Agents that reverse the effect of direct oral anticoagulants (DOACs) are highly effective in patients with severe bleeding, but mortality rates remain high despite their use, a meta-analysis shows.
Effective hemostasis was achieved in 78.5% of patients treated with a reversal agent, whereas failure to achieve hemostasis was associated with more than a threefold higher relative risk for death (relative risk, 3.63; 95% confidence interval, 2.56-5.16).
“This has implications in practice because it emphasizes the need for achieving effective hemostasis, if not with only one agent, trying other agents or treatment modalities, because it is a strong predictor of survival,” lead author Antonio Gómez-Outes, MD, PhD, said in an interview.
The bad news, he said, is that the mortality rate was still significant, at 17.7%, and approximately half of patients with DOAC-related severe intracranial bleeding survived with long-term moderate/severe disability.
“The lesson is to prevent these bleeding events because once they appear, even if you give an antidote, the outcome is poor, particularly for intracranial bleeding,” said Dr. Gómez-Outes, division of pharmacology and clinical drug evaluation, Spanish Agency for Medicines and Medical Devices, Madrid.
To put this in context, mortality rates were close to 50% after intracranial bleeding a decade ago when there were no antidotes or reversal agents, he observed. “So to some extent, patient care has improved, and the outcome has improved, but there is a long road to improve regarding disability.”
More than 100,000 DOAC-related major bleeding cases occur each year in the United States and European Union, Dr. Gómez-Outes said, and about half are severe enough to require hospitalization and potentially the use of a reversal agent. These include idarucizumab (Praxbind) for dabigatran reversal and prothombin complex concentrates (4CCC) or andexanet alpha (Andexxa) for reversal of direct factor Xa inhibitors like rivaroxaban, apixaban, and edoxaban.
As reported in the June 22 issue of the Journal of the American College of Cardiology, the meta-analysis comprised 4,735 patients (mean age, 77 years; 57% male) with severe DOAC-related bleeding who received 4PCC (n = 2,688), idarucizumab (n = 1,111), or andexanet (n = 936) in 60 studies between January 2010 and December 2020.
Atrial fibrillation (AFib) was the most common reason for use of a DOAC (82%), followed by venous thromboembolism (14%). Rivaroxaban was used in 36%, apixaban in 32%, dabigatran in 31%, and edoxaban in 1%.
The index bleeding event was intracranial hemorrhage (ICH) in 55%. Anticoagulation was restarted in 57% of patients an average of 11 days after admission.
Mortality rates were 20.2% in patients with ICH and 15.4% in those with extracranial bleeding. There were no differences in death rates by reversal agent used, type of study, risk for bias, or study sponsorship in meta-regression analysis.
Rebleeding occurred in 13.2% of patients; 82.0% of these events were described as an ICH, and 78.0% occurred after anticoagulation was restarted.
The overall rate of thromboembolism was 4.6%. The risk was particularly high with andexanet, at 10.7%, and relatively low with idarucizumab (3.8%) and 4PCC (4.3%), the authors note.
“Our meta-analysis suggests specific reversal with andexanet is not superior to unspecific reversal with 4PCC, and that’s good news because many centers, in many countries, have no access to specific antidotes that are more costly,” Dr. Gómez-Outes said. “4PCC is an effective and relatively safe drug, so it’s still a good option for these patients.”
Labeling for andexanet includes a warning for thromboembolic events, but in the absence of direct comparisons, the findings should be interpreted with caution, he added. Further insights are expected from an ongoing randomized trial of andexanet and standard of care in 900 patients who present with acute ICH less than 15 hours after taking an oral factor Xa inhibitor. The preliminary completion date is set for 2023.
“The meta-analysis raises awareness about the rates of mortality and thromboembolism after reversal agent administration, although understanding the implications of these data is challenging,” Christopher Granger, MD, and Sean P. Pokomey, MD, MBA, Duke University Medical Center, Durham, N.C., say in an accompanying editorial.
The fact that failure to achieve hemostasis was associated with death is expected and might be related to the way hemostasis was defined, rather than the actual failure of the hemostatic treatments, they suggest. “The prothrombotic effects of each agent, including andexanet, need to be better understood, as clinicians work toward including reversal agents into algorithms for bleeding management.”
Effective hemostasis was defined in the studies through various methods as: “Excellent/good” using the Sarode and ANNEXA-4 scales; “yes” in the International Society on Thrombosis and Hemostasis Scale; and with other scales and through clinical judgment.
Although the size of the meta-analysis dwarfs previous reviews, the editorialists and authors point out that 47 of the 60 studies were retrospective, only two had control groups, and 45 had a high risk for bias.
In general, there was also poor reporting of key clinical data, such as postbleeding anticoagulation management, and a limitation of the mortality analysis is that it was based in selected patients with effective hemostasis assessed within 48 hours, which may not capture early deaths, the authors note.
“The morbidity and mortality from ischemic strokes as a result of undertreatment of stroke prevention in patients with AFib continue to dwarf the bleeding related mortality among patients with AFib and on DOACs, and thus the number one priority is to treat nearly all patients with AFib with a DOAC,” Dr. Granger and Dr. Pokomey conclude. “The availability of reversal agents for DOACs should provide reassurance, with another tool in our armamentarium, to providers to prescribe OACs for stroke prevention.”
No funding/grant support was received to conduct the study. Coauthor Ramón Lecumberri has received personal fees from Boehringer Ingelheim and Bristol Myers Squibb outside the submitted work. All other authors report no relevant financial relationships. Dr. Granger has received research and consulting fees from Bristol Myers Squibb, Pfizer, Boehringer Ingelheim, Bayer, Janssen, Boston Scientific, Apple, AstraZeneca, Novartis, AbbVie, Biomed, CeleCor, GSK, Novartis, Medtronic, Merck, Novo Nordisk, Philips, Rho, and the U.S. Food and Drug Administration. Dr. Pokomey has received modest consulting support from Bristol Myers Squibb, Pfizer, Boston Scientific, Medtronic, Janssen, and Zoll; modest research support from Gilead, Boston Scientific, Bristol Myers Squibb, Pfizer, and Janssen; and significant research support from the FDA.
A version of this article first appeared on.