Optimal Time to Start Anticoagulation After Stroke Remains Unclear

LOS ANGELES, CA—To balance risks of hemorrhage and recurrent ischemic events, researchers continue to try to refine the optimal timing for initiation of oral anticoagulation after stroke in patients with A-fib, with no clear answer emerging, as highlighted by two studies presented here at the International Stroke Conference last week.

In one study presented by Shadi Yaghi, MD (NYU Langone Health, New York, NY), there were no significant differences in risks of recurrent ischemic events or symptomatic intracranial hemorrhage (SICH) based on whether anticoagulation was started within the first few days after a cardioembolic stroke, during an intermediate period up to 2 weeks, or beyond that time frame. There were suggestions, however, that starting treatment at 7 to 10 days carried the lowest risks.

And in the other study, a meta-analysis presented by Gian Marco De Marchis, MD (University Hospital Basel, Switzerland), risks of recurrent ischemic stroke and SICH did not differ between patients who were started on a direct oral anticoagulant (DOAC) within the first 5 days versus later.

During his presentation, Yaghi said “the timing of anticoagulation did not seem to affect the risk of early recurrence. Other factors may be more important.” The study did, however, show that bridging with heparin/low-molecular-weight heparin (LMWH) was associated with an increased SICH risk, without any reduction in ischemic events, and that using a DOAC instead of warfarin was associated with fewer recurrent ischemic events.

“It seems reasonable to avoid bridging unless absolutely necessary and initiating a DOAC unless there’s a contraindication,” Yaghi said. “And it may be reasonable to start a DOAC day 7 to 10 in most patients. Ongoing RCTs will inform us on the optimal timing.”

Limited Evidence Breeds Uncertainty

When to initiate oral anticoagulation after stroke among patients with A-fib is a controversial issue, Yaghi told TCTMD. Start too early and there’s a risk of hemorrhagic transformation. Wait too long and there’s a greater risk of a recurrent event.

Guidelines from the American Heart Association/American Stroke Association state that it is reasonable to start anticoagulation 4 to 14 days after an acute ischemic stroke in the setting of A-fib. That advice is based largely on the results of the RAF study, a prospective, observational analysis showing that 90-day risk of composite ischemic and bleeding outcomes was lowest when treatment was started within that window.

But RAF was limited in that patients were mostly treated with warfarin rather than newer DOACs and it did not take into consideration many high-risk features, like the presence of cardiac thrombus, that would cause clinicians to err on the side of initiating anticoagulation earlier, Yaghi said. The treatment window is also very wide, creating uncertainty and a lot of variability in practice, he added.

To try to get a better handle on when to start anticoagulation, Yaghi and colleagues conducted the IAC study, which pooled data from the registries of eight comprehensive stroke centers in the United States. The analysis included 1,289 patients with A-fib who had had a cardioembolic stroke. Timing of anticoagulation initiation was initially divided into three time periods: 0 to 3 days, 4 to 14 days, and beyond 14 days.

Risk of a recurrent ischemic event (ischemic stroke, TIA, or symptomatic systemic embolism) at 90 days did not differ across these periods (7.3%, 6,0%, and 7.2%, respectively, P = 0.651), and neither did risk of SICH after initiating oral anticoagulation (1.1%, 1.7%, and 2.9%, respectively, P = 0.295).

A composite endpoint consisting of recurrent ischemic events, SICH, and major extracranial hemorrhage occurred in 10.1% of patients overall, with no differences according to the timing of treatment (P = 0.933).

Adjusted regression models that included tighter treatment windows provided similar results, although initiation at 7 to 10 days was associated with numerically—but nonsignificantly—lower risks of all outcomes (ORs ranging from 0.64 to 0.72). In particular, starting a DOAC within that time frame was associated with a numerically lower risk of the combined endpoint (OR 0.50; 95% CI 0.23-1.06).

Take Findings ‘With a Grain of Salt’

De Marchis and his colleagues looked at the timing question by pooling individual patient-level data from seven European and Japanese prospective, observational cohort studies. The analysis included 2,550 patients who had had an ischemic stroke/TIA related to A-fib and were started on a DOAC within 30 days. Treatment was started within 5 days in 53% and later in the rest; average time to initiation was 14 days in the latter group.

Similar to Yaghi’s study, this analysis showed that outcomes did not significantly differ based on when anticoagulation was started. That applied to both recurrent ischemic stroke risk at 30 days (1.7% early vs 1.2% late; adjusted HR 1.24; 95% CI 0.51-2.99) and intracranial hemorrhage (0.1% vs 1.2%; adjusted HR 6.46; 95% CI 0.68-61.61).

“Our results do not support the concerns that [an] early DOAC start is associated with intracranial hemorrhage, and the sevenfold higher risk of recurrent ischemic events compared to the risk of intracranial hemorrhage argues for an early start of DOACs after an ischemic stroke due to A-fib,” De Marchis concluded. “However, we encourage randomizing these patients in the ongoing clinical trials comparing an early to a late DOAC start.”

Yaghi agreed that randomized data are needed to sort this out, pointing to limitations inherent to observational analyses like these. Of note, these types of investigations cannot fully account for the factors causing clinicians to choose between early and late starts to anticoagulation in individual patients, which introduces bias. “I think we should take all these studies with a grain of salt,” he told TCTMD. “They definitely don’t adequately answer the question.”

While awaiting further data, however, Yaghi said that the results of his study and others have influenced his practice.

“I think each patient is different, and I think we need to individualize, but perhaps for most patients, the 7- to 10-day time frame would be a reasonable time period to start anticoagulation,” he advised. For a patient with A-fib who has a TIA and normal MRI findings, oral anticoagulation can probably be started immediately, Yaghi added. “But if they have a sizable stroke, then I would probably want to wait for at least 7 days to start anticoagulation.”