Oral Anticoagulation in ICH Patients With AF: New Data

New information has become available to help guide clinicians on the thorny issue of oral anticoagulation in patients with atrial fibrillation (AF) who have had an intracranial hemorrhage (ICH).

The data come from a new meta-analysis of four small, randomized trials, and a warning from a fifth ongoing study about possible harm in certain patients.

"While the issue of whether to give oral anticoagulation to ICH patients with AF is still very uncertain, I think this new information will provide some helpful guidance on decision-making in these patients," lead author of the meta-analysis, Rustam Al-Shahi Salman, MD, professor of clinical neurology at the University of Edinburgh, UK, told theheart.org | Medscape Cardiology.

The meta-analysis shows oral anticoagulation was associated with a clear reduction in major ischemic cardiovascular events, and ischemic stroke, but, because of uncertainties about bleeding risk, and small numbers of patients, it was not possible to show a net clinical benefit.

Salman presented the meta-analysis at the World Stroke Conference (WSC) 2023, held last week in Toronto. It was also simultaneously published online in Lancet Neurology.

At the same time, investigators of a fifth still-ongoing study of oral anticoagulation in ICH patients with AF – the ENRICH-AF study – announced that the trial's data safety monitoring board has recommended that patients with two particular types of ICH (lobar intracranial hemorrhage and convexity subarachnoid hemorrhage) stop receiving the anticoagulant being tested (edoxaban) and that no further patients with these types of ICH be enrolled into the study because of an "unacceptably high risk of recurrent hemorrhagic stroke" in those assigned to the edoxaban group.

The ENRICH-AF investigators reported this in a letter to The Lancet published online on October 12.

Salman explained that while oral anticoagulation has been known for many years to have a clear and powerful benefit in preventing ischemic strokes for patients with AF, whether oral anticoagulation would also be beneficial in AF patients who have had an ICH is not at all clear.

"One in five patients who have an ICH also have AF, so this is a large group of patients. Doctors are particularly worried about bleeding risk in ICH patients, and so, by and large, these patients are not treated with anticoagulation," he noted.

Several small pilot trials have now been done looking at the use of oral anticoagulants in ICH patients with AF. Four trials have been completed so far, and another five trials are ongoing.

"Because the results from all the trials are not going to be available for several years, we thought this was a good time to look at the data we have so far with an individual patient meta-analysis of the first four studies," Salman said.

The meta-analysis was conducted by the investigators involved in the four completed trials as a group known as COCROACH (Collaboration of Controlled Randomized Trials of Long-Term Oral Antithrombotic Agents After Spontaneous Intracranial Hemorrhage).

The meta-analysis included 412 patients with AF and ICH from four trials — SoSTART, APACHE-AF, NASPAF-ICH and ELDERCARE-AF — comparing oral anticoagulation (99% direct oral anticoagulant and 1% vitamin K antagonist) with control patients who avoided oral anticoagulation (67% no antithrombotic therapy and 33% antiplatelet monotherapy).

Meta-analysis Results

The primary outcome of any stroke or cardiovascular death occurred in 14% of those on oral anticoagulation vs 22% of controls, with a pooled hazard ratio (HR) of 0.68 (95% CI, 0.42 - 1.10).

Oral anticoagulation reduced the risk for ischemic major adverse cardiovascular events, which occurred in 4% of the treated patients vs 19% of controls (HR, 0.27; 95% CI, 0.13 - 0.56).

There was no significant increase in hemorrhagic major adverse cardiovascular events, which occurred in 7% of the oral anticoagulant group vs 5% of controls (pooled HR, 1.80; 95% CI, 0.77 - 4.21).

There was also no statistical difference in death from any cause: 18% of the oral anticoagulation group vs 15% controls (HR, 1.29; 95% CI, 0.78 - 2.11); or in death or dependence after 1 year, 53% vs 51% (pooled odds ratio, 1.12 (95% CI, 0.70 - 1.79).

"Our main finding was that oral anticoagulation reduces the risk of ischemic cardiovascular events in this patient group," Salman commented. "We showed a clear benefit there. But all the other findings were neutral, and we couldn't show a definite net effect on any stroke or cardiovascular death. But nor did we find a significant increase in the risk of bleeding, which is important because we first worry about harm."

These new data shift "us from a position of not knowing what the effects of oral anticoagulation are in this population on either clotting or bleeding to one where we now know the effects on clotting, but we are still uncertain where the overall net balance lies," he added. "While the results provide us with some encouragement in that we have confirmed a beneficial effect on ischemic events, they are not practice changing.

"The fact we haven't shown an increase in ICH with oral anticoagulation in this meta-analysis is encouraging, but we have to remember that an ICH is generally more severe than an ischemic event, so the overall balance when it comes to death and disability is still uncertain," Salman noted.

While the authors of the meta-analysis looked at subgroups, the sample size was too small to find any meaningful results. 

"We didn’t identify a major hazard or a major benefit in any particular subgroup, and we did not see any hints of the harms found in the ENRICH-AF trial in patients with lobar or convexity subarachnoid hemorrhage," he added. 

Weighing Risks and Benefits in Each Individual

Salman suggested that applying the evidence to individuals will involve weighing up risks and benefits.

"There are known risk factors for increased bleeding risk and there are also known risk factors for higher risks of clotting. Patients with lobar hemorrhages over the surfaces of [the] brain and convexity subarachnoid hemorrhage are at higher risk of recurrent ICH, so clinicians will be reluctant to use oral anticoagulants in those groups, especially given the announcement from the ENRICH-AF trial," he noted.

"But we might now feel a bit more secure in using an oral anticoagulant in patients with higher CHA₂DS₂-VASc scores (showing a higher risk of ischemic stroke), but who do not have lobar or convexity subarachnoid hemorrhage," he added.

Salman noted that it is important to continue to randomly assign patients into the ongoing trials so a clearer result will be available in future.

"In my own practice and for many other practices, the sensible thing is to randomize patients into the ongoing trials. If those trials are not available, then clinicians now have a bit more guidance on how to apply the data in a nuanced way to individuals thinking about their risk of bleeding and clotting," he concluded.

In a commentary accompanying publication of the meta-analysis, Shinichiro Uchiyama, MD, International University of Health and Welfare, Sanno Medical Center, Tokyo, Japan, suggests that among individuals with AF and ICH, direct oral anticoagulants should be used for those at low risk for bleeding, or they should be given at low doses for individuals at high risk of bleeding.

"Knowledge of the risk of stroke, major cardiovascular events, and poor outcome will be imperative to decide on the contraindications for anticoagulation and the indications for left atrial appendage closure in people with atrial fibrillation and intracranial hemorrhage," Uchiyama writes. "The development of a useful risk score for stroke and poor outcome in these individuals is keenly anticipated."

The COCROACH meta-analysis was funded by the British Heart Foundation. The APACHE-AF trial was funded by the Dutch Heart Foundation. NASPAF-ICH was funded by the Heart and Stroke Foundation of Canada, Canadian Stroke Prevention Intervention Network and the Population Health Research Institute. SoSTART was funded by the Medical Research Council, Chest Heart & Stroke Scotland, and the British Heart Foundation.

World Stroke Congress 2023. Presented October 12, 2023.

Lancet Neurology. Published online October 12, 2023.  Abstract, Editorial

Lancet. Published online October 12, 2023. ENRICH-AF Letter

For more from theheart.org | Medscape Cardiology, join us on X (formerly Twitter) and Facebook