95% CI, 0.84-4.2) and lacunar (HR, 1.5; 95% CI, 0.76-2.9) stroke, although none of these differences were statistically significant.

    Territories of the qualifying ESUS as well as the recurrent ischemic stroke were classified as left hemispheric, right hemispheric, brainstem and/or cerebellar, or multiple infarcts. Single-territory qualifying infarcts (n = 6445) and single-territory recurrent ischemic strokes (n = 260) were more often located in the left (2983 [46%] and 141 [54%], respectively) than in the right hemisphere (2570 [40%] and 95 [37%], respectively) or brainstem/cerebellum (892 [14%] and 24 [9%], respectively) (Table 4). Patients with a recurrent ischemic stroke also had a qualifying ESUS in the left hemisphere more often than those who did not have a recurrent stroke (141 of 260 [54%] vs 2842 of 6185 [46%]). Of those with single-territory qualifying ESUS and recurrent stroke, 61% (95% CI, 54%-68%) of recurrent ischemic stroke (eTable 2 in Supplement 2) and 62% (95% CI, 53%-71%) of recurrent ESUS (eTable 3 in Supplement 2) occurred in the same territory as the qualifying ESUS. Multiple territories were observed in 763 of 7208 (11%) qualifying ESUS and in 24 of 156 (15%) recurrent strokes classified as ESUS (Table 4). Multiple-territory qualifying ESUS occurred in 49 of 309 patients with a recurrent stroke (16%) and 714 of 6899 qualifying ESUS without recurrence (10%) (eTable 4 in Supplement 2).

    Discussion

    These exploratory analyses of the NAVIGATE-ESUS trial yield 3 major new findings. First, most of the recurrent ischemic strokes after ESUS met the criteria for ESUS. No source of embolus was identified in about three-quarters of cases, despite repeated diagnostic workup at the time of recurrent stroke. Second, analysis of location of recurrent ischemic strokes underscores the coexistence of several embolic mechanisms for recurrent strokes. Third, recurrences were rarely associated with atrial fibrillation, but atrial fibrillation–related recurrent ischemic strokes had particularly grave consequences and were prevented by rivaroxaban better than by aspirin.

    The predominance of embolic features in most of the recurrent strokes after ESUS found in our analysis supports the validity of the ESUS construct in terms of embolism as the pathogenic mechanism. However, even with the additional knowledge of recurrent stroke subtype in the present analysis, it is not possible to identify patient characteristics that are associated with the mechanism of recurrence at the time of the qualifying ESUS. A previous analysis of NAVIGATE-ESUS data15 reported prior stroke or transient ischemic attack, current tobacco use, higher age, diabetes, multiple acute infarcts on neuroimaging, and aspirin use before the qualifying stroke as independently predictive of all recurrent stroke, but covariates associated with specific stroke subtypes were not investigated. The persisting uncertainty regarding the embolic source after stroke recurrence in a large proportion of our patients with ESUS suggests that the search for an underlying source may remain futile in many cases with ESUS, given the limitations of widely used diagnostic testing at present. Consequently, addressing a specific embolic source by targeted antithrombotic stroke prevention remains an unresolved dilemma for many patients with ESUS.

    The NAVIGATE-ESUS trial did not show an overall benefit for stroke prevention with rivaroxaban vs aspirin. Subsequent exploratory subgroup analyses suggested that patients with ESUS and a patent foramen ovale or with a dilated left atrium at baseline were less likely to have a stroke when allocated to rivaroxaban, but independent confirmation of these findings is needed.12,13 Based on the classification of the recurrent event taken in the present analysis, only statistically nonsignificant trends for better prevention of cardioembolic events with rivaroxaban and for more effective prevention of atherosclerotic and lacunar strokes by aspirin, respectively, were observed. Corresponding to the analysis by Healey and colleagues,12 atrial diameter was larger in patients who experienced recurrent ischemic stroke associated with atrial fibrillation than those without atrial fibrillation in our analysis.

    When the ESUS construct was originally proposed, common thinking was that covert atrial fibrillation may frequently underlie ESUS and that anticoagulation may be beneficial for stroke prevention after ESUS because of this association. In the meantime, several observational studies have shown that the characteristics of ESUS differ from those of stroke in patients with atrial fibrillation. On average, patients with ESUS are younger and have lower baseline stroke severity, lower burden of cardiovascular risk factors, and lower mortality compared with patients with cardioembolism.4,5,16,17 Moreover, the lack of success of anticoagulation in the NAVIGATE-ESUS6 and RESPECT-ESUS7 trials contradicts the assumption of a major quantitative role of atrial fibrillation–related cardioembolism in ESUS. However, an important finding of our analysis is that recurrent ischemic strokes after ESUS in patients with a first diagnosis of atrial fibrillation during follow-up in the NAVIGATE-ESUS trial had much more severe consequences in terms of disability and mortality than other recurrent stroke types. Therefore, although covert atrial fibrillation may not underlie most recurrent strokes after ESUS, its particularly grave consequences warrant a more extensive search for atrial fibrillation in patients with ESUS than mandated in the originally proposed criteria for ESUS.

    Our finding that qualifying ESUS and recurrent ischemic stroke were more frequently located in the left than in the right hemisphere is probably the result of more sensitive recognition of left hemispheric symptoms.18,19 Less likely, there may be a predilection of cerebral blood flow carrying emboli from the heart to the brain via the left than via the right carotid territory.20 Interestingly, after excluding recurrences in multiple locations, 62% (95% CI, 53%-71%) of recurrent ESUS in our study occurred in the same location as the qualifying stroke. This was significantly more frequent than expected if the recurrent stroke emboli had originated from a cardiac source where random distribution of emboli follow cerebral blood flow distribution (ie, a ratio of left carotid vs right carotid vs vertebrobasilar of 40:40:20). In that case, the theoretically expected consistency of location following from the observed distribution of the qualifying ESUS would be 38% (eResults in Supplement 2). Instead, recurrent strokes in some cases may have originated from the same vascular, originally nonstenotic culprit lesion in a carotid or vertebrobasilar artery in qualifying and recurrent strokes.21 One possibility is that lesion progression to a hemodynamically relevant stenosis may have occurred, although this is not likely in most cases during a median follow-up of 11 months.22 Alternatively, even without progression of the degree of stenosis, a nonstenotic plaque may cause recurrent stroke.23,24 Although the degree of stenosis can be easily defined by conventional luminal imaging, it seems to have limited predictive value compared with alternative markers of plaque instability, such as ulceration, intraplaque hemorrhage, vessel wall inflammation, and microemboli signal detection, that may identify a culprit lesion. In the setting of ESUS, thrombogenesis triggered by other ESUS-related constellations, such as cancer or atrial cardiomyopathy, may activate a vulnerable plaque.25,26 However, although a consistency of location in 62% of ESUS recurrent strokes differs from the expected distribution in case of a cardiac source, it does not support an overwhelming pathogenetic role of atherosclerosis in stroke recurrence but rather underscores the competition and interaction of various potential embolic sources in patients with ESUS. Some evidence suggests that different sources of emboli may have a negative association serving as competitors.27,28 In a recent study,27 atrial fibrillation was less frequently detected in patients with ESUS and nonstenotic carotid plaques compared with those without ESUS and such plaques. Another study28 showed a negative association between patent foramen ovale and nonstenotic lesions in young patients with cryptogenic stroke.

    Strengths and Limitations

    Our present analysis has limitations and strengths. First, this was an exploratory analysis of the large, randomized NAVIGATE-ESUS trial, which was halted early and yielded neutral results. Moreover, the classification of subtypes of recurrent ischemic events and the location of infarcts on brain imaging relied on information provided by site investigators and was based on routine assessment rather than a defined set of diagnostic tests based on specific guidance. For example, a more extensive search for atrial fibrillation may have yielded a larger prevalence of the arrhythmia than reported in this paper. On the other hand, all incident strokes during the trial were adjudicated by a panel of experts who were blinded to treatment allocation. Another limitation is that 13% of stroke recurrences were unclassifiable because of insufficient diagnostic evaluation, and location of recurrent infarcts was not reported in 11% of all recurrent strokes. Finally, the number of ischemic strokes for each of the non-ESUS stroke subtypes was somewhat small, which limited our ability to detect differences in patient characteristics and effect of treatment.

    Conclusions

    This secondary analysis found that most of the recurrent ischemic strokes after ESUS in the NAVIGATE-ESUS trial were embolic in nature and frequently of undetermined source. This finding emphasizes the need for a more extensive search for specific sources of embolism in individual patients or, alternatively, the establishment of another unifying strategy for antithrombotic therapy addressing different pathways of embolus formation.

    Back to top
    Article Information

    Accepted for Publication: February 28, 2020.

    Published Online: July 6, 2020. doi:10.1001/jamaneurol.2020.1995

    Correction: This article was corrected on September 8, 2020, to remove the middle initial “S” from author Danilo Toni’s name in the byline.

    Corresponding Author: Roland Veltkamp, MD, Department of Neurology, Alfried Krupp Krankenhaus, Alfried-Krupp Str 21, 45131 Essen, Germany (r.veltkamp@imperial.ac.uk).

    Author Contributions: Drs Veltkamp and Pearce 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: Veltkamp, Sharma, Kasner, Ameriso, Mundl, Tatlisumak, Hankey, Lindgren, Perera, Shoamanesh, Connolly, Hart.

    Acquisition, analysis, or interpretation of data: Veltkamp, Pearce, Korompoki, Sharma, Kasner, Toni, Ameriso, Mundl, Tatlisumak, Lindgren, Berkowitz, Arauz, Ozturk, Muir, Chamorro, Shuaib, Rudilosso, Shoamanesh, Connolly, Hart.

    Drafting of the manuscript: Veltkamp, Korompoki, Ozturk.

    Critical revision of the manuscript for important intellectual content: All authors.

    Statistical analysis: Pearce.

    Obtained funding: Berkowitz, Connolly, Hart.

    Administrative, technical, or material support: Korompoki, Tatlisumak, Berkowitz, Ozturk, Muir, Shoamanesh, Connolly, Hart.

    Supervision: Veltkamp, Ameriso, Lindgren, Berkowitz, Connolly, Hart.

    Conflict of Interest Disclosures: Dr Veltkamp reported receiving fees for consulting and speaker honoraria from Bayer AG, Boehringer Ingelheim, Bristol-Myers Squibb, Pfizer, Inc, Daiichi Sankyo Company, Limited, Portola Pharmaceuticals, Biogen, Inc, Medtronic plc, MorphoSys AG, Amgen, Inc, and Javelin Biotech and research support from Bayer AG, Boehringer Ingelheim, Bristol-Myers Squibb/Pfizer, Inc, Daiichi Sankyo Company, Limited, Medtronic plc, and Biogen Inc outside of present work; and serving as an investigator of Imperial Biomedical Research Center and partially funded by the European Union’s Horizon 2020 research and innovation programme under grant agreement 754517 (PRESTIGE-AF). Dr Korompoki reported receiving speaker fees, serving on the advisory board, and receiving travel grants from Amgen, Inc, Bayer AG, and Bristol-Myers Squibb and Pfizer, Inc, outside the submitted work. Dr Sharma reported receiving research support from Bayer AG, Bristol-Myers Squibb, and Boehringer Ingelheim and serving on the advisory board and/or receiving speaker’s honoraria from Bayer AG, Boehringer Ingelheim, Bristol-Myers Squibb, and Daiichi Sankyo Company, Limited. Dr Kasner reported receiving research support from Bayer AG, Bristol-Myers Squibb, WL Gore and Associates, and Medtronic plc and consulting for Bristol-Myers Squibb, Boehringer Ingelheim, and Portola Pharmaceuticals. Dr Toni reported receiving speaker fees or serving on the advisory board of Abbott Laboratories, Boehringer Ingelheim, Bayer AG, Daiichi Sankyo Company, Limited, Medtronic plc, and Pfizer, Inc. Dr Ameriso reported receiving speaker fees, serving on steering committees, and receiving travel grants and clinical research honorarium from Bayer AG, Boehringer Ingelheim, and Bristol-Myers Squibb/Pfizer, Inc. Dr Tatlisumak reported serving on the advisory board of Bayer AG, Boehringer Ingelheim, Brainsgate, Bristol-Myers Squibb, Lumosa Pharma, Inc, and Portola Pharmaceuticals and receiving research contracts with Bayer AG, Boehringer Ingelheim, Brainsgate, Pfizer, Inc, Portola Pharmaceuticals, and Sanofi Aventis. Dr Hankey reported receiving personal honoraria outside the submitted work from Bayer AG, Bristol-Myers Squibb, Medscape, and the American Heart Association. Dr Lindgren reported receiving grants from the Swedish Heart and Lung Foundation, Region Skåne, Skåne University Hospital, Freemasons Lodge of Instruction Eos in Lund, Lund University, and the Foundation of Färs & Frosta (one of Sparbanken Skåne’s ownership foundations) and personal fees from Bayer AG, AstraZeneca plc, Bristol-Myers Squibb/Pfizer, Inc, and Portola Pharmaceuticals outside the submitted work. Dr Arauz reported receiving speaker fees, serving on steering committees, and receiving travel grants from Bayer AG, Boehringer Ingelheim, and Pfizer, Inc. Dr Muir reported receiving honoraria outside the submitted work from Bayer AG, Daiichi Sankyo Company, Limited, Boehringer Ingelheim and research support from Boehringer Ingelheim. Dr Perera reported receiving speaker fees and research grants and serving on advisory boards from Bayer AG and speaker fees from Abbott Laboratories. Dr Shoamanesh reported receiving research support and honoraria from Bayer AG. Dr Connolly reported receiving research funding and consulting honoraria from Bayer AG, Bristol-Myers Squibb, Daiichi Sankyo Company, Limited, Portola Pharmaceuticals, Javelin Biotech, and Boehringer Ingelheim. Dr Hart reported receiving a stipend for research and honoraria for advisory board participation from Bayer AG. No other disclosures were reported.

    Funding/Support: This study was supported by Bayer AG and Janssen Research and Development (NAVIGATE-ESUS trial).

    Role of the Funder/Sponsor: The sponsors 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 the manuscript for publication.

    Additional Contributions: We thank all NAVIGATE-ESUS investigators and all patients participating in the trial.

    References
    1.
    Hart  RG, Pearce  LA, Aguilar  MI.  Meta-analysis: antithrombotic therapy to prevent stroke in patients who have nonvalvular atrial fibrillation.   Ann Intern Med. 2007;146(12):857-867. doi:10.7326/0003-4819-146-12-200706190-00007 PubMedGoogle ScholarCrossref
    2.
    Connolly  SJ, Eikelboom  J, Joyner  C,  et al; AVERROES Steering Committee and Investigators.  Apixaban in patients with atrial fibrillation.   N Engl J Med. 2011;364(9):806-817. doi:10.1056/NEJMoa1007432 PubMedGoogle ScholarCrossref
    3.
    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 PubMedGoogle ScholarCrossref
    4.
    Hart  RG, Catanese  L, Perera  KS, Ntaios  G, Connolly  SJ.  Embolic stroke of undetermined source: a systematic review and clinical update.   Stroke. 2017;48(4):867-872. doi:10.1161/STROKEAHA.116.016414 PubMedGoogle ScholarCrossref
    5.
    Ntaios  G, Vemmos  K, Lip  GY,  et al.  Risk stratification for recurrence and mortality in embolic stroke of undetermined source.   Stroke. 2016;47(9):2278-2285. doi:10.1161/STROKEAHA.116.013713 PubMedGoogle ScholarCrossref
    6.
    Hart  RG, Sharma  M, Mundl  H,  et al; NAVIGATE ESUS Investigators.  Rivaroxaban for stroke prevention after embolic stroke of undetermined source.   N Engl J Med. 2018;378(23):2191-2201. doi:10.1056/NEJMoa1802686PubMedGoogle ScholarCrossref
    7.
    Diener  HC, Sacco  RL, Easton  JD,  et al; RE-SPECT ESUS Steering Committee and Investigators.  Dabigatran for prevention of stroke after embolic stroke of undetermined source.   N Engl J Med. 2019;380(20):1906-1917. doi:10.1056/NEJMoa1813959PubMedGoogle ScholarCrossref
    8.
    Wachter  R, Freedman  B.  The role of atrial fibrillation in patients with an embolic stroke of unknown source (ESUS).   Thromb Haemost. 2017;117(10):1833-1835. doi:10.1160/TH17-08-0592 PubMedGoogle ScholarCrossref
    9.
    Poli  S, Bombach  P, Geisler  T.  Atrial fibrosis and its implications on a revised ESUS concept.   Neurology. 2019;93(4):141-142. doi:10.1212/WNL.0000000000007823 PubMedGoogle ScholarCrossref
    10.
    Kamel  H, Merkler  AE, Iadecola  C, Gupta  A, Navi  BB.  Tailoring the approach to embolic stroke of undetermined source: a review.   JAMA Neurol. 2019;76(7):855-861. doi:10.1001/jamaneurol.2019.0591 PubMedGoogle ScholarCrossref
    11.
    Tsivgoulis  G, Katsanos  AH, Köhrmann  M,  et al.  Embolic strokes of undetermined source: theoretical construct or useful clinical tool?   Ther Adv Neurol Disord. Published online May 24, 2019. doi:10.1177/1756286419851381PubMedGoogle Scholar
    12.
    Healey  JS, Gladstone  DJ, Swaminathan  B,  et al.  Recurrent stroke with rivaroxaban compared with aspirin according to predictors of atrial fibrillation: secondary analysis of the NAVIGATE ESUS randomized clinical trial.   JAMA Neurol. 2019;76(7):764-773. doi:10.1001/jamaneurol.2019.0617 PubMedGoogle ScholarCrossref
    13.
    Kasner  SE, Swaminathan  B, Lavados  P,  et al; NAVIGATE ESUS Investigators.  Rivaroxaban or aspirin for patent foramen ovale and embolic stroke of undetermined source: a prespecified subgroup analysis from the NAVIGATE ESUS trial.   Lancet Neurol. 2018;17(12):1053-1060. doi:10.1016/S1474-4422(18)30319-3 PubMedGoogle ScholarCrossref
    14.
    Hart  RG, Sharma  M, Mundl  H,  et al.  Rivaroxaban for secondary stroke prevention in patients with embolic strokes of undetermined source: design of the NAVIGATE ESUS randomized trial.   Eur Stroke J. 2016;1(3):146-154. doi:10.1177/2396987316663049 PubMedGoogle ScholarCrossref
    15.
    Hart  RG, Veltkamp  RC, Sheridan  P,  et al; NAVIGATE ESUS Investigators.  Predictors of recurrent ischemic stroke in patients with embolic strokes of undetermined source and effects of rivaroxaban versus aspirin according to risk status: the NAVIGATE ESUS Trial.   J Stroke Cerebrovasc Dis. 2019;28(8):2273-2279. doi:10.1016/j.jstrokecerebrovasdis.2019.05.014PubMedGoogle ScholarCrossref
    16.
    Ntaios  G, Papavasileiou  V, Milionis  H,  et al.  Embolic strokes of undetermined source in the Athens stroke registry: an outcome analysis.   Stroke. 2015;46(8):2087-2093. doi:10.1161/STROKEAHA.115.009334 PubMedGoogle ScholarCrossref
    17.
    Perera  KS, Vanassche  T, Bosch  J,  et al; ESUS Global Registry Investigators.  Embolic strokes of undetermined source: prevalence and patient features in the ESUS global registry.   Int J Stroke. 2016;11(5):526-533. doi:10.1177/1747493016641967 PubMedGoogle ScholarCrossref
    18.
    Portegies  ML, Selwaness  M, Hofman  A, Koudstaal  PJ, Vernooij  MW, Ikram  MA.  Left-sided strokes are more often recognized than right-sided strokes: the Rotterdam study.   Stroke. 2015;46(1):252-254. doi:10.1161/STROKEAHA.114.007385 PubMedGoogle ScholarCrossref
    19.
    Hedna  VS, Bodhit  AN, Ansari  S,  et al.  Hemispheric differences in ischemic stroke: is left-hemisphere stroke more common?   J Clin Neurol. 2013;9(2):97-102. doi:10.3988/jcn.2013.9.2.97 PubMedGoogle ScholarCrossref
    20.
    Rodríguez Hernández  SA, Kroon  AA, van Boxtel  MP,  et al.  Is there a side predilection for cerebrovascular disease?   Hypertension. 2003;42(1):56-60. doi:10.1161/01.HYP.0000077983.66161.6F PubMedGoogle ScholarCrossref
    21.
    Ntaios  G, Swaminathan  B, Berkowitz  SD,  et al; NAVIGATE ESUS Investigators.  Efficacy and safety of rivaroxaban versus aspirin in embolic stroke of undetermined source and carotid atherosclerosis.   Stroke. 2019;50(9):2477-2485. doi:10.1161/STROKEAHA.119.025168 PubMedGoogle ScholarCrossref
    22.
    Gupta  A, Gialdini  G, Lerario  MP,  et al.  Magnetic resonance angiography detection of abnormal carotid artery plaque in patients with cryptogenic stroke.   J Am Heart Assoc. 2015;4(6):e002012. doi:10.1161/JAHA.115.002012 PubMedGoogle Scholar
    23.
    Hyafil  F, Schindler  A, Sepp  D,  et al.  High-risk plaque features can be detected in non-stenotic carotid plaques of patients with ischaemic stroke classified as cryptogenic using combined (18)F-FDG PET/MR imaging.   Eur J Nucl Med Mol Imaging. 2016;43(2):270-279. doi:10.1007/s00259-015-3201-8 PubMedGoogle ScholarCrossref
    24.
    Bentzon  JF, Otsuka  F, Virmani  R, Falk  E.  Mechanisms of plaque formation and rupture.   Circ Res. 2014;114(12):1852-1866. doi:10.1161/CIRCRESAHA.114.302721 PubMedGoogle ScholarCrossref
    25.
    Simes  J, Robledo  KP, White  HD,  et al; LIPID Study Investigators.  D-dimer predicts long-term cause-specific mortality, cardiovascular events, and cancer in patients with stable coronary heart disease: LIPID study.   Circulation. 2018;138(7):712-723. doi:10.1161/CIRCULATIONAHA.117.029901 PubMedGoogle ScholarCrossref
    26.
    Goldberger  JJ, Arora  R, Green  D,  et al.  Evaluating the atrial myopathy underlying atrial fibrillation: identifying the arrhythmogenic and thrombogenic substrate.   Circulation. 2015;132(4):278-291. doi:10.1161/CIRCULATIONAHA.115.016795 PubMedGoogle ScholarCrossref
    27.
    Ntaios  G, Perlepe  K, Sirimarco  G,  et al.  Carotid plaques and detection of atrial fibrillation in embolic stroke of undetermined source.   Neurology. 2019;92(23):e2644-e2652. doi:10.1212/WNL.0000000000007611 PubMedGoogle ScholarCrossref
    28.
    Jaffre  A, Guidolin  B, Ruidavets  JB, Nasr  N, Larrue  V.  Non-obstructive carotid atherosclerosis and patent foramen ovale in young adults with cryptogenic stroke.   Eur J Neurol. 2017;24(5):663-666. doi:10.1111/ene.13275 PubMedGoogle ScholarCrossref
    ×