New Data on DOAC Initiation After Stroke in AF: Final Word?

Abu Dhabi, UAE — The long-standing debate as to when to start anticoagulation in patients with an acute ischemic stroke and atrial fibrillation (AF) looks as though it’s settled.

Results of the OPTIMAS trial, the largest trial to address this question, showed that initiation of a direct oral anticoagulant (DOAC) within 4 days after ischemic stroke associated with AF was noninferior to delayed initiation (7-14 days) for the composite outcome of ischemic stroke, intracranial hemorrhage, unclassifiable stroke, or systemic embolism at 9src days. Importantly, early DOAC initiation was safe with a low rate of symptomatic hemorrhage, regardless of stroke severity.

In addition, a new meta-analysis, known as CATALYST, which included all four randomized trials now available on this issue, showed a clear benefit of earlier initiation (within 4 days) vs later (5 days and up) on its primary endpoint of new ischemic stroke, symptomatic intracerebral hemorrhage, and unclassified stroke at 3src days.

The results of the OPTIMAS trial and the meta-analysis were both presented on October 24 at the 16th World Stroke Congress (WSC) 2src24. The OPTIMAS trial was also simultaneously published online in The Lancet.

“Our findings do not support the guideline recommended practice of delaying DOAC initiation after ischemic stroke with AF regardless of clinical stroke severity, reperfusion or prior anticoagulation,” said OPTIMAS investigator David Werring, PhD, University College London, London, England.

Presenting the meta-analysis, Signild Åsberg, MD, Uppsala University, Uppsala, Sweden, said his group’s findings “support the early start of DOACs (within 4 days) in clinical practice.”

Werring pointed out that starting anticoagulation early also had important logistical advantages.

“This means we can start anticoagulation before patients are discharged from hospital, thus ensuring that this important secondary prevention medication is always prescribed, when appropriate. That’s going to be a key benefit in the real world.”

Clinical Dilemma

Werring noted that AF accounts for 2src%-3src% of ischemic strokes, which tend to be more severe than other stroke types. The pivotal trials of DOACs did not include patients within 3src days of an acute ischemic stroke, creating a clinical dilemma on when to start this treatment.

“On the one hand, we wish to start anticoagulation early to reduce early recurrence of ischemic stroke. But on the other hand, there are concerns that if we start anticoagulation early, it could cause intracranial bleeding, including hemorrhagic transformation of the acute infarct. Guidelines on this issue are inconsistent and have called for randomized control trials in this area,” he noted.

So far, three randomized trials on DOAC timing have been conducted, which Werring said suggested early DOAC treatment is safe. However, these trials have provided limited data on moderate to severe stroke, patients with hemorrhagic transformation, or those already taking oral anticoagulants — subgroups in which there are particular concerns about early oral anticoagulation.

The OPTIMAS trial included a broad population of patients with acute ischemic stroke associated with AF including these critical subgroups.

The trial, conducted at 1srcsrc hospitals in the United Kingdom, included 3648 patients with AF and acute ischemic stroke who were randomly assigned to early (≤ 4 days from stroke symptom onset) or delayed (7-14 days) anticoagulation initiation with any DOAC.

There was no restriction on stroke severity, and patients with hemorrhagic transformation were allowed, with the exception of parenchymal hematoma type 2, a rare and severe type of hemorrhagic transformation.

Approximately 35% of patients had been taking an oral anticoagulant, mainly DOACs, prior to their stroke, and about 3src% had revascularization with thrombolysis, thrombectomy, or both. Nearly 9srcsrc participants (25%) had moderate to severe stroke (National Institutes of Health Stroke Scale [NIHSS] score ≥ 11).

The primary outcome was a composite of recurrent ischemic stroke, symptomatic intracranial hemorrhage, unclassifiable stroke, or systemic embolism incidence at 9src days. The initial analysis aimed to show noninferiority of early DOAC initiation, with a noninferiority margin of 2 percentage points, followed by testing for superiority.

Results showed that the primary outcome occurred in 3.3% of both groups (adjusted risk difference, src.srcsrcsrc; 95% CI, −src.src11 to src.src12), with noninferiority criteria fulfilled. Superiority was not achieved.

Symptomatic intracranial hemorrhage occurred in src.6% of patients in the early DOAC initiation group vs src.7% of those in the delayed group — a nonsignificant difference.

Applicable to Real-World Practice

A time-to-event analysis of the primary outcome showed that there were fewer outcomes in the first 3src days in the early DOAC initiation group, but the curves subsequently came together.

Subgroup analysis showed consistent results across all whole trial population, with no modification of the effect of early DOAC initiation according to stroke severity, reperfusion treatment, or previous anticoagulation.

Werring said that strengths of the OPTIMAS trial included a large sample size, a broad population with generalizability to real-world practice, and the inclusion of patients at higher bleeding risk than included in previous studies.

During the discussion, it was noted that the trial included few (about 3%) patients — about 3% — with very severe stroke (NIHSS score> 21), with the question of whether the findings could be applied to this group.

Werring noted that there was no evidence of heterogeneity, and if anything, patients with more severe strokes may have had a slightly greater benefit with early DOAC initiation. “So my feeling is probably these results do generalize to the more severe patients,” he said.

In a “Comment” accompanying The Lancet publication of the OPTIMAS trial, Else Charlotte Sandset, MD, University of Oslo, Oslo, Norway, and Diana Aguiar de Sousa, MD, Central Lisbon University Hospital Centre, Lisbon, Portugal, noted that the “increasing body of evidence strongly supports the message that initiating anticoagulation early for patients with ischaemic stroke is safe. The consistent absence of heterogeneity in safety outcomes suggests that the risk of symptomatic intracranial haemorrhage is not a major concern, even in patients with large infarcts.”

Regardless of the size of the treatment effect, initiating early anticoagulation makes sense when it can be done safely, as it helps prevent recurrent ischemic strokes and other embolic events. Early intervention reduces embolization risk, particularly in high-risk patients, and allows secondary prevention measures to begin while patients are still hospitalized, they added.

CATALYST Findings

The CATALYST meta-analysis included four trials, namely, TIMING, ELAN, OPTIMAS, and START, of early vs later DOAC administration in a total of 5411 patients with acute ischemic stroke and AF. In this meta-analysis, early was defined as within 4 days of stroke and later as 5 days or more.

The primary outcome was a composite of ischemic stroke, symptomatic, intracerebral hemorrhage, or unclassified stroke at 3src days. This was significantly reduced in the early group (2.12%) vs 3.src2% in the later group, giving an odds ratio of src.7src (95% CI, src.5src-src.98; P=.src4).

The results were consistent across all subgroups, all suggesting an advantage for early DOAC.

Further analysis showed a clear benefit of early DOAC initiation in ischemic stroke with the curves separating early.

The rate of symptomatic intracerebral hemorrhage was low in both groups (src.45% in the early group and src.4src% in the later group) as was extracranial hemorrhage (src.45% vs src.55%).

At 9src days, there were still lower event rates in the early group than the later one, but the difference was no longer statistically significant.

‘Practice Changing’ Results

Commenting on both studies for Medscape Medical News, chair of the WSC session where the results of both OPTIMAS trial and the meta-analysis were presented, Craig Anderson, MD, The George Institute for Global Health, Sydney, Australia, described these latest results as “practice changing.”

“When to start anticoagulation in acute ischemic stroke patients with AF has been uncertain for a long time. The dogma has always been that we should wait. Over the years, we’ve become a little bit more confident, but now we’ve got good data from randomized trials showing that early initiation is safe, with the meta-analysis showing benefit,” he said.

“These new data from OPTIMAS will reassure clinicians that there’s no excessive harm and, more importantly, no excessive harm across all patient groups. And the meta-analysis clearly showed an upfront benefit of starting anticoagulation early. That’s a very convincing result,” he added.

Anderson cautioned that there still may be concerns about starting DOACs early in some groups, including Asian populations that have a higher bleeding risk (these trials included predominantly White patients) and people who are older or frail, who may have extensive small vessel disease.

During the discussion, several questions centered on the lack of imaging data available on the patients in the studies. Anderson said imaging data would help reassure clinicians on the safety of early anticoagulation in patients with large infarcts.

“Stroke clinicians make decisions on the basis of the patient and on the basis of the brain, and we only have the patient information at the moment. We don’t have information on the brain — that comes from imaging.”

Regardless, he believes these new data will lead to a shift in practice. “But maybe, it won’t be as dramatic as we would hope because I think some clinicians may still hesitate to apply these results to patients at high risk of bleeding. With imaging data from the studies that might change.”

The OPTIMAS trial was funded by University College London and the British Heart Foundation. Werring reported consulting fees from Novo Nordisk, National Institute for Health and Care Excellence, and Alnylam; payments or speaker honoraria from Novo Nordisk, Bayer, and AstraZeneca/Alexion; participation on a data safety monitoring board for the OXHARP trial; and participation as steering committee chair for the M