Brain Bleed Survivors Walk a Razor-Thin Line Taking DOACs

LOS ANGELES — In the PRESTIGE-AF trial, survivors of intracerebral hemorrhage (ICH) had good protection against future ischemic events using direct oral anticoagulants (DOACs) — though, as feared, bleeding side effects reared their ugly heads.

Study participants, people weeks or months out from their index spontaneous ICH who also had atrial fibrillation (Afib), had a major reduction in their risk of a first ischemic stroke (adjusted HR 0.05, 95% CI 0.01-0.38) over a mean follow-up of 1.43 years if they had been randomized to DOAC therapy instead of no anticoagulation (OAC) in the trial.

Altogether, ischemic strokes occurred at a low 0.83 per 100 patient-years with DOACs versus 8.60 per 100 patient-years with no OAC. The number needed to treat (to prevent one ischemic stroke per year) was just 13, reported Roland Veltkamp, MD, of Imperial College London, at the International Stroke Conference.

The bad news was that there was also a major risk of recurrent ICH associated with DOAC therapy (adjusted HR 11.2, 95% CI 2.01-62.86). The incidence of all ICH events reached 5.00 per 100 patient-years versus 0.82 per 100 patient-years, between DOAC and controls, and the number needed to harm (to cause one more ICH per year) was 24.

Calculations of net clinical benefit ultimately did not tilt towards either study group in the 319-person trial.

Thus, nearly 6 years since the start of the study, it remains a conundrum how ICH survivors with Afib are supposed to balance their opposing risks of recurrent bleeding and ischemic events. Looking forward, the data from the ongoing phase III trials ENRICH-AF and ASPIRE may clarify the risk-benefit profile of DOACs in this setting.

For one, the sheer amount of bleeding that occurred in PRESTIGE-AF came as a surprise to Joao Gomes, MD, of Cleveland Clinic.

“Observational data and subsequent meta-analyses had suggested no increase in recurrent ICH risk in patients treated with vitamin K antagonist agents. Similarly, preliminary pilot data on DOACs for ischemic stroke prevention in Afib patients following ICH had showed encouraging safety trends. It was a bit surprising that the risk of subsequent ICH was as high as it was reported in PRESTIGE-AF. Similarly, one would have also expected a higher recurrent ICH risk early on that hopefully stabilized over time, but that did not seem to be the case,” Gomes told MedPage Today in an email.

Meanwhile, there is also interest in alternative therapies such as left atrial appendage (LAA) closure — which had been among the exclusion criteria of PRESTIGE-AF. Veltkamp said the long-term goal is to find imaging and other markers that help personalize the care of each ICH survivor, as some may be better suited to DOACs, and some to other interventions.

“It is becoming increasingly clear that a more individualized approach (i.e., biomarkers that can more accurately predict recurrent ICH risk) is needed for optimal risk stratification of patients facing this conundrum,” agreed Gomes.

“Clinicians who treat stroke survivors face this challenging situation on a frequent basis, and thorough discussions with patients and their surrogates are required to properly inform them of potential risks and benefits of any approach, while assisting in the decision-making process and until more definitive data are available,” he urged.

PRESTIGE-AF was a phase III open-label trial conducted in 63 European sites. Investigators recruited adults who were between 14 days and a year out from their index spontaneous ICH, were diagnosed with Afib, and had an indication for anticoagulation and no contraindication to therapy.

Participants were randomized 1:1 to a standard-dose DOAC or control without any OAC. The choice of the DOAC was left to each site. Ultimately, the study ended up with patients receiving mostly apixaban (53.8%), followed by dabigatran (20.9%) and edoxaban (18.4%).

The final 319-person cohort had a median age of 78-79 and was 35% women. Nearly all participants were white. The median CHA2DS2-VASc score was 4 and the HAS-BLED score a 3. It typically took people just under 50 days between index ICH and study enrollment. Median ICH volumes had been 4.2 mL and 3.2 mL between DOAC and control groups, respectively. The index ICH had been non-lobar in about 70% of cases, the rest lobar.

Outside the bleeding outcomes clearly favoring the control arm, secondary endpoints like mortality and major adverse cardiac events did not suggest that either strategy was better.

Veltkamp and colleagues acknowledged the limited generalizability of the trial beyond the largely white populations studied. The authors had also limited their pool to people with small index ICH volumes and no severe disability.

The PRESTIGE-AF group had started out expecting 654 participants; due to slow enrollment, the authors revised their power calculations later on, settling on a goal of 312 participants.

Please enable JavaScript to view the

comments powered by Disqus.