Andexanet alfa and the risk of haematoma expansion in patients with non-traumatic intracerebral haemorrhage – a pooled individual patient data analysis of ANNEXA-4 ICH and TICH-NOAC

Lead Investigator: David Seiffge, Insel Gruppe AG
Title of Proposal Research: Andexanet alfa and the risk of haematoma expansion in patients with non-traumatic intracerebral haemorrhage – a pooled individual patient data analysis of ANNEXA-4 ICH and TICH-NOAC
Vivli Data Request: 8440
Funding Source: This project will receive funding from AstraZeneca to support the analysis.
Potential Conflicts of Interest: None.

Summary of the Proposed Research:

Bleeding in the brain is a terrible disease leading to many patients dying or surviving with severe disability. A special group of blood thinners, so called factor Xa-inhibitors, are valuable drugs to protect patients with certain heart conditions from having a stroke. If patients taking this type of blood thinners bleed in the brain, we fear that the bleeding might grow fast (so called haematoma expansion) which will result in larger bleedings in the brain causing more death and disability. Although it is a rare event (<1% of patients/year), many patients die from the disease (mortality around 30-50%) and others remain severely disabled.

The best therapy to prevent haematoma expansion is not known. Historically, drugs known to stop bleedings in general working in different steps of the blood clotting (coagulation) cascade (so called “non-specific reversal agents”) were used to treat this condition. These unspecific approaches include prothrombin complex concentrate (PCC) and tranexamic acid (TXA). A recent randomized controlled trial assessed treatment with TXA (TICH-NOAC) in combination with PCC in this condition. Recently, a novel drug that specifically eliminates the blood thinner (andexanet alfa) has been developed. We currently do not know, which approach works best to stop the bleeding. Any drug used to stop bleeding might also have side effects, specifically blood clotting and causing stroke, heart attack or other conditions usually summarized as “thrombembolic complications.” There is currently no high-class evidence available which treatment should be given to patients who bleed in the brain while taking this type of blood thinner.

In this study, we will compare data from 2 different studies that used different treatment approaches (andexanet alfa vs. non-specific reversal agents) in patients who bleed in the brain (unrelated to trauma) while taking factor Xa therapy to compare the effect on haematoma expansion and thrombembolic events to better understand, which treatment might be best for patients with this condition. The results of this study will highlight the optimal treatment of patients taking a specific type of blood thinner and who bleed in the brain. The results have the potential to influence future guideline recommendations and improve patient treatment and ultimately patient outcomes.

Statistical Analysis Plan:

We will compare patients enrolled in ANNEXA-4 (Andexanet group) with patients enrolled in TICH-NOAC (non-specific treatment group). All patients will be analysed as intention-to-treat analysis. We will report baseline characteristics according to both groups using appropriate descriptive statistics. For the main analysis, we will compare rate of HE between both groups using binary logistic regression analysis adjusting for pre-defined confounders (baseline haematoma volume, time since last intake of FXaI, age and time-since-symptom-onset) onset reporting adjusted odds ratios (aOR) with 95% confidence intervals (95%CI). For the main safety analysis, we will compare the rate of patients with any thromboembolic complication at 30 days between both groups using unadjusted binary regression analysis. Secondary analyses will compare secondary efficacy and safety outcomes using appropriate regression analyses adjusted for the aforementioned variables.
We will perform pre-specified subgroup analysis repeating the aforementioned analyses in the following subgroups: a) patients fulfilling the inclusion/exclusion criteria for ANNEXA-I and b) patients with calibrated anti-FXa-activity ≥75ng/ml on admission and sensitivity analysis using the ANNEXA-4 definition of HE (≥35%).

For sensitivity analysis, we will perform propensity score–adjusted analyses using the inverse probability of treatment weighting (IPTW) approach6 as done in prior research5.

Handling of missing values: missing values will be imputed for baseline variables. We will not impute outcomes.

Requested Studies:

A Study in Participants With Acute Major Bleeding to Evaluate the Ability of Andexanet Alfa to Reverse the Anticoagulation Effect of Direct and Indirect Oral Anticoagulants (Extension Study)
Data Contributor: AstraZeneca
Study ID: NCT02329327

Treatment of Intracerebral Hemorrhage in Patients on Non-vitamin K Antagonist (TICH-NOAC) This is a randomized controlled double-blinded trial assessing tranexamic acid an addition to best medical management in patients with factor-Xa inhibitor associated intracerebral haemorrhage
Data Contributor: I WILL BRING MY OWN
Study ID: NCT02866838

Public disclosures:

Siepen, B.M., Polymeris, A., Shoamanesh, A., Connolly, S., Steiner, T., Poli, S., Lemmens, R., Goeldlin, M.B., Müller, M., Branca, M. and Rauch, J., 2023. Andexanet alfa versus non-specific treatments for intracerebral hemorrhage in patients taking factor Xa inhibitors—Individual patient data analysis of ANNEXA-4 and TICH-NOAC. International Journal of Stroke, p.17474930 241230209. Doi: 10.1177/17474930241230209