One in five Australian patients with factor Xa inhibitor-related major bleeding die in hospital, with intracranial haemorrhage outcomes particularly poor regardless of the haemostatic strategy chosen, delegates at an international conference have heard.
Presenting at the International Society on Thrombosis and Haemostasis (ISTH) 2026 Congress in Paris, Melbourne investigator Angela Huang said major bleeding associated with oral anticoagulant use remained a significant clinical challenge and the optimal management strategy wasn’t clear-cut.
“Potentially the greatest impact may come from preventing these bleeds in the first place… maybe… we need to identify better predictors of bleeding,” said Ms Huang, a PhD candidate from Monash University and research assistant at the Baker Heart and Diabetes Institute.
The FIBROMA study, which included Melbourne haematologists Associate Professor James McFadyen and Professor Huyen Tran from Alfred Health among the study authors and was conducted with funding from AstraZeneca, is the largest multisite cohort study of factor Xa inhibitor-related major bleeding in Australia.
The researchers looked at 627 cases across nine major hospitals in Victoria and NSW – the country’s most populous states – between January 2019 and June 2023.
Ms Huang pointed out that as there was no specific reversal agent readily available in Australia, haemostatic agents such as prothrombin complex concentrates (PCC) were used off label to manage cases. During the study period, three-factor PCC (3F-PCC) (minimal Factor VII) was the only available PCC formulation.
“Currently, there is a clear evidence gap as there are no randomised controlled trials to guide the use of PCC in this context, and the existing evidence is largely observational and derived predominantly from the four-factor formulation.”
“Consequently, there has been a lack of outcome data specific to the Australian context.”
The study cohort was elderly (53% male, median age 80) and most received no haemostatic agent. Of the 206 who received a haemostatic agent, 187 patients received PCC as either monotherapy or in combination with other agents.
The type of anticoagulation was mostly apixaban (65%) followed by rivaroxaban (35%) and the original indication for anticoagulation was mostly for atrial fibrillation (77%), followed by venous thromboembolism (20%) and other conditions (3.2%).
Half of the major bleeds were ICH and 38% were GI bleeding.
Findings showed the use of haemostatic agents was not associated with improvements in mortality.
Overall the haemostatic efficacy was:
- Excellent: 15% of the cohort
- Good: 35%
- Effective: 30%
- Ineffective: 20%
For ICH cases versus GI bleeding cases, haemostatic efficacy was:
- Excellent: 15% versus 13%
- Good: 41% versus 17%
- Effective: 11% versus 65%
- Ineffective: 33% versus 6%
The mortality rate overall was 23.8%, with ICH associated with a much higher mortality rate than GI bleeds (37.2% versus 11.7%; 4-fold higher odds).
Just over half of the cohort was discharged home and 19.8% of the cohort was discharged to a rehabilitation centre. But again, outcomes differed markedly by bleeding site (29.8% versus 80% and 31% versus 5% for ICH versus GI bleeds discharged home and discharged to a rehabilitation centre, respectively).
Along with ICH, PCC monotherapy was independently associated with in-hospital mortality (1.63-fold higher odds versus no PCC). There was no significant association for age, sex, haemoglobin levels, eGFR and procedural intervention.
Regarding hospital stay, ICH was associated with a substantially longer admission (86% longer than GI bleeding), whereas undergoing an intervention was associated with a 54% shorter stay and increasing age with a marginally shorter stay.
Regarding safety, the overall thrombosis rate at 30 days was low at 2.8% and for PCC users specifically, the rate was 1.9%, indicating no increased thrombosis risk.