skip_previous play_arrow skip_next
00:00 00:00
playlist_play chevron_left


ELAN trial – Early vs late anticoagulation in Post-ischaemic stroke with AF

Dr Swapnil Pawar October 4, 2023 215

share close
  • cover play_arrow

    ELAN trial – Early vs late anticoagulation in Post-ischaemic stroke with AF
    Dr Swapnil Pawar


Blog written by – Dr Jose Chacko

Early versus Later Anticoagulation for Stroke with Atrial Fibrillation

(Fischer U, Koga M, Strbian D, et al; ELAN Investigators. Early versus Later Anticoagulation for Stroke with Atrial Fibrillation. N Engl J Med. 2023 Jun 29;388(26):2411-2421)


Considering the lack of high-quality evidence, guideline recommendations vary regarding the timing of initiation of anticoagulation following ischemic stroke. The European Stroke Organization and the American Heart Association/American Stroke Association currently recommend delaying anticoagulation after ischemic stroke for more than 48 hours and up to 14 days, based on stroke severity and the risk of hemorrhagic conversion. The TIMING trial revealed noninferiority of initiation with direct-acting oral anticoagulants (DOAC) at ≤4 days after ischemic stroke. Early commencement of anticoagulation may increase the risk of intracranial hemorrhage; however, delayed initiation may lead to early stroke recurrence.

Population and design

The ELAN trial was conducted across 103 centers in Europe, Middle East, and Asia between 2017–2022. The study aimed to compare the effect of early with later initiation of anticoagulation following acute ischemic stroke in patients with atrial fibrillation. Patients were eligible if they suffered an acute ischemic stroke and had non-valvar atrial fibrillation – persistent, paroxysmal, or diagnosed during hospitalization. Ischemic stroke was confirmed by evidence of acute cerebral infarction on MRI or CT imaging, with persisting symptoms for >24 hours. The infarct size was classified based on the appearance on imaging into minor, moderate, or major stroke.

Minor stroke: infarct size ≤1.5cm

Moderate stroke: Infarct in the distribution of a cortical superficial branch of the middle, anterior, or posterior cerebral artery

Major stroke: Larger infarcts in the distribution of the above arteries OR a brainstem or cerebellar infarct ≥ 1.5cm


Patients who received intravenous thrombolysis or underwent mechanical thrombectomy before randomization were eligible; therapeutic anticoagulation at the onset of stroke onset was not allowed. Prophylaxis against deep vein thrombosis was allowed.



  • Patients with prosthetic valves
  • Moderate to severe mitral stenosis
  • Patients with other indications for anticoagulant therapy
  • Therapeutic anticoagulation at baseline
  • Serious bleeding in the last 6 months or patients with high risk of bleeding
  • Severe renal impairment
  • Intraparenchymal bleed


Study groups

Patients were randomized to early or later initiation of direct-acting oral anticoagulants (DOACs) in a 1:1 ratio. Randomization was stratified based on age (<70 years or more), infarct size (minor, moderate, or major), the NIHSS score (<10 or more), and trial site. Earlier and later treatments were defined as follows.

Early treatment

DOAC was commenced within 48 hours of symptom onset in minor or moderate stroke; in major stroke, treatment was commenced on day 6 or 7.

Later treatment

DOAS was commenced on day 3 or 4 for minor stroke, day 6 or 7 in moderate stroke, and day 12, 13, or 14 in major stroke. Both groups received stroke treatment according to local protocols. The choice of DOAC was left to clinician discretion.


Statistical methodology

At the time of commencement of the trial, no unbiased data were available to form a hypothesis regarding potential superiority, or non-inferiority margin between early compared with later treatment. Hence, no hypothesis-based testing was carried out. The authors aimed to provide ranges for outcome differences between trial groups based on the confidence intervals. The results were meant solely to provide qualitative data to clinicians. Sample size calculation was entirely based on the expected width of the confidence interval. A sample size of 1802 patients was calculated based on the primary outcome event of 5% in the later and 4.5% in the early treatment group with 95% confidence interval width of at least 2.0 percentage points. The authors planned to enroll 2000 patients, allowing for missing outcome data.



Baseline characteristics

Overall, 2032 patients were randomized, full data were available for 2013 – 1006 in the early, and 1007 in the late treatment arm. Baseline characteristics were similar in both groups – the median age was 70 years; nearly 40% of patients experienced a minor stroke, while another 40% had a moderate stroke; 20% suffered a major stroke. Thrombolysis was carried out in approximately 40% and mechanical thrombectomy in 20% of patients.

The primary outcome – at 30 days

The primary outcome was a composite of recurrent ischemic stroke, symptomatic intracranial hemorrhage, major external bleeding, systemic embolism, and vascular death at 30 ± 3 days after randomization. The primary-outcome event was lower in the early compared with the later treatment group [29 (2.9%) vs. 41 (4.1%) patients]. The odds ratio of a primary outcome event in the early vs. later treatment group was 0.7 (95% confidence interval, 0.44–1.14). This corresponds to a range that lies between a reduction by approximately 2.8 percentage points to an increase by 0.5 percentage points in the risk of a primary outcome event.

Components of the primary outcome at 30 days


Endpoint Early Late OR (95% CI)
Recurrent ischemic stroke 14/984 (1.4%) 25/991 (2.5%) 0.57 (0.29–1.07)
Systemic embolism 4/984 (0.4%) 9/991 (0.9) 0.48 (0.14 to 1.42)
Vascular death 11/984 (1.1%) 10/991 (1.0%) 1.12 (0.47–2.65)
Symptomatic intracranial bleed 2/984 (0.2%) 2/991 (0.2%) 1.02 (0.16 to 6.59)
Major extracranial bleed 3/984 (0.3) 5/991 (0.5) 0.63 (0.15 to 2.38)


Patients with a score on the modified Rankin scale of 0–2 (able to look after own affairs without assistance) were nearly equal – approximately 66% in both groups.

Secondary outcomes – at 90 days

The composite outcome at 90 days occurred in 36/968 (3.7%) of patients in the early compared with 54/965 (5.6%) of patients in the late group [OR, 0.65; 95% CI, 0.42–0.99). Recurrent ischemic events were lower in the early group; other outcomes were similar between the two groups.


The study employed an image-based classification of the severity of stroke into minor, moderate, or severe. Based on this classification, the incidence of the composite outcome of recurrent stroke, systemic embolism, hemorrhage, or death at 30 days was in the range from 2.8% points lower to 0.5% points higher with the early use of DOACS compared to later use. No increased risk of bleeding was observed with the early use of DOACs. The study supports the early use of DOACs following acute ischemic stroke in patients with patients with atrial fibrillation. The study suggests that early anticoagulation may be better and unlikely to cause harm.


  • Multinational randomized controlled trial
  • The authors addressed a relevant clinical question considering the lack of robust evidence and variable guideline recommendations
  • Loss to follow-up was minimal


  • The study excluded patients who were on therapeutic anticoagulation at baseline
  • Most patients were enrolled from Europe – validation in other ethnic groups cannot be confirmed
  • Patients with parenchymal hemorrhage were excluded
  • Stroke severity was low based on the NIHSS stroke scale
  • The study was not designed to test superiority or non-inferiority; hence, it can only offer guidance to clinicians on the most appropriate timing of anticoagulation



Rate it
Previous episode
eCritCare Podcast
share playlist_add
  • 402


Steroids in Community Acquired Pneumonia

Dr Swapnil Pawar August 21, 2023

play_arrow Steroids in Community Acquired Pneumonia Dr Swapnil Pawar In this episode, we discuss the controversy surrounding the usage of steroids in Community-acquired pneumonia. This controversy arose after the CAPE-COD […]

Read more trending_flat