Timing of anticoagulation after emergency endovascular therapy for acute ischaemic stroke with atrial fibrillation (TIMERS): protocol for a randomised controlled trial Authors: Yibing Guo, Xiaole Jia, Wanqing Liu, Wanwan Zhang, Xiao Dong, Ran Meng, Hongrui Ma, Guiyun Cui, Yanbo Cheng, Tingyu Yi, Jiheng Hao, Yufeng Jiang, Xunming Ji, Yu Feng, Chuanjie Wu Published: 2026 ABSTRACT AbstractIntroductionThe optimal timing for direct oral anticoagulants (DOACs) initiation after acute ischaemic stroke in patients with atrial fibrillation (AF) remains unclear, particularly in those undergoing endovascular thrombectomy (EVT), who are at higher risk of both haemorrhagic transformation and early embolic recurrence. Patients undergoing thrombectomy have been under-represented in previous randomised trials, resulting in limited evidence for guiding management in this high-risk population. The TIMERS trial aims to evaluate the efficacy and safety of early versus delayed initiation of DOACs therapy after EVT in patients with large vessel occlusion stroke and AF.Methods and analysisThis is a multicentre, prospective, randomised, open-label trial with a blinded outcome assessment. In total, 438 patients with acute ischaemic stroke and AF treated with EVT will be enrolled in this study. Patients will be randomised in a 1:1 ratio within 72 hours after stroke onset to early initiation (≤4 days) or delayed initiation (5–14 days). The primary outcome is a composite of recurrent ischaemic stroke, symptomatic intracranial haemorrhage or all-cause death within 90 days. Secondary outcomes include individual components of the primary endpoint, venous thromboembolism, systemic embolism, myocardial infarction, functional outcome (modified Rankin Scale), health-related quality of life (EuroQol 5-Dimension 5-Level questionnaire (EQ-5D-5L)) and major extracranial bleeding events.Ethics and disseminationThe study is approved by the Ethics Committee of Xuanwu Hospital of Capital Medical University, the ethics approval document number is (2025)259-001. The trial findings will be disseminated through peer-reviewed publications and presentations at scientific conferences.Trial registration numberNCT07139301. STRENGTHS AND LIMITATIONS OF THIS STUDY The multicentre design increases the representativeness of the study population across participating centres and supports the external validity of the trial. The multicentre design increases the representativeness of the study population across participating centres and supports the external validity of the trial. Centralised randomisation with blinded outcome assessment enhances internal validity and minimises outcome assessment bias. Centralised randomisation with blinded outcome assessment enhances internal validity and minimises outcome assessment bias. Blinding of treating clinicians and participants is not feasible due to the nature of the intervention, which may introduce performance bias. Blinding of treating clinicians and participants is not feasible due to the nature of the intervention, which may introduce performance bias. Introduction Atrial fibrillation (AF) is a leading cause of ischaemic stroke. It is associated with a substantially increased risk of recurrence and unfavourable functional outcome.1 2 Oral anticoagulation is the cornerstone of secondary prevention in patients with AF-related stroke. This intervention reduces the risk of recurrence by approximately two-thirds.3 4 However, the optimal timing of anticoagulation initiation during the acute phase remains unclear. A substantial proportion of AF-related strokes are caused by large vessel occlusion, which accounts for approximately 15–30% of all ischaemic strokes.5 Endovascular thrombectomy (EVT) has become the standard of care for this patient population.6 Individuals undergoing EVT typically present with more severe neurological deficits and higher baseline National Institutes of Health Stroke Scale (NIHSS) scores.7 In addition, reperfusion after EVT may cause blood-brain barrier disruption, ischaemia-reperfusion injury and procedure-related vascular damage. This may substantially increase the risk of intracranial haemorrhage, including haemorrhagic transformation and parenchymal haematoma.8 9 Consequently, the haemorrhagic risk after EVT may differ from that of patients who do not undergo reperfusion therapy. Thus, determining the optimal timing of anticoagulation after EVT represents a complex clinical challenge. Current guideline recommendations for the timing of anticoagulation therapy initiation after AF-related stroke are based on observational studies and expert consensus.10–12 Several recent randomised controlled trials, including TIMING, ELAN, OPTIMAS and START, have suggested that the early initiation of direct oral anticoagulants (DOACs) after acute ischaemic stroke may be feasible. However, these trials primarily included patients who did not undergo EVT, which limits the generalisability of their findings to patients treated with thrombectomy.13–16 Furthermore, evidence regarding the optimal timing of anticoagulation therapy in EVT-treated patients remains inconsistent. Early initiation of anticoagulation therapy was associated with fewer outcome events in some reperfusion subgroup analyses of randomised trials.17 In contrast, initiation within 4 days was associated with a higher rate of the composite outcome of recurrent ischaemic stroke, symptomatic intracranial haemorrhage or death than later initiation in a reperfusion subgroup analysis from the TIMING trial.13 Similarly, a retrospective cohort study conducted by our group demonstrated that the initiation of anticoagulation therapy within 4 days after EVT was associated with an increased risk of intracranial haemorrhage. In contrast, initiation between 5 and 14 days was associated with a more favourable balance between ischaemic and haemorrhagic risks.18 These inconsistent findings underscore the limitations of existing study designs and the potential effect of selection bias. Therefore, a dedicated randomised controlled trial is warranted. In this study, we aim to compare the safety and efficacy of early and delayed DOACs initiation after EVT in patients with AF and large vessel occlusion stroke. We further aim to provide sufficient evidence to inform the establishment of anticoagulation strategies for this high-risk population. Methods Design TIMERS is a multicentre, prospective, randomised, parallel, open-label trial with a blinded endpoint assessment conducted in China. The flowchart of this study is shown in figure 1. The trial has been registered at ClinicalTrials.gov (NCT07139301; 17 August 2025). Ethical approval will be obtained from all participating centres, and written informed consent will be obtained from all patients or their legally authorised representatives before enrolment (see online supplemental file). Any substantial amendments to the study protocol will require review and approval by the relevant ethics committees before implementation. Study flowchart. AE, adverse event; AF, atrial fibrillation; DOAC, direct oral anticoagulant; EQ-5D-5L, EuroQol 5-Dimension 5-Level questionnaire; EVT, endovascular thrombectomy; mRS, modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale; SAE, serious adverse event. Patient population TIMERS will enrol 438 patients within 72 hours of stroke onset across multiple stroke centres in China. Inclusion criteria Aged ≥18 years. Aged ≥18 years. Clinical diagnosis of acute ischaemic stroke related to large vessel occlusion. Clinical diagnosis of acute ischaemic stroke related to large vessel occlusion. Emergency EVT performed within 24 hours of stroke onset. Emergency EVT performed within 24 hours of stroke onset. AF (including paroxysmal, persistent or permanent AF), confirmed by at least one of the following: (1) 12-lead ECG recording; (2) inpatient ECG telemetry; (3) prolonged ECG monitoring (eg, Holter monitor); and (4) previously established diagnosis of AF verified by medical records. AF (including paroxysmal, persistent or permanent AF), confirmed by at least one of the following: (1) 12-lead ECG recording; (2) inpatient ECG telemetry; (3) prolonged ECG monitoring (eg, Holter monitor); and (4) previously established diagnosis of AF verified by medical records. CT or MRI demonstrating one of the following findings: (1) no haemorrhagic transformation; (2) haemorrhagic infarction type 1, defined as small petechiae along the margins of the infarct (Heidelberg classification); and (3) haemorrhagic infarction type 2, defined as confluent petechiae within the infarcted area without a space-occupying effect (Heidelberg classification). CT or MRI demonstrating one of the following findings: (1) no haemorrhagic transformation; (2) haemorrhagic infarction type 1, defined as small petechiae along the margins of the infarct (Heidelberg classification); and (3) haemorrhagic infarction type 2, defined as confluent petechiae within the infarcted area without a space-occupying effect (Heidelberg classification). Randomisation within 72 hours of stroke onset. Randomisation within 72 hours of stroke onset. Written informed consent obtained from the patient or a legally authorised representative. Written informed consent obtained from the patient or a legally authorised representative. Exclusion criteria AF due to reversible causes (eg, thyrotoxicosis, pericarditis, recent surgery or acute myocardial infarction). AF due to reversible causes (eg, thyrotoxicosis, pericarditis, recent surgery or acute myocardial infarction). Contraindication to the use of DOACs: (1) known allergy or intolerance to both factor Xa inhibitors and direct thrombin inhibitors; (2) definite indication for vitamin K antagonist (VKA) treatment (eg, mechanical heart valve and valvular AF); (3) severe renal impairment (defined as creatinine >1.5 times the upper limit of normal) and significant hepatic dysfunction (defined as alanine aminotransferase or aspartate aminotransferase >2 times the upper limit of normal); (4) concomitant use of medications with significant interactions with DOACs, including azole antifungals, HIV protease inhibitors or strong CYP3A4 inducers; (5) baseline platelet count <100×109/L; and (6) history of coagulopathy or systemic haemorrhage. Contraindication to the use of DOACs: (1) known allergy or intolerance to both factor Xa inhibitors and direct thrombin inhibitors; (2) definite indication for vitamin K antagonist (VKA) treatment (eg, mechanical heart valve and valvular AF); (3) severe renal impairment (defined as creatinine >1.5 times the upper limit of normal) and significant hepatic dysfunction (defined as alanine aminotransferase or aspartate aminotransferase >2 times the upper limit of normal); (4) concomitant use of medications with significant interactions with DOACs, including azole antifungals, HIV protease inhibitors or strong CYP3A4 inducers; (5) baseline platelet count <100×109/L; and (6) history of coagulopathy or systemic haemorrhage. Use of DOACs within 48 hours before stroke onset, or recent treatment with VKA leading to an International Normalised Ratio ≥1.7 at the time of randomisation. Use of DOACs within 48 hours before stroke onset, or recent treatment with VKA leading to an International Normalised Ratio ≥1.7 at the time of randomisation. Pregnant, breastfeeding or a positive pregnancy test at admission. Pregnant, breastfeeding or a positive pregnancy test at admission. History of major surgery or severe trauma within 1 month before stroke onset. History of major surgery or severe trauma within 1 month before stroke onset. History of active bleeding within 1 month before stroke onset (eg, gastrointestinal bleeding and urinary tract bleeding). History of active bleeding within 1 month before stroke onset (eg, gastrointestinal bleeding and urinary tract bleeding). Dual antiplatelet therapy at baseline or high likelihood of requiring dual antiplatelet therapy during the trial. Dual antiplatelet therapy at baseline or high likelihood of requiring dual antiplatelet therapy during the trial. Evidence of cerebral amyloid angiopathy. Evidence of cerebral amyloid angiopathy. CT or MRI evidence of non-stroke pathology that is likely to account for the presenting clinical symptoms (eg, mass lesions and encephalitis). CT or MRI evidence of non-stroke pathology that is likely to account for the presenting clinical symptoms (eg, mass lesions and encephalitis). Modified Rankin scale (mRS) score >1 before stroke onset. Modified Rankin scale (mRS) score >1 before stroke onset. Inability to complete the 90-day follow-up. Inability to complete the 90-day follow-up. Current participation in another drug clinical trial. Current participation in another drug clinical trial. Any other reason deemed by the investigator to make the patient unsuitable for participation in the trial. Any other reason deemed by the investigator to make the patient unsuitable for participation in the trial. Randomisation Eligible patients will be randomised within 72 hours of stroke onset using a centralised, independent web-based randomisation system. Randomisation will be performed by the enrolling investigator at each participating centre in a 1:1 ratio for the early and delayed anticoagulation groups. The allocation sequence uses a minimisation algorithm to ensure a balanced distribution between groups with regard to the NIHSS score at randomisation (0–10 or ≥10), sex, age, study centre and presence of haemorrhagic transformation. Both the patients and treating physicians are aware of the assigned treatments. All outcome assessments will be performed by independent evaluators blinded to the treatment assignment. Intervention DOAC therapy will be initiated within 4 days of stroke onset in the early anticoagulation group. In the delayed anticoagulation group, DOAC therapy will be initiated between 5 and 14 days after stroke onset; if the exact time of stroke onset cannot be determined, the time last known well is used as the reference. All enrolled patients undergo EVT before randomisation. The endovascular strategy, including mechanical thrombectomy, thrombus aspiration or balloon angioplasty, is determined by the treating physician according to standard clinical practice. Permitted anticoagulant medications include rivaroxaban, dabigatran, apixaban and edoxaban. The agent will be selected by the treating physician according to approved indications, and dose adjustment will be permitted in accordance with the prescribing information. Routine systemic heparin administration is not permitted during the periprocedural period of EVT. All other clinical management guidelines follow the recommendations of the American Heart Association and American Stroke Association guidelines.11 19 The allocated anticoagulation strategy will be modified or discontinued at the discretion of the treating physician in cases of major bleeding or serious adverse events. All treatment modifications, interruptions and permanent discontinuations are documented in the trial records. Adherence to the allocated anticoagulation timing strategy is assessed by documenting the actual date of DOACs initiation, selected agent and dosage and continuation of therapy throughout the follow-up. Primary outcome The primary outcome is the incidence of a composite endpoint comprising recurrent ischaemic stroke, symptomatic intracranial haemorrhage and all cause death within 90 days after randomisation. Recurrent ischaemic stroke is defined as the occurrence of a new, sudden-onset focal neurological deficit arising more than 24 hours after the index stroke. Neurological deficits that persist for more than 24 hours and cannot be explained by cerebral oedema, mass effect, brain shift or haemorrhagic transformation of the initial infarction are considered. The diagnosis is confirmed by neuroimaging (CT or MRI).20 Intracranial haemorrhage will be classified as symptomatic if it is associated with an increase of ≥4 points in the NIHSS score compared with the neurological status immediately before worsening, an increase of ≥2 points in one NIHSS category, the need for major medical or surgical intervention and an absence of alternative explanation for the neurological worsening.21 All primary outcome events are adjudicated by an independent clinical endpoint committee blinded to the treatment allocation. Secondary outcomes Secondary outcomes are as follows: (1) recurrent ischaemic stroke within 90 days; (2) venous thromboembolism within 90 days; (3) systemic embolism within 90 days; (4) myocardial infarction within 90 days; (5) proportion of patients achieving an mRS score of 0–1 at 90 days; (6) proportion of patients achieving an mRS score of 0–2 at 90 days; (7) health-related quality of life (EuroQol 5-Dimension 5-Level questionnaire (EQ-5D-5L) score) at 90 days; (8) stroke-related length of hospital stay. The safety outcomes include (1) incidence of symptomatic intracranial haemorrhage within 90 days; (2) incidence of major extracranial bleeding within 90 days; (3) proportion of patients with all-cause death within 90 days; (4) proportion of patients with vascular death within 90 days. All outcome events and adverse events are adjudicated by an independent endpoint assessment committee blinded to the treatment allocation. Data collection Data are collected at each participating centre using a secure web-based electronic data capture system. Each participant is assigned a unique study identification number, and identifiable information is accessible only to authorised study personnel to ensure confidentiality. Follow-up assessments are conducted at 7, 14 and 90 days after randomisation in accordance with the study schedule (table 1). Table 1Schedule of study assessments and interventionsTime pointScreening and enrolment periodTreatment and follow-up period−3 days to 0 days0 daysDay 7±1 day or at dischargeDay 14±1 dayDay 90±7 daysScreening Demographic data× Vital signs× Medical history× Medical history× Concomitant medications× CHA2 DS2-VASc× Pre stroke mRS× NIHSS after EVT× NIHSS before EVT× Blood tests× ECG or cardiac monitoring× CT/MRI× ASPECTS× mTICI× EVT details×Enrolment Inclusion and exclusion criteria× Informed consent× Randomisation×Intervention Early anticoagulation groupAnticoagulation initiated on days 1–4 after stroke onset Delayed anticoagulation groupAnticoagulation initiated on days 5–14 after stroke onsetAssessment mRS× DOAC treatment-related information××× Concomitant medications××× EQ-5D-5L× Serious adverse events××× Adverse events××× New outcome events××× Stroke-related length of hospital stay××× Outpatient or telephone follow-up××ASPECTS, Alberta Stroke Program Early CT Score; CHA₂DS₂-VASc, Congestive heart failure, Hypertension, Age ≥75 years, Diabetes mellitus, Stroke/transient ischaemic attack/thromboembolism, Vascular disease, Age 65–74 years, and Sex category (female); DOAC, direct oral anticoagulant; EQ-5D-5L, EuroQol 5-Dimension 5-Level questionnaire; EVT, endovascular thrombectomy; mRS, modified Rankin Scale; mTICI, modified Thrombolysis in Cerebral Infarction score; NIHSS, National Institutes of Health Stroke Scale. Schedule of study assessments and interventions ASPECTS, Alberta Stroke Program Early CT Score; CHA₂DS₂-VASc, Congestive heart failure, Hypertension, Age ≥75 years, Diabetes mellitus, Stroke/transient ischaemic attack/thromboembolism, Vascular disease, Age 65–74 years, and Sex category (female); DOAC, direct oral anticoagulant; EQ-5D-5L, EuroQol 5-Dimension 5-Level questionnaire; EVT, endovascular thrombectomy; mRS, modified Rankin Scale; mTICI, modified Thrombolysis in Cerebral Infarction score; NIHSS, National Institutes of Health Stroke Scale. Efforts are made to minimise the loss to follow-up. In particular, investigators at each centre are available to address the participants’ questions and concerns. Moreover, participants who are unable to attend in-person visits are contacted via telephone when feasible. Participants who discontinue or deviate from the allocated intervention are encouraged to continue follow-up, and outcome data will be collected whenever possible. Data safety and monitoring The trial is overseen by an independent data monitoring committee (DMC). The DMC reviews accumulated safety data at regular intervals. Early termination may be considered if the DMC identifies an unacceptable risk to participant safety, insufficient recruitment that threatens trial completion, serious or persistent deviations from the approved protocol or compelling evidence of benefit or harm associated with either treatment strategy. These criteria will prompt formal DMC consideration rather than automatic termination. Any recommendation to suspend, modify or terminate the trial will be based on the totality of the available evidence and clinical judgement. The final decision will be made by the steering committee after consideration of the DMC recommendation. Sample size estimates In a previous study,18 the 90-day incidence of the primary composite outcome was 26.4% among patients who initiated anticoagulation therapy within 4 days of stroke onset and 13.8% among those who initiated anticoagulation therapy between 5 and 14 days. Assuming a two-sided type 1 error (α) of 0.05 and 90% power (β=0.1), with a 1:1 allocation ratio between the two groups, 208 patients were required in each group. To account for an anticipated 5% loss to follow-up or withdrawal, the sample size was increased to 219 patients per group, resulting in a total planned enrolment of 438 participants. Statistical analyses The primary analysis of this study will follow the intention-to-treat principle, with all participants analysed according to a randomised treatment allocation. Cox proportional hazard regression models will be used to estimate HRs and corresponding 95% CIs for the primary composite outcome of recurrent ischaemic stroke, all-cause death and symptomatic intracranial haemorrhage within 90 days. The models will be adjusted for prespecified covariates, including age, sex, NIHSS score at randomisation and the presence of haemorrhagic transformation. Time-to-event curves will be constructed using the Kaplan-Meier method and compared using the log-rank test. Survival outcomes, such as 90-day recurrent stroke, myocardial infarction and death, will also be analysed using Cox proportional hazard regression models. Binary outcomes, such as the proportion of patients achieving specific mRS categories will be analysed using modified Poisson regression models. Continuous outcomes, including EQ-5D-5L scores and length of hospital stay, will be analysed using linear regression models if normally distributed; a win-ratio approach will be applied if normality assumptions are not met. Missing covariate data will be handled according to the extent of missingness. When the proportion of missing data is below 5%, a single imputation based on the distribution of the observed data will be applied. For higher levels of missingness, multiple imputations will be performed using Markov chain Monte Carlo methods. All statistical tests will be two-sided, p<0.05 will be considered statistically significant. Study conduct The first patient was enrolled on 4 September 2025. The estimated study completion date is 31 December 2026. As of 28 February 2026, 150 patients have been enrolled in this study. Patient and public involvement The patients and members of the public were consulted during the study design phase. Two public representatives (one woman and one man) reviewed the proposed study procedures, follow-up schedule and outcome assessments to ensure that the protocol was feasible and acceptable from a patient perspective. However, they will not be involved in participant recruitment, data collection or data analysis. The trial findings will be disseminated through peer-reviewed publications and presentations at scientific conferences. Discussion This randomised controlled trial is designed to investigate the optimal timing of anticoagulation therapy initiation after EVT in patients with AF and large vessel occlusion stroke. This patient population has been under-represented in previous randomised studies. Compared with non-reperfused patients, individuals undergoing EVT typically present with more severe neurological deficits, larger infarct volumes and a higher likelihood of blood-brain barrier disruption, all of which may increase the risk of haemorrhagic transformation.7 As a result, the balance between recurrent ischaemia and haemorrhage may differ in this population compared with that in patients who do not receive reperfusion therapy. In our previous real-world cohort study18 comprising patients treated with EVT and a median baseline NIHSS score of 15, the early initiation of anticoagulation was associated with a higher incidence of intracranial haemorrhage. In contrast, initiation between 5 and 14 days was more likely to achieve a favourable balance between recurrent ischaemia and haemorrhagic complications. Although residual confounding factors cannot be excluded given the observational design, our findings suggest that the risk profile of patients treated with EVT may differ from that observed in previous randomised trials. These data highlight the need for a dedicated randomised trial involving this specific population. In this study, we do not stratify patients according to infarct size. In the ELAN trial, risk stratification was based on infarct location and extent on neuroimaging. However, no significant interaction was observed between infarct extent and the timing of anticoagulation therapy initiation with respect to clinical outcomes.22 Based on this finding and considering the practical challenges of standardised quantitative imaging assessment in an acute, time-sensitive EVT setting, we adopt a time-based randomisation strategy without stratification by infarct size. This approach is intended to enhance the feasibility, simplify enrolment procedures and improve the generalisability of the findings to real-world clinical practice. Supplementary material